From 21efadc62272cabee9bec27777ae75d84a9ca8a8 Mon Sep 17 00:00:00 2001 From: =?UTF-8?q?Alvin=20=C5=A0ipraga?= Date: Tue, 18 Aug 2026 18:00:02 +0200 Subject: [PATCH 001/592] Input: adp5588-keys - cache GPIO state before registering the gpiochip MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit So as not to clobber any pre-programmed GPIO state in the execution of its gpiochip ops, the driver caches things during probe time. However, since those ops can be called both during and immediately after the call to devm_gpiochip_add_data(), it is imperative that things are cached before that. That's not the case right now, so reorder the two steps to prevent any clobbering. In the concrete example which motivated this change, a bootloader was preconfiguring an important GPIO output to HIGH before booting the kernel. Linux would then inadvertently set that output to LOW while configuring a GPIO hog on a discrete GPIO line within the same 8-bit bank (because the cached value was 0=LOW). Fixes: ba9f507a1bea ("Input: adp5588-keys - export unused GPIO pins") Signed-off-by: Alvin Šipraga Reviewed-by: Nuno Sá Link: https://patch.msgid.link/20260818-adp5588-gpio-cache-v1-1-650a2674fc0d@analog.com Cc: stable@vger.kernel.org Signed-off-by: Dmitry Torokhov --- drivers/input/keyboard/adp5588-keys.c | 12 ++++++------ 1 file changed, 6 insertions(+), 6 deletions(-) diff --git a/drivers/input/keyboard/adp5588-keys.c b/drivers/input/keyboard/adp5588-keys.c index 40371f5bd9ba..4f0ddff5baba 100644 --- a/drivers/input/keyboard/adp5588-keys.c +++ b/drivers/input/keyboard/adp5588-keys.c @@ -446,12 +446,6 @@ static int adp5588_gpio_add(struct adp5588_kpad *kpad) mutex_init(&kpad->gpio_lock); - error = devm_gpiochip_add_data(dev, &kpad->gc, kpad); - if (error) { - dev_err(dev, "gpiochip_add failed: %d\n", error); - return error; - } - for (i = 0; i <= ADP5588_BANK(ADP5588_MAXGPIO); i++) { kpad->dat_out[i] = adp5588_read(kpad->client, GPIO_DAT_OUT1 + i); @@ -459,6 +453,12 @@ static int adp5588_gpio_add(struct adp5588_kpad *kpad) kpad->pull_dis[i] = adp5588_read(kpad->client, GPIO_PULL1 + i); } + error = devm_gpiochip_add_data(dev, &kpad->gc, kpad); + if (error) { + dev_err(dev, "gpiochip_add failed: %d\n", error); + return error; + } + return 0; } From 4d7fa28e151a6a6f24098d9e60f558f8ea61ddaa Mon Sep 17 00:00:00 2001 From: Roman Vivchar Date: Fri, 14 Aug 2026 16:13:52 +0300 Subject: [PATCH 002/592] dt-bindings: input: mediatek,mt6779-keypad: add mt6572 Add a compatible string for the mt6572 keypad, that is compatible with mt6779. Signed-off-by: Roman Vivchar Reviewed-by: Mattijs Korpershoek Acked-by: Krzysztof Kozlowski Link: https://patch.msgid.link/20260814-6572-dt-keypad-v1-1-2ceff2373211@protonmail.com Signed-off-by: Dmitry Torokhov --- .../devicetree/bindings/input/mediatek,mt6779-keypad.yaml | 1 + 1 file changed, 1 insertion(+) diff --git a/Documentation/devicetree/bindings/input/mediatek,mt6779-keypad.yaml b/Documentation/devicetree/bindings/input/mediatek,mt6779-keypad.yaml index 914dd3283df3..20fa41f691ed 100644 --- a/Documentation/devicetree/bindings/input/mediatek,mt6779-keypad.yaml +++ b/Documentation/devicetree/bindings/input/mediatek,mt6779-keypad.yaml @@ -26,6 +26,7 @@ properties: - const: mediatek,mt6779-keypad - items: - enum: + - mediatek,mt6572-keypad - mediatek,mt6873-keypad - mediatek,mt8183-keypad - mediatek,mt8365-keypad From aefbda23eeba234c3ff0f135dc5be6e294bd25a6 Mon Sep 17 00:00:00 2001 From: Alexei Turtanov <9alexei9@gmail.com> Date: Fri, 28 Aug 2026 14:22:39 +0300 Subject: [PATCH 003/592] Input: atkbd - skip deactivate for Xiaomi Redmi Book Pro 16 2026 The internal keyboard of the Xiaomi Redmi Book Pro 16 2026 (board TM2425) does not work: atkbd_probe() succeeds and every command is ACKed, but no scancodes ever arrive afterwards. Testing on the hardware through serio_raw shows that ATKBD_CMD_RESET_DIS (0xF5) is the culprit. After 0xF5 the embedded controller keeps ACKing commands but stops delivering scancodes, and neither ATKBD_CMD_ENABLE (0xF4) nor ATKBD_CMD_RESET_BAT (0xFF) bring them back. Only re-enabling the keyboard interface at the controller level (i8042 command 0xAE, or rewriting the command byte as i8042_port_close() does) revives it. Running the init sequence without 0xF5 (0xED 0x00, 0xF3 0x00, 0xF4) keeps the keyboard working. 'i8042.dumbkbd=1' also works around this, but then the driver never writes to the keyboard and the LEDs cannot be controlled. Use the existing atkbd_deactivate_fixup quirk instead, as done for the sibling TM2424 by commit 3a046db33bb9 ("Input: atkbd - skip deactivate for Xiaomi Book Pro 14's internal keyboard"). Tested on v7.2: keyboard, Caps Lock LED and s2idle suspend/resume all work. DMI: XIAOMI REDMI Book Pro 16 2026/TM2425, BIOS RMAPT6B0P0909 05/22/2026 Fixes: 9cf6e24c9fbf ("Input: atkbd - do not skip atkbd_deactivate() when skipping ATKBD_CMD_GETID") Cc: stable@vger.kernel.org Signed-off-by: Alexei Turtanov <9alexei9@gmail.com> Link: https://patch.msgid.link/20260828112239.18081-1-9alexei9@gmail.com Signed-off-by: Dmitry Torokhov --- drivers/input/keyboard/atkbd.c | 8 ++++++++ 1 file changed, 8 insertions(+) diff --git a/drivers/input/keyboard/atkbd.c b/drivers/input/keyboard/atkbd.c index 5736f4bc5a50..2627434099ff 100644 --- a/drivers/input/keyboard/atkbd.c +++ b/drivers/input/keyboard/atkbd.c @@ -1946,6 +1946,14 @@ static const struct dmi_system_id atkbd_dmi_quirk_table[] __initconst = { }, .callback = atkbd_deactivate_fixup, }, + { + /* Xiaomi Redmi Book Pro 16 2026 (TM2425) */ + .matches = { + DMI_MATCH(DMI_SYS_VENDOR, "XIAOMI"), + DMI_MATCH(DMI_PRODUCT_NAME, "REDMI Book Pro 16 2026"), + }, + .callback = atkbd_deactivate_fixup, + }, { } }; From 51cfe54f815ae175c7d1126b983d4d7c89715004 Mon Sep 17 00:00:00 2001 From: Dmitry Torokhov Date: Tue, 4 Aug 2026 22:08:54 -0700 Subject: [PATCH 004/592] Input: rmi_smbus - fix out-of-bounds read in rmi_smb_write_block() When chunking writes into SMBus blocks in rmi_smb_write_block(), the loop calculates block_len using the original total length (len) instead of the remaining length (cur_len). If len is greater than 32 bytes (SMB_MAX_COUNT), block_len remains 32 for every iteration, even on the final partial chunk where fewer than 32 bytes remain. This causes smb_block_write() to read 32 bytes from the advanced data buffer pointer, reading past the end of the input buffer. Fix this by calculating block_len using cur_len and advancing the buffer and address pointers by block_len. Fixes: 82264d0cf7ae ("Input: synaptics-rmi4 - add SMBus support") Cc: stable@vger.kernel.org Reported-by: sashiko-bot@kernel.org Assisted-by: LLM Link: https://patch.msgid.link/anLFSMKSoKyyZ272@google.com Signed-off-by: Dmitry Torokhov --- drivers/input/rmi4/rmi_smbus.c | 10 +++++----- 1 file changed, 5 insertions(+), 5 deletions(-) diff --git a/drivers/input/rmi4/rmi_smbus.c b/drivers/input/rmi4/rmi_smbus.c index 3160714a514a..8b61c2382297 100644 --- a/drivers/input/rmi4/rmi_smbus.c +++ b/drivers/input/rmi4/rmi_smbus.c @@ -140,7 +140,7 @@ static int rmi_smb_write_block(struct rmi_transport_dev *xport, u16 rmiaddr, u8 commandcode; struct rmi_smb_xport *rmi_smb = container_of(xport, struct rmi_smb_xport, xport); - int cur_len = (int)len; + size_t cur_len = len; mutex_lock(&rmi_smb->page_mutex); @@ -148,7 +148,7 @@ static int rmi_smb_write_block(struct rmi_transport_dev *xport, u16 rmiaddr, /* * break into 32 bytes chunks to write get command code */ - int block_len = min_t(int, len, SMB_MAX_COUNT); + int block_len = min_t(size_t, cur_len, SMB_MAX_COUNT); retval = rmi_smb_get_command_code(xport, rmiaddr, block_len, false, &commandcode); @@ -161,9 +161,9 @@ static int rmi_smb_write_block(struct rmi_transport_dev *xport, u16 rmiaddr, goto exit; /* prepare to write next block of bytes */ - cur_len -= SMB_MAX_COUNT; - databuff += SMB_MAX_COUNT; - rmiaddr += SMB_MAX_COUNT; + cur_len -= block_len; + databuff += block_len; + rmiaddr += block_len; } exit: mutex_unlock(&rmi_smb->page_mutex); From fe10579b6dc3f0dac61e51e1797cacbba5039ac2 Mon Sep 17 00:00:00 2001 From: Dmitry Torokhov Date: Wed, 5 Aug 2026 22:44:23 -0700 Subject: [PATCH 005/592] Input: synaptics-rmi4 - fix GPF in suspend and resume when unbound Transport drivers (such as rmi_i2c and rmi_spi) invoke rmi_driver_suspend() and rmi_driver_resume() on their child rmi_dev device during system power management events. However, transport drivers are fully registered and operational even if the physical RMI driver failed to bind or probe the rmi_dev device. When rmi_driver_suspend() or rmi_driver_resume() is called on an unbound rmi_dev, dev_get_drvdata() returns NULL. Calling rmi_disable_irq() or rmi_enable_irq() without driver data attached causes a NULL pointer dereference and General Protection Fault when attempting to lock data->enabled_mutex. Fix this by checking if driver data is attached to rmi_dev in rmi_driver_suspend() and rmi_driver_resume(), exiting early if no driver data is present. Fixes: 2b6a321da9a2 ("Input: synaptics-rmi4 - add support for Synaptics RMI4 devices") Reported-by: syzbot+09103639e39c989e3ed3@syzkaller.appspotmail.com Closes: https://syzkaller.appspot.com/bug?extid=09103639e39c989e3ed3 Cc: stable@vger.kernel.org Assisted-by: LLM Link: https://patch.msgid.link/anQe8UiyUR4x0flD@google.com Signed-off-by: Dmitry Torokhov --- drivers/input/rmi4/rmi_driver.c | 13 +++++++++++++ 1 file changed, 13 insertions(+) diff --git a/drivers/input/rmi4/rmi_driver.c b/drivers/input/rmi4/rmi_driver.c index 5d49a9021c7d..a349dfd17519 100644 --- a/drivers/input/rmi4/rmi_driver.c +++ b/drivers/input/rmi4/rmi_driver.c @@ -991,6 +991,15 @@ int rmi_driver_suspend(struct rmi_device *rmi_dev, bool enable_wake) { int retval; + /* + * Transport driver will try to suspend RMI device even if physical + * driver did not bind to the RMI device, because transport device + * (I2C, SPI) is fully registered and operational. Exit early if + * there is no driver data attached to the RMI device. + */ + if (!dev_get_drvdata(&rmi_dev->dev)) + return 0; + retval = rmi_suspend_functions(rmi_dev); if (retval) dev_warn(&rmi_dev->dev, "Failed to suspend functions: %d\n", @@ -1005,6 +1014,10 @@ int rmi_driver_resume(struct rmi_device *rmi_dev, bool clear_wake) { int retval; + /* Skip if not fully bound to RMI driver */ + if (!dev_get_drvdata(&rmi_dev->dev)) + return 0; + rmi_enable_irq(rmi_dev, clear_wake); retval = rmi_resume_functions(rmi_dev); From 882de800bf583dd15969836088789edbf035a944 Mon Sep 17 00:00:00 2001 From: Lucas Tanure Date: Sun, 23 Aug 2026 12:53:34 +0100 Subject: [PATCH 006/592] arm64: dts: amlogic: t7: use the real UART pclk uart_a listed the 24MHz crystal for all three of its clocks because the T7 clock controller driver did not exist when these boards were added. That leaves the real UART bus clock without a user, so the kernel turns it off when it disables unused clocks at the end of boot, and the board hangs. Update the board DTS files to point uart_a's pclk to CLKID_SYS_UART_A instead of the dummy crystal clock. Fixes: 4fef056588f5 ("arm64: dts: amlogic-t7-a311d2-khadas-vim4: add initial device-tree") Fixes: 6f048cc7a635 ("arm64: dts: add board AN400") Signed-off-by: Lucas Tanure Assisted-by: Claude:claude-fable-5 Reviewed-by: Neil Armstrong Link: https://patch.msgid.link/20260823115335.102219-2-tanure@linux.com Signed-off-by: Neil Armstrong --- arch/arm64/boot/dts/amlogic/amlogic-t7-a311d2-an400.dts | 2 +- arch/arm64/boot/dts/amlogic/amlogic-t7-a311d2-khadas-vim4.dts | 2 +- 2 files changed, 2 insertions(+), 2 deletions(-) diff --git a/arch/arm64/boot/dts/amlogic/amlogic-t7-a311d2-an400.dts b/arch/arm64/boot/dts/amlogic/amlogic-t7-a311d2-an400.dts index cab2ee9ea0d3..ca7536f772ff 100644 --- a/arch/arm64/boot/dts/amlogic/amlogic-t7-a311d2-an400.dts +++ b/arch/arm64/boot/dts/amlogic/amlogic-t7-a311d2-an400.dts @@ -33,7 +33,7 @@ xtal: xtal-clk { }; &uart_a { - clocks = <&xtal>, <&xtal>, <&xtal>; + clocks = <&xtal>, <&clkc_periphs CLKID_SYS_UART_A>, <&xtal>; clock-names = "xtal", "pclk", "baud"; status = "okay"; }; diff --git a/arch/arm64/boot/dts/amlogic/amlogic-t7-a311d2-khadas-vim4.dts b/arch/arm64/boot/dts/amlogic/amlogic-t7-a311d2-khadas-vim4.dts index c41525a34b72..75d81ad6830d 100644 --- a/arch/arm64/boot/dts/amlogic/amlogic-t7-a311d2-khadas-vim4.dts +++ b/arch/arm64/boot/dts/amlogic/amlogic-t7-a311d2-khadas-vim4.dts @@ -266,6 +266,6 @@ &sd_emmc_c { &uart_a { status = "okay"; - clocks = <&xtal>, <&xtal>, <&xtal>; + clocks = <&xtal>, <&clkc_periphs CLKID_SYS_UART_A>, <&xtal>; clock-names = "xtal", "pclk", "baud"; }; From c793a084ab3088744516aef437b4239d0d98623f Mon Sep 17 00:00:00 2001 From: Lucas Tanure Date: Sun, 23 Aug 2026 12:53:35 +0100 Subject: [PATCH 007/592] arm64: dts: amlogic: t7: khadas-vim4: allow the SD card to be power cycled SD cards start at 3.3V and switch to 1.8V to reach UHS-I speeds. Some cards refuse that switch, and the SD specification says the only way to recover is to power the card off and start again. SD_3V3 is marked regulator-always-on, so the supply never goes off, the card stays stuck half way through the switch, and the MMC core retries forever: mmc1: error -95 whilst initialising SD card Drop regulator-always-on. regulator-boot-on still turns the supply on at boot, and a card that refuses the switch now falls back to high speed instead of failing to initialise. Fixes: 8c45bf9ae4ef ("arm64: dts: amlogic: t7: khadas-vim4: Add power regulators") Signed-off-by: Lucas Tanure Assisted-by: Claude:claude-fable-5 Reviewed-by: Neil Armstrong Link: https://patch.msgid.link/20260823115335.102219-3-tanure@linux.com Signed-off-by: Neil Armstrong --- arch/arm64/boot/dts/amlogic/amlogic-t7-a311d2-khadas-vim4.dts | 1 - 1 file changed, 1 deletion(-) diff --git a/arch/arm64/boot/dts/amlogic/amlogic-t7-a311d2-khadas-vim4.dts b/arch/arm64/boot/dts/amlogic/amlogic-t7-a311d2-khadas-vim4.dts index 75d81ad6830d..c94afaaf8826 100644 --- a/arch/arm64/boot/dts/amlogic/amlogic-t7-a311d2-khadas-vim4.dts +++ b/arch/arm64/boot/dts/amlogic/amlogic-t7-a311d2-khadas-vim4.dts @@ -71,7 +71,6 @@ sd_3v3: regulator-sdcard-3v3 { vin-supply = <&vddao_3v3>; gpio = <&gpio GPIOD_11 GPIO_ACTIVE_LOW>; regulator-boot-on; - regulator-always-on; }; sdio_pwrseq: sdio-pwrseq { From ae7be5c58a819259e5eb98a3abfa60842478f2ae Mon Sep 17 00:00:00 2001 From: Lucas Tanure Date: Sat, 29 Aug 2026 10:47:56 +0100 Subject: [PATCH 008/592] arm64: dts: amlogic: t7: fix the pin groups of two PWM outputs Two of the PWM outputs can each appear on more than one pin, but the description named a single group that does not exist, so anything using it refused to start. Name the real groups instead, one entry per pin, the same way the other multi-pin PWM output is already described. Fixes: 2a2a7b9701a7 ("arm64: dts: amlogic: t7: Add PWM pinctrl nodes") Assisted-by: Claude:claude-opus-5 Signed-off-by: Lucas Tanure Reviewed-by: Neil Armstrong Link: https://patch.msgid.link/20260829094758.23248-3-tanure@linux.com Signed-off-by: Neil Armstrong --- arch/arm64/boot/dts/amlogic/amlogic-t7.dtsi | 32 ++++++++++++++++++--- 1 file changed, 28 insertions(+), 4 deletions(-) diff --git a/arch/arm64/boot/dts/amlogic/amlogic-t7.dtsi b/arch/arm64/boot/dts/amlogic/amlogic-t7.dtsi index c3dc479b137d..8c4bd683cce0 100644 --- a/arch/arm64/boot/dts/amlogic/amlogic-t7.dtsi +++ b/arch/arm64/boot/dts/amlogic/amlogic-t7.dtsi @@ -458,9 +458,25 @@ mux { }; }; - pwm_ao_g_pins: pwm-ao-g { + pwm_ao_g_d11_pins: pwm-ao-g-d11 { mux { - groups = "pwm_ao_g"; + groups = "pwm_ao_g_d11"; + function = "pwm_ao_g"; + bias-disable; + }; + }; + + pwm_ao_g_d7_pins: pwm-ao-g-d7 { + mux { + groups = "pwm_ao_g_d7"; + function = "pwm_ao_g"; + bias-disable; + }; + }; + + pwm_ao_g_e_pins: pwm-ao-g-e { + mux { + groups = "pwm_ao_g_e"; function = "pwm_ao_g"; bias-disable; }; @@ -474,9 +490,17 @@ mux { }; }; - pwm_ao_h_pins: pwm-ao-h { + pwm_ao_h_d5_pins: pwm-ao-h-d5 { mux { - groups = "pwm_ao_h"; + groups = "pwm_ao_h_d5"; + function = "pwm_ao_h"; + bias-disable; + }; + }; + + pwm_ao_h_d10_pins: pwm-ao-h-d10 { + mux { + groups = "pwm_ao_h_d10"; function = "pwm_ao_h"; bias-disable; }; From 1e5a53bd16ac501e68463e84023d1bff543cc696 Mon Sep 17 00:00:00 2001 From: Lucas Tanure Date: Sat, 29 Aug 2026 10:47:57 +0100 Subject: [PATCH 009/592] arm64: dts: amlogic: t7: khadas-vim4: add the PWM-driven supplies The board powers its two CPU clusters, the GPU, the NPU, the memory and the always-on domain from regulators steered by PWM outputs. None of them were described, so Linux treated those outputs as unused and switched them off part way through boot. The supplies then drifted away from the levels the bootloader had set, which showed up as random hangs and memory corruption. Describe each supply so it has an owner and is left alone. The voltage ranges are read off the feedback networks on the board schematic. VDDNPU is deliberately not the range in Amlogic's own device tree: this board fits a different feedback resistor, which puts it about 40mV higher at both ends. Fixes: 8c45bf9ae4ef ("arm64: dts: amlogic: t7: khadas-vim4: Add power regulators") Assisted-by: Claude:claude-opus-5 Signed-off-by: Lucas Tanure Reviewed-by: Neil Armstrong Link: https://patch.msgid.link/20260829094758.23248-4-tanure@linux.com Signed-off-by: Neil Armstrong --- .../amlogic/amlogic-t7-a311d2-khadas-vim4.dts | 110 +++++++++++++++++- 1 file changed, 109 insertions(+), 1 deletion(-) diff --git a/arch/arm64/boot/dts/amlogic/amlogic-t7-a311d2-khadas-vim4.dts b/arch/arm64/boot/dts/amlogic/amlogic-t7-a311d2-khadas-vim4.dts index c94afaaf8826..77abfd555ec5 100644 --- a/arch/arm64/boot/dts/amlogic/amlogic-t7-a311d2-khadas-vim4.dts +++ b/arch/arm64/boot/dts/amlogic/amlogic-t7-a311d2-khadas-vim4.dts @@ -104,6 +104,66 @@ vcc5v0_usb: regulator-vcc-usb { enable-active-high; }; + vdd_ddr: regulator-vddddr { + /* + * SY8003ADFC Regulator. + */ + compatible = "pwm-regulator"; + regulator-name = "VDDDDR"; + regulator-min-microvolt = <690000>; + regulator-max-microvolt = <890000>; + pwm-supply = <&dc_in>; + pwms = <&pwm_ao_gh 0 1500 0>; + pwm-dutycycle-range = <100 0>; + regulator-boot-on; + regulator-always-on; + }; + + vdd_ee: regulator-vddee { + /* + * MP8756GD Regulator. + */ + compatible = "pwm-regulator"; + regulator-name = "VDDEE"; + regulator-min-microvolt = <700000>; + regulator-max-microvolt = <922000>; + pwm-supply = <&dc_in>; + pwms = <&pwm_ao_ab 0 1500 0>; + pwm-dutycycle-range = <100 0>; + regulator-boot-on; + regulator-always-on; + }; + + vdd_gpu: regulator-vddgpu { + /* + * SY8003ADFC Regulator. + */ + compatible = "pwm-regulator"; + regulator-name = "VDDGPU"; + regulator-min-microvolt = <700000>; + regulator-max-microvolt = <922000>; + pwm-supply = <&dc_in>; + pwms = <&pwm_ao_ef 0 1500 0>; + pwm-dutycycle-range = <100 0>; + regulator-boot-on; + regulator-always-on; + }; + + vdd_npu: regulator-vddnpu { + /* + * SY8003ADFC Regulator. + */ + compatible = "pwm-regulator"; + regulator-name = "VDDNPU"; + regulator-min-microvolt = <733000>; + regulator-max-microvolt = <933000>; + pwm-supply = <&dc_in>; + pwms = <&pwm_ao_ef 1 1500 0>; + pwm-dutycycle-range = <100 0>; + regulator-boot-on; + regulator-always-on; + }; + vddao_1v8: regulator-vddao-1v8 { compatible = "regulator-fixed"; regulator-name = "VDDAO_1V8"; @@ -122,6 +182,36 @@ vddao_3v3: regulator-vddao-3v3 { regulator-always-on; }; + vddcpu_a: regulator-vddcpu-a { + /* + * MP8756GD Regulator. + */ + compatible = "pwm-regulator"; + regulator-name = "VDDCPU_A"; + regulator-min-microvolt = <689000>; + regulator-max-microvolt = <1049000>; + pwm-supply = <&dc_in>; + pwms = <&pwm_ao_cd 1 1500 0>; + pwm-dutycycle-range = <100 0>; + regulator-boot-on; + regulator-always-on; + }; + + vddcpu_b: regulator-vddcpu-b { + /* + * MP8756GD Regulator. + */ + compatible = "pwm-regulator"; + regulator-name = "VDDCPU_B"; + regulator-min-microvolt = <689000>; + regulator-max-microvolt = <1049000>; + pwm-supply = <&dc_in>; + pwms = <&pwm_ao_ab 1 1500 0>; + pwm-dutycycle-range = <100 0>; + regulator-boot-on; + regulator-always-on; + }; + vddio_1v8: regulator-vddio-1v8 { compatible = "regulator-fixed"; regulator-name = "VDDIO_1V8"; @@ -172,9 +262,27 @@ &pwm_ab { pinctrl-names = "default"; }; +&pwm_ao_ab { + status = "okay"; + pinctrl-0 = <&pwm_ao_a_pins>, <&pwm_ao_b_pins>; + pinctrl-names = "default"; +}; + &pwm_ao_cd { status = "okay"; - pinctrl-0 = <&pwm_ao_c_d_pins>; + pinctrl-0 = <&pwm_ao_c_d_pins>, <&pwm_ao_d_pins>; + pinctrl-names = "default"; +}; + +&pwm_ao_ef { + status = "okay"; + pinctrl-0 = <&pwm_ao_e_pins>, <&pwm_ao_f_pins>; + pinctrl-names = "default"; +}; + +&pwm_ao_gh { + status = "okay"; + pinctrl-0 = <&pwm_ao_g_e_pins>; pinctrl-names = "default"; }; From 406292fd75f95aa3010fec95b5beb5a8b7e3ba3a Mon Sep 17 00:00:00 2001 From: Lucas Tanure Date: Sat, 29 Aug 2026 10:47:58 +0100 Subject: [PATCH 010/592] arm64: dts: amlogic: t7: fix the pin groups of the vsync PWM The vsync PWM output can appear on either of two pins, but the description named a single group that does not exist, so anything using it would refuse to start. Name the real groups instead, one entry per pin. No board describes this output yet, so nothing changes today. Fixes: 2a2a7b9701a7 ("arm64: dts: amlogic: t7: Add PWM pinctrl nodes") Assisted-by: Claude:claude-opus-5 Signed-off-by: Lucas Tanure Reviewed-by: Neil Armstrong Link: https://patch.msgid.link/20260829094758.23248-5-tanure@linux.com Signed-off-by: Neil Armstrong --- arch/arm64/boot/dts/amlogic/amlogic-t7.dtsi | 12 ++++++++++-- 1 file changed, 10 insertions(+), 2 deletions(-) diff --git a/arch/arm64/boot/dts/amlogic/amlogic-t7.dtsi b/arch/arm64/boot/dts/amlogic/amlogic-t7.dtsi index 8c4bd683cce0..719e111bc3dd 100644 --- a/arch/arm64/boot/dts/amlogic/amlogic-t7.dtsi +++ b/arch/arm64/boot/dts/amlogic/amlogic-t7.dtsi @@ -546,9 +546,17 @@ mux { }; }; - pwm_vs_pins: pwm-vs { + pwm_vs_y_pins: pwm-vs-y { mux { - groups = "pwm_vs"; + groups = "pwm_vs_y"; + function = "pwm_vs"; + bias-disable; + }; + }; + + pwm_vs_h_pins: pwm-vs-h { + mux { + groups = "pwm_vs_h"; function = "pwm_vs"; bias-disable; }; From cfc1e9a543e3589ba200795b6e7fd8ef4314efdf Mon Sep 17 00:00:00 2001 From: Pengpeng Hou Date: Sun, 30 Aug 2026 22:22:44 +0800 Subject: [PATCH 011/592] ARM: socfpga: select the PL310 erratum 753970 workaround ARCH_INTEL_SOCFPGA selects CACHE_L2X0 and several PL310 erratum workarounds. The 753970 workaround is still conditioned on PL310, but that Kconfig symbol no longer exists, so this one selection is always disabled. Select PL310_ERRATA_753970 directly, consistently with the other PL310 workarounds required by the platform. Fixes: fbc125afdc50 ("ARM: socfpga: Turn on ARM errata for L2 cache") Signed-off-by: Pengpeng Hou Signed-off-by: Dinh Nguyen --- arch/arm/mach-socfpga/Kconfig | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/arch/arm/mach-socfpga/Kconfig b/arch/arm/mach-socfpga/Kconfig index eb72c240c248..528c5c1368c3 100644 --- a/arch/arm/mach-socfpga/Kconfig +++ b/arch/arm/mach-socfpga/Kconfig @@ -16,7 +16,7 @@ menuconfig ARCH_INTEL_SOCFPGA select ARM_ERRATA_775420 select PL310_ERRATA_588369 select PL310_ERRATA_727915 - select PL310_ERRATA_753970 if PL310 + select PL310_ERRATA_753970 select PL310_ERRATA_769419 select RESET_CONTROLLER From 34b8b2d78b6276dc2dc4ebc06625a39956f266e4 Mon Sep 17 00:00:00 2001 From: Shawn Guo Date: Sat, 1 Aug 2026 09:17:31 +0800 Subject: [PATCH 012/592] remoteproc: qcom: q6v5_pas: Don't enable handover IRQ on attach qcom_pas_attach() unmasks the handover IRQ and marks handover_issued even though this driver instance never runs qcom_q6v5_prepare() for the boot it is attaching to. This was believed necessary to flush a stale interrupt latched at the interrupt controller while masked, but the handover IRQ is a Qualcomm SMP2P soft IRQ, not a real edge-latched hardware interrupt. The Linux SMP2P driver updates its cached value unconditionally on every notification and only delivers the nested IRQ for bits currently enabled in its own software bitmap, so a transition that happens while masked is simply dropped, never replayed on a later unmask. Since there is nothing to flush, and this driver instance never takes the proxy power-domain/clock/regulator votes that the handover callback would tear down, there is no need to unmask the IRQ in attach() at all. Drop the enable_irq()/disable_irq() pair; setting handover_issued = true is sufficient to keep the flag consistent for the eventual qcom_q6v5_unprepare()/qcom_q6v5_prepare() cycle. It fixes the following unbalanced runtime PM usage and IRQ enable warnings seen on Nord ADSP (probed as attached), after commit bb7c5d6f5b41 ("remoteproc: qcom: q6v5: Make handover IRQ one-shot") comes in place. root@iq10-rrd:~# cat /sys/class/remoteproc/remoteproc0/state attached root@iq10-rrd:~# echo stop > /sys/class/remoteproc/remoteproc0/state [ 40.004874] genpd genpd:0:4c00000.remoteproc: Runtime PM usage count underflow! [ 40.012409] genpd genpd:1:4c00000.remoteproc: Runtime PM usage count underflow! [ 40.050074] remoteproc remoteproc0: stopped remote processor adsp root@iq10-rrd:~# echo start > /sys/class/remoteproc/remoteproc0/state [ 44.350298] remoteproc remoteproc0: powering up adsp [ 44.375769] remoteproc remoteproc0: Booting fw image qcom/nord/adsp.mbn, size 8241816 [ 44.389850] PDM: no support for the platform, userspace daemon might be required. [ 44.397864] ------------[ cut here ]------------ [ 44.402633] Unbalanced enable for IRQ 363 [ 44.406779] WARNING: kernel/irq/manage.c:775 at __enable_irq+0x4c/0x7c, CPU#9: sh/791 ... Suggested-by: Stephan Gerhold Signed-off-by: Shawn Guo Reviewed-by: Abel Vesa Link: https://lore.kernel.org/r/20260801011731.1084591-1-shengchao.guo@oss.qualcomm.com Signed-off-by: Bjorn Andersson --- drivers/remoteproc/qcom_q6v5_pas.c | 2 -- 1 file changed, 2 deletions(-) diff --git a/drivers/remoteproc/qcom_q6v5_pas.c b/drivers/remoteproc/qcom_q6v5_pas.c index ca8e61254c44..8cff0684fe2f 100644 --- a/drivers/remoteproc/qcom_q6v5_pas.c +++ b/drivers/remoteproc/qcom_q6v5_pas.c @@ -524,7 +524,6 @@ static int qcom_pas_attach(struct rproc *rproc) int ret; pas->q6v5.handover_issued = true; - enable_irq(pas->q6v5.handover_irq); pas->q6v5.running = true; ret = irq_get_irqchip_state(pas->q6v5.fatal_irq, @@ -570,7 +569,6 @@ static int qcom_pas_attach(struct rproc *rproc) pas->rproc->state = RPROC_OFFLINE; ret = -EINVAL; disable_running: - disable_irq(pas->q6v5.handover_irq); pas->q6v5.running = false; return ret; From b853857293584dd1f52183f70a2bdeca692a1e71 Mon Sep 17 00:00:00 2001 From: Paul Hollinsky Date: Fri, 21 Aug 2026 01:15:40 -0700 Subject: [PATCH 013/592] remoteproc: qcom_q6v5_mss: Don't require PAS for memory protection Commit f3b1357673dd ("remoteproc: qcom_q6v5_mss: Switch to generic PAS TZ APIs") changed the probe-time gate for need_mem_protection platforms from qcom_scm_is_available() to qcom_pas_is_available(). Memory protection in this driver is implemented with qcom_scm_assign_mem(), which is a TZ service distinct from PAS. The only PAS call in the driver is qcom_pas_mem_setup(), and it is already guarded by need_pas_mem_setup. No descriptor sets both flags: sc7180, sc7280, sdm660, sdm845, msm8996 and msm8998 set need_mem_protection only, while msm8937, msm8940 and msm8953 set need_pas_mem_setup only. On TrustZone firmware that does not implement PAS - for example SC7180 Chromebooks, where call-availability queries report every QCOM_SCM_SVC_PIL command as unavailable - the modem consequently never probes: platform 4080000.remoteproc: deferred probe pending: (reason unknown) On those machines the modem is also what loads the WLAN firmware, so ath10k never receives QMI and wifi does not come up either. Gate memory protection on SCM availability as it was before, and require PAS only where a PAS call is actually issued. Keeping the SCM check matters: qcom_scm_assign_mem() passes __scm->mempool to qcom_tzmem_alloc() without testing __scm, so dropping the gate entirely would allow a NULL dereference when qcom_scm has not yet probed. Fixes: f3b1357673dd ("remoteproc: qcom_q6v5_mss: Switch to generic PAS TZ APIs") Link: https://lore.kernel.org/r/20260808034716.58888-1-phollinsky@holtechnik.com Signed-off-by: Paul Hollinsky Reviewed-by: Abel Vesa Reviewed-by: Konrad Dybcio Link: https://lore.kernel.org/r/20260821081540.89137-1-phollinsky@holtechnik.com Signed-off-by: Bjorn Andersson --- drivers/remoteproc/qcom_q6v5_mss.c | 11 ++++++++++- 1 file changed, 10 insertions(+), 1 deletion(-) diff --git a/drivers/remoteproc/qcom_q6v5_mss.c b/drivers/remoteproc/qcom_q6v5_mss.c index bef198b9ee63..2f71ed2feff6 100644 --- a/drivers/remoteproc/qcom_q6v5_mss.c +++ b/drivers/remoteproc/qcom_q6v5_mss.c @@ -2079,7 +2079,16 @@ static int q6v5_probe(struct platform_device *pdev) if (!desc) return -EINVAL; - if (desc->need_mem_protection && !qcom_pas_is_available()) + /* + * Memory protection is done through qcom_scm_assign_mem(), which needs + * SCM but not PAS. Only the memory setup path issues PAS calls, so + * requiring PAS for every need_mem_protection platform prevents the + * modem from probing at all on TZ firmware that offers no PAS. + */ + if (desc->need_mem_protection && !qcom_scm_is_available()) + return -EPROBE_DEFER; + + if (desc->need_pas_mem_setup && !qcom_pas_is_available()) return -EPROBE_DEFER; mba_image = desc->hexagon_mba_image; From 0d8e2195bce6f08c1c53c5ef4d7347fe46418101 Mon Sep 17 00:00:00 2001 From: Mostafa Saleh Date: Thu, 27 Aug 2026 20:30:55 +0000 Subject: [PATCH 014/592] remoteproc: qcom_q6v5_adsp: Fix iommu_unmap() usage During adsp_map_carveout, the IOVA is computed by combining the physical address and the SID: iova = adsp->mem_phys | (sid << 32); However, adsp_unmap_carveout() uses the physical address and not the IOVA in iommu_unmap(), causing the unmap to fail or leak mappings because the address doesn't match the original IOVA. Cache the constructed IOVA within the qcom_adsp device struct during mapping and use it during unmapping. Fixes: f22eedff28af ("remoteproc: qcom: Add support for memory sandbox") Signed-off-by: Mostafa Saleh Link: https://lore.kernel.org/r/20260827203055.640116-1-smostafa@google.com Signed-off-by: Bjorn Andersson --- drivers/remoteproc/qcom_q6v5_adsp.c | 8 ++++---- 1 file changed, 4 insertions(+), 4 deletions(-) diff --git a/drivers/remoteproc/qcom_q6v5_adsp.c b/drivers/remoteproc/qcom_q6v5_adsp.c index 39654206781d..2e5fb5954fa9 100644 --- a/drivers/remoteproc/qcom_q6v5_adsp.c +++ b/drivers/remoteproc/qcom_q6v5_adsp.c @@ -104,6 +104,7 @@ struct qcom_adsp { struct completion stop_done; phys_addr_t mem_phys; + unsigned long iova; phys_addr_t mem_reloc; void __iomem *mem_region; size_t mem_size; @@ -333,7 +334,7 @@ static void adsp_unmap_carveout(struct rproc *rproc) struct qcom_adsp *adsp = rproc->priv; if (adsp->has_iommu) - iommu_unmap(rproc->domain, adsp->mem_phys, adsp->mem_size); + iommu_unmap(rproc->domain, adsp->iova, adsp->mem_size); } static int adsp_map_carveout(struct rproc *rproc) @@ -341,7 +342,6 @@ static int adsp_map_carveout(struct rproc *rproc) struct qcom_adsp *adsp = rproc->priv; struct of_phandle_args args; long long sid; - unsigned long iova; int ret; if (!adsp->has_iommu) @@ -358,9 +358,9 @@ static int adsp_map_carveout(struct rproc *rproc) of_node_put(args.np); /* Add SID configuration for ADSP Firmware to SMMU */ - iova = adsp->mem_phys | (sid << 32); + adsp->iova = adsp->mem_phys | (sid << 32); - ret = iommu_map(rproc->domain, iova, adsp->mem_phys, + ret = iommu_map(rproc->domain, adsp->iova, adsp->mem_phys, adsp->mem_size, IOMMU_READ | IOMMU_WRITE, GFP_KERNEL); if (ret) { From 9db31edf92dde1f3c98008bd5f88e859d14324cc Mon Sep 17 00:00:00 2001 From: Vignesh Viswanathan Date: Wed, 19 Aug 2026 11:00:30 +0530 Subject: [PATCH 015/592] remoteproc: qcom_q6v5_pas: Fix error masking in qcom_pas_stop() In qcom_pas_stop function, return value of qcom_pas_shutdown for pas_id is overwritten by the return value of qcom_pas_shutdown for dtb_pas_id. This causes errors seen on qcom_pas_shutdown failures for pas_id to be masked to the caller. This might lead to issues where the memory regions locked by PAS, as part of qcom_pas_auth_and_reset, are not released for access by linux and rproc_coredump flow will end up accessing the locked memory, leading to an access violation. Fix this by using a separate variable for the dtb_pas_id shutdown call and only overriding the main return value if the pas_id shutdown succeeded but dtb_pas_id shutdown failed. Fixes: 29814986b82e ("remoteproc: qcom_q6v5_pas: add support for dtb co-firmware loading") Signed-off-by: Vignesh Viswanathan Reviewed-by: Konrad Dybcio Reviewed-by: Mukesh Ojha Link: https://lore.kernel.org/r/20260819-rproc_dtb_fix-v1-1-cb6be7f93bc9@oss.qualcomm.com Signed-off-by: Bjorn Andersson --- drivers/remoteproc/qcom_q6v5_pas.c | 10 +++++++--- 1 file changed, 7 insertions(+), 3 deletions(-) diff --git a/drivers/remoteproc/qcom_q6v5_pas.c b/drivers/remoteproc/qcom_q6v5_pas.c index 8cff0684fe2f..a005546c265d 100644 --- a/drivers/remoteproc/qcom_q6v5_pas.c +++ b/drivers/remoteproc/qcom_q6v5_pas.c @@ -405,6 +405,7 @@ static int qcom_pas_stop(struct rproc *rproc) { struct qcom_pas *pas = rproc->priv; int handover; + int dtb_ret; int ret; ret = qcom_q6v5_request_stop(&pas->q6v5, pas->sysmon); @@ -419,9 +420,12 @@ static int qcom_pas_stop(struct rproc *rproc) dev_err(pas->dev, "failed to shutdown: %d\n", ret); if (pas->dtb_pas_id) { - ret = qcom_pas_shutdown(pas->dtb_pas_id); - if (ret) - dev_err(pas->dev, "failed to shutdown dtb: %d\n", ret); + dtb_ret = qcom_pas_shutdown(pas->dtb_pas_id); + if (dtb_ret) + dev_err(pas->dev, "failed to shutdown dtb: %d\n", dtb_ret); + + if (!ret && dtb_ret) + ret = dtb_ret; qcom_pas_unmap_carveout(rproc, pas->dtb_mem_phys, pas->dtb_mem_size); } From 5f8cb07d7a8573753c5d31dc76b51cab8e5e9460 Mon Sep 17 00:00:00 2001 From: Dinh Nguyen Date: Fri, 28 Aug 2026 10:31:29 +0800 Subject: [PATCH 016/592] arm64: dts: socfpga: change access permission from 755 to 644 These files have an incorrect access permission of 755 instead of 644. Change them to the correct access permission of 644. Fixes: 4bc04eb90b7c ("arm64: dts: socfpga: stratix10: Add emmc support") Cc: # v7.1+ Signed-off-by: Dinh Nguyen --- arch/arm64/boot/dts/altera/socfpga_stratix10_socdk.dtsi | 0 arch/arm64/boot/dts/altera/socfpga_stratix10_socdk_emmc.dts | 0 2 files changed, 0 insertions(+), 0 deletions(-) mode change 100755 => 100644 arch/arm64/boot/dts/altera/socfpga_stratix10_socdk.dtsi mode change 100755 => 100644 arch/arm64/boot/dts/altera/socfpga_stratix10_socdk_emmc.dts diff --git a/arch/arm64/boot/dts/altera/socfpga_stratix10_socdk.dtsi b/arch/arm64/boot/dts/altera/socfpga_stratix10_socdk.dtsi old mode 100755 new mode 100644 diff --git a/arch/arm64/boot/dts/altera/socfpga_stratix10_socdk_emmc.dts b/arch/arm64/boot/dts/altera/socfpga_stratix10_socdk_emmc.dts old mode 100755 new mode 100644 From 65320b642025c53063372bd34a376bec7bf15598 Mon Sep 17 00:00:00 2001 From: Xixin Liu Date: Tue, 28 Jul 2026 08:50:00 +0800 Subject: [PATCH 017/592] firmware: arm_scpi: fix device_node leak in scpi_dev_domain_id of_parse_phandle_with_args() takes a reference on clkspec.np that must be released with of_node_put(). scpi_dev_domain_id() returned clkspec.args[0] without dropping that reference, so every domain lookup leaked a device node. Paths such as scpi_dvfs_info() / cpufreq init call this per CPU, so the leak accumulates over time. Save the domain id, of_node_put(clkspec.np), then return the saved value. Signed-off-by: Xixin Liu Link: https://patch.msgid.link/84fdd490495b.v2.1785200642.git.liuxixin@kylinos.cn Signed-off-by: Sudeep Holla --- drivers/firmware/arm_scpi.c | 5 ++++- 1 file changed, 4 insertions(+), 1 deletion(-) diff --git a/drivers/firmware/arm_scpi.c b/drivers/firmware/arm_scpi.c index 2acad5fa5a28..43d6d9bbc7a9 100644 --- a/drivers/firmware/arm_scpi.c +++ b/drivers/firmware/arm_scpi.c @@ -661,12 +661,15 @@ static struct scpi_dvfs_info *scpi_dvfs_get_info(u8 domain) static int scpi_dev_domain_id(struct device *dev) { struct of_phandle_args clkspec; + int domain; if (of_parse_phandle_with_args(dev->of_node, "clocks", "#clock-cells", 0, &clkspec)) return -EINVAL; - return clkspec.args[0]; + domain = clkspec.args[0]; + of_node_put(clkspec.np); + return domain; } static struct scpi_dvfs_info *scpi_dvfs_info(struct device *dev) From 32471d84a487c7fd74532bc96be56f8028cf4a3f Mon Sep 17 00:00:00 2001 From: Xixin Liu Date: Tue, 28 Jul 2026 08:50:00 +0800 Subject: [PATCH 018/592] firmware: arm_scpi: reject DVFS OPP count above MAX_DVFS_OPPS scpi_dvfs_get_info() already rejected a zero opp_count, but still trusted any larger value from the SCP firmware. The shared-memory reply only holds MAX_DVFS_OPPS entries in buf.opps[]; a bigger count over-reads that array and then sizes the allocated OPP table incorrectly (garbage OPPs / OOB). The missing upper bound dates back to the original SCPI DVFS support. Reject zero and out-of-range counts in one check and return -EINVAL. Fixes: 8cb7cf56c9fe ("firmware: add support for ARM System Control and Power Interface(SCPI) protocol") Signed-off-by: Xixin Liu Link: https://patch.msgid.link/022802f0b38f.v2.1785200642.git.liuxixin@kylinos.cn Signed-off-by: Sudeep Holla --- drivers/firmware/arm_scpi.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/drivers/firmware/arm_scpi.c b/drivers/firmware/arm_scpi.c index 43d6d9bbc7a9..68a730d22037 100644 --- a/drivers/firmware/arm_scpi.c +++ b/drivers/firmware/arm_scpi.c @@ -631,8 +631,8 @@ static struct scpi_dvfs_info *scpi_dvfs_get_info(u8 domain) if (ret) return ERR_PTR(ret); - if (!buf.opp_count) - return ERR_PTR(-ENOENT); + if (!buf.opp_count || buf.opp_count > MAX_DVFS_OPPS) + return ERR_PTR(-EINVAL); info = kmalloc_obj(*info); if (!info) From 70f4b78d560e592cbf3325b162424737d032fc1d Mon Sep 17 00:00:00 2001 From: Xixin Liu Date: Tue, 28 Jul 2026 08:50:00 +0800 Subject: [PATCH 019/592] clk: scpi: bound-check DVFS index in scpi_dvfs_recalc_rate dvfs_get_idx() may return an out-of-range index if the SCP firmware is buggy or returns a stale value. Only negative indexes were rejected, so a large index walked past info->opps and could treat garbage as a clock rate (KASAN OOB / wrong frequency to consumers). The missing upper bound dates back to the original SCPI clock driver. Treat indexes >= opp count as invalid and return 0, same as idx < 0. Fixes: cd52c2a4b5c4 ("clk: add support for clocks provided by SCP(System Control Processor)") Signed-off-by: Xixin Liu Link: https://patch.msgid.link/04f9ab766e07.v2.1785200642.git.liuxixin@kylinos.cn Signed-off-by: Sudeep Holla --- drivers/clk/clk-scpi.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/drivers/clk/clk-scpi.c b/drivers/clk/clk-scpi.c index 2328d2abf6d8..fa4349c8178a 100644 --- a/drivers/clk/clk-scpi.c +++ b/drivers/clk/clk-scpi.c @@ -73,7 +73,7 @@ static unsigned long scpi_dvfs_recalc_rate(struct clk_hw *hw, int idx = clk->scpi_ops->dvfs_get_idx(clk->id); const struct scpi_opp *opp; - if (idx < 0) + if (idx < 0 || idx >= clk->info->count) return 0; opp = clk->info->opps + idx; From ab06cf8152dace327cd873188e4a036c4e0b5944 Mon Sep 17 00:00:00 2001 From: Xixin Liu Date: Tue, 28 Jul 2026 08:50:00 +0800 Subject: [PATCH 020/592] clk: scpi: register scpi-cpufreq once and clear on failure scpi_clocks_probe() walks clock children and, for each DVFS provider, calls platform_device_register_simple("scpi-cpufreq", -1, ...). Two related bugs: Since all DVFS providers register the fixed scpi-cpufreq device using PLATFORM_DEVID_NONE, a second registration fails with -EEXIST and overwrites the pointer to the successfully registered device. The first device can then no longer be unregistered. Register the virtual device only once. If registration fails, reset the pointer to NULL so a subsequent DVFS provider can retry and the global pointer only represents a successfully registered device. Fixes: 9490f01e2471 ("clk: scpi: add support for cpufreq virtual device") Fixes: 67bcc2c5f1da ("clk: scpi: don't add cpufreq device if the scpi dvfs node is disabled") Signed-off-by: Xixin Liu Link: https://patch.msgid.link/fd1b9199a9c3.v2.1785200642.git.liuxixin@kylinos.cn (sudeep.holla: reworded the commit message to improve readability) Signed-off-by: Sudeep Holla --- drivers/clk/clk-scpi.c | 6 +++++- 1 file changed, 5 insertions(+), 1 deletion(-) diff --git a/drivers/clk/clk-scpi.c b/drivers/clk/clk-scpi.c index fa4349c8178a..f8182175b483 100644 --- a/drivers/clk/clk-scpi.c +++ b/drivers/clk/clk-scpi.c @@ -272,10 +272,14 @@ static int scpi_clocks_probe(struct platform_device *pdev) if (match->data != &scpi_dvfs_ops) continue; /* Add the virtual cpufreq device if it's DVFS clock provider */ + if (cpufreq_dev) + continue; cpufreq_dev = platform_device_register_simple("scpi-cpufreq", -1, NULL, 0); - if (IS_ERR(cpufreq_dev)) + if (IS_ERR(cpufreq_dev)) { pr_warn("unable to register cpufreq device"); + cpufreq_dev = NULL; + } } return 0; } From 86d923a882f48b047b079a293abb6ebca3eaf23f Mon Sep 17 00:00:00 2001 From: Xixin Liu Date: Tue, 28 Jul 2026 09:27:31 +0800 Subject: [PATCH 021/592] clk: scpi: use PLATFORM_DEVID_NONE for scpi-cpufreq Replace the magic -1 passed to platform_device_register_simple() with PLATFORM_DEVID_NONE. This is a readability cleanup only and does not change behavior. Signed-off-by: Xixin Liu Link: https://patch.msgid.link/fb51cfbfbb41.v2.1785200642.git.liuxixin@kylinos.cn Signed-off-by: Sudeep Holla --- drivers/clk/clk-scpi.c | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-) diff --git a/drivers/clk/clk-scpi.c b/drivers/clk/clk-scpi.c index f8182175b483..b2d412ca1d22 100644 --- a/drivers/clk/clk-scpi.c +++ b/drivers/clk/clk-scpi.c @@ -275,7 +275,8 @@ static int scpi_clocks_probe(struct platform_device *pdev) if (cpufreq_dev) continue; cpufreq_dev = platform_device_register_simple("scpi-cpufreq", - -1, NULL, 0); + PLATFORM_DEVID_NONE, + NULL, 0); if (IS_ERR(cpufreq_dev)) { pr_warn("unable to register cpufreq device"); cpufreq_dev = NULL; From 8b852965b8eaf910c314dc346967ed82c8d4f235 Mon Sep 17 00:00:00 2001 From: =?UTF-8?q?Iv=C3=A1n=20Ezequiel=20Rodriguez?= Date: Tue, 1 Sep 2026 10:06:27 -0300 Subject: [PATCH 022/592] Input: evdev - zero absinfo before partial copy in EVIOCSABS MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit The EVIOCSABS handler copies at most the user supplied ioctl size into an uninitialized on-stack struct input_absinfo: if (copy_from_user(&abs, p, min_t(size_t, size, sizeof(struct input_absinfo)))) The size comes from _IOC_SIZE() of the ioctl command and is therefore fully controlled by userspace. A short size leaves the trailing part of the structure holding whatever was on the kernel stack, and the whole structure is then stored into the device: dev->absinfo[t] = abs; EVIOCGABS hands that back to userspace, disclosing the stale stack bytes. Only the resolution field is currently cleared, which covers the legacy struct layout but not an arbitrarily short size. Zero the structure before the copy so any part not supplied by the caller reads back as zero. The existing resolution fixup is kept, since it also handles a size that partially overlaps that field. Fixes: 448cd1664a57 ("Input: evdev - rearrange ioctl handling") Cc: stable@vger.kernel.org Signed-off-by: Iván Ezequiel Rodriguez Link: https://patch.msgid.link/20260901130629.24078-2-ivanrwcm25@gmail.com Signed-off-by: Dmitry Torokhov --- drivers/input/evdev.c | 2 ++ 1 file changed, 2 insertions(+) diff --git a/drivers/input/evdev.c b/drivers/input/evdev.c index 3a718d600006..8bfaaa45e0b9 100644 --- a/drivers/input/evdev.c +++ b/drivers/input/evdev.c @@ -1229,6 +1229,8 @@ static long evdev_do_ioctl(struct file *file, unsigned int cmd, t = _IOC_NR(cmd) & ABS_MAX; + memset(&abs, 0, sizeof(abs)); + if (copy_from_user(&abs, p, min_t(size_t, size, sizeof(struct input_absinfo)))) return -EFAULT; From f84819ef8d66931ee3998fee3c4f03230f4cb6cc Mon Sep 17 00:00:00 2001 From: =?UTF-8?q?Iv=C3=A1n=20Ezequiel=20Rodriguez?= Date: Tue, 1 Sep 2026 10:06:28 -0300 Subject: [PATCH 023/592] Input: zero ff_effect before compat copy in input_ff_effect_from_user MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit In the compat path input_ff_effect_from_user() aliases the caller's native struct ff_effect with the smaller struct ff_effect_compat and copies only the compat sized prefix: compat_effect = (struct ff_effect_compat *)effect; if (copy_from_user(compat_effect, buffer, sizeof(struct ff_effect_compat))) The tail of the native structure is never written. Callers pass an uninitialized on-stack object, for example evdev_do_ioctl() for EVIOCSFF, so those bytes keep their previous stack contents. input_ff_upload() then stores the full native structure in ff->effects[id], from where a uinput based force feedback daemon can read it back via UI_BEGIN_FF_UPLOAD, disclosing kernel stack memory to userspace. Zero the effect before the compat copy. Fixes: 2d56f3a32c0e ("Input: refactor evdev 32bit compat to be shareable with uinput") Cc: stable@vger.kernel.org Signed-off-by: Iván Ezequiel Rodriguez Link: https://patch.msgid.link/20260901130629.24078-3-ivanrwcm25@gmail.com Signed-off-by: Dmitry Torokhov --- drivers/input/input-compat.c | 2 ++ 1 file changed, 2 insertions(+) diff --git a/drivers/input/input-compat.c b/drivers/input/input-compat.c index a5043193ead8..8860a0730294 100644 --- a/drivers/input/input-compat.c +++ b/drivers/input/input-compat.c @@ -76,6 +76,8 @@ int input_ff_effect_from_user(const char __user *buffer, size_t size, */ compat_effect = (struct ff_effect_compat *)effect; + memset(effect, 0, sizeof(*effect)); + if (copy_from_user(compat_effect, buffer, sizeof(struct ff_effect_compat))) return -EFAULT; From 85f080fb87ed5cd3e46121be677f52c82f26a0ab Mon Sep 17 00:00:00 2001 From: Linkai Gong Date: Tue, 1 Sep 2026 20:26:49 +0800 Subject: [PATCH 024/592] Input: cyttsp5 - clamp the HID report size before memcpy The size field comes from the device and is used as the memcpy() length into response_buf, which is CY_MAX_INPUT bytes. Fixes: 5b0c03e24a06 ("Input: Add driver for Cypress Generation 5 touchscreen") Signed-off-by: Linkai Gong Link: https://patch.msgid.link/20260901122649.1173066-1-gonglinkai@kylinos.cn Cc: stable@vger.kernel.org Signed-off-by: Dmitry Torokhov --- drivers/input/touchscreen/cyttsp5.c | 1 + 1 file changed, 1 insertion(+) diff --git a/drivers/input/touchscreen/cyttsp5.c b/drivers/input/touchscreen/cyttsp5.c index 9266c07314be..e878a02dc9b7 100644 --- a/drivers/input/touchscreen/cyttsp5.c +++ b/drivers/input/touchscreen/cyttsp5.c @@ -710,6 +710,7 @@ static irqreturn_t cyttsp5_handle_irq(int irq, void *handle) size = 2; } else { report_id = ts->input_buf[2]; + size = min(size, CY_MAX_INPUT); } switch (report_id) { From 3bb3e80faf21e432f6e6d89c55fb587c323b8813 Mon Sep 17 00:00:00 2001 From: Sudeep Holla Date: Tue, 1 Sep 2026 14:11:12 +0100 Subject: [PATCH 025/592] firmware: arm_ffa: Tear down driver during shutdown The platform core invokes a driver's shutdown callback, rather than its remove callback, while preparing devices for a normal kexec. Without a shutdown callback, the FF-A driver leaves notifications, partition devices, and the RX/TX mapping active before the replacement kernel is booted. Use ffa_remove() for shutdown so the existing cleanup runs before a normal kexec and other orderly system shutdowns. Reported-by: Nat Gurumoorthy Closes: https://lore.kernel.org/all/20260729162731.1383875-1-natg@google.com/ Reported-by: Carol L Soto Closes: https://lore.kernel.org/all/20260818224404.3694580-1-csoto@nvidia.com Reported-by: Maxi Saparov Closes: https://lore.kernel.org/all/20260826222337.73480-1-maxi.saparov@gmail.com Link: https://patch.msgid.link/20260901131112.3437516-1-sudeep.holla@kernel.org Tested-by: Carol L Soto Signed-off-by: Sudeep Holla --- drivers/firmware/arm_ffa/driver.c | 1 + 1 file changed, 1 insertion(+) diff --git a/drivers/firmware/arm_ffa/driver.c b/drivers/firmware/arm_ffa/driver.c index 8654b3365c9b..28abc808fd32 100644 --- a/drivers/firmware/arm_ffa/driver.c +++ b/drivers/firmware/arm_ffa/driver.c @@ -2225,6 +2225,7 @@ static void ffa_remove(struct platform_device *pdev) static struct platform_driver ffa_driver = { .probe = ffa_probe, .remove = ffa_remove, + .shutdown = ffa_remove, .driver = { .name = FFA_PLATFORM_NAME, }, From 6d49beec658f61801e79019fd73691b17252dee3 Mon Sep 17 00:00:00 2001 From: Ian Luites Date: Mon, 31 Aug 2026 08:25:47 +0000 Subject: [PATCH 026/592] soundwire: dmi-quirks: Disable ghost Realtek on Asus GX651AX The Asus ROG Zephyrus Duo GX651AX exposes a Realtek RT722 device in ACPI which does not exist in the physical hardware. The device remains unattached while the CS42L43 and both CS35L56 devices attach successfully. This confuses the function topology machine driver into creating duplicate DAI links named SDW3-Playback-SimpleJack, and the sof_sdw probe fails with error -12. Add a model-specific quirk to remove the ghost RT722 device. Fixes: 45cf24da0a10 ("ASoC: Intel: soc-acpi-intel-ptl-match: Remove unnecessary cs42l43 match") Cc: stable@vger.kernel.org # 7.2.x Assisted-by: LLM Signed-off-by: Ian Luites Reviewed-by: Charles Keepax Link: https://patch.msgid.link/20260831082534.224716-1-ian@luites.com Signed-off-by: Vinod Koul --- drivers/soundwire/dmi-quirks.c | 7 +++++++ 1 file changed, 7 insertions(+) diff --git a/drivers/soundwire/dmi-quirks.c b/drivers/soundwire/dmi-quirks.c index 768255dd12db..62fa64b22412 100644 --- a/drivers/soundwire/dmi-quirks.c +++ b/drivers/soundwire/dmi-quirks.c @@ -200,6 +200,13 @@ static const struct dmi_system_id adr_remap_quirk_table[] = { }, .driver_data = (void *)ghost_realtek, }, + { + .matches = { + DMI_MATCH(DMI_SYS_VENDOR, "ASUS"), + DMI_MATCH(DMI_BOARD_NAME, "GX651AX"), + }, + .driver_data = (void *)ghost_realtek, + }, { .matches = { DMI_MATCH(DMI_SYS_VENDOR, "ASUS"), From 33da68f61d25ef8411489d06514ff627c1f88152 Mon Sep 17 00:00:00 2001 From: Ovidiu Panait Date: Wed, 22 Jul 2026 08:53:49 +0000 Subject: [PATCH 027/592] arm64: dts: renesas: r9a09g057: Switch GBETH TX queue scheduling to WRR The GBETH ethernet nodes don't specify a TX scheduling policy, so stmmac falls back to Strict Priority. In this configuration the queue with the highest priority gets all the traffic, starving the others under load. Under sustained UDP TX load with multiple data streams, this starvation triggers spurious adapter resets due to TX queue timeouts: iperf3 -c -i0 -t60 --bind-dev end0 -u -b0 -P4 end0: NETDEV WATCHDOG: CPU: 1: transmit queue 1 timed out 5228 ms end0: Reset adapter. Investigation shows that only the highest priority queue is advancing while the others stall for more than 5 seconds, causing a netdev watchdog reset. Switch the TX scheduling policy to Weighted-Round-Robin (WRR) so that traffic is processed across all queues, eliminating the stalls. Fixes: 050ee38d0002 ("arm64: dts: renesas: r9a09g057: Add GBETH nodes") Signed-off-by: Ovidiu Panait Reviewed-by: Geert Uytterhoeven Link: https://patch.msgid.link/20260722085353.136986-2-ovidiu.panait.rb@renesas.com Signed-off-by: Geert Uytterhoeven --- arch/arm64/boot/dts/renesas/r9a09g057.dtsi | 10 ++++++++++ 1 file changed, 10 insertions(+) diff --git a/arch/arm64/boot/dts/renesas/r9a09g057.dtsi b/arch/arm64/boot/dts/renesas/r9a09g057.dtsi index 639693d464a7..188ce9f9c7c2 100644 --- a/arch/arm64/boot/dts/renesas/r9a09g057.dtsi +++ b/arch/arm64/boot/dts/renesas/r9a09g057.dtsi @@ -1715,23 +1715,28 @@ queue3 { mtl_tx_setup0: tx-queues-config { snps,tx-queues-to-use = <4>; + snps,tx-sched-wrr; queue0 { + snps,weight = <0x10>; snps,dcb-algorithm; snps,priority = <0x1>; }; queue1 { + snps,weight = <0x12>; snps,dcb-algorithm; snps,priority = <0x2>; }; queue2 { + snps,weight = <0x14>; snps,dcb-algorithm; snps,priority = <0x4>; }; queue3 { + snps,weight = <0x18>; snps,dcb-algorithm; snps,priority = <0x8>; }; @@ -1816,23 +1821,28 @@ queue3 { mtl_tx_setup1: tx-queues-config { snps,tx-queues-to-use = <4>; + snps,tx-sched-wrr; queue0 { + snps,weight = <0x10>; snps,dcb-algorithm; snps,priority = <0x1>; }; queue1 { + snps,weight = <0x12>; snps,dcb-algorithm; snps,priority = <0x2>; }; queue2 { + snps,weight = <0x14>; snps,dcb-algorithm; snps,priority = <0x4>; }; queue3 { + snps,weight = <0x18>; snps,dcb-algorithm; snps,priority = <0x8>; }; From 66fcbdbeca0118b8aeac218b33fa18c394513543 Mon Sep 17 00:00:00 2001 From: Ovidiu Panait Date: Wed, 22 Jul 2026 08:53:50 +0000 Subject: [PATCH 028/592] arm64: dts: renesas: r9a09g056: Switch GBETH TX queue scheduling to WRR The GBETH ethernet nodes don't specify a TX scheduling policy, so stmmac falls back to Strict Priority. In this configuration the queue with the highest priority gets all the traffic, starving the others under load. Under sustained UDP TX load with multiple data streams, this starvation triggers spurious adapter resets due to TX queue timeouts: iperf3 -c -i0 -t60 --bind-dev end0 -u -b0 -P4 end0: NETDEV WATCHDOG: CPU: 1: transmit queue 1 timed out 5228 ms end0: Reset adapter. Investigation shows that only the highest priority queue is advancing while the others stall for more than 5 seconds, causing a netdev watchdog reset. Switch the TX scheduling policy to Weighted-Round-Robin (WRR) so that traffic is processed across all queues, eliminating the stalls. Fixes: c8c8a57c5b40 ("arm64: dts: renesas: r9a09g056: Add GBETH nodes") Signed-off-by: Ovidiu Panait Reviewed-by: Geert Uytterhoeven Link: https://patch.msgid.link/20260722085353.136986-3-ovidiu.panait.rb@renesas.com Signed-off-by: Geert Uytterhoeven --- arch/arm64/boot/dts/renesas/r9a09g056.dtsi | 10 ++++++++++ 1 file changed, 10 insertions(+) diff --git a/arch/arm64/boot/dts/renesas/r9a09g056.dtsi b/arch/arm64/boot/dts/renesas/r9a09g056.dtsi index 76fa34ff3d07..77c2221a9e2a 100644 --- a/arch/arm64/boot/dts/renesas/r9a09g056.dtsi +++ b/arch/arm64/boot/dts/renesas/r9a09g056.dtsi @@ -1585,23 +1585,28 @@ queue3 { mtl_tx_setup0: tx-queues-config { snps,tx-queues-to-use = <4>; + snps,tx-sched-wrr; queue0 { + snps,weight = <0x10>; snps,dcb-algorithm; snps,priority = <0x1>; }; queue1 { + snps,weight = <0x12>; snps,dcb-algorithm; snps,priority = <0x2>; }; queue2 { + snps,weight = <0x14>; snps,dcb-algorithm; snps,priority = <0x4>; }; queue3 { + snps,weight = <0x18>; snps,dcb-algorithm; snps,priority = <0x8>; }; @@ -1686,23 +1691,28 @@ queue3 { mtl_tx_setup1: tx-queues-config { snps,tx-queues-to-use = <4>; + snps,tx-sched-wrr; queue0 { + snps,weight = <0x10>; snps,dcb-algorithm; snps,priority = <0x1>; }; queue1 { + snps,weight = <0x12>; snps,dcb-algorithm; snps,priority = <0x2>; }; queue2 { + snps,weight = <0x14>; snps,dcb-algorithm; snps,priority = <0x4>; }; queue3 { + snps,weight = <0x18>; snps,dcb-algorithm; snps,priority = <0x8>; }; From 63016c3a91f2c458ca75869c8c782e899591f22d Mon Sep 17 00:00:00 2001 From: Ovidiu Panait Date: Wed, 22 Jul 2026 08:53:51 +0000 Subject: [PATCH 029/592] arm64: dts: renesas: r9a09g047: Switch GBETH TX queue scheduling to WRR The GBETH ethernet nodes don't specify a TX scheduling policy, so stmmac falls back to Strict Priority. In this configuration the queue with the highest priority gets all the traffic, starving the others under load. Under sustained UDP TX load with multiple data streams, this starvation triggers spurious adapter resets due to TX queue timeouts: iperf3 -c -i0 -t60 --bind-dev end0 -u -b0 -P4 end0: NETDEV WATCHDOG: CPU: 1: transmit queue 1 timed out 5228 ms end0: Reset adapter. Investigation shows that only the highest priority queue is advancing while the others stall for more than 5 seconds, causing a netdev watchdog reset. Switch the TX scheduling policy to Weighted-Round-Robin (WRR) so that traffic is processed across all queues, eliminating the stalls. Fixes: 41ffbb1c42d3 ("arm64: dts: renesas: r9a09g047: Add GBETH nodes") Signed-off-by: Ovidiu Panait Reviewed-by: Geert Uytterhoeven Tested-by: Tommaso Merciai Link: https://patch.msgid.link/20260722085353.136986-4-ovidiu.panait.rb@renesas.com Signed-off-by: Geert Uytterhoeven --- arch/arm64/boot/dts/renesas/r9a09g047.dtsi | 10 ++++++++++ 1 file changed, 10 insertions(+) diff --git a/arch/arm64/boot/dts/renesas/r9a09g047.dtsi b/arch/arm64/boot/dts/renesas/r9a09g047.dtsi index 73757e8e2197..060405d4a2dd 100644 --- a/arch/arm64/boot/dts/renesas/r9a09g047.dtsi +++ b/arch/arm64/boot/dts/renesas/r9a09g047.dtsi @@ -1913,23 +1913,28 @@ queue3 { mtl_tx_setup0: tx-queues-config { snps,tx-queues-to-use = <4>; + snps,tx-sched-wrr; queue0 { + snps,weight = <0x10>; snps,dcb-algorithm; snps,priority = <0x1>; }; queue1 { + snps,weight = <0x12>; snps,dcb-algorithm; snps,priority = <0x2>; }; queue2 { + snps,weight = <0x14>; snps,dcb-algorithm; snps,priority = <0x4>; }; queue3 { + snps,weight = <0x18>; snps,dcb-algorithm; snps,priority = <0x8>; }; @@ -2013,23 +2018,28 @@ queue3 { mtl_tx_setup1: tx-queues-config { snps,tx-queues-to-use = <4>; + snps,tx-sched-wrr; queue0 { + snps,weight = <0x10>; snps,dcb-algorithm; snps,priority = <0x1>; }; queue1 { + snps,weight = <0x12>; snps,dcb-algorithm; snps,priority = <0x2>; }; queue2 { + snps,weight = <0x14>; snps,dcb-algorithm; snps,priority = <0x4>; }; queue3 { + snps,weight = <0x18>; snps,dcb-algorithm; snps,priority = <0x8>; }; From f9d9a1913c09366bf7b967d39575e06cb80128dc Mon Sep 17 00:00:00 2001 From: Ovidiu Panait Date: Wed, 22 Jul 2026 08:53:52 +0000 Subject: [PATCH 030/592] arm64: dts: renesas: r9a09g077: Switch GBETH TX queue scheduling to WRR The GBETH ethernet nodes don't specify a TX scheduling policy, so stmmac falls back to Strict Priority. In this configuration the queue with the highest priority gets all the traffic, starving the others under load. Under sustained UDP TX load with multiple data streams, this starvation triggers spurious adapter resets due to TX queue timeouts: iperf3 -c -i0 -t60 --bind-dev end0 -u -b0 -P4 end0: NETDEV WATCHDOG: CPU: 1: transmit queue 1 timed out 5228 ms end0: Reset adapter. Investigation shows that only the highest priority queue is advancing while the others stall for more than 5 seconds, causing a netdev watchdog reset. Switch the TX scheduling policy to Weighted-Round-Robin (WRR) so that traffic is processed across all queues, eliminating the stalls. Fixes: 394c1e24a4cf ("arm64: dts: renesas: r9a09g077: Add GMAC nodes") Signed-off-by: Ovidiu Panait Reviewed-by: Geert Uytterhoeven Link: https://patch.msgid.link/20260722085353.136986-5-ovidiu.panait.rb@renesas.com Signed-off-by: Geert Uytterhoeven --- arch/arm64/boot/dts/renesas/r9a09g077.dtsi | 27 ++++++++++++++++++++++ 1 file changed, 27 insertions(+) diff --git a/arch/arm64/boot/dts/renesas/r9a09g077.dtsi b/arch/arm64/boot/dts/renesas/r9a09g077.dtsi index 40494159831d..bac39390ead7 100644 --- a/arch/arm64/boot/dts/renesas/r9a09g077.dtsi +++ b/arch/arm64/boot/dts/renesas/r9a09g077.dtsi @@ -642,36 +642,45 @@ queue7 { mtl_tx_setup0: tx-queues-config { snps,tx-queues-to-use = <8>; + snps,tx-sched-wrr; queue0 { + snps,weight = <0x10>; snps,dcb-algorithm; }; queue1 { + snps,weight = <0x11>; snps,dcb-algorithm; }; queue2 { + snps,weight = <0x12>; snps,dcb-algorithm; }; queue3 { + snps,weight = <0x13>; snps,dcb-algorithm; }; queue4 { + snps,weight = <0x14>; snps,dcb-algorithm; }; queue5 { + snps,weight = <0x15>; snps,dcb-algorithm; }; queue6 { + snps,weight = <0x16>; snps,dcb-algorithm; }; queue7 { + snps,weight = <0x17>; snps,dcb-algorithm; }; }; @@ -788,36 +797,45 @@ queue7 { mtl_tx_setup1: tx-queues-config { snps,tx-queues-to-use = <8>; + snps,tx-sched-wrr; queue0 { + snps,weight = <0x10>; snps,dcb-algorithm; }; queue1 { + snps,weight = <0x11>; snps,dcb-algorithm; }; queue2 { + snps,weight = <0x12>; snps,dcb-algorithm; }; queue3 { + snps,weight = <0x13>; snps,dcb-algorithm; }; queue4 { + snps,weight = <0x14>; snps,dcb-algorithm; }; queue5 { + snps,weight = <0x15>; snps,dcb-algorithm; }; queue6 { + snps,weight = <0x16>; snps,dcb-algorithm; }; queue7 { + snps,weight = <0x17>; snps,dcb-algorithm; }; }; @@ -934,36 +952,45 @@ queue7 { mtl_tx_setup2: tx-queues-config { snps,tx-queues-to-use = <8>; + snps,tx-sched-wrr; queue0 { + snps,weight = <0x10>; snps,dcb-algorithm; }; queue1 { + snps,weight = <0x11>; snps,dcb-algorithm; }; queue2 { + snps,weight = <0x12>; snps,dcb-algorithm; }; queue3 { + snps,weight = <0x13>; snps,dcb-algorithm; }; queue4 { + snps,weight = <0x14>; snps,dcb-algorithm; }; queue5 { + snps,weight = <0x15>; snps,dcb-algorithm; }; queue6 { + snps,weight = <0x16>; snps,dcb-algorithm; }; queue7 { + snps,weight = <0x17>; snps,dcb-algorithm; }; }; From 2ac7bad110be6ebe478d6bd57821f7f5259a1f54 Mon Sep 17 00:00:00 2001 From: Ovidiu Panait Date: Wed, 22 Jul 2026 08:53:53 +0000 Subject: [PATCH 031/592] arm64: dts: renesas: r9a09g087: Switch GBETH TX queue scheduling to WRR The GBETH ethernet nodes don't specify a TX scheduling policy, so stmmac falls back to Strict Priority. In this configuration the queue with the highest priority gets all the traffic, starving the others under load. Under sustained UDP TX load with multiple data streams, this starvation triggers spurious adapter resets due to TX queue timeouts: iperf3 -c -i0 -t60 --bind-dev end0 -u -b0 -P4 end0: NETDEV WATCHDOG: CPU: 1: transmit queue 1 timed out 5228 ms end0: Reset adapter. Investigation shows that only the highest priority queue is advancing while the others stall for more than 5 seconds, causing a netdev watchdog reset. Switch the TX scheduling policy to Weighted-Round-Robin (WRR) so that traffic is processed across all queues, eliminating the stalls. Fixes: c4698a34993b ("arm64: dts: renesas: r9a09g087: Add GMAC nodes") Signed-off-by: Ovidiu Panait Reviewed-by: Geert Uytterhoeven Link: https://patch.msgid.link/20260722085353.136986-6-ovidiu.panait.rb@renesas.com Signed-off-by: Geert Uytterhoeven --- arch/arm64/boot/dts/renesas/r9a09g087.dtsi | 27 ++++++++++++++++++++++ 1 file changed, 27 insertions(+) diff --git a/arch/arm64/boot/dts/renesas/r9a09g087.dtsi b/arch/arm64/boot/dts/renesas/r9a09g087.dtsi index e8d4f76949cc..03b976d93e10 100644 --- a/arch/arm64/boot/dts/renesas/r9a09g087.dtsi +++ b/arch/arm64/boot/dts/renesas/r9a09g087.dtsi @@ -643,36 +643,45 @@ queue7 { mtl_tx_setup0: tx-queues-config { snps,tx-queues-to-use = <8>; + snps,tx-sched-wrr; queue0 { + snps,weight = <0x10>; snps,dcb-algorithm; }; queue1 { + snps,weight = <0x11>; snps,dcb-algorithm; }; queue2 { + snps,weight = <0x12>; snps,dcb-algorithm; }; queue3 { + snps,weight = <0x13>; snps,dcb-algorithm; }; queue4 { + snps,weight = <0x14>; snps,dcb-algorithm; }; queue5 { + snps,weight = <0x15>; snps,dcb-algorithm; }; queue6 { + snps,weight = <0x16>; snps,dcb-algorithm; }; queue7 { + snps,weight = <0x17>; snps,dcb-algorithm; }; }; @@ -790,36 +799,45 @@ queue7 { mtl_tx_setup1: tx-queues-config { snps,tx-queues-to-use = <8>; + snps,tx-sched-wrr; queue0 { + snps,weight = <0x10>; snps,dcb-algorithm; }; queue1 { + snps,weight = <0x11>; snps,dcb-algorithm; }; queue2 { + snps,weight = <0x12>; snps,dcb-algorithm; }; queue3 { + snps,weight = <0x13>; snps,dcb-algorithm; }; queue4 { + snps,weight = <0x14>; snps,dcb-algorithm; }; queue5 { + snps,weight = <0x15>; snps,dcb-algorithm; }; queue6 { + snps,weight = <0x16>; snps,dcb-algorithm; }; queue7 { + snps,weight = <0x17>; snps,dcb-algorithm; }; }; @@ -937,36 +955,45 @@ queue7 { mtl_tx_setup2: tx-queues-config { snps,tx-queues-to-use = <8>; + snps,tx-sched-wrr; queue0 { + snps,weight = <0x10>; snps,dcb-algorithm; }; queue1 { + snps,weight = <0x11>; snps,dcb-algorithm; }; queue2 { + snps,weight = <0x12>; snps,dcb-algorithm; }; queue3 { + snps,weight = <0x13>; snps,dcb-algorithm; }; queue4 { + snps,weight = <0x14>; snps,dcb-algorithm; }; queue5 { + snps,weight = <0x15>; snps,dcb-algorithm; }; queue6 { + snps,weight = <0x16>; snps,dcb-algorithm; }; queue7 { + snps,weight = <0x17>; snps,dcb-algorithm; }; }; From 90af7fde083e1b22c349c3a8b1626728e44e474c Mon Sep 17 00:00:00 2001 From: Yifei Gao Date: Tue, 4 Aug 2026 21:34:56 +0000 Subject: [PATCH 032/592] memstick: ms_block: destroy io_queue workqueue on removal msb_init_disk() creates the per-card ordered workqueue msb->io_queue with alloc_ordered_workqueue(). It is torn down with destroy_workqueue() only on the init error path; msb_remove() never destroys it. msb_stop() merely flushes the queue, and neither msb_data_clear() nor put_disk() free it. As a result every card insert/remove cycle leaks the workqueue and its kworker, exhausting kernel memory over repeated cycles. Destroy the workqueue in msb_remove() after the disk has been removed and the queue drained. Fixes: 0ab30494bc4f ("memstick: add support for legacy memorysticks") Cc: stable@vger.kernel.org Assisted-by: Claude:claude-opus-4-8 Signed-off-by: Yifei Gao Signed-off-by: Ulf Hansson --- drivers/memstick/core/ms_block.c | 2 ++ 1 file changed, 2 insertions(+) diff --git a/drivers/memstick/core/ms_block.c b/drivers/memstick/core/ms_block.c index ce33907bfc24..65154e569a9e 100644 --- a/drivers/memstick/core/ms_block.c +++ b/drivers/memstick/core/ms_block.c @@ -2204,6 +2204,8 @@ static void msb_remove(struct memstick_dev *card) msb_data_clear(msb); mutex_unlock(&msb_disk_lock); + destroy_workqueue(msb->io_queue); + put_disk(msb->disk); memstick_set_drvdata(card, NULL); } From d1fc569fcbc7be6de06b034cf954a95e30ca1fb2 Mon Sep 17 00:00:00 2001 From: Troy Mitchell Date: Mon, 27 Jul 2026 01:26:15 -0700 Subject: [PATCH 033/592] dmaengine: mmp_pdma: fix wrong extended DRCMR base for SpacemiT K3 The extended DRCMR window on SpacemiT K3 starts at 0x1100. Commit 6587b8661a0b ("dmaengine: mmp_pdma: add SpacemiT K3 support") incorrectly set it to 0x1000, causing DRCMR accesses for request IDs >= 64 to target offsets 0x100 too low. The 0x1100 base has been verified on K3 silicon using real SPI and QSPI DMA transactions. The K3 DMA documentation [1] was updated on June 24, 2026, to reflect the corrected register addresses. Drop the bogus DRCMR_EXT_BASE_K3 macro and reuse DRCMR_EXT_BASE_DEFAULT for the K3 ops. Fixes: 6587b8661a0b ("dmaengine: mmp_pdma: add SpacemiT K3 support") Link: https://www.spacemit.com/community/document/info?nodepath=hardware/key_stone/k3/k3_docs/k3_usermanual/16_peripherals/dma.md&lang=en [1] Signed-off-by: Troy Mitchell Reviewed-by: Frank Li Link: https://patch.msgid.link/20260727-k3-pdma-fix-drcmr-base-v2-1-afba55cba1f3@linux.spacemit.com Signed-off-by: Vinod Koul --- drivers/dma/mmp_pdma.c | 3 +-- 1 file changed, 1 insertion(+), 2 deletions(-) diff --git a/drivers/dma/mmp_pdma.c b/drivers/dma/mmp_pdma.c index 386e85cd4882..78e3e07e681d 100644 --- a/drivers/dma/mmp_pdma.c +++ b/drivers/dma/mmp_pdma.c @@ -52,7 +52,6 @@ #define DCSR_EORINTR BIT(9) /* The end of Receive */ #define DRCMR_BASE 0x0100 -#define DRCMR_EXT_BASE_K3 0x1000 #define DRCMR_EXT_BASE_DEFAULT 0x1100 #define DRCMR_REQ_LIMIT 64 #define DRCMR_MAPVLD BIT(7) /* Map Valid (read / write) */ @@ -1219,7 +1218,7 @@ static const struct mmp_pdma_ops spacemit_k3_pdma_ops = { .get_desc_dst_addr = get_desc_dst_addr_64, .run_bits = (DCSR_RUN | DCSR_LPAEEN | DCSR_EORIRQEN | DCSR_EORSTOPEN), .dma_width = 64, - .drcmr_ext_base = DRCMR_EXT_BASE_K3, + .drcmr_ext_base = DRCMR_EXT_BASE_DEFAULT, }; static const struct of_device_id mmp_pdma_dt_ids[] = { From 0294b6dd515256c03ea2dbf508ddd3826d788579 Mon Sep 17 00:00:00 2001 From: Alexander Chesnokov Date: Wed, 12 Aug 2026 08:34:26 +0300 Subject: [PATCH 034/592] dmaengine: ti: k3-udma-glue: fix NULL dereference in k3_udma_glue_release_rx_chn() If devm_kcalloc() for rx_chn->flows fails in a channel request function, the error path calls k3_udma_glue_release_rx_chn(), which dereferences the NULL rx_chn->flows pointer in k3_udma_glue_release_rx_flow(). Skip the flow release loop in k3_udma_glue_release_rx_chn() when rx_chn->flows is not allocated. Found by Linux Verification Center (linuxtesting.org) with SVACE. Fixes: d70241913413 ("dmaengine: ti: k3-udma: Add glue layer for non DMAengine users") Cc: stable@vger.kernel.org Reported-by: Pavel Zhigulin Signed-off-by: Alexander Chesnokov Reviewed-by: Frank Li Link: https://patch.msgid.link/20260812053426.3521589-1-Alexander.Chesnokov@kaspersky.com Signed-off-by: Vinod Koul --- drivers/dma/ti/k3-udma-glue.c | 5 +++-- 1 file changed, 3 insertions(+), 2 deletions(-) diff --git a/drivers/dma/ti/k3-udma-glue.c b/drivers/dma/ti/k3-udma-glue.c index 686dc140293e..70eaf7ee57e6 100644 --- a/drivers/dma/ti/k3-udma-glue.c +++ b/drivers/dma/ti/k3-udma-glue.c @@ -1243,8 +1243,9 @@ void k3_udma_glue_release_rx_chn(struct k3_udma_glue_rx_channel *rx_chn) rx_chn->psil_paired = false; } - for (i = 0; i < rx_chn->flow_num; i++) - k3_udma_glue_release_rx_flow(rx_chn, i); + if (rx_chn->flows) + for (i = 0; i < rx_chn->flow_num; i++) + k3_udma_glue_release_rx_flow(rx_chn, i); if (xudma_rflow_is_gp(rx_chn->common.udmax, rx_chn->flow_id_base)) xudma_free_gp_rflow_range(rx_chn->common.udmax, From a7df136ec529ee49a789c5029bc37b98b0d4bedd Mon Sep 17 00:00:00 2001 From: Ruoyu Wang Date: Thu, 13 Aug 2026 23:31:49 +0800 Subject: [PATCH 035/592] dmaengine: sprd: Fix runtime PM reference leak in probe pm_runtime_get_sync() increments a device's usage counter even when it fails. sprd_dma_probe() currently jumps directly to controller clock cleanup on that error, bypassing both pm_runtime_put_noidle() and pm_runtime_disable(). This can happen if the preceding unchecked pm_runtime_set_active() fails and the following runtime-resume attempt also returns an error. Enter the existing runtime-PM unwind path instead. This drops the reference without idling the partially initialized device, disables runtime PM, and then releases the controller clocks. The success path and propagated error code are unchanged. This issue was found by a static analysis checker and confirmed by manual source review. Fixes: 9b3b8171f7f4 ("dmaengine: sprd: Add Spreadtrum DMA driver") Signed-off-by: Ruoyu Wang Reviewed-by: Frank Li Reviewed-by: Baolin Wang Link: https://patch.msgid.link/20260813153149.3953497-1-ruoyuw560@gmail.com Signed-off-by: Vinod Koul --- drivers/dma/sprd-dma.c | 3 +-- 1 file changed, 1 insertion(+), 2 deletions(-) diff --git a/drivers/dma/sprd-dma.c b/drivers/dma/sprd-dma.c index 087fea3af2e4..19b32a23c882 100644 --- a/drivers/dma/sprd-dma.c +++ b/drivers/dma/sprd-dma.c @@ -1212,7 +1212,7 @@ static int sprd_dma_probe(struct platform_device *pdev) ret = pm_runtime_get_sync(&pdev->dev); if (ret < 0) - goto err_rpm; + goto err_register; ret = dma_async_device_register(&sdev->dma_dev); if (ret < 0) { @@ -1234,7 +1234,6 @@ static int sprd_dma_probe(struct platform_device *pdev) err_register: pm_runtime_put_noidle(&pdev->dev); pm_runtime_disable(&pdev->dev); -err_rpm: sprd_dma_disable(sdev); return ret; } From 6feadbecdae60a6324c967f3b1493741083793a3 Mon Sep 17 00:00:00 2001 From: Fan Wu Date: Thu, 6 Aug 2026 13:02:33 +0000 Subject: [PATCH 036/592] mmc: core: Cancel SDIO IRQ work before freeing host A host controller that uses sdio_signal_irq() schedules host->sdio_irq_work from its interrupt handler. That work is only cancelled on the suspend path (mmc_sdio_suspend()), not on the remove/free path, so a worker armed just before the controller freed its IRQ can run after mmc_host_classdev_release() has freed the host and dereference it through container_of(). Cancel host->sdio_irq_work in mmc_free_host(), like the existing host->detect drain added by commit 1036f69e2513 ("mmc: core: Cancel delayed work before releasing host"). This issue was found by an in-house static analysis tool. Fixes: 682696605c70 ("mmc: sdio: Add API to manage SDIO IRQs from a workqueue") Cc: stable@vger.kernel.org Assisted-by: Codex:gpt-5.6 Signed-off-by: Fan Wu Signed-off-by: Ulf Hansson --- drivers/mmc/core/host.c | 1 + 1 file changed, 1 insertion(+) diff --git a/drivers/mmc/core/host.c b/drivers/mmc/core/host.c index b7ce3137d452..828078b3f962 100644 --- a/drivers/mmc/core/host.c +++ b/drivers/mmc/core/host.c @@ -698,6 +698,7 @@ EXPORT_SYMBOL(mmc_remove_host); void mmc_free_host(struct mmc_host *host) { cancel_delayed_work_sync(&host->detect); + cancel_work_sync(&host->sdio_irq_work); mmc_pwrseq_free(host); put_device(&host->class_dev); } From 2b19cf3e50cddaff07b657dae1a8f30f06032852 Mon Sep 17 00:00:00 2001 From: Fan Wu Date: Fri, 14 Aug 2026 08:25:50 +0000 Subject: [PATCH 037/592] mmc: mmci: Fix use-after-free in busy-timeout work ux500_busy_complete() can queue ux500_busy_timeout_work for an R1b command, but mmci_remove() never cancels it. The work can subsequently dereference the devm-allocated mmci_host after it has been released. Mask the controller interrupts and disable the delayed work during removal. This drains any queued instance and stops an IRQ handler that is still in progress from queueing the work again once it has been disabled. This issue was found by an in-house static analysis tool. Fixes: b1a665932dc2 ("mmc: mmci: Add support for SW busy-end timeouts") Cc: stable@vger.kernel.org # v6.10+ Assisted-by: Codex:gpt-5.6 Signed-off-by: Fan Wu Reviewed-by: Linus Walleij Signed-off-by: Ulf Hansson --- drivers/mmc/host/mmci.c | 3 +++ 1 file changed, 3 insertions(+) diff --git a/drivers/mmc/host/mmci.c b/drivers/mmc/host/mmci.c index e500051bd572..416bdb184ed4 100644 --- a/drivers/mmc/host/mmci.c +++ b/drivers/mmc/host/mmci.c @@ -2511,6 +2511,9 @@ static void mmci_remove(struct amba_device *dev) writel(0, host->base + MMCICOMMAND); writel(0, host->base + MMCIDATACTRL); + if (variant->busy_detect) + disable_delayed_work_sync(&host->ux500_busy_timeout_work); + mmci_dma_release(host); clk_disable_unprepare(host->clk); } From d3a421c82412344022982d5b91ba23194a0a6f29 Mon Sep 17 00:00:00 2001 From: Fan Wu Date: Fri, 7 Aug 2026 03:26:54 +0000 Subject: [PATCH 038/592] mmc: mxcmmc: cancel data work and watchdog on remove mxcmci_remove() frees the host through the devm tail, but neither it nor mmc_remove_host() drains the driver's own asynchronous state. host->watchdog, a 10 s timer armed on the DMA path in mxcmci_setup_data(), is deleted only by the DMA- and IRQ-complete paths, which the remove path does not explicitly drain; it can therefore fire after the host is freed and dereference it in mxcmci_watchdog(). host->datawork, armed from the IRQ handler on the PIO path, is not cancelled by the remove path either. Free the devm-registered IRQ, then cancel datawork and delete the watchdog in mxcmci_remove(), before dma_release_channel(). Freeing the IRQ first keeps a trailing handler from re-arming datawork between the cancel and the host free. Both callbacks are non-self-rearming. This issue was found by an in-house static analysis tool. Fixes: f6ad0a481342 ("mmc: mxcmmc: fix bug that may block a data transfer forever") Cc: stable@vger.kernel.org Assisted-by: Codex:gpt-5.6 Signed-off-by: Fan Wu Signed-off-by: Ulf Hansson --- drivers/mmc/host/mxcmmc.c | 4 ++++ 1 file changed, 4 insertions(+) diff --git a/drivers/mmc/host/mxcmmc.c b/drivers/mmc/host/mxcmmc.c index c405cfb8b269..097498a3f8ff 100644 --- a/drivers/mmc/host/mxcmmc.c +++ b/drivers/mmc/host/mxcmmc.c @@ -1173,6 +1173,10 @@ static void mxcmci_remove(struct platform_device *pdev) mmc_remove_host(mmc); + devm_free_irq(&pdev->dev, platform_get_irq(pdev, 0), host); + cancel_work_sync(&host->datawork); + timer_delete_sync(&host->watchdog); + if (host->pdata && host->pdata->exit) host->pdata->exit(&pdev->dev, mmc); From 5d132990475f02cfa1debe03d50b479432864ebd Mon Sep 17 00:00:00 2001 From: Fan Wu Date: Fri, 14 Aug 2026 08:23:54 +0000 Subject: [PATCH 039/592] mmc: hsq: Fix use-after-free in retry work mmc_hsq_pump_requests() queues retry_work when request_atomic() returns -EBUSY; today sdhci-sprd is the only consumer that implements request_atomic(). The work is embedded in a devm-allocated mmc_hsq, but is never cancelled during driver removal. Work still pending at unbind can therefore run after the devm allocation has been released and dereference hsq->mmc and hsq->mrq. Use devm_work_autocancel() to cancel and drain retry_work before the devm allocation is released. By the time devres cleanup begins, mmc_remove_host() has already stopped the host, so no new requests can arm the work. This issue was found by an in-house static analysis tool. Fixes: 6db96e5810e0 ("mmc: host: Introduce the request_atomic() for the host") Cc: stable@vger.kernel.org Assisted-by: Codex:gpt-5.6 Signed-off-by: Fan Wu Signed-off-by: Ulf Hansson --- drivers/mmc/host/mmc_hsq.c | 8 +++++++- 1 file changed, 7 insertions(+), 1 deletion(-) diff --git a/drivers/mmc/host/mmc_hsq.c b/drivers/mmc/host/mmc_hsq.c index 79836705c176..57e172bd3487 100644 --- a/drivers/mmc/host/mmc_hsq.c +++ b/drivers/mmc/host/mmc_hsq.c @@ -7,6 +7,7 @@ * Author: Baolin Wang */ +#include #include #include #include @@ -345,6 +346,7 @@ static const struct mmc_cqe_ops mmc_hsq_ops = { int mmc_hsq_init(struct mmc_hsq *hsq, struct mmc_host *mmc) { + int ret; int i; hsq->num_slots = HSQ_NUM_SLOTS; hsq->next_tag = HSQ_INVALID_TAG; @@ -363,7 +365,11 @@ int mmc_hsq_init(struct mmc_hsq *hsq, struct mmc_host *mmc) for (i = 0; i < HSQ_NUM_SLOTS; i++) hsq->tag_slot[i] = HSQ_INVALID_TAG; - INIT_WORK(&hsq->retry_work, mmc_hsq_retry_handler); + ret = devm_work_autocancel(mmc_dev(mmc), &hsq->retry_work, + mmc_hsq_retry_handler); + if (ret) + return ret; + spin_lock_init(&hsq->lock); init_waitqueue_head(&hsq->wait_queue); From ff894dced1a7ad7523f9c65dbdb53d02474cca0f Mon Sep 17 00:00:00 2001 From: "Diogo Ivo (Schneider Electric)" Date: Fri, 7 Aug 2026 13:06:57 +0200 Subject: [PATCH 040/592] mmc: sdhci_am654: Move tuning_loop to local variable The tuning_loop field in struct sdhci_am654_data is only used within sdhci_am654_platform_execute_tuning() as a loop counter that is initialized to 0 in sdhci_am654_init(). Since it shouldn't persist across function calls, otherwise every failure expends its "budget", move it to a local variable and remove the struct field along with the now-unnecessary initialization. Signed-off-by: Diogo Ivo (Schneider Electric) Reviewed-by: Judith Mendez Acked-by: Adrian Hunter Fixes: de31f6ab68a3 ("mmc: sdhci_am654: Reset Command and Data line after tuning") Cc: stable@vger.kernel.org Signed-off-by: Ulf Hansson --- drivers/mmc/host/sdhci_am654.c | 7 ++----- 1 file changed, 2 insertions(+), 5 deletions(-) diff --git a/drivers/mmc/host/sdhci_am654.c b/drivers/mmc/host/sdhci_am654.c index 2a27db2f558b..4b74a4115509 100644 --- a/drivers/mmc/host/sdhci_am654.c +++ b/drivers/mmc/host/sdhci_am654.c @@ -151,7 +151,6 @@ struct sdhci_am654_data { u32 flags; u32 quirks; bool dll_enable; - u32 tuning_loop; #define SDHCI_AM654_QUIRK_FORCE_CDTEST BIT(0) #define SDHCI_AM654_QUIRK_SUPPRESS_V1P8_ENA BIT(1) @@ -576,13 +575,14 @@ static int sdhci_am654_platform_execute_tuning(struct sdhci_host *host, struct sdhci_am654_data *sdhci_am654 = sdhci_pltfm_priv(pltfm_host); unsigned char timing = host->mmc->ios.timing; struct device *dev = mmc_dev(host->mmc); + unsigned int tuning_loop = 0; int itapdly; do { itapdly = sdhci_am654_do_tuning(host, opcode); if (itapdly >= 0) break; - } while (++sdhci_am654->tuning_loop < RETRY_TUNING_MAX); + } while (++tuning_loop < RETRY_TUNING_MAX); if (itapdly < 0) { dev_err(dev, "Failed to find itapdly, fail tuning\n"); @@ -806,9 +806,6 @@ static int sdhci_am654_init(struct sdhci_host *host) regmap_update_bits(sdhci_am654->base, CTL_CFG_3, TUNINGFORSDR50_MASK, TUNINGFORSDR50_MASK); - /* Use to re-execute tuning */ - sdhci_am654->tuning_loop = 0; - ret = sdhci_setup_host(host); if (ret) return ret; From 7197d9107d9545730153b82ea5a411c5208b443f Mon Sep 17 00:00:00 2001 From: "Diogo Ivo (Schneider Electric)" Date: Fri, 7 Aug 2026 13:06:58 +0200 Subject: [PATCH 041/592] mmc: sdhci_am654: Reset command and data lines on failed tuning The CMD/DATA reset after tuning should be performed regardless of whether tuning succeeded or failed, since tuning data may remain in the buffer in either case. Move the error return after the reset so that the controller is always cleaned up. Fixes: de31f6ab68a3 ("mmc: sdhci_am654: Reset Command and Data line after tuning") Cc: stable@vger.kernel.org Signed-off-by: Diogo Ivo (Schneider Electric) Reviewed-by: Judith Mendez Acked-by: Adrian Hunter Signed-off-by: Ulf Hansson --- drivers/mmc/host/sdhci_am654.c | 4 +--- 1 file changed, 1 insertion(+), 3 deletions(-) diff --git a/drivers/mmc/host/sdhci_am654.c b/drivers/mmc/host/sdhci_am654.c index 4b74a4115509..6abe206ed292 100644 --- a/drivers/mmc/host/sdhci_am654.c +++ b/drivers/mmc/host/sdhci_am654.c @@ -442,15 +442,13 @@ static int sdhci_am654_execute_tuning(struct mmc_host *mmc, u32 opcode) struct sdhci_host *host = mmc_priv(mmc); int err = sdhci_execute_tuning(mmc, opcode); - if (err) - return err; /* * Tuning data remains in the buffer after tuning. * Do a command and data reset to get rid of it */ sdhci_reset(host, SDHCI_RESET_CMD | SDHCI_RESET_DATA); - return 0; + return err; } static u32 sdhci_am654_cqhci_irq(struct sdhci_host *host, u32 intmask) From c9f47cc8c37f7659897142ffe216c250fbc1d4ed Mon Sep 17 00:00:00 2001 From: "Diogo Ivo (Schneider Electric)" Date: Fri, 7 Aug 2026 13:06:59 +0200 Subject: [PATCH 042/592] mmc: sdhci_am654: Clear ITAPDLY on tuning failure When tuning fails, stale ITAPDLY values can persist and interfere with subsequent I/O accesses, for example in DDR50 mode in cards with no tuning support. Move the ITAPDLY enable setting out of the tuning loop to after successful tuning, and explicitly clear ITAPDLY (delay and enable) when tuning fails so that we are sure only working values are actually left in hardware. Fixes: 901d16e46296 ("mmc: sdhci_am654: Add retry tuning") Cc: stable@vger.kernel.org Signed-off-by: Diogo Ivo (Schneider Electric) Reviewed-by: Judith Mendez Acked-by: Adrian Hunter Signed-off-by: Ulf Hansson --- drivers/mmc/host/sdhci_am654.c | 12 +++++++----- 1 file changed, 7 insertions(+), 5 deletions(-) diff --git a/drivers/mmc/host/sdhci_am654.c b/drivers/mmc/host/sdhci_am654.c index 6abe206ed292..49a46583c8bc 100644 --- a/drivers/mmc/host/sdhci_am654.c +++ b/drivers/mmc/host/sdhci_am654.c @@ -527,7 +527,6 @@ static int sdhci_am654_do_tuning(struct sdhci_host *host, { struct sdhci_pltfm_host *pltfm_host = sdhci_priv(host); struct sdhci_am654_data *sdhci_am654 = sdhci_pltfm_priv(pltfm_host); - unsigned char timing = host->mmc->ios.timing; struct window fail_window[ITAPDLY_LENGTH]; struct device *dev = mmc_dev(host->mmc); u8 curr_pass, itap; @@ -536,11 +535,8 @@ static int sdhci_am654_do_tuning(struct sdhci_host *host, memset(fail_window, 0, sizeof(fail_window)); - /* Enable ITAPDLY */ - sdhci_am654->itap_del_ena[timing] = 0x1; - for (itap = 0; itap < ITAPDLY_LENGTH; itap++) { - sdhci_am654_write_itapdly(sdhci_am654, itap, sdhci_am654->itap_del_ena[timing]); + sdhci_am654_write_itapdly(sdhci_am654, itap, 0x1); curr_pass = !mmc_send_tuning(host->mmc, opcode, NULL); @@ -584,10 +580,16 @@ static int sdhci_am654_platform_execute_tuning(struct sdhci_host *host, if (itapdly < 0) { dev_err(dev, "Failed to find itapdly, fail tuning\n"); + sdhci_am654_write_itapdly(sdhci_am654, 0, 0); + sdhci_am654->itap_del_ena[timing] = 0; + sdhci_am654->itap_del_sel[timing] = 0; return -1; } dev_dbg(dev, "Passed tuning, final itapdly=%d\n", itapdly); + + /* Enable ITAPDLY */ + sdhci_am654->itap_del_ena[timing] = 0x1; sdhci_am654_write_itapdly(sdhci_am654, itapdly, sdhci_am654->itap_del_ena[timing]); /* Save ITAPDLY */ sdhci_am654->itap_del_sel[timing] = itapdly; From 308d05225281d86150d88141990d6caf8c902349 Mon Sep 17 00:00:00 2001 From: "Diogo Ivo (Schneider Electric)" Date: Fri, 7 Aug 2026 13:07:00 +0200 Subject: [PATCH 043/592] mmc: sdhci_am654: Fallback to DT-provided itap delay on DDR50 tuning failure DDR50 mode is not required to support the tuning command CMD19, meaning that calibration may fail on cards that do not implement it, in which case a known-good itap delay value should be programmed into the host controller. Do this by reading the (already defined) itap delay DT property for DDR50 and, if tuning fails for this mode, fall back to the DT-provided itap delay value. If the DT does not provide a value for DDR50 fallback then this simply disables using itapdly. Fixes: 901d16e46296 ("mmc: sdhci_am654: Add retry tuning") Cc: stable@vger.kernel.org Signed-off-by: Diogo Ivo (Schneider Electric) Acked-by: Adrian Hunter Reviewed-by: Judith Mendez Signed-off-by: Ulf Hansson --- drivers/mmc/host/sdhci_am654.c | 26 +++++++++++++++++++++----- 1 file changed, 21 insertions(+), 5 deletions(-) diff --git a/drivers/mmc/host/sdhci_am654.c b/drivers/mmc/host/sdhci_am654.c index 49a46583c8bc..2e332f52c301 100644 --- a/drivers/mmc/host/sdhci_am654.c +++ b/drivers/mmc/host/sdhci_am654.c @@ -126,7 +126,7 @@ static const struct timing_data td[] = { NULL, MMC_CAP_UHS_SDR104}, [MMC_TIMING_UHS_DDR50] = {"ti,otap-del-sel-ddr50", - NULL, + "ti,itap-del-sel-ddr50", MMC_CAP_UHS_DDR50}, [MMC_TIMING_MMC_DDR52] = {"ti,otap-del-sel-ddr52", "ti,itap-del-sel-ddr52", @@ -144,6 +144,8 @@ struct sdhci_am654_data { u32 otap_del_sel[ARRAY_SIZE(td)]; u32 itap_del_sel[ARRAY_SIZE(td)]; u32 itap_del_ena[ARRAY_SIZE(td)]; + u32 itap_del_sel_dt_ddr50; + u32 itap_del_ena_dt_ddr50; int clkbuf_sel; int trm_icp; int drv_strength; @@ -579,10 +581,19 @@ static int sdhci_am654_platform_execute_tuning(struct sdhci_host *host, } while (++tuning_loop < RETRY_TUNING_MAX); if (itapdly < 0) { - dev_err(dev, "Failed to find itapdly, fail tuning\n"); - sdhci_am654_write_itapdly(sdhci_am654, 0, 0); - sdhci_am654->itap_del_ena[timing] = 0; - sdhci_am654->itap_del_sel[timing] = 0; + if (timing == MMC_TIMING_UHS_DDR50) { + dev_dbg(dev, "Failed DDR50 tuning, fallback to DT ITAP\n"); + sdhci_am654->itap_del_sel[timing] = sdhci_am654->itap_del_sel_dt_ddr50; + sdhci_am654->itap_del_ena[timing] = sdhci_am654->itap_del_ena_dt_ddr50; + } else { + dev_err(dev, "Failed to find itapdly, fail tuning\n"); + sdhci_am654->itap_del_ena[timing] = 0; + sdhci_am654->itap_del_sel[timing] = 0; + } + + sdhci_am654_write_itapdly(sdhci_am654, + sdhci_am654->itap_del_sel[timing], + sdhci_am654->itap_del_ena[timing]); return -1; } @@ -758,6 +769,11 @@ static int sdhci_am654_get_otap_delay(struct sdhci_host *host, } } + sdhci_am654->itap_del_sel_dt_ddr50 = + sdhci_am654->itap_del_sel[MMC_TIMING_UHS_DDR50]; + sdhci_am654->itap_del_ena_dt_ddr50 = + sdhci_am654->itap_del_ena[MMC_TIMING_UHS_DDR50]; + return 0; } From 44caf1844a258534e891d2c4071b011097e19235 Mon Sep 17 00:00:00 2001 From: Hemanth Selam Date: Fri, 4 Sep 2026 16:07:31 +0530 Subject: [PATCH 044/592] firmware: arm_scmi: Fix typo "upto" in comment Correct "upto" to "up to", reported by scripts/checkpatch.pl using the misspelling list in scripts/spelling.txt. Only touches comments, no code changes. Assisted-by: Cursor:claude-opus-5 Signed-off-by: Hemanth Selam Link: https://patch.msgid.link/20260904103732.7320-1-hemanth.selam@gmail.com Signed-off-by: Sudeep Holla --- drivers/firmware/arm_scmi/driver.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/drivers/firmware/arm_scmi/driver.c b/drivers/firmware/arm_scmi/driver.c index ef29fd223287..fb45b8e6459d 100644 --- a/drivers/firmware/arm_scmi/driver.c +++ b/drivers/firmware/arm_scmi/driver.c @@ -477,7 +477,7 @@ void *scmi_notification_instance_data_get(const struct scmi_handle *handle) * - exactly 'next_token' may be NOT available so pick xfer_id >= next_token * using find_next_zero_bit() starting from candidate next_token bit * - * - all tokens ahead upto (MSG_TOKEN_ID_MASK - 1) are used in-flight but we + * - all tokens ahead up to (MSG_TOKEN_ID_MASK - 1) are used in-flight but we * are plenty of free tokens at start, so try a second pass using * find_next_zero_bit() and starting from 0. * From cdb669a3b8f844aca71fc3224990157d61562165 Mon Sep 17 00:00:00 2001 From: Junjie Cao Date: Mon, 24 Aug 2026 11:14:19 +0800 Subject: [PATCH 045/592] HID: quirks: add ALWAYS_POLL quirk for SDINNOVATION gaming keyboard The SDINNOVATION gaming keyboard (USB ID 36ae:feab) stops reporting input events after its RGB lighting mode is switched about twice. Disabling USB autosuspend and unbinding the other HID interfaces make no difference; the issue does not occur on Windows. HID_QUIRK_ALWAYS_POLL alone resolves it, verified on 7.1.8 via usbhid.quirks=0x36ae:0xfeab:0x400. Reported-by: Marco Carvalho Link: https://bugzilla.redhat.com/show_bug.cgi?id=2514627 Cc: stable@vger.kernel.org Signed-off-by: Junjie Cao Signed-off-by: Benjamin Tissoires --- drivers/hid/hid-ids.h | 3 +++ drivers/hid/hid-quirks.c | 1 + 2 files changed, 4 insertions(+) diff --git a/drivers/hid/hid-ids.h b/drivers/hid/hid-ids.h index b3aca5aa9176..8b4f4b02aec0 100644 --- a/drivers/hid/hid-ids.h +++ b/drivers/hid/hid-ids.h @@ -1296,6 +1296,9 @@ #define USB_DEVICE_ID_SAMSUNG_WIRELESS_UNIVERSAL_KBD 0xa006 #define USB_DEVICE_ID_SAMSUNG_WIRELESS_MULTI_HOGP_KBD 0xa064 +#define USB_VENDOR_ID_SDINNOVATION 0x36ae +#define USB_DEVICE_ID_SDINNOVATION_GAMING_KBD 0xfeab + #define USB_VENDOR_ID_SEMICO 0x1a2c #define USB_DEVICE_ID_SEMICO_USB_KEYKOARD 0x0023 #define USB_DEVICE_ID_SEMICO_USB_KEYKOARD2 0x0027 diff --git a/drivers/hid/hid-quirks.c b/drivers/hid/hid-quirks.c index 8a0b51d47040..c8c6b29fc04d 100644 --- a/drivers/hid/hid-quirks.c +++ b/drivers/hid/hid-quirks.c @@ -186,6 +186,7 @@ static const struct hid_device_id hid_quirks[] = { { HID_USB_DEVICE(USB_VENDOR_ID_SAITEK, USB_DEVICE_ID_SAITEK_X52_2), HID_QUIRK_INCREMENT_USAGE_ON_DUPLICATE }, { HID_USB_DEVICE(USB_VENDOR_ID_SAITEK, USB_DEVICE_ID_SAITEK_X52_PRO), HID_QUIRK_INCREMENT_USAGE_ON_DUPLICATE }, { HID_USB_DEVICE(USB_VENDOR_ID_SAITEK, USB_DEVICE_ID_SAITEK_X65), HID_QUIRK_INCREMENT_USAGE_ON_DUPLICATE }, + { HID_USB_DEVICE(USB_VENDOR_ID_SDINNOVATION, USB_DEVICE_ID_SDINNOVATION_GAMING_KBD), HID_QUIRK_ALWAYS_POLL }, { HID_USB_DEVICE(USB_VENDOR_ID_SEMICO, USB_DEVICE_ID_SEMICO_USB_KEYKOARD2), HID_QUIRK_NO_INIT_REPORTS }, { HID_USB_DEVICE(USB_VENDOR_ID_SEMICO, USB_DEVICE_ID_SEMICO_USB_KEYKOARD), HID_QUIRK_NO_INIT_REPORTS }, { HID_USB_DEVICE(USB_VENDOR_ID_SENNHEISER, USB_DEVICE_ID_SENNHEISER_BTD500USB), HID_QUIRK_NOGET }, From 3b55f350c68a0aceff108f47f9d31f47ebffaf7b Mon Sep 17 00:00:00 2001 From: Daniel Borkmann Date: Mon, 7 Sep 2026 14:10:24 +0200 Subject: [PATCH 046/592] bpf: Fix bpf_skb_change_tail wrt csum partial skbs Cilium generates ICMP "frag needed" replies from BPF when a LB DSR packet exceeds the egress MTU. The reply is built by first trimming the packet down to target size via bpf_skb_change_tail(), and then pushing the ICMP error headers in front of it. The trim is rejected for skbs which carry a checksum offload, e.g. TCP packets aggregated by GRO on ingress where tcp_gro_complete() leaves the skb as CHECKSUM_PARTIAL. __bpf_skb_min_len() raises the minimum length to the end of the L4 checksum field, so a trim to 42 bytes bails out with -EINVAL given a min_len of 52 in this case, and due to that the ICMP generator fails. This is not the case if GRO is turned off. Fix this bpf_skb_change_tail() restriction and drop the checksum offload when the new length no longer covers the checksum field. The BPF program rewrites the skb into an ICMP error and computes the checksum itself anyway. Fixes: 5293efe62df8 ("bpf: add bpf_skb_change_tail helper") Reported-by: Tom Hadlaw Reported-by: Yusuke Suzuki Signed-off-by: Daniel Borkmann Link: https://lore.kernel.org/r/20260907121025.1923656-1-daniel@iogearbox.net Signed-off-by: Alexei Starovoitov --- net/core/filter.c | 11 +++++------ 1 file changed, 5 insertions(+), 6 deletions(-) diff --git a/net/core/filter.c b/net/core/filter.c index 61940e753552..8513167a858a 100644 --- a/net/core/filter.c +++ b/net/core/filter.c @@ -3961,12 +3961,6 @@ static u32 __bpf_skb_min_len(const struct sk_buff *skb) if (offset > 0) min_len = offset; } - if (skb->ip_summed == CHECKSUM_PARTIAL) { - offset = skb_checksum_start_offset(skb) + - skb->csum_offset + sizeof(__sum16); - if (offset > 0) - min_len = offset; - } return min_len; } @@ -3983,6 +3977,11 @@ static int bpf_skb_grow_rcsum(struct sk_buff *skb, unsigned int new_len) static int bpf_skb_trim_rcsum(struct sk_buff *skb, unsigned int new_len) { + if (skb->ip_summed == CHECKSUM_PARTIAL && + new_len < skb_checksum_start_offset(skb) + skb->csum_offset + + sizeof(__sum16)) + skb->ip_summed = CHECKSUM_NONE; + return __skb_trim_rcsum(skb, new_len); } From 15e2565f1c43771af0bc5324971cabaad79ac286 Mon Sep 17 00:00:00 2001 From: Daniel Borkmann Date: Mon, 7 Sep 2026 14:10:25 +0200 Subject: [PATCH 047/592] selftests/bpf: Add test for bpf_skb_change_tail on csum partial skbs Add a test which builds an ICMP error out of a TCP segment. A tcx prog on the client's egress side trims the first data segment down to the target size and pushes the ICMP error headers in front of it to then reflect the packet back to the sender. # LDLIBS=-static PKG_CONFIG='pkg-config --static' ./vmtest.sh -- ./test_progs -t tc_change_tail [...] #509 tc_change_tail:OK #510 tc_change_tail_pmtu:OK Summary: 2/0 PASSED, 0 SKIPPED, 0/0 FAILED Signed-off-by: Daniel Borkmann Link: https://lore.kernel.org/r/20260907121025.1923656-2-daniel@iogearbox.net Signed-off-by: Alexei Starovoitov --- .../bpf/prog_tests/tc_change_tail_pmtu.c | 125 +++++++++++++++++ .../bpf/progs/test_tc_change_tail_pmtu.c | 129 ++++++++++++++++++ 2 files changed, 254 insertions(+) create mode 100644 tools/testing/selftests/bpf/prog_tests/tc_change_tail_pmtu.c create mode 100644 tools/testing/selftests/bpf/progs/test_tc_change_tail_pmtu.c diff --git a/tools/testing/selftests/bpf/prog_tests/tc_change_tail_pmtu.c b/tools/testing/selftests/bpf/prog_tests/tc_change_tail_pmtu.c new file mode 100644 index 000000000000..7acdbd5757a9 --- /dev/null +++ b/tools/testing/selftests/bpf/prog_tests/tc_change_tail_pmtu.c @@ -0,0 +1,125 @@ +// SPDX-License-Identifier: GPL-2.0 + +#include + +#include "test_progs.h" +#include "network_helpers.h" +#include "test_tc_change_tail_pmtu.skel.h" + +#define CLIENT_NS "tc-change-tail-cli-ns" +#define SERVER_NS "tc-change-tail-srv-ns" +#define CLIENT_IP "192.168.1.1" +#define SERVER_IP "192.168.1.2" + +#define TEST_PMTU 1000 +#define TEST_MSS_MAX (TEST_PMTU - 20 - 20) +#define TIMEOUT_MS 3000 +#define XFER_BYTES 8192 + +void test_tc_change_tail_pmtu(void) +{ + LIBBPF_OPTS(bpf_tcx_opts, tcx_opts); + int mss_before = 0, mss_after = 0, ifindex, port; + int srv_fd = -1, srv_conn_fd = -1, cli_fd = -1; + struct test_tc_change_tail_pmtu *skel = NULL; + struct nstoken *nstoken = NULL; + static char buf[XFER_BYTES]; + socklen_t optlen; + ssize_t bytes; + size_t total; + + if (!ASSERT_OK(make_netns(CLIENT_NS), "make client ns")) + return; + if (!ASSERT_OK(make_netns(SERVER_NS), "make server ns")) + goto out_client_ns; + + nstoken = open_netns(CLIENT_NS); + if (!ASSERT_OK_PTR(nstoken, "open client ns")) + goto out; + SYS(out, "ip link add veth1 type veth peer name veth2 netns " SERVER_NS); + SYS(out, "ip -4 addr add " CLIENT_IP "/24 dev veth1"); + SYS(out, "ip link set veth1 up"); + ifindex = if_nametoindex("veth1"); + if (!ASSERT_NEQ(ifindex, 0, "if_nametoindex")) + goto out; + close_netns(nstoken); + nstoken = NULL; + + nstoken = open_netns(SERVER_NS); + if (!ASSERT_OK_PTR(nstoken, "open server ns")) + goto out; + SYS(out, "ip -4 addr add " SERVER_IP "/24 dev veth2"); + SYS(out, "ip link set veth2 up"); + srv_fd = start_server(AF_INET, SOCK_STREAM, SERVER_IP, 0, TIMEOUT_MS); + if (!ASSERT_OK_FD(srv_fd, "start server")) + goto out; + close_netns(nstoken); + nstoken = NULL; + + skel = test_tc_change_tail_pmtu__open_and_load(); + if (!ASSERT_OK_PTR(skel, "open and load skeleton")) + goto out; + + port = get_socket_local_port(srv_fd); + if (!ASSERT_GE(port, 0, "get server port")) + goto out; + + skel->bss->server_port = port; + skel->bss->pmtu = TEST_PMTU; + + nstoken = open_netns(CLIENT_NS); + if (!ASSERT_OK_PTR(nstoken, "open client ns")) + goto out; + + skel->links.change_tail_icmp = + bpf_program__attach_tcx(skel->progs.change_tail_icmp, ifindex, + &tcx_opts); + if (!ASSERT_OK_PTR(skel->links.change_tail_icmp, "attach tcx")) + goto out; + + cli_fd = connect_to_fd(srv_fd, TIMEOUT_MS); + if (!ASSERT_OK_FD(cli_fd, "connect to server")) + goto out; + srv_conn_fd = accept(srv_fd, NULL, NULL); + if (!ASSERT_OK_FD(srv_conn_fd, "accept connection")) + goto out; + if (!ASSERT_OK(settimeo(srv_conn_fd, TIMEOUT_MS), "set server timeout")) + goto out; + + optlen = sizeof(mss_before); + if (!ASSERT_OK(getsockopt(cli_fd, IPPROTO_TCP, TCP_MAXSEG, &mss_before, + &optlen), "get mss before")) + goto out; + + bytes = send(cli_fd, buf, sizeof(buf), 0); + if (!ASSERT_EQ(bytes, (ssize_t)sizeof(buf), "send data")) + goto out; + + for (total = 0; total < sizeof(buf); total += bytes) { + bytes = recv(srv_conn_fd, buf, sizeof(buf), 0); + if (bytes <= 0) + break; + } + + ASSERT_EQ(total, sizeof(buf), "receive data"); + ASSERT_OK(skel->data->change_tail_ret, "change tail"); + ASSERT_OK(skel->bss->adjust_room_ret, "adjust room"); + ASSERT_TRUE(skel->bss->icmp_sent, "icmp sent"); + + optlen = sizeof(mss_after); + if (!ASSERT_OK(getsockopt(cli_fd, IPPROTO_TCP, TCP_MAXSEG, &mss_after, + &optlen), "get mss after")) + goto out; + + ASSERT_LT(mss_after, mss_before, "mss reduced"); + ASSERT_LE(mss_after, TEST_MSS_MAX, "mss below pmtu"); +out: + close(srv_conn_fd); + close(cli_fd); + close(srv_fd); + test_tc_change_tail_pmtu__destroy(skel); + close_netns(nstoken); + remove_netns(SERVER_NS); +out_client_ns: + remove_netns(CLIENT_NS); +} diff --git a/tools/testing/selftests/bpf/progs/test_tc_change_tail_pmtu.c b/tools/testing/selftests/bpf/progs/test_tc_change_tail_pmtu.c new file mode 100644 index 000000000000..5c4c07545bc9 --- /dev/null +++ b/tools/testing/selftests/bpf/progs/test_tc_change_tail_pmtu.c @@ -0,0 +1,129 @@ +// SPDX-License-Identifier: GPL-2.0 + +#include +#include + +#include +#include +#include +#include +#include +#include + +#include +#include + +#define ICMP_SAMPLE_LEN (sizeof(struct iphdr) + 8) +#define ICMP_HDRS_LEN (sizeof(struct iphdr) + sizeof(struct icmphdr)) + +__be16 server_port = 0; +__u16 pmtu = 0; + +long change_tail_ret = 1; +long adjust_room_ret = 0; +bool icmp_sent = false; +bool icmp_err = false; + +static __always_inline __sum16 csum_fold(__wsum csum) +{ + csum = (csum & 0xffff) + (csum >> 16); + csum = (csum & 0xffff) + (csum >> 16); + + return (__sum16)~csum; +} + +SEC("tc/egress") +int change_tail_icmp(struct __sk_buff *skb) +{ + __u8 smac[ETH_ALEN], dmac[ETH_ALEN]; + void *data, *data_end; + struct icmphdr *icmp; + struct ethhdr *eth; + struct tcphdr *tcp; + __be32 saddr, daddr; + struct iphdr *ip; + __wsum csum; + + if (icmp_sent || icmp_err) + return TCX_PASS; + + data = (void *)(long)skb->data; + data_end = (void *)(long)skb->data_end; + + eth = data; + if ((void *)(eth + 1) > data_end) + return TCX_PASS; + if (eth->h_proto != bpf_htons(ETH_P_IP)) + return TCX_PASS; + + ip = (void *)(eth + 1); + if ((void *)(ip + 1) > data_end) + return TCX_PASS; + if (ip->ihl != 5 || ip->protocol != IPPROTO_TCP) + return TCX_PASS; + + tcp = (void *)(ip + 1); + if ((void *)(tcp + 1) > data_end) + return TCX_PASS; + if (tcp->dest != server_port) + return TCX_PASS; + if (bpf_ntohs(ip->tot_len) <= sizeof(*ip) + tcp->doff * 4) + return TCX_PASS; + + __builtin_memcpy(smac, eth->h_source, ETH_ALEN); + __builtin_memcpy(dmac, eth->h_dest, ETH_ALEN); + saddr = ip->saddr; + daddr = ip->daddr; + + change_tail_ret = bpf_skb_change_tail(skb, ETH_HLEN + ICMP_SAMPLE_LEN, 0); + if (change_tail_ret) { + icmp_err = true; + return TCX_PASS; + } + + adjust_room_ret = bpf_skb_adjust_room(skb, ICMP_HDRS_LEN, + BPF_ADJ_ROOM_MAC, + BPF_F_ADJ_ROOM_NO_CSUM_RESET); + if (adjust_room_ret) { + icmp_err = true; + return TCX_DROP; + } + + data = (void *)(long)skb->data; + data_end = (void *)(long)skb->data_end; + + eth = data; + ip = (void *)(eth + 1); + icmp = (void *)(ip + 1); + if ((void *)icmp + sizeof(*icmp) + ICMP_SAMPLE_LEN > data_end) { + icmp_err = true; + return TCX_DROP; + } + + __builtin_memcpy(eth->h_dest, smac, ETH_ALEN); + __builtin_memcpy(eth->h_source, dmac, ETH_ALEN); + + __builtin_memset(icmp, 0, sizeof(*icmp)); + icmp->type = ICMP_DEST_UNREACH; + icmp->code = ICMP_FRAG_NEEDED; + icmp->un.frag.mtu = bpf_htons(pmtu); + + __builtin_memset(ip, 0, sizeof(*ip)); + ip->version = 4; + ip->ihl = 5; + ip->ttl = 64; + ip->protocol = IPPROTO_ICMP; + ip->tot_len = bpf_htons(ICMP_HDRS_LEN + ICMP_SAMPLE_LEN); + ip->saddr = daddr; + ip->daddr = saddr; + + csum = bpf_csum_diff(NULL, 0, (__be32 *)icmp, + sizeof(*icmp) + ICMP_SAMPLE_LEN, 0); + icmp->checksum = csum_fold(csum); + csum = bpf_csum_diff(NULL, 0, (__be32 *)ip, sizeof(*ip), 0); + ip->check = csum_fold(csum); + icmp_sent = true; + return bpf_redirect(skb->ifindex, BPF_F_INGRESS); +} + +char _license[] SEC("license") = "GPL"; From b7c0f8436f077e7f66c9f07714bd57068e2a0c31 Mon Sep 17 00:00:00 2001 From: Paul Hollinsky Date: Fri, 21 Aug 2026 01:13:25 -0700 Subject: [PATCH 048/592] drm/msm/adreno: Only check for PAS when a zap shader is present Commit 0be72be03ca7 ("drm/msm: Switch to generic PAS TZ APIs") replaced the qcom_scm_is_available() check in adreno_zap_shader_load() with qcom_pas_is_available(). These are not equivalent: the former reports whether the SCM transport is up, the latter whether the TrustZone firmware implements the peripheral authentication service. On SC7180 Chromebooks (trogdor) TZ does not implement PAS at all. SCM call-availability queries return 0 for every PAS command while other services answer normally: svc 0x06 cmd 0x01 IS_CALL_AVAIL -> 1 svc 0x02 cmd 0x01 PAS_INIT_IMAGE -> 0 svc 0x02 cmd 0x05 PAS_AUTH_RESET -> 0 svc 0x02 cmd 0x07 PAS_IS_SUPPORTED -> 0 svc 0x0c cmd 0x16 MP_ASSIGN -> 1 svc 0x05 cmd 0x01 IO_READ -> 1 so qcom_scm_probe() never registers a PAS backend and qcom_pas_is_available() is false for the lifetime of the boot. That on its own need not matter, because sc7180-trogdor.dtsi does /delete-node/ &gpu_zap_shader;, and the intended path for such a board is for zap_shader_load_mdt() to find no zap-shader child, clear zap_available, return -ENODEV, and let the caller fall back to SECVID_TRUST_CNTL. The problem is the ordering. zap_available is a static initialised to true and is only ever cleared inside zap_shader_load_mdt(), but adreno_zap_shader_load() consults PAS before calling it. The discovery that decides whether a zap shader is needed at all can therefore never run, the flag is never cleared, and every call returns -EPROBE_DEFER: adreno 5000000.gpu: [drm:adreno_zap_shader_load] *ERROR* PAS is not available msm_dpu ae01000.display-controller: [drm:adreno_load_gpu] *ERROR* gpu hw init failed: -517 Nothing retries that deferral, either. adreno_zap_shader_load() is called from a6xx_hw_init() rather than from probe, so the -EPROBE_DEFER is not a probe return value: it propagates up until adreno_load_gpu() returns NULL. load_gpu() re-attempts on every DRM open while priv->gpu is NULL, each open fails identically, and PAS cannot become available in between - which is why the error repeats and userspace stays on llvmpipe. Move the availability check into zap_shader_load_mdt(), behind the zap-shader node lookup, so the driver only consults PAS once it knows it needs PAS. Boards with no zap-shader node take the intended -ENODEV fallback without ever asking, and boards that do have one keep the qcom_pas_is_available() gate. Fixes: 0be72be03ca7 ("drm/msm: Switch to generic PAS TZ APIs") Link: https://lore.kernel.org/r/20260808034716.58888-1-phollinsky@holtechnik.com Signed-off-by: Paul Hollinsky Reviewed-by: Konrad Dybcio Patchwork: https://patchwork.freedesktop.org/patch/747583/ Message-ID: <20260821081325.89088-1-phollinsky@holtechnik.com> Signed-off-by: Rob Clark --- drivers/gpu/drm/msm/adreno/adreno_gpu.c | 13 ++++++------- 1 file changed, 6 insertions(+), 7 deletions(-) diff --git a/drivers/gpu/drm/msm/adreno/adreno_gpu.c b/drivers/gpu/drm/msm/adreno/adreno_gpu.c index 8cd2020d4b7e..ca5e4e560cde 100644 --- a/drivers/gpu/drm/msm/adreno/adreno_gpu.c +++ b/drivers/gpu/drm/msm/adreno/adreno_gpu.c @@ -52,6 +52,12 @@ static int zap_shader_load_mdt(struct msm_gpu *gpu, const char *fwname, return -ENODEV; } + /* We need PAS to be able to load the firmware */ + if (!qcom_pas_is_available()) { + DRM_DEV_ERROR(dev, "PAS is not available\n"); + return -EPROBE_DEFER; + } + ret = of_reserved_mem_region_to_resource(np, 0, &r); if (ret) { zap_available = false; @@ -170,18 +176,11 @@ static int zap_shader_load_mdt(struct msm_gpu *gpu, const char *fwname, int adreno_zap_shader_load(struct msm_gpu *gpu, u32 pasid) { struct adreno_gpu *adreno_gpu = to_adreno_gpu(gpu); - struct platform_device *pdev = gpu->pdev; /* Short cut if we determine the zap shader isn't available/needed */ if (!zap_available) return -ENODEV; - /* We need PAS to be able to load the firmware */ - if (!qcom_pas_is_available()) { - DRM_DEV_ERROR(&pdev->dev, "PAS is not available\n"); - return -EPROBE_DEFER; - } - return zap_shader_load_mdt(gpu, adreno_gpu->info->zapfw, pasid); } From aba7b41faeecb7692458095ce6fafc341fe0b80e Mon Sep 17 00:00:00 2001 From: Bard Liao Date: Tue, 1 Sep 2026 11:10:19 +0800 Subject: [PATCH 049/592] soundwire: cadence_master: wait and cancel cdns->work before clock stop A peripheral event could happen during the clock stop process. We need to wait for the event be handled before stopping the bus clock. Otherwise, we will get the IO transfer timed out issue. Fixes: af4cc917826f ("soundwire: cadence: mask Slave interrupt before stopping clock") Signed-off-by: Bard Liao Reviewed-by: David Lin Reviewed-by: Shuming Fan Reviewed-by: Pierre-Louis Bossart Link: https://patch.msgid.link/20260901031019.233254-1-yung-chuan.liao@linux.intel.com Signed-off-by: Vinod Koul --- drivers/soundwire/cadence_master.c | 7 +++++++ 1 file changed, 7 insertions(+) diff --git a/drivers/soundwire/cadence_master.c b/drivers/soundwire/cadence_master.c index e690237fe981..a0a63b76793e 100644 --- a/drivers/soundwire/cadence_master.c +++ b/drivers/soundwire/cadence_master.c @@ -1701,6 +1701,13 @@ int sdw_cdns_clock_stop(struct sdw_cdns *cdns, bool block_wake) return 0; } + /* + * wait for any in-flight peripheral event handling to complete before stopping the clock. + * No need to disable peripheral interrupts before canceling the work, as the peripheral + * interrupts are already masked before the work is scheduled. + */ + cancel_work_sync(&cdns->work); + /* * Before entering clock stop we mask the Slave * interrupts. This helps avoid having to deal with e.g. a From fb5a08026a68be0ec5f660588311fb1091bf5d58 Mon Sep 17 00:00:00 2001 From: Julian Braha Date: Sun, 16 Aug 2026 00:51:09 +0100 Subject: [PATCH 050/592] mips: econet: fix unmet dependencies for ECONET ECONET selects EARLY_PRINTK_8250, SERIAL_OF_PLATFORM, and SERIAL_8250 without ensuring their dependencies, EARLY_PRINTK and TTY are met. This causes unmet dependencies: WARNING: unmet direct dependencies detected for SERIAL_8250 Depends on [n]: TTY [=n] && HAS_IOMEM [=y] && !S390 Selected by [y]: - ECONET [=y] WARNING: unmet direct dependencies detected for EARLY_PRINTK_8250 Depends on [n]: EARLY_PRINTK [=n] && USE_GENERIC_EARLY_PRINTK_8250 [=y] Selected by [y]: - ECONET [=y] WARNING: unmet direct dependencies detected for SERIAL_OF_PLATFORM Depends on [n]: TTY [=n] && HAS_IOMEM [=y] && SERIAL_8250 [=y] && OF [=y] Selected by [y]: - ECONET [=y] However, in the discussion of v1 of this patch, Arnd pointed out that these selects don't belong here in the first place. [1] Let's remove them to resolve the unmet dependencies. These unmet dependencies were found by kconfirm, a static analysis tool for Kconfig. Fixes: 79ee1d20e37c ("mips: econet: Fix incorrect Kconfig dependencies") Fixes: 35fb26f94dfa ("mips: Add EcoNet MIPS platform support") Suggested-by: Arnd Bergmann Signed-off-by: Julian Braha Tested-by: Caleb James DeLisle Signed-off-by: Thomas Bogendoerfer --- arch/mips/Kconfig | 3 --- 1 file changed, 3 deletions(-) diff --git a/arch/mips/Kconfig b/arch/mips/Kconfig index e2eb9627bd14..c3f69def2008 100644 --- a/arch/mips/Kconfig +++ b/arch/mips/Kconfig @@ -396,10 +396,7 @@ config ECONET bool "EcoNet MIPS family" select BOOT_RAW select DEBUG_ZBOOT if DEBUG_KERNEL - select EARLY_PRINTK_8250 select ECONET_EN751221_TIMER - select SERIAL_8250 - select SERIAL_OF_PLATFORM select SYS_SUPPORTS_BIG_ENDIAN select SYS_HAS_CPU_MIPS32_R1 select SYS_HAS_CPU_MIPS32_R2 From e261bfd3c042bccd25c82b04de079bdf8bdd8c3a Mon Sep 17 00:00:00 2001 From: Caleb James DeLisle Date: Mon, 17 Aug 2026 22:03:44 +0000 Subject: [PATCH 051/592] MIPS: config: Add EcoNet EN751221 defconfig Add config based on OpenWrt for EN751221 boards. Of note: EcoNet bootloader has a 7.2MB kernel size limit so ZBOOT is assumed to be needed in all cases. Signed-off-by: Caleb James DeLisle Tested-by: Randy Dunlap Signed-off-by: Thomas Bogendoerfer --- MAINTAINERS | 1 + arch/mips/configs/econet_en751221_defconfig | 264 ++++++++++++++++++++ 2 files changed, 265 insertions(+) create mode 100644 arch/mips/configs/econet_en751221_defconfig diff --git a/MAINTAINERS b/MAINTAINERS index 3a19da74d00c..9b9e20d039f6 100644 --- a/MAINTAINERS +++ b/MAINTAINERS @@ -9324,6 +9324,7 @@ F: Documentation/devicetree/bindings/interrupt-controller/econet,en751221-intc.y F: Documentation/devicetree/bindings/mips/econet.yaml F: Documentation/devicetree/bindings/timer/econet,en751221-timer.yaml F: arch/mips/boot/dts/econet/ +F: arch/mips/configs/econet_en751221_defconfig F: arch/mips/econet/ F: drivers/clocksource/timer-econet-en751221.c F: drivers/irqchip/irq-econet-en751221.c diff --git a/arch/mips/configs/econet_en751221_defconfig b/arch/mips/configs/econet_en751221_defconfig new file mode 100644 index 000000000000..b35d33fe4044 --- /dev/null +++ b/arch/mips/configs/econet_en751221_defconfig @@ -0,0 +1,264 @@ +# CONFIG_LOCALVERSION_AUTO is not set +CONFIG_KERNEL_XZ=y +CONFIG_SYSVIPC=y +CONFIG_POSIX_MQUEUE=y +# CONFIG_CROSS_MEMORY_ATTACH is not set +CONFIG_HIGH_RES_TIMERS=y +CONFIG_BPF_SYSCALL=y +CONFIG_BPF_JIT=y +CONFIG_PREEMPT=y +CONFIG_CGROUPS=y +CONFIG_MEMCG=y +CONFIG_BLK_CGROUP=y +CONFIG_CGROUP_SCHED=y +CONFIG_CFS_BANDWIDTH=y +CONFIG_RT_GROUP_SCHED=y +CONFIG_CGROUP_PIDS=y +CONFIG_CGROUP_RDMA=y +CONFIG_CPUSETS=y +CONFIG_CGROUP_CPUACCT=y +CONFIG_CGROUP_BPF=y +CONFIG_NAMESPACES=y +# CONFIG_TIME_NS is not set +CONFIG_USER_NS=y +CONFIG_BLK_DEV_INITRD=y +# CONFIG_RD_GZIP is not set +# CONFIG_RD_BZIP2 is not set +# CONFIG_RD_LZMA is not set +# CONFIG_RD_XZ is not set +# CONFIG_RD_LZO is not set +# CONFIG_RD_LZ4 is not set +# CONFIG_RD_ZSTD is not set +# CONFIG_INITRAMFS_PRESERVE_MTIME is not set +CONFIG_LD_DEAD_CODE_DATA_ELIMINATION=y +CONFIG_EXPERT=y +# CONFIG_SGETMASK_SYSCALL is not set +# CONFIG_IO_URING is not set +# CONFIG_RSEQ is not set +# CONFIG_CACHESTAT_SYSCALL is not set +CONFIG_ECONET=y +CONFIG_CPU_MIPS32_R2=y +CONFIG_ZBOOT_LOAD_ADDRESS=0x80020000 +CONFIG_ARCH_FORCE_MAX_ORDER=11 +# CONFIG_MIPS_FP_SUPPORT is not set +CONFIG_NR_CPUS=2 +CONFIG_HZ_100=y +CONFIG_MIPS_RAW_APPENDED_DTB=y +# CONFIG_SCHED_SMT is not set +CONFIG_JUMP_LABEL=y +# CONFIG_STACKPROTECTOR_STRONG is not set +# CONFIG_GCC_PLUGINS is not set +CONFIG_MODULES=y +CONFIG_MODULE_UNLOAD=y +# CONFIG_BLOCK_LEGACY_AUTOLOAD is not set +CONFIG_BLK_DEV_THROTTLING=y +# CONFIG_BLK_DEBUG_FS is not set +CONFIG_PARTITION_ADVANCED=y +CONFIG_OF_PARTITION=y +# CONFIG_MQ_IOSCHED_DEADLINE is not set +# CONFIG_MQ_IOSCHED_KYBER is not set +# CONFIG_CORE_DUMP_DEFAULT_ELF_HEADERS is not set +CONFIG_SLAB_FREELIST_RANDOM=y +CONFIG_SLAB_FREELIST_HARDENED=y +# CONFIG_COMPAT_BRK is not set +CONFIG_LRU_GEN=y +CONFIG_LRU_GEN_ENABLED=y +CONFIG_NET=y +CONFIG_PACKET=y +CONFIG_UNIX=y +CONFIG_INET=y +CONFIG_IP_MULTICAST=y +CONFIG_IP_ADVANCED_ROUTER=y +CONFIG_IP_MULTIPLE_TABLES=y +CONFIG_IP_ROUTE_MULTIPATH=y +CONFIG_IP_ROUTE_VERBOSE=y +CONFIG_IP_MROUTE=y +CONFIG_IP_MROUTE_MULTIPLE_TABLES=y +CONFIG_IP_PIMSM_V1=y +CONFIG_IP_PIMSM_V2=y +CONFIG_SYN_COOKIES=y +# CONFIG_INET_DIAG is not set +CONFIG_TCP_CONG_ADVANCED=y +# CONFIG_TCP_CONG_BIC is not set +# CONFIG_TCP_CONG_WESTWOOD is not set +# CONFIG_TCP_CONG_HTCP is not set +# CONFIG_IPV6_SIT is not set +CONFIG_IPV6_SUBTREES=y +CONFIG_IPV6_MROUTE=y +CONFIG_IPV6_MROUTE_MULTIPLE_TABLES=y +CONFIG_IPV6_PIMSM_V2=y +CONFIG_IPV6_SEG6_LWTUNNEL=y +CONFIG_MPTCP=y +CONFIG_NETFILTER=y +# CONFIG_NETFILTER_EGRESS is not set +CONFIG_NF_CONNTRACK=m +CONFIG_NF_CONNTRACK_MARK=y +CONFIG_NF_CONNTRACK_ZONES=y +CONFIG_NF_CONNTRACK_PROCFS=y +CONFIG_NF_CONNTRACK_TIMEOUT=y +# CONFIG_NF_CT_PROTO_SCTP is not set +CONFIG_NF_TABLES=m +CONFIG_NF_TABLES_INET=y +CONFIG_NF_TABLES_NETDEV=y +CONFIG_NFT_NUMGEN=m +CONFIG_NFT_CT=m +CONFIG_NFT_FLOW_OFFLOAD=m +CONFIG_NFT_LOG=m +CONFIG_NFT_LIMIT=m +CONFIG_NFT_MASQ=m +CONFIG_NFT_REDIR=m +CONFIG_NFT_NAT=m +CONFIG_NFT_QUOTA=m +CONFIG_NFT_REJECT=m +CONFIG_NFT_HASH=m +CONFIG_NFT_FIB_INET=m +CONFIG_NF_FLOW_TABLE_INET=m +CONFIG_NF_FLOW_TABLE=m +CONFIG_NFT_FIB_IPV4=m +CONFIG_NF_TABLES_ARP=y +CONFIG_NF_LOG_IPV4=m +CONFIG_NFT_FIB_IPV6=m +CONFIG_NF_LOG_IPV6=m +CONFIG_NF_TABLES_BRIDGE=m +CONFIG_BRIDGE=y +CONFIG_BRIDGE_VLAN_FILTERING=y +CONFIG_VLAN_8021Q=y +CONFIG_NET_SCHED=y +CONFIG_NET_SCH_FQ_CODEL=y +CONFIG_NET_SCH_DEFAULT=y +CONFIG_DEFAULT_FQ_CODEL=y +CONFIG_NET_SWITCHDEV=y +CONFIG_NET_L3_MASTER_DEV=y +CONFIG_RFKILL=y +# CONFIG_LWTUNNEL_BPF is not set +CONFIG_PCI=y +CONFIG_PCIEPORTBUS=y +CONFIG_PCIEAER=y +# CONFIG_PCIEASPM is not set +CONFIG_PCI_MSI=y +# CONFIG_VGA_ARB is not set +CONFIG_PCIE_MEDIATEK=y +CONFIG_MTD=y +CONFIG_MTD_BLOCK=y +CONFIG_MTD_CFI=y +CONFIG_MTD_CFI_INTELEXT=y +CONFIG_MTD_CFI_AMDSTD=y +CONFIG_MTD_COMPLEX_MAPPINGS=y +CONFIG_MTD_SPI_NAND=y +CONFIG_MTD_UBI=y +CONFIG_MTD_UBI_BEB_LIMIT=13 +CONFIG_MTD_UBI_BLOCK=y +CONFIG_NETDEVICES=y +# CONFIG_NET_VENDOR_ASIX is not set +# CONFIG_NET_VENDOR_ENGLEDER is not set +# CONFIG_NET_VENDOR_FUNGIBLE is not set +# CONFIG_NET_VENDOR_LITEX is not set +# CONFIG_NET_VENDOR_MICROSOFT is not set +# CONFIG_NET_VENDOR_VERTEXCOM is not set +# CONFIG_NET_VENDOR_WANGXUN is not set +CONFIG_PPP=m +CONFIG_PPP_FILTER=y +CONFIG_PPP_MULTILINK=y +CONFIG_PPPOE=m +CONFIG_PPP_ASYNC=m +# CONFIG_WLAN_VENDOR_ADMTEK is not set +# CONFIG_WLAN_VENDOR_ATH is not set +# CONFIG_WLAN_VENDOR_ATMEL is not set +# CONFIG_WLAN_VENDOR_BROADCOM is not set +# CONFIG_WLAN_VENDOR_INTEL is not set +# CONFIG_WLAN_VENDOR_INTERSIL is not set +# CONFIG_WLAN_VENDOR_MARVELL is not set +# CONFIG_WLAN_VENDOR_MEDIATEK is not set +# CONFIG_WLAN_VENDOR_MICROCHIP is not set +# CONFIG_WLAN_VENDOR_PURELIFI is not set +# CONFIG_WLAN_VENDOR_RALINK is not set +# CONFIG_WLAN_VENDOR_REALTEK is not set +# CONFIG_WLAN_VENDOR_RSI is not set +# CONFIG_WLAN_VENDOR_SILABS is not set +# CONFIG_WLAN_VENDOR_ST is not set +# CONFIG_WLAN_VENDOR_TI is not set +# CONFIG_WLAN_VENDOR_ZYDAS is not set +# CONFIG_WLAN_VENDOR_QUANTENNA is not set +# CONFIG_INPUT is not set +# CONFIG_SERIO is not set +# CONFIG_VT is not set +# CONFIG_LEGACY_PTYS is not set +# CONFIG_LEGACY_TIOCSTI is not set +CONFIG_SERIAL_8250=y +CONFIG_SERIAL_8250_CONSOLE=y +CONFIG_SERIAL_OF_PLATFORM=y +# CONFIG_HW_RANDOM is not set +# CONFIG_DEVMEM is not set +CONFIG_SPI=y +# CONFIG_PTP_1588_CLOCK is not set +CONFIG_GPIOLIB=y +# CONFIG_HWMON is not set +CONFIG_WATCHDOG=y +CONFIG_MFD_SYSCON=y +# CONFIG_USB_PCI is not set +CONFIG_NEW_LEDS=y +CONFIG_LEDS_CLASS=y +CONFIG_LEDS_CLASS_MULTICOLOR=y +CONFIG_LEDS_GPIO=m +CONFIG_LEDS_TRIGGERS=y +CONFIG_LEDS_TRIGGER_TIMER=y +CONFIG_LEDS_TRIGGER_HEARTBEAT=y +CONFIG_LEDS_TRIGGER_DEFAULT_ON=y +CONFIG_LEDS_TRIGGER_NETDEV=y +CONFIG_STAGING=y +# CONFIG_MIPS_PLATFORM_DEVICES is not set +CONFIG_COMMON_CLK_EN7523=y +# CONFIG_IOMMU_SUPPORT is not set +CONFIG_RESET_CONTROLLER=y +CONFIG_PHY_ECONET_PCIE=y +# CONFIG_DNOTIFY is not set +CONFIG_FANOTIFY=y +CONFIG_OVERLAY_FS=y +# CONFIG_OVERLAY_FS_REDIRECT_ALWAYS_FOLLOW is not set +# CONFIG_PROC_PAGE_MONITOR is not set +CONFIG_TMPFS=y +CONFIG_TMPFS_POSIX_ACL=y +CONFIG_JFFS2_FS=y +CONFIG_JFFS2_SUMMARY=y +CONFIG_JFFS2_FS_XATTR=y +CONFIG_JFFS2_COMPRESSION_OPTIONS=y +CONFIG_UBIFS_FS=y +CONFIG_SQUASHFS=y +CONFIG_SQUASHFS_FILE_DIRECT=y +CONFIG_SQUASHFS_COMPILE_DECOMP_MULTI_PERCPU=y +# CONFIG_SQUASHFS_ZLIB is not set +CONFIG_SQUASHFS_XZ=y +CONFIG_SQUASHFS_EMBEDDED=y +CONFIG_KEYS=y +CONFIG_INIT_STACK_NONE=y +CONFIG_FORTIFY_SOURCE=y +CONFIG_HARDENED_USERCOPY=y +CONFIG_LIST_HARDENED=y +CONFIG_CRYPTO_NULL=y +CONFIG_CRYPTO_AES=y +CONFIG_CRYPTO_CCM=y +CONFIG_CRYPTO_GCM=y +# CONFIG_CRYPTO_HW is not set +# CONFIG_XZ_DEC_X86 is not set +# CONFIG_XZ_DEC_POWERPC is not set +# CONFIG_XZ_DEC_ARM is not set +# CONFIG_XZ_DEC_ARMTHUMB is not set +# CONFIG_XZ_DEC_ARM64 is not set +# CONFIG_XZ_DEC_SPARC is not set +# CONFIG_XZ_DEC_RISCV is not set +CONFIG_PRINTK_TIME=y +# CONFIG_DEBUG_MISC is not set +CONFIG_DEBUG_INFO_DWARF_TOOLCHAIN_DEFAULT=y +CONFIG_DEBUG_INFO_REDUCED=y +CONFIG_FRAME_WARN=1024 +CONFIG_STRIP_ASM_SYMS=y +CONFIG_MAGIC_SYSRQ=y +# CONFIG_MAGIC_SYSRQ_SERIAL is not set +CONFIG_DEBUG_FS=y +# CONFIG_SLUB_DEBUG is not set +CONFIG_SCHED_STACK_END_CHECK=y +CONFIG_PANIC_ON_OOPS=y +CONFIG_PANIC_TIMEOUT=1 +CONFIG_RCU_CPU_STALL_TIMEOUT=60 +# CONFIG_RCU_TRACE is not set +# CONFIG_FTRACE is not set From bbc448c541ed90d52f9ec22f17394304e3bf170e Mon Sep 17 00:00:00 2001 From: =?UTF-8?q?Th=C3=A9o=20Lebrun?= Date: Fri, 28 Aug 2026 16:07:33 +0200 Subject: [PATCH 052/592] mips: select CONFIG_WEAK_REORDERING_BEYOND_LLSC from CONFIG_EYEQ MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit On I6500 CPU cores, lld and scd give no ordering guarantees (same as all other instructions). To respect the assumption that arch_cmpxchg() is fully ordered, we must inject sync instructions above and below our lld/scd loops using the already in place WEAK_REORDERING_BEYOND_LLSC infrastructure. Otherwise, bad things can happen: [ 34.054496] CPU 3 Unable to handle kernel paging request at virtual address 0000000000000000, epc == a80000080838e01c, ra == a80000080838dfc4 [ 34.054559] Oops[#1]: [ 34.069561] CPU: 3 UID: 0 PID: 170 Comm: pipe_race Not tainted 7.2.0-rc6-01553-gb73c35220968-dirty #103 VOLUNTARY [ 34.079932] Hardware name: Mobile EyeQ5 MP5 Evaluation board [ 34.085592] $ 0 : 0000000000000000 0000000000000001 0000000000000000 0000000000000000 [ 34.093616] $ 4 : a800000808ee2618 000000000b7a879d 0000000000001000 0000000000000000 [ 34.101638] $ 8 : 0000000000e3f2c9 0000000000000000 a800000808a2a9f8 0000000000000000 [ 34.109660] $12 : a8000008139ffcd8 ffffffff84080018 a80000080837fae0 7878787878787878 [ 34.117682] $16 : a800000807e82940 0000000000001000 0000000000000000 0000000000000000 [ 34.125704] $20 : a800000802920e00 a8000008139ffdf8 a800000802649400 0000000000e3f2c9 [ 34.133726] $24 : 0000000000000006 00000001200406e0 [ 34.141783] $28 : a8000008139fc000 a8000008139ffd10 0000000000e3f2c8 a80000080838dfc4 [ 34.149837] epc : a80000080838e01c anon_pipe_read+0xd4/0x428 [ 34.155697] ra : a80000080838dfc4 anon_pipe_read+0x7c/0x428 [ 34.161549] Status: 140000e3 KX SX UX KERNEL EXL IE [ 34.166551] Cause : 40800408 (ExcCode 02) [ 34.170574] BadVA : 0000000000000000 [ 34.174161] PrId : 0001b028 (MIPS I6500) [ 34.178183] Process pipe_race (pid: 170, threadinfo=000000005ca35720, task=00000000e1013890, tls=000000014ebbb780) [ 34.188568] Stack : a800000802649400 0000000000000000 0000000000000000 a8000008139ffdd0 [ 34.196623] 0000000000000fba a800000808ee0000 0000000000000001 a8000008130c3e80 [ 34.204676] a8000008080d1280 a8000008139ffd58 a8000008139ffd58 1dbd2b22ea1dd500 [ 34.212729] a800000802649400 a800000808ee0000 ffffffffffffffea 0000000000000001 [ 34.220783] 0000000000001000 0000000000000000 00000001200ae518 ffffffffffffffff [ 34.228836] 000000fffbe0e530 a80000080837edf4 000000fffbe0e530 0000000000000000 [ 34.236890] 0000000000000000 0000000000000000 000000014ebb55a0 0000000000001000 [ 34.244943] 0000000000000001 a800000802649400 0000000000000000 0000000000000000 [ 34.252996] 0000000000000000 0000400400000000 0000000000000000 1dbd2b22ea1dd500 [ 34.261049] 00000000140000e3 a800000802649400 a800000802649400 a800000808ee0000 [ 34.269103] ... [ 34.271568] Call Trace: [ 34.274026] [] anon_pipe_read+0xd4/0x428 [ 34.279533] [] vfs_read+0x25c/0x318 [ 34.284607] [] ksys_read+0x104/0x138 [ 34.289763] [] syscall_common+0x44/0x68 [ 34.295187] [ 34.296689] Code: f84000cf 02209825 de020010 d8400004 02002825 0040f809 02802025 f84000c3 [ 34.306504] [ 34.308099] ---[ end trace 0000000000000000 ]--- My initial reproducer was the xdp-tools test suite. A standalone reproducer would be an lld/scd loop that, when the read is reordered by the CPU, triggers a fault. We can achieve this from userspace by stressing an anonymous pipe, which uses a mutex. Program used: // SPDX-License-Identifier: GPL-2.0 // pipe_race.c - reproducer for MIPS LL/SC reordering vs fs/pipe.c // // Two userspace processes on an anonymous pipe: // parent = writer: tight write() loop // child = reader: tight read() loop #define _GNU_SOURCE #include #include #include #include #include #include #include #include #include #include #include int main(void) { long wrsize = 70; // bytes per write() long rdsize = 4096; // bytes per read() int pfd[2]; char *buf; pid_t pid; int ret; ret = pipe(pfd); assert(!ret); pid = fork(); assert(pid >= 0); if (pid == 0) { /* reader */ close(pfd[1]); buf = malloc(rdsize); assert(buf); for (;;) { ssize_t n = read(pfd[0], buf, rdsize); if (n < 0 && errno == EINTR) continue; if (n <= 0) _exit(n < 0 ? 1 : 0); } } close(pfd[0]); /* writer */ buf = malloc(wrsize); assert(buf); memset(buf, 'x', wrsize); for (;;) { ssize_t n = write(pfd[1], buf, wrsize); if (n < 0 && errno == EINTR) continue; if (n != wrsize) break; } kill(pid, SIGKILL); wait(NULL); return 0; } Fixes: fbe0fae601b7 ("MIPS: mobileye: Add EyeQ6H support") Cc: stable@vger.kernel.org Signed-off-by: Théo Lebrun Reviewed-by: Jiaxun Yang Signed-off-by: Thomas Bogendoerfer --- arch/mips/Kconfig | 1 + 1 file changed, 1 insertion(+) diff --git a/arch/mips/Kconfig b/arch/mips/Kconfig index c3f69def2008..d7c67cebe065 100644 --- a/arch/mips/Kconfig +++ b/arch/mips/Kconfig @@ -658,6 +658,7 @@ config EYEQ select USB_UHCI_BIG_ENDIAN_MMIO if CPU_BIG_ENDIAN select USE_OF select HOTPLUG_PARALLEL if HOTPLUG_CPU + select WEAK_REORDERING_BEYOND_LLSC help Select this to build a kernel supporting EyeQ SoC from Mobileye. From 126f16e0a1b353c2ba5c7e2c8626cfa865934f9f Mon Sep 17 00:00:00 2001 From: Orgad Shaneh Date: Tue, 1 Sep 2026 19:33:55 +0000 Subject: [PATCH 053/592] MIPS: Octeon: apply USB FDT fixups also when USB is modular The uctl/usbn device-tree fixups in octeon_prune_device_tree() - which set the board's USB reference-clock frequency and type from __cvmx_helper_board_usb_get_clock_type() - are guarded by "#ifdef CONFIG_USB", which is false when USB is built as a module. The fixups then silently disappear and octeon-hcd sees whatever default the DTS carries (12MHz crystal in octeon_3xxx.dts), leaving the PHY dead or the bus erroring on boards with a different reference clock. Use IS_ENABLED() so USB=m gets the same fixups as USB=y. Fixes: 7fd57ab9d9cf ("MIPS: Octeon: Fix compile error when USB is not enabled.") Assisted-by: Claude:claude-opus-5 Signed-off-by: Orgad Shaneh Signed-off-by: Thomas Bogendoerfer --- arch/mips/cavium-octeon/octeon-platform.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/arch/mips/cavium-octeon/octeon-platform.c b/arch/mips/cavium-octeon/octeon-platform.c index 47677b5d7ed0..f53c98372e0b 100644 --- a/arch/mips/cavium-octeon/octeon-platform.c +++ b/arch/mips/cavium-octeon/octeon-platform.c @@ -18,7 +18,7 @@ #include #include -#ifdef CONFIG_USB +#if IS_ENABLED(CONFIG_USB) #include #include #include @@ -1080,7 +1080,7 @@ int __init octeon_prune_device_tree(void) ; } -#ifdef CONFIG_USB +#if IS_ENABLED(CONFIG_USB) /* OHCI/UHCI USB */ alias_prop = fdt_getprop(initial_boot_params, aliases, "uctl", NULL); From 8b0cc8707f65e0f51912e764e1b309b2559db1ec Mon Sep 17 00:00:00 2001 From: Xu Rao Date: Tue, 18 Aug 2026 19:31:53 +0800 Subject: [PATCH 054/592] mmc: spi: reset bytes_xfered before retrying CRC failures mmc_spi_data_do() updates data->bytes_xfered after each block has been transferred successfully. If a later block in the same data request fails with a CRC error, data->bytes_xfered may therefore contain the number of bytes completed before the failing block. mmc_spi_request() has a private recovery path for such CRC failures. It sends STOP_TRANSMISSION, clears data->error and jumps back to crc_recover to issue the same command and data request again. However, it does not clear data->bytes_xfered before the retry. If the retry succeeds, the request is completed with the bytes from the failed attempt still included in data->bytes_xfered. For a multi-block request this can make the completed request report more bytes than were transferred by the successful retry, and can even exceed the request size when most blocks completed before the CRC error. This is most likely to be observed on MMC-over-SPI systems where long multi-block transfers occasionally hit a data CRC error but the mmc_spi-internal retry succeeds. The data itself is retried, but the completion accounting is not. Clear data->bytes_xfered together with data->error before repeating the request so the final completion reports only the bytes transferred by the successful attempt. Fixes: 061c6c847eeb ("mmc_spi: Recover from CRC errors for r/w operation over SPI.") Cc: stable@vger.kernel.org Signed-off-by: Xu Rao Signed-off-by: Ulf Hansson --- drivers/mmc/host/mmc_spi.c | 1 + 1 file changed, 1 insertion(+) diff --git a/drivers/mmc/host/mmc_spi.c b/drivers/mmc/host/mmc_spi.c index b471a7795b4d..5217d713c826 100644 --- a/drivers/mmc/host/mmc_spi.c +++ b/drivers/mmc/host/mmc_spi.c @@ -952,6 +952,7 @@ static void mmc_spi_request(struct mmc_host *mmc, struct mmc_request *mrq) status = mmc_spi_command_send(host, mrq, &stop, 0); crc_retry--; mrq->data->error = 0; + mrq->data->bytes_xfered = 0; goto crc_recover; } From 53823e25793a97d07e6e98e0904bbf74cac8bc76 Mon Sep 17 00:00:00 2001 From: Felix Gu Date: Sat, 22 Aug 2026 02:58:47 +0800 Subject: [PATCH 055/592] mmc: sdio_uart: fix xmit_fifo leak when the port table is full sdio_uart_add_port() allocates the transmit fifo before claiming a slot in sdio_uart_table[]. When all UART_NR slots are taken, it returns -EBUSY with the fifo still allocated, but the probe error path only kfree()s the port, leaking the transmit fifo. Free the fifo in the failure path of sdio_uart_add_port() itself so the function retains nothing on error. Fixes: 8b197a5ce7a7 ("sdio_uart: Use kfifo instead of the messy circ stuff") Signed-off-by: Felix Gu Cc: stable@vger.kernel.org Signed-off-by: Ulf Hansson --- drivers/mmc/core/sdio_uart.c | 3 +++ 1 file changed, 3 insertions(+) diff --git a/drivers/mmc/core/sdio_uart.c b/drivers/mmc/core/sdio_uart.c index 7fd5dedf3ac0..705163c42972 100644 --- a/drivers/mmc/core/sdio_uart.c +++ b/drivers/mmc/core/sdio_uart.c @@ -104,6 +104,9 @@ static int sdio_uart_add_port(struct sdio_uart_port *port) } spin_unlock(&sdio_uart_table_lock); + if (ret) + kfifo_free(&port->xmit_fifo); + return ret; } From 424103642d009f5a7dc33300a49d5606d1bf4fb5 Mon Sep 17 00:00:00 2001 From: Shivank Garg Date: Sat, 22 Aug 2026 19:22:03 +0000 Subject: [PATCH 056/592] dmaengine: add dma_device_get() helper Add dma_device_get() helper to match dma_device_put() to make code symmetric. It wraps open-coded kref_get_unless_zero() and asserts that dma_list_mutex is held, matching its put counterpart. No functional change intended. Suggested-by: Frank Li Reviewed-by: Logan Gunthorpe Signed-off-by: Shivank Garg Link: https://patch.msgid.link/20260822-dmaengine-kref-fix-v5-1-d4a4ee47d927@amd.com Signed-off-by: Vinod Koul --- drivers/dma/dmaengine.c | 9 +++++++-- 1 file changed, 7 insertions(+), 2 deletions(-) diff --git a/drivers/dma/dmaengine.c b/drivers/dma/dmaengine.c index 6ffd8bd82154..6b8af8607e5c 100644 --- a/drivers/dma/dmaengine.c +++ b/drivers/dma/dmaengine.c @@ -433,6 +433,12 @@ static void dma_device_release(struct kref *ref) device->device_release(device); } +static int __must_check dma_device_get(struct dma_device *device) +{ + lockdep_assert_held(&dma_list_mutex); + return kref_get_unless_zero(&device->ref); +} + static void dma_device_put(struct dma_device *device) { lockdep_assert_held(&dma_list_mutex); @@ -460,8 +466,7 @@ static int dma_chan_get(struct dma_chan *chan) if (!try_module_get(owner)) return -ENODEV; - ret = kref_get_unless_zero(&chan->device->ref); - if (!ret) { + if (!dma_device_get(chan->device)) { ret = -ENODEV; goto module_put_out; } From 44dab659064eb5c10adb0306510eebe848ed592d Mon Sep 17 00:00:00 2001 From: Shivank Garg Date: Sat, 22 Aug 2026 19:22:04 +0000 Subject: [PATCH 057/592] dmaengine: Fix device kref underflow in dma_chan_put() dma_chan_get() takes chan->device->ref only on the slow path: /* no kref on fast path */ if (chan->client_count) { __module_get(owner); chan->client_count++; return 0; } if (!try_module_get(owner)) return -ENODEV; if (!dma_device_get(chan->device)) { // calls kref_get_unless_zero() dma_chan_put() drops the ref unconditionally, so every fast-path get/put pair drops one extra device reference. The bug fires when two conditions hold together: a non-private provider has a persistent client holding chan->client_count > 0 and another client cycles dmaengine_get()/dmaengine_put(). When the kref hits zero, the subsequent dma_find_channel() returns NULL even though the provider module is still loaded. Fix this by dropping device->ref only on the last put, matching the single slow-path get. Fixes: 8ad342a86359 ("dmaengine: Add reference counting to dma_device struct") Reviewed-by: Frank Li Reviewed-by: Logan Gunthorpe Signed-off-by: Shivank Garg Link: https://patch.msgid.link/20260822-dmaengine-kref-fix-v5-2-d4a4ee47d927@amd.com Signed-off-by: Vinod Koul --- drivers/dma/dmaengine.c | 4 +++- 1 file changed, 3 insertions(+), 1 deletion(-) diff --git a/drivers/dma/dmaengine.c b/drivers/dma/dmaengine.c index 6b8af8607e5c..e380b801df73 100644 --- a/drivers/dma/dmaengine.c +++ b/drivers/dma/dmaengine.c @@ -520,7 +520,9 @@ static void dma_chan_put(struct dma_chan *chan) chan->route_data = NULL; } - dma_device_put(chan->device); + /* This channel is not in use anymore, drop the device ref */ + if (!chan->client_count) + dma_device_put(chan->device); module_put(dma_chan_to_owner(chan)); } From e873c74132f0c5f1452816cd9bb26208f0bba1e1 Mon Sep 17 00:00:00 2001 From: Shivank Garg Date: Sat, 22 Aug 2026 19:22:05 +0000 Subject: [PATCH 058/592] dmaengine: fix use-after-free in dma_chan_put() and dma_release_channel() When dma_device_put() drops the last reference on chan->device->ref, dma_device_release() runs and may free the dma_device along with its channels. dma_chan_put() then still reads chan->device->owner via dma_chan_to_owner() for the trailing module_put(). KASAN catches it: slab-use-after-free in dma_chan_put+0x3e6/0x4c0 Read of size 8 by task insmod/6319 Freed by task 6319: kfree+0x225/0x470 dma_chan_put+0x395/0x4c0 dmaengine_put+0xf8/0x160 Cache the module owner in dma_chan_put() before the put so the trailing module_put() does not need chan->device. Fixes: 8ad342a86359 ("dmaengine: Add reference counting to dma_device struct") Suggested-by: Sashiko Link: https://sashiko.dev/#/patchset/20260518-dmaengine-kref-fix-v1-1-4d6125048fb7@amd.com Reviewed-by: Frank Li Reviewed-by: Logan Gunthorpe Signed-off-by: Shivank Garg Link: https://patch.msgid.link/20260822-dmaengine-kref-fix-v5-3-d4a4ee47d927@amd.com Signed-off-by: Vinod Koul --- drivers/dma/dmaengine.c | 5 ++++- 1 file changed, 4 insertions(+), 1 deletion(-) diff --git a/drivers/dma/dmaengine.c b/drivers/dma/dmaengine.c index e380b801df73..872ae0d2ed95 100644 --- a/drivers/dma/dmaengine.c +++ b/drivers/dma/dmaengine.c @@ -500,10 +500,13 @@ static int dma_chan_get(struct dma_chan *chan) */ static void dma_chan_put(struct dma_chan *chan) { + struct module *owner; + /* This channel is not in use, bail out */ if (!chan->client_count) return; + owner = dma_chan_to_owner(chan); chan->client_count--; /* This channel is not in use anymore, free it */ @@ -523,7 +526,7 @@ static void dma_chan_put(struct dma_chan *chan) /* This channel is not in use anymore, drop the device ref */ if (!chan->client_count) dma_device_put(chan->device); - module_put(dma_chan_to_owner(chan)); + module_put(owner); } enum dma_status dma_sync_wait(struct dma_chan *chan, dma_cookie_t cookie) From dc750422170a563c7a81f6e49d36bb02c62ae37f Mon Sep 17 00:00:00 2001 From: Shivank Garg Date: Sat, 22 Aug 2026 19:22:06 +0000 Subject: [PATCH 059/592] dmaengine: wait for RCU readers before releasing dma_device dma_issue_pending_all() walks the dma_device_list with list_for_each_entry_rcu() under rcu_read_lock(). dma_device_release() unlinks the device with list_del_rcu() and then calls device->device_release() (which in many drivers, such as plx_dma.c, directly calls kfree()). Because there is no grace period between unlinking the device and freeing it, concurrent RCU readers in dma_issue_pending_all() can access the device after it has been freed. The lockless walk originally relied on clients holding a dmaengine reference to pin the provider module, and therefore the device, for as long as they might traverse the list. Commit 8ad342a86359 ("dmaengine: Add reference counting to dma_device struct") decoupled the dma_device lifetime from the module reference, so the device can now be released while a reader is still walking the list. Add synchronize_rcu() before the device is freed, so RCU readers are guaranteed to have finished. Keep it unconditional: providers that do not implement device_release() free the device themselves once dma_async_device_unregister() returns. This call will delay for a grace period with dma_list_mutex held, which is safe and only teardown path is delayed. Fixes: 2ba05622b8b1 ("dmaengine: provide a common 'issue_pending_all' implementation") Suggested-by: Sashiko Link: https://sashiko.dev/#/patchset/20260526-dmaengine-kref-fix-v2-0-3df60afac01d@amd.com Reviewed-by: Frank Li Reviewed-by: Logan Gunthorpe Signed-off-by: Shivank Garg Link: https://patch.msgid.link/20260822-dmaengine-kref-fix-v5-4-d4a4ee47d927@amd.com Signed-off-by: Vinod Koul --- drivers/dma/dmaengine.c | 1 + 1 file changed, 1 insertion(+) diff --git a/drivers/dma/dmaengine.c b/drivers/dma/dmaengine.c index 872ae0d2ed95..c71763047126 100644 --- a/drivers/dma/dmaengine.c +++ b/drivers/dma/dmaengine.c @@ -428,6 +428,7 @@ static void dma_device_release(struct kref *ref) list_del_rcu(&device->global_node); dma_channel_rebalance(); + synchronize_rcu(); if (device->device_release) device->device_release(device); From 955d86e5f3b95b731991fdb84966c50b16314629 Mon Sep 17 00:00:00 2001 From: Bradley Morgan Date: Sun, 9 Aug 2026 21:36:15 +0000 Subject: [PATCH 060/592] arm64: hibernate: pass HVC_SET_VECTORS args to the resume hvc swsusp_arch_suspend_exit() reinstalls the restored kernel's hyp stub vectors with an hvc, but never passes the arguments. x0 is not set to HVC_SET_VECTORS and x1 is not set to the vector address, so the stub dispatch falls through and returns without writing vbar_el2. EL2 is left pointing at the trans_pgd copy of the vectors, a page that swsusp_free() releases right after resume. Set the arguments up the same way __hyp_set_vectors() does. Without this fix, Vladimir was able to trigger a hang when resuming from hibernation with CONFIG_PAGE_POISONING=y and page_poison=on. Fixes: 788bfdd97434 ("arm64: trans_pgd: hibernate: Add trans_pgd_copy_el2_vectors") Cc: stable@vger.kernel.org Signed-off-by: Bradley Morgan Reviewed-by: Vladimir Murzin Tested-by: Vladimir Murzin Acked-by: Mark Rutland Signed-off-by: Will Deacon --- arch/arm64/kernel/hibernate-asm.S | 2 ++ 1 file changed, 2 insertions(+) diff --git a/arch/arm64/kernel/hibernate-asm.S b/arch/arm64/kernel/hibernate-asm.S index 0e1d9c3c6a93..2baefe7a82d3 100644 --- a/arch/arm64/kernel/hibernate-asm.S +++ b/arch/arm64/kernel/hibernate-asm.S @@ -89,6 +89,8 @@ alternative_insn "dc cvau, x4", "dc civac, x4", ARM64_WORKAROUND_CLEAN_CACHE isb cbz x24, 3f /* Do we need to re-initialise EL2? */ + mov x1, x24 + mov x0, #HVC_SET_VECTORS hvc #0 3: ret SYM_CODE_END(swsusp_arch_suspend_exit) From 3fc9f0558c95152b14d31badd5b27e5978bfb204 Mon Sep 17 00:00:00 2001 From: Rob Clark Date: Wed, 9 Sep 2026 07:41:54 -0700 Subject: [PATCH 061/592] drm/ci: Update xfails for kms_cursor_legacy regression These four started failing with some change not merged thru the msm tree. Update xfails to reflect reality. Signed-off-by: Rob Clark --- drivers/gpu/drm/ci/xfails/msm-apq8016-fails.txt | 7 +++++++ 1 file changed, 7 insertions(+) diff --git a/drivers/gpu/drm/ci/xfails/msm-apq8016-fails.txt b/drivers/gpu/drm/ci/xfails/msm-apq8016-fails.txt index 4546363447ff..0f3d85e4845e 100644 --- a/drivers/gpu/drm/ci/xfails/msm-apq8016-fails.txt +++ b/drivers/gpu/drm/ci/xfails/msm-apq8016-fails.txt @@ -7,3 +7,10 @@ kms_hdmi_inject@inject-4k,Fail kms_lease@lease-uevent,Fail msm/msm_mapping@memptrs,Fail msm/msm_mapping@ring,Fail + +# Started failing with v7.3-rc2 backmerge +# https://gitlab.freedesktop.org/drm/msm/-/work_items/104 +kms_cursor_legacy@single-move,Fail +kms_cursor_legacy@torture-bo,Fail +kms_cursor_legacy@forked-bo,Fail +kms_cursor_legacy@torture-move,Fail From e4a62833adff6ef0fe7c0b90393204fe3c26b5c5 Mon Sep 17 00:00:00 2001 From: Weiming Shi Date: Wed, 9 Sep 2026 12:08:08 +0800 Subject: [PATCH 062/592] bpf: Disallow bpf_skb_pull_data() for LWT_SEG6LOCAL An LWT_SEG6LOCAL program can invalidate its cached SRH with bpf_lwt_seg6_adjust_srh() and then call bpf_skb_pull_data(). The latter may reallocate skb->head, leaving the per-CPU SRH pointer dangling. Post-program SRH validation then writes through that pointer. Disallow bpf_skb_pull_data() for LWT_SEG6LOCAL programs so the verifier rejects this unsafe helper combination. Other LWT program types continue to expose the helper through lwt_out_func_proto(). Fixes: 004d4b274e2a ("ipv6: sr: Add seg6local action End.BPF") Reported-by: co+adfca3e91be95776@bugs.sh Suggested-by: Alexei Starovoitov Signed-off-by: Weiming Shi Signed-off-by: Daniel Borkmann Reviewed-by: Emil Tsalapatis Closes: https://lore.kernel.org/all/GCy0KRM2IcQGoJQTjJEU9D0maBxXzEDHuQpq@bugs.sh/ Link: https://lore.kernel.org/bpf/DL9COXZQXX4V.1FN45QO2Q77ZH@gmail.com/ Link: https://lore.kernel.org/bpf/20260909040807.3885815-2-bestswngs@gmail.com --- net/core/filter.c | 2 ++ 1 file changed, 2 insertions(+) diff --git a/net/core/filter.c b/net/core/filter.c index 8513167a858a..2a84f9d01131 100644 --- a/net/core/filter.c +++ b/net/core/filter.c @@ -9044,6 +9044,8 @@ static const struct bpf_func_proto * lwt_seg6local_func_proto(enum bpf_func_id func_id, const struct bpf_prog *prog) { switch (func_id) { + case BPF_FUNC_skb_pull_data: + return NULL; #if IS_ENABLED(CONFIG_IPV6_SEG6_BPF) case BPF_FUNC_lwt_seg6_store_bytes: return &bpf_lwt_seg6_store_bytes_proto; From 264bf9655c3d067d775a46f05eb8c871c488a864 Mon Sep 17 00:00:00 2001 From: Karl Mehltretter Date: Sun, 6 Sep 2026 19:09:25 +0200 Subject: [PATCH 063/592] scsi: pm80xx: Fix the use_msix, use_tasklet and read_wwn parameter descriptions The MODULE_PARM_DESC() lines of use_msix, use_tasklet and read_wwn all name a parameter zoned, which does not exist, and the use_tasklet one repeats the use_msix text. modinfo shows three "zoned" entries and no description for the real parameters. Name the right parameters and describe use_tasklet. Fixes: efa1fca45082 ("scsi: pm8001: Remove PM8001_USE_MSIX") Fixes: 205430290ad0 ("scsi: pm8001: Remove PM8001_USE_TASKLET") Fixes: 80975adc79dd ("scsi: pm8001: Remove PM8001_READ_VPD") Assisted-by: LLM Signed-off-by: Karl Mehltretter Reviewed-by: Damien Le Moal Link: https://patch.msgid.link/20260906170925.2524-1-kmehltretter@gmail.com Signed-off-by: Martin K. Petersen (Oracle) --- drivers/scsi/pm8001/pm8001_init.c | 6 +++--- 1 file changed, 3 insertions(+), 3 deletions(-) diff --git a/drivers/scsi/pm8001/pm8001_init.c b/drivers/scsi/pm8001/pm8001_init.c index 54b35893261a..5af81c73a8f8 100644 --- a/drivers/scsi/pm8001/pm8001_init.c +++ b/drivers/scsi/pm8001/pm8001_init.c @@ -58,15 +58,15 @@ MODULE_PARM_DESC(link_rate, "Enable link rate.\n" bool pm8001_use_msix = true; module_param_named(use_msix, pm8001_use_msix, bool, 0444); -MODULE_PARM_DESC(zoned, "Use MSIX interrupts. Default: true"); +MODULE_PARM_DESC(use_msix, "Use MSIX interrupts. Default: true"); static bool pm8001_use_tasklet = true; module_param_named(use_tasklet, pm8001_use_tasklet, bool, 0444); -MODULE_PARM_DESC(zoned, "Use MSIX interrupts. Default: true"); +MODULE_PARM_DESC(use_tasklet, "Use tasklets for interrupt handling. Default: true"); static bool pm8001_read_wwn = true; module_param_named(read_wwn, pm8001_read_wwn, bool, 0444); -MODULE_PARM_DESC(zoned, "Get WWN from the controller. Default: true"); +MODULE_PARM_DESC(read_wwn, "Get WWN from the controller. Default: true"); uint pcs_event_log_severity = 0x03; module_param(pcs_event_log_severity, int, 0644); From 779f202a92ef10a426efc07d0f4267918cb07ca3 Mon Sep 17 00:00:00 2001 From: Karl Mehltretter Date: Sun, 6 Sep 2026 19:10:09 +0200 Subject: [PATCH 064/592] scsi: qla2xxx: Fix the ql2xfc2target parameter description The module parameter is ql2xfc2target, but its MODULE_PARM_DESC() names qla2xfc2target, so modinfo describes a parameter that does not exist and shows no description for the real one. Use the parameter name in the description. Fixes: 877b03795fcf ("scsi: qla2xxx: Add option to disable FC2 Target support") Assisted-by: LLM Signed-off-by: Karl Mehltretter Link: https://patch.msgid.link/20260906171009.2560-1-kmehltretter@gmail.com Signed-off-by: Martin K. Petersen (Oracle) --- drivers/scsi/qla2xxx/qla_os.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/drivers/scsi/qla2xxx/qla_os.c b/drivers/scsi/qla2xxx/qla_os.c index c0efdbff5da7..3c412c7fb6fe 100644 --- a/drivers/scsi/qla2xxx/qla_os.c +++ b/drivers/scsi/qla2xxx/qla_os.c @@ -352,7 +352,7 @@ MODULE_PARM_DESC(ql2xnvme_queues, int ql2xfc2target = 1; module_param(ql2xfc2target, int, 0444); -MODULE_PARM_DESC(qla2xfc2target, +MODULE_PARM_DESC(ql2xfc2target, "Enables FC2 Target support. " "0 - FC2 Target support is disabled. " "1 - FC2 Target support is enabled (default)."); From 1274045b0eda1df5a5cdd3e63ed48cf013b3b4ea Mon Sep 17 00:00:00 2001 From: Venkat Rao Bagalkote Date: Thu, 3 Sep 2026 13:13:29 +0530 Subject: [PATCH 065/592] scsi: ibmvfc: Add Kconfig dependency to fix link failure when NVME_FC=m Building with CONFIG_SCSI_IBMVFC=y and CONFIG_NVME_FC=m results in a link failure: ibmvfc-nvme.o: undefined reference to `nvme_fc_register_localport' ibmvfc-nvme.o: undefined reference to `nvme_fc_register_remoteport' ibmvfc-nvme.o: undefined reference to `nvme_fc_unregister_localport' ibmvfc-nvme.o: undefined reference to `nvme_fc_unregister_remoteport' ibmvfc-core.o: undefined reference to `nvme_fc_rescan_remoteport' IS_ENABLED() evaluates to 1 for both =y and =m, so the nvme_fc_* call sites are kept in the object file. When SCSI_IBMVFC=y (built-in) but NVME_FC=m (loadable module), the linker cannot resolve these symbols at vmlinux link time. Add the same "depends on NVME_FC || NVME_FC=n" constraint already used by SCSI_LPFC to prevent this configuration. Reported-by: Pavithra Closes: https://lore.kernel.org/all/327877a29337aa526cc50ac88fbddb86@linux.ibm.com/ Signed-off-by: Venkat Rao Bagalkote Tested-by: Pavithra Acked-by: Tyrel Datwyler Link: https://patch.msgid.link/20260903074329.6705-1-venkat88@linux.ibm.com Signed-off-by: Martin K. Petersen (Oracle) --- drivers/scsi/Kconfig | 1 + 1 file changed, 1 insertion(+) diff --git a/drivers/scsi/Kconfig b/drivers/scsi/Kconfig index 4a2af0f702e1..1eec66195cf4 100644 --- a/drivers/scsi/Kconfig +++ b/drivers/scsi/Kconfig @@ -753,6 +753,7 @@ config SCSI_IBMVFC tristate "IBM Virtual FC support" depends on PPC_PSERIES && SCSI depends on SCSI_FC_ATTRS + depends on NVME_FC || NVME_FC=n help This is the IBM POWER Virtual FC Client From 0cb1fd924126f1f581621a5e804df98a02be9dff Mon Sep 17 00:00:00 2001 From: Arun Easi Date: Thu, 3 Sep 2026 10:55:47 -0700 Subject: [PATCH 066/592] scsi: fnic: Fix missed link-up when critical IRQ targets offline CPU When CPU Hyper Threading is disabled, sibling CPUs remain present but are reported offline. Managed MSI-X IRQs can still receive affinity masks that include those offline CPUs. If a driver-critical vector is managed, it can be parked on an offline CPU and the driver may miss critical events such as link-up. Keep driver-critical vectors unmanaged so they can be migrated by the IRQ core when their target CPU is offlined. Since HWQ-0 is unmanaged now, in some queue combinations there can be no mappings to it in mq_map. So without the blk-mq fix mentioned below, system may crash during cpu offline/online tests. Fixes: 8a8449ca5e33 ("scsi: fnic: Modify ISRs to support multiqueue (MQ)") Cc: stable@vger.kernel.org Depends-on: commit 10845a105bbc ("blk-mq: skip CPU offline notify on unmapped hctx") Reviewed-by: Sesidhar Baddela Reviewed-by: Arulprabhu Ponnusamy Reviewed-by: Gian Carlo Boffa Reviewed-by: Karan Tilak Kumar Signed-off-by: Arun Easi Reviewed-by: Laurence Oberman Link: https://patch.msgid.link/20260903175547.57971-1-aeasi@cisco.com Signed-off-by: Martin K. Petersen (Oracle) --- drivers/scsi/fnic/fnic.h | 2 +- drivers/scsi/fnic/fnic_isr.c | 13 ++++++++--- drivers/scsi/fnic/fnic_main.c | 41 ++++++++++++++++++++++++++++++++++- 3 files changed, 51 insertions(+), 5 deletions(-) diff --git a/drivers/scsi/fnic/fnic.h b/drivers/scsi/fnic/fnic.h index c576a7f5083e..3ba1592940ca 100644 --- a/drivers/scsi/fnic/fnic.h +++ b/drivers/scsi/fnic/fnic.h @@ -31,7 +31,7 @@ #define DRV_NAME "fnic" #define DRV_DESCRIPTION "Cisco FCoE HBA Driver" -#define DRV_VERSION "1.9.0.0" +#define DRV_VERSION "1.9.0.1" #define PFX DRV_NAME ": " #define DFX DRV_NAME "%d: " diff --git a/drivers/scsi/fnic/fnic_isr.c b/drivers/scsi/fnic/fnic_isr.c index 02856745580f..43149d8312ec 100644 --- a/drivers/scsi/fnic/fnic_isr.c +++ b/drivers/scsi/fnic/fnic_isr.c @@ -245,7 +245,14 @@ int fnic_set_intr_mode_msix(struct fnic *fnic) unsigned int m = ARRAY_SIZE(fnic->wq); unsigned int o = ARRAY_SIZE(fnic->hw_copy_wq); unsigned int min_irqs = n + m + 1 + 1; /*rq, raw wq, wq, err*/ - + /* + * Make driver critical vectors unmanaged, or else it can get tied + * to an offline CPU. This can happen when hyper-threading is off. + */ + struct irq_affinity affd = { + .pre_vectors = n + m + 1, /* rq, raw wq, 1 ioq */ + .post_vectors = 1, /* err */ + }; /* * We need n RQs, m WQs, o Copy WQs, n+m+o CQs, and n+m+o+1 INTRs * (last INTR is used for WQ/RQ errors and notification area) @@ -263,8 +270,8 @@ int fnic_set_intr_mode_msix(struct fnic *fnic) int vec_count = 0; int vecs = fnic->rq_count + fnic->raw_wq_count + fnic->wq_copy_count + 1; - vec_count = pci_alloc_irq_vectors(fnic->pdev, min_irqs, vecs, - PCI_IRQ_MSIX | PCI_IRQ_AFFINITY); + vec_count = pci_alloc_irq_vectors_affinity(fnic->pdev, min_irqs, + vecs, PCI_IRQ_MSIX|PCI_IRQ_AFFINITY, &affd); FNIC_ISR_DBG(KERN_INFO, fnic, "allocated %d MSI-X vectors\n", vec_count); diff --git a/drivers/scsi/fnic/fnic_main.c b/drivers/scsi/fnic/fnic_main.c index 9b3025007075..f13c381a66d7 100644 --- a/drivers/scsi/fnic/fnic_main.c +++ b/drivers/scsi/fnic/fnic_main.c @@ -744,8 +744,19 @@ static int fnic_nvme_drv_init(struct fnic *fnic) return ret; } +static void fnic_mq_init_queue_map(struct fnic *fnic, + struct blk_mq_queue_map *qmap) +{ + unsigned int cpu; + + for_each_possible_cpu(cpu) + qmap->mq_map[cpu] = 0; +} + void fnic_mq_map_queues_cpus(struct Scsi_Host *host) { + const struct cpumask *mask; + unsigned int queue, cpu; struct fnic *fnic = *((struct fnic **) shost_priv(host)); struct pci_dev *l_pdev = fnic->pdev; int intr_mode = fnic->config.intr_mode; @@ -766,7 +777,35 @@ void fnic_mq_map_queues_cpus(struct Scsi_Host *host) return; } - blk_mq_map_hw_queues(qmap, &l_pdev->dev, FNIC_PCI_OFFSET); + fnic_mq_init_queue_map(fnic, qmap); + + /* + * Setup CPU to Queue mapping for all managed MSI-X IRQs. + * Q0 is driver critical and non-managed, hence start from Q1. + */ + for (queue = 1; queue < qmap->nr_queues; queue++) { + int irq_num = pci_irq_vector(fnic->pdev, + queue + FNIC_PCI_OFFSET); + + if (irq_num < 0) + continue; + + mask = pci_irq_get_affinity(fnic->pdev, + queue + FNIC_PCI_OFFSET); + if (!mask) { + shost_printk(KERN_ERR, host, + "failed to get irq_affinity map for queue:%d\n", irq_num); + continue; + } + FNIC_MAIN_DBG(KERN_INFO, fnic, + "got irq_affinity map for %d:\n", irq_num); + for_each_cpu(cpu, mask) { + qmap->mq_map[cpu] = qmap->queue_offset + queue; + FNIC_MAIN_DBG(KERN_INFO, fnic, + "[Q%d] cpu:%d <=> irq:%d\n", + queue, cpu, irq_num); + } + } } static int fnic_probe(struct pci_dev *pdev, const struct pci_device_id *ent) From 3d676e458fe0c566f5a62753dc696b6a862fc412 Mon Sep 17 00:00:00 2001 From: Alberto Carboneri Date: Fri, 4 Sep 2026 13:54:37 +0000 Subject: [PATCH 067/592] scsi: core: Validate MODE SENSE lengths in scsi_cdl_enable() scsi_cdl_enable() uses length fields returned by MODE SENSE to locate the ATA feature mode page in a 64-byte stack buffer. A target can report a total length shorter than its mode header and block descriptors. The unsigned subtraction used for the MODE SELECT length can wrap, and the separately computed buf_data can point beyond buf. During automatic scan, enable is false, so the read-modify-write of buf_data[4] can clear the low two bits of a target-selected out-of-bounds stack byte. scsi_mode_select() can then copy up to 64 bytes from outside the buffer into the outgoing MODE SELECT payload, disclosing stack contents to the target. This is reachable while scanning a USB storage device that identifies as an ATA device and advertises CDL support. No filesystem mount or userspace access to the block device is required. On upstream commit cee9395acd80 ("Linux 7.3-rc1"), a build-specific, one-vCPU QEMU/Raw Gadget proof using QEMU-only multi-UDC allocator sampling executed a fixed proof command inside the guest and created a UID-0-owned marker during automatic enumeration, with KASLR and NX enabled. The issue was independently found during security research at Drivesec S.r.l. Cap the available length to the buffer size. Validate and consume the mode header and block descriptor lengths before using the page, and require the five bytes needed to access the CDL field. Fixes: 1b22cfb14142 ("scsi: core: Allow enabling and disabling command duration limits") Reported-by: Sashiko AI Review Closes: https://lore.kernel.org/linux-scsi/20260717192313.93D791F000E9@smtp.kernel.org/ Link: https://lore.kernel.org/linux-scsi/20260717222931.AC4EE1F000E9@smtp.kernel.org/ Link: https://lore.kernel.org/linux-scsi/df13ec87ac9b28e3b0a2d9eb26477e276ff0278a.camel@HansenPartnership.com/ Cc: stable@vger.kernel.org Assisted-by: LLM Co-developed-by: Pimen Flavian Dei (Drivesec S.r.l.) Signed-off-by: Pimen Flavian Dei (Drivesec S.r.l.) Signed-off-by: Alberto Carboneri (Drivesec S.r.l.) Link: https://lore.kernel.org/linux-scsi/20260717192313.93D791F000E9@smtp.kernel.org/ Reviewed-by: Damien Le Moal Link: https://patch.msgid.link/20260904135410.360314-1-acarboneri@drivesec.com Signed-off-by: Martin K. Petersen (Oracle) --- drivers/scsi/scsi.c | 24 +++++++++++++++++++----- 1 file changed, 19 insertions(+), 5 deletions(-) diff --git a/drivers/scsi/scsi.c b/drivers/scsi/scsi.c index 76cdad063f7b..f285521d9de6 100644 --- a/drivers/scsi/scsi.c +++ b/drivers/scsi/scsi.c @@ -727,6 +727,7 @@ int scsi_cdl_enable(struct scsi_device *sdev, bool enable) struct scsi_mode_data data; struct scsi_sense_hdr sshdr; char *buf_data; + size_t avail, offset; int len; ret = scsi_mode_sense(sdev, 0x08, 0x0a, 0xf2, buf, sizeof(buf), @@ -735,11 +736,24 @@ int scsi_cdl_enable(struct scsi_device *sdev, bool enable) return -EINVAL; /* Enable or disable CDL using the ATA feature page */ - len = min_t(size_t, sizeof(buf), - data.length - data.header_length - - data.block_descriptor_length); - buf_data = buf + data.header_length + - data.block_descriptor_length; + avail = min_t(size_t, data.length, sizeof(buf)); + if (data.header_length > avail) + return -EINVAL; + + offset = data.header_length; + avail -= data.header_length; + + if (data.block_descriptor_length > avail) + return -EINVAL; + + offset += data.block_descriptor_length; + avail -= data.block_descriptor_length; + + if (avail < 5) + return -EINVAL; + + buf_data = buf + offset; + len = avail; /* * If we want to enable CDL and CDL is already enabled on the From 9c182bc5d7817437a7d04ab96133f9191846d93d Mon Sep 17 00:00:00 2001 From: Florian Maillard Date: Mon, 24 Aug 2026 08:57:55 +0200 Subject: [PATCH 068/592] mmc: rtsx_pci_sdmmc: ignore broken write-protect on ThinkPad X260 The Realtek RTS522A card reader in the Lenovo ThinkPad X260 (subsystem 17aa:504a) incorrectly reports inserted SD cards as write-protected. This causes the MMC core to expose the card as read-only: mmcblk0: mmc0:aaaa SN256 238 GiB (ro) and /sys/block/mmcblk0/ro reports 1. Setting MMC_CAP2_NO_WRITE_PROTECT makes the card writable again. Limit the quirk to the affected Lenovo subsystem. Assisted-by: ChatGPT:GPT-5.6 Sol Signed-off-by: Florian Maillard Cc: stable@vger.kernel.org Signed-off-by: Ulf Hansson --- drivers/mmc/host/rtsx_pci_sdmmc.c | 5 +++++ 1 file changed, 5 insertions(+) diff --git a/drivers/mmc/host/rtsx_pci_sdmmc.c b/drivers/mmc/host/rtsx_pci_sdmmc.c index 8dfbc62f165b..d25d4cb59bc9 100644 --- a/drivers/mmc/host/rtsx_pci_sdmmc.c +++ b/drivers/mmc/host/rtsx_pci_sdmmc.c @@ -1425,6 +1425,11 @@ static void realtek_init_host(struct realtek_pci_sdmmc *host) mmc->caps = mmc->caps | MMC_CAP_AGGRESSIVE_PM; mmc->caps2 = MMC_CAP2_NO_PRESCAN_POWERUP | MMC_CAP2_FULL_PWR_CYCLE | MMC_CAP2_NO_SDIO; + + if (pcr->pci->device == 0x522a && + pcr->pci->subsystem_vendor == PCI_VENDOR_ID_LENOVO && + pcr->pci->subsystem_device == 0x504a) + mmc->caps2 |= MMC_CAP2_NO_WRITE_PROTECT; mmc->max_current_330 = 400; mmc->max_current_180 = 800; mmc->ops = &realtek_pci_sdmmc_ops; From 08b54e16d547d5c1aa61bf7a3595bb1620975eeb Mon Sep 17 00:00:00 2001 From: Zhu Ling Date: Fri, 4 Sep 2026 17:07:46 +0800 Subject: [PATCH 069/592] mmc: core: Fix OF node reference leak on card add failure mmc_of_find_child_device() returns a device node with its reference count incremented. mmc_add_card() stores the reference before calling device_add(), while the card is marked present only after device_add() succeeds. If device_add() fails, the callers release the card through mmc_remove_card(). However, mmc_remove_card() only drops the OF node reference for a present card, leaking the reference on this error path. Move of_node_put() outside the present-card conditional so the reference is released for both registered cards and card-add failures. Fixes: 25185f3f31c9 ("mmc: Add SDIO function devicetree subnode parsing") Cc: stable@vger.kernel.org Signed-off-by: Zhu Ling Reviewed-by: Shawn Lin Signed-off-by: Ulf Hansson --- drivers/mmc/core/bus.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/drivers/mmc/core/bus.c b/drivers/mmc/core/bus.c index be5cf338bdeb..92ad39c0cabc 100644 --- a/drivers/mmc/core/bus.c +++ b/drivers/mmc/core/bus.c @@ -417,8 +417,8 @@ void mmc_remove_card(struct mmc_card *card) mmc_hostname(card->host), card->rca); } device_del(&card->dev); - of_node_put(card->dev.of_node); } + of_node_put(card->dev.of_node); if (host->cqe_enabled) { host->cqe_ops->cqe_disable(host); From cee9c863ee68cb27d66745eb03f60e357f4f8ad2 Mon Sep 17 00:00:00 2001 From: Alex Bereza Date: Mon, 17 Aug 2026 11:23:55 +0200 Subject: [PATCH 070/592] dmaengine: xilinx_dma: Fix hardware buffer descriptor reuse order xilinx_dma_alloc_chan_resources() builds a static ring of hardware buffer descriptors once and the driver uses this ring throughout the lifetime of a channel. This requires the allocation order of hardware buffer descriptors from chan->free_seg_list to stay in sync with the hardware buffer descriptor ring built at channel allocation time by returning oldest descriptors to chan->free_seg_list first. When chan->pending_list is not empty e.g. during xilinx_dma_terminate_all() the chan->free_seg_list and the order of the static hardware buffer descriptor ring get out of sync. Descriptors age in this order: pending -> active -> done. So freeing pending_list first returns the newest buffer descriptors to the chan->free_seg_list first and thus breaks the order required by the static hardware buffer descriptor ring. Then when the channel is reused, after a wrap around of the free_seg_list the DMA will find a hardware buffer descriptor with a length field that is still zeroed and stop with something like this: xilinx-vdma 86000000.dma: Channel 000000003a21d7b8 has errors 10, cdr 6de4c000 tdr 6de4c000 After this no more descriptors are completed and a consumer potentially blocks and waits forever. The only way to get out of this error state is to rebuild the static hardware buffer descriptor ring and the free_seg_list by releasing and re-acquiring the channel. Fix the order in which hardware buffer descriptors are returned to free_seg_list to ensure the mentioned requirement holds. Fixes: 23059408b6a3 ("dmaengine: xilinx_dma: Fix race condition in the driver for multiple descriptor scenario") Signed-off-by: Alex Bereza Reviewed-by: Frank Li Reviewed-by: Suraj Gupta Link: https://patch.msgid.link/20260817-fix-hw-buf-desc-reuse-v1-1-d79827a844c7@bereza.email Signed-off-by: Vinod Koul --- drivers/dma/xilinx/xilinx_dma.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/drivers/dma/xilinx/xilinx_dma.c b/drivers/dma/xilinx/xilinx_dma.c index bef2b031dba1..0817b74f7450 100644 --- a/drivers/dma/xilinx/xilinx_dma.c +++ b/drivers/dma/xilinx/xilinx_dma.c @@ -920,9 +920,9 @@ static void xilinx_dma_free_descriptors(struct xilinx_dma_chan *chan) spin_lock_irqsave(&chan->lock, flags); - xilinx_dma_free_desc_list(chan, &chan->pending_list); xilinx_dma_free_desc_list(chan, &chan->done_list); xilinx_dma_free_desc_list(chan, &chan->active_list); + xilinx_dma_free_desc_list(chan, &chan->pending_list); spin_unlock_irqrestore(&chan->lock, flags); } From c90b6973daa37f4c283342dff881ae001dea4fe6 Mon Sep 17 00:00:00 2001 From: Christian Lugnberg Date: Mon, 17 Aug 2026 15:51:22 +0200 Subject: [PATCH 071/592] dmaengine: sun6i: fix non-atomic read of DMA position registers sun6i_get_chan_size() reads DMA_CHAN_LLI_ADDR and DMA_CHAN_CUR_CNT in two separate readl() calls with no synchronisation between them: pos = readl(pchan->base + DMA_CHAN_LLI_ADDR); bytes = readl(pchan->base + DMA_CHAN_CUR_CNT); DMA_CHAN_LLI_ADDR holds the physical address of the *next* descriptor the engine will load once the current one completes. DMA_CHAN_CUR_CNT holds the remaining byte count for the *current* descriptor. If the DMA engine advances to the next LLI entry between the two reads, pos becomes stale: it still points to what was the next descriptor at the time of the first read, but that descriptor is now the current one and CUR_CNT reflects its initial (full) byte count. The subsequent virtual-chain walk starts one entry too early and accumulates an extra full period's worth of bytes into the residue estimate. Fix this by re-reading DMA_CHAN_LLI_ADDR after DMA_CHAN_CUR_CNT and retrying if the value changed. This double-read pattern guarantees that both registers were sampled during the same descriptor interval. The cost is at most one extra readl() pair per call in the racy case, which occurs only at descriptor boundaries (~every 2 ms) and is negligible. Fixes: a90e173f3faf ("dmaengine: sun6i: Add cyclic capability") Cc: stable@vger.kernel.org Assisted-by: Claude:claude-sonnet-4-6 Signed-off-by: Christian Lugnberg Reviewed-by: Frank Li Link: https://patch.msgid.link/20260817135723.12807-2-christian.lugnberg@soundtrack.io Signed-off-by: Vinod Koul --- drivers/dma/sun6i-dma.c | 6 ++++-- 1 file changed, 4 insertions(+), 2 deletions(-) diff --git a/drivers/dma/sun6i-dma.c b/drivers/dma/sun6i-dma.c index f47a326dd7ff..04fe1f5042e9 100644 --- a/drivers/dma/sun6i-dma.c +++ b/drivers/dma/sun6i-dma.c @@ -354,8 +354,10 @@ static size_t sun6i_get_chan_size(struct sun6i_pchan *pchan) size_t bytes; dma_addr_t pos; - pos = readl(pchan->base + DMA_CHAN_LLI_ADDR); - bytes = readl(pchan->base + DMA_CHAN_CUR_CNT); + do { + pos = readl(pchan->base + DMA_CHAN_LLI_ADDR); + bytes = readl(pchan->base + DMA_CHAN_CUR_CNT); + } while (pos != readl(pchan->base + DMA_CHAN_LLI_ADDR)); if (pos == LLI_LAST_ITEM) return bytes; From 9096bdc8d930147f7c39a493a859acbd3a8485d8 Mon Sep 17 00:00:00 2001 From: Christian Lugnberg Date: Mon, 17 Aug 2026 15:51:23 +0200 Subject: [PATCH 072/592] dmaengine: sun6i: fix undefined behaviour in sun6i_dma_tx_status sun6i_dma_tx_status() calls vchan_find_desc() to look up the virtual descriptor for a given cookie, before checking whether the pointer vd is NULL: vd = vchan_find_desc(&vchan->vc, cookie); txd = to_sun6i_desc(&vd->tx); /* vd may be NULL here */ if (vd) { for (lli = txd->v_lli; ...) vchan_find_desc() returns NULL when the descriptor has already been completed or is in-flight on a physical channel and no longer present in the virtual channel's descriptor list. When vd is NULL, to_sun6i_desc() is called unconditionally on &vd->tx before the NULL check, which is undefined behaviour. Move the call inside the if (vd) guard to ensure it is only reached with a valid pointer. vd = vchan_find_desc(&vchan->vc, cookie); if (vd) { struct sun6i_desc *txd = to_sun6i_desc(&vd->tx); for (lli = txd->v_lli; ...) Fixes: 555859308723 ("dmaengine: sun6i: Add driver for the Allwinner A31 DMA controller") Cc: stable@vger.kernel.org Assisted-by: Claude:claude-sonnet-4-6 Signed-off-by: Christian Lugnberg Reviewed-by: Frank Li Link: https://patch.msgid.link/20260817135723.12807-3-christian.lugnberg@soundtrack.io Signed-off-by: Vinod Koul --- drivers/dma/sun6i-dma.c | 3 +-- 1 file changed, 1 insertion(+), 2 deletions(-) diff --git a/drivers/dma/sun6i-dma.c b/drivers/dma/sun6i-dma.c index 04fe1f5042e9..7704b016aed8 100644 --- a/drivers/dma/sun6i-dma.c +++ b/drivers/dma/sun6i-dma.c @@ -981,7 +981,6 @@ static enum dma_status sun6i_dma_tx_status(struct dma_chan *chan, struct sun6i_pchan *pchan = vchan->phy; struct sun6i_dma_lli *lli; struct virt_dma_desc *vd; - struct sun6i_desc *txd; enum dma_status ret; unsigned long flags; size_t bytes = 0; @@ -993,9 +992,9 @@ static enum dma_status sun6i_dma_tx_status(struct dma_chan *chan, spin_lock_irqsave(&vchan->vc.lock, flags); vd = vchan_find_desc(&vchan->vc, cookie); - txd = to_sun6i_desc(&vd->tx); if (vd) { + struct sun6i_desc *txd = to_sun6i_desc(&vd->tx); for (lli = txd->v_lli; lli != NULL; lli = lli->v_lli_next) bytes += lli->len; } else if (!pchan || !pchan->desc) { From f6504be006aa4bb4bd26285f410a885c17920d65 Mon Sep 17 00:00:00 2001 From: Sascha Hauer Date: Mon, 17 Aug 2026 22:44:33 +0200 Subject: [PATCH 073/592] dmaengine: pxa: fix double counting of the hw descriptors pxad_alloc_desc() was converted from kzalloc(struct_size(sw_desc, hw_desc, nb_hw_desc), GFP_NOWAIT) to kzalloc_flex(), which sets the __counted_by() counter sw_desc->nb_desc itself - but only where the compiler has __builtin_counted_by_ref(), so from gcc 15.1 or clang 22.1 on. The loop below it still increments nb_desc, which makes it come out doubled there and correct elsewhere. nb_desc is what pxad_free_desc() iterates over and what set_updater_desc() indexes from, so set it explicitly and drop the increment. The error path has to lower it to the number of descriptors allocated so far, otherwise pxad_free_desc() would free entries that were never allocated. Fixes: 69050f8d6d075 ("treewide: Replace kmalloc with kmalloc_obj for non-scalar types") Assisted-by: Claude:claude-opus-5 Signed-off-by: Sascha Hauer Reviewed-by: Frank Li Link: https://lore.kernel.org/r/20260817-dmaengine-pxa-v1-1-850c215c1196@pengutronix.de Link: https://patch.msgid.link/20260817-dmaengine-pxa-v2-1-f42ab0569a48@pengutronix.de Signed-off-by: Vinod Koul --- drivers/dma/pxa_dma.c | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-) diff --git a/drivers/dma/pxa_dma.c b/drivers/dma/pxa_dma.c index fa2ee0b3e09f..fc43124fefa8 100644 --- a/drivers/dma/pxa_dma.c +++ b/drivers/dma/pxa_dma.c @@ -744,6 +744,7 @@ pxad_alloc_desc(struct pxad_chan *chan, unsigned int nb_hw_desc) sw_desc = kzalloc_flex(*sw_desc, hw_desc, nb_hw_desc, GFP_NOWAIT); if (!sw_desc) return NULL; + sw_desc->nb_desc = nb_hw_desc; sw_desc->desc_pool = chan->desc_pool; for (i = 0; i < nb_hw_desc; i++) { @@ -752,10 +753,10 @@ pxad_alloc_desc(struct pxad_chan *chan, unsigned int nb_hw_desc) dev_err(&chan->vc.chan.dev->device, "%s(): Couldn't allocate the %dth hw_desc from dma_pool %p\n", __func__, i, sw_desc->desc_pool); + sw_desc->nb_desc = i; goto err; } - sw_desc->nb_desc++; sw_desc->hw_desc[i] = desc; if (i == 0) From 7ed1e3070c9b4bbd67d5519e14711038dd53ab13 Mon Sep 17 00:00:00 2001 From: Alex Bereza Date: Tue, 18 Aug 2026 09:36:29 +0200 Subject: [PATCH 074/592] dmaengine: xilinx_dma: Fix hardware buffer descriptor chain after cyclic DMA Using the DMA in cyclic mode modifies the hardware buffer descriptor chain in xilinx_dma_prep_dma_cyclic so that the last descriptor used by the cyclic transfer points back to the first descriptor, but it never restores the original descriptor ring. This breaks using non-cyclic mode after cyclic mode with an error like: xilinx-vdma 86000000.dma: Channel 00000000354d5c8d has errors 100, cdr 6de40000 tdr 6de40400 The only way to get out of this error state is to rebuild the hardware buffer descriptor ring by releasing and re-acquiring the channel. Fix using non-cyclic mode after cyclic mode by always restoring the original buffer descriptor ring in the same manner as it is set up by xilinx_dma_alloc_chan_resources(). Fixes: 23059408b6a3 ("dmaengine: xilinx_dma: Fix race condition in the driver for multiple descriptor scenario") Signed-off-by: Alex Bereza Reviewed-by: Frank Li Reviewed-by: Suraj Gupta Link: https://patch.msgid.link/20260817-fix-hw-buf-desc-after-cyclic-mode-v1-1-1fe47e701d6c@bereza.email Link: https://patch.msgid.link/20260818-fix-hw-buf-desc-after-cyclic-mode-v2-1-530ff44c6a81@bereza.email Signed-off-by: Vinod Koul --- drivers/dma/xilinx/xilinx_dma.c | 24 +++++++++++++++++------- 1 file changed, 17 insertions(+), 7 deletions(-) diff --git a/drivers/dma/xilinx/xilinx_dma.c b/drivers/dma/xilinx/xilinx_dma.c index 0817b74f7450..cffe7c6fa640 100644 --- a/drivers/dma/xilinx/xilinx_dma.c +++ b/drivers/dma/xilinx/xilinx_dma.c @@ -756,15 +756,25 @@ xilinx_aximcdma_alloc_tx_segment(struct xilinx_dma_chan *chan) return segment; } -static void xilinx_dma_clean_hw_desc(struct xilinx_axidma_desc_hw *hw) +static void xilinx_dma_clean_hw_desc(struct xilinx_dma_chan *chan, + struct xilinx_axidma_tx_segment *segment) { - u32 next_desc = hw->next_desc; - u32 next_desc_msb = hw->next_desc_msb; + dma_addr_t next; + u32 i; - memset(hw, 0, sizeof(struct xilinx_axidma_desc_hw)); + /* + * Restore the buffer descriptor's next descriptor pointer to the value + * set up in xilinx_dma_alloc_chan_resources(). Otherwise using the DMA + * in cyclic mode leaves the next descriptor pointer altered and + * prevents subsequent non-cyclic transfers. + */ + i = segment - chan->seg_v; + next = chan->seg_p + + sizeof(*chan->seg_v) * ((i + 1) % XILINX_DMA_NUM_DESCS); - hw->next_desc = next_desc; - hw->next_desc_msb = next_desc_msb; + memset(&segment->hw, 0, sizeof(segment->hw)); + segment->hw.next_desc = lower_32_bits(next); + segment->hw.next_desc_msb = upper_32_bits(next); } static void xilinx_mcdma_clean_hw_desc(struct xilinx_aximcdma_desc_hw *hw) @@ -786,7 +796,7 @@ static void xilinx_mcdma_clean_hw_desc(struct xilinx_aximcdma_desc_hw *hw) static void xilinx_dma_free_tx_segment(struct xilinx_dma_chan *chan, struct xilinx_axidma_tx_segment *segment) { - xilinx_dma_clean_hw_desc(&segment->hw); + xilinx_dma_clean_hw_desc(chan, segment); list_add_tail(&segment->node, &chan->free_seg_list); } From 3db7d7d583419f7b1f2e141e36418802dbb25cf8 Mon Sep 17 00:00:00 2001 From: Vadim Nikitushkin Date: Wed, 9 Sep 2026 23:50:28 +0300 Subject: [PATCH 075/592] drm/ttm: fix swapped-out resources never leaving their bulk_move range MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit ttm_tt_swapout() returns the number of pages swapped out on success and a negative error code on failure; for a populated ttm it never returns zero. Commit b2ed01e7ad3d ("drm/ttm: Fix ttm_bo_swapout() infinite LRU walk on swapout failure") moved the bulk_move bookkeeping in ttm_bo_swapout_cb() under "if (!ret)", so the ttm_resource_del_bulk_move_unevictable() / ttm_resource_move_to_lru_tail() pair is now skipped on every successful swapout. The equivalent change for the shrinker in commit 1d59f36e95f7 ("drm/ttm: Fix ttm_bo_shrink() infinite LRU walk on backup failure") tests "lret > 0", which is what was intended here as well. Before b2ed01e7ad3d the resource was taken off the bulk_move before the swapout; since then a swapped-out resource stays inside its BO's bulk_move range (and on the manager LRU) although it is unevictable. When it is later freed or the BO leaves the bulk_move (ttm_resource_free(), ttm_bo_set_bulk_move() via amdgpu_vm_bo_del()), ttm_resource_del_bulk_move() skips it because of its !ttm_resource_unevictable() guard, so a range endpoint in pos->first / pos->last is left pointing at freed memory. The next ttm_lru_bulk_move_tail() or ttm_resource_add_bulk_move() on that cursor is a use-after-free, seen as the resv WARN in ttm_lru_bulk_move_add(), "list_del corruption" in ttm_resource_move_to_lru_tail() or a NULL dereference in ttm_resource_manager_next() -- minutes to hours after a hibernation, or at process exit / reboot following one. Samuel Ainsworth's analysis of drm/amd issue 5387 (see Link) identified the dangling cursor; the missing removal at swapout time is the reason it dangles. Testing the condition for success restores the removal. On an AMD Phoenix APU (ASUS UM3406GA, gfx1103) running suspend-then-hibernate on a 7.0.y stable kernel carrying the backport (Ubuntu 7.0.0-31) the bug crashed 5 of 18 hibernation cycles; a function profile of one hibernation showed 336 ttm_tt_swapout() calls and zero ttm_resource_del_bulk_move_unevictable() calls. With this change the removal happens for every swapped-out resource and 12 further cycles were clean. Fixes: b2ed01e7ad3d ("drm/ttm: Fix ttm_bo_swapout() infinite LRU walk on swapout failure") Cc: stable@vger.kernel.org # v7.1+ Closes: https://gitlab.freedesktop.org/drm/amd/-/issues/5387 Link: https://lore.kernel.org/dri-devel/CAHYiNPa6aVacJoLOje-qZ1GyYx-9p0tN4NuP8D_eSL+UJeevXw@mail.gmail.com/ Signed-off-by: Vadim Nikitushkin Reviewed-by: Thomas Hellström Reviewed-by: Christian König Signed-off-by: Christian König Link: https://lore.kernel.org/r/20260909205028.13799-1-bub4z0r@gmail.com --- drivers/gpu/drm/ttm/ttm_bo.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/drivers/gpu/drm/ttm/ttm_bo.c b/drivers/gpu/drm/ttm/ttm_bo.c index ef56c18ded1b..a12af5b38a31 100644 --- a/drivers/gpu/drm/ttm/ttm_bo.c +++ b/drivers/gpu/drm/ttm/ttm_bo.c @@ -532,7 +532,7 @@ static int ttm_bo_alloc_at_place(struct ttm_buffer_object *bo, ret = ttm_resource_try_charge(bo, place, &alloc_state->charge_pool, force_space ? &alloc_state->limit_pool : NULL); - if (ret) { + if (ret > 0) { /* * -EAGAIN means the charge failed, which we treat * like an allocation failure. Therefore, return an From fcfe64715b425262af1b36f498f9197f3537ceed Mon Sep 17 00:00:00 2001 From: Vadim Nikitushkin Date: Thu, 10 Sep 2026 17:34:51 +0300 Subject: [PATCH 076/592] drm/ttm: apply the swapout bulk_move fix to the intended condition MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit Commit 3db7d7d58341 ("drm/ttm: fix swapped-out resources never leaving their bulk_move range") landed in drm-misc-fixes with its one-line change applied to the wrong "if": the "if (ret)" after ttm_resource_try_charge() in ttm_bo_alloc_at_place() became "if (ret > 0)", while the "if (!ret)" after ttm_tt_swapout() in ttm_bo_swapout_cb() that the patch targeted was left untouched. ttm_resource_try_charge() returns 0 or a negative error code, so with "ret > 0" a failed dmem cgroup charge no longer fails the allocation. Restore that check and apply the intended change: ttm_tt_swapout() returns the number of pages swapped out on success, so the bulk_move removal must run for ret > 0. Fixes: 3db7d7d58341 ("drm/ttm: fix swapped-out resources never leaving their bulk_move range") Cc: stable@vger.kernel.org # v7.1+ Signed-off-by: Vadim Nikitushkin Reviewed-by: Christian König Signed-off-by: Christian König Link: https://lore.kernel.org/r/20260910143451.65853-1-bub4z0r@gmail.com --- drivers/gpu/drm/ttm/ttm_bo.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/drivers/gpu/drm/ttm/ttm_bo.c b/drivers/gpu/drm/ttm/ttm_bo.c index a12af5b38a31..9b85b5f388d4 100644 --- a/drivers/gpu/drm/ttm/ttm_bo.c +++ b/drivers/gpu/drm/ttm/ttm_bo.c @@ -532,7 +532,7 @@ static int ttm_bo_alloc_at_place(struct ttm_buffer_object *bo, ret = ttm_resource_try_charge(bo, place, &alloc_state->charge_pool, force_space ? &alloc_state->limit_pool : NULL); - if (ret > 0) { + if (ret) { /* * -EAGAIN means the charge failed, which we treat * like an allocation failure. Therefore, return an @@ -1434,7 +1434,7 @@ ttm_bo_swapout_cb(struct ttm_lru_walk *walk, struct ttm_buffer_object *bo) if (ttm_tt_is_populated(tt)) { ret = ttm_tt_swapout(bdev, tt, swapout_walk->gfp_flags); - if (!ret) { + if (ret > 0) { spin_lock(&bdev->lru_lock); ttm_resource_del_bulk_move_unevictable(bo->resource, bo); ttm_resource_move_to_lru_tail(bo->resource); From adf5967331318bcb436fc80069915231ec039352 Mon Sep 17 00:00:00 2001 From: Karl Mehltretter Date: Sun, 6 Sep 2026 19:03:47 +0200 Subject: [PATCH 077/592] drm/msm: Fix the separate_gpu_kms parameter description The module parameter is separate_gpu_kms, but its MODULE_PARM_DESC() names separate_gpu_drm, so modinfo describes a parameter that does not exist and shows no description for the real one. Use the parameter name in the description. Fixes: 217ed15bd399 ("drm/msm: enable separate binding of GPU and display devices") Assisted-by: LLM Signed-off-by: Karl Mehltretter Patchwork: https://patchwork.freedesktop.org/patch/751407/ Message-ID: <20260906170347.2427-1-kmehltretter@gmail.com> Signed-off-by: Rob Clark --- drivers/gpu/drm/msm/msm_drv.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/drivers/gpu/drm/msm/msm_drv.c b/drivers/gpu/drm/msm/msm_drv.c index db1b655dd055..f3d2eaa04f14 100644 --- a/drivers/gpu/drm/msm/msm_drv.c +++ b/drivers/gpu/drm/msm/msm_drv.c @@ -55,7 +55,7 @@ MODULE_PARM_DESC(modeset, "Use kernel modesetting [KMS] (1=on (default), 0=disab module_param(modeset, bool, 0600); static bool separate_gpu_kms; -MODULE_PARM_DESC(separate_gpu_drm, "Use separate DRM device for the GPU (0=single DRM device for both GPU and display (default), 1=two DRM devices)"); +MODULE_PARM_DESC(separate_gpu_kms, "Use separate DRM device for the GPU (0=single DRM device for both GPU and display (default), 1=two DRM devices)"); module_param(separate_gpu_kms, bool, 0400); DECLARE_FAULT_ATTR(fail_gem_alloc); From 8061ee61b9426fe38350fa9eead2d9c50b03deb6 Mon Sep 17 00:00:00 2001 From: Karl Mehltretter Date: Sun, 6 Sep 2026 19:03:01 +0200 Subject: [PATCH 078/592] drm/msm/adreno: Fix the skip_gpu parameter description The module parameter is skip_gpu, but its MODULE_PARM_DESC() names no_gpu, so modinfo describes a parameter that does not exist and shows no description for the real one. Use the parameter name in the description. Fixes: 3f17991488af ("drm/msm/adreno: Add a modparam to skip GPU") Assisted-by: LLM Signed-off-by: Karl Mehltretter Patchwork: https://patchwork.freedesktop.org/patch/751406/ Message-ID: <20260906170301.2393-1-kmehltretter@gmail.com> Signed-off-by: Rob Clark --- drivers/gpu/drm/msm/adreno/adreno_device.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/drivers/gpu/drm/msm/adreno/adreno_device.c b/drivers/gpu/drm/msm/adreno/adreno_device.c index 7f20320ef66a..05c77fe27e62 100644 --- a/drivers/gpu/drm/msm/adreno/adreno_device.c +++ b/drivers/gpu/drm/msm/adreno/adreno_device.c @@ -25,7 +25,7 @@ MODULE_PARM_DESC(disable_acd, "Forcefully disable GPU ACD"); module_param_unsafe(disable_acd, bool, 0400); static bool skip_gpu; -MODULE_PARM_DESC(no_gpu, "Disable GPU driver register (0=enable GPU driver register (default), 1=skip GPU driver register"); +MODULE_PARM_DESC(skip_gpu, "Disable GPU driver register (0=enable GPU driver register (default), 1=skip GPU driver register"); module_param(skip_gpu, bool, 0400); extern const struct adreno_gpulist a2xx_gpulist; From 01c8d1f385f788f1bbbbb7687c4386d614281218 Mon Sep 17 00:00:00 2001 From: "Jonghyuk Kim(MalHyuk)" Date: Wed, 2 Sep 2026 10:27:20 +0900 Subject: [PATCH 079/592] drm/msm: RCU-free the scheduler-containing ring and VM objects Both struct msm_ringbuffer and struct msm_gem_vm embed a struct drm_gpu_scheduler. msm_ringbuffer_destroy() and the VM free callback msm_gem_vm_free() call drm_sched_fini() on the embedded scheduler and then free the containing object with plain kfree(). drm_sched_fence_get_timeline_name() returns fence->sched->name, and the scheduler fence keeps a .release callback so it is not ops-detached on signalling. A finished fence exported to userspace (the submit out-fence, or a VM_BIND fence, via sync_file / drm_syncobj) keeps pointing at the embedded scheduler after the ring/VM is freed, so a later get_timeline_name() -- reachable unprivileged through SYNC_IOC_FILE_INFO -- dereferences freed slab memory (KASAN slab-use-after-free read). Per the dma-fence lifetime contract the exporter must keep the data backing a signalled fence alive for an RCU grace period. Free the scheduler-containing objects with kfree_rcu() instead of kfree(). Fixes: 1d8a5ca436ee ("drm/msm: Conversion to drm scheduler") Fixes: 92395af63a99 ("drm/msm: Add VM_BIND submitqueue") Cc: stable@vger.kernel.org Signed-off-by: Jonghyuk Kim(MalHyuk) Patchwork: https://patchwork.freedesktop.org/patch/750234/ Message-ID: <20260902012720.880783-1-malhyuk97@gmail.com> Signed-off-by: Rob Clark --- drivers/gpu/drm/msm/msm_gem.h | 3 +++ drivers/gpu/drm/msm/msm_gem_vma.c | 2 +- drivers/gpu/drm/msm/msm_ringbuffer.c | 2 +- drivers/gpu/drm/msm/msm_ringbuffer.h | 1 + 4 files changed, 6 insertions(+), 2 deletions(-) diff --git a/drivers/gpu/drm/msm/msm_gem.h b/drivers/gpu/drm/msm/msm_gem.h index dff60cbc9d95..7c6a8c01f910 100644 --- a/drivers/gpu/drm/msm/msm_gem.h +++ b/drivers/gpu/drm/msm/msm_gem.h @@ -68,6 +68,9 @@ struct msm_gem_vm { /** @base: Inherit from drm_gpuvm. */ struct drm_gpuvm base; + /** @rcu: RCU-delayed free so an exported sched fence->sched stays valid. */ + struct rcu_head rcu; + /** * @sched: Scheduler used for asynchronous VM_BIND request. * diff --git a/drivers/gpu/drm/msm/msm_gem_vma.c b/drivers/gpu/drm/msm/msm_gem_vma.c index c11d021581e0..1badec3caa7b 100644 --- a/drivers/gpu/drm/msm/msm_gem_vma.c +++ b/drivers/gpu/drm/msm/msm_gem_vma.c @@ -166,7 +166,7 @@ msm_gem_vm_free(struct drm_gpuvm *gpuvm) dma_fence_put(vm->last_fence); put_pid(vm->pid); kfree(vm->log); - kfree(vm); + kfree_rcu(vm, rcu); } /** diff --git a/drivers/gpu/drm/msm/msm_ringbuffer.c b/drivers/gpu/drm/msm/msm_ringbuffer.c index 59c69aa75649..38e1e6866301 100644 --- a/drivers/gpu/drm/msm/msm_ringbuffer.c +++ b/drivers/gpu/drm/msm/msm_ringbuffer.c @@ -140,5 +140,5 @@ void msm_ringbuffer_destroy(struct msm_ringbuffer *ring) msm_gem_kernel_put(ring->bo, ring->gpu->vm); - kfree(ring); + kfree_rcu(ring, rcu); } diff --git a/drivers/gpu/drm/msm/msm_ringbuffer.h b/drivers/gpu/drm/msm/msm_ringbuffer.h index 3631ec283c6e..0bfb6b0f42fa 100644 --- a/drivers/gpu/drm/msm/msm_ringbuffer.h +++ b/drivers/gpu/drm/msm/msm_ringbuffer.h @@ -55,6 +55,7 @@ struct msm_ringbuffer { /* * The job scheduler for this ring. */ + struct rcu_head rcu; struct drm_gpu_scheduler sched; bool sched_initialized; From 88ce88e933e40f4ab4b81a5e9d0a10328583f593 Mon Sep 17 00:00:00 2001 From: Daniel Borkmann Date: Thu, 10 Sep 2026 23:35:07 +0200 Subject: [PATCH 080/592] bpf: Add KF_PERFMON kfunc flag Tracing related BPF helpers e.g. under bpf_base_func_proto() are gated behind CAP_PERFMON. However, the same is currently not true for kfuncs and they are accessible via plain CAP_BPF. Add a new KF_PERFMON flag which can be used such that check_kfunc_call() ensures env->allow_ptr_leaks is permitted. This follows similar pattern to existing KF_DESTRUCTIVE flag. The rejection returns -EPERM to match the other CAP_PERFMON gates in the verifier, that is, check_ptr_to_btf_access() and check_ptr_to_map_access(), which report the very same policy to user space. Signed-off-by: Daniel Borkmann Link: https://lore.kernel.org/r/20260910213510.49358-1-daniel@iogearbox.net Signed-off-by: Alexei Starovoitov --- Documentation/bpf/kfuncs.rst | 10 ++++++++++ include/linux/btf.h | 1 + kernel/bpf/verifier.c | 14 ++++++++++++++ 3 files changed, 25 insertions(+) diff --git a/Documentation/bpf/kfuncs.rst b/Documentation/bpf/kfuncs.rst index 85f73e0bbd0f..6691fe8a32c3 100644 --- a/Documentation/bpf/kfuncs.rst +++ b/Documentation/bpf/kfuncs.rst @@ -486,6 +486,16 @@ Example usage in BPF program: /* note that the last argument is omitted */ bpf_task_work_schedule_signal(task, &work->tw, &arrmap, task_work_callback); +2.5.10 KF_PERFMON flag +---------------------- + +The KF_PERFMON flag is used for kfuncs that can expose kernel memory or kernel +addresses to the BPF program, for example by reading through a pointer that the +verifier does not check. Calling such a kfunc requires CAP_PERFMON, or +CAP_SYS_ADMIN, in the same way that the equivalent BPF helpers are gated in +bpf_base_func_proto(). A program loaded with CAP_BPF alone is rejected at load +time. + 2.6 Registering the kfuncs -------------------------- diff --git a/include/linux/btf.h b/include/linux/btf.h index 89d5a5c4f117..7c62ea17b116 100644 --- a/include/linux/btf.h +++ b/include/linux/btf.h @@ -80,6 +80,7 @@ #define KF_ARENA_ARG2 (1 << 15) /* kfunc takes an arena pointer as its second argument */ #define KF_IMPLICIT_ARGS (1 << 16) /* kfunc has implicit arguments supplied by the verifier */ #define KF_SPINLOCK_SAFE (1 << 17) /* kfunc is allowed inside bpf_spin_lock-ed region */ +#define KF_PERFMON (1 << 18) /* kfunc requires CAP_PERFMON */ /* * Tag marking a kernel function as a kfunc. This is meant to minimize the diff --git a/kernel/bpf/verifier.c b/kernel/bpf/verifier.c index 72a3f5998dd2..939e535a3442 100644 --- a/kernel/bpf/verifier.c +++ b/kernel/bpf/verifier.c @@ -11356,6 +11356,11 @@ static bool is_kfunc_destructive(struct bpf_call_arg_meta *meta) return meta->kfunc_flags & KF_DESTRUCTIVE; } +static bool is_kfunc_perfmon(struct bpf_call_arg_meta *meta) +{ + return meta->kfunc_flags & KF_PERFMON; +} + static bool is_kfunc_rcu(struct bpf_call_arg_meta *meta) { return meta->kfunc_flags & KF_RCU; @@ -13834,6 +13839,15 @@ static int check_kfunc_call(struct bpf_verifier_env *env, struct bpf_insn *insn, return -EACCES; } + if (is_kfunc_perfmon(&meta) && !env->allow_ptr_leaks) { + verbose(env, "%s is allowed only to CAP_PERFMON and CAP_SYS_ADMIN\n", + func_name); + operation = bpf_diag_fmt(env, "kfunc %s", func_name); + bpf_diag_policy(env, insn_idx, operation, "the kfunc requires CAP_PERFMON", + "Load the program with CAP_PERFMON, or avoid the kfunc."); + return -EPERM; + } + sleepable = bpf_is_kfunc_sleepable(&meta); if (sleepable && !in_sleepable(env)) { verbose(env, "program must be sleepable to call sleepable kfunc %s\n", func_name); From 81c975aae375d2053fa2926ee3730f02e713140c Mon Sep 17 00:00:00 2001 From: Daniel Borkmann Date: Thu, 10 Sep 2026 23:35:08 +0200 Subject: [PATCH 081/592] bpf: Require CAP_PERFMON for kfuncs reading memory Mark fault-safe probe reading kfuncs as KF_PERFMON, similarly as we do for the old-style BPF helper equivalents. bpf_rdonly_cast() is included in this list as well as it returns PTR_TO_MEM | MEM_RDONLY | PTR_UNTRUSTED for an unchecked object and is using fault-safe BPF_PROBE_MEM. Note that only the void form of bpf_rdonly_cast() produced a register that was readable without CAP_PERFMON. For a struct type id the kfunc returns PTR_TO_BTF_ID | PTR_UNTRUSTED, whose dereference has always been gated in check_ptr_to_btf_access(). The flag is not conditional on the type id, so for the latter it only moves the rejection from the dereference to the call itself, which is the better place to report it anyway. The bpf_stream_vprintk() and bpf_stream_print_stack() kfuncs are marked as well. The former ends up in the same bpf_bprintf_prepare() as the bpf_snprintf() helper, where %pks, %pus and %pI4 read through a program- supplied address and %pB resolves one into a symbol. The latter walks the stack and prints each instruction pointer via %pS. Lastly, bpf_get_kmem_cache() takes a raw scalar address that the verifier does not constrain and dereferences the page/slab metadata for it. Field reads on the returned pointer are still blocked (PTR_TO_BTF_ID | PTR_UNTRUSTED -> check_ptr_to_btf_access() results in -EPERM without the CAP), but the NULL/non-NULL result is observable. Reported-by: STAR Labs SG Signed-off-by: Daniel Borkmann Link: https://lore.kernel.org/r/20260910213510.49358-2-daniel@iogearbox.net Signed-off-by: Alexei Starovoitov --- kernel/bpf/helpers.c | 60 ++++++++++++++++++++++---------------------- 1 file changed, 30 insertions(+), 30 deletions(-) diff --git a/kernel/bpf/helpers.c b/kernel/bpf/helpers.c index b3cc5c8fc875..712dca5a2c5b 100644 --- a/kernel/bpf/helpers.c +++ b/kernel/bpf/helpers.c @@ -4883,7 +4883,7 @@ BTF_ID(func, bpf_cgroup_release_dtor) BTF_KFUNCS_START(common_btf_ids) BTF_ID_FLAGS(func, bpf_cast_to_kern_ctx, KF_FASTCALL) -BTF_ID_FLAGS(func, bpf_rdonly_cast, KF_FASTCALL) +BTF_ID_FLAGS(func, bpf_rdonly_cast, KF_FASTCALL | KF_PERFMON) BTF_ID_FLAGS(func, bpf_rcu_read_lock) BTF_ID_FLAGS(func, bpf_rcu_read_unlock) BTF_ID_FLAGS(func, bpf_dynptr_slice, KF_RET_NULL) @@ -4920,26 +4920,26 @@ BTF_ID_FLAGS(func, bpf_wq_set_callback, KF_IMPLICIT_ARGS) BTF_ID_FLAGS(func, bpf_wq_start) BTF_ID_FLAGS(func, bpf_preempt_disable) BTF_ID_FLAGS(func, bpf_preempt_enable) -BTF_ID_FLAGS(func, bpf_iter_bits_new, KF_ITER_NEW) +BTF_ID_FLAGS(func, bpf_iter_bits_new, KF_ITER_NEW | KF_PERFMON) BTF_ID_FLAGS(func, bpf_iter_bits_next, KF_ITER_NEXT | KF_RET_NULL) BTF_ID_FLAGS(func, bpf_iter_bits_destroy, KF_ITER_DESTROY) -BTF_ID_FLAGS(func, bpf_copy_from_user_str, KF_SLEEPABLE) -BTF_ID_FLAGS(func, bpf_copy_from_user_task_str, KF_SLEEPABLE) -BTF_ID_FLAGS(func, bpf_get_kmem_cache) +BTF_ID_FLAGS(func, bpf_copy_from_user_str, KF_SLEEPABLE | KF_PERFMON) +BTF_ID_FLAGS(func, bpf_copy_from_user_task_str, KF_SLEEPABLE | KF_PERFMON) +BTF_ID_FLAGS(func, bpf_get_kmem_cache, KF_PERFMON) BTF_ID_FLAGS(func, bpf_iter_kmem_cache_new, KF_ITER_NEW | KF_SLEEPABLE) BTF_ID_FLAGS(func, bpf_iter_kmem_cache_next, KF_ITER_NEXT | KF_RET_NULL | KF_SLEEPABLE) BTF_ID_FLAGS(func, bpf_iter_kmem_cache_destroy, KF_ITER_DESTROY | KF_SLEEPABLE) BTF_ID_FLAGS(func, bpf_local_irq_save) BTF_ID_FLAGS(func, bpf_local_irq_restore) #ifdef CONFIG_BPF_EVENTS -BTF_ID_FLAGS(func, bpf_probe_read_user_dynptr) -BTF_ID_FLAGS(func, bpf_probe_read_kernel_dynptr) -BTF_ID_FLAGS(func, bpf_probe_read_user_str_dynptr) -BTF_ID_FLAGS(func, bpf_probe_read_kernel_str_dynptr) -BTF_ID_FLAGS(func, bpf_copy_from_user_dynptr, KF_SLEEPABLE) -BTF_ID_FLAGS(func, bpf_copy_from_user_str_dynptr, KF_SLEEPABLE) -BTF_ID_FLAGS(func, bpf_copy_from_user_task_dynptr, KF_SLEEPABLE) -BTF_ID_FLAGS(func, bpf_copy_from_user_task_str_dynptr, KF_SLEEPABLE) +BTF_ID_FLAGS(func, bpf_probe_read_user_dynptr, KF_PERFMON) +BTF_ID_FLAGS(func, bpf_probe_read_kernel_dynptr, KF_PERFMON) +BTF_ID_FLAGS(func, bpf_probe_read_user_str_dynptr, KF_PERFMON) +BTF_ID_FLAGS(func, bpf_probe_read_kernel_str_dynptr, KF_PERFMON) +BTF_ID_FLAGS(func, bpf_copy_from_user_dynptr, KF_SLEEPABLE | KF_PERFMON) +BTF_ID_FLAGS(func, bpf_copy_from_user_str_dynptr, KF_SLEEPABLE | KF_PERFMON) +BTF_ID_FLAGS(func, bpf_copy_from_user_task_dynptr, KF_SLEEPABLE | KF_PERFMON) +BTF_ID_FLAGS(func, bpf_copy_from_user_task_str_dynptr, KF_SLEEPABLE | KF_PERFMON) #endif #ifdef CONFIG_DMA_SHARED_BUFFER BTF_ID_FLAGS(func, bpf_iter_dmabuf_new, KF_ITER_NEW | KF_SLEEPABLE) @@ -4947,26 +4947,26 @@ BTF_ID_FLAGS(func, bpf_iter_dmabuf_next, KF_ITER_NEXT | KF_RET_NULL | KF_SLEEPAB BTF_ID_FLAGS(func, bpf_iter_dmabuf_destroy, KF_ITER_DESTROY | KF_SLEEPABLE) #endif BTF_ID_FLAGS(func, __bpf_trap) -BTF_ID_FLAGS(func, bpf_strcmp); -BTF_ID_FLAGS(func, bpf_strcasecmp); -BTF_ID_FLAGS(func, bpf_strncasecmp); -BTF_ID_FLAGS(func, bpf_strchr); -BTF_ID_FLAGS(func, bpf_strchrnul); -BTF_ID_FLAGS(func, bpf_strnchr); -BTF_ID_FLAGS(func, bpf_strrchr); -BTF_ID_FLAGS(func, bpf_strlen); -BTF_ID_FLAGS(func, bpf_strnlen); -BTF_ID_FLAGS(func, bpf_strspn); -BTF_ID_FLAGS(func, bpf_strcspn); -BTF_ID_FLAGS(func, bpf_strstr); -BTF_ID_FLAGS(func, bpf_strcasestr); -BTF_ID_FLAGS(func, bpf_strnstr); -BTF_ID_FLAGS(func, bpf_strncasestr); +BTF_ID_FLAGS(func, bpf_strcmp, KF_PERFMON); +BTF_ID_FLAGS(func, bpf_strcasecmp, KF_PERFMON); +BTF_ID_FLAGS(func, bpf_strncasecmp, KF_PERFMON); +BTF_ID_FLAGS(func, bpf_strchr, KF_PERFMON); +BTF_ID_FLAGS(func, bpf_strchrnul, KF_PERFMON); +BTF_ID_FLAGS(func, bpf_strnchr, KF_PERFMON); +BTF_ID_FLAGS(func, bpf_strrchr, KF_PERFMON); +BTF_ID_FLAGS(func, bpf_strlen, KF_PERFMON); +BTF_ID_FLAGS(func, bpf_strnlen, KF_PERFMON); +BTF_ID_FLAGS(func, bpf_strspn, KF_PERFMON); +BTF_ID_FLAGS(func, bpf_strcspn, KF_PERFMON); +BTF_ID_FLAGS(func, bpf_strstr, KF_PERFMON); +BTF_ID_FLAGS(func, bpf_strcasestr, KF_PERFMON); +BTF_ID_FLAGS(func, bpf_strnstr, KF_PERFMON); +BTF_ID_FLAGS(func, bpf_strncasestr, KF_PERFMON); #if defined(CONFIG_BPF_LSM) && defined(CONFIG_CGROUPS) BTF_ID_FLAGS(func, bpf_cgroup_read_xattr, KF_RCU) #endif -BTF_ID_FLAGS(func, bpf_stream_vprintk, KF_IMPLICIT_ARGS | KF_SPINLOCK_SAFE) -BTF_ID_FLAGS(func, bpf_stream_print_stack, KF_IMPLICIT_ARGS | KF_SPINLOCK_SAFE) +BTF_ID_FLAGS(func, bpf_stream_vprintk, KF_IMPLICIT_ARGS | KF_SPINLOCK_SAFE | KF_PERFMON) +BTF_ID_FLAGS(func, bpf_stream_print_stack, KF_IMPLICIT_ARGS | KF_SPINLOCK_SAFE | KF_PERFMON) BTF_ID_FLAGS(func, bpf_task_work_schedule_signal, KF_IMPLICIT_ARGS) BTF_ID_FLAGS(func, bpf_task_work_schedule_resume, KF_IMPLICIT_ARGS) BTF_ID_FLAGS(func, bpf_dynptr_from_file) From f9191460cd805e84eace0884c75c67ec719fa1d8 Mon Sep 17 00:00:00 2001 From: Daniel Borkmann Date: Thu, 10 Sep 2026 23:35:09 +0200 Subject: [PATCH 082/592] bpf: Require CAP_PERFMON for untrusted read-only memory reads Marking bpf_rdonly_cast() KF_PERFMON CAP-limits one producer of PTR_TO_MEM | MEM_RDONLY | PTR_UNTRUSTED, but not the type itself. A global subprogram argument tagged __arg_untrusted results in the same register with no kfunc call. Reported-by: STAR Labs SG Signed-off-by: Daniel Borkmann Link: https://lore.kernel.org/r/20260910213510.49358-3-daniel@iogearbox.net Signed-off-by: Alexei Starovoitov --- kernel/bpf/verifier.c | 9 +++++++++ 1 file changed, 9 insertions(+) diff --git a/kernel/bpf/verifier.c b/kernel/bpf/verifier.c index 939e535a3442..8058f684a9ea 100644 --- a/kernel/bpf/verifier.c +++ b/kernel/bpf/verifier.c @@ -6453,6 +6453,15 @@ static int check_mem_access(struct bpf_verifier_env *env, int insn_idx, struct b return -EACCES; } + if (rdonly_untrusted && !env->allow_ptr_leaks) { + verbose(env, "%s access is allowed only to CAP_PERFMON and CAP_SYS_ADMIN\n", + reg_type_str(env, reg->type)); + bpf_diag_policy(env, insn_idx, "read from untrusted read-only memory", + "the access requires CAP_PERFMON", + "Load the program with CAP_PERFMON, or avoid dereferencing untrusted pointers."); + return -EPERM; + } + /* * Accesses to untrusted PTR_TO_MEM are done through probe * instructions, hence no need to check bounds in that case. From a903f145a8914d9ecba009b398d135154e4a5d94 Mon Sep 17 00:00:00 2001 From: Daniel Borkmann Date: Thu, 10 Sep 2026 23:35:10 +0200 Subject: [PATCH 083/592] selftests/bpf: Add tests for the KF_PERFMON gates Add test cases where each one loads with CAP_BPF alone and checks that the program is correctly rejected. # LDLIBS=-static PKG_CONFIG='pkg-config --static' ./vmtest.sh -- ./test_progs -t verifier_kfunc_perfmon [...] #627/1 verifier_kfunc_perfmon/rdonly_cast_noperfmon:OK #627/2 verifier_kfunc_perfmon/rdonly_cast_noperfmon @unpriv:OK #627/3 verifier_kfunc_perfmon/probe_read_kernel_dynptr_noperfmon:OK #627/4 verifier_kfunc_perfmon/probe_read_kernel_dynptr_noperfmon @unpriv:OK #627/5 verifier_kfunc_perfmon/stream_vprintk_noperfmon:OK #627/6 verifier_kfunc_perfmon/stream_vprintk_noperfmon @unpriv:OK #627/7 verifier_kfunc_perfmon/get_kmem_cache_noperfmon:OK #627/8 verifier_kfunc_perfmon/get_kmem_cache_noperfmon @unpriv:OK #627/9 verifier_kfunc_perfmon/arg_untrusted_read_noperfmon:OK #627/10 verifier_kfunc_perfmon/arg_untrusted_read_noperfmon @unpriv:OK #627 verifier_kfunc_perfmon:OK Summary: 1/10 PASSED, 0 SKIPPED, 0/0 FAILED Signed-off-by: Daniel Borkmann Link: https://lore.kernel.org/r/20260910213510.49358-4-daniel@iogearbox.net Signed-off-by: Alexei Starovoitov --- .../selftests/bpf/prog_tests/verifier.c | 2 + .../bpf/progs/verifier_kfunc_perfmon.c | 75 +++++++++++++++++++ 2 files changed, 77 insertions(+) create mode 100644 tools/testing/selftests/bpf/progs/verifier_kfunc_perfmon.c diff --git a/tools/testing/selftests/bpf/prog_tests/verifier.c b/tools/testing/selftests/bpf/prog_tests/verifier.c index 64ac49ad67e6..ec9e3907c75d 100644 --- a/tools/testing/selftests/bpf/prog_tests/verifier.c +++ b/tools/testing/selftests/bpf/prog_tests/verifier.c @@ -53,6 +53,7 @@ #include "verifier_iterating_callbacks.skel.h" #include "verifier_jeq_infer_not_null.skel.h" #include "verifier_jit_convergence.skel.h" +#include "verifier_kfunc_perfmon.skel.h" #include "verifier_ld_ind.skel.h" #include "verifier_ldsx.skel.h" #include "verifier_leak_ptr.skel.h" @@ -216,6 +217,7 @@ void test_verifier_int_ptr(void) { RUN(verifier_int_ptr); } void test_verifier_iterating_callbacks(void) { RUN(verifier_iterating_callbacks); } void test_verifier_jeq_infer_not_null(void) { RUN(verifier_jeq_infer_not_null); } void test_verifier_jit_convergence(void) { RUN(verifier_jit_convergence); } +void test_verifier_kfunc_perfmon(void) { RUN(verifier_kfunc_perfmon); } void test_verifier_load_acquire(void) { RUN(verifier_load_acquire); } void test_verifier_ld_ind(void) { RUN(verifier_ld_ind); } void test_verifier_ldsx(void) { RUN(verifier_ldsx); } diff --git a/tools/testing/selftests/bpf/progs/verifier_kfunc_perfmon.c b/tools/testing/selftests/bpf/progs/verifier_kfunc_perfmon.c new file mode 100644 index 000000000000..76c39ef30e96 --- /dev/null +++ b/tools/testing/selftests/bpf/progs/verifier_kfunc_perfmon.c @@ -0,0 +1,75 @@ +// SPDX-License-Identifier: GPL-2.0 + +#include +#include +#include "bpf_misc.h" + +void *user_ptr; +char dynptr_buf[8]; +u64 kaddr; + +extern struct kmem_cache *bpf_get_kmem_cache(u64 addr) __ksym; + +SEC("socket") +__success +__caps_unpriv(CAP_BPF) +__failure_unpriv +__msg_unpriv("bpf_rdonly_cast is allowed only to CAP_PERFMON and CAP_SYS_ADMIN") +int rdonly_cast_noperfmon(void *ctx) +{ + char *p = bpf_rdonly_cast(0, 0); + + return p[0x7fff]; +} + +SEC("socket") +__success +__caps_unpriv(CAP_BPF) +__failure_unpriv +__msg_unpriv("bpf_probe_read_kernel_dynptr is allowed only to CAP_PERFMON and CAP_SYS_ADMIN") +int probe_read_kernel_dynptr_noperfmon(void *ctx) +{ + struct bpf_dynptr dptr; + + bpf_dynptr_from_mem(dynptr_buf, sizeof(dynptr_buf), 0, &dptr); + bpf_probe_read_kernel_dynptr(&dptr, 0, sizeof(dynptr_buf), user_ptr); + return 0; +} + +SEC("socket") +__success +__caps_unpriv(CAP_BPF) +__failure_unpriv +__msg_unpriv("bpf_stream_vprintk is allowed only to CAP_PERFMON and CAP_SYS_ADMIN") +int stream_vprintk_noperfmon(void *ctx) +{ + bpf_stream_printk(BPF_STDOUT, "%pB", (void *)kaddr); + return 0; +} + +SEC("socket") +__success +__caps_unpriv(CAP_BPF) +__failure_unpriv +__msg_unpriv("bpf_get_kmem_cache is allowed only to CAP_PERFMON and CAP_SYS_ADMIN") +int get_kmem_cache_noperfmon(void *ctx) +{ + return !!bpf_get_kmem_cache(kaddr); +} + +__weak int subprog_untrusted_read(void *p __arg_untrusted) +{ + return *(char *)p; +} + +SEC("socket") +__success +__caps_unpriv(CAP_BPF) +__failure_unpriv +__msg_unpriv("rdonly_untrusted_mem access is allowed only to CAP_PERFMON and CAP_SYS_ADMIN") +int arg_untrusted_read_noperfmon(void *ctx) +{ + return subprog_untrusted_read(0); +} + +char _license[] SEC("license") = "GPL"; From 2936aed9b02162df7fbe08fcd6bfbb3270ad9f95 Mon Sep 17 00:00:00 2001 From: Daniel Borkmann Date: Thu, 10 Sep 2026 20:56:04 +0200 Subject: [PATCH 084/592] bpf: Clear scalar delta on narrowing stack spill check_stack_write_fixed_off() breaks the scalar link on a narrowing spill by zeroing the id of the destination stack slot, but leaves the delta in place. save_register_state() has just copied the source register verbatim, so the slot keeps the BPF_ADD_CONST delta of a register it is no longer linked to. This is the same shape as the case addressed in 1b327732c846 ("bpf: Clear delta when clearing reg id for non-{add,sub} ops"). Unlike the latter, no miscomputation seems reachable, so mainly consistency. Signed-off-by: Daniel Borkmann Acked-by: Eduard Zingerman Link: https://lore.kernel.org/r/20260910185604.44949-1-daniel@iogearbox.net Signed-off-by: Alexei Starovoitov --- kernel/bpf/verifier.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/kernel/bpf/verifier.c b/kernel/bpf/verifier.c index 8058f684a9ea..1b0b1fb62878 100644 --- a/kernel/bpf/verifier.c +++ b/kernel/bpf/verifier.c @@ -3582,7 +3582,7 @@ static int check_stack_write_fixed_off(struct bpf_verifier_env *env, save_register_state(env, state, spi, reg, size); /* Break the relation on a narrowing spill. */ if (!reg_value_fits) - state->stack[spi].spilled_ptr.id = 0; + clear_scalar_id(&state->stack[spi].spilled_ptr); } else if (!reg && !(off % BPF_REG_SIZE) && is_bpf_st_mem(insn) && env->bpf_capable) { struct bpf_reg_state *tmp_reg = &env->fake_reg[0]; From b0b3dc66529676228cb938cbcad66920f735c223 Mon Sep 17 00:00:00 2001 From: Jiayuan Chen Date: Thu, 10 Sep 2026 20:22:55 +0800 Subject: [PATCH 085/592] bpf: Fix divide-by-zero in btf_struct_walk() When an access goes past the struct and the last member is a flexible array, btf_struct_walk() folds the offset back into a single element with (off - moff) % t->size, but never checks that the element type has a size. BTF takes an empty struct, so this in program BTF /* event could be empty */ struct event { #ifdef HAVE_TIMESTAMP __u64 ts; #endif }; struct batch { int nr; struct event events[]; }; divides by zero at prog load time. Getting there needs a PTR_TO_BTF_ID that is not MEM_ALLOC, e.g. a plain read of a local kptr stashed in a map from a sleepable program. Oops: divide error: 0000 [#1] SMP KASAN PTI RIP: 0010:btf_struct_walk+0x53f/0x1570 Call Trace: btf_struct_access+0x42a/0xcd0 check_ptr_to_btf_access+0x4dc/0x1160 check_mem_access+0x3a45/0x8740 check_load_mem+0x36a/0xd10 do_check_common+0x3ef0/0xb210 bpf_check+0x6d3b/0x8580 bpf_prog_load+0xf7c/0x2720 __sys_bpf+0xa83/0x3690 __x64_sys_bpf+0xc7/0x150 x64_sys_call+0x1f3f/0x27e0 do_syscall_64+0xe5/0x610 entry_SYSCALL_64_after_hwframe+0x76/0x7e Reject a zero-sized element type. The fixed array path in the same function already bails out on the same thing: btf_struct_walk() ... /* skip empty array */ if (moff == mtrue_end) continue; msize /= total_nelems; Fixes: 9c5f8a1008a1 ("bpf: Support variable length array in tracing programs") Signed-off-by: Jiayuan Chen Acked-by: Eduard Zingerman Link: https://lore.kernel.org/r/20260910122316.186384-1-jiayuan.chen@linux.dev Signed-off-by: Alexei Starovoitov --- kernel/bpf/btf.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/kernel/bpf/btf.c b/kernel/bpf/btf.c index 9f33e95d5741..d67a169ba2bb 100644 --- a/kernel/bpf/btf.c +++ b/kernel/bpf/btf.c @@ -7187,7 +7187,7 @@ static int btf_struct_walk(struct bpf_verifier_log *log, const struct btf *btf, if (btf_type_is_int(t)) return WALK_SCALAR; - if (!btf_type_is_struct(t)) + if (!btf_type_is_struct(t) || !t->size) goto error; off = (off - moff) % t->size; From f77d21245710690f3fb02d19d8dd4dc17ee58c2c Mon Sep 17 00:00:00 2001 From: Jiayuan Chen Date: Thu, 10 Sep 2026 20:22:56 +0800 Subject: [PATCH 086/592] selftests/bpf: Test BTF walk into a flexible array of zero-sized elements The program stashes a bpf_obj_new() object whose type ends with a flexible array of empty structs, then reads it back as an untrusted kptr. Without the previous patch this divides by zero in btf_struct_walk() instead of being rejected. # ./test_progs -t verifier_btf_flex_array ... #602 verifier_btf_flex_array:OK Summary: 1/1 PASSED, 0 SKIPPED, 0/0 FAILED Signed-off-by: Jiayuan Chen Link: https://lore.kernel.org/r/20260910122316.186384-2-jiayuan.chen@linux.dev Signed-off-by: Alexei Starovoitov --- .../selftests/bpf/prog_tests/verifier.c | 2 + .../bpf/progs/verifier_btf_flex_array.c | 56 +++++++++++++++++++ 2 files changed, 58 insertions(+) create mode 100644 tools/testing/selftests/bpf/progs/verifier_btf_flex_array.c diff --git a/tools/testing/selftests/bpf/prog_tests/verifier.c b/tools/testing/selftests/bpf/prog_tests/verifier.c index ec9e3907c75d..8b439e194bcc 100644 --- a/tools/testing/selftests/bpf/prog_tests/verifier.c +++ b/tools/testing/selftests/bpf/prog_tests/verifier.c @@ -23,6 +23,7 @@ #include "verifier_bpf_trap.skel.h" #include "verifier_bswap.skel.h" #include "verifier_btf_ctx_access.skel.h" +#include "verifier_btf_flex_array.skel.h" #include "verifier_btf_unreliable_prog.skel.h" #include "verifier_call_large_imm.skel.h" #include "verifier_cfg.skel.h" @@ -187,6 +188,7 @@ void test_verifier_bpf_get_stack(void) { RUN(verifier_bpf_get_stack); } void test_verifier_bpf_trap(void) { RUN(verifier_bpf_trap); } void test_verifier_bswap(void) { RUN(verifier_bswap); } void test_verifier_btf_ctx_access(void) { RUN(verifier_btf_ctx_access); } +void test_verifier_btf_flex_array(void) { RUN(verifier_btf_flex_array); } void test_verifier_btf_unreliable_prog(void) { RUN(verifier_btf_unreliable_prog); } void test_verifier_call_large_imm(void) { RUN(verifier_call_large_imm); } void test_verifier_cfg(void) { RUN(verifier_cfg); } diff --git a/tools/testing/selftests/bpf/progs/verifier_btf_flex_array.c b/tools/testing/selftests/bpf/progs/verifier_btf_flex_array.c new file mode 100644 index 000000000000..59b84261f622 --- /dev/null +++ b/tools/testing/selftests/bpf/progs/verifier_btf_flex_array.c @@ -0,0 +1,56 @@ +// SPDX-License-Identifier: GPL-2.0 + +#include +#include + +#include "bpf_experimental.h" +#include "bpf_misc.h" + +struct test_empty_event {}; + +struct test_flex_batch { + int nr; + struct test_empty_event events[]; +}; + +struct map_value { + struct test_flex_batch __kptr *batch; +}; + +struct { + __uint(type, BPF_MAP_TYPE_ARRAY); + __type(key, int); + __type(value, struct map_value); + __uint(max_entries, 1); +} batches SEC(".maps"); + +SEC("syscall") +__description("btf walk into flexible array of zero-sized elements") +__failure __msg("access beyond struct test_flex_batch at off 4 size 1") +int stash_and_peek(void *ctx) +{ + struct test_flex_batch *b, *old; + struct map_value *v; + int key = 0; + + v = bpf_map_lookup_elem(&batches, &key); + if (!v) + return 0; + + b = bpf_obj_new(struct test_flex_batch); + if (!b) + return 0; + b->nr = 1; + + old = bpf_kptr_xchg(&v->batch, b); + if (old) + bpf_obj_drop(old); + + b = v->batch; + if (!b) + return 0; + + return b->nr + *(char *)&b->events[0]; +} + +char _license[] SEC("license") = "GPL"; From 01b245ba016d44861690594e10f67e026ce8552f Mon Sep 17 00:00:00 2001 From: Jiayuan Chen Date: Thu, 10 Sep 2026 19:26:26 +0800 Subject: [PATCH 087/592] bpf: Fix out-of-bounds read of sk_protocol in bpf_sock_destroy() sk_protocol lives in struct sock, not in struct sock_common. A timewait or request sock handed to bpf_sock_destroy() by the tcp iterator is neither, so reading sk->sk_protocol runs past the object: ================================================================== BUG: KASAN: slab-out-of-bounds in bpf_sock_destroy+0xc7/0xe0 Read of size 2 at addr ffff8881047d11b4 by task test_progs/428 Tainted: [W]=WARN Call Trace: dump_stack_lvl+0x91/0xf0 print_report+0xd1/0x630 kasan_report+0xf3/0x130 __asan_report_load2_noabort+0x14/0x30 bpf_sock_destroy+0xc7/0xe0 bpf_prog_c3dd61f9d9cd9f37_iter_tcp6_timewait+0x9f/0xb7 bpf_iter_run_prog+0x538/0xde0 bpf_iter_tcp_seq_show+0x26b/0x4b0 bpf_seq_read+0x424/0x1210 vfs_read+0x197/0xe40 ksys_read+0x119/0x240 __x64_sys_read+0x72/0xc0 x64_sys_call+0x647/0x27e0 do_syscall_64+0xe5/0x610 entry_SYSCALL_64_after_hwframe+0x76/0x7e Only check sk_protocol on full socks. tcp_abort() already knows how to deal with TIME_WAIT and NEW_SYN_RECV socks. Also fix the comment, it never matched the code. Fixes: 4ddbcb886268 ("bpf: Add bpf_sock_destroy kfunc") Reported-by: Xiang Mei (Microsoft) Closes: https://lore.kernel.org/bpf/20260702224519.800135-1-xmei5@asu.edu/ Signed-off-by: Jiayuan Chen Reviewed-by: Kuniyuki Iwashima Link: https://lore.kernel.org/r/20260910112634.152195-1-jiayuan.chen@linux.dev Signed-off-by: Alexei Starovoitov --- net/core/filter.c | 13 +++++++++---- 1 file changed, 9 insertions(+), 4 deletions(-) diff --git a/net/core/filter.c b/net/core/filter.c index 2a84f9d01131..cae43b999162 100644 --- a/net/core/filter.c +++ b/net/core/filter.c @@ -12913,8 +12913,9 @@ __bpf_kfunc_start_defs(); * @sock: Pointer to socket to be destroyed * * Return: - * On error, may return EPROTONOSUPPORT, EINVAL. - * EPROTONOSUPPORT if protocol specific destroy handler is not supported. + * On error, may return EOPNOTSUPP, or whatever the protocol specific + * destroy handler returns. + * EOPNOTSUPP if protocol specific destroy handler is not supported. * 0 otherwise */ __bpf_kfunc int bpf_sock_destroy(struct sock_common *sock) @@ -12926,8 +12927,12 @@ __bpf_kfunc int bpf_sock_destroy(struct sock_common *sock) * Supporting protocols will need to acquire sock lock in the BPF context * prior to invoking this kfunc. */ - if (!sk->sk_prot->diag_destroy || (sk->sk_protocol != IPPROTO_TCP && - sk->sk_protocol != IPPROTO_UDP)) + if (!sk->sk_prot->diag_destroy) + return -EOPNOTSUPP; + + if (sk_fullsock(sk) && + sk->sk_protocol != IPPROTO_TCP && + sk->sk_protocol != IPPROTO_UDP) return -EOPNOTSUPP; return sk->sk_prot->diag_destroy(sk, ECONNABORTED); From eaab8cab451b9502ce224cd202550375b894a467 Mon Sep 17 00:00:00 2001 From: Jiayuan Chen Date: Thu, 10 Sep 2026 19:27:28 +0800 Subject: [PATCH 088/592] tcp: Skip cond_resched() in inet_csk_listen_stop() under BPF context bpf_sock_destroy() runs from the tcp iterator, under rcu_read_lock(). If the sock is a listener that still has children in its accept queue, tcp_abort() ends up in inet_csk_listen_stop() and the cond_resched() there trips the debug check: BUG: sleeping function called from invalid context at net/ipv4/inet_connection_sock.c:1523 in_atomic(): 0, irqs_disabled(): 0, non_block: 0, pid: 628, name: test_progs preempt_count: 0, expected: 0 RCU nest depth: 1, expected: 0 locks held by test_progs/628: 3, last CPU#3: #0: ffff8881158cee18 (&p->lock){+.+.}-{4:4}, at: bpf_seq_read+0x56/0x1210 #1: ffff8881106bb858 (sk_lock-AF_INET6){+.+.}-{0:0}, at: bpf_iter_tcp_seq_show+0x32b/0x4b0 #2: ffffffffb435af20 (rcu_read_lock){....}-{1:3}, at: bpf_iter_run_prog+0x46b/0xde0 CPU: 3 UID: 0 PID: 628 Comm: test_progs Tainted: G W 7.2.0+ #65 PREEMPT Tainted: [W]=WARN Call Trace: dump_stack_lvl+0xc1/0xf0 dump_stack+0x10/0x20 __might_resched+0x3d2/0x610 inet_csk_listen_stop+0x7b/0xbf0 tcp_abort+0x23b/0x3b0 bpf_sock_destroy+0xfc/0x140 bpf_prog_448133d24601754f_iter_tcp6_server+0x81/0x8a bpf_iter_run_prog+0x538/0xde0 bpf_iter_tcp_seq_show+0x26b/0x4b0 bpf_seq_read+0x424/0x1210 vfs_read+0x197/0xe40 ksys_read+0x119/0x240 __x64_sys_read+0x72/0xc0 x64_sys_call+0x647/0x27e0 do_syscall_64+0xe5/0x610 entry_SYSCALL_64_after_hwframe+0x76/0x7e RIP: 0033:0x7fad39b28aca RSP: 002b:00007ffc381c61c0 EFLAGS: 00000246 ORIG_RAX: 0000000000000000 RAX: ffffffffffffffda RBX: 00007ffc381c6a88 RCX: 00007fad39b28aca RDX: 0000000000000032 RSI: 00007ffc381c6250 RDI: 0000000000000014 RBP: 00007ffc381c61e0 R08: 0000000000000000 R09: 0000000000000000 R10: 0000000000000000 R11: 0000000000000246 R12: 0000000000000003 R13: 0000000000000000 R14: 000055f077c1bbb0 R15: 00007fad3a0f3000 The commit that added the kfunc already guards lock_sock() in tcp_abort() and udp_abort() with has_current_bpf_ctx(), but missed the listener path. Do the same for the cond_resched(). The loop runs inside the iterator's rcu_read_lock(), it must not reschedule or report a quiescent state there. Fixes: 4ddbcb886268 ("bpf: Add bpf_sock_destroy kfunc") Signed-off-by: Jiayuan Chen Link: https://lore.kernel.org/r/20260910112736.153710-1-jiayuan.chen@linux.dev Signed-off-by: Alexei Starovoitov --- net/ipv4/inet_connection_sock.c | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-) diff --git a/net/ipv4/inet_connection_sock.c b/net/ipv4/inet_connection_sock.c index 6257459bcee2..6a30f1138454 100644 --- a/net/ipv4/inet_connection_sock.c +++ b/net/ipv4/inet_connection_sock.c @@ -1520,7 +1520,8 @@ void inet_csk_listen_stop(struct sock *sk) local_bh_enable(); sock_put(child); - cond_resched(); + if (!has_current_bpf_ctx()) + cond_resched(); } if (queue->fastopenq.rskq_rst_head) { /* Free all the reqs queued in rskq_rst_head. */ From 8036d3a5a6589ef0721d15e7c03976bc3996f7a8 Mon Sep 17 00:00:00 2001 From: Jiayuan Chen Date: Thu, 10 Sep 2026 19:28:02 +0800 Subject: [PATCH 089/592] selftests/bpf: Test bpf_sock_destroy() on TIME_WAIT and listener socks Add two subtests. tcp_timewait: the client shuts down first and the server closes after it, so the client sock ends up in TIME_WAIT. A tcp iterator then finds the timewait sock by the cookie it inherited from the client sock and destroys it. Iterate once more to make sure it is gone. Without the first fix bpf_sock_destroy() reads past the timewait sock and KASAN complains. tcp_listen_pending: connect to a listener but never accept, so the child sits in the accept queue, then destroy the listener. Without the second fix the cond_resched() in inet_csk_listen_stop() trips the might_sleep check under rcu_read_lock(). ./test_progs -a sock_destroy #444/1 sock_destroy/tcp_client:OK #444/2 sock_destroy/tcp_server:OK #444/3 sock_destroy/tcp_listen_pending:OK #444/4 sock_destroy/tcp_timewait:OK #444/5 sock_destroy/udp_client:OK #444/6 sock_destroy/udp_server:OK #444/7 sock_destroy/trace_tcp_destroy_sock:OK #444 sock_destroy:OK Summary: 1/7 PASSED, 0 SKIPPED, 0/0 FAILED Signed-off-by: Jiayuan Chen Link: https://lore.kernel.org/r/20260910112810.153801-1-jiayuan.chen@linux.dev Signed-off-by: Alexei Starovoitov --- .../selftests/bpf/prog_tests/sock_destroy.c | 121 ++++++++++++++++++ .../selftests/bpf/progs/sock_destroy_prog.c | 30 +++++ 2 files changed, 151 insertions(+) diff --git a/tools/testing/selftests/bpf/prog_tests/sock_destroy.c b/tools/testing/selftests/bpf/prog_tests/sock_destroy.c index 9c11938fe597..78d642a02bdb 100644 --- a/tools/testing/selftests/bpf/prog_tests/sock_destroy.c +++ b/tools/testing/selftests/bpf/prog_tests/sock_destroy.c @@ -1,4 +1,5 @@ // SPDX-License-Identifier: GPL-2.0 +#include #include #include @@ -110,6 +111,122 @@ static void test_tcp_server(struct sock_destroy_prog *skel) close(serv); } +static void test_tcp_listen_pending(struct sock_destroy_prog *skel) +{ + int serv = -1, clien = -1, accept_serv = -1, n, serv_port; + struct pollfd pfd = { .events = POLLIN }; + char buf[1]; + + serv = start_server(AF_INET6, SOCK_STREAM, NULL, 0, 0); + if (!ASSERT_GE(serv, 0, "start_server")) + goto cleanup; + serv_port = get_socket_local_port(serv); + if (!ASSERT_GE(serv_port, 0, "get_sock_local_port")) + goto cleanup; + skel->bss->serv_port = (__be16)serv_port; + + /* + * Connect but never accept, so the child sits in the accept queue + * of the listener. Wait until it's actually there. + */ + clien = connect_to_fd(serv, 0); + if (!ASSERT_GE(clien, 0, "connect_to_fd")) + goto cleanup; + pfd.fd = serv; + if (!ASSERT_EQ(poll(&pfd, 1, -1), 1, "poll listener")) + goto cleanup; + + /* Run iterator program that destroys server sockets. */ + start_iter_sockets(skel->progs.iter_tcp6_server); + + accept_serv = accept(serv, NULL, NULL); + if (!ASSERT_LT(accept_serv, 0, "accept on destroyed listener")) + goto cleanup; + ASSERT_EQ(errno, EINVAL, "error code on destroyed listener"); + + /* The unaccepted child was reset along with the listener. */ + n = recv(clien, buf, sizeof(buf), 0); + if (!ASSERT_LT(n, 0, "client recv on reset child")) + goto cleanup; + ASSERT_EQ(errno, ECONNRESET, "error code on reset child"); + +cleanup: + if (clien != -1) + close(clien); + if (accept_serv != -1) + close(accept_serv); + if (serv != -1) + close(serv); +} + +static void test_tcp_timewait(struct sock_destroy_prog *skel) +{ + int serv = -1, clien = -1, accept_serv = -1, n; + struct timeval tv = {}; + char buf[1]; + + serv = start_server(AF_INET6, SOCK_STREAM, NULL, 0, 0); + if (!ASSERT_GE(serv, 0, "start_server")) + goto cleanup; + + clien = connect_to_fd(serv, 0); + if (!ASSERT_GE(clien, 0, "connect_to_fd")) + goto cleanup; + + accept_serv = accept(serv, NULL, NULL); + if (!ASSERT_GE(accept_serv, 0, "serv accept")) + goto cleanup; + + /* + * Active close from the client, then close the server side. Once + * recv() sees EOF the server FIN has been processed and the client + * sock is in TIME_WAIT. Block without timeout so a loaded CI box + * can't race us. + */ + if (!ASSERT_OK(setsockopt(clien, SOL_SOCKET, SO_RCVTIMEO, &tv, + sizeof(tv)), "clear rcvtimeo")) + goto cleanup; + if (!ASSERT_OK(shutdown(clien, SHUT_WR), "client shutdown")) + goto cleanup; + + /* + * Make sure the server has seen the client FIN before it closes, + * so the two FINs never cross. + */ + n = recv(accept_serv, buf, sizeof(buf), 0); + if (!ASSERT_EQ(n, 0, "server recv EOF")) + goto cleanup; + + close(accept_serv); + accept_serv = -1; + + /* block until return EOF */ + n = recv(clien, buf, sizeof(buf), 0); + if (!ASSERT_EQ(n, 0, "client recv EOF")) + goto cleanup; + + /* Run iterator program that destroys the timewait client sock. */ + skel->bss->tw_found = 0; + start_iter_sockets(skel->progs.iter_tcp6_timewait); + if (!ASSERT_EQ(skel->bss->tw_found, 1, "timewait sock found")) + goto cleanup; + + ASSERT_OK(skel->bss->tw_destroy_err, "destroy timewait sock"); + + /* The destroyed timewait sock must be gone. */ + skel->bss->tw_found = 0; + start_iter_sockets(skel->progs.iter_tcp6_timewait); + ASSERT_EQ(skel->bss->tw_found, 0, "timewait sock destroyed"); + +cleanup: + if (clien != -1) + close(clien); + if (accept_serv != -1) + close(accept_serv); + if (serv != -1) + close(serv); +} + static void test_udp_client(struct sock_destroy_prog *skel) { int serv = -1, clien = -1, n = 0; @@ -204,6 +321,10 @@ void test_sock_destroy(void) test_tcp_client(skel); if (test__start_subtest("tcp_server")) test_tcp_server(skel); + if (test__start_subtest("tcp_listen_pending")) + test_tcp_listen_pending(skel); + if (test__start_subtest("tcp_timewait")) + test_tcp_timewait(skel); if (test__start_subtest("udp_client")) test_udp_client(skel); if (test__start_subtest("udp_server")) diff --git a/tools/testing/selftests/bpf/progs/sock_destroy_prog.c b/tools/testing/selftests/bpf/progs/sock_destroy_prog.c index 9e0bf7a54cec..0a8887543218 100644 --- a/tools/testing/selftests/bpf/progs/sock_destroy_prog.c +++ b/tools/testing/selftests/bpf/progs/sock_destroy_prog.c @@ -7,6 +7,8 @@ #include "bpf_tracing_net.h" __be16 serv_port = 0; +int tw_found = 0; +int tw_destroy_err = 0; int bpf_sock_destroy(struct sock_common *sk) __ksym; @@ -100,6 +102,34 @@ int iter_tcp6_server(struct bpf_iter__tcp *ctx) return 0; } +SEC("iter/tcp") +int iter_tcp6_timewait(struct bpf_iter__tcp *ctx) +{ + struct sock_common *sk_common = ctx->sk_common; + __u64 *val; + int key = 0; + + if (!sk_common) + return 0; + + if (sk_common->skc_family != AF_INET6) + return 0; + + if (!bpf_skc_to_tcp_timewait_sock(sk_common)) + return 0; + + val = bpf_map_lookup_elem(&tcp_conn_sockets, &key); + if (!val) + return 0; + /* The timewait sock inherits the cookie of the closed client sock. */ + if (bpf_get_socket_cookie(sk_common) != *val) + return 0; + + tw_found++; + tw_destroy_err = bpf_sock_destroy(sk_common); + + return 0; +} SEC("iter/udp") int iter_udp6_client(struct bpf_iter__udp *ctx) From 143755bdabaa96776c24f878014608e9cb44f930 Mon Sep 17 00:00:00 2001 From: Karl Mehltretter Date: Mon, 7 Sep 2026 04:00:15 +0200 Subject: [PATCH 090/592] dma-buf: Make DMABUF_DEBUG default to y on DEBUG_KERNEL kernels MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit Commit 646013f513f3 ("dma-buf: enable DMABUF_DEBUG by default on DEBUG kernels") changed the default of DMABUF_DEBUG to "y if DEBUG", but no Kconfig symbol DEBUG exists, so the option has had no default since. Use DEBUG_KERNEL, the Kconfig symbol for a debug kernel. Fixes: 646013f513f3 ("dma-buf: enable DMABUF_DEBUG by default on DEBUG kernels") Assisted-by: LLM Signed-off-by: Karl Mehltretter Reviewed-by: Christian König Signed-off-by: Christian König Link: https://lore.kernel.org/r/20260907020015.24719-1-kmehltretter@gmail.com --- drivers/dma-buf/Kconfig | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/drivers/dma-buf/Kconfig b/drivers/dma-buf/Kconfig index 7efc0f0d0712..e4f078a326a4 100644 --- a/drivers/dma-buf/Kconfig +++ b/drivers/dma-buf/Kconfig @@ -43,7 +43,7 @@ config UDMABUF config DMABUF_DEBUG bool "DMA-BUF debug checks" depends on DMA_SHARED_BUFFER - default y if DEBUG + default y if DEBUG_KERNEL help This option enables additional checks for DMA-BUF importers and exporters. Specifically it validates that importers do not peek at the From b344ca94e8cc85796f16ea25e2e5a8e0303fe813 Mon Sep 17 00:00:00 2001 From: David Hu Date: Tue, 1 Sep 2026 17:08:48 +0000 Subject: [PATCH 091/592] dma-buf: Fix silent overflow for phys vec to sgt MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit In case MMIO size is bigger than 4G and peer2peer DMA goes through host bridge, we trigger a code path that assigns the total linked IOVA (which is greater than 4G) to mapped_len. Previously, `mapped_len` was declared as 32-bit `unsigned int`. When accumulating `size_t` lengths, this leads to a silent wrap-around. This truncation causes truncated lengths to be passed to functions like `fill_sg_entry()`. Fix this by changing `mapped_len` to `size_t` (64-bit). While at it, fix similar potential overflow issues in `calc_sg_nents` by using `check_add_overflow()` for `nents` and using `unsigned int` for the loop iterator in `fill_sg_entry` to match. Fixes: 3aa31a8bb11e ("dma-buf: provide phys_vec to scatter-gather mapping routine") Cc: stable@vger.kernel.org Cc: iommu@lists.linux.dev Reviewed-by: Pranjal Shrivastava Reviewed-by: Kevin Tian Reviewed-by: Leon Romanovsky Signed-off-by: David Hu Signed-off-by: Christian König Link: https://lore.kernel.org/r/20260901170849.4052816-2-dhu@x6u.co --- drivers/dma-buf/dma-buf-mapping.c | 16 ++++++++++++---- 1 file changed, 12 insertions(+), 4 deletions(-) diff --git a/drivers/dma-buf/dma-buf-mapping.c b/drivers/dma-buf/dma-buf-mapping.c index 794acff2546a..80f6ab2f4809 100644 --- a/drivers/dma-buf/dma-buf-mapping.c +++ b/drivers/dma-buf/dma-buf-mapping.c @@ -5,12 +5,13 @@ */ #include #include +#include static struct scatterlist *fill_sg_entry(struct scatterlist *sgl, size_t length, dma_addr_t addr) { unsigned int len, nents; - int i; + unsigned int i; nents = DIV_ROUND_UP(length, UINT_MAX); for (i = 0; i < nents; i++) { @@ -40,8 +41,12 @@ static unsigned int calc_sg_nents(struct dma_iova_state *state, size_t i; if (!state || !dma_use_iova(state)) { - for (i = 0; i < nr_ranges; i++) - nents += DIV_ROUND_UP(phys_vec[i].len, UINT_MAX); + for (i = 0; i < nr_ranges; i++) { + unsigned int added = DIV_ROUND_UP(phys_vec[i].len, UINT_MAX); + + if (check_add_overflow(nents, added, &nents)) + return 0; + } } else { /* * In IOVA case, there is only one SG entry which spans @@ -95,9 +100,10 @@ struct sg_table *dma_buf_phys_vec_to_sgt(struct dma_buf_attachment *attach, size_t nr_ranges, size_t size, enum dma_data_direction dir) { - unsigned int nents, mapped_len = 0; struct dma_buf_dma *dma; struct scatterlist *sgl; + size_t mapped_len = 0; + unsigned int nents; dma_addr_t addr; size_t i; int ret; @@ -133,6 +139,8 @@ struct sg_table *dma_buf_phys_vec_to_sgt(struct dma_buf_attachment *attach, } nents = calc_sg_nents(dma->state, phys_vec, nr_ranges, size); + + /* sg_alloc_table will cleanly fail and return -EINVAL if nents == 0 */ ret = sg_alloc_table(&dma->sgt, nents, GFP_KERNEL | __GFP_ZERO); if (ret) goto err_free_state; From 06dd5e1ae8ce4e129791087c8c66594950f0ba03 Mon Sep 17 00:00:00 2001 From: David Hu Date: Tue, 1 Sep 2026 17:08:49 +0000 Subject: [PATCH 092/592] dma-buf: Split sgl by largest page-aligned chunk MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit Currently, `fill_sg_entry()` splits the scatterlist using `UINT_MAX`. This creates a non-page-aligned DMA length (`0xFFFFFFFF`) for the first entry, resulting in non-page-aligned DMA addresses for all subsequent entries. While the underlying IOMMU mapping may be contiguous, hardware DMA engines often require explicit address alignment (e.g., page, cacheline, or storage sector boundaries). Passing unaligned addresses and lengths can cause explicit failures in DMA descriptor creation or silent data corruption if lower unaligned bits are truncated. In addition, a non-page-aligned sgl length will trigger an edge case in `ib_umem_find_best_pgsz()`. In case of a discontinuity in later buffers, we will have a `va` with lowest bit set to 1. That will lead to `ib_umem_find_best_pgsz()` always return 0, and break the promise to find best page size for the mapping on the NIC side. Fix this by splitting the scatterlist by the largest possible page aligned chunk within `UINT_MAX` (`ALIGN_DOWN(UINT_MAX, PAGE_SIZE)`). This ensures all scatterlist DMA addresses and lengths remain page aligned, while minimizing the total number of sgl entries. Page-aligned entries allow the system to cleanly chunk payloads into PCIe MaxPayloadSize (MPS) (e.g., 128 bytes, 256 bytes, 512 bytes). As a result, this may help reduce TLP fragmentation in P2P transfers and alleviate potential congestion within a logical PCIe switch partition, especially when Relaxed Ordering is not possible due to hardware constraints. Reported-by: sashiko-bot Closes: https://lore.kernel.org/all/20260609165431.778061F00893@smtp.kernel.org/ Fixes: 3aa31a8bb11e ("dma-buf: provide phys_vec to scatter-gather mapping routine") Cc: stable@vger.kernel.org Reviewed-by: Leon Romanovsky Signed-off-by: David Hu Signed-off-by: Christian König Link: https://lore.kernel.org/r/20260901170849.4052816-3-dhu@x6u.co --- drivers/dma-buf/dma-buf-mapping.c | 19 +++++++++++-------- 1 file changed, 11 insertions(+), 8 deletions(-) diff --git a/drivers/dma-buf/dma-buf-mapping.c b/drivers/dma-buf/dma-buf-mapping.c index 80f6ab2f4809..833be519e1e6 100644 --- a/drivers/dma-buf/dma-buf-mapping.c +++ b/drivers/dma-buf/dma-buf-mapping.c @@ -6,16 +6,17 @@ #include #include #include +#include + +#define MAX_SG_ENT_SZ ALIGN_DOWN(UINT_MAX, PAGE_SIZE) static struct scatterlist *fill_sg_entry(struct scatterlist *sgl, size_t length, dma_addr_t addr) { - unsigned int len, nents; - unsigned int i; + size_t len; - nents = DIV_ROUND_UP(length, UINT_MAX); - for (i = 0; i < nents; i++) { - len = min_t(size_t, length, UINT_MAX); + while (length) { + len = min(length, MAX_SG_ENT_SZ); length -= len; /* * DMABUF abuses scatterlist to create a scatterlist @@ -25,8 +26,10 @@ static struct scatterlist *fill_sg_entry(struct scatterlist *sgl, size_t length, * does not require the CPU list for mapping or unmapping. */ sg_set_page(sgl, NULL, 0, 0); - sg_dma_address(sgl) = addr + (dma_addr_t)i * UINT_MAX; + sg_dma_address(sgl) = addr; sg_dma_len(sgl) = len; + addr += len; + /* Unconditionally advance. On last segment, this becomes NULL */ sgl = sg_next(sgl); } @@ -42,7 +45,7 @@ static unsigned int calc_sg_nents(struct dma_iova_state *state, if (!state || !dma_use_iova(state)) { for (i = 0; i < nr_ranges; i++) { - unsigned int added = DIV_ROUND_UP(phys_vec[i].len, UINT_MAX); + unsigned int added = DIV_ROUND_UP(phys_vec[i].len, MAX_SG_ENT_SZ); if (check_add_overflow(nents, added, &nents)) return 0; @@ -53,7 +56,7 @@ static unsigned int calc_sg_nents(struct dma_iova_state *state, * for whole IOVA address space, but we need to make sure * that it fits sg->length, maybe we need more. */ - nents = DIV_ROUND_UP(size, UINT_MAX); + nents = DIV_ROUND_UP(size, MAX_SG_ENT_SZ); } return nents; From 49daa3d668b69a5454b5aba0078848a479f79f1c Mon Sep 17 00:00:00 2001 From: Shouping Wang Date: Thu, 10 Sep 2026 19:46:01 +0800 Subject: [PATCH 093/592] perf/arm-cmn: Fix wp_dev_sel2 setting for multi-DTM configurations When MXP_MULTIPLE_DTM_EN is TRUE, each DTM will monitor at most two device ports. In this case, {wp_dev_sel2, wp_dev_sel} will only use values 2'b00 and 2'b01 per DTM. Previously the setting allowed values beyond the supported range per DTM, which could cause each DTM to select invalid ports when MXP_MULTIPLE_DTM_EN is TRUE. Fix this by only setting CMN_DTM_WPn_CONFIG_WP_DEV_SEL2 when !multi_dtm. Fixes: 60d1504070c2 ("perf/arm-cmn: Support new IP features") Signed-off-by: Shouping Wang Reviewed-by: Robin Murphy Signed-off-by: Will Deacon --- drivers/perf/arm-cmn.c | 10 +++++++--- 1 file changed, 7 insertions(+), 3 deletions(-) diff --git a/drivers/perf/arm-cmn.c b/drivers/perf/arm-cmn.c index b162de3d9d16..33ee2be9b386 100644 --- a/drivers/perf/arm-cmn.c +++ b/drivers/perf/arm-cmn.c @@ -1582,13 +1582,14 @@ static void arm_cmn_claim_wp_idx(struct arm_cmn_dtm *dtm, static u32 arm_cmn_wp_config(struct perf_event *event, int wp_idx) { + struct arm_cmn *cmn = to_cmn(event->pmu); u32 config; u32 dev = CMN_EVENT_WP_DEV_SEL(event); u32 chn = CMN_EVENT_WP_CHN_SEL(event); u32 grp = CMN_EVENT_WP_GRP(event); u32 exc = CMN_EVENT_WP_EXCLUSIVE(event); u32 combine = CMN_EVENT_WP_COMBINE(event); - bool is_cmn600 = to_cmn(event->pmu)->part == PART_CMN600; + bool is_cmn600 = cmn->part == PART_CMN600; /* CMN-600 supports only primary and secondary matching groups */ if (is_cmn600) @@ -1596,8 +1597,11 @@ static u32 arm_cmn_wp_config(struct perf_event *event, int wp_idx) config = FIELD_PREP(CMN_DTM_WPn_CONFIG_WP_DEV_SEL, dev) | FIELD_PREP(CMN_DTM_WPn_CONFIG_WP_CHN_SEL, chn) | - FIELD_PREP(CMN_DTM_WPn_CONFIG_WP_GRP, grp) | - FIELD_PREP(CMN_DTM_WPn_CONFIG_WP_DEV_SEL2, dev >> 1); + FIELD_PREP(CMN_DTM_WPn_CONFIG_WP_GRP, grp); + + if (!cmn->multi_dtm) + config |= FIELD_PREP(CMN_DTM_WPn_CONFIG_WP_DEV_SEL2, dev >> 1); + if (exc) config |= is_cmn600 ? CMN600_WPn_CONFIG_WP_EXCLUSIVE : CMN_DTM_WPn_CONFIG_WP_EXCLUSIVE; From e4a6f57d22e079e23fafac51057fad534160b269 Mon Sep 17 00:00:00 2001 From: Breno Leitao Date: Thu, 10 Sep 2026 06:53:27 -0700 Subject: [PATCH 094/592] arm64: hibernate: clone only the linear map that exists at runtime This is similar to commit 1537e55728ec2 ("arm64: trans_pgd: clone only the linear map that exists at runtime"), but in a different place. swsusp_arch_resume() clones the kernel linear map with trans_pgd_create_copy(..., PAGE_OFFSET, PAGE_END). PAGE_OFFSET comes from the compile-time VA_BITS, so a CONFIG_ARM64_VA_BITS_52 kernel booting on hardware without LPA2 -- vabits_actual is 48 and the fifth level is folded -- hands the walk a 3.9PB window while its linear map only spans the top 128TB. On a VA_BITS_52 4k kernel with CONFIG_KASAN_GENERIC in a 4GB VM, I see: swapper/0: page allocation failure: order:0, mode:0x920(GFP_ATOMIC|__GFP_ZERO) hibernate_page_alloc+0x10/0x1c swsusp_arch_resume+0x70/0x320 hibernation_restore+0xa4/0x138 software_resume+0x15c/0x270 PM: hibernation: Failed to load image, recovering. PM: hibernation: resume failed (-12) Fix it by copying the linear map that is the actual one, not the compiled one. Fixes: a6bbf5d4d9d1 ("arm64: mm: Add definitions to support 5 levels of paging") Signed-off-by: Breno Leitao Reviewed-by: Ard Biesheuvel Signed-off-by: Will Deacon --- arch/arm64/kernel/hibernate.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/arch/arm64/kernel/hibernate.c b/arch/arm64/kernel/hibernate.c index 7bf117427777..424291c547f0 100644 --- a/arch/arm64/kernel/hibernate.c +++ b/arch/arm64/kernel/hibernate.c @@ -423,8 +423,8 @@ int __nocfi swsusp_arch_resume(void) * Create a second copy of just the linear map, and use this when * restoring. */ - rc = trans_pgd_create_copy(&trans_info, &tmp_pg_dir, PAGE_OFFSET, - PAGE_END); + rc = trans_pgd_create_copy(&trans_info, &tmp_pg_dir, + _PAGE_OFFSET(vabits_actual), PAGE_END); if (rc) return rc; From 885bff055a0f251a51a0d4fd4f0a7b525582a3de Mon Sep 17 00:00:00 2001 From: Mark Rutland Date: Tue, 8 Sep 2026 16:17:21 +0100 Subject: [PATCH 095/592] arm64: percpu: Fix this_cpu_write() casting The arm64 implementation of this_cpu_write() casts 'val' to unsigned long. This is necessary to handle cases where 'val' is a pointer type, and to avoid spurious compiler warnings for the (unreachable!) cases where the pointer type would be cast to a smaller integer type. Unfortunately, the cast is applied to 'val' rather than '(val)', which won't always generate the expected value when 'val' is an expression. For example, for this_cpu_write(pcp, zero - 1), where 'pcp' is a u64 and 'zero' is a u32: * 'zero' ===> (u32) 0x00000000 * 'zero - 1' ===> (u32) 0xffffffff * '(unsigned long)zero - 1' ===> (u64) 0xffffffffffffffff * '(unsigned long)(zero - 1)' ===> (u64) 0x00000000ffffffff Fix this by adding brackets around 'val'. Fixes: 959bf2fd03b5 ("arm64: percpu: Rewrite per-cpu ops to allow use of LSE atomics") Reported-by: David Laight Signed-off-by: Mark Rutland Reviewed-by: David Laight Reviewed-by: Jinjie Ruan Tested-by: Muhammad Usama Anjum Acked-by: Christopher Lameter (Ampere) Cc: Ada Couprie Diaz Cc: Ard Biesheuvel Cc: Catalin Marinas Cc: James Morse Cc: Marc Zyngier Cc: Peter Zijlstra Cc: Vladimir Murzin Cc: Will Deacon Cc: Yang Shi Cc: stable@vger.kernel.org Reviewed-by: Lorenzo Stoakes (ARM) Signed-off-by: Will Deacon --- arch/arm64/include/asm/percpu.h | 8 ++++---- 1 file changed, 4 insertions(+), 4 deletions(-) diff --git a/arch/arm64/include/asm/percpu.h b/arch/arm64/include/asm/percpu.h index b57b2bb00967..63bbfd4944a3 100644 --- a/arch/arm64/include/asm/percpu.h +++ b/arch/arm64/include/asm/percpu.h @@ -179,13 +179,13 @@ PERCPU_RET_OP(add, add, ldadd) _pcp_protect_return(__percpu_read_64, pcp) #define this_cpu_write_1(pcp, val) \ - _pcp_protect(__percpu_write_8, pcp, (unsigned long)val) + _pcp_protect(__percpu_write_8, pcp, (unsigned long)(val)) #define this_cpu_write_2(pcp, val) \ - _pcp_protect(__percpu_write_16, pcp, (unsigned long)val) + _pcp_protect(__percpu_write_16, pcp, (unsigned long)(val)) #define this_cpu_write_4(pcp, val) \ - _pcp_protect(__percpu_write_32, pcp, (unsigned long)val) + _pcp_protect(__percpu_write_32, pcp, (unsigned long)(val)) #define this_cpu_write_8(pcp, val) \ - _pcp_protect(__percpu_write_64, pcp, (unsigned long)val) + _pcp_protect(__percpu_write_64, pcp, (unsigned long)(val)) #define this_cpu_add_1(pcp, val) \ _pcp_protect(__percpu_add_case_8, pcp, val) From 44274c657256b4911de82f8104e9e22f054cf742 Mon Sep 17 00:00:00 2001 From: Mark Rutland Date: Tue, 8 Sep 2026 16:17:22 +0100 Subject: [PATCH 096/592] arm64: percpu: Fix this_cpu_and() mask generation The arm64 implementation of this_cpu_and(pcp, val) is built in terms of ANDNOT operations, which requires the 'val' argument to be bitwise negated. The bitwise negation is not implemented correctly, with two bugs described below. (1) The bitwise negation is performed as '~val' rather than '~(val)'. This won't always generate the expected value when 'val' is an expression. For example, for this_cpu_and(pcp, 1 - 1): * 'val' is '1 - 1' ===> (int) 0x00000000 * '~val' is '~1 - 1' ===> (int) 0xfffffffd * '~(val)' is '~(1 - 1)' ===> (int) 0xffffffff ... and thus bit[1] of 'pcp' would be preserved unexpectedly by the ANDNOT operation. (2) The bitwise negation is performed on 'val' before it has been cast to (at least) the width of 'pcp'. This won't always generate the expected value for the upper bits. For example, for this_cpu_and(pcp, zero), where 'pcp' is a u64 and 'zero' is a u32: * 'zero' ===> (u32) 0x00000000 * '~(zero)' ===> (u32) 0xffffffff * '(u64)~(zero)' ===> (u64) 0x00000000ffffffff * '~((u64)(zero))' ===> (u64) 0xffffffffffffffff ... and thus bits[63:32] of 'pcp' would be preserved unexpectedly by the ANDNOT operation. Fix these issues by adding brackets around 'val', and by casting 'val' to an appropriately-sized type before bitwise negation. Fixes: 959bf2fd03b5 ("arm64: percpu: Rewrite per-cpu ops to allow use of LSE atomics") Signed-off-by: Mark Rutland Reviewed-by: Jinjie Ruan Tested-by: Muhammad Usama Anjum Acked-by: Christopher Lameter (Ampere) Cc: Ada Couprie Diaz Cc: Ard Biesheuvel Cc: Catalin Marinas Cc: James Morse Cc: Marc Zyngier Cc: Peter Zijlstra Cc: Vladimir Murzin Cc: Will Deacon Cc: Yang Shi Cc: stable@vger.kernel.org Signed-off-by: Will Deacon --- arch/arm64/include/asm/percpu.h | 8 ++++---- 1 file changed, 4 insertions(+), 4 deletions(-) diff --git a/arch/arm64/include/asm/percpu.h b/arch/arm64/include/asm/percpu.h index 63bbfd4944a3..31193bcf89a2 100644 --- a/arch/arm64/include/asm/percpu.h +++ b/arch/arm64/include/asm/percpu.h @@ -206,13 +206,13 @@ PERCPU_RET_OP(add, add, ldadd) _pcp_protect_return(__percpu_add_return_case_64, pcp, val) #define this_cpu_and_1(pcp, val) \ - _pcp_protect(__percpu_andnot_case_8, pcp, ~val) + _pcp_protect(__percpu_andnot_case_8, pcp, ~(u8)(val)) #define this_cpu_and_2(pcp, val) \ - _pcp_protect(__percpu_andnot_case_16, pcp, ~val) + _pcp_protect(__percpu_andnot_case_16, pcp, ~(u16)(val)) #define this_cpu_and_4(pcp, val) \ - _pcp_protect(__percpu_andnot_case_32, pcp, ~val) + _pcp_protect(__percpu_andnot_case_32, pcp, ~(u32)(val)) #define this_cpu_and_8(pcp, val) \ - _pcp_protect(__percpu_andnot_case_64, pcp, ~val) + _pcp_protect(__percpu_andnot_case_64, pcp, ~(u64)(val)) #define this_cpu_or_1(pcp, val) \ _pcp_protect(__percpu_or_case_8, pcp, val) From 8cf2093f5372952a9ebc805c418d45df7112cd14 Mon Sep 17 00:00:00 2001 From: Mark Rutland Date: Tue, 8 Sep 2026 16:17:23 +0100 Subject: [PATCH 097/592] arm64: percpu: Fix LSE operations on {8,16}-bit types The assembly for __percpu_##name##_case_##sz() and __percpu_##name##_return_case_##sz() doesn't use the 'sfx' macro argument to form the LSE instruction. Without 'sfx', a W register argument will imply a 32-bit memory location, and consequently {8,16}-bit ops will erroneously read and write 32 bits of memory when the LSE instruction is used. Fix this by appending 'sfx' to 'op_lse' to LSE instruction. It is not necessary (and not valid) to append 'sfx' to 'op_llsc', as 'op_llsc' is a register-register operation which does not access memory (and does not take a size suffix). Fixes: 959bf2fd03b5 ("arm64: percpu: Rewrite per-cpu ops to allow use of LSE atomics") Signed-off-by: Mark Rutland Reviewed-by: Jinjie Ruan Cc: Ada Couprie Diaz Cc: Ard Biesheuvel Cc: Catalin Marinas Cc: James Morse Cc: Marc Zyngier Cc: Peter Zijlstra Cc: Vladimir Murzin Cc: Will Deacon Cc: Yang Shi Cc: stable@vger.kernel.org Reviewed-by: Vladimir Murzin Signed-off-by: Will Deacon --- arch/arm64/include/asm/percpu.h | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/arch/arm64/include/asm/percpu.h b/arch/arm64/include/asm/percpu.h index 31193bcf89a2..8cf4068ce1b5 100644 --- a/arch/arm64/include/asm/percpu.h +++ b/arch/arm64/include/asm/percpu.h @@ -77,7 +77,7 @@ __percpu_##name##_case_##sz(void *ptr, unsigned long val) \ " stxr" #sfx "\t%w[loop], %" #w "[tmp], %[ptr]\n" \ " cbnz %w[loop], 1b", \ /* LSE atomics */ \ - #op_lse "\t%" #w "[val], %" #w "[tmp], %[ptr]\n" \ + #op_lse #sfx "\t%" #w "[val], %" #w "[tmp], %[ptr]\n" \ __nops(3)) \ : [loop] "=&r" (loop), [tmp] "=&r" (tmp), \ [ptr] "+Q"(*(u##sz *)ptr) \ @@ -98,7 +98,7 @@ __percpu_##name##_return_case_##sz(void *ptr, unsigned long val) \ " stxr" #sfx "\t%w[loop], %" #w "[ret], %[ptr]\n" \ " cbnz %w[loop], 1b", \ /* LSE atomics */ \ - #op_lse "\t%" #w "[val], %" #w "[ret], %[ptr]\n" \ + #op_lse #sfx "\t%" #w "[val], %" #w "[ret], %[ptr]\n" \ #op_llsc "\t%" #w "[ret], %" #w "[ret], %" #w "[val]\n" \ __nops(2)) \ : [loop] "=&r" (loop), [ret] "=&r" (ret), \ From 3d1ba5cbfb622025690c218d8f20da92a9ecb383 Mon Sep 17 00:00:00 2001 From: Thomas Huth Date: Wed, 9 Sep 2026 17:57:07 +0200 Subject: [PATCH 098/592] kselftest/arm64: Fix size of thread_data values for pthread_join() pthread_join() stores the thread's return value (a "void *", i.e. 8 bytes on 64 bit computers) into the address that is passed as second parameter. However, the entries of thread_data are only normal "int"s, i.e. only 4 bytes. The additional 4 bytes of the return value clobber whatever is adjacent on the stack, i.e. other members of the thread_data array (which will be re-written in the next iteration of the for-loop, so that nobody noticed this problem), or another other local variable on the stack for the last iteration. Use "intptr_t" to declare the thread_data array entries with the correct size. Fixes: 29f080881601c ("kselftest/arm64: check GCR_EL1 after context switch") Cc: stable@vger.kernel.org Signed-off-by: Thomas Huth Signed-off-by: Will Deacon --- tools/testing/selftests/arm64/mte/check_gcr_el1_cswitch.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/tools/testing/selftests/arm64/mte/check_gcr_el1_cswitch.c b/tools/testing/selftests/arm64/mte/check_gcr_el1_cswitch.c index d23f154d3288..5d9dc8bcfbf5 100644 --- a/tools/testing/selftests/arm64/mte/check_gcr_el1_cswitch.c +++ b/tools/testing/selftests/arm64/mte/check_gcr_el1_cswitch.c @@ -69,7 +69,7 @@ void *execute_thread(void *x) int execute_test(pid_t pid) { pthread_t thread_id[MAX_THREADS]; - int thread_data[MAX_THREADS]; + intptr_t thread_data[MAX_THREADS]; for (int i = 0; i < MAX_THREADS; i++) pthread_create(&thread_id[i], NULL, From fdb9ebc7fb7788b8371fbef6c17dc5c8291e1429 Mon Sep 17 00:00:00 2001 From: Catalin Marinas Date: Mon, 7 Sep 2026 12:30:25 +0100 Subject: [PATCH 099/592] arm64: mte: Fix PTRACE_{PEEK,POKE}MTETAGS error documentation PTRACE_{PEEK,POKE}MTETAGS return -EIO rather than -EOPNOTSUPP (as documented) when no tags are copied from/to a mapping without PROT_MTE. This has been the behaviour since the interface was introduced, though the original intent was to distinguish between address not being accessible and mapped as untagged. Update the documentation to match the implementation (de-facto ABI). Since -EOPNOTSUPP was never returned, change the error assignment to -EIO as well to avoid confusion. Fixes: df9d7a22dd21 ("arm64: mte: Add Memory Tagging Extension documentation") Fixes: 18ddbaa02b7a ("arm64: mte: ptrace: Add PTRACE_{PEEK,POKE}MTETAGS support") Reported-by: Yury Khrustalev Cc: Will Deacon Cc: Mark Rutland Signed-off-by: Catalin Marinas Signed-off-by: Will Deacon --- Documentation/arch/arm64/memory-tagging-extension.rst | 5 ++--- arch/arm64/kernel/mte.c | 2 +- 2 files changed, 3 insertions(+), 4 deletions(-) diff --git a/Documentation/arch/arm64/memory-tagging-extension.rst b/Documentation/arch/arm64/memory-tagging-extension.rst index e6fe428f0e2a..1d32fc5df183 100644 --- a/Documentation/arch/arm64/memory-tagging-extension.rst +++ b/Documentation/arch/arm64/memory-tagging-extension.rst @@ -208,11 +208,10 @@ will use the corresponding aligned address. tracer's space cannot be accessed or does not have valid tags. - ``-EPERM`` - the specified process cannot be traced. - ``-EIO`` - the tracee's address range cannot be accessed (e.g. invalid - address) and no tags copied. ``iov_len`` not updated. + address) or does not have valid tags (not mapped with the ``PROT_MTE`` + flag) and no tags copied. ``iov_len`` not updated. - ``-EFAULT`` - fault on accessing the tracer's memory (``struct iovec`` or ``iov_base`` buffer) and no tags copied. ``iov_len`` not updated. -- ``-EOPNOTSUPP`` - the tracee's address does not have valid tags (never - mapped with the ``PROT_MTE`` flag). ``iov_len`` not updated. **Note**: There are no transient errors for the requests above, so user programs should not retry in case of a non-zero system call return. diff --git a/arch/arm64/kernel/mte.c b/arch/arm64/kernel/mte.c index 1a9aad6ef22a..31f5c6b0510e 100644 --- a/arch/arm64/kernel/mte.c +++ b/arch/arm64/kernel/mte.c @@ -476,7 +476,7 @@ static int __access_remote_tags(struct mm_struct *mm, unsigned long addr, * was never mapped with PROT_MTE. */ if (!(vma->vm_flags & VM_MTE)) { - err = -EOPNOTSUPP; + err = -EIO; put_page(page); break; } From 9aa237cf66495b2426ddde8532e9b08a0ed83aaa Mon Sep 17 00:00:00 2001 From: Wei Jie LAW <98lawweijie@gmail.com> Date: Wed, 9 Sep 2026 09:15:13 +0800 Subject: [PATCH 100/592] HID: wacom: fix OOB read in wacom_wac_pen_serial_enforce() The 'wacom_wac_pen_serial_enforce()' function may calculate and pass an invalid offset to hid_field_extract(), resulting in memory reads at incorrect addresses -- possibly beyond the end of the report. If a field in the HID descriptor lists more usages than its Report Count actually reserves space for, the function's inner 'j' will walk past the end of the field: for (i = 0; i < report->maxfield; i++) { for (j = 0; j < report->field[i]->maxusage; j++) { ... value = hid_field_extract(hdev, raw_data + 1, offset + j * size, size); A descriptor listing 12288 usages against Report Count 1 has the loop extract the usage at index 12287 from bit offset 98296 -- about 12 KB past a 2-byte received report. The value is stored in wacom_wac->serial[0] and can reach userspace as an MSC_SERIAL event, making this an information disclosure. Clamp the loop to field->report_count, the number of value slots the report holds. Value slots past the last declared usage are still scanned; they reuse that usage (HID 1.11, 6.2.2.8). Verified on v6.12.105 with a UHID reproducer: a 2-byte report from such a descriptor trips KASAN before the patch and not after it. Fixes: 83417206427b ("HID: wacom: Queue events with missing type/serial data for later processing") Suggested-by: Jason Gerecke Cc: stable@vger.kernel.org Assisted-by: Claude:claude-opus-5 Assisted-by: GLM:glm-5.3 Signed-off-by: Wei Jie Law <98lawweijie@gmail.com> Reviewed-by: Jason Gerecke Signed-off-by: Jiri Kosina --- drivers/hid/wacom_sys.c | 5 +++-- 1 file changed, 3 insertions(+), 2 deletions(-) diff --git a/drivers/hid/wacom_sys.c b/drivers/hid/wacom_sys.c index 0eafa483b7f7..40770affdbde 100644 --- a/drivers/hid/wacom_sys.c +++ b/drivers/hid/wacom_sys.c @@ -113,8 +113,9 @@ static int wacom_wac_pen_serial_enforce(struct hid_device *hdev, /* Queue events which have invalid tool type or serial number */ for (i = 0; i < report->maxfield; i++) { - for (j = 0; j < report->field[i]->maxusage; j++) { - struct hid_field *field = report->field[i]; + struct hid_field *field = report->field[i]; + + for (j = 0; j < field->report_count; j++) { struct hid_usage *usage = &field->usage[j]; unsigned int equivalent_usage = wacom_equivalent_usage(usage->hid); unsigned int offset; From aaaea79efba5a27cb9e0a5a628046d829d3f2cbb Mon Sep 17 00:00:00 2001 From: Lovekesh Solanki Date: Wed, 5 Aug 2026 01:50:31 +0530 Subject: [PATCH 101/592] HID: multitouch: Add report ID mismatch quirk for ASUS ROG Z13 Folio Commit e716edafedad ("HID: multitouch: Check to ensure report responses match the request") introduced validating GET_FEATURE responses return the requested report ID. ASUS ROG Z13 Flow (2025) GZ302EA touchpad (USB 0b05:1a30) returns a different report ID for Win8 feature request. Before this check, the response was still processed and allowed device to switch into its full Touchpad Precision mode. After the validation, the response is discarded before hid_report_raw_event() processes it and device remains in fallback mode and no longer exposes ABS_MT_SLOT, ABS_MT_TOOL_TYPE or the multi-finger BTN_TOOL_* capabilities for palm rejection. Add a device quirk to allow the known firmware behavior for this device while preserving report ID validation for all other devices. The device previously matched the generic MT_CLS_WIN_8 entry, so base the new class on MT_CLS_WIN_8 to keep it on the same quirk set as before the regression. MT_QUIRK_CONFIDENCE must be set explicitly: it is normally enabled by the class name check in mt_touch_input_mapping(), which only matches the MT_CLS_WIN_8* names, and it is what makes ABS_MT_TOOL_TYPE available for touchpads. Fixes: e716edafedad ("HID: multitouch: Check to ensure report responses match the request") Signed-off-by: Lovekesh Solanki Reported-by: mayhemandcoffee Closes: https://bugzilla.kernel.org/show_bug.cgi?id=221774 Tested-by: mayhemandcoffee Link: https://bugzilla.kernel.org/show_bug.cgi?id=221774 Signed-off-by: Jiri Kosina --- drivers/hid/hid-multitouch.c | 29 +++++++++++++++++++++++++++-- 1 file changed, 27 insertions(+), 2 deletions(-) diff --git a/drivers/hid/hid-multitouch.c b/drivers/hid/hid-multitouch.c index 451c7324e6a0..ab4ac261fe76 100644 --- a/drivers/hid/hid-multitouch.c +++ b/drivers/hid/hid-multitouch.c @@ -79,6 +79,7 @@ MODULE_LICENSE("GPL"); #define MT_QUIRK_APPLE_TOUCHBAR BIT(23) #define MT_QUIRK_YOGABOOK9I BIT(24) #define MT_QUIRK_KEEP_LATENCY_ON_CLOSE BIT(25) +#define MT_QUIRK_IGNORE_FEATURE_ID_MISMATCH BIT(26) #define MT_INPUTMODE_TOUCHSCREEN 0x02 #define MT_INPUTMODE_TOUCHPAD 0x03 @@ -235,6 +236,7 @@ static void mt_post_parse(struct mt_device *td, struct mt_application *app); #define MT_CLS_APPLE_TOUCHBAR 0x0114 #define MT_CLS_YOGABOOK9I 0x0115 #define MT_CLS_EGALAX_P80H84 0x0116 +#define MT_CLS_ASUS_ROG_Z13_FOLIO 0x0117 #define MT_CLS_SIS 0x0457 #define MT_DEFAULT_MAXCONTACT 10 @@ -405,6 +407,16 @@ static const struct mt_class mt_classes[] = { .quirks = MT_QUIRK_ALWAYS_VALID | MT_QUIRK_CONTACT_CNT_ACCURATE | MT_QUIRK_ASUS_CUSTOM_UP }, + { .name = MT_CLS_ASUS_ROG_Z13_FOLIO, + .quirks = MT_QUIRK_ALWAYS_VALID | + MT_QUIRK_IGNORE_DUPLICATES | + MT_QUIRK_HOVERING | + MT_QUIRK_CONTACT_CNT_ACCURATE | + MT_QUIRK_STICKY_FINGERS | + MT_QUIRK_WIN8_PTP_BUTTONS | + MT_QUIRK_CONFIDENCE | + MT_QUIRK_IGNORE_FEATURE_ID_MISMATCH, + .export_all_inputs = true }, { .name = MT_CLS_VTL, .quirks = MT_QUIRK_ALWAYS_VALID | MT_QUIRK_CONTACT_CNT_ACCURATE | @@ -507,6 +519,7 @@ static const struct attribute_group mt_attribute_group = { static void mt_get_feature(struct hid_device *hdev, struct hid_report *report) { + struct mt_device *td = hid_get_drvdata(hdev); int ret; u32 size = hid_report_len(report); u8 *buf; @@ -528,8 +541,14 @@ static void mt_get_feature(struct hid_device *hdev, struct hid_report *report) dev_warn(&hdev->dev, "failed to fetch feature %d\n", report->id); } else { - /* The report ID in the request and the response should match */ - if (report->id != buf[0]) { + /* + * The report ID in the request and the response should match. + * Some firmware (e.g. the ASUS ROG Z13 Folio + * touchpad) returns a mismatched ID on this specific fetch; + * tolerate it only for devices explicitly flagged as such. + */ + if (report->id != buf[0] && + !(td->mtclass.quirks & MT_QUIRK_IGNORE_FEATURE_ID_MISMATCH)) { hid_err(hdev, "Returned feature report did not match the request\n"); goto free; } @@ -2715,6 +2734,12 @@ static const struct hid_device_id mt_devices[] = { HID_DEVICE(BUS_I2C, HID_GROUP_MULTITOUCH_WIN_8, I2C_VENDOR_ID_HANTICK, I2C_PRODUCT_ID_HANTICK_5288) }, + /* Asus ROG Flow Z13 (2025) GZ302EA keyboard-cover touchpad */ + { .driver_data = MT_CLS_ASUS_ROG_Z13_FOLIO, + HID_DEVICE(BUS_USB, HID_GROUP_MULTITOUCH_WIN_8, + USB_VENDOR_ID_ASUSTEK, + USB_DEVICE_ID_ASUSTEK_ROG_Z13_FOLIO) }, + /* Generic MT device */ { HID_DEVICE(HID_BUS_ANY, HID_GROUP_MULTITOUCH, HID_ANY_ID, HID_ANY_ID) }, From 9eb1a393c89a79c4210230d23e7d88d239c61d7b Mon Sep 17 00:00:00 2001 From: Thadeu Lima de Souza Cascardo Date: Wed, 26 Aug 2026 08:46:58 -0300 Subject: [PATCH 102/592] drm: Fix drm_pending_vblank_event leak in error path for out_fence_ptr When an out_fence_ptr is provided but DRM_MODE_PAGE_FLIP_EVENT is not set, a drm_pending_vblank_event will be allocated. If later, there is an allocation failure or another failure at setup_out_fence(), that event will not have base.fence set and it will not be released at complete_signaling(). Release the event and set crtc_state->event to NULL just like in the DRM_MODE_PAGE_FLIP_EVENT case when there is a failure at drm_event_reserve_init(). That is, prepare_signaling() releases the event and there is nothing to be done at complete_signaling(). Use drm_event_cancel_free() as that will also undo drm_event_reserve_init() in case it has been called. Reported-by: sashiko-bot@kernel.org Closes: https://sashiko.dev/#/patchset/20260727-drm_crtc_atomic_commit_leak-v1-1-23d9948a9d7c@igalia.com?part=1 Fixes: 92c715fca907 ("drm/atomic: Fix double free in drm_atomic_state_default_clear") Signed-off-by: Thadeu Lima de Souza Cascardo Reviewed-by: Melissa Wen Signed-off-by: Melissa Wen Link: https://patch.msgid.link/20260826-drm_pending_vblank_event_leak-v4-1-f8de8b996b9d@igalia.com --- drivers/gpu/drm/drm_atomic_uapi.c | 13 ++++++++++--- 1 file changed, 10 insertions(+), 3 deletions(-) diff --git a/drivers/gpu/drm/drm_atomic_uapi.c b/drivers/gpu/drm/drm_atomic_uapi.c index 5ea593b3a98e..68d59e17ffc2 100644 --- a/drivers/gpu/drm/drm_atomic_uapi.c +++ b/drivers/gpu/drm/drm_atomic_uapi.c @@ -1462,10 +1462,12 @@ static int prepare_signaling(struct drm_device *dev, struct dma_fence *fence; struct drm_out_fence_state *f; + ret = -ENOMEM; + f = krealloc(*fence_state, sizeof(**fence_state) * (*num_fences + 1), GFP_KERNEL); if (!f) - return -ENOMEM; + goto err_free_event; memset(&f[*num_fences], 0, sizeof(*f)); @@ -1474,12 +1476,12 @@ static int prepare_signaling(struct drm_device *dev, fence = drm_crtc_create_fence(crtc); if (!fence) - return -ENOMEM; + goto err_free_event; ret = setup_out_fence(&f[(*num_fences)++], fence); if (ret) { dma_fence_put(fence); - return ret; + goto err_free_event; } crtc_state->event->base.fence = fence; @@ -1535,6 +1537,11 @@ static int prepare_signaling(struct drm_device *dev, } return 0; + +err_free_event: + drm_event_cancel_free(dev, &crtc_state->event->base); + crtc_state->event = NULL; + return ret; } static void complete_signaling(struct drm_device *dev, From a1a5ad37e50ceb192c07ac7e2d7143638cd4d110 Mon Sep 17 00:00:00 2001 From: =?UTF-8?q?Ren=C3=A9=20Onier?= Date: Wed, 12 Aug 2026 13:13:59 -0400 Subject: [PATCH 103/592] HID: winwing: fix use-after-free in force feedback teardown MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit winwing_init_ff() passes the driver's private data, allocated with devm_kzalloc() in winwing_probe(), as the effect context to input_ff_create_memless(). The memoryless force-feedback core takes ownership of that pointer and frees it with kfree() from input_ff_destroy() (ml_ff_destroy()) when the input device is destroyed. Freeing a devm-managed allocation with kfree() is an invalid free, and the same object is then released again by devres when the HID device is torn down, a double free. As the allocation also embeds the LED class devices, their timers and work item live on freed memory and the slab gets corrupted. This triggers on unbind, rmmod, hot-unplug and on system suspend, where the firmware cache walks the now-corrupt devres list. KASAN reports: BUG: KASAN: invalid-free in input_ff_destroy Allocated by task N: winwing_probe Pass NULL as the memless context instead and fetch the driver data from the input device in winwing_play_effect(): the HID core already stores the hid_device as the input device's drvdata. The force-feedback core then owns nothing that it must not free. Fixes: 42d020b54edc ("HID: winwing: Enable rumble effects") Signed-off-by: René Onier Signed-off-by: Jiri Kosina --- drivers/hid/hid-winwing.c | 10 ++++++++-- 1 file changed, 8 insertions(+), 2 deletions(-) diff --git a/drivers/hid/hid-winwing.c b/drivers/hid/hid-winwing.c index 9cd25a77999e..19b92c2c6579 100644 --- a/drivers/hid/hid-winwing.c +++ b/drivers/hid/hid-winwing.c @@ -315,7 +315,8 @@ static void winwing_haptic_rumble_cb(struct work_struct *work) static int winwing_play_effect(struct input_dev *dev, void *context, struct ff_effect *effect) { - struct winwing_drv_data *data = (struct winwing_drv_data *) context; + struct hid_device *hdev = input_get_drvdata(dev); + struct winwing_drv_data *data = hid_get_drvdata(hdev); if (effect->type != FF_RUMBLE) return 0; @@ -342,7 +343,12 @@ static int winwing_init_ff(struct hid_device *hdev, struct hid_input *hidinput) input_set_capability(hidinput->input, EV_FF, FF_RUMBLE); - return input_ff_create_memless(hidinput->input, data, + /* + * input_ff_create_memless() takes ownership of the context pointer + * and frees it on teardown; do not hand it the devm-managed drvdata. + * winwing_play_effect() fetches it from the input device instead. + */ + return input_ff_create_memless(hidinput->input, NULL, winwing_play_effect); } From aa9dde93e05a837645fdfd577eea71f0733f0694 Mon Sep 17 00:00:00 2001 From: Chen Changcheng Date: Fri, 14 Aug 2026 15:06:20 +0800 Subject: [PATCH 104/592] HID: alps: unregister DualPoint Stick input device on remove alps_input_configured() allocates a second input device ("DualPoint Stick") with input_allocate_device() and registers it, but the alps_driver struct has no .remove handler and input2 is not tracked in hdev->inputs. The default remove path (hid_hw_stop -> hidinput_disconnect) only iterates hdev->inputs, so input2 is never unregistered and leaks on every device removal. Add a .remove handler that stops the device first (preventing URB callbacks from touching input2 during teardown) and then unregisters input2. Fixes: 2562756dde55 ("HID: add Alps I2C HID Touchpad-Stick support") Cc: stable@vger.kernel.org Signed-off-by: Chen Changcheng Signed-off-by: Jiri Kosina --- drivers/hid/hid-alps.c | 19 +++++++++++++++++++ 1 file changed, 19 insertions(+) diff --git a/drivers/hid/hid-alps.c b/drivers/hid/hid-alps.c index 67179e3fe39b..370635f5b704 100644 --- a/drivers/hid/hid-alps.c +++ b/drivers/hid/hid-alps.c @@ -823,6 +823,24 @@ static int alps_probe(struct hid_device *hdev, const struct hid_device_id *id) return 0; } +static void alps_remove(struct hid_device *hdev) +{ + struct alps_dev *data = hid_get_drvdata(hdev); + + /* + * input2 ("DualPoint Stick") is allocated separately and is not + * tracked in hdev->inputs, so the default remove path + * (hid_hw_stop -> hidinput_disconnect) does not unregister it. + * + * Stop the device first so that no URB callback can touch input2 + * while it is being unregistered, then drop it explicitly. + */ + hid_hw_stop(hdev); + + if (data->input2) + input_unregister_device(data->input2); +} + static const struct hid_device_id alps_id[] = { { HID_DEVICE(HID_BUS_ANY, HID_GROUP_ANY, USB_VENDOR_ID_ALPS_JP, HID_DEVICE_ID_ALPS_U1_DUAL) }, @@ -845,6 +863,7 @@ static struct hid_driver alps_driver = { .input_configured = alps_input_configured, .resume = pm_ptr(alps_post_resume), .reset_resume = pm_ptr(alps_post_reset), + .remove = alps_remove, }; module_hid_driver(alps_driver); From d3aba3442798ce4a4c8ce3104d7b286d61e605f9 Mon Sep 17 00:00:00 2001 From: Chen Changcheng Date: Fri, 14 Aug 2026 15:06:21 +0800 Subject: [PATCH 105/592] HID: alps: fix use-after-free on input2 registration failure alps_input_configured() stores data->input2 before calling input_register_device(). If registration fails, input_free_device() frees the input device but data->input2 still points to the freed memory. alps_input_configured() calls hid_hw_open() before allocating input2, so URBs are already active and raw_event can fire during the failure window. A U1_SP_ABSOLUTE_REPORT_ID report arriving then causes u1_raw_event() to dereference the freed data->input2 -> use-after-free. Fix by only storing input2 into drvdata after successful registration and adding a NULL guard in the raw_event path. Fixes: 2562756dde55 ("HID: add Alps I2C HID Touchpad-Stick support") Cc: stable@vger.kernel.org Signed-off-by: Chen Changcheng Signed-off-by: Jiri Kosina --- drivers/hid/hid-alps.c | 6 ++++-- 1 file changed, 4 insertions(+), 2 deletions(-) diff --git a/drivers/hid/hid-alps.c b/drivers/hid/hid-alps.c index 370635f5b704..0556cb5645eb 100644 --- a/drivers/hid/hid-alps.c +++ b/drivers/hid/hid-alps.c @@ -407,6 +407,8 @@ static int u1_raw_event(struct alps_dev *hdata, u8 *data, int size) return 1; case U1_SP_ABSOLUTE_REPORT_ID: + if (!hdata->input2) + return 0; sp_x = get_unaligned_le16(data+2); sp_y = get_unaligned_le16(data+4); @@ -738,7 +740,6 @@ static int alps_input_configured(struct hid_device *hdev, struct hid_input *hi) goto exit; } - data->input2 = input2; input2->phys = input->phys; input2->name = "DualPoint Stick"; input2->id.bustype = BUS_I2C; @@ -762,11 +763,12 @@ static int alps_input_configured(struct hid_device *hdev, struct hid_input *hi) __set_bit(INPUT_PROP_POINTER, input2->propbit); __set_bit(INPUT_PROP_POINTING_STICK, input2->propbit); - if (input_register_device(data->input2)) { + if (input_register_device(input2)) { input_free_device(input2); ret = -ENOENT; goto exit; } + data->input2 = input2; } exit: From 31fe2cb51133f00e1138c617fef8ee808a37714d Mon Sep 17 00:00:00 2001 From: Julia Lawall Date: Thu, 27 Aug 2026 13:29:49 +0200 Subject: [PATCH 106/592] HID: fix semantic patch and improve its performance Replace "expression" with "identifier" in the declaration of hdev. This is necessary because hdev is used as the name of a function parameter. Move the two uses of @p2 to the relevant function names. Convert <... ...>, meaning that the contained pattern is optional, to use ..., when any, and exists. This requires that the function contain calls to hid_hw_start, etc, which reduces the set of files that are considered for matching against this pattern. Reported-by: Ricardo Ribalda Signed-off-by: Julia Lawall Signed-off-by: Jiri Kosina --- scripts/coccinelle/hid/ff_race.cocci | 11 ++++++----- 1 file changed, 6 insertions(+), 5 deletions(-) diff --git a/scripts/coccinelle/hid/ff_race.cocci b/scripts/coccinelle/hid/ff_race.cocci index 479f5d1e3184..e861de00cb10 100644 --- a/scripts/coccinelle/hid/ff_race.cocci +++ b/scripts/coccinelle/hid/ff_race.cocci @@ -7,18 +7,19 @@ virtual report -@r@ +@r exists@ identifier probe_fn; -expression hdev, flags; +identifier hdev; +expression flags; position p1, p2; @@ probe_fn(struct hid_device *hdev, ...) { - <... + ... when any hid_hw_start@p1(hdev, flags) ... - \(input_ff_create\|input_ff_create_memless\)@p2(...) - ...> + \(input_ff_create@p2\|input_ff_create_memless@p2\)(...) + ... when any } @script:python depends on report@ From d5ea0d226e8f0801d78702142a124d78c317d822 Mon Sep 17 00:00:00 2001 From: Runyu Xiao Date: Wed, 2 Sep 2026 22:09:26 +0800 Subject: [PATCH 107/592] mmc: sh_mmcif: initialize IRQ-thread mutex before requesting interrupt The threaded IRQ handler can run before devm_request_threaded_irq() returns, but thread_lock was initialized afterwards. Initialize it before requesting either interrupt. Fixes: 8047310ee984 ("mmc: sh_mmcif: fix a race, causing an Oops on SMP") Cc: stable@vger.kernel.org Assisted-by: Codex:GPT-5 Signed-off-by: Runyu Xiao Signed-off-by: Ulf Hansson --- drivers/mmc/host/sh_mmcif.c | 3 +-- 1 file changed, 1 insertion(+), 2 deletions(-) diff --git a/drivers/mmc/host/sh_mmcif.c b/drivers/mmc/host/sh_mmcif.c index 7706d01b149e..8bd9cdedafa4 100644 --- a/drivers/mmc/host/sh_mmcif.c +++ b/drivers/mmc/host/sh_mmcif.c @@ -1461,6 +1461,7 @@ static int sh_mmcif_probe(struct platform_device *pdev) host->pd = pdev; spin_lock_init(&host->lock); + mutex_init(&host->thread_lock); mmc->ops = &sh_mmcif_ops; sh_mmcif_init_ocr(host); @@ -1515,8 +1516,6 @@ static int sh_mmcif_probe(struct platform_device *pdev) goto err_clk; } - mutex_init(&host->thread_lock); - ret = mmc_add_host(mmc); if (ret < 0) goto err_clk; From fce551616dfa4523c15fe50c27aa3baf8afda955 Mon Sep 17 00:00:00 2001 From: Andres Diaz Date: Mon, 31 Aug 2026 23:01:47 -0600 Subject: [PATCH 108/592] HID: logitech-hidpp: Add support for G502 X Lightspeed USB mouse The G502 X Lightspeed enumerates as 046d:c098 when connected with its cable. Add the id so the driver handles the wired device. Signed-off-by: Andres Diaz Signed-off-by: Jiri Kosina --- drivers/hid/hid-logitech-hidpp.c | 2 ++ 1 file changed, 2 insertions(+) diff --git a/drivers/hid/hid-logitech-hidpp.c b/drivers/hid/hid-logitech-hidpp.c index 1504de32b1c8..493763a12518 100644 --- a/drivers/hid/hid-logitech-hidpp.c +++ b/drivers/hid/hid-logitech-hidpp.c @@ -4924,6 +4924,8 @@ static const struct hid_device_id hidpp_devices[] = { HID_USB_DEVICE(USB_VENDOR_ID_LOGITECH, 0xC08D) }, { /* Logitech G502 X Plus Wireless Gaming Mouse over USB */ HID_USB_DEVICE(USB_VENDOR_ID_LOGITECH, 0xC095) }, + { /* Logitech G502 X Lightspeed Wireless Gaming Mouse over USB */ + HID_USB_DEVICE(USB_VENDOR_ID_LOGITECH, 0xC098) }, { /* Logitech G703 Gaming Mouse over USB */ HID_USB_DEVICE(USB_VENDOR_ID_LOGITECH, 0xC087) }, { /* Logitech G703 Hero Gaming Mouse over USB */ From 7e749a7972829a53fd1e41568cf2018c180627d1 Mon Sep 17 00:00:00 2001 From: =?UTF-8?q?Erik=20H=C3=A5kansson?= Date: Tue, 1 Sep 2026 22:30:44 +0200 Subject: [PATCH 109/592] HID: steelseries: Add support for Arctis 7 (2018) MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit The headset reports connection and battery status on HID interface 5. When the device is disconnected, ignore incoming battery reports as they will incorrectly report battery level 0. Clamp overreported battery values to 100% and add USB ID 1038:12ad to the driver's device tables. Signed-off-by: Erik Håkansson Signed-off-by: Jiri Kosina --- drivers/hid/Kconfig | 3 +- drivers/hid/hid-ids.h | 1 + drivers/hid/hid-quirks.c | 1 + drivers/hid/hid-steelseries-arctis.c | 44 ++++++++++++++++++++++++++++ 4 files changed, 48 insertions(+), 1 deletion(-) diff --git a/drivers/hid/Kconfig b/drivers/hid/Kconfig index a81bf51cbcf1..c43e824428f0 100644 --- a/drivers/hid/Kconfig +++ b/drivers/hid/Kconfig @@ -1210,7 +1210,8 @@ config HID_STEELSERIES depends on USB_HID help Support for Steelseries SRW-S1 steering wheel, and the Steelseries - Arctis 1 Wireless for XBox headset. + Arctis 1 Wireless for XBox, Arctis 7 (2018), Arctis 9, and Arctis + Nova headsets. config HID_SUNPLUS tristate "Sunplus wireless desktop" diff --git a/drivers/hid/hid-ids.h b/drivers/hid/hid-ids.h index 8b4f4b02aec0..874ed8e49302 100644 --- a/drivers/hid/hid-ids.h +++ b/drivers/hid/hid-ids.h @@ -1398,6 +1398,7 @@ #define USB_VENDOR_ID_STEELSERIES 0x1038 #define USB_DEVICE_ID_STEELSERIES_SRWS1 0x1410 #define USB_DEVICE_ID_STEELSERIES_ARCTIS_1_X 0x12b6 +#define USB_DEVICE_ID_STEELSERIES_ARCTIS_7_2018 0x12ad #define USB_DEVICE_ID_STEELSERIES_ARCTIS_9 0x12c2 #define USB_DEVICE_ID_STEELSERIES_ARCTIS_NOVA_5_X 0x2253 #define USB_DEVICE_ID_STEELSERIES_ARCTIS_NOVA_7 0x2202 diff --git a/drivers/hid/hid-quirks.c b/drivers/hid/hid-quirks.c index c8c6b29fc04d..9ac90b7efcb1 100644 --- a/drivers/hid/hid-quirks.c +++ b/drivers/hid/hid-quirks.c @@ -752,6 +752,7 @@ static const struct hid_device_id hid_have_special_driver[] = { #if IS_ENABLED(CONFIG_HID_STEELSERIES) { HID_USB_DEVICE(USB_VENDOR_ID_STEELSERIES, USB_DEVICE_ID_STEELSERIES_SRWS1) }, { HID_USB_DEVICE(USB_VENDOR_ID_STEELSERIES, USB_DEVICE_ID_STEELSERIES_ARCTIS_1_X) }, + { HID_USB_DEVICE(USB_VENDOR_ID_STEELSERIES, USB_DEVICE_ID_STEELSERIES_ARCTIS_7_2018) }, { HID_USB_DEVICE(USB_VENDOR_ID_STEELSERIES, USB_DEVICE_ID_STEELSERIES_ARCTIS_9) }, { HID_USB_DEVICE(USB_VENDOR_ID_STEELSERIES, USB_DEVICE_ID_STEELSERIES_ARCTIS_NOVA_5_X) }, { HID_USB_DEVICE(USB_VENDOR_ID_STEELSERIES, USB_DEVICE_ID_STEELSERIES_ARCTIS_NOVA_7) }, diff --git a/drivers/hid/hid-steelseries-arctis.c b/drivers/hid/hid-steelseries-arctis.c index 23fb0cebd72a..fa33fbd94889 100644 --- a/drivers/hid/hid-steelseries-arctis.c +++ b/drivers/hid/hid-steelseries-arctis.c @@ -101,6 +101,19 @@ static int steelseries_arctis_1_request_status(struct hid_device *hdev) return steelseries_send_output_report(hdev, data, sizeof(data)); } +static int steelseries_arctis_7_2018_request_status(struct hid_device *hdev) +{ + const u8 connection[] = { 0x06, 0x14 }; + const u8 battery[] = { 0x06, 0x18 }; + int ret; + + ret = steelseries_send_output_report(hdev, connection, sizeof(connection)); + if (ret) + return ret; + + return steelseries_send_output_report(hdev, battery, sizeof(battery)); +} + static int steelseries_arctis_9_request_status(struct hid_device *hdev) { const u8 data[] = { 0x00, 0x20 }; @@ -152,6 +165,27 @@ static void steelseries_arctis_1_parse_status(struct steelseries_device *sd, sd->battery_capacity = data[3]; } +static void steelseries_arctis_7_2018_parse_status(struct steelseries_device *sd, + u8 *data, int size) +{ + if (size < 3 || data[0] != 0x06) + return; + + switch (data[1]) { + case 0x14: + /* 0x03 means that the headset is connected to the transmitter. */ + sd->headset_connected = data[2] == 0x03; + break; + case 0x18: + if (!sd->headset_connected) + break; + + /* The Arctis 7 sometimes overreports battery. Cap to 100. */ + sd->battery_capacity = steelseries_map_capacity(data[2], 0, 100); + break; + } +} + static void steelseries_arctis_9_parse_status(struct steelseries_device *sd, u8 *data, int size) { @@ -232,6 +266,13 @@ static const struct steelseries_device_info arctis_1_info = { .parse_status = steelseries_arctis_1_parse_status, }; +static const struct steelseries_device_info arctis_7_2018_info = { + .sync_interface = 5, + .capabilities = SS_CAP_BATTERY, + .request_status = steelseries_arctis_7_2018_request_status, + .parse_status = steelseries_arctis_7_2018_parse_status, +}; + static const struct steelseries_device_info arctis_9_info = { .sync_interface = 0, .capabilities = SS_CAP_BATTERY, @@ -653,6 +694,9 @@ static const struct hid_device_id steelseries_arctis_devices[] = { { HID_USB_DEVICE(USB_VENDOR_ID_STEELSERIES, USB_DEVICE_ID_STEELSERIES_ARCTIS_1_X), .driver_data = (unsigned long)&arctis_1_info }, + { HID_USB_DEVICE(USB_VENDOR_ID_STEELSERIES, + USB_DEVICE_ID_STEELSERIES_ARCTIS_7_2018), + .driver_data = (unsigned long)&arctis_7_2018_info }, { HID_USB_DEVICE(USB_VENDOR_ID_STEELSERIES, USB_DEVICE_ID_STEELSERIES_ARCTIS_9), .driver_data = (unsigned long)&arctis_9_info }, From d76994443eed0af297cb82bd038ae9d76327ef90 Mon Sep 17 00:00:00 2001 From: Dmitry Antipov Date: Wed, 2 Sep 2026 12:45:49 +0300 Subject: [PATCH 110/592] HID: roccat: fix locking in roccat_connect() and roccat_disconnect() Extend critical section in roccat_connect() to ensure that partially initialized 'struct roccat_device' is never exposed in 'devices' list, and do the same in roccat_disconnect() to avoid racy 'devices' access against roccat_release(). Signed-off-by: Dmitry Antipov Signed-off-by: Jiri Kosina --- drivers/hid/hid-roccat.c | 8 +++----- 1 file changed, 3 insertions(+), 5 deletions(-) diff --git a/drivers/hid/hid-roccat.c b/drivers/hid/hid-roccat.c index 4f15eb951039..5deb6da8d4f7 100644 --- a/drivers/hid/hid-roccat.c +++ b/drivers/hid/hid-roccat.c @@ -344,8 +344,6 @@ int roccat_connect(const struct class *klass, struct hid_device *hid, int report return temp; } - mutex_unlock(&devices_lock); - init_waitqueue_head(&device->wait); INIT_LIST_HEAD(&device->readers); mutex_init(&device->readers_lock); @@ -356,6 +354,7 @@ int roccat_connect(const struct class *klass, struct hid_device *hid, int report device->cbuf_end = 0; device->report_size = report_size; + mutex_unlock(&devices_lock); return minor; } EXPORT_SYMBOL_GPL(roccat_connect); @@ -369,15 +368,12 @@ void roccat_disconnect(int minor) mutex_lock(&devices_lock); device = devices[minor]; - mutex_unlock(&devices_lock); device->exist = 0; /* TODO exist maybe not needed */ device_destroy(device->dev->class, MKDEV(roccat_major, minor)); - mutex_lock(&devices_lock); devices[minor] = NULL; - mutex_unlock(&devices_lock); if (device->open) { hid_hw_close(device->hid); @@ -385,6 +381,8 @@ void roccat_disconnect(int minor) } else { roccat_free_device(device); } + + mutex_unlock(&devices_lock); } EXPORT_SYMBOL_GPL(roccat_disconnect); From 39e8e085710a2d36d9ceade11739fb7b87f5aeb0 Mon Sep 17 00:00:00 2001 From: Oleg Keri Date: Thu, 10 Sep 2026 08:30:50 +0200 Subject: [PATCH 111/592] HID: i2c-hid: add reset quirk for Lenovo Yoga Slim 7x Gen 11 keyboard The ITE controller behind the keyboard of the Lenovo Yoga Slim 7x Gen 11 (048d:83db) carries out a reset but never raises the interrupt that acknowledges it. i2c_hid_finish_hwreset() therefore waits out its full one second timeout and logs "device did not ack reset within 1000 ms" on every probe and every resume, before the keyboard comes up regardless. Set I2C_HID_QUIRK_NO_IRQ_AFTER_RESET for it, as is already done for several other ITE parts, so the reset is followed by a fixed 100 ms sleep instead. Signed-off-by: Oleg Keri Signed-off-by: Jiri Kosina --- drivers/hid/hid-ids.h | 1 + drivers/hid/i2c-hid/i2c-hid-core.c | 2 ++ 2 files changed, 3 insertions(+) diff --git a/drivers/hid/hid-ids.h b/drivers/hid/hid-ids.h index 874ed8e49302..06a1309589e3 100644 --- a/drivers/hid/hid-ids.h +++ b/drivers/hid/hid-ids.h @@ -765,6 +765,7 @@ #define I2C_DEVICE_ID_ITE_LENOVO_LEGION_Y720 0x837a #define USB_DEVICE_ID_ITE_LENOVO_YOGA900 0x8396 #define I2C_DEVICE_ID_ITE_LENOVO_YOGA_SLIM_7X_KEYBOARD 0x8987 +#define I2C_DEVICE_ID_ITE_LENOVO_YOGA_SLIM_7X_G11_KEYBOARD 0x83db #define USB_DEVICE_ID_ITE8595 0x8595 #define USB_DEVICE_ID_ITE_MEDION_E1239T 0xce50 diff --git a/drivers/hid/i2c-hid/i2c-hid-core.c b/drivers/hid/i2c-hid/i2c-hid-core.c index 0ff07fdab442..ad8d9f329404 100644 --- a/drivers/hid/i2c-hid/i2c-hid-core.c +++ b/drivers/hid/i2c-hid/i2c-hid-core.c @@ -126,6 +126,8 @@ static const struct i2c_hid_quirks { I2C_HID_QUIRK_NO_IRQ_AFTER_RESET }, { I2C_VENDOR_ID_ITE, I2C_DEVICE_ID_ITE_VOYO_WINPAD_A15, I2C_HID_QUIRK_NO_IRQ_AFTER_RESET }, + { USB_VENDOR_ID_ITE, I2C_DEVICE_ID_ITE_LENOVO_YOGA_SLIM_7X_G11_KEYBOARD, + I2C_HID_QUIRK_NO_IRQ_AFTER_RESET }, { I2C_VENDOR_ID_RAYDIUM, I2C_PRODUCT_ID_RAYDIUM_3118, I2C_HID_QUIRK_NO_IRQ_AFTER_RESET }, { USB_VENDOR_ID_ALPS_JP, HID_ANY_ID, From abd24922c2a9797d6184be0561dd85e7bcdfd091 Mon Sep 17 00:00:00 2001 From: Tristan Madani Date: Fri, 4 Sep 2026 10:58:00 +0000 Subject: [PATCH 112/592] HID: hid-oxp: use cancel_delayed_work_sync() in remove oxp_hid_remove() uses cancel_delayed_work() for all three delayed work items. cancel_delayed_work() only dequeues a pending work item without waiting for a currently executing callback to finish. If any of the work callbacks (oxp_rgb_queue_fn, oxp_btn_queue_fn, oxp_mcu_init_fn) is running at the time of removal, the callback continues executing concurrently with hid_hw_close() and hid_hw_stop(), accessing the HID device after it has been closed and stopped. Use cancel_delayed_work_sync() instead to ensure that any in-progress work callback completes before device teardown proceeds. Fixes: 84910c459d65 ("HID: hid-oxp: Add OneXPlayer configuration driver") Cc: stable@vger.kernel.org Signed-off-by: Tristan Madani Reviewed-by: Derek J. Clark Link: https://lore.kernel.org/r/20260804-oxp-fix-v2-1-b2d56e4c8a2c@cherr.cc Link: https://lore.kernel.org/r/20260804-oxp-fix-v1-1-51a4fe787167@cherr.cc Signed-off-by: Jiri Kosina --- drivers/hid/hid-oxp.c | 6 +++--- 1 file changed, 3 insertions(+), 3 deletions(-) diff --git a/drivers/hid/hid-oxp.c b/drivers/hid/hid-oxp.c index d2ded6b08ce9..1e691ebc1199 100644 --- a/drivers/hid/hid-oxp.c +++ b/drivers/hid/hid-oxp.c @@ -1552,9 +1552,9 @@ static int oxp_hid_probe(struct hid_device *hdev, static void oxp_hid_remove(struct hid_device *hdev) { - cancel_delayed_work(&drvdata.oxp_rgb_queue); - cancel_delayed_work(&drvdata.oxp_btn_queue); - cancel_delayed_work(&drvdata.oxp_mcu_init); + cancel_delayed_work_sync(&drvdata.oxp_rgb_queue); + cancel_delayed_work_sync(&drvdata.oxp_btn_queue); + cancel_delayed_work_sync(&drvdata.oxp_mcu_init); hid_hw_close(hdev); hid_hw_stop(hdev); } From 58d97b45f3f43af0afd9c74e8480d22d36fd3f72 Mon Sep 17 00:00:00 2001 From: Youth Cao Date: Fri, 4 Sep 2026 00:14:44 +0800 Subject: [PATCH 113/592] HID: i2c-hid: Add i2c-hid-quirk-bad-input-size quirk for 0911:5288 device I have recently acquired a cheap Apollo Lake-based laptop that uses a Hynitron CST128-A touchpad controller. While booting from a Debian LiveCD, the kernel log is flooded with the following error (though the touchpad works well): i2c_hid_acpi i2c-ALPS0001:00: i2c_hid_get_input: incomplete report (27/42405) The CST128-A was identified via ACPI as ALPS0001:00, and the I2C HID device ID (0911:5288) was shared with the Hantick 5288. Add the I2C_HID_QUIRK_BAD_INPUT_SIZE quirk option to the existing Hantick 5288 quirk entry to suppress the kernel log flood. Signed-off-by: Youth Cao Signed-off-by: Jiri Kosina --- drivers/hid/i2c-hid/i2c-hid-core.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/drivers/hid/i2c-hid/i2c-hid-core.c b/drivers/hid/i2c-hid/i2c-hid-core.c index ad8d9f329404..a61fd1de5203 100644 --- a/drivers/hid/i2c-hid/i2c-hid-core.c +++ b/drivers/hid/i2c-hid/i2c-hid-core.c @@ -123,7 +123,7 @@ static const struct i2c_hid_quirks { __u32 quirks; } i2c_hid_quirks[] = { { I2C_VENDOR_ID_HANTICK, I2C_PRODUCT_ID_HANTICK_5288, - I2C_HID_QUIRK_NO_IRQ_AFTER_RESET }, + I2C_HID_QUIRK_NO_IRQ_AFTER_RESET | I2C_HID_QUIRK_BAD_INPUT_SIZE }, { I2C_VENDOR_ID_ITE, I2C_DEVICE_ID_ITE_VOYO_WINPAD_A15, I2C_HID_QUIRK_NO_IRQ_AFTER_RESET }, { USB_VENDOR_ID_ITE, I2C_DEVICE_ID_ITE_LENOVO_YOGA_SLIM_7X_G11_KEYBOARD, From 1d00442cc4699accfaa2317fb18259ddd7556f9a Mon Sep 17 00:00:00 2001 From: Stuart Hayhurst Date: Tue, 18 Aug 2026 14:00:53 +0100 Subject: [PATCH 114/592] HID: corsair-void: Fix firmware event packet description The size was incorrectly stated as 4 bytes since the ID was missed out. Add the ID in and correct the indices for the firmware versions. Signed-off-by: Stuart Hayhurst Signed-off-by: Jiri Kosina --- drivers/hid/hid-corsair-void.c | 15 +++++++++------ 1 file changed, 9 insertions(+), 6 deletions(-) diff --git a/drivers/hid/hid-corsair-void.c b/drivers/hid/hid-corsair-void.c index 071a663a6c26..7be8c974d11e 100644 --- a/drivers/hid/hid-corsair-void.c +++ b/drivers/hid/hid-corsair-void.c @@ -51,20 +51,23 @@ /* Receiver report information: (ID 102) */ /* -------------------------------------------------------------------------- */ /* - * When queried, the recevier responds with 4 bytes to describe the firmware - * The first 2 bytes are for the receiver, the second 2 are the headset + * When queried, the receiver responds with 5 bytes to describe the firmware + * The first byte is the ID, then 2 bytes for the receiver, then 2 for the headset * The headset firmware version will be 0 if no headset is connected * - * 0: Recevier firmware major version + * 0: Report ID + * 102 for the firmware packet + * + * 1: Receiver firmware major version * Major version of the receiver's firmware * - * 1: Recevier firmware minor version + * 2: Receiver firmware minor version * Minor version of the receiver's firmware * - * 2: Headset firmware major version + * 3: Headset firmware major version * Major version of the headset's firmware * - * 3: Headset firmware minor version + * 4: Headset firmware minor version * Minor version of the headset's firmware */ /* -------------------------------------------------------------------------- */ From f3c2b266b8fd7cf816d36168d19f67e716f2c080 Mon Sep 17 00:00:00 2001 From: Berke Durak Date: Mon, 7 Sep 2026 11:47:19 -0700 Subject: [PATCH 115/592] HID: elecom: Add support for ELECOM M-XT4DRBK (018E) The 2025 revision of the left-handed EX-G wireless trackball (M-XT4DRBK-G) reports USB ID 056e:018e instead of 056e:00fd. Its report descriptor declares an 8-bit button field (Report Count 8) but only five usages (Usage Maximum 5), so the sixth (Fn) button ends up as a duplicate of button 5 and is unusable. The report descriptor has the same layout as the M-XT3DRBK 018C (button usage maximum at offset 16, button report count at 22, button report size at 24, padding report size at 30), so reuse that fixup. Rename the existing M_XT4DRBK define to M_XT4DRBK_00FD to match the convention used for the other EX-G revisions. Tested on a Raspberry Pi 5 with the same fixup in an out-of-tree module on 6.12 (6.12.96+rpt-rpi-2712); all six buttons are reported after the fix, and they work (tested with xev as well). Report descriptor as sent by the device, pre-fix (056e:018e, 215 bytes): 05 01 09 02 A1 01 09 01 A1 00 85 01 05 09 19 01 29 05 15 00 25 01 95 08 75 01 81 02 95 01 75 00 81 01 05 01 09 30 09 31 16 00 80 26 FF 7F 75 10 95 02 81 06 C0 A1 00 05 01 09 38 15 81 25 7F 75 08 95 01 81 06 C0 A1 00 05 0C 0A 38 02 95 01 75 08 15 81 25 7F 81 06 C0 C0 06 01 FF 09 00 A1 01 85 02 09 00 15 00 26 FF 00 75 08 95 07 81 02 C0 05 0C 09 01 A1 01 85 05 15 00 26 3C 02 19 00 2A 3C 02 75 10 95 01 81 00 C0 05 01 09 80 A1 01 85 03 19 81 29 83 15 00 25 01 95 03 75 01 81 02 95 01 75 05 81 01 C0 06 BC FF 09 88 A1 01 85 04 95 01 75 08 15 00 26 FF 00 19 00 2A FF 00 81 00 C0 06 02 FF 09 02 A1 01 85 06 09 02 15 00 26 FF 00 75 08 95 07 B1 02 C0 Assisted-by: LLM Signed-off-by: Berke Durak Signed-off-by: Jiri Kosina --- drivers/hid/hid-elecom.c | 6 ++++-- drivers/hid/hid-ids.h | 3 ++- drivers/hid/hid-quirks.c | 3 ++- 3 files changed, 8 insertions(+), 4 deletions(-) diff --git a/drivers/hid/hid-elecom.c b/drivers/hid/hid-elecom.c index 37d88ce57f67..d21ead40bb07 100644 --- a/drivers/hid/hid-elecom.c +++ b/drivers/hid/hid-elecom.c @@ -79,7 +79,7 @@ static const __u8 *elecom_report_fixup(struct hid_device *hdev, __u8 *rdesc, case USB_DEVICE_ID_ELECOM_M_XT3URBK_00FB: case USB_DEVICE_ID_ELECOM_M_XT3URBK_018F: case USB_DEVICE_ID_ELECOM_M_XT3DRBK_00FC: - case USB_DEVICE_ID_ELECOM_M_XT4DRBK: + case USB_DEVICE_ID_ELECOM_M_XT4DRBK_00FD: /* * Report descriptor format: * 12: button bit count @@ -104,6 +104,7 @@ static const __u8 *elecom_report_fixup(struct hid_device *hdev, __u8 *rdesc, mouse_button_fixup(hdev, rdesc, *rsize, 12, 30, 14, 20, 8); break; case USB_DEVICE_ID_ELECOM_M_XT3DRBK_018C: + case USB_DEVICE_ID_ELECOM_M_XT4DRBK_018E: /* * Report descriptor format: * 22: button bit count @@ -148,7 +149,8 @@ static const struct hid_device_id elecom_devices[] = { { HID_USB_DEVICE(USB_VENDOR_ID_ELECOM, USB_DEVICE_ID_ELECOM_M_XT3URBK_018F) }, { HID_USB_DEVICE(USB_VENDOR_ID_ELECOM, USB_DEVICE_ID_ELECOM_M_XT3DRBK_00FC) }, { HID_USB_DEVICE(USB_VENDOR_ID_ELECOM, USB_DEVICE_ID_ELECOM_M_XT3DRBK_018C) }, - { HID_USB_DEVICE(USB_VENDOR_ID_ELECOM, USB_DEVICE_ID_ELECOM_M_XT4DRBK) }, + { HID_USB_DEVICE(USB_VENDOR_ID_ELECOM, USB_DEVICE_ID_ELECOM_M_XT4DRBK_00FD) }, + { HID_USB_DEVICE(USB_VENDOR_ID_ELECOM, USB_DEVICE_ID_ELECOM_M_XT4DRBK_018E) }, { HID_USB_DEVICE(USB_VENDOR_ID_ELECOM, USB_DEVICE_ID_ELECOM_M_DT1URBK) }, { HID_USB_DEVICE(USB_VENDOR_ID_ELECOM, USB_DEVICE_ID_ELECOM_M_DT1DRBK) }, { HID_USB_DEVICE(USB_VENDOR_ID_ELECOM, USB_DEVICE_ID_ELECOM_M_DT2DRBK) }, diff --git a/drivers/hid/hid-ids.h b/drivers/hid/hid-ids.h index 06a1309589e3..9a473f72c8cb 100644 --- a/drivers/hid/hid-ids.h +++ b/drivers/hid/hid-ids.h @@ -477,7 +477,8 @@ #define USB_DEVICE_ID_ELECOM_M_XT3URBK_018F 0x018f #define USB_DEVICE_ID_ELECOM_M_XT3DRBK_00FC 0x00fc #define USB_DEVICE_ID_ELECOM_M_XT3DRBK_018C 0x018c -#define USB_DEVICE_ID_ELECOM_M_XT4DRBK 0x00fd +#define USB_DEVICE_ID_ELECOM_M_XT4DRBK_00FD 0x00fd +#define USB_DEVICE_ID_ELECOM_M_XT4DRBK_018E 0x018e #define USB_DEVICE_ID_ELECOM_M_DT1URBK 0x00fe #define USB_DEVICE_ID_ELECOM_M_DT1DRBK 0x00ff #define USB_DEVICE_ID_ELECOM_M_DT2DRBK 0x018d diff --git a/drivers/hid/hid-quirks.c b/drivers/hid/hid-quirks.c index 9ac90b7efcb1..8c4b7580a6ee 100644 --- a/drivers/hid/hid-quirks.c +++ b/drivers/hid/hid-quirks.c @@ -432,7 +432,8 @@ static const struct hid_device_id hid_have_special_driver[] = { { HID_USB_DEVICE(USB_VENDOR_ID_ELECOM, USB_DEVICE_ID_ELECOM_M_XT3URBK_018F) }, { HID_USB_DEVICE(USB_VENDOR_ID_ELECOM, USB_DEVICE_ID_ELECOM_M_XT3DRBK_00FC) }, { HID_USB_DEVICE(USB_VENDOR_ID_ELECOM, USB_DEVICE_ID_ELECOM_M_XT3DRBK_018C) }, - { HID_USB_DEVICE(USB_VENDOR_ID_ELECOM, USB_DEVICE_ID_ELECOM_M_XT4DRBK) }, + { HID_USB_DEVICE(USB_VENDOR_ID_ELECOM, USB_DEVICE_ID_ELECOM_M_XT4DRBK_00FD) }, + { HID_USB_DEVICE(USB_VENDOR_ID_ELECOM, USB_DEVICE_ID_ELECOM_M_XT4DRBK_018E) }, { HID_USB_DEVICE(USB_VENDOR_ID_ELECOM, USB_DEVICE_ID_ELECOM_M_DT1URBK) }, { HID_USB_DEVICE(USB_VENDOR_ID_ELECOM, USB_DEVICE_ID_ELECOM_M_DT1DRBK) }, { HID_USB_DEVICE(USB_VENDOR_ID_ELECOM, USB_DEVICE_ID_ELECOM_M_DT2DRBK) }, From 8e2a4b458ad25e13422bb059758c30a6562aa9cf Mon Sep 17 00:00:00 2001 From: Oscar Priego Verdugo Date: Mon, 17 Aug 2026 06:05:46 -0600 Subject: [PATCH 116/592] HID: elecom: fix bus type for M-XGL20DLBK The M-XGL20DLBK is matched as a USB device by hid-elecom, but its entry in hid_have_special_driver[] uses HID_BLUETOOTH_DEVICE. This prevents the special-driver quirk entry from matching the USB device handled by hid-elecom. Use HID_USB_DEVICE there as well. Fixes: 55633e681afb ("HID: elecom: add support for EX-G M-XGL20DLBK wireless mouse") Signed-off-by: Oscar Priego Verdugo Signed-off-by: Jiri Kosina --- drivers/hid/hid-quirks.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/drivers/hid/hid-quirks.c b/drivers/hid/hid-quirks.c index 8c4b7580a6ee..96a36c5ba04f 100644 --- a/drivers/hid/hid-quirks.c +++ b/drivers/hid/hid-quirks.c @@ -426,7 +426,7 @@ static const struct hid_device_id hid_have_special_driver[] = { #endif #if IS_ENABLED(CONFIG_HID_ELECOM) { HID_BLUETOOTH_DEVICE(USB_VENDOR_ID_ELECOM, USB_DEVICE_ID_ELECOM_BM084) }, - { HID_BLUETOOTH_DEVICE(USB_VENDOR_ID_ELECOM, USB_DEVICE_ID_ELECOM_M_XGL20DLBK) }, + { HID_USB_DEVICE(USB_VENDOR_ID_ELECOM, USB_DEVICE_ID_ELECOM_M_XGL20DLBK) }, { HID_BLUETOOTH_DEVICE(USB_VENDOR_ID_ELECOM, USB_DEVICE_ID_ELECOM_M_HT1MRBK_01AC) }, { HID_USB_DEVICE(USB_VENDOR_ID_ELECOM, USB_DEVICE_ID_ELECOM_M_XT3URBK_00FB) }, { HID_USB_DEVICE(USB_VENDOR_ID_ELECOM, USB_DEVICE_ID_ELECOM_M_XT3URBK_018F) }, From a5c41fa7f925fda2db394329fa0b26243fa63a81 Mon Sep 17 00:00:00 2001 From: Lyude Paul Date: Fri, 14 Aug 2026 15:43:48 -0400 Subject: [PATCH 117/592] Revert "nouveau/gsp: fix suspend/resume regression on r570 firmware" This reverts commit 8302d0afeaec0bc57d951dd085e0cffe997d4d18. It turns out this looked like the right fix on some systems, but it's not - as this causes runtime PM to actually fail on many a laptop. Fixes: 8302d0afeaec ("nouveau/gsp: fix suspend/resume regression on r570 firmware") Cc: # v6.19+ Signed-off-by: Lyude Paul Reviewed-by: Dave Airlie Link: https://patch.msgid.link/20260814194542.781955-2-lyude@redhat.com (cherry picked from commit 94097122bfd701976bc1a62ccd434c13f3f67cde) Signed-off-by: Lyude Paul --- drivers/gpu/drm/nouveau/nvkm/subdev/gsp/rm/r535/fbsr.c | 2 +- drivers/gpu/drm/nouveau/nvkm/subdev/gsp/rm/r535/gsp.c | 2 +- drivers/gpu/drm/nouveau/nvkm/subdev/gsp/rm/r570/fbsr.c | 8 ++++---- drivers/gpu/drm/nouveau/nvkm/subdev/gsp/rm/rm.h | 2 +- 4 files changed, 7 insertions(+), 7 deletions(-) diff --git a/drivers/gpu/drm/nouveau/nvkm/subdev/gsp/rm/r535/fbsr.c b/drivers/gpu/drm/nouveau/nvkm/subdev/gsp/rm/r535/fbsr.c index f128330f30d7..40bf83ea33ac 100644 --- a/drivers/gpu/drm/nouveau/nvkm/subdev/gsp/rm/r535/fbsr.c +++ b/drivers/gpu/drm/nouveau/nvkm/subdev/gsp/rm/r535/fbsr.c @@ -208,7 +208,7 @@ r535_fbsr_resume(struct nvkm_gsp *gsp) } static int -r535_fbsr_suspend(struct nvkm_gsp *gsp, bool runtime) +r535_fbsr_suspend(struct nvkm_gsp *gsp) { struct nvkm_subdev *subdev = &gsp->subdev; struct nvkm_device *device = subdev->device; diff --git a/drivers/gpu/drm/nouveau/nvkm/subdev/gsp/rm/r535/gsp.c b/drivers/gpu/drm/nouveau/nvkm/subdev/gsp/rm/r535/gsp.c index f544afa12b6b..4a3b771ded25 100644 --- a/drivers/gpu/drm/nouveau/nvkm/subdev/gsp/rm/r535/gsp.c +++ b/drivers/gpu/drm/nouveau/nvkm/subdev/gsp/rm/r535/gsp.c @@ -1749,7 +1749,7 @@ r535_gsp_fini(struct nvkm_gsp *gsp, enum nvkm_suspend_state suspend) sr->sysmemAddrOfSuspendResumeData = gsp->sr.radix3.lvl0.addr; sr->sizeOfSuspendResumeData = len; - ret = rm->api->fbsr->suspend(gsp, suspend == NVKM_RUNTIME_SUSPEND); + ret = rm->api->fbsr->suspend(gsp); if (ret) { nvkm_gsp_mem_dtor(&gsp->sr.meta); nvkm_gsp_radix3_dtor(gsp, &gsp->sr.radix3); diff --git a/drivers/gpu/drm/nouveau/nvkm/subdev/gsp/rm/r570/fbsr.c b/drivers/gpu/drm/nouveau/nvkm/subdev/gsp/rm/r570/fbsr.c index 8ef8b4f65588..2945d5b4e570 100644 --- a/drivers/gpu/drm/nouveau/nvkm/subdev/gsp/rm/r570/fbsr.c +++ b/drivers/gpu/drm/nouveau/nvkm/subdev/gsp/rm/r570/fbsr.c @@ -62,7 +62,7 @@ r570_fbsr_resume(struct nvkm_gsp *gsp) } static int -r570_fbsr_init(struct nvkm_gsp *gsp, struct sg_table *sgt, u64 size, bool runtime) +r570_fbsr_init(struct nvkm_gsp *gsp, struct sg_table *sgt, u64 size) { NV2080_CTRL_INTERNAL_FBSR_INIT_PARAMS *ctrl; struct nvkm_gsp_object memlist; @@ -81,7 +81,7 @@ r570_fbsr_init(struct nvkm_gsp *gsp, struct sg_table *sgt, u64 size, bool runtim ctrl->hClient = gsp->internal.client.object.handle; ctrl->hSysMem = memlist.handle; ctrl->sysmemAddrOfSuspendResumeData = gsp->sr.meta.addr; - ctrl->bEnteringGcoffState = runtime ? 1 : 0; + ctrl->bEnteringGcoffState = 1; ret = nvkm_gsp_rm_ctrl_wr(&gsp->internal.device.subdevice, ctrl); if (ret) @@ -92,7 +92,7 @@ r570_fbsr_init(struct nvkm_gsp *gsp, struct sg_table *sgt, u64 size, bool runtim } static int -r570_fbsr_suspend(struct nvkm_gsp *gsp, bool runtime) +r570_fbsr_suspend(struct nvkm_gsp *gsp) { struct nvkm_subdev *subdev = &gsp->subdev; struct nvkm_device *device = subdev->device; @@ -133,7 +133,7 @@ r570_fbsr_suspend(struct nvkm_gsp *gsp, bool runtime) return ret; /* Initialise FBSR on RM. */ - ret = r570_fbsr_init(gsp, &gsp->sr.fbsr, size, runtime); + ret = r570_fbsr_init(gsp, &gsp->sr.fbsr, size); if (ret) { nvkm_gsp_sg_free(device, &gsp->sr.fbsr); return ret; diff --git a/drivers/gpu/drm/nouveau/nvkm/subdev/gsp/rm/rm.h b/drivers/gpu/drm/nouveau/nvkm/subdev/gsp/rm/rm.h index fcd0221dcea1..e9ac47d86b69 100644 --- a/drivers/gpu/drm/nouveau/nvkm/subdev/gsp/rm/rm.h +++ b/drivers/gpu/drm/nouveau/nvkm/subdev/gsp/rm/rm.h @@ -79,7 +79,7 @@ struct nvkm_rm_api { } *device; const struct nvkm_rm_api_fbsr { - int (*suspend)(struct nvkm_gsp *, bool runtime); + int (*suspend)(struct nvkm_gsp *); void (*resume)(struct nvkm_gsp *); } *fbsr; From 12f6eff11ccf9cad3b2dfcdd94184fdb9ffface2 Mon Sep 17 00:00:00 2001 From: Lyude Paul Date: Fri, 14 Aug 2026 15:43:49 -0400 Subject: [PATCH 118/592] drm/nouveau/gsp/r570: Set GcOff = 0 in fbsr Previously, it looked as if we were able to fix suspend/resume on some desktops by setting Gcoff based on whether or not we were entering runtime PM. This was a mistake though - the only time suspend/resume would end up actually working was if Gcoff = 0. It seems like it's likely the main reason for this is the FBSR GcOff argument actually controls GSP's behavior with regards to which buffers it decides to save across suspend/resume. When GcOff = 1, RM reserved regions are saved unless they are marked as LOST_ON_SUSPEND, and RM channel-context and kernel-client buffers are also saved -including- when they are LOST_ON_SUSPEND. This means with GcOff = 1, we end up having GSP save and restore buffers that actually need to be reinitialized on resume - causing the failures we're setting. Thanks to John Hubbard from Nvidia for providing some background on what these options do in the GSP firmware do! Signed-off-by: Lyude Paul Fixes: 53dac0623853 ("drm/nouveau/gsp: add support for 570.144") Cc: # v6.16+ Reviewed-by: Dave Airlie Link: https://patch.msgid.link/20260814194542.781955-3-lyude@redhat.com (cherry picked from commit c7abe771e013848970421e5ca29c6b2f05c31965) Signed-off-by: Lyude Paul --- drivers/gpu/drm/nouveau/nvkm/subdev/gsp/rm/r570/fbsr.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/drivers/gpu/drm/nouveau/nvkm/subdev/gsp/rm/r570/fbsr.c b/drivers/gpu/drm/nouveau/nvkm/subdev/gsp/rm/r570/fbsr.c index 2945d5b4e570..af5aa5065c3d 100644 --- a/drivers/gpu/drm/nouveau/nvkm/subdev/gsp/rm/r570/fbsr.c +++ b/drivers/gpu/drm/nouveau/nvkm/subdev/gsp/rm/r570/fbsr.c @@ -81,7 +81,7 @@ r570_fbsr_init(struct nvkm_gsp *gsp, struct sg_table *sgt, u64 size) ctrl->hClient = gsp->internal.client.object.handle; ctrl->hSysMem = memlist.handle; ctrl->sysmemAddrOfSuspendResumeData = gsp->sr.meta.addr; - ctrl->bEnteringGcoffState = 1; + ctrl->bEnteringGcoffState = 0; ret = nvkm_gsp_rm_ctrl_wr(&gsp->internal.device.subdevice, ctrl); if (ret) From 24fbd6d4bcf3363ef13ebe0d36dea93f30396c6d Mon Sep 17 00:00:00 2001 From: Lyude Paul Date: Fri, 14 Aug 2026 15:43:50 -0400 Subject: [PATCH 119/592] drm/nouveau/gsp/r570: Enable S/R Display workaround in GSP There's two flags that we've never been setting when asking GSP to suspend the GPU, which OpenRM does set: GPU_STATE_FLAGS_PRESERVING GPU_STATE_FLAGS_PM_TRANSITION These flags aren't -supposed- to do much in GSP, they're mostly used by OpenRM itself for state tracking. The only thing they do from GSP's side is control whether or not a single display related workaround is applied during suspend. But as it turns out, that single workaround is actually quite crucial for getting runtime PM working with nouveau - and without it set we end up seeing a lot more failures with runtime PM resume. So, let's start setting it. Signed-off-by: Lyude Paul Fixes: 53dac0623853 ("drm/nouveau/gsp: add support for 570.144") Cc: # v6.16+ Reviewed-by: Dave Airlie Link: https://patch.msgid.link/20260814194542.781955-4-lyude@redhat.com (cherry picked from commit ca57629b3eb912c77bc4357178a2130ea6c2d6df) Signed-off-by: Lyude Paul --- drivers/gpu/drm/nouveau/nvkm/subdev/gsp/rm/r570/gsp.c | 3 ++- .../gpu/drm/nouveau/nvkm/subdev/gsp/rm/r570/nvrm/gsp.h | 8 ++++++++ 2 files changed, 10 insertions(+), 1 deletion(-) diff --git a/drivers/gpu/drm/nouveau/nvkm/subdev/gsp/rm/r570/gsp.c b/drivers/gpu/drm/nouveau/nvkm/subdev/gsp/rm/r570/gsp.c index 1488771c63fc..b45781cd0dfd 100644 --- a/drivers/gpu/drm/nouveau/nvkm/subdev/gsp/rm/r570/gsp.c +++ b/drivers/gpu/drm/nouveau/nvkm/subdev/gsp/rm/r570/gsp.c @@ -207,7 +207,8 @@ r570_gsp_set_rmargs(struct nvkm_gsp *gsp, bool resume) args->srInitArguments.bInPMTransition = 0; } else { args->srInitArguments.oldLevel = NV2080_CTRL_GPU_SET_POWER_STATE_GPU_LEVEL_3; - args->srInitArguments.flags = 0; + args->srInitArguments.flags = + GPU_STATE_FLAGS_PRESERVING | GPU_STATE_FLAGS_PM_TRANSITION; args->srInitArguments.bInPMTransition = 1; } diff --git a/drivers/gpu/drm/nouveau/nvkm/subdev/gsp/rm/r570/nvrm/gsp.h b/drivers/gpu/drm/nouveau/nvkm/subdev/gsp/rm/r570/nvrm/gsp.h index b6075021e74f..c458569af9d7 100644 --- a/drivers/gpu/drm/nouveau/nvkm/subdev/gsp/rm/r570/nvrm/gsp.h +++ b/drivers/gpu/drm/nouveau/nvkm/subdev/gsp/rm/r570/nvrm/gsp.h @@ -523,6 +523,14 @@ typedef struct #define NV2080_CTRL_GPU_SET_POWER_STATE_GPU_LEVEL_3 (0x00000003U) +#define GPU_STATE_FLAGS_PRESERVING BIT(0) // GPU state is preserved +#define GPU_STATE_FLAGS_VGA_TRANSITION BIT(1) // To be used with GPU_STATE_FLAGS_PRESERVING. +#define GPU_STATE_FLAGS_PM_TRANSITION BIT(2) // To be used with GPU_STATE_FLAGS_PRESERVING. +#define GPU_STATE_FLAGS_PM_SUSPEND BIT(3) +#define GPU_STATE_FLAGS_PM_HIBERNATE BIT(4) +#define GPU_STATE_FLAGS_GC6_TRANSITION BIT(5) // To be used with GPU_STATE_FLAGS_PRESERVING. +#define GPU_STATE_FLAGS_FAST_UNLOAD BIT(6) // Used during windows restart, skips stateDestroy steps + typedef struct { // Magic for verification by secure ucode From bbb9293c9bb792f3f16c842f223b8c97bdbaf227 Mon Sep 17 00:00:00 2001 From: Lyude Paul Date: Fri, 14 Aug 2026 15:43:51 -0400 Subject: [PATCH 120/592] drm/nouveau/gsp: Increase delay for magic sleep in r535_gsp_fini() As it turns out, Turing isn't the only architecture that needs this. On this Dell Precision 7780 with an AD103 GPU, along with pretty much every other laptop I tested, runtime PM is still somewhat unreliable. At first glance it seems as if it's fixed, but lowering the autosuspend delay to 500ms and then doing a stress test of suspend/resume cycles on the GPU ends up causing everything to start timing out. After quite a lot of digging, I eventually landed back on this magic timeout in r535_gsp_fini(). As it turns out, increasing the timeout ends up fixing the runtime PM issues as far as I can tell, even during intense stress testing. Unfortunately after spending quite a bit of time trying to dig through OpenRM to figure out what this magic sleep is actually doing, I've also come up short with any reasonable explanation. In lieu of that, I'm going to include the observations I did make while trying to figure this out in hopes someone eventually does figure this out: * The magic sleep has to occur after fbsr is initialized. Performing it at any time before that doesn't appear to work. * In situations where runtime PM starts getting flaky, some rather interesting visual effects end up happening on occasion before the GPU fully falls over. In particular, squares that look like the result of an incomplete blitting operation to a tiled buffer end up showing up on applications like vkcube. Interestingly enough, they remain in precisely the same place between runtime PM cycles until the GPU falls over - even when restarting vkcube multiple times, and even when vkcube is actively updating the screen. Even more interestingly, they're not limited to a specific framebuffer - you can see the squares changing as the cube rotates around. We cannot however, say that this is likely to be a incomplete fbsr operation. The magic sleep happens before fbsr is actually saved (which happens on the GSP unload), so it's something else. * During a short bit of testing with a desktop that I have, the magic sleep seemed to make no difference to whether or not suspend/resume works. It seems to generally work almost always. So we can assume this is likely exclusive to runtime PM, not S3. As well, here's a list of the things I tried before settling on the magic sleep: * Hooking up NV2080_CTRL_CMD_INTERNAL_GCX_ENTRY_PREREQUISITE and then blocking runtime PM until OpenRM signals that GC6/GCOFF is ready appears to make no difference. * Hooking up some (maybe not all, unsure about that part) bits of comptag saving including: * Fetching static memsys information from GSP * Adding the size of the comptag storage to the fbsr data * Adding a GA103+ workaround for disabling raw compression mode during fbsr (it doesn't seem like it applies for any systems I tried it on anyhow) * Setting bPreserveVideoMemoryAllocations=1 in GspSystemInfo So, until we can figure this out properly - just sleep for longer. Signed-off-by: Lyude Paul Fixes: 53dac0623853 ("drm/nouveau/gsp: add support for 570.144") Cc: # v6.16+ Reviewed-by: Dave Airlie Link: https://patch.msgid.link/20260814194542.781955-5-lyude@redhat.com (cherry picked from commit 09b47186a4164f3aaa3591313f80794443117342) Signed-off-by: Lyude Paul --- drivers/gpu/drm/nouveau/nvkm/subdev/gsp/rm/r535/gsp.c | 6 +++++- 1 file changed, 5 insertions(+), 1 deletion(-) diff --git a/drivers/gpu/drm/nouveau/nvkm/subdev/gsp/rm/r535/gsp.c b/drivers/gpu/drm/nouveau/nvkm/subdev/gsp/rm/r535/gsp.c index 4a3b771ded25..94925f1590ea 100644 --- a/drivers/gpu/drm/nouveau/nvkm/subdev/gsp/rm/r535/gsp.c +++ b/drivers/gpu/drm/nouveau/nvkm/subdev/gsp/rm/r535/gsp.c @@ -1761,8 +1761,12 @@ r535_gsp_fini(struct nvkm_gsp *gsp, enum nvkm_suspend_state suspend) * TODO: Debug the GSP firmware / RPC handling to find out why * without this Turing (but none of the other architectures) * ends up resetting all channels after resume. + * Additionally, runtime suspend on other architectures quickly + * becomes unreliable without this sleep. If you're experiencing + * issues with runtime suspend, try bumping this delay up and + * sending a patch if it fixes your GPU. */ - msleep(50); + msleep(200); } ret = r535_gsp_rpc_unloading_guest_driver(gsp, suspend); From 23c240d9509e15f72e4112fc95f0160ab32ec430 Mon Sep 17 00:00:00 2001 From: =?UTF-8?q?J=C3=A9r=C3=A9my=20Jean?= Date: Thu, 10 Sep 2026 18:37:43 +0000 Subject: [PATCH 121/592] smb: client: validate absolute native symlink targets before NT fixups MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit With symlinkroot unset, an absolute target is copied without conversion to an NT drive path. Later code still assumes an NT prefix is present when modifying the target and calculating the print name length. For "/ab", this causes two failures: sym[5] and path[5] are written past their allocations, and plen -= 2 * poff subtracts an assumed 8-byte prefix from a 6-byte UTF-16 target, wrapping u16 plen to 65534. That underflow causes another overflow: memcpy() copies 65534 bytes into a 24-byte buffer. A user with write access to a mounted share can trigger these bugs with default settings. Validate the NT drive prefix, including an ASCII drive letter, before accessing fixed offsets or subtracting the prefix length. Fixes: 3363da82e02f ("smb: client: fix native SMB symlink traversal") Assisted-by: Codex:gpt-5 Signed-off-by: Jérémy Jean Reviewed-by: Namjae Jeon Signed-off-by: Paulo Alcantara --- fs/smb/client/reparse.c | 24 +++++++++++++++++------- 1 file changed, 17 insertions(+), 7 deletions(-) diff --git a/fs/smb/client/reparse.c b/fs/smb/client/reparse.c index 8a1b9e8be5ba..9e31fce7e0a5 100644 --- a/fs/smb/client/reparse.c +++ b/fs/smb/client/reparse.c @@ -3,6 +3,7 @@ * Copyright (c) 2024 Paulo Alcantara */ +#include #include #include #include @@ -159,15 +160,24 @@ static int create_native_symlink(const unsigned int xid, struct inode *inode, convert_delimiter(sym, sep); /* - * For absolute NT symlinks it is required to pass also leading - * backslash and to not mangle NT object prefix "\\??\\" and not to - * mangle colon in drive letter. But cifs_convert_path_to_utf16() - * removes leading backslash and replaces '?' and ':'. So temporary - * mask these characters in NT object prefix by '_' and then change - * them back. + * Absolute NT symlinks must retain the leading backslash, "\\??\\" + * prefix and drive-letter colon. cifs_convert_path_to_utf16() strips + * the leading backslash and maps '?' and ':', so temporarily mask + * these characters with '_' and restore them after conversion. + * + * When symlinkroot is unset, sym comes directly from the caller. + * Validate the complete "\\??\\X:" prefix before using fixed offsets + * or subtracting the NT prefix length below. Require an ASCII drive + * letter so the prefix occupies six characters in UTF-16 too. */ - if (!(sbflags & CIFS_MOUNT_POSIX_PATHS) && symname[0] == '/') + if (!(sbflags & CIFS_MOUNT_POSIX_PATHS) && symname[0] == '/') { + if (!strstarts(sym, "\\??\\") || !isascii(sym[4]) || + !isalpha(sym[4]) || sym[5] != ':') { + rc = -EINVAL; + goto out; + } sym[0] = sym[1] = sym[2] = sym[5] = '_'; + } /* * On a POSIX paths mount the symlink target is stored verbatim, so From 3163cd7253432f262c2fa22edb51474d7afc76a0 Mon Sep 17 00:00:00 2001 From: George Emmanuel Thomas Date: Sat, 15 Aug 2026 22:13:35 +0530 Subject: [PATCH 122/592] drm/msm: remove stale perf counter XML TODO The adreno_perfcntrs macro already takes the XML file stem as its second argument, allowing each perf counter JSON file to select the appropriate register XML. The a2xx and a5xx entries already pass their respective XML file stems. Remove the stale TODO comment. Signed-off-by: George Emmanuel Thomas Patchwork: https://patchwork.freedesktop.org/patch/746745/ Message-ID: <20260815164335.158958-1-georgeemmanuelthomas@gmail.com> Signed-off-by: Rob Clark --- drivers/gpu/drm/msm/Makefile | 1 - 1 file changed, 1 deletion(-) diff --git a/drivers/gpu/drm/msm/Makefile b/drivers/gpu/drm/msm/Makefile index d0c3a4c6703b..0b8f2cafed19 100644 --- a/drivers/gpu/drm/msm/Makefile +++ b/drivers/gpu/drm/msm/Makefile @@ -177,7 +177,6 @@ quiet_cmd_headergen = GENHDR $@ cmd_headergen = mkdir -p $(obj)/generated && $(PYTHON3) $(src)/registers/gen_header.py \ $(headergen-opts) --rnn $(src)/registers --xml $< c-defines > $@ -# TODO how to do this for a2xx/a5xx which have different .xml arg? quiet_cmd_headergen_json = GENHDRJSN $@ cmd_headergen_json = mkdir -p $(obj)/generated && $(PYTHON3) $(src)/registers/gen_header.py \ $(headergen-opts) --rnn $(src)/registers --xml $(filter %.xml,$^) perfcntrs --json $< > $@ From ed9ad3d4183053d4a8c43960a3d23ea8db71c30d Mon Sep 17 00:00:00 2001 From: Neil Armstrong Date: Wed, 9 Sep 2026 15:07:37 +0200 Subject: [PATCH 123/592] drm/msm/a6xx: Add CX AO Counter registers used for a750 GPUs The a750 uses the CX AO Counters instead of the GMU_ALWAYS_ON_COUNTER register on A6xx and other A7xx GPUs. Signed-off-by: Neil Armstrong Patchwork: https://patchwork.freedesktop.org/patch/752234/ Message-ID: <20260909-topic-sm8650-gmu-a750-timestamp-reg-v2-1-091d74958951@linaro.org> Signed-off-by: Rob Clark --- drivers/gpu/drm/msm/registers/adreno/a6xx_gmu.xml | 2 ++ 1 file changed, 2 insertions(+) diff --git a/drivers/gpu/drm/msm/registers/adreno/a6xx_gmu.xml b/drivers/gpu/drm/msm/registers/adreno/a6xx_gmu.xml index 33404eb18fd0..b3082738c99d 100644 --- a/drivers/gpu/drm/msm/registers/adreno/a6xx_gmu.xml +++ b/drivers/gpu/drm/msm/registers/adreno/a6xx_gmu.xml @@ -141,6 +141,8 @@ xsi:schemaLocation="https://gitlab.freedesktop.org/freedreno/ rules-fd.xsd"> + + From a2b65837980a513cf24825bc4a50b0430e3391e7 Mon Sep 17 00:00:00 2001 From: Neil Armstrong Date: Wed, 9 Sep 2026 15:07:38 +0200 Subject: [PATCH 124/592] drm/msm/a6xx: Use CX AO Counter register for timestamp on a750 GPUs The a750 uses the GMU CX AO Counters instead of the GMU_ALWAYS_ON_COUNTER register on A6xx and other A7xx GPUs, use it when running a A750 GPU. The GMU_ALWAYS_ON_COUNTER at offset 0x1f888 doesn't seem to exist on the SM8650 A750 GMU and returns 0, but the CX AO counter at offset 0x1f880 returns some proper timestamp data. Signed-off-by: Neil Armstrong Patchwork: https://patchwork.freedesktop.org/patch/752235/ Message-ID: <20260909-topic-sm8650-gmu-a750-timestamp-reg-v2-2-091d74958951@linaro.org> Signed-off-by: Rob Clark --- drivers/gpu/drm/msm/adreno/a6xx_gpu.c | 12 +++++++++--- 1 file changed, 9 insertions(+), 3 deletions(-) diff --git a/drivers/gpu/drm/msm/adreno/a6xx_gpu.c b/drivers/gpu/drm/msm/adreno/a6xx_gpu.c index f9de9329dee3..081e79ea4652 100644 --- a/drivers/gpu/drm/msm/adreno/a6xx_gpu.c +++ b/drivers/gpu/drm/msm/adreno/a6xx_gpu.c @@ -23,9 +23,15 @@ static u64 a6xx_gmu_get_timestamp(struct msm_gpu *gpu) u64 count_hi, count_lo, temp; do { - count_hi = gmu_read(&a6xx_gpu->gmu, REG_A6XX_GMU_ALWAYS_ON_COUNTER_H); - count_lo = gmu_read(&a6xx_gpu->gmu, REG_A6XX_GMU_ALWAYS_ON_COUNTER_L); - temp = gmu_read(&a6xx_gpu->gmu, REG_A6XX_GMU_ALWAYS_ON_COUNTER_H); + if (adreno_is_a750_family(adreno_gpu)) { + count_hi = gmu_read(&a6xx_gpu->gmu, REG_A7XX_GMU_CX_AO_COUNTER_H); + count_lo = gmu_read(&a6xx_gpu->gmu, REG_A7XX_GMU_CX_AO_COUNTER_L); + temp = gmu_read(&a6xx_gpu->gmu, REG_A7XX_GMU_CX_AO_COUNTER_H); + } else { + count_hi = gmu_read(&a6xx_gpu->gmu, REG_A6XX_GMU_ALWAYS_ON_COUNTER_H); + count_lo = gmu_read(&a6xx_gpu->gmu, REG_A6XX_GMU_ALWAYS_ON_COUNTER_L); + temp = gmu_read(&a6xx_gpu->gmu, REG_A6XX_GMU_ALWAYS_ON_COUNTER_H); + } } while (unlikely(count_hi != temp)); return (count_hi << 32) | count_lo; From ebc5660132ddd244b57f03ed324922013a3d7363 Mon Sep 17 00:00:00 2001 From: April Cardenas Date: Thu, 10 Sep 2026 00:16:49 -0500 Subject: [PATCH 125/592] smb/client: send lease break ACKs thru correct session for multiuser mounts Currently, when cifs_oplock_break handles a break request from the server it searches for the appropriate tlink to handle the request but incorrectly uses the current fsuid as the search key, eventually causing read errors for users with multiuser mounts on NetApp. Fix this by using the tlink from the cfile struct instead to respond through the correct session. As breaks are handled in a worker thread, the current fsuid isn't guaranteed to match the session that the break is intended for. This means that cifs_sb_tlink may search the rbtree using the wrong fsuid, and return a tlink with an incorrect session than the lease break was intended for. As a result, the breaks may be ACKed through an incorrect session. While it seems that Samba/Windows Servers 2016-2025 ignore this as long as the lease key is correct, we ran into a case where if you're using NetApp ONTAP or Azure NetApp Files they will reject the ACK and return `STATUS_LOCK_NOT_GRANTED` errors on any future read requests a user may initiate through their still held open file handle, and the server will eventually close the file. In the dmesg logs, the user may see errors like these: CIFS: Status code returned 0xc0000128 STATUS_FILE_CLOSED CIFS: VFS: Send error in read = -9 With a multiuser mount using NetApp, this issue is really easy for users to hit on a wide variety of kernel versions by attempting to copy a file from the share to the local machine through GNOME Files/Nautilus. This copy will always result in Nautilus throwing a `Bad File Descriptor` error to the user and fail. With this fix, you can copy files through Nautilus without issue. >From looking at the traces, it seems that glib will open the file first, and call listxattr before actually attempting to copy the file data. The listxattr call always triggers a break, causing the copy to fail. The proposed fix returns to the way the client grabbed the tlink before commit e8f5f849ffce2 ("cifs: fix potential oops in cifs_oplock_break"). The bulk of that commit (checking for list empty) remains untouched, and I think the change to using cifs_sb_tlink was intended to avoid a NULL/ERR deference on the tlink as well as update the reference count. I believe this fix should preserve those safety properties, but of course I'd appreciate any corrections here. Fixes: e8f5f849ffce2 ("cifs: fix potential oops in cifs_oplock_break") Cc: stable@vger.kernel.org Signed-off-by: April Cardenas Reviewed-by: Namjae Jeon Reviewed-by: Bharath S M Signed-off-by: Paulo Alcantara --- fs/smb/client/file.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/fs/smb/client/file.c b/fs/smb/client/file.c index 1aa4844f8b8a..0d428517f454 100644 --- a/fs/smb/client/file.c +++ b/fs/smb/client/file.c @@ -3354,8 +3354,8 @@ void cifs_oplock_break(struct work_struct *work) wait_on_bit(&cinode->flags, CIFS_INODE_PENDING_WRITERS, TASK_UNINTERRUPTIBLE); - tlink = cifs_sb_tlink(cifs_sb); - if (IS_ERR(tlink)) { + tlink = cifs_get_tlink(cfile->tlink); + if (IS_ERR_OR_NULL(tlink)) { /* drop the reference taken when the break was queued */ _cifsFileInfo_put(cfile, false /* do not wait for ourself */, false); goto out; From 2725ab3f5ad1c5f375c7c9fee4af02a9b138f701 Mon Sep 17 00:00:00 2001 From: Karl Mehltretter Date: Fri, 21 Aug 2026 04:53:27 +0200 Subject: [PATCH 126/592] keys: fix lost wakeup when reaping a dead key type clear_bit() is atomic with respect to the word it modifies, but it is an unordered operation: it implies no memory barrier on either side (Documentation/atomic_bitops.txt). key_garbage_collector() clears KEY_GC_REAPING_KEYTYPE with clear_bit() and calls wake_up_bit() after reaping a dead key type. wake_up_bit() uses a lockless waitqueue check and requires a full barrier after the clear. The existing smp_mb() is before clear_bit(), so nothing orders the clear against that check. The GC can see an empty waitqueue while unregister_key_type() still sees the bit set. The final wakeup is then lost, leaving module unload stuck in wait_on_bit(). Use clear_and_wake_up_bit(). Its clear_bit_unlock() has RELEASE semantics, so the completed GC work stays ordered before the clear, and its smp_mb__after_atomic() orders the clear before the waitqueue check. Fixes: 0c061b5707ab ("KEYS: Correctly destroy key payloads when their keytype is removed") Assisted-by: Claude:claude-fable-5 Signed-off-by: Karl Mehltretter Link: https://lore.kernel.org/r/20260821025327.61488-1-kmehltretter@gmail.com Reviewed-by: Jarkko Sakkinen Signed-off-by: Jarkko Sakkinen --- security/keys/gc.c | 4 +--- 1 file changed, 1 insertion(+), 3 deletions(-) diff --git a/security/keys/gc.c b/security/keys/gc.c index 748e83818a76..eda445f815d4 100644 --- a/security/keys/gc.c +++ b/security/keys/gc.c @@ -318,9 +318,7 @@ static void key_garbage_collector(struct work_struct *work) if (unlikely(gc_state & KEY_GC_REAPING_DEAD_3)) { kdebug("dead wake"); - smp_mb(); - clear_bit(KEY_GC_REAPING_KEYTYPE, &key_gc_flags); - wake_up_bit(&key_gc_flags, KEY_GC_REAPING_KEYTYPE); + clear_and_wake_up_bit(KEY_GC_REAPING_KEYTYPE, &key_gc_flags); } if (gc_state & KEY_GC_REAP_AGAIN) From 0d6a4268b06084baafd8ee5d66955c7e1c2e053b Mon Sep 17 00:00:00 2001 From: Maoyi Xie Date: Fri, 21 Aug 2026 17:59:35 +0800 Subject: [PATCH 127/592] keys: translate request_key_auth pid for the reading procfs instance request_key_auth_describe() prints rka->pid into /proc/keys as a raw pid_t in the initial pid namespace. A reader can open /proc/keys through a mount in another pid namespace. That reader sees a number with no meaning there. The number can even name an unrelated task. The line needs VIEW on the key. So the reader either shares the key owner's uid or possesses the key. The fix keeps a struct pid. Commit 4f82f45730c6 ("net ip6 flowlabel: Make owner a union of struct pid * and kuid_t") gave /proc/net/ip6_flowlabel the same storage. The print goes through pid_nr_ns(). It renders against the pid namespace of the procfs instance the line is read through. Commit ad08978ab41c ("ipv6/flowlabel: simplify pid namespace lookup") moved that print to the same anchor. Output through an initial namespace /proc does not change. The line shows 0 for a requestor with no number in that namespace. Translating at read time was the alternative. find_pid_ns() can resolve a recycled number. The line would then name a live task with no connection to the key. A stored struct pid gives 0 instead when the requestor has no number there. Link: https://lore.kernel.org/keyrings/20260809110202.2180410-1-maoyixie.tju@gmail.com/ Fixes: 78b7280cce23 ("KEYS: Improve /proc/keys") Cc: stable@vger.kernel.org # v5.10+ Assisted-by: Claude:claude-opus-5 codeql Signed-off-by: Maoyi Xie Link: https://lore.kernel.org/r/20260821095935.1864998-1-maoyixie.tju@gmail.com Reviewed-by: Jarkko Sakkinen Signed-off-by: Jarkko Sakkinen --- include/keys/request_key_auth-type.h | 2 +- security/keys/request_key_auth.c | 12 +++++++++--- 2 files changed, 10 insertions(+), 4 deletions(-) diff --git a/include/keys/request_key_auth-type.h b/include/keys/request_key_auth-type.h index 01e42ee5f409..464636278c4f 100644 --- a/include/keys/request_key_auth-type.h +++ b/include/keys/request_key_auth-type.h @@ -22,7 +22,7 @@ struct request_key_auth { const struct cred *cred; void *callout_info; size_t callout_len; - pid_t pid; + struct pid *pid; char op[8]; } __randomize_layout; diff --git a/security/keys/request_key_auth.c b/security/keys/request_key_auth.c index 282e09d8fa46..ed6f55b9cdd9 100644 --- a/security/keys/request_key_auth.c +++ b/security/keys/request_key_auth.c @@ -9,6 +9,8 @@ #include #include +#include +#include #include #include #include @@ -73,7 +75,10 @@ static void request_key_auth_describe(const struct key *key, seq_puts(m, "key:"); seq_puts(m, key->description); if (key_is_positive(key)) - seq_printf(m, " pid:%d ci:%zu", rka->pid, rka->callout_len); + seq_printf(m, " pid:%d ci:%zu", + pid_nr_ns(rka->pid, + proc_pid_ns(file_inode(m->file)->i_sb)), + rka->callout_len); } /* @@ -113,6 +118,7 @@ static void free_request_key_auth(struct request_key_auth *rka) if (rka->cred) put_cred(rka->cred); kfree(rka->callout_info); + put_pid(rka->pid); kfree(rka); } @@ -226,14 +232,14 @@ struct key *request_key_auth_new(struct key *target, const char *op, irka = cred->request_key_auth->payload.data[0]; rka->cred = get_cred(irka->cred); - rka->pid = irka->pid; + rka->pid = get_pid(irka->pid); up_read(&cred->request_key_auth->sem); } else { /* it isn't - use this process as the context */ rka->cred = get_cred(cred); - rka->pid = current->pid; + rka->pid = get_pid(task_pid(current)); } rka->target_key = key_get(target); From 114f00d738f15dd8c7318369edcdc53dd6d08763 Mon Sep 17 00:00:00 2001 From: Jarkko Sakkinen Date: Tue, 1 Sep 2026 23:58:06 +0300 Subject: [PATCH 128/592] KEYS: trusted: Fix tpm2_load_cmd() boundary check tpm2_load_cmd() does boundary checks against the ASN.1 size i.e., payload->blob_len. Address this by passing the decoded blob size to tpm2_load_cmd(), and use it for the boundary checks. Cc: stable@vger.kernel.org # v5.13+ Fixes: f2219745250f ("security: keys: trusted: use ASN.1 TPM2 key format for the blobs") Reported-by: co+6a581c4284f721d4@bugs.sh Closes: https://bugs.sh/b/6a581c4284f721d4/ Reviewed-by: Stefano Garzarella Tested-by: Srish Srinivasan Link: https://lore.kernel.org/r/20260901205809.2028454-1-jarkko@kernel.org Signed-off-by: Jarkko Sakkinen --- security/keys/trusted-keys/trusted_tpm2.c | 12 +++++++----- 1 file changed, 7 insertions(+), 5 deletions(-) diff --git a/security/keys/trusted-keys/trusted_tpm2.c b/security/keys/trusted-keys/trusted_tpm2.c index 67225dd562a9..01f18bb37047 100644 --- a/security/keys/trusted-keys/trusted_tpm2.c +++ b/security/keys/trusted-keys/trusted_tpm2.c @@ -99,7 +99,7 @@ struct tpm2_key_context { static int tpm2_key_decode(struct trusted_key_payload *payload, struct trusted_key_options *options, - u8 **buf) + u8 **buf, unsigned int *blob_len) { int ret; struct tpm2_key_context ctx; @@ -120,6 +120,7 @@ static int tpm2_key_decode(struct trusted_key_payload *payload, return -ENOMEM; *buf = blob; + *blob_len = ctx.priv_len + ctx.pub_len; options->keyhandle = ctx.parent; memcpy(blob, ctx.priv, ctx.priv_len); @@ -384,10 +385,11 @@ static int tpm2_load_cmd(struct tpm_chip *chip, int rc; u32 attrs; - rc = tpm2_key_decode(payload, options, &blob); + rc = tpm2_key_decode(payload, options, &blob, &blob_len); if (rc) { /* old form */ blob = payload->blob; + blob_len = payload->blob_len; payload->old_format = 1; } else { /* Bind for cleanup: */ @@ -399,17 +401,17 @@ static int tpm2_load_cmd(struct tpm_chip *chip, return -EINVAL; /* must be big enough for at least the two be16 size counts */ - if (payload->blob_len < 4) + if (blob_len < 4) return -EINVAL; private_len = get_unaligned_be16(blob); /* must be big enough for following public_len */ - if (private_len + 2 + 2 > (payload->blob_len)) + if (private_len + 2 + 2 > blob_len) return -E2BIG; public_len = get_unaligned_be16(blob + 2 + private_len); - if (private_len + 2 + public_len + 2 > payload->blob_len) + if (private_len + 2 + public_len + 2 > blob_len) return -E2BIG; pub = blob + 2 + private_len + 2; From 8697c431e297eb0d0ab13dda6bc172b48a34f05c Mon Sep 17 00:00:00 2001 From: Cen Zhang Date: Sat, 12 Sep 2026 00:31:16 +0300 Subject: [PATCH 129/592] KEYS: encrypted: fix integer overflow of datablob_len encrypted_key_alloc() stores datablob_len in a u16. It is computed from multiple string and payload lengths. If the result exceeds U16_MAX, the assignment truncates the allocation size. KASAN reports a 32760-byte slab-out-of-bounds write when __ekey_init() copies the master key description into the undersized buffer. The total payload length stored in key->datalen is also a u16. Use check_add_overflow() to reject values that do not fit either destination, and use kzalloc_flex() for the flexible-array allocation. Fixes: 7e70cb497850 ("keys: add new key-type encrypted") Cc: stable@vger.kernel.org Assisted-by: GitHub-Copilot:claude-opus-4.6 Signed-off-by: Cen Zhang Signed-off-by: Francis Perron Reviewed-by: Jarkko Sakkinen Tested-by: R Nageswara Sastry Link: https://lore.kernel.org/r/20260909153433.83117-1-cenzhang@linux.microsoft.com Signed-off-by: Jarkko Sakkinen --- security/keys/encrypted-keys/encrypted.c | 20 ++++++++++++++------ 1 file changed, 14 insertions(+), 6 deletions(-) diff --git a/security/keys/encrypted-keys/encrypted.c b/security/keys/encrypted-keys/encrypted.c index 59cb77b237b3..e07092ea301a 100644 --- a/security/keys/encrypted-keys/encrypted.c +++ b/security/keys/encrypted-keys/encrypted.c @@ -19,6 +19,7 @@ #include #include #include +#include #include #include #include @@ -579,6 +580,7 @@ static struct encrypted_key_payload *encrypted_key_alloc(struct key *key, { struct encrypted_key_payload *epayload = NULL; unsigned short datablob_len; + unsigned short payload_totallen; unsigned short decrypted_datalen; unsigned short payload_datalen; unsigned int encrypted_datalen; @@ -632,16 +634,22 @@ static struct encrypted_key_payload *encrypted_key_alloc(struct key *key, encrypted_datalen = roundup(decrypted_datalen, blksize); - datablob_len = format_len + 1 + strlen(master_desc) + 1 - + strlen(datalen) + 1 + ivsize + 1 + encrypted_datalen; + if (check_add_overflow(format_len + 1 + strlen(master_desc) + 1 + + strlen(datalen) + 1 + ivsize + 1, + encrypted_datalen, &datablob_len)) + return ERR_PTR(-EINVAL); - ret = key_payload_reserve(key, payload_datalen + datablob_len - + HASH_SIZE + 1); + if (check_add_overflow(datablob_len, + payload_datalen + HASH_SIZE + 1, + &payload_totallen)) + return ERR_PTR(-EINVAL); + + ret = key_payload_reserve(key, payload_totallen); if (ret < 0) return ERR_PTR(ret); - epayload = kzalloc(sizeof(*epayload) + payload_datalen + - datablob_len + HASH_SIZE + 1, GFP_KERNEL); + epayload = kzalloc_flex(*epayload, payload_data, payload_totallen, + GFP_KERNEL); if (!epayload) return ERR_PTR(-ENOMEM); From effce1cb87ee0d8b3a8cbe7722968f4ea7efd360 Mon Sep 17 00:00:00 2001 From: Sophie D Date: Wed, 9 Sep 2026 21:49:10 -0400 Subject: [PATCH 130/592] drm/gud: Ignore damage clips in full update mode When running in full update mode, previously small updates (such as moving the mouse across the screen) would cause many full frames to be generated. This would bog down the bus and lower the effective framerate significantly - I was seeing a drop from 60 FPS to 2 FPS. Set ignore_damage_clips in full update mode so the damage iterator yields a single full-plane rectangle instead of one per clip. Fixes: 73cfd166e045 ("drm/gud: Replace simple display pipe with DRM atomic helpers") Cc: # 6.18.x Signed-off-by: Sophie D Reviewed-by: Thomas Zimmermann Acked-by: Ruben Wauters Signed-off-by: Ruben Wauters Link: https://patch.msgid.link/20260910014910.8564-1-patches@scd31.com --- drivers/gpu/drm/gud/gud_pipe.c | 3 +++ 1 file changed, 3 insertions(+) diff --git a/drivers/gpu/drm/gud/gud_pipe.c b/drivers/gpu/drm/gud/gud_pipe.c index 5ef887d8485a..3388fdc8ea7b 100644 --- a/drivers/gpu/drm/gud/gud_pipe.c +++ b/drivers/gpu/drm/gud/gud_pipe.c @@ -482,6 +482,9 @@ int gud_plane_atomic_check(struct drm_plane *plane, if (!new_plane_state->visible) return 0; + if (gdrm->flags & GUD_DISPLAY_FLAG_FULL_UPDATE) + new_plane_state->ignore_damage_clips = true; + if (old_plane_state->rotation != new_plane_state->rotation) crtc_state->mode_changed = true; From 59ced288fcba9e91bd38e61a972ad782c4edb7d0 Mon Sep 17 00:00:00 2001 From: Sajal Gupta Date: Wed, 2 Sep 2026 18:00:57 +0530 Subject: [PATCH 131/592] drm/gud: fix out-of-bounds write in gud_plane_atomic_check() MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit The plane property loop uses req->properties[num_properties + i] as write index while simultaneously incrementing `num_properties` inside the loop. At iteration i, num_properties has also incremented by i, so the write is done at `initial_num_properties + 2*i`, skipping every other index and advancing by 2 per iteration. With just 2 connector and 32 plane properties the last write happens at index 64, one slot past the end of the 64-slot (indices 0–63) allocation. A USB device can trigger OOB by advertising the maximum number of properties. Fix by dropping the redundant `+ i`; num_properties is already the correct running index, as gud_connector_fill_properties() fills the preceding slots. Fixes: 40e1a70b4aed ("drm: Add GUD USB Display driver") Reported-by: Sashiko Link: https://sashiko.dev/#/patchset/20260821071812.16500-1-sajal2005gupta%40gmail.com?part=1 Signed-off-by: Sajal Gupta Cc: Acked-by: Ruben Wauters Signed-off-by: Ruben Wauters Link: https://patch.msgid.link/20260902123254.36987-1-sajal2005gupta@gmail.com --- drivers/gpu/drm/gud/gud_pipe.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/drivers/gpu/drm/gud/gud_pipe.c b/drivers/gpu/drm/gud/gud_pipe.c index 3388fdc8ea7b..aa7792966287 100644 --- a/drivers/gpu/drm/gud/gud_pipe.c +++ b/drivers/gpu/drm/gud/gud_pipe.c @@ -565,8 +565,8 @@ int gud_plane_atomic_check(struct drm_plane *plane, goto out; } - req->properties[num_properties + i].prop = cpu_to_le16(prop); - req->properties[num_properties + i].val = cpu_to_le64(val); + req->properties[num_properties].prop = cpu_to_le16(prop); + req->properties[num_properties].val = cpu_to_le64(val); num_properties++; } From ba970587a0e1203b6a0934b5b9174886b4c4d24c Mon Sep 17 00:00:00 2001 From: Rob Clark Date: Sat, 12 Sep 2026 08:09:14 -0700 Subject: [PATCH 132/592] drm/msm/a6xx+: Increase GMU FW init timeout We were using 10ms, kgsl uses 100ms. In practice it is usually takes less than 10ms, but very occasionally goes a bit above 10ms, leading to a "GMU firmware inialization timed out", from which point things go south. There are probably some things that could be done to speed up init, like increasing GMU freq. But to be safe, increase the timeout to match kgsl. Signed-off-by: Rob Clark Reviewed-by: Akhil P Oommen Patchwork: https://patchwork.freedesktop.org/patch/753014/ Message-ID: <20260912150915.28700-1-robin.clark@oss.qualcomm.com> --- drivers/gpu/drm/msm/adreno/a6xx_gmu.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/drivers/gpu/drm/msm/adreno/a6xx_gmu.c b/drivers/gpu/drm/msm/adreno/a6xx_gmu.c index 27cac853975f..6d49c51df1a2 100644 --- a/drivers/gpu/drm/msm/adreno/a6xx_gmu.c +++ b/drivers/gpu/drm/msm/adreno/a6xx_gmu.c @@ -320,7 +320,7 @@ static int a6xx_gmu_start(struct a6xx_gmu *gmu) gmu_write(gmu, REG_A6XX_GMU_CM3_SYSRESET, 0); ret = gmu_poll_timeout(gmu, REG_A6XX_GMU_CM3_FW_INIT_RESULT, val, - (val & mask) == reset_val, 100, 10000); + (val & mask) == reset_val, 100, 100000); if (ret) DRM_DEV_ERROR(gmu->dev, "GMU firmware initialization timed out\n"); From 0d7823cd4cda35f2060a685fa276ff7709915abc Mon Sep 17 00:00:00 2001 From: Siddharth Chintamaneni Date: Wed, 2 Sep 2026 17:14:13 +0000 Subject: [PATCH 133/592] bpf: Allow terminal gotox instructions check_subprogs() treats gotox as a direct jump and validates its reserved zero offset. When gotox is the final instruction, this produces a synthetic successor one instruction past the end of the subprogram and rejects an otherwise valid program. Skip direct-offset validation for gotox and accept it as a non-fallthrough terminal instruction. Its actual targets remain validated from the instruction-array jump table during CFG construction. Fixes: 493d9e0d6083 ("bpf, x86: add support for indirect jumps") Signed-off-by: Siddharth Chintamaneni Reviewed-by: Anton Protopopov Link: https://lore.kernel.org/r/20260902171414.96165-1-sidchintamaneni@gmail.com Signed-off-by: Alexei Starovoitov --- kernel/bpf/verifier.c | 5 ++++- 1 file changed, 4 insertions(+), 1 deletion(-) diff --git a/kernel/bpf/verifier.c b/kernel/bpf/verifier.c index 1b0b1fb62878..ddba53eaa333 100644 --- a/kernel/bpf/verifier.c +++ b/kernel/bpf/verifier.c @@ -3042,6 +3042,8 @@ static int check_subprogs(struct bpf_verifier_env *env) subprog[cur_subprog].exit_idx = i; goto next; } + if (insn_is_gotox(&insn[i])) + goto next; off = i + bpf_jmp_offset(&insn[i]) + 1; if (off < subprog_start || off >= subprog_end) { verbose(env, "jump out of range from insn %d to %d\n", i, off); @@ -3061,7 +3063,8 @@ static int check_subprogs(struct bpf_verifier_env *env) */ if (code != (BPF_JMP | BPF_EXIT) && code != (BPF_JMP32 | BPF_JA) && - code != (BPF_JMP | BPF_JA)) { + code != (BPF_JMP | BPF_JA) && + !insn_is_gotox(&insn[i])) { verbose(env, "last insn is not an exit or jmp\n"); bpf_diag_program_structure( env, i, "subprogram can fall through", From ac781acaef4904b9be64a7b5755c778e5d76ede6 Mon Sep 17 00:00:00 2001 From: Siddharth Chintamaneni Date: Wed, 2 Sep 2026 17:14:14 +0000 Subject: [PATCH 134/592] selftests/bpf: Test terminal gotox instructions Add tests that place gotox at the end of the main program and a subprogram, with each jump-table target preceding the gotox instruction. This tests gotox as a valid non-fallthrough terminal instruction. Signed-off-by: Siddharth Chintamaneni Reviewed-by: Anton Protopopov Link: https://lore.kernel.org/r/20260902171414.96165-2-sidchintamaneni@gmail.com Signed-off-by: Alexei Starovoitov --- .../selftests/bpf/progs/verifier_gotox.c | 48 +++++++++++++++++++ 1 file changed, 48 insertions(+) diff --git a/tools/testing/selftests/bpf/progs/verifier_gotox.c b/tools/testing/selftests/bpf/progs/verifier_gotox.c index 5b18c9a27717..0e27c2c79c57 100644 --- a/tools/testing/selftests/bpf/progs/verifier_gotox.c +++ b/tools/testing/selftests/bpf/progs/verifier_gotox.c @@ -47,6 +47,54 @@ DEFINE_SIMPLE_JUMP_TABLE_PROG(reserved_field_src_reg, BPF_REG_1, 0, 0, __fa DEFINE_SIMPLE_JUMP_TABLE_PROG(reserved_field_non_zero_off, BPF_REG_0, 1, 0, __failure __msg("BPF_JA|BPF_X uses reserved fields")) DEFINE_SIMPLE_JUMP_TABLE_PROG(reserved_field_non_zero_imm, BPF_REG_0, 0, 1, __failure __msg("BPF_JA|BPF_X uses reserved fields")) +#define DEFINE_TERMINAL_GOTOX_PROG(NAME, BASE) \ + __naked void NAME(void) \ + { \ + asm volatile (" \ + .pushsection .jumptables,\"\",@progbits; \ +jt0_%=: \ + .quad ret0_%= - " BASE "; \ + .size jt0_%=, 8; \ + .global jt0_%=; \ + .popsection; \ + \ + r0 = jt0_%= ll; \ + r0 = *(u64 *)(r0 + 0); \ + goto end_%=; \ +ret0_%=: \ + r0 = 0; \ + exit; \ +end_%=: \ + .8byte %[gotox_r0]; \ +" : \ + : __imm_insn(gotox_r0, BPF_RAW_INSN(BPF_JMP | BPF_JA | BPF_X, \ + BPF_REG_0, 0, 0, 0)) \ + : __clobber_all); \ + } + +SEC("socket") +__success __retval(0) +DEFINE_TERMINAL_GOTOX_PROG(jump_table_terminal_gotox, "socket") + +static __noinline __used +DEFINE_TERMINAL_GOTOX_PROG(terminal_gotox_subprog1, ".text") + +static __noinline __used int terminal_gotox_subprog2(void) +{ + return 0; +} + +SEC("socket") +__success __retval(0) +__naked void jump_table_terminal_gotox_subprog(void) +{ + asm volatile (" \ + call terminal_gotox_subprog1; \ + call terminal_gotox_subprog2; \ + exit; \ +" ::: __clobber_all); +} + /* * Gotox is forbidden when there is no jump table loaded * which points to the sub-function where the gotox is used From 85136bf22404474a815fc0ed26ec0d1cbc1bc3f9 Mon Sep 17 00:00:00 2001 From: "Jose Fernandez (Anthropic)" Date: Wed, 9 Sep 2026 17:51:04 +0000 Subject: [PATCH 135/592] bpf: Avoid soft lockup in __htab_map_lookup_and_delete_batch() __htab_map_lookup_and_delete_batch() has no rescheduling point. The batch count bounds how many entries are copied out, not how many buckets are visited, so one BPF_MAP_LOOKUP_BATCH call can walk the map end to end. The empty-bucket fast path is worse: it stays inside a single rcu_read_lock() / bpf_disable_instrumentation() section for any run of consecutive empty buckets. That holds up on small maps, but it falls apart at scale. On a 144-CPU arm64 host running a CONFIG_PREEMPT_NONE kernel, periodic BPF_MAP_LOOKUP_BATCH calls against an LRU hash map with 16,777,216 buckets held a CPU inside the batch op for 77+ seconds and triggered the soft lockup watchdog. Commit 75134f16e7dd ("bpf: Add schedule points in batch ops") fixed this same problem in the generic batch ops, but not in this htab-native path, which every htab-based hash map variant uses for its lookup[_and_delete] batch ops. Complete that fix here. Leave the critical section after 64 consecutive empty buckets, call cond_resched_tasks_rcu_qs(), and resume at the saved bucket cursor. No locks are held at that point, and resuming from the cursor is already the function's behavior for non-empty buckets. Add the same call to the per-bucket loop after copy_to_user(), where every lock has been dropped. cond_resched_rcu() is not enough here: sleeping with bpf_prog_active elevated makes tracing programs on that CPU silently skip their invocations. Plain cond_resched() is not enough either. It is a no-op under PREEMPT and PREEMPT_LAZY, the only models arm64 and x86 have offered since commit 7dadeaa6e851 ("sched: Further restrict the preemption modes"). It is also never a Tasks RCU quiescent state, in any model: the reschedule counts as a preemption. The walking task stays a holdout and stalls every synchronize_rcu_tasks() caller, ftrace and BPF trampoline teardown included, until the syscall returns [1]. cond_resched_tasks_rcu_qs() is the usual tool for that [2]. It reports the quiescent state at each yield and still reschedules as cond_resched() does on PREEMPT_NONE and PREEMPT_VOLUNTARY kernels. Fixes: 057996380a42 ("bpf: Add batch ops to all htab bpf map") Cc: "Paul E. McKenney" Cc: Rik van Riel Link: https://lore.kernel.org/bpf/20260715215314.44423f47@fangorn/ [1] Link: https://lore.kernel.org/bpf/9d444098-7c03-4163-af12-bd0a79a51443@paulmck-laptop/ [2] Assisted-by: LLM Signed-off-by: Jose Fernandez (Anthropic) Signed-off-by: Josef Bacik Reviewed-by: Rik van Riel Link: https://lore.kernel.org/r/20260909-b4-htab-batch-resched-v2-1-0cb529d8f95a@toxicpanda.com Signed-off-by: Alexei Starovoitov --- kernel/bpf/hashtab.c | 24 +++++++++++++++++++++--- 1 file changed, 21 insertions(+), 3 deletions(-) diff --git a/kernel/bpf/hashtab.c b/kernel/bpf/hashtab.c index 6f331c80130d..a72dc5b9f184 100644 --- a/kernel/bpf/hashtab.c +++ b/kernel/bpf/hashtab.c @@ -1772,6 +1772,12 @@ static int htab_lru_percpu_map_lookup_and_delete_elem(struct bpf_map *map, flags); } +/* + * Max consecutive empty buckets to walk in one RCU + + * instrumentation-disabled section before rescheduling. + */ +#define HTAB_BATCH_EMPTY_RESCHED 64 + static int __htab_map_lookup_and_delete_batch(struct bpf_map *map, const union bpf_attr *attr, @@ -1793,6 +1799,7 @@ __htab_map_lookup_and_delete_batch(struct bpf_map *map, unsigned long flags = 0; bool locked = false; struct htab_elem *l; + u32 empty_cnt = 0; struct bucket *b; int ret = 0; @@ -1971,12 +1978,21 @@ __htab_map_lookup_and_delete_batch(struct bpf_map *map, } next_batch: - /* If we are not copying data, we can go to next bucket and avoid - * unlocking the rcu. + /* + * If we are not copying data, we can go to next bucket and avoid + * unlocking the rcu. Bound the walk though: after + * HTAB_BATCH_EMPTY_RESCHED consecutive empty buckets, fully exit + * the critical section (no locks are held here) and reschedule. */ if (!bucket_cnt && (batch + 1 < htab->n_buckets)) { batch++; - goto again_nocopy; + if (++empty_cnt < HTAB_BATCH_EMPTY_RESCHED) + goto again_nocopy; + empty_cnt = 0; + rcu_read_unlock(); + bpf_enable_instrumentation(); + cond_resched_tasks_rcu_qs(); + goto again; } rcu_read_unlock(); @@ -1990,11 +2006,13 @@ __htab_map_lookup_and_delete_batch(struct bpf_map *map, } total += bucket_cnt; + empty_cnt = 0; batch++; if (batch >= htab->n_buckets) { ret = -ENOENT; goto after_loop; } + cond_resched_tasks_rcu_qs(); goto again; after_loop: From 75f8cf22463d82bb1fb0239a3d485fc8f4c8ef03 Mon Sep 17 00:00:00 2001 From: Jiayuan Chen Date: Thu, 3 Sep 2026 18:09:20 +0800 Subject: [PATCH 136/592] bpf: Fix out-of-bounds read of rtt_min in sock_ops A sockops prog reading skops->rtt_min never checks the sk type: on the tcp_conn_request() path sock_ops->sk is a request_sock (non-full), and the ctx rewrite casts it to a tcp_sock (full) and reads rtt_min past the end of the request_sock, returning dirty adjacent memory. SEC("sockops") int prog(struct bpf_sock_ops *skops) { switch (skops->op) { case BPF_SOCK_OPS_RWND_INIT: leak = skops->rtt_min; /* reads the request_sock OOB */ ... } } For instance one such read returned rtt_min=0xffff8881, the high half of a leaked kernel pointer. Guarding that cast is exactly what SOCK_OPS_GET_FIELD() does -- it checks is_locked_tcp_sock and returns 0 when sock_ops->sk is not a locked full socket. Every other tcp_sock field in sock_ops goes through it; rtt_min is the only one open-coded, so it skips the check. Read rtt_min through SOCK_OPS_GET_FIELD() too. rtt_min is a bit special: it is a struct minmax and we only want the current min, so pass rtt_min.s[0].v. That is equivalent to the old hand-computed offset offsetof(struct tcp_sock, rtt_min) + sizeof_field(struct minmax_sample, t) (s[0] sits at rtt_min + 0 and .v at + sizeof(.t), i.e. what minmax_get() returns), so the loaded field is unchanged and only the full-sock guard is added. The two BUILD_BUG_ON()s that protected the hand-computed offset are no longer needed. Before patch: 0: r1 = *(u64 *)(r1 +0) ; r1 = skops->sk 1: r1 = *(u32 *)(r1 +2324) ; ((tcp_sock *)sk)->rtt_min.s[0].v After patch: 0: *(u64 *)(r1 +56) = r9 1: r9 = *(u8 *)(r1 +50) ; is_locked_tcp_sock 2: if r9 == 0 goto pc+4 ; not a locked full sock -> 0 3: r9 = *(u64 *)(r1 +56) 4: r1 = *(u64 *)(r1 +0) ; r1 = skops->sk 5: r1 = *(u32 *)(r1 +2324) ; rtt_min.s[0].v 6: goto pc+2 7: r9 = *(u64 *)(r1 +56) 8: r1 = 0 Fixes: 44f0e43037d3 ("bpf: Add support for reading sk_state and more") Reported-by: VEGA Signed-off-by: Jiayuan Chen Reviewed-by: Emil Tsalapatis Link: https://lore.kernel.org/r/20260903100921.113374-1-jiayuan.chen@linux.dev Signed-off-by: Alexei Starovoitov --- net/core/filter.c | 13 +------------ 1 file changed, 1 insertion(+), 12 deletions(-) diff --git a/net/core/filter.c b/net/core/filter.c index cae43b999162..532405988fd9 100644 --- a/net/core/filter.c +++ b/net/core/filter.c @@ -11106,18 +11106,7 @@ static u32 sock_ops_convert_ctx_access(enum bpf_access_type type, break; case offsetof(struct bpf_sock_ops, rtt_min): - BUILD_BUG_ON(sizeof_field(struct tcp_sock, rtt_min) != - sizeof(struct minmax)); - BUILD_BUG_ON(sizeof(struct minmax) < - sizeof(struct minmax_sample)); - - *insn++ = BPF_LDX_MEM(BPF_FIELD_SIZEOF( - struct bpf_sock_ops_kern, sk), - si->dst_reg, si->src_reg, - offsetof(struct bpf_sock_ops_kern, sk)); - *insn++ = BPF_LDX_MEM(BPF_W, si->dst_reg, si->dst_reg, - offsetof(struct tcp_sock, rtt_min) + - sizeof_field(struct minmax_sample, t)); + SOCK_OPS_GET_FIELD(rtt_min, rtt_min.s[0].v, struct tcp_sock); break; case offsetof(struct bpf_sock_ops, bpf_sock_ops_cb_flags): From 7d70a0b02d262971201fdd1e221586bdd95c2910 Mon Sep 17 00:00:00 2001 From: Zhixing Chen Date: Thu, 3 Sep 2026 18:43:58 +0800 Subject: [PATCH 137/592] bpf: Use kvfree() in xdp_test_run_teardown() xdp_test_run_setup() allocates xdp->frames and xdp->skbs with kvmalloc_array(). The setup error path already releases both arrays with kvfree(), while the normal teardown path still uses kfree(). Use kvfree() in xdp_test_run_teardown() as well, so the release helper matches the allocator on both paths. Signed-off-by: Zhixing Chen Reviewed-by: Emil Tsalapatis Link: https://lore.kernel.org/r/20260903104358.29228-1-running910@gmail.com Signed-off-by: Alexei Starovoitov --- net/bpf/test_run.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/net/bpf/test_run.c b/net/bpf/test_run.c index 5d51f6cb7d15..513354e928cb 100644 --- a/net/bpf/test_run.c +++ b/net/bpf/test_run.c @@ -205,8 +205,8 @@ static void xdp_test_run_teardown(struct xdp_test_data *xdp) { xdp_unreg_mem_model(&xdp->mem); page_pool_destroy(xdp->pp); - kfree(xdp->frames); - kfree(xdp->skbs); + kvfree(xdp->frames); + kvfree(xdp->skbs); } static bool frame_was_changed(const struct xdp_page_head *head) From 6d8c197c9992659a65525a07de4368c8401fdda7 Mon Sep 17 00:00:00 2001 From: Namjae Jeon Date: Tue, 8 Sep 2026 15:52:46 +0900 Subject: [PATCH 138/592] ntfs: use dynamic MFT tail reservation The ntfs MFT allocator historically treated records below 64 as a permanent extension area and stopped searching for $MFT extent records after record 400. Windows and ntfs3 do not maintain that on-disk layout, so an NTFS volume can have free MFT records while ntfs returns -ENOSPC when $MFT:$DATA needs another mapping-pairs extent. Use record 24 as the first normal record and maintain an in-memory tail reserve of up to four initialized records. Normal allocations skip the reserve, while $MFT metadata extent allocations consume it. When a new tail is initialized, allocate at least two records and reserve the following records to avoid recursive allocation during MFT extension. Existing free runs can seed the reserve on volumes mounted without one. For $MFT/$DATA, constrain an extent record to a record whose byte offset is below the new extent lowest VCN byte offset. This preserves bootstrap reachability without an arbitrary record 400 limit. If no safe record is available, validate and use reserved records 15, 12, 13, and 14 as bootstrap candidates while keeping their MFT bitmap entries in use. This allows existing Windows volumes to extend $MFT using their actual free records and prevents normal file allocation from consuming the metadata reserve. Fixes: 115380f9a2f9 ("ntfs: update mft operations") Reviewed-by: Hyunchul Lee Signed-off-by: Namjae Jeon --- fs/ntfs/attrib.c | 11 +- fs/ntfs/mft.c | 418 +++++++++++++++++++++++++++++++++++++---------- fs/ntfs/mft.h | 2 +- fs/ntfs/namei.c | 2 +- fs/ntfs/volume.h | 6 + 5 files changed, 348 insertions(+), 91 deletions(-) diff --git a/fs/ntfs/attrib.c b/fs/ntfs/attrib.c index 848a0d338b89..81b9e0971b3e 100644 --- a/fs/ntfs/attrib.c +++ b/fs/ntfs/attrib.c @@ -2917,7 +2917,7 @@ int ntfs_attr_add(struct ntfs_inode *ni, __le32 type, attr_ni = NULL; /* Allocate new extent. */ - err = ntfs_mft_record_alloc(ni->vol, 0, &attr_ni, ni, NULL); + err = ntfs_mft_record_alloc(ni->vol, 0, &attr_ni, ni, NULL, -1); if (err) { ntfs_error(sb, "Failed to allocate extent record"); goto err_out; @@ -3550,7 +3550,7 @@ int ntfs_attr_record_move_away(struct ntfs_attr_search_ctx *ctx, int extra) * new extent and move attribute to it. */ ni = NULL; - err = ntfs_mft_record_alloc(base_ni->vol, 0, &ni, base_ni, NULL); + err = ntfs_mft_record_alloc(base_ni->vol, 0, &ni, base_ni, NULL, -1); if (err) { ntfs_error(sb, "Couldn't allocate MFT record, err : %d", err); return err; @@ -3976,7 +3976,10 @@ int ntfs_attr_update_mapping_pairs(struct ntfs_inode *ni, s64 from_vcn) unsigned int de_cnt = 0; /* Allocate new mft record. */ - err = ntfs_mft_record_alloc(ni->vol, 0, &ext_ni, base_ni, NULL); + err = ntfs_mft_record_alloc(ni->vol, 0, &ext_ni, base_ni, NULL, + base_ni->mft_no == FILE_MFT && + ni->type == AT_DATA && + ni->name == AT_UNNAMED ? stop_vcn : -1); if (err) { ntfs_error(sb, "Failed to allocate extent record"); goto put_err_out; @@ -4836,7 +4839,7 @@ static int ntfs_resident_attr_resize(struct ntfs_inode *attr_ni, const s64 newsi } /* Allocate new mft record. */ - err = ntfs_mft_record_alloc(base_ni->vol, 0, &ext_ni, base_ni, NULL); + err = ntfs_mft_record_alloc(base_ni->vol, 0, &ext_ni, base_ni, NULL, -1); if (err) { ntfs_error(sb, "Couldn't allocate MFT record"); goto put_err_out; diff --git a/fs/ntfs/mft.c b/fs/ntfs/mft.c index 7e58c99f1728..61d4ee3a57fd 100644 --- a/fs/ntfs/mft.c +++ b/fs/ntfs/mft.c @@ -841,12 +841,126 @@ static bool ntfs_may_write_mft_record(struct ntfs_volume *vol, const u64 mft_no, static const char *es = " Leaving inconsistent metadata. Unmount and run chkdsk."; -#define RESERVED_MFT_RECORDS 64 +#define FIRST_NORMAL_MFT_RECORD 24 +#define MFT_RECORD_RESERVE 4 /* - * ntfs_mft_bitmap_find_and_alloc_free_rec_nolock - see name + * Records 12-15 are marked in use by Windows but normally have no name + * and no links. Keep them as the last bootstrap option when a volume + * mounted without an in-memory tail reserve needs its first $MFT metadata + * extent. + */ +static bool mft_reserved_is_free(struct ntfs_volume *vol, + struct ntfs_inode *mft_ni, s64 mft_no) +{ + struct attr_record *a; + struct mft_record *m; + struct folio *folio; + void *mapped; + pgoff_t index = NTFS_MFT_NR_TO_PIDX(vol, mft_no); + unsigned int ofs = NTFS_MFT_NR_TO_POFS(vol, mft_no); + u32 attrs_offset, bytes_in_use; + bool available = false, have_std = false; + int i; + + for (i = 0; i < mft_ni->nr_extents; i++) { + if (mft_ni->ext.extent_ntfs_inos[i] && + mft_ni->ext.extent_ntfs_inos[i]->mft_no == mft_no) + return false; + } + m = kmalloc(vol->mft_record_size, GFP_NOFS); + if (!m) + return false; + + folio = read_mapping_folio(vol->mft_ino->i_mapping, index, NULL); + if (IS_ERR(folio)) + goto free_m; + + folio_lock(folio); + mapped = kmap_local_folio(folio, 0); + memcpy(m, (u8 *)mapped + ofs, vol->mft_record_size); + kunmap_local(mapped); + folio_unlock(folio); + folio_put(folio); + if (post_read_mst_fixup((struct ntfs_record *)m, vol->mft_record_size)) + goto free_m; + + if (!ntfs_is_mft_record(m->magic) || + !(m->flags & MFT_RECORD_IN_USE) || m->base_mft_record || + m->link_count) + goto out; + + attrs_offset = le16_to_cpu(m->attrs_offset); + bytes_in_use = le32_to_cpu(m->bytes_in_use); + if (attrs_offset > bytes_in_use || bytes_in_use > vol->mft_record_size || + bytes_in_use - attrs_offset < sizeof(a->type)) + goto out; + + for (a = (struct attr_record *)((u8 *)m + attrs_offset); + (u8 *)a + sizeof(a->type) <= (u8 *)m + bytes_in_use;) { + u32 len; + + if (a->type == AT_END) { + if ((u8 *)a + sizeof(a->type) + sizeof(a->length) > + (u8 *)m + bytes_in_use) + break; + /* Also accept a record emptied by an earlier bootstrap. */ + available = have_std || + (u8 *)a == (u8 *)m + attrs_offset; + break; + } + if (a->type == AT_FILE_NAME) + break; + len = le32_to_cpu(a->length); + if (len < offsetof(struct attr_record, data) || + (u8 *)a + len > (u8 *)m + bytes_in_use) + break; + if (a->type == AT_STANDARD_INFORMATION) { + u32 value_len, value_ofs; + + if (have_std || a->non_resident || + len < offsetof(struct attr_record, + data.resident.reserved) + 1) + break; + value_len = le32_to_cpu(a->data.resident.value_length); + value_ofs = le16_to_cpu(a->data.resident.value_offset); + if (value_ofs > len || value_len > len - value_ofs) + break; + have_std = true; + } + a = (struct attr_record *)((u8 *)a + len); + } +out: + kfree(m); + return available; +free_m: + kfree(m); + return false; +} + +static s64 mft_reserve_end(const u8 *buf, s64 buf_start, s64 buf_end, + s64 start, s64 pass_end, s64 initialized_mft_records) +{ + s64 end = start + 1; + s64 limit = min_t(s64, start + MFT_RECORD_RESERVE, pass_end); + + if (limit > initialized_mft_records) + limit = initialized_mft_records; + if (limit > buf_end) + limit = buf_end; + while (end < limit && + !(buf[(end - buf_start) >> 3] & + (1 << ((end - buf_start) & 7)))) + end++; + return end; +} + +/* + * mft_bitmap_alloc_free_rec - find and allocate a free MFT record * @vol: volume on which to search for a free mft record * @base_ni: open base inode if allocating an extent mft record or NULL + * @max_mft_no: first record which must not be allocated, or -1 + * @new_reserve_end: if not NULL, end of a free run starting after the result * * Search for a free mft record in the mft bitmap attribute on the ntfs volume * @vol. @@ -862,10 +976,12 @@ static const char *es = " Leaving inconsistent metadata. Unmount and run chkds * * Locking: Caller must hold vol->mftbmp_lock for writing. */ -static s64 ntfs_mft_bitmap_find_and_alloc_free_rec_nolock(struct ntfs_volume *vol, - struct ntfs_inode *base_ni) +static s64 mft_bitmap_alloc_free_rec(struct ntfs_volume *vol, + struct ntfs_inode *base_ni, + s64 max_mft_no, s64 *new_reserve_end) { s64 pass_end, ll, data_pos, pass_start, ofs, bit; + s64 initialized_mft_records; unsigned long flags; struct address_space *mftbmp_mapping; u8 *buf = NULL, *byte; @@ -882,30 +998,36 @@ static s64 ntfs_mft_bitmap_find_and_alloc_free_rec_nolock(struct ntfs_volume *vo read_lock_irqsave(&NTFS_I(vol->mft_ino)->size_lock, flags); pass_end = NTFS_I(vol->mft_ino)->allocated_size >> vol->mft_record_size_bits; + initialized_mft_records = NTFS_I(vol->mft_ino)->initialized_size >> + vol->mft_record_size_bits; read_unlock_irqrestore(&NTFS_I(vol->mft_ino)->size_lock, flags); read_lock_irqsave(&NTFS_I(vol->mftbmp_ino)->size_lock, flags); ll = NTFS_I(vol->mftbmp_ino)->initialized_size << 3; read_unlock_irqrestore(&NTFS_I(vol->mftbmp_ino)->size_lock, flags); if (pass_end > ll) pass_end = ll; - pass = 1; - if (!base_ni) - data_pos = vol->mft_data_pos; - else - data_pos = base_ni->mft_no + 1; - if (data_pos < RESERVED_MFT_RECORDS) - data_pos = RESERVED_MFT_RECORDS; - if (data_pos >= pass_end) { - data_pos = RESERVED_MFT_RECORDS; + if (max_mft_no >= 0 && pass_end > max_mft_no) + pass_end = max_mft_no; + if (base_ni && base_ni->mft_no == FILE_MFT) { + data_pos = FILE_first_user; pass = 2; - /* This happens on a freshly formatted volume. */ if (data_pos >= pass_end) return -ENOSPC; - } - - if (base_ni && base_ni->mft_no == FILE_MFT) { - data_pos = 0; - pass = 2; + } else { + pass = 1; + if (!base_ni) + data_pos = vol->mft_data_pos; + else + data_pos = base_ni->mft_no + 1; + if (data_pos < FIRST_NORMAL_MFT_RECORD) + data_pos = FIRST_NORMAL_MFT_RECORD; + if (data_pos >= pass_end) { + data_pos = FIRST_NORMAL_MFT_RECORD; + pass = 2; + /* This happens on a freshly formatted volume. */ + if (data_pos >= pass_end) + return -ENOSPC; + } } pass_start = data_pos; @@ -940,38 +1062,28 @@ static s64 ntfs_mft_bitmap_find_and_alloc_free_rec_nolock(struct ntfs_volume *vo size, data_pos, bit); for (; bit < size && data_pos + bit < pass_end; bit &= ~7ull, bit += 8) { - /* - * If we're extending $MFT and running out of the first - * mft record (base record) then give up searching since - * no guarantee that the found record will be accessible. - */ - if (base_ni && base_ni->mft_no == FILE_MFT && bit > 400) { - folio_unlock(folio); - kunmap_local(buf); - folio_put(folio); - return -ENOSPC; - } - byte = buf + (bit >> 3); if (*byte == 0xff) continue; - b = ffz((unsigned long)*byte); - if (b < 8 && b >= (bit & 7)) { + b = bit & 7; + for (; b < 8; b++) { + if (*byte & (1 << b)) + continue; ll = data_pos + (bit & ~7ull) + b; + if (ll >= pass_end) + break; + /* Keep the dynamic tail reserve for $MFT metadata. */ + if ((!base_ni || base_ni->mft_no != FILE_MFT) && + ll >= vol->mft_record_reserve_pos && + ll < vol->mft_record_reserve_end) + continue; if (unlikely(ll >= (1ll << 32))) { folio_unlock(folio); kunmap_local(buf); folio_put(folio); return -ENOSPC; } - *byte |= 1 << b; - folio_mark_dirty(folio); - folio_unlock(folio); - kunmap_local(buf); - folio_put(folio); - ntfs_debug("Done. (Found and allocated mft record 0x%llx.)", - ll); - return ll; + goto found; } } ntfs_debug("After inner for loop: size 0x%x, data_pos 0x%llx, bit 0x%llx", @@ -994,7 +1106,8 @@ static s64 ntfs_mft_bitmap_find_and_alloc_free_rec_nolock(struct ntfs_volume *vo * part of the zone which we omitted earlier. */ pass_end = pass_start; - data_pos = pass_start = RESERVED_MFT_RECORDS; + data_pos = FIRST_NORMAL_MFT_RECORD; + pass_start = FIRST_NORMAL_MFT_RECORD; ntfs_debug("pass %i, pass_start 0x%llx, pass_end 0x%llx.", pass, pass_start, pass_end); if (data_pos >= pass_end) @@ -1004,6 +1117,18 @@ static s64 ntfs_mft_bitmap_find_and_alloc_free_rec_nolock(struct ntfs_volume *vo /* No free mft records in currently initialized mft bitmap. */ ntfs_debug("Done. (No free mft records left in currently initialized mft bitmap.)"); return -ENOSPC; +found: + if (new_reserve_end) + *new_reserve_end = mft_reserve_end(buf, data_pos, + data_pos + size, ll, pass_end, + initialized_mft_records); + *byte |= 1 << b; + folio_mark_dirty(folio); + folio_unlock(folio); + kunmap_local(buf); + folio_put(folio); + ntfs_debug("Done. (Found and allocated mft record 0x%llx.)", ll); + return ll; } static int ntfs_mft_attr_extend(struct ntfs_inode *ni) @@ -1471,8 +1596,9 @@ static int ntfs_mft_bitmap_extend_initialized_nolock(struct ntfs_volume *vol) * @vol: volume on which to extend the mft data attribute * * Extend the mft data attribute on the ntfs volume @vol by 16 mft records - * worth of clusters or if not enough space for this by one mft record worth - * of clusters. + * worth of clusters or if not enough space for this by two mft records worth + * of clusters. Keeping at least two new records breaks the recursion between + * extending $MFT and allocating a record for a new $MFT attribute extent. * * Note: Only changes allocated_size, i.e. does not touch initialized_size or * data_size. @@ -1526,10 +1652,8 @@ static int ntfs_mft_data_extend_allocation_nolock(struct ntfs_volume *vol) } lcn = rl->lcn + rl->length; ntfs_debug("Last lcn of mft data attribute is 0x%llx.", lcn); - /* Minimum allocation is one mft record worth of clusters. */ - min_nr = NTFS_B_TO_CLU(vol, vol->mft_record_size); - if (!min_nr) - min_nr = 1; + /* Keep room for the allocating record and at least one MFT reserve. */ + min_nr = DIV_ROUND_UP_ULL((u64)vol->mft_record_size * 2, vol->cluster_size); /* Want to allocate 16 mft records worth of clusters. */ nr = vol->mft_record_size << 4 >> vol->cluster_size_bits; if (!nr) @@ -1928,6 +2052,7 @@ static int ntfs_mft_record_format(const struct ntfs_volume *vol, const s64 mft_n * @ni: [OUT] on success, set to the allocated ntfs inode * @base_ni: [IN] open base inode if allocating an extent mft record or NULL * @ni_mrec: [OUT] on successful return this is the mapped mft record + * @mft_data_vcn: [IN] lowest VCN of a new $MFT/$DATA extent, or -1 * * Allocate an mft record in $MFT/$DATA of an open ntfs volume @vol. * @@ -1955,30 +2080,23 @@ static int ntfs_mft_record_format(const struct ntfs_volume *vol, const s64 mft_n * optimize this we start scanning at the place specified by @base_ni or if * @base_ni is NULL we start where we last stopped and we perform wrap around * when we reach the end. Note, we do not try to allocate mft records below - * number 64 because numbers 0 to 15 are the defined system files anyway and 16 - * to 64 are special in that they are used for storing extension mft records - * for the $DATA attribute of $MFT. This is required to avoid the possibility - * of creating a runlist with a circular dependency which once written to disk - * can never be read in again. Windows will only use records 16 to 24 for - * normal files if the volume is completely out of space. We never use them - * which means that when the volume is really out of space we cannot create any - * more files while Windows can still create up to 8 small files. We can start - * doing this at some later time, it does not matter much for now. + * number 24 because numbers 0 to 15 are the defined system files and records + * 16 to 23 are kept for metadata compatibility. Records reserved dynamically + * at the initialized MFT tail are skipped by normal allocation and consumed by + * $MFT metadata extent allocation. * * When scanning the mft bitmap, we only search up to the last allocated mft - * record. If there are no free records left in the range 64 to number of + * record. If there are no free records left in the range 24 to number of * allocated mft records, then we extend the $MFT/$DATA attribute in order to * create free mft records. We extend the allocated size of $MFT/$DATA by 16 * records at a time or one cluster, if cluster size is above 16kiB. If there - * is not sufficient space to do this, we try to extend by a single mft record - * or one cluster, if cluster size is above the mft record size. + * is not sufficient space to do this, we try to extend by two mft records or + * one cluster, if a cluster already contains at least two mft records. * - * No matter how many mft records we allocate, we initialize only the first - * allocated mft record, incrementing mft data size and initialized size - * accordingly, open an struct ntfs_inode for it and return it to the caller, unless - * there are less than 64 mft records, in which case we allocate and initialize - * mft records until we reach record 64 which we consider as the first free mft - * record for use by normal files. + * When extending the initialized MFT tail, we also initialize up to four + * additional records and reserve them in memory for future $MFT metadata + * extents. If there are less than 24 mft records, records are initialized + * until record 24, which is the first record used for normal files. * * If during any stage we overflow the initialized data in the mft bitmap, we * extend the initialized size (and data size) by 8 bytes, allocating another @@ -2014,9 +2132,12 @@ static int ntfs_mft_record_format(const struct ntfs_volume *vol, const s64 mft_n */ int ntfs_mft_record_alloc(struct ntfs_volume *vol, const int mode, struct ntfs_inode **ni, struct ntfs_inode *base_ni, - struct mft_record **ni_mrec) + struct mft_record **ni_mrec, const s64 mft_data_vcn) { s64 ll, bit, old_data_initialized, old_data_size; + s64 max_mft_no = -1, reserve_start = -1, reserve_end = -1; + s64 candidate_reserve_end = -1; + s64 *reserve_endp; unsigned long flags; struct folio *folio; struct ntfs_inode *mft_ni, *mftbmp_ni; @@ -2027,7 +2148,9 @@ int ntfs_mft_record_alloc(struct ntfs_volume *vol, const int mode, unsigned int ofs; int err; __le16 seq_no, usn; - bool record_formatted = false; + bool record_formatted = false, from_reserve = false, tail_alloc = false; + bool reserve_created = false; + bool forced_reserved_record = false; unsigned int memalloc_flags; if (base_ni && *ni) @@ -2036,6 +2159,21 @@ int ntfs_mft_record_alloc(struct ntfs_volume *vol, const int mode, /* @mode and @base_ni are mutually exclusive. */ if (mode && base_ni) return -EINVAL; + if (mft_data_vcn >= 0 && + (!base_ni || base_ni->mft_no != FILE_MFT)) + return -EINVAL; + if (mft_data_vcn >= 0) { + u64 vbo; + + if ((u64)mft_data_vcn > (U64_MAX >> vol->cluster_size_bits)) + return -EOVERFLOW; + vbo = (u64)mft_data_vcn << vol->cluster_size_bits; + /* + * The whole extent record must be reachable without this + * extent, including when an MFT record spans multiple clusters. + */ + max_mft_no = vbo >> vol->mft_record_size_bits; + } if (base_ni) ntfs_debug("Entering (allocating an extent mft record for base mft record 0x%llx).", @@ -2050,10 +2188,39 @@ int ntfs_mft_record_alloc(struct ntfs_volume *vol, const int mode, mutex_lock(&mft_ni->mrec_lock); mftbmp_ni = NTFS_I(vol->mftbmp_ino); search_free_rec: + from_reserve = false; + reserve_created = false; + candidate_reserve_end = -1; if (!base_ni || base_ni->mft_no != FILE_MFT) down_write(&vol->mftbmp_lock); - bit = ntfs_mft_bitmap_find_and_alloc_free_rec_nolock(vol, base_ni); + if (base_ni && base_ni->mft_no == FILE_MFT && + vol->mft_record_reserve_pos < vol->mft_record_reserve_end && + (max_mft_no < 0 || vol->mft_record_reserve_pos < max_mft_no)) { + bit = vol->mft_record_reserve_pos; + err = ntfs_bitmap_set_bit(vol->mftbmp_ino, bit); + if (unlikely(err)) { + ntfs_error(vol->sb, + "Failed to allocate reserved MFT record 0x%llx.", + bit); + goto err_out; + } + vol->mft_record_reserve_pos++; + from_reserve = true; + ntfs_debug("Allocated MFT metadata record 0x%llx from tail reserve.", + bit); + goto have_alloc_rec; + } + reserve_endp = vol->mft_record_reserve_pos >= + vol->mft_record_reserve_end ? &candidate_reserve_end : NULL; + bit = mft_bitmap_alloc_free_rec(vol, base_ni, max_mft_no, reserve_endp); if (bit >= 0) { + if (candidate_reserve_end > bit + 1) { + vol->mft_record_reserve_pos = bit + 1; + vol->mft_record_reserve_end = candidate_reserve_end; + reserve_created = true; + ntfs_debug("Reserved free MFT records [0x%llx, 0x%llx) for metadata.", + bit + 1, candidate_reserve_end); + } ntfs_debug("Found and allocated free record (#1), bit 0x%llx.", (long long)bit); goto have_alloc_rec; @@ -2068,6 +2235,24 @@ int ntfs_mft_record_alloc(struct ntfs_volume *vol, const int mode, } if (base_ni && base_ni->mft_no == FILE_MFT) { + static const u8 bootstrap_records[] = { + FILE_reserved15, FILE_reserved12, FILE_reserved13, + FILE_reserved14, + }; + int i; + + for (i = 0; i < ARRAY_SIZE(bootstrap_records); i++) { + if (max_mft_no >= 0 && bootstrap_records[i] >= max_mft_no) + continue; + if (!mft_reserved_is_free(vol, mft_ni, + bootstrap_records[i])) + continue; + bit = bootstrap_records[i]; + forced_reserved_record = true; + ntfs_debug("Using reserved MFT record %lld to bootstrap metadata extension.", + bit); + goto have_alloc_rec; + } memalloc_nofs_restore(memalloc_flags); return bit; } @@ -2087,10 +2272,10 @@ int ntfs_mft_record_alloc(struct ntfs_volume *vol, const int mode, old_data_initialized = mftbmp_ni->initialized_size; read_unlock_irqrestore(&mftbmp_ni->size_lock, flags); if (old_data_initialized << 3 > ll && - old_data_initialized > RESERVED_MFT_RECORDS / 8) { + old_data_initialized << 3 > FIRST_NORMAL_MFT_RECORD) { bit = ll; - if (bit < RESERVED_MFT_RECORDS) - bit = RESERVED_MFT_RECORDS; + if (bit < FIRST_NORMAL_MFT_RECORD) + bit = FIRST_NORMAL_MFT_RECORD; if (unlikely(bit >= (1ll << 32))) goto max_err_out; ntfs_debug("Found free record (#2), bit 0x%llx.", @@ -2176,6 +2361,11 @@ int ntfs_mft_record_alloc(struct ntfs_volume *vol, const int mode, read_lock_irqsave(&mft_ni->size_lock, flags); old_data_initialized = mft_ni->initialized_size; read_unlock_irqrestore(&mft_ni->size_lock, flags); + tail_alloc = (!base_ni || base_ni->mft_no != FILE_MFT) && + bit >= (old_data_initialized >> vol->mft_record_size_bits) && + vol->mft_record_reserve_pos >= vol->mft_record_reserve_end; + if (tail_alloc) + ll = (bit + 2) << vol->mft_record_size_bits; if (ll <= old_data_initialized) { ntfs_debug("Allocated mft record already initialized."); goto mft_rec_already_initialized; @@ -2208,6 +2398,29 @@ int ntfs_mft_record_alloc(struct ntfs_volume *vol, const int mode, mft_ni->initialized_size); } read_unlock_irqrestore(&mft_ni->size_lock, flags); + if (tail_alloc) { + s64 bitmap_records; + + read_lock_irqsave(&mft_ni->size_lock, flags); + reserve_end = mft_ni->allocated_size >> + vol->mft_record_size_bits; + read_unlock_irqrestore(&mft_ni->size_lock, flags); + read_lock_irqsave(&mftbmp_ni->size_lock, flags); + bitmap_records = mftbmp_ni->initialized_size << 3; + read_unlock_irqrestore(&mftbmp_ni->size_lock, flags); + if (reserve_end > bitmap_records) + reserve_end = bitmap_records; + if (reserve_end > bit + 1 + MFT_RECORD_RESERVE) + reserve_end = bit + 1 + MFT_RECORD_RESERVE; + reserve_start = bit + 1; + if (reserve_end > reserve_start) { + ll = reserve_end << vol->mft_record_size_bits; + } else { + reserve_start = -1; + reserve_end = -1; + ll = (bit + 1) << vol->mft_record_size_bits; + } + } } else if (ll > mft_ni->allocated_size) { err = -ENOSPC; goto undo_mftbmp_alloc_nolock; @@ -2276,6 +2489,12 @@ int ntfs_mft_record_alloc(struct ntfs_volume *vol, const int mode, mark_mft_record_dirty(ctx->ntfs_ino); ntfs_attr_put_search_ctx(ctx); unmap_mft_record(mft_ni); + if (reserve_start >= 0 && reserve_end > reserve_start) { + vol->mft_record_reserve_pos = reserve_start; + vol->mft_record_reserve_end = reserve_end; + ntfs_debug("Reserved MFT records [0x%llx, 0x%llx) for metadata.", + reserve_start, reserve_end); + } read_lock_irqsave(&mft_ni->size_lock, flags); ntfs_debug("Status of mft data after mft record initialization: allocated_size 0x%llx, data_size 0x%llx, initialized_size 0x%llx.", mft_ni->allocated_size, i_size_read(vol->mft_ino), @@ -2315,8 +2534,8 @@ int ntfs_mft_record_alloc(struct ntfs_volume *vol, const int mode, /* If we just formatted the mft record no need to do it again. */ if (!record_formatted) { /* Sanity check that the mft record is really not in use. */ - if (ntfs_is_file_record(m->magic) && - (m->flags & MFT_RECORD_IN_USE)) { + if (!forced_reserved_record && ntfs_is_file_record(m->magic) && + (m->flags & MFT_RECORD_IN_USE)) { ntfs_warning(vol->sb, "Mft record 0x%llx was marked free in mft bitmap but is marked used itself. Unmount and run chkdsk.", bit); @@ -2389,9 +2608,13 @@ int ntfs_mft_record_alloc(struct ntfs_volume *vol, const int mode, ntfs_error(vol->sb, "Failed to map allocated extent mft record 0x%llx.", bit); err = PTR_ERR(m_tmp); - /* Set the mft record itself not in use. */ - m->flags &= cpu_to_le16( - ~le16_to_cpu(MFT_RECORD_IN_USE)); + if (forced_reserved_record) { + m->base_mft_record = 0; + m->flags |= MFT_RECORD_IN_USE; + } else { + /* Set the mft record itself not in use. */ + m->flags &= cpu_to_le16(~le16_to_cpu(MFT_RECORD_IN_USE)); + } /* Make sure the mft record is written out to disk. */ ntfs_mft_mark_dirty(folio); folio_unlock(folio); @@ -2463,7 +2686,8 @@ int ntfs_mft_record_alloc(struct ntfs_volume *vol, const int mode, (*ni)->mft_no = bit; if (ni_mrec) *ni_mrec = (*ni)->mrec; - ntfs_dec_free_mft_records(vol, 1); + if (!forced_reserved_record) + ntfs_dec_free_mft_records(vol, 1); return 0; undo_data_init: write_lock_irqsave(&mft_ni->size_lock, flags); @@ -2475,10 +2699,13 @@ int ntfs_mft_record_alloc(struct ntfs_volume *vol, const int mode, if (!base_ni || base_ni->mft_no != FILE_MFT) down_write(&vol->mftbmp_lock); undo_mftbmp_alloc_nolock: - if (ntfs_bitmap_clear_bit(vol->mftbmp_ino, bit)) { + if (!forced_reserved_record && ntfs_bitmap_clear_bit(vol->mftbmp_ino, bit)) { ntfs_error(vol->sb, "Failed to clear bit in mft bitmap.%s", es); NVolSetErrors(vol); } + if ((from_reserve || reserve_created) && + vol->mft_record_reserve_pos == bit + 1) + vol->mft_record_reserve_pos = bit; if (!base_ni || base_ni->mft_no != FILE_MFT) up_write(&vol->mftbmp_lock); err_out: @@ -2514,9 +2741,11 @@ int ntfs_mft_record_free(struct ntfs_volume *vol, struct ntfs_inode *ni) int err; u16 seq_no; __le16 old_seq_no; + __le64 old_base_mft_record; struct mft_record *ni_mrec; unsigned int memalloc_flags; struct ntfs_inode *base_ni; + bool keep_reserved; if (!vol || !ni) return -EINVAL; @@ -2529,9 +2758,23 @@ int ntfs_mft_record_free(struct ntfs_volume *vol, struct ntfs_inode *ni) /* Cache the mft reference for later. */ mft_no = ni->mft_no; + if (likely(ni->nr_extents >= 0)) + base_ni = ni; + else + base_ni = ni->ext.base_ntfs_ino; + keep_reserved = mft_no >= FILE_reserved12 && + mft_no <= FILE_reserved15 && + base_ni->mft_no == FILE_MFT; - /* Mark the mft record as not in use. */ - ni_mrec->flags &= ~MFT_RECORD_IN_USE; + old_base_mft_record = ni_mrec->base_mft_record; + if (keep_reserved) { + /* Restore the special, unnamed form used by reserved records. */ + ni_mrec->base_mft_record = 0; + ni_mrec->flags |= MFT_RECORD_IN_USE; + } else { + /* Mark the mft record as not in use. */ + ni_mrec->flags &= ~MFT_RECORD_IN_USE; + } /* Increment the sequence number, skipping zero, if it is not zero. */ old_seq_no = ni_mrec->sequence_number; @@ -2560,16 +2803,20 @@ int ntfs_mft_record_free(struct ntfs_volume *vol, struct ntfs_inode *ni) if (err) goto sync_rollback; - if (likely(ni->nr_extents >= 0)) - base_ni = ni; - else - base_ni = ni->ext.base_ntfs_ino; + if (keep_reserved) { + unmap_mft_record(ni); + return 0; + } /* Clear the bit in the $MFT/$BITMAP corresponding to this record. */ memalloc_flags = memalloc_nofs_save(); if (base_ni->mft_no != FILE_MFT) down_write(&vol->mftbmp_lock); err = ntfs_bitmap_clear_bit(vol->mftbmp_ino, mft_no); + if (!err && base_ni->mft_no == FILE_MFT && + mft_no + 1 == vol->mft_record_reserve_pos && + mft_no < vol->mft_record_reserve_end) + vol->mft_record_reserve_pos = mft_no; if (base_ni->mft_no != FILE_MFT) up_write(&vol->mftbmp_lock); memalloc_nofs_restore(memalloc_flags); @@ -2595,6 +2842,7 @@ int ntfs_mft_record_free(struct ntfs_volume *vol, struct ntfs_inode *ni) "Eeek! Rollback failed in %s. Leaving inconsistent metadata!\n", __func__); ni_mrec->flags |= MFT_RECORD_IN_USE; ni_mrec->sequence_number = old_seq_no; + ni_mrec->base_mft_record = old_base_mft_record; NInoSetDirty(ni); write_mft_record(ni, ni_mrec, 0); unmap_mft_record(ni); diff --git a/fs/ntfs/mft.h b/fs/ntfs/mft.h index 75a51a98d0f6..ed5c1d595c0d 100644 --- a/fs/ntfs/mft.h +++ b/fs/ntfs/mft.h @@ -78,7 +78,7 @@ static inline int write_mft_record(struct ntfs_inode *ni, struct mft_record *m, int ntfs_mft_record_alloc(struct ntfs_volume *vol, const int mode, struct ntfs_inode **ni, struct ntfs_inode *base_ni, - struct mft_record **ni_mrec); + struct mft_record **ni_mrec, const s64 mft_data_vcn); int ntfs_mft_record_free(struct ntfs_volume *vol, struct ntfs_inode *ni); int ntfs_mft_records_write(const struct ntfs_volume *vol, const u64 mref, const s64 count, struct mft_record *b); diff --git a/fs/ntfs/namei.c b/fs/ntfs/namei.c index 7091b2496fac..fdf52fac4329 100644 --- a/fs/ntfs/namei.c +++ b/fs/ntfs/namei.c @@ -480,7 +480,7 @@ static struct ntfs_inode *__ntfs_create(struct mnt_idmap *idmap, struct inode *d mark_inode_dirty(dir); err = ntfs_mft_record_alloc(dir_ni->vol, mode, &ni, NULL, - &ni_mrec); + &ni_mrec, -1); if (err) { iput(vi); return ERR_PTR(err); diff --git a/fs/ntfs/volume.h b/fs/ntfs/volume.h index 65fd3908af26..2b60d14fc7ef 100644 --- a/fs/ntfs/volume.h +++ b/fs/ntfs/volume.h @@ -55,6 +55,10 @@ * @attrdef_size: Size of the attribute definition table in bytes. * @attrdef: Table of attribute definitions. Obtained from FILE_AttrDef. * @mft_data_pos: Mft record number at which to allocate the next mft record. + * @mft_record_reserve_pos: First record in the in-memory MFT metadata reserve + * (protected by mftbmp_lock). + * @mft_record_reserve_end: First record beyond the MFT metadata reserve + * (protected by mftbmp_lock). * @mft_zone_start: First cluster of the mft zone. * @mft_zone_end: First cluster beyond the mft zone. * @mft_zone_pos: Current position in the mft zone. @@ -119,6 +123,8 @@ struct ntfs_volume { s32 attrdef_size; struct attr_def *attrdef; s64 mft_data_pos; + s64 mft_record_reserve_pos; + s64 mft_record_reserve_end; s64 mft_zone_start; s64 mft_zone_end; s64 mft_zone_pos; From b1d732e62a5b3942546e4edaab8976258e779287 Mon Sep 17 00:00:00 2001 From: Namjae Jeon Date: Sun, 6 Sep 2026 11:12:47 +0900 Subject: [PATCH 139/592] ntfs: repack $MFT/$ATTRIBUTE LIST Repack the non-resident $MFT/$ATTRIBUTE_LIST into a contiguous run when its mapping pairs no longer fit in the base MFT record. Propagate allocation and writeback errors, and check synchronous replacement writes. Fixes: 495e90fa3348 ("ntfs: update attrib operations") Reviewed-by: Hyunchul Lee Signed-off-by: Namjae Jeon --- fs/ntfs/attrib.c | 47 ++++++++--- fs/ntfs/attrlist.c | 202 ++++++++++++++++++++++++++++++++++++++++++--- fs/ntfs/inode.c | 10 ++- 3 files changed, 236 insertions(+), 23 deletions(-) diff --git a/fs/ntfs/attrib.c b/fs/ntfs/attrib.c index 81b9e0971b3e..d1696ce6acd5 100644 --- a/fs/ntfs/attrib.c +++ b/fs/ntfs/attrib.c @@ -3844,13 +3844,13 @@ int ntfs_attr_update_mapping_pairs(struct ntfs_inode *ni, s64 from_vcn) */ if (ni->type == AT_ATTRIBUTE_LIST) { ntfs_attr_put_search_ctx(ctx); - if (ntfs_inode_free_space(base_ni, mp_size - - cur_max_mp_size)) { - ntfs_debug("Attribute list is too big. Defragment the volume\n"); - return -ENOSPC; - } - if (ntfs_attrlist_update(base_ni)) - return -EIO; + err = ntfs_inode_free_space(base_ni, mp_size - + cur_max_mp_size); + if (err) + return err; + err = ntfs_attrlist_update(base_ni); + if (err) + return err; goto retry; } @@ -4522,13 +4522,39 @@ static int ntfs_non_resident_attr_expand(struct ntfs_inode *ni, const s64 newsiz ntfs_bytes_to_cluster(vol, ni->allocated_size), first_free_vcn - ntfs_bytes_to_cluster(vol, ni->allocated_size), - lcn_seek_from, DATA_ZONE, false, false, false); + lcn_seek_from, DATA_ZONE, false, + ni->type == AT_ATTRIBUTE_LIST, false); if (IS_ERR(rl)) { ntfs_debug("Cluster allocation failed (%lld)", (long long)first_free_vcn - ntfs_bytes_to_cluster(vol, ni->allocated_size)); return PTR_ERR(rl); } + /* + * A contiguous ATTRIBUTE_LIST allocation keeps its mapping + * pairs small enough to fit in the base MFT record. The + * allocator can return a short run when contiguity was + * requested, so discard it and retry normally if necessary. + */ + if (ni->type == AT_ATTRIBUTE_LIST && + (rl->vcn != ntfs_bytes_to_cluster(vol, + ni->allocated_size) || + rl->length != first_free_vcn - + ntfs_bytes_to_cluster(vol, ni->allocated_size) || + rl[1].length)) { + ntfs_cluster_free_from_rl(vol, rl); + kvfree(rl); + rl = ntfs_cluster_alloc(vol, + ntfs_bytes_to_cluster(vol, + ni->allocated_size), + first_free_vcn - + ntfs_bytes_to_cluster(vol, + ni->allocated_size), + lcn_seek_from, DATA_ZONE, false, + false, false); + if (IS_ERR(rl)) + return PTR_ERR(rl); + } } if (!NInoCompressed(ni)) { @@ -4921,7 +4947,8 @@ int ntfs_attr_expand(struct ntfs_inode *ni, const s64 newsize, const s64 preallo ntfs_debug("Entering for inode 0x%llx, attr 0x%x, size %lld\n", (unsigned long long)ni->mft_no, ni->type, newsize); - if (ni->data_size == newsize) { + if (ni->data_size == newsize && + (!prealloc_size || prealloc_size <= ni->allocated_size)) { ntfs_debug("Size is already ok\n"); return 0; } @@ -4936,7 +4963,7 @@ int ntfs_attr_expand(struct ntfs_inode *ni, const s64 newsize, const s64 preallo } if (NInoNonResident(ni)) { - if (newsize > ni->data_size) + if (newsize > ni->data_size || prealloc_size > ni->allocated_size) err = ntfs_non_resident_attr_expand(ni, newsize, prealloc_size, NVolDisableSparse(ni->vol) ? HOLES_NO : HOLES_OK, true); diff --git a/fs/ntfs/attrlist.c b/fs/ntfs/attrlist.c index be3086d34338..4e60f7e93010 100644 --- a/fs/ntfs/attrlist.c +++ b/fs/ntfs/attrlist.c @@ -12,6 +12,9 @@ #include "mft.h" #include "attrib.h" #include "attrlist.h" +#include "lcnalloc.h" + +#define NTFS_MAX_ATTR_LIST_SIZE (256 * 1024) /* * ntfs_attrlist_need - check whether inode need attribute list @@ -51,11 +54,151 @@ int ntfs_attrlist_need(struct ntfs_inode *ni) return 0; } +/* + * Repack the $MFT/$ATTRIBUTE_LIST data into one run. + * + * The mapping pairs for an $ATTRIBUTE_LIST must remain in the base MFT + * record. Once that record has no room left, extending a fragmented list + * can require one more mapping-pairs byte than the record can hold. There + * is no attribute that can legally be moved out in that state: $STANDARD_ + * INFORMATION, $ATTRIBUTE_LIST, and the first $MFT/$DATA extent all have to + * stay in the base record. Move the list data to one contiguous run. The + * caller supplies the minimum allocation size so a recovery can use the + * smallest useful run while normal updates can still request the maximum + * legal list size as a reserve. + */ +static int ntfs_attrlist_repack(struct inode *attr_vi, + struct ntfs_inode *attr_ni, s64 min_alloc_size) +{ + struct ntfs_volume *vol = attr_ni->vol; + struct runlist_element *old_rl, *new_rl; + u8 *data = NULL; + s64 data_size, alloc_size, nr_clusters, written; + s64 old_alloc_size; + size_t old_rl_count, new_rl_count; + unsigned long flags; + int err, restore_err; + + if (attr_ni->mft_no != FILE_MFT || !NInoNonResident(attr_ni) || + min_alloc_size < 0) + return -EINVAL; + + err = ntfs_attr_map_whole_runlist(attr_ni); + if (err) + return err; + + data_size = attr_ni->data_size; + if (data_size < 0) + return -EIO; + + if (data_size) { + data = kvmalloc(data_size, GFP_NOFS); + if (!data) + return -ENOMEM; + + written = ntfs_inode_attr_pread(attr_vi, 0, data_size, data); + if (written != data_size) { + err = written < 0 ? (int)written : -EIO; + goto out_free_data; + } + } + + old_alloc_size = attr_ni->allocated_size; + alloc_size = max_t(s64, old_alloc_size, min_alloc_size); + nr_clusters = ntfs_bytes_to_cluster(vol, + alloc_size + vol->cluster_size - 1); + if (nr_clusters <= 0) { + err = -EFBIG; + goto out_free_data; + } + + /* A single run keeps the mapping pairs at the minimum size. */ + new_rl = ntfs_cluster_alloc(vol, 0, nr_clusters, -1, DATA_ZONE, + true, true, false); + if (IS_ERR(new_rl)) { + err = PTR_ERR(new_rl); + goto out_free_data; + } + + new_rl_count = 0; + if (new_rl->vcn == 0 && new_rl->length == nr_clusters && + !new_rl[1].length) + new_rl_count = 2; + + if (new_rl_count != 2) { + ntfs_cluster_free_from_rl(vol, new_rl); + kvfree(new_rl); + err = -ENOSPC; + goto out_free_data; + } + + old_rl = attr_ni->runlist.rl; + old_rl_count = attr_ni->runlist.count; + down_write(&attr_ni->runlist.lock); + attr_ni->runlist.rl = new_rl; + attr_ni->runlist.count = new_rl_count; + up_write(&attr_ni->runlist.lock); + + write_lock_irqsave(&attr_ni->size_lock, flags); + attr_ni->allocated_size = ntfs_cluster_to_bytes(vol, nr_clusters); + write_unlock_irqrestore(&attr_ni->size_lock, flags); + + /* Populate the replacement extent before publishing its mapping pairs. */ + if (data_size) { + written = ntfs_inode_attr_pwrite(attr_vi, 0, data_size, data, true); + if (written != data_size) { + err = written < 0 ? (int)written : -EIO; + goto restore_old_runlist; + } + } + + err = ntfs_attr_update_mapping_pairs(attr_ni, 0); + if (err) + goto restore_old_runlist; + + /* The new mapping is now authoritative; release the old data runs. */ + if (ntfs_cluster_free_from_rl(vol, old_rl)) { + ntfs_error(vol->sb, + "Failed to free old ATTRIBUTE_LIST extent: inode %#llx", + (long long)attr_ni->mft_no); + NVolSetErrors(vol); + } + kvfree(old_rl); + kvfree(data); + return 0; + +restore_old_runlist: + down_write(&attr_ni->runlist.lock); + attr_ni->runlist.rl = old_rl; + attr_ni->runlist.count = old_rl_count; + up_write(&attr_ni->runlist.lock); + + write_lock_irqsave(&attr_ni->size_lock, flags); + attr_ni->allocated_size = old_alloc_size; + write_unlock_irqrestore(&attr_ni->size_lock, flags); + + restore_err = ntfs_attr_update_mapping_pairs(attr_ni, 0); + if (restore_err) { + ntfs_error(vol->sb, "Failed to restore ATTRIBUTE_LIST mapping pairs (%d)", + restore_err); + NVolSetErrors(vol); + } + + ntfs_cluster_free_from_rl(vol, new_rl); + kvfree(new_rl); + err = err ? err : restore_err; + +out_free_data: + kvfree(data); + return err; +} + int ntfs_attrlist_update(struct ntfs_inode *base_ni) { struct inode *attr_vi; struct ntfs_inode *attr_ni; - int err; + s64 written; + int err, retry_err; /* * generic_shutdown_super() clears SB_ACTIVE before evicting cached @@ -74,21 +217,55 @@ int ntfs_attrlist_update(struct ntfs_inode *base_ni) attr_ni = NTFS_I(attr_vi); err = ntfs_attr_truncate_i(attr_ni, base_ni->attr_list_size, HOLES_NO); - if (err == -ENOSPC && attr_ni->mft_no == FILE_MFT) { - err = ntfs_attr_truncate(attr_ni, 0); - if (err || ntfs_attr_truncate_i(attr_ni, base_ni->attr_list_size, HOLES_NO) != 0) { + if (err == -ENOSPC && attr_ni->mft_no == FILE_MFT && + NInoNonResident(attr_ni)) { + retry_err = ntfs_attrlist_repack(attr_vi, attr_ni, + base_ni->attr_list_size); + if (retry_err) { + ntfs_error(base_ni->vol->sb, "Failed to repack attribute list"); iput(attr_vi); + return retry_err; + } + + retry_err = ntfs_attr_truncate_i(attr_ni, base_ni->attr_list_size, + HOLES_NO); + if (retry_err) { ntfs_error(base_ni->vol->sb, - "Failed to truncate attribute list of inode %#llx", - (long long)base_ni->mft_no); - return -EIO; + "Failed to resize attribute list after repack"); + iput(attr_vi); + return retry_err; } } else if (err) { iput(attr_vi); ntfs_error(base_ni->vol->sb, "Failed to truncate attribute list of inode %#llx", (long long)base_ni->mft_no); - return -EIO; + return err; + } + + /* + * Reserve the maximum legal list size while the MFT metadata area is + * still easy to allocate contiguously. This prevents a later list entry + * from needing another mapping-pairs byte in the full base MFT record. + * Failure to obtain the optional reserve must not reject the current + * metadata update; the repack retry above remains available if needed. + */ + if (base_ni->mft_no == FILE_MFT && NInoNonResident(attr_ni) && + attr_ni->allocated_size < NTFS_MAX_ATTR_LIST_SIZE) { + retry_err = ntfs_attr_expand(attr_ni, base_ni->attr_list_size, + NTFS_MAX_ATTR_LIST_SIZE); + if (retry_err == -ENOSPC) { + retry_err = ntfs_attrlist_repack(attr_vi, attr_ni, + NTFS_MAX_ATTR_LIST_SIZE); + if (retry_err == -ENOSPC) + retry_err = 0; + } + if (retry_err) { + ntfs_error(base_ni->vol->sb, + "Failed to reserve attribute list space"); + iput(attr_vi); + return retry_err; + } } i_size_write(attr_vi, base_ni->attr_list_size); @@ -96,14 +273,15 @@ int ntfs_attrlist_update(struct ntfs_inode *base_ni) if (NInoNonResident(attr_ni) && !NInoAttrListNonResident(base_ni)) NInoSetAttrListNonResident(base_ni); - if (ntfs_inode_attr_pwrite(attr_vi, 0, base_ni->attr_list_size, - base_ni->attr_list, false) != - base_ni->attr_list_size) { + written = ntfs_inode_attr_pwrite(attr_vi, 0, base_ni->attr_list_size, + base_ni->attr_list, false); + if (written != base_ni->attr_list_size) { + err = written < 0 ? (int)written : -EIO; iput(attr_vi); ntfs_error(base_ni->vol->sb, "Failed to write attribute list of inode %#llx", (long long)base_ni->mft_no); - return -EIO; + return err; } NInoSetAttrListDirty(base_ni); diff --git a/fs/ntfs/inode.c b/fs/ntfs/inode.c index 5aedc045f65a..332825db477a 100644 --- a/fs/ntfs/inode.c +++ b/fs/ntfs/inode.c @@ -3745,6 +3745,7 @@ static s64 __ntfs_inode_non_resident_attr_pwrite(struct inode *vi, u64 rl_length = 0; s64 vcn; struct runlist_element *rl; + int bio_err; lcn_count = max_t(s64, 1, ntfs_bytes_to_cluster(vol, attr_len)); vcn = ntfs_pidx_to_cluster(vol, folio->index); @@ -3787,8 +3788,15 @@ static s64 __ntfs_inode_non_resident_attr_pwrite(struct inode *vi, goto err_unlock_folio; } - submit_bio_wait(bio); + bio_err = submit_bio_wait(bio); bio_put(bio); + if (bio_err) { + ntfs_error(vi->i_sb, + "Synchronous attribute write failed (%d)", + bio_err); + ret = bio_err; + goto err_unlock_folio; + } vcn += rl_length; offset += length; } while (lcn_count != 0); From 631946431ddc66a472c5cc629cd654e62dfa1f88 Mon Sep 17 00:00:00 2001 From: Namjae Jeon Date: Fri, 4 Sep 2026 14:46:38 +0900 Subject: [PATCH 140/592] ntfs: account for MFT records added during allocation When no free MFT record is available in the initialized $MFT/$BITMAP, ntfs_mft_record_alloc() extends $MFT/$DATA and formats the requested record together with a dynamically sized tail reserve. Those records become visible through the $MFT file size before charging the requested record to the free-record counter. Account for all newly visible records before releasing the MFT allocation lock, then subtract the one record being allocated. Keep MFT counter updates independent of the asynchronous free-cluster scan and update the counter when a record is successfully cleared in the MFT bitmap. Store the clamped result of the MFT bitmap scan and keep statfs from exposing an invalid cached count if an accounting error occurs. Fixes: 115380f9a2f9 ("ntfs: update mft operations") Reviewed-by: Hyunchul Lee Signed-off-by: Namjae Jeon --- fs/ntfs/mft.c | 11 ++++++++--- fs/ntfs/super.c | 12 +++++++++--- fs/ntfs/volume.h | 6 ------ 3 files changed, 17 insertions(+), 12 deletions(-) diff --git a/fs/ntfs/mft.c b/fs/ntfs/mft.c index 61d4ee3a57fd..f0656d0bbeeb 100644 --- a/fs/ntfs/mft.c +++ b/fs/ntfs/mft.c @@ -1539,7 +1539,6 @@ static int ntfs_mft_bitmap_extend_initialized_nolock(struct ntfs_volume *vol) ret = ntfs_attr_set(mftbmp_ni, old_initialized_size, 8, 0); if (likely(!ret)) { ntfs_debug("Done. (Wrote eight initialized bytes to mft bitmap."); - ntfs_inc_free_mft_records(vol, 8 * 8); return 0; } ntfs_error(vol->sb, "Failed to write to mft bitmap."); @@ -2135,6 +2134,7 @@ int ntfs_mft_record_alloc(struct ntfs_volume *vol, const int mode, struct mft_record **ni_mrec, const s64 mft_data_vcn) { s64 ll, bit, old_data_initialized, old_data_size; + s64 nr_new_mft_records = 0; s64 max_mft_no = -1, reserve_start = -1, reserve_end = -1; s64 candidate_reserve_end = -1; s64 *reserve_endp; @@ -2501,8 +2501,13 @@ int ntfs_mft_record_alloc(struct ntfs_volume *vol, const int mode, mft_ni->initialized_size); WARN_ON(i_size_read(vol->mft_ino) > mft_ni->allocated_size); WARN_ON(mft_ni->initialized_size > i_size_read(vol->mft_ino)); + nr_new_mft_records = (i_size_read(vol->mft_ino) - old_data_size) >> + vol->mft_record_size_bits; read_unlock_irqrestore(&mft_ni->size_lock, flags); mft_rec_already_initialized: + /* Account for newly visible MFT records before dropping the lock. */ + if (nr_new_mft_records > 0) + ntfs_inc_free_mft_records(vol, nr_new_mft_records); /* * We can finally drop the mft bitmap lock as the mft data attribute * has been fully updated. The only disparity left is that the @@ -2813,6 +2818,8 @@ int ntfs_mft_record_free(struct ntfs_volume *vol, struct ntfs_inode *ni) if (base_ni->mft_no != FILE_MFT) down_write(&vol->mftbmp_lock); err = ntfs_bitmap_clear_bit(vol->mftbmp_ino, mft_no); + if (!err) + ntfs_inc_free_mft_records(vol, 1); if (!err && base_ni->mft_no == FILE_MFT && mft_no + 1 == vol->mft_record_reserve_pos && mft_no < vol->mft_record_reserve_end) @@ -2822,9 +2829,7 @@ int ntfs_mft_record_free(struct ntfs_volume *vol, struct ntfs_inode *ni) memalloc_nofs_restore(memalloc_flags); if (err) goto bitmap_rollback; - unmap_mft_record(ni); - ntfs_inc_free_mft_records(vol, 1); return 0; /* Rollback what we did... */ diff --git a/fs/ntfs/super.c b/fs/ntfs/super.c index 60d43339c590..0867d6a82ebe 100644 --- a/fs/ntfs/super.c +++ b/fs/ntfs/super.c @@ -2064,8 +2064,7 @@ static unsigned long __get_nr_free_mft_records(struct ntfs_volume *vol, /* If errors occurred we may well have gone below zero, fix this. */ if (nr_free < 0) nr_free = 0; - else - atomic64_set(&vol->free_mft_records, nr_free); + atomic64_set(&vol->free_mft_records, nr_free); ntfs_debug("Exiting."); return nr_free; @@ -2131,7 +2130,14 @@ static int ntfs_statfs(struct dentry *dentry, struct kstatfs *sfs) read_unlock_irqrestore(&mft_ni->size_lock, flags); /* Free inodes in fs (based on current total count). */ - sfs->f_ffree = atomic64_read(&vol->free_mft_records); + size = atomic64_read(&vol->free_mft_records); + if (unlikely(size < 0 || size > (s64)sfs->f_files)) + ntfs_warning(vol->sb, "Invalid free MFT record count %lld.", size); + if (size < 0) + size = 0; + else if (size > (s64)sfs->f_files) + size = sfs->f_files; + sfs->f_ffree = size; /* * File system id. This is extremely *nix flavour dependent and even diff --git a/fs/ntfs/volume.h b/fs/ntfs/volume.h index 2b60d14fc7ef..bc85a9592245 100644 --- a/fs/ntfs/volume.h +++ b/fs/ntfs/volume.h @@ -258,17 +258,11 @@ static inline void ntfs_dec_free_clusters(struct ntfs_volume *vol, s64 nr) static inline void ntfs_inc_free_mft_records(struct ntfs_volume *vol, s64 nr) { - if (!NVolFreeClusterKnown(vol)) - return; - atomic64_add(nr, &vol->free_mft_records); } static inline void ntfs_dec_free_mft_records(struct ntfs_volume *vol, s64 nr) { - if (!NVolFreeClusterKnown(vol)) - return; - atomic64_sub(nr, &vol->free_mft_records); } From 91709ba5d6d709b2b663287b7e871e2c6b480502 Mon Sep 17 00:00:00 2001 From: Namjae Jeon Date: Sun, 6 Sep 2026 21:37:28 +0900 Subject: [PATCH 141/592] ntfs: protect runlist updates with the runlist lock ntfs_non_resident_attr_shrink() calls runlist helpers that require the runlist write lock, but did not hold it while freeing clusters and truncating the runlist. Serialize those operations and the resident conversion with the runlist lock. ntfs_attr_map_cluster() can merge a newly allocated run before updating mapping pairs. If the update fails, free the clusters and restore both the in-memory runlist and on-disk mapping pairs from a saved runlist. Mark the volume in error if either rollback step fails. Fixes: 495e90fa3348 ("ntfs: update attrib operations") Reviewed-by: Hyunchul Lee Signed-off-by: Namjae Jeon --- fs/ntfs/attrib.c | 205 +++++++++++++++++++++++++++++++++++---------- fs/ntfs/attrib.h | 9 ++ fs/ntfs/attrlist.c | 46 ++++++---- fs/ntfs/attrlist.h | 2 + fs/ntfs/compress.c | 2 +- fs/ntfs/file.c | 3 +- fs/ntfs/inode.c | 2 +- fs/ntfs/mft.c | 13 +-- 8 files changed, 216 insertions(+), 66 deletions(-) diff --git a/fs/ntfs/attrib.c b/fs/ntfs/attrib.c index d1696ce6acd5..c949ff765075 100644 --- a/fs/ntfs/attrib.c +++ b/fs/ntfs/attrib.c @@ -3574,7 +3574,8 @@ int ntfs_attr_record_move_away(struct ntfs_attr_search_ctx *ctx, int extra) * update allocated and compressed size. */ static int ntfs_attr_update_meta(struct attr_record *a, struct ntfs_inode *ni, - struct mft_record *m, struct ntfs_attr_search_ctx *ctx) + struct mft_record *m, struct ntfs_attr_search_ctx *ctx, + struct ntfs_inode *locked_ni, bool defer_attrlist) { int sparse, err = 0; struct ntfs_inode *base_ni; @@ -3610,6 +3611,8 @@ static int ntfs_attr_update_meta(struct attr_record *a, struct ntfs_inode *ni, le16_to_cpu(a->data.non_resident.mapping_pairs_offset) == 8) && !(le32_to_cpu(m->bytes_allocated) - le32_to_cpu(m->bytes_in_use))) { + if (defer_attrlist) + return -ENOSPC; if (!NInoAttrList(base_ni)) { err = ntfs_inode_add_attrlist(base_ni); if (err) @@ -3623,7 +3626,7 @@ static int ntfs_attr_update_meta(struct attr_record *a, struct ntfs_inode *ni, goto out; } - err = ntfs_attrlist_update(base_ni); + err = ntfs_attrlist_update_locked(base_ni, locked_ni); if (err) goto out; err = -EAGAIN; @@ -3703,6 +3706,8 @@ static int ntfs_attr_update_meta(struct attr_record *a, struct ntfs_inode *ni, * ntfs_attr_update_mapping_pairs - update mapping pairs for ntfs attribute * @ni: non-resident ntfs inode for which we need update * @from_vcn: update runlist starting this VCN + * @locked_ni: inode whose runlist write lock is already held + * @defer_attrlist: return -ENOSPC instead of updating an attribute list * * Build mapping pairs from @na->rl and write them to the disk. Also, this * function updates sparse bit, allocated and compressed size (allocates/frees @@ -3712,7 +3717,10 @@ static int ntfs_attr_update_meta(struct attr_record *a, struct ntfs_inode *ni, * call to this function. Vice-versa @na->compressed_size will be calculated and * set to correct value during this function. */ -int ntfs_attr_update_mapping_pairs(struct ntfs_inode *ni, s64 from_vcn) +static int __ntfs_attr_update_mapping_pairs(struct ntfs_inode *ni, + s64 from_vcn, + struct ntfs_inode *locked_ni, + bool defer_attrlist) { struct ntfs_attr_search_ctx *ctx; struct ntfs_inode *base_ni; @@ -3804,7 +3812,8 @@ int ntfs_attr_update_mapping_pairs(struct ntfs_inode *ni, s64 from_vcn) continue; } - err = ntfs_attr_update_meta(a, ni, m, ctx); + err = ntfs_attr_update_meta(a, ni, m, ctx, locked_ni, + defer_attrlist); if (err < 0) { if (err == -EAGAIN) { ntfs_attr_put_search_ctx(ctx); @@ -3844,11 +3853,17 @@ int ntfs_attr_update_mapping_pairs(struct ntfs_inode *ni, s64 from_vcn) */ if (ni->type == AT_ATTRIBUTE_LIST) { ntfs_attr_put_search_ctx(ctx); + ctx = NULL; + if (locked_ni == ni || defer_attrlist) { + err = -ENOSPC; + goto put_err_out; + } err = ntfs_inode_free_space(base_ni, mp_size - cur_max_mp_size); if (err) return err; - err = ntfs_attrlist_update(base_ni); + err = ntfs_attrlist_update_locked( + base_ni, locked_ni); if (err) return err; goto retry; @@ -3856,6 +3871,10 @@ int ntfs_attr_update_mapping_pairs(struct ntfs_inode *ni, s64 from_vcn) /* Add attribute list if it isn't present, and retry. */ if (!NInoAttrList(base_ni)) { + if (defer_attrlist) { + err = -ENOSPC; + goto put_err_out; + } ntfs_attr_put_search_ctx(ctx); if (ntfs_inode_add_attrlist(base_ni)) { ntfs_error(sb, "Can not add attrlist"); @@ -3883,13 +3902,21 @@ int ntfs_attr_update_mapping_pairs(struct ntfs_inode *ni, s64 from_vcn) } } + if (defer_attrlist && + (ctx->ntfs_ino->nr_extents == -1 || + NInoAttrList(ctx->ntfs_ino)) && + ctx->attr->type != AT_ATTRIBUTE_LIST) { + err = -ENOSPC; + goto put_err_out; + } + /* Update lowest vcn. */ a->data.non_resident.lowest_vcn = cpu_to_le64(stop_vcn); mark_mft_record_dirty(ctx->ntfs_ino); if ((ctx->ntfs_ino->nr_extents == -1 || NInoAttrList(ctx->ntfs_ino)) && ctx->attr->type != AT_ATTRIBUTE_LIST) { ctx->al_entry->lowest_vcn = cpu_to_le64(stop_vcn); - err = ntfs_attrlist_update(base_ni); + err = ntfs_attrlist_update_locked(base_ni, locked_ni); if (err) goto put_err_out; } @@ -4064,6 +4091,19 @@ int ntfs_attr_update_mapping_pairs(struct ntfs_inode *ni, s64 from_vcn) return err; } +int ntfs_attr_update_mapping_pairs_locked(struct ntfs_inode *ni, + s64 from_vcn, + struct ntfs_inode *locked_ni) +{ + return __ntfs_attr_update_mapping_pairs(ni, from_vcn, locked_ni, + false); +} + +int ntfs_attr_update_mapping_pairs(struct ntfs_inode *ni, s64 from_vcn) +{ + return ntfs_attr_update_mapping_pairs_locked(ni, from_vcn, NULL); +} + /* * ntfs_attr_make_resident - convert a non-resident to a resident attribute * @ni: open ntfs attribute to make resident @@ -4197,7 +4237,9 @@ static int ntfs_attr_make_resident(struct ntfs_inode *ni, struct ntfs_attr_searc * * Reduce the size of a non-resident, open ntfs attribute @na to @newsize bytes. */ -static int ntfs_non_resident_attr_shrink(struct ntfs_inode *ni, const s64 newsize) +static int ntfs_non_resident_attr_shrink(struct ntfs_inode *ni, + const s64 newsize, + struct ntfs_inode *locked_ni) { struct ntfs_volume *vol; struct ntfs_attr_search_ctx *ctx; @@ -4205,6 +4247,7 @@ static int ntfs_non_resident_attr_shrink(struct ntfs_inode *ni, const s64 newsiz s64 nr_freed_clusters; int err; struct ntfs_inode *base_ni; + bool runlist_locked = locked_ni == ni; ntfs_debug("Inode 0x%llx attr 0x%x new size %lld\n", (unsigned long long)ni->mft_no, ni->type, (long long)newsize); @@ -4250,18 +4293,24 @@ static int ntfs_non_resident_attr_shrink(struct ntfs_inode *ni, const s64 newsiz * clusters if there is a change. */ if (ntfs_bytes_to_cluster(vol, ni->allocated_size) != first_free_vcn) { - struct ntfs_attr_search_ctx *ctx; + /* + * ntfs_cluster_free() and ntfs_rl_truncate_nolock() + * both require this lock. + */ + if (!runlist_locked) + down_write(&ni->runlist.lock); err = ntfs_attr_map_whole_runlist(ni); if (err) { ntfs_debug("Eeek! ntfs_attr_map_whole_runlist failed.\n"); - return err; + goto unlock_runlist; } ctx = ntfs_attr_get_search_ctx(ni, NULL); if (!ctx) { ntfs_error(vol->sb, "%s: Failed to get search context", __func__); - return -ENOMEM; + err = -ENOMEM; + goto unlock_runlist; } /* Deallocate all clusters starting with the first free one. */ @@ -4269,7 +4318,8 @@ static int ntfs_non_resident_attr_shrink(struct ntfs_inode *ni, const s64 newsiz if (nr_freed_clusters < 0) { ntfs_debug("Eeek! Freeing of clusters failed. Aborting...\n"); ntfs_attr_put_search_ctx(ctx); - return (int)nr_freed_clusters; + err = (int)nr_freed_clusters; + goto unlock_runlist; } ntfs_attr_put_search_ctx(ctx); @@ -4282,7 +4332,8 @@ static int ntfs_non_resident_attr_shrink(struct ntfs_inode *ni, const s64 newsiz kvfree(ni->runlist.rl); ni->runlist.rl = NULL; ntfs_error(vol->sb, "Eeek! Run list truncation failed.\n"); - return -EIO; + err = -EIO; + goto unlock_runlist; } /* Prepare to mapping pairs update. */ @@ -4298,11 +4349,13 @@ static int ntfs_non_resident_attr_shrink(struct ntfs_inode *ni, const s64 newsiz VFS_I(base_ni)->i_blocks = ni->allocated_size >> 9; /* Write mapping pairs for new runlist. */ - err = ntfs_attr_update_mapping_pairs(ni, 0 /*first_free_vcn*/); + err = ntfs_attr_update_mapping_pairs_locked(ni, 0, ni); if (err) { ntfs_debug("Eeek! Mapping pairs update failed. Leaving inconstant metadata. Run chkdsk.\n"); - return err; + goto unlock_runlist; } + if (!runlist_locked) + up_write(&ni->runlist.lock); } /* Get the first attribute record. */ @@ -4344,7 +4397,11 @@ static int ntfs_non_resident_attr_shrink(struct ntfs_inode *ni, const s64 newsiz /* If the attribute now has zero size, make it resident. */ if (!newsize && !NInoEncrypted(ni) && !NInoCompressed(ni)) { + if (!runlist_locked) + down_write(&ni->runlist.lock); err = ntfs_attr_make_resident(ni, ctx); + if (!runlist_locked) + up_write(&ni->runlist.lock); if (err) { /* If couldn't make resident, just continue. */ if (err != -EPERM) @@ -4361,6 +4418,11 @@ static int ntfs_non_resident_attr_shrink(struct ntfs_inode *ni, const s64 newsiz put_err_out: ntfs_attr_put_search_ctx(ctx); return err; + +unlock_runlist: + if (!runlist_locked) + up_write(&ni->runlist.lock); + return err; } /* @@ -4369,13 +4431,14 @@ static int ntfs_non_resident_attr_shrink(struct ntfs_inode *ni, const s64 newsiz * @prealloc_size: preallocation size (in bytes) to which to expand the attribute * @newsize: new size (in bytes) to which to expand the attribute * @holes: how to create a hole if expanding - * @need_lock: whether mrec lock is needed or not + * @locked_ni: inode whose runlist lock is already held * * Expand the size of a non-resident, open ntfs attribute @na to @newsize bytes, * by allocating new clusters. */ static int ntfs_non_resident_attr_expand(struct ntfs_inode *ni, const s64 newsize, - const s64 prealloc_size, unsigned int holes, bool need_lock) + const s64 prealloc_size, unsigned int holes, + struct ntfs_inode *locked_ni) { s64 lcn_seek_from; s64 first_free_vcn; @@ -4573,7 +4636,8 @@ static int ntfs_non_resident_attr_expand(struct ntfs_inode *ni, const s64 newsiz /* Prepare to mapping pairs update. */ ni->allocated_size = ntfs_cluster_to_bytes(vol, first_free_vcn); - err = ntfs_attr_update_mapping_pairs(ni, 0); + err = ntfs_attr_update_mapping_pairs_locked( + ni, 0, locked_ni); if (err) { ntfs_debug("Mapping pairs update failed"); goto rollback; @@ -4617,11 +4681,11 @@ static int ntfs_non_resident_attr_expand(struct ntfs_inode *ni, const s64 newsiz ntfs_debug("Leaking clusters"); /* Now, truncate the runlist itself. */ - if (need_lock) + if (ni != locked_ni) down_write(&ni->runlist.lock); err2 = ntfs_rl_truncate_nolock(vol, &ni->runlist, ntfs_bytes_to_cluster(vol, org_alloc_size)); - if (need_lock) + if (ni != locked_ni) up_write(&ni->runlist.lock); if (err2) { /* @@ -4635,11 +4699,11 @@ static int ntfs_non_resident_attr_expand(struct ntfs_inode *ni, const s64 newsiz /* Prepare to mapping pairs update. */ ni->allocated_size = org_alloc_size; /* Restore mapping pairs. */ - if (need_lock) + if (ni != locked_ni) down_read(&ni->runlist.lock); - if (ntfs_attr_update_mapping_pairs(ni, 0)) + if (__ntfs_attr_update_mapping_pairs(ni, 0, locked_ni, true)) ntfs_error(sb, "Failed to restore old mapping pairs"); - if (need_lock) + if (ni != locked_ni) up_read(&ni->runlist.lock); if (NInoSparse(ni) || NInoCompressed(ni)) { @@ -4744,7 +4808,8 @@ static int ntfs_resident_attr_resize(struct ntfs_inode *attr_ni, const s64 newsi mark_mft_record_dirty(ctx->ntfs_ino); ntfs_attr_put_search_ctx(ctx); /* Resize non-resident attribute */ - return ntfs_non_resident_attr_expand(attr_ni, newsize, prealloc_size, holes, true); + return ntfs_non_resident_attr_expand( + attr_ni, newsize, prealloc_size, holes, NULL); } else if (err != -ENOSPC && err != -EPERM) { ntfs_error(sb, "Failed to make attribute non-resident"); goto put_err_out; @@ -4919,13 +4984,14 @@ int __ntfs_attr_truncate_vfs(struct ntfs_inode *ni, const s64 newsize, if (NInoNonResident(ni)) { if (newsize > i_size) { down_write(&ni->runlist.lock); - err = ntfs_non_resident_attr_expand(ni, newsize, 0, - NVolDisableSparse(ni->vol) ? - HOLES_NO : HOLES_OK, - false); + err = ntfs_non_resident_attr_expand( + ni, newsize, 0, + NVolDisableSparse(ni->vol) ? + HOLES_NO : HOLES_OK, ni); up_write(&ni->runlist.lock); } else - err = ntfs_non_resident_attr_shrink(ni, newsize); + err = ntfs_non_resident_attr_shrink( + ni, newsize, NULL); } else err = ntfs_resident_attr_resize(ni, newsize, 0, NVolDisableSparse(ni->vol) ? @@ -4934,7 +5000,9 @@ int __ntfs_attr_truncate_vfs(struct ntfs_inode *ni, const s64 newsize, return err; } -int ntfs_attr_expand(struct ntfs_inode *ni, const s64 newsize, const s64 prealloc_size) +int ntfs_attr_expand_locked(struct ntfs_inode *ni, const s64 newsize, + const s64 prealloc_size, + struct ntfs_inode *locked_ni) { int err = 0; @@ -4964,9 +5032,10 @@ int ntfs_attr_expand(struct ntfs_inode *ni, const s64 newsize, const s64 preallo if (NInoNonResident(ni)) { if (newsize > ni->data_size || prealloc_size > ni->allocated_size) - err = ntfs_non_resident_attr_expand(ni, newsize, prealloc_size, - NVolDisableSparse(ni->vol) ? - HOLES_NO : HOLES_OK, true); + err = ntfs_non_resident_attr_expand( + ni, newsize, prealloc_size, + NVolDisableSparse(ni->vol) ? + HOLES_NO : HOLES_OK, locked_ni); } else err = ntfs_resident_attr_resize(ni, newsize, prealloc_size, NVolDisableSparse(ni->vol) ? @@ -4977,6 +5046,12 @@ int ntfs_attr_expand(struct ntfs_inode *ni, const s64 newsize, const s64 preallo return err; } +int ntfs_attr_expand(struct ntfs_inode *ni, const s64 newsize, + const s64 prealloc_size) +{ + return ntfs_attr_expand_locked(ni, newsize, prealloc_size, NULL); +} + /* * ntfs_attr_truncate_i - resize an ntfs attribute * @ni: open ntfs inode to resize @@ -4989,7 +5064,9 @@ int ntfs_attr_expand(struct ntfs_inode *ni, const s64 newsize, const s64 preallo * newly allocated space is marked as not initialised and no real allocation * on disk is performed. */ -int ntfs_attr_truncate_i(struct ntfs_inode *ni, const s64 newsize, unsigned int holes) +int ntfs_attr_truncate_i_locked(struct ntfs_inode *ni, const s64 newsize, + unsigned int holes, + struct ntfs_inode *locked_ni) { int err; @@ -5023,15 +5100,23 @@ int ntfs_attr_truncate_i(struct ntfs_inode *ni, const s64 newsize, unsigned int if (NInoNonResident(ni)) { if (newsize > ni->data_size) - err = ntfs_non_resident_attr_expand(ni, newsize, 0, holes, true); + err = ntfs_non_resident_attr_expand( + ni, newsize, 0, holes, locked_ni); else - err = ntfs_non_resident_attr_shrink(ni, newsize); + err = ntfs_non_resident_attr_shrink( + ni, newsize, locked_ni); } else err = ntfs_resident_attr_resize(ni, newsize, 0, holes); ntfs_debug("Return status %d\n", err); return err; } +int ntfs_attr_truncate_i(struct ntfs_inode *ni, const s64 newsize, + unsigned int holes) +{ + return ntfs_attr_truncate_i_locked(ni, newsize, holes, NULL); +} + /* * Resize an attribute, creating a hole if relevant */ @@ -5049,10 +5134,11 @@ int ntfs_attr_map_cluster(struct ntfs_inode *ni, s64 vcn_start, s64 *lcn_start, struct ntfs_volume *vol = ni->vol; struct ntfs_attr_search_ctx *ctx; struct runlist_element *rl, *rlc; + struct runlist_element *old_rl = NULL; s64 vcn = vcn_start, lcn, clu_count; s64 lcn_seek_from = -1; int err = 0; - size_t new_rl_count; + size_t new_rl_count, old_rl_count; err = ntfs_attr_map_whole_runlist(ni); if (err) @@ -5145,6 +5231,19 @@ int ntfs_attr_map_cluster(struct ntfs_inode *ni, s64 vcn_start, s64 *lcn_start, WARN_ON(rlc->vcn != vcn); lcn = rlc->lcn; clu_count = rlc->length; + old_rl_count = ni->runlist.count; + old_rl = kmemdup(ni->runlist.rl, + old_rl_count * sizeof(*old_rl), GFP_NOFS); + if (!old_rl) { + err = -ENOMEM; + if (ntfs_cluster_free_from_rl(vol, rlc)) { + ntfs_error(vol->sb, + "Failed to free cluster allocation after runlist backup failure."); + NVolSetErrors(vol); + } + kvfree(rlc); + goto out; + } rl = ntfs_runlists_merge(&ni->runlist, rlc, 0, &new_rl_count); if (IS_ERR(rl)) { @@ -5168,15 +5267,32 @@ int ntfs_attr_map_cluster(struct ntfs_inode *ni, s64 vcn_start, s64 *lcn_start, if (update_mp) { ntfs_attr_reinit_search_ctx(ctx); - err = ntfs_attr_update_mapping_pairs(ni, 0); + err = ntfs_attr_update_mapping_pairs_locked(ni, 0, ni); if (err) { int err2; err2 = ntfs_cluster_free(ni, vcn, clu_count, ctx); - if (err2 < 0) + if (err2 < 0 || err2 != clu_count) { ntfs_error(vol->sb, - "Failed to free cluster allocation. Leaving inconstant metadata.\n"); - goto out; + "Failed to free cluster allocation. Leaving inconsistent metadata.\n"); + NVolSetErrors(vol); + goto out; + } + + /* + * Restore the runlist before repairing the on-disk + * mapping pairs. + */ + kvfree(ni->runlist.rl); + ni->runlist.rl = old_rl; + ni->runlist.count = old_rl_count; + old_rl = NULL; + if (ntfs_attr_update_mapping_pairs_locked( + ni, 0, ni)) { + ntfs_error(vol->sb, + "Failed to restore mapping pairs after allocation rollback.\n"); + NVolSetErrors(vol); + } } } else { VFS_I(ni)->i_blocks += clu_count << (vol->cluster_size_bits - 9); @@ -5188,6 +5304,7 @@ int ntfs_attr_map_cluster(struct ntfs_inode *ni, s64 vcn_start, s64 *lcn_start, *lcn_count = clu_count; *balloc = true; out: + kvfree(old_rl); ntfs_attr_put_search_ctx(ctx); return err; } @@ -5431,7 +5548,7 @@ int ntfs_non_resident_attr_insert_range(struct ntfs_inode *ni, s64 start_vcn, s6 ni->data_size += ntfs_cluster_to_bytes(vol, len); if (ntfs_cluster_to_bytes(vol, start_vcn) < ni->initialized_size) ni->initialized_size += ntfs_cluster_to_bytes(vol, len); - ret = ntfs_attr_update_mapping_pairs(ni, 0); + ret = ntfs_attr_update_mapping_pairs_locked(ni, 0, ni); up_write(&ni->runlist.lock); if (ret) return ret; @@ -5516,7 +5633,7 @@ int ntfs_non_resident_attr_collapse_range(struct ntfs_inode *ni, s64 start_vcn, } if (ni->allocated_size > 0) { - ret = ntfs_attr_update_mapping_pairs(ni, 0); + ret = ntfs_attr_update_mapping_pairs_locked(ni, 0, ni); if (ret) { up_write(&ni->runlist.lock); goto out_rl; @@ -5594,7 +5711,7 @@ int ntfs_non_resident_attr_punch_hole(struct ntfs_inode *ni, s64 start_vcn, s64 ni->runlist.rl = rl; ni->runlist.count = new_rl_count; - ret = ntfs_attr_update_mapping_pairs(ni, 0); + ret = ntfs_attr_update_mapping_pairs_locked(ni, 0, ni); up_write(&ni->runlist.lock); if (ret) { kvfree(punch_rl); @@ -5770,7 +5887,7 @@ int ntfs_attr_fallocate(struct ntfs_inode *ni, loff_t start, loff_t byte_len, bo if (NInoRunlistDirty(ni)) { mutex_lock_nested(&ni->mrec_lock, NTFS_INODE_MUTEX_NORMAL); down_write(&ni->runlist.lock); - err = ntfs_attr_update_mapping_pairs(ni, 0); + err = ntfs_attr_update_mapping_pairs_locked(ni, 0, ni); if (err) ntfs_error(ni->vol->sb, "Updating mapping pairs failed"); else diff --git a/fs/ntfs/attrib.h b/fs/ntfs/attrib.h index e2224fbfaabe..6b4fa9f57640 100644 --- a/fs/ntfs/attrib.h +++ b/fs/ntfs/attrib.h @@ -112,7 +112,13 @@ int ntfs_non_resident_attr_punch_hole(struct ntfs_inode *ni, s64 start_vcn, s64 int __ntfs_attr_truncate_vfs(struct ntfs_inode *ni, const s64 newsize, const s64 i_size); int ntfs_attr_expand(struct ntfs_inode *ni, const s64 newsize, const s64 prealloc_size); +int ntfs_attr_expand_locked(struct ntfs_inode *ni, const s64 newsize, + const s64 prealloc_size, + struct ntfs_inode *locked_ni); int ntfs_attr_truncate_i(struct ntfs_inode *ni, const s64 newsize, unsigned int holes); +int ntfs_attr_truncate_i_locked(struct ntfs_inode *ni, const s64 newsize, + unsigned int holes, + struct ntfs_inode *locked_ni); int ntfs_attr_truncate(struct ntfs_inode *ni, const s64 newsize); int ntfs_attr_rm(struct ntfs_inode *ni); int ntfs_attr_exist(struct ntfs_inode *ni, const __le32 type, __le16 *name, @@ -133,6 +139,9 @@ int ntfs_resident_attr_record_add(struct ntfs_inode *ni, __le32 type, __le16 *name, u8 name_len, u8 *val, u32 size, __le16 flags); int ntfs_attr_update_mapping_pairs(struct ntfs_inode *ni, s64 from_vcn); +int ntfs_attr_update_mapping_pairs_locked(struct ntfs_inode *ni, + s64 from_vcn, + struct ntfs_inode *locked_ni); struct runlist_element *ntfs_attr_vcn_to_rl(struct ntfs_inode *ni, s64 vcn, s64 *lcn); /* diff --git a/fs/ntfs/attrlist.c b/fs/ntfs/attrlist.c index 4e60f7e93010..bb191953dcb1 100644 --- a/fs/ntfs/attrlist.c +++ b/fs/ntfs/attrlist.c @@ -68,7 +68,8 @@ int ntfs_attrlist_need(struct ntfs_inode *ni) * legal list size as a reserve. */ static int ntfs_attrlist_repack(struct inode *attr_vi, - struct ntfs_inode *attr_ni, s64 min_alloc_size) + struct ntfs_inode *attr_ni, s64 min_alloc_size, + struct ntfs_inode *locked_ni) { struct ntfs_volume *vol = attr_ni->vol; struct runlist_element *old_rl, *new_rl; @@ -78,10 +79,12 @@ static int ntfs_attrlist_repack(struct inode *attr_vi, size_t old_rl_count, new_rl_count; unsigned long flags; int err, restore_err; - if (attr_ni->mft_no != FILE_MFT || !NInoNonResident(attr_ni) || min_alloc_size < 0) return -EINVAL; + /* The buffered I/O below can reacquire the attribute runlist lock. */ + if (attr_ni == locked_ni) + return -ENOSPC; err = ntfs_attr_map_whole_runlist(attr_ni); if (err) @@ -131,7 +134,6 @@ static int ntfs_attrlist_repack(struct inode *attr_vi, err = -ENOSPC; goto out_free_data; } - old_rl = attr_ni->runlist.rl; old_rl_count = attr_ni->runlist.count; down_write(&attr_ni->runlist.lock); @@ -152,7 +154,7 @@ static int ntfs_attrlist_repack(struct inode *attr_vi, } } - err = ntfs_attr_update_mapping_pairs(attr_ni, 0); + err = ntfs_attr_update_mapping_pairs_locked(attr_ni, 0, locked_ni); if (err) goto restore_old_runlist; @@ -177,7 +179,8 @@ static int ntfs_attrlist_repack(struct inode *attr_vi, attr_ni->allocated_size = old_alloc_size; write_unlock_irqrestore(&attr_ni->size_lock, flags); - restore_err = ntfs_attr_update_mapping_pairs(attr_ni, 0); + restore_err = ntfs_attr_update_mapping_pairs_locked( + attr_ni, 0, locked_ni); if (restore_err) { ntfs_error(vol->sb, "Failed to restore ATTRIBUTE_LIST mapping pairs (%d)", restore_err); @@ -193,7 +196,8 @@ static int ntfs_attrlist_repack(struct inode *attr_vi, return err; } -int ntfs_attrlist_update(struct ntfs_inode *base_ni) +int ntfs_attrlist_update_locked(struct ntfs_inode *base_ni, + struct ntfs_inode *locked_ni) { struct inode *attr_vi; struct ntfs_inode *attr_ni; @@ -215,20 +219,27 @@ int ntfs_attrlist_update(struct ntfs_inode *base_ni) return err; } attr_ni = NTFS_I(attr_vi); + /* Truncation and page-cache writes can reacquire this runlist lock. */ + if (attr_ni == locked_ni) { + iput(attr_vi); + return -ENOSPC; + } - err = ntfs_attr_truncate_i(attr_ni, base_ni->attr_list_size, HOLES_NO); + err = ntfs_attr_truncate_i_locked( + attr_ni, base_ni->attr_list_size, HOLES_NO, locked_ni); if (err == -ENOSPC && attr_ni->mft_no == FILE_MFT && NInoNonResident(attr_ni)) { retry_err = ntfs_attrlist_repack(attr_vi, attr_ni, - base_ni->attr_list_size); + base_ni->attr_list_size, locked_ni); if (retry_err) { ntfs_error(base_ni->vol->sb, "Failed to repack attribute list"); iput(attr_vi); return retry_err; } - retry_err = ntfs_attr_truncate_i(attr_ni, base_ni->attr_list_size, - HOLES_NO); + retry_err = ntfs_attr_truncate_i_locked( + attr_ni, base_ni->attr_list_size, + HOLES_NO, locked_ni); if (retry_err) { ntfs_error(base_ni->vol->sb, "Failed to resize attribute list after repack"); @@ -252,11 +263,13 @@ int ntfs_attrlist_update(struct ntfs_inode *base_ni) */ if (base_ni->mft_no == FILE_MFT && NInoNonResident(attr_ni) && attr_ni->allocated_size < NTFS_MAX_ATTR_LIST_SIZE) { - retry_err = ntfs_attr_expand(attr_ni, base_ni->attr_list_size, - NTFS_MAX_ATTR_LIST_SIZE); + retry_err = ntfs_attr_expand_locked( + attr_ni, base_ni->attr_list_size, + NTFS_MAX_ATTR_LIST_SIZE, locked_ni); if (retry_err == -ENOSPC) { - retry_err = ntfs_attrlist_repack(attr_vi, attr_ni, - NTFS_MAX_ATTR_LIST_SIZE); + retry_err = ntfs_attrlist_repack( + attr_vi, attr_ni, + NTFS_MAX_ATTR_LIST_SIZE, locked_ni); if (retry_err == -ENOSPC) retry_err = 0; } @@ -289,6 +302,11 @@ int ntfs_attrlist_update(struct ntfs_inode *base_ni) return 0; } +int ntfs_attrlist_update(struct ntfs_inode *base_ni) +{ + return ntfs_attrlist_update_locked(base_ni, NULL); +} + /* * ntfs_attrlist_entry_add - add an attribute list attribute entry * @ni: opened ntfs inode, which contains that attribute diff --git a/fs/ntfs/attrlist.h b/fs/ntfs/attrlist.h index 1892a3934d3a..10cc2cc8e208 100644 --- a/fs/ntfs/attrlist.h +++ b/fs/ntfs/attrlist.h @@ -16,5 +16,7 @@ int ntfs_attrlist_need(struct ntfs_inode *ni); int ntfs_attrlist_entry_add(struct ntfs_inode *ni, struct attr_record *attr); int ntfs_attrlist_entry_rm(struct ntfs_attr_search_ctx *ctx); int ntfs_attrlist_update(struct ntfs_inode *base_ni); +int ntfs_attrlist_update_locked(struct ntfs_inode *base_ni, + struct ntfs_inode *locked_ni); #endif /* defined _NTFS_ATTRLIST_H */ diff --git a/fs/ntfs/compress.c b/fs/ntfs/compress.c index 197d8607fc63..6927d115d6af 100644 --- a/fs/ntfs/compress.c +++ b/fs/ntfs/compress.c @@ -1450,7 +1450,7 @@ static int ntfs_write_cb(struct ntfs_inode *ni, loff_t pos, struct page **pages, ni->runlist.rl = rl; rlc = NULL; - err = ntfs_attr_update_mapping_pairs(ni, 0); + err = ntfs_attr_update_mapping_pairs_locked(ni, 0, ni); up_write(&ni->runlist.lock); if (err) err = -EIO; diff --git a/fs/ntfs/file.c b/fs/ntfs/file.c index 8164326b7812..007d1614b9ac 100644 --- a/fs/ntfs/file.c +++ b/fs/ntfs/file.c @@ -111,7 +111,8 @@ static int ntfs_trim_prealloc(struct inode *vi) ntfs_error(vol->sb, "Preallocated block rollback failed"); } else { ni->allocated_size = ntfs_cluster_to_bytes(vol, vcn_tr); - err = ntfs_attr_update_mapping_pairs(ni, 0); + err = ntfs_attr_update_mapping_pairs_locked( + ni, 0, ni); if (err) ntfs_error(vol->sb, "Failed to rollback mapping pairs for prealloc"); diff --git a/fs/ntfs/inode.c b/fs/ntfs/inode.c index 332825db477a..5d9d48135ecd 100644 --- a/fs/ntfs/inode.c +++ b/fs/ntfs/inode.c @@ -2772,7 +2772,7 @@ int __ntfs_write_inode(struct inode *vi, int sync) if (NInoNonResident(ni) && NInoRunlistDirty(ni)) { down_write(&ni->runlist.lock); - err = ntfs_attr_update_mapping_pairs(ni, 0); + err = ntfs_attr_update_mapping_pairs_locked(ni, 0, ni); if (!err) NInoClearRunlistDirty(ni); up_write(&ni->runlist.lock); diff --git a/fs/ntfs/mft.c b/fs/ntfs/mft.c index f0656d0bbeeb..5b9723abb10f 100644 --- a/fs/ntfs/mft.c +++ b/fs/ntfs/mft.c @@ -1131,7 +1131,8 @@ static s64 mft_bitmap_alloc_free_rec(struct ntfs_volume *vol, return ll; } -static int ntfs_mft_attr_extend(struct ntfs_inode *ni) +static int ntfs_mft_attr_extend(struct ntfs_inode *ni, + struct ntfs_inode *locked_ni) { int ret = 0; struct ntfs_inode *base_ni; @@ -1152,7 +1153,7 @@ static int ntfs_mft_attr_extend(struct ntfs_inode *ni) } } - ret = ntfs_attr_update_mapping_pairs(ni, 0); + ret = ntfs_attr_update_mapping_pairs_locked(ni, 0, locked_ni); if (ret) pr_err("MP update failed\n"); @@ -1340,7 +1341,7 @@ static int ntfs_mft_bitmap_extend_allocation_nolock(struct ntfs_volume *vol) ret = ntfs_attr_record_resize(ctx->mrec, a, mp_size + le16_to_cpu(a->data.non_resident.mapping_pairs_offset)); if (unlikely(ret)) { - ret = ntfs_mft_attr_extend(mftbmp_ni); + ret = ntfs_mft_attr_extend(mftbmp_ni, mftbmp_ni); if (!ret) goto extended_ok; if (ret != -EAGAIN) @@ -1451,7 +1452,9 @@ static int ntfs_mft_bitmap_extend_allocation_nolock(struct ntfs_volume *vol) NVolSetErrors(vol); } mark_mft_record_dirty(ctx->ntfs_ino); - } else if (status.mp_extended && ntfs_attr_update_mapping_pairs(mftbmp_ni, 0)) { + } else if (status.mp_extended && + ntfs_attr_update_mapping_pairs_locked(mftbmp_ni, 0, + mftbmp_ni)) { ntfs_error(vol->sb, "Failed to restore mapping pairs.%s", es); NVolSetErrors(vol); } @@ -1776,7 +1779,7 @@ static int ntfs_mft_data_extend_allocation_nolock(struct ntfs_volume *vol) ret = ntfs_attr_record_resize(ctx->mrec, a, mp_size + le16_to_cpu(a->data.non_resident.mapping_pairs_offset)); if (unlikely(ret)) { - ret = ntfs_mft_attr_extend(mft_ni); + ret = ntfs_mft_attr_extend(mft_ni, NULL); if (!ret) goto extended_ok; if (ret != -EAGAIN) From 1923eeffa63edeff427d76fc302bc5eb835771ce Mon Sep 17 00:00:00 2001 From: Namjae Jeon Date: Mon, 7 Sep 2026 13:34:01 +0900 Subject: [PATCH 142/592] ntfs: propagate folio errors Return the error from __filemap_get_folio() instead of replacing it with -ENOMEM. Fixes: af0db57d4293 ("ntfs: update inode operations") Reviewed-by: Hyunchul Lee Signed-off-by: Namjae Jeon --- fs/ntfs/inode.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/fs/ntfs/inode.c b/fs/ntfs/inode.c index 5d9d48135ecd..0a0b7c5547f7 100644 --- a/fs/ntfs/inode.c +++ b/fs/ntfs/inode.c @@ -3713,7 +3713,7 @@ static s64 __ntfs_inode_non_resident_attr_pwrite(struct inode *vi, FGP_CREAT | FGP_LOCK, mapping_gfp_mask(mapping)); if (IS_ERR(folio)) { - ret = -ENOMEM; + ret = PTR_ERR(folio); break; } } else { From 8c5dc7587fdd45f957af81a9adc1e16863f300fc Mon Sep 17 00:00:00 2001 From: Namjae Jeon Date: Mon, 7 Sep 2026 13:34:25 +0900 Subject: [PATCH 143/592] ntfs: ignore interrupted inode reads as corruption Do not mark the volume in error or report an inode as corrupt when reading it was interrupted by a signal. -EINTR and -ERESTARTSYS indicate a transient read failure rather than on-disk NTFS corruption. Fixes: 115380f9a2f9 ("ntfs: update mft operations") Reviewed-by: Hyunchul Lee Signed-off-by: Namjae Jeon --- fs/ntfs/inode.c | 21 +++++++++++++-------- fs/ntfs/mft.c | 3 ++- 2 files changed, 15 insertions(+), 9 deletions(-) diff --git a/fs/ntfs/inode.c b/fs/ntfs/inode.c index 0a0b7c5547f7..210adf589319 100644 --- a/fs/ntfs/inode.c +++ b/fs/ntfs/inode.c @@ -1241,7 +1241,8 @@ static int ntfs_read_locked_inode(struct inode *vi) if (m) unmap_mft_record(ni); err_out: - if (err != -EOPNOTSUPP && err != -ENOMEM && vol_err == true) { + if (err != -EOPNOTSUPP && err != -ENOMEM && + err != -EINTR && err != -ERESTARTSYS && vol_err == true) { ntfs_error(vol->sb, "Failed with error code %i. Marking corrupt inode 0x%llx as bad. Run chkdsk.", err, ni->mft_no); @@ -1467,12 +1468,13 @@ static int ntfs_read_locked_attr_inode(struct inode *base_vi, struct inode *vi) ntfs_attr_put_search_ctx(ctx); unmap_mft_record(base_ni); err_out: - if (err != -ENOENT) + if (err != -ENOENT && err != -EINTR && err != -ERESTARTSYS) ntfs_error(vol->sb, "Failed with error code %i while reading attribute inode (mft_no 0x%llx, type 0x%x, name_len %i). Marking corrupt inode and base inode 0x%llx as bad. Run chkdsk.", err, ni->mft_no, ni->type, ni->name_len, base_ni->mft_no); - if (err != -ENOENT && err != -ENOMEM) + if (err != -ENOENT && err != -ENOMEM && + err != -EINTR && err != -ERESTARTSYS) NVolSetErrors(vol); return err; } @@ -1676,8 +1678,9 @@ static int ntfs_read_locked_index_inode(struct inode *base_vi, struct inode *vi) /* Get the index bitmap attribute inode. */ bvi = ntfs_attr_iget(base_vi, AT_BITMAP, ni->name, ni->name_len); if (IS_ERR(bvi)) { - ntfs_error(vi->i_sb, "Failed to get bitmap attribute."); err = PTR_ERR(bvi); + if (err != -EINTR && err != -ERESTARTSYS) + ntfs_error(vi->i_sb, "Failed to get bitmap attribute."); goto unm_err_out; } bni = NTFS_I(bvi); @@ -1721,10 +1724,12 @@ static int ntfs_read_locked_index_inode(struct inode *base_vi, struct inode *vi) if (m) unmap_mft_record(base_ni); err_out: - ntfs_error(vi->i_sb, - "Failed with error code %i while reading index inode (mft_no 0x%llx, name_len %i.", - err, ni->mft_no, ni->name_len); - if (err != -EOPNOTSUPP && err != -ENOMEM) + if (err != -EINTR && err != -ERESTARTSYS) + ntfs_error(vi->i_sb, + "Failed with error code %i while reading index inode (mft_no 0x%llx, name_len %i.", + err, ni->mft_no, ni->name_len); + if (err != -EOPNOTSUPP && err != -ENOMEM && + err != -EINTR && err != -ERESTARTSYS) NVolSetErrors(vol); return err; } diff --git a/fs/ntfs/mft.c b/fs/ntfs/mft.c index 5b9723abb10f..6d5e56378afd 100644 --- a/fs/ntfs/mft.c +++ b/fs/ntfs/mft.c @@ -213,7 +213,8 @@ struct mft_record *map_mft_record(struct ntfs_inode *ni) return m; atomic_dec(&ni->count); - ntfs_error(ni->vol->sb, "Failed with error code %lu.", -PTR_ERR(m)); + if (PTR_ERR(m) != -EINTR && PTR_ERR(m) != -ERESTARTSYS) + ntfs_error(ni->vol->sb, "Failed with error code %lu.", -PTR_ERR(m)); return m; } From fc440366c47000b768d013e347f60e81d60328ca Mon Sep 17 00:00:00 2001 From: Namjae Jeon Date: Mon, 7 Sep 2026 13:34:44 +0900 Subject: [PATCH 144/592] ntfs: discard inodes that fail initialization Discard a newly allocated normal, attribute, or index inode when initialization fails. Keeping an incompletely initialized inode in the inode cache can expose stale sequence data to later directory lookups. Fixes: d7aa04984f12 ("ntfs: return errors from inode initialization") Reviewed-by: Hyunchul Lee Signed-off-by: Namjae Jeon --- fs/ntfs/inode.c | 27 +++++++++++++++------------ 1 file changed, 15 insertions(+), 12 deletions(-) diff --git a/fs/ntfs/inode.c b/fs/ntfs/inode.c index 210adf589319..b1f9a72ea963 100644 --- a/fs/ntfs/inode.c +++ b/fs/ntfs/inode.c @@ -170,17 +170,18 @@ struct inode *ntfs_iget(struct super_block *sb, u64 mft_no) /* If this is a freshly allocated inode, need to read it now. */ if (inode_state_read_once(vi) & I_NEW) { err = ntfs_read_locked_inode(vi); - unlock_new_inode(vi); + if (err) + discard_new_inode(vi); + else + unlock_new_inode(vi); } /* * There is no point in keeping bad inodes around. This also * simplifies things in that we never need to check for bad inodes * elsewhere. */ - if (unlikely(err)) { - iput(vi); + if (unlikely(err)) vi = ERR_PTR(err); - } return vi; } @@ -231,17 +232,18 @@ struct inode *ntfs_attr_iget(struct inode *base_vi, __le32 type, /* If this is a freshly allocated inode, need to read it now. */ if (inode_state_read_once(vi) & I_NEW) { err = ntfs_read_locked_attr_inode(base_vi, vi); - unlock_new_inode(vi); + if (err) + discard_new_inode(vi); + else + unlock_new_inode(vi); } /* * There is no point in keeping bad attribute inodes around. This also * simplifies things in that we never need to check for bad attribute * inodes elsewhere. */ - if (unlikely(err)) { - iput(vi); + if (unlikely(err)) vi = ERR_PTR(err); - } return vi; } @@ -286,17 +288,18 @@ struct inode *ntfs_index_iget(struct inode *base_vi, __le16 *name, /* If this is a freshly allocated inode, need to read it now. */ if (inode_state_read_once(vi) & I_NEW) { err = ntfs_read_locked_index_inode(base_vi, vi); - unlock_new_inode(vi); + if (err) + discard_new_inode(vi); + else + unlock_new_inode(vi); } /* * There is no point in keeping bad index inodes around. This also * simplifies things in that we never need to check for bad index * inodes elsewhere. */ - if (unlikely(err)) { - iput(vi); + if (unlikely(err)) vi = ERR_PTR(err); - } return vi; } From 0c32a42fd96a0e7c06c8a648a6dd8f1ec0bf643b Mon Sep 17 00:00:00 2001 From: Namjae Jeon Date: Mon, 7 Sep 2026 13:38:17 +0900 Subject: [PATCH 145/592] ntfs: unhash failed inode reads Remove a newly allocated inode from the inode hash before discarding it. NTFS may set i_nlink before a later initialization step fails, in which case iput alone can retain the incomplete inode in the cache. Fixes: bf6be898fbc5 ("ntfs: discard inodes that fail initialization") Reviewed-by: Hyunchul Lee Signed-off-by: Namjae Jeon --- fs/ntfs/inode.c | 15 +++++++++------ 1 file changed, 9 insertions(+), 6 deletions(-) diff --git a/fs/ntfs/inode.c b/fs/ntfs/inode.c index b1f9a72ea963..a777de8a80c7 100644 --- a/fs/ntfs/inode.c +++ b/fs/ntfs/inode.c @@ -170,9 +170,10 @@ struct inode *ntfs_iget(struct super_block *sb, u64 mft_no) /* If this is a freshly allocated inode, need to read it now. */ if (inode_state_read_once(vi) & I_NEW) { err = ntfs_read_locked_inode(vi); - if (err) + if (err) { + remove_inode_hash(vi); discard_new_inode(vi); - else + } else unlock_new_inode(vi); } /* @@ -232,9 +233,10 @@ struct inode *ntfs_attr_iget(struct inode *base_vi, __le32 type, /* If this is a freshly allocated inode, need to read it now. */ if (inode_state_read_once(vi) & I_NEW) { err = ntfs_read_locked_attr_inode(base_vi, vi); - if (err) + if (err) { + remove_inode_hash(vi); discard_new_inode(vi); - else + } else unlock_new_inode(vi); } /* @@ -288,9 +290,10 @@ struct inode *ntfs_index_iget(struct inode *base_vi, __le16 *name, /* If this is a freshly allocated inode, need to read it now. */ if (inode_state_read_once(vi) & I_NEW) { err = ntfs_read_locked_index_inode(base_vi, vi); - if (err) + if (err) { + remove_inode_hash(vi); discard_new_inode(vi); - else + } else unlock_new_inode(vi); } /* From 229e8188307b9724cc676a49e9600c4acd24b571 Mon Sep 17 00:00:00 2001 From: Zhu Tianhao Date: Tue, 8 Sep 2026 05:41:00 +0900 Subject: [PATCH 146/592] ntfs: fix $MFTMirr write offset when it spans multiple folios When commit 115380f9a2f9 ("ntfs: update mft operations") refactored the MFT code to use folios, it assumed $MFTMirr is allocated contiguously, which is indeed what mkfs arranges. However, the folio rewrite kept a leftover from the old buffer-head/runlist based implementation: vol->cluster_size_mask is still applied to the mirror write offset. In fact it is no longer needed -- and in some configurations it is actively wrong. The bug triggers whenever the four mirror records do not fit in a single folio, i.e. when mft_record_size exceeds PAGE_SIZE / 4 (for example mft_record_size > 1KiB on 4KiB pages). For example, on the built-in 4Kn SSD (Apple SSD AP0256J) of a MacBookPro14,1 with 4096-byte MFT records, mirror record 3 is still written to the location of record 0. $MFTMirr therefore gets out of sync with $MFT and the following warning is printed on remount: ntfs: (device nvme0n1p3): check_mft_mirror(): $MFT and $MFTMirr record 0 do not match. Run chkdsk. Fix it by always adding the folio offset (folio->index << PAGE_SHIFT) to the base LCN address instead of applying the cluster_size_mask truncation. This is the standard file-offset calculation and is correct for every combination of cluster size, page size and MFT record size, provided $MFTMirr data is contiguous from mftmirr_lcn (which mkfs always arranges). Tested on the volume above: before the change mirror records 1-3 are all written to record 0's sector; after the change the write-back probe reports: ntfs: NTFSDBG pre mft_no=0x3 folio_idx=0x3 ofs=0x0 mftmirr_lcn=0x2540c5 clu_bits=12 rec_bits=12 PAGE_SHIFT=12 sect=0x12a0640 ntfs: NTFSDBG pre mft_no=0x0 folio_idx=0x0 ofs=0x0 mftmirr_lcn=0x2540c5 clu_bits=12 rec_bits=12 PAGE_SHIFT=12 sect=0x12a0628 which matches the expected sectors computed as: (NTFS_CLU_TO_B(mftmirr_lcn) + (folio->index << PAGE_SHIFT)) >> SECTOR_SHIFT record 0: (0x2540c5 << 12) + 0x0000 = 0x2540c5000 >> 9 = 0x12a0628 record 3: (0x2540c5 << 12) + 0x3000 = 0x2540c8000 >> 9 = 0x12a0640 A 13 MB file write succeeds on this volume and the file is byte-for-byte identical after a umount/mount cycle. Fixes: 115380f9a2f9 ("ntfs: update mft operations") Assisted-by: UOS-AI Assisted-by: CodeBuddy:Hy4 preview Signed-off-by: Zhu Tianhao Reviewed-by: Baolin Liu Signed-off-by: Namjae Jeon --- fs/ntfs/mft.c | 10 +++------- 1 file changed, 3 insertions(+), 7 deletions(-) diff --git a/fs/ntfs/mft.c b/fs/ntfs/mft.c index 6d5e56378afd..a09496622c7a 100644 --- a/fs/ntfs/mft.c +++ b/fs/ntfs/mft.c @@ -463,7 +463,7 @@ int ntfs_sync_mft_mirror(struct ntfs_volume *vol, const u64 mft_no, { u8 *kmirr; struct folio *folio; - unsigned int folio_ofs, lcn_folio_off = 0; + unsigned int folio_ofs; int err = 0; struct bio *bio; @@ -493,15 +493,11 @@ int ntfs_sync_mft_mirror(struct ntfs_volume *vol, const u64 mft_no, memcpy(kmirr, m, vol->mft_record_size); kunmap_local(kmirr); - if (vol->cluster_size_bits > PAGE_SHIFT) { - lcn_folio_off = folio->index << PAGE_SHIFT; - lcn_folio_off &= vol->cluster_size_mask; - } - bio = bio_alloc(vol->sb->s_bdev, 1, REQ_OP_WRITE, GFP_NOIO); bio->bi_iter.bi_sector = ntfs_bytes_to_bio_sector(NTFS_CLU_TO_B(vol, vol->mftmirr_lcn) + - lcn_folio_off + folio_ofs); + ((u64)folio->index << PAGE_SHIFT) + + folio_ofs); if (bio_add_folio(bio, folio, vol->mft_record_size, folio_ofs)) err = submit_bio_wait(bio); From 48e97c59a49c9e90270f5e3d221db253b76de5da Mon Sep 17 00:00:00 2001 From: Claudiu Beznea Date: Thu, 16 Jul 2026 21:32:46 +0300 Subject: [PATCH 147/592] phy: renesas: rcar-gen3-usb2: Avoid long delay in atomic context The OTG PHY initialization sequence needs to wait for 20 ms at a specific step, as described in commit 72c0339c115b ("phy: renesas: rcar-gen3-usb2: follow the hardware manual procedure"). Commit 55a387ebb921 ("phy: renesas: rcar-gen3-usb2: Lock around hardware registers and driver data") tried to address various problems in the rcar-gen3-usb2 driver and converted the mutex protecting HW register accesses to a spin lock, leaving, however, a long delay in the critical section protected by the spin lock. This may become a problem, especially on RT kernels. To address this, release the spin lock before sleeping for 20 ms as required by the HW manual and reacquire it afterwards. To avoid other threads entering the critical section and configuring the HW while the software is waiting for the OTG initialization to complete, introduce the otg_initializing variable alongside the otg_init_done wait queue. Any other thread trying to configure the HW while the OTG PHY initialization is in progress waits for the wait queue instead of immediately returning errors to PHY users. The IRQs were also disabled while waiting for the OTG PHY initialization to complete, as the interrupt handler may also apply HW settings. The OTG can only be initialized once. It is initialized by the first PHY that calls struct phy_ops::rcar_gen3_phy_usb2_init(). To avoid failures when multiple PHYs call struct phy_ops::rcar_gen3_phy_usb2_init() simultaneously, and the PHY responsible for initializing the OTG either fails or deinit quiqly and another PHY takes over the PHY init role), the code waiting for the channel->otg_init_done wait queue retries up to NUM_OF_PHYS times. Fixes: 55a387ebb921 ("phy: renesas: rcar-gen3-usb2: Lock around hardware registers and driver data") Cc: stable@vger.kernel.org Reported-by: Pavel Machek Closes: https://lore.kernel.org/all/afhkX2Ys2BG1gnqy@duo.ucw.cz Reported-by: Nobuhiro Iwamatsu Closes: https://lore.kernel.org/all/afhkX2Ys2BG1gnqy@duo.ucw.cz Signed-off-by: Claudiu Beznea Reviewed-by: Manivannan Sadhasivam Link: https://lore.kernel.org/all/afhkX2Ys2BG1gnqy@duo.ucw.cz Link: https://patch.msgid.link/20260716183246.3183877-1-claudiu.beznea+renesas@tuxon.dev Signed-off-by: Vinod Koul --- drivers/phy/renesas/phy-rcar-gen3-usb2.c | 310 +++++++++++++++++++---- 1 file changed, 265 insertions(+), 45 deletions(-) diff --git a/drivers/phy/renesas/phy-rcar-gen3-usb2.c b/drivers/phy/renesas/phy-rcar-gen3-usb2.c index 9ae9975d3255..b5ba751805aa 100644 --- a/drivers/phy/renesas/phy-rcar-gen3-usb2.c +++ b/drivers/phy/renesas/phy-rcar-gen3-usb2.c @@ -27,6 +27,7 @@ #include #include #include +#include #include /******* USB2.0 Host registers (original offset is +0x200) *******/ @@ -106,6 +107,13 @@ /* RZ/G2L specific */ #define USB2_LINECTRL1_USB2_IDMON BIT(0) +/* + * The OTG initialization is expected to finish in 20ms. Choose a large enough + * timeout to avoid waiters exit prematurely the waiting section under heavy + * CPU load. + */ +#define USB2_OTG_INIT_TIMEOUT msecs_to_jiffies(120) + #define NUM_OF_PHYS 4 enum rcar_gen3_phy_index { PHY_INDEX_BOTH_HC, @@ -138,12 +146,20 @@ struct rcar_gen3_chan { struct rcar_gen3_phy rphys[NUM_OF_PHYS]; struct regulator *vbus; struct work_struct work; + wait_queue_head_t otg_init_done; spinlock_t lock; /* protects access to hardware and driver data structure. */ enum usb_dr_mode dr_mode; bool extcon_host; bool is_otg_channel; bool uses_otg_pins; bool otg_internal_reg; + /* + * The OTG can be initialized only once and needs to release the spinlock + * and wait for 20 ms due to hardware constraints. If a thread executes + * PHY configuration code while the OTG PHY is waiting for the 20 ms, the + * thread will have to wait for the OTG PHY initialization to complete. + */ + bool otg_initializing; }; struct rcar_gen3_phy_drv_data { @@ -392,26 +408,58 @@ static ssize_t role_store(struct device *dev, struct device_attribute *attr, struct rcar_gen3_chan *ch = dev_get_drvdata(dev); bool is_b_device; enum phy_mode cur_mode, new_mode; + int retries = NUM_OF_PHYS; + unsigned long flags; + int ret = -EIO; - guard(spinlock_irqsave)(&ch->lock); + spin_lock_irqsave(&ch->lock, flags); - if (!ch->is_otg_channel || !rcar_gen3_is_any_otg_rphy_initialized(ch)) - return -EIO; + if (!ch->is_otg_channel) + goto unlock; - if (sysfs_streq(buf, "host")) + while (retries-- && ch->otg_initializing) { + spin_unlock_irqrestore(&ch->lock, flags); + + ret = wait_event_timeout(ch->otg_init_done, !ch->otg_initializing, + USB2_OTG_INIT_TIMEOUT); + ret = ret ? 0 : -ETIMEDOUT; + if (ret && !retries) + goto exit; + + spin_lock_irqsave(&ch->lock, flags); + } + + /* If another thread started a new initialization just return -EBUSY. */ + if (ch->otg_initializing) { + ret = -EBUSY; + goto unlock; + } else { + ret = 0; + } + + if (!rcar_gen3_is_any_otg_rphy_initialized(ch)) { + ret = -EIO; + goto unlock; + } + + if (sysfs_streq(buf, "host")) { new_mode = PHY_MODE_USB_HOST; - else if (sysfs_streq(buf, "peripheral")) + } else if (sysfs_streq(buf, "peripheral")) { new_mode = PHY_MODE_USB_DEVICE; - else - return -EINVAL; + } else { + ret = -EINVAL; + goto unlock; + } /* is_b_device: true is B-Device. false is A-Device. */ is_b_device = rcar_gen3_check_id(ch); cur_mode = rcar_gen3_get_phy_mode(ch); /* If current and new mode is the same, this returns the error */ - if (cur_mode == new_mode) - return -EINVAL; + if (cur_mode == new_mode) { + ret = -EINVAL; + goto unlock; + } if (new_mode == PHY_MODE_USB_HOST) { /* And is_host must be false */ if (!is_b_device) /* A-Peripheral */ @@ -425,7 +473,10 @@ static ssize_t role_store(struct device *dev, struct device_attribute *attr, rcar_gen3_init_for_peri(ch); } - return count; +unlock: + spin_unlock_irqrestore(&ch->lock, flags); +exit: + return ret ?: count; } static ssize_t role_show(struct device *dev, struct device_attribute *attr, @@ -441,14 +492,11 @@ static ssize_t role_show(struct device *dev, struct device_attribute *attr, } static DEVICE_ATTR_RW(role); -static void rcar_gen3_init_otg(struct rcar_gen3_chan *ch) +static void rcar_gen3_init_otg_phase0(struct rcar_gen3_chan *ch) { void __iomem *usb2_base = ch->base; u32 val; - if (!ch->is_otg_channel || rcar_gen3_is_any_otg_rphy_initialized(ch)) - return; - /* Should not use functions of read-modify-write a register */ val = readl(usb2_base + USB2_LINECTRL1); val = (val & ~USB2_LINECTRL1_DP_RPD) | USB2_LINECTRL1_DPRPD_EN | @@ -471,7 +519,11 @@ static void rcar_gen3_init_otg(struct rcar_gen3_chan *ch) writel(val | USB2_ADPCTRL_IDPULLUP, usb2_base + USB2_ADPCTRL); } } - mdelay(20); +} + +static void rcar_gen3_init_otg_phase1(struct rcar_gen3_chan *ch) +{ + void __iomem *usb2_base = ch->base; writel(0xffffffff, usb2_base + USB2_OBINTSTA); writel(ch->phy_data->obint_enable_bits, usb2_base + USB2_OBINTEN); @@ -502,6 +554,7 @@ static irqreturn_t rcar_gen3_phy_usb2_irq(int irq, void *_ch) void __iomem *usb2_base = ch->base; struct device *dev = ch->dev; irqreturn_t ret = IRQ_NONE; + unsigned long flags; u32 status; pm_runtime_get_noresume(dev); @@ -509,33 +562,102 @@ static irqreturn_t rcar_gen3_phy_usb2_irq(int irq, void *_ch) if (pm_runtime_suspended(dev)) goto rpm_put; - scoped_guard(spinlock, &ch->lock) { - status = readl(usb2_base + USB2_OBINTSTA); - if (status & ch->phy_data->obint_enable_bits) { - dev_vdbg(dev, "%s: %08x\n", __func__, status); - if (ch->phy_data->vblvl_ctrl) - writel(USB2_OBINTSTA_CLEAR, usb2_base + USB2_OBINTSTA); - else - writel(ch->phy_data->obint_enable_bits, usb2_base + USB2_OBINTSTA); - rcar_gen3_device_recognition(ch); - rcar_gen3_configure_vblvl_ctrl(ch); - ret = IRQ_HANDLED; - } + spin_lock_irqsave(&ch->lock, flags); + + status = readl(usb2_base + USB2_OBINTSTA); + if (status & ch->phy_data->obint_enable_bits) { + dev_vdbg(dev, "%s: %08x\n", __func__, status); + if (ch->phy_data->vblvl_ctrl) + writel(USB2_OBINTSTA_CLEAR, usb2_base + USB2_OBINTSTA); + else + writel(ch->phy_data->obint_enable_bits, usb2_base + USB2_OBINTSTA); + + ret = IRQ_HANDLED; + + /* This should not happen! */ + if (ch->otg_initializing) + goto unlock; + + rcar_gen3_device_recognition(ch); + rcar_gen3_configure_vblvl_ctrl(ch); } +unlock: + spin_unlock_irqrestore(&ch->lock, flags); rpm_put: pm_runtime_put_noidle(dev); return ret; } +static void rcar_gen3_phy_usb2_irqs_mask_all(struct rcar_gen3_chan *channel, + u32 *masked_irqs_bits) +{ + u32 val, bitmask = USB2_INT_ENABLE_UCOM_INTEN; + void __iomem *usb2_base = channel->base; + + for (unsigned int i = 0; i < NUM_OF_PHYS; i++) + bitmask |= channel->rphys[i].int_enable_bits; + + val = readl(usb2_base + USB2_INT_ENABLE); + *masked_irqs_bits = val & bitmask; + val &= ~bitmask; + writel(val, usb2_base + USB2_INT_ENABLE); + + /* + * Don't report channel->phy_data->obint_enable_bits IRQs. These are + * unmasked anyway in rcar_gen3_init_otg_phase1(). + */ + val = readl(usb2_base + USB2_OBINTEN); + val &= ~channel->phy_data->obint_enable_bits; + writel(val, usb2_base + USB2_OBINTEN); +} + +static void rcar_gen3_phy_usb2_irqs_unmask(struct rcar_gen3_chan *channel, + u32 irqs_bits) +{ + u32 val, bitmask = USB2_INT_ENABLE_UCOM_INTEN; + void __iomem *usb2_base = channel->base; + + for (unsigned int i = 0; i < NUM_OF_PHYS; i++) + bitmask |= channel->rphys[i].int_enable_bits; + + val = readl(usb2_base + USB2_INT_ENABLE); + val &= ~bitmask; + val |= irqs_bits; + writel(val, usb2_base + USB2_INT_ENABLE); +} + static int rcar_gen3_phy_usb2_init(struct phy *p) { struct rcar_gen3_phy *rphy = phy_get_drvdata(p); struct rcar_gen3_chan *channel = rphy->ch; void __iomem *usb2_base = channel->base; + int retries = NUM_OF_PHYS; + unsigned long flags; u32 val; + int ret; - guard(spinlock_irqsave)(&channel->lock); + spin_lock_irqsave(&channel->lock, flags); + + while (retries-- && channel->otg_initializing) { + spin_unlock_irqrestore(&channel->lock, flags); + + ret = wait_event_timeout(channel->otg_init_done, !channel->otg_initializing, + USB2_OTG_INIT_TIMEOUT); + ret = ret ? 0 : -ETIMEDOUT; + if (ret && !retries) + return ret; + + spin_lock_irqsave(&channel->lock, flags); + } + + /* If another thread started a new initialization just return -EBUSY. */ + if (channel->otg_initializing) { + ret = -EBUSY; + goto unlock; + } else { + ret = 0; + } /* Initialize USB2 part */ val = readl(usb2_base + USB2_INT_ENABLE); @@ -548,8 +670,23 @@ static int rcar_gen3_phy_usb2_init(struct phy *p) } /* Initialize otg part (only if we initialize a PHY with IRQs). */ - if (rphy->int_enable_bits) - rcar_gen3_init_otg(channel); + if (rphy->int_enable_bits && channel->is_otg_channel && + !rcar_gen3_is_any_otg_rphy_initialized(channel)) { + u32 masked_irq_bits = 0; + + rcar_gen3_init_otg_phase0(channel); + rcar_gen3_phy_usb2_irqs_mask_all(channel, &masked_irq_bits); + channel->otg_initializing = true; + spin_unlock_irqrestore(&channel->lock, flags); + + fsleep(20000); + + spin_lock_irqsave(&channel->lock, flags); + rcar_gen3_phy_usb2_irqs_unmask(channel, masked_irq_bits); + rcar_gen3_init_otg_phase1(channel); + channel->otg_initializing = false; + wake_up_all(&channel->otg_init_done); + } if (channel->phy_data->vblvl_ctrl) { /* SIDDQ mode release */ @@ -568,7 +705,10 @@ static int rcar_gen3_phy_usb2_init(struct phy *p) rphy->initialized = true; - return 0; +unlock: + spin_unlock_irqrestore(&channel->lock, flags); + + return ret; } static int rcar_gen3_phy_usb2_exit(struct phy *p) @@ -576,9 +716,32 @@ static int rcar_gen3_phy_usb2_exit(struct phy *p) struct rcar_gen3_phy *rphy = phy_get_drvdata(p); struct rcar_gen3_chan *channel = rphy->ch; void __iomem *usb2_base = channel->base; + int retries = NUM_OF_PHYS; + unsigned long flags; u32 val; + int ret; - guard(spinlock_irqsave)(&channel->lock); + spin_lock_irqsave(&channel->lock, flags); + + while (retries-- && channel->otg_initializing) { + spin_unlock_irqrestore(&channel->lock, flags); + + ret = wait_event_timeout(channel->otg_init_done, !channel->otg_initializing, + USB2_OTG_INIT_TIMEOUT); + ret = ret ? 0 : -ETIMEDOUT; + if (ret && !retries) + return ret; + + spin_lock_irqsave(&channel->lock, flags); + } + + /* If another thread started a new initialization just return -EBUSY. */ + if (channel->otg_initializing) { + ret = -EBUSY; + goto unlock; + } else { + ret = 0; + } rphy->initialized = false; @@ -588,7 +751,9 @@ static int rcar_gen3_phy_usb2_exit(struct phy *p) val &= ~USB2_INT_ENABLE_UCOM_INTEN; writel(val, usb2_base + USB2_INT_ENABLE); - return 0; +unlock: + spin_unlock_irqrestore(&channel->lock, flags); + return ret; } static int rcar_gen3_phy_usb2_power_on(struct phy *p) @@ -596,8 +761,10 @@ static int rcar_gen3_phy_usb2_power_on(struct phy *p) struct rcar_gen3_phy *rphy = phy_get_drvdata(p); struct rcar_gen3_chan *channel = rphy->ch; void __iomem *usb2_base = channel->base; + int retries = NUM_OF_PHYS; + unsigned long flags; u32 val; - int ret = 0; + int ret; if (channel->vbus && !channel->otg_internal_reg) { ret = regulator_enable(channel->vbus); @@ -605,7 +772,27 @@ static int rcar_gen3_phy_usb2_power_on(struct phy *p) return ret; } - guard(spinlock_irqsave)(&channel->lock); + spin_lock_irqsave(&channel->lock, flags); + + while (retries-- && channel->otg_initializing) { + spin_unlock_irqrestore(&channel->lock, flags); + + ret = wait_event_timeout(channel->otg_init_done, !channel->otg_initializing, + USB2_OTG_INIT_TIMEOUT); + ret = ret ? 0 : -ETIMEDOUT; + if (ret && !retries) + goto disable_regulator; + + spin_lock_irqsave(&channel->lock, flags); + } + + /* If another thread started a new initialization just return -EBUSY. */ + if (channel->otg_initializing) { + ret = -EBUSY; + goto unlock; + } else { + ret = 0; + } if (!rcar_gen3_are_all_rphys_power_off(channel)) goto out; @@ -620,27 +807,59 @@ static int rcar_gen3_phy_usb2_power_on(struct phy *p) /* The powered flag should be set for any other phys anyway */ rphy->powered = true; - return 0; +unlock: + spin_unlock_irqrestore(&channel->lock, flags); + +disable_regulator: + if (ret && channel->vbus && !channel->otg_internal_reg) + regulator_disable(channel->vbus); + + return ret; } static int rcar_gen3_phy_usb2_power_off(struct phy *p) { struct rcar_gen3_phy *rphy = phy_get_drvdata(p); struct rcar_gen3_chan *channel = rphy->ch; - int ret = 0; + int retries = NUM_OF_PHYS; + unsigned long flags; + int ret; - scoped_guard(spinlock_irqsave, &channel->lock) { - rphy->powered = false; + spin_lock_irqsave(&channel->lock, flags); - if (rcar_gen3_are_all_rphys_power_off(channel)) { - u32 val = readl(channel->base + USB2_USBCTR); + while (retries-- && channel->otg_initializing) { + spin_unlock_irqrestore(&channel->lock, flags); - val |= USB2_USBCTR_PLL_RST; - writel(val, channel->base + USB2_USBCTR); - } + ret = wait_event_timeout(channel->otg_init_done, !channel->otg_initializing, + USB2_OTG_INIT_TIMEOUT); + ret = ret ? 0 : -ETIMEDOUT; + if (ret && !retries) + return ret; + + spin_lock_irqsave(&channel->lock, flags); } - if (channel->vbus && !channel->otg_internal_reg) + /* If another thread started a new initialization just return -EBUSY. */ + if (channel->otg_initializing) { + ret = -EBUSY; + goto unlock; + } else { + ret = 0; + } + + rphy->powered = false; + + if (rcar_gen3_are_all_rphys_power_off(channel)) { + u32 val = readl(channel->base + USB2_USBCTR); + + val |= USB2_USBCTR_PLL_RST; + writel(val, channel->base + USB2_USBCTR); + } + +unlock: + spin_unlock_irqrestore(&channel->lock, flags); + + if (!ret && channel->vbus && !channel->otg_internal_reg) ret = regulator_disable(channel->vbus); return ret; @@ -1022,6 +1241,7 @@ static int rcar_gen3_phy_usb2_probe(struct platform_device *pdev) return ret; spin_lock_init(&channel->lock); + init_waitqueue_head(&channel->otg_init_done); for (i = 0; i < NUM_OF_PHYS; i++) { channel->rphys[i].phy = devm_phy_create(dev, NULL, channel->phy_data->phy_usb2_ops); From de7f29a1fe1dc2864d8a47f8c39d508442cae167 Mon Sep 17 00:00:00 2001 From: AngeloGioacchino Del Regno Date: Fri, 11 Sep 2026 09:40:14 +0200 Subject: [PATCH 148/592] phy: mediatek: phy-mtk-hdmi-mt8195: Fix PLL calc divisor overflow When trying to calculate a PLL rate for target display resolutions above 2560x1440, 24bpp, 30Hz, the pixel clock value will be more than 32-bits long but the division to finally calculate the digital clock divider is being done with div_u64(), which expects a 32bit unsigned divisor. Fix the overflow by using div64_u64() instead. Fixes: 9d9ff3d2a4a5 ("phy: mediatek: hdmi: mt8195: fix wrong pll calculus") Reviewed-by: Manivannan Sadhasivam Signed-off-by: AngeloGioacchino Del Regno Link: https://patch.msgid.link/20260911074015.9994-2-angelogioacchino.delregno@collabora.com Signed-off-by: Vinod Koul --- drivers/phy/mediatek/phy-mtk-hdmi-mt8195.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/drivers/phy/mediatek/phy-mtk-hdmi-mt8195.c b/drivers/phy/mediatek/phy-mtk-hdmi-mt8195.c index 1426a2db984d..e6ee8e080022 100644 --- a/drivers/phy/mediatek/phy-mtk-hdmi-mt8195.c +++ b/drivers/phy/mediatek/phy-mtk-hdmi-mt8195.c @@ -290,7 +290,7 @@ static int mtk_hdmi_pll_calc(struct mtk_hdmi_phy *hdmi_phy, struct clk_hw *hw, posdiv2 = 1; /* Digital clk divider, max /32 */ - digital_div = div_u64(ns_hdmipll_ck, posdiv1 * posdiv2 * pixel_clk); + digital_div = div64_u64(ns_hdmipll_ck, posdiv1 * posdiv2 * pixel_clk); if (!(digital_div <= 32 && digital_div >= 1)) return -EINVAL; From 486a70ef848264dcf9a57f0bb0452848db9537de Mon Sep 17 00:00:00 2001 From: AngeloGioacchino Del Regno Date: Fri, 11 Sep 2026 09:40:15 +0200 Subject: [PATCH 149/592] phy: mediatek: phy-mtk-hdmi-mt8195: Fix TMDS clk bit ratio setting The comment in the mtk_phy_tmds_clk_ratio() function clearly and correctly explains that the TMDS ratio has to be 1/10 for data rates under 3.4Gbps, and 1/40 over that. Unfortunately though, the TXC_DIV register setting was wrong, as in value 3 means to divide by 8 and, in order to achieve the in spec 1/40 (tmds) data rate, this has to divide by 4 instead! Add definitions for the TXC_DIV register values clearly explaining the meanings (DIV2, DIV4, DIV8), and program the correct, DIV 4, value to the register in mtk_phy_tmds_clk_ratio(). This fixes out of spec clocking and, with this change, SoCs using the MT8195 class HDMI PHYs can now successfully be configured to output 3840x2160@60Hz over HDMI. Fixes: 45810d486bb4 ("phy: mediatek: add support for phy-mtk-hdmi-mt8195") Reviewed-by: Manivannan Sadhasivam Signed-off-by: AngeloGioacchino Del Regno Link: https://patch.msgid.link/20260911074015.9994-3-angelogioacchino.delregno@collabora.com Signed-off-by: Vinod Koul --- drivers/phy/mediatek/phy-mtk-hdmi-mt8195.c | 2 +- drivers/phy/mediatek/phy-mtk-hdmi-mt8195.h | 3 +++ 2 files changed, 4 insertions(+), 1 deletion(-) diff --git a/drivers/phy/mediatek/phy-mtk-hdmi-mt8195.c b/drivers/phy/mediatek/phy-mtk-hdmi-mt8195.c index e6ee8e080022..a4bc1268946d 100644 --- a/drivers/phy/mediatek/phy-mtk-hdmi-mt8195.c +++ b/drivers/phy/mediatek/phy-mtk-hdmi-mt8195.c @@ -36,7 +36,7 @@ mtk_phy_tmds_clk_ratio(struct mtk_hdmi_phy *hdmi_phy, bool enable) * clock bit ratio 1:40, under 3.4Gbps, clock bit ratio 1:10 */ if (enable) - mtk_phy_update_field(regs + HDMI20_CLK_CFG, REG_TXC_DIV, 3); + mtk_phy_update_field(regs + HDMI20_CLK_CFG, REG_TXC_DIV, VAL_TXC_DIV4); else mtk_phy_clear_bits(regs + HDMI20_CLK_CFG, REG_TXC_DIV); } diff --git a/drivers/phy/mediatek/phy-mtk-hdmi-mt8195.h b/drivers/phy/mediatek/phy-mtk-hdmi-mt8195.h index e26caaf4d104..58800d7659ca 100644 --- a/drivers/phy/mediatek/phy-mtk-hdmi-mt8195.h +++ b/drivers/phy/mediatek/phy-mtk-hdmi-mt8195.h @@ -17,6 +17,9 @@ #define HDMI20_CLK_CFG 0x70 #define REG_TXC_DIV GENMASK(31, 30) +#define VAL_TXC_DIV2 1 +#define VAL_TXC_DIV4 2 +#define VAL_TXC_DIV8 3 #define HDMI_1_CFG_0 0x00 #define RG_HDMITX21_DRV_IBIAS_CLK GENMASK(10, 5) From c65eae6f61d1778ff7a82e4aae4080e26f486af1 Mon Sep 17 00:00:00 2001 From: Paulo Alcantara Date: Fri, 11 Sep 2026 22:38:04 -0300 Subject: [PATCH 150/592] smb: client: cancel reconnect work in clean_demultiplex_info() clean_demultiplex_info() cancels server->echo delayed work but not server->reconnect, which can cause a use-after-free when the demultiplex thread exits while a reconnect work is still queued: cifs_demultiplex_thread() cifs_readv_from_socket() cifs_reconnect() __cifs_reconnect() cifs_queue_server_reconn() mod_delayed_work(cifsiod_wq, &server->reconnect, 0) clean_demultiplex_info() cancel_delayed_work_sync(&server->echo) // echo canceled // reconnect NOT canceled kfree_sensitive(server) // server freed ...later, on cifsiod_wq: smb2_reconnect_server() server->srv_count // UAF read of freed server Fix this by canceling server->reconnect delayed work in clean_demultiplex_info() before the server is freed, the same way cifs_put_tcp_session() already does. Reported-by: syzbot+5003556314abc915a71f@syzkaller.appspotmail.com Closes: https://lore.kernel.org/r/6aa4a12d.f81106d8.2ab401.0023.GAE@google.com Fixes: 53e0e11efe92 ("CIFS: Fix a possible memory corruption during reconnect") Reviewed-by: Namjae Jeon Signed-off-by: Paulo Alcantara Cc: David Howells Cc: Shyam Prasad N Cc: Ronnie Sahlberg Cc: Tom Talpey Cc: Bharath SM Cc: Namjae Jeon Cc: stable@vger.kernel.org --- fs/smb/client/connect.c | 1 + 1 file changed, 1 insertion(+) diff --git a/fs/smb/client/connect.c b/fs/smb/client/connect.c index b6e98eb31673..bef710d9a3c3 100644 --- a/fs/smb/client/connect.c +++ b/fs/smb/client/connect.c @@ -1067,6 +1067,7 @@ clean_demultiplex_info(struct TCP_Server_Info *server) spin_unlock(&server->srv_lock); cancel_delayed_work_sync(&server->echo); + cancel_delayed_work_sync(&server->reconnect); spin_lock(&server->srv_lock); server->tcpStatus = CifsExiting; From 5f270f091256da1338c3631083e15d7f83cc05e1 Mon Sep 17 00:00:00 2001 From: Paulo Alcantara Date: Wed, 9 Sep 2026 13:52:14 -0300 Subject: [PATCH 151/592] smb: client: fix rlist race and missing initialization TCP_Server_Info.rlist is allocated via kzalloc which zeros both ->next and ->prev to NULL instead of pointing to itself, making list_empty() always return false and list_add() dereference a NULL ->prev pointer. Also, cifs_signal_cifsd_for_reconnect() can be called concurrently from multiple cifsd threads, allowing the same server's rlist node to be added twice into the local list, corrupting it. Closes: https://sashiko.dev/#/patchset/20260911204446.1719356-1-pc%40manguebit.org Fixes: df0e03a4fb94 ("smb: client: fix potential deadlock when reconnecting channels") Reviewed-by: Namjae Jeon Signed-off-by: Paulo Alcantara Cc: David Howells Cc: Shyam Prasad N Cc: Ronnie Sahlberg Cc: Tom Talpey Cc: Bharath SM Cc: Namjae Jeon Cc: stable@vger.kernel.org --- fs/smb/client/connect.c | 7 +++++++ 1 file changed, 7 insertions(+) diff --git a/fs/smb/client/connect.c b/fs/smb/client/connect.c index bef710d9a3c3..4bd9f9fb93bb 100644 --- a/fs/smb/client/connect.c +++ b/fs/smb/client/connect.c @@ -174,6 +174,8 @@ cifs_signal_cifsd_for_reconnect(struct TCP_Server_Info *server, nserver = ses->chans[i].server; if (!nserver) continue; + if (!list_empty(&nserver->rlist)) + continue; nserver->srv_count++; list_add(&nserver->rlist, &reco); } @@ -182,11 +184,15 @@ cifs_signal_cifsd_for_reconnect(struct TCP_Server_Info *server, } } + spin_lock(&cifs_tcp_ses_lock); list_for_each_entry_safe(server, nserver, &reco, rlist) { list_del_init(&server->rlist); set_need_reco(server); + spin_unlock(&cifs_tcp_ses_lock); cifs_put_tcp_session(server, 0); + spin_lock(&cifs_tcp_ses_lock); } + spin_unlock(&cifs_tcp_ses_lock); } /* @@ -1824,6 +1830,7 @@ cifs_get_tcp_session(struct smb3_fs_context *ctx, spin_lock_init(&tcp_ses->mid_counter_lock); INIT_LIST_HEAD(&tcp_ses->tcp_ses_list); INIT_LIST_HEAD(&tcp_ses->smb_ses_list); + INIT_LIST_HEAD(&tcp_ses->rlist); INIT_DELAYED_WORK(&tcp_ses->echo, cifs_echo_request); INIT_DELAYED_WORK(&tcp_ses->reconnect, smb2_reconnect_server); mutex_init(&tcp_ses->reconnect_mutex); From e75c96157d45e498970158c8f7373d90102e33b9 Mon Sep 17 00:00:00 2001 From: Paulo Alcantara Date: Sat, 12 Sep 2026 14:20:08 -0300 Subject: [PATCH 152/592] smb: client: fix smbd_connection leak on cifs_get_tcp_session() error When an RDMA connection is successfully established via smbd_get_connection() but cifs_get_tcp_session() later fails (e.g. kthread_create() returns an error), the error path frees tcp_ses without first destroying the smbd_connection. Fix this by calling smbd_destroy() in the out_err cleanup path before kfree(tcp_ses). smbd_destroy() safely handles the case where smbd_conn is NULL, so it can be called unconditionally. Closes: https://sashiko.dev/#/patchset/20260912165503.521597-1-pc%40manguebit.org Fixes: 2f8946464b11 ("CIFS: SMBD: Upper layer connects to SMBDirect session") Reviewed-by: Namjae Jeon Signed-off-by: Paulo Alcantara Cc: Tom Talpey Cc: Stefan Metzmacher Cc: Shyam Prasad N Cc: Ronnie Sahlberg Cc: Bharath SM Cc: Namjae Jeon Cc: stable@vger.kernel.org --- fs/smb/client/connect.c | 1 + 1 file changed, 1 insertion(+) diff --git a/fs/smb/client/connect.c b/fs/smb/client/connect.c index 4bd9f9fb93bb..28e1ddeb6182 100644 --- a/fs/smb/client/connect.c +++ b/fs/smb/client/connect.c @@ -1934,6 +1934,7 @@ cifs_get_tcp_session(struct smb3_fs_context *ctx, kfree(tcp_ses->leaf_fullpath); if (tcp_ses->ssocket) sock_release(tcp_ses->ssocket); + smbd_destroy(tcp_ses); kfree(tcp_ses); } return ERR_PTR(rc); From ea9dadeac79cef509a4b8b4a3e3b39a741e63313 Mon Sep 17 00:00:00 2001 From: Dmitry Baryshkov Date: Thu, 30 Jul 2026 16:05:13 +0300 Subject: [PATCH 153/592] drm/msm: mark the fbdev framebuffer as system memory msm_fbdev_driver_fbdev_probe() points screen_buffer at a kernel virtual mapping of the GEM object and uses the deferred sysmem fb ops, but never sets FBINFO_VIRTFB. The framebuffer core then assumes the memory is not in the virtual address space and warns on the first console draw: fb0: sys_fillrect: framebuffer is not in virtual address space. The drm_fbdev_dma, drm_fbdev_shmem and drm_fbdev_ttm helpers all set the flag for system memory. Do the same here. Signed-off-by: Dmitry Baryshkov Assisted-by: Claude:claude-opus-5 Patchwork: https://patchwork.freedesktop.org/patch/743293/ Link: https://lore.kernel.org/r/20260730-drm-msm-fbinfo-virt-v1-1-a27099a6dc58@oss.qualcomm.com Acked-by: Rob Clark # on IRC Signed-off-by: Dmitry Baryshkov --- drivers/gpu/drm/msm/msm_fbdev.c | 1 + 1 file changed, 1 insertion(+) diff --git a/drivers/gpu/drm/msm/msm_fbdev.c b/drivers/gpu/drm/msm/msm_fbdev.c index 89ca9da3e1f2..dd6d6c507d77 100644 --- a/drivers/gpu/drm/msm/msm_fbdev.c +++ b/drivers/gpu/drm/msm/msm_fbdev.c @@ -155,6 +155,7 @@ int msm_fbdev_driver_fbdev_probe(struct drm_fb_helper *helper, helper->fb = buffer->fb; fbi->fbops = &msm_fb_ops; + fbi->flags |= FBINFO_VIRTFB; /* system memory */ drm_fb_helper_fill_info(fbi, helper, sizes); From e249a6e2a130c08bb4d8b0a55cbe29754307e5c9 Mon Sep 17 00:00:00 2001 From: Jesse Casco Date: Sat, 8 Aug 2026 13:13:25 -0400 Subject: [PATCH 154/592] drm/msm/dp: skip PUSH_IDLE when the link was never enabled msm_dp_display_atomic_enable() returns early when link training fails, leaving ->power_on false and the main link down. msm_dp_display_atomic_disable() nevertheless writes DP_STATE_CTRL_PUSH_IDLE and waits for an idle-pattern completion that cannot arrive, so every failed enable is followed by "PUSH_IDLE pattern timedout". Every other step of the teardown is already gated on that flag: msm_dp_display_disable(), called from .atomic_post_disable(), returns early on !power_on. The PUSH_IDLE write is the only one that is not, so the controller's runtime-PM reference is then dropped without the link having been taken down. On glymur (Snapdragon X2 Elite) the consequence is not a warning. The SoC does not survive it: TrustZone force-stops the SOCCP and ADSP remote processors and the machine resets silently about 50 ms later, with no oops and no panic. On an ASUS Zenbook A16 (UX3607OA), whose eDP panel does not currently train, this reproduces without any compositor or GPU involvement: # eDP enable has already failed with "Failed link training (rc=-104)" echo 1 > /sys/class/graphics/fb0/blank [535.645455] === marker === [535.694833] qcom_q6v5_pas d00000.remoteproc: fatal error received: \ sys_m_smsm.c:512:TZ force stop [535.694875] remoteproc remoteproc0: crash detected in soccp: type fatal error [535.728857] qcom_q6v5_pas 6800000.remoteproc: fatal error received: \ sys_m_smsm.c:783:err fatal notification received from TZ Gate the PUSH_IDLE write on ->power_on so the disable path is consistent with the rest of the teardown. With this applied the same sequence is harmless and the machine stays up; without it, it resets every time. The unconditional write dates back to the original DP driver (c943b4948b58 ("drm/msm/dp: add displayPort driver support")), but the surrounding code has been restructured several times since, so no Fixes: tag is offered. Note that the eDP link-training failure that exposes this on the A16 is a separate problem in the glymur eDP PHY and is reported separately; this change is about not damaging the machine when training fails, for whatever reason. Tested on ASUS Zenbook A16 (UX3607OA), Snapdragon X2 Elite Extreme, on linux-next next-20260803 and next-20260807. The machine has since been running next-20260807 with this patch as its daily driver. Assisted-by: Anthropic:Claude-Opus-5 Signed-off-by: Jesse Casco Reviewed-by: Dmitry Baryshkov Patchwork: https://patchwork.freedesktop.org/patch/745167/ Link: https://lore.kernel.org/r/20260808171325.133041-1-jesse.casco@gmail.com Signed-off-by: Dmitry Baryshkov --- drivers/gpu/drm/msm/dp/dp_display.c | 14 ++++++++++++++ 1 file changed, 14 insertions(+) diff --git a/drivers/gpu/drm/msm/dp/dp_display.c b/drivers/gpu/drm/msm/dp/dp_display.c index bc646d172abe..5d2ddf1808fe 100644 --- a/drivers/gpu/drm/msm/dp/dp_display.c +++ b/drivers/gpu/drm/msm/dp/dp_display.c @@ -1458,6 +1458,20 @@ void msm_dp_display_atomic_disable(struct msm_dp *dp) msm_dp_display = container_of(dp, struct msm_dp_display_private, msm_dp_display); + /* + * If .atomic_enable() bailed out - link training failure is the common + * case - the mainlink was never brought up and ->power_on stayed false. + * Driving the PUSH_IDLE pattern into a controller that was never + * enabled times out, and .atomic_post_disable() then drops the + * controller's runtime-PM reference without tearing the PHY back down, + * because msm_dp_display_disable() returns early on !power_on. On + * glymur (Snapdragon X2 Elite) that combination is answered by a + * TrustZone-level SOCCP/ADSP force-stop and a silent SoC reset. + * There is nothing to push idle, so leave it alone. + */ + if (!dp->power_on) + return; + msm_dp_ctrl_push_idle(msm_dp_display->ctrl); } From 6fbbf1e152f34ad3913e4a6476680aba672c5068 Mon Sep 17 00:00:00 2001 From: Guangshuo Li Date: Sat, 8 Aug 2026 21:16:24 +0800 Subject: [PATCH 155/592] drm/msm/adreno: fix autosuspend cleanup during teardown adreno_gpu_init() calls pm_runtime_use_autosuspend(), but adreno_gpu_cleanup() does not call the matching pm_runtime_dont_use_autosuspend() during teardown. If the autosuspend delay is set to a negative value while autosuspend is enabled, the runtime PM core increments usage_count to prevent runtime suspend. Without calling pm_runtime_dont_use_autosuspend() during teardown, this reference is not dropped and usage_count remains unbalanced. The documentation for pm_runtime_use_autosuspend() also notes that it is important to undo it with pm_runtime_dont_use_autosuspend() at driver exit time, unless runtime PM was initially enabled with devm_pm_runtime_enable(). Add the missing pm_runtime_dont_use_autosuspend() call to adreno_gpu_cleanup(). This issue was found by manual code inspection. Fixes: eeb754746b14 ("drm/msm/gpu: use pm-runtime") Cc: stable@vger.kernel.org Signed-off-by: Guangshuo Li Reviewed-by: Dmitry Baryshkov Patchwork: https://patchwork.freedesktop.org/patch/745110/ Link: https://lore.kernel.org/r/20260808131624.2854412-1-lgs201920130244@gmail.com Signed-off-by: Dmitry Baryshkov --- drivers/gpu/drm/msm/adreno/adreno_gpu.c | 2 ++ 1 file changed, 2 insertions(+) diff --git a/drivers/gpu/drm/msm/adreno/adreno_gpu.c b/drivers/gpu/drm/msm/adreno/adreno_gpu.c index ca5e4e560cde..5832dc25d6bf 100644 --- a/drivers/gpu/drm/msm/adreno/adreno_gpu.c +++ b/drivers/gpu/drm/msm/adreno/adreno_gpu.c @@ -1261,6 +1261,8 @@ void adreno_gpu_cleanup(struct adreno_gpu *adreno_gpu) for (i = 0; i < ARRAY_SIZE(adreno_gpu->info->fw); i++) release_firmware(adreno_gpu->fw[i]); + pm_runtime_dont_use_autosuspend(&gpu->pdev->dev); + if (priv && pm_runtime_enabled(&priv->gpu_pdev->dev)) pm_runtime_disable(&priv->gpu_pdev->dev); From 58995b11dfb7dda095d23f22fa4dc79b923b5adf Mon Sep 17 00:00:00 2001 From: William Bright Date: Wed, 12 Aug 2026 11:05:52 +0100 Subject: [PATCH 156/592] drm/msm/dp: fix link bandwidth check when wide bus is enabled msm_dp_display_mode_valid() halves the pixel clock when either YUV420 or wide bus is in use, then uses that halved value both for the controller pixel clock limit and for the DP link bandwidth check. Only YUV420 halves the data crossing the link. Wide bus widens the internal DPU to DP interface to two pixels per clock, halving the controller clock. Every pixel is still transmitted, so the link bandwidth requirement remains. As a result, modes needing up to twice the available link bandwidth pass validation. On the IMDT QCS8550 SBC (rev5 with CYPD6125), where DP runs over USB-C alt mode where only two lanes are available, 3840x2160@60 was accepted despite needing 9.6 Gbps against the 8.64 Gbps the link can carry. Use a separate link pixel clock that is only halved for YUV420 for the bandwidth calculation, leaving the wide bus halving to apply solely to the controller pixel clock limit. With this, 4k@60 is correctly rejected and 4k@30 selected instead. Fixes: df9cf852ca30 ("drm/msm/dp: account for widebus and yuv420 during mode validation") Assisted-by: Claude:claude-opus-5 Signed-off-by: William Bright Reviewed-by: Dmitry Baryshkov Patchwork: https://patchwork.freedesktop.org/patch/746145/ Link: https://lore.kernel.org/r/20260812-msm-dp-link-bw-v1-1-b0e3ce1190be@imd-tec.com [DB: dropped useless comment] Signed-off-by: Dmitry Baryshkov --- drivers/gpu/drm/msm/dp/dp_display.c | 7 +++++-- 1 file changed, 5 insertions(+), 2 deletions(-) diff --git a/drivers/gpu/drm/msm/dp/dp_display.c b/drivers/gpu/drm/msm/dp/dp_display.c index 5d2ddf1808fe..4dcbd9b99d06 100644 --- a/drivers/gpu/drm/msm/dp/dp_display.c +++ b/drivers/gpu/drm/msm/dp/dp_display.c @@ -756,6 +756,7 @@ enum drm_mode_status msm_dp_display_mode_valid(struct msm_dp *dp, struct msm_dp_link_info *link_info; u32 mode_rate_khz = 0, supported_rate_khz = 0, mode_bpp = 0; int mode_pclk_khz = mode->clock; + int link_pclk_khz; bool is_yuv_420; if (!dp || !mode_pclk_khz || !dp->connector) { @@ -775,6 +776,8 @@ enum drm_mode_status msm_dp_display_mode_valid(struct msm_dp *dp, if (is_yuv_420 && !msm_dp_display->panel->vsc_sdp_supported) return MODE_NO_420; + link_pclk_khz = is_yuv_420 ? mode_pclk_khz / 2 : mode_pclk_khz; + if (is_yuv_420 || msm_dp_display->wide_bus_supported) mode_pclk_khz /= 2; @@ -786,9 +789,9 @@ enum drm_mode_status msm_dp_display_mode_valid(struct msm_dp *dp, mode_bpp = default_bpp; mode_bpp = msm_dp_panel_get_mode_bpp(msm_dp_display->panel, - mode_bpp, mode_pclk_khz); + mode_bpp, link_pclk_khz); - mode_rate_khz = mode_pclk_khz * mode_bpp; + mode_rate_khz = link_pclk_khz * mode_bpp; supported_rate_khz = link_info->num_lanes * link_info->rate * 8; if (mode_rate_khz > supported_rate_khz) From a5b5cc909931572aec446e129c035b76b3f0c1fa Mon Sep 17 00:00:00 2001 From: Saim Shujah Date: Fri, 28 Aug 2026 11:54:40 +0500 Subject: [PATCH 157/592] drm/msm/dpu: clear pending peripheral flush state dpu_hw_ctl_clear_pending_flush() resets the cached per-block state after a flush transaction, but misses pending_periph_flush_mask. The peripheral flush updater accumulates interface bits in this mask. A later transaction which sets the top-level peripheral flush bit can write stale interface bits to CTL_PERIPH_FLUSH together with the current state. Peripheral flush support was added after the helper started clearing every individual pending flush mask. Clear the peripheral mask together with the other cached child masks. Fixes: 64f7b81f0358 ("drm/msm/dpu: add support of new peripheral flush mechanism") Cc: stable@vger.kernel.org Signed-off-by: Saim Shujah Patchwork: https://patchwork.freedesktop.org/patch/748968/ Link: https://lore.kernel.org/r/20260828065440.140410-1-saimzst@gmail.com Signed-off-by: Dmitry Baryshkov --- drivers/gpu/drm/msm/disp/dpu1/dpu_hw_ctl.c | 1 + 1 file changed, 1 insertion(+) diff --git a/drivers/gpu/drm/msm/disp/dpu1/dpu_hw_ctl.c b/drivers/gpu/drm/msm/disp/dpu1/dpu_hw_ctl.c index 36a497f1d6c1..e8729fe1cfef 100644 --- a/drivers/gpu/drm/msm/disp/dpu1/dpu_hw_ctl.c +++ b/drivers/gpu/drm/msm/disp/dpu1/dpu_hw_ctl.c @@ -118,6 +118,7 @@ static inline void dpu_hw_ctl_clear_pending_flush(struct dpu_hw_ctl *ctx) ctx->pending_intf_flush_mask = 0; ctx->pending_wb_flush_mask = 0; ctx->pending_cwb_flush_mask = 0; + ctx->pending_periph_flush_mask = 0; ctx->pending_merge_3d_flush_mask = 0; ctx->pending_dsc_flush_mask = 0; ctx->pending_cdm_flush_mask = 0; From 26eb3d92c7a4d7adb1ae1740ca6e8e100b11d1ec Mon Sep 17 00:00:00 2001 From: =?UTF-8?q?Rapha=C3=ABl=20Larocque?= Date: Thu, 10 Sep 2026 12:44:25 -0400 Subject: [PATCH 158/592] Input: synaptics - disable InterTouch on ThinkPad T440p (board id 2722) MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit The Lenovo ThinkPad T440p (PNP ID LEN0036, board id 2722) has a Synaptics touchpad whose SMBus companion is not ready at boot and takes roughly 200 seconds to appear. During this window the touchpad and TrackPoint are completely unresponsive on approximately 50% of boots, making the machine unusable until the companion finally registers. The device is in the topbuttonpad_pnp_ids[] SMBus allowlist, so the kernel attempts to use SMBus/RMI4 mode by default. When the companion is not ready, psmouse_smbus_init() leaves breadcrumbs and returns -EAGAIN, the PS/2 fallback path is taken, but the device does not function properly until the companion appears and RMI4 takes over. Disable SMBus InterTouch for board id 2722 so the touchpad and TrackPoint work immediately via PS/2 from boot. Users can still force SMBus with psmouse.synaptics_intertouch=1 if needed. Tested-by: Raphaël Larocque Signed-off-by: Raphaël Larocque Link: https://patch.msgid.link/20260910164425.12832-1-rlarocque@disroot.org Cc: stable@vger.kernel.org Signed-off-by: Dmitry Torokhov --- drivers/input/mouse/synaptics.c | 8 ++++++++ 1 file changed, 8 insertions(+) diff --git a/drivers/input/mouse/synaptics.c b/drivers/input/mouse/synaptics.c index 2170bbe4c589..a8ce89d41faa 100644 --- a/drivers/input/mouse/synaptics.c +++ b/drivers/input/mouse/synaptics.c @@ -1838,6 +1838,14 @@ static int synaptics_setup_intertouch(struct psmouse *psmouse, return -ENXIO; } + + /* Disable intertouch on known-broken board revisions */ + if (info->board_id == 2722) { + psmouse_info(psmouse, + "Disabling intertouch for board id %u\n", + info->board_id); + return -ENXIO; + } } psmouse_info(psmouse, "Trying to set up SMBus access\n"); From 25e424eb4ae1a662d9c3573218d06ac32f797fc5 Mon Sep 17 00:00:00 2001 From: Chris Sommers Date: Mon, 7 Sep 2026 11:27:23 -0700 Subject: [PATCH 159/592] Input: i8042 - add quirk for Acer Aspire Go 15 AG15-42P On the Acer Aspire Go 15 (AG15-42P), the internal keyboard drops out ~5 seconds after boot on both Linux and Linux-LTS kernels. Keystrokes on the built-in keyboard stop registering while the trackpad and external keyboards remain functional. Testing confirms that booting with the i8042.reset kernel parameter resolves the issue and keeps the internal keyboard responsive. Add SERIO_QUIRK_RESET_ALWAYS to i8042_dmi_quirk_table for the Acer Aspire AG15-42P to automatically apply this quirk on boot. Signed-off-by: Chris Sommers Link: https://patch.msgid.link/20260907182723.2709981-1-chris.sommers@icloud.com Cc: stable@vger.kernel.org Signed-off-by: Dmitry Torokhov --- drivers/input/serio/i8042-acpipnpio.h | 7 +++++++ 1 file changed, 7 insertions(+) diff --git a/drivers/input/serio/i8042-acpipnpio.h b/drivers/input/serio/i8042-acpipnpio.h index 9ecb0eed48c4..54eb3687000a 100644 --- a/drivers/input/serio/i8042-acpipnpio.h +++ b/drivers/input/serio/i8042-acpipnpio.h @@ -259,6 +259,13 @@ static const struct dmi_system_id i8042_dmi_quirk_table[] __initconst = { }, .driver_data = (void *)(SERIO_QUIRK_RESET_ALWAYS) }, + { + .matches = { + DMI_MATCH(DMI_SYS_VENDOR, "Acer"), + DMI_MATCH(DMI_PRODUCT_NAME, "Aspire AG15-42P"), + }, + .driver_data = (void *)(SERIO_QUIRK_RESET_ALWAYS) + }, { .matches = { DMI_MATCH(DMI_SYS_VENDOR, "Acer"), From 930a7312c946bf4731721cadd82bb9a2ada496ca Mon Sep 17 00:00:00 2001 From: Gabor Juhos Date: Wed, 9 Sep 2026 16:08:25 +0200 Subject: [PATCH 160/592] spi: spi-qpic-snand: avoid writing QPIC_EBI2_ECC_BUF_CFG register The description of commit bfb34eced559 ("mtd: rawnand: qcom: avoid writing to obsolete register") says this: "QPIC_EBI2_ECC_BUF_CFG register got obsolete from QPIC V2.0 onwards. Avoid writing this register if QPIC version is V2.0 or newer." Although the referenced commit is related to the 'qcom-nandc' driver, however the hardware supported by the current driver is also based on QPIC v2.0 so we should avoid writing that register here as well. Remove the register writing code to avoid undefined behaviour. Fixes: 7304d1909080 ("spi: spi-qpic: add driver for QCOM SPI NAND flash Interface") Signed-off-by: Gabor Juhos Reviewed-by: Md Sadre Alam Link: https://patch.msgid.link/20260909-qpic-snand-avoid-ebi2-reg-write-v1-1-9b1b1466cc75@gmail.com Signed-off-by: Mark Brown --- drivers/spi/spi-qpic-snand.c | 4 ---- 1 file changed, 4 deletions(-) diff --git a/drivers/spi/spi-qpic-snand.c b/drivers/spi/spi-qpic-snand.c index 61b1f2eb19ce..05efe6313b7f 100644 --- a/drivers/spi/spi-qpic-snand.c +++ b/drivers/spi/spi-qpic-snand.c @@ -765,8 +765,6 @@ static int qcom_spi_read_cw_raw(struct qcom_nand_controller *snandc, u8 *data_bu qcom_write_reg_dma(snandc, &snandc->regs->addr0, NAND_ADDR0, 2, 0); qcom_write_reg_dma(snandc, &snandc->regs->cfg0, NAND_DEV0_CFG0, 3, 0); - qcom_write_reg_dma(snandc, &snandc->regs->ecc_buf_cfg, NAND_EBI2_ECC_BUF_CFG, 1, 0); - qcom_write_reg_dma(snandc, &snandc->regs->erased_cw_detect_cfg_clr, NAND_ERASED_CW_DETECT_CFG, 1, 0); qcom_write_reg_dma(snandc, &snandc->regs->erased_cw_detect_cfg_set, @@ -1104,8 +1102,6 @@ static void qcom_spi_config_page_write(struct qcom_nand_controller *snandc) { qcom_write_reg_dma(snandc, &snandc->regs->addr0, NAND_ADDR0, 2, 0); qcom_write_reg_dma(snandc, &snandc->regs->cfg0, NAND_DEV0_CFG0, 3, 0); - qcom_write_reg_dma(snandc, &snandc->regs->ecc_buf_cfg, NAND_EBI2_ECC_BUF_CFG, - 1, NAND_BAM_NEXT_SGL); } static void qcom_spi_config_cw_write(struct qcom_nand_controller *snandc) From ed22ad5fdbdbf9b4cb4ad3003f60314b5a5eb89d Mon Sep 17 00:00:00 2001 From: Hans de Goede Date: Wed, 9 Sep 2026 11:39:33 +0200 Subject: [PATCH 161/592] Input: soc_button_array - fix MS Surface Pro 11 probe failure On the MS Surface Pro 11 soc_button_array probing races with the GPIO driver probing. If soc_button_array wins the race then gpiod_get() returns EPROBE_DEFER, which should normally take care of retrying later, but the soc_button_array code deliberately ignores EPROBE_DEFER causing it to fail its probe() which causes the volume and power buttons to now work. The ignoring of EPROBE_DEFER is there to deal with a problem specific to older Bay Trail (BYT) and Cherry Trail (CHT) tablets which often use this driver. Modify the error handling to only ignore EPROBE_DEFER on BYT and CHT platforms and propagate EPROBE_DEFER normally on other platforms. Fixes: bcf059578980 ("Input: soc_button_array - partial revert of support for newer surface devices") Cc: stable@vger.kernel.org Reported-by: Sergey Lebedev Closes: https://lore.kernel.org/lkml/20260830141355.55898-1-lsa.uz@pm.me/ Signed-off-by: Hans de Goede Tested-by: Sergey Lebedev Link: https://patch.msgid.link/20260909093934.29411-1-johannes.goede@oss.qualcomm.com Signed-off-by: Dmitry Torokhov --- drivers/input/misc/soc_button_array.c | 14 +++++++++++--- 1 file changed, 11 insertions(+), 3 deletions(-) diff --git a/drivers/input/misc/soc_button_array.c b/drivers/input/misc/soc_button_array.c index b8cad415c62c..264c41f80d2b 100644 --- a/drivers/input/misc/soc_button_array.c +++ b/drivers/input/misc/soc_button_array.c @@ -16,6 +16,7 @@ #include #include #include +#include #include static bool use_low_level_irq; @@ -160,7 +161,7 @@ soc_button_device_create(struct platform_device *pdev, struct gpio_keys_platform_data *gpio_keys_pdata; const struct dmi_system_id *dmi_id; int invalid_acpi_index = -1; - int error, gpio, irq; + int error, gpio, irq = 0; int n_buttons = 0; for (info = button_info; info->name; info++) @@ -191,8 +192,9 @@ soc_button_device_create(struct platform_device *pdev, error = soc_button_lookup_gpio(&pdev->dev, info->acpi_index, &gpio, &irq); if (error || irq < 0) { /* - * Skip GPIO if not present. Note we deliberately - * ignore -EPROBE_DEFER errors here. On some devices + * Propagate -EPROBE_DEFER, skip button on other errors. + * + * -EPROBE_DEFER is ignored on Bay & Cherry Trail. Here * Intel is using so called virtual GPIOs which are not * GPIOs at all but some way for AML code to check some * random status bits without need a custom opregion. @@ -201,6 +203,12 @@ soc_button_device_create(struct platform_device *pdev, * we do not have a driver for these so they will never * show up, therefore we ignore -EPROBE_DEFER. */ + if ((error == -EPROBE_DEFER || irq == -EPROBE_DEFER) && + !(soc_intel_is_byt() || soc_intel_is_cht())) { + error = -EPROBE_DEFER; + goto err_free_mem; + } + continue; } From fb5022278b6ea7f1838e3ef78028d5d5e3375f65 Mon Sep 17 00:00:00 2001 From: Hans de Goede Date: Wed, 9 Sep 2026 11:39:34 +0200 Subject: [PATCH 162/592] Input: soc_button_array - check btns_desc->package.count Check that btns_desc->package.count is not 0 before accessing btns_desc->package.elements[0]. Fixes: 4c3362f44980 ("Input: soc_button_array - add support for ACPI 6.0 Generic Button Device") Cc: stable@vger.kernel.org Reported-by: Shashiko Closes: https://lore.kernel.org/linux-input/20260909091440.3384C1F00A3A@smtp.kernel.org/ Signed-off-by: Hans de Goede Link: https://patch.msgid.link/20260909093934.29411-2-johannes.goede@oss.qualcomm.com Signed-off-by: Dmitry Torokhov --- drivers/input/misc/soc_button_array.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/drivers/input/misc/soc_button_array.c b/drivers/input/misc/soc_button_array.c index 264c41f80d2b..f08e29af531f 100644 --- a/drivers/input/misc/soc_button_array.c +++ b/drivers/input/misc/soc_button_array.c @@ -377,7 +377,7 @@ static struct soc_button_info *soc_button_get_button_info(struct device *dev) } } - if (!btns_desc) { + if (!btns_desc || !btns_desc->package.count) { dev_err(dev, "ACPI Button Descriptors not found\n"); button_info = ERR_PTR(-ENODEV); goto out; From ea48250a0dc9708c198e8665391973bfc6c86fd2 Mon Sep 17 00:00:00 2001 From: Dmitry Torokhov Date: Sun, 13 Sep 2026 17:19:48 -0700 Subject: [PATCH 163/592] Input: document that no new LED codes should be added Add a comment to input-event-codes.h clarifying that no new LED definitions should be added to the input subsystem. The existing LED_* definitions are legacy and grandfathered for backwards compatibility with userspace via evdev. Any new LED indicators should instead use the dedicated LED subsystem. Signed-off-by: Dmitry Torokhov --- include/uapi/linux/input-event-codes.h | 6 ++++++ 1 file changed, 6 insertions(+) diff --git a/include/uapi/linux/input-event-codes.h b/include/uapi/linux/input-event-codes.h index 3528168f7c6d..4217950e5d16 100644 --- a/include/uapi/linux/input-event-codes.h +++ b/include/uapi/linux/input-event-codes.h @@ -970,6 +970,12 @@ /* * LEDs + * + * Do not add any new LED definitions to the input subsystem. The existing + * definitions are legacy and grandfathered for backwards compatibility with + * userspace (via evdev). Any new LEDs should be implemented using the + * LED subsystem (struct led_classdev). The input core provides a bridge to + * the LED subsystem in drivers/input/input-leds.c. */ #define LED_NUML 0x00 From 7bc369cb3d3f3656eb77285628ee264264d28ad4 Mon Sep 17 00:00:00 2001 From: Jeremy Nyberg Date: Sun, 13 Sep 2026 17:43:02 -0700 Subject: [PATCH 164/592] Input: xpad - fix PDP Marvel Xbox 360 controller The PDP Marvel Xbox 360 controller with USB ID 0e6f:0147 is incorrectly classified as an Xbox One controller. With the current XTYPE_XBOXONE classification, the controller is detected but produces no input, while its four player LEDs continue blinking indefinitely. Classify USB ID 0e6f:0147 as an Xbox 360 controller instead. Tested on a PDP Marvel Xbox 360 controller with USB ID 0e6f:0147. All inputs register correctly and the player LED indicates the current player. Fixes: c225370e01b8 ("Input: xpad - sync supported devices with 360Controller") Cc: stable@vger.kernel.org Signed-off-by: Jeremy Nyberg Link: https://patch.msgid.link/20260910071627.236014-1-SlickStretch3.0@gmail.com Signed-off-by: Dmitry Torokhov --- drivers/input/joystick/xpad.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/drivers/input/joystick/xpad.c b/drivers/input/joystick/xpad.c index 2da0b7f1722a..b7a9a940d250 100644 --- a/drivers/input/joystick/xpad.c +++ b/drivers/input/joystick/xpad.c @@ -215,7 +215,7 @@ static const struct xpad_device { { 0x0e6f, 0x0139, "Afterglow Prismatic Wired Controller", 0, XTYPE_XBOXONE }, { 0x0e6f, 0x013a, "PDP Xbox One Controller", 0, XTYPE_XBOXONE }, { 0x0e6f, 0x0146, "Rock Candy Wired Controller for Xbox One", 0, XTYPE_XBOXONE }, - { 0x0e6f, 0x0147, "PDP Marvel Xbox One Controller", 0, XTYPE_XBOXONE }, + { 0x0e6f, 0x0147, "PDP Marvel Xbox 360 Controller", 0, XTYPE_XBOX360 }, { 0x0e6f, 0x015c, "PDP Xbox One Arcade Stick", MAP_TRIGGERS_TO_BUTTONS, XTYPE_XBOXONE }, { 0x0e6f, 0x015d, "PDP Mirror's Edge Official Wired Controller for Xbox One", 0, XTYPE_XBOXONE }, { 0x0e6f, 0x0161, "PDP Xbox One Controller", 0, XTYPE_XBOXONE }, From cba76c0f47af1a389d718c5bb69e75cbd67bba98 Mon Sep 17 00:00:00 2001 From: Roberts Kursitis Date: Sun, 6 Sep 2026 17:30:40 +0300 Subject: [PATCH 165/592] Input: xpad - add support for Azeron devices Azeron controllers (Cyro, Cyborg, Classic/Compact, Cyro Lefty, Cyborg II and Keyzen) present a standard Xbox 360 controller interface, so they work with the existing xpad driver once their USB IDs are added. The 0x16d0 vendor ID is a shared block, but this is safe because xpad only binds interfaces that match the Xbox 360 signature. Tested with an Azeron Keyzen. Signed-off-by: Roberts Kursitis Cc: stable@vger.kernel.org Link: https://patch.msgid.link/20260906143040.162418-1-roberts.kursitis@azeron.eu Signed-off-by: Dmitry Torokhov --- drivers/input/joystick/xpad.c | 7 +++++++ 1 file changed, 7 insertions(+) diff --git a/drivers/input/joystick/xpad.c b/drivers/input/joystick/xpad.c index b7a9a940d250..8e4e06b53c2a 100644 --- a/drivers/input/joystick/xpad.c +++ b/drivers/input/joystick/xpad.c @@ -292,6 +292,12 @@ static const struct xpad_device { { 0x1689, 0xfd00, "Razer Onza Tournament Edition", 0, XTYPE_XBOX360 }, { 0x1689, 0xfd01, "Razer Onza Classic Edition", 0, XTYPE_XBOX360 }, { 0x1689, 0xfe00, "Razer Sabertooth", 0, XTYPE_XBOX360 }, + { 0x16d0, 0x1103, "Azeron Cyro", 0, XTYPE_XBOX360 }, + { 0x16d0, 0x113c, "Azeron Cyborg", 0, XTYPE_XBOX360 }, + { 0x16d0, 0x1192, "Azeron Classic/Compact", 0, XTYPE_XBOX360 }, + { 0x16d0, 0x1212, "Azeron Cyro Lefty", 0, XTYPE_XBOX360 }, + { 0x16d0, 0x12f7, "Azeron Cyborg II", 0, XTYPE_XBOX360 }, + { 0x16d0, 0x13ea, "Azeron Keyzen", 0, XTYPE_XBOX360 }, { 0x17ef, 0x6182, "Lenovo Legion Controller for Windows", 0, XTYPE_XBOX360 }, { 0x1949, 0x041a, "Amazon Game Controller", 0, XTYPE_XBOX360 }, { 0x1a86, 0xe310, "Legion Go S", 0, XTYPE_XBOX360 }, @@ -534,6 +540,7 @@ static const struct usb_device_id xpad_table[] = { XPAD_XBOX360_VENDOR(0x15e4), /* Numark Xbox 360 controllers */ XPAD_XBOX360_VENDOR(0x162e), /* Joytech Xbox 360 controllers */ XPAD_XBOX360_VENDOR(0x1689), /* Razer Onza */ + XPAD_XBOX360_VENDOR(0x16d0), /* Azeron controllers */ XPAD_XBOX360_VENDOR(0x17ef), /* Lenovo */ XPAD_XBOX360_VENDOR(0x1949), /* Amazon controllers */ XPAD_XBOX360_VENDOR(0x1a86), /* Nanjing Qinheng Microelectronics (WCH) */ From a52ae68a937efc353251aec27fc995ff66cbe1ca Mon Sep 17 00:00:00 2001 From: "hpp.iscas" Date: Sat, 5 Sep 2026 21:40:04 +0800 Subject: [PATCH 166/592] Input: eeti_ts - publish the OF module alias The EETI driver matches eeti,exc3000-i2c Device Tree clients, but only publishes the legacy eeti_ts I2C ID. The I2C core emits an OF modalias for a Device Tree client. Publish the existing OF match table within its CONFIG_OF guard. Fixes: e32d7f1b246c ("Input: eeti - add device tree matching table") Signed-off-by: hpp.iscas Link: https://patch.msgid.link/20260905134004.66336-1-hppiscas@163.com Signed-off-by: Dmitry Torokhov --- drivers/input/touchscreen/eeti_ts.c | 1 + 1 file changed, 1 insertion(+) diff --git a/drivers/input/touchscreen/eeti_ts.c b/drivers/input/touchscreen/eeti_ts.c index b12602bc368e..b870a939508b 100644 --- a/drivers/input/touchscreen/eeti_ts.c +++ b/drivers/input/touchscreen/eeti_ts.c @@ -276,6 +276,7 @@ static const struct of_device_id of_eeti_ts_match[] = { { .compatible = "eeti,exc3000-i2c", }, { } }; +MODULE_DEVICE_TABLE(of, of_eeti_ts_match); #endif static struct i2c_driver eeti_ts_driver = { From 45b0037899704caf9078be2be4de69361ca7d933 Mon Sep 17 00:00:00 2001 From: Karl Mehltretter Date: Sat, 5 Sep 2026 12:20:38 +0200 Subject: [PATCH 167/592] Input: trackpoint - fix the inertia attribute name in the ABI document The attribute is created as "inertia" (TRACKPOINT_INT_ATTR(inertia, ...) in drivers/input/mouse/trackpoint.c); the ABI file spells the path "intertia". The description below it already says inertia. Fix the spelling. Fixes: aebb47d4e7a9 ("Input: trackpoint: document sysfs interface") Assisted-by: LLM Signed-off-by: Karl Mehltretter Link: https://patch.msgid.link/20260905102038.42882-1-kmehltretter@gmail.com Signed-off-by: Dmitry Torokhov --- Documentation/ABI/testing/sysfs-devices-platform-trackpoint | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/Documentation/ABI/testing/sysfs-devices-platform-trackpoint b/Documentation/ABI/testing/sysfs-devices-platform-trackpoint index df11901a6b3d..7954434b0434 100644 --- a/Documentation/ABI/testing/sysfs-devices-platform-trackpoint +++ b/Documentation/ABI/testing/sysfs-devices-platform-trackpoint @@ -5,7 +5,7 @@ Contact: linux-input@vger.kernel.org Description: (RW) Trackpoint sensitivity. -What: /sys/devices/platform/i8042/.../intertia +What: /sys/devices/platform/i8042/.../inertia Date: Aug, 2005 KernelVersion: 2.6.14 Contact: linux-input@vger.kernel.org From 7e61560628d17ea6b1d8ee370f6d42694cff8758 Mon Sep 17 00:00:00 2001 From: Ulises Mendez Martinez Date: Fri, 4 Sep 2026 15:07:09 +0000 Subject: [PATCH 168/592] objtool: Validate disassembler headers in libopcodes probe commit 3f2de814c059 ("objtool: Fix libopcodes linking with static libraries") tested for libopcodes availability by linking a test snippet with a forward declaration of disassemble_init_for_target(). However, testing symbol linkage with an extern declaration only verifies the presence of the library (.so/.a) and bypasses checking for development headers (binutils-dev). On systems where libopcodes is present without development headers installed, the probe succeeds, enabling BUILD_DISAS. Subsequent compilation of objtool then fails: fatal error: 'bfd.h' file not found 113 | #include Additionally, the probe invokes $(HOSTCC) without $(HOSTCFLAGS), ignoring any sysroot or include flags specified for the host compiler. Fix this by including and directly in the test snippet, passing $(HOSTCFLAGS) so host compiler options are respected, and defining PACKAGE="objtool" to satisfy the configuration check in . Fixes: 3f2de814c059 ("objtool: Fix libopcodes linking with static libraries") Fixes: 436326bc525d ("objtool: fix build failure due to missing libopcodes check") Reported-by: Alice Ryhl Assisted-by: Antigravity:Gemini-Next Signed-off-by: Ulises Mendez Martinez Link: https://patch.msgid.link/20260904150710.2997558-1-umendez@google.com Signed-off-by: Josh Poimboeuf --- tools/objtool/Makefile | 8 +++++--- 1 file changed, 5 insertions(+), 3 deletions(-) diff --git a/tools/objtool/Makefile b/tools/objtool/Makefile index a4484fd22a96..4cc2e756af84 100644 --- a/tools/objtool/Makefile +++ b/tools/objtool/Makefile @@ -89,9 +89,11 @@ LIBOPCODES_LIBS := $(shell \ "-lopcodes -lbfd" \ "-lopcodes -lbfd -liberty" \ "-lopcodes -lbfd -liberty -lz"; do \ - echo 'extern void disassemble_init_for_target(void *);' \ - 'int main(void) { disassemble_init_for_target(0); return 0; }' | \ - $(HOSTCC) -xc - -o /dev/null $$libs 2>/dev/null && \ + printf '%s\n' \ + '$(pound)include ' \ + '$(pound)include ' \ + 'int main(void) { disassemble_init_for_target(0); return 0; }' | \ + $(HOSTCC) $(HOSTCFLAGS) -DPACKAGE='"objtool"' -xc - -o /dev/null $$libs 2>/dev/null && \ echo "$$libs" && break; \ done) From 55fc280e951ab2b39f3dcb640edc1b1eebc6d173 Mon Sep 17 00:00:00 2001 From: "Rob Herring (Arm)" Date: Mon, 31 Aug 2026 14:43:51 -0500 Subject: [PATCH 169/592] Input: tsc2007 - read "ti,poll-period" as u32 The "ti,poll-period" property is documented as a normal uint32 cell. The driver used a u64 helper, which makes the helper type disagree with the schema even though the stored value is still small. Read "ti,poll-period" with the u32 helper matching the documented DT cell size. Assisted-by: Codex:gpt-5-5 Signed-off-by: Rob Herring (Arm) Link: https://patch.msgid.link/20260831194352.1185860-1-robh@kernel.org Signed-off-by: Dmitry Torokhov --- drivers/input/touchscreen/tsc2007_core.c | 5 ++--- 1 file changed, 2 insertions(+), 3 deletions(-) diff --git a/drivers/input/touchscreen/tsc2007_core.c b/drivers/input/touchscreen/tsc2007_core.c index e4d7da0f4434..e2f49b37e18c 100644 --- a/drivers/input/touchscreen/tsc2007_core.c +++ b/drivers/input/touchscreen/tsc2007_core.c @@ -221,7 +221,6 @@ static int tsc2007_get_pendown_state_gpio(struct device *dev) static int tsc2007_probe_properties(struct device *dev, struct tsc2007 *ts) { u32 val32; - u64 val64; if (!device_property_read_u32(dev, "ti,max-rt", &val32)) ts->max_rt = val32; @@ -237,8 +236,8 @@ static int tsc2007_probe_properties(struct device *dev, struct tsc2007 *ts) if (!device_property_read_u32(dev, "ti,fuzzz", &val32)) ts->fuzzz = val32; - if (!device_property_read_u64(dev, "ti,poll-period", &val64)) - ts->poll_period = msecs_to_jiffies(val64); + if (!device_property_read_u32(dev, "ti,poll-period", &val32)) + ts->poll_period = msecs_to_jiffies(val32); else ts->poll_period = msecs_to_jiffies(1); From 971fa7ea8621e123feb9c8d7dc61be1c656bd945 Mon Sep 17 00:00:00 2001 From: Erich Sartison Date: Thu, 3 Sep 2026 12:31:37 +0200 Subject: [PATCH 170/592] Input: xpad - add support for Victrix Pro BFG Controller The controller doesn't currently work via USB-cable. Signed-off-by: Erich Sartison Link: https://patch.msgid.link/20260903103137.630170-1-byt.es@mailbox.org Cc: stable@vger.kernel.org Signed-off-by: Dmitry Torokhov --- drivers/input/joystick/xpad.c | 1 + 1 file changed, 1 insertion(+) diff --git a/drivers/input/joystick/xpad.c b/drivers/input/joystick/xpad.c index 8e4e06b53c2a..6406ceab988a 100644 --- a/drivers/input/joystick/xpad.c +++ b/drivers/input/joystick/xpad.c @@ -227,6 +227,7 @@ static const struct xpad_device { { 0x0e6f, 0x0213, "Afterglow Gamepad for Xbox 360", 0, XTYPE_XBOX360 }, { 0x0e6f, 0x021f, "Rock Candy Gamepad for Xbox 360", 0, XTYPE_XBOX360 }, { 0x0e6f, 0x0246, "Rock Candy Gamepad for Xbox One 2015", 0, XTYPE_XBOXONE }, + { 0x0e6f, 0x024c, "PDP Victrix Pro BFG Wired Controller for Xbox", 0, XTYPE_XBOXONE }, { 0x0e6f, 0x02a0, "PDP Xbox One Controller", 0, XTYPE_XBOXONE }, { 0x0e6f, 0x02a1, "PDP Xbox One Controller", 0, XTYPE_XBOXONE }, { 0x0e6f, 0x02a2, "PDP Wired Controller for Xbox One - Crimson Red", 0, XTYPE_XBOXONE }, From 309731e95917125bbd13626a7a5600490a5bf44f Mon Sep 17 00:00:00 2001 From: Runyu Xiao Date: Wed, 2 Sep 2026 23:40:04 +0800 Subject: [PATCH 171/592] Input: hp_sdc - shut down kicker timer on module exit hp_sdc_kicker() rearms hp_sdc.kicker with mod_timer() after scheduling the tasklet. The module exit path uses timer_delete_sync(). That waits for a callback already running but can still leave the timer rearmed. A callback can therefore leave the timer pending while hp_sdc_exit() tears down the driver, allowing timer activity to access dismantled driver state. Use timer_shutdown_sync() for final teardown. It waits for a running callback and prevents rearming after module exit begins. Fixes: 1da177e4c3f4 ("Linux-2.6.12-rc2") Cc: stable@vger.kernel.org Assisted-by: Codex:GPT-5 Signed-off-by: Runyu Xiao Acked-by: Helge Deller Link: https://patch.msgid.link/20260902154004.3595416-1-runyu.xiao@seu.edu.cn Signed-off-by: Dmitry Torokhov --- drivers/input/serio/hp_sdc.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/drivers/input/serio/hp_sdc.c b/drivers/input/serio/hp_sdc.c index 1461ef319f92..ecf5347b63ab 100644 --- a/drivers/input/serio/hp_sdc.c +++ b/drivers/input/serio/hp_sdc.c @@ -981,7 +981,7 @@ static void hp_sdc_exit(void) free_irq(hp_sdc.irq, &hp_sdc); write_unlock_irq(&hp_sdc.lock); - timer_delete_sync(&hp_sdc.kicker); + timer_shutdown_sync(&hp_sdc.kicker); tasklet_kill(&hp_sdc.task); From 1c21452d02eec2f008e2c5535820f85adbd7587a Mon Sep 17 00:00:00 2001 From: Pu Lehui Date: Sat, 5 Sep 2026 02:11:39 +0000 Subject: [PATCH 172/592] bpf: Fix UAF due to concurrent consumption of ttrace lists in alloc_bulk Syzkaller repeatedly triggered UAF splats related to nodes in waiting_for_gp_ttrace within the bpf memalloc: BUG: KASAN: slab-use-after-free in llist_del_first+0x85/0x110 lib/llist.c:61 Read of size 8 at addr ffff8881572cd080 by task syz.4.470/5112 ... llist_del_first+0x85/0x110 lib/llist.c:61 alloc_bulk+0x193/0x460 kernel/bpf/memalloc.c:229 bpf_mem_refill+0x386/0x560 kernel/bpf/memalloc.c:436 Freed by task 14: ... __free_rcu kernel/bpf/memalloc.c:281 [inline] __free_rcu_tasks_trace+0x48/0xd0 kernel/bpf/memalloc.c:291 rcu_tasks_invoke_cbs+0x1ec/0x3e0 kernel/rcu/tasks.h:571 rcu_tasks_one_gp+0x13d/0x220 kernel/rcu/tasks.h:621 rcu_tasks_kthread+0xf3/0x120 kernel/rcu/tasks.h:651 The reason is that the UAF occurs after the RCU Tasks Trace GP expires: when the __free_rcu() callback runs, there is no synchronization protecting llist_del_all() against concurrent alloc_bulk() operating on waiting_for_gp_ttrace, leading to the race condition below: CPU0 CPU1 __free_rcu (RCU Tasks Trace callback) alloc_bulk llist_del_first(&c->waiting_for_gp_ttrace) entry = smp_load_acquire(&head->first); do { if (entry == NULL) return NULL; free_all(llist_del_all(&c->waiting_for_gp_ttrace)) llist_for_each_safe(pos, t, llnode) free_one(pos); next = READ_ONCE(entry->next); <-- trigger UAF } while (!try_cmpxchg(&head->first, &entry, next)); In addition, there is also a theoretical race condition on the free_by_rcu_ttrace list. This race requires two preconditions: an in-flight Tasks Trace GP keeping c->call_rcu_ttrace_in_progress == 1, and concurrent cross-CPU frees repopulating c->free_by_rcu_ttrace with new nodes. Under these conditions, the following scenario triggers UAF: // CPU0 // irq work is still busy (on PREEMPT_RT) alloc_bulk() llist_del_first(&c->free_by_rcu_ttrace) entry = smp_load_acquire(&head->first); do { if (entry == NULL) return NULL; // CPU1 bpf_mem_alloc_destroy() WRITE_ONCE(c->draining, true) // wait for CPU0 irq_work_sync() // CPU2 do_call_rcu_ttrace(tgt(CPU0)) if (c->draining) { llist_del_all(&c->free_by_rcu_ttrace) free_all() } // CPU0 continue next = READ_ONCE(entry->next); <-- trigger UAF while (!try_cmpxchg(&head->first, &entry, next)); Fix this by introducing a raw spinlock to synchronize the concurrent consumption on waiting_for_gp_ttrace and free_by_rcu_ttrace. Fixes: 04fabf00b4d3 ("bpf: Allow reuse from waiting_for_gp_ttrace list.") Suggested-by: Alexei Starovoitov Suggested-by: Hou Tao Signed-off-by: Pu Lehui Acked-by: Hou Tao Link: https://lore.kernel.org/r/20260905021139.4116529-1-pulehui@huaweicloud.com Signed-off-by: Alexei Starovoitov --- kernel/bpf/memalloc.c | 50 ++++++++++++++++++++++++------------------- 1 file changed, 28 insertions(+), 22 deletions(-) diff --git a/kernel/bpf/memalloc.c b/kernel/bpf/memalloc.c index e9662db7198f..8a8f088e83e6 100644 --- a/kernel/bpf/memalloc.c +++ b/kernel/bpf/memalloc.c @@ -119,6 +119,7 @@ struct bpf_mem_cache { struct llist_head waiting_for_gp_ttrace; struct rcu_head rcu_ttrace; atomic_t call_rcu_ttrace_in_progress; + raw_spinlock_t lock; }; struct bpf_mem_caches { @@ -214,25 +215,24 @@ static void alloc_bulk(struct bpf_mem_cache *c, int cnt, int node, bool atomic) gfp = __GFP_NOWARN | __GFP_ACCOUNT; gfp |= atomic ? GFP_NOWAIT : GFP_KERNEL; - for (i = 0; i < cnt; i++) { - /* - * For every 'c' llist_del_first(&c->free_by_rcu_ttrace); is - * done only by one CPU == current CPU. Other CPUs might - * llist_add() and llist_del_all() in parallel. - */ - obj = llist_del_first(&c->free_by_rcu_ttrace); - if (!obj) - break; - add_obj_to_free_list(c, obj); - } - if (i >= cnt) - return; + /* + * c->lock serializes concurrent llist_del_first() against + * llist_del_all() in __free_rcu() and do_call_rcu_ttrace(). + */ + scoped_guard(raw_spinlock_irqsave, &c->lock) { + for (i = 0; i < cnt; i++) { + obj = llist_del_first(&c->free_by_rcu_ttrace); + if (!obj) + break; + add_obj_to_free_list(c, obj); + } - for (; i < cnt; i++) { - obj = llist_del_first(&c->waiting_for_gp_ttrace); - if (!obj) - break; - add_obj_to_free_list(c, obj); + for (; i < cnt; i++) { + obj = llist_del_first(&c->waiting_for_gp_ttrace); + if (!obj) + break; + add_obj_to_free_list(c, obj); + } } if (i >= cnt) return; @@ -279,8 +279,12 @@ static int free_all(struct bpf_mem_cache *c, struct llist_node *llnode, bool per static void __free_rcu(struct rcu_head *head) { struct bpf_mem_cache *c = container_of(head, struct bpf_mem_cache, rcu_ttrace); + struct llist_node *llnode; - free_all(c, llist_del_all(&c->waiting_for_gp_ttrace), !!c->percpu_size); + scoped_guard(raw_spinlock_irqsave, &c->lock) + llnode = llist_del_all(&c->waiting_for_gp_ttrace); + + free_all(c, llnode, !!c->percpu_size); atomic_set(&c->call_rcu_ttrace_in_progress, 0); } @@ -300,7 +304,8 @@ static void do_call_rcu_ttrace(struct bpf_mem_cache *c) if (atomic_xchg(&c->call_rcu_ttrace_in_progress, 1)) { if (unlikely(READ_ONCE(c->draining))) { - llnode = llist_del_all(&c->free_by_rcu_ttrace); + scoped_guard(raw_spinlock_irqsave, &c->lock) + llnode = llist_del_all(&c->free_by_rcu_ttrace); free_all(c, llnode, !!c->percpu_size); } return; @@ -535,6 +540,7 @@ int bpf_mem_alloc_init(struct bpf_mem_alloc *ma, int size, bool percpu) c->objcg = objcg; c->percpu_size = percpu_size; c->tgt = c; + raw_spin_lock_init(&c->lock); init_refill_work(c); prefill_mem_cache(c, cpu); } @@ -557,7 +563,7 @@ int bpf_mem_alloc_init(struct bpf_mem_alloc *ma, int size, bool percpu) c->objcg = objcg; c->percpu_size = percpu_size; c->tgt = c; - + raw_spin_lock_init(&c->lock); init_refill_work(c); prefill_mem_cache(c, cpu); } @@ -609,7 +615,7 @@ int bpf_mem_alloc_percpu_unit_init(struct bpf_mem_alloc *ma, int size) c->objcg = objcg; c->percpu_size = percpu_size; c->tgt = c; - + raw_spin_lock_init(&c->lock); init_refill_work(c); prefill_mem_cache(c, cpu); } From 451b1c19dc7cbbad194a6e717b6a732c443d7dd5 Mon Sep 17 00:00:00 2001 From: Rong Zhang Date: Fri, 11 Sep 2026 02:39:25 +0800 Subject: [PATCH 173/592] hwmon: (k10temp) Fix model id range of Zen5 Turin MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit Model 20h-2Fh are mobile processors with single CCD. For example, model 24h is Strix Point, i.e., Ryzen AI 7 (PRO) (H/HX) 360/365/370. Including mobile processors in the model id range of Zen5 Turin processors leads to bogus reporting: k10temp-pci-00c3 Adapter: PCI adapter Tctl: +54.1°C Tccd4: +148.6°C Tccd6: +148.4°C Tccd7: +149.1°C Tccd8: +149.2°C Tccd9: +149.2°C Tccd12: +149.1°C Tccd14: +22.0°C Tccd15: +22.0°C Tccd16: +22.0°C Fix it by removing the said range. Fixes: 8440d5aca227 ("hwmon: (k10temp) Add per-CCD temperature monitoring for Zen5 Turin") Signed-off-by: Rong Zhang Reviewed-by: Mario Limonciello (AMD) Link: https://patch.msgid.link/20260911-k10temp-fix-zen5-epyc-v1-1-643f5a248ae1@rong.moe Signed-off-by: Guenter Roeck --- drivers/hwmon/k10temp.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/drivers/hwmon/k10temp.c b/drivers/hwmon/k10temp.c index 75a45010d687..3e7e63edc6a3 100644 --- a/drivers/hwmon/k10temp.c +++ b/drivers/hwmon/k10temp.c @@ -523,7 +523,7 @@ static int k10temp_probe(struct pci_dev *pdev, const struct pci_device_id *id) } } else if (boot_cpu_data.x86 == 0x1a) { switch (boot_cpu_data.x86_model) { - case 0x00 ... 0x2f: /* Zen5 Turin */ + case 0x00 ... 0x1f: /* Zen5 Turin */ data->ccd_offset = 0x1F0; k10temp_get_ccd_support(data, 16); break; From 26d5ff79768548efb1e604bb6e8697c101e06269 Mon Sep 17 00:00:00 2001 From: Yibo Tan Date: Fri, 11 Sep 2026 15:18:09 +0800 Subject: [PATCH 174/592] hwmon: (pwm-fan) Stop RPM timer before freeing tach data sample_timer() rearms the RPM timer and accesses the devm-managed ctx->tachs and ctx->pulses_per_revolution arrays. The cleanup action which stops the timer is registered before those arrays are allocated. Since devres releases entries in reverse order, driver detach can free the arrays before pwm_fan_cleanup() shuts down the timer. A timer expiry in that window accesses the freed tach data. With a KASAN kernel, a test-only kprobe delayed entry to pwm_fan_cleanup() while normal sysfs unbind ran. Each of three runs reported three four-byte reads and two four-byte writes in sample_timer() after its backing devm allocations had been freed. The helper did not invoke the timer callback, cleanup actions or free functions. With the fix, three matching unbind runs completed without KASAN, BUG, WARNING, Oops or panic. Instrumentation confirmed that timer retirement completed before the first timer backing allocation was released. Split timer retirement from the power cleanup and register its devres action after the timer backing data and IRQ actions are installed. This preserves the early power rollback action while ensuring the timer is retired before its backing data is released. Use timer_shutdown_sync() because the callback can rearm itself. Fixes: 01695410d452 ("hwmon: (pwm-fan) Store tach data separately") Cc: stable@vger.kernel.org Assisted-by: Codex:GPT-5 Signed-off-by: Yibo Tan Link: https://patch.msgid.link/20260911071809.130151-1-lhfff@tju.edu.cn Signed-off-by: Guenter Roeck --- drivers/hwmon/pwm-fan.c | 13 ++++++++++++- 1 file changed, 12 insertions(+), 1 deletion(-) diff --git a/drivers/hwmon/pwm-fan.c b/drivers/hwmon/pwm-fan.c index 3b87f65bae05..c633d7f6464c 100644 --- a/drivers/hwmon/pwm-fan.c +++ b/drivers/hwmon/pwm-fan.c @@ -483,7 +483,6 @@ static void pwm_fan_cleanup(void *__ctx) { struct pwm_fan_ctx *ctx = __ctx; - timer_delete_sync(&ctx->rpm_timer); if (ctx->pwm_shutdown) { ctx->enable_mode = pwm_enable_reg_enable; __set_pwm(ctx, ctx->pwm_shutdown); @@ -494,6 +493,13 @@ static void pwm_fan_cleanup(void *__ctx) } } +static void pwm_fan_timer_cleanup(void *__ctx) +{ + struct pwm_fan_ctx *ctx = __ctx; + + timer_shutdown_sync(&ctx->rpm_timer); +} + static int pwm_fan_probe(struct platform_device *pdev) { struct thermal_cooling_device *cdev; @@ -644,6 +650,10 @@ static int pwm_fan_probe(struct platform_device *pdev) } if (ctx->tach_count > 0) { + ret = devm_add_action_or_reset(dev, pwm_fan_timer_cleanup, ctx); + if (ret) + return ret; + ctx->sample_start = ktime_get(); mod_timer(&ctx->rpm_timer, jiffies + HZ); @@ -700,6 +710,7 @@ static void pwm_fan_shutdown(struct platform_device *pdev) { struct pwm_fan_ctx *ctx = platform_get_drvdata(pdev); + pwm_fan_timer_cleanup(ctx); pwm_fan_cleanup(ctx); } From 7bae83ffb133bc373d098fa6828cef2ef4da49fe Mon Sep 17 00:00:00 2001 From: "Thomas Richard (congatec GmbH)" Date: Fri, 11 Sep 2026 19:31:58 +0200 Subject: [PATCH 175/592] hwmon: (cgbc-hwmon) Fix current sensors ID lookup Current sensors on the Congatec Board Controller don't use consecutive IDs, unlike other sensor types (voltage, temperature, fan). The driver assumed consecutive IDs and performed a simple lookup, which caused an unknown sensor warning. Define current sensor IDs explicitly. Changes the warning on conga-SA7 (type and channel are correct now). Before: Board Controller returned an unknown sensor (type=2, channel=17), ignore it After: Board Controller returned an unknown sensor (bc_type=1, bc_id=11), ignore it Cc: stable@kernel.org Fixes: 08ebc9def79f ("hwmon: Add Congatec Board Controller monitoring driver") Signed-off-by: Thomas Richard (congatec GmbH) Link: https://patch.msgid.link/20260911-cgbc-hwmon-fix-and-new-sensors-v2-1-0c6bf078d173@bootlin.com Signed-off-by: Guenter Roeck --- drivers/hwmon/cgbc-hwmon.c | 95 ++++++++++++++++++++++---------------- 1 file changed, 55 insertions(+), 40 deletions(-) diff --git a/drivers/hwmon/cgbc-hwmon.c b/drivers/hwmon/cgbc-hwmon.c index 3aff4e092132..2effa0b56286 100644 --- a/drivers/hwmon/cgbc-hwmon.c +++ b/drivers/hwmon/cgbc-hwmon.c @@ -52,30 +52,35 @@ static const char * const cgbc_hwmon_labels_temp[] = { "BOTTOMDIM Temperature", }; -static const struct { - enum hwmon_sensor_types type; - const char *label; -} cgbc_hwmon_labels_in[] = { - { hwmon_in, "CPU Voltage" }, - { hwmon_in, "DC Runtime Voltage" }, - { hwmon_in, "DC Standby Voltage" }, - { hwmon_in, "CMOS Battery Voltage" }, - { hwmon_in, "Battery Voltage" }, - { hwmon_in, "AC Voltage" }, - { hwmon_in, "Other Voltage" }, - { hwmon_in, "5V Voltage" }, - { hwmon_in, "5V Standby Voltage" }, - { hwmon_in, "3V3 Voltage" }, - { hwmon_in, "3V3 Standby Voltage" }, - { hwmon_in, "VCore A Voltage" }, - { hwmon_in, "VCore B Voltage" }, - { hwmon_in, "12V Voltage" }, - { hwmon_curr, "DC Current" }, - { hwmon_curr, "5V Current" }, - { hwmon_curr, "12V Current" }, +static const char * const cgbc_hwmon_labels_in[] = { + "CPU Voltage", + "DC Runtime Voltage", + "DC Standby Voltage", + "CMOS Battery Voltage", + "Battery Voltage", + "AC Voltage", + "Other Voltage", + "5V Voltage", + "5V Standby Voltage", + "3V3 Voltage", + "3V3 Standby Voltage", + "VCore A Voltage", + "VCore B Voltage", + "12V Voltage", }; -#define CGBC_HWMON_NB_IN_SENSORS 14 +/* + * Current sensors are a bit special, they don't have consecutive IDs like + * other types of sensors. So they need to be defined explicitly. + */ +static const struct { + const char *label; + int id; +} cgbc_hwmon_labels_curr[] = { + { "DC Current", 0x12 }, + { "5V Current", 0x18 }, + { "12V Current", 0x1E }, +}; static const char * const cgbc_hwmon_labels_fan[] = { "CPU Fan", @@ -114,7 +119,8 @@ static int cgbc_hwmon_probe_sensors(struct device *dev, struct cgbc_hwmon_data * for (i = 0; i < nb_sensors; i++) { enum cgbc_sensor_types type; - unsigned int channel; + unsigned int channel, id; + int j; /* * No need to request data for the first sensor. @@ -128,32 +134,49 @@ static int cgbc_hwmon_probe_sensors(struct device *dev, struct cgbc_hwmon_data * } type = FIELD_GET(CGBC_HWMON_TYPE_MASK, data[1]); - channel = FIELD_GET(CGBC_HWMON_ID_MASK, data[1]) - 1; + id = FIELD_GET(CGBC_HWMON_ID_MASK, data[1]); + channel = id - 1; if (type == CGBC_HWMON_TYPE_TEMP && channel < ARRAY_SIZE(cgbc_hwmon_labels_temp)) { sensor->type = hwmon_temp; sensor->label = cgbc_hwmon_labels_temp[channel]; - } else if (type == CGBC_HWMON_TYPE_IN && - channel < ARRAY_SIZE(cgbc_hwmon_labels_in)) { + } else if (type == CGBC_HWMON_TYPE_IN) { /* * The Board Controller doesn't differentiate current and voltage sensors. - * Get the sensor type from cgbc_hwmon_labels_in[channel].type instead. + * First check if it is a current sensor. */ - sensor->type = cgbc_hwmon_labels_in[channel].type; - sensor->label = cgbc_hwmon_labels_in[channel].label; + for (j = 0; j < ARRAY_SIZE(cgbc_hwmon_labels_curr); j++) { + if (id == cgbc_hwmon_labels_curr[j].id) { + sensor->type = hwmon_curr; + sensor->label = cgbc_hwmon_labels_curr[j].label; + channel = j; + } + } + + /* If it's not a current sensor, it may be a voltage sensor. */ + if (!sensor->label && channel < ARRAY_SIZE(cgbc_hwmon_labels_in)) { + sensor->type = hwmon_in; + sensor->label = cgbc_hwmon_labels_in[channel]; + } } else if (type == CGBC_HWMON_TYPE_FAN && channel < ARRAY_SIZE(cgbc_hwmon_labels_fan)) { sensor->type = hwmon_fan; sensor->label = cgbc_hwmon_labels_fan[channel]; - } else { - dev_warn(dev, "Board Controller returned an unknown sensor (type=%d, channel=%d), ignore it", - type, channel); + } + + if (!sensor->label) { + dev_warn(dev, "Board Controller returned an unknown sensor (bc_type=%d, bc_id=%d), ignore it", + type, id); continue; } sensor->active = FIELD_GET(CGBC_HWMON_ACTIVE_BIT, data[1]); sensor->channel = channel; sensor->index = i; + + dev_dbg(dev, "Found sensor: bc_type=%d, bc_id=%d, hwmon_type=%d, hwmon_channel=%d, hwmon_label='%s', active=%d\n", + type, id, sensor->type, sensor->channel, sensor->label, sensor->active); + sensor++; hwmon->nb_sensors++; } @@ -167,14 +190,6 @@ static struct cgbc_hwmon_sensor *cgbc_hwmon_find_sensor(struct cgbc_hwmon_data * struct cgbc_hwmon_sensor *sensor = NULL; int i; - /* - * The Board Controller doesn't differentiate current and voltage sensors. - * The channel value (from the Board Controller point of view) shall be computed for current - * sensors. - */ - if (type == hwmon_curr) - channel += CGBC_HWMON_NB_IN_SENSORS; - for (i = 0; i < hwmon->nb_sensors; i++) { if (hwmon->sensors[i].type == type && hwmon->sensors[i].channel == channel) { sensor = &hwmon->sensors[i]; From 3550d1dbbcb9f51b77e077e8958423ee2c401c6c Mon Sep 17 00:00:00 2001 From: "Thomas Richard (congatec GmbH)" Date: Fri, 11 Sep 2026 19:31:59 +0200 Subject: [PATCH 176/592] hwmon: (cgbc-hwmon) Add missing sensors Add the following sensors: - Alternate Board Temperature (temp11_input) - Top DIMM 1-7 Temperature (temp12_input to temp18_input) - Bottom DIMM 1 Temperature (temp19_input) - 12V Standby Voltage (in14_input) This fixes the following warning on conga-SA7: Board Controller returned an unknown sensor (bc_type=1, bc_id=11), ignore it Also update existing labels to match Congatec documentation. Cc: stable@kernel.org Fixes: 08ebc9def79f ("hwmon: Add Congatec Board Controller monitoring driver") Signed-off-by: Thomas Richard (congatec GmbH) Link: https://patch.msgid.link/20260911-cgbc-hwmon-fix-and-new-sensors-v2-2-0c6bf078d173@bootlin.com Signed-off-by: Guenter Roeck --- Documentation/hwmon/cgbc-hwmon.rst | 42 +++++++++++++++--------- drivers/hwmon/cgbc-hwmon.c | 52 +++++++++++++++++++----------- 2 files changed, 60 insertions(+), 34 deletions(-) diff --git a/Documentation/hwmon/cgbc-hwmon.rst b/Documentation/hwmon/cgbc-hwmon.rst index 3a5e6e6e8639..c6d09232392a 100644 --- a/Documentation/hwmon/cgbc-hwmon.rst +++ b/Documentation/hwmon/cgbc-hwmon.rst @@ -28,34 +28,44 @@ system. Name Description ============= ====================== temp1_input CPU temperature -temp2_input Box temperature +temp2_input Case temperature temp3_input Ambient temperature -temp4_input Board temperature -temp5_input Carrier temperature -temp6_input Chipset temperature -temp7_input Video temperature +temp4_input CPU Board temperature +temp5_input Carrier Board temperature +temp6_input System Chipset temperature +temp7_input Video Controller/Board temperature temp8_input Other temperature -temp9_input TOPDIM temperature -temp10_input BOTTOMDIM temperature -in0_input CPU voltage +temp9_input Top DIMM 0 temperature +temp10_input Bottom DIMM 0 temperature +temp11_input Alternate Board temperature +temp12_input Top DIMM 1 temperature +temp13_input Top DIMM 2 temperature +temp14_input Top DIMM 3 temperature +temp15_input Top DIMM 4 temperature +temp16_input Top DIMM 5 temperature +temp17_input Top DIMM 6 temperature +temp18_input Top DIMM 7 temperature +temp19_input Bottom DIMM 1 temperature +in0_input CPU Core voltage in1_input DC Runtime voltage in2_input DC Standby voltage in3_input CMOS Battery voltage -in4_input Battery voltage +in4_input Battery Supply voltage in5_input AC voltage in6_input Other voltage -in7_input 5V voltage +in7_input 5V Runtime voltage in8_input 5V Standby voltage -in9_input 3V3 voltage +in9_input 3V3 Runtime voltage in10_input 3V3 Standby voltage in11_input VCore A voltage in12_input VCore B voltage -in13_input 12V voltage -curr1_input DC current -curr2_input 5V current -curr3_input 12V current +in13_input 12V Runtime voltage +in14_input 12V Standby voltage +curr1_input DC Runtime current +curr2_input 5V Runtime current +curr3_input 12V Runtime current fan1_input CPU fan -fan2_input Box fan +fan2_input Case fan fan3_input Ambient fan fan4_input Chiptset fan fan5_input Video fan diff --git a/drivers/hwmon/cgbc-hwmon.c b/drivers/hwmon/cgbc-hwmon.c index 2effa0b56286..230062a46907 100644 --- a/drivers/hwmon/cgbc-hwmon.c +++ b/drivers/hwmon/cgbc-hwmon.c @@ -41,32 +41,42 @@ enum cgbc_sensor_types { static const char * const cgbc_hwmon_labels_temp[] = { "CPU Temperature", - "Box Temperature", + "Case Temperature", "Ambient Temperature", - "Board Temperature", - "Carrier Temperature", - "Chipset Temperature", - "Video Temperature", + "CPU Board Temperature", + "Carrier Board Temperature", + "System Chipset Temperature", + "Video Controller/Board Temperature", "Other Temperature", - "TOPDIM Temperature", - "BOTTOMDIM Temperature", + "Top DIMM 0 Temperature", + "Bottom DIMM 0 Temperature", + "Alternate Board Temperature", + "Top DIMM 1 Temperature", + "Top DIMM 2 Temperature", + "Top DIMM 3 Temperature", + "Top DIMM 4 Temperature", + "Top DIMM 5 Temperature", + "Top DIMM 6 Temperature", + "Top DIMM 7 Temperature", + "Bottom DIMM 1 Temperature", }; static const char * const cgbc_hwmon_labels_in[] = { - "CPU Voltage", + "CPU Core Voltage", "DC Runtime Voltage", "DC Standby Voltage", "CMOS Battery Voltage", - "Battery Voltage", + "Battery Supply Voltage", "AC Voltage", "Other Voltage", - "5V Voltage", + "5V Runtime Voltage", "5V Standby Voltage", - "3V3 Voltage", + "3V3 Runtime Voltage", "3V3 Standby Voltage", "VCore A Voltage", "VCore B Voltage", - "12V Voltage", + "12V Runtime Voltage", + "12V Standby Voltage", }; /* @@ -77,14 +87,14 @@ static const struct { const char *label; int id; } cgbc_hwmon_labels_curr[] = { - { "DC Current", 0x12 }, - { "5V Current", 0x18 }, - { "12V Current", 0x1E }, + { "DC Runtime Current", 0x12 }, + { "5V Runtime Current", 0x18 }, + { "12V Runtime Current", 0x1E }, }; static const char * const cgbc_hwmon_labels_fan[] = { "CPU Fan", - "Box Fan", + "Case Fan", "Ambient Fan", "Chipset Fan", "Video Fan", @@ -255,7 +265,12 @@ static const struct hwmon_channel_info * const cgbc_hwmon_info[] = { HWMON_T_INPUT | HWMON_T_LABEL, HWMON_T_INPUT | HWMON_T_LABEL, HWMON_T_INPUT | HWMON_T_LABEL, HWMON_T_INPUT | HWMON_T_LABEL, HWMON_T_INPUT | HWMON_T_LABEL, HWMON_T_INPUT | HWMON_T_LABEL, - HWMON_T_INPUT | HWMON_T_LABEL, HWMON_T_INPUT | HWMON_T_LABEL), + HWMON_T_INPUT | HWMON_T_LABEL, HWMON_T_INPUT | HWMON_T_LABEL, + HWMON_T_INPUT | HWMON_T_LABEL, HWMON_T_INPUT | HWMON_T_LABEL, + HWMON_T_INPUT | HWMON_T_LABEL, HWMON_T_INPUT | HWMON_T_LABEL, + HWMON_T_INPUT | HWMON_T_LABEL, HWMON_T_INPUT | HWMON_T_LABEL, + HWMON_T_INPUT | HWMON_T_LABEL, HWMON_T_INPUT | HWMON_T_LABEL, + HWMON_T_INPUT | HWMON_T_LABEL), HWMON_CHANNEL_INFO(in, HWMON_I_INPUT | HWMON_I_LABEL, HWMON_I_INPUT | HWMON_I_LABEL, HWMON_I_INPUT | HWMON_I_LABEL, HWMON_I_INPUT | HWMON_I_LABEL, @@ -263,7 +278,8 @@ static const struct hwmon_channel_info * const cgbc_hwmon_info[] = { HWMON_I_INPUT | HWMON_I_LABEL, HWMON_I_INPUT | HWMON_I_LABEL, HWMON_I_INPUT | HWMON_I_LABEL, HWMON_I_INPUT | HWMON_I_LABEL, HWMON_I_INPUT | HWMON_I_LABEL, HWMON_I_INPUT | HWMON_I_LABEL, - HWMON_I_INPUT | HWMON_I_LABEL, HWMON_I_INPUT | HWMON_I_LABEL), + HWMON_I_INPUT | HWMON_I_LABEL, HWMON_I_INPUT | HWMON_I_LABEL, + HWMON_I_INPUT | HWMON_I_LABEL), HWMON_CHANNEL_INFO(curr, HWMON_C_INPUT | HWMON_C_LABEL, HWMON_C_INPUT | HWMON_C_LABEL, HWMON_C_INPUT | HWMON_C_LABEL), From 06bd6794b5fd2163880ac3bfe973d4cc61f359f3 Mon Sep 17 00:00:00 2001 From: =?UTF-8?q?Nuno=20S=C3=A1?= Date: Fri, 11 Sep 2026 14:53:37 +0100 Subject: [PATCH 177/592] hwmon: (pmbus/core) increase number of phases and add new mask MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit Increase the number of phases to 16 as a new upcoming device supports such a number. While at it, add a new mask for controlling the source of the output voltage. Note (groeck): This patch was meant to prepare for support of MAX20826 and compatible devices, which support more than 10 phases per page. However, Sashiko reports that the mp2975 driver already supports up to 14 phases, and the mp2856 driver supports up to 12 phases. This already has the potential for out-of-bounds writes when probing the affected chips, making this patch a bug fix. Fixes: 2c6fcbb21149 ("hwmon: (pmbus) Add support for MPS Multi-phase mp2975 controller") Fixes: f9e5f289b686 ("hwmon: (pmbus) Add support for MPS Multi-phase mp2856/mp2857 controller") Signed-off-by: Nuno Sá Link: https://patch.msgid.link/20260911-hwmon-max20826-support-v2-1-5e30cbd97d84@analog.com Cc: stable@vger.kernel.org Signed-off-by: Guenter Roeck --- drivers/hwmon/pmbus/pmbus.h | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-) diff --git a/drivers/hwmon/pmbus/pmbus.h b/drivers/hwmon/pmbus/pmbus.h index 2cd3216b3cd9..920c1102ab6d 100644 --- a/drivers/hwmon/pmbus/pmbus.h +++ b/drivers/hwmon/pmbus/pmbus.h @@ -242,6 +242,7 @@ enum pmbus_regs { /* * OPERATION */ +#define PB_OPERATION_CONTROL_V_SRC GENMASK(5, 4) #define PB_OPERATION_CONTROL_ON BIT(7) /* @@ -386,7 +387,7 @@ enum pmbus_sensor_classes { }; #define PMBUS_PAGES 32 /* Per PMBus specification */ -#define PMBUS_PHASES 10 /* Maximum number of phases per page */ +#define PMBUS_PHASES 16 /* Maximum number of phases per page */ /* Functionality bit mask */ #define PMBUS_HAVE_VIN BIT(0) From 490a83d6386eec1d29f470c8d7331677fb46c3b7 Mon Sep 17 00:00:00 2001 From: Geliang Tang Date: Tue, 8 Sep 2026 17:08:32 +0800 Subject: [PATCH 178/592] bpf, sockmap: Fix self-redirect copied_seq double-counting When a BPF stream_verdict program redirects an skb back to the same socket (self-redirect with BPF_F_INGRESS), sk_psock_verdict_apply() calls tcp_eat_skb() which advances tcp_sk->copied_seq. However, the skb is then delivered to the socket's psock ingress queue and later read by tcp_bpf_recvmsg_parser(), which also advances copied_seq via the copied_from_self accounting path. This double-counting causes copied_seq to advance by 2x the actual data length, triggering: TCP recvmsg seq # bug 2: copied BF2E806, seq BF2E7FD, \ rcvnxt BF2E806, fl 0 WARNING: net/ipv4/tcp.c:2745 at tcp_recvmsg_locked+0x72b/0x2640 Call Trace: tcp_recvmsg+0x10a/0x500 sock_recvmsg+0x168/0x1d0 __sys_recvfrom+0x19a/0x2a0 __x64_sys_recvfrom+0xe4/0x1f0 do_syscall_64+0xf7/0x530 entry_SYSCALL_64_after_hwframe+0x77/0x7f cleanup rbuf bug: copied BF2E806 seq BF2E806 rcvnxt BF2E806 WARNING: net/ipv4/tcp.c:1609 at tcp_cleanup_rbuf+0xf2/0x1c0 Call Trace: tcp_recvmsg_locked+0x8d1/0x2640 tcp_recvmsg+0x10a/0x500 sock_recvmsg+0x168/0x1d0 __sys_recvfrom+0x19a/0x2a0 __x64_sys_recvfrom+0xe4/0x1f0 do_syscall_64+0xf7/0x530 entry_SYSCALL_64_after_hwframe+0x77/0x7f Fix this by converting self-redirect verdict to __SK_PASS at the beginning of sk_psock_verdict_apply(). This bypasses the __SK_REDIRECT case entirely (which calls sk_psock_eat_skb), letting the __SK_PASS path queue the skb to the psock ingress queue. The data is then read via tcp_bpf_recvmsg_parser(), which advances copied_seq exactly once through copied_from_self. Cross-socket redirects continue through __SK_REDIRECT with sk_psock_eat_skb() unchanged. Fixes: e5c6de5fa025 ("bpf, sockmap: Incorrectly handling copied_seq") Suggested-by: Jakub Sitnicki Suggested-by: Jiayuan Chen Signed-off-by: Geliang Tang Reviewed-by: Emil Tsalapatis Reviewed-by: Jiayuan Chen Link: https://lore.kernel.org/r/1a8e797a1b26e2f695aaac22ac644c2862f63466.1788858299.git.tanggeliang@kylinos.cn Signed-off-by: Alexei Starovoitov --- net/core/skmsg.c | 4 ++++ 1 file changed, 4 insertions(+) diff --git a/net/core/skmsg.c b/net/core/skmsg.c index 2521b643fa05..df385a5a961e 100644 --- a/net/core/skmsg.c +++ b/net/core/skmsg.c @@ -1000,6 +1000,10 @@ static int sk_psock_verdict_apply(struct sk_psock *psock, struct sk_buff *skb, int err = 0; u32 len, off; + if (verdict == __SK_REDIRECT && skb_bpf_ingress(skb) && + skb_bpf_redirect_fetch(skb) == psock->sk) + verdict = __SK_PASS; + switch (verdict) { case __SK_PASS: err = -EIO; From 953824e508b27d12837e32ef37ef6248e1f6fc7a Mon Sep 17 00:00:00 2001 From: Masoud Aghasi Date: Thu, 3 Sep 2026 09:27:34 +0100 Subject: [PATCH 179/592] bpf: Fix u32 overflow issue in map batch operations Several map batch operation implementations such as generic_map_lookup_batch() use calculations in the form of "values + cp * map->value_size" to compute the desired userspace memory address for reading or writing. This can overflow the u32 type (the result of "cp * map->value_size") when the map size exceeds 4GB. generic_map_lookup_batch() may corrupt values for some keys in userspace memory, and in some cases it mismatches values for some keys while still reporting success. Other batch operations may fail to delete or update some keys, or the syscall may return unexpected errors. Add size_t casts to prevent the affected offset and size calculations from overflowing. Fixes: cb4d03ab499d ("bpf: Add generic support for lookup batch op") Fixes: aa2e93b8e58e ("bpf: Add generic support for update and delete batch ops") Fixes: 057996380a42 ("bpf: Add batch ops to all htab bpf map") Signed-off-by: Masoud Aghasi Link: https://lore.kernel.org/r/20260903082734.623904-1-maghasi@disroot.org Signed-off-by: Alexei Starovoitov --- kernel/bpf/hashtab.c | 8 ++++---- kernel/bpf/syscall.c | 10 +++++----- 2 files changed, 9 insertions(+), 9 deletions(-) diff --git a/kernel/bpf/hashtab.c b/kernel/bpf/hashtab.c index a72dc5b9f184..4f495dcbf670 100644 --- a/kernel/bpf/hashtab.c +++ b/kernel/bpf/hashtab.c @@ -1997,10 +1997,10 @@ __htab_map_lookup_and_delete_batch(struct bpf_map *map, rcu_read_unlock(); bpf_enable_instrumentation(); - if (bucket_cnt && (copy_to_user(ukeys + total * key_size, keys, - key_size * bucket_cnt) || - copy_to_user(uvalues + total * value_size, values, - value_size * bucket_cnt))) { + if (bucket_cnt && (copy_to_user(ukeys + (size_t)total * key_size, keys, + (size_t)key_size * bucket_cnt) || + copy_to_user(uvalues + (size_t)total * value_size, values, + (size_t)value_size * bucket_cnt))) { ret = -EFAULT; goto after_loop; } diff --git a/kernel/bpf/syscall.c b/kernel/bpf/syscall.c index c7bc9ba9b331..853b47f81384 100644 --- a/kernel/bpf/syscall.c +++ b/kernel/bpf/syscall.c @@ -2036,7 +2036,7 @@ int generic_map_delete_batch(struct bpf_map *map, for (cp = 0; cp < max_count; cp++) { err = -EFAULT; - if (copy_from_user(key, keys + cp * map->key_size, + if (copy_from_user(key, keys + (size_t)cp * map->key_size, map->key_size)) break; @@ -2098,9 +2098,9 @@ int generic_map_update_batch(struct bpf_map *map, struct file *map_file, for (cp = 0; cp < max_count; cp++) { err = -EFAULT; - if (copy_from_user(key, keys + cp * map->key_size, + if (copy_from_user(key, keys + (size_t)cp * map->key_size, map->key_size) || - copy_from_user(value, values + cp * value_size, value_size)) + copy_from_user(value, values + (size_t)cp * value_size, value_size)) break; err = bpf_map_update_value(map, map_file, key, value, @@ -2179,12 +2179,12 @@ int generic_map_lookup_batch(struct bpf_map *map, if (err) goto free_buf; - if (copy_to_user(keys + cp * map->key_size, key, + if (copy_to_user(keys + (size_t)cp * map->key_size, key, map->key_size)) { err = -EFAULT; goto free_buf; } - if (copy_to_user(values + cp * value_size, value, value_size)) { + if (copy_to_user(values + (size_t)cp * value_size, value, value_size)) { err = -EFAULT; goto free_buf; } From ef1fb82f12186dd26153b14d9fbcf4ec98db81b3 Mon Sep 17 00:00:00 2001 From: Donggeun Yoo Date: Mon, 7 Sep 2026 22:06:23 +0900 Subject: [PATCH 180/592] bpf, arm64: set up the frame pointer for the exception callback A program acting as exception boundary saves all callee-saved registers, so build_prologue() takes the exception_cb path and never calls push_callee_regs(). That is the only place find_used_callee_regs() runs, and with it the only place ctx->fp_used is set, so the callback prologue does not emit the mov x25, sp that points BPF_REG_FP at the frame the callback runs on. x25 keeps whatever it held when bpf_throw() was called. If the throw came from a subprogram that uses its own BPF stack, that is the subprogram's frame pointer, and since the subprogram never returns it never restores x25 either. Stack accesses through BPF_REG_FP are rewritten to be stack pointer relative, so those still land in the callback's own frame. Materializing the register does not: a callback that passes the address of a local variable to a helper hands over an address in the dead subprogram's frame. That address is below the callback's stack pointer by then, and the helper's own call chain covers it, so the helper can write over its own return address. 0x1234 below is the value the helper was asked to store: pc : 0x1234 lr : 0x1234 Call trace: 0x1234 (P) bpf_test_run+0x188/0x3e0 bpf_prog_test_run_skb+0x47c/0x998 __sys_bpf+0xbdc/0xdd8 Kernel panic - not syncing: Oops: Fatal exception in interrupt Set ctx->fp_used on the exception callback path so that the existing code further down sets x25 from the stack pointer. The epilogue restores it from the main program's save area along with the other callee-saved registers, as it already does. x86 sets the frame pointer for the callback from the argument it is passed, and powerpc computes it from the stack pointer. Fixes: 5d4fa9ec5643 ("bpf, arm64: Avoid blindly saving/restoring all callee-saved registers") Acked-by: Xu Kuohai Signed-off-by: Donggeun Yoo Link: https://lore.kernel.org/r/20260907130624.611942-2-donggeunyoo.kernel@gmail.com Signed-off-by: Alexei Starovoitov --- arch/arm64/net/bpf_jit_comp.c | 2 ++ 1 file changed, 2 insertions(+) diff --git a/arch/arm64/net/bpf_jit_comp.c b/arch/arm64/net/bpf_jit_comp.c index c18e005a41db..c5f55d6161fe 100644 --- a/arch/arm64/net/bpf_jit_comp.c +++ b/arch/arm64/net/bpf_jit_comp.c @@ -600,6 +600,8 @@ static int build_prologue(struct jit_ctx *ctx, bool ebpf_from_cbpf) * 12 registers are on the stack */ emit(A64_SUB_I(1, A64_SP, A64_FP, 96), ctx); + /* The callback may use its own BPF stack, set up fp for it. */ + ctx->fp_used = true; } /* Stack must be multiples of 16B */ From 26a43a5f8c31b9132dc40a449d0f6c7ecfaa1f40 Mon Sep 17 00:00:00 2001 From: Donggeun Yoo Date: Mon, 7 Sep 2026 22:06:24 +0900 Subject: [PATCH 181/592] selftests/bpf: cover the exception callback using its own BPF stack The existing exception tests do not reach a callback that materializes BPF_REG_FP into a register. They either throw from the main program, where BPF_REG_FP already holds the value the callback needs, or use a callback whose only stack accesses are frame pointer relative, which the arm64 JIT rewrites to be stack pointer relative. Add a test that throws from a subprogram using its own BPF stack, with a callback that hands the address of a local variable to bpf_probe_read_kernel(). The helper and the callback have to name the same slot for the value read back to be the one the helper stored. Signed-off-by: Donggeun Yoo Link: https://lore.kernel.org/r/20260907130624.611942-3-donggeunyoo.kernel@gmail.com Signed-off-by: Alexei Starovoitov --- .../selftests/bpf/prog_tests/exceptions.c | 1 + .../testing/selftests/bpf/progs/exceptions.c | 30 +++++++++++++++++++ 2 files changed, 31 insertions(+) diff --git a/tools/testing/selftests/bpf/prog_tests/exceptions.c b/tools/testing/selftests/bpf/prog_tests/exceptions.c index 3588d6f97fd4..639866ce09a9 100644 --- a/tools/testing/selftests/bpf/prog_tests/exceptions.c +++ b/tools/testing/selftests/bpf/prog_tests/exceptions.c @@ -55,6 +55,7 @@ static void test_exceptions_success(void) RUN_SUCCESS(exception_ext, 0); RUN_SUCCESS(exception_ext_mod_cb_runtime, 35); RUN_SUCCESS(exception_throw_subprog, 1); + RUN_SUCCESS(exception_throw_subprog_stack_cb, 0x1234); RUN_SUCCESS(exception_assert_nz_gfunc, 1); RUN_SUCCESS(exception_assert_zero_gfunc, 1); RUN_SUCCESS(exception_assert_neg_gfunc, 1); diff --git a/tools/testing/selftests/bpf/progs/exceptions.c b/tools/testing/selftests/bpf/progs/exceptions.c index c8d716fbd419..91c81971e58c 100644 --- a/tools/testing/selftests/bpf/progs/exceptions.c +++ b/tools/testing/selftests/bpf/progs/exceptions.c @@ -212,6 +212,36 @@ int exception_throw_subprog(struct __sk_buff *ctx) return 0; } +u64 exception_cb_stack_src = 0x1234; + +/* + * The address handed to the helper has to be this callback's own stack + * slot, not one from a frame that is already gone. + */ +__noinline int exception_cb_stack(u64 cookie) +{ + volatile u64 val = 0xdead; + + bpf_probe_read_kernel((void *)&val, sizeof(val), &exception_cb_stack_src); + return val; +} + +/* Throws from a subprogram that has a stack of its own. */ +__noinline static int throwing_subprog_stack(struct __sk_buff *ctx) +{ + volatile u64 pad[4] = {}; + + bpf_throw(pad[0]); + return 0; +} + +SEC("tc") +__exception_cb(exception_cb_stack) +int exception_throw_subprog_stack_cb(struct __sk_buff *ctx) +{ + return throwing_subprog_stack(ctx); +} + __noinline int assert_nz_gfunc(u64 c) { volatile u64 cookie = c; From 8dc2615d5702059b2b71fca6f93c0d7d10ae54cb Mon Sep 17 00:00:00 2001 From: Koichiro Den Date: Fri, 11 Sep 2026 16:30:58 +0900 Subject: [PATCH 182/592] arm64: dts: renesas: r8a779f0: Set UFS lane count Since commit e72323f3b09f ("scsi: ufs: core: Configure only active lanes during link"), the following error is observed on R-Car S4: ufshcd-renesas e6860000.ufs: Tx lane mismatch [config,reported] [2,1] ufshcd-renesas e6860000.ufs: link startup failed -67 ufshcd-renesas e6860000.ufs: error -ENOLINK: Initialization failed with error -67 ufshcd-renesas e6860000.ufs: probe with driver ufshcd-renesas failed with error -67 R-Car S4 has one UFS lane per direction, as described in section 152.1 of its hardware manual. Without lanes-per-direction, the UFS platform driver defaults to two lanes. Previously, the core used PA_CONNECTEDRXDATALANES and PA_CONNECTEDTXDATALANES to configure the link without checking them against lanes-per-direction, so the missing property did not prevent initialization. Explicitly set lanes-per-direction to 1, now that the validation is in place. Fixes: 5235d551779d ("arm64: dts: renesas: r8a779f0: Add UFS node") Cc: stable@vger.kernel.org # 7.2+ Signed-off-by: Koichiro Den Reviewed-by: Geert Uytterhoeven Tested-by: Geert Uytterhoeven Link: https://patch.msgid.link/20260911073058.253000-1-den@valinux.co.jp Signed-off-by: Geert Uytterhoeven --- arch/arm64/boot/dts/renesas/r8a779f0.dtsi | 1 + 1 file changed, 1 insertion(+) diff --git a/arch/arm64/boot/dts/renesas/r8a779f0.dtsi b/arch/arm64/boot/dts/renesas/r8a779f0.dtsi index cbb161c863ac..a5118c2f2742 100644 --- a/arch/arm64/boot/dts/renesas/r8a779f0.dtsi +++ b/arch/arm64/boot/dts/renesas/r8a779f0.dtsi @@ -901,6 +901,7 @@ ufs: ufs@e6860000 { clocks = <&cpg CPG_MOD 1514>, <&ufs30_clk>; clock-names = "fck", "ref_clk"; freq-table-hz = <200000000 200000000>, <38400000 38400000>; + lanes-per-direction = <1>; power-domains = <&sysc R8A779F0_PD_ALWAYS_ON>; resets = <&cpg 1514>; status = "disabled"; From 692f32609a30f75ca3401e25b504bfd06bd5662a Mon Sep 17 00:00:00 2001 From: Sean Anderson Date: Mon, 17 Aug 2026 12:22:11 -0400 Subject: [PATCH 183/592] pinctrl: meson: Fix typo in s4 group name One of the i2c pin groups has some junk at the end. The name should be i2c2_scl_h1, and indeed that's the name used by i2c2_pins3 in meson-s4.dtsi. Fixes: 775214d389c25 ("pinctrl: meson: add pinctrl driver support for Meson-S4 Soc") Signed-off-by: Sean Anderson Reviewed-by: Neil Armstrong Signed-off-by: Linus Walleij --- drivers/pinctrl/meson/pinctrl-meson-s4.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/drivers/pinctrl/meson/pinctrl-meson-s4.c b/drivers/pinctrl/meson/pinctrl-meson-s4.c index 872948699e9f..365dafe457a9 100644 --- a/drivers/pinctrl/meson/pinctrl-meson-s4.c +++ b/drivers/pinctrl/meson/pinctrl-meson-s4.c @@ -854,7 +854,7 @@ static const char * const i2c1_groups[] = { static const char * const i2c2_groups[] = { "i2c2_sda_d", "i2c2_scl_d", "i2c2_sda_h8", "i2c2_scl_h9", - "i2c2_sda_h0", "i2c2_scl_h1l," + "i2c2_sda_h0", "i2c2_scl_h1", }; static const char * const i2c3_groups[] = { From 51ae99659469edba2e83931efa26b77d36ca02c2 Mon Sep 17 00:00:00 2001 From: Sarah Emery Date: Fri, 28 Aug 2026 17:58:17 +0200 Subject: [PATCH 184/592] pinctrl: generic: serialise pinctrl_generic_dt_node_to_map() pinctrl_generic_add_group() documents that the caller must take care of locking, and pinmux_generic_add_function() needs it too, but pinctrl_generic_dt_node_to_map() calls them without holding pctldev->mutex, and the core caller in create_pinctrl() does not take it either. The driver core calls pinctrl_bind_pins() before probing a device, so two devices that reference the same pin controller can run pinctrl_generic_dt_node_to_map() on one pctldev at the same time. Both `add` functions take the new selector from pctldev->num_groups or pctldev->num_functions, and radix_tree_insert() at that index. Two racing callers can read the same selector before either has inserted, so the second insert collides and fails: k1-pinctrl d401e000.pinctrl: error -EEXIST: error adding function pcie2-0-cfg k1-pinctrl d401e000.pinctrl: does not have pin group pcie0-0-cfg.pcie0-0-pins leaving one consumer without its pin configuration. This was hit on a SpacemiT K3 board, where PCIe devices probe in parallel against the single shared pin controller. Take pctldev->mutex across the whole function, so that the groups and the function referring are in a single critical section. Fixes: 43722575e5cd ("pinctrl: add generic functions + pins mapper") Signed-off-by: Sarah Emery Signed-off-by: Linus Walleij --- drivers/pinctrl/pinctrl-generic.c | 4 ++++ 1 file changed, 4 insertions(+) diff --git a/drivers/pinctrl/pinctrl-generic.c b/drivers/pinctrl/pinctrl-generic.c index fd6bdb74028a..4277c8748513 100644 --- a/drivers/pinctrl/pinctrl-generic.c +++ b/drivers/pinctrl/pinctrl-generic.c @@ -3,8 +3,10 @@ #define pr_fmt(fmt) "generic pinconfig core: " fmt #include +#include #include #include +#include #include #include @@ -196,6 +198,8 @@ static int pinctrl_generic_dt_node_to_map(struct pinctrl_dev *pctldev, int ngroups = 0; int ret; + guard(mutex)(&pctldev->mutex); + *maps = NULL; *num_maps = 0; From e6c5d6c589764a41218cbd81bd7d6c9b906e2cf0 Mon Sep 17 00:00:00 2001 From: Conor Dooley Date: Mon, 31 Aug 2026 11:50:58 +0100 Subject: [PATCH 185/592] pinctrl: mpfs-mssio: fix width of unused bank voltage setting The bank voltages are only 4 bits wide, so when a pin was unused the driver was not correctly interpreting it as being at zero volts, because the driver's value for unused had two extra set bits. CC: stable@vger.kernel.org Fixes: 488d704ed7b7 ("pinctrl: add polarfire soc mssio pinctrl driver") Signed-off-by: Conor Dooley Signed-off-by: Linus Walleij --- drivers/pinctrl/microchip/pinctrl-mpfs-mssio.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/drivers/pinctrl/microchip/pinctrl-mpfs-mssio.c b/drivers/pinctrl/microchip/pinctrl-mpfs-mssio.c index ea1026a0d22c..dafca82f3e54 100644 --- a/drivers/pinctrl/microchip/pinctrl-mpfs-mssio.c +++ b/drivers/pinctrl/microchip/pinctrl-mpfs-mssio.c @@ -86,7 +86,7 @@ static struct mpfs_pinctrl_bank_voltage mpfs_pinctrl_bank_voltages[8] = { { .uv = 1800000, .val = 4 }, { .uv = 2500000, .val = 6 }, { .uv = 3300000, .val = 8 }, - { .uv = 0, .val = 0x3f }, // pin unused + { .uv = 0, .val = 0xf }, // pin unused }; static int mpfs_pinctrl_get_drive_strength_ma(u32 drive_strength) From 8039b75af5808572ff3656eaed07d348aed3989a Mon Sep 17 00:00:00 2001 From: Conor Dooley Date: Mon, 31 Aug 2026 11:50:59 +0100 Subject: [PATCH 186/592] pinctrl: mpfs-mssio: use correct regmap function to set bank voltage regmap_assign_bits() is not the correct function to use for an RMW operation, as it maps to regmap_set_bits() or regmap_clear_bits() and the former will never zero a bit. Use regmap_update_bits() instead, which will actually set the bank voltages to what have been requested. CC: stable@vger.kernel.org Fixes: 488d704ed7b7 ("pinctrl: add polarfire soc mssio pinctrl driver") Signed-off-by: Conor Dooley Signed-off-by: Linus Walleij --- drivers/pinctrl/microchip/pinctrl-mpfs-mssio.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/drivers/pinctrl/microchip/pinctrl-mpfs-mssio.c b/drivers/pinctrl/microchip/pinctrl-mpfs-mssio.c index dafca82f3e54..92d38e5336de 100644 --- a/drivers/pinctrl/microchip/pinctrl-mpfs-mssio.c +++ b/drivers/pinctrl/microchip/pinctrl-mpfs-mssio.c @@ -156,10 +156,10 @@ static void mpfs_pinctrl_set_bank_voltage(struct mpfs_pinctrl *pctrl, unsigned i u32 val = FIELD_PREP(MPFS_PINCTRL_BANK_VOLTAGE_MASK, bank_voltage); if (pin < MPFS_PINCTRL_BANK2_START) - regmap_assign_bits(pctrl->sysreg_regmap, MPFS_PINCTRL_MSSIO_BANK4_CFG_CR, + regmap_update_bits(pctrl->sysreg_regmap, MPFS_PINCTRL_MSSIO_BANK4_CFG_CR, MPFS_PINCTRL_BANK_VOLTAGE_MASK, val); else - regmap_assign_bits(pctrl->sysreg_regmap, MPFS_PINCTRL_MSSIO_BANK2_CFG_CR, + regmap_update_bits(pctrl->sysreg_regmap, MPFS_PINCTRL_MSSIO_BANK2_CFG_CR, MPFS_PINCTRL_BANK_VOLTAGE_MASK, val); } From 4396d70bb7fec531bcf934fed016b2f3300c670b Mon Sep 17 00:00:00 2001 From: Myeonghun Pak Date: Sun, 13 Sep 2026 18:35:38 -0400 Subject: [PATCH 187/592] mmc: sdhci-of-aspeed: Remove children before releasing SDC resources Probe failure and removal leave SDHCI child devices registered after the parent clock and managed resources are released. Unregister the OF children in reverse order before disabling the parent clock on both paths. Use of_platform_device_destroy() because manual child creation does not set the flag required by of_platform_depopulate(). This issue was identified during our ongoing static-analysis research while reviewing kernel code. Fixes: bb7b8ec62dfb ("mmc: sdhci-of-aspeed: Add support for the ASPEED SD controller") Co-developed-by: Ijae Kim Signed-off-by: Ijae Kim Signed-off-by: Myeonghun Pak Assisted-by: OpenAI:GPT-5.6 Cc: stable@vger.kernel.org Signed-off-by: Ulf Hansson --- drivers/mmc/host/sdhci-of-aspeed.c | 5 ++++- 1 file changed, 4 insertions(+), 1 deletion(-) diff --git a/drivers/mmc/host/sdhci-of-aspeed.c b/drivers/mmc/host/sdhci-of-aspeed.c index f5d973783cbe..d317626feb4a 100644 --- a/drivers/mmc/host/sdhci-of-aspeed.c +++ b/drivers/mmc/host/sdhci-of-aspeed.c @@ -560,12 +560,14 @@ static int aspeed_sdc_probe(struct platform_device *pdev) cpdev = of_platform_device_create(child, NULL, &pdev->dev); if (!cpdev) { ret = -ENODEV; - goto err_clk; + goto err_children; } } return 0; +err_children: + device_for_each_child_reverse(&pdev->dev, NULL, of_platform_device_destroy); err_clk: clk_disable_unprepare(sdc->clk); return ret; @@ -575,6 +577,7 @@ static void aspeed_sdc_remove(struct platform_device *pdev) { struct aspeed_sdc *sdc = dev_get_drvdata(&pdev->dev); + device_for_each_child_reverse(&pdev->dev, NULL, of_platform_device_destroy); clk_disable_unprepare(sdc->clk); } From d7fd1f98607f2cd358e583f9548bdf0173090a86 Mon Sep 17 00:00:00 2001 From: Namjae Jeon Date: Thu, 3 Sep 2026 14:06:02 +0900 Subject: [PATCH 188/592] ksmbd: follow SMB2 session expiration semantics ksmbd_session_register() destroys valid sessions after ten seconds of inactivity whenever a client starts another SessionSetup exchange. This confuses Session.IdleTime with Session.ExpirationTime. Windows can create additional authenticated sessions on an existing connection, so deleting the older session invalidates its tree connects and makes mapped drives fail with STATUS_NETWORK_NAME_DELETED. The session expiration rules require the server to change a valid session to expired only after its credential expiration time passes. A valid or expired session otherwise keeps its connection from being scavenged. Connections that have not negotiated a dialect, have no sessions, or have only InProgress sessions are disconnected after an implementation-specific timeout. Use the Windows-compatible 45 second value and run the expiration check periodically for both TCP and SMB Direct. Keep zero as an infinite credential expiration time, and count each Valid-to-Expired transition. Set expired sessions to InProgress when they reauthenticate. If authentication fails, remove the session from the global and per-connection tables immediately, including SMB3 multichannel connections. Retain protection against abandoned SessionId-zero exchanges by allowing only one InProgress authentication per connection. Additional exchanges fail with STATUS_INSUFFICIENT_RESOURCES, and stale InProgress sessions are reaped after the same 45 second setup timeout. This bounds the original unauthenticated memory-exhaustion path without evicting established sessions. Fixes: ea174a918939 ("ksmbd: destroy expired sessions") Reported-by: Mobin Aydinfar Signed-off-by: Namjae Jeon --- fs/smb/server/connection.c | 72 +++++++++++++++++ fs/smb/server/connection.h | 3 + fs/smb/server/mgmt/user_session.c | 124 ++++++++++++++++++++++++++++-- fs/smb/server/mgmt/user_session.h | 8 +- fs/smb/server/proc.c | 2 + fs/smb/server/server.c | 2 +- fs/smb/server/smb2pdu.c | 37 ++++++--- fs/smb/server/smb2pdu.h | 2 - fs/smb/server/stats.h | 1 + 9 files changed, 227 insertions(+), 24 deletions(-) diff --git a/fs/smb/server/connection.c b/fs/smb/server/connection.c index 4cb92d6599ee..d211861ff86f 100644 --- a/fs/smb/server/connection.c +++ b/fs/smb/server/connection.c @@ -22,6 +22,8 @@ static DEFINE_MUTEX(init_lock); static struct ksmbd_conn_ops default_conn_ops; +static struct delayed_work session_expiration_work; +static bool stopping_session_expiration_work; DEFINE_HASHTABLE(conn_list, CONN_HASH_BITS); DECLARE_RWSEM(conn_list_lock); @@ -158,18 +160,28 @@ static void delete_proc_clients(void) {} static struct workqueue_struct *ksmbd_conn_wq; +static void ksmbd_session_expiration_worker(struct work_struct *work); + int ksmbd_conn_wq_init(void) { ksmbd_conn_wq = alloc_workqueue("ksmbd-conn-release", WQ_UNBOUND | WQ_MEM_RECLAIM, 0); if (!ksmbd_conn_wq) return -ENOMEM; + + WRITE_ONCE(stopping_session_expiration_work, false); + INIT_DELAYED_WORK(&session_expiration_work, + ksmbd_session_expiration_worker); + queue_delayed_work(ksmbd_conn_wq, &session_expiration_work, + KSMBD_SESSION_EXPIRATION_INTERVAL); return 0; } void ksmbd_conn_wq_destroy(void) { if (ksmbd_conn_wq) { + WRITE_ONCE(stopping_session_expiration_work, true); + cancel_delayed_work_sync(&session_expiration_work); destroy_workqueue(ksmbd_conn_wq); ksmbd_conn_wq = NULL; } @@ -279,6 +291,7 @@ struct ksmbd_conn *ksmbd_conn_alloc(void) return NULL; conn->need_neg = true; + conn->creation_time = jiffies; ksmbd_conn_set_new(conn); conn->local_nls = load_nls("utf8"); if (!conn->local_nls) @@ -588,6 +601,20 @@ bool ksmbd_conn_alive(struct ksmbd_conn *conn) if (kthread_should_stop()) return false; + /* + * Stale connections that have not completed NEGOTIATE and SESSION_SETUP + * must be disconnected. Do not race a request that is currently + * completing authentication. + */ + if (!atomic_read(&conn->req_running) && + time_after(jiffies, conn->creation_time + + KSMBD_UNAUTHENTICATED_CONN_TIMEOUT) && + (READ_ONCE(conn->need_neg) || + !ksmbd_conn_has_valid_or_expired_session(conn))) { + ksmbd_debug(CONN, "Connection setup timed out\n"); + return false; + } + if (atomic_read(&conn->stats.open_files_count) > 0) return true; @@ -605,6 +632,51 @@ bool ksmbd_conn_alive(struct ksmbd_conn *conn) return true; } +static void ksmbd_session_expiration_worker(struct work_struct *work) +{ + struct ksmbd_conn *conn, *target; + int bkt; + + if (!ksmbd_server_running()) + goto reschedule; + + ksmbd_expire_sessions(); + + /* + * An old connection without a Valid or Expired session must be + * disconnected. Process one connection at a time without holding + * conn_list_lock across transport shutdown. + */ +again: + target = NULL; + down_read(&conn_list_lock); + hash_for_each(conn_list, bkt, conn, hlist) { + if (ksmbd_conn_exiting(conn) || ksmbd_conn_releasing(conn) || + atomic_read(&conn->req_running) || + time_before_eq(jiffies, conn->creation_time + + KSMBD_UNAUTHENTICATED_CONN_TIMEOUT) || + (!READ_ONCE(conn->need_neg) && + ksmbd_conn_has_valid_or_expired_session(conn))) + continue; + + target = ksmbd_conn_get(conn); + break; + } + up_read(&conn_list_lock); + + if (target) { + ksmbd_debug(CONN, "Connection setup timed out\n"); + ksmbd_conn_abort(target); + ksmbd_conn_put(target); + goto again; + } + +reschedule: + if (!READ_ONCE(stopping_session_expiration_work)) + queue_delayed_work(ksmbd_conn_wq, &session_expiration_work, + KSMBD_SESSION_EXPIRATION_INTERVAL); +} + /* "+2" for BCC field (ByteCount, 2 bytes) */ #define SMB1_MIN_SUPPORTED_PDU_SIZE (sizeof(struct smb_hdr) + 2) #define SMB2_MIN_SUPPORTED_PDU_SIZE (sizeof(struct smb2_pdu)) diff --git a/fs/smb/server/connection.h b/fs/smb/server/connection.h index 63484c8efbbd..371f17b4f02a 100644 --- a/fs/smb/server/connection.h +++ b/fs/smb/server/connection.h @@ -77,6 +77,7 @@ struct ksmbd_conn { struct rw_semaphore session_lock; /* smb session 1 per user */ struct xarray sessions; + unsigned long creation_time; unsigned long last_active; /* How many request are running currently */ atomic_t req_running; @@ -192,6 +193,8 @@ struct ksmbd_transport { #define KSMBD_TCP_RECV_TIMEOUT (7 * HZ) #define KSMBD_TCP_SEND_TIMEOUT (5 * HZ) +#define KSMBD_SESSION_EXPIRATION_INTERVAL (5 * HZ) +#define KSMBD_UNAUTHENTICATED_CONN_TIMEOUT (45 * HZ) #define KSMBD_TCP_PEER_SOCKADDR(c) ((struct sockaddr *)&((c)->peer_addr)) #define CONN_HASH_BITS 12 diff --git a/fs/smb/server/mgmt/user_session.c b/fs/smb/server/mgmt/user_session.c index 2eb8f730e99e..44dc3f800cd4 100644 --- a/fs/smb/server/mgmt/user_session.c +++ b/fs/smb/server/mgmt/user_session.c @@ -22,6 +22,7 @@ static DEFINE_IDA(session_ida); #define SESSION_HASH_BITS 12 +#define KSMBD_MAX_PENDING_SESSIONS 1 static DEFINE_HASHTABLE(sessions_table, SESSION_HASH_BITS); static DECLARE_RWSEM(sessions_table_lock); @@ -432,26 +433,31 @@ struct ksmbd_session *__session_lookup(unsigned long long id) return NULL; } -static void ksmbd_expire_session(struct ksmbd_conn *conn) +static bool ksmbd_too_many_session_setups(struct ksmbd_conn *conn) { unsigned long id; struct ksmbd_session *sess; + unsigned int pending = 0; down_write(&sessions_table_lock); down_write(&conn->session_lock); xa_for_each(&conn->sessions, id, sess) { + if (READ_ONCE(sess->state) != SMB2_SESSION_IN_PROGRESS) + continue; + if (atomic_read(&sess->refcnt) <= 1 && - (sess->state != SMB2_SESSION_VALID || - time_after(jiffies, - sess->last_active + SMB2_SESSION_TIMEOUT))) { + time_after(jiffies, sess->last_active + + KSMBD_UNAUTHENTICATED_CONN_TIMEOUT)) { xa_erase(&conn->sessions, sess->id); ksmbd_session_remove_from_table(sess); ksmbd_session_destroy(sess); continue; } + pending++; } up_write(&conn->session_lock); up_write(&sessions_table_lock); + return pending >= KSMBD_MAX_PENDING_SESSIONS; } int ksmbd_session_register(struct ksmbd_conn *conn, @@ -461,9 +467,12 @@ int ksmbd_session_register(struct ksmbd_conn *conn, sess->dialect = conn->dialect; memcpy(sess->ClientGUID, conn->ClientGUID, SMB2_CLIENT_GUID_SIZE); - ksmbd_expire_session(conn); - ret = xa_err(xa_store(&conn->sessions, sess->id, sess, - KSMBD_DEFAULT_GFP)); + /* Bound abandoned SessionId-zero authentication exchanges. */ + if (ksmbd_too_many_session_setups(conn)) + ret = -ENOSPC; + else + ret = xa_err(xa_store(&conn->sessions, sess->id, sess, + KSMBD_DEFAULT_GFP)); if (ret) { down_write(&sessions_table_lock); ksmbd_session_remove_from_table(sess); @@ -474,6 +483,105 @@ int ksmbd_session_register(struct ksmbd_conn *conn, return ret; } +void ksmbd_session_unregister(struct ksmbd_conn *conn, + struct ksmbd_session *sess) +{ + struct ksmbd_conn *session_conns[KSMBD_MAX_CHANNELS]; + struct channel *chann; + unsigned long index; + unsigned int nr_conns = 0, i; + bool removed = false; + + down_write(&sessions_table_lock); + if (!hlist_unhashed(&sess->hlist)) { + /* Keep each channel connection stable under sessions_table_lock. */ + down_read(&sess->chann_lock); + xa_for_each(&sess->ksmbd_chann_list, index, chann) { + if (nr_conns == ARRAY_SIZE(session_conns)) + break; + session_conns[nr_conns++] = chann->conn; + } + up_read(&sess->chann_lock); + + ksmbd_session_remove_from_table(sess); + removed = true; + } + + down_write(&conn->session_lock); + if (xa_load(&conn->sessions, sess->id) == sess) + xa_erase(&conn->sessions, sess->id); + up_write(&conn->session_lock); + for (i = 0; i < nr_conns; i++) { + if (session_conns[i] == conn) + continue; + down_write(&session_conns[i]->session_lock); + if (xa_load(&session_conns[i]->sessions, sess->id) == sess) + xa_erase(&session_conns[i]->sessions, sess->id); + up_write(&session_conns[i]->session_lock); + } + up_write(&sessions_table_lock); + + if (removed) + ksmbd_user_session_put(sess); +} + +bool ksmbd_conn_has_valid_or_expired_session(struct ksmbd_conn *conn) +{ + struct ksmbd_session *sess; + unsigned long id; + int state, bkt; + bool found = false; + + down_read(&conn->session_lock); + xa_for_each(&conn->sessions, id, sess) { + state = READ_ONCE(sess->state); + if (state == SMB2_SESSION_VALID || + state == SMB2_SESSION_EXPIRED) { + found = true; + break; + } + } + up_read(&conn->session_lock); + if (found) + return true; + + /* A session bound through SMB3 multichannel is not in conn->sessions. */ + down_read(&sessions_table_lock); + hash_for_each(sessions_table, bkt, sess, hlist) { + state = READ_ONCE(sess->state); + if (state != SMB2_SESSION_VALID && + state != SMB2_SESSION_EXPIRED) + continue; + + down_read(&sess->chann_lock); + found = xa_load(&sess->ksmbd_chann_list, (long)conn); + up_read(&sess->chann_lock); + if (found) + break; + } + up_read(&sessions_table_lock); + return found; +} + +void ksmbd_expire_sessions(void) +{ + struct ksmbd_session *sess; + u64 now = ktime_get_real_seconds(); + int bkt; + + down_read(&sessions_table_lock); + hash_for_each(sessions_table, bkt, sess, hlist) { + if (READ_ONCE(sess->state) != SMB2_SESSION_VALID || + !sess->kerberos_expiry || now < sess->kerberos_expiry) + continue; + + if (cmpxchg(&sess->state, SMB2_SESSION_VALID, + SMB2_SESSION_EXPIRED) == SMB2_SESSION_VALID) + ksmbd_counter_inc(KSMBD_COUNTER_SESSION_TIMEOUTS); + } + up_read(&sessions_table_lock); +} + static int ksmbd_chann_del(struct ksmbd_conn *conn, struct ksmbd_session *sess) { struct channel *chann; @@ -488,7 +596,7 @@ static int ksmbd_chann_del(struct ksmbd_conn *conn, struct ksmbd_session *sess) return 0; } -void ksmbd_sessions_deregister(struct ksmbd_conn *conn) +void ksmbd_conn_sessions_cleanup(struct ksmbd_conn *conn) { struct ksmbd_session *sess; unsigned long id; diff --git a/fs/smb/server/mgmt/user_session.h b/fs/smb/server/mgmt/user_session.h index 3e52d4cc1324..217258551d6d 100644 --- a/fs/smb/server/mgmt/user_session.h +++ b/fs/smb/server/mgmt/user_session.h @@ -72,6 +72,8 @@ struct ksmbd_session { struct rw_semaphore rpc_lock; }; +#define KSMBD_MAX_CHANNELS 32 + static inline int test_session_flag(struct ksmbd_session *sess, int bit) { return sess->flags & bit; @@ -98,7 +100,11 @@ bool is_ksmbd_session_in_connection(struct ksmbd_conn *conn, unsigned long long id); int ksmbd_session_register(struct ksmbd_conn *conn, struct ksmbd_session *sess); -void ksmbd_sessions_deregister(struct ksmbd_conn *conn); +void ksmbd_session_unregister(struct ksmbd_conn *conn, + struct ksmbd_session *sess); +void ksmbd_conn_sessions_cleanup(struct ksmbd_conn *conn); +bool ksmbd_conn_has_valid_or_expired_session(struct ksmbd_conn *conn); +void ksmbd_expire_sessions(void); struct ksmbd_session *__session_lookup(unsigned long long id); struct ksmbd_session *ksmbd_session_lookup_all(struct ksmbd_conn *conn, unsigned long long id); diff --git a/fs/smb/server/proc.c b/fs/smb/server/proc.c index 826353ed0553..19f0f2cfbf54 100644 --- a/fs/smb/server/proc.c +++ b/fs/smb/server/proc.c @@ -178,6 +178,8 @@ static int proc_show_ksmbd_stats(struct seq_file *m, void *v) proc_show_runtime_totals(m); seq_printf(m, "sessions:\t%lld\n", ksmbd_counter_sum(KSMBD_COUNTER_SESSIONS)); + seq_printf(m, "session_timeouts:\t%lld\n", + ksmbd_counter_sum(KSMBD_COUNTER_SESSION_TIMEOUTS)); seq_printf(m, "tree_connects:\t%lld\n", ksmbd_counter_sum(KSMBD_COUNTER_TREE_CONNS)); seq_printf(m, "requests:\t%lld\n", diff --git a/fs/smb/server/server.c b/fs/smb/server/server.c index 0069d4e6a60a..0827c8c51006 100644 --- a/fs/smb/server/server.c +++ b/fs/smb/server/server.c @@ -414,7 +414,7 @@ static int ksmbd_server_process_request(struct ksmbd_conn *conn) static int ksmbd_server_terminate_conn(struct ksmbd_conn *conn) { - ksmbd_sessions_deregister(conn); + ksmbd_conn_sessions_cleanup(conn); destroy_lease_table(conn); return 0; } diff --git a/fs/smb/server/smb2pdu.c b/fs/smb/server/smb2pdu.c index b7ce67094626..8acc5174530b 100644 --- a/fs/smb/server/smb2pdu.c +++ b/fs/smb/server/smb2pdu.c @@ -85,8 +85,6 @@ struct channel *lookup_chann_list(struct ksmbd_session *sess, struct ksmbd_conn return chann; } -#define KSMBD_MAX_CHANNELS 32 - static int register_session_channel(struct ksmbd_session *sess, struct ksmbd_conn *conn, const char *sess_key) @@ -933,8 +931,14 @@ static bool smb2_session_expired_cmd_allowed(struct ksmbd_work *work, static bool smb2_session_kerberos_expired(struct ksmbd_session *sess) { - return sess->kerberos_expiry && - ktime_get_real_seconds() >= sess->kerberos_expiry; + if (!sess->kerberos_expiry || + ktime_get_real_seconds() < sess->kerberos_expiry) + return false; + + if (cmpxchg(&sess->state, SMB2_SESSION_VALID, + SMB2_SESSION_EXPIRED) == SMB2_SESSION_VALID) + ksmbd_counter_inc(KSMBD_COUNTER_SESSION_TIMEOUTS); + return true; } /** @@ -969,9 +973,8 @@ int smb2_check_user_session(struct ksmbd_work *work) if (!work->next_smb2_rcv_hdr_off && sess_id) work->sess = ksmbd_session_lookup_all_states(conn, sess_id); if (work->sess) { - if (smb2_session_kerberos_expired(work->sess)) { - work->sess->state = SMB2_SESSION_EXPIRED; - } else if (work->sess->state != SMB2_SESSION_VALID) { + if (!smb2_session_kerberos_expired(work->sess) && + work->sess->state != SMB2_SESSION_VALID) { ksmbd_user_session_put(work->sess); work->sess = NULL; } @@ -996,8 +999,7 @@ int smb2_check_user_session(struct ksmbd_work *work) sess_id, work->sess->id); return -EINVAL; } - if (smb2_session_kerberos_expired(work->sess)) - work->sess->state = SMB2_SESSION_EXPIRED; + smb2_session_kerberos_expired(work->sess); if (work->sess->state != SMB2_SESSION_VALID) { pr_err("compound request on a non-valid session (state %d)\n", work->sess->state); @@ -1014,7 +1016,6 @@ int smb2_check_user_session(struct ksmbd_work *work) work->sess = ksmbd_session_lookup_all_states(conn, sess_id); if (work->sess) { if (smb2_session_kerberos_expired(work->sess)) { - work->sess->state = SMB2_SESSION_EXPIRED; return smb2_session_expired_cmd_allowed(work, cmd) ? 1 : -EKEYEXPIRED; } @@ -2436,7 +2437,7 @@ int smb2_sess_setup(struct ksmbd_work *work) struct ksmbd_conn *conn = work->conn; struct smb2_sess_setup_req *req; struct smb2_sess_setup_rsp *rsp; - struct ksmbd_session *sess; + struct ksmbd_session *sess = NULL; struct negotiate_message *negblob; unsigned int negblob_len, negblob_off; int rc = 0; @@ -2592,6 +2593,9 @@ int smb2_sess_setup(struct ksmbd_work *work) goto out_err; } + if (work->session_setup_reauth) + WRITE_ONCE(sess->state, SMB2_SESSION_IN_PROGRESS); + conn->binding = false; } work->sess = sess; @@ -2703,6 +2707,14 @@ int smb2_sess_setup(struct ksmbd_work *work) } if (rc < 0) { + bool setup_in_progress = sess && + READ_ONCE(sess->state) == SMB2_SESSION_IN_PROGRESS && + !(req->Flags & SMB2_SESSION_REQ_FLAG_BINDING); + + /* Authentication errors must not leave the new session published. */ + if (setup_in_progress) + ksmbd_session_unregister(conn, sess); + if (sess && conn->dialect == SMB311_PROT_ID && (req->Flags & SMB2_SESSION_REQ_FLAG_BINDING)) { struct preauth_session *preauth_sess; @@ -2736,7 +2748,8 @@ int smb2_sess_setup(struct ksmbd_work *work) * For binding requests, session belongs to another * connection. Do not expire it. */ - if (!(req->Flags & SMB2_SESSION_REQ_FLAG_BINDING)) { + if (!(req->Flags & SMB2_SESSION_REQ_FLAG_BINDING) && + !setup_in_progress) { sess->last_active = jiffies; sess->kerberos_expiry = 0; sess->state = SMB2_SESSION_EXPIRED; diff --git a/fs/smb/server/smb2pdu.h b/fs/smb/server/smb2pdu.h index 3f08d1ca5a38..a6200d8630e2 100644 --- a/fs/smb/server/smb2pdu.h +++ b/fs/smb/server/smb2pdu.h @@ -61,8 +61,6 @@ struct preauth_integrity_info { #define SMB2_SESSION_IN_PROGRESS BIT(0) #define SMB2_SESSION_VALID BIT(1) -#define SMB2_SESSION_TIMEOUT (10 * HZ) - /* Apple Defined Contexts */ #define SMB2_CREATE_AAPL "AAPL" diff --git a/fs/smb/server/stats.h b/fs/smb/server/stats.h index bc864efa0d46..8b32b8b4e8be 100644 --- a/fs/smb/server/stats.h +++ b/fs/smb/server/stats.h @@ -15,6 +15,7 @@ enum { KSMBD_COUNTER_SESSIONS = 0, + KSMBD_COUNTER_SESSION_TIMEOUTS, KSMBD_COUNTER_TREE_CONNS, KSMBD_COUNTER_REQUESTS, KSMBD_COUNTER_STATUS_SUCCESS, From 65bcc5f89704efe5b9d69d4ea2c1002d90c31382 Mon Sep 17 00:00:00 2001 From: Slawomir Stepien Date: Mon, 14 Sep 2026 11:50:07 +0200 Subject: [PATCH 189/592] HID: amd_sfh: Validate PCI BAR size before mapping The amd_sfh driver maps PCI BAR 2 using pcim_iomap_regions() and subsequently accesses MMIO registers at offsets up to 0x10958 (e.g., AMD_P2C_MSG3 at 0x1068C). However, the driver never validates that the BAR size is large enough to cover these accesses. If the driver is bound to a device with a smaller BAR 2, this leads to an out-of-bounds memory access and a page fault during the probe function. For example, a page fault can occur when reading from privdata->mmio + AMD_P2C_MSG3 in mp2_select_ops(): BUG: unable to handle page fault for address: ffffc9000390368c PGD 100000067 P4D 100000067 PUD 1012c1067 PMD 105b64067 PTE 0 Oops: Oops: 0000 [#1] SMP KASAN NOPTI RIP: 0010:readl arch/x86/include/asm/io.h:59 [inline] RIP: 0010:mp2_select_ops drivers/hid/amd-sfh-hid/amd_sfh_pcie.c:282 [inline] RIP: 0010:amd_mp2_pci_probe+0x337/0x5f0 drivers/hid/amd-sfh-hid/amd_sfh_pcie.c:487 Call Trace: local_pci_probe drivers/pci/pci-driver.c:332 [inline] pci_call_probe drivers/pci/pci-driver.c:394 [inline] __pci_device_probe drivers/pci/pci-driver.c:455 [inline] pci_device_probe+0x431/0xc90 drivers/pci/pci-driver.c:489 Fix this by verifying that the length of BAR 2 is at least 128KB before attempting to map it. Since the maximum accessed offset is 0x10958, and PCI BAR sizes are powers of 2, any legitimate hardware will have a BAR size of at least 128KB. Fixes: 4f567b9f8141 ("SFH: PCIe driver to add support of AMD sensor fusion hub") Assisted-by: Gemini:gemini-3.7-flash Gemini:gemini-3.1-pro-preview syzbot Reported-by: syzbot+4eadd4dfe9e66522bae8@syzkaller.appspotmail.com Closes: https://syzkaller.appspot.com/bug?extid=4eadd4dfe9e66522bae8 Link: https://syzkaller.appspot.com/ai_job?id=3bc1c45c-548f-4ab5-8243-d2c8ec321d6c Signed-off-by: Slawomir Stepien Acked-by: Basavaraj Natikar Link: https://syzkaller.appspot.com/bug?extid=4eadd4dfe9e66522bae8 Signed-off-by: Jiri Kosina --- drivers/hid/amd-sfh-hid/amd_sfh_common.h | 4 ++++ drivers/hid/amd-sfh-hid/amd_sfh_pcie.c | 10 ++++++++++ 2 files changed, 14 insertions(+) diff --git a/drivers/hid/amd-sfh-hid/amd_sfh_common.h b/drivers/hid/amd-sfh-hid/amd_sfh_common.h index f8c6b7fc34fb..3d29119a3765 100644 --- a/drivers/hid/amd-sfh-hid/amd_sfh_common.h +++ b/drivers/hid/amd-sfh-hid/amd_sfh_common.h @@ -13,11 +13,15 @@ #include #include #include +#include #include "amd_sfh_hid.h" #define PCI_DEVICE_ID_AMD_MP2 0x15E4 #define PCI_DEVICE_ID_AMD_MP2_1_1 0x164A +/* The BAR 2 size must cover the highest register offset (0x10958) */ +#define AMD_SFH_MIN_BAR_SIZE SZ_128K + #define AMD_C2P_MSG(regno) (0x10500 + ((regno) * 4)) #define AMD_P2C_MSG(regno) (0x10680 + ((regno) * 4)) diff --git a/drivers/hid/amd-sfh-hid/amd_sfh_pcie.c b/drivers/hid/amd-sfh-hid/amd_sfh_pcie.c index eda26a094d3f..061a63519d44 100644 --- a/drivers/hid/amd-sfh-hid/amd_sfh_pcie.c +++ b/drivers/hid/amd-sfh-hid/amd_sfh_pcie.c @@ -497,6 +497,16 @@ static int amd_mp2_pci_probe(struct pci_dev *pdev, const struct pci_device_id *i if (rc) return rc; + if (!(pci_resource_flags(pdev, 2) & IORESOURCE_MEM)) { + dev_err(&pdev->dev, "BAR 2 is not IORESOURCE_MEM\n"); + return -ENODEV; + } + + if (pci_resource_len(pdev, 2) < AMD_SFH_MIN_BAR_SIZE) { + dev_err(&pdev->dev, "BAR 2 is too small\n"); + return -EINVAL; + } + rc = pcim_iomap_regions(pdev, BIT(2), DRIVER_NAME); if (rc) return rc; From 247a82da6f563dcfd9074a68f99a0c0997d0679c Mon Sep 17 00:00:00 2001 From: Ilia Levi Date: Tue, 8 Sep 2026 17:50:48 +0100 Subject: [PATCH 190/592] drm/xe/mmio_gem: forbid VMA split The fault handler assumes it always operates on a VMA spanning the entire GEM object. This does not hold when the VMA has been split, e.g. by a partial munmap or mprotect. In that case the handler may map wrong physical pages or cause SIGBUS. Handle this by forbidding VMA split, as partial unmaps are not deemed useful for MMIO GEMs. Suggested-by: Matthew Auld Signed-off-by: Ilia Levi Fixes: 1ffcf8b8ae8a ("drm/xe: Support for mmap-ing mmio regions") Reviewed-by: Matthew Auld Signed-off-by: Matthew Auld Link: https://patch.msgid.link/20260908165046.1393557-11-matthew.auld@intel.com (cherry picked from commit f3391a0b12d7bf826a0b21600d2f294f3dce4c14) Signed-off-by: Rodrigo Vivi --- drivers/gpu/drm/xe/xe_mmio_gem.c | 10 ++++++++++ 1 file changed, 10 insertions(+) diff --git a/drivers/gpu/drm/xe/xe_mmio_gem.c b/drivers/gpu/drm/xe/xe_mmio_gem.c index 3741ae60f532..d54477b93b6e 100644 --- a/drivers/gpu/drm/xe/xe_mmio_gem.c +++ b/drivers/gpu/drm/xe/xe_mmio_gem.c @@ -39,10 +39,20 @@ struct xe_mmio_gem { phys_addr_t phys_addr; }; +static int xe_mmio_gem_vm_may_split(struct vm_area_struct *area, unsigned long addr) +{ + /* + * Forbid splitting. Together with VM_DONTEXPAND, this keeps the VMA + * matching the GEM object exactly. + */ + return -EINVAL; +} + static const struct vm_operations_struct vm_ops = { .open = drm_gem_vm_open, .close = drm_gem_vm_close, .fault = xe_mmio_gem_vm_fault, + .may_split = xe_mmio_gem_vm_may_split, }; static const struct drm_gem_object_funcs xe_mmio_gem_funcs = { From 819f189265a5955da743e78e20a713a038982e89 Mon Sep 17 00:00:00 2001 From: Ilia Levi Date: Tue, 8 Sep 2026 17:50:49 +0100 Subject: [PATCH 191/592] drm/xe/mmio_gem: use write-back mapping for dummy page Currently vmf_insert_pfn() maps the dummy page as UC, inheriting the VMA's page protection which was set for the real MMIO region. This conflicts with the direct map's WB mapping of the same page, creating a cache type alias which is architecturally undefined on some platforms. Use vmf_insert_pfn_prot() with a WB pgprot instead. Also simplify to fault in the requested page instead of the whole VMA. Fixes: 1ffcf8b8ae8a ("drm/xe: Support for mmap-ing mmio regions") Reported-by: Sashiko Closes: https://sashiko.dev/#/patchset/20260525125801.975038-6-ilia.levi%40intel.com Assisted-by: GitHub-Copilot:claude-opus-4.6 Signed-off-by: Ilia Levi Reviewed-by: Matthew Auld Signed-off-by: Matthew Auld Link: https://patch.msgid.link/20260908165046.1393557-12-matthew.auld@intel.com (cherry picked from commit 1e8e28e35df0e77ae1b22fc091c1f422f62fa5e9) Signed-off-by: Rodrigo Vivi --- drivers/gpu/drm/xe/xe_mmio_gem.c | 19 +++++-------------- 1 file changed, 5 insertions(+), 14 deletions(-) diff --git a/drivers/gpu/drm/xe/xe_mmio_gem.c b/drivers/gpu/drm/xe/xe_mmio_gem.c index d54477b93b6e..96f3bf46fd75 100644 --- a/drivers/gpu/drm/xe/xe_mmio_gem.c +++ b/drivers/gpu/drm/xe/xe_mmio_gem.c @@ -172,14 +172,13 @@ static void xe_mmio_gem_release_dummy_page(struct drm_device *dev, void *res) __free_page((struct page *)res); } -static vm_fault_t xe_mmio_gem_vm_fault_dummy_page(struct vm_area_struct *vma) +static vm_fault_t xe_mmio_gem_vm_fault_dummy_page(struct vm_fault *vmf) { + struct vm_area_struct *vma = vmf->vma; struct drm_gem_object *base = vma->vm_private_data; struct drm_device *dev = base->dev; - vm_fault_t ret = VM_FAULT_NOPAGE; struct page *page; unsigned long pfn; - unsigned long i; page = alloc_page(GFP_KERNEL | __GFP_ZERO); if (!page) @@ -190,16 +189,8 @@ static vm_fault_t xe_mmio_gem_vm_fault_dummy_page(struct vm_area_struct *vma) pfn = page_to_pfn(page); - /* Map the entire VMA to the same dummy page */ - for (i = 0; i < base->size; i += PAGE_SIZE) { - unsigned long addr = vma->vm_start + i; - - ret = vmf_insert_pfn(vma, addr, pfn); - if (ret & VM_FAULT_ERROR) - break; - } - - return ret; + return vmf_insert_pfn_prot(vma, vmf->address, pfn, + vm_get_page_prot(vma->vm_flags)); } static vm_fault_t xe_mmio_gem_vm_fault(struct vm_fault *vmf) @@ -219,7 +210,7 @@ static vm_fault_t xe_mmio_gem_vm_fault(struct vm_fault *vmf) * It is assumed the userspace will receive the notification via some * other channel (e.g. drm uevent). */ - return xe_mmio_gem_vm_fault_dummy_page(vma); + return xe_mmio_gem_vm_fault_dummy_page(vmf); } for (i = 0; i < base->size; i += PAGE_SIZE) { From 0c50663403b7aa271b1974ba32ee6c8296711142 Mon Sep 17 00:00:00 2001 From: Ilia Levi Date: Tue, 8 Sep 2026 17:50:50 +0100 Subject: [PATCH 192/592] drm/xe/mmio_gem: simplify fault handler loop Make the iteration over the addresses in the VMA more explicit. No functional change, as the VMA matches the GEM object exactly. Signed-off-by: Ilia Levi Reviewed-by: Matthew Auld Signed-off-by: Matthew Auld Link: https://patch.msgid.link/20260908165046.1393557-13-matthew.auld@intel.com (cherry picked from commit 6666ca9192f3bdf839aad33b9e1c9ebb7a222a29) Signed-off-by: Rodrigo Vivi --- drivers/gpu/drm/xe/xe_mmio_gem.c | 12 ++++++------ 1 file changed, 6 insertions(+), 6 deletions(-) diff --git a/drivers/gpu/drm/xe/xe_mmio_gem.c b/drivers/gpu/drm/xe/xe_mmio_gem.c index 96f3bf46fd75..3b6d04efe81a 100644 --- a/drivers/gpu/drm/xe/xe_mmio_gem.c +++ b/drivers/gpu/drm/xe/xe_mmio_gem.c @@ -200,7 +200,7 @@ static vm_fault_t xe_mmio_gem_vm_fault(struct vm_fault *vmf) struct xe_mmio_gem *obj = to_xe_mmio_gem(base); struct drm_device *dev = base->dev; vm_fault_t ret = VM_FAULT_NOPAGE; - unsigned long i; + unsigned long addr, pfn; int idx; if (!drm_dev_enter(dev, &idx)) { @@ -213,13 +213,13 @@ static vm_fault_t xe_mmio_gem_vm_fault(struct vm_fault *vmf) return xe_mmio_gem_vm_fault_dummy_page(vmf); } - for (i = 0; i < base->size; i += PAGE_SIZE) { - unsigned long addr = vma->vm_start + i; - unsigned long phys_addr = obj->phys_addr + i; - - ret = vmf_insert_pfn(vma, addr, PHYS_PFN(phys_addr)); + pfn = PHYS_PFN(obj->phys_addr); + for (addr = vma->vm_start; addr < vma->vm_end; addr += PAGE_SIZE) { + ret = vmf_insert_pfn(vma, addr, pfn); if (ret & VM_FAULT_ERROR) break; + + pfn++; } drm_dev_exit(idx); From 37fcbd7b2f8996d783932dab11bc668e169b0de6 Mon Sep 17 00:00:00 2001 From: Shuicheng Lin Date: Tue, 8 Sep 2026 17:50:51 +0100 Subject: [PATCH 193/592] drm/xe/mmio_gem: Revoke drm_vma_node on xe_mmio_gem destroy xe_mmio_gem_create() calls drm_vma_node_allow() but nothing ever calls drm_vma_node_revoke(). The drm_vma_offset_file rb-tree entry allocated by drm_vma_node_allow() is not freed by drm_gem_object_release(), so it is leaked on every create/destroy cycle. Add a struct drm_file * parameter to xe_mmio_gem_destroy() and call drm_vma_node_revoke() from there, mirroring the drm_vma_node_allow() call in xe_mmio_gem_create(). Fixes: 1ffcf8b8ae8a ("drm/xe: Support for mmap-ing mmio regions") Suggested-by: Ilia Levi Assisted-by: Claude:claude-opus-4.6 Signed-off-by: Shuicheng Lin Reviewed-by: Ilia Levi Signed-off-by: Matthew Auld Link: https://patch.msgid.link/20260908165046.1393557-14-matthew.auld@intel.com (cherry picked from commit 32f0cb250598456d812fb7ca57a040282858323d) Signed-off-by: Rodrigo Vivi --- drivers/gpu/drm/xe/xe_mmio_gem.c | 4 +++- drivers/gpu/drm/xe/xe_mmio_gem.h | 2 +- 2 files changed, 4 insertions(+), 2 deletions(-) diff --git a/drivers/gpu/drm/xe/xe_mmio_gem.c b/drivers/gpu/drm/xe/xe_mmio_gem.c index 3b6d04efe81a..3c42d8358c7d 100644 --- a/drivers/gpu/drm/xe/xe_mmio_gem.c +++ b/drivers/gpu/drm/xe/xe_mmio_gem.c @@ -138,14 +138,16 @@ static void xe_mmio_gem_free(struct drm_gem_object *base) /** * xe_mmio_gem_destroy - Destroy the GEM object that exposes an MMIO region * @gem: the GEM object to destroy + * @file: DRM file descriptor previously passed to xe_mmio_gem_create() * * This function releases resources associated with the GEM object created by * xe_mmio_gem_create(). * * See: "Exposing MMIO regions to userspace" */ -void xe_mmio_gem_destroy(struct xe_mmio_gem *gem) +void xe_mmio_gem_destroy(struct xe_mmio_gem *gem, struct drm_file *file) { + drm_vma_node_revoke(&gem->base.vma_node, file); xe_mmio_gem_free(&gem->base); } diff --git a/drivers/gpu/drm/xe/xe_mmio_gem.h b/drivers/gpu/drm/xe/xe_mmio_gem.h index 4b76d5586ebb..80d7795f07c8 100644 --- a/drivers/gpu/drm/xe/xe_mmio_gem.h +++ b/drivers/gpu/drm/xe/xe_mmio_gem.h @@ -15,6 +15,6 @@ struct xe_mmio_gem; struct xe_mmio_gem *xe_mmio_gem_create(struct xe_device *xe, struct drm_file *file, phys_addr_t phys_addr, size_t size); u64 xe_mmio_gem_mmap_offset(struct xe_mmio_gem *gem); -void xe_mmio_gem_destroy(struct xe_mmio_gem *gem); +void xe_mmio_gem_destroy(struct xe_mmio_gem *gem, struct drm_file *file); #endif /* _XE_MMIO_GEM_H_ */ From de40d31275cd57408ff1fdd97ea117a3f8c50cc5 Mon Sep 17 00:00:00 2001 From: Ilia Levi Date: Tue, 8 Sep 2026 17:50:52 +0100 Subject: [PATCH 194/592] drm/xe/mmio_gem: cache the dummy page per object Currently, when the fault handler provides a dummy page, it allocates a new one on every invocation and ties its lifetime to the drm_device via drmm_add_action_or_reset(). Concurrent faults after hot-unplug therefore accumulate pages that persist until device teardown. Cache a single dummy page in the xe_mmio_gem object and use dma_resv lock to protect its allocation. Free it with the object. v2: use dma_resv lock to protect the allocation (Matt Auld) Assisted-by: GitHub-Copilot:claude-opus-4.6 Signed-off-by: Ilia Levi Reviewed-by: Matthew Auld Signed-off-by: Matthew Auld Link: https://patch.msgid.link/20260908165046.1393557-15-matthew.auld@intel.com (cherry picked from commit 8bf6213f9831e46313af4722a1ee6db1b7596378) Signed-off-by: Rodrigo Vivi --- drivers/gpu/drm/xe/xe_mmio_gem.c | 27 ++++++++++++++++----------- 1 file changed, 16 insertions(+), 11 deletions(-) diff --git a/drivers/gpu/drm/xe/xe_mmio_gem.c b/drivers/gpu/drm/xe/xe_mmio_gem.c index 3c42d8358c7d..970b1e2f4981 100644 --- a/drivers/gpu/drm/xe/xe_mmio_gem.c +++ b/drivers/gpu/drm/xe/xe_mmio_gem.c @@ -5,9 +5,9 @@ #include "xe_mmio_gem.h" +#include #include #include -#include #include "xe_device_types.h" @@ -37,6 +37,7 @@ static vm_fault_t xe_mmio_gem_vm_fault(struct vm_fault *); struct xe_mmio_gem { struct drm_gem_object base; phys_addr_t phys_addr; + struct page *dummy_page; /* protected by the GEM's dma_resv */ }; static int xe_mmio_gem_vm_may_split(struct vm_area_struct *area, unsigned long addr) @@ -131,6 +132,8 @@ static void xe_mmio_gem_free(struct drm_gem_object *base) { struct xe_mmio_gem *obj = to_xe_mmio_gem(base); + if (obj->dummy_page) + __free_page(obj->dummy_page); drm_gem_object_release(base); kfree(obj); } @@ -169,27 +172,29 @@ static int xe_mmio_gem_mmap(struct drm_gem_object *base, struct vm_area_struct * return 0; } -static void xe_mmio_gem_release_dummy_page(struct drm_device *dev, void *res) +static int alloc_dummy_page_if_needed(struct drm_gem_object *base) { - __free_page((struct page *)res); + struct xe_mmio_gem *obj = to_xe_mmio_gem(base); + + dma_resv_lock(base->resv, NULL); + if (!obj->dummy_page) + obj->dummy_page = alloc_page(GFP_KERNEL | __GFP_ZERO); + dma_resv_unlock(base->resv); + + return obj->dummy_page ? 0 : -ENOMEM; } static vm_fault_t xe_mmio_gem_vm_fault_dummy_page(struct vm_fault *vmf) { struct vm_area_struct *vma = vmf->vma; struct drm_gem_object *base = vma->vm_private_data; - struct drm_device *dev = base->dev; - struct page *page; + struct xe_mmio_gem *obj = to_xe_mmio_gem(base); unsigned long pfn; - page = alloc_page(GFP_KERNEL | __GFP_ZERO); - if (!page) + if (alloc_dummy_page_if_needed(base)) return VM_FAULT_OOM; - if (drmm_add_action_or_reset(dev, xe_mmio_gem_release_dummy_page, page)) - return VM_FAULT_OOM; - - pfn = page_to_pfn(page); + pfn = page_to_pfn(obj->dummy_page); return vmf_insert_pfn_prot(vma, vmf->address, pfn, vm_get_page_prot(vma->vm_flags)); From d0c09528781938362655d9c336364e777119bd6b Mon Sep 17 00:00:00 2001 From: Ilia Levi Date: Tue, 8 Sep 2026 17:50:53 +0100 Subject: [PATCH 195/592] drm/xe/mmio_gem: fix destroy flow xe_mmio_gem_destroy() currently frees the GEM object directly, bypassing reference counting. Since existing VMAs hold a reference and the fault handler accesses the object through vma->vm_private_data, this is use-after-free. Additionally, nothing prevents the fault handler from installing PTEs to the real MMIO after destroy. Fix this with proper synchronization and refcounting. Also, do not set vm_pgoff to zero. Many DRM drivers do this because helpers like dma_mmap_pages() interpret vm_pgoff as an intra-buffer page offset; leaving the DRM fake offset there would break these helpers. Those drivers can get away with zeroing it because they map eagerly - all PTEs are established before mmap returns, so vm_pgoff is never consulted again. Our driver does not use such helpers and the newly introduced call to drm_vma_node_unmap() relies on vm_pgoff being untouched. v2: (Matt Auld) - use dma_resv lock to serialize fault handler with destroy - SIGBUS on access after destroy Fixes: 1ffcf8b8ae8a ("drm/xe: Support for mmap-ing mmio regions") Assisted-by: GitHub-Copilot:claude-opus-4.6 Signed-off-by: Ilia Levi Reviewed-by: Matthew Auld Signed-off-by: Matthew Auld Link: https://patch.msgid.link/20260908165046.1393557-16-matthew.auld@intel.com (cherry picked from commit fb2ee38bab8025ad6a7a9cbb4635c5a178e4a7bc) Signed-off-by: Rodrigo Vivi --- drivers/gpu/drm/xe/xe_mmio_gem.c | 42 ++++++++++++++++++++++++++------ 1 file changed, 35 insertions(+), 7 deletions(-) diff --git a/drivers/gpu/drm/xe/xe_mmio_gem.c b/drivers/gpu/drm/xe/xe_mmio_gem.c index 970b1e2f4981..5ffe03d36190 100644 --- a/drivers/gpu/drm/xe/xe_mmio_gem.c +++ b/drivers/gpu/drm/xe/xe_mmio_gem.c @@ -38,6 +38,7 @@ struct xe_mmio_gem { struct drm_gem_object base; phys_addr_t phys_addr; struct page *dummy_page; /* protected by the GEM's dma_resv */ + bool destroyed; /* protected by the GEM's dma_resv */ }; static int xe_mmio_gem_vm_may_split(struct vm_area_struct *area, unsigned long addr) @@ -150,8 +151,22 @@ static void xe_mmio_gem_free(struct drm_gem_object *base) */ void xe_mmio_gem_destroy(struct xe_mmio_gem *gem, struct drm_file *file) { - drm_vma_node_revoke(&gem->base.vma_node, file); - xe_mmio_gem_free(&gem->base); + struct drm_gem_object *base = &gem->base; + struct drm_device *dev = base->dev; + + drm_vma_node_revoke(&base->vma_node, file); + + dma_resv_lock(base->resv, NULL); + gem->destroyed = true; + dma_resv_unlock(base->resv); + /* + * Setting 'destroyed' under lock takes care of the subsequent faults. + * Zap the existing PTEs to cut off access to the real MMIO through + * currently mapped pages. + */ + drm_vma_node_unmap(&base->vma_node, dev->anon_inode->i_mapping); + + drm_gem_object_put(base); } static int xe_mmio_gem_mmap(struct drm_gem_object *base, struct vm_area_struct *vma) @@ -162,8 +177,6 @@ static int xe_mmio_gem_mmap(struct drm_gem_object *base, struct vm_area_struct * if ((vma->vm_flags & VM_SHARED) == 0) return -EINVAL; - /* Set vm_pgoff (used as a fake buffer offset by DRM) to 0 */ - vma->vm_pgoff = 0; vma->vm_page_prot = pgprot_noncached(vma_get_page_prot(vma)); vm_flags_set(vma, VM_IO | VM_PFNMAP | VM_DONTEXPAND | VM_DONTDUMP | VM_DONTCOPY | VM_NORESERVE); @@ -176,10 +189,9 @@ static int alloc_dummy_page_if_needed(struct drm_gem_object *base) { struct xe_mmio_gem *obj = to_xe_mmio_gem(base); - dma_resv_lock(base->resv, NULL); + dma_resv_assert_held(base->resv); if (!obj->dummy_page) obj->dummy_page = alloc_page(GFP_KERNEL | __GFP_ZERO); - dma_resv_unlock(base->resv); return obj->dummy_page ? 0 : -ENOMEM; } @@ -200,7 +212,7 @@ static vm_fault_t xe_mmio_gem_vm_fault_dummy_page(struct vm_fault *vmf) vm_get_page_prot(vma->vm_flags)); } -static vm_fault_t xe_mmio_gem_vm_fault(struct vm_fault *vmf) +static vm_fault_t xe_mmio_gem_vm_fault_locked(struct vm_fault *vmf) { struct vm_area_struct *vma = vmf->vma; struct drm_gem_object *base = vma->vm_private_data; @@ -210,6 +222,10 @@ static vm_fault_t xe_mmio_gem_vm_fault(struct vm_fault *vmf) unsigned long addr, pfn; int idx; + dma_resv_assert_held(base->resv); + if (obj->destroyed) + return VM_FAULT_SIGBUS; + if (!drm_dev_enter(dev, &idx)) { /* * Provide a dummy page to avoid SIGBUS for events such as hot-unplug. @@ -232,3 +248,15 @@ static vm_fault_t xe_mmio_gem_vm_fault(struct vm_fault *vmf) drm_dev_exit(idx); return ret; } + +static vm_fault_t xe_mmio_gem_vm_fault(struct vm_fault *vmf) +{ + struct vm_area_struct *vma = vmf->vma; + struct drm_gem_object *base = vma->vm_private_data; + vm_fault_t ret; + + dma_resv_lock(base->resv, NULL); + ret = xe_mmio_gem_vm_fault_locked(vmf); + dma_resv_unlock(base->resv); + return ret; +} From 3c90e42a01426262f0cd166bc01b45c05562640d Mon Sep 17 00:00:00 2001 From: Shuicheng Lin Date: Wed, 9 Sep 2026 16:21:01 +0000 Subject: [PATCH 196/592] drm/xe/shrinker: Return the freed page count through a parameter MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit __xe_shrinker_walk() and xe_shrinker_walk() return either the number of pages freed or a negative error, so the two cannot be reported at once. On error the pages already freed are dropped, and since xe_shrinker_scan() only accumulates non-negative returns while *scanned is updated by pointer, the shrinker tells mm that it scanned without freeing. Accumulate the count into a caller-provided counter and return only the status, so an error no longer discards what the walk had freed. Fixes: 00c8efc3180f ("drm/xe: Add a shrinker for xe bos") Assisted-by: Claude:claude-opus-5 Reviewed-by: Thomas Hellström Cc: Matthew Brost Link: https://patch.msgid.link/20260909162102.1097006-2-shuicheng.lin@intel.com Signed-off-by: Shuicheng Lin (cherry picked from commit d7aac1a0235a6ce41e30cec385e2db8c33dad12d) Signed-off-by: Rodrigo Vivi --- drivers/gpu/drm/xe/xe_shrinker.c | 62 ++++++++++++++------------------ 1 file changed, 26 insertions(+), 36 deletions(-) diff --git a/drivers/gpu/drm/xe/xe_shrinker.c b/drivers/gpu/drm/xe/xe_shrinker.c index 83374cd57660..89445cd20238 100644 --- a/drivers/gpu/drm/xe/xe_shrinker.c +++ b/drivers/gpu/drm/xe/xe_shrinker.c @@ -54,13 +54,14 @@ xe_shrinker_mod_pages(struct xe_shrinker *shrinker, long shrinkable, long purgea write_unlock(&shrinker->lock); } -static s64 __xe_shrinker_walk(struct xe_device *xe, +static int __xe_shrinker_walk(struct xe_device *xe, struct ttm_operation_ctx *ctx, const struct xe_bo_shrink_flags flags, - unsigned long to_scan, unsigned long *scanned) + unsigned long to_scan, unsigned long *scanned, + unsigned long *freed) { unsigned int mem_type; - s64 freed = 0, lret; + s64 lret; for (mem_type = XE_PL_SYSTEM; mem_type <= XE_PL_TT; ++mem_type) { struct ttm_resource_manager *man = ttm_manager_type(&xe->ttm, mem_type); @@ -82,7 +83,7 @@ static s64 __xe_shrinker_walk(struct xe_device *xe, if (lret < 0) return lret; - freed += lret; + *freed += lret; if (*scanned >= to_scan) break; } @@ -90,7 +91,7 @@ static s64 __xe_shrinker_walk(struct xe_device *xe, xe_assert(xe, !IS_ERR(ttm_bo)); } - return freed; + return 0; } /* @@ -99,40 +100,35 @@ static s64 __xe_shrinker_walk(struct xe_device *xe, * add writeback. This avoids stalls and explicit writebacks with light or * moderate memory pressure. */ -static s64 xe_shrinker_walk(struct xe_device *xe, +static int xe_shrinker_walk(struct xe_device *xe, struct ttm_operation_ctx *ctx, const struct xe_bo_shrink_flags flags, - unsigned long to_scan, unsigned long *scanned) + unsigned long to_scan, unsigned long *scanned, + unsigned long *freed) { bool no_wait_gpu = true; struct xe_bo_shrink_flags save_flags = flags; - s64 lret, freed; + int ret; swap(no_wait_gpu, ctx->no_wait_gpu); save_flags.writeback = false; - lret = __xe_shrinker_walk(xe, ctx, save_flags, to_scan, scanned); + ret = __xe_shrinker_walk(xe, ctx, save_flags, to_scan, scanned, freed); swap(no_wait_gpu, ctx->no_wait_gpu); - if (lret < 0 || *scanned >= to_scan) - return lret; + if (ret || *scanned >= to_scan) + return ret; - freed = lret; if (!ctx->no_wait_gpu) { - lret = __xe_shrinker_walk(xe, ctx, save_flags, to_scan, scanned); - if (lret < 0) - return lret; - freed += lret; - if (*scanned >= to_scan) - return freed; + ret = __xe_shrinker_walk(xe, ctx, save_flags, to_scan, scanned, + freed); + if (ret || *scanned >= to_scan) + return ret; } - if (flags.writeback) { - lret = __xe_shrinker_walk(xe, ctx, flags, to_scan, scanned); - if (lret < 0) - return lret; - freed += lret; - } + if (flags.writeback) + ret = __xe_shrinker_walk(xe, ctx, flags, to_scan, scanned, + freed); - return freed; + return ret; } static unsigned long @@ -214,7 +210,6 @@ static unsigned long xe_shrinker_scan(struct shrinker *shrink, struct shrink_con bool runtime_pm; bool purgeable; bool can_backup = !!(sc->gfp_mask & __GFP_FS); - s64 lret; nr_to_scan = sc->nr_to_scan; @@ -225,12 +220,9 @@ static unsigned long xe_shrinker_scan(struct shrinker *shrink, struct shrink_con /* Might need runtime PM. Try to wake early if it looks like it. */ runtime_pm = xe_shrinker_runtime_pm_get(shrinker, false, nr_to_scan, can_backup); - if (purgeable && nr_scanned < nr_to_scan) { - lret = xe_shrinker_walk(shrinker->xe, &ctx, shrink_flags, - nr_to_scan, &nr_scanned); - if (lret >= 0) - freed += lret; - } + if (purgeable && nr_scanned < nr_to_scan) + xe_shrinker_walk(shrinker->xe, &ctx, shrink_flags, + nr_to_scan, &nr_scanned, &freed); sc->nr_scanned = nr_scanned; if (nr_scanned >= nr_to_scan || !can_backup) @@ -242,10 +234,8 @@ static unsigned long xe_shrinker_scan(struct shrinker *shrink, struct shrink_con shrink_flags.purge = false; - lret = xe_shrinker_walk(shrinker->xe, &ctx, shrink_flags, - nr_to_scan, &nr_scanned); - if (lret >= 0) - freed += lret; + xe_shrinker_walk(shrinker->xe, &ctx, shrink_flags, + nr_to_scan, &nr_scanned, &freed); sc->nr_scanned = nr_scanned; out: From 985862be16c7e4da808c51f393d631fb60c0be5c Mon Sep 17 00:00:00 2001 From: Shuicheng Lin Date: Wed, 9 Sep 2026 16:21:02 +0000 Subject: [PATCH 197/592] drm/xe/shrinker: Take a runtime PM ref before shrinking non-system memory MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit __xe_shrinker_walk() walks the SYSTEM and TT LRUs without a runtime PM reference. Shrinking a bo outside system memory invalidates its GPU mappings, which needs the device resumed, so while it is runtime suspended the page table zap trips an assert and the TLB invalidation returns -ENODEV: WARNING: drivers/gpu/drm/xe/xe_bo.c:770 at xe_bo_move_notify+0x1fc/0x450 [xe] xe_bo_shrink+0x20f/0x2b0 [xe] __xe_shrinker_walk+0x174/0x410 [xe] xe_shrinker_scan+0x10c/0x1e0 [xe] do_shrink_slab+0x176/0x7e0 drop_caches_sysctl_handler+0x9c/0xf0 Take a reference before walking a memory type other than XE_PL_SYSTEM and stop there if it cannot be acquired. Reuse the shrinker's existing acquire path, which resumes the device directly where reclaim allows that and otherwise queues the PM worker for a later scan. Stop the walk once the scan target is met, so a satisfied scan does not wake the device. System memory is still reclaimed while the device is suspended. Gate this on xe_device_is_l2_flush_optimized(), the same condition under which xe_bo_trigger_rebind() issues the invalidation for a non-fault-mode vm, so reclaim is unaffected elsewhere. The System CCS copy already has its own reference in xe_bo_shrink(). Only a non-fault-mode vm can reach this, since a fault-mode vm requires LR mode and that holds a runtime PM reference for the vm's lifetime. Reproduced with igt@xe_madvise@dontneed-before-exec while the GPU is runtime suspended. v2: simplify needs_rpm check. (Matt) retarget Fixes tag since the issue occurs with the non-fault-mode path added by 4e7ebff69aed. v3: handle this in xe_shrinker.c instead of xe_bo.c (Thomas) v4: stop the walk once the scan target is met. (Sashiko) v5: rebase on the freed page accounting fix. (Sashiko) v6: reuse the shrinker acquire path so runtime pm can be resumed directly instead of always queueing a worker. (Thomas) v7: replace xe_pm_runtime_put() with xe_shrinker_runtime_pm_put(). (Thomas) Fixes: 4e7ebff69aed ("drm/xe/xe3p_lpg: flush shrinker bo cachelines manually") Assisted-by: Claude:claude-opus-5 Cc: Tejas Upadhyay Cc: Matthew Brost Reviewed-by: Thomas Hellström Link: https://patch.msgid.link/20260909162102.1097006-3-shuicheng.lin@intel.com Signed-off-by: Shuicheng Lin (cherry picked from commit 628f92b28bf4c371c10207daf6fc4caee0c0db2e) Signed-off-by: Rodrigo Vivi --- drivers/gpu/drm/xe/xe_shrinker.c | 78 +++++++++++++++++++++----------- 1 file changed, 51 insertions(+), 27 deletions(-) diff --git a/drivers/gpu/drm/xe/xe_shrinker.c b/drivers/gpu/drm/xe/xe_shrinker.c index 89445cd20238..deb4378c1ec1 100644 --- a/drivers/gpu/drm/xe/xe_shrinker.c +++ b/drivers/gpu/drm/xe/xe_shrinker.c @@ -54,13 +54,39 @@ xe_shrinker_mod_pages(struct xe_shrinker *shrinker, long shrinkable, long purgea write_unlock(&shrinker->lock); } -static int __xe_shrinker_walk(struct xe_device *xe, +static bool __xe_shrinker_runtime_pm_get(struct xe_shrinker *shrinker) +{ + struct xe_device *xe = shrinker->xe; + + if (xe_pm_runtime_get_if_active(xe)) + return true; + + if (xe_rpm_reclaim_safe(xe) && !ttm_bo_shrink_avoid_wait()) { + xe_pm_runtime_get(xe); + return true; + } + + queue_work(xe->unordered_wq, &shrinker->pm_worker); + + return false; +} + +static void xe_shrinker_runtime_pm_put(struct xe_shrinker *shrinker, bool runtime_pm) +{ + if (runtime_pm) + xe_pm_runtime_put(shrinker->xe); +} + +static int __xe_shrinker_walk(struct xe_shrinker *shrinker, struct ttm_operation_ctx *ctx, const struct xe_bo_shrink_flags flags, unsigned long to_scan, unsigned long *scanned, unsigned long *freed) { + struct xe_device *xe = shrinker->xe; unsigned int mem_type; + bool rpm = false; + int ret = 0; s64 lret; for (mem_type = XE_PL_SYSTEM; mem_type <= XE_PL_TT; ++mem_type) { @@ -75,23 +101,35 @@ static int __xe_shrinker_walk(struct xe_device *xe, if (!man || !man->use_tt) continue; + if (mem_type != XE_PL_SYSTEM && !rpm && + xe_device_is_l2_flush_optimized(xe)) { + if (!__xe_shrinker_runtime_pm_get(shrinker)) + break; + rpm = true; + } + ttm_bo_lru_for_each_reserved_guarded(&curs, man, &arg, ttm_bo) { if (!ttm_bo_shrink_suitable(ttm_bo, ctx)) continue; lret = xe_bo_shrink(ctx, ttm_bo, flags, scanned); - if (lret < 0) - return lret; + if (lret < 0) { + ret = lret; + goto out; + } *freed += lret; if (*scanned >= to_scan) - break; + goto out; } /* Trylocks should never error, just fail. */ xe_assert(xe, !IS_ERR(ttm_bo)); } - return 0; +out: + xe_shrinker_runtime_pm_put(shrinker, rpm); + + return ret; } /* @@ -100,7 +138,7 @@ static int __xe_shrinker_walk(struct xe_device *xe, * add writeback. This avoids stalls and explicit writebacks with light or * moderate memory pressure. */ -static int xe_shrinker_walk(struct xe_device *xe, +static int xe_shrinker_walk(struct xe_shrinker *shrinker, struct ttm_operation_ctx *ctx, const struct xe_bo_shrink_flags flags, unsigned long to_scan, unsigned long *scanned, @@ -112,20 +150,21 @@ static int xe_shrinker_walk(struct xe_device *xe, swap(no_wait_gpu, ctx->no_wait_gpu); save_flags.writeback = false; - ret = __xe_shrinker_walk(xe, ctx, save_flags, to_scan, scanned, freed); + ret = __xe_shrinker_walk(shrinker, ctx, save_flags, to_scan, scanned, + freed); swap(no_wait_gpu, ctx->no_wait_gpu); if (ret || *scanned >= to_scan) return ret; if (!ctx->no_wait_gpu) { - ret = __xe_shrinker_walk(xe, ctx, save_flags, to_scan, scanned, + ret = __xe_shrinker_walk(shrinker, ctx, save_flags, to_scan, scanned, freed); if (ret || *scanned >= to_scan) return ret; } if (flags.writeback) - ret = __xe_shrinker_walk(xe, ctx, flags, to_scan, scanned, + ret = __xe_shrinker_walk(shrinker, ctx, flags, to_scan, scanned, freed); return ret; @@ -176,22 +215,7 @@ static bool xe_shrinker_runtime_pm_get(struct xe_shrinker *shrinker, bool force, return false; } - if (!xe_pm_runtime_get_if_active(xe)) { - if (xe_rpm_reclaim_safe(xe) && !ttm_bo_shrink_avoid_wait()) { - xe_pm_runtime_get(xe); - return true; - } - queue_work(xe->unordered_wq, &shrinker->pm_worker); - return false; - } - - return true; -} - -static void xe_shrinker_runtime_pm_put(struct xe_shrinker *shrinker, bool runtime_pm) -{ - if (runtime_pm) - xe_pm_runtime_put(shrinker->xe); + return __xe_shrinker_runtime_pm_get(shrinker); } static unsigned long xe_shrinker_scan(struct shrinker *shrink, struct shrink_control *sc) @@ -221,7 +245,7 @@ static unsigned long xe_shrinker_scan(struct shrinker *shrink, struct shrink_con runtime_pm = xe_shrinker_runtime_pm_get(shrinker, false, nr_to_scan, can_backup); if (purgeable && nr_scanned < nr_to_scan) - xe_shrinker_walk(shrinker->xe, &ctx, shrink_flags, + xe_shrinker_walk(shrinker, &ctx, shrink_flags, nr_to_scan, &nr_scanned, &freed); sc->nr_scanned = nr_scanned; @@ -234,7 +258,7 @@ static unsigned long xe_shrinker_scan(struct shrinker *shrink, struct shrink_con shrink_flags.purge = false; - xe_shrinker_walk(shrinker->xe, &ctx, shrink_flags, + xe_shrinker_walk(shrinker, &ctx, shrink_flags, nr_to_scan, &nr_scanned, &freed); sc->nr_scanned = nr_scanned; From 073a30d75f309812ed61af134f24ffef4107b13a Mon Sep 17 00:00:00 2001 From: Karl Mehltretter Date: Sat, 22 Aug 2026 16:31:10 +0200 Subject: [PATCH 198/592] drm/vc4: Use managed KMS polling to fix UAF on unbind MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit vc4_kms_load() calls drm_kms_helper_poll_init() but the driver provides no matching drm_kms_helper_poll_fini(). The output poll work stays scheduled after unbind and runs on the freed drm_device: # modprobe vc4; rmmod vc4; sleep 10 BUG: KASAN: slab-use-after-free in delayed_work_timer_fn BUG: KASAN: slab-use-after-free in drm_client_dev_hotplug [drm] Workqueue: events output_poll_execute [drm_kms_helper] Allocated by task 171: __devm_drm_dev_alloc Freed by task 262 (rmmod): drm_dev_put / component_del Use drmm_kms_helper_poll_init() so polling is finalized with the device, as other drivers do. Fixes: c8b75bca92cb ("drm/vc4: Add KMS support for Raspberry Pi.") Assisted-by: Claude:claude-fable-5 Signed-off-by: Karl Mehltretter Link: https://patch.msgid.link/20260822143110.68594-1-kmehltretter@gmail.com Reviewed-by: Maíra Canal Signed-off-by: Maíra Canal --- drivers/gpu/drm/vc4/vc4_kms.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/drivers/gpu/drm/vc4/vc4_kms.c b/drivers/gpu/drm/vc4/vc4_kms.c index b17e73bce384..82a7f2ac4c7e 100644 --- a/drivers/gpu/drm/vc4/vc4_kms.c +++ b/drivers/gpu/drm/vc4/vc4_kms.c @@ -1188,7 +1188,7 @@ int vc4_kms_load(struct drm_device *dev) drm_mode_config_reset(dev); - drm_kms_helper_poll_init(dev); + drmm_kms_helper_poll_init(dev); return 0; } From bdf5f731957de48acada392f28e82bc019713adb Mon Sep 17 00:00:00 2001 From: Cong Nguyen Date: Mon, 14 Sep 2026 17:41:36 +0700 Subject: [PATCH 199/592] hwmon: (gpio-fan) return IRQ_HANDLED from the shared alarm IRQ handler fan_alarm_irq_handler() always schedules alarm_work but returns IRQ_NONE, so the kernel treats every alarm interrupt as unhandled. On a shared line that risks the whole line being disabled as spurious. v1 just fixed that, but it was still IRQF_SHARED, and always returning IRQ_HANDLED there defeats spurious-interrupt detection for the line -- if the interrupt ever fires without a real event, nothing catches it, and a fault could spin the CPU in the handler. Sashiko flagged this in v1, and Guenter confirmed: this interrupt must not be shared. So v2 drops IRQF_SHARED too. Fixes: d6fe1360f42e ("hwmon: add generic GPIO fan driver") Reported-by: Sashiko AI review Link: https://lore.kernel.org/r/20260901160931.DD3811F00A3D@smtp.kernel.org Assisted-by: Claude:claude-opus-4 Signed-off-by: Cong Nguyen Link: https://patch.msgid.link/20260914104136.1797979-1-congnt264@gmail.com Signed-off-by: Guenter Roeck --- drivers/hwmon/gpio-fan.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/drivers/hwmon/gpio-fan.c b/drivers/hwmon/gpio-fan.c index df8bd9707605..3f78375eeb44 100644 --- a/drivers/hwmon/gpio-fan.c +++ b/drivers/hwmon/gpio-fan.c @@ -68,7 +68,7 @@ static irqreturn_t fan_alarm_irq_handler(int irq, void *dev_id) schedule_work(&fan_data->alarm_work); - return IRQ_NONE; + return IRQ_HANDLED; } static ssize_t fan1_alarm_show(struct device *dev, @@ -103,7 +103,7 @@ static int fan_alarm_init(struct gpio_fan_data *fan_data) irq_set_irq_type(alarm_irq, IRQ_TYPE_EDGE_BOTH); return devm_request_irq(dev, alarm_irq, fan_alarm_irq_handler, - IRQF_SHARED, "GPIO fan alarm", fan_data); + 0, "GPIO fan alarm", fan_data); } /* From 0ff9c7775e51ac6d47b1bb5c46f06b1434fe58a8 Mon Sep 17 00:00:00 2001 From: Guangshuo Li Date: Mon, 14 Sep 2026 14:28:09 +0800 Subject: [PATCH 200/592] hwmon: (w83791d) remove fan/pwm 4-5 sysfs group on remove When the fan/pwm 4-5 pins are not used as GPIO, w83791d_probe() creates the w83791d_group_fanpwm45 sysfs group on the I2C client device. The probe error path removes this group when a later initialization step fails, but the normal remove path only removes w83791d_group. As a result, the optional fan/pwm 4-5 sysfs files can remain after the driver is unbound. The callbacks associated with these files access the driver data, which is devm allocated and released after driver unbind. Leaving the sysfs files behind can therefore result in accesses to stale driver data. Remove w83791d_group_fanpwm45 during normal teardown as well. This issue was found by manual code inspection. Fixes: 6e1ecd9b8f13 ("hwmon: (w83791d) fan 4/5 pins can also be used for gpio") Cc: stable@vger.kernel.org Signed-off-by: Guangshuo Li Link: https://patch.msgid.link/20260914062809.1650538-1-lgs201920130244@gmail.com Signed-off-by: Guenter Roeck --- drivers/hwmon/w83791d.c | 1 + 1 file changed, 1 insertion(+) diff --git a/drivers/hwmon/w83791d.c b/drivers/hwmon/w83791d.c index 4a777430af5c..4b07a25ae59e 100644 --- a/drivers/hwmon/w83791d.c +++ b/drivers/hwmon/w83791d.c @@ -1415,6 +1415,7 @@ static void w83791d_remove(struct i2c_client *client) struct w83791d_data *data = i2c_get_clientdata(client); hwmon_device_unregister(data->hwmon_dev); + sysfs_remove_group(&client->dev.kobj, &w83791d_group_fanpwm45); sysfs_remove_group(&client->dev.kobj, &w83791d_group); } From c702a5f18b780e477eccbbab558e590e9673e4cb Mon Sep 17 00:00:00 2001 From: Guangshuo Li Date: Mon, 14 Sep 2026 15:36:38 +0800 Subject: [PATCH 201/592] hwmon: (w83793) release probe data through kref w83793_probe() initializes data->kref to manage the lifetime of the driver data. The normal remove path drops the driver-owned reference with kref_put(), while watchdog users take and release additional references through the same kref. However, the probe error path still frees data directly with kfree(). This bypasses the kref-managed lifetime and discards the initial reference without a matching kref_put(), leaving the reference accounting unbalanced. Drop the probe-owned reference with kref_put() instead and let w83793_release_resources() perform the final free, matching the normal remove path. This issue was found by manual code inspection. Fixes: 5852f9609d21 ("hwmon: (w83793) Add watchdog functionality") Cc: stable@vger.kernel.org Signed-off-by: Guangshuo Li Link: https://patch.msgid.link/20260914073638.1662500-1-lgs201920130244@gmail.com Signed-off-by: Guenter Roeck --- drivers/hwmon/w83793.c | 4 +++- 1 file changed, 3 insertions(+), 1 deletion(-) diff --git a/drivers/hwmon/w83793.c b/drivers/hwmon/w83793.c index a548586369e1..c6ef04c69856 100644 --- a/drivers/hwmon/w83793.c +++ b/drivers/hwmon/w83793.c @@ -1928,7 +1928,9 @@ static int w83793_probe(struct i2c_client *client) for (i = 0; i < ARRAY_SIZE(w83793_temp); i++) device_remove_file(dev, &w83793_temp[i].dev_attr); free_mem: - kfree(data); + mutex_lock(&watchdog_data_mutex); + kref_put(&data->kref, w83793_release_resources); + mutex_unlock(&watchdog_data_mutex); exit: return err; } From 9d1e523b92d8cb11a4a7e5a2655d6824c2d0f6e3 Mon Sep 17 00:00:00 2001 From: Benjamin Tissoires Date: Fri, 4 Sep 2026 14:52:59 +0200 Subject: [PATCH 202/592] selftests/hid: add define for commonly used buf size If we want to add another report descriptor without report IDs with a report size bigger than 10, we have multiple magic values to replace. Put a #define once and for all, so we don't have dangling ones. Signed-off-by: Benjamin Tissoires --- tools/testing/selftests/hid/hid_bpf.c | 38 ++++++++++++------------ tools/testing/selftests/hid/hid_common.h | 3 +- 2 files changed, 21 insertions(+), 20 deletions(-) diff --git a/tools/testing/selftests/hid/hid_bpf.c b/tools/testing/selftests/hid/hid_bpf.c index 7ab86296ff23..0d03ad5245fc 100644 --- a/tools/testing/selftests/hid/hid_bpf.c +++ b/tools/testing/selftests/hid/hid_bpf.c @@ -5,7 +5,7 @@ #include struct hid_hw_request_syscall_args { - __u8 data[10]; + __u8 data[MAX_BUF_SIZE]; unsigned int hid; int retval; size_t size; @@ -175,7 +175,7 @@ TEST_F(hid_bpf, raw_event) const struct test_program progs[] = { { .name = "hid_first_event" }, }; - __u8 buf[10] = {0}; + __u8 buf[MAX_BUF_SIZE] = {0}; int err; LOAD_PROGRAMS(progs); @@ -226,7 +226,7 @@ TEST_F(hid_bpf, subprog_raw_event) const struct test_program progs[] = { { .name = "hid_subprog_first_event" }, }; - __u8 buf[10] = {0}; + __u8 buf[MAX_BUF_SIZE] = {0}; int err; LOAD_PROGRAMS(progs); @@ -284,7 +284,7 @@ TEST_F(hid_bpf, test_attach_detach) { .name = "hid_second_event" }, }; struct bpf_link *link; - __u8 buf[10] = {0}; + __u8 buf[MAX_BUF_SIZE] = {0}; int err, link_fd; LOAD_PROGRAMS(progs); @@ -369,7 +369,7 @@ TEST_F(hid_bpf, test_hid_change_report) const struct test_program progs[] = { { .name = "hid_change_report_id" }, }; - __u8 buf[10] = {0}; + __u8 buf[MAX_BUF_SIZE] = {0}; int err; LOAD_PROGRAMS(progs); @@ -396,13 +396,13 @@ TEST_F(hid_bpf, test_hid_user_input_report_call) { struct hid_hw_request_syscall_args args = { .retval = -1, - .size = 10, + .size = MAX_BUF_SIZE, }; DECLARE_LIBBPF_OPTS(bpf_test_run_opts, tattrs, .ctx_in = &args, .ctx_size_in = sizeof(args), ); - __u8 buf[10] = {0}; + __u8 buf[MAX_BUF_SIZE] = {0}; int err, prog_fd; LOAD_BPF; @@ -442,7 +442,7 @@ TEST_F(hid_bpf, test_hid_user_output_report_call) { struct hid_hw_request_syscall_args args = { .retval = -1, - .size = 10, + .size = MAX_BUF_SIZE, }; DECLARE_LIBBPF_OPTS(bpf_test_run_opts, tattrs, .ctx_in = &args, @@ -491,7 +491,7 @@ TEST_F(hid_bpf, test_hid_user_raw_request_call) .retval = -1, .type = HID_FEATURE_REPORT, .request_type = HID_REQ_GET_REPORT, - .size = 10, + .size = MAX_BUF_SIZE, }; DECLARE_LIBBPF_OPTS(bpf_test_run_opts, tattrs, .ctx_in = &args, @@ -524,7 +524,7 @@ TEST_F(hid_bpf, test_hid_filter_raw_request_call) const struct test_program progs[] = { { .name = "hid_test_filter_raw_request" }, }; - __u8 buf[10] = {0}; + __u8 buf[MAX_BUF_SIZE] = {0}; int err; LOAD_PROGRAMS(progs); @@ -577,7 +577,7 @@ TEST_F(hid_bpf, test_hid_change_raw_request_call) const struct test_program progs[] = { { .name = "hid_test_hidraw_raw_request" }, }; - __u8 buf[10] = {0}; + __u8 buf[MAX_BUF_SIZE] = {0}; int err; LOAD_PROGRAMS(progs); @@ -603,7 +603,7 @@ TEST_F(hid_bpf, test_hid_infinite_loop_raw_request_call) const struct test_program progs[] = { { .name = "hid_test_infinite_loop_raw_request" }, }; - __u8 buf[10] = {0}; + __u8 buf[MAX_BUF_SIZE] = {0}; int err; LOAD_PROGRAMS(progs); @@ -626,7 +626,7 @@ TEST_F(hid_bpf, test_hid_filter_output_report_call) const struct test_program progs[] = { { .name = "hid_test_filter_output_report" }, }; - __u8 buf[10] = {0}; + __u8 buf[MAX_BUF_SIZE] = {0}; int err; LOAD_PROGRAMS(progs); @@ -679,7 +679,7 @@ TEST_F(hid_bpf, test_hid_change_output_report_call) const struct test_program progs[] = { { .name = "hid_test_hidraw_output_report" }, }; - __u8 buf[10] = {0}; + __u8 buf[MAX_BUF_SIZE] = {0}; int err; LOAD_PROGRAMS(progs); @@ -703,7 +703,7 @@ TEST_F(hid_bpf, test_hid_infinite_loop_output_report_call) const struct test_program progs[] = { { .name = "hid_test_infinite_loop_output_report" }, }; - __u8 buf[10] = {0}; + __u8 buf[MAX_BUF_SIZE] = {0}; int err; LOAD_PROGRAMS(progs); @@ -729,7 +729,7 @@ TEST_F(hid_bpf, test_multiply_events_wq) const struct test_program progs[] = { { .name = "hid_test_multiply_events_wq" }, }; - __u8 buf[10] = {0}; + __u8 buf[MAX_BUF_SIZE] = {0}; int err; LOAD_PROGRAMS(progs); @@ -767,7 +767,7 @@ TEST_F(hid_bpf, test_multiply_events) const struct test_program progs[] = { { .name = "hid_test_multiply_events" }, }; - __u8 buf[10] = {0}; + __u8 buf[MAX_BUF_SIZE] = {0}; int err; LOAD_PROGRAMS(progs); @@ -801,7 +801,7 @@ TEST_F(hid_bpf, test_hid_infinite_loop_input_report_call) const struct test_program progs[] = { { .name = "hid_test_infinite_loop_input_report" }, }; - __u8 buf[10] = {0}; + __u8 buf[MAX_BUF_SIZE] = {0}; int err; LOAD_PROGRAMS(progs); @@ -855,7 +855,7 @@ TEST_F(hid_bpf, test_hid_attach_flags) .insert_head = 0, }, }; - __u8 buf[10] = {0}; + __u8 buf[MAX_BUF_SIZE] = {0}; int err; LOAD_PROGRAMS(progs); diff --git a/tools/testing/selftests/hid/hid_common.h b/tools/testing/selftests/hid/hid_common.h index e3b267446fa0..4567336f131d 100644 --- a/tools/testing/selftests/hid/hid_common.h +++ b/tools/testing/selftests/hid/hid_common.h @@ -13,6 +13,7 @@ #include #define SHOW_UHID_DEBUG 0 +#define MAX_BUF_SIZE 10 #define min(a, b) \ ({ __typeof__(a) _a = (a); \ @@ -110,7 +111,7 @@ static pthread_cond_t uhid_started = PTHREAD_COND_INITIALIZER; static pthread_mutex_t uhid_output_mtx = PTHREAD_MUTEX_INITIALIZER; static pthread_cond_t uhid_output_cond = PTHREAD_COND_INITIALIZER; -static unsigned char output_report[10]; +static unsigned char output_report[MAX_BUF_SIZE]; /* no need to protect uhid_stopped, only one thread accesses it */ static bool uhid_stopped; From c4afa4862b878d56e0cc1021298794ac1b45bc49 Mon Sep 17 00:00:00 2001 From: Benjamin Tissoires Date: Fri, 4 Sep 2026 14:53:00 +0200 Subject: [PATCH 203/592] HID: bpf: fix __hid_bpf_hw_check_params report length Turns out that USB, I2C and other transport drivers (except uhid which just passes the data) still need to have the report ID in the first byte. Because they expect the first byte to be the report ID or 0, when the report ID is 0, they strip that first byte before forwarding to the device. This means that the transport layer forwards a buffer of size N-1 to the device, which gets rejected. Fixes: 5599f8019661 ("HID: bpf: export hid_hw_output_report as a BPF kfunc") Signed-off-by: Benjamin Tissoires --- drivers/hid/bpf/hid_bpf_dispatch.c | 12 ++++++++---- 1 file changed, 8 insertions(+), 4 deletions(-) diff --git a/drivers/hid/bpf/hid_bpf_dispatch.c b/drivers/hid/bpf/hid_bpf_dispatch.c index 536f6d01fd14..b1de1dd0f21d 100644 --- a/drivers/hid/bpf/hid_bpf_dispatch.c +++ b/drivers/hid/bpf/hid_bpf_dispatch.c @@ -359,7 +359,7 @@ hid_bpf_release_context(struct hid_bpf_ctx *ctx) static int __hid_bpf_hw_check_params(struct hid_bpf_ctx *ctx, __u8 *buf, size_t *buf__sz, - enum hid_report_type rtype) + enum hid_report_type rtype, bool hw_request) { struct hid_report_enum *report_enum; struct hid_report *report; @@ -388,6 +388,10 @@ __hid_bpf_hw_check_params(struct hid_bpf_ctx *ctx, __u8 *buf, size_t *buf__sz, report_len = hid_report_len(report); + /* unnumbered reports need to have a report ID reserved in the first byte */ + if (hw_request && report_enum->numbered == 0) + report_len += 1; + if (*buf__sz > report_len) *buf__sz = report_len; @@ -420,7 +424,7 @@ hid_bpf_hw_request(struct hid_bpf_ctx *ctx, __u8 *buf, size_t buf__sz, return -EDEADLOCK; /* check arguments */ - ret = __hid_bpf_hw_check_params(ctx, buf, &size, rtype); + ret = __hid_bpf_hw_check_params(ctx, buf, &size, rtype, true); if (ret) return ret; @@ -480,7 +484,7 @@ hid_bpf_hw_output_report(struct hid_bpf_ctx *ctx, __u8 *buf, size_t buf__sz) return -EDEADLOCK; /* check arguments */ - ret = __hid_bpf_hw_check_params(ctx, buf, &size, HID_OUTPUT_REPORT); + ret = __hid_bpf_hw_check_params(ctx, buf, &size, HID_OUTPUT_REPORT, true); if (ret) return ret; @@ -506,7 +510,7 @@ __hid_bpf_input_report(struct hid_bpf_ctx *ctx, enum hid_report_type type, u8 *b return -EDEADLOCK; /* check arguments */ - ret = __hid_bpf_hw_check_params(ctx, buf, &size, type); + ret = __hid_bpf_hw_check_params(ctx, buf, &size, type, false); if (ret) return ret; From b6f69097c8271cca30314fd6e4c802f90737bfcc Mon Sep 17 00:00:00 2001 From: Benjamin Tissoires Date: Fri, 4 Sep 2026 14:53:01 +0200 Subject: [PATCH 204/592] selftests/hid: add unnumbered variant to the hid_bpf tests A bug appeared in hid_bpf_dispatch.c where it wasn't properly handling unnumbered reports. Add a device variant without report IDs so we can also test them. Signed-off-by: Benjamin Tissoires --- tools/testing/selftests/hid/hid_bpf.c | 53 ++++++++++++++++++++---- tools/testing/selftests/hid/hid_common.h | 24 ++++++++++- tools/testing/selftests/hid/progs/hid.c | 2 +- 3 files changed, 68 insertions(+), 11 deletions(-) diff --git a/tools/testing/selftests/hid/hid_bpf.c b/tools/testing/selftests/hid/hid_bpf.c index 0d03ad5245fc..32d81ba15a25 100644 --- a/tools/testing/selftests/hid/hid_bpf.c +++ b/tools/testing/selftests/hid/hid_bpf.c @@ -54,11 +54,27 @@ FIXTURE_TEARDOWN(hid_bpf) { hid_bpf_teardown(_metadata, self, variant); \ } while (0) +FIXTURE_VARIANT(hid_bpf) { + __u8 *rdesc; + size_t rdesc_size; +}; + +FIXTURE_VARIANT_ADD(hid_bpf, numbered) { + .rdesc = rdesc, + .rdesc_size = sizeof(rdesc), +}; + +FIXTURE_VARIANT_ADD(hid_bpf, unnumbered) { + .rdesc = fido2_rdesc, + .rdesc_size = sizeof(fido2_rdesc), +}; + FIXTURE_SETUP(hid_bpf) { int err; - err = setup_uhid(_metadata, &self->hid, BUS_USB, 0x0001, 0x0a36, rdesc, sizeof(rdesc)); + err = setup_uhid(_metadata, &self->hid, BUS_USB, 0x0001, 0x0a36, + variant->rdesc, variant->rdesc_size); ASSERT_OK(err); } @@ -409,8 +425,11 @@ TEST_F(hid_bpf, test_hid_user_input_report_call) args.hid = self->hid.hid_id; args.data[0] = 1; /* report ID */ - args.data[1] = 2; /* report ID */ - args.data[2] = 42; /* report ID */ + args.data[1] = 2; + args.data[2] = 42; + + if (variant->rdesc == fido2_rdesc) + args.data[0] = 0; prog_fd = bpf_program__fd(self->skel->progs.hid_user_input_report); @@ -428,8 +447,13 @@ TEST_F(hid_bpf, test_hid_user_input_report_call) /* read the data from hidraw */ memset(buf, 0, sizeof(buf)); err = read(self->hidraw_fd, buf, sizeof(buf)); - ASSERT_EQ(err, 6) TH_LOG("read_hidraw"); - ASSERT_EQ(buf[0], 1); + if (variant->rdesc == rdesc) { + ASSERT_EQ(err, 6) TH_LOG("read_hidraw"); + } else { + ASSERT_EQ(err, 64) + TH_LOG("read_hidraw"); + } + ASSERT_EQ(buf[0], args.data[0]); ASSERT_EQ(buf[1], 2); ASSERT_EQ(buf[2], 42); } @@ -455,8 +479,11 @@ TEST_F(hid_bpf, test_hid_user_output_report_call) args.hid = self->hid.hid_id; args.data[0] = 1; /* report ID */ - args.data[1] = 2; /* report ID */ - args.data[2] = 42; /* report ID */ + args.data[1] = 2; + args.data[2] = 42; + + if (variant->rdesc == fido2_rdesc) + args.data[0] = 0; prog_fd = bpf_program__fd(self->skel->progs.hid_user_output_report); @@ -472,9 +499,14 @@ TEST_F(hid_bpf, test_hid_user_output_report_call) ASSERT_OK(err) TH_LOG("error while calling bpf_prog_test_run_opts"); ASSERT_OK(cond_err) TH_LOG("error while calling waiting for the condition"); - ASSERT_EQ(args.retval, 3); + if (variant->rdesc == rdesc) { + ASSERT_EQ(args.retval, 3); + } else if (variant->rdesc == fido2_rdesc) { + ASSERT_EQ(args.retval, 65) + TH_LOG("report size error, should have 64 + 1 extra byte for the report ID 0"); + } - ASSERT_EQ(output_report[0], 1); + ASSERT_EQ(output_report[0], args.data[0]); ASSERT_EQ(output_report[1], 2); ASSERT_EQ(output_report[2], 42); @@ -886,6 +918,9 @@ TEST_F(hid_bpf, test_rdesc_fixup) }; int err, desc_size; + if (variant->rdesc != rdesc) + SKIP(return, "not compatible report descriptor"); + LOAD_PROGRAMS(progs); /* check that hid_rdesc_fixup() was executed */ diff --git a/tools/testing/selftests/hid/hid_common.h b/tools/testing/selftests/hid/hid_common.h index 4567336f131d..b7890ba2878f 100644 --- a/tools/testing/selftests/hid/hid_common.h +++ b/tools/testing/selftests/hid/hid_common.h @@ -13,7 +13,7 @@ #include #define SHOW_UHID_DEBUG 0 -#define MAX_BUF_SIZE 10 +#define MAX_BUF_SIZE 128 #define min(a, b) \ ({ __typeof__(a) _a = (a); \ @@ -98,6 +98,28 @@ static unsigned char rdesc[] = { static __u8 feature_data[] = { 1, 2 }; +static __maybe_unused unsigned char fido2_rdesc[] = { + 0x06, 0xd0, 0xf1, /* Usage Page (FIDO Alliance) */ + 0x09, 0x01, /* Usage (U2F Authenticator Device) */ + 0xa1, 0x01, /* Collection (Application) */ + 0x09, 0x20, /* Usage (Input Report Data) */ + 0x15, 0x00, /* Logical Minimum (0) */ + 0x26, 0xff, 0x00, /* Logical Maximum (255) */ + 0x75, 0x08, /* Report Size (8) */ + 0x95, 0x40, /* Report Count (64) */ + 0x81, 0x02, /* Input (Data,Var,Abs) */ + 0x09, 0x21, /* Usage (Output Report Data) */ + 0x15, 0x00, /* Logical Minimum (0) */ + 0x26, 0xff, 0x00, /* Logical Maximum (255) */ + 0x75, 0x08, /* Report Size (8) */ + 0x95, 0x40, /* Report Count (64) */ + 0x91, 0x02, /* Output (Data,Var,Abs) */ + 0x06, 0x00, 0xff, /* Usage Page (Vendor Defined Page 1) */ + 0x09, 0x22, /* Usage (Vendor Usage 0x22) */ + 0xb1, 0x02, /* Feature (Data,Var,Abs) */ + 0xc0, /* End Collection */ +}; + #define ASSERT_OK(data) ASSERT_FALSE(data) #define ASSERT_OK_PTR(ptr) ASSERT_NE(NULL, ptr) diff --git a/tools/testing/selftests/hid/progs/hid.c b/tools/testing/selftests/hid/progs/hid.c index 361dc7eaad22..48aa8088cc53 100644 --- a/tools/testing/selftests/hid/progs/hid.c +++ b/tools/testing/selftests/hid/progs/hid.c @@ -98,7 +98,7 @@ struct hid_bpf_ops change_report_id = { struct hid_hw_request_syscall_args { /* data needs to come at offset 0 so we can use it in calls */ - __u8 data[10]; + __u8 data[128]; unsigned int hid; int retval; size_t size; From 3a35e787ac1acf94d33eebb26fd7248811cc66c9 Mon Sep 17 00:00:00 2001 From: Johannes Thumshirn Date: Mon, 24 Aug 2026 18:19:10 +0200 Subject: [PATCH 205/592] btrfs: zoned: handle RAID profiles in btrfs_can_activate_zone() btrfs_can_activate_zone() only accounts for the single and DUP profiles. For a RAID0, RAID1, RAID1C3, RAID1C4 or RAID10 block group the profile switch matches no case, so 'ret' stays false and the function reports that no zone can be activated, even when the devices have plenty of active zones left. As a side effect BTRFS_FS_NEED_ZONE_FINISH gets set and, since btrfs_can_activate_zone() bails out early once that bit is set, data allocations will fail permanently: writers loop on -EAGAIN and hang in btrfs_new_extent_direct() waiting for the bit to clear. Each of these profiles needs one active zone per device, just like single, so handle them the same way. Reviewed-by: Boris Burkov Signed-off-by: Johannes Thumshirn Signed-off-by: David Sterba --- fs/btrfs/zoned.c | 5 +++++ 1 file changed, 5 insertions(+) diff --git a/fs/btrfs/zoned.c b/fs/btrfs/zoned.c index 9cc2c9c1a606..08a15465a087 100644 --- a/fs/btrfs/zoned.c +++ b/fs/btrfs/zoned.c @@ -2688,6 +2688,11 @@ bool btrfs_can_activate_zone(struct btrfs_fs_devices *fs_devices, u64 flags) switch (flags & BTRFS_BLOCK_GROUP_PROFILE_MASK) { case 0: /* single */ + case BTRFS_BLOCK_GROUP_RAID0: + case BTRFS_BLOCK_GROUP_RAID1: + case BTRFS_BLOCK_GROUP_RAID1C3: + case BTRFS_BLOCK_GROUP_RAID1C4: + case BTRFS_BLOCK_GROUP_RAID10: ret = (atomic_read(&zinfo->active_zones_left) >= (1 + reserved)); break; case BTRFS_BLOCK_GROUP_DUP: From 0594e3423f4ba3137c734371169491f9a98e9af4 Mon Sep 17 00:00:00 2001 From: Hongling Zeng Date: Mon, 31 Aug 2026 13:38:01 +0800 Subject: [PATCH 206/592] btrfs: take commit root semaphore when iterating in mark_block_group_to_copy() mark_block_group_to_copy() iterates over the commit root with skip_locking=true. A concurrent transaction commit can swap and free the commit root during iteration, causing use-after-free when accessing extent buffers. Fix it by using path->need_commit_sem to protect the commit root search. Fixes: 78ce9fc269af ("btrfs: zoned: mark block groups to copy for device-replace") CC: stable@vger.kernel.org Assisted-by: Codex:gpt-5.5 Reviewed-by: Johannes Thumshirn Signed-off-by: Hongling Zeng Reviewed-by: David Sterba Signed-off-by: David Sterba --- fs/btrfs/dev-replace.c | 1 + 1 file changed, 1 insertion(+) diff --git a/fs/btrfs/dev-replace.c b/fs/btrfs/dev-replace.c index bf0b78790171..1894c7ebe9d9 100644 --- a/fs/btrfs/dev-replace.c +++ b/fs/btrfs/dev-replace.c @@ -493,6 +493,7 @@ static int mark_block_group_to_copy(struct btrfs_fs_info *fs_info, path->reada = READA_FORWARD; path->search_commit_root = true; path->skip_locking = true; + path->need_commit_sem = true; key.objectid = src_dev->devid; key.type = BTRFS_DEV_EXTENT_KEY; From a1167d9420474ab9ed9efca99d86aeb6217c0265 Mon Sep 17 00:00:00 2001 From: Filipe Manana Date: Thu, 10 Sep 2026 17:37:48 +0100 Subject: [PATCH 207/592] btrfs: tree-checker: print dev extent offset in error message If a dev extent's offset is not sector size aligned, the error message is printing the dev extent's objectid instead of the offset. This is a copy paste error, as before this check we check the objectid field. Fixes: 008e2512dc56 ("btrfs: tree-checker: add dev extent item checks") Reviewed-by: Qu Wenruo Signed-off-by: Filipe Manana Reviewed-by: David Sterba Signed-off-by: David Sterba --- fs/btrfs/tree-checker.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/fs/btrfs/tree-checker.c b/fs/btrfs/tree-checker.c index ab5abbb475e2..8b1b706e0996 100644 --- a/fs/btrfs/tree-checker.c +++ b/fs/btrfs/tree-checker.c @@ -2129,7 +2129,7 @@ static int check_dev_extent_item(const struct extent_buffer *leaf, sectorsize))) { generic_err(leaf, slot, "invalid dev extent chunk offset, has %llu not aligned to %u", - btrfs_dev_extent_chunk_objectid(leaf, de), + btrfs_dev_extent_chunk_offset(leaf, de), sectorsize); return -EUCLEAN; } From f59d86d25e41c5ac511584d9d6808ee448fd42b0 Mon Sep 17 00:00:00 2001 From: Filipe Manana Date: Thu, 10 Sep 2026 17:48:06 +0100 Subject: [PATCH 208/592] btrfs: tree-checker: fix error message regarding free space extent items The error message mentions a free space info item, but we are processing a free space extent item, so fix the message. Reviewed-by: Qu Wenruo Signed-off-by: Filipe Manana Reviewed-by: David Sterba Signed-off-by: David Sterba --- fs/btrfs/tree-checker.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/fs/btrfs/tree-checker.c b/fs/btrfs/tree-checker.c index 8b1b706e0996..4b1e47173c63 100644 --- a/fs/btrfs/tree-checker.c +++ b/fs/btrfs/tree-checker.c @@ -2306,7 +2306,7 @@ static int check_free_space_extent(struct extent_buffer *leaf, struct btrfs_key if (unlikely(btrfs_item_size(leaf, slot) != 0)) { generic_err(leaf, slot, - "invalid item size for free space info, has %u expect 0", + "invalid item size for free space extent, has %u expect 0", btrfs_item_size(leaf, slot)); return -EUCLEAN; } From 2028280686f4fa78e2f1f6dede4b6c1fd782b9e3 Mon Sep 17 00:00:00 2001 From: Dmitry Baryshkov Date: Thu, 3 Sep 2026 15:19:10 +0300 Subject: [PATCH 209/592] drm/msm/dsi: round the byte clock rate after reparenting to the PHY PLL DSI 6G v2.9 hosts (SM8650, SM8750, Kaanapali, etc.) reparent the byte and pixel RCGs to the DSI PHY PLL at runtime from dsi_link_clk_set_rate_6g_v2_9(), after the PHY has been enabled. However dsi_calc_clk_rate_6g() runs earlier, in order to compute the bit clock request for the PHY. At that point the byte RCG still has its reset parent (XO), so clk_round_rate() returns a bogus rate, which then ends up in the PHY bit clock request and the PLL gets programmed to a wrong frequency, breaking the panel. Move the rounding to dsi_link_clk_set_rate_6g(), which is called after the RCGs have been reparented to the PLL. Storing the rounded rate at this point still makes later link_clk_set_rate() calls no-ops in the CCF. Derive the byte interface clock rate from the rounded byte clock rate, otherwise it would keep requesting the idealized rate and retrigger the PLL on every transfer. Reported-by: Abel Vesa Reported-by: Krzysztof Kozlowski Fixes: 6cd33b6f4155 ("drm/msm/dsi: round 6G byte clock rate to the PLL-achievable value") Assisted-by: LLM Signed-off-by: Dmitry Baryshkov Reviewed-by: Konrad Dybcio Tested-by: Konrad Dybcio # SM6115P J606F Tested-by: Abel Vesa Reviewed-by: Abel Vesa Patchwork: https://patchwork.freedesktop.org/patch/750496/ Link: https://lore.kernel.org/r/20260903-fix-eliza-dsi-v1-1-3474a6c9f2e0@oss.qualcomm.com --- drivers/gpu/drm/msm/dsi/dsi_host.c | 36 ++++++++++++++++-------------- 1 file changed, 19 insertions(+), 17 deletions(-) diff --git a/drivers/gpu/drm/msm/dsi/dsi_host.c b/drivers/gpu/drm/msm/dsi/dsi_host.c index 7e4e3718b536..b292dfd266d1 100644 --- a/drivers/gpu/drm/msm/dsi/dsi_host.c +++ b/drivers/gpu/drm/msm/dsi/dsi_host.c @@ -129,7 +129,7 @@ struct msm_dsi_host { struct clk *dsi_pll_pixel_clk; unsigned long byte_clk_rate; - unsigned long byte_intf_clk_rate; + bool byte_intf_clk_div_2; unsigned long pixel_clk_rate; unsigned long esc_clk_rate; @@ -382,8 +382,20 @@ int msm_dsi_runtime_resume(struct device *dev) int dsi_link_clk_set_rate_6g(struct msm_dsi_host *msm_host) { + unsigned long byte_intf_clk_rate; + long rounded_byte_clk_rate; int ret; + rounded_byte_clk_rate = clk_round_rate(msm_host->byte_clk, + msm_host->byte_clk_rate); + if (rounded_byte_clk_rate < 0) { + pr_err("%s: failed to round byte clock rate, %ld\n", + __func__, rounded_byte_clk_rate); + return rounded_byte_clk_rate; + } + + msm_host->byte_clk_rate = rounded_byte_clk_rate; + DBG("Set clk rates: pclk=%lu, byteclk=%lu", msm_host->pixel_clk_rate, msm_host->byte_clk_rate); @@ -401,7 +413,11 @@ int dsi_link_clk_set_rate_6g(struct msm_dsi_host *msm_host) } if (msm_host->byte_intf_clk) { - ret = clk_set_rate(msm_host->byte_intf_clk, msm_host->byte_intf_clk_rate); + byte_intf_clk_rate = msm_host->byte_clk_rate; + if (msm_host->byte_intf_clk_div_2) + byte_intf_clk_rate /= 2; + + ret = clk_set_rate(msm_host->byte_intf_clk, byte_intf_clk_rate); if (ret) { pr_err("%s: Failed to set rate byte intf clk, %d\n", __func__, ret); @@ -669,24 +685,12 @@ static void dsi_calc_pclk(struct msm_dsi_host *msm_host, bool is_bonded_dsi) int dsi_calc_clk_rate_6g(struct msm_dsi_host *msm_host, bool is_bonded_dsi) { - long rounded_byte_clk_rate; - if (!msm_host->mode) { pr_err("%s: mode not set\n", __func__); return -EINVAL; } dsi_calc_pclk(msm_host, is_bonded_dsi); - - rounded_byte_clk_rate = clk_round_rate(msm_host->byte_clk, - msm_host->byte_clk_rate); - if (rounded_byte_clk_rate < 0) { - pr_err("%s: failed to round byte clock rate, %ld\n", - __func__, rounded_byte_clk_rate); - return rounded_byte_clk_rate; - } - - msm_host->byte_clk_rate = rounded_byte_clk_rate; msm_host->esc_clk_rate = clk_get_rate(msm_host->esc_clk); return 0; } @@ -2495,9 +2499,7 @@ int msm_dsi_host_power_on(struct mipi_dsi_host *host, goto unlock_ret; } - msm_host->byte_intf_clk_rate = msm_host->byte_clk_rate; - if (phy_shared_timings->byte_intf_clk_div_2) - msm_host->byte_intf_clk_rate /= 2; + msm_host->byte_intf_clk_div_2 = phy_shared_timings->byte_intf_clk_div_2; msm_dsi_sfpb_config(msm_host, true); From f4fae975db08a9aeec0b15e145c7d4d0fe02a0ec Mon Sep 17 00:00:00 2001 From: Guangshuo Li Date: Sun, 13 Sep 2026 16:58:14 +0800 Subject: [PATCH 210/592] drm/msm/hdmi_phy: fix runtime PM cleanup on probe failure msm_hdmi_phy_probe() enables runtime PM before enabling the PHY resources and initializing the PLL, but failures from either operation return without calling the matching pm_runtime_disable(). The remove path disables runtime PM, but it is not called when probe fails. As a result, runtime PM remains enabled after an unsuccessful probe. Route failures after pm_runtime_enable() through a common error path and disable runtime PM before returning. This issue was found by manual code inspection. Fixes: 15b4a4523859 ("drm/msm/hdmi: Create a separate HDMI PHY driver") Cc: stable@vger.kernel.org Signed-off-by: Guangshuo Li Reviewed-by: Krzysztof Kozlowski Reviewed-by: Dmitry Baryshkov Patchwork: https://patchwork.freedesktop.org/patch/753043/ Link: https://lore.kernel.org/r/20260913085814.1509352-1-lgs201920130244@gmail.com Signed-off-by: Dmitry Baryshkov --- drivers/gpu/drm/msm/hdmi/hdmi_phy.c | 8 ++++++-- 1 file changed, 6 insertions(+), 2 deletions(-) diff --git a/drivers/gpu/drm/msm/hdmi/hdmi_phy.c b/drivers/gpu/drm/msm/hdmi/hdmi_phy.c index eb1088755cb3..77dce35cd45e 100644 --- a/drivers/gpu/drm/msm/hdmi/hdmi_phy.c +++ b/drivers/gpu/drm/msm/hdmi/hdmi_phy.c @@ -168,13 +168,13 @@ static int msm_hdmi_phy_probe(struct platform_device *pdev) ret = msm_hdmi_phy_resource_enable(phy); if (ret) - return ret; + goto err_pm_disable; ret = msm_hdmi_phy_pll_init(pdev, phy->cfg->type); if (ret) { DRM_DEV_ERROR(dev, "couldn't init PLL\n"); msm_hdmi_phy_resource_disable(phy); - return ret; + goto err_pm_disable; } msm_hdmi_phy_resource_disable(phy); @@ -182,6 +182,10 @@ static int msm_hdmi_phy_probe(struct platform_device *pdev) platform_set_drvdata(pdev, phy); return 0; + +err_pm_disable: + pm_runtime_disable(dev); + return ret; } static void msm_hdmi_phy_remove(struct platform_device *pdev) From a15fac810c76397ec9f62a6fc26c4d7ab6e238a7 Mon Sep 17 00:00:00 2001 From: Krzysztof Kozlowski Date: Sun, 13 Sep 2026 14:33:33 +0200 Subject: [PATCH 211/592] dt-bindings: display/msm: Use consistent indentation in the example Correct indentation in the examples to consistent 2- or 4-spaces indentation to fix dt-check-style warnings ("example 0 [indent-consistent] indent mismatch ..."). Preferred is 4-spaces, but re-indenting entire example just for that is too much churn. Signed-off-by: Krzysztof Kozlowski Patchwork: https://patchwork.freedesktop.org/patch/753054/ Link: https://lore.kernel.org/r/20260913123331.100293-4-krzysztof.kozlowski@oss.qualcomm.com Signed-off-by: Dmitry Baryshkov --- .../display/msm/qcom,qcs8300-mdss.yaml | 6 +- .../display/msm/qcom,sar2130p-mdss.yaml | 4 +- .../bindings/display/msm/qcom,sm6150-dpu.yaml | 14 ++--- .../display/msm/qcom,sm8350-mdss.yaml | 2 +- .../display/msm/qcom,x1e80100-mdss.yaml | 60 +++++++++---------- 5 files changed, 43 insertions(+), 43 deletions(-) diff --git a/Documentation/devicetree/bindings/display/msm/qcom,qcs8300-mdss.yaml b/Documentation/devicetree/bindings/display/msm/qcom,qcs8300-mdss.yaml index c41a86203e78..287cce4dbddc 100644 --- a/Documentation/devicetree/bindings/display/msm/qcom,qcs8300-mdss.yaml +++ b/Documentation/devicetree/bindings/display/msm/qcom,qcs8300-mdss.yaml @@ -150,7 +150,7 @@ examples: reg = <0>; dpu_intf0_out: endpoint { - remote-endpoint = <&mdss_dp0_in>; + remote-endpoint = <&mdss_dp0_in>; }; }; @@ -352,9 +352,9 @@ examples: }; port@1 { - reg = <1>; + reg = <1>; - mdss_dp_out: endpoint { }; + mdss_dp_out: endpoint { }; }; }; diff --git a/Documentation/devicetree/bindings/display/msm/qcom,sar2130p-mdss.yaml b/Documentation/devicetree/bindings/display/msm/qcom,sar2130p-mdss.yaml index 44c1bb9e4109..fb3d24bbac7d 100644 --- a/Documentation/devicetree/bindings/display/msm/qcom,sar2130p-mdss.yaml +++ b/Documentation/devicetree/bindings/display/msm/qcom,sar2130p-mdss.yaml @@ -248,9 +248,9 @@ examples: remote-endpoint = <&usb_dp_qmpphy_dp_in>; }; }; - }; + }; - dp_opp_table: opp-table { + dp_opp_table: opp-table { compatible = "operating-points-v2"; opp-162000000 { diff --git a/Documentation/devicetree/bindings/display/msm/qcom,sm6150-dpu.yaml b/Documentation/devicetree/bindings/display/msm/qcom,sm6150-dpu.yaml index b4f437172218..8cbf99f9fcba 100644 --- a/Documentation/devicetree/bindings/display/msm/qcom,sm6150-dpu.yaml +++ b/Documentation/devicetree/bindings/display/msm/qcom,sm6150-dpu.yaml @@ -81,7 +81,7 @@ examples: port@1 { reg = <1>; dpu_intf1_out: endpoint { - remote-endpoint = <&mdss_dsi0_in>; + remote-endpoint = <&mdss_dsi0_in>; }; }; }; @@ -90,18 +90,18 @@ examples: compatible = "operating-points-v2"; opp-19200000 { - opp-hz = /bits/ 64 <19200000>; - required-opps = <&rpmhpd_opp_low_svs>; + opp-hz = /bits/ 64 <19200000>; + required-opps = <&rpmhpd_opp_low_svs>; }; opp-25600000 { - opp-hz = /bits/ 64 <25600000>; - required-opps = <&rpmhpd_opp_svs>; + opp-hz = /bits/ 64 <25600000>; + required-opps = <&rpmhpd_opp_svs>; }; opp-307200000 { - opp-hz = /bits/ 64 <307200000>; - required-opps = <&rpmhpd_opp_nom>; + opp-hz = /bits/ 64 <307200000>; + required-opps = <&rpmhpd_opp_nom>; }; }; }; diff --git a/Documentation/devicetree/bindings/display/msm/qcom,sm8350-mdss.yaml b/Documentation/devicetree/bindings/display/msm/qcom,sm8350-mdss.yaml index 68176de854b3..f5acf528d0e4 100644 --- a/Documentation/devicetree/bindings/display/msm/qcom,sm8350-mdss.yaml +++ b/Documentation/devicetree/bindings/display/msm/qcom,sm8350-mdss.yaml @@ -223,7 +223,7 @@ examples: phys = <&mdss_dsi0_phy>; ports { - #address-cells = <1>; + #address-cells = <1>; #size-cells = <0>; port@0 { diff --git a/Documentation/devicetree/bindings/display/msm/qcom,x1e80100-mdss.yaml b/Documentation/devicetree/bindings/display/msm/qcom,x1e80100-mdss.yaml index 8d698a2e055a..8c65a80b46e4 100644 --- a/Documentation/devicetree/bindings/display/msm/qcom,x1e80100-mdss.yaml +++ b/Documentation/devicetree/bindings/display/msm/qcom,x1e80100-mdss.yaml @@ -208,47 +208,47 @@ examples: #sound-dai-cells = <0>; ports { - #address-cells = <1>; - #size-cells = <0>; + #address-cells = <1>; + #size-cells = <0>; - port@0 { - reg = <0>; + port@0 { + reg = <0>; - mdss_dp0_in: endpoint { - remote-endpoint = <&mdss_intf0_out>; - }; - }; + mdss_dp0_in: endpoint { + remote-endpoint = <&mdss_intf0_out>; + }; + }; - port@1 { - reg = <1>; + port@1 { + reg = <1>; - mdss_dp0_out: endpoint { - }; - }; + mdss_dp0_out: endpoint { + }; + }; }; mdss_dp0_opp_table: opp-table { - compatible = "operating-points-v2"; + compatible = "operating-points-v2"; - opp-160000000 { - opp-hz = /bits/ 64 <160000000>; - required-opps = <&rpmhpd_opp_low_svs>; - }; + opp-160000000 { + opp-hz = /bits/ 64 <160000000>; + required-opps = <&rpmhpd_opp_low_svs>; + }; - opp-270000000 { - opp-hz = /bits/ 64 <270000000>; - required-opps = <&rpmhpd_opp_svs>; - }; + opp-270000000 { + opp-hz = /bits/ 64 <270000000>; + required-opps = <&rpmhpd_opp_svs>; + }; - opp-540000000 { - opp-hz = /bits/ 64 <540000000>; - required-opps = <&rpmhpd_opp_svs_l1>; - }; + opp-540000000 { + opp-hz = /bits/ 64 <540000000>; + required-opps = <&rpmhpd_opp_svs_l1>; + }; - opp-810000000 { - opp-hz = /bits/ 64 <810000000>; - required-opps = <&rpmhpd_opp_nom>; - }; + opp-810000000 { + opp-hz = /bits/ 64 <810000000>; + required-opps = <&rpmhpd_opp_nom>; + }; }; }; }; From e1aeaf79dea51e6065da56924bc07e22d59012ac Mon Sep 17 00:00:00 2001 From: Paulo Alcantara Date: Sun, 13 Sep 2026 21:09:15 -0300 Subject: [PATCH 212/592] smb: client: fix unaligned access in WSL reparse point parser When wsl_to_fattr() parses WSL extended attributes, it computes a payload pointer from ea->ea_data + ea_name_length + 1. Since the smb2_file_full_ea_info struct is __packed and all WSL xattr names are 6 bytes long, the value pointer always lands at an odd byte offset, never satisfying __le32 or __le64 alignment requirements. The code then casts this pointer to __le32 * or __le64 * and dereferences it directly, which may cause alignment faults on some architectures. Replace all such casts with get_unaligned_le32() and get_unaligned_le64() in reparse_mkdev(), wsl_make_kuid(), wsl_make_kgid() and wsl_to_fattr(). Closes: https://sashiko.dev/#/patchset/20260906200517.725015-1-pc%40manguebit.org Fixes: 78e26bec4d6d ("smb: client: parse uid, gid, mode and dev from WSL reparse points") Reviewed-by: Namjae Jeon Signed-off-by: Paulo Alcantara Cc: David Howells Cc: Tom Talpey Cc: Shyam Prasad N Cc: Ronnie Sahlberg Cc: Bharath SM Cc: Namjae Jeon Cc: stable@vger.kernel.org --- fs/smb/client/reparse.c | 4 ++-- fs/smb/client/reparse.h | 7 ++++--- 2 files changed, 6 insertions(+), 5 deletions(-) diff --git a/fs/smb/client/reparse.c b/fs/smb/client/reparse.c index 9e31fce7e0a5..6ac69f4d391a 100644 --- a/fs/smb/client/reparse.c +++ b/fs/smb/client/reparse.c @@ -1201,9 +1201,9 @@ static bool wsl_to_fattr(struct cifs_open_info_data *data, fattr->cf_gid = wsl_make_kgid(cifs_sb, v); } else if (!strncmp(name, SMB2_WSL_XATTR_MODE, nlen)) { /* File type in reparse point tag and in xattr mode must match. */ - if (S_DT(fattr->cf_mode) != S_DT(le32_to_cpu(*(__le32 *)v))) + if (S_DT(fattr->cf_mode) != S_DT(get_unaligned_le32(v))) return false; - fattr->cf_mode = (umode_t)le32_to_cpu(*(__le32 *)v); + fattr->cf_mode = (umode_t)get_unaligned_le32(v); } else if (!strncmp(name, SMB2_WSL_XATTR_DEV, nlen)) { fattr->cf_rdev = reparse_mkdev(v); have_xattr_dev = true; diff --git a/fs/smb/client/reparse.h b/fs/smb/client/reparse.h index 49efd85b1e94..05b2cecb4495 100644 --- a/fs/smb/client/reparse.h +++ b/fs/smb/client/reparse.h @@ -9,6 +9,7 @@ #include #include #include +#include #include "fs_context.h" #include "cifsglob.h" #include "../common/smbfsctl.h" @@ -23,7 +24,7 @@ static inline dev_t reparse_mkdev(void *ptr) { - u64 v = le64_to_cpu(*(__le64 *)ptr); + u64 v = get_unaligned_le64(ptr); return MKDEV(v & 0xffffffff, v >> 32); } @@ -31,7 +32,7 @@ static inline dev_t reparse_mkdev(void *ptr) static inline kuid_t wsl_make_kuid(struct cifs_sb_info *cifs_sb, void *ptr) { - u32 uid = le32_to_cpu(*(__le32 *)ptr); + u32 uid = get_unaligned_le32(ptr); if (cifs_sb_flags(cifs_sb) & CIFS_MOUNT_OVERR_UID) return cifs_sb->ctx->linux_uid; @@ -41,7 +42,7 @@ static inline kuid_t wsl_make_kuid(struct cifs_sb_info *cifs_sb, static inline kgid_t wsl_make_kgid(struct cifs_sb_info *cifs_sb, void *ptr) { - u32 gid = le32_to_cpu(*(__le32 *)ptr); + u32 gid = get_unaligned_le32(ptr); if (cifs_sb_flags(cifs_sb) & CIFS_MOUNT_OVERR_GID) return cifs_sb->ctx->linux_gid; From e1253a82bb4c0fed6706a5839fc8b6e01be1abe2 Mon Sep 17 00:00:00 2001 From: Paulo Alcantara Date: Sun, 13 Sep 2026 21:15:41 -0300 Subject: [PATCH 213/592] smb: client: fix fattr leaking on wsl_to_fattr() failure wsl_to_fattr() mutates fattr fields as it parses each WSL EA. If validation later fails, the function returns false with partially mutated fattr fields that callers do not reset. Fix this by parsing into local variables and only committing them to fattr on success. Closes: https://sashiko.dev/#/patchset/20260906200517.725015-1-pc%40manguebit.org Fixes: 78e26bec4d6d ("smb: client: parse uid, gid, mode and dev from WSL reparse points") Reviewed-by: Namjae Jeon Signed-off-by: Paulo Alcantara Cc: David Howells Cc: Tom Talpey Cc: Shyam Prasad N Cc: Ronnie Sahlberg Cc: Bharath SM Cc: Namjae Jeon Cc: stable@vger.kernel.org --- fs/smb/client/reparse.c | 34 +++++++++++++++++++--------------- 1 file changed, 19 insertions(+), 15 deletions(-) diff --git a/fs/smb/client/reparse.c b/fs/smb/client/reparse.c index 6ac69f4d391a..3a27773186ae 100644 --- a/fs/smb/client/reparse.c +++ b/fs/smb/client/reparse.c @@ -1149,29 +1149,30 @@ static bool wsl_to_fattr(struct cifs_open_info_data *data, u32 tag, struct cifs_fattr *fattr) { unsigned int sbflags = cifs_sb_flags(cifs_sb); + kuid_t uid = cifs_sb->ctx->linux_uid; + kgid_t gid = cifs_sb->ctx->linux_gid; struct smb2_file_full_ea_info *ea; bool have_xattr_dev = false; + dev_t rdev = 0; + umode_t mode; u32 next = 0; - fattr->cf_uid = cifs_sb->ctx->linux_uid; - fattr->cf_gid = cifs_sb->ctx->linux_gid; - - fattr->cf_mode &= ~S_IFMT; + mode = fattr->cf_mode & ~S_IFMT; switch (tag) { case IO_REPARSE_TAG_LX_SYMLINK: - fattr->cf_mode |= S_IFLNK; + mode |= S_IFLNK; break; case IO_REPARSE_TAG_LX_FIFO: - fattr->cf_mode |= S_IFIFO; + mode |= S_IFIFO; break; case IO_REPARSE_TAG_AF_UNIX: - fattr->cf_mode |= S_IFSOCK; + mode |= S_IFSOCK; break; case IO_REPARSE_TAG_LX_CHR: - fattr->cf_mode |= S_IFCHR; + mode |= S_IFCHR; break; case IO_REPARSE_TAG_LX_BLK: - fattr->cf_mode |= S_IFBLK; + mode |= S_IFBLK; break; } @@ -1195,26 +1196,29 @@ static bool wsl_to_fattr(struct cifs_open_info_data *data, if (!strncmp(name, SMB2_WSL_XATTR_UID, nlen)) { if (!(sbflags & CIFS_MOUNT_OVERR_UID)) - fattr->cf_uid = wsl_make_kuid(cifs_sb, v); + uid = wsl_make_kuid(cifs_sb, v); } else if (!strncmp(name, SMB2_WSL_XATTR_GID, nlen)) { if (!(sbflags & CIFS_MOUNT_OVERR_GID)) - fattr->cf_gid = wsl_make_kgid(cifs_sb, v); + gid = wsl_make_kgid(cifs_sb, v); } else if (!strncmp(name, SMB2_WSL_XATTR_MODE, nlen)) { /* File type in reparse point tag and in xattr mode must match. */ - if (S_DT(fattr->cf_mode) != S_DT(get_unaligned_le32(v))) + if (S_DT(mode) != S_DT(get_unaligned_le32(v))) return false; - fattr->cf_mode = (umode_t)get_unaligned_le32(v); + mode = get_unaligned_le32(v); } else if (!strncmp(name, SMB2_WSL_XATTR_DEV, nlen)) { - fattr->cf_rdev = reparse_mkdev(v); + rdev = reparse_mkdev(v); have_xattr_dev = true; } } while (next); out: - /* Major and minor numbers for char and block devices are mandatory. */ if (!have_xattr_dev && (tag == IO_REPARSE_TAG_LX_CHR || tag == IO_REPARSE_TAG_LX_BLK)) return false; + fattr->cf_uid = uid; + fattr->cf_gid = gid; + fattr->cf_mode = mode; + fattr->cf_rdev = rdev; return true; } From 3ed11c671ff7ec58c8fd96410233c677df23f407 Mon Sep 17 00:00:00 2001 From: =?UTF-8?q?Christian=20K=C3=B6nig?= Date: Fri, 4 Sep 2026 10:26:13 +0200 Subject: [PATCH 214/592] dma-buf/dma-fence: fix checking signaling bit for timeline and driver name v3 MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit The patch "dma-buf: dma-fence: Fix potential NULL pointer dereference" changed the check to test for the ops pointer instead of the signaled bit to avoid a potential NULL dereference when the ops pointer has been cleared. The problem is now that the ops pointer is cleared only when neither the release nor the wait callback is implemented and this isn't true for a lot of dma_fence implementations yet. So those implementations lost the RCU protection after signaling of the returned string resulting in potential use after free. Add the signaling check additional to the ops pointer check so that we have both the protection against NULL dereference as well as the RCU protection after signaling for the returned string. v2: improve comments to note RCU protection and explain why we check both signaling state and ops pointer v3: some comment improvements suggested by Philip Signed-off-by: Christian König Fixes: 035219a760ed ("dma-buf: dma-fence: Fix potential NULL pointer dereference") CC: stable@vger.kernel.org # 7.2+ Reported-by: Jonghyuk Kim(MalHyuk) Tested-by: Jonghyuk Kim(MalHyuk) Reviewed-by: Philipp Stanner Link: https://lore.kernel.org/r/20260914182740.1587-1-christian.koenig@amd.com --- drivers/dma-buf/dma-fence.c | 14 ++++++++++++-- include/linux/dma-fence.h | 6 ++++++ 2 files changed, 18 insertions(+), 2 deletions(-) diff --git a/drivers/dma-buf/dma-fence.c b/drivers/dma-buf/dma-fence.c index 05090fb0fd5a..bd58688b81a7 100644 --- a/drivers/dma-buf/dma-fence.c +++ b/drivers/dma-buf/dma-fence.c @@ -1170,7 +1170,12 @@ const char __rcu *dma_fence_driver_name(struct dma_fence *fence) /* RCU protection is required for safe access to returned string */ ops = rcu_dereference(fence->ops); - if (ops) + + /* + * Make load ordering irrelevant by checking both signaled state and ops + * pointer and ops pointer is only set to NULL on newer implementations. + */ + if (!dma_fence_test_signaled_flag(fence) && ops) return (const char __rcu *)ops->get_driver_name(fence); else return (const char __rcu *)"detached-driver"; @@ -1203,7 +1208,12 @@ const char __rcu *dma_fence_timeline_name(struct dma_fence *fence) /* RCU protection is required for safe access to returned string */ ops = rcu_dereference(fence->ops); - if (ops) + + /* + * Make load ordering irrelevant by checking both signaled state and ops + * pointer and ops pointer is only set to NULL on newer implementations. + */ + if (!dma_fence_test_signaled_flag(fence) && ops) return (const char __rcu *)ops->get_timeline_name(fence); else return (const char __rcu *)"signaled-timeline"; diff --git a/include/linux/dma-fence.h b/include/linux/dma-fence.h index 158cd609f103..ffa99b930843 100644 --- a/include/linux/dma-fence.h +++ b/include/linux/dma-fence.h @@ -141,6 +141,9 @@ struct dma_fence_ops { * compute the name at runtime, without having it to store permanently * for each fence, or build a cache of some sort. * + * The returned string is RCU protected and can be freed after the fence + * signaled and a RCU grace period passed. + * * This callback is mandatory. */ const char * (*get_driver_name)(struct dma_fence *fence); @@ -153,6 +156,9 @@ struct dma_fence_ops { * having it to store permanently for each fence, or build a cache of * some sort. * + * The returned string is RCU protected and can be freed after the fence + * signaled and a RCU grace period passed. + * * This callback is mandatory. */ const char * (*get_timeline_name)(struct dma_fence *fence); From 5b644229bd677d52c4efa5973c1fab842c57f896 Mon Sep 17 00:00:00 2001 From: "Darrick J. Wong" Date: Wed, 9 Sep 2026 22:59:59 -0700 Subject: [PATCH 215/592] xfs: guard against igrab failure in xrep_findparent_from_dcache LOLLM suggests that we need to handle igrab returning NULL here. I don't think it's possible for the inode to enter I_FREEING or I_WILL_FREE while we have an active reference to the corresponding dentry, but we can code defensively anyway. Signed-off-by: Darrick J. Wong Assisted-by: LOLLM # finding obvious bugs Reviewed-by: Christoph Hellwig Signed-off-by: Carlos Maiolino --- fs/xfs/scrub/findparent.c | 3 +++ 1 file changed, 3 insertions(+) diff --git a/fs/xfs/scrub/findparent.c b/fs/xfs/scrub/findparent.c index eab3ac2704be..d921fe5a9b0c 100644 --- a/fs/xfs/scrub/findparent.c +++ b/fs/xfs/scrub/findparent.c @@ -473,6 +473,9 @@ xrep_findparent_from_dcache( pip = igrab(d_inode(parent)); dput(parent); + if (!pip) + goto out_dput; + if (S_ISDIR(pip->i_mode)) { ret = pip->i_ino; trace_xrep_findparent_from_dcache(sc->ip, ret); From afbccf99f7f82117cba9ad4b0b006692030f49e8 Mon Sep 17 00:00:00 2001 From: "Darrick J. Wong" Date: Wed, 9 Sep 2026 23:00:16 -0700 Subject: [PATCH 216/592] xfs: don't assert when XFS_SCRUB_TYPE_HEALTHY scans return corruption XFS_SCRUB_TYPE_HEALTHY is a synthentic scrub type so that xfs_scrub can tell the kernel "Hey, I finished a scan and saw no problems" and have the kernel forget that it saw indirect evidence of corruption. Unfortunately, as LOLLM points out, it's possible for the health system to record a new corruption just before xfs_scrub gets to XFS_SCRUB_TYPE_HEALTHY. In this case, the existing logic doesn't return early and instead wanders into unknown regions of type_to_health_flag and trips the assert because HEALTHY doesn't have a group assignment. Fix the logic so that we always return early for a HEALTHY scrub type, even if we decide not to call xchk_mark_all_healthy. Cc: stable@vger.kernel.org # v6.9 Fixes: a1f3e0cca41036 ("xfs: update health status if we get a clean bill of health") Signed-off-by: Darrick J. Wong Assisted-by: LOLLM # finding obvious bugs Reviewed-by: Christoph Hellwig Signed-off-by: Carlos Maiolino --- fs/xfs/scrub/health.c | 6 +++--- 1 file changed, 3 insertions(+), 3 deletions(-) diff --git a/fs/xfs/scrub/health.c b/fs/xfs/scrub/health.c index 2171bcf0f6c1..487ecc5f9f3c 100644 --- a/fs/xfs/scrub/health.c +++ b/fs/xfs/scrub/health.c @@ -202,9 +202,9 @@ xchk_update_health( * there's no sick flag defined for it, so we branch here ahead of the * mask check. */ - if (sc->sm->sm_type == XFS_SCRUB_TYPE_HEALTHY && - !(sc->sm->sm_flags & XFS_SCRUB_OFLAG_CORRUPT)) { - xchk_mark_all_healthy(sc->mp); + if (sc->sm->sm_type == XFS_SCRUB_TYPE_HEALTHY) { + if (!(sc->sm->sm_flags & XFS_SCRUB_OFLAG_CORRUPT)) + xchk_mark_all_healthy(sc->mp); return; } From bb991b7f79dd34cc5f24db0f736bf75630c970e7 Mon Sep 17 00:00:00 2001 From: "Darrick J. Wong" Date: Wed, 9 Sep 2026 23:00:31 -0700 Subject: [PATCH 217/592] xfs: fix attr fork block count checks in xrep_inode_blockcounts LOLLM points out that a file has an attr fork, it will call xchk_inode_count_blocks to set @ablocks to the number of fsblocks mapped by the attr fork; but then it'll compare @blocks (aka the count of fsblocks mapped by the data fork). We already checked that and we never do anything with @acount, so I think this is clearly a bug. Fix the comparison. Cc: stable@vger.kernel.org # v6.8 Fixes: 2d295fe65776d1 ("xfs: repair inode records") Signed-off-by: Darrick J. Wong Assisted-by: LOLLM # finding obvious bugs Reviewed-by: Christoph Hellwig Signed-off-by: Carlos Maiolino --- fs/xfs/scrub/inode_repair.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/fs/xfs/scrub/inode_repair.c b/fs/xfs/scrub/inode_repair.c index 8bc508336aa5..b87c22146233 100644 --- a/fs/xfs/scrub/inode_repair.c +++ b/fs/xfs/scrub/inode_repair.c @@ -1702,7 +1702,7 @@ xrep_inode_blockcounts( &acount); if (error) return error; - if (count >= sc->mp->m_sb.sb_dblocks) + if (acount >= sc->mp->m_sb.sb_dblocks) return -EFSCORRUPTED; error = xrep_ino_ensure_extent_count(sc, XFS_ATTR_FORK, nextents); From 1c32cdc986467eaffeedb6c5334852809555b82d Mon Sep 17 00:00:00 2001 From: "Darrick J. Wong" Date: Wed, 9 Sep 2026 23:00:47 -0700 Subject: [PATCH 218/592] xfs: release orphanage dir inode if chown fails LOLLM points out that we leak the igrab'd reference to the orphanage directory inode if chowning it fails. Fix that. Cc: stable@vger.kernel.org # v6.10 Fixes: 1e58a8ccf2597c ("xfs: move orphan files to the orphanage") Signed-off-by: Darrick J. Wong Assisted-by: LOLLM # finding obvious bugs Reviewed-by: Christoph Hellwig Signed-off-by: Carlos Maiolino --- fs/xfs/scrub/orphanage.c | 6 +++++- 1 file changed, 5 insertions(+), 1 deletion(-) diff --git a/fs/xfs/scrub/orphanage.c b/fs/xfs/scrub/orphanage.c index 3aca66869b80..21e31eeaa042 100644 --- a/fs/xfs/scrub/orphanage.c +++ b/fs/xfs/scrub/orphanage.c @@ -192,12 +192,16 @@ xrep_orphanage_create( /* Make sure the orphanage is owned by root. */ error = xrep_chown_orphanage(sc, XFS_I(orphanage_inode)); if (error) - goto out_dput_orphanage; + goto out_rele_orphanage; /* Stash the reference for later and bail out. */ sc->orphanage = XFS_I(orphanage_inode); sc->orphanage_ilock_flags = 0; + orphanage_inode = NULL; +out_rele_orphanage: + if (orphanage_inode) + xchk_irele(sc, XFS_I(orphanage_inode)); out_dput_orphanage: end_creating(orphanage_dentry); out_dput_root: From 8b4ad2814274d43ed8bdfcf857efb8c79815d9c6 Mon Sep 17 00:00:00 2001 From: "Darrick J. Wong" Date: Wed, 9 Sep 2026 23:01:03 -0700 Subject: [PATCH 219/592] xfs: release AGFL after walking it during rmapbt repair LOLLM suggests that we not hold the AGFL locked to the scrub transaction for any longer than we have to when we're rebuilding the rmapbt. Since we only took it to generate an rmap record for the AGFL blocks, I think we can safely release it. Signed-off-by: Darrick J. Wong Assisted-by: LOLLM # finding obvious bugs Reviewed-by: Christoph Hellwig Signed-off-by: Carlos Maiolino --- fs/xfs/scrub/rmap_repair.c | 1 + 1 file changed, 1 insertion(+) diff --git a/fs/xfs/scrub/rmap_repair.c b/fs/xfs/scrub/rmap_repair.c index 590f9f41856e..725035bf4903 100644 --- a/fs/xfs/scrub/rmap_repair.c +++ b/fs/xfs/scrub/rmap_repair.c @@ -1109,6 +1109,7 @@ xrep_rmap_try_reserve( return error; error = xfs_agfl_walk(sc->mp, agf, agfl_bp, xrep_rmap_walk_agfl, &ra); + xfs_trans_brelse(sc->tp, agfl_bp); if (error) return error; From 984aab2d905a8557fafb27cd9e8713d6d12b3437 Mon Sep 17 00:00:00 2001 From: "Darrick J. Wong" Date: Wed, 9 Sep 2026 23:01:18 -0700 Subject: [PATCH 220/592] xfs: use correct jiffies comparison function in xchk_maybe_relax LOLLM points out that we're supposed to use time_after_eq, not a raw >= operation here, or else jiffies wraps can go unnoticed. Fix this. Cc: stable@vger.kernel.org # v6.10 Fixes: 271557de7cbfde ("xfs: reduce the rate of cond_resched calls inside scrub") Signed-off-by: Darrick J. Wong Assisted-by: LOLLM # finding obvious bugs Reviewed-by: Christoph Hellwig Signed-off-by: Carlos Maiolino --- fs/xfs/scrub/scrub.h | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/fs/xfs/scrub/scrub.h b/fs/xfs/scrub/scrub.h index 737a5d6db15f..b093945f3631 100644 --- a/fs/xfs/scrub/scrub.h +++ b/fs/xfs/scrub/scrub.h @@ -40,7 +40,7 @@ static inline int xchk_maybe_relax(struct xchk_relax *widget) return 0; widget->resched_nr = 0; - if (unlikely(widget->next_resched <= jiffies)) { + if (unlikely(time_after_eq(jiffies, widget->next_resched))) { cond_resched(); widget->next_resched = XCHK_RELAX_NEXT; } From 3083ba8dde765a9ab2337f3db68d00724a6b1202 Mon Sep 17 00:00:00 2001 From: "Darrick J. Wong" Date: Thu, 10 Sep 2026 22:53:39 -0700 Subject: [PATCH 221/592] xfs: check padding field in xfs_ioc_commit_range LOLLM points out that we don't check the ioctl padding field here, so let's do that. I don't think there are many users yet since exchrange requires a new feature flag, so it's a good time to try to plug this hole. Cc: stable@vger.kernel.org # v6.12 Fixes: 398597c3ef7fb1 ("xfs: introduce new file range commit ioctls") Signed-off-by: Darrick J. Wong Assisted-by: LOLLM # finding obvious bugs Reviewed-by: Christoph Hellwig Signed-off-by: Carlos Maiolino --- fs/xfs/xfs_exchrange.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/fs/xfs/xfs_exchrange.c b/fs/xfs/xfs_exchrange.c index c69ecd6a19de..07090487c581 100644 --- a/fs/xfs/xfs_exchrange.c +++ b/fs/xfs/xfs_exchrange.c @@ -902,7 +902,7 @@ xfs_ioc_commit_range( if (copy_from_user(&args, argp, sizeof(args))) return -EFAULT; - if (args.flags & ~XFS_EXCHANGE_RANGE_ALL_FLAGS) + if (args.pad || (args.flags & ~XFS_EXCHANGE_RANGE_ALL_FLAGS)) return -EINVAL; if (kern_f->magic != XCR_FRESH_MAGIC) return -EBUSY; From 8fc18580ec17f90beac4c933fbe4c74dcd3b7f36 Mon Sep 17 00:00:00 2001 From: "Darrick J. Wong" Date: Thu, 10 Sep 2026 22:53:55 -0700 Subject: [PATCH 222/592] xfs: don't call xfs_exchange_range_finish for a dry run LOLLM noticed that we strip file privileges and whatnot even for a dry run. We also shouldn't flush dirty data to disk or trim COW staging events for a dry run. Neither of those behaviors are allowed by the manpage, so fix that by exiting early on DRY_RUN in various functions. Cc: stable@vger.kernel.org # v6.10 Fixes: 42672471f938cd ("xfs: bind together the front and back ends of the file range exchange code") Signed-off-by: Darrick J. Wong Assisted-by: LOLLM # finding obvious bugs Reviewed-by: Christoph Hellwig Signed-off-by: Carlos Maiolino --- fs/xfs/xfs_exchrange.c | 11 ++++++++--- 1 file changed, 8 insertions(+), 3 deletions(-) diff --git a/fs/xfs/xfs_exchrange.c b/fs/xfs/xfs_exchrange.c index 07090487c581..fafb4e3f065c 100644 --- a/fs/xfs/xfs_exchrange.c +++ b/fs/xfs/xfs_exchrange.c @@ -633,6 +633,9 @@ xfs_exchrange_prep( if (error) return error; + if (fxr->flags & XFS_EXCHANGE_RANGE_DRY_RUN) + return 0; + trace_xfs_exchrange_flush(fxr, ip1, ip2); /* Flush the relevant ranges of both files. */ @@ -709,9 +712,11 @@ xfs_exchrange_contents( * other file write would do. This may involve turning on support for * logged xattrs if either file has security capabilities. */ - error = xfs_exchange_range_finish(fxr); - if (error) - goto out_unlock; + if (!(fxr->flags & XFS_EXCHANGE_RANGE_DRY_RUN)) { + error = xfs_exchange_range_finish(fxr); + if (error) + goto out_unlock; + } out_unlock: xfs_iunlock2_io_mmap(ip1, ip2); From 471e0b6e2ddac9e16b8dc2153d6e5575fefb8a2f Mon Sep 17 00:00:00 2001 From: "Darrick J. Wong" Date: Thu, 10 Sep 2026 22:54:11 -0700 Subject: [PATCH 223/592] xfs: use the correct reservations for rtrmap/refcount recovery LOLLM noticed that we might reserve the wrong number of blocks for recovering rtrmap and rtrefcount updates after a crash. Fix that. Cc: stable@vger.kernel.org # v6.14 Fixes: 5e0679d1c62f25 ("xfs: support recovering rmap intent items targetting realtime extents") Signed-off-by: Darrick J. Wong Assisted-by: LOLLM # finding obvious bugs Reviewed-by: Christoph Hellwig Signed-off-by: Carlos Maiolino --- fs/xfs/xfs_refcount_item.c | 8 ++++++-- fs/xfs/xfs_rmap_item.c | 8 ++++++-- 2 files changed, 12 insertions(+), 4 deletions(-) diff --git a/fs/xfs/xfs_refcount_item.c b/fs/xfs/xfs_refcount_item.c index 8bccf89a7766..682c6e1b45e3 100644 --- a/fs/xfs/xfs_refcount_item.c +++ b/fs/xfs/xfs_refcount_item.c @@ -508,6 +508,7 @@ xfs_refcount_recover_work( struct xfs_cui_log_item *cuip = CUI_ITEM(lip); struct xfs_trans *tp; struct xfs_mount *mp = lip->li_log->l_mp; + unsigned int dblocks; bool isrt = xfs_cui_item_isrt(lip); int i; int error = 0; @@ -543,8 +544,11 @@ xfs_refcount_recover_work( * full btree split on either end of the refcount range. */ resv = xlog_recover_resv(&M_RES(mp)->tr_itruncate); - error = xfs_trans_alloc(mp, &resv, mp->m_refc_maxlevels * 2, 0, - XFS_TRANS_RESERVE, &tp); + if (isrt) + dblocks = mp->m_rtrefc_maxlevels * 2; + else + dblocks = mp->m_refc_maxlevels * 2; + error = xfs_trans_alloc(mp, &resv, dblocks, 0, XFS_TRANS_RESERVE, &tp); if (error) return error; diff --git a/fs/xfs/xfs_rmap_item.c b/fs/xfs/xfs_rmap_item.c index 2a3a73a8566d..000cff1ce324 100644 --- a/fs/xfs/xfs_rmap_item.c +++ b/fs/xfs/xfs_rmap_item.c @@ -573,6 +573,7 @@ xfs_rmap_recover_work( struct xfs_rui_log_item *ruip = RUI_ITEM(lip); struct xfs_trans *tp; struct xfs_mount *mp = lip->li_log->l_mp; + unsigned int dblocks; bool isrt = xfs_rui_item_isrt(lip); int i; int error = 0; @@ -596,8 +597,11 @@ xfs_rmap_recover_work( } resv = xlog_recover_resv(&M_RES(mp)->tr_itruncate); - error = xfs_trans_alloc(mp, &resv, mp->m_rmap_maxlevels, 0, - XFS_TRANS_RESERVE, &tp); + if (isrt) + dblocks = mp->m_rtrmap_maxlevels; + else + dblocks = mp->m_rmap_maxlevels; + error = xfs_trans_alloc(mp, &resv, dblocks, 0, XFS_TRANS_RESERVE, &tp); if (error) return error; From 46c1b6674a7b3a8385e7fa27f4efc6f45eddf967 Mon Sep 17 00:00:00 2001 From: "Darrick J. Wong" Date: Thu, 10 Sep 2026 22:54:26 -0700 Subject: [PATCH 224/592] xfs: fix integer overflows in xbitmap set functions LOLLM complains that the xbitmap set functions can suffer an integer underflow or overflow and thereby return the wrong left and right pointers. Fix that logic bomb, even though (AFAICT) we never actually try to set the *entire* bitmap. Signed-off-by: Darrick J. Wong Assisted-by: LOLLM # finding obvious bugs Reviewed-by: Christoph Hellwig Signed-off-by: Carlos Maiolino --- fs/xfs/scrub/bitmap.c | 26 ++++++++++++++++++-------- 1 file changed, 18 insertions(+), 8 deletions(-) diff --git a/fs/xfs/scrub/bitmap.c b/fs/xfs/scrub/bitmap.c index c7fa908d92b2..08f216d26ec6 100644 --- a/fs/xfs/scrub/bitmap.c +++ b/fs/xfs/scrub/bitmap.c @@ -122,8 +122,8 @@ xbitmap64_set( uint64_t start, uint64_t len) { - struct xbitmap64_node *left; - struct xbitmap64_node *right; + struct xbitmap64_node *left = NULL; + struct xbitmap64_node *right = NULL; uint64_t last = start + len - 1; int error; @@ -131,6 +131,7 @@ xbitmap64_set( left = xbitmap64_tree_iter_first(&bitmap->xb_root, start, last); if (left && left->bn_start <= start && left->bn_last >= last) return 0; + left = NULL; /* Clear out everything in the range we want to set. */ error = xbitmap64_clear(bitmap, start, len); @@ -138,11 +139,15 @@ xbitmap64_set( return error; /* Do we have a left-adjacent extent? */ - left = xbitmap64_tree_iter_first(&bitmap->xb_root, start - 1, start - 1); + if (start > 0) + left = xbitmap64_tree_iter_first(&bitmap->xb_root, start - 1, + start - 1); ASSERT(!left || left->bn_last + 1 == start); /* Do we have a right-adjacent extent? */ - right = xbitmap64_tree_iter_first(&bitmap->xb_root, last + 1, last + 1); + if (last < U64_MAX) + right = xbitmap64_tree_iter_first(&bitmap->xb_root, last + 1, + last + 1); ASSERT(!right || right->bn_start == last + 1); if (left && right) { @@ -397,8 +402,8 @@ xbitmap32_set( uint32_t start, uint32_t len) { - struct xbitmap32_node *left; - struct xbitmap32_node *right; + struct xbitmap32_node *left = NULL; + struct xbitmap32_node *right = NULL; uint32_t last = start + len - 1; int error; @@ -406,6 +411,7 @@ xbitmap32_set( left = xbitmap32_tree_iter_first(&bitmap->xb_root, start, last); if (left && left->bn_start <= start && left->bn_last >= last) return 0; + left = NULL; /* Clear out everything in the range we want to set. */ error = xbitmap32_clear(bitmap, start, len); @@ -413,11 +419,15 @@ xbitmap32_set( return error; /* Do we have a left-adjacent extent? */ - left = xbitmap32_tree_iter_first(&bitmap->xb_root, start - 1, start - 1); + if (start > 0) + left = xbitmap32_tree_iter_first(&bitmap->xb_root, start - 1, + start - 1); ASSERT(!left || left->bn_last + 1 == start); /* Do we have a right-adjacent extent? */ - right = xbitmap32_tree_iter_first(&bitmap->xb_root, last + 1, last + 1); + if (last < U32_MAX) + right = xbitmap32_tree_iter_first(&bitmap->xb_root, last + 1, + last + 1); ASSERT(!right || right->bn_start == last + 1); if (left && right) { From c54110d814c3e8ed6bbbb5e02874994ed9f668ac Mon Sep 17 00:00:00 2001 From: "Darrick J. Wong" Date: Thu, 10 Sep 2026 22:54:42 -0700 Subject: [PATCH 225/592] xfs: only flag zero padding for dir3 data blocks, not dir3 block blocks LOLLM complains that xchk_directory_data_bestfree can be passed a directory block that is either in "block" or "data" format, but the check here unconditionally treats the dir3_block and dir3_data blocks as if they have the same header format (they don't). Consequently, we can incorrectly set the preen state on dir3_block blocks, which of course we can't preen away because dir3_block blocks do not have a padding field. Fix this. Cc: stable@vger.kernel.org # v7.1-rc4 Fixes: 939919ccddfcc3 ("xfs: check directory data block header padding in scrub") Signed-off-by: Darrick J. Wong Assisted-by: LOLLM # finding obvious bugs Reviewed-by: Christoph Hellwig Signed-off-by: Carlos Maiolino --- fs/xfs/scrub/dir.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/fs/xfs/scrub/dir.c b/fs/xfs/scrub/dir.c index 2a037aae904d..19d974c7e2b7 100644 --- a/fs/xfs/scrub/dir.c +++ b/fs/xfs/scrub/dir.c @@ -492,7 +492,7 @@ xchk_directory_data_bestfree( goto out; xchk_buffer_recheck(sc, bp); - if (xfs_has_crc(sc->mp)) { + if (!is_block && xfs_has_crc(sc->mp)) { struct xfs_dir3_data_hdr *hdr3 = bp->b_addr; if (hdr3->pad) From e9193f2f1ce32d02b9230094ffdbfab715ab6137 Mon Sep 17 00:00:00 2001 From: "Darrick J. Wong" Date: Thu, 10 Sep 2026 22:54:58 -0700 Subject: [PATCH 226/592] xfs: check di_forkoff correctly in scrub The di_forkoff check in xchk_dinode is incorrect, according to LOLLM. XFS_DFORK_BOFF returns a byte count relative to the start of the literal area, not the start of the inode. Therefore, this check won't flag di_forkoff values that are larger than the literal area but not the inode size itself. Fix this check; sadly the old APTR code was correct. Cc: stable@vger.kernel.org # v6.8 Fixes: 6b5d917780219d ("xfs: dont cast to char * for XFS_DFORK_*PTR macros") Signed-off-by: Darrick J. Wong Assisted-by: LOLLM # finding obvious bugs Reviewed-by: Christoph Hellwig Signed-off-by: Carlos Maiolino --- fs/xfs/scrub/inode.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/fs/xfs/scrub/inode.c b/fs/xfs/scrub/inode.c index 65b13e311916..46e9bf4a4317 100644 --- a/fs/xfs/scrub/inode.c +++ b/fs/xfs/scrub/inode.c @@ -607,7 +607,7 @@ xchk_dinode( } /* di_forkoff */ - if (XFS_DFORK_BOFF(dip) >= mp->m_sb.sb_inodesize) + if (dip->di_forkoff >= (XFS_LITINO(mp) >> 3)) xchk_ino_set_corrupt(sc, ino); if (naextents != 0 && dip->di_forkoff == 0) xchk_ino_set_corrupt(sc, ino); From 14e379600d3e57ab0872b049c28cfe5ef519d439 Mon Sep 17 00:00:00 2001 From: Christoph Hellwig Date: Mon, 14 Sep 2026 13:35:38 +0200 Subject: [PATCH 227/592] xfs: don't try to get a reference to a NULL oz in xfs_get_cached_zone oz can be NULL when we resample it after taking i_flags_lock, so account for that. Fixes: 2d829cc76777 ("xfs: fix racy open zone caching") Signed-off-by: Christoph Hellwig Reviewed-by: Hans Holmberg Reviewed-by: Carlos Maiolino Signed-off-by: Carlos Maiolino --- fs/xfs/xfs_zone_alloc.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/fs/xfs/xfs_zone_alloc.c b/fs/xfs/xfs_zone_alloc.c index 28c1e48909fa..d7ec055a9a06 100644 --- a/fs/xfs/xfs_zone_alloc.c +++ b/fs/xfs/xfs_zone_alloc.c @@ -820,7 +820,7 @@ xfs_get_cached_zone( spin_unlock(&ip->i_flags_lock); } - if (!atomic_inc_not_zero(&oz->oz_ref)) + if (oz && !atomic_inc_not_zero(&oz->oz_ref)) oz = NULL; out_unlock: rcu_read_unlock(); From ee415ce8cba154d07d02a6d2fbb27ff518264c3a Mon Sep 17 00:00:00 2001 From: Luca Coelho Date: Tue, 8 Sep 2026 13:06:51 +0300 Subject: [PATCH 228/592] drm/i915/display: check configuration index before shifting The calc_allowed_config_filter() function passes the return value of iter_pos_to_idx() directly to BIT(), but the helper can return -1 for an invalid iterator. The iterator already rejects negative indices before doing a configuration, so this should not matter in normal flows. In any case, for robustness, check the index explicitly and warn if it is negative, avoiding an undefined shift. Fixes: 39e30bdf2f92 ("drm/i915/dp_link_caps: Add link configuration iterator") Reviewed-by: Imre Deak Link: https://patch.msgid.link/20260908100659.113555-1-luciano.coelho@intel.com Signed-off-by: Luca Coelho (cherry picked from commit fe05cb9b9fb0ecc10409c4c6133257214b6cd8c8) Signed-off-by: Jani Nikula --- drivers/gpu/drm/i915/display/intel_dp_link_caps.c | 9 ++++++++- 1 file changed, 8 insertions(+), 1 deletion(-) diff --git a/drivers/gpu/drm/i915/display/intel_dp_link_caps.c b/drivers/gpu/drm/i915/display/intel_dp_link_caps.c index 7b6cc6055da8..98657aa4d3d5 100644 --- a/drivers/gpu/drm/i915/display/intel_dp_link_caps.c +++ b/drivers/gpu/drm/i915/display/intel_dp_link_caps.c @@ -426,12 +426,15 @@ calc_allowed_config_filter(struct intel_dp_link_caps *link_caps, const struct intel_dp_link_config *forced_params) { struct intel_dp_link_caps_filter allowed_configs = INTEL_DP_LINK_CAPS_FILTER_NONE; + struct intel_display *display = to_intel_display(link_caps->dp); struct intel_dp_link_caps_order order = bw_desc_config_order(); struct intel_dp_link_caps_iter iter; struct intel_dp_link_config config; iter_start(&iter, link_caps, order, enabled_configs); for_each_dp_link_config(&iter, &config) { + int config_idx; + if (forced_params->rate && forced_params->rate != config.rate) continue; @@ -446,7 +449,11 @@ calc_allowed_config_filter(struct intel_dp_link_caps *link_caps, if (config.lane_count > max_limits->lane_count) continue; - allowed_configs.config_mask |= BIT(iter_pos_to_idx(link_caps, order, iter.pos)); + config_idx = iter_pos_to_idx(link_caps, order, iter.pos); + if (drm_WARN_ON(display->drm, config_idx < 0)) + continue; + + allowed_configs.config_mask |= BIT(config_idx); } intel_dp_link_caps_iter_end(&iter); From ce2b91bebc7bc0495fe3ad5ee47e4977fbb77fc5 Mon Sep 17 00:00:00 2001 From: Jiangshan Yi Date: Mon, 14 Sep 2026 16:21:11 +0800 Subject: [PATCH 229/592] xfs: remove duplicate INO1_WRITTEN check Commit a23eca88448e ("xfs: fix exchange-range reflink flag clearing issue with INO1_WRITTEN") duplicated commit b2d5a81dae38 ("xfs: fix exchange-range reflink flag clearing issue with INO1_WRITTEN"), so xmi_can_exchange_reflink_flags() ended up with two identical XFS_EXCHMAPS_INO1_WRITTEN checks. The second one is dead code, since the first one already returned false. Remove it. Signed-off-by: Jiangshan Yi Reviewed-by: Darrick J. Wong Signed-off-by: Carlos Maiolino --- fs/xfs/libxfs/xfs_exchmaps.c | 10 ---------- 1 file changed, 10 deletions(-) diff --git a/fs/xfs/libxfs/xfs_exchmaps.c b/fs/xfs/libxfs/xfs_exchmaps.c index 49eda8d0994d..3efed37cb98a 100644 --- a/fs/xfs/libxfs/xfs_exchmaps.c +++ b/fs/xfs/libxfs/xfs_exchmaps.c @@ -959,16 +959,6 @@ xmi_can_exchange_reflink_flags( { struct xfs_mount *mp = req->ip1->i_mount; - /* - * The INO1_WRITTEN optimization can skip exchanging hole and - * unwritten mappings, which means we cannot guarantee that all - * shared extents actually moved to the other file. Clearing the - * reflink flag of an inode that still holds shared extents breaks - * the CoW write path, so refuse to exchange the flags in that case. - */ - if (req->flags & XFS_EXCHMAPS_INO1_WRITTEN) - return false; - /* * The INO1_WRITTEN optimization can skip exchanging hole and * unwritten mappings, which means we cannot guarantee that all From c16b885ad6b589b233534ee99ade82110d2bc381 Mon Sep 17 00:00:00 2001 From: Anuj Gupta Date: Fri, 4 Sep 2026 18:06:41 +0530 Subject: [PATCH 230/592] xfs: remove unused xfs_reflink_remap_range declaration The implementation was inlined into xfs_file_remap_range() by commit 3fc9f5e40931 ("xfs: remove xfs_reflink_remap_range"), leaving this declaration orphaned. Signed-off-by: Anuj Gupta Reviewed-by: Darrick J. Wong Signed-off-by: Carlos Maiolino --- fs/xfs/xfs_reflink.h | 3 --- 1 file changed, 3 deletions(-) diff --git a/fs/xfs/xfs_reflink.h b/fs/xfs/xfs_reflink.h index 9d1ed9bb0bee..683c1841e640 100644 --- a/fs/xfs/xfs_reflink.h +++ b/fs/xfs/xfs_reflink.h @@ -48,9 +48,6 @@ extern int xfs_reflink_end_cow(struct xfs_inode *ip, xfs_off_t offset, int xfs_reflink_end_atomic_cow(struct xfs_inode *ip, xfs_off_t offset, xfs_off_t count); extern int xfs_reflink_recover_cow(struct xfs_mount *mp); -extern loff_t xfs_reflink_remap_range(struct file *file_in, loff_t pos_in, - struct file *file_out, loff_t pos_out, loff_t len, - unsigned int remap_flags); extern int xfs_reflink_inode_has_shared_extents(struct xfs_trans *tp, struct xfs_inode *ip, bool *has_shared); extern int xfs_reflink_clear_inode_flag(struct xfs_inode *ip, From 2f3c2a6f963e57cf4f0f34cbf24ab11abb64d118 Mon Sep 17 00:00:00 2001 From: Hemanth Selam Date: Fri, 11 Sep 2026 13:00:07 +0530 Subject: [PATCH 231/592] xfs: fix typos and repeated words in comments Correct ten misspellings and eight accidentally doubled words in comments. No code changes. The doubled word in xfs_zone_alloc.c was not a duplicate: "so that is is reused" is "it is" misspelt, so that one reads "so that it is reused" rather than dropping a word. Signed-off-by: Hemanth Selam Reviewed-by: Darrick J. Wong Reviewed-by: Carlos Maiolino Signed-off-by: Carlos Maiolino --- fs/xfs/libxfs/xfs_ag.h | 2 +- fs/xfs/libxfs/xfs_alloc.c | 4 ++-- fs/xfs/libxfs/xfs_attr_leaf.c | 2 +- fs/xfs/libxfs/xfs_errortag.h | 2 +- fs/xfs/libxfs/xfs_exchmaps.c | 2 +- fs/xfs/libxfs/xfs_format.h | 2 +- fs/xfs/libxfs/xfs_inode_buf.c | 2 +- fs/xfs/scrub/agheader_repair.c | 2 +- fs/xfs/scrub/alloc_repair.c | 2 +- fs/xfs/scrub/dirtree.c | 2 +- fs/xfs/scrub/reap.c | 2 +- fs/xfs/xfs_bmap_item.c | 2 +- fs/xfs/xfs_inode.c | 2 +- fs/xfs/xfs_log_cil.c | 2 +- fs/xfs/xfs_platform.h | 2 +- fs/xfs/xfs_zone_alloc.c | 2 +- fs/xfs/xfs_zone_gc.c | 2 +- 17 files changed, 18 insertions(+), 18 deletions(-) diff --git a/fs/xfs/libxfs/xfs_ag.h b/fs/xfs/libxfs/xfs_ag.h index fd22fe598931..ee636b66a72f 100644 --- a/fs/xfs/libxfs/xfs_ag.h +++ b/fs/xfs/libxfs/xfs_ag.h @@ -207,7 +207,7 @@ xfs_perag_next( } /* - * Per-ag geometry infomation and validation + * Per-ag geometry information and validation */ xfs_agblock_t xfs_ag_block_count(struct xfs_mount *mp, xfs_agnumber_t agno); void xfs_agino_range(struct xfs_mount *mp, xfs_agnumber_t agno, diff --git a/fs/xfs/libxfs/xfs_alloc.c b/fs/xfs/libxfs/xfs_alloc.c index d99602bcc16f..f762dcce8d13 100644 --- a/fs/xfs/libxfs/xfs_alloc.c +++ b/fs/xfs/libxfs/xfs_alloc.c @@ -3487,7 +3487,7 @@ xfs_alloc_read_agf( } /* - * Pre-proces allocation arguments to set initial state that we don't require + * Pre-process allocation arguments to set initial state that we don't require * callers to set up correctly, as well as bounds check the allocation args * that are set up. */ @@ -3608,7 +3608,7 @@ xfs_alloc_vextent_finish( * ABBA AGF deadlocks because a future allocation attempt in this * transaction may attempt to lock a lower number AGF. * - * We can't release the AGF until the transaction is commited, so at + * We can't release the AGF until the transaction is committed, so at * this point we must update the "first allocation" tracker to point at * this AG if the tracker is empty or points to a lower AG. This allows * the next allocation attempt to be modified appropriately to avoid diff --git a/fs/xfs/libxfs/xfs_attr_leaf.c b/fs/xfs/libxfs/xfs_attr_leaf.c index b6288395f853..2c80f4fd0b78 100644 --- a/fs/xfs/libxfs/xfs_attr_leaf.c +++ b/fs/xfs/libxfs/xfs_attr_leaf.c @@ -1715,7 +1715,7 @@ xfs_attr3_leaf_add_work( /* * This freemap entry starts at the old end of the * leaf entry array, so we need to adjust its base - * upward to accomodate the larger array. + * upward to accommodate the larger array. */ diff = sizeof(struct xfs_attr_leaf_entry); } else if (ichdr->freemap[i].size > 0 && diff --git a/fs/xfs/libxfs/xfs_errortag.h b/fs/xfs/libxfs/xfs_errortag.h index 6de207fed2d8..f0c83f1f0b3b 100644 --- a/fs/xfs/libxfs/xfs_errortag.h +++ b/fs/xfs/libxfs/xfs_errortag.h @@ -83,7 +83,7 @@ #define XFS_RANDOM_DEFAULT 100 /* - * Table of errror injection knobs. The parameters to the XFS_ERRTAG macro are: + * Table of error injection knobs. The parameters to the XFS_ERRTAG macro are: * 1. The XFS_ERRTAG_ flag but without the prefix; * 2. The name of the sysfs knob; and * 3. The default value for the knob. diff --git a/fs/xfs/libxfs/xfs_exchmaps.c b/fs/xfs/libxfs/xfs_exchmaps.c index 3efed37cb98a..6a66b6075e0a 100644 --- a/fs/xfs/libxfs/xfs_exchmaps.c +++ b/fs/xfs/libxfs/xfs_exchmaps.c @@ -395,7 +395,7 @@ xfs_exchmaps_one_step( /* * Re-add both mappings. We exchange the file offsets between the two * maps and add the opposite map, which has the effect of filling the - * logical offsets we just unmapped, but with with the physical mapping + * logical offsets we just unmapped, but with the physical mapping * information exchanged. */ swap(irec1->br_startoff, irec2->br_startoff); diff --git a/fs/xfs/libxfs/xfs_format.h b/fs/xfs/libxfs/xfs_format.h index dd0ed046fbe9..1a7a7e60a170 100644 --- a/fs/xfs/libxfs/xfs_format.h +++ b/fs/xfs/libxfs/xfs_format.h @@ -1051,7 +1051,7 @@ enum xfs_dinode_fmt { * block is 1KB in size. * * With XFS_MAX_EXTCNT_DATA_FORK_SMALL representing maximum extent count and - * with 1KB sized blocks, a file can reach upto, + * with 1KB sized blocks, a file can reach up to, * 1KB * (2^31) = 2TB * * This is much larger than the theoretical maximum size of a directory diff --git a/fs/xfs/libxfs/xfs_inode_buf.c b/fs/xfs/libxfs/xfs_inode_buf.c index e4c3f7b24e95..0340e2189921 100644 --- a/fs/xfs/libxfs/xfs_inode_buf.c +++ b/fs/xfs/libxfs/xfs_inode_buf.c @@ -626,7 +626,7 @@ xfs_dinode_verify( * have di_nlink track the link count, even if the actual filesystem * only supported V1 inodes (i.e. di_onlink). When writing out the * ondisk inode, it would set both the ondisk di_nlink and di_onlink to - * the the incore di_nlink value, which is why we cannot check for + * the incore di_nlink value, which is why we cannot check for * di_nlink==0 on a V1 inode. V2/3 inodes would get written out with * di_onlink==0, so we can check that. */ diff --git a/fs/xfs/scrub/agheader_repair.c b/fs/xfs/scrub/agheader_repair.c index a66b611588c4..ff1b4b361cf2 100644 --- a/fs/xfs/scrub/agheader_repair.c +++ b/fs/xfs/scrub/agheader_repair.c @@ -1369,7 +1369,7 @@ xrep_iunlink_mark_ondisk( /* * Walk an iunlink bucket's inode list. For each inode that should be on this - * chain, clear its entry in in iunlink_bmp because it's ok and we don't need + * chain, clear its entry in iunlink_bmp because it's ok and we don't need * to touch it further. */ STATIC int diff --git a/fs/xfs/scrub/alloc_repair.c b/fs/xfs/scrub/alloc_repair.c index 95e318e4f3a6..2398e3819597 100644 --- a/fs/xfs/scrub/alloc_repair.c +++ b/fs/xfs/scrub/alloc_repair.c @@ -338,7 +338,7 @@ xrep_cntbt_extent_cmp( } /* - * Sort the free extents by length so so that we can put the records into the + * Sort the free extents by length so that we can put the records into the * cntbt in the correct order. Don't let userspace kill us if we're resorting * after allocating btree blocks. */ diff --git a/fs/xfs/scrub/dirtree.c b/fs/xfs/scrub/dirtree.c index 9b0ab2316612..887383d2e941 100644 --- a/fs/xfs/scrub/dirtree.c +++ b/fs/xfs/scrub/dirtree.c @@ -1021,7 +1021,7 @@ xchk_dirtree( return error; } -/* Does the directory targetted by this scrub have no parents? */ +/* Does the directory targeted by this scrub have no parents? */ bool xchk_dirtree_parentless(const struct xchk_dirtree *dl) { diff --git a/fs/xfs/scrub/reap.c b/fs/xfs/scrub/reap.c index f698b9be3dd1..0dfe61bafc6a 100644 --- a/fs/xfs/scrub/reap.c +++ b/fs/xfs/scrub/reap.c @@ -172,7 +172,7 @@ static inline bool xreap_is_dirty(const struct xreap_state *rs) } /* - * Decide if we need to roll the transaction to clear out the the log + * Decide if we need to roll the transaction to clear out the log * reservation that we allocated to buffer invalidations. */ static inline bool xreap_want_binval_roll(const struct xreap_state *rs) diff --git a/fs/xfs/xfs_bmap_item.c b/fs/xfs/xfs_bmap_item.c index 89f6e79a955f..aa5b41629747 100644 --- a/fs/xfs/xfs_bmap_item.c +++ b/fs/xfs/xfs_bmap_item.c @@ -339,7 +339,7 @@ xfs_bmap_update_get_group( /* * Bump the intent count on behalf of the deferred rmap and refcount - * intent items that that we can queue when we finish this bmap work. + * intent items that we can queue when we finish this bmap work. * This new intent item will bump the intent count before the bmap * intent drops the intent count, ensuring that the intent count * remains nonzero across the transaction roll. diff --git a/fs/xfs/xfs_inode.c b/fs/xfs/xfs_inode.c index 030a7c8f2c12..621513d7215e 100644 --- a/fs/xfs/xfs_inode.c +++ b/fs/xfs/xfs_inode.c @@ -2669,7 +2669,7 @@ xfs_irele( } /* - * Ensure all commited transactions touching the inode are written to the log. + * Ensure all committed transactions touching the inode are written to the log. */ int xfs_log_force_inode( diff --git a/fs/xfs/xfs_log_cil.c b/fs/xfs/xfs_log_cil.c index f9e07a32f60f..9446ac44ba88 100644 --- a/fs/xfs/xfs_log_cil.c +++ b/fs/xfs/xfs_log_cil.c @@ -1370,7 +1370,7 @@ xlog_cil_cleanup_whiteouts( * allocation context. However, we do not want to block on memory reclaim * recursing back into the filesystem because this push may have been triggered * by memory reclaim itself. Hence we really need to run under full GFP_NOFS - * contraints here. + * constraints here. */ static void xlog_cil_push_work( diff --git a/fs/xfs/xfs_platform.h b/fs/xfs/xfs_platform.h index 5d542e95fe44..745d715b4c64 100644 --- a/fs/xfs/xfs_platform.h +++ b/fs/xfs/xfs_platform.h @@ -153,7 +153,7 @@ static inline void delay(long ticks) /* * XFS wrapper structure for sysfs support. It depends on external data * structures and is embedded in various internal data structures to implement - * the XFS sysfs object heirarchy. Define it here for broad access throughout + * the XFS sysfs object hierarchy. Define it here for broad access throughout * the codebase. */ struct xfs_kobj { diff --git a/fs/xfs/xfs_zone_alloc.c b/fs/xfs/xfs_zone_alloc.c index d7ec055a9a06..b75cf3bfe33c 100644 --- a/fs/xfs/xfs_zone_alloc.c +++ b/fs/xfs/xfs_zone_alloc.c @@ -828,7 +828,7 @@ xfs_get_cached_zone( } /* - * Stash our zone in the inode so that is is reused for future allocations. + * Stash our zone in the inode so that it is reused for future allocations. * * The open_zone structure will be pinned until either the inode is freed or * until the cached open zone is replaced with a different one because the diff --git a/fs/xfs/xfs_zone_gc.c b/fs/xfs/xfs_zone_gc.c index 5fdcf98a2133..54b70ed2922f 100644 --- a/fs/xfs/xfs_zone_gc.c +++ b/fs/xfs/xfs_zone_gc.c @@ -46,7 +46,7 @@ * before remapping. * * Once a zone does not contain any valid data, be that through GC or user - * block removal, it is queued for for a zone reset. The reset operation + * block removal, it is queued for a zone reset. The reset operation * carefully ensures that the RT device cache is flushed and all transactions * referencing the rmap have been committed to disk. */ From 82e47533221d4746947b74d2e79a478c36c6433a Mon Sep 17 00:00:00 2001 From: Niklas Cassel Date: Fri, 4 Sep 2026 15:43:11 +0200 Subject: [PATCH 232/592] ata: libahci: clear PxCLBU and PxFBU for AHCI_HFLAG_32BIT_ONLY A user reported that commit 105c42566a55 ("ata: ahci: force 32-bit DMA for JMicron JMB582/JMB585") made the JMicron JMB585 unusable on his board. The failure is seen as soon as the ahci driver is probed, and booting with iommu=off does not solve the problem. Looking at the AHCI specification, PxCLBU and PxFBU are both read only '0' for HBAs that do not support 64-bit addressing. For HBAs that do support 64-bit addressing, the registers are read write, with a reset value that is Implementation Specific. When using the AHCI_HFLAG_32BIT_ONLY flag, the HBA does support 64-bit addressing, and a 32-bit DMA mask is set by simply clearing HOST_CAP_64. Thus, in this case, we need to explicitly clear the registers to 0. Fixes: 105c42566a55 ("ata: ahci: force 32-bit DMA for JMicron JMB582/JMB585") Fixes: c7a42156d99b ("ahci: disable 64bit dma on sb600") Cc: stable@vger.kernel.org Reported-by: Roland Waltersson Closes: https://lore.kernel.org/linux-ide/IA0PR17MB668730A4ECCD65F7A1DC3EDC9EB62@IA0PR17MB6687.namprd17.prod.outlook.com/ Reviewed-by: Damien Le Moal Link: https://lore.kernel.org/r/20260904134310.1465051-2-cassel@kernel.org Signed-off-by: Niklas Cassel --- drivers/ata/libahci.c | 15 ++++++++++++++- 1 file changed, 14 insertions(+), 1 deletion(-) diff --git a/drivers/ata/libahci.c b/drivers/ata/libahci.c index 6d72eb017b49..08ab56e95566 100644 --- a/drivers/ata/libahci.c +++ b/drivers/ata/libahci.c @@ -744,15 +744,28 @@ void ahci_start_fis_rx(struct ata_port *ap) struct ahci_port_priv *pp = ap->private_data; u32 tmp; - /* set FIS registers */ + /* + * On HBAs that only support 32-bit addressing PxCLBU is read only '0'. + * When applying the AHCI_HFLAG_32BIT_ONLY quirk, PxCLBU is RW, and the + * reset value is Implementation Specific, so we need to clear it to 0. + */ if (hpriv->cap & HOST_CAP_64) writel((pp->cmd_slot_dma >> 16) >> 16, port_mmio + PORT_LST_ADDR_HI); + else if (hpriv->flags & AHCI_HFLAG_32BIT_ONLY) + writel(0, port_mmio + PORT_LST_ADDR_HI); writel(pp->cmd_slot_dma & 0xffffffff, port_mmio + PORT_LST_ADDR); + /* + * On HBAs that only support 32-bit addressing PxFBU is read only '0'. + * When applying the AHCI_HFLAG_32BIT_ONLY quirk, PxFBU is RW, and the + * reset value is Implementation Specific, so we need to clear it to 0. + */ if (hpriv->cap & HOST_CAP_64) writel((pp->rx_fis_dma >> 16) >> 16, port_mmio + PORT_FIS_ADDR_HI); + else if (hpriv->flags & AHCI_HFLAG_32BIT_ONLY) + writel(0, port_mmio + PORT_FIS_ADDR_HI); writel(pp->rx_fis_dma & 0xffffffff, port_mmio + PORT_FIS_ADDR); /* enable FIS reception */ From ba13f1f32d96d7ce9069ae4f32404ecde8da89c4 Mon Sep 17 00:00:00 2001 From: Meijing Zhao Date: Tue, 15 Sep 2026 20:46:41 +0800 Subject: [PATCH 233/592] mm: memblock: add missing HugeTLB flag name Commit 7d163a75f821 ("memblock: make HugeTLB bootmem allocation work with KHO") added MEMBLOCK_RSRV_HUGETLB but did not add the corresponding entry to flagname[]. As a result, memblock debugfs cannot report the flag by name. Add the missing RSV_HUGETLB entry. Fixes: 7d163a75f821 ("memblock: make HugeTLB bootmem allocation work with KHO") Signed-off-by: Meijing Zhao Signed-off-by: Wandun Chen Link: https://lore.kernel.org/lkml/20260821020910.3428585-2-zhaomeijing100@gmail.com/ Link: https://patch.msgid.link/20260915124641.2498280-1-chenwandun1@gmail.com Signed-off-by: Mike Rapoport (Microsoft) --- mm/memblock.c | 1 + 1 file changed, 1 insertion(+) diff --git a/mm/memblock.c b/mm/memblock.c index 021db49eb7fc..e6fdf95a0534 100644 --- a/mm/memblock.c +++ b/mm/memblock.c @@ -2886,6 +2886,7 @@ static const char * const flagname[] = { [ilog2(MEMBLOCK_RSRV_NOINIT)] = "RSV_NIT", [ilog2(MEMBLOCK_RSRV_KERN)] = "RSV_KERN", [ilog2(MEMBLOCK_KHO_SCRATCH)] = "KHO_SCRATCH", + [ilog2(MEMBLOCK_RSRV_HUGETLB)] = "RSV_HUGETLB", }; static int memblock_debug_show(struct seq_file *m, void *private) From f4fafaf02174c32bce2f9bb4196fadf13f1fd96e Mon Sep 17 00:00:00 2001 From: Namjae Jeon Date: Wed, 9 Sep 2026 09:58:22 +0900 Subject: [PATCH 234/592] ksmbd: fix partial normalized name responses Windows may request FILE_NORMALIZED_NAME_INFORMATION with an output buffer that only fits the fixed portion of the variable-length response. Treat the fixed portion as FILE_NORMALIZED_NAME_INFORMATION_SIZE so ksmbd returns STATUS_BUFFER_OVERFLOW instead of STATUS_INFO_LENGTH_MISMATCH. This avoids rejecting valid partial normalized-name responses. Fixes: 6b8b79226bc3 ("ksmbd: fix partial file information responses") Reported-by: Mobin Aydinfar Reviewed-by: ChenXiaoSong Signed-off-by: Namjae Jeon --- fs/smb/server/smb2pdu.c | 3 +++ fs/smb/server/smb2pdu.h | 1 + 2 files changed, 4 insertions(+) diff --git a/fs/smb/server/smb2pdu.c b/fs/smb/server/smb2pdu.c index 8acc5174530b..0436b7c898b1 100644 --- a/fs/smb/server/smb2pdu.c +++ b/fs/smb/server/smb2pdu.c @@ -7312,6 +7312,9 @@ static int smb2_get_info_file(struct ksmbd_work *work, case FILE_ALTERNATE_NAME_INFORMATION: fixed_len = FILE_ALTERNATE_NAME_INFORMATION_SIZE; break; + case FILE_NORMALIZED_NAME_INFORMATION: + fixed_len = FILE_NORMALIZED_NAME_INFORMATION_SIZE; + break; case FILE_STREAM_INFORMATION: fixed_len = FILE_STREAM_INFORMATION_SIZE; break; diff --git a/fs/smb/server/smb2pdu.h b/fs/smb/server/smb2pdu.h index a6200d8630e2..ca8e27f7b712 100644 --- a/fs/smb/server/smb2pdu.h +++ b/fs/smb/server/smb2pdu.h @@ -212,6 +212,7 @@ struct file_sparse { #define FILE_ALLOCATION_INFORMATION_SIZE 19 #define FILE_END_OF_FILE_INFORMATION_SIZE 20 #define FILE_ALTERNATE_NAME_INFORMATION_SIZE 8 +#define FILE_NORMALIZED_NAME_INFORMATION_SIZE 8 #define FILE_STREAM_INFORMATION_SIZE 32 #define FILE_PIPE_INFORMATION_SIZE 23 #define FILE_PIPE_LOCAL_INFORMATION_SIZE 24 From 9fa26285ae70ac2d3d1b47459a6b4463ab053e1c Mon Sep 17 00:00:00 2001 From: Namjae Jeon Date: Wed, 9 Sep 2026 09:58:48 +0900 Subject: [PATCH 235/592] ksmbd: keep compound responses on query info errors Do not reset the RFC1002 length of the complete response when a query info buffer is too small. The current command will add its error response through ksmbd_iov_pin_rsp(), while resetting the base length can truncate earlier responses in a compound request. This lets ksmbd return the earlier responses and the query-info error response together. Remove the now-unused rsp_org parameter from the pipe query-info helpers. Fixes: e2b76ab8b5c9 ("ksmbd: add support for read compound") Reported-by: Mobin Aydinfar Reviewed-by: ChenXiaoSong Signed-off-by: Namjae Jeon --- fs/smb/server/smb2pdu.c | 31 ++++++++++++------------------- 1 file changed, 12 insertions(+), 19 deletions(-) diff --git a/fs/smb/server/smb2pdu.c b/fs/smb/server/smb2pdu.c index 0436b7c898b1..6b8809f67b92 100644 --- a/fs/smb/server/smb2pdu.c +++ b/fs/smb/server/smb2pdu.c @@ -6275,21 +6275,18 @@ int smb2_query_dir(struct ksmbd_work *work) * @reqOutputBufferLength: max buffer length expected in command response * @fixed_len: minimum fixed response length * @rsp: query info response buffer contains output buffer length - * @rsp_org: base response buffer pointer in case of chained response * * Return: 0 on success, otherwise error */ static int buffer_check_err(int reqOutputBufferLength, unsigned int fixed_len, - struct smb2_query_info_rsp *rsp, - void *rsp_org) + struct smb2_query_info_rsp *rsp) { unsigned int output_len = le32_to_cpu(rsp->OutputBufferLength); if (reqOutputBufferLength < fixed_len) { pr_err("Invalid Buffer Size Requested\n"); rsp->hdr.Status = STATUS_INFO_LENGTH_MISMATCH; - *(__be32 *)rsp_org = cpu_to_be32(sizeof(struct smb2_hdr)); return -EINVAL; } @@ -6300,8 +6297,7 @@ static int buffer_check_err(int reqOutputBufferLength, return 0; } -static void get_standard_info_pipe(struct smb2_query_info_rsp *rsp, - void *rsp_org) +static void get_standard_info_pipe(struct smb2_query_info_rsp *rsp) { struct smb2_file_standard_info *sinfo; @@ -6316,8 +6312,7 @@ static void get_standard_info_pipe(struct smb2_query_info_rsp *rsp, cpu_to_le32(sizeof(struct smb2_file_standard_info)); } -static void get_internal_info_pipe(struct smb2_query_info_rsp *rsp, u64 num, - void *rsp_org) +static void get_internal_info_pipe(struct smb2_query_info_rsp *rsp, u64 num) { struct smb2_file_internal_info *file_info; @@ -6331,8 +6326,7 @@ static void get_internal_info_pipe(struct smb2_query_info_rsp *rsp, u64 num, static int smb2_get_info_file_pipe(struct ksmbd_session *sess, struct smb2_query_info_req *req, - struct smb2_query_info_rsp *rsp, - void *rsp_org) + struct smb2_query_info_rsp *rsp) { u64 id; int rc; @@ -6357,16 +6351,16 @@ static int smb2_get_info_file_pipe(struct ksmbd_session *sess, switch (req->FileInfoClass) { case FILE_STANDARD_INFORMATION: - get_standard_info_pipe(rsp, rsp_org); + get_standard_info_pipe(rsp); rc = buffer_check_err(le32_to_cpu(req->OutputBufferLength), le32_to_cpu(rsp->OutputBufferLength), - rsp, rsp_org); + rsp); break; case FILE_INTERNAL_INFORMATION: - get_internal_info_pipe(rsp, id, rsp_org); + get_internal_info_pipe(rsp, id); rc = buffer_check_err(le32_to_cpu(req->OutputBufferLength), le32_to_cpu(rsp->OutputBufferLength), - rsp, rsp_org); + rsp); break; default: ksmbd_debug(SMB, "smb2_info_file_pipe for %u not supported\n", @@ -7202,8 +7196,7 @@ static int smb2_get_info_file(struct ksmbd_work *work, if (test_share_config_flag(work->tcon->share_conf, KSMBD_SHARE_FLAG_PIPE)) { /* smb2 info file called for pipe */ - rc = smb2_get_info_file_pipe(work->sess, req, rsp, - work->response_buf); + rc = smb2_get_info_file_pipe(work->sess, req, rsp); goto iov_pin_out; } @@ -7321,7 +7314,7 @@ static int smb2_get_info_file(struct ksmbd_work *work, } rc = buffer_check_err(le32_to_cpu(req->OutputBufferLength), fixed_len, - rsp, work->response_buf); + rsp); } ksmbd_fd_put(work, fp); @@ -7592,7 +7585,7 @@ static int smb2_get_info_filesystem(struct ksmbd_work *work, } rc = buffer_check_err(le32_to_cpu(req->OutputBufferLength), fixed_len, - rsp, work->response_buf); + rsp); path_put(&path); if (!rc) @@ -7706,7 +7699,7 @@ static int smb2_get_info_sec(struct ksmbd_work *work, rsp->OutputBufferLength = cpu_to_le32(secdesclen); rc = buffer_check_err(le32_to_cpu(req->OutputBufferLength), le32_to_cpu(rsp->OutputBufferLength), - rsp, work->response_buf); + rsp); if (rc) goto err_out; From 075bc7b1d3dde5ed43fbaabbc1a69f09b7fc3a47 Mon Sep 17 00:00:00 2001 From: Baineng Shou Date: Thu, 10 Sep 2026 10:16:52 +0800 Subject: [PATCH 236/592] dmaengine: mmp_pdma: fix wrong sg length in mmp_pdma_prep_slave_sg() In mmp_pdma_prep_slave_sg(), for_each_sg() iterates the scatterlist putting each entry into 'sg', but the entry length is read from 'sgl' (the list head) instead of 'sg' (the current entry): for_each_sg(sgl, sg, sg_len, i) { addr = sg_dma_address(sg); avail = sg_dma_len(sgl); /* should be 'sg' */ Consequently 'avail' is always the length of the first entry. For multi-sg lists this causes out-of-bounds reads when a later entry is shorter than the first, and silent data loss when it is longer. Single-sg or uniformly-sized lists happen to mask the issue. Fixes: c8acd6aa6bed3 ("dmaengine: mmp-pdma support") Signed-off-by: Baineng Shou Reviewed-by: Frank Li Link: https://patch.msgid.link/20260910021652.1296640-1-shoubaineng@gmail.com Signed-off-by: Vinod Koul --- drivers/dma/mmp_pdma.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/drivers/dma/mmp_pdma.c b/drivers/dma/mmp_pdma.c index 78e3e07e681d..ed520737882b 100644 --- a/drivers/dma/mmp_pdma.c +++ b/drivers/dma/mmp_pdma.c @@ -712,7 +712,7 @@ mmp_pdma_prep_slave_sg(struct dma_chan *dchan, struct scatterlist *sgl, for_each_sg(sgl, sg, sg_len, i) { addr = sg_dma_address(sg); - avail = sg_dma_len(sgl); + avail = sg_dma_len(sg); do { len = min_t(size_t, avail, PDMA_MAX_DESC_BYTES); From 095858324f063dba830041f067872f0a08765d2f Mon Sep 17 00:00:00 2001 From: Hao-Qun Huang Date: Sun, 13 Sep 2026 03:20:49 +0800 Subject: [PATCH 237/592] spi: virtio: Use the per-transfer bits per word virtio_spi_transfer_one() puts spi->bits_per_word into the request header, so a transfer that sets its own word size reaches the backend with the device default instead. The SPI core has already copied that default into xfer->bits_per_word when the transfer leaves it at zero, the same way it does for xfer->speed_hz, which this function already uses. Per-transfer word sizes are ordinary SPI usage. mipi_dbi, for one, sends a 9-bit command and reads the reply as 8-bit data in the same message. With a 16-bit device default, a one-byte transfer asking for 8 bits goes out as a partial 16-bit word, which the backend may reject. Fixes: f98cabe3f6cf ("SPI: Add virtio SPI driver") Cc: stable@vger.kernel.org Assisted-by: LLM Signed-off-by: Hao-Qun Huang Link: https://patch.msgid.link/20260913032049.11209.alvinhuang0603@gmail.com Signed-off-by: Mark Brown --- drivers/spi/spi-virtio.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/drivers/spi/spi-virtio.c b/drivers/spi/spi-virtio.c index 2256dfec5407..3e181bd8bc94 100644 --- a/drivers/spi/spi-virtio.c +++ b/drivers/spi/spi-virtio.c @@ -168,7 +168,7 @@ static int virtio_spi_transfer_one(struct spi_controller *ctrl, /* Fill struct spi_transfer_head */ th->chip_select_id = spi_get_chipselect(spi, 0); - th->bits_per_word = spi->bits_per_word; + th->bits_per_word = xfer->bits_per_word; th->cs_change = xfer->cs_change; th->tx_nbits = xfer->tx_nbits; th->rx_nbits = xfer->rx_nbits; From 8861db305103107199b1426f25fde1fb6d465583 Mon Sep 17 00:00:00 2001 From: =?UTF-8?q?Christian=20G=C3=B6ttsche?= Date: Thu, 3 Sep 2026 13:43:38 +0200 Subject: [PATCH 238/592] selinux: always fill AVC decision in avc_has_perm_noaudit() MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit avc_has_perm_noaudit() is documented to return a copy of the access decision in @avd, but its early return for an empty requested permission set leaves the buffer untouched. All callers pass an uninitialized stack variable and afterwards feed it to avc_audit(), and the inode hook even stores it in the per-task decision cache. Fill in a deny-all, audit-all decision, similar to avd_init(), so every caller receives a defined value at no cost on the hot path. Cc: stable@vger.kernel.org Fixes: e6f2f381e4015386 ("selinux: replace BUG_ONs with WARN_ONs in avc.c") Signed-off-by: Christian Göttsche Reviewed-by: Stephen Smalley Signed-off-by: Paul Moore --- security/selinux/avc.c | 5 ++++- 1 file changed, 4 insertions(+), 1 deletion(-) diff --git a/security/selinux/avc.c b/security/selinux/avc.c index a9401d6c2e5f..560a82c6d682 100644 --- a/security/selinux/avc.c +++ b/security/selinux/avc.c @@ -1149,8 +1149,11 @@ inline int avc_has_perm_noaudit(u32 ssid, u32 tsid, u32 denied; struct avc_node *node; - if (WARN_ON(!requested)) + if (WARN_ON(!requested)) { + /* Provide a deny-all, audit-all decision to the caller. */ + *avd = (struct av_decision){ .auditdeny = 0xffffffff }; return -EACCES; + } rcu_read_lock(); node = avc_lookup(ssid, tsid, tclass); From 8c0c602202b9a4909b00bc3354e3c0355bc69e65 Mon Sep 17 00:00:00 2001 From: Karl Mehltretter Date: Sat, 29 Aug 2026 23:32:55 +0200 Subject: [PATCH 239/592] selinux: preserve user SID across nested backing files SELinux saves the user file SID in a backing-file security blob so it remains available after mmap() replaces vma->vm_file with a backing file. For nested backing files (overlayfs over overlayfs, or FUSE passthrough backed by overlayfs), user_file may itself be a backing file. Its fsec->sid is the SID of the mounter that opened it, rather than the user that opened the top-level file. mprotect() then checks fd { use } against the mounter SID. This can incorrectly deny access without a domain transition, or check the wrong target SID after one. Copy the saved user SID when user_file is a backing file. Keep using the regular file SID for the first backing layer. With two nested overlayfs mounts and SELinux enforcing, mprotect(PROT_READ) returns EACCES with an fd { use } denial against the mounter SID. With this change, mprotect() succeeds. Tested on arm64 QEMU with a small BusyBox initramfs and a purpose-built SELinux policy. The original test was also repeated with Fedora Cloud Base 44 userspace and gave the same result. Cc: stable@vger.kernel.org Fixes: 82544d36b172 ("selinux: fix overlayfs mmap() and mprotect() access checks") Assisted-by: LLM Signed-off-by: Karl Mehltretter Reviewed-by: Amir Goldstein Reviewed-by: Stephen Smalley Signed-off-by: Paul Moore --- security/selinux/hooks.c | 9 ++++++++- security/selinux/include/objsec.h | 2 +- 2 files changed, 9 insertions(+), 2 deletions(-) diff --git a/security/selinux/hooks.c b/security/selinux/hooks.c index e5e17f100aae..b0bb5f8c90bf 100644 --- a/security/selinux/hooks.c +++ b/security/selinux/hooks.c @@ -3843,13 +3843,20 @@ static int selinux_file_alloc_security(struct file *file) return 0; } +static inline u32 selinux_file_user_sid(const struct file *file) +{ + if (unlikely(file->f_mode & FMODE_BACKING)) + return selinux_backing_file(file)->uf_sid; + return selinux_file(file)->sid; +} + static int selinux_backing_file_alloc(struct file *backing_file, const struct file *user_file) { struct backing_file_security_struct *bfsec; bfsec = selinux_backing_file(backing_file); - bfsec->uf_sid = selinux_file(user_file)->sid; + bfsec->uf_sid = selinux_file_user_sid(user_file); return 0; } diff --git a/security/selinux/include/objsec.h b/security/selinux/include/objsec.h index 3c0a16ec978b..853f7266ed18 100644 --- a/security/selinux/include/objsec.h +++ b/security/selinux/include/objsec.h @@ -87,7 +87,7 @@ struct file_security_struct { }; struct backing_file_security_struct { - u32 uf_sid; /* associated user file fsec->sid */ + u32 uf_sid; /* top-level user file fsec->sid */ }; struct superblock_security_struct { From 78fc54b934bfb2c18aad8154c7302067146946f9 Mon Sep 17 00:00:00 2001 From: Karl Mehltretter Date: Sat, 29 Aug 2026 23:32:56 +0200 Subject: [PATCH 240/592] selinux: recheck intermediate backing files on mprotect() mprotect() can be used to bypass the SELinux checks that mmap() performs against the intermediate layers of a stacked filesystem. mmap() checks every backing layer as the request descends through the stack. mprotect() only has the lowest backing file in vma->vm_file, so it rechecks the top-level user and the lowest mounter, but skips the mounters of every layer in between. With two nested overlayfs mounts and a policy denying mounter_t -> middle_file_t:file { execute }, a direct mmap(PROT_EXEC) is denied: avc: denied { execute } for pid=71 comm="nested_exec" path="/payload" dev="overlay" ino=9 scontext=user_u:base_r:mounter_t tcontext=user_u:object_r:middle_file_t tclass=file permissive=0 while mmap(PROT_NONE) followed by mprotect(PROT_EXEC) succeeds. Preserve each intermediate path, mounter SID and file-description SID in the backing-file security blob, copying the saved entries when another backing layer is opened. Allocate the array only for nested backing files, and release it and the path references in the backing_file_free hook. During mprotect(), recheck fd { use } and the requested inode permissions for every saved mounter, and include the intermediate layers in the execmod checks. Policy for nested stacking may then need to grant intermediate mounters what a direct mmap() already requires, and execmod on intermediate labels for binaries using text relocations. Tested on arm64 QEMU with a small BusyBox initramfs and a purpose-built SELinux policy, on a mainline tree containing commit f2381b546e7e ("fs: fix user path of nested backing files"). Cc: stable@vger.kernel.org Fixes: 82544d36b172 ("selinux: fix overlayfs mmap() and mprotect() access checks") Assisted-by: LLM Signed-off-by: Karl Mehltretter Reviewed-by: Stephen Smalley [PM: subject tweak] Signed-off-by: Paul Moore --- security/selinux/hooks.c | 149 ++++++++++++++++++++++++++---- security/selinux/include/objsec.h | 8 ++ 2 files changed, 137 insertions(+), 20 deletions(-) diff --git a/security/selinux/hooks.c b/security/selinux/hooks.c index b0bb5f8c90bf..3f4a6ddee322 100644 --- a/security/selinux/hooks.c +++ b/security/selinux/hooks.c @@ -1674,24 +1674,30 @@ static int cred_has_capability(const struct cred *cred, return rc; } -/* Check whether a task has a particular permission to an inode. - The 'adp' parameter is optional and allows other audit - data to be passed (e.g. the dentry). */ +/* + * Check whether a SID has a particular permission to an inode. The 'adp' + * parameter is optional and allows other audit data to be passed (e.g. the + * dentry). + */ +static int inode_sid_has_perm(u32 sid, struct inode *inode, u32 perms, + struct common_audit_data *adp) +{ + struct inode_security_struct *isec; + + if (unlikely(IS_PRIVATE(inode))) + return 0; + + isec = selinux_inode(inode); + + return avc_has_perm(sid, isec->sid, isec->sclass, perms, adp); +} + static int inode_has_perm(const struct cred *cred, struct inode *inode, u32 perms, struct common_audit_data *adp) { - struct inode_security_struct *isec; - u32 sid; - - if (unlikely(IS_PRIVATE(inode))) - return 0; - - sid = cred_sid(cred); - isec = selinux_inode(inode); - - return avc_has_perm(sid, isec->sid, isec->sclass, perms, adp); + return inode_sid_has_perm(cred_sid(cred), inode, perms, adp); } /* Same as inode_has_perm, but pass explicit audit data containing @@ -3854,13 +3860,63 @@ static int selinux_backing_file_alloc(struct file *backing_file, const struct file *user_file) { struct backing_file_security_struct *bfsec; + const struct backing_file_security_struct *ubfsec; + struct backing_file_security_layer *layer; + u32 i; bfsec = selinux_backing_file(backing_file); bfsec->uf_sid = selinux_file_user_sid(user_file); + if (!(user_file->f_mode & FMODE_BACKING)) + return 0; + + ubfsec = selinux_backing_file(user_file); + /* a wrapped count would make kmalloc_array() return ZERO_SIZE_PTR */ + if (unlikely(ubfsec->layer_count == U32_MAX)) + return -EOVERFLOW; + + /* + * The final VMA only retains the lowest backing file, so record the + * whole chain here rather than in the mmap hook, where concurrent + * mappings would have to be serialized. Size it dynamically: erofs + * inode sharing adds a backing file without bumping s_stack_depth. + */ + bfsec->layers = kmalloc_array(ubfsec->layer_count + 1, + sizeof(*bfsec->layers), GFP_KERNEL); + if (!bfsec->layers) + return -ENOMEM; + + for (i = 0; i < ubfsec->layer_count; i++) { + layer = &bfsec->layers[i]; + *layer = ubfsec->layers[i]; + path_get(&layer->path); + } + + /* f_path, not file_user_path(): this layer, not the top-level file */ + layer = &bfsec->layers[i]; + layer->path = user_file->f_path; + layer->mounter_sid = cred_sid(user_file->f_cred); + layer->fd_sid = selinux_file(user_file)->sid; + path_get(&layer->path); + bfsec->layer_count = ubfsec->layer_count + 1; return 0; } +static void selinux_backing_file_free(struct file *backing_file) +{ + struct backing_file_security_struct *bfsec; + + /* security_backing_file_free() may be called twice after an error */ + if (!backing_file_security(backing_file)) + return; + + bfsec = selinux_backing_file(backing_file); + while (bfsec->layer_count) + path_put(&bfsec->layers[--bfsec->layer_count].path); + kfree(bfsec->layers); + bfsec->layers = NULL; +} + /* * Check whether a task has the ioctl permission and cmd * operation to an inode. @@ -3978,6 +4034,53 @@ static int selinux_file_ioctl_compat(struct file *file, unsigned int cmd, static int default_noexec __ro_after_init; +static u32 file_map_prot_to_av(unsigned long prot, bool shared) +{ + u32 av = FILE__READ; + + if (shared && (prot & PROT_WRITE)) + av |= FILE__WRITE; + if (prot & PROT_EXEC) + av |= FILE__EXECUTE; + + return av; +} + +static int backing_mounters_has_perm(const struct file *file, u32 av) +{ + const struct backing_file_security_struct *bfsec; + const struct backing_file_security_layer *layer; + struct common_audit_data ad; + struct inode *inode; + u32 i; + int rc; + + if (WARN_ON_ONCE(!(file->f_mode & FMODE_BACKING))) + return -EIO; + + bfsec = selinux_backing_file(file); + for (i = 0; i < bfsec->layer_count; i++) { + layer = &bfsec->layers[i]; + inode = d_inode(layer->path.dentry); + + ad.type = LSM_AUDIT_DATA_PATH; + ad.u.path = layer->path; + + if (layer->mounter_sid != layer->fd_sid) { + rc = avc_has_perm(layer->mounter_sid, layer->fd_sid, + SECCLASS_FD, FD__USE, &ad); + if (rc) + return rc; + } + + rc = inode_sid_has_perm(layer->mounter_sid, inode, av, &ad); + if (rc) + return rc; + } + + return 0; +} + static int __file_map_prot_check(const struct file *file, unsigned long prot, bool shared, bool mounter_check, bool bf_user_file) @@ -4011,14 +4114,10 @@ static int __file_map_prot_check(const struct file *file, unsigned long prot, if (file) { const struct cred *cred = mounter_check ? file->f_cred : current_cred(); - /* "read" always possible, "write" only if shared */ - u32 av = FILE__READ; - if (shared && prot_write) - av |= FILE__WRITE; - if (prot_exec) - av |= FILE__EXECUTE; - return __file_has_perm(cred, file, av, bf_user_file); + return __file_has_perm(cred, file, + file_map_prot_to_av(prot, shared), + bf_user_file); } return 0; @@ -4113,6 +4212,7 @@ static int selinux_file_mprotect(struct vm_area_struct *vma, int rc; const struct cred *cred = current_cred(); u32 sid = cred_sid(cred); + u32 av; const struct file *file = vma->vm_file; bool backing_file; bool shared = vma->vm_flags & VM_SHARED; @@ -4156,6 +4256,10 @@ static int selinux_file_mprotect(struct vm_area_struct *vma, if (rc) return rc; if (backing_file) { + rc = backing_mounters_has_perm(file, + FILE__EXECMOD); + if (rc) + return rc; rc = file_has_perm(file->f_cred, file, FILE__EXECMOD); if (rc) @@ -4168,6 +4272,10 @@ static int selinux_file_mprotect(struct vm_area_struct *vma, if (rc) return rc; if (backing_file) { + av = file_map_prot_to_av(prot, shared); + rc = backing_mounters_has_perm(file, av); + if (rc) + return rc; rc = file_map_prot_check(file, prot, shared, true); if (rc) return rc; @@ -7626,6 +7734,7 @@ static struct security_hook_list selinux_hooks[] __ro_after_init = { LSM_HOOK_INIT(file_permission, selinux_file_permission), LSM_HOOK_INIT(file_alloc_security, selinux_file_alloc_security), LSM_HOOK_INIT(backing_file_alloc, selinux_backing_file_alloc), + LSM_HOOK_INIT(backing_file_free, selinux_backing_file_free), LSM_HOOK_INIT(file_ioctl, selinux_file_ioctl), LSM_HOOK_INIT(file_ioctl_compat, selinux_file_ioctl_compat), LSM_HOOK_INIT(mmap_file, selinux_mmap_file), diff --git a/security/selinux/include/objsec.h b/security/selinux/include/objsec.h index 853f7266ed18..2f21568251ff 100644 --- a/security/selinux/include/objsec.h +++ b/security/selinux/include/objsec.h @@ -86,8 +86,16 @@ struct file_security_struct { u32 pseqno; /* Policy seqno at the time of file open */ }; +struct backing_file_security_layer { + struct path path; /* this layer's real path */ + u32 mounter_sid; /* SID of the mounter that opened it */ + u32 fd_sid; /* SID of its open file description */ +}; + struct backing_file_security_struct { u32 uf_sid; /* top-level user file fsec->sid */ + u32 layer_count; /* number of intermediate backing files */ + struct backing_file_security_layer *layers; }; struct superblock_security_struct { From e0c3e9d76adbe522dd420a766ce42d03ce887c29 Mon Sep 17 00:00:00 2001 From: Guangshuo Li Date: Mon, 14 Sep 2026 17:15:44 +0800 Subject: [PATCH 241/592] i2c: imx: disable autosuspend on remove i2c_imx_probe() enables runtime PM autosuspend with pm_runtime_use_autosuspend(). The probe error path correctly undoes this setting with pm_runtime_dont_use_autosuspend(), but the normal remove path only disables runtime PM. The runtime PM API requires pm_runtime_use_autosuspend() to be undone with pm_runtime_dont_use_autosuspend() at driver exit unless runtime PM was enabled with devm_pm_runtime_enable(). Leaving the autosuspend flag set therefore leaves the runtime PM state incompletely cleaned up after the driver is unbound. Add the missing pm_runtime_dont_use_autosuspend() call to the remove path. This issue was found by manual code inspection. Fixes: 588eb93ea49f ("i2c: imx: add runtime pm support to improve the performance") Signed-off-by: Guangshuo Li Cc: # v4.5+ Reviewed-by: Frank Li Signed-off-by: Andi Shyti Link: https://patch.msgid.link/20260914091544.1667137-1-lgs201920130244@gmail.com --- drivers/i2c/busses/i2c-imx.c | 1 + 1 file changed, 1 insertion(+) diff --git a/drivers/i2c/busses/i2c-imx.c b/drivers/i2c/busses/i2c-imx.c index 19ec056b00af..4dc3df6a41a4 100644 --- a/drivers/i2c/busses/i2c-imx.c +++ b/drivers/i2c/busses/i2c-imx.c @@ -1920,6 +1920,7 @@ static void i2c_imx_remove(struct platform_device *pdev) pm_runtime_put_noidle(&pdev->dev); pm_runtime_disable(&pdev->dev); + pm_runtime_dont_use_autosuspend(&pdev->dev); } static int i2c_imx_runtime_suspend(struct device *dev) From 0d1cb83337f13af082afb68b28d3fdfe29cde7fb Mon Sep 17 00:00:00 2001 From: Wentao Liang Date: Tue, 15 Sep 2026 06:59:33 +0000 Subject: [PATCH 242/592] ata: libahci_platform: Fix device reference leak in ahci_platform_get_resources() of_find_device_by_node() takes a reference on the port platform device, which is only used to look up its port regulator and is never released, neither on success nor on the error paths. Drop the reference with put_device() once the regulator has been obtained, which covers both the success and error paths. Fixes: c7d7ddee7e24 ("ata: libahci: Allow using multiple regulators") Cc: stable@vger.kernel.org Signed-off-by: Wentao Liang Link: https://lore.kernel.org/r/20260915065933.1733061-1-vulab@iscas.ac.cn Reviewed-by: Damien Le Moal Signed-off-by: Niklas Cassel --- drivers/ata/libahci_platform.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/drivers/ata/libahci_platform.c b/drivers/ata/libahci_platform.c index 6e072d681341..f6524135ea11 100644 --- a/drivers/ata/libahci_platform.c +++ b/drivers/ata/libahci_platform.c @@ -620,10 +620,10 @@ struct ahci_host_priv *ahci_platform_get_resources(struct platform_device *pdev, of_platform_device_create(child, NULL, NULL); port_dev = of_find_device_by_node(child); - if (port_dev) { rc = ahci_platform_get_regulator(hpriv, port, &port_dev->dev); + put_device(&port_dev->dev); if (rc == -EPROBE_DEFER) goto err_out; } From ad34235808b63a70ca4989b7a2852923193d06ef Mon Sep 17 00:00:00 2001 From: Linkai Gong Date: Mon, 7 Sep 2026 15:11:02 +0800 Subject: [PATCH 243/592] i2c: atr: fix dangling adapter pointer on add failure i2c_atr_add_adapter() stores atr->adapter[chan_id] before i2c_add_adapter() so that the I2C bus notifier can match child clients during registration. On failure the channel is freed but the slot was left pointing at freed memory, which can lead to use-after-free in i2c_atr_del_adapter() / cleanup and also block reuse with -EEXIST. Clear the slot on the i2c_add_adapter() error path before freeing chan. Fixes: a076a860acae ("media: i2c: add I2C Address Translator (ATR) support") Signed-off-by: Linkai Gong Cc: # v6.6+ Reviewed-by: Andy Shevchenko Signed-off-by: Andi Shyti Link: https://patch.msgid.link/20260907071102.1080840-1-gonglinkai@kylinos.cn --- drivers/i2c/i2c-atr.c | 1 + 1 file changed, 1 insertion(+) diff --git a/drivers/i2c/i2c-atr.c b/drivers/i2c/i2c-atr.c index e6d2af659d81..ca29633dcd62 100644 --- a/drivers/i2c/i2c-atr.c +++ b/drivers/i2c/i2c-atr.c @@ -855,6 +855,7 @@ int i2c_atr_add_adapter(struct i2c_atr *atr, struct i2c_atr_adap_desc *desc) ret = i2c_add_adapter(&chan->adap); if (ret) { + atr->adapter[chan_id] = NULL; dev_err(dev, "failed to add atr-adapter %u (error=%d)\n", chan_id, ret); goto err_free_alias_pool; From 2ab510e63197360945f915dd5631a77c63ac6b27 Mon Sep 17 00:00:00 2001 From: Tvrtko Ursulin Date: Tue, 15 Sep 2026 16:05:57 +0100 Subject: [PATCH 244/592] drm/sched: Fix virtual runtime race MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit Prevent pushing a new job to an entity seeing it being the first in the queue, and hence entering the drm_sched_rq_add_entity() path, if the pop side in drm_sched_entity_pop_job() has just de-queued the job but not yet updated the saved virtual time. Restoring the unsaved virtual time, which is at this point not a delta but still an absolute value, pushes the said entity to the rear of the run queue for a potentially very long time. We close this race by pulling the locked sections out to encompass both the queue push/pop and corresponding rbtree management. This is aligned with the future direction to replace the current lockless job queue with one of the fully locked standard list primitives. Signed-off-by: Tvrtko Ursulin Fixes: 2fa4d8e2c109 ("drm/sched: Add fair scheduling policy") Suggested-by: Luke.Wildhardt@proton.me # via Claude Opus Tested-by: Luke.Wildhardt@proton.me Cc: Christian König Cc: Danilo Krummrich Cc: Philipp Stanner Cc: Pierre-Eric Pelloux-Prayer Cc: Matthew Brost Cc: Vitaly Prosyak Cc: stable@vger.kernel.org # v7.2+ [phasta: commit title] Signed-off-by: Philipp Stanner Link: https://patch.msgid.link/20260915150557.62847-1-tvrtko.ursulin@igalia.com --- drivers/gpu/drm/scheduler/sched_entity.c | 8 +++++++- drivers/gpu/drm/scheduler/sched_rq.c | 20 +++++++++----------- 2 files changed, 16 insertions(+), 12 deletions(-) diff --git a/drivers/gpu/drm/scheduler/sched_entity.c b/drivers/gpu/drm/scheduler/sched_entity.c index a4a7efdbf229..673ca9cbf362 100644 --- a/drivers/gpu/drm/scheduler/sched_entity.c +++ b/drivers/gpu/drm/scheduler/sched_entity.c @@ -559,9 +559,10 @@ struct drm_sched_job *drm_sched_entity_pop_job(struct drm_sched_entity *entity) */ smp_wmb(); + spin_lock(&entity->lock); spsc_queue_pop(&entity->job_queue); - drm_sched_rq_pop_entity(entity); + spin_unlock(&entity->lock); /* Jobs and entities might have different lifecycles. Since we're * removing the job from the entities queue, set the jobs entity pointer @@ -647,6 +648,9 @@ void drm_sched_entity_push_job(struct drm_sched_job *sched_job) * Make sure to set the submit_ts first, to avoid a race. */ sched_job->submit_ts = submit_ts = ktime_get(); + + spin_lock(&entity->lock); + first = spsc_queue_push(&entity->job_queue, &sched_job->queue_node); /* first job wakes up scheduler */ @@ -657,5 +661,7 @@ void drm_sched_entity_push_job(struct drm_sched_job *sched_job) if (sched) drm_sched_wakeup(sched); } + + spin_unlock(&entity->lock); } EXPORT_SYMBOL(drm_sched_entity_push_job); diff --git a/drivers/gpu/drm/scheduler/sched_rq.c b/drivers/gpu/drm/scheduler/sched_rq.c index 0464d324d98d..23f46ec610e7 100644 --- a/drivers/gpu/drm/scheduler/sched_rq.c +++ b/drivers/gpu/drm/scheduler/sched_rq.c @@ -257,19 +257,17 @@ static ktime_t drm_sched_entity_get_job_ts(struct drm_sched_entity *entity) struct drm_gpu_scheduler * drm_sched_rq_add_entity(struct drm_sched_entity *entity, ktime_t ts) { + struct drm_sched_rq *rq = entity->rq; struct drm_gpu_scheduler *sched; - struct drm_sched_rq *rq; /* Add the entity to the run queue */ - spin_lock(&entity->lock); - if (entity->stopped) { - spin_unlock(&entity->lock); + lockdep_assert_held(&entity->lock); + if (entity->stopped) { DRM_ERROR("Trying to push to a killed entity\n"); return NULL; } - rq = entity->rq; spin_lock(&rq->lock); sched = rq->sched; @@ -289,7 +287,6 @@ drm_sched_rq_add_entity(struct drm_sched_entity *entity, ktime_t ts) drm_sched_rq_update_fifo_locked(entity, rq, ts); spin_unlock(&rq->lock); - spin_unlock(&entity->lock); return sched; } @@ -343,16 +340,17 @@ drm_sched_rq_next_rr_ts(struct drm_sched_rq *rq, */ void drm_sched_rq_pop_entity(struct drm_sched_entity *entity) { + struct drm_sched_rq *rq = entity->rq; struct drm_sched_job *next_job; - struct drm_sched_rq *rq; + + lockdep_assert_held(&entity->lock); + + spin_lock(&rq->lock); /* * Update the entity's location in the min heap according to * the timestamp of the next job, if any. */ - spin_lock(&entity->lock); - rq = entity->rq; - spin_lock(&rq->lock); next_job = drm_sched_entity_queue_peek(entity); if (next_job) { ktime_t ts; @@ -375,8 +373,8 @@ void drm_sched_rq_pop_entity(struct drm_sched_entity *entity) drm_sched_entity_save_vruntime(entity, min_vruntime); } } + spin_unlock(&rq->lock); - spin_unlock(&entity->lock); } /** From 5a83606d9f31a38273ee803f3aae2e39247032c5 Mon Sep 17 00:00:00 2001 From: Icenowy Zheng Date: Thu, 10 Sep 2026 17:49:04 +0800 Subject: [PATCH 245/592] drm/verisilicon: set blend mode for the cursor plane Blend mode properties are now required to expose pixel formats w/ alpha. Experiments show that the fixed blending mode for the cursor seems to be COVERAGE: - With a cursor plane filled with R=G=0, B=0xff, A=0x40, the cursor is visible on a pure-white background, which means the background is multiplied. - With a cursor plane filled with R=G=B=0xff, A=0x40, the cursor isn't pure white and non-white patterns can be see through, which means the cursor is multiplied. Add a fixed COVERAGE blend mode property for the cursor plane. Signed-off-by: Icenowy Zheng Reviewed-by: Thomas Zimmermann Link: https://patch.msgid.link/20260910094904.3502741-1-zhengxingda@iscas.ac.cn --- drivers/gpu/drm/verisilicon/vs_cursor_plane.c | 2 ++ 1 file changed, 2 insertions(+) diff --git a/drivers/gpu/drm/verisilicon/vs_cursor_plane.c b/drivers/gpu/drm/verisilicon/vs_cursor_plane.c index fa4f601dd0c8..da456dced88a 100644 --- a/drivers/gpu/drm/verisilicon/vs_cursor_plane.c +++ b/drivers/gpu/drm/verisilicon/vs_cursor_plane.c @@ -11,6 +11,7 @@ #include #include +#include #include #include #include @@ -267,6 +268,7 @@ struct drm_plane *vs_cursor_plane_init(struct drm_device *drm_dev, return plane; drm_plane_helper_add(plane, &vs_cursor_plane_helper_funcs); + drm_plane_create_blend_mode_property(plane, BIT(DRM_MODE_BLEND_COVERAGE)); return plane; } From e5d43d7e924d7e1d0c815a5c7407f2a40a6dd2f2 Mon Sep 17 00:00:00 2001 From: Icenowy Zheng Date: Thu, 10 Sep 2026 17:49:59 +0800 Subject: [PATCH 246/592] drm/verisilicon: add primary modifier for format tables Currently the format tables are only used for the primary plane. Add primary modifiers to names related to the tables. Signed-off-by: Icenowy Zheng Reviewed-by: Thomas Zimmermann Link: https://patch.msgid.link/20260910095000.3505878-1-zhengxingda@iscas.ac.cn --- drivers/gpu/drm/verisilicon/vs_hwdb.c | 12 ++++++------ drivers/gpu/drm/verisilicon/vs_hwdb.h | 4 ++-- drivers/gpu/drm/verisilicon/vs_primary_plane.c | 4 ++-- 3 files changed, 10 insertions(+), 10 deletions(-) diff --git a/drivers/gpu/drm/verisilicon/vs_hwdb.c b/drivers/gpu/drm/verisilicon/vs_hwdb.c index 2a0f7c59afa3..56aa45044306 100644 --- a/drivers/gpu/drm/verisilicon/vs_hwdb.c +++ b/drivers/gpu/drm/verisilicon/vs_hwdb.c @@ -10,7 +10,7 @@ #include "vs_dc_top_regs.h" #include "vs_hwdb.h" -static const u32 vs_formats_array_no_yuv444[] = { +static const u32 vs_primary_formats_array_no_yuv444[] = { DRM_FORMAT_XRGB4444, DRM_FORMAT_XBGR4444, DRM_FORMAT_RGBX4444, @@ -44,7 +44,7 @@ static const u32 vs_formats_array_no_yuv444[] = { /* TODO: non-RGB formats */ }; -static const u32 vs_formats_array_with_yuv444[] = { +static const u32 vs_primary_formats_array_with_yuv444[] = { DRM_FORMAT_XRGB4444, DRM_FORMAT_XBGR4444, DRM_FORMAT_RGBX4444, @@ -79,13 +79,13 @@ static const u32 vs_formats_array_with_yuv444[] = { }; static const struct vs_formats vs_formats_no_yuv444 = { - .array = vs_formats_array_no_yuv444, - .num = ARRAY_SIZE(vs_formats_array_no_yuv444) + .primary_array = vs_primary_formats_array_no_yuv444, + .primary_num = ARRAY_SIZE(vs_primary_formats_array_no_yuv444) }; static const struct vs_formats vs_formats_with_yuv444 = { - .array = vs_formats_array_with_yuv444, - .num = ARRAY_SIZE(vs_formats_array_with_yuv444) + .primary_array = vs_primary_formats_array_with_yuv444, + .primary_num = ARRAY_SIZE(vs_primary_formats_array_with_yuv444) }; static struct vs_chip_identity vs_chip_identities[] = { diff --git a/drivers/gpu/drm/verisilicon/vs_hwdb.h b/drivers/gpu/drm/verisilicon/vs_hwdb.h index 2065ecb73043..616076d931a5 100644 --- a/drivers/gpu/drm/verisilicon/vs_hwdb.h +++ b/drivers/gpu/drm/verisilicon/vs_hwdb.h @@ -10,8 +10,8 @@ #include struct vs_formats { - const u32 *array; - unsigned int num; + const u32 *primary_array; + unsigned int primary_num; }; struct vs_chip_identity { diff --git a/drivers/gpu/drm/verisilicon/vs_primary_plane.c b/drivers/gpu/drm/verisilicon/vs_primary_plane.c index 1f2be41ae496..8e9449169301 100644 --- a/drivers/gpu/drm/verisilicon/vs_primary_plane.c +++ b/drivers/gpu/drm/verisilicon/vs_primary_plane.c @@ -168,8 +168,8 @@ struct drm_plane *vs_primary_plane_init(struct drm_device *drm_dev, struct vs_dc plane = drmm_universal_plane_alloc(drm_dev, struct drm_plane, dev, 0, &vs_primary_plane_funcs, - dc->identity.formats->array, - dc->identity.formats->num, + dc->identity.formats->primary_array, + dc->identity.formats->primary_num, NULL, DRM_PLANE_TYPE_PRIMARY, NULL); From 6c62dfd2820f0c32b0083668edbe0baca6e10b92 Mon Sep 17 00:00:00 2001 From: Icenowy Zheng Date: Thu, 10 Sep 2026 17:50:00 +0800 Subject: [PATCH 247/592] drm/verisilicon: remove ARGB formats from primary plane As the blending of the primary plane is currently explicitly disabled (and it's not possible on DC8000), remove the ARGB formats from the primary plane format tables. Signed-off-by: Icenowy Zheng Reviewed-by: Thomas Zimmermann Link: https://patch.msgid.link/20260910095000.3505878-2-zhengxingda@iscas.ac.cn --- drivers/gpu/drm/verisilicon/vs_hwdb.c | 32 --------------------------- 1 file changed, 32 deletions(-) diff --git a/drivers/gpu/drm/verisilicon/vs_hwdb.c b/drivers/gpu/drm/verisilicon/vs_hwdb.c index 56aa45044306..ebf6f843bc88 100644 --- a/drivers/gpu/drm/verisilicon/vs_hwdb.c +++ b/drivers/gpu/drm/verisilicon/vs_hwdb.c @@ -15,32 +15,16 @@ static const u32 vs_primary_formats_array_no_yuv444[] = { DRM_FORMAT_XBGR4444, DRM_FORMAT_RGBX4444, DRM_FORMAT_BGRX4444, - DRM_FORMAT_ARGB4444, - DRM_FORMAT_ABGR4444, - DRM_FORMAT_RGBA4444, - DRM_FORMAT_BGRA4444, DRM_FORMAT_XRGB1555, DRM_FORMAT_XBGR1555, DRM_FORMAT_RGBX5551, DRM_FORMAT_BGRX5551, - DRM_FORMAT_ARGB1555, - DRM_FORMAT_ABGR1555, - DRM_FORMAT_RGBA5551, - DRM_FORMAT_BGRA5551, DRM_FORMAT_RGB565, DRM_FORMAT_BGR565, DRM_FORMAT_XRGB8888, DRM_FORMAT_XBGR8888, DRM_FORMAT_RGBX8888, DRM_FORMAT_BGRX8888, - DRM_FORMAT_ARGB8888, - DRM_FORMAT_ABGR8888, - DRM_FORMAT_RGBA8888, - DRM_FORMAT_BGRA8888, - DRM_FORMAT_ARGB2101010, - DRM_FORMAT_ABGR2101010, - DRM_FORMAT_RGBA1010102, - DRM_FORMAT_BGRA1010102, /* TODO: non-RGB formats */ }; @@ -49,32 +33,16 @@ static const u32 vs_primary_formats_array_with_yuv444[] = { DRM_FORMAT_XBGR4444, DRM_FORMAT_RGBX4444, DRM_FORMAT_BGRX4444, - DRM_FORMAT_ARGB4444, - DRM_FORMAT_ABGR4444, - DRM_FORMAT_RGBA4444, - DRM_FORMAT_BGRA4444, DRM_FORMAT_XRGB1555, DRM_FORMAT_XBGR1555, DRM_FORMAT_RGBX5551, DRM_FORMAT_BGRX5551, - DRM_FORMAT_ARGB1555, - DRM_FORMAT_ABGR1555, - DRM_FORMAT_RGBA5551, - DRM_FORMAT_BGRA5551, DRM_FORMAT_RGB565, DRM_FORMAT_BGR565, DRM_FORMAT_XRGB8888, DRM_FORMAT_XBGR8888, DRM_FORMAT_RGBX8888, DRM_FORMAT_BGRX8888, - DRM_FORMAT_ARGB8888, - DRM_FORMAT_ABGR8888, - DRM_FORMAT_RGBA8888, - DRM_FORMAT_BGRA8888, - DRM_FORMAT_ARGB2101010, - DRM_FORMAT_ABGR2101010, - DRM_FORMAT_RGBA1010102, - DRM_FORMAT_BGRA1010102, /* TODO: non-RGB formats */ }; From 53cf0f26eece095a3752642151150404bc3351cc Mon Sep 17 00:00:00 2001 From: Rosen Penev Date: Tue, 1 Sep 2026 13:37:36 -0700 Subject: [PATCH 248/592] crypto: caam - map job ring registers without claiming region devm_platform_ioremap_resource() ends up calling request_mem_region(), which fails with -EBUSY on i.MX SoCs: There the job rings are modelled as sub-regions of their parent fsl,sec-v4.0 register window, and caam_probe() already reserves the whole window exclusively via devm_of_iomap() before the children are populated. Every job ring therefore collides with its own parent and fails to probe, taking the hardware RNG offline (seen on i.MX6, i.MX7 and i.MX8 boards such as colibri-imx7 and verdin-imx8mp). Map the job ring registers with devm_ioremap() instead, which does not claim the (already owned) region. Fixes: 9a955c0a7d11 ("crypto: caam - simplify probe resource and IRQ handling") Assisted-by: opencode:big-pickle Signed-off-by: Rosen Penev Reported-by: Emanuele Ghidoli Tested-by: Emanuele Ghidoli Tested-by: Richard Leitner # i.MX8MP Tested-by: Sahil Malhotra # i.MX8MM-EVK Signed-off-by: Herbert Xu --- drivers/crypto/caam/jr.c | 23 ++++++++++++++++++++--- 1 file changed, 20 insertions(+), 3 deletions(-) diff --git a/drivers/crypto/caam/jr.c b/drivers/crypto/caam/jr.c index 239469f6882e..cf725a6ac0fb 100644 --- a/drivers/crypto/caam/jr.c +++ b/drivers/crypto/caam/jr.c @@ -583,12 +583,29 @@ static int caam_jr_probe(struct platform_device *pdev) struct caam_drv_private_jr *jrpriv; static int total_jobrs; void __iomem *ctrl; + struct resource *r; int error; int irq; - ctrl = devm_platform_ioremap_resource(pdev, 0); - if (IS_ERR(ctrl)) - return PTR_ERR(ctrl); + /* + * The job rings live inside the register window of their parent + * fsl,sec-v4.0 node, which caam_probe() already reserves (and maps) + * via devm_of_iomap(). A requested region that overlaps that + * reservation, e.g. from devm_platform_ioremap_resource(), would + * therefore fail with -EBUSY, so map the registers without claiming + * the region here. + */ + r = platform_get_resource(pdev, IORESOURCE_MEM, 0); + if (!r) { + dev_err(&pdev->dev, "platform_get_resource() failed\n"); + return -EINVAL; + } + + ctrl = devm_ioremap(&pdev->dev, r->start, resource_size(r)); + if (!ctrl) { + dev_err(&pdev->dev, "devm_ioremap() failed\n"); + return -ENOMEM; + } irq = platform_get_irq(pdev, 0); if (irq < 0) From a26204be587c57bd5c54fa513be26c4fd7bf252d Mon Sep 17 00:00:00 2001 From: Nemesa Garg Date: Wed, 9 Sep 2026 16:33:31 +0530 Subject: [PATCH 249/592] Revert "drm/i915/display: Clear SEL_FETCH_PLANE_CTL on plane disable" MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit This reverts commit 7f1172a2ac0d7e50850785e2e65789c8aac8411a. This commit replaced the crtc_state->enable_psr2_sel_fetch guard in icl_plane_disable_sel_fetch_arm() and i9xx_cursor_disable_sel_fetch_arm() with HAS_PSR2_SEL_FETCH(). This is a display version check and says nothing about the pipe, so every plane and cursor disable on a display 12+ platform started writing SEL_FETCH_PLANE_CTL() / SEL_FETCH_CUR_CTL(), including on pipes that do not implement them. It shows up as an unclaimed register access on pipes driving HDMI where selective fetch was never enabled. The stale selective fetch enable bit that commit addressed is handled in the next patch. Fixes: 7f1172a2ac0d ("drm/i915/display: Clear SEL_FETCH_PLANE_CTL on plane disable") Closes: https://gitlab.freedesktop.org/drm/i915/kernel/-/work_items/16876 Signed-off-by: Nemesa Garg Reviewed-by: Jouni Högander Signed-off-by: Suraj Kandpal Link: https://patch.msgid.link/20260909110332.3528029-2-nemesa.garg@intel.com (cherry picked from commit d393529394167e0f5f706657eebe84d8529ce4fc) Signed-off-by: Jani Nikula --- drivers/gpu/drm/i915/display/intel_cursor.c | 15 +++++---------- .../gpu/drm/i915/display/skl_universal_plane.c | 15 +++++---------- 2 files changed, 10 insertions(+), 20 deletions(-) diff --git a/drivers/gpu/drm/i915/display/intel_cursor.c b/drivers/gpu/drm/i915/display/intel_cursor.c index 86bb96ac449b..0673f16f6fd0 100644 --- a/drivers/gpu/drm/i915/display/intel_cursor.c +++ b/drivers/gpu/drm/i915/display/intel_cursor.c @@ -530,18 +530,13 @@ static int i9xx_check_cursor(struct intel_crtc_state *crtc_state, } static void i9xx_cursor_disable_sel_fetch_arm(struct intel_dsb *dsb, - struct intel_plane *plane) + struct intel_plane *plane, + const struct intel_crtc_state *crtc_state) { struct intel_display *display = to_intel_display(plane); enum pipe pipe = plane->pipe; - /* - * Clear this whenever the hardware has selective fetch, not just when - * the current state uses it. The cursor may have been enabled with - * selective fetch earlier and had its enable bit orphaned when the - * feature was switched off. - */ - if (!HAS_PSR2_SEL_FETCH(display)) + if (!crtc_state->enable_psr2_sel_fetch) return; intel_de_write_dsb(display, dsb, SEL_FETCH_CUR_CTL(pipe), 0); @@ -591,7 +586,7 @@ static void i9xx_cursor_update_sel_fetch_arm(struct intel_dsb *dsb, if (crtc_state->enable_psr2_su_region_et) wa_16021440873(dsb, plane, crtc_state, plane_state); else - i9xx_cursor_disable_sel_fetch_arm(dsb, plane); + i9xx_cursor_disable_sel_fetch_arm(dsb, plane, crtc_state); } } @@ -700,7 +695,7 @@ static void i9xx_cursor_update_arm(struct intel_dsb *dsb, if (plane_state) i9xx_cursor_update_sel_fetch_arm(dsb, plane, crtc_state, plane_state); else - i9xx_cursor_disable_sel_fetch_arm(dsb, plane); + i9xx_cursor_disable_sel_fetch_arm(dsb, plane, crtc_state); if (plane->cursor.base != base || plane->cursor.size != fbc_ctl || diff --git a/drivers/gpu/drm/i915/display/skl_universal_plane.c b/drivers/gpu/drm/i915/display/skl_universal_plane.c index 5cda1ab90e40..07a683293352 100644 --- a/drivers/gpu/drm/i915/display/skl_universal_plane.c +++ b/drivers/gpu/drm/i915/display/skl_universal_plane.c @@ -879,18 +879,13 @@ skl_plane_disable_arm(struct intel_dsb *dsb, } static void icl_plane_disable_sel_fetch_arm(struct intel_dsb *dsb, - struct intel_plane *plane) + struct intel_plane *plane, + const struct intel_crtc_state *crtc_state) { struct intel_display *display = to_intel_display(plane); enum pipe pipe = plane->pipe; - /* - * Clear this whenever the hardware has selective fetch, not just when - * the current state uses it. The plane may have been enabled with - * selective fetch earlier and had its enable bit orphaned when the - * feature was switched off. - */ - if (!HAS_PSR2_SEL_FETCH(display)) + if (!crtc_state->enable_psr2_sel_fetch) return; intel_de_write_dsb(display, dsb, SEL_FETCH_PLANE_CTL(pipe, plane->id), 0); @@ -926,7 +921,7 @@ icl_plane_disable_arm(struct intel_dsb *dsb, skl_write_plane_wm(dsb, plane, crtc_state); - icl_plane_disable_sel_fetch_arm(dsb, plane); + icl_plane_disable_sel_fetch_arm(dsb, plane, crtc_state); if (plane_has_normalizer(plane)) intel_de_write_dsb(display, dsb, @@ -1646,7 +1641,7 @@ static void icl_plane_update_sel_fetch_arm(struct intel_dsb *dsb, intel_de_write_dsb(display, dsb, SEL_FETCH_PLANE_CTL(pipe, plane->id), SEL_FETCH_PLANE_CTL_ENABLE); else - icl_plane_disable_sel_fetch_arm(dsb, plane); + icl_plane_disable_sel_fetch_arm(dsb, plane, crtc_state); } static void From b61b6f95d6722ddbbbd09e689fa41b55fd36f9a5 Mon Sep 17 00:00:00 2001 From: Peter Zijlstra Date: Fri, 11 Sep 2026 11:04:47 +0200 Subject: [PATCH 250/592] futex: Also allocate private hash on vfork() As Jann demonstrated, it is entirely feasible to access the mm through vfork(). Therefore we need to allocate a private hash on vfork() as well as any other CLONE_VM user. Specifically, it must be avoided to have (private) futex waiters before allocating the private hash. Fixes: ee9dce44362b ("futex: Drop CLONE_THREAD requirement for private default hash alloc") Reported-by: Jann Horn Signed-off-by: Peter Zijlstra (Intel) Link: https://patch.msgid.link/20260911090447.GT788244@noisy.programming.kicks-ass.net --- kernel/fork.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/kernel/fork.c b/kernel/fork.c index a5934a317634..5ef413368912 100644 --- a/kernel/fork.c +++ b/kernel/fork.c @@ -1996,9 +1996,9 @@ static bool need_futex_hash_allocate_default(u64 clone_flags) { /* * Allocate a default futex hash for any sibling that will - * share the parent's mm, except vfork. + * share the parent's mm. */ - return (clone_flags & (CLONE_VM | CLONE_VFORK)) == CLONE_VM; + return clone_flags & CLONE_VM; } /* From f0e9f963a3d209d7dc7ddd61116118ab5da2797d Mon Sep 17 00:00:00 2001 From: Raag Jadav Date: Fri, 11 Sep 2026 17:45:47 +0530 Subject: [PATCH 251/592] drm/xe/i2c: Disable IRQ on unbind Currently, struct xe_i2c is freed before SGUnit IRQ is disabled in unbind path, leaving a potential UAF in case I2C IRQ is hit during this small window. Explicitly disable I2C IRQ in xe_i2c_remove() and fix this. Fixes: 0bb78ce09926 ("drm/xe/i2c: Wire up reset/postinstall for I2C IRQ") Signed-off-by: Raag Jadav Reviewed-by: Heikki Krogerus Link: https://patch.msgid.link/20260911121547.2407261-1-raag.jadav@intel.com Signed-off-by: Matt Roper (cherry picked from commit 8ba5c8b8ab3fd362267c11df2cd5a90ee46f6e24) Signed-off-by: Rodrigo Vivi --- drivers/gpu/drm/xe/xe_i2c.c | 3 +++ 1 file changed, 3 insertions(+) diff --git a/drivers/gpu/drm/xe/xe_i2c.c b/drivers/gpu/drm/xe/xe_i2c.c index 399099ff0dbc..f4f381988289 100644 --- a/drivers/gpu/drm/xe/xe_i2c.c +++ b/drivers/gpu/drm/xe/xe_i2c.c @@ -318,8 +318,10 @@ void xe_i2c_pm_resume(struct xe_device *xe, bool d3cold) static void xe_i2c_remove(void *data) { struct xe_i2c *i2c = data; + struct xe_device *xe = tile_to_xe(i2c->mmio->tile); unsigned int i; + xe_i2c_irq_reset(xe); xe_amc_exit(i2c); for (i = 0; i < XE_I2C_MAX_CLIENTS; i++) { @@ -329,6 +331,7 @@ static void xe_i2c_remove(void *data) bus_unregister_notifier(&i2c_bus_type, &i2c->bus_notifier); xe_i2c_unregister_adapter(i2c); + xe->i2c = NULL; } /** From a13f7f5d14af9baf34eb12c25962f8d3542b281d Mon Sep 17 00:00:00 2001 From: Ilya Titov Date: Thu, 3 Sep 2026 12:17:18 +0300 Subject: [PATCH 252/592] pinctrl: sunxi: keep a shadow copy of the data register output latches MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit On Allwinner SoCs, reading a bank's data register returns the pin level, not the output latch, for pins that are muxed as inputs. Writing a GPIO therefore corrupts the output latches of all input-muxed pins in the same bank: the read-modify-write in sunxi_pinctrl_gpio_set() reads back their pin levels and writes those into their latches. This breaks emulated open-drain lines (e.g. a bit-banged I2C bus from i2c-gpio). Such a line is released high by muxing it as input and letting the pull-up raise it, so any concurrent GPIO write in the same bank stores 1 into its latch. Driving the line low afterwards is a non-atomic data-then-mux sequence in sunxi_pinctrl_gpio_direction_output(); if the poisoning write lands between the two steps, the pin actively drives high (push-pull) instead of low. Observed in practice as sporadic glitches on a T507 board bit-banging I2C on port E while other PE GPIOs are toggled. On a scope the failure is unmistakable: on a clock pulse where SCL should fall to GND, the line instead steps *above* its idle high level for the whole low phase — the pad drives a strong push-pull 3.3 V high, higher than the level the pull-up sustains on the loaded bus — before the next transition recovers it. The same can hit SDA, corrupting data instead of clocks. Steps to reproduce on any sunxi board with a bit-banged (i2c-gpio) bus: # background: toggle any other GPIO of the same bank, e.g. line 21 gpioset -c --toggle 100us 21=0 & # foreground: keep the bit-banged bus busy while :; do i2cdetect -y 0x50 0x57; done # watch SCL/SDA with a scope or logic analyzer: sporadic clock-low # phases driven high (above the pull-up level) instead of low The bank spinlock cannot help: the racing write is a perfectly valid whole-register RMW that faithfully writes back what the hardware returned. There are no set/clear registers on this IP to write a single bit atomically. Fix it the same way gpio-mmio handles hardware whose data register read does not return the output latch: keep a shadow copy of each bank's latches, base the read-modify-write on the shadow, and only write the register. The shadow is seeded from the hardware at probe time so pins left in output mode by the bootloader keep their state. Pins that reach output mode through the gpiolib paths write their value (and thereby their shadow bit) before the mux switch in sunxi_pinctrl_gpio_direction_output(); pins muxed to gpio_out directly through a pinmux node bypass that path, so sunxi_pmx_set() refreshes their shadow bit from the latch (readable once the pin is in output mode) to keep them driving their pre-existing level. Seeding the shadow reads the PIO registers at probe time, which requires the bus clock to be enabled. The clock was only requested at the very end of probe, after devm_pinctrl_register() had already claimed the pin hogs described in the device tree - which mux pins, and thus access registers, with the clock still gated. Move the request ahead of both. Boards whose bootloader leaves the PIO clock running are unaffected, which is why the pre-existing hog problem has gone unnoticed since commit 950707c0eb5c ("pinctrl: sunxi: add clock support"). Fixes: df7b34f4c3d2 ("pinctrl: sunxi: Fix gpio_set behaviour") Cc: stable@vger.kernel.org Signed-off-by: Ilya Titov Signed-off-by: Linus Walleij --- drivers/pinctrl/sunxi/pinctrl-sunxi.c | 76 +++++++++++++++++++++------ drivers/pinctrl/sunxi/pinctrl-sunxi.h | 7 +++ 2 files changed, 68 insertions(+), 15 deletions(-) diff --git a/drivers/pinctrl/sunxi/pinctrl-sunxi.c b/drivers/pinctrl/sunxi/pinctrl-sunxi.c index 25489beeb312..2881a83be99c 100644 --- a/drivers/pinctrl/sunxi/pinctrl-sunxi.c +++ b/drivers/pinctrl/sunxi/pinctrl-sunxi.c @@ -837,6 +837,21 @@ static void sunxi_pmx_set(struct pinctrl_dev *pctldev, writel((readl(pctl->membase + reg) & ~mask) | config << shift, pctl->membase + reg); + /* + * A pin muxed to gpio_out directly through a pinmux node bypasses + * sunxi_pinctrl_gpio_set() and drives whatever its output latch + * holds. Now that the pin is in output mode the data register + * reads back the latch, so refresh the shadow to keep such pins + * driving their pre-existing level. + */ + if (config == SUN4I_FUNC_OUTPUT) { + u32 *shadow = &pctl->dat_shadow[pin / PINS_PER_BANK]; + + sunxi_data_reg(pctl, pin, ®, &shift, &mask); + *shadow = (*shadow & ~mask) | + (readl(pctl->membase + reg) & mask); + } + raw_spin_unlock_irqrestore(&pctl->lock, flags); } @@ -1017,21 +1032,29 @@ static int sunxi_pinctrl_gpio_set(struct gpio_chip *chip, unsigned int offset, int value) { struct sunxi_pinctrl *pctl = gpiochip_get_data(chip); - u32 reg, shift, mask, val; + u32 *shadow = &pctl->dat_shadow[offset / PINS_PER_BANK]; + u32 reg, shift, mask; unsigned long flags; sunxi_data_reg(pctl, offset, ®, &shift, &mask); raw_spin_lock_irqsave(&pctl->lock, flags); - val = readl(pctl->membase + reg); - + /* + * Reading the data register returns the pin level, not the output + * latch, for pins muxed as inputs. A read-modify-write based on + * the register would therefore corrupt the latches of input-muxed + * pins in the same bank (e.g. an emulated open-drain I2C line + * released high), making them drive the wrong level once switched + * to output. Base the read-modify-write on a shadow copy of the + * latches instead. + */ if (value) - val |= mask; + *shadow |= mask; else - val &= ~mask; + *shadow &= ~mask; - writel(val, pctl->membase + reg); + writel(*shadow, pctl->membase + reg); raw_spin_unlock_irqrestore(&pctl->lock, flags); @@ -1572,7 +1595,7 @@ int sunxi_pinctrl_init_with_flags(struct platform_device *pdev, struct pinctrl_pin_desc *pins; struct sunxi_pinctrl *pctl; struct pinmux_ops *pmxops; - int i, ret, last_pin, pin_idx; + int i, ret, last_pin, pin_idx, nbanks; struct clk *clk; pctl = devm_kzalloc(&pdev->dev, sizeof(*pctl), GFP_KERNEL); @@ -1610,6 +1633,37 @@ int sunxi_pinctrl_init_with_flags(struct platform_device *pdev, if (!pctl->irq_array) return -ENOMEM; + /* + * The bus clock has to be enabled before the pinctrl device + * registers, as the pin hogs claimed from there access registers. + */ + ret = of_clk_get_parent_count(node); + clk = devm_clk_get_enabled(&pdev->dev, ret == 1 ? NULL : "apb"); + if (IS_ERR(clk)) + return PTR_ERR(clk); + + /* + * Seed the output latch shadow from the hardware so pins the + * bootloader left in output mode keep their state; see + * sunxi_pinctrl_gpio_set() for why a shadow is needed. This must + * happen before the pinctrl device registers, as pin hogs can mux + * pins to gpio_out and thereby update the shadow. + */ + last_pin = pctl->desc->pins[pctl->desc->npins - 1].pin.number; + nbanks = DIV_ROUND_UP(last_pin + 1 - pctl->desc->pin_base, + PINS_PER_BANK); + pctl->dat_shadow = devm_kcalloc(&pdev->dev, nbanks, + sizeof(*pctl->dat_shadow), GFP_KERNEL); + if (!pctl->dat_shadow) + return -ENOMEM; + + for (i = 0; i < nbanks; i++) { + u32 reg, shift, mask; + + sunxi_data_reg(pctl, i * PINS_PER_BANK, ®, &shift, &mask); + pctl->dat_shadow[i] = readl(pctl->membase + reg); + } + ret = sunxi_pinctrl_build_state(pdev); if (ret) { dev_err(&pdev->dev, "dt probe failed: %d\n", ret); @@ -1665,7 +1719,6 @@ int sunxi_pinctrl_init_with_flags(struct platform_device *pdev, if (!pctl->chip) return -ENOMEM; - last_pin = pctl->desc->pins[pctl->desc->npins - 1].pin.number; pctl->chip->owner = THIS_MODULE; pctl->chip->request = gpiochip_generic_request; pctl->chip->free = gpiochip_generic_free; @@ -1699,13 +1752,6 @@ int sunxi_pinctrl_init_with_flags(struct platform_device *pdev, goto gpiochip_error; } - ret = of_clk_get_parent_count(node); - clk = devm_clk_get_enabled(&pdev->dev, ret == 1 ? NULL : "apb"); - if (IS_ERR(clk)) { - ret = PTR_ERR(clk); - goto gpiochip_error; - } - pctl->irq = devm_kcalloc(&pdev->dev, pctl->desc->irq_banks, sizeof(*pctl->irq), diff --git a/drivers/pinctrl/sunxi/pinctrl-sunxi.h b/drivers/pinctrl/sunxi/pinctrl-sunxi.h index 0daf7600e2fb..498e88a19f71 100644 --- a/drivers/pinctrl/sunxi/pinctrl-sunxi.h +++ b/drivers/pinctrl/sunxi/pinctrl-sunxi.h @@ -85,6 +85,7 @@ #define IO_BIAS_MASK GENMASK(3, 0) #define SUN4I_FUNC_INPUT 0 +#define SUN4I_FUNC_OUTPUT 1 #define SUN4I_FUNC_IRQ 6 #define SUN4I_FUNC_DISABLED_OLD 7 #define SUN4I_FUNC_DISABLED_NEW 15 @@ -175,6 +176,12 @@ struct sunxi_pinctrl { int *irq; unsigned *irq_array; raw_spinlock_t lock; + /* + * Output latch shadow, one word per bank. Seeded lockless at + * probe before the pinctrl device registers, protected by @lock + * afterwards. + */ + u32 *dat_shadow; struct pinctrl_dev *pctl_dev; unsigned long flags; u32 bank_mem_size; From cc337324cc6be1ce0ae7e5a068dcb7e99ae1b59c Mon Sep 17 00:00:00 2001 From: Filipe Manana Date: Mon, 14 Sep 2026 18:11:30 +0100 Subject: [PATCH 253/592] btrfs: fix creation of compressed inline extents that don't save space If the compressed data of an inline extent is larger than or equals to the size of the uncompressed data, we are still allowing the creation of the compressed inline extent, which does not result in any benefits, quite the contrary as we waste metadata space and have to decompress when reading. This is a recent regression introduced in commit 3eaf5f082c4c ("btrfs: extract inlined creation into a dedicated delalloc helper"). It happens because we are passing the block size to btrfs_compress_bio(), so we don't get -E2BIG from the compression code anymore, but we can not pass i_size either, because if i_size is smaller than sector size, we end up never creating lzo compressed inline extent for such small i_size values. So refuse the compressed result at run_delalloc_inline() if its size is not smaller than the uncompressed size (i_size). Reported-by: Hanabishi Link: https://lore.kernel.org/linux-btrfs/c97652a5-ac6b-4de6-aa23-3cdebc01d00b@gmail.com/ Fixes: 3eaf5f082c4c ("btrfs: extract inlined creation into a dedicated delalloc helper") CC: stable@vger.kernel.org # 7.1+ Reviewed-by: Qu Wenruo Signed-off-by: Filipe Manana Signed-off-by: David Sterba --- fs/btrfs/inode.c | 15 +++++++++++++++ 1 file changed, 15 insertions(+) diff --git a/fs/btrfs/inode.c b/fs/btrfs/inode.c index 3668cbc7598e..4f976b52996d 100644 --- a/fs/btrfs/inode.c +++ b/fs/btrfs/inode.c @@ -2339,12 +2339,27 @@ static int run_delalloc_inline(struct btrfs_inode *inode, struct folio *locked_f } else if (inode->prop_compress) { compress_type = inode->prop_compress; } + /* + * We need to pass blocksize and not i_size, otherwise we can't + * create compressed inline extents for data smaller than sector + * size with lzo. + */ cb = btrfs_compress_bio(inode, 0, blocksize, compress_type, compress_level, 0); if (IS_ERR(cb)) { cb = NULL; /* Just fall back to non-compressed case. */ } else { compressed_size = cb->bbio.bio.bi_iter.bi_size; + /* + * If we did not save space, it's pointless and wasteful + * to have an inline compressed extent, so fallback to + * an uncompressed inline extent. + */ + if (compressed_size >= i_size) { + cleanup_compressed_bio(cb); + cb = NULL; + compressed_size = 0; + } } } if (!can_cow_file_range_inline(inode, 0, i_size, compressed_size)) { From 76bf149cd0298544631e756670b89c399c7acbca Mon Sep 17 00:00:00 2001 From: Daniel Linjama Date: Wed, 16 Sep 2026 09:15:56 +0300 Subject: [PATCH 254/592] btrfs: handle lack of space when cleaning up verity items When enable_verity() hits the qgroup limit, rollback_verity() needs its own metadata reservation. When the qgroup limit or lack of space refuses the rollback, the whole filesystem is forced read-only even though the qgroup limit was for one subvolume only. Also orphan cleanup at the next mount fails the same way, so the leftover items are never removed: with -EDQUOT the subvolume stays unreachable, and with -ENOSPC on a full filesystem the next read-write mount fails. Start transactions with btrfs_start_transaction_fallback_global_rsv() in btrfs_orphan_cleanup(), drop_verity_items() and rollback_verity(). Those calls only delete items and free the space in the end, so they may use the global reserve and skip the qgroup limit, which avoids -ENOSPC and -EDQUOT. Fixes: 146054090b08 ("btrfs: initial fsverity support") Reviewed-by: Qu Wenruo Signed-off-by: Daniel Linjama Signed-off-by: David Sterba --- fs/btrfs/inode.c | 3 ++- fs/btrfs/verity.c | 18 ++++++++++++++++-- 2 files changed, 18 insertions(+), 3 deletions(-) diff --git a/fs/btrfs/inode.c b/fs/btrfs/inode.c index 4f976b52996d..e97446fc9e10 100644 --- a/fs/btrfs/inode.c +++ b/fs/btrfs/inode.c @@ -3892,7 +3892,8 @@ int btrfs_orphan_cleanup(struct btrfs_root *root) if (ret) goto out; } - trans = btrfs_start_transaction(root, 1); + /* Only deletes the orphan. */ + trans = btrfs_start_transaction_fallback_global_rsv(root, 1); if (IS_ERR(trans)) { ret = PTR_ERR(trans); goto out; diff --git a/fs/btrfs/verity.c b/fs/btrfs/verity.c index 4e0ab5842274..600337a84fbe 100644 --- a/fs/btrfs/verity.c +++ b/fs/btrfs/verity.c @@ -93,6 +93,20 @@ static loff_t merkle_file_pos(const struct inode *inode) return rounded; } +/* + * Start a transaction for removing verity items or the verity orphan. + * + * Like unlink, this only deletes items and frees space in the end, so the + * reservation may come from the global reserve when the filesystem is full + * (-ENOSPC) and is not subject to the qgroup limit (-EDQUOT). Otherwise a + * failed enable could never be cleaned up in either situation. + */ +static struct btrfs_trans_handle *start_verity_cleanup_trans(struct btrfs_root *root, + unsigned int num_items) +{ + return btrfs_start_transaction_fallback_global_rsv(root, num_items); +} + /* * Drop all the items for this inode with this key_type. * @@ -120,7 +134,7 @@ static int drop_verity_items(struct btrfs_inode *inode, u8 key_type) while (1) { /* 1 for the item being dropped */ - trans = btrfs_start_transaction(root, 1); + trans = start_verity_cleanup_trans(root, 1); if (IS_ERR(trans)) return PTR_ERR(trans); @@ -466,7 +480,7 @@ static int rollback_verity(struct btrfs_inode *inode) * 1 for updating the inode flag * 1 for deleting the orphan */ - trans = btrfs_start_transaction(root, 2); + trans = start_verity_cleanup_trans(root, 2); if (IS_ERR(trans)) { ret = PTR_ERR(trans); trans = NULL; From 8d836581f9b1f57ceaa2b47654754ef1260b410b Mon Sep 17 00:00:00 2001 From: Niklas Cassel Date: Wed, 16 Sep 2026 15:00:32 +0200 Subject: [PATCH 255/592] ata: libata-scsi: fix ata_dsm_trim_pages() kernel-doc make htmldocs fails with: Documentation/driver-api/libata:604: ./drivers/ata/libata-scsi.c:2301: ERROR: Unexpected indentation. [docutils] The Return: section of ata_dsm_trim_pages() ends its first sentence with a colon and continues with an indented bullet list. reStructuredText requires a blank line before an indented block, so docutils chokes on the list. Simply adding the missing blank line does not work either: Return: is a kernel-doc "special section", which is terminated by the first blank line, so the bullet list would end up in the Description section, detached from the sentence introducing it. Spell the two bounds out as prose instead, so that the Return: section stays self-contained. Documentation-only change. Fixes: e64e6b5dc867 ("ata: libata-scsi: scale DSM TRIM payload by MAX PAGES PER DSM COMMAND") Reported-by: Thomas Huth Closes: https://lore.kernel.org/linux-ide/b0a0b8e8-cc4f-4c7b-8bc5-fee0712d405a@redhat.com/ Reviewed-by: Damien Le Moal Reviewed-by: Thomas Huth Link: https://lore.kernel.org/r/20260916130031.29990-2-cassel@kernel.org Signed-off-by: Niklas Cassel --- drivers/ata/libata-scsi.c | 7 +++---- 1 file changed, 3 insertions(+), 4 deletions(-) diff --git a/drivers/ata/libata-scsi.c b/drivers/ata/libata-scsi.c index b3666519b648..7e22bbc38238 100644 --- a/drivers/ata/libata-scsi.c +++ b/drivers/ata/libata-scsi.c @@ -2297,10 +2297,9 @@ static unsigned int ata_scsiop_inq_89(struct ata_device *dev, * logical block, so it can hold at most sector_size / 512 pages. * * Return: the maximum number of 512-byte pages a single translated WRITE SAME - * command may send to @dev (never less than one), that is the smaller of: - * - MAX PAGES PER DSM COMMAND (IDENTIFY DEVICE word 105), when the device - * reports a non-zero limit; and - * - the logical sector size expressed in 512-byte pages (see above). + * command may send to @dev, that is the smaller of MAX PAGES PER DSM COMMAND + * (IDENTIFY DEVICE word 105, when the device reports a non-zero limit) and + * the logical sector size expressed in 512-byte pages; never less than one. */ static unsigned int ata_dsm_trim_pages(struct ata_device *dev) { From e6cb0b4d4ecb8e71fd2200d907ab2e9663356f69 Mon Sep 17 00:00:00 2001 From: Muhammad Bilal Date: Wed, 16 Sep 2026 05:29:26 +0500 Subject: [PATCH 256/592] hwmon: (hp-wmi-sensors) Fix use-after-free in fungible_show() nsensor->current_state is dynamically replaced as the sensor's state changes. update_numeric_sensor_from_wobj() does this by freeing the old string and installing a new one: if (strcmp(trimmed, nsensor->current_state)) { new_string = hp_wmi_strdup(dev, trimmed); if (new_string) { devm_kfree(dev, nsensor->current_state); nsensor->current_state = new_string; } } This function is only ever called from hp_wmi_update_info() while state->lock is held, so the free-and-replace itself is properly serialized against concurrent updates. fungible_show(), however, reads the same pointer after the lock has already been dropped: err = hp_wmi_update_info(state, info); if (err) return err; switch (prop) { ... case HP_WMI_PROPERTY_CURRENT_STATE: seq_printf(seqf, "%s\n", nsensor->current_state); break; hp_wmi_update_info() takes state->lock internally and releases it before returning, so by the time fungible_show() dereferences nsensor->current_state in seq_printf(), no lock is held. Two processes reading a sensor's current_state debugfs entry at overlapping times (or one reading it while another read of the same sensor triggers a refresh) can race: one thread's seq_printf() can be part-way through printing the string at the moment another thread's call into update_numeric_sensor_from_wobj() frees it with devm_kfree() and installs a new pointer, causing a use-after-free read. Take state->lock around the read in fungible_show() as well, so it can never run concurrently with the free-and-replace in update_numeric_sensor_from_wobj(). Fixes: 23902f98f8d4 ("hwmon: add HP WMI Sensors driver") Cc: stable@vger.kernel.org Signed-off-by: Muhammad Bilal Acked-by: James Seo Link: https://patch.msgid.link/20260916002926.161595-1-meatuni001@gmail.com Signed-off-by: Guenter Roeck --- drivers/hwmon/hp-wmi-sensors.c | 2 ++ 1 file changed, 2 insertions(+) diff --git a/drivers/hwmon/hp-wmi-sensors.c b/drivers/hwmon/hp-wmi-sensors.c index 03c684ba83bd..55aee16df57d 100644 --- a/drivers/hwmon/hp-wmi-sensors.c +++ b/drivers/hwmon/hp-wmi-sensors.c @@ -1247,7 +1247,9 @@ static int fungible_show(struct seq_file *seqf, enum hp_wmi_property prop) break; case HP_WMI_PROPERTY_CURRENT_STATE: + mutex_lock(&state->lock); seq_printf(seqf, "%s\n", nsensor->current_state); + mutex_unlock(&state->lock); break; case HP_WMI_PROPERTY_UNIT_MODIFIER: From 1d12fb94ac0975566545871dda100df34df5f845 Mon Sep 17 00:00:00 2001 From: Sanman Pradhan Date: Tue, 15 Sep 2026 16:48:35 +0000 Subject: [PATCH 257/592] hwmon: (pmbus/tps53679) Fix TPS53676 phase page decoding tps53676_identify() reads the USER_DATA_03 phase configuration to count the phases assigned to each channel and derive the number of PMBus pages. In each 16-bit phase descriptor the channel (PAGE) is encoded in bit 4 and the firing order in bits 3:0, but the code tested bit 3 (0x08), which is part of the firing-order field. TPS53676 supports up to seven phases, so firing-order bit 3 is never set. As a result the existing test classifies every enabled phase as channel A. On a dual-channel configuration the phases assigned to channel B are therefore miscounted as channel A and page 1 is not exposed. Test the PAGE field (bit 4) instead. Fixes: cb3d37b59012 ("hwmon: (pmbus/tps53679) Add support for TI TPS53676") Cc: stable@vger.kernel.org Signed-off-by: Sanman Pradhan Link: https://patch.msgid.link/20260915164823.160977-2-sanman.pradhan@hpe.com Signed-off-by: Guenter Roeck --- drivers/hwmon/pmbus/tps53679.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/drivers/hwmon/pmbus/tps53679.c b/drivers/hwmon/pmbus/tps53679.c index 31e54608b3c9..9f27832703ff 100644 --- a/drivers/hwmon/pmbus/tps53679.c +++ b/drivers/hwmon/pmbus/tps53679.c @@ -187,7 +187,7 @@ static int tps53676_identify(struct i2c_client *client, return -EIO; for (i = 0; i < 2 * TPS53676_MAX_PHASES; i += 2) { if (buf[i + 1] & 0x80) { - if (buf[i] & 0x08) + if (buf[i] & BIT(4)) phases_b++; else phases_a++; From d2710c8d938ae6a825a6463158e6e6f31eac792a Mon Sep 17 00:00:00 2001 From: Thomas Gleixner Date: Fri, 11 Sep 2026 11:09:12 +0200 Subject: [PATCH 258/592] signal: Prevent exec() race Hyunwoo debugged the following KASAN UAF splat: BUG: KASAN: slab-use-after-free in __send_signal_locked+0xb27/0xba0 Write of size 8 at addr ffff888007ed80c8 by task poc/79 ... Call Trace: __send_signal_locked+0xb27/0xba0 do_send_sig_info+0xa7/0x160 do_send_specific+0x76/0xa0 __x64_sys_tgkill+0x193/0x270 ... Allocated by task 80: do_timer_create+0x1a4/0x1030 __x64_sys_timer_create+0x145/0x190 ... Freed by task 12: kmem_cache_free_bulk+0x1f8/0x4a0 kvfree_rcu_bulk+0x14f/0x1c0 kfree_rcu_work+0x128/0x1a0 ... Last potentially related work creation: kvfree_call_rcu+0x39/0x390 __flush_itimer_signals+0x211/0x320 flush_itimer_signals+0x47/0x90 begin_new_exec+0xa6b/0x28c0 It turned out that this happens with a non-leader exec() as Hyunwoo explained: de_thread() calls exchange_tids() before release_task(leader), so the struct pid held by a SIGEV_THREAD_ID timer created against the leader's tid now points to the thread which called execve(). pid_task() returns that thread and lock_task_sighand() on it succeeds. If the timer signal is blocked, its sigqueue stays queued on the leader's task::pending. The next expiry of that timer can then run while release_task() flushes the queue. posixtimer_send_sigqueue() checks whether the sigqueue is already queued with a plain list_empty(), which only reads list_head::next. list_del_init() is not atomic and INIT_LIST_HEAD() stores list_head::next before list_head::prev, so the check can pass in between. list_add_tail() queues the entry on the task::pending of the live thread, and the list_head::prev store from the flush then overwrites the list_head::prev link that list_add_tail() has just set. __flush_itimer_signals() does not undo that either. With list_head::prev pointing at the entry itself, its list_del_init() only stores the same values again, so the entry is not removed from the list. It is still there after the last reference is dropped and the timer is freed by RCU, and the list_add_tail() of a later tgkill() follows that list_head::prev into the freed timer. This problem surfaced with the recent commit which moved the sigqueue flush out of the sighand lock held region. Hyonwoo proposed to fix this by using list_del_init_careful(), but that just papers over the problem. After some disucssions and various attempts to solve it, Eric pointed out that there is no reason to flush task::pending late in release_task() and it should be done in exit_signals() already. As nothing can collect and deliver signals which are queued in a dying task's pending queue, there is no reason to delay it further. But it has to be ensured that no signals can be queued into it after that point. exit_signals() sets PF_EXITING in task::flags, which can be used as an indicator for this. Cure it by: - Preventing signal queueing for task private signals (PIDTYPE_PID) when the task has PF_EXITING set in __send_signal_locked() and in posixtimer_send_sigqueue(). - Protecting the unlocked setting of PF_EXITING in exit_signals() for the task group empty and the group exit case with sighand lock - Flushing task::pending signals right there. Optimize that by moving the whole pending list to an on-stack list head under sighand lock and free the signals without the lock held. There has been quite some discussion about the lockless flush and the non-leader exec case on weakly ordered systems. The problem is that a third party which tries to send a posix timer signal relies on the PID lookup to find the target task and that lookup might result in the new leader when the signal was originaly directed to the old leader. In case that the signal was queued on the old leader then the lockless flush raised a concern over the following situation: old_leader new_leader third party A: flush_list() // list_del_init() stores to sigqueue LOCK (tasklist) old_leader->exit_state = EXIT_ZOMBIE; B: UNLOCK (tasklist) C: LOCK (tasklist) if (old_leader->exit_state) transfer_tids() D: store PID posix_timer_send_sigqueue() // Observes #D so t = new_leader E: t = get_target() F: LOCK (sighand) G: if (list_empty(sigqueue)) list_add(sigqueue) The concern was that the third party might observe #D but not observe #A and therefore would proceed to #G while the list_del() stores (#A) in flush_list() are not visible yet, which could result in list corruption. That would be possible if looking at it solely from a RELEASE+ACQUIRE ordering point of view, but B-C is a UNLOCK+LOCK hand-over, which is not the same as RELEASE+ACQUIRE: RELEASE+ACQUIRE: RCpc, only the CPUs involved agree on the ordering UNLOCK+LOCK: RCtso, the hand-over is store-ordering As B-C is UNLOCK+LOCK, which is RCtso and that does impose store order, A stores must happen before the D store. Combine with E-F, which has a data dependency from the LOAD to the LOCK and thereby constraints later LOADs, those sigqueue loads in G that come after F must in fact observe the A stores. Fixes: fb3bbcfe344e ("exit: change the release_task() paths to call flush_sigqueue() lockless") Reported-by: Hyunwoo Kim Debugged-by: Hyunwoo Kim Suggested-by: "Eric W. Biederman" Signed-off-by: Thomas Gleixner Tested-by: Kijo Park Reviewed-by: Oleg Nesterov Reviewed-by: Frederic Weisbecker Cc: stable@vger.kernel.org Link: https://patch.msgid.link/20260911090541.572536604@kernel.org Closes: https://patch.msgid.link/aok1rdkBgZsynHZB@v4bel --- kernel/exit.c | 11 ++--- kernel/signal.c | 117 +++++++++++++++++++++++++++++++++++------------- 2 files changed, 92 insertions(+), 36 deletions(-) diff --git a/kernel/exit.c b/kernel/exit.c index 4e028f157597..424c44a42a4d 100644 --- a/kernel/exit.c +++ b/kernel/exit.c @@ -302,12 +302,13 @@ void release_task(struct task_struct *p) free_pids(post.pids); release_thread(p); /* - * This task was already removed from the process/thread/pid lists - * and lock_task_sighand(p) can't succeed. Nobody else can touch - * ->pending or, if group dead, signal->shared_pending. We can call - * flush_sigqueue() lockless. + * This task was already removed from the process/thread/pid lists and + * lock_task_sighand(p) can't succeed. If it's the group leader then + * flush tsk->signal->shared_pending. tsk->pending has been flushed + * already in exit_signals(). Nothing else can touch + * signal->shared_pending anymore, so flush_sigqueue() can be invoked + * lockless. */ - flush_sigqueue(&p->pending); if (thread_group_leader(p)) flush_sigqueue(&p->signal->shared_pending); diff --git a/kernel/signal.c b/kernel/signal.c index ec30550951ec..d31ebcb6ed4d 100644 --- a/kernel/signal.c +++ b/kernel/signal.c @@ -457,18 +457,42 @@ static void __sigqueue_free(struct sigqueue *q) kmem_cache_free(sigqueue_cachep, q); } -void flush_sigqueue(struct sigpending *queue) +/* + * flush_sigqueue_list() can only be invoked without holding sighand::siglock in + * the following cases: + * + * 1) When flushing task::pending _after_ setting task::flags PF_EXITING + * + * All functions which try to send a signal to @task will observe PF_EXITING + * and drop the signal. + * + * 2) When flushing task::signal::shared_pending _after_ the last task in a + * thread group was unhashed and task::sighand is NULL. + * + * Nothing can queue a signal anymore because sighand is NULL. + */ +static void flush_sigqueue_list(struct list_head *head) { - struct sigqueue *q; + struct sigqueue *q, *tmp; - sigemptyset(&queue->signal); - while (!list_empty(&queue->list)) { - q = list_entry(queue->list.next, struct sigqueue , list); + list_for_each_entry_safe(q, tmp, head, list) { list_del_init(&q->list); __sigqueue_free(q); } } +void flush_sigqueue(struct sigpending *queue) +{ + sigemptyset(&queue->signal); + flush_sigqueue_list(&queue->list); +} + +static void sigqueue_dequeue_pending(struct sigpending *queue, struct list_head *head) +{ + sigemptyset(&queue->signal); + list_splice_init(&queue->list, head); +} + /* * Flush all pending signals for this kthread. */ @@ -1019,6 +1043,21 @@ static inline bool legacy_queue(struct sigpending *signals, int sig) return (sig < SIGRTMIN) && sigismember(&signals->signal, sig); } +/* + * When PF_EXITING is set the task is on the way out and has t::pending + * flushed already. Prevent queueing of PIDTYPE_PID signals as they would + * be leaked. + */ +static inline bool task_can_queue_signal(struct task_struct *t, enum pid_type type) +{ + lockdep_assert_held(&t->sighand->siglock); + + if (!(t->flags & PF_EXITING)) + return true; + + return type != PIDTYPE_PID; +} + static int __send_signal_locked(int sig, struct kernel_siginfo *info, struct task_struct *t, enum pid_type type, bool force) { @@ -1030,6 +1069,10 @@ static int __send_signal_locked(int sig, struct kernel_siginfo *info, lockdep_assert_held(&t->sighand->siglock); result = TRACE_SIGNAL_IGNORED; + + if (!task_can_queue_signal(t, type)) + goto ret; + if (!prepare_signal(sig, t, force)) goto ret; @@ -1980,11 +2023,25 @@ static inline struct task_struct *posixtimer_get_target(struct k_itimer *tmr) struct task_struct *t = pid_task(tmr->it_pid, tmr->it_pid_type); if (t && tmr->it_pid_type != PIDTYPE_PID && - same_thread_group(t, current) && !current->exit_state) + same_thread_group(t, current) && !(current->flags & PF_EXITING)) t = current; return t; } +/* + * Find the target task for the POSIX timer signal and prevent that a + * PIDTYPE_PID signal is queued on a task which has PF_EXITING set. + */ +static inline struct task_struct *posixtimer_get_unignore_target(struct k_itimer *tmr) +{ + struct task_struct *t = posixtimer_get_target(tmr); + + if (t && task_can_queue_signal(t, tmr->it_pid_type)) + return t; + + return NULL; +} + void posixtimer_send_sigqueue(struct k_itimer *tmr) { struct sigqueue *q = &tmr->sigq; @@ -2002,6 +2059,9 @@ void posixtimer_send_sigqueue(struct k_itimer *tmr) if (!likely(lock_task_sighand(t, &flags))) return; + if (!task_can_queue_signal(t, tmr->it_pid_type)) + goto unlock; + /* * Update @tmr::sigqueue_seq for posix timer signals with sighand * locked to prevent a race against dequeue_signal(). @@ -2093,6 +2153,7 @@ void posixtimer_send_sigqueue(struct k_itimer *tmr) result = TRACE_SIGNAL_DELIVERED; out: trace_signal_generate(sig, &q->info, t, tmr->it_pid_type != PIDTYPE_PID, result); +unlock: unlock_task_sighand(t, &flags); } @@ -2148,7 +2209,7 @@ static void posixtimer_sig_unignore(struct task_struct *tsk, int sig) * has exited by now, drop the reference count. */ guard(rcu)(); - target = posixtimer_get_target(tmr); + target = posixtimer_get_unignore_target(tmr); if (target) posixtimer_queue_sigqueue(&tmr->sigq, target, tmr->it_pid_type); else @@ -3132,42 +3193,36 @@ static void retarget_shared_pending(struct task_struct *tsk, sigset_t *which) void exit_signals(struct task_struct *tsk) { + LIST_HEAD(sigq_list); int group_stop = 0; - sigset_t unblocked; /* * @tsk is about to have PF_EXITING set - lock out users which - * expect stable threadgroup. + * expect a stable threadgroup. */ cgroup_threadgroup_change_begin(tsk); - if (thread_group_empty(tsk) || (tsk->signal->flags & SIGNAL_GROUP_EXIT)) { + scoped_guard(spinlock_irq, &tsk->sighand->siglock) { tsk->flags |= PF_EXITING; - cgroup_threadgroup_change_end(tsk); - return; - } - spin_lock_irq(&tsk->sighand->siglock); - /* - * From now this task is not visible for group-wide signals, - * see wants_signal(), do_signal_stop(). - */ - tsk->flags |= PF_EXITING; + sigqueue_dequeue_pending(&tsk->pending, &sigq_list); + + if (task_sigpending(tsk) && !thread_group_empty(tsk) && + !(tsk->signal->flags & SIGNAL_GROUP_EXIT)) { + sigset_t unblocked = tsk->blocked; + + signotset(&unblocked); + retarget_shared_pending(tsk, &unblocked); + + if (unlikely(tsk->jobctl & JOBCTL_STOP_PENDING) && + task_participate_group_stop(tsk)) + group_stop = CLD_STOPPED; + } + } cgroup_threadgroup_change_end(tsk); - if (!task_sigpending(tsk)) - goto out; - - unblocked = tsk->blocked; - signotset(&unblocked); - retarget_shared_pending(tsk, &unblocked); - - if (unlikely(tsk->jobctl & JOBCTL_STOP_PENDING) && - task_participate_group_stop(tsk)) - group_stop = CLD_STOPPED; -out: - spin_unlock_irq(&tsk->sighand->siglock); + flush_sigqueue_list(&sigq_list); /* * If group stop has completed, deliver the notification. This From acb03d3881818581052924a9bbbe92b8741ed448 Mon Sep 17 00:00:00 2001 From: Hyunwoo Kim Date: Fri, 11 Sep 2026 11:09:17 +0200 Subject: [PATCH 259/592] exec: Cleanup POSIX timers right after de_thread() A per-thread CPU timer holds a reference to the PID of the thread it is attached to and, while it is armed, its node is queued in that thread's posix_cputimers. The task is looked up by that PID. When a non-leader thread exec()s, de_thread() changes which task owns that PID. pid_task(timer->it.cpu.pid, PIDTYPE_PID) then returns NULL, but the node is still queued on tsk, which is alive. timer_lock_sighand() takes a failed lookup to mean that the node is already dequeued, so it has nothing to undo. begin_new_exec() calls posix_cpu_timers_exit(me) right after exec_task_namespaces() and that removes the leftover node, so the state normally stays invisible. But bprm->point_of_no_return is set before de_thread(), so if unshare_files(), set_mm_exe_file(), exec_mmap() or exec_task_namespaces() fails, the task dies before it gets there. exit_itimers() then frees the k_itimer while its node is still queued, and reaping tsk later erases that freed node from the rbtree. In short: the non-leader thread B the parent timer_create(CLOCK_THREAD_CPUTIME_ID) timer_settime() arm_timer() // the node is queued on B execve() de_thread(B) exchange_tids(B, leader) // B's PID now belongs to the leader release_task(leader) __exit_signal(leader) posix_cpu_timers_exit(leader) // cleans leader's queue, not B's __unhash_process(leader) // that PID has no task anymore exec_mmap() mmap_read_lock_killable(old_mm) kill(B, SIGKILL) // -EINTR get_signal() do_exit() exit_itimers() posix_timer_delete() posix_cpu_timer_del() posix_timer_unhash_and_free() // freed while still queued wait4() release_task(B) posix_cpu_timers_exit(B) cleanup_timerqueue() timerqueue_del() // use-after-free Move the POSIX timer cleanup right after de_thread() before any of the later failure conditions brings the task into do_exit(). [ tglx: Move the cleanup right after de_thread() ] Fixes: 55e8c8eb2c7b ("posix-cpu-timers: Store a reference to a pid not a task") Signed-off-by: Hyunwoo Kim Signed-off-by: Thomas Gleixner Tested-by: Kijo Park Reviewed-by: Oleg Nesterov Reviewed-by: Frederic Weisbecker Cc: stable@vger.kernel.org Link: https://patch.msgid.link/ao7Q8miiuLAPVnWv@v4bel Link: https://patch.msgid.link/20260911090541.627712075@kernel.org --- fs/exec.c | 29 +++++++++++++++++++++-------- 1 file changed, 21 insertions(+), 8 deletions(-) diff --git a/fs/exec.c b/fs/exec.c index d3081c8f7c10..819643408e6d 100644 --- a/fs/exec.c +++ b/fs/exec.c @@ -1115,6 +1115,17 @@ static struct file *bprm_identity_file(const struct linux_binprm *bprm) return bprm->file; } +static void posixtimer_exec(struct task_struct *me) +{ +#ifdef CONFIG_POSIX_TIMERS + spin_lock_irq(&me->sighand->siglock); + posix_cpu_timers_exit(me); + spin_unlock_irq(&me->sighand->siglock); + exit_itimers(me); + flush_itimer_signals(); +#endif +} + /* * Calling this is the point of no return. None of the failures will be * seen by userspace since either the process is already taking a fatal @@ -1152,6 +1163,16 @@ int begin_new_exec(struct linux_binprm * bprm) retval = de_thread(me); if (retval) goto out; + + /* + * This must be done here to ensure that POSIX CPU timers which were + * armed on the current task are dequeued from me::posix_cputimers. + * Otherwise in case of a TID switch the deletion of the related POSIX + * timer would not remove an enqueued timer because the TID lookup + * of the old TID fails. + */ + posixtimer_exec(me); + /* see the comment in check_unsafe_exec() */ current->fs->in_exec = 0; /* @@ -1206,14 +1227,6 @@ int begin_new_exec(struct linux_binprm * bprm) if (retval) goto out_unlock; -#ifdef CONFIG_POSIX_TIMERS - spin_lock_irq(&me->sighand->siglock); - posix_cpu_timers_exit(me); - spin_unlock_irq(&me->sighand->siglock); - exit_itimers(me); - flush_itimer_signals(); -#endif - /* * Make the signal table private. */ From ef56085dfd1df3a53ecce8a4ff440cf6d67f430d Mon Sep 17 00:00:00 2001 From: Andre Przywara Date: Mon, 14 Sep 2026 12:07:44 +0200 Subject: [PATCH 260/592] pinctrl: sunxi: A523: fix voltage withstand encoding The Allwinner A523 uses the same GPIO voltage "withstand" programming (setting the input level voltage thresholds) as the previous SoCs, but for some odd reason inverts the encoding of 1.8V vs. 3.3V. Add a new bias voltage type to note this difference, and select it for the A523. At the same time also use the newer "CTL" version, which in addition allows to turn off the withstand programming for I/O voltages other than exact 1.8V or 3.3V (for instance for 2.5V sometimes used for Ethernet PHYs). The A523 has that enable register, but didn't use it so far. This fixes eMMC and reportedly Ethernet operation on some A523 boards. Fixes: 648be4cd9517 ("pinctrl: sunxi: Add support for the Allwinner A523") Signed-off-by: Andre Przywara Tested-by: Per Larsson Tested-by: Juan Manuel Lopez Carrillo Reviewed-by: Chen-Yu Tsai Tested-by: Chen-Yu Tsai # Fixes eMMC on Orange Pi 4A Signed-off-by: Linus Walleij --- drivers/pinctrl/sunxi/pinctrl-sun55i-a523-r.c | 2 +- drivers/pinctrl/sunxi/pinctrl-sun55i-a523.c | 2 +- drivers/pinctrl/sunxi/pinctrl-sunxi.c | 6 ++++++ drivers/pinctrl/sunxi/pinctrl-sunxi.h | 2 ++ 4 files changed, 10 insertions(+), 2 deletions(-) diff --git a/drivers/pinctrl/sunxi/pinctrl-sun55i-a523-r.c b/drivers/pinctrl/sunxi/pinctrl-sun55i-a523-r.c index 462aa1c4a5fa..e27e4945def2 100644 --- a/drivers/pinctrl/sunxi/pinctrl-sun55i-a523-r.c +++ b/drivers/pinctrl/sunxi/pinctrl-sun55i-a523-r.c @@ -26,7 +26,7 @@ static const u8 a523_r_irq_bank_muxes[SUNXI_PINCTRL_MAX_BANKS] = static struct sunxi_pinctrl_desc a523_r_pinctrl_data = { .irq_banks = ARRAY_SIZE(a523_r_irq_bank_map), .irq_bank_map = a523_r_irq_bank_map, - .io_bias_cfg_variant = BIAS_VOLTAGE_PIO_POW_MODE_SEL, + .io_bias_cfg_variant = BIAS_VOLTAGE_PIO_POW_MODE_CTL_INV, .pin_base = PL_BASE, }; diff --git a/drivers/pinctrl/sunxi/pinctrl-sun55i-a523.c b/drivers/pinctrl/sunxi/pinctrl-sun55i-a523.c index b6f78f1f30ac..88d8acd5bc24 100644 --- a/drivers/pinctrl/sunxi/pinctrl-sun55i-a523.c +++ b/drivers/pinctrl/sunxi/pinctrl-sun55i-a523.c @@ -26,7 +26,7 @@ static const u8 a523_irq_bank_muxes[SUNXI_PINCTRL_MAX_BANKS] = static struct sunxi_pinctrl_desc a523_pinctrl_data = { .irq_banks = ARRAY_SIZE(a523_irq_bank_map), .irq_bank_map = a523_irq_bank_map, - .io_bias_cfg_variant = BIAS_VOLTAGE_PIO_POW_MODE_SEL, + .io_bias_cfg_variant = BIAS_VOLTAGE_PIO_POW_MODE_CTL_INV, }; static int a523_pinctrl_probe(struct platform_device *pdev) diff --git a/drivers/pinctrl/sunxi/pinctrl-sunxi.c b/drivers/pinctrl/sunxi/pinctrl-sunxi.c index 2881a83be99c..31cd142ce0f7 100644 --- a/drivers/pinctrl/sunxi/pinctrl-sunxi.c +++ b/drivers/pinctrl/sunxi/pinctrl-sunxi.c @@ -728,6 +728,7 @@ static int sunxi_pinctrl_set_io_bias_cfg(struct sunxi_pinctrl *pctl, { unsigned short bank; unsigned long flags; + bool inverted = false; u32 val, reg; int uV; @@ -766,6 +767,9 @@ static int sunxi_pinctrl_set_io_bias_cfg(struct sunxi_pinctrl *pctl, reg &= ~IO_BIAS_MASK; writel(reg | val, pctl->membase + sunxi_grp_config_reg(pin)); return 0; + case BIAS_VOLTAGE_PIO_POW_MODE_CTL_INV: + inverted = true; + fallthrough; case BIAS_VOLTAGE_PIO_POW_MODE_CTL: val = uV > 1800000 && uV <= 2500000 ? BIT(bank) : 0; @@ -780,6 +784,8 @@ static int sunxi_pinctrl_set_io_bias_cfg(struct sunxi_pinctrl *pctl, fallthrough; case BIAS_VOLTAGE_PIO_POW_MODE_SEL: val = uV <= 1800000 ? 1 : 0; + if (inverted) + val = !val; raw_spin_lock_irqsave(&pctl->lock, flags); reg = readl(pctl->membase + pctl->pow_mod_sel_offset); diff --git a/drivers/pinctrl/sunxi/pinctrl-sunxi.h b/drivers/pinctrl/sunxi/pinctrl-sunxi.h index 498e88a19f71..8bd00c6ff628 100644 --- a/drivers/pinctrl/sunxi/pinctrl-sunxi.h +++ b/drivers/pinctrl/sunxi/pinctrl-sunxi.h @@ -117,8 +117,10 @@ enum sunxi_desc_bias_voltage { * Bias voltage is set through PIO_POW_MOD_SEL_REG * and PIO_POW_MOD_CTL_REG register, as seen on * A100 and D1 SoC, for example. + * Some SoCs invert the encoding for 1.8V vs. 3.3V. */ BIAS_VOLTAGE_PIO_POW_MODE_CTL, + BIAS_VOLTAGE_PIO_POW_MODE_CTL_INV, }; struct sunxi_desc_function { From 1d9bb9c870632adea249cfdec49ac37e6f069164 Mon Sep 17 00:00:00 2001 From: Myeonghun Pak Date: Sun, 13 Sep 2026 00:03:12 -0400 Subject: [PATCH 261/592] pinctrl: single: free the IRQ on domain creation failure pcs_irq_init_chained_handler() requests a shared IRQ on affected SoCs, but its domain creation failure path only removes a chained handler. That does not release the action installed by request_irq(). The probe can continue without interrupt support while leaving the shared IRQ action registered. Use pcs_irq_free() to undo the appropriate type of handler registration. At this point pcs->domain is NULL, so the helper only releases the parent IRQ handler. Then mark the IRQ invalid, as the other initialization error paths already do, to prevent another release from a later probe unwind or remove. This issue was identified during our ongoing static-analysis research while reviewing kernel code. Fixes: 3e6cee1786a1 ("pinctrl: single: Add support for wake-up interrupts") Cc: stable@vger.kernel.org Assisted-by: LLM Co-developed-by: Ijae Kim Signed-off-by: Ijae Kim Signed-off-by: Myeonghun Pak Signed-off-by: Linus Walleij --- drivers/pinctrl/pinctrl-single.c | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-) diff --git a/drivers/pinctrl/pinctrl-single.c b/drivers/pinctrl/pinctrl-single.c index 4d5f85b7e6bb..e0eb7240a985 100644 --- a/drivers/pinctrl/pinctrl-single.c +++ b/drivers/pinctrl/pinctrl-single.c @@ -1628,7 +1628,8 @@ static int pcs_irq_init_chained_handler(struct pcs_device *pcs, &pcs_irqdomain_ops, pcs_soc); if (!pcs->domain) { - irq_set_chained_handler(pcs_soc->irq, NULL); + pcs_irq_free(pcs); + pcs_soc->irq = -1; return -EINVAL; } From e922bad8b2d5028c51a096d083fea41cd0987154 Mon Sep 17 00:00:00 2001 From: Itai Handler Date: Thu, 10 Sep 2026 20:48:32 +0300 Subject: [PATCH 262/592] spi: spi-zynqmp-gqspi: stop the controller on shutdown The driver has no ->shutdown, and platform_drv_shutdown() has no fallback of its own. Unlike pci_device_shutdown(), which clears bus mastering when kexec_in_progress, nothing on the platform bus disarms a device that can still write to memory. The normal kexec path never calls ->suspend either, so the quiesce in zynqmp_qspi_suspend() is not reached. A controller that is still executing a DMA read may therefore keep writing to memory across a kexec. QSPIDMA_DST_ADDR still points at memory owned by the kernel that called kexec, DST_SIZE is non-zero and the flash is still clocked, so data can keep landing in RAM while the new kernel is being relocated, and after it has started executing. That destination is a physical address which means nothing to the new kernel, so the writes can corrupt whatever now occupies it: kernel text or data, page tables, or the initrd. Nothing reports an error and the resulting behaviour is undefined. This can be observed by reading GQSPI_EN (offset 0x114) and QSPIDMA_DST_ADDR/SIZE/STS/CTRL (offsets 0x800 to 0x80c) early in the new kernel, before the driver probes: without this patch GQSPI_EN reads 1 and QSPIDMA_DST_ADDR still points into the previous kernel's memory. Add a ->shutdown that stops the controller the way zynqmp_qspi_suspend() already does. spi_controller_suspend() stops the queue, waits for a message that is already executing and makes any later transfer fail with -ESHUTDOWN, so nothing can be cut short by the register write that follows. It may sleep, which is fine here: device_shutdown() runs in process context. Unlike ->suspend this cannot abort on error, because a controller left mastering the bus is worse than a truncated transfer, so a failure to drain is only logged. GQSPI_EN_OFST is then cleared, as zynqmp_qspi_remove() and zynqmp_qspi_suspend() already do. Skip that write only when pm_runtime_get_if_in_use() returns 0, i.e. runtime suspended: the clocks are gated, so the registers are unreachable and the controller cannot be mastering the bus. A negative return is not the same thing - it is what the CONFIG_PM=n stub always returns, and there probe() has enabled pclk and refclk for good, so the controller is running and must be stopped. Fixes: dfe11a11d523 ("spi: Add support for Zynq Ultrascale+ MPSoC GQSPI controller") Cc: stable@vger.kernel.org Signed-off-by: Itai Handler Link: https://patch.msgid.link/20260910174832.873352-1-itai.handler@gmail.com Signed-off-by: Mark Brown --- drivers/spi/spi-zynqmp-gqspi.c | 34 ++++++++++++++++++++++++++++++++++ 1 file changed, 34 insertions(+) diff --git a/drivers/spi/spi-zynqmp-gqspi.c b/drivers/spi/spi-zynqmp-gqspi.c index 4d55090fa443..15e9d3ef8839 100644 --- a/drivers/spi/spi-zynqmp-gqspi.c +++ b/drivers/spi/spi-zynqmp-gqspi.c @@ -1373,11 +1373,45 @@ static void zynqmp_qspi_remove(struct platform_device *pdev) clk_disable_unprepare(xqspi->pclk); } +static void zynqmp_qspi_shutdown(struct platform_device *pdev) +{ + struct zynqmp_qspi *xqspi = platform_get_drvdata(pdev); + int ret; + + /* + * Stop the queue and reject any later transfer first, so the write + * below cannot cut into a message that is still being executed. + * Unlike ->suspend this cannot abort on error: a controller left + * mastering the bus is worse than a truncated transfer. + */ + ret = spi_controller_suspend(xqspi->ctlr); + if (ret) + dev_warn(&pdev->dev, "could not stop the queue: %d\n", ret); + + /* + * Only a runtime suspended controller can be left alone: its clocks + * are gated, so it cannot be mastering the bus, and its registers + * must not be accessed either. Any other answer means it may be + * running and has to be stopped. In particular, on a kernel built + * without runtime PM this returns -EINVAL, and there the clocks + * enabled in probe() are never gated at all. + */ + ret = pm_runtime_get_if_in_use(&pdev->dev); + if (!ret) + return; + + zynqmp_gqspi_write(xqspi, GQSPI_EN_OFST, 0x0); + + if (ret > 0) + pm_runtime_put_noidle(&pdev->dev); +} + MODULE_DEVICE_TABLE(of, zynqmp_qspi_of_match); static struct platform_driver zynqmp_qspi_driver = { .probe = zynqmp_qspi_probe, .remove = zynqmp_qspi_remove, + .shutdown = zynqmp_qspi_shutdown, .driver = { .name = "zynqmp-qspi", .of_match_table = zynqmp_qspi_of_match, From f033482d76a9f18080c7a40c5f9c678bd7adc8f3 Mon Sep 17 00:00:00 2001 From: Christiano Amora Date: Wed, 16 Sep 2026 10:46:22 -0300 Subject: [PATCH 263/592] Bluetooth: SMP: reject Security Request over BR/EDR Bose QC Ultra Headphones (dual-mode, same public address on both transports) occasionally send an SMP Security Request on the BR/EDR SMP fixed channel right after the ACL link is encrypted. The kernel handles it as if it were an LE link: smp_cmd_security_req() has no transport check, smp_ltk_encrypt() looks up an LTK with the ACL connection's dst_type, and hci_find_ltk() matches the peer's LE LTK because the LE public address type is stored as ADDR_LE_DEV_PUBLIC (0), the same value as BDADDR_BREDR. HCI_OP_LE_START_ENC is then issued on the ACL handle, the controller rejects it with Invalid HCI Command Parameters, and hci_cs_le_start_enc() disconnects the link with HCI_ERROR_AUTH_FAILURE. The headphones drop within a second of connecting, before any profile is up; a manual reconnect works. btmon (MediaTek MT7922, kernel 7.0.12): > HCI Event: Encryption Change (0x08) plen 4 Status: Success (0x00) Handle: 50 Address: BC:87:FA:47:73:5E (Bose Corporation) Encryption: Enabled with AES-CCM (0x02) > ACL Data RX: Handle 50 flags 0x02 dlen 6 BR/EDR SMP: Security Request (0x0b) len 1 Authentication requirement: No bonding, No MITM, SC (0x08) < HCI Command: LE Start Encryption (0x08|0x0019) plen 28 Handle: 50 Address: BC:87:FA:47:73:5E (Bose Corporation) > HCI Event: Command Status (0x0f) plen 4 LE Start Encryption (0x08|0x0019) ncmd 1 Status: Invalid HCI Command Parameters (0x12) < HCI Command: Disconnect (0x01|0x0006) plen 3 Handle: 50 Address: BC:87:FA:47:73:5E (Bose Corporation) Reason: Authentication Failure (0x05) SMP over BR/EDR is limited to cross-transport key derivation; the Security Request procedure (Core Specification Vol 3, Part H, Section 2.4.6, PDU in Section 3.6.7) has no BR/EDR counterpart. Reply with Pairing Failed / Command Not Supported on a non-LE link, before the PDU is parsed, and keep the connection. The reply is sent directly rather than through smp_failure(): rejecting a command on the wrong transport is not an authentication failure, and MGMT_EV_AUTH_FAILED would make bluetoothd disconnect the device. Tested on the affected host (kernel 7.0.12, MediaTek MT7922, Bose QC Ultra) with the patched module built out of tree: 7 days and 49 reconnects without a drop, against 2 drops in the 3 days before the patch. Every disconnect in that week had a userspace or remote reason. Fixes: b5ae344d4c0f ("Bluetooth: Add full SMP BR/EDR support") Assisted-by: LLM Signed-off-by: Christiano Amora Signed-off-by: Luiz Augusto von Dentz --- net/bluetooth/smp.c | 17 +++++++++++++++++ 1 file changed, 17 insertions(+) diff --git a/net/bluetooth/smp.c b/net/bluetooth/smp.c index 6091c47cb002..d23f9d0729c4 100644 --- a/net/bluetooth/smp.c +++ b/net/bluetooth/smp.c @@ -2269,6 +2269,23 @@ static u8 smp_cmd_security_req(struct l2cap_conn *conn, struct sk_buff *skb) bt_dev_dbg(hdev, "conn %p", conn); + /* SMP over BR/EDR only covers cross-transport key derivation; the + * Security Request procedure has no BR/EDR counterpart. Reject it + * here, otherwise smp_ltk_encrypt() finds the peer's LE LTK + * (ADDR_LE_DEV_PUBLIC and BDADDR_BREDR are both 0) and issues + * HCI_OP_LE_START_ENC on the ACL handle, which the controller + * rejects and hci_cs_le_start_enc() turns into a disconnect. Reply + * without smp_failure(): this is not an authentication failure, and + * MGMT_EV_AUTH_FAILED would make bluetoothd drop the device. + */ + if (hcon->type != LE_LINK) { + u8 reason = SMP_CMD_NOTSUPP; + + smp_send_cmd(conn, SMP_CMD_PAIRING_FAIL, sizeof(reason), + &reason); + return 0; + } + if (skb->len < sizeof(*rp)) return SMP_INVALID_PARAMS; From f2bbb36426581045a8bf7793da5419b9375e4348 Mon Sep 17 00:00:00 2001 From: Zijun Hu Date: Tue, 15 Sep 2026 19:17:18 -0700 Subject: [PATCH 264/592] Bluetooth: btnxpuart: Fix skb leak in nxp_process_fw_dump() When CONFIG_DEV_COREDUMP=n, hci_devcd_append() returns -EOPNOTSUPP without freeing its skb argument. This leaks the cloned skb and also prevents nxp_set_ind_reset() from being called to perform recovery. Fix by guarding the hci_devcd_append(hdev, skb_clone(skb, GFP_ATOMIC)) call with IS_ENABLED(CONFIG_DEV_COREDUMP). Fixes: 998e447f443f ("Bluetooth: btnxpuart: Add support for HCI coredump feature") Signed-off-by: Zijun Hu Signed-off-by: Luiz Augusto von Dentz --- drivers/bluetooth/btnxpuart.c | 8 +++++--- 1 file changed, 5 insertions(+), 3 deletions(-) diff --git a/drivers/bluetooth/btnxpuart.c b/drivers/bluetooth/btnxpuart.c index 25e7b41b349f..4e23d71d00e8 100644 --- a/drivers/bluetooth/btnxpuart.c +++ b/drivers/bluetooth/btnxpuart.c @@ -1388,9 +1388,11 @@ static int nxp_process_fw_dump(struct hci_dev *hdev, struct sk_buff *skb) msecs_to_jiffies(20000)); } - err = hci_devcd_append(hdev, skb_clone(skb, GFP_ATOMIC)); - if (err < 0) - goto free_skb; + if (IS_ENABLED(CONFIG_DEV_COREDUMP)) { + err = hci_devcd_append(hdev, skb_clone(skb, GFP_ATOMIC)); + if (err < 0) + goto free_skb; + } if (buf_len == 0) { bt_dev_warn(hdev, "==== FW dump complete ==="); From 71af682ba4692c2ed9ace4c3d4ca462ae368c029 Mon Sep 17 00:00:00 2001 From: Lee Jones Date: Tue, 15 Sep 2026 12:08:22 +0000 Subject: [PATCH 265/592] Bluetooth: mgmt: Dequeue pending mesh_send_sync entries on cancel In send_cancel(), pending mesh_tx objects are removed from the hdev->mesh_pending list and freed via mesh_send_complete(). However, if a mesh transmission was already queued onto hdev->cmd_sync_work_list via mesh_next(), the queued entry retains a raw pointer to mesh_tx. When hci_cmd_sync_work later processes the entry, it attempts to execute mesh_send_sync and its destroy callback mesh_send_start_complete using the already freed mesh_tx pointer, leading to a use-after-free. Fix this by invoking hci_cmd_sync_dequeue() for mesh_send_sync on the target mesh_tx before completing it. If the entry is found and dequeued, its destroy callback will complete and free the object; otherwise, mesh_send_complete() is called directly. Additionally, ensure the transmission queue advances after cancellation or errors. In mesh_send_start_complete(), call mesh_next() on error unless err is -ECANCELED, because hci_cmd_sync_dequeue() holds hdev->cmd_sync_work_lock and calling mesh_next() synchronously would deadlock. Instead, advance the queue in send_cancel() once the lock is released and if no transmission is in progress. Fixes: b338d91703fa ("Bluetooth: Implement support for Mesh") Signed-off-by: Lee Jones Signed-off-by: Luiz Augusto von Dentz --- net/bluetooth/mgmt.c | 19 +++++++++++++++---- 1 file changed, 15 insertions(+), 4 deletions(-) diff --git a/net/bluetooth/mgmt.c b/net/bluetooth/mgmt.c index ac4864e56ec7..f740e745ae78 100644 --- a/net/bluetooth/mgmt.c +++ b/net/bluetooth/mgmt.c @@ -2316,6 +2316,8 @@ static void mesh_send_start_complete(struct hci_dev *hdev, void *data, int err) hci_dev_clear_flag(hdev, HCI_MESH_SENDING); /* Send Complete Error Code for handle */ mesh_send_complete(hdev, mesh_tx, false); + if (err != -ECANCELED) + mesh_next(hdev, NULL, 0); return; } @@ -2425,19 +2427,28 @@ static int send_cancel(struct hci_dev *hdev, void *data) do { mesh_tx = mgmt_mesh_next(hdev, cmd->sk); - if (mesh_tx) - mesh_send_complete(hdev, mesh_tx, false); + if (mesh_tx) { + if (!hci_cmd_sync_dequeue(hdev, mesh_send_sync, + mesh_tx, NULL)) + mesh_send_complete(hdev, mesh_tx, false); + } } while (mesh_tx); } else { mesh_tx = mgmt_mesh_find(hdev, cancel->handle); - if (mesh_tx && mesh_tx->sk == cmd->sk) - mesh_send_complete(hdev, mesh_tx, false); + if (mesh_tx && mesh_tx->sk == cmd->sk) { + if (!hci_cmd_sync_dequeue(hdev, mesh_send_sync, + mesh_tx, NULL)) + mesh_send_complete(hdev, mesh_tx, false); + } } mgmt_cmd_complete(cmd->sk, hdev->id, MGMT_OP_MESH_SEND_CANCEL, 0, NULL, 0); + if (!hci_dev_test_flag(hdev, HCI_MESH_SENDING)) + mesh_next(hdev, NULL, 0); + return 0; } From e06d549fcd4a0ba381ed67ddf1ab3c7a6ca4314c Mon Sep 17 00:00:00 2001 From: Aldo Ariel Panzardo Date: Tue, 15 Sep 2026 13:04:29 -0300 Subject: [PATCH 266/592] Bluetooth: hci_conn: fix CIS hold ownership on reuse Commit 69997d50ec57 ("Bluetooth: ISO: handle bound CIS cleanup via hci_conn") made hci_bind_cis() and hci_connect_cis() return a connection with one hold for the ISO layer. hci_bind_cis() currently takes that hold only after configuring a CIS, so its BT_CONNECTED and matching BT_BOUND paths return a bare lookup result. Its configuration failure path can likewise call hci_conn_drop() before taking a hold. Take the hold before any state-dependent return or configuration error so every successful return follows the documented ownership contract and every error drop is balanced. hci_connect_cis() also assumes hci_conn_link() always takes a new CIS hold before dropping the one returned by hci_bind_cis(). However, the helper returns an existing link without taking another hold. In that case, preserve the CIS hold for the caller and drop the redundant LE hold because the existing link already owns its parent hold. Returning early also avoids changing an existing CIS back to BT_CONNECT. Fixes: 69997d50ec57 ("Bluetooth: ISO: handle bound CIS cleanup via hci_conn") Cc: stable@vger.kernel.org Signed-off-by: Aldo Ariel Panzardo Signed-off-by: Luiz Augusto von Dentz --- net/bluetooth/hci_conn.c | 9 ++++++++- 1 file changed, 8 insertions(+), 1 deletion(-) diff --git a/net/bluetooth/hci_conn.c b/net/bluetooth/hci_conn.c index fa72cf8aaa7a..e32bb9c342a5 100644 --- a/net/bluetooth/hci_conn.c +++ b/net/bluetooth/hci_conn.c @@ -2079,6 +2079,8 @@ struct hci_conn *hci_bind_cis(struct hci_dev *hdev, bdaddr_t *dst, cis->conn_timeout = timeout; } + hci_conn_hold(cis); + if (cis->state == BT_CONNECTED) return cis; @@ -2120,7 +2122,6 @@ struct hci_conn *hci_bind_cis(struct hci_dev *hdev, bdaddr_t *dst, return ERR_PTR(-EINVAL); } - hci_conn_hold(cis); cis->state = BT_BOUND; return cis; @@ -2497,6 +2498,12 @@ struct hci_conn *hci_connect_cis(struct hci_dev *hdev, bdaddr_t *dst, return cis; } + /* The existing link already owns the hold on its parent. */ + if (cis->link) { + hci_conn_drop(le); + return cis; + } + link = hci_conn_link(le, cis); hci_conn_drop(cis); if (!link) { From 0fcd4dad555c96e0bd3a1b8c569f989be85c7341 Mon Sep 17 00:00:00 2001 From: Aldo Ariel Panzardo Date: Tue, 15 Sep 2026 13:04:30 -0300 Subject: [PATCH 267/592] Bluetooth: ISO: release unused CIS holds after channel attach hci_bind_cis() and hci_connect_cis() return one hci_conn hold for the ISO layer. A new channel association consumes that hold, which is eventually released by iso_conn_free(). There are two cases where iso_chan_add() does not create an association: it returns success when the socket is already attached to the same iso_conn, and it returns -EBUSY when another socket is attached. The hold returned for the current call is unused in both cases. This occurs when deferred setup calls iso_connect_cis() again for its existing socket, or when another socket attempts to reuse the CIS. Detect the idempotent case while the connection is locked and release the unused hold after iso_chan_add(). Also release it on -EBUSY. Do not drop it for other errors: a newly allocated iso_conn releases the transferred hold when its last temporary reference is put. Fixes: 69997d50ec57 ("Bluetooth: ISO: handle bound CIS cleanup via hci_conn") Cc: stable@vger.kernel.org Signed-off-by: Aldo Ariel Panzardo Signed-off-by: Luiz Augusto von Dentz --- net/bluetooth/iso.c | 7 +++++++ 1 file changed, 7 insertions(+) diff --git a/net/bluetooth/iso.c b/net/bluetooth/iso.c index eb99653f33f9..7657c2a0abbf 100644 --- a/net/bluetooth/iso.c +++ b/net/bluetooth/iso.c @@ -496,6 +496,7 @@ static int iso_connect_cis(struct sock *sk) struct hci_dev *hdev; bdaddr_t src, dst; u8 src_type; + bool already_attached; int err; lock_sock(sk); @@ -568,8 +569,14 @@ static int iso_connect_cis(struct sock *sk) goto unlock; } + iso_conn_lock(conn); + already_attached = iso_pi(sk)->conn == conn && conn->sk == sk; + iso_conn_unlock(conn); + err = iso_chan_add(conn, sk, NULL); iso_conn_put(conn); + if (already_attached || err == -EBUSY) + hci_conn_drop(hcon); if (err) goto unlock; From e93fad891c72deb84cae49430163b384ebcc92b1 Mon Sep 17 00:00:00 2001 From: Aldo Ariel Panzardo Date: Tue, 15 Sep 2026 13:03:58 -0300 Subject: [PATCH 268/592] Bluetooth: hci_sock: reject out-of-range OCF values The raw HCI socket security filter has 128 OCF bits per supported OGF, but masks the 10-bit OCF with 127 before looking up the command. An unprivileged socket can therefore submit a reserved OCF that aliases an allowlisted command modulo 128. A conforming controller should reject reserved opcodes. Nevertheless, the security decision must apply to the opcode that will actually be sent, especially since controller-specific behavior is outside the host stack's control. Reject OCF values that cannot be represented by the security filter instead of aliasing them onto an unrelated command. Fixes: 1da177e4c3f4 ("Linux-2.6.12-rc2") Cc: stable@vger.kernel.org Signed-off-by: Aldo Ariel Panzardo Signed-off-by: Luiz Augusto von Dentz --- net/bluetooth/hci_sock.c | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-) diff --git a/net/bluetooth/hci_sock.c b/net/bluetooth/hci_sock.c index 070ca388f9ac..6413593eee51 100644 --- a/net/bluetooth/hci_sock.c +++ b/net/bluetooth/hci_sock.c @@ -1881,7 +1881,8 @@ static int hci_sock_sendmsg(struct socket *sock, struct msghdr *msg, u16 ocf = hci_opcode_ocf(opcode); if (((ogf > HCI_SFLT_MAX_OGF) || - !hci_test_bit(ocf & HCI_FLT_OCF_BITS, + (ocf > HCI_FLT_OCF_BITS) || + !hci_test_bit(ocf, &hci_sec_filter.ocf_mask[ogf])) && !capable(CAP_NET_RAW)) { err = -EPERM; From b0a6cf99afd57a39598b1beca0e86ef5004980de Mon Sep 17 00:00:00 2001 From: Aldo Ariel Panzardo Date: Tue, 15 Sep 2026 13:03:07 -0300 Subject: [PATCH 269/592] Bluetooth: hci_sock: validate event length before filtering is_filtered_packet() reads the event code from skb->data[0] without first checking that the skb is nonempty. When an opcode filter is configured, it also reads the command opcode at offsets 3 or 4 without checking that a Command Complete or Command Status event is long enough. hci_send_to_sock() invokes the filter before hci_event_packet() validates the event header. A malformed event supplied by a controller or a vhci device can therefore cause an out-of-bounds read. Keep the unmasked event code for the opcode checks. The masked value is needed for the 64-bit event bitmap, but using it to identify command events aliases event codes above 0x3f. In particular, Synchronous Train Complete (0x4f) was treated as Command Status (0x0f) even though its payload has no opcode. Reject actual command events that are too short for the field being inspected. A truncated command event cannot match a configured opcode. Fixes: 1da177e4c3f4 ("Linux-2.6.12-rc2") Cc: stable@vger.kernel.org Signed-off-by: Aldo Ariel Panzardo Signed-off-by: Luiz Augusto von Dentz --- net/bluetooth/hci_sock.c | 17 ++++++++++++++--- 1 file changed, 14 insertions(+), 3 deletions(-) diff --git a/net/bluetooth/hci_sock.c b/net/bluetooth/hci_sock.c index 6413593eee51..6d56c77741e1 100644 --- a/net/bluetooth/hci_sock.c +++ b/net/bluetooth/hci_sock.c @@ -164,6 +164,7 @@ static bool is_filtered_packet(struct sock *sk, struct sk_buff *skb) { struct hci_filter *flt; int flt_type, flt_event; + u8 event; /* Apply filter */ flt = &hci_pi(sk)->filter; @@ -177,7 +178,11 @@ static bool is_filtered_packet(struct sock *sk, struct sk_buff *skb) if (hci_skb_pkt_type(skb) != HCI_EVENT_PKT) return false; - flt_event = (*(__u8 *)skb->data & HCI_FLT_EVENT_BITS); + if (skb->len < 1) + return true; + + event = *(__u8 *)skb->data; + flt_event = event & HCI_FLT_EVENT_BITS; if (!hci_test_bit(flt_event, &flt->event_mask)) return true; @@ -186,11 +191,17 @@ static bool is_filtered_packet(struct sock *sk, struct sk_buff *skb) if (!flt->opcode) return false; - if (flt_event == HCI_EV_CMD_COMPLETE && + if (event == HCI_EV_CMD_COMPLETE && skb->len < 5) + return true; + + if (event == HCI_EV_CMD_COMPLETE && flt->opcode != get_unaligned((__le16 *)(skb->data + 3))) return true; - if (flt_event == HCI_EV_CMD_STATUS && + if (event == HCI_EV_CMD_STATUS && skb->len < 6) + return true; + + if (event == HCI_EV_CMD_STATUS && flt->opcode != get_unaligned((__le16 *)(skb->data + 4))) return true; From 4c94557dd02569efa6c1072a0439addaef9a5224 Mon Sep 17 00:00:00 2001 From: Aldo Ariel Panzardo Date: Tue, 15 Sep 2026 13:03:32 -0300 Subject: [PATCH 270/592] Bluetooth: ISO: balance the parent hold in hci_bind_bis() hci_conn_link() takes a lifetime reference to its parent with hci_conn_get(), but only takes an operational hold on the child. hci_conn_unlink() later balances both a hold and a reference on the parent. The SCO and CIS paths pass a parent acquired from a connect helper, so it already has a hold. For an additional BIS, hci_bind_bis() obtains the parent from hci_conn_hash_lookup_big(), which returns a bare pointer. Unlinking the child then drops the parent's existing hold and can schedule it for disconnection while its socket is still using it. Take a hold on the parent before linking it and drop that hold if linking fails. A successful link transfers the hold to hci_conn_unlink(). Fixes: fa224d0c094a ("Bluetooth: ISO: Reassociate a socket with an active BIS") Cc: stable@vger.kernel.org Signed-off-by: Aldo Ariel Panzardo Signed-off-by: Luiz Augusto von Dentz --- net/bluetooth/hci_conn.c | 5 ++++- 1 file changed, 4 insertions(+), 1 deletion(-) diff --git a/net/bluetooth/hci_conn.c b/net/bluetooth/hci_conn.c index e32bb9c342a5..96195d2fd10f 100644 --- a/net/bluetooth/hci_conn.c +++ b/net/bluetooth/hci_conn.c @@ -2375,10 +2375,13 @@ struct hci_conn *hci_bind_bis(struct hci_dev *hdev, bdaddr_t *dst, __u8 sid, parent = hci_conn_hash_lookup_big(hdev, conn->iso_qos.bcast.big); if (parent && parent != conn) { + hci_conn_hold(parent); link = hci_conn_link(parent, conn); hci_conn_drop(conn); - if (!link) + if (!link) { + hci_conn_drop(parent); return ERR_PTR(-ENOLINK); + } } return conn; From 6c78a213d9070b610c7f418af2c25b66180b7e37 Mon Sep 17 00:00:00 2001 From: Aldo Ariel Panzardo Date: Tue, 15 Sep 2026 13:02:39 -0300 Subject: [PATCH 271/592] Bluetooth: L2CAP: validate frame length before control and FCS access l2cap_data_rcv() unpacks either a two-byte or four-byte control field without first ensuring that it is present. A short ERTM or streaming-mode frame can therefore cause an out-of-bounds read. There is a second short-frame case when CRC16 is enabled. After the control field is pulled, l2cap_check_fcs() subtracts two from skb->len without checking it. If fewer than two bytes remain, the subtraction wraps; skb_trim() leaves the buffer unchanged and the subsequent FCS load reads past the logical end of the frame. Validate that the frame contains both its control field and, when enabled, its FCS before either field is accessed. Fixes: 1c2acffb76d4 ("Bluetooth: Add initial support for ERTM packets transfers") Fixes: fcc203c30d72 ("Bluetooth: Add support for FCS option to L2CAP") Cc: stable@vger.kernel.org Signed-off-by: Aldo Ariel Panzardo Signed-off-by: Luiz Augusto von Dentz --- net/bluetooth/l2cap_core.c | 10 +++++++++- 1 file changed, 9 insertions(+), 1 deletion(-) diff --git a/net/bluetooth/l2cap_core.c b/net/bluetooth/l2cap_core.c index 644e31160d55..aaa2a1cd489a 100644 --- a/net/bluetooth/l2cap_core.c +++ b/net/bluetooth/l2cap_core.c @@ -6702,9 +6702,17 @@ static int l2cap_stream_rx(struct l2cap_chan *chan, struct l2cap_ctrl *control, static int l2cap_data_rcv(struct l2cap_chan *chan, struct sk_buff *skb) { struct l2cap_ctrl *control = &bt_cb(skb)->l2cap; - u16 len; + u16 len, min_len; u8 event; + min_len = test_bit(FLAG_EXT_CTRL, &chan->flags) ? + L2CAP_EXT_CTRL_SIZE : L2CAP_ENH_CTRL_SIZE; + if (chan->fcs == L2CAP_FCS_CRC16) + min_len += L2CAP_FCS_SIZE; + + if (skb->len < min_len) + goto drop; + __unpack_control(chan, skb); len = skb->len; From b5dbb41b212c50c095a4dbee3017a84fe94f033b Mon Sep 17 00:00:00 2001 From: Aldo Ariel Panzardo Date: Tue, 15 Sep 2026 12:59:52 -0300 Subject: [PATCH 272/592] Bluetooth: mgmt: fix race in read_unconf_index_list() read_unconf_index_list() counts unconfigured controllers before allocating its response, then checks the device flags again while filling it. hci_dev_list_lock stabilizes list membership, but it does not serialize the per-device flags. During asynchronous controller setup, the worker can set HCI_UNCONFIGURED and clear HCI_SETUP between the two passes. A controller omitted from the allocation count can then become eligible for the fill pass, causing an out-of-bounds write to rp->index[]. Allocate space for every device on hci_dev_list. Since list membership cannot change while hci_dev_list_lock is held, the response remains large enough regardless of flag transitions. The reported count and response length still include only eligible unconfigured controllers. Fixes: 73d1df2a7a10 ("Bluetooth: Add support for Read Unconfigured Index List command") Cc: stable@vger.kernel.org Signed-off-by: Aldo Ariel Panzardo Signed-off-by: Luiz Augusto von Dentz --- net/bluetooth/mgmt.c | 8 ++------ 1 file changed, 2 insertions(+), 6 deletions(-) diff --git a/net/bluetooth/mgmt.c b/net/bluetooth/mgmt.c index f740e745ae78..41956cdde982 100644 --- a/net/bluetooth/mgmt.c +++ b/net/bluetooth/mgmt.c @@ -496,13 +496,9 @@ static int read_unconf_index_list(struct sock *sk, struct hci_dev *hdev, read_lock(&hci_dev_list_lock); - count = 0; - list_for_each_entry(d, &hci_dev_list, list) { - if (hci_dev_test_flag(d, HCI_UNCONFIGURED)) - count++; - } + count = list_count_nodes(&hci_dev_list); - rp_len = sizeof(*rp) + (2 * count); + rp_len = sizeof(*rp) + (sizeof(__le16) * count); rp = kmalloc(rp_len, GFP_ATOMIC); if (!rp) { read_unlock(&hci_dev_list_lock); From d034e836eefd7ce75e588f7031cffbeec594f5ac Mon Sep 17 00:00:00 2001 From: Joseph Qi Date: Tue, 1 Sep 2026 09:04:13 +0800 Subject: [PATCH 273/592] smb: client: fix use-after-free of iface in cifs_try_adding_channels() cifs_try_adding_channels() iterates ses->iface_list with list_for_each_entry_safe_from(), which captures the next entry (niface) under iface_lock. The loop body then drops iface_lock for the whole duration of cifs_ses_add_channel(). A concurrent interface refresh (SMB3_request_interfaces() -> parse_server_interfaces()) marks all ifaces inactive and removes and frees any that are not re-advertised via list_del() + kref_put(), where release_iface() is a bare kfree(). Since niface typically has no channel holding a reference, the list reference is its last and it can be freed inside the unlocked window. On continue, the iterator advance step then dereferences niface->iface_head.next, and the loop body reads iface->rdma_capable/is_active, both on freed memory. Fix this by never keeping an unreferenced list pointer across the unlocked window. Each channel attempt now re-scans the list from the head under iface_lock, takes a kref on the selected candidate, and passes only that referenced candidate to cifs_ses_add_channel(). weight_fulfilled still tracks selection progress, so restarting the scan preserves the original weighted distribution and the weight_fulfilled-before-kref_put ordering on the failure path. Add a per-pass attempts cap so a flapping interface refresh cannot keep the inner loop spinning within a single tries increment. Fixes: aa45dadd34e4 ("cifs: change iface_list from array to sorted linked list") Cc: stable@vger.kernel.org Assisted-by: Qoder:Qwen3.8-Max Signed-off-by: Joseph Qi Acked-by: Shyam Prasad N Signed-off-by: Paulo Alcantara --- fs/smb/client/sess.c | 104 ++++++++++++++++++++++++++----------------- 1 file changed, 63 insertions(+), 41 deletions(-) diff --git a/fs/smb/client/sess.c b/fs/smb/client/sess.c index 7cf7dd104f7c..e095f41b5882 100644 --- a/fs/smb/client/sess.c +++ b/fs/smb/client/sess.c @@ -149,9 +149,9 @@ int cifs_try_adding_channels(struct cifs_ses *ses) int old_chan_count, new_chan_count; int left; int rc = 0; - int tries = 0; + int tries = 0, attempts; size_t iface_weight = 0, iface_min_speed = 0; - struct cifs_server_iface *iface = NULL, *niface = NULL; + struct cifs_server_iface *iface = NULL, *candidate = NULL; struct cifs_server_iface *last_iface = NULL; spin_lock(&ses->chan_lock); @@ -197,67 +197,89 @@ int cifs_try_adding_channels(struct cifs_ses *ses) break; } - if (!iface) - iface = list_first_entry(&ses->iface_list, struct cifs_server_iface, - iface_head); last_iface = list_last_entry(&ses->iface_list, struct cifs_server_iface, iface_head); iface_min_speed = last_iface->speed; + spin_unlock(&ses->iface_lock); - list_for_each_entry_safe_from(iface, niface, &ses->iface_list, - iface_head) { - /* do not mix rdma and non-rdma interfaces */ - if (iface->rdma_capable != ses->server->rdma) - continue; + attempts = 0; + while (left > 0) { + spin_lock(&ses->iface_lock); - /* skip ifaces that are unusable */ - if (!iface->is_active || - (is_ses_using_iface(ses, iface) && - !iface->rss_capable)) - continue; + /* + * iface_lock must be dropped while opening a channel, + * and a concurrent interface refresh may remove and + * free entries during that window, so no list entry + * may be kept across it without a reference. Scan + * the list from the beginning each time and only pass + * a referenced candidate to cifs_ses_add_channel(); + * weight_fulfilled tracks the progress so that no + * iface is selected beyond its weight. + */ + candidate = NULL; + list_for_each_entry(iface, &ses->iface_list, iface_head) { + /* do not mix rdma and non-rdma interfaces */ + if (iface->rdma_capable != ses->server->rdma) + continue; - /* check if we already allocated enough channels */ - iface_weight = iface->speed / iface_min_speed; + /* skip ifaces that are unusable */ + if (!iface->is_active || + (is_ses_using_iface(ses, iface) && + !iface->rss_capable)) + continue; - if (iface->weight_fulfilled >= iface_weight) - continue; + /* check if we already allocated enough channels */ + iface_weight = iface->speed / iface_min_speed; - /* take ref before unlock */ - kref_get(&iface->refcount); + if (iface->weight_fulfilled >= iface_weight) + continue; + + /* take ref before unlock */ + kref_get(&iface->refcount); + candidate = iface; + break; + } + + if (!candidate) { + /* no usable iface. reset weight_fulfilled and start over */ + list_for_each_entry(iface, &ses->iface_list, iface_head) + iface->weight_fulfilled = 0; + spin_unlock(&ses->iface_lock); + break; + } + + attempts++; + if (attempts > 3 * ses->chan_max) { + kref_put(&candidate->refcount, release_iface); + spin_unlock(&ses->iface_lock); + break; + } spin_unlock(&ses->iface_lock); - rc = cifs_ses_add_channel(ses, iface); + rc = cifs_ses_add_channel(ses, candidate); spin_lock(&ses->iface_lock); if (rc) { cifs_dbg(VFS, "failed to open extra channel on iface:%pIS rc=%d\n", - &iface->sockaddr, + &candidate->sockaddr, rc); /* failure to add chan should increase weight */ - iface->weight_fulfilled++; - kref_put(&iface->refcount, release_iface); + candidate->weight_fulfilled++; + kref_put(&candidate->refcount, release_iface); + spin_unlock(&ses->iface_lock); continue; } - iface->num_channels++; - iface->weight_fulfilled++; + candidate->num_channels++; + candidate->weight_fulfilled++; cifs_info("successfully opened new channel on iface:%pIS\n", - &iface->sockaddr); + &candidate->sockaddr); + spin_unlock(&ses->iface_lock); + + left--; + new_chan_count++; break; } - - /* reached end of list. reset weight_fulfilled and start over */ - if (list_entry_is_head(iface, &ses->iface_list, iface_head)) { - list_for_each_entry(iface, &ses->iface_list, iface_head) - iface->weight_fulfilled = 0; - spin_unlock(&ses->iface_lock); - iface = NULL; - continue; - } - spin_unlock(&ses->iface_lock); - - left--; - new_chan_count++; } return new_chan_count - old_chan_count; From 089070b51ccbac411462a30a454690274c6e4270 Mon Sep 17 00:00:00 2001 From: Sanman Pradhan Date: Wed, 16 Sep 2026 23:54:17 +0000 Subject: [PATCH 274/592] hwmon: (pmbus/tps53679) Select page 0 for single-page TPS53676 tps53676_identify() derives the number of PMBus pages but does not ensure that page 0 is selected for single-page configurations. pmbus_set_page() does not update the PAGE register when info->pages is 1, so if boot firmware leaves PAGE set to another value subsequent register accesses may target the wrong page. For single-page devices, select page 0 explicitly. Fixes: cb3d37b59012 ("hwmon: (pmbus/tps53679) Add support for TI TPS53676") Cc: stable@vger.kernel.org Signed-off-by: Sanman Pradhan Link: https://patch.msgid.link/20260916235406.681131-2-sanman.pradhan@hpe.com Signed-off-by: Guenter Roeck --- drivers/hwmon/pmbus/tps53679.c | 9 +++++++++ 1 file changed, 9 insertions(+) diff --git a/drivers/hwmon/pmbus/tps53679.c b/drivers/hwmon/pmbus/tps53679.c index 9f27832703ff..6c25701b36fc 100644 --- a/drivers/hwmon/pmbus/tps53679.c +++ b/drivers/hwmon/pmbus/tps53679.c @@ -200,6 +200,15 @@ static int tps53676_identify(struct i2c_client *client, if (phases_b > 0) { info->pages = 2; info->phases[1] = phases_b; + } else { + /* + * pmbus_set_page() does not update the PAGE register on + * single-page devices, so select page 0 explicitly in case + * the boot firmware left the device on another page. + */ + ret = i2c_smbus_write_byte_data(client, PMBUS_PAGE, 0); + if (ret < 0) + return ret; } return 0; } From 92b68492eae701e5b0e9d142ffe229921af7b1fa Mon Sep 17 00:00:00 2001 From: James Seo Date: Wed, 16 Sep 2026 15:19:15 -0700 Subject: [PATCH 275/592] hwmon: (hp-wmi-sensors) Improve raw WMI string handling Commit c9ba59258094 ("hwmon: (hp-wmi-sensors) Fix failure to load on EliteDesk 800 G6") left out some logic for recognizing raw WMI strings in check_numeric_sensor_wobj(). This issue was reported by a user along with an incomplete and unsuitable proposed solution [1]. Add the missing logic and properly remedy the issue. Also slightly refactor how raw WMI strings are recognized elsewhere to make the intent that they should be treated as regular ACPI strings clearer. Reported-by: Muhammad Bilal Link: https://lore.kernel.org/linux-hwmon/20260916002907.161210-1-meatuni001@gmail.com/ [1] Fixes: c9ba59258094 ("hwmon: (hp-wmi-sensors) Fix failure to load on EliteDesk 800 G6") Signed-off-by: James Seo Link: https://patch.msgid.link/20260916221912.434119-5-james@equiv.tech Signed-off-by: Guenter Roeck --- drivers/hwmon/hp-wmi-sensors.c | 30 ++++++++++++++++++++++-------- 1 file changed, 22 insertions(+), 8 deletions(-) diff --git a/drivers/hwmon/hp-wmi-sensors.c b/drivers/hwmon/hp-wmi-sensors.c index 55aee16df57d..cfa0d6ad0fb3 100644 --- a/drivers/hwmon/hp-wmi-sensors.c +++ b/drivers/hwmon/hp-wmi-sensors.c @@ -526,14 +526,12 @@ static int check_wobj(const union acpi_object *wobj, for (prop = 0; prop <= last_prop; prop++) { type = elements[prop].type; valid_type = property_map[prop]; - if (type != valid_type) { - if (type == ACPI_TYPE_BUFFER && - valid_type == ACPI_TYPE_STRING && - is_raw_wmi_string(elements[prop].buffer.pointer, - elements[prop].buffer.length)) - continue; + if (type == ACPI_TYPE_BUFFER && + is_raw_wmi_string(elements[prop].buffer.pointer, + elements[prop].buffer.length)) + type = ACPI_TYPE_STRING; + if (type != valid_type) return -EINVAL; - } } return 0; @@ -579,6 +577,7 @@ static int check_numeric_sensor_wobj(const union acpi_object *wobj, int prop = HP_WMI_PROPERTY_NAME; acpi_object_type valid_type; union acpi_object *elements; + union acpi_object *element; u32 elem_count; int last_prop; bool is_new; @@ -602,13 +601,20 @@ static int check_numeric_sensor_wobj(const union acpi_object *wobj, elem_count > HP_WMI_MAX_PROPERTIES) return -EINVAL; - type = elements[HP_WMI_PROPERTY_SIZE].type; + element = &elements[HP_WMI_PROPERTY_SIZE]; + type = element->type; switch (type) { case ACPI_TYPE_INTEGER: is_new = true; last_prop = HP_WMI_PROPERTY_RATE_UNITS; break; + case ACPI_TYPE_BUFFER: + if (!is_raw_wmi_string(element->buffer.pointer, + element->buffer.length)) + return -EINVAL; + fallthrough; + case ACPI_TYPE_STRING: is_new = false; last_prop = HP_WMI_PROPERTY_CURRENT_READING; @@ -631,6 +637,10 @@ static int check_numeric_sensor_wobj(const union acpi_object *wobj, for (i = 0; i < elem_count && prop <= last_prop; i++, prop++) { type = elements[i].type; valid_type = hp_wmi_property_map[prop]; + if (type == ACPI_TYPE_BUFFER && + is_raw_wmi_string(elements[i].buffer.pointer, + elements[i].buffer.length)) + type = ACPI_TYPE_STRING; if (type != valid_type) return -EINVAL; @@ -651,6 +661,10 @@ static int check_numeric_sensor_wobj(const union acpi_object *wobj, /* PossibleStates[0] has already been type-checked. */ for (j = 0; i + 1 < elem_count && j + 1 < count; j++) { type = elements[++i].type; + if (type == ACPI_TYPE_BUFFER && + is_raw_wmi_string(elements[i].buffer.pointer, + elements[i].buffer.length)) + type = ACPI_TYPE_STRING; if (type != valid_type) return -EINVAL; } From 88f113634028ca90a857031837d8061d1a9e1a7b Mon Sep 17 00:00:00 2001 From: Wentao Liang Date: Wed, 16 Sep 2026 17:05:11 +0000 Subject: [PATCH 276/592] watchdog: sp5100_tco: Fix pci_dev reference leak in sp5100_tco_init() sp5100_tco_init() stores the PCI device matched by for_each_pci_dev() in the global sp5100_tco_pci and keeps its reference for the lifetime of the driver, but neither sp5100_tco_exit() nor the error paths of sp5100_tco_init() call pci_dev_put(), leaking the reference on driver registration failure and on every module load/unload cycle. Drop the reference when the platform driver or device registration fails and when the module is unloaded. Fixes: 15e28bf13008 ("watchdog: Add support for sp5100 chipset TCO") Cc: stable@vger.kernel.org Signed-off-by: Wentao Liang Link: https://patch.msgid.link/20260916170511.2086199-1-vulab@iscas.ac.cn Signed-off-by: Guenter Roeck --- drivers/watchdog/sp5100_tco.c | 6 +++++- 1 file changed, 5 insertions(+), 1 deletion(-) diff --git a/drivers/watchdog/sp5100_tco.c b/drivers/watchdog/sp5100_tco.c index 7e99c3b1f367..72ad59649ccc 100644 --- a/drivers/watchdog/sp5100_tco.c +++ b/drivers/watchdog/sp5100_tco.c @@ -605,8 +605,10 @@ static int __init sp5100_tco_init(void) pr_info("SP5100/SB800 TCO WatchDog Timer Driver\n"); err = platform_driver_register(&sp5100_tco_driver); - if (err) + if (err) { + pci_dev_put(sp5100_tco_pci); return err; + } sp5100_tco_platform_device = platform_device_register_simple(TCO_DRIVER_NAME, -1, NULL, 0); @@ -619,6 +621,7 @@ static int __init sp5100_tco_init(void) unreg_platform_driver: platform_driver_unregister(&sp5100_tco_driver); + pci_dev_put(sp5100_tco_pci); return err; } @@ -626,6 +629,7 @@ static void __exit sp5100_tco_exit(void) { platform_device_unregister(sp5100_tco_platform_device); platform_driver_unregister(&sp5100_tco_driver); + pci_dev_put(sp5100_tco_pci); } module_init(sp5100_tco_init); From 400cb663ca019bae6eb878f06f1094ddf7c0b0df Mon Sep 17 00:00:00 2001 From: Tzung-Bi Shih Date: Sun, 13 Sep 2026 14:48:49 +0800 Subject: [PATCH 277/592] watchdog: digicolor: Avoid division by zero clk_get_rate() could return 0. Avoid a division by zero panic. Since get_timeleft() cannot propagate errors, check the clock rate early in probe() and cache the rate in the driver data as it is unlikely to change at runtime. Fixes: 336694a01dae ("watchdog: digicolor: driver for Conexant Digicolor CX92755 SoC") Cc: stable@vger.kernel.org Signed-off-by: Tzung-Bi Shih Acked-by: Baruch Siach Link: https://patch.msgid.link/20260913064851.8239-2-tzungbi@kernel.org Signed-off-by: Guenter Roeck --- drivers/watchdog/digicolor_wdt.c | 13 +++++++++---- 1 file changed, 9 insertions(+), 4 deletions(-) diff --git a/drivers/watchdog/digicolor_wdt.c b/drivers/watchdog/digicolor_wdt.c index 073d37867f47..de1a3267a972 100644 --- a/drivers/watchdog/digicolor_wdt.c +++ b/drivers/watchdog/digicolor_wdt.c @@ -25,6 +25,7 @@ struct dc_wdt { void __iomem *base; struct clk *clk; spinlock_t lock; + unsigned long rate; }; static unsigned timeout; @@ -61,7 +62,7 @@ static int dc_wdt_start(struct watchdog_device *wdog) { struct dc_wdt *wdt = watchdog_get_drvdata(wdog); - dc_wdt_set(wdt, wdog->timeout * clk_get_rate(wdt->clk)); + dc_wdt_set(wdt, wdog->timeout * wdt->rate); return 0; } @@ -79,7 +80,7 @@ static int dc_wdt_set_timeout(struct watchdog_device *wdog, unsigned int t) { struct dc_wdt *wdt = watchdog_get_drvdata(wdog); - dc_wdt_set(wdt, t * clk_get_rate(wdt->clk)); + dc_wdt_set(wdt, t * wdt->rate); wdog->timeout = t; return 0; @@ -90,7 +91,7 @@ static unsigned int dc_wdt_get_timeleft(struct watchdog_device *wdog) struct dc_wdt *wdt = watchdog_get_drvdata(wdog); uint32_t count = readl_relaxed(wdt->base + TIMER_A_COUNT); - return count / clk_get_rate(wdt->clk); + return count / wdt->rate; } static const struct watchdog_ops dc_wdt_ops = { @@ -130,7 +131,11 @@ static int dc_wdt_probe(struct platform_device *pdev) wdt->clk = devm_clk_get(dev, NULL); if (IS_ERR(wdt->clk)) return PTR_ERR(wdt->clk); - dc_wdt_wdd.max_timeout = U32_MAX / clk_get_rate(wdt->clk); + + wdt->rate = clk_get_rate(wdt->clk); + if (!wdt->rate) + return -EINVAL; + dc_wdt_wdd.max_timeout = U32_MAX / wdt->rate; dc_wdt_wdd.timeout = dc_wdt_wdd.max_timeout; dc_wdt_wdd.parent = dev; From 5af7d2cbd20f893def03c8310a460ade66a5d822 Mon Sep 17 00:00:00 2001 From: Tzung-Bi Shih Date: Sun, 13 Sep 2026 14:48:50 +0800 Subject: [PATCH 278/592] watchdog: rtd119x: Avoid division by zero clk_get_rate() could return 0. Avoid a division by zero panic. Fixes: 2bdf6acbfead ("watchdog: Add Realtek RTD1295") Cc: stable@vger.kernel.org Signed-off-by: Tzung-Bi Shih Link: https://patch.msgid.link/20260913064851.8239-3-tzungbi@kernel.org Signed-off-by: Guenter Roeck --- drivers/watchdog/rtd119x_wdt.c | 7 ++++++- 1 file changed, 6 insertions(+), 1 deletion(-) diff --git a/drivers/watchdog/rtd119x_wdt.c b/drivers/watchdog/rtd119x_wdt.c index 984905695dde..0bfadb58917b 100644 --- a/drivers/watchdog/rtd119x_wdt.c +++ b/drivers/watchdog/rtd119x_wdt.c @@ -98,6 +98,7 @@ static int rtd119x_wdt_probe(struct platform_device *pdev) { struct device *dev = &pdev->dev; struct rtd119x_watchdog_device *data; + unsigned long rate; data = devm_kzalloc(dev, sizeof(*data), GFP_KERNEL); if (!data) @@ -111,10 +112,14 @@ static int rtd119x_wdt_probe(struct platform_device *pdev) if (IS_ERR(data->clk)) return PTR_ERR(data->clk); + rate = clk_get_rate(data->clk); + if (!rate) + return -EINVAL; + data->wdt_dev.info = &rtd119x_wdt_info; data->wdt_dev.ops = &rtd119x_wdt_ops; data->wdt_dev.timeout = 120; - data->wdt_dev.max_timeout = 0xffffffff / clk_get_rate(data->clk); + data->wdt_dev.max_timeout = 0xffffffff / rate; data->wdt_dev.min_timeout = 1; data->wdt_dev.parent = dev; From 6274281c41efa8dd1aa5234c59ad904ff89d7af4 Mon Sep 17 00:00:00 2001 From: Tzung-Bi Shih Date: Sun, 13 Sep 2026 14:48:51 +0800 Subject: [PATCH 279/592] watchdog: rzv2h: Avoid division by zero clk_get_rate() could return 0. Avoid a division by zero panic. Fixes: f6febd0a30b6 ("watchdog: Add Watchdog Timer driver for RZ/V2H(P)") Cc: stable@vger.kernel.org Signed-off-by: Tzung-Bi Shih Link: https://patch.msgid.link/20260913064851.8239-4-tzungbi@kernel.org Signed-off-by: Guenter Roeck --- drivers/watchdog/rzv2h_wdt.c | 7 ++++++- 1 file changed, 6 insertions(+), 1 deletion(-) diff --git a/drivers/watchdog/rzv2h_wdt.c b/drivers/watchdog/rzv2h_wdt.c index 3b6abb66a1da..83540dd9a37b 100644 --- a/drivers/watchdog/rzv2h_wdt.c +++ b/drivers/watchdog/rzv2h_wdt.c @@ -278,6 +278,7 @@ static int rzv2h_wdt_probe(struct platform_device *pdev) struct device *dev = &pdev->dev; struct rzv2h_wdt_priv *priv; struct clk *count_clk; + unsigned long rate; int ret; priv = devm_kzalloc(dev, sizeof(*priv), GFP_KERNEL); @@ -314,8 +315,12 @@ static int rzv2h_wdt_probe(struct platform_device *pdev) return dev_err_probe(dev, -EINVAL, "Invalid count source\n"); } + rate = clk_get_rate(count_clk); + if (!rate) + return dev_err_probe(dev, -EINVAL, "Invalid clock rate\n"); + priv->wdev.max_hw_heartbeat_ms = (MILLI * priv->of_data->timeout_cycles * - priv->of_data->cks_div) / clk_get_rate(count_clk); + priv->of_data->cks_div) / rate; dev_dbg(dev, "max hw timeout of %dms\n", priv->wdev.max_hw_heartbeat_ms); ret = devm_pm_runtime_enable(dev); From 1d9763f34a85680db1e8233d654fdb85e5f897cc Mon Sep 17 00:00:00 2001 From: Tzung-Bi Shih Date: Sun, 13 Sep 2026 00:33:34 +0800 Subject: [PATCH 280/592] watchdog: msc313e: Propagate error code in resume() If msc313e_wdt_start() fails during system resume, the error is currently ignored. Consequently, the watchdog isn't running without the user's knowledge. Propagate the error code and print a message if msc313e_wdt_start() fails. Signed-off-by: Tzung-Bi Shih Fixes: e9800b7994642 ("watchdog: Add Mstar MSC313e WDT driver") Cc: stable@vger.kernel.org Link: https://patch.msgid.link/20260912163334.28636-1-tzungbi@kernel.org Signed-off-by: Guenter Roeck --- drivers/watchdog/msc313e_wdt.c | 10 +++++++--- 1 file changed, 7 insertions(+), 3 deletions(-) diff --git a/drivers/watchdog/msc313e_wdt.c b/drivers/watchdog/msc313e_wdt.c index 4a5cce2a16b1..9a400562d1ae 100644 --- a/drivers/watchdog/msc313e_wdt.c +++ b/drivers/watchdog/msc313e_wdt.c @@ -191,11 +191,15 @@ static int __maybe_unused msc313e_wdt_suspend(struct device *dev) static int __maybe_unused msc313e_wdt_resume(struct device *dev) { struct msc313e_wdt_priv *priv = dev_get_drvdata(dev); + int ret = 0; - if (watchdog_active(&priv->wdev) || watchdog_hw_running(&priv->wdev)) - msc313e_wdt_start(&priv->wdev); + if (watchdog_active(&priv->wdev) || watchdog_hw_running(&priv->wdev)) { + ret = msc313e_wdt_start(&priv->wdev); + if (ret) + dev_err(dev, "Failed to restart watchdog (err=%d)\n", ret); + } - return 0; + return ret; } static SIMPLE_DEV_PM_OPS(msc313e_wdt_pm_ops, msc313e_wdt_suspend, msc313e_wdt_resume); From 22737cfced627ffcb4b5c36d63bb3d4476f63213 Mon Sep 17 00:00:00 2001 From: Tzung-Bi Shih Date: Sun, 13 Sep 2026 14:51:26 +0800 Subject: [PATCH 281/592] watchdog: msc313e: Fix premature reset during timeout update Updating the 32-bit hardware timeout requires writing to two 16-bit registers sequentially. If the watchdog is actively running, this non-atomic update might trigger a premature system reset. Clear the watchdog counter before updating the registers to prevent the timer from timing out prematurely against an intermediate threshold. Fixes: e9800b799464 ("watchdog: Add Mstar MSC313e WDT driver") Cc: stable@vger.kernel.org Signed-off-by: Tzung-Bi Shih Link: https://patch.msgid.link/20260913065126.8350-1-tzungbi@kernel.org Signed-off-by: Guenter Roeck --- drivers/watchdog/msc313e_wdt.c | 6 ++++++ 1 file changed, 6 insertions(+) diff --git a/drivers/watchdog/msc313e_wdt.c b/drivers/watchdog/msc313e_wdt.c index 9a400562d1ae..d62586f6e09d 100644 --- a/drivers/watchdog/msc313e_wdt.c +++ b/drivers/watchdog/msc313e_wdt.c @@ -46,6 +46,9 @@ static void msc313e_wdt_set_hw_timeout(struct msc313e_wdt_priv *priv, { u32 t = timeout * clk_get_rate(priv->clk); + /* Clear before to prevent premature reset during non-atomic updates. */ + writew(1, priv->base + REG_WDT_CLR); + writew(t & 0xffff, priv->base + REG_WDT_MAX_PRD_L); writew((t >> 16) & 0xffff, priv->base + REG_WDT_MAX_PRD_H); writew(1, priv->base + REG_WDT_CLR); @@ -76,6 +79,9 @@ static int msc313e_wdt_stop(struct watchdog_device *wdev) { struct msc313e_wdt_priv *priv = watchdog_get_drvdata(wdev); + /* Clear before to prevent premature reset during non-atomic updates. */ + writew(1, priv->base + REG_WDT_CLR); + writew(0, priv->base + REG_WDT_MAX_PRD_L); writew(0, priv->base + REG_WDT_MAX_PRD_H); writew(0, priv->base + REG_WDT_CLR); From 5071122bf5a628494db16d98d253f622a5aab074 Mon Sep 17 00:00:00 2001 From: Li Jun Date: Mon, 14 Sep 2026 14:23:53 +0800 Subject: [PATCH 282/592] watchdog: da9062: fix suspend/resume handling of HW_RUNNING watchdog MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit da9062_wdt_suspend() and da9062_wdt_resume() only check watchdog_active(), when the watchdog is left running by the driver sets WDOG_HW_RUNNING in da9062_wdt_probe() but userspace never opens the device, so WDOG_ACTIVE remains cleared, the wdt_disable() will not be executed in da9062_wdt_suspend. In this case, the suspend callback is a no-op and the watchdog keeps counting during system suspend, leading to an unexpected system reset. Check WDOG_HW_RUNNING and wdt->wdd，can fix this issue. Fixes: f6c98b08381c7 ("watchdog: da9062: add power management ops") Cs: stable@vger.kernel.org Signed-off-by: Li Jun Link: https://patch.msgid.link/20260914062353.582205-1-lijun01@kylinos.cn Signed-off-by: Guenter Roeck --- drivers/watchdog/da9062_wdt.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/drivers/watchdog/da9062_wdt.c b/drivers/watchdog/da9062_wdt.c index 426962547df1..4d558652e9e7 100644 --- a/drivers/watchdog/da9062_wdt.c +++ b/drivers/watchdog/da9062_wdt.c @@ -256,7 +256,7 @@ static int __maybe_unused da9062_wdt_suspend(struct device *dev) if (!wdt->use_sw_pm) return 0; - if (watchdog_active(wdd)) + if (watchdog_active(wdd) || watchdog_hw_running(wdd)) return da9062_wdt_stop(wdd); return 0; @@ -270,7 +270,7 @@ static int __maybe_unused da9062_wdt_resume(struct device *dev) if (!wdt->use_sw_pm) return 0; - if (watchdog_active(wdd)) + if (watchdog_active(wdd) || watchdog_hw_running(wdd)) return da9062_wdt_start(wdd); return 0; From 8f0ca55016a7647109ae2bc91bcb346fc8b13785 Mon Sep 17 00:00:00 2001 From: Wentao Liang Date: Wed, 16 Sep 2026 17:07:04 +0000 Subject: [PATCH 283/592] watchdog: starfive-wdt: Fix runtime PM leak in starfive_wdt_pm_start() starfive_wdt_pm_start() takes a runtime PM reference with pm_runtime_get_sync(), which increments the usage counter even when it fails, and returns the error without dropping it again. The watchdog core does not invoke the stop callback when start fails, so the reference taken on the error path is leaked. Use pm_runtime_resume_and_get() instead, which keeps the usage counter balanced when the resume fails. Fixes: db728ea9c7be ("drivers: watchdog: Add StarFive Watchdog driver") Cc: stable@vger.kernel.org Signed-off-by: Wentao Liang Link: https://patch.msgid.link/20260916170704.2086331-1-vulab@iscas.ac.cn Signed-off-by: Guenter Roeck --- drivers/watchdog/starfive-wdt.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/drivers/watchdog/starfive-wdt.c b/drivers/watchdog/starfive-wdt.c index af55adc4a3c6..5a3254c83d3b 100644 --- a/drivers/watchdog/starfive-wdt.c +++ b/drivers/watchdog/starfive-wdt.c @@ -371,7 +371,7 @@ static void starfive_wdt_stop(struct starfive_wdt *wdt) static int starfive_wdt_pm_start(struct watchdog_device *wdd) { struct starfive_wdt *wdt = watchdog_get_drvdata(wdd); - int ret = pm_runtime_get_sync(wdd->parent); + int ret = pm_runtime_resume_and_get(wdd->parent); if (ret < 0) return ret; From 5535d5e61a77ad79118ea7665ce1c14f857f0f12 Mon Sep 17 00:00:00 2001 From: Huacai Chen Date: Thu, 3 Sep 2026 16:42:12 +0800 Subject: [PATCH 284/592] drm/loongson: Create blend mode property for cursor plane After commit 860e748bddcc929 ("drm: ensure blend mode supported if pixel format with alpha exposed") we get warnings at boot: loongson 0000:00:06.1: [drm] [PLANE:41:ls-cursor-plane-0] pixel format with alpha exposed but blend mode not setup. Please fix. loongson 0000:00:06.1: [drm] [PLANE:46:ls-cursor-plane-1] pixel format with alpha exposed but blend mode not setup. Please fix. The reason is the cursor plane supports color formats with alpha but the driver doesn't create blend mode property, which triggers the warning in validate_blend_mode_for_alpha_formats(). The loongson DC HW doesn't support DRM_MODE_BLEND_PREMULTI, so create blend mode property with DRM_MODE_BLEND_COVERAGE for cursor planes since it is the only one implemented in the driver. Signed-off-by: Huacai Chen Reviewed-by: Jianmin Lv Signed-off-by: Icenowy Zheng Link: https://patch.msgid.link/20260903084212.3621540-1-chenhuacai@loongson.cn --- drivers/gpu/drm/loongson/lsdc_plane.c | 5 +++-- 1 file changed, 3 insertions(+), 2 deletions(-) diff --git a/drivers/gpu/drm/loongson/lsdc_plane.c b/drivers/gpu/drm/loongson/lsdc_plane.c index bea42215796d..bcc0ffa17bdf 100644 --- a/drivers/gpu/drm/loongson/lsdc_plane.c +++ b/drivers/gpu/drm/loongson/lsdc_plane.c @@ -7,6 +7,7 @@ #include #include +#include #include #include #include @@ -765,7 +766,7 @@ int ls7a1000_cursor_plane_init(struct drm_device *ddev, drm_plane_helper_add(plane, &ls7a1000_cursor_plane_helper_funcs); - return 0; + return drm_plane_create_blend_mode_property(plane, BIT(DRM_MODE_BLEND_COVERAGE)); } int ls7a2000_cursor_plane_init(struct drm_device *ddev, @@ -790,5 +791,5 @@ int ls7a2000_cursor_plane_init(struct drm_device *ddev, drm_plane_helper_add(plane, &ls7a2000_cursor_plane_helper_funcs); - return 0; + return drm_plane_create_blend_mode_property(plane, BIT(DRM_MODE_BLEND_COVERAGE)); } From a5f7a5bb3b7f28ba7e4fa246775b29a0e5537255 Mon Sep 17 00:00:00 2001 From: Jinke Han Date: Tue, 8 Sep 2026 15:37:42 +0800 Subject: [PATCH 285/592] x86/kprobes: Fix crash when probing CS CALL instructions When using eBPF to probe CS CALL instructions within a function, a crash can be triggered. The eBPF tool probes offset 257 of the __hrtimer_run_queues() function: <__hrtimer_run_queues+249>: nopl 0x0(%rax,%rax,1) <__hrtimer_run_queues+254>: mov %r14,%rdi <__hrtimer_run_queues+257>: cs call <__x86_indirect_thunk_r12> <__hrtimer_run_queues+263>: mov %eax,%r12d <__hrtimer_run_queues+266>: xchg %ax,%ax <__hrtimer_run_queues+268>: mov %r13,%rdi Which triggers this crash: BUG: unable to handle page fault for address: 00000000000f41c9 #PF: supervisor write access in kernel mode #PF: error_code(0x0002) - not-present page PGD 0 P4D 0 Oops: 0002 [#1] SMP NOPTI CPU: 1 PID: 0 Comm: swapper/1 Kdump: loaded Tainted: P RIP: 0010:__hrtimer_run_queues+0x106/0x230 Note that __hrtimer_run_queues+0x106 is __hrtimer_run_queues+262, which is at the 6th byte of the above CS CALL instruction. Since the CS CALL instruction occupies 6 bytes, the exception occurred in the middle of that call instruction. The root cause is that when using eBPF tools to probe in the middle of a function, a kprobe with INT3 is used as the underlying implementation. During single-step emulation of the original CALL instruction, int3_emulate_call() assumes that the probed CALL instruction is 5 bytes long. However, the actual CS-prefixed CALL instruction occupies 6 bytes, so it constructs an incorrect exception return address. When the CPU returns from the kprobe handler, the next instruction to be executed is at the address of the last byte of that CS CALL instruction. Coincidentally, starting from that address, the CPU fetches and decodes a completely different instruction, which ultimately triggers a kernel crash. Fix the issue by using the actual instruction length obtained from the instruction decoder when constructing the exception return address, rather than relying on the hardcoded CALL_INSN_SIZE macro. [ mingo: Refined the changelog ] Fixes: 6256e668b7af ("x86/kprobes: Use int3 instead of debug trap for single-step") Suggested-by: Masami Hiramatsu (Google) Signed-off-by: Jinke Han Signed-off-by: Ingo Molnar Reviewed-by: Masami Hiramatsu (Google) Acked-by: Yafang Shao Acked-by: Borislav Petkov Cc: Peter Zijlstra Link: https://patch.msgid.link/20260908073742.GA10517@didi-ThinkCentre-M920t-N000 --- arch/x86/include/asm/text-patching.h | 4 ++-- arch/x86/kernel/alternative.c | 6 ++++-- arch/x86/kernel/kprobes/core.c | 5 ++--- 3 files changed, 8 insertions(+), 7 deletions(-) diff --git a/arch/x86/include/asm/text-patching.h b/arch/x86/include/asm/text-patching.h index f2d142a0a862..ea09381070e8 100644 --- a/arch/x86/include/asm/text-patching.h +++ b/arch/x86/include/asm/text-patching.h @@ -164,9 +164,9 @@ unsigned long int3_emulate_pop(struct pt_regs *regs) } static __always_inline -void int3_emulate_call(struct pt_regs *regs, unsigned long func) +void int3_emulate_call(struct pt_regs *regs, unsigned long ip, unsigned long func) { - int3_emulate_push(regs, regs->ip - INT3_INSN_SIZE + CALL_INSN_SIZE); + int3_emulate_push(regs, ip); int3_emulate_jmp(regs, func); } diff --git a/arch/x86/kernel/alternative.c b/arch/x86/kernel/alternative.c index 741d8767ddf8..582c6d8307d1 100644 --- a/arch/x86/kernel/alternative.c +++ b/arch/x86/kernel/alternative.c @@ -2176,6 +2176,7 @@ int3_exception_notify(struct notifier_block *self, unsigned long val, void *data unsigned long selftest = (unsigned long)&int3_selftest_asm; struct die_args *args = data; struct pt_regs *regs = args->regs; + unsigned long ip; OPTIMIZER_HIDE_VAR(selftest); @@ -2188,7 +2189,8 @@ int3_exception_notify(struct notifier_block *self, unsigned long val, void *data if (regs->ip - INT3_INSN_SIZE != selftest) return NOTIFY_DONE; - int3_emulate_call(regs, (unsigned long)&int3_selftest_callee); + ip = regs->ip - INT3_INSN_SIZE + CALL_INSN_SIZE; + int3_emulate_call(regs, ip, (unsigned long)&int3_selftest_callee); return NOTIFY_STOP; } @@ -2758,7 +2760,7 @@ noinstr int smp_text_poke_int3_handler(struct pt_regs *regs) break; case CALL_INSN_OPCODE: - int3_emulate_call(regs, (long)ip + tpl->disp); + int3_emulate_call(regs, (long)ip, (long)ip + tpl->disp); break; case JMP32_INSN_OPCODE: diff --git a/arch/x86/kernel/kprobes/core.c b/arch/x86/kernel/kprobes/core.c index 4e5f8c1736ec..133ff20caccd 100644 --- a/arch/x86/kernel/kprobes/core.c +++ b/arch/x86/kernel/kprobes/core.c @@ -510,10 +510,9 @@ NOKPROBE_SYMBOL(kprobe_emulate_ret); static void kprobe_emulate_call(struct kprobe *p, struct pt_regs *regs) { - unsigned long func = regs->ip - INT3_INSN_SIZE + p->ainsn.size; + unsigned long ip = regs->ip - INT3_INSN_SIZE + p->ainsn.size; - func += p->ainsn.rel32; - int3_emulate_call(regs, func); + int3_emulate_call(regs, ip, ip + p->ainsn.rel32); } NOKPROBE_SYMBOL(kprobe_emulate_call); From 065f3ce5936e68da75f3dc18201d290073d78f3c Mon Sep 17 00:00:00 2001 From: "Darrick J. Wong" Date: Mon, 14 Sep 2026 22:37:05 -0700 Subject: [PATCH 286/592] xfs: call xfs_dquot_set_prealloc_limits if we installed default rtb limits Now that we have quotas for the realtime volume, we also have precomputed watermark limits for the realtime block counts. These precomputations should be done any time we change the rtb limits, which means that xfs_qm_adjust_dqlimits needs to ensure that if we installed a default rtb limit. Cc: stable@vger.kernel.org # v6.13 Fixes: 5dd70852b03901 ("xfs: create quota preallocation watermarks for realtime quota") Signed-off-by: Darrick J. Wong Reviewed-by: Christoph Hellwig Signed-off-by: Carlos Maiolino --- fs/xfs/xfs_dquot.c | 8 ++++++-- 1 file changed, 6 insertions(+), 2 deletions(-) diff --git a/fs/xfs/xfs_dquot.c b/fs/xfs/xfs_dquot.c index b4f6c594808c..e696ee36c2e8 100644 --- a/fs/xfs/xfs_dquot.c +++ b/fs/xfs/xfs_dquot.c @@ -139,10 +139,14 @@ xfs_qm_adjust_dqlimits( dq->q_ino.softlimit = defq->ino.soft; if (!dq->q_ino.hardlimit) dq->q_ino.hardlimit = defq->ino.hard; - if (!dq->q_rtb.softlimit) + if (!dq->q_rtb.softlimit) { dq->q_rtb.softlimit = defq->rtb.soft; - if (!dq->q_rtb.hardlimit) + prealloc = 1; + } + if (!dq->q_rtb.hardlimit) { dq->q_rtb.hardlimit = defq->rtb.hard; + prealloc = 1; + } if (prealloc) xfs_dquot_set_prealloc_limits(dq); From 41c4c41cf6c44f98db2916e1781f537d9ba6461a Mon Sep 17 00:00:00 2001 From: "Darrick J. Wong" Date: Mon, 14 Sep 2026 22:37:20 -0700 Subject: [PATCH 287/592] xfs: fix rtgroup repair estimations When I added online fsck for realtime reflink, I forgot to update xrep_calc_rtgroup_resblks to factor in the size of the refcount btree when it guesses how much space we need to start a repair. This hasn't been a huge problem in practice because there are few filesystems with (a) realtime, (b) rtgroups, (c) reflink, and (d) no rmap. But let's fix this before someone stumbles upon it, especially since LOLLM flagged this for me. Cc: stable@vger.kernel.org # v6.14 Fixes: 83ccffc489975d ("xfs: online repair of the realtime refcount btree") Signed-off-by: Darrick J. Wong Assisted-by: LOLLM # finding obvious bugs Reviewed-by: Christoph Hellwig Signed-off-by: Carlos Maiolino --- fs/xfs/scrub/repair.c | 21 ++++++++++++++++++--- fs/xfs/scrub/trace.h | 12 ++++++++---- 2 files changed, 26 insertions(+), 7 deletions(-) diff --git a/fs/xfs/scrub/repair.c b/fs/xfs/scrub/repair.c index 11697a8b2a1d..c2a437416227 100644 --- a/fs/xfs/scrub/repair.c +++ b/fs/xfs/scrub/repair.c @@ -399,6 +399,7 @@ xrep_calc_rtgroup_resblks( struct xfs_mount *mp = sc->mp; struct xfs_scrub_metadata *sm = sc->sm; uint64_t usedlen; + xfs_extlen_t refcbt_sz = 0; xfs_extlen_t rmapbt_sz = 0; if (!(sm->sm_flags & XFS_SCRUB_IFLAG_REPAIR)) @@ -411,13 +412,27 @@ xrep_calc_rtgroup_resblks( usedlen = xfs_rtbxlen_to_blen(mp, xfs_rtgroup_extents(mp, sm->sm_agno)); ASSERT(usedlen <= XFS_MAX_RGBLOCKS); + if (xfs_has_reflink(mp)) + refcbt_sz = xfs_rtrefcountbt_calc_size(mp, usedlen); + if (xfs_has_rmapbt(mp)) rmapbt_sz = xfs_rtrmapbt_calc_size(mp, usedlen); - trace_xrep_calc_rtgroup_resblks_btsize(mp, sm->sm_agno, usedlen, - rmapbt_sz); + /* + * Guess how many blocks we need to rebuild the rmapbt. For + * non-reflink filesystems we can't have more records than used blocks. + * However, with reflink it's possible to have more than one rmap + * record per rtgroup block. We don't know how many rmaps there could + * be in the rtgroup, so we start off with what we hope is an generous + * over-estimation. + */ + if (refcbt_sz > 0 && rmapbt_sz > 0) + rmapbt_sz *= 2; - return rmapbt_sz; + trace_xrep_calc_rtgroup_resblks_btsize(mp, sm->sm_agno, usedlen, + rmapbt_sz, refcbt_sz); + + return max(rmapbt_sz, refcbt_sz); } #endif /* CONFIG_XFS_RT */ diff --git a/fs/xfs/scrub/trace.h b/fs/xfs/scrub/trace.h index 0f5adc293962..cb85f75ce101 100644 --- a/fs/xfs/scrub/trace.h +++ b/fs/xfs/scrub/trace.h @@ -2376,25 +2376,29 @@ TRACE_EVENT(xrep_calc_ag_resblks_btsize, #ifdef CONFIG_XFS_RT TRACE_EVENT(xrep_calc_rtgroup_resblks_btsize, TP_PROTO(struct xfs_mount *mp, xfs_rgnumber_t rgno, - xfs_rgblock_t usedlen, xfs_rgblock_t rmapbt_sz), - TP_ARGS(mp, rgno, usedlen, rmapbt_sz), + xfs_rgblock_t usedlen, xfs_rgblock_t rmapbt_sz, + xfs_rgblock_t refcbt_sz), + TP_ARGS(mp, rgno, usedlen, rmapbt_sz, refcbt_sz), TP_STRUCT__entry( __field(dev_t, dev) __field(xfs_rgnumber_t, rgno) __field(xfs_rgblock_t, usedlen) __field(xfs_rgblock_t, rmapbt_sz) + __field(xfs_rgblock_t, refcbt_sz) ), TP_fast_assign( __entry->dev = mp->m_super->s_dev; __entry->rgno = rgno; __entry->usedlen = usedlen; __entry->rmapbt_sz = rmapbt_sz; + __entry->refcbt_sz = refcbt_sz; ), - TP_printk("dev %d:%d rgno 0x%x usedlen %u rmapbt %u", + TP_printk("dev %d:%d rgno 0x%x usedlen %u rmapbt %u refcountbt %u", MAJOR(__entry->dev), MINOR(__entry->dev), __entry->rgno, __entry->usedlen, - __entry->rmapbt_sz) + __entry->rmapbt_sz, + __entry->refcbt_sz) ); #endif /* CONFIG_XFS_RT */ From d7b92cbe566f6fe54368f62e4b515d9f80c43a18 Mon Sep 17 00:00:00 2001 From: "Darrick J. Wong" Date: Mon, 14 Sep 2026 22:37:36 -0700 Subject: [PATCH 288/592] xfs: don't cross reference rmapbt with bitmaps if they're incomplete LOLLM points out that runtime errors (e.g. ENOMEM) when we're trying to compute space usag bitmaps are silently dropped by the rmapbt scrubber. We ought to flag that as an incomplete scrub instead of reporting cross-referencing errors based on faulty data. Cc: stable@vger.kernel.org # v6.4 Fixes: fed050f3452da0 ("xfs: cross-reference rmap records with ag btrees") Signed-off-by: Darrick J. Wong Assisted-by: LOLLM # finding obvious bugs Reviewed-by: Christoph Hellwig Signed-off-by: Carlos Maiolino --- fs/xfs/scrub/rmap.c | 20 ++++++++++++++------ 1 file changed, 14 insertions(+), 6 deletions(-) diff --git a/fs/xfs/scrub/rmap.c b/fs/xfs/scrub/rmap.c index 0cd3eecd2ca5..68e2847c962b 100644 --- a/fs/xfs/scrub/rmap.c +++ b/fs/xfs/scrub/rmap.c @@ -493,11 +493,18 @@ xchk_rmapbt_walk_ag_metadata( * If there's an error, set XFAIL and disable the bitmap * cross-referencing checks, but proceed with the scrub anyway. */ - if (error) - xchk_btree_xref_process_error(sc, sc->sa.rmap_cur, - sc->sa.rmap_cur->bc_nlevels - 1, &error); - else - cr->bitmaps_complete = true; + if (error) { + if (!xchk_btree_xref_process_error(sc, sc->sa.rmap_cur, + sc->sa.rmap_cur->bc_nlevels - 1, &error)) { + /* only set incomplete if we didn't set xfail */ + if (error) + xchk_set_incomplete(sc); + } + + return 0; + } + + cr->bitmaps_complete = true; return 0; } @@ -567,7 +574,8 @@ xchk_rmapbt( if (error) goto out; - xchk_rmapbt_check_bitmaps(sc, cr); + if (cr->bitmaps_complete) + xchk_rmapbt_check_bitmaps(sc, cr); out: xagb_bitmap_destroy(&cr->refcbt_owned); From ab1c416d2377cdc16123ef521aac4da1c468c3d4 Mon Sep 17 00:00:00 2001 From: "Darrick J. Wong" Date: Mon, 14 Sep 2026 22:37:52 -0700 Subject: [PATCH 289/592] xfs: don't let memory failures leak blocks and kill repairs LOLLM complains that a memory allocation failure in xrep_newbt_add_blocks results in online repair leaking blocks that were previously allocated to write a new btree, but the problem is worse than that -- a limitation of the codebase is that the callers cannot undo the transaction /and/ return the error -- either you undo all changes and commit the transaction, or you error out and the filesystem goes down. However, the new btree space reservation object isn't that big (~48 bytes). Let's just do a NOFAIL allocation and the problem goes away. Cc: stable@vger.kernel.org # v6.8 Fixes: be408417630427 ("xfs: implement block reservation accounting for btrees we're staging") Signed-off-by: Darrick J. Wong Assisted-by: LOLLM # finding obvious bugs Reviewed-by: Christoph Hellwig Signed-off-by: Carlos Maiolino --- fs/xfs/scrub/newbt.c | 8 +++++--- 1 file changed, 5 insertions(+), 3 deletions(-) diff --git a/fs/xfs/scrub/newbt.c b/fs/xfs/scrub/newbt.c index c82f4631fd9c..584076b2a6ee 100644 --- a/fs/xfs/scrub/newbt.c +++ b/fs/xfs/scrub/newbt.c @@ -193,9 +193,11 @@ xrep_newbt_add_blocks( struct xrep_newbt_resv *resv; int error; - resv = kmalloc_obj(struct xrep_newbt_resv, XCHK_GFP_FLAGS); - if (!resv) - return -ENOMEM; + /* + * We have no way to clean up the allocated space *and* return an + * ENOMEM if we fail to allocate this control structure. + */ + resv = kmalloc_obj(struct xrep_newbt_resv, GFP_KERNEL | __GFP_NOFAIL); INIT_LIST_HEAD(&resv->list); resv->agbno = XFS_FSB_TO_AGBNO(mp, args->fsbno); From d80993655f7be2a461c0a63e2022da5757f47dac Mon Sep 17 00:00:00 2001 From: "Darrick J. Wong" Date: Mon, 14 Sep 2026 22:38:07 -0700 Subject: [PATCH 290/592] xfs: don't merge different file IO error types LOLLM noticed that we can accidentally merge file range health monitoring events even if they have different errors. We shouldn't do that. Cc: stable@vger.kernel.org # v7.0 Fixes: dfa8bad3a8796c ("xfs: convey file I/O errors to the health monitor") Signed-off-by: Darrick J. Wong Assisted-by: LOLLM # finding obvious bugs Reviewed-by: Christoph Hellwig Signed-off-by: Carlos Maiolino --- fs/xfs/xfs_healthmon.c | 4 +++- 1 file changed, 3 insertions(+), 1 deletion(-) diff --git a/fs/xfs/xfs_healthmon.c b/fs/xfs/xfs_healthmon.c index c3749675ef19..2bdd747f1ead 100644 --- a/fs/xfs/xfs_healthmon.c +++ b/fs/xfs/xfs_healthmon.c @@ -247,7 +247,9 @@ xfs_healthmon_merge_events( case XFS_HEALTHMON_DIOWRITE: case XFS_HEALTHMON_DATALOST: /* logically adjacent file ranges can merge */ - if (existing->fino != new->fino || existing->fgen != new->fgen) + if (existing->fino != new->fino || + existing->fgen != new->fgen || + existing->error != new->error) return false; if (existing->fpos + existing->flen == new->fpos) { From f8f6382ff13109e19d0fc1d0224ff7d29641e56a Mon Sep 17 00:00:00 2001 From: "Darrick J. Wong" Date: Mon, 14 Sep 2026 22:38:23 -0700 Subject: [PATCH 291/592] xfs: fix blockgc group quota scanning when usrquota isn't enforced LOLLM noticed the copy-paste error here -- if user quotas aren't enforced but we're near the group quota limit, we fail to set FLAG_GID and hence we might not actually free any preallocations, causing unnecessary EDQUOT. Fix that. Cc: stable@vger.kernel.org # v5.12 Fixes: c237dd7c709432 ("xfs: flush eof/cowblocks if we can't reserve quota for inode creation") Signed-off-by: Darrick J. Wong Assisted-by: LOLLM # finding obvious bugs Reviewed-by: Christoph Hellwig Signed-off-by: Carlos Maiolino --- fs/xfs/xfs_icache.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/fs/xfs/xfs_icache.c b/fs/xfs/xfs_icache.c index 82dac88e3c4c..de8be344e987 100644 --- a/fs/xfs/xfs_icache.c +++ b/fs/xfs/xfs_icache.c @@ -1653,7 +1653,7 @@ xfs_blockgc_free_dquots( do_work = true; } - if (XFS_IS_UQUOTA_ENFORCED(mp) && gdqp && xfs_dquot_lowsp(gdqp)) { + if (XFS_IS_GQUOTA_ENFORCED(mp) && gdqp && xfs_dquot_lowsp(gdqp)) { icw.icw_gid = make_kgid(mp->m_super->s_user_ns, gdqp->q_id); icw.icw_flags |= XFS_ICWALK_FLAG_GID; do_work = true; From 65f39d09d73718611cee40399179323b5d4ead00 Mon Sep 17 00:00:00 2001 From: "Darrick J. Wong" Date: Mon, 14 Sep 2026 22:38:38 -0700 Subject: [PATCH 292/592] xfs: fix cursor and pointer handling when recovering iunlink buckets LOLLM pointed out a bug in xlog_recover_iunlink_bucket: 1. We don't null out prev_ip after releasing it, which can lead to UAF problems if the inodegc flush call in the loop fails. at which point I noticed even more bugs: 2. If the inodegc flush inside the loop fails, we also leak @ip. 3. We set prev_agino to agino having already advanced agino, which results in inodes with i_prev_unlinked set to itself. 4. If we exit the bottom of the loop with prev_ip set, then prev_ip aliases ip and we also set its i_prev_unlinked to itself. Bugs 3 and 4 introduce loops into the unlinked list, though these loops don't surface because we immediately flush each unlinked inode after loading it. Fix all of these issues. Cc: stable@vger.kernel.org # v6.0 Fixes: 04755d2e5821b3 ("xfs: refactor xlog_recover_process_iunlinks()") Signed-off-by: Darrick J. Wong Assisted-by: LOLLM # finding obvious bugs Reviewed-by: Christoph Hellwig Signed-off-by: Carlos Maiolino --- fs/xfs/xfs_log_recover.c | 13 ++++++++----- 1 file changed, 8 insertions(+), 5 deletions(-) diff --git a/fs/xfs/xfs_log_recover.c b/fs/xfs/xfs_log_recover.c index e7e49529658b..cf0d610265fe 100644 --- a/fs/xfs/xfs_log_recover.c +++ b/fs/xfs/xfs_log_recover.c @@ -2736,12 +2736,13 @@ xlog_recover_iunlink_bucket( { struct xfs_mount *mp = pag_mount(pag); struct xfs_inode *prev_ip = NULL; - struct xfs_inode *ip; xfs_agino_t prev_agino, agino; int error = 0; agino = be32_to_cpu(agi->agi_unlinked[bucket]); while (agino != NULLAGINO) { + struct xfs_inode *ip; + error = xfs_iget(mp, NULL, xfs_agino_to_ino(pag, agino), 0, 0, &ip); if (error) @@ -2750,11 +2751,11 @@ xlog_recover_iunlink_bucket( ASSERT(VFS_I(ip)->i_nlink == 0); ASSERT(VFS_I(ip)->i_mode != 0); xfs_iflags_clear(ip, XFS_IRECOVERY); - agino = ip->i_next_unlinked; if (prev_ip) { ip->i_prev_unlinked = prev_agino; xfs_irele(prev_ip); + prev_ip = NULL; /* * Ensure the inode is removed from the unlinked list @@ -2766,18 +2767,20 @@ xlog_recover_iunlink_bucket( * complete. */ error = xfs_inodegc_flush(mp); - if (error) - break; + if (error) { + xfs_irele(ip); + return error; + } } prev_agino = agino; + agino = ip->i_next_unlinked; prev_ip = ip; } if (prev_ip) { int error2; - ip->i_prev_unlinked = prev_agino; xfs_irele(prev_ip); error2 = xfs_inodegc_flush(mp); From ffb48dccce1960a9ea24463a2f3c21d124d6b672 Mon Sep 17 00:00:00 2001 From: "Darrick J. Wong" Date: Mon, 14 Sep 2026 22:38:54 -0700 Subject: [PATCH 293/592] xfs: drop dquot flush lock when we can't find a buffer to flush LOLLM noticed that xfs_qm_flush_one fails to drop the dquot flush lock if it can't grab the buffer associated with the dquot. Since there's no buffer, nobody else is going to drop the dqflock, so we need to do it ourselves. Cc: stable@vger.kernel.org # v6.13 Fixes: ca378189fdfa89 ("xfs: convert quotacheck to attach dquot buffers") Signed-off-by: Darrick J. Wong Assisted-by: LOLLM # finding obvious bugs Reviewed-by: Christoph Hellwig Signed-off-by: Carlos Maiolino --- fs/xfs/xfs_qm.c | 10 ++++++++-- 1 file changed, 8 insertions(+), 2 deletions(-) diff --git a/fs/xfs/xfs_qm.c b/fs/xfs/xfs_qm.c index 99a82107b8e6..54d00d543b51 100644 --- a/fs/xfs/xfs_qm.c +++ b/fs/xfs/xfs_qm.c @@ -1432,16 +1432,22 @@ xfs_qm_flush_one( error = xfs_dquot_use_attached_buf(dqp, &bp); if (error) - goto out_unlock; + goto out_dqflock; if (!bp) { error = -EFSCORRUPTED; - goto out_unlock; + goto out_dqflock; } error = xfs_qm_dqflush(dqp, bp); if (!error) xfs_buf_delwri_queue(bp, buffer_list); xfs_buf_relse(bp); + mutex_unlock(&dqp->q_qlock); + xfs_qm_dqrele(dqp); + return error; + +out_dqflock: + xfs_dqfunlock(dqp); out_unlock: mutex_unlock(&dqp->q_qlock); xfs_qm_dqrele(dqp); From 476582d754cdc5110f806001417fea6c77824c13 Mon Sep 17 00:00:00 2001 From: "Darrick J. Wong" Date: Mon, 14 Sep 2026 22:39:10 -0700 Subject: [PATCH 294/592] xfs: don't let hidden_space go negative in xfs_metafile_resv_init LOLLM points out that if the amount of fdblocks that we can reserve for a metadata btree file goes below the space already used by that file, then the hidden_space subtraction can underflow, causing xfs_dec_fdblocks to subtract a huge amount of space. We never want the target to be less than the used sapce, so fix the logic that adjusts dblocks_avail downwards. Also fix an error in the adjacent comment. Cc: stable@vger.kernel.org # v6.15 Fixes: 1df8d75030b787 ("xfs: make metabtree reservations global") Signed-off-by: Darrick J. Wong Assisted-by: LOLLM # finding obvious bugs Reviewed-by: Christoph Hellwig Signed-off-by: Carlos Maiolino --- fs/xfs/libxfs/xfs_metafile.c | 8 ++++---- 1 file changed, 4 insertions(+), 4 deletions(-) diff --git a/fs/xfs/libxfs/xfs_metafile.c b/fs/xfs/libxfs/xfs_metafile.c index 71f004e9dc64..1f54d39003c2 100644 --- a/fs/xfs/libxfs/xfs_metafile.c +++ b/fs/xfs/libxfs/xfs_metafile.c @@ -297,14 +297,14 @@ xfs_metafile_resv_init( goto out_unlock; /* - * Space taken by the per-AG metadata btrees are accounted on-disk as - * used space. We therefore only hide the space that is reserved but - * not used by the trees. + * Space taken by metadata btrees are accounted on-disk as used space. + * We therefore only hide the space that is reserved but not used by + * the trees. */ if (used > target) target = used; else if (target > dblocks_avail) - target = dblocks_avail; + target = max(dblocks_avail, used); hidden_space = target - used; error = xfs_dec_fdblocks(mp, hidden_space, true); From fe2f9135df43db849e74f03956d322ac20b59af7 Mon Sep 17 00:00:00 2001 From: "Darrick J. Wong" Date: Mon, 14 Sep 2026 22:39:25 -0700 Subject: [PATCH 295/592] xfs: fix wild memcpy access when formatting ondisk rtrefcount btree roots LOLLM noticed that the inode btree root formatting methods copy too many bytes -- there's only one set of keys in node blocks, not two. This causes memory corruption of whatever's beyond the buffers. Cc: stable@vger.kernel.org # v6.14 Fixes: f0415af60f482a ("xfs: wire up a new metafile type for the realtime refcount") Signed-off-by: Darrick J. Wong Assisted-by: LOLLM # finding obvious bugs Reviewed-by: Christoph Hellwig Signed-off-by: Carlos Maiolino --- fs/xfs/libxfs/xfs_rtrefcount_btree.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/fs/xfs/libxfs/xfs_rtrefcount_btree.c b/fs/xfs/libxfs/xfs_rtrefcount_btree.c index e2950dbe2068..dcc89b8e149b 100644 --- a/fs/xfs/libxfs/xfs_rtrefcount_btree.c +++ b/fs/xfs/libxfs/xfs_rtrefcount_btree.c @@ -617,7 +617,7 @@ xfs_rtrefcountbt_from_disk( fpp = xfs_rtrefcount_droot_ptr_addr(dblock, 1, maxrecs); tpp = xfs_rtrefcount_broot_ptr_addr(mp, rblock, 1, rblocklen); numrecs = be16_to_cpu(dblock->bb_numrecs); - memcpy(tkp, fkp, 2 * sizeof(*fkp) * numrecs); + memcpy(tkp, fkp, sizeof(*fkp) * numrecs); memcpy(tpp, fpp, sizeof(*fpp) * numrecs); } else { frp = xfs_rtrefcount_droot_rec_addr(dblock, 1); @@ -703,7 +703,7 @@ xfs_rtrefcountbt_to_disk( fpp = xfs_rtrefcount_broot_ptr_addr(mp, rblock, 1, rblocklen); tpp = xfs_rtrefcount_droot_ptr_addr(dblock, 1, maxrecs); numrecs = be16_to_cpu(rblock->bb_numrecs); - memcpy(tkp, fkp, 2 * sizeof(*fkp) * numrecs); + memcpy(tkp, fkp, sizeof(*fkp) * numrecs); memcpy(tpp, fpp, sizeof(*fpp) * numrecs); } else { frp = xfs_rtrefcount_rec_addr(rblock, 1); From 9413959fa9fe2d94d4814f8cc2b60409f4cd46b5 Mon Sep 17 00:00:00 2001 From: Kevin Wang Date: Tue, 25 Aug 2026 23:09:32 +0800 Subject: [PATCH 296/592] drm/amd/pm: report energy accumulator for smu 14.0.3 add energy accumulator on pmfw 0x00685000 and above version. Signed-off-by: Kevin Wang Reviewed-by: Kenneth Feng Signed-off-by: Alex Deucher (cherry picked from commit 4aa733ab15b303e2a40e2985ac21a0e01f24cc4a) --- drivers/gpu/drm/amd/pm/swsmu/smu14/smu_v14_0_2_ppt.c | 3 +++ 1 file changed, 3 insertions(+) diff --git a/drivers/gpu/drm/amd/pm/swsmu/smu14/smu_v14_0_2_ppt.c b/drivers/gpu/drm/amd/pm/swsmu/smu14/smu_v14_0_2_ppt.c index 56a5c11bc196..ed99e61f18e1 100644 --- a/drivers/gpu/drm/amd/pm/swsmu/smu14/smu_v14_0_2_ppt.c +++ b/drivers/gpu/drm/amd/pm/swsmu/smu14/smu_v14_0_2_ppt.c @@ -2142,6 +2142,7 @@ static void smu_v14_0_2_init_msg_ctl(struct smu_context *smu) static ssize_t smu_v14_0_2_get_gpu_metrics(struct smu_context *smu, void **table) { + uint32_t mp1_ver = amdgpu_ip_version(smu->adev, MP1_HWIP, 0); struct gpu_metrics_v1_3 *gpu_metrics = (struct gpu_metrics_v1_3 *)smu_driver_table_ptr( smu, SMU_DRIVER_TABLE_GPU_METRICS); @@ -2171,6 +2172,8 @@ static ssize_t smu_v14_0_2_get_gpu_metrics(struct smu_context *smu, metrics->Vcn1ActivityPercentage); gpu_metrics->average_socket_power = metrics->AverageSocketPower; + if (mp1_ver == IP_VERSION(14, 0, 3) && smu->smc_fw_version >= 0x00685000) + gpu_metrics->energy_accumulator = metrics->EnergyAccumulator; if (metrics->AverageGfxActivity <= SMU_14_0_2_BUSY_THRESHOLD) gpu_metrics->average_gfxclk_frequency = metrics->AverageGfxclkFrequencyPostDs; From 63e19ef3ddab806c472748c825f4dc88dcd994e8 Mon Sep 17 00:00:00 2001 From: Leo Li Date: Tue, 25 Aug 2026 23:21:53 -0400 Subject: [PATCH 297/592] drm/amd/display: Atomize IRQ register read/modify/write ops [Why] The OTG_GLOBAL_SYNC_STATUS register controls various HW IRQ sources for the output timing generator (OTG). VUPDATE_NO_LOCK is one of them. To enable the IRQ, driver sets the VUPDATE_NO_LOCK_EN bit in the GLOBAL_SYNC_STATUS register. To ack the IRQ after it fires, the driver sets the VUPDATE_NO_LOCK_CLEAR bit in the same GLOBAL_SYNC_STATUS register. The bit sets are done through read/modify/write operations, which are not atomic. Thus, the following race is possible: Thread A: IRQ handler: *HW IRQ fires* # IRQ disable val = read(GLOBAL_SYNC_STATUS) unset(val, VUPDATE_NO_LOCK_EN) write(val, GLOBAL_SYNC_STATUS) # ACK reads VUPDATE_NO_LOCK_EN unset val1 = read(GLOBAL_SYNC_STATUS) set(val1, VUPDATE_NO_LOCK_CLEAR) # IRQ enable val = read(GLOBAL_SYNC_STATUS) set(val, VUPDATE_NO_LOCK_EN) write(val, GLOBAL_SYNC_STATUS) # BAD! clears VUPDATE_NO_LOCK_EN write(val1, GLOBAL_SYNC_STATUS) Regarding the tagged Fixes: change, it appears the change made this race more likely to occur. Since VUPDATE_NO_LOCK is now the sole IRQ source for vblank handling, a single race on high refresh panels can lead to a time out. [How] The GLOBAL_SYNC_STATUS register is only one example, other IRQ control registers also share the same scheme. On top of GLOBAL_SYNC_STATUS, let's clean up those as well. To keep things simple, Let's atomize the IRQ rmw ops via a single driver-wide spinlock. Due to the small scope of this lock, it is unlikely to cause noticeable overhead on top of all the existing locking within the IRQ set/handle paths. Since DM is responsible for locking, wrap dc_interrupt_set/ack with the spinlock in the new amdgpu_dm_irq_set/ack functions. Migrate/drop all references in DM to dc_interrupt_set/ack to use amdgpu_dm_irq_set/ack instead. Closes: https://gitlab.freedesktop.org/drm/amd/-/work_items/5616 Fixes: c87e6635d2db ("drm/amd/display: consolidate DCN vblank/flip handling onto vupdate_no_lock") Reviewed-by: Mario Limonciello Signed-off-by: Leo Li Signed-off-by: Chenyu Chen Tested-by: Daniel Wheeler Signed-off-by: Alex Deucher (cherry picked from commit 70de0a0216583a53c946155f8c8adedfdca6b4e7) Cc: stable@vger.kernel.org --- .../gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.c | 4 +- .../gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.h | 12 ++++ .../amd/display/amdgpu_dm/amdgpu_dm_crtc.c | 3 +- .../amd/display/amdgpu_dm/amdgpu_dm_helpers.c | 3 +- .../drm/amd/display/amdgpu_dm/amdgpu_dm_irq.c | 64 +++++++++++-------- .../drm/amd/display/amdgpu_dm/amdgpu_dm_irq.h | 28 ++++++++ .../amdgpu_dm/tests/amdgpu_dm_crtc_test.c | 16 +++-- .../amdgpu_dm/tests/amdgpu_dm_helpers_test.c | 8 ++- .../amdgpu_dm/tests/amdgpu_dm_irq_test.c | 34 +++++----- 9 files changed, 117 insertions(+), 55 deletions(-) diff --git a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.c b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.c index 08b8605029ab..36d2f86f000a 100644 --- a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.c +++ b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.c @@ -1420,7 +1420,7 @@ static void dm_gpureset_toggle_interrupts(struct amdgpu_device *adev, if (acrtc && state->stream_status[i].plane_count != 0 && amdgpu_ip_version(adev, DCE_HWIP, 0) == 0) { irq_source = IRQ_TYPE_PFLIP + acrtc->otg_inst; - rc = dc_interrupt_set(adev->dm.dc, irq_source, enable) ? 0 : -EBUSY; + rc = amdgpu_dm_irq_set(adev, irq_source, enable) ? 0 : -EBUSY; if (rc) drm_warn(adev_to_drm(adev), "Failed to %s pflip interrupts\n", enable ? "enable" : "disable"); @@ -1444,7 +1444,7 @@ static void dm_gpureset_toggle_interrupts(struct amdgpu_device *adev, /* During gpu-reset we disable and then enable vblank irq, so * don't use amdgpu_irq_get/put() to avoid refcount change. */ - if (!dc_interrupt_set(adev->dm.dc, irq_source, enable)) + if (!amdgpu_dm_irq_set(adev, irq_source, enable)) drm_warn(adev_to_drm(adev), "Failed to %sable vblank interrupt\n", enable ? "en" : "dis"); } else if (acrtc && state->stream_status[i].plane_count != 0) { diff --git a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.h b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.h index 3524931451c8..881c8d1c3cc0 100644 --- a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.h +++ b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.h @@ -552,6 +552,18 @@ struct amdgpu_display_manager { struct common_irq_params vupdate_params[DC_IRQ_SOURCE_VUPDATE6 - DC_IRQ_SOURCE_VUPDATE1 + 1]; + /** + * @irq_reg_lock: + * + * Serializes the read-modify-writes of the HW interrupt control + * registers. Several interrupt sources share one register - e.g. the + * enable and clear bits of both VSTARTUP (vblank) and VUPDATE_NO_LOCK + * live in OTG_GLOBAL_SYNC_STATUS. Therefore, enabling one source must + * not race with acking another. Held only across amdgpu_dm_irq_set() + * and amdgpu_dm_irq_ack(). + */ + spinlock_t irq_reg_lock; + /** * @dmub_trace_params: * diff --git a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_crtc.c b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_crtc.c index 62eac6e65334..1d941be73561 100644 --- a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_crtc.c +++ b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_crtc.c @@ -31,6 +31,7 @@ #include "amdgpu_dm_psr.h" #include "amdgpu_dm_replay.h" #include "amdgpu_dm_crtc.h" +#include "amdgpu_dm_irq.h" #include "amdgpu_dm_plane.h" #include "amdgpu_dm_trace.h" #include "amdgpu_dm_debugfs.h" @@ -91,7 +92,7 @@ int amdgpu_dm_crtc_set_vupdate_irq(struct drm_crtc *crtc, bool enable) irq_source = IRQ_TYPE_VUPDATE + acrtc->otg_inst; - rc = dc_interrupt_set(adev->dm.dc, irq_source, enable) ? 0 : -EBUSY; + rc = amdgpu_dm_irq_set(adev, irq_source, enable) ? 0 : -EBUSY; DRM_DEBUG_VBL("crtc %d - vupdate irq %sabling: r=%d\n", acrtc->crtc_id, enable ? "en" : "dis", rc); diff --git a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_helpers.c b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_helpers.c index 298de7b75ca8..ced8b3d2d762 100644 --- a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_helpers.c +++ b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_helpers.c @@ -1439,12 +1439,13 @@ void dm_helpers_free_gpu_mem( bool dm_helpers_dmub_outbox_interrupt_control(struct dc_context *ctx, bool enable) { + struct amdgpu_device *adev = ctx->driver_context; enum dc_irq_source irq_source; bool ret; irq_source = DC_IRQ_SOURCE_DMCUB_OUTBOX; - ret = dc_interrupt_set(ctx->dc, irq_source, enable); + ret = amdgpu_dm_irq_set(adev, irq_source, enable); DRM_DEBUG_DRIVER("Dmub trace irq %sabling: r=%d\n", enable ? "en" : "dis", ret); diff --git a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_irq.c b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_irq.c index d0239a3de2e1..74a8735168aa 100644 --- a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_irq.c +++ b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_irq.c @@ -396,6 +396,7 @@ int amdgpu_dm_irq_init(struct amdgpu_device *adev) DRM_DEBUG_KMS("DM_IRQ\n"); spin_lock_init(&adev->dm.irq_handler_list_table_lock); + spin_lock_init(&adev->dm.irq_reg_lock); adev->dm.irq_wq = alloc_workqueue("amdgpu_dm_irq", WQ_UNBOUND | WQ_HIGHPRI, 0); @@ -530,7 +531,7 @@ void amdgpu_dm_irq_suspend(struct amdgpu_device *adev) */ for (src = DC_IRQ_SOURCE_HPD1; src <= DC_IRQ_SOURCE_HPD6RX; src++) { hnd_list_l = &adev->dm.irq_handler_list_low_tab[src]; - dc_interrupt_set(adev->dm.dc, src, false); + amdgpu_dm_irq_set(adev, src, false); DM_IRQ_TABLE_UNLOCK(adev, irq_table_flags); @@ -568,7 +569,7 @@ void amdgpu_dm_irq_resume_early(struct amdgpu_device *adev) hnd_list_l = &adev->dm.irq_handler_list_low_tab[src]; hnd_list_h = &adev->dm.irq_handler_list_high_tab[src]; if (!list_empty(hnd_list_l) || !list_empty(hnd_list_h)) - dc_interrupt_set(adev->dm.dc, src, true); + amdgpu_dm_irq_set(adev, src, true); } DM_IRQ_TABLE_UNLOCK(adev, irq_table_flags); @@ -594,7 +595,7 @@ void amdgpu_dm_irq_resume_late(struct amdgpu_device *adev) hnd_list_l = &adev->dm.irq_handler_list_low_tab[src]; hnd_list_h = &adev->dm.irq_handler_list_high_tab[src]; if (!list_empty(hnd_list_l) || !list_empty(hnd_list_h)) - dc_interrupt_set(adev->dm.dc, src, true); + amdgpu_dm_irq_set(adev, src, true); } DM_IRQ_TABLE_UNLOCK(adev, irq_table_flags); @@ -690,6 +691,23 @@ STATIC_IFN_KUNIT void amdgpu_dm_irq_immediate_work(struct amdgpu_device *adev, } EXPORT_IF_KUNIT(amdgpu_dm_irq_immediate_work); +bool amdgpu_dm_irq_set(struct amdgpu_device *adev, enum dc_irq_source src, + bool enable) +{ + guard(spinlock_irqsave)(&adev->dm.irq_reg_lock); + + return dc_interrupt_set(adev->dm.dc, src, enable); +} +EXPORT_IF_KUNIT(amdgpu_dm_irq_set); + +void amdgpu_dm_irq_ack(struct amdgpu_device *adev, enum dc_irq_source src) +{ + guard(spinlock_irqsave)(&adev->dm.irq_reg_lock); + + dc_interrupt_ack(adev->dm.dc, src); +} +EXPORT_IF_KUNIT(amdgpu_dm_irq_ack); + /** * amdgpu_dm_irq_handler - Generic DM IRQ handler * @adev: amdgpu base driver device containing the DM device @@ -710,7 +728,7 @@ STATIC_IFN_KUNIT int amdgpu_dm_irq_handler(struct amdgpu_device *adev, entry->src_id, entry->src_data[0]); - dc_interrupt_ack(adev->dm.dc, src); + amdgpu_dm_irq_ack(adev, src); /* Call high irq work immediately */ amdgpu_dm_irq_immediate_work(adev, src); @@ -750,7 +768,7 @@ STATIC_IFN_KUNIT int amdgpu_dm_set_hpd_irq_state(struct amdgpu_device *adev, enum dc_irq_source src = amdgpu_dm_hpd_to_dal_irq_source(type); bool st = (state == AMDGPU_IRQ_STATE_ENABLE); - dc_interrupt_set(adev->dm.dc, src, st); + amdgpu_dm_irq_set(adev, src, st); return 0; } EXPORT_IF_KUNIT(amdgpu_dm_set_hpd_irq_state); @@ -785,7 +803,7 @@ static inline int dm_irq_state(struct amdgpu_device *adev, if (dc && dc->caps.ips_support && dc->idle_optimizations_allowed) dc_allow_idle_optimizations(dc, false); - dc_interrupt_set(adev->dm.dc, irq_source, st); + amdgpu_dm_irq_set(adev, irq_source, st); return 0; } @@ -842,7 +860,7 @@ STATIC_IFN_KUNIT int amdgpu_dm_set_dmub_outbox_irq_state(struct amdgpu_device *a enum dc_irq_source irq_source = DC_IRQ_SOURCE_DMCUB_OUTBOX; bool st = (state == AMDGPU_IRQ_STATE_ENABLE); - dc_interrupt_set(adev->dm.dc, irq_source, st); + amdgpu_dm_irq_set(adev, irq_source, st); return 0; } EXPORT_IF_KUNIT(amdgpu_dm_set_dmub_outbox_irq_state); @@ -870,7 +888,7 @@ STATIC_IFN_KUNIT int amdgpu_dm_set_dmub_trace_irq_state(struct amdgpu_device *ad enum dc_irq_source irq_source = DC_IRQ_SOURCE_DMCUB_OUTBOX0; bool st = (state == AMDGPU_IRQ_STATE_ENABLE); - dc_interrupt_set(adev->dm.dc, irq_source, st); + amdgpu_dm_irq_set(adev, irq_source, st); return 0; } EXPORT_IF_KUNIT(amdgpu_dm_set_dmub_trace_irq_state); @@ -937,9 +955,7 @@ EXPORT_IF_KUNIT(amdgpu_dm_set_irq_funcs); void amdgpu_dm_outbox_init(struct amdgpu_device *adev) { - dc_interrupt_set(adev->dm.dc, - DC_IRQ_SOURCE_DMCUB_OUTBOX, - true); + amdgpu_dm_irq_set(adev, DC_IRQ_SOURCE_DMCUB_OUTBOX, true); } EXPORT_IF_KUNIT(amdgpu_dm_outbox_init); @@ -962,7 +978,7 @@ void amdgpu_dm_hpd_init(struct amdgpu_device *adev) /* First, clear all hpd and hpdrx interrupts */ for (i = DC_IRQ_SOURCE_HPD1; i <= DC_IRQ_SOURCE_HPD6RX; i++) { - if (!dc_interrupt_set(adev->dm.dc, i, false)) + if (!amdgpu_dm_irq_set(adev, i, false)) drm_err(dev, "Failed to clear hpd(rx) source=%d on init\n", i); } @@ -991,7 +1007,7 @@ void amdgpu_dm_hpd_init(struct amdgpu_device *adev) * of dm. Note that only hpd interrupt types are registered with * base driver; hpd_rx types aren't. IOW, amdgpu_irq_get/put on * hpd_rx isn't available. DM currently controls hpd_rx - * explicitly with dc_interrupt_set() + * explicitly with amdgpu_dm_irq_set() */ if (dc_link->irq_source_hpd != DC_IRQ_SOURCE_INVALID) { irq_type = dc_link->irq_source_hpd - DC_IRQ_SOURCE_HPD1; @@ -1000,23 +1016,21 @@ void amdgpu_dm_hpd_init(struct amdgpu_device *adev) * and what bios reports as the # of connectors with hpd * sources. Since the # of hpd source types registered * with base driver == mode_info.num_hpd, we have to - * fallback to dc_interrupt_set for the remaining types. + * fallback to amdgpu_dm_irq_set for the remaining types. */ if (irq_type < adev->mode_info.num_hpd) { if (amdgpu_irq_get(adev, &adev->hpd_irq, irq_type)) drm_err(dev, "DM_IRQ: Failed get HPD for source=%d)!\n", dc_link->irq_source_hpd); } else { - dc_interrupt_set(adev->dm.dc, - dc_link->irq_source_hpd, - true); + amdgpu_dm_irq_set(adev, dc_link->irq_source_hpd, + true); } } if (dc_link->irq_source_hpd_rx != DC_IRQ_SOURCE_INVALID) { - dc_interrupt_set(adev->dm.dc, - dc_link->irq_source_hpd_rx, - true); + amdgpu_dm_irq_set(adev, dc_link->irq_source_hpd_rx, + true); } } drm_connector_list_iter_end(&iter); @@ -1061,16 +1075,14 @@ void amdgpu_dm_hpd_fini(struct amdgpu_device *adev) drm_err(dev, "DM_IRQ: Failed put HPD for source=%d!\n", dc_link->irq_source_hpd); } else { - dc_interrupt_set(adev->dm.dc, - dc_link->irq_source_hpd, - false); + amdgpu_dm_irq_set(adev, dc_link->irq_source_hpd, + false); } } if (dc_link->irq_source_hpd_rx != DC_IRQ_SOURCE_INVALID) { - dc_interrupt_set(adev->dm.dc, - dc_link->irq_source_hpd_rx, - false); + amdgpu_dm_irq_set(adev, dc_link->irq_source_hpd_rx, + false); } } drm_connector_list_iter_end(&iter); diff --git a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_irq.h b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_irq.h index 4c200a9614a7..bc16ecc67329 100644 --- a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_irq.h +++ b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_irq.h @@ -89,6 +89,34 @@ void amdgpu_dm_irq_unregister_interrupt(struct amdgpu_device *adev, enum dc_irq_source irq_source, void *ih_index); +/** + * amdgpu_dm_irq_set - enable or disable a DC interrupt source. + * + * @adev: AMD DRM device + * @src: DC interrupt source to toggle + * @enable: true to enable the source, false to disable it + * + * DM-wide replacement for dc_interrupt_set(). As locking is DM's + * responsibility, this is a thin wrapper serializes the underlying + * read-modify-write against the other interrupt sources sharing HW control + * registers with @src, so DM must never call dc_interrupt_set() directly. + * + * Returns: true if the source was toggled. + */ +bool amdgpu_dm_irq_set(struct amdgpu_device *adev, enum dc_irq_source src, + bool enable); + +/** + * amdgpu_dm_irq_ack - acknowledge a DC interrupt source. + * + * @adev: AMD DRM device + * @src: DC interrupt source to acknowledge + * + * DM-wide replacement for dc_interrupt_ack(), serialized the same way as + * amdgpu_dm_irq_set(). + */ +void amdgpu_dm_irq_ack(struct amdgpu_device *adev, enum dc_irq_source src); + void amdgpu_dm_set_irq_funcs(struct amdgpu_device *adev); void amdgpu_dm_outbox_init(struct amdgpu_device *adev); diff --git a/drivers/gpu/drm/amd/display/amdgpu_dm/tests/amdgpu_dm_crtc_test.c b/drivers/gpu/drm/amd/display/amdgpu_dm/tests/amdgpu_dm_crtc_test.c index 0d998f204250..ae0f4da96252 100644 --- a/drivers/gpu/drm/amd/display/amdgpu_dm/tests/amdgpu_dm_crtc_test.c +++ b/drivers/gpu/drm/amd/display/amdgpu_dm/tests/amdgpu_dm_crtc_test.c @@ -436,7 +436,7 @@ static void dm_test_crtc_set_vupdate_irq_no_otg(struct kunit *test) * dm_test_crtc_set_vupdate_irq_dc_busy - Test vupdate irq when DC rejects request * @test: The KUnit test context * - * With an OTG instance assigned but no DC attached, dc_interrupt_set() returns + * With an OTG instance assigned but no DC attached, amdgpu_dm_irq_set() returns * false and the function must report the request as busy (-EBUSY). */ static void dm_test_crtc_set_vupdate_irq_dc_busy(struct kunit *test) @@ -453,12 +453,12 @@ static void dm_test_crtc_set_vupdate_irq_dc_busy(struct kunit *test) acrtc->base.dev = &adev->ddev; acrtc->otg_inst = 0; - /* adev->dm.dc is NULL, so dc_interrupt_set() returns false. */ + /* adev->dm.dc is NULL, so amdgpu_dm_irq_set() returns false. */ KUNIT_EXPECT_EQ(test, amdgpu_dm_crtc_set_vupdate_irq(&acrtc->base, true), -EBUSY); } -/* Per-source funcs let dc_interrupt_set() succeed without register access. */ +/* Per-source funcs let amdgpu_dm_irq_set() succeed without register access. */ static bool dm_test_vupdate_irq_src_set(struct irq_service *irq_service, const struct irq_source_info *info, bool enable) @@ -477,7 +477,7 @@ static struct irq_source_info_funcs dm_test_vupdate_irq_src_funcs = { .ack = dm_test_vupdate_irq_src_ack, }; -/* A .set that fails so dc_interrupt_set() reports the source as busy. */ +/* A .set that fails so amdgpu_dm_irq_set() reports the source as busy. */ static bool dm_test_vupdate_irq_src_set_busy(struct irq_service *irq_service, const struct irq_source_info *info, bool enable) @@ -519,7 +519,9 @@ static void dm_test_crtc_set_vupdate_irq_enable(struct kunit *test) irqs = kunit_kzalloc(test, sizeof(*irqs), GFP_KERNEL); KUNIT_ASSERT_NOT_ERR_OR_NULL(test, irqs); - /* Populate the per-source info table so dc_interrupt_set() succeeds. */ + /* + * Populate the per-source info table so amdgpu_dm_irq_set() succeeds. + */ info = kunit_kzalloc(test, sizeof(*info) * DAL_IRQ_SOURCES_NUMBER, GFP_KERNEL); KUNIT_ASSERT_NOT_ERR_OR_NULL(test, info); @@ -1018,7 +1020,9 @@ static void dm_test_crtc_enable_vblank_vupdate_busy(struct kunit *test) irqs = kunit_kzalloc(test, sizeof(*irqs), GFP_KERNEL); KUNIT_ASSERT_NOT_ERR_OR_NULL(test, irqs); - /* Per-source .set fails so dc_interrupt_set() reports the source busy. */ + /* + * Per-source .set fails so amdgpu_dm_irq_set() reports the source busy. + */ info = kunit_kzalloc(test, sizeof(*info) * DAL_IRQ_SOURCES_NUMBER, GFP_KERNEL); KUNIT_ASSERT_NOT_ERR_OR_NULL(test, info); diff --git a/drivers/gpu/drm/amd/display/amdgpu_dm/tests/amdgpu_dm_helpers_test.c b/drivers/gpu/drm/amd/display/amdgpu_dm/tests/amdgpu_dm_helpers_test.c index 82e0c984693c..639512bea275 100644 --- a/drivers/gpu/drm/amd/display/amdgpu_dm/tests/amdgpu_dm_helpers_test.c +++ b/drivers/gpu/drm/amd/display/amdgpu_dm/tests/amdgpu_dm_helpers_test.c @@ -2442,17 +2442,21 @@ static void dm_test_is_dp_sink_present_null_priv(struct kunit *test) * dm_test_dmub_outbox_interrupt_control_null_dc - Test outbox irq control with NULL dc * @test: The KUnit test context * - * dc_interrupt_set() is NULL-safe and returns false when dc is NULL, so the + * amdgpu_dm_irq_set() is NULL-safe and returns false when dc is NULL, so the * helper returns false without touching real interrupt hardware. */ static void dm_test_dmub_outbox_interrupt_control_null_dc(struct kunit *test) { + struct amdgpu_device *adev; struct dc_context *ctx; + adev = kunit_kzalloc(test, sizeof(*adev), GFP_KERNEL); + KUNIT_ASSERT_NOT_NULL(test, adev); ctx = kunit_kzalloc(test, sizeof(*ctx), GFP_KERNEL); KUNIT_ASSERT_NOT_NULL(test, ctx); + ctx->driver_context = adev; - /* ctx->dc is NULL → dc_interrupt_set returns false */ + /* adev->dm.dc is NULL → amdgpu_dm_irq_set returns false */ KUNIT_EXPECT_FALSE(test, dm_helpers_dmub_outbox_interrupt_control(ctx, true)); KUNIT_EXPECT_FALSE(test, dm_helpers_dmub_outbox_interrupt_control(ctx, false)); } diff --git a/drivers/gpu/drm/amd/display/amdgpu_dm/tests/amdgpu_dm_irq_test.c b/drivers/gpu/drm/amd/display/amdgpu_dm/tests/amdgpu_dm_irq_test.c index 861ee9eaa032..95322d8c8613 100644 --- a/drivers/gpu/drm/amd/display/amdgpu_dm/tests/amdgpu_dm_irq_test.c +++ b/drivers/gpu/drm/amd/display/amdgpu_dm/tests/amdgpu_dm_irq_test.c @@ -268,7 +268,7 @@ static bool dm_test_irq_src_ack(struct irq_service *irq_service, return true; } -/* Per-source funcs let dc_interrupt_set() succeed without register access. */ +/* Per-source funcs let amdgpu_dm_irq_set() succeed without register access. */ static struct irq_source_info_funcs dm_test_irq_src_funcs = { .set = dm_test_irq_src_set, .ack = dm_test_irq_src_ack, @@ -290,7 +290,7 @@ static struct dc *dm_test_alloc_dc_with_irq_service(struct kunit *test, KUNIT_ASSERT_NOT_ERR_OR_NULL(test, irqs); /* - * Populate the per-source info table so dc_interrupt_set()/_ack() + * Populate the per-source info table so amdgpu_dm_irq_set()/_ack() * succeed without touching hardware registers. */ info = kunit_kzalloc(test, sizeof(*info) * DAL_IRQ_SOURCES_NUMBER, @@ -1227,7 +1227,7 @@ static void dm_test_irq_suspend_empty(struct kunit *test) KUNIT_ASSERT_EQ(test, amdgpu_dm_irq_init(adev), 0); /* - * With no registered handlers the HW dc_interrupt_set() calls are + * With no registered handlers the amdgpu_dm_irq_set() calls are * skipped, so suspend must complete without touching the (absent) DC. */ amdgpu_dm_irq_suspend(adev); @@ -1275,11 +1275,11 @@ static void dm_test_irq_resume_late_empty(struct kunit *test) } /** - * dm_test_irq_suspend_registered - Test suspend reaches the dc_interrupt_set path + * dm_test_irq_suspend_registered - Test suspend reaches the irq set path * @test: The KUnit test context * * Registers a low-context HPD handler so the handler list is non-empty, - * forcing amdgpu_dm_irq_suspend() to call dc_interrupt_set() (NULL-safe with + * forcing amdgpu_dm_irq_suspend() to call amdgpu_dm_irq_set() (NULL-safe with * no DC) and flush_work() on the registered handler. */ static void dm_test_irq_suspend_registered(struct kunit *test) @@ -1330,11 +1330,11 @@ static void dm_test_irq_suspend_disables_polling(struct kunit *test) } /** - * dm_test_irq_resume_early_registered - Test early resume reaches dc_interrupt_set + * dm_test_irq_resume_early_registered - Test early resume reaches irq set * @test: The KUnit test context * * Registers a low-context HPD RX handler so early resume calls - * dc_interrupt_set() for the short-pulse interrupt source. + * amdgpu_dm_irq_set() for the short-pulse interrupt source. */ static void dm_test_irq_resume_early_registered(struct kunit *test) { @@ -1358,10 +1358,10 @@ static void dm_test_irq_resume_early_registered(struct kunit *test) } /** - * dm_test_irq_resume_late_registered - Test late resume reaches dc_interrupt_set + * dm_test_irq_resume_late_registered - Test late resume reaches irq set * @test: The KUnit test context * - * Registers a low-context HPD handler so late resume calls dc_interrupt_set() + * Registers a low-context HPD handler so late resume calls amdgpu_dm_irq_set() * for the HPD interrupt source. */ static void dm_test_irq_resume_late_registered(struct kunit *test) @@ -1592,7 +1592,7 @@ static void dm_test_set_crtc_irq_state_enable(struct kunit *test) /* * otg_inst >= 0 computes the irq source and reaches the NULL-safe - * dc_interrupt_set(); the ips_support branch is skipped (dc == NULL). + * amdgpu_dm_irq_set(); the ips_support branch is skipped (dc == NULL). */ acrtc->otg_inst = 3; adev->mode_info.crtcs[0] = acrtc; @@ -1671,8 +1671,8 @@ static void dm_test_set_vupdate_irq_state_enable(struct kunit *test) * * With a non-NULL DC that advertises IPS support and currently allows idle * optimizations, dm_irq_state() must call dc_allow_idle_optimizations() before - * dc_interrupt_set(). disable_idle_power_optimizations makes that call a safe - * early return, and per-source stub funcs let dc_interrupt_set() succeed. + * amdgpu_dm_irq_set(). disable_idle_power_optimizations makes that call a safe + * early return, and per-source stub funcs let amdgpu_dm_irq_set() succeed. */ static void dm_test_set_crtc_irq_state_allows_idle(struct kunit *test) { @@ -1891,7 +1891,7 @@ static void dm_test_set_hpd_irq_state_null_dc(struct kunit *test) adev = kunit_kzalloc(test, sizeof(*adev), GFP_KERNEL); KUNIT_ASSERT_NOT_ERR_OR_NULL(test, adev); - /* dc_interrupt_set() is a no-op when dc is NULL, so both states + /* amdgpu_dm_irq_set() is a no-op when dc is NULL, so both states * return 0 without dereferencing the (absent) DC. */ KUNIT_EXPECT_EQ(test, amdgpu_dm_set_hpd_irq_state(adev, NULL, AMDGPU_HPD_1, @@ -1951,7 +1951,7 @@ static void dm_test_outbox_init_null_dc(struct kunit *test) adev = kunit_kzalloc(test, sizeof(*adev), GFP_KERNEL); KUNIT_ASSERT_NOT_ERR_OR_NULL(test, adev); - /* Single dc_interrupt_set() call must be skipped when dc is NULL. */ + /* Single amdgpu_dm_irq_set() call must be skipped when dc is NULL. */ amdgpu_dm_outbox_init(adev); } @@ -1969,7 +1969,7 @@ static void dm_test_hpd_init_empty_connectors(struct kunit *test) /* * With an empty connector list the per-connector loop is skipped and - * the initial clear loop relies on dc_interrupt_set() being a no-op + * the initial clear loop relies on amdgpu_dm_irq_set() being a no-op * for a NULL dc, so init must complete without touching the DC. */ amdgpu_dm_hpd_init(adev); @@ -2004,7 +2004,7 @@ static void dm_test_hpd_init_fini_with_connectors(struct kunit *test) /* * num_hpd = 0 forces irq_type >= num_hpd so the loop takes the HW - * fallback (dc_interrupt_set()) instead of amdgpu_irq_get(); with a + * fallback (amdgpu_dm_irq_set()) instead of amdgpu_irq_get(); with a * NULL dc that fallback is a safe no-op. */ adev->mode_info.num_hpd = 0; @@ -2090,7 +2090,7 @@ static void dm_test_hpd_init_fini_irq_ref(struct kunit *test) /* * num_hpd >= 1 makes irq_type (0) < num_hpd, so the loop takes the * amdgpu_irq_get()/amdgpu_irq_put() branch instead of the - * dc_interrupt_set() fallback. The mock device has irq.installed == + * amdgpu_dm_irq_set() fallback. The mock device has irq.installed == * false, so both calls fail early with -ENOENT (logging an error) * without touching the base-driver irq state. */ From 5f28bb1c2cd9dcdb76a20d61b3ea069b85893c59 Mon Sep 17 00:00:00 2001 From: Vladimir Marioukhine Date: Wed, 12 Aug 2026 13:19:46 -0400 Subject: [PATCH 298/592] drm/amdkfd: implement restore_mqd callbacks for GFX12/12.1 kfd_mqd_manager_v12.c (GFX 12.0) and kfd_mqd_manager_v12_1.c (GFX 12.1) do not implement restore_mqd callbacks, leaving the function pointers NULL and causing CRIU restore to return -EOPNOTSUPP on GFX12. Implement restore_mqd for both compute and SDMA queues in kfd_mqd_manager_v12.c and kfd_mqd_manager_v12_1.c, modeled after the GFX 11 implementation with the following improvements: - update cp_mqd_base_addr_lo/hi to the newly allocated MQD address, fixing a pre-existing gap shared with v11 where the in-MQD copy still pointed at the old checkpoint-time address after restore - memset the full allocation before memcpy for compute queues to avoid stale data in the GTT sub-allocator tail; SDMA MQDs use sizeof(*m) since they are packed at mqd_size stride in a shared BO checkpoint_mqd registration is deferred to a follow-up patch that also implements get_checkpoint_info, so that checkpoint and restore are enabled together as a complete and testable unit. Note: GFX12.1 restore handles XCC0 only. Multi-XCC CRIU restore is currently unreachable due to a separate validation issue in kfd_criu_restore_queue(). A pr_warn_once() is emitted if a multi-XCC device is encountered. Signed-off-by: Vladimir Marioukhine Reviewed-by: Alex Deucher Signed-off-by: Alex Deucher (cherry picked from commit b1f9601237d050f5df478464cf51bf1fff29a256) Cc: stable@vger.kernel.org --- .../drm/amd/amdkfd/kfd_device_queue_manager.c | 7 +- .../gpu/drm/amd/amdkfd/kfd_mqd_manager_v12.c | 59 ++++++++++++++++ .../drm/amd/amdkfd/kfd_mqd_manager_v12_1.c | 68 +++++++++++++++++++ 3 files changed, 132 insertions(+), 2 deletions(-) diff --git a/drivers/gpu/drm/amd/amdkfd/kfd_device_queue_manager.c b/drivers/gpu/drm/amd/amdkfd/kfd_device_queue_manager.c index 9811e4e10291..2f78395a0c31 100644 --- a/drivers/gpu/drm/amd/amdkfd/kfd_device_queue_manager.c +++ b/drivers/gpu/drm/amd/amdkfd/kfd_device_queue_manager.c @@ -770,10 +770,12 @@ static int create_queue_nocpsch(struct device_queue_manager *dqm, mqd_mgr = dqm->mqd_mgrs[get_mqd_type_from_queue_type( q->properties.type)]; if (qd && !mqd_mgr->restore_mqd) { - pr_debug("restore_mqd not implemented for this GPU\n"); + pr_debug("restore_mqd not implemented for queue type %d\n", + q->properties.type); retval = -EOPNOTSUPP; goto deallocate_vmid; } + if (q->properties.type == KFD_QUEUE_TYPE_COMPUTE) { retval = allocate_hqd(dqm, q); if (retval) @@ -2250,7 +2252,8 @@ static int create_queue_cpsch(struct device_queue_manager *dqm, struct queue *q, mqd_mgr = dqm->mqd_mgrs[get_mqd_type_from_queue_type( q->properties.type)]; if (qd && !mqd_mgr->restore_mqd) { - pr_debug("restore_mqd not implemented for this GPU\n"); + pr_debug("restore_mqd not implemented for queue type %d\n", + q->properties.type); retval = -EOPNOTSUPP; goto out_deallocate_doorbell; } diff --git a/drivers/gpu/drm/amd/amdkfd/kfd_mqd_manager_v12.c b/drivers/gpu/drm/amd/amdkfd/kfd_mqd_manager_v12.c index 7c387fa90076..63f25a60baa5 100644 --- a/drivers/gpu/drm/amd/amdkfd/kfd_mqd_manager_v12.c +++ b/drivers/gpu/drm/amd/amdkfd/kfd_mqd_manager_v12.c @@ -380,6 +380,63 @@ static int debugfs_show_mqd_sdma(struct seq_file *m, void *data) #endif +static void restore_mqd(struct mqd_manager *mm, void **mqd, + struct kfd_mem_obj *mqd_mem_obj, uint64_t *gart_addr, + struct queue_properties *qp, const void *mqd_src, + const void *ctl_stack_src, const u32 ctl_stack_size) +{ + u64 addr; + struct v12_compute_mqd *m; + + m = (struct v12_compute_mqd *)mqd_mem_obj->cpu_ptr; + addr = mqd_mem_obj->gpu_addr; + + memset(m, 0, AMDGPU_MQD_SIZE_ALIGN(mm->mqd_size)); + memcpy(m, mqd_src, sizeof(*m)); + + /* Update MQD base address to the newly allocated location */ + m->cp_mqd_base_addr_lo = lower_32_bits(addr); + m->cp_mqd_base_addr_hi = upper_32_bits(addr); + + m->cp_hqd_pq_doorbell_control &= + ~CP_HQD_PQ_DOORBELL_CONTROL__DOORBELL_OFFSET_MASK; + m->cp_hqd_pq_doorbell_control |= + qp->doorbell_off << CP_HQD_PQ_DOORBELL_CONTROL__DOORBELL_OFFSET__SHIFT; + pr_debug("cp_hqd_pq_doorbell_control 0x%x\n", m->cp_hqd_pq_doorbell_control); + + *mqd = m; + if (gart_addr) + *gart_addr = addr; + + qp->is_active = 0; +} + +static void restore_mqd_sdma(struct mqd_manager *mm, void **mqd, + struct kfd_mem_obj *mqd_mem_obj, uint64_t *gart_addr, + struct queue_properties *qp, + const void *mqd_src, + const void *ctl_stack_src, + const u32 ctl_stack_size) +{ + u64 addr; + struct v12_sdma_mqd *m; + + m = (struct v12_sdma_mqd *)mqd_mem_obj->cpu_ptr; + addr = mqd_mem_obj->gpu_addr; + + memset(m, 0, AMDGPU_MQD_SIZE_ALIGN(mm->mqd_size)); + memcpy(m, mqd_src, sizeof(*m)); + + m->sdmax_rlcx_doorbell_offset = + qp->doorbell_off << SDMA0_QUEUE0_DOORBELL_OFFSET__OFFSET__SHIFT; + + *mqd = m; + if (gart_addr) + *gart_addr = addr; + + qp->is_active = 0; +} + struct mqd_manager *mqd_manager_init_v12(enum KFD_MQD_TYPE type, struct kfd_node *dev) { @@ -407,6 +464,7 @@ struct mqd_manager *mqd_manager_init_v12(enum KFD_MQD_TYPE type, mqd->mqd_size = sizeof(struct v12_compute_mqd); mqd->get_wave_state = get_wave_state; mqd->mqd_stride = kfd_mqd_stride; + mqd->restore_mqd = restore_mqd; #if defined(CONFIG_DEBUG_FS) mqd->debugfs_show_mqd = debugfs_show_mqd; #endif @@ -453,6 +511,7 @@ struct mqd_manager *mqd_manager_init_v12(enum KFD_MQD_TYPE type, mqd->is_occupied = kfd_is_occupied_sdma; mqd->mqd_size = sizeof(struct v12_sdma_mqd); mqd->mqd_stride = kfd_mqd_stride; + mqd->restore_mqd = restore_mqd_sdma; #if defined(CONFIG_DEBUG_FS) mqd->debugfs_show_mqd = debugfs_show_mqd_sdma; #endif diff --git a/drivers/gpu/drm/amd/amdkfd/kfd_mqd_manager_v12_1.c b/drivers/gpu/drm/amd/amdkfd/kfd_mqd_manager_v12_1.c index 431a940f91f3..708bbb08c089 100644 --- a/drivers/gpu/drm/amd/amdkfd/kfd_mqd_manager_v12_1.c +++ b/drivers/gpu/drm/amd/amdkfd/kfd_mqd_manager_v12_1.c @@ -641,6 +641,72 @@ static int debugfs_show_mqd_sdma(struct seq_file *m, void *data) #endif +static void restore_mqd_v12_1(struct mqd_manager *mm, void **mqd, + struct kfd_mem_obj *mqd_mem_obj, uint64_t *gart_addr, + struct queue_properties *qp, const void *mqd_src, + const void *ctl_stack_src, const u32 ctl_stack_size) +{ + u64 addr; + struct v12_1_compute_mqd *m; + + /* + * GFX12.1 is multi-XCC capable but this restore handles XCC0 only. + * Multi-XCC CRIU restore is currently unreachable because + * kfd_criu_restore_queue() validates against unscaled mqd_size. + */ + if (NUM_XCC(mm->dev->xcc_mask) > 1) + pr_warn_once("GFX12.1 multi-XCC CRIU restore not fully supported\n"); + + m = (struct v12_1_compute_mqd *)mqd_mem_obj->cpu_ptr; + addr = mqd_mem_obj->gpu_addr; + + memset(m, 0, AMDGPU_MQD_SIZE_ALIGN(mm->mqd_size) * + NUM_XCC(mm->dev->xcc_mask)); + memcpy(m, mqd_src, sizeof(*m)); + + /* Update MQD base address to the newly allocated location */ + m->cp_mqd_base_addr_lo = lower_32_bits(addr); + m->cp_mqd_base_addr_hi = upper_32_bits(addr); + + m->cp_hqd_pq_doorbell_control &= + ~CP_HQD_PQ_DOORBELL_CONTROL__DOORBELL_OFFSET_MASK; + m->cp_hqd_pq_doorbell_control |= + qp->doorbell_off << CP_HQD_PQ_DOORBELL_CONTROL__DOORBELL_OFFSET__SHIFT; + pr_debug("cp_hqd_pq_doorbell_control 0x%x\n", m->cp_hqd_pq_doorbell_control); + + *mqd = m; + if (gart_addr) + *gart_addr = addr; + + qp->is_active = 0; +} + +static void restore_mqd_sdma_v12_1(struct mqd_manager *mm, void **mqd, + struct kfd_mem_obj *mqd_mem_obj, uint64_t *gart_addr, + struct queue_properties *qp, + const void *mqd_src, + const void *ctl_stack_src, + const u32 ctl_stack_size) +{ + u64 addr; + struct v12_sdma_mqd *m; + + m = (struct v12_sdma_mqd *)mqd_mem_obj->cpu_ptr; + addr = mqd_mem_obj->gpu_addr; + + memset(m, 0, AMDGPU_MQD_SIZE_ALIGN(mm->mqd_size)); + memcpy(m, mqd_src, sizeof(*m)); + + m->sdmax_rlcx_doorbell_offset = + qp->doorbell_off << SDMA0_SDMA_QUEUE0_DOORBELL_OFFSET__OFFSET__SHIFT; + + *mqd = m; + if (gart_addr) + *gart_addr = addr; + + qp->is_active = 0; +} + struct mqd_manager *mqd_manager_init_v12_1(enum KFD_MQD_TYPE type, struct kfd_node *dev) { @@ -668,6 +734,7 @@ struct mqd_manager *mqd_manager_init_v12_1(enum KFD_MQD_TYPE type, mqd->mqd_size = sizeof(struct v12_1_compute_mqd); mqd->get_wave_state = get_wave_state_v12_1; mqd->mqd_stride = kfd_mqd_stride; + mqd->restore_mqd = restore_mqd_v12_1; #if defined(CONFIG_DEBUG_FS) mqd->debugfs_show_mqd = debugfs_show_mqd; #endif @@ -714,6 +781,7 @@ struct mqd_manager *mqd_manager_init_v12_1(enum KFD_MQD_TYPE type, mqd->is_occupied = kfd_is_occupied_sdma; mqd->mqd_size = sizeof(struct v12_sdma_mqd); mqd->mqd_stride = kfd_mqd_stride; + mqd->restore_mqd = restore_mqd_sdma_v12_1; #if defined(CONFIG_DEBUG_FS) mqd->debugfs_show_mqd = debugfs_show_mqd_sdma; #endif From 636139603b99d2e3a18a46cf3f8d39313ce8042e Mon Sep 17 00:00:00 2001 From: Mike Lothian Date: Sat, 12 Sep 2026 00:29:08 +0100 Subject: [PATCH 299/592] drm/amdgpu: hold a runtime PM reference for P2P dma-buf attachments MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit amdgpu_dma_buf_map() adds VRAM to the allowed domains for a peer2peer attachment. GTT is only a fallback placement when VRAM is preferred, so ttm_bo_validate() migrates the buffer from GTT into VRAM. While the exporting device is runtime suspended its SDMA rings are down and the move fails: amdgpu: Move buffer fallback to memcpy unavailable An importer on a second GPU reaches this holding no runtime PM reference on the exporter, e.g. a compositor on the APU submitting a frame that references a buffer exported by an idle dGPU: amdgpu_cs_ioctl -> amdgpu_cs_parser_bos -> amdgpu_cs_bo_validate -> ttm_bo_validate -> amdgpu_bo_move -> dma_buf_map_attachment -> amdgpu_dma_buf_map -> ttm_bo_validate -> amdgpu_bo_move Pinning a dma-buf into VRAM has the same requirement, which commit 030631e97b20 ("drm/amdgpu: revert "take runtime pm reference when we attach a buffer" v2") called out as the one case that would need the reference back. Take it in attach and drop it in detach. pm_runtime_get_if_active() never resumes the device, so it cannot deadlock against the reservation taken during resume, which is why the old pm_runtime_get_sync() had to go. If the device is not active, clear peer2peer instead: the buffer then stays in GTT, which remains accessible while the GPU is powered down. If runtime PM is disabled, take a plain reference so the put in detach stays balanced. Fixes: 030631e97b20 ("drm/amdgpu: revert "take runtime pm reference when we attach a buffer" v2") Suggested-by: Christian König Reviewed-by: Christian König Signed-off-by: Mike Lothian Assisted-by: Claude:Opus-5 [Claude Code] Signed-off-by: Alex Deucher (cherry picked from commit 062ff15e30a48d14fb7d7558eba84f8dc97197f0) Cc: stable@vger.kernel.org --- drivers/gpu/drm/amd/amdgpu/amdgpu_dma_buf.c | 43 ++++++++++++++++++++- 1 file changed, 42 insertions(+), 1 deletion(-) diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_dma_buf.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_dma_buf.c index b33c300e26e2..9adf3eed8822 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_dma_buf.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_dma_buf.c @@ -43,6 +43,7 @@ #include #include #include +#include static const struct dma_buf_attach_ops amdgpu_dma_buf_attach_ops; @@ -100,15 +101,54 @@ static int amdgpu_dma_buf_attach(struct dma_buf *dmabuf, pci_p2pdma_distance(adev->pdev, attach->dev, false) < 0) attach->peer2peer = false; + /* + * Only allow P2P while the exporter is active, and keep it active + * until detach. With runtime PM disabled take a plain reference so + * the put in detach stays balanced. + */ + if (attach->peer2peer) { + struct device *dev = adev_to_drm(adev)->dev; + int ret = pm_runtime_get_if_active(dev); + + if (!ret) + attach->peer2peer = false; + else if (ret < 0) + pm_runtime_get_noresume(dev); + } + r = dma_resv_lock(bo->tbo.base.resv, NULL); if (r) - return r; + goto err_pm_put; amdgpu_vm_bo_update_shared(bo); dma_resv_unlock(bo->tbo.base.resv); return 0; + +err_pm_put: + if (attach->peer2peer) + pm_runtime_put_autosuspend(adev_to_drm(adev)->dev); + return r; +} + +/** + * amdgpu_dma_buf_detach - &dma_buf_ops.detach implementation + * + * @dmabuf: DMA-buf where we remove the attachment from + * @attach: the attachment to remove + * + * Drop the runtime PM reference taken in amdgpu_dma_buf_attach(). + */ +static void amdgpu_dma_buf_detach(struct dma_buf *dmabuf, + struct dma_buf_attachment *attach) +{ + struct drm_gem_object *obj = dmabuf->priv; + struct amdgpu_bo *bo = gem_to_amdgpu_bo(obj); + struct amdgpu_device *adev = amdgpu_ttm_adev(bo->tbo.bdev); + + if (attach->peer2peer) + pm_runtime_put_autosuspend(adev_to_drm(adev)->dev); } /** @@ -350,6 +390,7 @@ static void amdgpu_dma_buf_vunmap(struct dma_buf *dma_buf, struct iosys_map *map const struct dma_buf_ops amdgpu_dmabuf_ops = { .attach = amdgpu_dma_buf_attach, + .detach = amdgpu_dma_buf_detach, .pin = amdgpu_dma_buf_pin, .unpin = amdgpu_dma_buf_unpin, .map_dma_buf = amdgpu_dma_buf_map, From 723d4dc628d764b19cf9efca14b82cca5ff020c9 Mon Sep 17 00:00:00 2001 From: Dmitriy Chumachenko Date: Mon, 14 Sep 2026 17:33:03 +0300 Subject: [PATCH 300/592] drm/amdgpu: check ras and obj before dereference nbio_v7_9_handle_ras_controller_intr_no_bifring() dereferences ras and obj without checking either for NULL. Both amdgpu_ras_get_context() and amdgpu_ras_find_obj() can return NULL, e.g. during the window between adev->nbio.ras being set (early in amdgpu_ras_init(), by design, to enable the fatal-error interrupt as soon as possible) and the PCIE_BIF ras object actually being created in RAS late_init. Any interrupt in that window crashes in hard-IRQ context. This is analogous to commit d190b459b2a4 ("drm/amdgpu: the warning dereferencing obj for nbio_v7_4"), which fixed the same issue in the nbio_v7_4 handler. Found by Linux Verification Center (linuxtesting.org) with SVACE. Fixes: 7692e1ee2446 ("drm/amdgpu: add RAS fatal error handler for NBIO v7.9") Reviewed-by: Tao Zhou Signed-off-by: Dmitriy Chumachenko Signed-off-by: Alex Deucher (cherry picked from commit c7071767a50a32ed727cf800ac84372429e3b4b3) --- drivers/gpu/drm/amd/amdgpu/nbio_v7_9.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/drivers/gpu/drm/amd/amdgpu/nbio_v7_9.c b/drivers/gpu/drm/amd/amdgpu/nbio_v7_9.c index bdfd2917e3ca..def02993b7cf 100644 --- a/drivers/gpu/drm/amd/amdgpu/nbio_v7_9.c +++ b/drivers/gpu/drm/amd/amdgpu/nbio_v7_9.c @@ -535,7 +535,7 @@ static void nbio_v7_9_handle_ras_controller_intr_no_bifring(struct amdgpu_device RAS_CNTLR_INTERRUPT_CLEAR, 1); WREG32_SOC15(NBIO, 0, regBIF_BX0_BIF_DOORBELL_INT_CNTL, bif_doorbell_intr_cntl); - if (!ras->disable_ras_err_cnt_harvest) { + if (ras && !ras->disable_ras_err_cnt_harvest && obj) { /* * clear error status after ras_controller_intr * according to hw team and count ue number From 04de4007d32385b8b6a5dd72bff3146dfdc592c3 Mon Sep 17 00:00:00 2001 From: Mario Limonciello Date: Tue, 15 Sep 2026 10:59:45 -0500 Subject: [PATCH 301/592] drm/amdgpu: Fix GPU PCIe link capability reporting MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit Commit eb53125a7ad9 ("drm/amd: Add dedicated helper for amdgpu_device_find_parent()") made amdgpu_device_gpu_bandwidth() query the first device outside the dGPU. That is the host side of the physical link, not the GPU side. As a result, the ASIC and platform capability masks can both be based on the host port. drm_amdgpu_info_device then exposes the host capabilities to userspace, such as Gen5 x16 for a Gen4 x8 GPU. Cache both ends of the physical link during device initialization. Use link_dev for the GPU capability and link_partner for the platform capability and _PR3 detection. Reported-by: "Marek Olšák" Closes: https://lore.kernel.org/amd-gfx/CAAxE2A4VhsAzzO1QjBjUg+NgnbD04ZzMyN6xsUJxjKJHH6hxiw@mail.gmail.com/ Suggested-by: Lijo Lazar Fixes: eb53125a7ad9 ("drm/amd: Add dedicated helper for amdgpu_device_find_parent()") Reviewed-by: Alex Deucher Signed-off-by: Mario Limonciello Signed-off-by: Alex Deucher (cherry picked from commit 7ea6a47224e2c6e89a3a682d7fbaace4817a55aa) Cc: stable@vger.kernel.org --- drivers/gpu/drm/amd/amdgpu/amdgpu.h | 3 ++ drivers/gpu/drm/amd/amdgpu/amdgpu_device.c | 48 ++++++++-------------- 2 files changed, 20 insertions(+), 31 deletions(-) diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu.h b/drivers/gpu/drm/amd/amdgpu/amdgpu.h index 7974f9b7944f..a9c6f5d4a639 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu.h +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu.h @@ -621,6 +621,9 @@ enum amdgpu_enforce_isolation_mode { struct amdgpu_device { struct device *dev; struct pci_dev *pdev; + /* The two ends of the physical PCIe link outside the device. */ + struct pci_dev *link_dev; + struct pci_dev *link_partner; struct drm_device ddev; #ifdef CONFIG_DRM_AMD_ACP diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_device.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_device.c index 104d1d2cbad9..933804349dbf 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_device.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_device.c @@ -1954,18 +1954,17 @@ static void amdgpu_uid_fini(struct amdgpu_device *adev) adev->uid_info = NULL; } -static struct pci_dev *amdgpu_device_find_parent(struct amdgpu_device *adev) +static void amdgpu_device_init_pcie_links(struct amdgpu_device *adev) { - struct pci_dev *parent = adev->pdev; + adev->link_dev = adev->pdev; + adev->link_partner = pci_upstream_bridge(adev->link_dev); - /* skip upstream/downstream switches internal to dGPU */ - while ((parent = pci_upstream_bridge(parent))) { - if (parent->vendor == PCI_VENDOR_ID_ATI) - continue; - break; + /* Skip upstream/downstream switches internal to the dGPU. */ + while (adev->link_partner && + adev->link_partner->vendor == PCI_VENDOR_ID_ATI) { + adev->link_dev = adev->link_partner; + adev->link_partner = pci_upstream_bridge(adev->link_dev); } - - return parent; } /** @@ -1981,7 +1980,6 @@ static struct pci_dev *amdgpu_device_find_parent(struct amdgpu_device *adev) static int amdgpu_device_ip_early_init(struct amdgpu_device *adev) { struct amdgpu_ip_block *ip_block; - struct pci_dev *parent; bool total, skip_bios, early_full_gpu_access = false; uint32_t bios_flags; int i, r; @@ -2077,10 +2075,9 @@ static int amdgpu_device_ip_early_init(struct amdgpu_device *adev) !dev_is_removable(&adev->pdev->dev)) adev->flags |= AMD_IS_PX; - if (!(adev->flags & AMD_IS_APU)) { - parent = amdgpu_device_find_parent(adev); - adev->has_pr3 = parent ? pci_pr3_present(parent) : false; - } + if (!(adev->flags & AMD_IS_APU)) + adev->has_pr3 = adev->link_partner && + pci_pr3_present(adev->link_partner); adev->pm.pp_feature = amdgpu_pp_feature_mask; if (amdgpu_sriov_vf(adev) || sched_policy == KFD_SCHED_POLICY_NO_HWS) @@ -3776,6 +3773,7 @@ int amdgpu_device_init(struct amdgpu_device *adev, adev->shutdown = false; adev->flags = flags; + amdgpu_device_init_pcie_links(adev); if (amdgpu_force_asic_type >= 0 && amdgpu_force_asic_type < CHIP_LAST) adev->asic_type = amdgpu_force_asic_type; @@ -5872,11 +5870,9 @@ static void amdgpu_device_partner_bandwidth(struct amdgpu_device *adev, *width = PCIE_LNK_WIDTH_UNKNOWN; if (amdgpu_device_pcie_dynamic_switching_supported(adev)) { - struct pci_dev *parent = amdgpu_device_find_parent(adev); - - if (parent) { - *speed = pcie_get_speed_cap(parent); - *width = pcie_get_width_cap(parent); + if (adev->link_partner) { + *speed = pcie_get_speed_cap(adev->link_partner); + *width = pcie_get_width_cap(adev->link_partner); } } else { /* use the current speeds rather than max if switching is not supported */ @@ -5898,21 +5894,11 @@ static void amdgpu_device_gpu_bandwidth(struct amdgpu_device *adev, enum pci_bus_speed *speed, enum pcie_link_width *width) { - struct pci_dev *parent = adev->pdev; - if (!speed || !width) return; - /* use the device itself */ - *speed = pcie_get_speed_cap(adev->pdev); - *width = pcie_get_width_cap(adev->pdev); - - /* use the link outside the device */ - parent = amdgpu_device_find_parent(adev); - if (parent) { - *speed = pcie_get_speed_cap(parent); - *width = pcie_get_width_cap(parent); - } + *speed = pcie_get_speed_cap(adev->link_dev); + *width = pcie_get_width_cap(adev->link_dev); } /** From c883d0a132d430ef7ebb23fd94323be94d0fbdb8 Mon Sep 17 00:00:00 2001 From: David Francis Date: Wed, 5 Aug 2026 09:51:35 -0400 Subject: [PATCH 302/592] drm/amdkfd: Avoid integer underflow with ffs in EOP ring size calc The low 6 bits of cp_hqd_eop_control store the base-2 logarithm of the EOP ring size. This was calculated as ffs(q->eop_ring_buffer_size / sizeof(unsigned int)) - 1 - 1 But ffs can in theory return 1 or 0, so this could underflow (although in practice the ring buffer size cannot be less than 4096). Change this to ffs(q->eop_ring_buffer_size / sizeof(unsigned int) / 4) using properties of logarithms. Reviewed-by: Kent Russell Signed-off-by: David Francis Signed-off-by: Alex Deucher (cherry picked from commit 4f18c56630383c14bfc6b2d65f88f2f895d2121a) Cc: stable@vger.kernel.org --- drivers/gpu/drm/amd/amdkfd/kfd_mqd_manager_v10.c | 2 +- drivers/gpu/drm/amd/amdkfd/kfd_mqd_manager_v11.c | 2 +- drivers/gpu/drm/amd/amdkfd/kfd_mqd_manager_v12.c | 2 +- drivers/gpu/drm/amd/amdkfd/kfd_mqd_manager_v12_1.c | 2 +- 4 files changed, 4 insertions(+), 4 deletions(-) diff --git a/drivers/gpu/drm/amd/amdkfd/kfd_mqd_manager_v10.c b/drivers/gpu/drm/amd/amdkfd/kfd_mqd_manager_v10.c index e034da638c07..4f8a8a1a6186 100644 --- a/drivers/gpu/drm/amd/amdkfd/kfd_mqd_manager_v10.c +++ b/drivers/gpu/drm/amd/amdkfd/kfd_mqd_manager_v10.c @@ -204,7 +204,7 @@ static void update_mqd(struct mqd_manager *mm, void *mqd, * is safe, giving a maximum field value of 0xA. */ m->cp_hqd_eop_control = q->eop_ring_buffer_size ? min(0xA, - ffs(q->eop_ring_buffer_size / sizeof(unsigned int)) - 1 - 1) : 0; + ffs(q->eop_ring_buffer_size / sizeof(unsigned int) / 4)) : 0; m->cp_hqd_eop_base_addr_lo = lower_32_bits(q->eop_ring_buffer_address >> 8); m->cp_hqd_eop_base_addr_hi = diff --git a/drivers/gpu/drm/amd/amdkfd/kfd_mqd_manager_v11.c b/drivers/gpu/drm/amd/amdkfd/kfd_mqd_manager_v11.c index 350fcbbba4b2..bf015dc5b868 100644 --- a/drivers/gpu/drm/amd/amdkfd/kfd_mqd_manager_v11.c +++ b/drivers/gpu/drm/amd/amdkfd/kfd_mqd_manager_v11.c @@ -242,7 +242,7 @@ static void update_mqd(struct mqd_manager *mm, void *mqd, * is safe, giving a maximum field value of 0xA. */ m->cp_hqd_eop_control = q->eop_ring_buffer_size ? min(0xA, - ffs(q->eop_ring_buffer_size / sizeof(unsigned int)) - 1 - 1) : 0; + ffs(q->eop_ring_buffer_size / sizeof(unsigned int) / 4)) : 0; m->cp_hqd_eop_base_addr_lo = lower_32_bits(q->eop_ring_buffer_address >> 8); m->cp_hqd_eop_base_addr_hi = diff --git a/drivers/gpu/drm/amd/amdkfd/kfd_mqd_manager_v12.c b/drivers/gpu/drm/amd/amdkfd/kfd_mqd_manager_v12.c index 63f25a60baa5..6ea09b031caf 100644 --- a/drivers/gpu/drm/amd/amdkfd/kfd_mqd_manager_v12.c +++ b/drivers/gpu/drm/amd/amdkfd/kfd_mqd_manager_v12.c @@ -217,7 +217,7 @@ static void update_mqd(struct mqd_manager *mm, void *mqd, * is safe, giving a maximum field value of 0xA. */ m->cp_hqd_eop_control = q->eop_ring_buffer_size ? min(0xA, - ffs(q->eop_ring_buffer_size / sizeof(unsigned int)) - 1 - 1) : 0; + ffs(q->eop_ring_buffer_size / sizeof(unsigned int) / 4)) : 0; m->cp_hqd_eop_base_addr_lo = lower_32_bits(q->eop_ring_buffer_address >> 8); m->cp_hqd_eop_base_addr_hi = diff --git a/drivers/gpu/drm/amd/amdkfd/kfd_mqd_manager_v12_1.c b/drivers/gpu/drm/amd/amdkfd/kfd_mqd_manager_v12_1.c index 708bbb08c089..c709db0210ce 100644 --- a/drivers/gpu/drm/amd/amdkfd/kfd_mqd_manager_v12_1.c +++ b/drivers/gpu/drm/amd/amdkfd/kfd_mqd_manager_v12_1.c @@ -295,7 +295,7 @@ static void update_mqd(struct mqd_manager *mm, void *mqd, * is safe, giving a maximum field value of 0xA. */ m->cp_hqd_eop_control = q->eop_ring_buffer_size ? min(0xA, - ffs(q->eop_ring_buffer_size / sizeof(unsigned int)) - 1 - 1) : 0; + ffs(q->eop_ring_buffer_size / sizeof(unsigned int) / 4)) : 0; m->cp_hqd_eop_base_addr_lo = lower_32_bits(q->eop_ring_buffer_address >> 8); m->cp_hqd_eop_base_addr_hi = From 8ee521b8b189799e361d4233c5180ba56656d4d4 Mon Sep 17 00:00:00 2001 From: David Francis Date: Wed, 5 Aug 2026 09:16:51 -0400 Subject: [PATCH 303/592] drm/amdkfd: Avoid integer underflow in EOP ring size calculation. The low 6 bits of cp_hqd_eop_control store the base-2 logarithm of the EOP ring size. This was calculated as order_base_2(q->eop_ring_buffer_size / 4) - 1 But order_base_2 can in theory return 0, so this could underflow (although in practice the ring buffer size cannot be less than 4096). Change this to order_base_2(q->eop_ring_buffer_size / 8) using properties of logarithms. Also add to the above comment to make the mathematics more clear. Reviewed-by: Kent Russell Signed-off-by: David Francis Signed-off-by: Alex Deucher (cherry picked from commit f0f43fcf8b2b3a924cad9444340921c96ed5f634) Cc: stable@vger.kernel.org --- drivers/gpu/drm/amd/amdkfd/kfd_mqd_manager_v9.c | 6 +++++- drivers/gpu/drm/amd/amdkfd/kfd_mqd_manager_vi.c | 5 ++++- 2 files changed, 9 insertions(+), 2 deletions(-) diff --git a/drivers/gpu/drm/amd/amdkfd/kfd_mqd_manager_v9.c b/drivers/gpu/drm/amd/amdkfd/kfd_mqd_manager_v9.c index b95720198e28..6e6bc1ec0b64 100644 --- a/drivers/gpu/drm/amd/amdkfd/kfd_mqd_manager_v9.c +++ b/drivers/gpu/drm/amd/amdkfd/kfd_mqd_manager_v9.c @@ -285,6 +285,10 @@ static void update_mqd(struct mqd_manager *mm, void *mqd, 1 << CP_HQD_IB_CONTROL__IB_EXE_DISABLE__SHIFT; /* + * The lowest 6 bits of eop_control store the EOP ring size. If + * their value is X, the ring size is 2^(X + 1) dwords, or + * 2^(X + 3) bytes. + * * HW does not clamp this field correctly. Maximum EOP queue size * is constrained by per-SE EOP done signal count, which is 8-bit. * Limit is 0xFF EOP entries (= 0x7F8 dwords). CP will not submit @@ -296,7 +300,7 @@ static void update_mqd(struct mqd_manager *mm, void *mqd, * */ m->cp_hqd_eop_control = q->eop_ring_buffer_size ? - min(0xA, order_base_2(q->eop_ring_buffer_size / 4) - 1) : 0; + min(0xA, order_base_2(q->eop_ring_buffer_size / 8)) : 0; m->cp_hqd_eop_base_addr_lo = lower_32_bits(q->eop_ring_buffer_address >> 8); diff --git a/drivers/gpu/drm/amd/amdkfd/kfd_mqd_manager_vi.c b/drivers/gpu/drm/amd/amdkfd/kfd_mqd_manager_vi.c index 60b87a500698..029572548c14 100644 --- a/drivers/gpu/drm/amd/amdkfd/kfd_mqd_manager_vi.c +++ b/drivers/gpu/drm/amd/amdkfd/kfd_mqd_manager_vi.c @@ -208,6 +208,9 @@ static void __update_mqd(struct mqd_manager *mm, void *mqd, mtype << CP_HQD_IB_CONTROL__MTYPE__SHIFT; /* + * The lowest 6 bits of eop_control store the EOP ring size. If + * their value is X, the ring size is 2^(X + 1) dwords, or + * 2^(X + 3) bytes. * HW does not clamp this field correctly. Maximum EOP queue size * is constrained by per-SE EOP done signal count, which is 8-bit. * Limit is 0xFF EOP entries (= 0x7F8 dwords). CP will not submit @@ -215,7 +218,7 @@ static void __update_mqd(struct mqd_manager *mm, void *mqd, * is safe, giving a maximum field value of 0xA. */ m->cp_hqd_eop_control |= q->eop_ring_buffer_size ? min(0xA, - order_base_2(q->eop_ring_buffer_size / 4) - 1) : 0; + order_base_2(q->eop_ring_buffer_size / 8)) : 0; m->cp_hqd_eop_base_addr_lo = lower_32_bits(q->eop_ring_buffer_address >> 8); m->cp_hqd_eop_base_addr_hi = From 0d2f4cfa564355fcbbc71498fd8ec09243036109 Mon Sep 17 00:00:00 2001 From: Srinivasan Shanmugam Date: Tue, 15 Sep 2026 14:07:39 +0530 Subject: [PATCH 304/592] drm/amd/display: Fix NULL dereference in dcn50/dcn60 init_hw dc->clk_mgr is checked for NULL earlier in dcn50_init_hw() and dcn60_init_hw(), but dcn50_initialize_min_clocks() and dcn401_initialize_min_clocks() are called without any guard, causing Smatch to report potential NULL dereferences. Guard both call sites with the same pattern used throughout both functions: if (dc->clk_mgr && dc->clk_mgr->funcs) Also fix dcn50_initialize_min_clocks() which calls get_dispclk_from_dentist without checking the function pointer, unlike the dcn401 equivalent which guards that call. Fix kernel-doc in dcn60_hwseq.c by adding missing parameter descriptions for @probe in dcn60_update_probe_status() and @type in is_probe_measurement_type_for_hubbub(). Fixes: 7f7d7ea1fa51 ("drm/amd/display: Add new sources for DCN6") Reported-by: Dan Carpenter Cc: Aurabindo Pillai Cc: Ivan Lipski Cc: Dan Wheeler Cc: Roman Li Cc: Alex Hung Cc: Tom Chung Signed-off-by: Srinivasan Shanmugam Reviewed-by: Alex Hung Signed-off-by: Alex Deucher (cherry picked from commit 325c9a827cdd748e126eafeadaffc556204773d2) --- drivers/gpu/drm/amd/display/dc/hwss/dcn50/dcn50_hwseq.c | 6 ++++-- drivers/gpu/drm/amd/display/dc/hwss/dcn60/dcn60_hwseq.c | 5 ++++- 2 files changed, 8 insertions(+), 3 deletions(-) diff --git a/drivers/gpu/drm/amd/display/dc/hwss/dcn50/dcn50_hwseq.c b/drivers/gpu/drm/amd/display/dc/hwss/dcn50/dcn50_hwseq.c index a7f8fd03faea..e549556b9679 100644 --- a/drivers/gpu/drm/amd/display/dc/hwss/dcn50/dcn50_hwseq.c +++ b/drivers/gpu/drm/amd/display/dc/hwss/dcn50/dcn50_hwseq.c @@ -67,7 +67,8 @@ static void dcn50_initialize_min_clocks(struct dc *dc) * audio corruption. Read current DISPCLK from DENTIST and request the same * freq to ensure that the timing is valid and unchanged. */ - clocks->dispclk_khz = dc->clk_mgr->funcs->get_dispclk_from_dentist(dc->clk_mgr); + if (dc->clk_mgr->funcs->get_dispclk_from_dentist) + clocks->dispclk_khz = dc->clk_mgr->funcs->get_dispclk_from_dentist(dc->clk_mgr); } clocks->ref_dtbclk_khz = dc->clk_mgr->bw_params->clk_table.entries[0].dtbclk_mhz * 1000; clocks->fclk_p_state_change_support = true; @@ -639,7 +640,8 @@ void dcn50_init_hw(struct dc *dc) dc->res_pool->hubbub->funcs->allow_self_refresh_control(dc->res_pool->hubbub, !dc->res_pool->hubbub->ctx->dc->debug.disable_stutter); - dcn50_initialize_min_clocks(dc); + if (dc->clk_mgr && dc->clk_mgr->funcs) + dcn50_initialize_min_clocks(dc); /* On HW init, allow idle optimizations after pipes have been turned off. * diff --git a/drivers/gpu/drm/amd/display/dc/hwss/dcn60/dcn60_hwseq.c b/drivers/gpu/drm/amd/display/dc/hwss/dcn60/dcn60_hwseq.c index 72c2d3ca52f6..61ad6efa7a74 100644 --- a/drivers/gpu/drm/amd/display/dc/hwss/dcn60/dcn60_hwseq.c +++ b/drivers/gpu/drm/amd/display/dc/hwss/dcn60/dcn60_hwseq.c @@ -643,7 +643,8 @@ void dcn60_init_hw(struct dc *dc) dc->res_pool->hubbub->funcs->allow_self_refresh_control(dc->res_pool->hubbub, !dc->res_pool->hubbub->ctx->dc->debug.disable_stutter); - dcn401_initialize_min_clocks(dc); + if (dc->clk_mgr && dc->clk_mgr->funcs) + dcn401_initialize_min_clocks(dc); /* On HW init, allow idle optimizations after pipes have been turned off. * @@ -1001,6 +1002,7 @@ static void dcn60_build_hubbub_perfmon_sequence( /** * dcn60_update_probe_status - Set the valid flag on a latched probe result. * @status: result sink whose u was written by the GET BLS step during execute + * @probe: current probe state used to determine measurement type and validity */ static void dcn60_update_probe_status(struct dc_probe_status *status) { @@ -1024,6 +1026,7 @@ static void dcn60_update_probe_status(struct dc_probe_status *status) /** * is_probe_measurement_type_for_hubbub - Returns true if the probe type is * served by the hubbub perfmon block on DCN60. + * @type: the probe measurement type to classify */ static bool is_probe_measurement_type_for_hubbub(enum dc_probe_type type) { From 7f9caa70aef0950e06d395ca0035831214d88187 Mon Sep 17 00:00:00 2001 From: Mario Limonciello Date: Tue, 15 Sep 2026 12:51:24 -0500 Subject: [PATCH 305/592] drm/amdgpu: Skip KFD mapping clear before initialization amdgpu_amdkfd_clear_kfd_mapping() assumes that a non-NULL kfd_dev has a fully populated node array. This is not true when KFD device initialization fails after probe. For example, kgd2kfd_device_init() sets num_nodes before checking PCIe atomics support. On Polaris systems without the required atomics, it returns before allocating nodes[0], but the kfd_dev remains attached to the amdgpu device. A later GPU reset then dereferences nodes[0]->id. Require the authoritative KFD initialization flag before walking the node array, matching the existing KFD reset and teardown paths. Fixes: 70cadefcc616 ("drm/amdgpu: unmap all user mappings of framebuffer and doorbell before mode1 reset") Closes: https://gitlab.freedesktop.org/drm/amd/-/work_items/5833 Reviewed-by: Alex Deucher Signed-off-by: Mario Limonciello Signed-off-by: Alex Deucher (cherry picked from commit 4ac1835823c47903fbb278bbf474773c46f59edc) Cc: stable@vger.kernel.org --- drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd.c index 816d8817f0b2..054870e9078d 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd.c @@ -330,7 +330,7 @@ void amdgpu_amdkfd_clear_kfd_mapping(struct amdgpu_device *adev) struct kfd_dev *kfd = adev->kfd.dev; unsigned int i; - if (!kfd) + if (!kfd || !kfd->init_complete) return; for (i = 0; i < kfd->num_nodes; i++) { From 5155002b03b24ba3ef91c5c313b8cf0171b24904 Mon Sep 17 00:00:00 2001 From: Chengjun Yao Date: Tue, 8 Sep 2026 10:15:43 +0800 Subject: [PATCH 306/592] drm/amdgpu: fix rmmio iounmap skipped on device removal amdgpu_pci_remove() calls drm_dev_unplug() before fini_sw(), so drm_dev_enter() is already false there and the iounmap() guarded by it is skipped. This .remove path runs on both hot-unplug and plain rmmod, so the register BAR ioremap mapping leaks one instance per unload. Unmap rmmio unconditionally (guard only on non-NULL) and drop the now unused idx. Fixes: 62d5f9f7110a ("drm/amdgpu: Unmap MMIO mappings when device is not unplugged") Signed-off-by: Chengjun Yao Reviewed-by: Asad Kamal Signed-off-by: Alex Deucher (cherry picked from commit dd6f86a97260e5207d3329ad03aa89fdad61b1e6) Cc: stable@vger.kernel.org --- drivers/gpu/drm/amd/amdgpu/amdgpu_device.c | 6 ++---- 1 file changed, 2 insertions(+), 4 deletions(-) diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_device.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_device.c index 933804349dbf..9269e780feb7 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_device.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_device.c @@ -4335,7 +4335,7 @@ void amdgpu_device_fini_hw(struct amdgpu_device *adev) void amdgpu_device_fini_sw(struct amdgpu_device *adev) { - int i, idx; + int i; bool px; amdgpu_device_ip_fini(adev); @@ -4377,11 +4377,9 @@ void amdgpu_device_fini_sw(struct amdgpu_device *adev) if ((adev->pdev->class >> 8) == PCI_CLASS_DISPLAY_VGA) vga_client_unregister(adev->pdev); - if (drm_dev_enter(adev_to_drm(adev), &idx)) { - + if (adev->rmmio) { iounmap(adev->rmmio); adev->rmmio = NULL; - drm_dev_exit(idx); } if (IS_ENABLED(CONFIG_PERF_EVENTS)) From 2ac2fe765ef475f409616ac0b57c4a3922749b0f Mon Sep 17 00:00:00 2001 From: Francis Marlou Pacaro Date: Fri, 4 Sep 2026 08:20:21 +0800 Subject: [PATCH 307/592] drm/amd/display: fix MALL hysteresis timer underflow at high refresh rates dcn30_apply_idle_power_optimizations() derives the MALL frame cache hysteresis timer with tmr_delay = (uint32_t)(div_u64(..., denom) - 64LL); div_u64() returns a u64, so when the quotient is smaller than 64 the subtraction wraps instead of going negative and tmr_delay ends up huge. The loop that follows tries to squeeze it into the 6 bit register field by doubling denom, but that only makes the quotient smaller, so tmr_delay can never converge. tmr_scale is bumped past 3 and the function gives up with /* Delay exceeds range of hysteresis timer */ ASSERT(false); even though the requested delay is too *short* to encode, not too long. With mall_additional_timer_percent left at its default of 0, the quotient drops below 64 once the refresh rate used for the calculation goes above ~243 Hz. Every DCN 3.0 display above that loses MALL static screen entirely and splats a WARN once per boot. Reproduced on Navi 23 (RX 6600) driving 1920x1080, resetting /sys/kernel/debug/clear_warn_once between modes: refresh MALL ASSERT 144 Hz enabled no 240 Hz enabled no 280 Hz skipped yes 360 Hz skipped yes Commit 3bb68cec4db8 ("drm/amd/display: Add Overflow check to skip MALL") already covered the other end of the range, where a large stutter period makes the delay too long to encode. Cover the short end by clamping to 0, which selects the shortest hysteresis the register can express, 65.28us * 64 = ~4.18ms. That is marginally longer than what the formula asks for at these refresh rates, and erring long is the safe direction: it only delays MALL entry, it can never enter early. The numerator does not change between iterations, only denom does, so compute it once and keep both call sites inside 100 columns. The genuinely out of range case at very low refresh rates still reaches the ASSERT, which is where it belongs. Fixes: 52f2e83e2fe5 ("drm/amdgpu/display: add MALL support (v2)") Signed-off-by: Francis Marlou Pacaro Reviewed-by: Leo Li Signed-off-by: Alex Deucher (cherry picked from commit 387550e53e1405f1f960b62b22f8783db17c8e1d) --- .../gpu/drm/amd/display/dc/hwss/dcn30/dcn30_hwseq.c | 13 +++++++------ 1 file changed, 7 insertions(+), 6 deletions(-) diff --git a/drivers/gpu/drm/amd/display/dc/hwss/dcn30/dcn30_hwseq.c b/drivers/gpu/drm/amd/display/dc/hwss/dcn30/dcn30_hwseq.c index cb163902e12e..d669b47af120 100644 --- a/drivers/gpu/drm/amd/display/dc/hwss/dcn30/dcn30_hwseq.c +++ b/drivers/gpu/drm/amd/display/dc/hwss/dcn30/dcn30_hwseq.c @@ -1064,10 +1064,12 @@ bool dcn30_apply_idle_power_optimizations(struct dc *dc, bool enable) */ unsigned int denom = refresh_hz * 6528; unsigned int stutter_period = dc->current_state->perf_params.stutter_period_us; + uint64_t num = (1000000LL + 2 * stutter_period * refresh_hz) * + (100LL + dc->debug.mall_additional_timer_percent); + uint64_t tmr_ticks; - tmr_delay = (uint32_t)(div_u64(((1000000LL + 2 * stutter_period * refresh_hz) * - (100LL + dc->debug.mall_additional_timer_percent) + denom - 1), - denom) - 64LL); + tmr_ticks = div_u64(num + denom - 1, denom); + tmr_delay = tmr_ticks > 64 ? (uint32_t)(tmr_ticks - 64) : 0; /* In some cases the stutter period is really big (tiny modes) in these * cases MALL cant be enabled, So skip these cases to avoid a ASSERT() @@ -1089,9 +1091,8 @@ bool dcn30_apply_idle_power_optimizations(struct dc *dc, bool enable) } denom *= 2; - tmr_delay = (uint32_t)(div_u64(((1000000LL + 2 * stutter_period * refresh_hz) * - (100LL + dc->debug.mall_additional_timer_percent) + denom - 1), - denom) - 64LL); + tmr_ticks = div_u64(num + denom - 1, denom); + tmr_delay = tmr_ticks > 64 ? (uint32_t)(tmr_ticks - 64) : 0; } /* Copy HW cursor */ From 3565893cc72cdf6b795cf6a33e7ff9605322334d Mon Sep 17 00:00:00 2001 From: Guanghui Yang <3497809730@qq.com> Date: Wed, 16 Sep 2026 05:16:38 +0000 Subject: [PATCH 308/592] btrfs: clear free space tree creation state on rebuild failure btrfs_rebuild_free_space_tree() sets BTRFS_FS_CREATING_FREE_SPACE_TREE before rebuilding the free space tree. Several error paths return without clearing this flag. The transaction restart failure path can leave the flag set on a live filesystem, causing delayed reference processing to be skipped. Clear it on all free space tree rebuild failure paths. Keep BTRFS_FS_FREE_SPACE_TREE_UNTRUSTED set, since a failed rebuild leaves the free space tree untrusted. Callers must fall back to extent-tree caching. Fixes: 882af9f13e83 ("btrfs: handle free space tree rebuild in multiple transactions") CC: stable@vger.kernel.org # 6.14+ Assisted-by: LLM Reviewed-by: Boris Burkov Reviewed-by: Qu Wenruo Signed-off-by: Guanghui Yang <3497809730@qq.com> Signed-off-by: David Sterba --- fs/btrfs/free-space-tree.c | 14 ++++++++++---- 1 file changed, 10 insertions(+), 4 deletions(-) diff --git a/fs/btrfs/free-space-tree.c b/fs/btrfs/free-space-tree.c index 1b3d82ae3de8..b7a4a6ade30f 100644 --- a/fs/btrfs/free-space-tree.c +++ b/fs/btrfs/free-space-tree.c @@ -1353,7 +1353,7 @@ int btrfs_rebuild_free_space_tree(struct btrfs_fs_info *fs_info) if (unlikely(ret)) { btrfs_abort_transaction(trans, ret); btrfs_end_transaction(trans); - return ret; + goto out_clear; } node = rb_first_cached(&fs_info->block_group_cache_tree); @@ -1371,14 +1371,16 @@ int btrfs_rebuild_free_space_tree(struct btrfs_fs_info *fs_info) if (unlikely(ret)) { btrfs_abort_transaction(trans, ret); btrfs_end_transaction(trans); - return ret; + goto out_clear; } next: if (btrfs_should_end_transaction(trans)) { btrfs_end_transaction(trans); trans = btrfs_start_transaction(free_space_root, 1); - if (IS_ERR(trans)) - return PTR_ERR(trans); + if (IS_ERR(trans)) { + ret = PTR_ERR(trans); + goto out_clear; + } } node = rb_next(node); } @@ -1390,6 +1392,10 @@ int btrfs_rebuild_free_space_tree(struct btrfs_fs_info *fs_info) ret = btrfs_commit_transaction(trans); clear_bit(BTRFS_FS_FREE_SPACE_TREE_UNTRUSTED, &fs_info->flags); return ret; + +out_clear: + clear_bit(BTRFS_FS_CREATING_FREE_SPACE_TREE, &fs_info->flags); + return ret; } static int __add_block_group_free_space(struct btrfs_trans_handle *trans, From 97fcd34aa9fd73cefe3120ac9a82ca9d7763922f Mon Sep 17 00:00:00 2001 From: Filipe Manana Date: Wed, 16 Sep 2026 15:43:41 +0100 Subject: [PATCH 309/592] btrfs: abort transaction on failure to update inode for hole punching and reflinking If we fail to update the inode we error out without aborting the transaction, which can result in a persistent inconsistency if after the failure the transaction is committed, as we have dropped file extent items from a range and either punched a hole or insert a new file extent item for that range (for reflinks). So add the missing transaction abort. Fixes: 2aaa66558172 ("Btrfs: add hole punching") Reviewed-by: Qu Wenruo Signed-off-by: Filipe Manana Signed-off-by: David Sterba --- fs/btrfs/file.c | 4 +++- 1 file changed, 3 insertions(+), 1 deletion(-) diff --git a/fs/btrfs/file.c b/fs/btrfs/file.c index 20e15dc30bfb..f978c6524aa0 100644 --- a/fs/btrfs/file.c +++ b/fs/btrfs/file.c @@ -2509,8 +2509,10 @@ int btrfs_replace_file_extents(struct btrfs_inode *inode, inode_set_ctime_current(&inode->vfs_inode)); ret = btrfs_update_inode(trans, inode); - if (ret) + if (unlikely(ret)) { + btrfs_abort_transaction(trans, ret); break; + } btrfs_end_transaction(trans); btrfs_btree_balance_dirty(fs_info); From aeab4c62875748ecfd390a47ac1d91ea7c9a6abb Mon Sep 17 00:00:00 2001 From: Filipe Manana Date: Wed, 16 Sep 2026 16:49:37 +0100 Subject: [PATCH 310/592] btrfs: check if there is space for chunk item when validating sys chunk array We checked if have enough remaining space for a key before dereferencing a key, but we then dereference a chunk item, to get the number of stripes, without checking if there is space for the item. So add a check to see if there is enough space for a chunk item before dereferencing the item to extract the stripe count. Fixes: 2a9bb78cfd36 ("btrfs: validate system chunk array at btrfs_validate_super()") Reviewed-by: Qu Wenruo Signed-off-by: Filipe Manana Reviewed-by: David Sterba Signed-off-by: David Sterba --- fs/btrfs/disk-io.c | 4 ++++ 1 file changed, 4 insertions(+) diff --git a/fs/btrfs/disk-io.c b/fs/btrfs/disk-io.c index 819727460bcf..dc7ad92876c0 100644 --- a/fs/btrfs/disk-io.c +++ b/fs/btrfs/disk-io.c @@ -2357,6 +2357,10 @@ static int validate_sys_chunk_array(const struct btrfs_fs_info *fs_info, key.type, cur); return -EUCLEAN; } + + if (unlikely(cur + sizeof(*chunk) > sys_array_size)) + goto short_read; + chunk = (struct btrfs_chunk *)(sb->sys_chunk_array + cur); num_stripes = btrfs_stack_chunk_num_stripes(chunk); if (unlikely(cur + btrfs_chunk_item_size(num_stripes) > sys_array_size)) From b797b52e88a66598a972111b02ef13edd8d03ed2 Mon Sep 17 00:00:00 2001 From: Qu Wenruo Date: Sat, 12 Sep 2026 18:12:06 +0930 Subject: [PATCH 311/592] btrfs: add "/dev/root" exception for device path update [BEHAVIOR CHANGE] Since commit 108cc8733989 ("btrfs: fix a lockdep caused by path resolution during device scan"), users with btrfs rootfs but without an initramfs are complaining that grub2 can no longer detect the rootfs device: /usr/sbin/grub-probe: error: cannot find a device for / (is /dev mounted?). [CAUSE] Although using btrfs without an initramfs is not recommended (if a new device is added to the rootfs, the system can no longer boot, as there is no way to register all devices), there is still a minority of users doing this. If there is no initramfs but the rootfs is on a block-device-based filesystem, the kernel boot sequence initializes a minimal ramfs/tmpfs, creates "/dev/root" with the proper device number for the rootfs, and then invokes mount using "/dev/root". That's why the end user will get the mount output: /dev/root on / rw To be honest, this is a user space problem: no one should trust the device path shown in mount, only the device number. E.g. one can even use "/proc/self/fd/*" to mount an fs, and that proc path will be registered, and no one else can mount that fs using that path. Before commit 108cc8733989 ("btrfs: fix a lockdep caused by path resolution during device scan"), btrfs had an internal path lookup workaround to address such weird paths, it works by checking if the existing device path can still resolve to the device number. But that path resolution is deadlock prone, thus it's replaced by a simple devt check. This works fine in most cases, as a btrfs device is registered by udev at boot time, thus all paths are sane. However this will not work for systems without an initramfs, causing the unreachable "/dev/root" path to exist forever without a way to rename it. [WORKAROUND] Add an exception to the device path rename requirement. If the device has the name "/dev/root", we know it's booted without an initramfs, and only for that case we allow device path update. And if someone intentionally created "/dev/root" after boot, the existing devt checks will reject that weird name as usual. This should satisfy the minority of users, and still keep most of the existing guards preventing unexpected/unnecessary device path updates. Fixes: 108cc8733989 ("btrfs: fix a lockdep caused by path resolution during device scan") Link: https://lore.kernel.org/linux-btrfs/CAKLYgeL7nrA4nXcewdv9Fqg_s=3GS=vmoypnEiZBKQ7rySZFuQ@mail.gmail.com/ Link: https://lore.kernel.org/linux-btrfs/dfbe1e27-dab8-4d55-8cf3-0b28eeac5df4@gmail.com/ Signed-off-by: Qu Wenruo Reviewed-by: David Sterba Signed-off-by: David Sterba --- fs/btrfs/volumes.c | 33 ++++++++++++++++++++++++++++++++- 1 file changed, 32 insertions(+), 1 deletion(-) diff --git a/fs/btrfs/volumes.c b/fs/btrfs/volumes.c index 427aa8e24fc3..7bbc8e745e6b 100644 --- a/fs/btrfs/volumes.c +++ b/fs/btrfs/volumes.c @@ -740,6 +740,36 @@ const u8 *btrfs_sb_fsid_ptr(const struct btrfs_super_block *sb) return has_metadata_uuid ? sb->metadata_uuid : sb->fsid; } +static bool should_rename_device(const struct btrfs_device *dev) +{ + bool ret; + const char *old_name; + + rcu_read_lock(); + old_name = rcu_dereference(dev->name); + /* + * For systems booted without an initramfs, the rootfs has the device + * name "/dev/root". + * + * Although using btrfs without an initramfs is not recommended (if a + * new device is added to the rootfs, the system can no longer boot, as + * there is no way to register all devices), there is still a minority + * of users doing this. + * + * And after the system is up, a later device scan on the real block + * device file will never get this device's name updated, as the + * device->devt is still the same. + * + * Here we add one and only one exception for "/dev/root", to allow the + * device name to be updated even if the new path points to the same + * block device. + */ + ret = (strcmp(old_name, "/dev/root") == 0); + rcu_read_unlock(); + + return ret; +} + /* * Add new device to list of registered devices * @@ -860,7 +890,8 @@ static noinline struct btrfs_device *device_list_add(const char *path, MAJOR(path_devt), MINOR(path_devt), current->comm, task_pid_nr(current)); - } else if (!device->name || device->devt != path_devt) { + } else if (!device->name || device->devt != path_devt || + should_rename_device(device)) { const char *old_name; /* From 72de4807ba84da485dda1a91572d66da9149e95a Mon Sep 17 00:00:00 2001 From: Anand Jain Date: Sun, 13 Sep 2026 02:06:28 +0800 Subject: [PATCH 312/592] btrfs: derive f_fsid with dev_t only when temp_fsid is active Commit c2a74ed0494c ("btrfs: derive f_fsid from on-disk fsid and dev_t") mixed dev_t into f_fsid for all single-device setups to avoid f_fsid collisions with cloned filesystems. However, doing this unconditionally breaks backward compatibility. statfs(2) f_fsid changes after a kernel upgrade, and also can shift across reboots or dev re-attaches as dev_t values change. Fix this by only mixing dev_t when temp_fsid is active. This means for non-temp_fsid setups or the original mount, we use the old method of deriving fsid based on the UUID. So in the case of a cloned Btrfs filesystem, we won't be able to maintain the same fsid across mount recycle if the mount order changes. Reported-by: Dave Hansen Link: https://lore.kernel.org/linux-btrfs/be0c08f5-2f31-40f5-8a3b-f2f58b3e00ff@intel.com Fixes: c2a74ed0494c ("btrfs: derive f_fsid from on-disk fsid and dev_t") CC: stable@vger.kernel.org # 7.2 Signed-off-by: Anand Jain Reviewed-by: David Sterba Signed-off-by: David Sterba --- fs/btrfs/super.c | 8 ++++++-- 1 file changed, 6 insertions(+), 2 deletions(-) diff --git a/fs/btrfs/super.c b/fs/btrfs/super.c index 464129b1b0d4..ddb620ac241b 100644 --- a/fs/btrfs/super.c +++ b/fs/btrfs/super.c @@ -1836,8 +1836,12 @@ static int btrfs_statfs(struct dentry *dentry, struct kstatfs *buf) f_fsid.val[0] ^= btrfs_root_id(BTRFS_I(d_inode(dentry))->root) >> 32; f_fsid.val[1] ^= btrfs_root_id(BTRFS_I(d_inode(dentry))->root); - /* Hash dev_t to avoid f_fsid collision with cloned filesystems. */ - if (fs_info->fs_devices->total_devices == 1) { + /* + * Hash dev_t to avoid f_fsid collisions with cloned filesystems. + * Only do this when a clone is present so the original filesystem + * (mounted first) maintains backward-compatible f_fsid behavior. + */ + if (fs_info->fs_devices->temp_fsid) { __kernel_fsid_t dev_fsid = u64_to_fsid(huge_encode_dev(fs_info->fs_devices->latest_dev->bdev->bd_dev)); From 40c2096961b4e8f48d1fc17406a90c5a36ac0a8d Mon Sep 17 00:00:00 2001 From: Kumar Kartikeya Dwivedi Date: Mon, 14 Sep 2026 15:19:20 +0200 Subject: [PATCH 313/592] bpf: Verify global subprogs in each sleepability context Global subprograms are verified independently with a fresh verifier root. do_check_common() currently seeds that root's in_sleepable state from the program, even though a global subprogram can also run from callbacks whose execution context differs from the program's main entry point. In particular, workqueue and task-work callbacks are sleepable even when the containing program is not. A global subprogram of that program is therefore verified as non-sleepable, making in_rcu_cs() true and allowing loads of RCU-protected kptrs to produce trusted MEM_RCU pointers. The same subprogram can then be called from a sleepable callback without a classic RCU reader. It can retain such a pointer while the object is freed and use it after free. The verifier's execution-context predicates are complementary. A state is sleepable only when in_sleepable is set and no RCU, preemption, IRQ, or lock region is active. Each condition which prevents sleeping also provides RCU protection, while in_rcu_cs() treats a non-sleepable state as implicitly protected. Use this relationship to represent a global subprogram caller with only the result of in_sleepable_context(). A protected sleepable caller is normalized to in_sleepable=false at the independent verification root. This both prevents sleepable operations and makes in_rcu_cs() true without copying caller-owned lock state. Track only the contexts in which each global subprogram is actually reached. Verify it once if all reachable calls use the same context, and twice only if both sleepable and non-sleepable calls reach it. Calls found while verifying globals or asynchronous callbacks mark further contexts for checking. Repeat the existing subprogram walk until all called contexts have been verified; unreachable global calls remain unchecked. Accumulate instruction counts over those verification passes. Preserve the total recorded before each pass, since path accounting has already added this pass's synchronous instructions and its root total must also include asynchronous subprograms. This makes an unprotected callback verify the global subprogram as sleepable, turning its RCU-protected kptr load into an untrusted pointer. Protected callers and global subprograms which do not depend on implicit RCU protection remain valid. Fixes: 81f1d7a583fa ("bpf: wq: add bpf_wq_set_callback_impl") Fixes: 38aa7003e369 ("bpf: task work scheduling kfuncs") Reported-by: Nicholas Carlini Suggested-by: Nicholas Carlini Signed-off-by: Kumar Kartikeya Dwivedi Acked-by: Eduard Zingerman Link: https://patch.msgid.link/20260914131923.2544250-2-memxor@gmail.com Signed-off-by: Eduard Zingerman --- include/linux/bpf.h | 5 +-- kernel/bpf/verifier.c | 71 ++++++++++++++++++++++--------------------- 2 files changed, 40 insertions(+), 36 deletions(-) diff --git a/include/linux/bpf.h b/include/linux/bpf.h index e57af902560c..1d2676782d70 100644 --- a/include/linux/bpf.h +++ b/include/linux/bpf.h @@ -1651,8 +1651,9 @@ static inline void bpf_trampoline_set_flags(struct bpf_trampoline *tr, u32 flags struct bpf_func_info_aux { u16 linkage; bool unreliable; - bool called : 1; - bool verified : 1; + /* Indexed by in_sleepable. */ + bool called[2]; + bool verified[2]; }; enum bpf_jit_poke_reason { diff --git a/kernel/bpf/verifier.c b/kernel/bpf/verifier.c index ddba53eaa333..5d7080c260d8 100644 --- a/kernel/bpf/verifier.c +++ b/kernel/bpf/verifier.c @@ -9931,6 +9931,7 @@ static int check_func_call(struct bpf_verifier_env *env, struct bpf_insn *insn, if (err == -EFAULT) return err; if (bpf_subprog_is_global(env, subprog)) { + struct bpf_func_info_aux *sub_aux = subprog_aux(env, subprog); const char *sub_name = bpf_subprog_name(env, subprog); const char *operation; bool returns_void; @@ -9962,11 +9963,10 @@ static int check_func_call(struct bpf_verifier_env *env, struct bpf_insn *insn, if (env->log.level & BPF_LOG_LEVEL) verbose(env, "Func#%d ('%s') is global and assumed valid.\n", subprog, sub_name); + sub_aux->called[in_sleepable_context(env)] = true; returns_void = subprog_returns_void(env, subprog); if (env->subprog_info[subprog].changes_pkt_data) clear_all_pkt_pointers(env); - /* mark global subprog for verifying after main prog */ - subprog_aux(env, subprog)->called = true; if (returns_void) bpf_diag_record_scrub(env, &caller->regs[BPF_REG_0], BPF_DIAG_MOD_CALLER_SAVED); else @@ -10784,11 +10784,7 @@ int bpf_get_helper_proto(struct bpf_verifier_env *env, int func_id, /* Check if we're in a sleepable context. */ static inline bool in_sleepable_context(struct bpf_verifier_env *env) { - return !env->cur_state->active_rcu_locks && - !env->cur_state->active_preempt_locks && - !env->cur_state->active_locks && - !env->cur_state->active_irq_id && - in_sleepable(env); + return !in_rcu_cs(env); } static const char *non_sleepable_context_description(struct bpf_verifier_env *env) @@ -19447,13 +19443,14 @@ static void free_states(struct bpf_verifier_env *env) } } -static int do_check_common(struct bpf_verifier_env *env, int subprog) +static int do_check_common(struct bpf_verifier_env *env, int subprog, bool is_sleepable) { bool pop_log = !(env->log.level & BPF_LOG_LEVEL2); struct bpf_subprog_info *sub = subprog_info(env, subprog); struct bpf_prog_aux *aux = env->prog->aux; struct bpf_verifier_state *state; struct bpf_reg_state *regs; + u32 old_insns_total = sub->insns_total; u32 insn_processed = env->insn_processed; int ret, i; @@ -19466,7 +19463,7 @@ static int do_check_common(struct bpf_verifier_env *env, int subprog) state->curframe = 0; state->speculative = false; state->branches = 1; - state->in_sleepable = env->prog->sleepable; + state->in_sleepable = is_sleepable; state->frame[0] = kzalloc_obj(struct bpf_func_state, GFP_KERNEL_ACCOUNT); if (!state->frame[0]) { kfree(state); @@ -19607,8 +19604,10 @@ static int do_check_common(struct bpf_verifier_env *env, int subprog) * not accounted as callees by account_current_path(). * Accumulate their total counts as total counts of the main or * global subprog hosting the async call. + * Start from the saved total of earlier contexts: adding to the current + * total would count this pass's synchronous paths twice. */ - env->subprog_info[subprog].insns_total = env->insn_processed - insn_processed; + sub->insns_total = old_insns_total + (env->insn_processed - insn_processed); return ret; } @@ -19636,14 +19635,19 @@ static int do_check_subprogs(struct bpf_verifier_env *env) { struct bpf_prog_aux *aux = env->prog->aux; struct bpf_func_info_aux *sub_aux; - int i, ret, new_cnt; + int context, i, ret, new_cnt; if (!aux->func_info) return 0; - /* exception callback is presumed to be always called */ - if (env->exception_callback_subprog) - subprog_aux(env, env->exception_callback_subprog)->called = true; + /* + * Callbacks cannot throw, so the exception callback always runs in the + * main program's context. It is presumed to be always called. + */ + if (env->exception_callback_subprog) { + sub_aux = subprog_aux(env, env->exception_callback_subprog); + sub_aux->called[env->prog->sleepable] = true; + } again: new_cnt = 0; @@ -19652,29 +19656,28 @@ static int do_check_subprogs(struct bpf_verifier_env *env) continue; sub_aux = subprog_aux(env, i); - if (!sub_aux->called || sub_aux->verified) - continue; + for (context = 0; context < ARRAY_SIZE(sub_aux->called); context++) { + if (!sub_aux->called[context] || sub_aux->verified[context]) + continue; - env->insn_idx = env->subprog_info[i].start; - WARN_ON_ONCE(env->insn_idx == 0); - ret = do_check_common(env, i); - if (ret) { - return ret; - } else if (env->log.level & BPF_LOG_LEVEL) { - verbose(env, "Func#%d ('%s') is safe for any args that match its prototype\n", - i, bpf_subprog_name(env, i)); + env->insn_idx = env->subprog_info[i].start; + WARN_ON_ONCE(env->insn_idx == 0); + ret = do_check_common(env, i, context); + if (ret) + return ret; + if (env->log.level & BPF_LOG_LEVEL) + verbose(env, "Func#%d ('%s') is safe for any args " + "that match its prototype\n", + i, bpf_subprog_name(env, i)); + + sub_aux->verified[context] = true; + new_cnt++; } - - /* We verified new global subprog, it might have called some - * more global subprogs that we haven't verified yet, so we - * need to do another pass over subprogs to verify those. - */ - sub_aux->verified = true; - new_cnt++; } - /* We can't loop forever as we verify at least one global subprog on - * each pass. + /* + * We can't loop forever as each pass verifies at least one new context, + * and there are only two contexts per global subprog. */ if (new_cnt) goto again; @@ -19687,7 +19690,7 @@ static int do_check_main(struct bpf_verifier_env *env) int ret; env->insn_idx = 0; - ret = do_check_common(env, 0); + ret = do_check_common(env, 0, env->prog->sleepable); if (!ret) env->prog->aux->stack_depth = env->subprog_info[0].stack_depth; return ret; From a452e729b7be317837bb2157d5d88aa3de9873e6 Mon Sep 17 00:00:00 2001 From: Kumar Kartikeya Dwivedi Date: Mon, 14 Sep 2026 15:19:21 +0200 Subject: [PATCH 314/592] selftests/bpf: Test global subprog callback contexts Exercise global subprogram verification from workqueue callbacks, which can run in a sleepable context even when the containing program is not sleepable. An unprotected callback must not let the global subprogram use implicit RCU protection inherited from the program. Add a negative case which loads an RCU-protected task kptr in a global subprogram reached from a workqueue callback. It fails on an unfixed kernel because the program is incorrectly accepted. Also cover a workqueue callback protected by an explicit RCU read-side critical section, where the same global subprogram remains valid. Call the same harmless global subprogram directly from the main program and from an unprotected callback. Mark it __weak __noinline so both calls survive optimization, and check that its instruction statistics account for both verification contexts. This also verifies that global calls from callbacks are not rejected wholesale. Workqueue callbacks return zero explicitly after the global call, as required by their contract. Signed-off-by: Kumar Kartikeya Dwivedi Acked-by: Eduard Zingerman Link: https://patch.msgid.link/20260914131923.2544250-3-memxor@gmail.com Signed-off-by: Eduard Zingerman --- .../bpf/progs/verifier_async_cb_context.c | 106 ++++++++++++++++++ 1 file changed, 106 insertions(+) diff --git a/tools/testing/selftests/bpf/progs/verifier_async_cb_context.c b/tools/testing/selftests/bpf/progs/verifier_async_cb_context.c index e0926767bbd3..1b653bfb63eb 100644 --- a/tools/testing/selftests/bpf/progs/verifier_async_cb_context.c +++ b/tools/testing/selftests/bpf/progs/verifier_async_cb_context.c @@ -9,6 +9,11 @@ char _license[] SEC("license") = "GPL"; +struct task_struct *bpf_task_acquire(struct task_struct *p) __ksym; +void bpf_task_release(struct task_struct *p) __ksym; +void bpf_rcu_read_lock(void) __ksym; +void bpf_rcu_read_unlock(void) __ksym; + /* Timer tests */ struct timer_elem { @@ -164,6 +169,7 @@ int syscall_btf_find_prog(void *ctx) struct wq_elem { struct bpf_wq w; + struct task_struct __kptr *task; }; struct { @@ -217,6 +223,106 @@ int wq_sleepable_prog(void *ctx) return 0; } +__noinline int wq_global_acquire(void) +{ + struct task_struct *task, *acquired; + struct wq_elem *val; + int key = 0; + + val = bpf_map_lookup_elem(&wq_map, &key); + if (!val) + return 0; + + task = val->task; + if (!task) + return 0; + + acquired = bpf_task_acquire(task); + if (acquired) + bpf_task_release(acquired); + return 0; +} + +static int wq_global_rcu_cb(void *map, int *key, void *value) +{ + wq_global_acquire(); + return 0; +} + +SEC("fentry/bpf_fentry_test1") +__failure __msg("R1 must be a rcu pointer") +int wq_global_rcu_prog(void *ctx) +{ + struct wq_elem *val; + int key = 0; + + val = bpf_map_lookup_elem(&wq_map, &key); + if (!val) + return 0; + + bpf_wq_init(&val->w, &wq_map, 0); + bpf_wq_set_callback(&val->w, wq_global_rcu_cb, 0); + return 0; +} + +static int wq_global_rcu_lock_cb(void *map, int *key, void *value) +{ + bpf_rcu_read_lock(); + wq_global_acquire(); + bpf_rcu_read_unlock(); + return 0; +} + +SEC("fentry/bpf_fentry_test1") +__success +int wq_global_rcu_lock_prog(void *ctx) +{ + struct wq_elem *val; + int key = 0; + + /* Verify the same global subprog in non-sleepable and protected contexts. */ + wq_global_acquire(); + + val = bpf_map_lookup_elem(&wq_map, &key); + if (!val) + return 0; + + bpf_wq_init(&val->w, &wq_map, 0); + bpf_wq_set_callback(&val->w, wq_global_rcu_lock_cb, 0); + return 0; +} + +__weak __noinline int wq_global_no_rcu(void) +{ + return 0; +} + +static int wq_global_no_rcu_cb(void *map, int *key, void *value) +{ + wq_global_no_rcu(); + return 0; +} + +SEC("fentry/bpf_fentry_test1") +__success __log_level(4) +__msg("subprog {{[0-9]+}} (wq_global_no_rcu) global insns_self 4 insns_total 4 stack 0") +int wq_global_no_rcu_prog(void *ctx) +{ + struct wq_elem *val; + int key = 0; + + /* Verify the same global in non-sleepable and unprotected contexts. */ + wq_global_no_rcu(); + + val = bpf_map_lookup_elem(&wq_map, &key); + if (!val) + return 0; + + bpf_wq_init(&val->w, &wq_map, 0); + bpf_wq_set_callback(&val->w, wq_global_no_rcu_cb, 0); + return 0; +} + /* Task work tests */ struct task_work_elem { From 70504de0bb627848667207bec7ccfd647deb8814 Mon Sep 17 00:00:00 2001 From: Zhiling Zou Date: Fri, 11 Sep 2026 00:18:24 +0800 Subject: [PATCH 315/592] xsk: Use a 32-bit compare in xsk_map_gen_lookup xsk_map_gen_lookup() loads a u32 key and compares it with max_entries using BPF_JMP_IMM. BPF immediates are sign-extended to 64 bits, so a max_entries value of 0x80000000 or higher becomes a threshold larger than every zero-extended 32-bit key. An out-of-range index then skips the bounds check and the generated lookup reads past xsk_map[]. Compare with BPF_JMP32_IMM so the check stays in 32-bit unsigned range. Fixes: e65650f291ee ("bpf: Implement map_gen_lookup() callback for XSKMAP") Reported-by: Vega Signed-off-by: Zhiling Zou Signed-off-by: Alexei Starovoitov Reviewed-by: Emil Tsalapatis Link: https://patch.msgid.link/7d2cb8e8dfaa9eb8fdff85156987a60960787dc3.1789056660.git.zhilinz@nebusec.ai Signed-off-by: Eduard Zingerman --- net/xdp/xskmap.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/net/xdp/xskmap.c b/net/xdp/xskmap.c index 3bff346308d0..bf00d6463c19 100644 --- a/net/xdp/xskmap.c +++ b/net/xdp/xskmap.c @@ -124,7 +124,7 @@ static int xsk_map_gen_lookup(struct bpf_map *map, struct bpf_insn *insn_buf) struct bpf_insn *insn = insn_buf; *insn++ = BPF_LDX_MEM(BPF_W, ret, index, 0); - *insn++ = BPF_JMP_IMM(BPF_JGE, ret, map->max_entries, 5); + *insn++ = BPF_JMP32_IMM(BPF_JGE, ret, map->max_entries, 5); *insn++ = BPF_ALU64_IMM(BPF_LSH, ret, ilog2(sizeof(struct xsk_sock *))); *insn++ = BPF_ALU64_IMM(BPF_ADD, mp, offsetof(struct xsk_map, xsk_map)); *insn++ = BPF_ALU64_REG(BPF_ADD, ret, mp); From 05762c5bc1cfdcac36747994fde2c04387a457f1 Mon Sep 17 00:00:00 2001 From: Frank Sorenson Date: Wed, 16 Sep 2026 16:33:52 -0500 Subject: [PATCH 316/592] smb: client: fix next_buffer UAF and NextCommand bounds in compound PDUs Fix several related bounds checking and pointer lifecycle issues in receive_encrypted_standard()'s handling of compound encrypted frames: - Clear next_buffer after assigning it to server->bigbuf. A stale next_buffer pointer can lead to a use-after-free on subsequent error paths. - Update pdu_length to the decrypted plaintext size (buf_size). Using the pre-decryption length allows NextCommand to point into stale ciphertext residue. - Reject next_cmd values smaller than MID_HEADER_SIZE(server). - Fix an integer overflow in the upper bound check by verifying pdu_length - next_cmd < MID_HEADER_SIZE(server), ensuring the trailing slice is large enough for a header. Fixes: b24df3e30cbf ("cifs: update receive_encrypted_standard to handle compounded responses") Cc: stable@vger.kernel.org Signed-off-by: Frank Sorenson Reviewed-by: David Howells Signed-off-by: Paulo Alcantara --- fs/smb/client/smb2ops.c | 11 ++++++++++- 1 file changed, 10 insertions(+), 1 deletion(-) diff --git a/fs/smb/client/smb2ops.c b/fs/smb/client/smb2ops.c index cb4fd09f996e..fcf7033889c7 100644 --- a/fs/smb/client/smb2ops.c +++ b/fs/smb/client/smb2ops.c @@ -5365,6 +5365,7 @@ receive_encrypted_standard(struct TCP_Server_Info *server, length = decrypt_raw_data(server, buf, buf_size, NULL, false); if (length) return length; + pdu_length = buf_size; next_is_large = server->large_buf; one_more: @@ -5377,8 +5378,15 @@ receive_encrypted_standard(struct TCP_Server_Info *server, } if (next_cmd) { - if (WARN_ON_ONCE(next_cmd > pdu_length)) + if (next_cmd < MID_HEADER_SIZE(server) || + next_cmd > pdu_length || + pdu_length - next_cmd < MID_HEADER_SIZE(server)) { + unsigned int max_next = pdu_length > (unsigned int)MID_HEADER_SIZE(server) ? + pdu_length - (unsigned int)MID_HEADER_SIZE(server) : 0; + cifs_server_dbg(VFS, "invalid NextCommand offset %u out of range [%zu, %u]\n", + next_cmd, MID_HEADER_SIZE(server), max_next); return -1; + } if (next_is_large) next_buffer = (char *)cifs_buf_get(); else @@ -5414,6 +5422,7 @@ receive_encrypted_standard(struct TCP_Server_Info *server, server->bigbuf = buf = next_buffer; else server->smallbuf = buf = next_buffer; + next_buffer = NULL; goto one_more; } else if (ret != 0) { /* From b4694f269e66dfcd66991446375285723b6957bd Mon Sep 17 00:00:00 2001 From: Frank Sorenson Date: Wed, 16 Sep 2026 16:33:53 -0500 Subject: [PATCH 317/592] smb: client: validate minimum PDU size before smb2_get_data_area_len() __smb2_calc_size() calls smb2_get_data_area_len(), which reads command-specific struct fields to locate the data area. However, smb2_check_message() only validates StructureSize2, meaning a truncated response could cause smb2_get_data_area_len() to read out-of-bounds. Replace has_smb2_data_area[] with smb2_min_pdu_len[], which is now used to indicate both whether a command's response has a data area and the size of that fixed response struct. A non-zero entry means the command has a data area, and is the minimum length required before the struct is read. For each command with a data area, PDUs shorter than this minimum size are rejected instead of parsed. The minimum is not applied to SMB2 error responses, which carry only the 9-byte error body, the same exemption the StructureSize2 check above it already makes. STATUS_MORE_PROCESSING_REQUIRED is treated as a normal reply, since an in-progress SESSION_SETUP response carries a full body and a security blob. Signed-off-by: Frank Sorenson Reviewed-by: David Howells Signed-off-by: Paulo Alcantara --- fs/smb/client/smb2misc.c | 69 ++++++++++++++++++++++++---------------- 1 file changed, 41 insertions(+), 28 deletions(-) diff --git a/fs/smb/client/smb2misc.c b/fs/smb/client/smb2misc.c index 9068175e57cd..0cfe60ae42c3 100644 --- a/fs/smb/client/smb2misc.c +++ b/fs/smb/client/smb2misc.c @@ -85,6 +85,36 @@ static const __le16 smb2_rsp_struct_sizes[NUMBER_OF_SMB2_COMMANDS] = { /* SMB2_OPLOCK_BREAK */ cpu_to_le16(24) }; +/* + * Minimum received PDU size for commands whose response carries a + * variable-length data area. A non-zero entry marks the command as + * having one, and gives the length smb2_check_message() requires + * before smb2_get_data_area_len() reads the offset and length fields + * out of the fixed response struct. + */ +static const size_t smb2_min_pdu_len[NUMBER_OF_SMB2_COMMANDS] = { + /* SMB2_NEGOTIATE */ sizeof(struct smb2_negotiate_rsp), + /* SMB2_SESSION_SETUP */ sizeof(struct smb2_sess_setup_rsp), + /* SMB2_LOGOFF */ 0, + /* SMB2_TREE_CONNECT */ 0, + /* SMB2_TREE_DISCONNECT */ 0, + /* SMB2_CREATE */ sizeof(struct smb2_create_rsp), + /* SMB2_CLOSE */ 0, + /* SMB2_FLUSH */ 0, + /* SMB2_READ */ sizeof(struct smb2_read_rsp), + /* SMB2_WRITE */ 0, + /* SMB2_LOCK */ 0, + /* SMB2_IOCTL */ sizeof(struct smb2_ioctl_rsp), + /* SMB2_CANCEL */ 0, + /* SMB2_ECHO */ 0, + /* SMB2_QUERY_DIRECTORY */ sizeof(struct smb2_query_directory_rsp), + /* SMB2_CHANGE_NOTIFY */ sizeof(struct smb2_change_notify_rsp), + /* SMB2_QUERY_INFO */ sizeof(struct smb2_query_info_rsp), + /* SMB2_SET_INFO */ 0, + /* SMB2_OPLOCK_BREAK */ 0, +}; + +#define smb2_has_data_area(cmd) (smb2_min_pdu_len[cmd] != 0) #define SMB311_NEGPROT_BASE_SIZE (sizeof(struct smb2_hdr) + sizeof(struct smb2_negotiate_rsp)) static __u32 get_neg_ctxt_len(struct smb2_hdr *hdr, __u32 len, @@ -233,6 +263,16 @@ smb2_check_message(char *buf, unsigned int pdu_len, unsigned int len, } } + if ((shdr->Status == STATUS_SUCCESS || + shdr->Status == STATUS_MORE_PROCESSING_REQUIRED || + pdu->StructureSize2 != SMB2_ERROR_STRUCTURE_SIZE2_LE) && + smb2_has_data_area(command) && + len < smb2_min_pdu_len[command]) { + cifs_server_dbg(VFS, "SMB2 command %d response too short: %u < %zu\n", + command, len, smb2_min_pdu_len[command]); + return 1; + } + have_data = false; data_area_overlap = false; calc_len = __smb2_calc_size(buf, &have_data, &data_area_overlap); @@ -298,33 +338,6 @@ smb2_check_message(char *buf, unsigned int pdu_len, unsigned int len, return 0; } -/* - * The size of the variable area depends on the offset and length fields - * located in different fields for various SMB2 responses. SMB2 responses - * with no variable length info, show an offset of zero for the offset field. - */ -static const bool has_smb2_data_area[NUMBER_OF_SMB2_COMMANDS] = { - /* SMB2_NEGOTIATE */ true, - /* SMB2_SESSION_SETUP */ true, - /* SMB2_LOGOFF */ false, - /* SMB2_TREE_CONNECT */ false, - /* SMB2_TREE_DISCONNECT */ false, - /* SMB2_CREATE */ true, - /* SMB2_CLOSE */ false, - /* SMB2_FLUSH */ false, - /* SMB2_READ */ true, - /* SMB2_WRITE */ false, - /* SMB2_LOCK */ false, - /* SMB2_IOCTL */ true, - /* SMB2_CANCEL */ false, /* BB CHECK this not listed in documentation */ - /* SMB2_ECHO */ false, - /* SMB2_QUERY_DIRECTORY */ true, - /* SMB2_CHANGE_NOTIFY */ true, - /* SMB2_QUERY_INFO */ true, - /* SMB2_SET_INFO */ false, - /* SMB2_OPLOCK_BREAK */ false -}; - /* * Returns the pointer to the beginning of the data area. Length of the data * area and the offset to it (from the beginning of the smb are also returned. @@ -451,7 +464,7 @@ __smb2_calc_size(void *buf, bool *have_data, bool *data_area_overlap) */ len += le16_to_cpu(pdu->StructureSize2); - if (has_smb2_data_area[le16_to_cpu(shdr->Command)] == false) + if (!smb2_has_data_area(le16_to_cpu(shdr->Command))) goto calc_size_exit; smb2_get_data_area_len(&offset, &data_length, shdr); From f73726b83e4756fdaa099e1bc1143293bd57ad79 Mon Sep 17 00:00:00 2001 From: Frank Sorenson Date: Wed, 16 Sep 2026 16:33:54 -0500 Subject: [PATCH 318/592] smb: client: fix server->total_read for compound encrypted PDUs In receive_encrypted_standard(), server->total_read is left at the full decrypted frame size when walking sub-PDUs of a compound encrypted frame. As a result, cifs_handle_standard() passes this full size to smb2_check_message(), causing the PDU length guards to incorrectly validate the entire compound frame instead of the current sub-PDU. This allows truncated non-last sub-PDUs to bypass length validation, leading to out-of-bounds reads in smb2_get_data_area_len(). Fix this by setting server->total_read to the true length of the current sub-PDU: next_cmd for non-last sub-PDUs, and the remaining pdu_length for the last one. Fixes: b24df3e30cbf ("cifs: update receive_encrypted_standard to handle compounded responses") Cc: stable@vger.kernel.org Signed-off-by: Frank Sorenson Reviewed-by: David Howells Signed-off-by: Paulo Alcantara --- fs/smb/client/smb2ops.c | 1 + 1 file changed, 1 insertion(+) diff --git a/fs/smb/client/smb2ops.c b/fs/smb/client/smb2ops.c index fcf7033889c7..7f2177f6fc01 100644 --- a/fs/smb/client/smb2ops.c +++ b/fs/smb/client/smb2ops.c @@ -5371,6 +5371,7 @@ receive_encrypted_standard(struct TCP_Server_Info *server, one_more: shdr = (struct smb2_hdr *)buf; next_cmd = le32_to_cpu(shdr->NextCommand); + server->total_read = next_cmd ? next_cmd : pdu_length; if (*num_mids >= MAX_COMPOUND) { cifs_server_dbg(VFS, "too many PDUs in compound\n"); From e83330c55edc0c3ac08aa6c95e49e4694c65523b Mon Sep 17 00:00:00 2001 From: Frank Sorenson Date: Wed, 16 Sep 2026 16:33:55 -0500 Subject: [PATCH 319/592] smb: client: fix missing lower-bound check on DFS referral string offsets parse_dfs_referrals() checks that DfsPathOffset and NetworkAddressOffset do not exceed the buffer end, but fails to check that they don't point inside the referral header itself. If a server provides an offset smaller than sizeof(struct dfs_referral_level_3), the derived string pointer overlaps with the struct fields, causing cifs_strndup_from_utf16() to interpret header data as UTF-16 strings. Fix this by enforcing that string offsets are at least sizeof(*ref). Fixes: 4ecce920e13a ("CIFS: move DFS response parsing out of SMB1 code") Cc: stable@vger.kernel.org Signed-off-by: Frank Sorenson Reviewed-by: David Howells Signed-off-by: Paulo Alcantara --- fs/smb/client/misc.c | 12 ++++++++++-- 1 file changed, 10 insertions(+), 2 deletions(-) diff --git a/fs/smb/client/misc.c b/fs/smb/client/misc.c index 945194fe7a97..05168284f205 100644 --- a/fs/smb/client/misc.c +++ b/fs/smb/client/misc.c @@ -788,7 +788,11 @@ parse_dfs_referrals(struct get_dfs_referral_rsp *rsp, u32 rsp_size, node->ref_flag = le16_to_cpu(ref->ReferralEntryFlags); /* copy DfsPath */ - if (le16_to_cpu(ref->DfsPathOffset) > data_end - (char *)ref) { + if (le16_to_cpu(ref->DfsPathOffset) < sizeof(*ref) || + le16_to_cpu(ref->DfsPathOffset) > data_end - (char *)ref) { + cifs_dbg(VFS, "%s: DfsPathOffset %u out of range [%zu, %td]\n", + __func__, le16_to_cpu(ref->DfsPathOffset), + sizeof(*ref), data_end - (char *)ref); rc = -EINVAL; goto parse_DFS_referrals_exit; } @@ -802,7 +806,11 @@ parse_dfs_referrals(struct get_dfs_referral_rsp *rsp, u32 rsp_size, } /* copy link target UNC */ - if (le16_to_cpu(ref->NetworkAddressOffset) > data_end - (char *)ref) { + if (le16_to_cpu(ref->NetworkAddressOffset) < sizeof(*ref) || + le16_to_cpu(ref->NetworkAddressOffset) > data_end - (char *)ref) { + cifs_dbg(VFS, "%s: NetworkAddressOffset %u out of range [%zu, %td]\n", + __func__, le16_to_cpu(ref->NetworkAddressOffset), + sizeof(*ref), data_end - (char *)ref); rc = -EINVAL; goto parse_DFS_referrals_exit; } From 1b3221bb121079ad79a1f3c3aa360ba649832e7a Mon Sep 17 00:00:00 2001 From: Frank Sorenson Date: Wed, 16 Sep 2026 16:33:56 -0500 Subject: [PATCH 320/592] smb: client: reject short Next offsets in parse_server_interfaces() In parse_server_interfaces(), the server-supplied Next offset is validated against bytes_left, but not against the size of the interface structure itself. A small, non-zero Next value can pass the bounds check but advance the pointer by less than sizeof(*p). This causes the next iteration of the loop to read misaligned, overlapping structure fields. Fix this by ensuring the Next offset is at least sizeof(*p). Fixes: 7d34ec36abb8 ("smb3: fix for slab out of bounds on mount to ksmbd") Cc: stable@vger.kernel.org Signed-off-by: Frank Sorenson Reviewed-by: David Howells Signed-off-by: Paulo Alcantara --- fs/smb/client/smb2ops.c | 6 +++--- 1 file changed, 3 insertions(+), 3 deletions(-) diff --git a/fs/smb/client/smb2ops.c b/fs/smb/client/smb2ops.c index 7f2177f6fc01..bda940cb3784 100644 --- a/fs/smb/client/smb2ops.c +++ b/fs/smb/client/smb2ops.c @@ -785,9 +785,9 @@ parse_server_interfaces(struct network_interface_info_ioctl_rsp *buf, break; } /* Validate that Next doesn't point beyond the buffer */ - if (next > bytes_left) { - cifs_dbg(VFS, "%s: invalid Next pointer %zu > %zd\n", - __func__, next, bytes_left); + if (next < sizeof(*p) || next > bytes_left) { + cifs_dbg(VFS, "%s: invalid Next pointer %zu out of range [%zu, %zd]\n", + __func__, next, sizeof(*p), bytes_left); rc = -EINVAL; goto out; } From eeb5ef6083e1cefa2ef75041b5597ff228b8d7bb Mon Sep 17 00:00:00 2001 From: Frank Sorenson Date: Wed, 16 Sep 2026 16:33:57 -0500 Subject: [PATCH 321/592] smb: client: fix OOB struct field reads in move_smb2_ea_to_cifs() In move_smb2_ea_to_cifs(), the while (src_size > 0) loop condition is insufficient. It allows iteration to continue even if the remaining src_size is too small to contain a complete smb2_ea_info structure. Consequently, reads of ea_name_length and ea_value_length can occur out-of-bounds. Fix this by ensuring src_size >= sizeof(*src) before attempting to read any structure fields. Additionally, reject any next_entry_offset that is smaller than sizeof(*src) or that would advance the pointer beyond the available buffer. Note that for calls where the server returns a malformed EA list, the error returned to userspace changes from -ENODATA (getxattr) or -ERANGE (listxattr) to -EIO. This correctly signals a server protocol error rather than misleadingly indicating "attribute not present" or "output buffer too small". Fixes: 95907fea4fd8 ("cifs: Add support for reading attributes on SMB2+") Cc: stable@vger.kernel.org Signed-off-by: Frank Sorenson Reviewed-by: David Howells Signed-off-by: Paulo Alcantara --- fs/smb/client/smb2ops.c | 25 +++++++++++++++++-------- fs/smb/client/trace.h | 1 + 2 files changed, 18 insertions(+), 8 deletions(-) diff --git a/fs/smb/client/smb2ops.c b/fs/smb/client/smb2ops.c index bda940cb3784..ee3c98e3f316 100644 --- a/fs/smb/client/smb2ops.c +++ b/fs/smb/client/smb2ops.c @@ -1053,8 +1053,9 @@ move_smb2_ea_to_cifs(char *dst, size_t dst_size, char *name, *value; size_t buf_size = dst_size; size_t name_len, value_len, user_name_len; + u32 next_off; - while (src_size > 0) { + while (src_size >= sizeof(*src)) { name_len = (size_t)src->ea_name_length; value_len = (size_t)le16_to_cpu(src->ea_value_length); @@ -1110,14 +1111,22 @@ move_smb2_ea_to_cifs(char *dst, size_t dst_size, if (!src->next_entry_offset) break; - if (src_size < le32_to_cpu(src->next_entry_offset)) { - /* stop before overrun buffer */ - rc = -ERANGE; - break; + next_off = le32_to_cpu(src->next_entry_offset); + if (next_off < sizeof(*src) || src_size < next_off) { + cifs_dbg(FYI, "EA next_entry_offset %u out of range [%zu, %zu]\n", + next_off, sizeof(*src), src_size); + rc = smb_EIO2(smb_eio_trace_ea_next_offset, + next_off, src_size); + goto out; + } + src_size -= next_off; + src = (void *)((char *)src + next_off); + if (src_size > 0 && src_size < sizeof(*src)) { + cifs_dbg(FYI, "EA next_entry_offset %u left truncated entry (%zu bytes)\n", + next_off, src_size); + rc = smb_EIO2(smb_eio_trace_ea_next_offset, next_off, src_size); + goto out; } - src_size -= le32_to_cpu(src->next_entry_offset); - src = (void *)((char *)src + - le32_to_cpu(src->next_entry_offset)); } /* didn't find the named attribute */ diff --git a/fs/smb/client/trace.h b/fs/smb/client/trace.h index b442cccd1530..bb8d0197cb54 100644 --- a/fs/smb/client/trace.h +++ b/fs/smb/client/trace.h @@ -27,6 +27,7 @@ EM(smb_eio_trace_copychunk_overcopy_c, "copychunk_overcopy_c") \ EM(smb_eio_trace_create_rsp_too_small, "create_rsp_too_small") \ EM(smb_eio_trace_dfsref_no_rsp, "dfsref_no_rsp") \ + EM(smb_eio_trace_ea_next_offset, "ea_next_offset") \ EM(smb_eio_trace_ea_overrun, "ea_overrun") \ EM(smb_eio_trace_extract_will_pin, "extract_will_pin") \ EM(smb_eio_trace_forced_shutdown, "forced_shutdown") \ From b09d092eb24ad0110f16a9b7c1ed5d2a0c1733dc Mon Sep 17 00:00:00 2001 From: Frank Sorenson Date: Wed, 16 Sep 2026 16:33:58 -0500 Subject: [PATCH 322/592] smb: client: fix missing iov bounds check in parse_posix_sids() In parse_posix_sids(), sidsbuf_end is calculated using the server-supplied out_len without being validated against the actual length of the received iov (iov_len). If a server provides an inflated out_len, sidsbuf_end will point past the end of the iov. This defeats the bounds guards in posix_info_sid_size(), allowing out-of-bounds reads into adjacent kernel memory. Fix this by rejecting responses where the calculated sidsbuf_end would exceed the received iov boundaries or cause pointer wraparound. Fixes: a90f37e3d7ac ("smb: client: parse owner/group when creating reparse points") Cc: stable@vger.kernel.org Signed-off-by: Frank Sorenson Reviewed-by: David Howells Signed-off-by: Paulo Alcantara --- fs/smb/client/smb2inode.c | 11 +++++++++++ 1 file changed, 11 insertions(+) diff --git a/fs/smb/client/smb2inode.c b/fs/smb/client/smb2inode.c index 96063e355186..13fe8e3b48f3 100644 --- a/fs/smb/client/smb2inode.c +++ b/fs/smb/client/smb2inode.c @@ -77,6 +77,17 @@ static int parse_posix_sids(struct cifs_open_info_data *data, sidsbuf = (u8 *)qi + le16_to_cpu(qi->OutputBufferOffset) + qi_len; sidsbuf_end = sidsbuf + out_len - qi_len; + if (sidsbuf_end < sidsbuf) { + cifs_dbg(VFS, "%s: server-supplied out_len %u caused pointer wraparound\n", + __func__, out_len); + return -EINVAL; + } + if (sidsbuf_end > (u8 *)rsp_iov->iov_base + rsp_iov->iov_len) { + cifs_dbg(VFS, "%s: server-supplied out_len %u overruns iov by %td bytes\n", + __func__, out_len, + sidsbuf_end - ((u8 *)rsp_iov->iov_base + rsp_iov->iov_len)); + return -EINVAL; + } owner_len = posix_info_sid_size(sidsbuf, sidsbuf_end); if (owner_len == -1) From 4775c3b7a597907e0b97556c7986fda238a377ae Mon Sep 17 00:00:00 2001 From: Frank Sorenson Date: Wed, 16 Sep 2026 16:33:59 -0500 Subject: [PATCH 323/592] smb: client: fix potential OOB read in smb3_enum_snapshots() If snapshot_array_size is smaller than GMT_TOKEN_SIZE, smb3_enum_snapshots() sets ret_data_len to sizeof(struct smb_snapshot_array) without verifying the actual length of the server's reply. Because SMB2_ioctl() places no lower bound on the server-supplied OutputCount and allocates retbuf to exactly that length, a short reply results in ret_data_len exceeding the size of retbuf. The subsequent copy_to_user() then reads past the end of retbuf, leaking adjacent slab memory to userspace. The subsequent clamp check is ineffective as it only reduces ret_data_len. Fix this by rejecting replies shorter than sizeof(struct smb_snapshot_array) with -EIO. Note that the bound is set to the 12-byte struct size rather than the 16-byte MIN_SNAPSHOT_ARRAY_SIZE defined in MS-SMB2 3.3.5.15.1, because 12 bytes is exactly what copy_to_user() attempts to read. Fixes: e02789a53d71 ("smb3: enumerating snapshots was leaving part of the data off end") Cc: stable@vger.kernel.org Signed-off-by: Frank Sorenson Reviewed-by: David Howells Signed-off-by: Paulo Alcantara --- fs/smb/client/smb2ops.c | 8 +++++++- 1 file changed, 7 insertions(+), 1 deletion(-) diff --git a/fs/smb/client/smb2ops.c b/fs/smb/client/smb2ops.c index ee3c98e3f316..3464470d3297 100644 --- a/fs/smb/client/smb2ops.c +++ b/fs/smb/client/smb2ops.c @@ -2463,8 +2463,14 @@ smb3_enum_snapshots(const unsigned int xid, struct cifs_tcon *tcon, * and retry the ioctl again with larger array size sufficient * to hold all of the snapshot GMT tokens on the second try. */ - if (snapshot_in.snapshot_array_size < GMT_TOKEN_SIZE) + if (snapshot_in.snapshot_array_size < GMT_TOKEN_SIZE) { + if (ret_data_len < sizeof(struct smb_snapshot_array)) { + rc = -EIO; + kfree(retbuf); + return rc; + } ret_data_len = sizeof(struct smb_snapshot_array); + } /* * We return struct SRV_SNAPSHOT_ARRAY, followed by From 5f0306e731e2f46e91419eae57eee3a241c055e0 Mon Sep 17 00:00:00 2001 From: Frank Sorenson Date: Wed, 16 Sep 2026 16:34:00 -0500 Subject: [PATCH 324/592] smb: client: fix reparse buffer bounds in cifs_query_reparse_point() In cifs_query_reparse_point(), the start >= end check before casting to struct reparse_data_buffer * only ensures the start pointer is within the response. It fails to verify that there is enough space remaining for the fixed 8-byte header of the structure. If a server provides a DataOffset that leaves less than 8 bytes remaining, the check passes, but subsequent reads of ReparseTag and ReparseDataLength will occur out-of-bounds. Fix this by ensuring the remaining space is at least the size of the reparse_data_buffer structure before accessing its fields. Fixes: 56e84c64fc25 ("cifs: Fix validation of SMB1 query reparse point response") Cc: stable@vger.kernel.org Signed-off-by: Frank Sorenson Reviewed-by: David Howells Signed-off-by: Paulo Alcantara --- fs/smb/client/cifssmb.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/fs/smb/client/cifssmb.c b/fs/smb/client/cifssmb.c index f9aff0712794..6dddbd84b93b 100644 --- a/fs/smb/client/cifssmb.c +++ b/fs/smb/client/cifssmb.c @@ -3080,7 +3080,7 @@ int cifs_query_reparse_point(const unsigned int xid, end = 2 + get_bcc(&io_rsp->hdr) + (__u8 *)&io_rsp->ByteCount; start = (__u8 *)&io_rsp->hdr.Protocol + data_offset; - if (start >= end) { + if (start >= end || (size_t)(end - start) < sizeof(*buf)) { rc = smb_EIO2(smb_eio_trace_qreparse_data_area, (unsigned long)start - (unsigned long)io_rsp, (unsigned long)end - (unsigned long)io_rsp); From 3d743adf090cd4c9a2120c1e02b0482e88aa0d2d Mon Sep 17 00:00:00 2001 From: Frieder Schrempf Date: Thu, 17 Sep 2026 16:10:15 +0200 Subject: [PATCH 325/592] spi: fsl-qspi: Reprogram the clock rate when the operation frequency changes fsl_qspi_select_mem() returns early when the chip select has not changed, which happens before it reaches clk_set_rate(). Since the rate is now taken from the spi-mem operation rather than from the SPI device, the controller honours op->max_freq exactly once per chip select and ignores it for every operation after that. q->selected is only reset to -1 in fsl_qspi_default_setup(), i.e. at probe and on resume, so on the common single chip select board the very first operation latches a rate that all subsequent operations inherit, whatever frequency they asked for. This results in operations being issued with the wrong frequency. Cache the operation frequency the clock was programmed for next to the selected chip select, and redo the clock setup when either changes. Fixes: 2438db5253eb ("spi: fsl-qspi: Support per spi-mem operation frequency switches") Cc: stable@vger.kernel.org Assisted-by: Claude:claude-opus-5 Signed-off-by: Frieder Schrempf Acked-by: Han Xu Link: https://patch.msgid.link/20260917-fsl-qspi-freq-op-fix-v1-1-5fbe6b02f738@kontron.de Signed-off-by: Mark Brown --- drivers/spi/spi-fsl-qspi.c | 5 ++++- 1 file changed, 4 insertions(+), 1 deletion(-) diff --git a/drivers/spi/spi-fsl-qspi.c b/drivers/spi/spi-fsl-qspi.c index 57358851029b..d2c2090442f8 100644 --- a/drivers/spi/spi-fsl-qspi.c +++ b/drivers/spi/spi-fsl-qspi.c @@ -289,6 +289,7 @@ struct fsl_qspi { struct pm_qos_request pm_qos_req; struct device *dev; int selected; + u32 selected_freq; u32 memmap_phy; }; @@ -551,7 +552,8 @@ static void fsl_qspi_select_mem(struct fsl_qspi *q, struct spi_device *spi, unsigned long rate = op->max_freq; int ret; - if (q->selected == spi_get_chipselect(spi, 0)) + if (q->selected == spi_get_chipselect(spi, 0) && + q->selected_freq == op->max_freq) return; if (needs_4x_clock(q)) @@ -571,6 +573,7 @@ static void fsl_qspi_select_mem(struct fsl_qspi *q, struct spi_device *spi, } q->selected = spi_get_chipselect(spi, 0); + q->selected_freq = op->max_freq; fsl_qspi_invalidate(q); } From c0078f4d8c8fcd8edfe8c53ffe77d48565c1957e Mon Sep 17 00:00:00 2001 From: Wei Wang Date: Tue, 15 Sep 2026 20:14:12 +0000 Subject: [PATCH 326/592] mailmap: add entry for Wei Wang My Meta email address is no longer active. Map it to my current address so that git and get_maintainer.pl stop pointing at a dead address for my contributions. Signed-off-by: Wei Wang Link: https://patch.msgid.link/20260915201412.2201757-1-weiwan@google.com Signed-off-by: Jakub Kicinski --- .mailmap | 1 + 1 file changed, 1 insertion(+) diff --git a/.mailmap b/.mailmap index 1f5540bc33f1..526217c88495 100644 --- a/.mailmap +++ b/.mailmap @@ -975,6 +975,7 @@ Vladimir Davydov Vlastimil Babka WangYuli WangYuli +Wei Wang Weiwen Hu WeiXiong Liao Wen Gong From c9dc7d730319ad64b51570c5387f1fee7b07b510 Mon Sep 17 00:00:00 2001 From: Richard Zhu Date: Thu, 13 Aug 2026 17:50:03 +0800 Subject: [PATCH 327/592] PCI: imx6: Move clock enable after core reset assertion Commit 610fa91d9863 ("PCI: imx6: Assert PERST# before enabling regulators") inadvertently moved clock enablement before core reset assertion, breaking PCI device initialization on i.MX6Q Apalis platforms with ASM1061/ASM1062 SATA controllers connected: imx6q-pcie 1ffc000.pcie: host bridge /soc/pcie@1ffc000 ranges: imx6q-pcie 1ffc000.pcie: IO 0x0001f80000..0x0001f8ffff -> 0x0000000000 imx6q-pcie 1ffc000.pcie: MEM 0x0001000000..0x0001efffff -> 0x0001000000 imx6q-pcie 1ffc000.pcie: config reg[1] 0x01f00000 == cpu 0x01f00000 imx6q-pcie 1ffc000.pcie: iATU: unroll F, 4 ob, 4 ib, align 64K, limit 4G imx6q-pcie 1ffc000.pcie: Link: Only Gen1 is enabled imx6q-pcie 1ffc000.pcie: Link failed to come up. LTSSM: POLL_CONFIG imx6q-pcie 1ffc000.pcie: probe with driver imx6q-pcie failed with error -110 NOTE: It is not 100% clear if the issue is specific to the ASM1061/ASM1062 device or on the specific power-up sequence (reset vs cold-power-on). To fix this regression, restore the original sequence where clocks are enabled after asserting core reset and configuring the controller type. Fixes: 610fa91d9863 ("PCI: imx6: Assert PERST# before enabling regulators") Reported-by: Leonardo Costa Closes: https://lore.kernel.org/all/bl7i3obu2clzsgeoct2a4mtfhv6typcjdqmgneropf3hpgwve6@n2m5uhlduw57/T/#u Reported-by: Franz Schnyder Closes: https://lore.kernel.org/all/t65y5d54axtksbfs7r4olcefqhwm6m4dz3njgnrnf7fcotj74i@o7avoznlafbj/ Signed-off-by: Richard Zhu Signed-off-by: Manivannan Sadhasivam [bhelgaas: move to pci/for-linus for v7.3] Signed-off-by: Bjorn Helgaas Cc: stable@vger.kernel.org # 7.2+ Link: https://patch.msgid.link/20260813095003.356062-1-hongxing.zhu@oss.nxp.com --- drivers/pci/controller/dwc/pci-imx6.c | 12 ++++++------ 1 file changed, 6 insertions(+), 6 deletions(-) diff --git a/drivers/pci/controller/dwc/pci-imx6.c b/drivers/pci/controller/dwc/pci-imx6.c index 39790e66b98d..f7a2eb257c16 100644 --- a/drivers/pci/controller/dwc/pci-imx6.c +++ b/drivers/pci/controller/dwc/pci-imx6.c @@ -1394,12 +1394,6 @@ static int imx_pcie_host_init(struct dw_pcie_rp *pp) } } - ret = imx_pcie_clk_enable(imx_pcie); - if (ret) { - dev_err(dev, "unable to enable pcie clocks: %d\n", ret); - goto err_pwrctrl_power_off; - } - if (pp->bridge && imx_check_flag(imx_pcie, IMX_PCIE_FLAG_HAS_LUT)) { pp->bridge->enable_device = imx_pcie_enable_device; pp->bridge->disable_device = imx_pcie_disable_device; @@ -1415,6 +1409,12 @@ static int imx_pcie_host_init(struct dw_pcie_rp *pp) imx_pcie_configure_type(imx_pcie); + ret = imx_pcie_clk_enable(imx_pcie); + if (ret) { + dev_err(dev, "unable to enable pcie clocks: %d\n", ret); + goto err_pwrctrl_power_off; + } + if (imx_pcie->phy) { ret = phy_init(imx_pcie->phy); if (ret) { From 50e80e2bb5e2be8515205b9c496b9640ddefa434 Mon Sep 17 00:00:00 2001 From: Weiming Shi Date: Tue, 15 Sep 2026 01:02:07 +0800 Subject: [PATCH 328/592] bpf: Skip unsettled links in link iterator bpf_link_prime() inserts a link into link_idr before anon_inode_getfile() succeeds and before bpf_link_settle() publishes the ID in link->id. bpf_link_by_id() treats such an ID-zero link as unsettled, but the link iterator takes a reference without this check. If anon_inode_getfile() then fails, the creator removes the ID and frees its still-private link directly. The iterator is left with a dangling reference and its next bpf_link_put() accesses freed memory. Treat ID-zero entries as transient in bpf_link_get_curr_or_next(), just as bpf_link_by_id() does. BUG: KASAN: slab-use-after-free in bpf_link_put Write of size 8 by task exp/384 Call Trace: bpf_link_put kernel/bpf/syscall.c:3372 bpf_link_seq_next kernel/bpf/link_iter.c:33 bpf_seq_read kernel/bpf/bpf_iter.c:158 vfs_read fs/read_write.c:572 ksys_read fs/read_write.c:716 do_syscall_64 arch/x86/entry/syscall_64.c:84 entry_SYSCALL_64_after_hwframe arch/x86/entry/entry_64.S:121 Kernel panic - not syncing: KASAN: panic_on_warn set ... Fixes: 9f8836127308 ("bpf: Add bpf_link iterator") Reported-by: Xiang Mei Signed-off-by: Weiming Shi Signed-off-by: Andrii Nakryiko Link: https://lore.kernel.org/bpf/20260914170206.170723-2-bestswngs@gmail.com --- kernel/bpf/syscall.c | 5 ++++- 1 file changed, 4 insertions(+), 1 deletion(-) diff --git a/kernel/bpf/syscall.c b/kernel/bpf/syscall.c index 853b47f81384..244a939b9d2d 100644 --- a/kernel/bpf/syscall.c +++ b/kernel/bpf/syscall.c @@ -6042,7 +6042,10 @@ struct bpf_link *bpf_link_get_curr_or_next(u32 *id) again: link = idr_get_next(&link_idr, id); if (link) { - link = bpf_link_inc_not_zero(link); + if (link->id) + link = bpf_link_inc_not_zero(link); + else + link = ERR_PTR(-EAGAIN); if (IS_ERR(link)) { (*id)++; goto again; From 717e0a25036b6c92cecace30913b2d874a4c22b8 Mon Sep 17 00:00:00 2001 From: Wentao Liang Date: Thu, 17 Sep 2026 16:34:39 +0000 Subject: [PATCH 329/592] cifs: Fix server use-after-free in cifs_chan_skip_or_disable() When a secondary channel is no longer supported by the server, cifs_chan_skip_or_disable() drops the channel reference with cifs_put_tcp_session() and then continues to use the server pointer by calling cifs_signal_cifsd_for_reconnect() on it and reading its primary_server pointer. cifs_put_tcp_session() can drop the last reference of the channel and tear it down, so both the channel and the primary server (whose reference is also dropped by cifs_put_tcp_session()) can be freed before they are signaled for reconnect. Signal the channel and the primary server and capture the primary server pointer before dropping the channel reference with cifs_put_tcp_session(). Fixes: f591062bdbf4 ("cifs: handle servers that still advertise multichannel after disabling") Cc: stable@vger.kernel.org Signed-off-by: Wentao Liang Signed-off-by: Paulo Alcantara --- fs/smb/client/smb2pdu.c | 13 +++++++------ 1 file changed, 7 insertions(+), 6 deletions(-) diff --git a/fs/smb/client/smb2pdu.c b/fs/smb/client/smb2pdu.c index dea05aeb53a1..880ce12f50c4 100644 --- a/fs/smb/client/smb2pdu.c +++ b/fs/smb/client/smb2pdu.c @@ -189,18 +189,19 @@ cifs_chan_skip_or_disable(struct cifs_ses *ses, spin_unlock(&ses->chan_lock); /* - * the above reference of server by channel - * needs to be dropped without holding chan_lock - * as cifs_put_tcp_session takes a higher lock - * i.e. cifs_tcp_ses_lock + * signal the channel and its primary server to + * reconnect before dropping the above reference of + * server by channel, which is done without holding + * chan_lock as cifs_put_tcp_session takes a higher + * lock i.e. cifs_tcp_ses_lock */ - cifs_put_tcp_session(server, from_reconnect); - cifs_signal_cifsd_for_reconnect(server, false); /* mark primary server as needing reconnect */ pserver = server->primary_server; cifs_signal_cifsd_for_reconnect(pserver, false); + + cifs_put_tcp_session(server, from_reconnect); skip_terminate: return -EHOSTDOWN; } From 8e0b235bd918d06f54ba8fddd2c3ddc36ca59c15 Mon Sep 17 00:00:00 2001 From: =?UTF-8?q?Bj=C3=B6rn=20T=C3=B6pel?= Date: Tue, 15 Sep 2026 12:49:15 +0200 Subject: [PATCH 330/592] eth: fbnic: Fix payload page pool error cleanup MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit The payload page pool pointer contains an error pointer when its allocation fails. The cleanup path passes that error pointer to page_pool_destroy() instead of destroying the header page pool. This can dereference the error pointer and leave the header page pool allocated. Destroy the header page pool instead. Fixes: 8a11010fdd96 ("eth: fbnic: allocate unreadable page pool for the payloads") Reported-by: Sashiko Link: https://lore.kernel.org/netdev/178915061000.219967.7726187707862333281@kernel.org/ Signed-off-by: Björn Töpel Reviewed-by: Simon Horman Link: https://patch.msgid.link/20260915104917.3978113-1-bjorn@kernel.org Signed-off-by: Jakub Kicinski --- drivers/net/ethernet/meta/fbnic/fbnic_txrx.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/drivers/net/ethernet/meta/fbnic/fbnic_txrx.c b/drivers/net/ethernet/meta/fbnic/fbnic_txrx.c index e7918d3f6aba..661dee1661af 100644 --- a/drivers/net/ethernet/meta/fbnic/fbnic_txrx.c +++ b/drivers/net/ethernet/meta/fbnic/fbnic_txrx.c @@ -1622,7 +1622,7 @@ fbnic_alloc_qt_page_pools(struct fbnic_net *fbn, struct fbnic_q_triad *qt, return 0; err_destroy_sub0: - page_pool_destroy(pp); + page_pool_destroy(qt->sub0.page_pool); return PTR_ERR(pp); } From e7d3e2f46dd5a69046e6d95a0f189155a5516b93 Mon Sep 17 00:00:00 2001 From: "Chang S. Bae" Date: Wed, 16 Sep 2026 22:59:39 +0000 Subject: [PATCH 331/592] x86/microcode/intel: Reject problematic loading on Granite Rapids systems Microcode updates can usually jump revisions. However, there is an erratum on Granite Rapids systems. If they "jump over" revision 0x1000405, they result in an #MC. Avoid it. Signed-off-by: Chang S. Bae Signed-off-by: Borislav Petkov (AMD) Reviewed-by: Dave Hansen Cc: stable@vger.kernel.org Link: https://patch.msgid.link/20260916225939.1144524-1-chang.seok.bae@intel.com --- arch/x86/kernel/cpu/microcode/intel.c | 26 ++++++++++++++++++++++++++ 1 file changed, 26 insertions(+) diff --git a/arch/x86/kernel/cpu/microcode/intel.c b/arch/x86/kernel/cpu/microcode/intel.c index 1142183c950c..9f09d388ed20 100644 --- a/arch/x86/kernel/cpu/microcode/intel.c +++ b/arch/x86/kernel/cpu/microcode/intel.c @@ -309,6 +309,26 @@ static void save_microcode_patch(struct microcode_intel *patch) pr_err("Unable to allocate microcode memory size: %u\n", size); } +static bool revision_is_safe(struct cpu_signature *sig, u32 rev) +{ + u32 vfm = IFM(x86_family(sig->sig), x86_model(sig->sig)); + + /* + * Erratum GNR98 can cause #MCs if "jumping over" revision 0x1000405. + * Avoid the jumps. + */ + if (vfm == INTEL_GRANITERAPIDS_X && + x86_stepping(sig->sig) == 1 && + sig->pf & 0x95 && + sig->rev < 0x1000405 && + rev > 0x1000405) { + pr_err_once("Erratum GNR98: skipping revision 0x%x.\n", rev); + return false; + } + + return true; +} + /* Scan blob for microcode matching the boot CPUs family, model, stepping */ static __init struct microcode_intel *scan_microcode(void *data, size_t size, struct ucode_cpu_info *uci, @@ -330,6 +350,9 @@ static __init struct microcode_intel *scan_microcode(void *data, size_t size, if (!intel_find_matching_signature(data, &uci->cpu_sig)) continue; + if (!revision_is_safe(&uci->cpu_sig, mc_header->rev)) + continue; + /* * For saving the early microcode, find the matching revision which * was loaded on the BSP. @@ -878,6 +901,9 @@ static enum ucode_state parse_microcode_blobs(int cpu, struct iov_iter *iter) if (!intel_find_matching_signature(mc, &uci->cpu_sig)) continue; + if (!revision_is_safe(&uci->cpu_sig, mc_header.rev)) + continue; + is_safe = ucode_validate_minrev(&mc_header); if (force_minrev && !is_safe) continue; From 0a5f5d9e94dead312d32c366b917c64e552b72f7 Mon Sep 17 00:00:00 2001 From: Jamal Hadi Salim Date: Wed, 16 Sep 2026 06:01:14 -0400 Subject: [PATCH 332/592] net/sched: cls_u32: fix manual hash table handle IDR aliasing A u32 hash table created with an explicit handle ('tc filter add ... handle 801: u32 divisor N') keys its IDR entry on the raw handle, while the destroy paths free it under handle2id(handle). The two key domains disagree for handles in the 0x800..0xFFF htid range: handle2id() folds them back into the auto-allocated id space (1..0x7FF). A manual table therefore leaves its raw-keyed IDR entry unreachable on delete (a permanent leak), and its delete can drop the idr entry of an unrelated live auto table. A later auto allocation can then hand out a handle that aliases the live manual table; u32_lookup_ht() first-match routes lookups and TCA_U32_LINK for that htid to the wrong table. Key the divisor-path alloc on handle2id(handle) so allocation and removal share one key domain. A manual handle that maps onto an id already in use is rejected with -ENOSPC, and auto allocation skips ids held by live manual tables. Conditions to recreate: ip link add test0 type dummy tc qdisc add dev test0 clsact tc filter add dev test0 ingress protocol ip pref 1 \ handle 801: u32 divisor 16 tc filter add dev test0 ingress protocol ip pref 2 u32 divisor 16 tc -d filter show dev test0 ingress | grep 'fh 801:' # unpatched: two live tables with handle 0x80100000 (the pref 2 root # hnode is auto-allocated id 1); patched: the auto hnode takes id 2. Also tested with a poc with a live u32 table on the block, add/delete a manual table 'handle 901: u32 divisor 1' twice; unpatched, the re-add fails with -ENOSPC because the raw key leaked on the first delete. Fixes: 73af53d82076 ("net: sched: cls_u32: Fix u32's systematic failure to free IDR entries for hnodes.") Reported-by: Sashiko (gemini + nipa) Closes: https://sashiko.dev/#/patchset/20260822222049.114526-1-jhs@mojatatu.com Reviewed-by: Victor Nogueira Tested-by: hybris Signed-off-by: Jamal Hadi Salim Reviewed-by: Simon Horman Link: https://patch.msgid.link/QDISC-LQFE.v1.20260911041746.1@mojatatu.com Signed-off-by: Jakub Kicinski --- net/sched/cls_u32.c | 12 ++++++++++-- 1 file changed, 10 insertions(+), 2 deletions(-) diff --git a/net/sched/cls_u32.c b/net/sched/cls_u32.c index a3e65c8cf29e..76ce2d124079 100644 --- a/net/sched/cls_u32.c +++ b/net/sched/cls_u32.c @@ -1003,8 +1003,16 @@ static int u32_change(struct net *net, struct sk_buff *in_skb, return -ENOMEM; } } else { - err = idr_alloc_u32(&tp_c->handle_idr, ht, &handle, - handle, GFP_KERNEL); + /* The IDR is keyed on the mapped id, and that is + * what the destroy paths remove. Ask for it here, + * so a manual handle colliding with the + * auto-allocated id space is rejected (-ENOSPC) + * instead of aliasing a future auto id. + */ + u32 id = handle2id(handle); + + err = idr_alloc_u32(&tp_c->handle_idr, ht, &id, id, + GFP_KERNEL); if (err) { kfree(ht); return err; From 960ab631f3d8789586db71b9914b361059ddcb6e Mon Sep 17 00:00:00 2001 From: Jamal Hadi Salim Date: Wed, 16 Sep 2026 06:01:15 -0400 Subject: [PATCH 333/592] selftests/tc-testing: add u32 manual table handle IDR tests 35fc: create a manual table with handle 801:, then add an auto-allocated table. Before the fix, the auto allocation reuses id 1 and hands out the same handle 0x80100000, aliasing the manual table; the test requires the manual 801: handle to keep exactly one entry in the dump. a6e8: with a live u32 table keeping the tc_u_common alive, add and delete a manual table with handle 901:, then re-add it. Unpatched, the delete leaks the raw-keyed IDR entry and the re-add fails with -ENOSPC; the test requires the re-add to succeed. Reviewed-by: Victor Nogueira Tested-by: hybris Signed-off-by: Jamal Hadi Salim Reviewed-by: Simon Horman Link: https://patch.msgid.link/QDISC-LQFE.v1.20260911041746.2@mojatatu.com Signed-off-by: Jakub Kicinski --- .../tc-testing/tc-tests/filters/u32.json | 48 +++++++++++++++++++ 1 file changed, 48 insertions(+) diff --git a/tools/testing/selftests/tc-testing/tc-tests/filters/u32.json b/tools/testing/selftests/tc-testing/tc-tests/filters/u32.json index e2b03f2b5e89..edc5148a8d97 100644 --- a/tools/testing/selftests/tc-testing/tc-tests/filters/u32.json +++ b/tools/testing/selftests/tc-testing/tc-tests/filters/u32.json @@ -376,5 +376,53 @@ "teardown": [ "$TC qdisc del dev $DUMMY clsact" ] + }, + { + "id": "35fc", + "name": "u32 manual table then auto table: auto allocation must not alias a live manual handle", + "category": [ + "filter", + "u32" + ], + "plugins": { + "requires": "nsPlugin" + }, + "setup": [ + "$TC qdisc add dev $DEV1 ingress", + "$TC filter add dev $DEV1 ingress protocol ip pref 1 handle 801: u32 divisor 16" + ], + "cmdUnderTest": "$TC filter add dev $DEV1 ingress protocol ip pref 2 u32 divisor 16", + "expExitCode": "0", + "verifyCmd": "$TC -d filter show dev $DEV1 ingress", + "matchPattern": "fh 801:", + "matchCount": "1", + "teardown": [ + "$TC qdisc del dev $DEV1 ingress" + ] + }, + { + "id": "a6e8", + "name": "u32 manual table add/del does not leak its idr entry (re-adding the same handle succeeds)", + "category": [ + "filter", + "u32" + ], + "plugins": { + "requires": "nsPlugin" + }, + "setup": [ + "$TC qdisc add dev $DEV1 ingress", + "$TC filter add dev $DEV1 ingress protocol ip pref 1 u32 divisor 16", + "$TC filter add dev $DEV1 ingress protocol ip pref 5 handle 901: u32 divisor 1", + "$TC filter del dev $DEV1 ingress protocol ip pref 5 handle 901: u32" + ], + "cmdUnderTest": "$TC filter add dev $DEV1 ingress protocol ip pref 6 handle 901: u32 divisor 1", + "expExitCode": "0", + "verifyCmd": "$TC -d filter show dev $DEV1 ingress", + "matchPattern": "fh 901: ht divisor 1", + "matchCount": "1", + "teardown": [ + "$TC qdisc del dev $DEV1 ingress" + ] } ] From daf677c2c6449011ee695d55b48b5b2977a36f88 Mon Sep 17 00:00:00 2001 From: Kyle Hendry Date: Tue, 15 Sep 2026 10:39:20 -0700 Subject: [PATCH 334/592] net: pcs: rzn1-miic: Fix config array initialization Fix memset parameters to initialize the entire DT value array Fixes: f39e968dc168a7bd ("net: pcs: rzn1-miic: Move configuration data to SoC-specific struct") Reviewed-by: Geert Uytterhoeven Signed-off-by: Kyle Hendry Reviewed-by: Lad Prabhakar Link: https://patch.msgid.link/20260915-rzn1-miic-fix-array-v5-1-b7173fd5b97d@reliablecontrols.com Signed-off-by: Jakub Kicinski --- drivers/net/pcs/pcs-rzn1-miic.c | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-) diff --git a/drivers/net/pcs/pcs-rzn1-miic.c b/drivers/net/pcs/pcs-rzn1-miic.c index 2b72fa98ddf1..cb74861e823c 100644 --- a/drivers/net/pcs/pcs-rzn1-miic.c +++ b/drivers/net/pcs/pcs-rzn1-miic.c @@ -683,7 +683,8 @@ static int miic_parse_dt(struct miic *miic, u32 *mode_cfg) if (!dt_val) return -ENOMEM; - memset(dt_val, MIIC_MODCTRL_CONF_NONE, sizeof(*dt_val)); + memset(dt_val, MIIC_MODCTRL_CONF_NONE, + sizeof(*dt_val) * miic->of_data->conf_conv_count); if (of_property_read_u32(np, "renesas,miic-switch-portin", &conf) == 0) dt_val[0] = conf; From 39c6580765dad6477fb2637f6f616e0d276aae65 Mon Sep 17 00:00:00 2001 From: Heyang Tan Date: Mon, 14 Sep 2026 10:05:21 +0800 Subject: [PATCH 335/592] octeontx2-af: use seq_file for rsrc_alloc debugfs The rsrc_alloc debugfs reader writes rows directly to userspace without respecting the caller's read count. It also uses the current row length as the userspace stride, which can corrupt output when rows have different widths. Use seq_file to handle userspace buffer sizes, offsets, and partial reads, and write output columns directly to the seq_file buffer. Fixes: 23205e6d06d4 ("octeontx2-af: Dump current resource provisioning status") Signed-off-by: Heyang Tan Reviewed-by: Ratheesh Kannoth Link: https://patch.msgid.link/20260914020521.146-1-thy15333007817@163.com Signed-off-by: Jakub Kicinski --- .../marvell/octeontx2/af/rvu_debugfs.c | 104 ++++++------------ 1 file changed, 33 insertions(+), 71 deletions(-) diff --git a/drivers/net/ethernet/marvell/octeontx2/af/rvu_debugfs.c b/drivers/net/ethernet/marvell/octeontx2/af/rvu_debugfs.c index 904374baae6f..2927633465d9 100644 --- a/drivers/net/ethernet/marvell/octeontx2/af/rvu_debugfs.c +++ b/drivers/net/ethernet/marvell/octeontx2/af/rvu_debugfs.c @@ -714,110 +714,72 @@ static int get_max_column_width(struct rvu *rvu) } /* Dumps current provisioning status of all RVU block LFs */ -static ssize_t rvu_dbg_rsrc_attach_status(struct file *filp, - char __user *buffer, - size_t count, loff_t *ppos) +static int rvu_dbg_rsrc_attach_status(struct seq_file *filp, void *unused) { - int index, off = 0, flag = 0, len = 0, i = 0; - struct rvu *rvu = filp->private_data; - int bytes_not_copied = 0; + struct rvu *rvu = filp->private; + int index, pf, vf, pcifunc; struct rvu_block block; - int pf, vf, pcifunc; - int buf_size = 2048; int lf_str_size; char *lfs; - char *buf; - /* don't allow partial reads */ - if (*ppos != 0) - return 0; - - buf = kzalloc(buf_size, GFP_KERNEL); - if (!buf) - return -ENOMEM; - - /* Get the maximum width of a column */ lf_str_size = get_max_column_width(rvu); + if (lf_str_size < 0) + return lf_str_size; lfs = kzalloc(lf_str_size, GFP_KERNEL); - if (!lfs) { - kfree(buf); + if (!lfs) return -ENOMEM; - } - off += scnprintf(&buf[off], buf_size - 1 - off, "%-*s", lf_str_size, - "pcifunc"); + + seq_printf(filp, "%-*s", lf_str_size, "pcifunc"); for (index = 0; index < BLK_COUNT; index++) - if (strlen(rvu->hw->block[index].name)) { - off += scnprintf(&buf[off], buf_size - 1 - off, - "%-*s", lf_str_size, - rvu->hw->block[index].name); - } + if (strlen(rvu->hw->block[index].name)) + seq_printf(filp, "%-*s", lf_str_size, + rvu->hw->block[index].name); - off += scnprintf(&buf[off], buf_size - 1 - off, "\n"); - bytes_not_copied = copy_to_user(buffer + (i * off), buf, off); - if (bytes_not_copied) - goto out; - - i++; - *ppos += off; + seq_putc(filp, '\n'); for (pf = 0; pf < rvu->hw->total_pfs; pf++) { for (vf = 0; vf <= rvu->hw->total_vfs; vf++) { - off = 0; - flag = 0; pcifunc = rvu_make_pcifunc(rvu->pdev, pf, vf); if (!pcifunc) continue; - if (vf) { - sprintf(lfs, "PF%d:VF%d", pf, vf - 1); - off = scnprintf(&buf[off], - buf_size - 1 - off, - "%-*s", lf_str_size, lfs); - } else { - sprintf(lfs, "PF%d", pf); - off = scnprintf(&buf[off], - buf_size - 1 - off, - "%-*s", lf_str_size, lfs); - } - for (index = 0; index < BLK_COUNT; index++) { block = rvu->hw->block[index]; if (!strlen(block.name)) continue; - len = 0; - lfs[len] = '\0'; + lfs[0] = '\0'; get_lf_str_list(&block, pcifunc, lfs); if (strlen(lfs)) - flag = 1; - - off += scnprintf(&buf[off], buf_size - 1 - off, - "%-*s", lf_str_size, lfs); + break; } - if (flag) { - off += scnprintf(&buf[off], - buf_size - 1 - off, "\n"); - bytes_not_copied = copy_to_user(buffer + - (i * off), - buf, off); - if (bytes_not_copied) - goto out; + if (index == BLK_COUNT) + continue; - i++; - *ppos += off; + if (vf) + sprintf(lfs, "PF%d:VF%d", pf, vf - 1); + else + sprintf(lfs, "PF%d", pf); + seq_printf(filp, "%-*s", lf_str_size, lfs); + + for (index = 0; index < BLK_COUNT; index++) { + block = rvu->hw->block[index]; + if (!strlen(block.name)) + continue; + + lfs[0] = '\0'; + get_lf_str_list(&block, pcifunc, lfs); + seq_printf(filp, "%-*s", lf_str_size, lfs); } + seq_putc(filp, '\n'); } } -out: kfree(lfs); - kfree(buf); - if (bytes_not_copied) - return -EFAULT; - return *ppos; + return 0; } -RVU_DEBUG_FOPS(rsrc_status, rsrc_attach_status, NULL); +RVU_DEBUG_SEQ_FOPS(rsrc_status, rsrc_attach_status, NULL); static int rvu_dbg_rvu_pf_cgx_map_display(struct seq_file *filp, void *unused) { From d09e8f64653c93da5793c16be19330968f2a32e6 Mon Sep 17 00:00:00 2001 From: Shay Drory Date: Tue, 15 Sep 2026 14:34:57 +0300 Subject: [PATCH 336/592] net/mlx5: devcom, Base component size on linked devices mlx5_devcom_comp_get_size() returns the component's kref count. That kref is bumped in mlx5_devcom_register_component() under comp_list_lock, before the comp_dev is linked onto comp_dev_list_head under comp->sem. The event broadcast (mlx5_devcom_locked_send_event()) walks that list. Hence, a caller can read the expected size, but send_event won't be sent to all peers. In the SD group registration path, this lets a member broadcast its role-election event over an incomplete list, electing a primary that never completes the group, is never marked ready, and leaves the group with a stale primary. Track the number of linked comp_devs in a dedicated counter, maintained under comp->sem together with the list add/remove, and return it from mlx5_devcom_comp_get_size(). Fixes: 9bb1ac80738a ("net/mlx5: devcom, Add component size getter") Signed-off-by: Shay Drory Reviewed-by: Akiva Goldberger Signed-off-by: Tariq Toukan Link: https://patch.msgid.link/20260915113459.3934760-2-tariqt@nvidia.com Signed-off-by: Jakub Kicinski --- drivers/net/ethernet/mellanox/mlx5/core/lib/devcom.c | 5 ++++- 1 file changed, 4 insertions(+), 1 deletion(-) diff --git a/drivers/net/ethernet/mellanox/mlx5/core/lib/devcom.c b/drivers/net/ethernet/mellanox/mlx5/core/lib/devcom.c index 64f92427602d..75855481522b 100644 --- a/drivers/net/ethernet/mellanox/mlx5/core/lib/devcom.c +++ b/drivers/net/ethernet/mellanox/mlx5/core/lib/devcom.c @@ -37,6 +37,7 @@ struct mlx5_devcom_comp { struct mlx5_devcom_key key; mlx5_devcom_event_handler_t handler; struct kref ref; + int nr_devs; bool ready; struct rw_semaphore sem; struct lock_class_key lock_key; @@ -170,6 +171,7 @@ devcom_alloc_comp_dev(struct mlx5_devcom_dev *devc, down_write(&comp->sem); list_add_tail(&devcom->list, &comp->comp_dev_list_head); + WRITE_ONCE(comp->nr_devs, comp->nr_devs + 1); up_write(&comp->sem); return devcom; @@ -182,6 +184,7 @@ devcom_free_comp_dev(struct mlx5_devcom_comp_dev *devcom) down_write(&comp->sem); list_del(&devcom->list); + WRITE_ONCE(comp->nr_devs, comp->nr_devs - 1); up_write(&comp->sem); kref_put(&devcom->devc->ref, mlx5_devcom_dev_release); @@ -284,7 +287,7 @@ int mlx5_devcom_comp_get_size(struct mlx5_devcom_comp_dev *devcom) { struct mlx5_devcom_comp *comp = devcom->comp; - return kref_read(&comp->ref); + return READ_ONCE(comp->nr_devs); } int mlx5_devcom_locked_send_event(struct mlx5_devcom_comp_dev *devcom, From e1e29ada2b938b13ba689a06a8bd8604564da2b3 Mon Sep 17 00:00:00 2001 From: Shay Drory Date: Tue, 15 Sep 2026 14:34:58 +0300 Subject: [PATCH 337/592] net/mlx5: SD, unload reps on shared FDB create error path mlx5_lag_shared_fdb_create() sets sd_fdb_active on every group member before reloading the representors, so mlx5_lag_is_active() is already true and the guard in mlx5_esw_offloads_rep_load() does not skip the VF/SF reps. If the reload then fails, the error path clears sd_fdb_active and destroys the shared FDB, leaving the reps loaded while SD LAG is inactive - the state cited commit was written to prevent. Unload the reps in the error path as well. Fixes: 68c2dd59a6c7 ("net/mlx5: E-Switch, Tie rep load/unload to SD LAG state") Signed-off-by: Shay Drory Reviewed-by: Akiva Goldberger Signed-off-by: Tariq Toukan Link: https://patch.msgid.link/20260915113459.3934760-3-tariqt@nvidia.com Signed-off-by: Jakub Kicinski --- drivers/net/ethernet/mellanox/mlx5/core/lag/shared_fdb.c | 1 + 1 file changed, 1 insertion(+) diff --git a/drivers/net/ethernet/mellanox/mlx5/core/lag/shared_fdb.c b/drivers/net/ethernet/mellanox/mlx5/core/lag/shared_fdb.c index 6b4ad3c53f2f..424040918fa3 100644 --- a/drivers/net/ethernet/mellanox/mlx5/core/lag/shared_fdb.c +++ b/drivers/net/ethernet/mellanox/mlx5/core/lag/shared_fdb.c @@ -270,6 +270,7 @@ int mlx5_lag_shared_fdb_create(struct mlx5_lag *ldev, pf->sd_fdb_active = false; } mlx5_lag_destroy_single_fdb_filter(ldev, group_id); + mlx5_lag_unload_reps_from_locked(ldev, filter); } err_add_devices: mlx5_lag_add_devices_filter(ldev, filter); From bae23d1ae62092c7f0ec6d5f7e1be5d164822638 Mon Sep 17 00:00:00 2001 From: Shay Drory Date: Tue, 15 Sep 2026 14:34:59 +0300 Subject: [PATCH 338/592] net/mlx5: LAG, reload IB reps of LAG master before the rest In a shared-FDB LAG the master device creates the bond IB device; the other LAG members do not create their own, they populate a port inside the master's IB device. mlx5_lag_reload_ib_reps_unlocked() reloaded the members' IB reps in iteration order, with no guarantee the master is reloaded first. When a non-master member is reloaded before the master, it tries to populate its port in an IB device that has not been recreated yet. Hence, reload the master's IB reps first, then every other member. Fixes: 2b204cdb1206 ("net/mlx5: LAG, use xa_alloc to manage LAG device indices") Signed-off-by: Shay Drory Reviewed-by: Akiva Goldberger Signed-off-by: Tariq Toukan Link: https://patch.msgid.link/20260915113459.3934760-4-tariqt@nvidia.com Signed-off-by: Jakub Kicinski --- .../net/ethernet/mellanox/mlx5/core/lag/lag.c | 44 ++++++++++++++----- 1 file changed, 32 insertions(+), 12 deletions(-) diff --git a/drivers/net/ethernet/mellanox/mlx5/core/lag/lag.c b/drivers/net/ethernet/mellanox/mlx5/core/lag/lag.c index c655f6e32e9b..dd14cdc378de 100644 --- a/drivers/net/ethernet/mellanox/mlx5/core/lag/lag.c +++ b/drivers/net/ethernet/mellanox/mlx5/core/lag/lag.c @@ -1266,25 +1266,45 @@ void mlx5_lag_remove_devices(struct mlx5_lag *ldev) mlx5_lag_remove_devices_filter(ldev, MLX5_LAG_FILTER_PORTS); } +static int mlx5_lag_reload_ib_reps_idx(struct mlx5_lag *ldev, int idx, + u32 flags) +{ + struct lag_func *pf = mlx5_lag_pf(ldev, idx); + struct mlx5_eswitch *esw; + int ret; + + if (pf->dev->priv.flags & flags) + return 0; + + esw = pf->dev->priv.eswitch; + mlx5_esw_reps_block(esw); + ret = mlx5_eswitch_reload_ib_reps(esw); + mlx5_esw_reps_unblock(esw); + + return ret; +} + static int mlx5_lag_reload_ib_reps_unlocked(struct mlx5_lag *ldev, u32 flags, u32 filter, bool cont_on_fail) { - struct lag_func *pf; + int master_idx = mlx5_lag_get_dev_index_by_seq_filter(ldev, MLX5_LAG_P1, + filter); int ret; int i; - mlx5_lag_for_each(i, 0, ldev, filter) { - pf = mlx5_lag_pf(ldev, i); - if (!(pf->dev->priv.flags & flags)) { - struct mlx5_eswitch *esw; + if (master_idx < 0) + return -EINVAL; - esw = pf->dev->priv.eswitch; - mlx5_esw_reps_block(esw); - ret = mlx5_eswitch_reload_ib_reps(esw); - mlx5_esw_reps_unblock(esw); - if (ret && !cont_on_fail) - return ret; - } + ret = mlx5_lag_reload_ib_reps_idx(ldev, master_idx, flags); + if (ret && !cont_on_fail) + return ret; + + mlx5_lag_for_each(i, 0, ldev, filter) { + if (i == master_idx) + continue; + ret = mlx5_lag_reload_ib_reps_idx(ldev, i, flags); + if (ret && !cont_on_fail) + return ret; } return 0; From 261b61d3735b042ae25634f795c4540be0fc140c Mon Sep 17 00:00:00 2001 From: Kumar Kartikeya Dwivedi Date: Fri, 18 Sep 2026 01:32:09 +0200 Subject: [PATCH 339/592] bpf: Make post-verification instruction rewrites killable After do_check() returns, the verifier runs several instruction rewrite passes. Some of them patch or remove one instruction at a time. Each operation moves the remaining instruction and auxiliary-data arrays and adjusts all branch offsets, making the overall work quadratic in the program length. A privileged loader can submit 131072 unconditional jumps by zero followed by a valid return. Verification finishes quickly, but bpf_opt_remove_nops() then spends a long time removing each jump separately. Since this post-verification work neither checks for signals nor reschedules, a pending SIGKILL cannot terminate the task until the rewrite finishes. Make bpf_patch_insn_data() and verifier_remove_insns() common cancellation and rescheduling points. These helpers run from BPF_PROG_LOAD process context, and bpf_patch_insn_data() can already sleep while reallocating auxiliary data. Report interrupted constant blinding as -EINTR and propagate it through both JIT paths, including kernels that permit interpreter fallback. Other blinding failures retain the existing fallback behavior. This does not reduce the quadratic cost of the rewrite passes, but it makes the work preemptible and allows a killed loader to be torn down promptly. Fixes: 52875a04f4b2 ("bpf: verifier: remove dead code") Reported-by: Nicholas Carlini Suggested-by: Nicholas Carlini Signed-off-by: Kumar Kartikeya Dwivedi Acked-by: Eduard Zingerman Link: https://patch.msgid.link/20260917233222.2542500-2-memxor@gmail.com Signed-off-by: Eduard Zingerman --- kernel/bpf/core.c | 21 +++++++++++++++++---- kernel/bpf/fixups.c | 24 ++++++++++++++++++++++-- 2 files changed, 39 insertions(+), 6 deletions(-) diff --git a/kernel/bpf/core.c b/kernel/bpf/core.c index 8b294dfc1ad4..2e3bf8113ae9 100644 --- a/kernel/bpf/core.c +++ b/kernel/bpf/core.c @@ -19,6 +19,7 @@ #include #include +#include #include #include #include @@ -1619,6 +1620,8 @@ struct bpf_prog *bpf_jit_blind_constants(struct bpf_verifier_env *env, struct bp * fix it up here on error. */ bpf_jit_prog_release_other(prog, clone); + if (env && fatal_signal_pending(current)) + return ERR_PTR(-EINTR); return IS_ERR(tmp) ? tmp : ERR_PTR(-ENOMEM); } @@ -2636,11 +2639,14 @@ static struct bpf_prog *bpf_prog_jit_compile(struct bpf_verifier_env *env, struc orig_prog = prog; prog = bpf_jit_blind_constants(env, prog); /* - * If blinding was requested and we failed during blinding, we must fall - * back to the interpreter. + * Fall back to the interpreter after blinding failures, except when + * the loader was killed. */ - if (IS_ERR(prog)) + if (IS_ERR(prog)) { + if (PTR_ERR(prog) == -EINTR) + return prog; goto out_restore; + } prog = bpf_int_jit_compile(env, prog); if (prog->jited) { @@ -2659,6 +2665,8 @@ static struct bpf_prog *bpf_prog_jit_compile(struct bpf_verifier_env *env, struc struct bpf_prog *__bpf_prog_select_runtime(struct bpf_verifier_env *env, struct bpf_prog *fp, int *err) { + struct bpf_prog *jit_prog; + /* In case of BPF to BPF calls, verifier did all the prep * work with regards to JITing, etc. */ @@ -2681,7 +2689,12 @@ struct bpf_prog *__bpf_prog_select_runtime(struct bpf_verifier_env *env, struct if (*err) return fp; - fp = bpf_prog_jit_compile(env, fp); + jit_prog = bpf_prog_jit_compile(env, fp); + if (IS_ERR(jit_prog)) { + *err = PTR_ERR(jit_prog); + return fp; + } + fp = jit_prog; bpf_prog_jit_attempt_done(fp); if (!fp->jited && jit_needed) { *err = -ENOTSUPP; diff --git a/kernel/bpf/fixups.c b/kernel/bpf/fixups.c index 52d3cec33672..d6f83521fc78 100644 --- a/kernel/bpf/fixups.c +++ b/kernel/bpf/fixups.c @@ -8,6 +8,7 @@ #include #include #include +#include #include #include "disasm.h" @@ -306,12 +307,28 @@ static void adjust_poke_descs(struct bpf_prog *prog, u32 off, u32 len) } } +/* + * Some post-verification instruction rewriting passes require an + * O(prog->len) operation per instruction. Keep their shared primitives + * killable and preemptible. + */ +static bool bpf_rewrite_must_abort(void) +{ + if (fatal_signal_pending(current)) + return true; + cond_resched(); + return false; +} + struct bpf_prog *bpf_patch_insn_data(struct bpf_verifier_env *env, u32 off, const struct bpf_insn *patch, u32 len) { struct bpf_prog *new_prog; struct bpf_insn_aux_data *new_data = NULL; + if (bpf_rewrite_must_abort()) + return NULL; + if (len > 1) { new_data = vrealloc(env->insn_aux_data, array_size(env->prog->len + len - 1, @@ -523,6 +540,9 @@ static int verifier_remove_insns(struct bpf_verifier_env *env, u32 off, u32 cnt) unsigned int orig_prog_len = env->prog->len; int err; + if (bpf_rewrite_must_abort()) + return -EINTR; + if (bpf_prog_is_offloaded(env->prog->aux)) bpf_prog_offload_remove_insns(env, off, cnt); @@ -1356,7 +1376,7 @@ int bpf_jit_subprogs(struct bpf_verifier_env *env) } prog = bpf_jit_blind_constants(env, prog); if (IS_ERR(prog)) { - err = -ENOMEM; + err = PTR_ERR(prog); prog = orig_prog; goto out_restore; } @@ -1433,7 +1453,7 @@ int bpf_fixup_call_args(struct bpf_verifier_env *env) err = bpf_jit_subprogs(env); if (err == 0) return 0; - if (err == -EFAULT) + if (err == -EFAULT || err == -EINTR) return err; } #ifndef CONFIG_BPF_JIT_ALWAYS_ON From fd16449a9b3b31a8f18944c2f0e29e4e218ca2cf Mon Sep 17 00:00:00 2001 From: Kumar Kartikeya Dwivedi Date: Fri, 18 Sep 2026 01:32:10 +0200 Subject: [PATCH 340/592] bpf: Preserve packet pointer class displacement in regsafe() regsafe() maps packet pointer IDs between states and checks that each current register range is a subset of the corresponding explored register range. It does not, however, preserve the displacement between registers that share a packet pointer ID. This is unsound because packet range is shared by ID. A bounds check on one class member updates every member, and a later access can consume the range through another member. Commit 022ac0750883 ("bpf: use reg->var_off instead of reg->off for pointers") folded the fixed pointer offset into r64 and removed the old off equality check, so two individually narrower registers can prune even when their displacement has changed. The explored path can then license an out-of-bounds packet access on the pruned path. Require matching range bases for packet pointers with an ID. Together with the existing ID mapping, this preserves the displacement between members of each packet-pointer class without adding per-ID state. Packet pointers without an ID remain unaffected. Fixes: 022ac0750883 ("bpf: use reg->var_off instead of reg->off for pointers") Reported-by: Nicholas Carlini Suggested-by: Nicholas Carlini Signed-off-by: Kumar Kartikeya Dwivedi Acked-by: Eduard Zingerman Link: https://patch.msgid.link/20260917233222.2542500-3-memxor@gmail.com Signed-off-by: Eduard Zingerman --- kernel/bpf/states.c | 3 +++ 1 file changed, 3 insertions(+) diff --git a/kernel/bpf/states.c b/kernel/bpf/states.c index 66fb11b6c6a7..6c88ad95b63b 100644 --- a/kernel/bpf/states.c +++ b/kernel/bpf/states.c @@ -635,6 +635,9 @@ static bool regsafe(struct bpf_verifier_env *env, struct bpf_reg_state *rold, /* id relations must be preserved */ if (!check_ids(rold->id, rcur->id, idmap)) return false; + /* Preserve displacements between pointers sharing an ID. */ + if (rold->id && rold->r64.base != rcur->r64.base) + return false; /* new val must satisfy old val knowledge */ return range_within(rold, rcur) && tnum_in(rold->var_off, rcur->var_off); From 2059d9af54f0d66deb237aa75d2ed28648df05a1 Mon Sep 17 00:00:00 2001 From: Kumar Kartikeya Dwivedi Date: Fri, 18 Sep 2026 01:32:11 +0200 Subject: [PATCH 341/592] selftests/bpf: Test packet pointer class displacement pruning Add two paths whose packet pointer ranges are individually compatible at a join but whose members have different relative displacements. The first path proves an eight-byte access through one member. On the second path, the same guard only proves that the access starts before data_end. An affected verifier prunes the second path and accepts the program. With packet pointer class displacement preserved, it explores that path and rejects the out-of-bounds access. Read the unknown offset and branch selector directly from XDP context fields, and force state checkpoints so the pruning attempt does not depend on the verifier checkpoint heuristics. Signed-off-by: Kumar Kartikeya Dwivedi Link: https://patch.msgid.link/20260917233222.2542500-4-memxor@gmail.com Signed-off-by: Eduard Zingerman --- .../progs/verifier_xdp_direct_packet_access.c | 35 +++++++++++++++++++ 1 file changed, 35 insertions(+) diff --git a/tools/testing/selftests/bpf/progs/verifier_xdp_direct_packet_access.c b/tools/testing/selftests/bpf/progs/verifier_xdp_direct_packet_access.c index 0b86d95a4133..9866bc154194 100644 --- a/tools/testing/selftests/bpf/progs/verifier_xdp_direct_packet_access.c +++ b/tools/testing/selftests/bpf/progs/verifier_xdp_direct_packet_access.c @@ -1719,4 +1719,39 @@ l0_%=: r0 = 0; \ : __clobber_all); } +SEC("xdp") +__description("XDP pkt regsafe preserves packet pointer class displacement") +__failure __msg("R2 min value is outside of the allowed memory range") +__flag(BPF_F_ANY_ALIGNMENT) __flag(BPF_F_TEST_STATE_FREQ) +__naked void pkt_regsafe_class_displacement(void) +{ + asm volatile (" \ + r8 = *(u32 *)(r1 + %[xdp_md_data_end]); \ + r9 = *(u32 *)(r1 + %[xdp_md_data]); \ + r4 = *(u32 *)(r1 + %[xdp_md_rx_queue_index]); \ + r4 &= 15; \ + r0 = *(u32 *)(r1 + %[xdp_md_ingress_ifindex]); \ + if r0 != 0 goto l0_%=; \ + r2 = r9; \ + r2 += r4; \ + r3 = r2; \ + r3 += 8; \ + goto l1_%=; \ +l0_%=: r4 &= 3; \ + r4 += 8; \ + r2 = r9; \ + r2 += r4; \ + r3 = r2; \ +l1_%=: if r3 > r8 goto l2_%=; \ + r0 = *(u64 *)(r2 + 0); \ +l2_%=: r0 = 0; \ + exit; \ +" : + : __imm_const(xdp_md_data, offsetof(struct xdp_md, data)), + __imm_const(xdp_md_data_end, offsetof(struct xdp_md, data_end)), + __imm_const(xdp_md_rx_queue_index, offsetof(struct xdp_md, rx_queue_index)), + __imm_const(xdp_md_ingress_ifindex, offsetof(struct xdp_md, ingress_ifindex)) + : __clobber_all); +} + char _license[] SEC("license") = "GPL"; From c26e97721b172163042b98572fead234797f2c3c Mon Sep 17 00:00:00 2001 From: Kumar Kartikeya Dwivedi Date: Fri, 18 Sep 2026 01:32:12 +0200 Subject: [PATCH 342/592] bpf: Apply CO-RE relocations before subprogram validation check_subprogs() verifies that each subprogram ends in an exit or an unconditional jump before in-kernel CO-RE relocations are applied. An unresolved relocation can then replace that terminal instruction with an invalid helper call. The resulting fall-through into another subprogram breaks the CFG invariant used by postorder and stack liveness analysis, which can write past their per-subprogram arrays. Apply CO-RE relocations immediately after preparing the program BTF, before subprogram discovery and validation. Keep func_info and line_info validation after subprogram discovery because those records depend on the complete subprogram layout. Reject an ldimm64 first slot at the end of the instruction stream before CO-RE can inspect its missing second slot. check_subprogs() previously rejected this form before relocation processing because it is not a valid subprogram terminator. Moving CO-RE ahead of check_subprogs() removes that implicit protection, so perform an explicit check before applying relocations. Include core_relo_cnt when deciding whether to prepare program BTF. A load that supplied only CO-RE relocation metadata previously skipped both BTF setup and relocation processing. Fixes: fbd94c7afcf9 ("bpf: Pass a set of bpf_core_relo-s to prog_load command.") Suggested-by: Andrii Nakryiko Suggested-by: Alexei Starovoitov Suggested-by: Eduard Zingerman Signed-off-by: Kumar Kartikeya Dwivedi Acked-by: Eduard Zingerman Link: https://patch.msgid.link/20260917233222.2542500-5-memxor@gmail.com Signed-off-by: Eduard Zingerman --- include/linux/bpf_verifier.h | 2 ++ kernel/bpf/check_btf.c | 12 ++++-------- kernel/bpf/verifier.c | 12 +++++++++++- 3 files changed, 17 insertions(+), 9 deletions(-) diff --git a/include/linux/bpf_verifier.h b/include/linux/bpf_verifier.h index 36b65797877d..bba5a727c651 100644 --- a/include/linux/bpf_verifier.h +++ b/include/linux/bpf_verifier.h @@ -1197,6 +1197,8 @@ static inline void bpf_trampoline_unpack_key(u64 key, u32 *obj_id, u32 *btf_id) int bpf_prepare_btf_info(struct bpf_verifier_env *env, const union bpf_attr *attr, bpfptr_t uattr); +int bpf_check_core_relo(struct bpf_verifier_env *env, + const union bpf_attr *attr, bpfptr_t uattr); int bpf_check_btf_info(struct bpf_verifier_env *env, const union bpf_attr *attr, bpfptr_t uattr); diff --git a/kernel/bpf/check_btf.c b/kernel/bpf/check_btf.c index 0e8b3ccc7a5b..4c1ed842f661 100644 --- a/kernel/bpf/check_btf.c +++ b/kernel/bpf/check_btf.c @@ -338,9 +338,9 @@ static int check_btf_line(struct bpf_verifier_env *env, #define MIN_CORE_RELO_SIZE sizeof(struct bpf_core_relo) #define MAX_CORE_RELO_SIZE MAX_FUNCINFO_REC_SIZE -static int check_core_relo(struct bpf_verifier_env *env, - const union bpf_attr *attr, - bpfptr_t uattr) +int bpf_check_core_relo(struct bpf_verifier_env *env, + const union bpf_attr *attr, + bpfptr_t uattr) { u32 i, nr_core_relo, ncopy, expected_size, rec_size; struct bpf_core_relo core_relo = {}; @@ -414,7 +414,7 @@ int bpf_prepare_btf_info(struct bpf_verifier_env *env, struct btf *btf; int err; - if (!attr->func_info_cnt && !attr->line_info_cnt) { + if (!attr->func_info_cnt && !attr->line_info_cnt && !attr->core_relo_cnt) { if (check_abnormal_return(env)) return -EINVAL; return 0; @@ -455,9 +455,5 @@ int bpf_check_btf_info(struct bpf_verifier_env *env, if (err) return err; - err = check_core_relo(env, attr, uattr); - if (err) - return err; - return 0; } diff --git a/kernel/bpf/verifier.c b/kernel/bpf/verifier.c index 5d7080c260d8..33161dc64568 100644 --- a/kernel/bpf/verifier.c +++ b/kernel/bpf/verifier.c @@ -21199,6 +21199,11 @@ int bpf_check(struct bpf_prog **prog, union bpf_attr *attr, bpfptr_t uattr, ret = bpf_diag_init(env); if (ret) goto err_prep; + if (env->prog->insnsi[env->prog->len - 1].code == (BPF_LD | BPF_IMM | BPF_DW)) { + verbose(env, "invalid bpf_ld_imm64 insn\n"); + ret = -EINVAL; + goto err_prep; + } if (env->signature) { ret = bpf_prog_calc_tag(env->prog); if (ret < 0) @@ -21274,6 +21279,11 @@ int bpf_check(struct bpf_prog **prog, union bpf_attr *attr, bpfptr_t uattr, if (ret < 0) goto skip_full_check; + /* Apply CO-RE before validating the program's instruction layout. */ + ret = bpf_check_core_relo(env, attr, uattr); + if (ret < 0) + goto skip_full_check; + /* Discover all subprograms before validating their layout and BTF. */ ret = add_subprogs(env); if (ret < 0) @@ -21283,7 +21293,7 @@ int bpf_check(struct bpf_prog **prog, union bpf_attr *attr, bpfptr_t uattr, if (ret < 0) goto skip_full_check; - /* Validate BTF against the complete subprogram layout and apply CO-RE. */ + /* Validate BTF against the complete subprogram layout. */ ret = bpf_check_btf_info(env, attr, uattr); if (ret < 0) goto skip_full_check; From 968ee7c06b62f1ac4597c351a45c2219a678a458 Mon Sep 17 00:00:00 2001 From: Kumar Kartikeya Dwivedi Date: Fri, 18 Sep 2026 01:32:13 +0200 Subject: [PATCH 343/592] selftests/bpf: Test early in-kernel CO-RE relocation Add a raw program load with CO-RE relocation metadata but no func_info or line_info. Place the relocation in dead code and require the poisoning log, proving that the kernel processes standalone CO-RE metadata instead of silently skipping it. Also give a subprogram a relocatable immediate as its terminal instruction. Require the relocation's poisoning log before check_subprogs() rejects the resulting fall-through. With the old ordering, check_subprogs() rejects the original terminal instruction before CO-RE can emit the substitution log, so the test continues to distinguish the ordering after relocation target validation is tightened. Submit a trailing ldimm64 first slot with CO-RE metadata and require the early structural diagnostic. This exercises the check that protects relocation processing instead of the later regular instruction validation. Load the standalone instruction stream without relocation metadata first to ensure that CO-RE processing causes its poisoning diagnostic. Encode the fixed BTF metadata directly with the selftest BTF helpers. Suggested-by: Eduard Zingerman Signed-off-by: Kumar Kartikeya Dwivedi Acked-by: Eduard Zingerman Link: https://patch.msgid.link/20260917233222.2542500-6-memxor@gmail.com Signed-off-by: Eduard Zingerman --- .../selftests/bpf/prog_tests/core_reloc_raw.c | 133 ++++++++++++++++++ 1 file changed, 133 insertions(+) diff --git a/tools/testing/selftests/bpf/prog_tests/core_reloc_raw.c b/tools/testing/selftests/bpf/prog_tests/core_reloc_raw.c index a18d3680fb16..bb19e49dd87d 100644 --- a/tools/testing/selftests/bpf/prog_tests/core_reloc_raw.c +++ b/tools/testing/selftests/bpf/prog_tests/core_reloc_raw.c @@ -14,6 +14,138 @@ static char log[16 * 1024]; +static int load_core_relo_insns(int btf_fd, struct bpf_insn *insns, int insn_cnt, + struct bpf_func_info *funcs, int func_cnt, + int enum_id, int access_str_off, int insn_idx, + bool relocate) +{ + struct bpf_core_relo relo = { + .insn_off = insn_idx * sizeof(struct bpf_insn), + .type_id = enum_id, + .access_str_off = access_str_off, + .kind = BPF_CORE_ENUMVAL_VALUE, + }; + union bpf_attr attr = { + .prog_type = BPF_PROG_TYPE_SOCKET_FILTER, + .insn_cnt = insn_cnt, + .insns = (__u64)insns, + .license = (__u64)"GPL", + .log_buf = (__u64)log, + .log_size = sizeof(log), + .log_level = 2, + .prog_btf_fd = btf_fd, + .func_info_rec_size = sizeof(struct bpf_func_info), + .func_info = (__u64)funcs, + .func_info_cnt = func_cnt, + }; + + if (relocate) { + attr.core_relo_cnt = 1; + attr.core_relos = (__u64)&relo; + attr.core_relo_rec_size = sizeof(relo); + } + memset(log, 0, sizeof(log)); + return sys_bpf_prog_load(&attr, sizeof(attr), 1); +} + +static void test_early_core_relo(void) +{ + struct test_btf { + struct btf_header hdr; + __u32 types[18]; + char strings[64]; + } raw_btf = { + .hdr = { + .magic = BTF_MAGIC, + .version = BTF_VERSION, + .hdr_len = sizeof(struct btf_header), + .type_off = 0, + .type_len = sizeof(raw_btf.types), + .str_off = offsetof(struct test_btf, strings) - + offsetof(struct test_btf, types), + .str_len = sizeof(raw_btf.strings), + }, + .types = { + BTF_TYPE_INT_ENC(1, BTF_INT_SIGNED, 0, 32, 4), /* [1] int */ + BTF_FUNC_PROTO_ENC(1, 0), /* [2] int (*)(void) */ + BTF_FUNC_ENC(5, 2), /* [3] main_fn */ + BTF_FUNC_ENC(13, 2), /* [4] sub_fn */ + BTF_TYPE_ENC(20, BTF_INFO_ENC(BTF_KIND_ENUM, 0, 1), 4), /* [5] enum */ + BTF_ENUM_ENC(45, 0), /* value = 0 */ + }, + .strings = "\0int\0main_fn\0sub_fn\0core_relo_poison_missing\0value\0" "0", + }; + struct bpf_func_info funcs[] = { + { .insn_off = 0, .type_id = 3 }, + { .insn_off = 3, .type_id = 4 }, + }; + struct bpf_insn core_only[] = { + BPF_MOV64_IMM(BPF_REG_0, 0), + BPF_JMP_IMM(BPF_JEQ, BPF_REG_0, 0, 1), + BPF_MOV64_IMM(BPF_REG_0, 0), + BPF_EXIT_INSN(), + }; + struct bpf_insn subprog[] = { + BPF_CALL_REL(2), + BPF_MOV64_IMM(BPF_REG_0, 0), + BPF_MOV64_IMM(BPF_REG_0, 0), + BPF_MOV64_IMM(BPF_REG_0, 0), + BPF_EXIT_INSN(), + }; + struct bpf_insn truncated_ldimm64[] = { + BPF_RAW_INSN(BPF_LD | BPF_IMM | BPF_DW, 0, 0, 0, 0), + }; + int access_str_off = 51; /* offset of "0" */ + int enum_id = 5; + int btf_fd, prog_fd = -1; + + btf_fd = bpf_btf_load(&raw_btf, sizeof(raw_btf), NULL); + if (!ASSERT_GE(btf_fd, 0, "btf_load")) + goto cleanup; + + if (test__start_subtest("without_func_info")) { + prog_fd = load_core_relo_insns(btf_fd, core_only, ARRAY_SIZE(core_only), NULL, 0, + enum_id, access_str_off, 2, false); + if (!ASSERT_GE(prog_fd, 0, "control_load")) + goto cleanup; + close(prog_fd); + prog_fd = load_core_relo_insns(btf_fd, core_only, ARRAY_SIZE(core_only), NULL, 0, + enum_id, access_str_off, 2, true); + if (!ASSERT_GE(prog_fd, 0, "poisoned_load")) + goto cleanup; + ASSERT_HAS_SUBSTR(log, "substituting insn #2", "poison_log"); + close(prog_fd); + prog_fd = -1; + } + + if (test__start_subtest("before_subprog_validation")) { + prog_fd = load_core_relo_insns(btf_fd, subprog, ARRAY_SIZE(subprog), funcs, 2, + enum_id, access_str_off, 2, true); + if (!ASSERT_LT(prog_fd, 0, "poisoned_load")) + goto cleanup; + ASSERT_HAS_SUBSTR(log, "substituting insn #2", "poison_log"); + ASSERT_HAS_SUBSTR(log, "last insn is not an exit or jmp", "poisoned_load_log"); + } + + if (test__start_subtest("truncated_ldimm64")) { + prog_fd = load_core_relo_insns(btf_fd, truncated_ldimm64, + ARRAY_SIZE(truncated_ldimm64), NULL, 0, + enum_id, access_str_off, 0, true); + if (!ASSERT_LT(prog_fd, 0, "truncated_load")) + goto cleanup; + ASSERT_HAS_SUBSTR(log, "invalid bpf_ld_imm64 insn", "truncated_load_log"); + } + +cleanup: + if (env.verbosity > VERBOSE_NORMAL && log[0]) { + printf("-------- program load log start --------\n"); + printf("%s", log); + printf("-------- program load log end ----------\n"); + } + close(prog_fd); + close(btf_fd); +} + /* Check that verifier rejects BPF program containing relocation * pointing to non-existent BTF type. */ @@ -120,6 +252,7 @@ static void test_bad_local_id(void) void test_core_reloc_raw(void) { + test_early_core_relo(); if (test__start_subtest("bad_local_id")) test_bad_local_id(); } From 394ae398337c5f87e567f6cd63b937fc2b2f6ddc Mon Sep 17 00:00:00 2001 From: Kumar Kartikeya Dwivedi Date: Fri, 18 Sep 2026 01:32:14 +0200 Subject: [PATCH 344/592] bpf: Restrict CO-RE poisoning to relocatable instructions CO-RE relocation records can name any instruction offset. When a relocation cannot be resolved, bpf_core_patch_insn() currently poisons its target before checking whether that instruction is a valid relocation target. Malformed metadata can therefore replace jumps, calls, exits, register-source arithmetic, or non-immediate loads instead of failing at the relocation step. Handle poisoning only after the instruction has passed the same class and operand-form checks used for a resolved relocation. Route invalid forms through the existing diagnostic and return a hard error. Keep poisoning supported instructions, including both halves of a plain ldimm64, so an unresolved relocation in dead code remains valid. Extend bpf_core_poison_insn() to poison both halves of ldimm64, and return its status directly from each validated instruction case. This avoids routing the success path through a common label and leaves the helper free to report errors. The shared relocation code applies this restriction to both libbpf and in-kernel CO-RE. Fixes: d7a252708dbc ("libbpf: Improve handling of failed CO-RE relocations") Reported-by: Nicholas Carlini Suggested-by: Nicholas Carlini Signed-off-by: Kumar Kartikeya Dwivedi Acked-by: Eduard Zingerman Link: https://patch.msgid.link/20260917233222.2542500-7-memxor@gmail.com Signed-off-by: Eduard Zingerman --- tools/lib/bpf/relo_core.c | 58 +++++++++++++++++++++------------------ 1 file changed, 31 insertions(+), 27 deletions(-) diff --git a/tools/lib/bpf/relo_core.c b/tools/lib/bpf/relo_core.c index 8ad2715721cf..2672623a4198 100644 --- a/tools/lib/bpf/relo_core.c +++ b/tools/lib/bpf/relo_core.c @@ -980,23 +980,30 @@ static int bpf_core_calc_relo(const char *prog_name, } /* - * Turn instruction for which CO_RE relocation failed into invalid one with + * Turn instruction for which CO-RE relocation failed into invalid one with * distinct signature. */ -static void bpf_core_poison_insn(const char *prog_name, int relo_idx, - int insn_idx, struct bpf_insn *insn) +static int bpf_core_poison_insn(const char *prog_name, int relo_idx, + struct bpf_insn *insn, int insn_idx) { - pr_debug("prog '%s': relo #%d: substituting insn #%d w/ invalid insn\n", - prog_name, relo_idx, insn_idx); - insn->code = BPF_JMP | BPF_CALL; - insn->dst_reg = 0; - insn->src_reg = 0; - insn->off = 0; - /* if this instruction is reachable (not a dead code), - * verifier will complain with the following message: - * invalid func unknown#195896080 - */ - insn->imm = 195896080; /* => 0xbad2310 => "bad relo" */ + int insn_cnt = is_ldimm64_insn(insn) ? 2 : 1; + int i; + + for (i = 0; i < insn_cnt; i++) { + pr_debug("prog '%s': relo #%d: substituting insn #%d w/ invalid insn\n", + prog_name, relo_idx, insn_idx + i); + insn[i].code = BPF_JMP | BPF_CALL; + insn[i].dst_reg = 0; + insn[i].src_reg = 0; + insn[i].off = 0; + /* + * If this instruction is reachable (not dead code), the verifier + * will complain with "invalid func unknown#195896080". + */ + insn[i].imm = 195896080; /* => 0xbad2310 => "bad relo" */ + } + + return 0; } static int insn_bpf_size_to_bytes(struct bpf_insn *insn) @@ -1047,17 +1054,6 @@ int bpf_core_patch_insn(const char *prog_name, struct bpf_insn *insn, class = BPF_CLASS(insn->code); - if (res->poison) { -poison: - /* poison second part of ldimm64 to avoid confusing error from - * verifier about "unknown opcode 00" - */ - if (is_ldimm64_insn(insn)) - bpf_core_poison_insn(prog_name, relo_idx, insn_idx + 1, insn + 1); - bpf_core_poison_insn(prog_name, relo_idx, insn_idx, insn); - return 0; - } - orig_val = res->orig_val; new_val = res->new_val; @@ -1065,7 +1061,9 @@ int bpf_core_patch_insn(const char *prog_name, struct bpf_insn *insn, case BPF_ALU: case BPF_ALU64: if (BPF_SRC(insn->code) != BPF_K) - return -EINVAL; + goto bad_insn; + if (res->poison) + return bpf_core_poison_insn(prog_name, relo_idx, insn, insn_idx); if (res->validate && insn->imm != orig_val) { pr_warn("prog '%s': relo #%d: unexpected insn #%d (ALU/ALU64) value: got %d, exp %llu -> %llu\n", prog_name, relo_idx, @@ -1082,6 +1080,8 @@ int bpf_core_patch_insn(const char *prog_name, struct bpf_insn *insn, case BPF_LDX: case BPF_ST: case BPF_STX: + if (res->poison) + return bpf_core_poison_insn(prog_name, relo_idx, insn, insn_idx); if (res->validate && insn->off != orig_val) { pr_warn("prog '%s': relo #%d: unexpected insn #%d (LDX/ST/STX) value: got %d, exp %llu -> %llu\n", prog_name, relo_idx, insn_idx, insn->off, (unsigned long long)orig_val, @@ -1097,7 +1097,7 @@ int bpf_core_patch_insn(const char *prog_name, struct bpf_insn *insn, pr_warn("prog '%s': relo #%d: insn #%d (LDX/ST/STX) accesses field incorrectly. " "Make sure you are accessing pointers, unsigned integers, or fields of matching type and size.\n", prog_name, relo_idx, insn_idx); - goto poison; + return bpf_core_poison_insn(prog_name, relo_idx, insn, insn_idx); } orig_val = insn->off; @@ -1140,6 +1140,9 @@ int bpf_core_patch_insn(const char *prog_name, struct bpf_insn *insn, return -EINVAL; } + if (res->poison) + return bpf_core_poison_insn(prog_name, relo_idx, insn, insn_idx); + imm = (__u32)insn[0].imm | ((__u64)insn[1].imm << 32); if (res->validate && imm != orig_val) { pr_warn("prog '%s': relo #%d: unexpected insn #%d (LDIMM64) value: got %llu, exp %llu -> %llu\n", @@ -1157,6 +1160,7 @@ int bpf_core_patch_insn(const char *prog_name, struct bpf_insn *insn, break; } default: +bad_insn: pr_warn("prog '%s': relo #%d: trying to relocate unrecognized insn #%d, code:0x%x, src:0x%x, dst:0x%x, off:0x%x, imm:0x%x\n", prog_name, relo_idx, insn_idx, insn->code, (unsigned)insn->src_reg, (unsigned)insn->dst_reg, (unsigned)insn->off, (unsigned)insn->imm); From 3440505aca926e726a26c0fd30356454334322bd Mon Sep 17 00:00:00 2001 From: Kumar Kartikeya Dwivedi Date: Fri, 18 Sep 2026 01:32:15 +0200 Subject: [PATCH 345/592] selftests/bpf: Test CO-RE instruction poisoning restrictions Add raw CO-RE relocations that fail to resolve their target enum value. Place each supported and unsupported instruction form in dead code. Unsupported targets must fail relocation with a diagnostic even when they are unreachable. Supported ALU immediates, memory accesses, and ldimm64 instructions must still be poisoned and removed as dead code, allowing the program to load. Check that both halves of ldimm64 are poisoned. Load every instruction stream without relocations first to ensure that rejection is caused by the relocation rather than the original program. Signed-off-by: Kumar Kartikeya Dwivedi Acked-by: Eduard Zingerman Link: https://patch.msgid.link/20260917233222.2542500-8-memxor@gmail.com Signed-off-by: Eduard Zingerman --- .../selftests/bpf/prog_tests/core_reloc_raw.c | 61 ++++++++++++++++++- 1 file changed, 60 insertions(+), 1 deletion(-) diff --git a/tools/testing/selftests/bpf/prog_tests/core_reloc_raw.c b/tools/testing/selftests/bpf/prog_tests/core_reloc_raw.c index bb19e49dd87d..51f42b02a267 100644 --- a/tools/testing/selftests/bpf/prog_tests/core_reloc_raw.c +++ b/tools/testing/selftests/bpf/prog_tests/core_reloc_raw.c @@ -50,6 +50,28 @@ static int load_core_relo_insns(int btf_fd, struct bpf_insn *insns, int insn_cnt static void test_early_core_relo(void) { + static const char unrecognized[] = "trying to relocate unrecognized insn #2"; + static const struct { + const char *name; + struct bpf_insn insns[2]; + const char *err_msg; + } tests[] = { + { "poison_exit", { BPF_EXIT_INSN() }, unrecognized }, + { "poison_ja", { BPF_JMP_A(1) }, unrecognized }, + { "poison_jmp", { BPF_JMP_IMM(BPF_JEQ, BPF_REG_0, 0, 1) }, unrecognized }, + { "poison_jmp32", { BPF_JMP32_IMM(BPF_JEQ, BPF_REG_0, 0, 1) }, unrecognized }, + { "poison_call", { BPF_EMIT_CALL(BPF_FUNC_get_prandom_u32) }, unrecognized }, + { "poison_alu_reg", { BPF_MOV32_REG(BPF_REG_0, BPF_REG_1) }, unrecognized }, + { "poison_alu64_reg", { BPF_MOV64_REG(BPF_REG_0, BPF_REG_1) }, unrecognized }, + { "poison_ld_abs", { BPF_LD_ABS(BPF_W, 0) }, + "insn #2 (LDIMM64) has unexpected form" }, + { "poison_alu_imm", { BPF_MOV32_IMM(BPF_REG_0, 0) } }, + { "poison_alu64_imm", { BPF_MOV64_IMM(BPF_REG_0, 0) } }, + { "poison_ldx", { BPF_LDX_MEM(BPF_W, BPF_REG_0, BPF_REG_1, 0) } }, + { "poison_st", { BPF_ST_MEM(BPF_W, BPF_REG_10, -4, 0) } }, + { "poison_stx", { BPF_STX_MEM(BPF_W, BPF_REG_10, BPF_REG_0, -4) } }, + { "poison_ldimm64", { BPF_LD_IMM64(BPF_REG_0, 0) } }, + }; struct test_btf { struct btf_header hdr; __u32 types[18]; @@ -97,7 +119,7 @@ static void test_early_core_relo(void) }; int access_str_off = 51; /* offset of "0" */ int enum_id = 5; - int btf_fd, prog_fd = -1; + int btf_fd, prog_fd = -1, i; btf_fd = bpf_btf_load(&raw_btf, sizeof(raw_btf), NULL); if (!ASSERT_GE(btf_fd, 0, "btf_load")) @@ -136,6 +158,43 @@ static void test_early_core_relo(void) ASSERT_HAS_SUBSTR(log, "invalid bpf_ld_imm64 insn", "truncated_load_log"); } + for (i = 0; i < ARRAY_SIZE(tests); i++) { + struct bpf_insn insns[] = { + BPF_MOV64_IMM(BPF_REG_0, 0), + BPF_JMP_IMM(BPF_JEQ, BPF_REG_0, 0, 1), + tests[i].insns[0], + BPF_MOV64_IMM(BPF_REG_0, 0), + BPF_EXIT_INSN(), + }; + bool is_ldimm64 = insns[2].code == (BPF_LD | BPF_DW | BPF_IMM); + + if (!test__start_subtest(tests[i].name)) + continue; + if (is_ldimm64) { + insns[1].off = 2; + insns[3] = tests[i].insns[1]; + } + prog_fd = load_core_relo_insns(btf_fd, insns, ARRAY_SIZE(insns), funcs, 1, + enum_id, access_str_off, 2, false); + if (!ASSERT_GE(prog_fd, 0, "control_load")) + goto cleanup; + close(prog_fd); + prog_fd = load_core_relo_insns(btf_fd, insns, ARRAY_SIZE(insns), funcs, 1, + enum_id, access_str_off, 2, true); + if (!tests[i].err_msg) { + ASSERT_GE(prog_fd, 0, "dead_poison_load"); + ASSERT_HAS_SUBSTR(log, "substituting insn #2", "poison_log"); + if (is_ldimm64) + ASSERT_HAS_SUBSTR(log, "substituting insn #3", "poison_ldimm64_log"); + } else { + ASSERT_LT(prog_fd, 0, "invalid_poison_load"); + ASSERT_HAS_SUBSTR(log, tests[i].err_msg, "invalid_poison_log"); + ASSERT_NULL(strstr(log, "substituting insn"), "invalid_poison_substitution"); + } + close(prog_fd); + prog_fd = -1; + } + cleanup: if (env.verbosity > VERBOSE_NORMAL && log[0]) { printf("-------- program load log start --------\n"); From 71919742c83c32afcccf06a7a28e28f3f55f21a2 Mon Sep 17 00:00:00 2001 From: Kumar Kartikeya Dwivedi Date: Fri, 18 Sep 2026 01:32:16 +0200 Subject: [PATCH 346/592] bpf: Assign lock identity to callback map values A nested bpf_for_each_map_elem() callback can unlock a different element of the same map: static long inner(void *map, int *key, struct value *v, struct value **outer_value) { bpf_spin_lock(&v->lock); bpf_spin_unlock(&(*outer_value)->lock); return 0; } static long outer(void *map, int *key, struct value *v, void *ctx) { bpf_for_each_map_elem(map, inner, &v, 0); return 0; } Both callback values currently have ID zero and the same map_ptr. process_spin_lock() compares those two fields, so it accepts the unlock even though the two callbacks can receive different map elements. Assign a fresh ID to every callback map value in the for-each, timer/workqueue, and task-work constructors. Copies of one callback argument retain its ID, so locking and unlocking through that argument continues to work. Distinct callbacks also get distinct IDs for single-element arrays, including inner arrays sharing inner_map_meta. Preserve map_uid for every inner-map lookup and compare it through check_ids() during state pruning. This preserves relationships between maps, keys, and values while allowing equivalent states with different lookup IDs to match. It avoids field-specific rules for when an inner map needs an identity. Move map_uid out of the metadata union and next to the other IDs, so register comparisons can use the existing memcmp() ranges and remap the IDs separately. Clear it when resetting a register or converting a map lookup result to a socket pointer. Shrink frameno to u8, which is enough for MAX_CALL_FRAMES, to make room without growing bpf_reg_state. Fixes: d0d78c1df9b1 ("bpf: Allow locking bpf_spin_lock global variables") Reported-by: Nicholas Carlini Suggested-by: Nicholas Carlini Signed-off-by: Kumar Kartikeya Dwivedi Acked-by: Eduard Zingerman Link: https://patch.msgid.link/20260917233222.2542500-9-memxor@gmail.com Signed-off-by: Eduard Zingerman --- include/linux/bpf_verifier.h | 25 +++++++++++++------------ kernel/bpf/states.c | 6 ++++-- kernel/bpf/verifier.c | 13 +++++++++---- 3 files changed, 26 insertions(+), 18 deletions(-) diff --git a/include/linux/bpf_verifier.h b/include/linux/bpf_verifier.h index bba5a727c651..a7202b44ab10 100644 --- a/include/linux/bpf_verifier.h +++ b/include/linux/bpf_verifier.h @@ -46,17 +46,11 @@ struct bpf_reg_state { /* valid when type == PTR_TO_PACKET */ int range; - /* valid when type == CONST_PTR_TO_MAP | PTR_TO_MAP_VALUE | - * PTR_TO_MAP_VALUE_OR_NULL + /* + * For CONST_PTR_TO_MAP, PTR_TO_MAP_KEY, PTR_TO_MAP_VALUE and + * PTR_TO_INSN. */ - struct { - struct bpf_map *map_ptr; - /* To distinguish map lookups from outer map - * the map_uid is non-zero for registers - * pointing to inner maps. - */ - u32 map_uid; - }; + struct bpf_map *map_ptr; /* for PTR_TO_BTF_ID */ struct { @@ -155,13 +149,20 @@ struct bpf_reg_state { * gets parent_id set to the dynptr's id. */ u32 parent_id; - /* Inside the callee two registers can be both PTR_TO_STACK like + /* + * Distinguishes inner-map lookups and their keys and values. Zero for + * other registers. Kept outside the metadata union for ID remapping + * during state comparisons. + */ + u32 map_uid; + /* + * Inside the callee two registers can be both PTR_TO_STACK like * R1=fp-8 and R2=fp-8, but one of them points to this function stack * while another to the caller's stack. To differentiate them 'frameno' * is used which is an index in bpf_verifier_state->frame[] array * pointing to bpf_func_state. */ - u32 frameno; + u8 frameno; /* if (!precise && SCALAR_VALUE) min/max/tnum don't affect safety */ bool precise; }; diff --git a/kernel/bpf/states.c b/kernel/bpf/states.c index 6c88ad95b63b..e4ec007f7fa6 100644 --- a/kernel/bpf/states.c +++ b/kernel/bpf/states.c @@ -491,7 +491,8 @@ static bool regs_exact(const struct bpf_reg_state *rold, { return memcmp(rold, rcur, offsetof(struct bpf_reg_state, id)) == 0 && check_ids(rold->id, rcur->id, idmap) && - check_ids(rold->parent_id, rcur->parent_id, idmap); + check_ids(rold->parent_id, rcur->parent_id, idmap) && + check_ids(rold->map_uid, rcur->map_uid, idmap); } enum exact_level { @@ -616,7 +617,8 @@ static bool regsafe(struct bpf_verifier_env *env, struct bpf_reg_state *rold, range_within(rold, rcur) && tnum_in(rold->var_off, rcur->var_off) && check_ids(rold->id, rcur->id, idmap) && - check_ids(rold->parent_id, rcur->parent_id, idmap); + check_ids(rold->parent_id, rcur->parent_id, idmap) && + check_ids(rold->map_uid, rcur->map_uid, idmap); case PTR_TO_PACKET_META: case PTR_TO_PACKET: /* We must have at least as much range as the old ptr diff --git a/kernel/bpf/verifier.c b/kernel/bpf/verifier.c index 33161dc64568..02be326f235c 100644 --- a/kernel/bpf/verifier.c +++ b/kernel/bpf/verifier.c @@ -1864,6 +1864,7 @@ static void __mark_reg_known(struct bpf_reg_state *reg, u64 imm) offsetof(struct bpf_reg_state, var_off) - sizeof(reg->type)); reg->id = 0; reg->parent_id = 0; + reg->map_uid = 0; ___mark_reg_known(reg, imm); } @@ -1925,17 +1926,18 @@ static void refine_map_lookup_value(struct bpf_reg_state *reg) if (map->inner_map_meta) { reg->type = CONST_PTR_TO_MAP | maybe_null; reg->map_ptr = map->inner_map_meta; - /* transfer reg's id which is unique for every map_lookup_elem + /* + * transfer reg's id which is unique for every map_lookup_elem * as UID of the inner map. */ - if (btf_record_has_field(map->inner_map_meta->record, - BPF_TIMER | BPF_WORKQUEUE | BPF_TASK_WORK)) - reg->map_uid = reg->id; + reg->map_uid = reg->id; } else if (map->map_type == BPF_MAP_TYPE_XSKMAP) { reg->type = PTR_TO_XDP_SOCK | maybe_null; + reg->map_uid = 0; } else if (map->map_type == BPF_MAP_TYPE_SOCKMAP || map->map_type == BPF_MAP_TYPE_SOCKHASH) { reg->type = PTR_TO_SOCKET | maybe_null; + reg->map_uid = 0; } } @@ -10048,6 +10050,7 @@ int map_set_for_each_callback_args(struct bpf_verifier_env *env, __mark_reg_known_zero(&callee->regs[BPF_REG_3]); callee->regs[BPF_REG_3].map_ptr = caller->regs[BPF_REG_1].map_ptr; callee->regs[BPF_REG_3].map_uid = caller->regs[BPF_REG_1].map_uid; + callee->regs[BPF_REG_3].id = ++env->id_gen; /* pointer to stack or null */ callee->regs[BPF_REG_4] = caller->regs[BPF_REG_3]; @@ -10144,6 +10147,7 @@ static int set_timer_callback_state(struct bpf_verifier_env *env, __mark_reg_known_zero(&callee->regs[BPF_REG_3]); callee->regs[BPF_REG_3].map_ptr = map_ptr; callee->regs[BPF_REG_3].map_uid = map_uid; + callee->regs[BPF_REG_3].id = ++env->id_gen; /* unused */ bpf_mark_reg_not_init(env, &callee->regs[BPF_REG_4]); @@ -10262,6 +10266,7 @@ static int set_task_work_schedule_callback_state(struct bpf_verifier_env *env, __mark_reg_known_zero(&callee->regs[BPF_REG_3]); callee->regs[BPF_REG_3].map_ptr = map_ptr; callee->regs[BPF_REG_3].map_uid = map_uid; + callee->regs[BPF_REG_3].id = ++env->id_gen; /* unused */ bpf_mark_reg_not_init(env, &callee->regs[BPF_REG_4]); From 04ae4ffc57a6b7a8215779f0e9c536cacda9f761 Mon Sep 17 00:00:00 2001 From: Kumar Kartikeya Dwivedi Date: Fri, 18 Sep 2026 01:32:17 +0200 Subject: [PATCH 347/592] selftests/bpf: Check callback map value lock identity Add a verifier test which retains a map value from an outer callback and then acquires a lock through an inner callback value before attempting to release the outer callback value. Both values can denote different elements, so the verifier must reject the mismatched unlock. Also exercise callbacks reached through two inner-map lookups. The lookup results share inner_map_meta but may refer to different one-element arrays, so their callback values must retain distinct lock identities. Extend the existing spin_lock failure table and reuse its array and inner-map fixtures to keep these cases alongside the other lock identity tests. Update the nested callback reference-leak expectation for the extra callback value ID. Signed-off-by: Kumar Kartikeya Dwivedi Link: https://patch.msgid.link/20260917233222.2542500-10-memxor@gmail.com Signed-off-by: Eduard Zingerman --- .../selftests/bpf/prog_tests/cb_refs.c | 2 +- .../selftests/bpf/prog_tests/spin_lock.c | 2 + .../selftests/bpf/progs/test_spin_lock_fail.c | 67 ++++++++++++++++++- 3 files changed, 68 insertions(+), 3 deletions(-) diff --git a/tools/testing/selftests/bpf/prog_tests/cb_refs.c b/tools/testing/selftests/bpf/prog_tests/cb_refs.c index 78566b817fd7..490e15e7126d 100644 --- a/tools/testing/selftests/bpf/prog_tests/cb_refs.c +++ b/tools/testing/selftests/bpf/prog_tests/cb_refs.c @@ -13,7 +13,7 @@ struct { } cb_refs_tests[] = { { "underflow_prog", "release kfunc bpf_kfunc_call_test_release expects referenced PTR_TO_BTF_ID passed to R1" }, { "leak_prog", "Possibly NULL pointer passed to helper R2" }, - { "nested_cb", "Unreleased reference id=4 alloc_insn=2" }, /* alloc_insn=2{4,5} */ + { "nested_cb", "Unreleased reference id=5 alloc_insn=2" }, /* alloc_insn=2{4,5} */ { "non_cb_transfer_ref", "Unreleased reference id=4 alloc_insn=1" }, /* alloc_insn=1{1,2} */ }; diff --git a/tools/testing/selftests/bpf/prog_tests/spin_lock.c b/tools/testing/selftests/bpf/prog_tests/spin_lock.c index 5c3579438427..e368370262c8 100644 --- a/tools/testing/selftests/bpf/prog_tests/spin_lock.c +++ b/tools/testing/selftests/bpf/prog_tests/spin_lock.c @@ -54,6 +54,8 @@ static struct { { "lock_global_sleepable_helper_subprog", "global function calls are not allowed while holding a lock" }, { "lock_global_sleepable_kfunc_subprog", "global function calls are not allowed while holding a lock" }, { "lock_global_sleepable_subprog_indirect", "global function calls are not allowed while holding a lock" }, + { "callback_value_lock_identity", "bpf_spin_unlock of different lock" }, + { "callback_inner_map_value_lock_identity", "bpf_spin_unlock of different lock" }, }; static int match_regex(const char *pattern, const char *string) diff --git a/tools/testing/selftests/bpf/progs/test_spin_lock_fail.c b/tools/testing/selftests/bpf/progs/test_spin_lock_fail.c index f678ee6bd7ea..55282f20fa32 100644 --- a/tools/testing/selftests/bpf/progs/test_spin_lock_fail.c +++ b/tools/testing/selftests/bpf/progs/test_spin_lock_fail.c @@ -14,17 +14,18 @@ struct array_map { __type(key, int); __type(value, struct foo); __uint(max_entries, 1); -} array_map SEC(".maps"); +} array_map SEC(".maps"), array_map_b SEC(".maps"); struct { __uint(type, BPF_MAP_TYPE_ARRAY_OF_MAPS); - __uint(max_entries, 1); + __uint(max_entries, 2); __type(key, int); __type(value, int); __array(values, struct array_map); } map_of_maps SEC(".maps") = { .values = { [0] = &array_map, + [1] = &array_map_b, }, }; @@ -314,4 +315,66 @@ int lock_global_sleepable_subprog_indirect(struct __sk_buff *ctx) return ret; } +struct { + __uint(type, BPF_MAP_TYPE_ARRAY); + __uint(max_entries, 2); + __type(key, int); + __type(value, struct foo); +} callback_array_map SEC(".maps"); + +struct callback_ctx { + struct foo *value; +}; + +static long lock_different_value(struct bpf_map *map, int *key, + struct foo *value, struct callback_ctx *ctx) +{ + bpf_spin_lock(&value->lock); + bpf_spin_unlock(&ctx->value->lock); + return 0; +} + +static long nest_lock_different_value(struct bpf_map *map, int *key, + struct foo *value, void *data) +{ + struct callback_ctx ctx = { .value = value }; + + bpf_for_each_map_elem(&callback_array_map, lock_different_value, &ctx, 0); + return 0; +} + +SEC("?tc") +int callback_value_lock_identity(void *ctx) +{ + bpf_for_each_map_elem(&callback_array_map, nest_lock_different_value, NULL, 0); + return 0; +} + +static long nest_lock_different_inner_value(struct bpf_map *map, int *key, + struct foo *value, void *data) +{ + struct callback_ctx ctx = { .value = value }; + int inner_key = 1; + void *inner_map; + + inner_map = bpf_map_lookup_elem(&map_of_maps, &inner_key); + if (!inner_map) + return 0; + bpf_for_each_map_elem(inner_map, lock_different_value, &ctx, 0); + return 0; +} + +SEC("?tc") +int callback_inner_map_value_lock_identity(void *ctx) +{ + int inner_key = 0; + void *inner_map; + + inner_map = bpf_map_lookup_elem(&map_of_maps, &inner_key); + if (!inner_map) + return 0; + bpf_for_each_map_elem(inner_map, nest_lock_different_inner_value, NULL, 0); + return 0; +} + char _license[] SEC("license") = "GPL"; From b4e875d397da451fb4e9c573ff4b86db53caba05 Mon Sep 17 00:00:00 2001 From: Kumar Kartikeya Dwivedi Date: Fri, 18 Sep 2026 01:32:18 +0200 Subject: [PATCH 348/592] libbpf: Reject truncated ldimm64 CO-RE relocations CO-RE relocation of an ldimm64 instruction operates on two instruction slots. A malformed BPF ELF can end a function after the first slot and attach a CO-RE relocation to it. libbpf allocates the instruction array according to the function symbol size, so the shared relocation code would then access beyond the allocation. Reject a terminal ldimm64 in libbpf's relocation loop, where the program length is available, before resolving or applying the relocation. Both resolved and unresolved relocations validate the absent second slot, and unresolved relocation poisoning would additionally write past the array. The in-kernel caller is protected by the verifier's early instruction-stream check before it applies CO-RE relocations. Fixes: eacaaed784e2 ("libbpf: Implement enum value-based CO-RE relocations") Reported-by: Sashiko Signed-off-by: Kumar Kartikeya Dwivedi Link: https://lore.kernel.org/20260914140852.03DA21F0089B@smtp.kernel.org Link: https://patch.msgid.link/20260917233222.2542500-11-memxor@gmail.com Signed-off-by: Eduard Zingerman --- tools/lib/bpf/libbpf.c | 7 +++++++ 1 file changed, 7 insertions(+) diff --git a/tools/lib/bpf/libbpf.c b/tools/lib/bpf/libbpf.c index b749c01742ee..bfa64ae6c94d 100644 --- a/tools/lib/bpf/libbpf.c +++ b/tools/lib/bpf/libbpf.c @@ -6206,6 +6206,13 @@ bpf_object__relocate_core(struct bpf_object *obj, const char *targ_btf_path) return -EINVAL; insn = &prog->insns[insn_idx]; + if (is_ldimm64_insn(insn) && (size_t)insn_idx + 1 >= prog->insns_cnt) { + pr_warn("prog '%s': relo #%d: insn #%d (LDIMM64) is truncated\n", + prog->name, i, insn_idx); + err = -EINVAL; + goto out; + } + err = record_relo_core(prog, rec, insn_idx); if (err) { pr_warn("prog '%s': relo #%d: failed to record relocation: %s\n", From 2ec28c09b320ba241bea8a70ee5cb9ccf4a099e8 Mon Sep 17 00:00:00 2001 From: Aldo Ariel Panzardo Date: Tue, 15 Sep 2026 14:30:50 -0300 Subject: [PATCH 349/592] vsock: ignore empty child namespace mode writes __vsock_net_mode_string() returns success without updating new_mode when the transfer length is zero. Its caller then reads the uninitialized enum and may permanently store a stack-derived value in the write-once child mode. Return before calling __vsock_net_mode_string() when *lenp is zero so that the helper is never invoked with nothing to parse and new_mode is never read uninitialized. This also prevents an empty write from locking the current mode. Fixes: eafb64f40ca4 ("vsock: add netns to vsock core") Cc: stable@vger.kernel.org Reviewed-by: Luigi Leonardi Signed-off-by: Aldo Ariel Panzardo Reviewed-by: Stefano Garzarella Reviewed-by: Bobby Eshleman Link: https://patch.msgid.link/20260915173050.3176344-1-qwe.aldo@gmail.com Signed-off-by: Jakub Kicinski --- net/vmw_vsock/af_vsock.c | 3 +++ 1 file changed, 3 insertions(+) diff --git a/net/vmw_vsock/af_vsock.c b/net/vmw_vsock/af_vsock.c index f840498b58af..9b71479a2b29 100644 --- a/net/vmw_vsock/af_vsock.c +++ b/net/vmw_vsock/af_vsock.c @@ -2889,6 +2889,9 @@ static int vsock_net_child_mode_string(const struct ctl_table *table, int write, net = container_of(table->data, struct net, vsock.child_ns_mode); + if (!*lenp) + return 0; + ret = __vsock_net_mode_string(table, write, buffer, lenp, ppos, vsock_net_child_mode(net), &new_mode); if (ret) From 651010592bdce7005c1179498327e51bfc4fe1a5 Mon Sep 17 00:00:00 2001 From: Zhang Yunfei Date: Fri, 11 Sep 2026 17:11:23 +0800 Subject: [PATCH 350/592] net: txgbe: fix FDIR filter restore for VF rules txgbe_fdir_filter_restore() reprograms every filter from txgbe->fdir_filter_list after a reset. It extracts the ring part of filter->action with ethtool_get_flow_spec_ring() and maps it onto a PF rx ring, silently dropping the VF part of the cookie that txgbe_add_ethtool_fdir_entry() stores there (input->action = fsp->ring_cookie). For a rule directed at a VF, restore therefore reprograms the filter to the PF queue with the same ring index: after any down/up or txgbe_reinit_locked(), traffic matching the rule is steered to the PF instead of the VF. Handle VF rules the same way txgbe_add_ethtool_fdir_entry() does: validate vf against wx->num_vfs and ring against wx->num_rx_queues_per_pool, and map the ring onto the absolute queue index ((vf - 1) * wx->num_rx_queues_per_pool) + ring. Fixes: 7a91722e0dd4 ("net: txgbe: Support the FDIR rules assigned to VFs") Cc: stable@vger.kernel.org Signed-off-by: Zhang Yunfei Link: https://patch.msgid.link/20260911091123.798931-1-zhangyunfei1@kylinos.cn Signed-off-by: Jakub Kicinski --- drivers/net/ethernet/wangxun/txgbe/txgbe_fdir.c | 13 +++++++++++-- 1 file changed, 11 insertions(+), 2 deletions(-) diff --git a/drivers/net/ethernet/wangxun/txgbe/txgbe_fdir.c b/drivers/net/ethernet/wangxun/txgbe/txgbe_fdir.c index a84010828551..59a47532618c 100644 --- a/drivers/net/ethernet/wangxun/txgbe/txgbe_fdir.c +++ b/drivers/net/ethernet/wangxun/txgbe/txgbe_fdir.c @@ -591,15 +591,24 @@ static void txgbe_fdir_filter_restore(struct wx *wx) queue = TXGBE_RDB_FDIR_DROP_QUEUE; } else { u32 ring = ethtool_get_flow_spec_ring(filter->action); + u8 vf = ethtool_get_flow_spec_ring_vf(filter->action); - if (ring >= wx->num_rx_queues) { + if (!vf && ring >= wx->num_rx_queues) { wx_err(wx, "FDIR restore failed, ring:%u\n", ring); continue; + } else if (vf && (vf > wx->num_vfs || + ring >= wx->num_rx_queues_per_pool)) { + wx_err(wx, "FDIR restore failed, vf:%u, ring:%u\n", + vf, ring); + continue; } /* Map the ring onto the absolute queue index */ - queue = wx->rx_ring[ring]->reg_idx; + if (!vf) + queue = wx->rx_ring[ring]->reg_idx; + else + queue = ((vf - 1) * wx->num_rx_queues_per_pool) + ring; } ret = txgbe_fdir_write_perfect_filter(wx, From 46bc52d13594848023e681860df8700c8db14354 Mon Sep 17 00:00:00 2001 From: Linkui Xiao Date: Wed, 16 Sep 2026 20:53:16 +0800 Subject: [PATCH 351/592] ipv4: fib: fix data-race and stale genid check around nh->nh_saddr fib_select_multipath() compares nexthop_nh->nh_saddr against the flow source address with no lock held, while fib_info_update_nhc_saddr() stores a new value from another CPU as soon as the preferred source address of the egress device changes. Commit 195374d89368 ("ipv4: fib: annotate races around nh->nh_saddr_genid and nh->nh_saddr") added WRITE_ONCE() on the store side and READ_ONCE() in fib_result_prefsrc() after syzbot reported BUG: KCSAN: data-race in fib_select_path / fib_select_path but it only covered that reader. fib_select_multipath(), reached from fib_select_path(), is a second lockless reader of nh->nh_saddr and was left bare. Moreover, nh_saddr is only meaningful when nh_saddr_genid matches dev_addr_genid, as established by commit 436c3b66ec98 ("ipv4: Invalidate nexthop cache nh_saddr more correctly."). fib_select_multipath() skips that validation, so it can score a nexthop using a stale source address and skew the ECMP selection. Annotate both reads with READ_ONCE() and refresh the cached source address via fib_info_update_nhc_saddr() when the genid does not match, mirroring fib_result_prefsrc(). Fixes: 32607a332cfe ("ipv4: prefer multipath nexthop that matches source address") Signed-off-by: Linkui Xiao Reviewed-by: Ido Schimmel Reviewed-by: Eric Dumazet Link: https://patch.msgid.link/20260916125316.988044-1-xiaolinkui@126.com Signed-off-by: Jakub Kicinski --- net/ipv4/fib_semantics.c | 13 ++++++++++++- 1 file changed, 12 insertions(+), 1 deletion(-) diff --git a/net/ipv4/fib_semantics.c b/net/ipv4/fib_semantics.c index 7a362f2e2c2b..5c9021ea3a79 100644 --- a/net/ipv4/fib_semantics.c +++ b/net/ipv4/fib_semantics.c @@ -2176,6 +2176,15 @@ static bool fib_good_nh(const struct fib_nh *nh) return !!(state & NUD_VALID); } +static __be32 fib_nh_saddr(struct net *net, const struct fib_info *fi, + struct fib_nh *nh, int genid) +{ + if (READ_ONCE(nh->nh_saddr_genid) == genid) + return READ_ONCE(nh->nh_saddr); + + return fib_info_update_nhc_saddr(net, &nh->nh_common, fi->fib_scope); +} + void fib_select_multipath(struct fib_result *res, int hash, const struct flowi4 *fl4) { @@ -2184,6 +2193,7 @@ void fib_select_multipath(struct fib_result *res, int hash, bool use_neigh; int score = -1; __be32 saddr; + int genid; if (unlikely(res->fi->nh)) { nexthop_path_fib_result(res, hash); @@ -2192,6 +2202,7 @@ void fib_select_multipath(struct fib_result *res, int hash, use_neigh = READ_ONCE(net->ipv4.sysctl_fib_multipath_use_neigh); saddr = fl4 ? fl4->saddr : 0; + genid = saddr ? atomic_read(&net->ipv4.dev_addr_genid) : 0; change_nexthops(fi) { int nh_upper_bound, nh_score = 0; @@ -2204,7 +2215,7 @@ void fib_select_multipath(struct fib_result *res, int hash, (use_neigh && !fib_good_nh(nexthop_nh))) continue; - if (saddr && nexthop_nh->nh_saddr == saddr) + if (saddr && fib_nh_saddr(net, fi, nexthop_nh, genid) == saddr) nh_score += 2; if (hash <= nh_upper_bound) nh_score++; From a363c62a653cc8b3e21da9545fa4e028ef50f9c3 Mon Sep 17 00:00:00 2001 From: Longlong Xia Date: Sun, 23 Aug 2026 12:40:51 +0800 Subject: [PATCH 352/592] mm/hugetlb: do not dissolve gigantic pages without runtime support dissolve_free_hugetlb_folio() doesn't check hstate_is_gigantic_no_runtime(h) though remove_hugetlb_folio()/ update_and_free_hugetlb_folio() silently bail for such folios, so it frees a still-listed folio and, on vmemmap restore failure, the add_hugetlb_folio() rollback corrupts the free list. Link: https://lore.kernel.org/20260823044118.1097121-2-xialonglong2025@163.com Fixes: 6eb4e88a6d27 ("hugetlb: create remove_hugetlb_page() to separate functionality") Signed-off-by: Longlong Xia Signed-off-by: Andrew Morton Assisted-by: Codex:gpt-5.6-sol Acked-by: Muchun Song Cc: David Hildenbrand Cc: Miaohe Lin Cc: Michal Hocko Cc: Oscar Salvador Cc: --- mm/hugetlb.c | 9 +++++++++ 1 file changed, 9 insertions(+) diff --git a/mm/hugetlb.c b/mm/hugetlb.c index 4f6f58bf3db6..d28972cd33f5 100644 --- a/mm/hugetlb.c +++ b/mm/hugetlb.c @@ -1967,6 +1967,15 @@ int dissolve_free_hugetlb_folio(struct folio *folio) struct hstate *h = folio_hstate(folio); bool adjust_surplus = false; + /* + * remove_hugetlb_folio()/update_and_free_hugetlb_folio() bail + * for gigantic hstates without runtime support, so dissolving one + * here would leave it on the free list and, on vmemmap restore + * failure, the add_hugetlb_folio() rollback corrupts that list. + */ + if (hstate_is_gigantic_no_runtime(h)) + goto out; + if (!available_huge_pages(h)) goto out; From d644b23afe1ef509c9961a6d84a093c2587edf02 Mon Sep 17 00:00:00 2001 From: =?UTF-8?q?J=C3=A9r=C3=A9my=20Jean?= Date: Tue, 18 Aug 2026 20:00:15 +0000 Subject: [PATCH 353/592] netfilter: flowtable: publish HW_DEAD after worker is done MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit flow_offload_work_del() sets NF_FLOW_HW_DEAD before the work handler clears NF_FLOW_HW_PENDING. Once a flow is both HW_DYING and HW_DEAD, a concurrent garbage collection pass can remove it and schedule it for RCU freeing. The offload worker holds neither an RCU read lock nor a reference to the flow. If it is preempted after publishing HW_DEAD, the RCU callback can free the flow before the worker resumes and clears HW_PENDING, resulting in a use-after-free. Move HW_DEAD publication to the common worker epilogue after the pending bit is cleared, making it the final flow access by destroy work. Order all preceding flow accesses before publishing the bit that allows garbage collection to free the object. Fixes: 2c8897953f3b ("netfilter: flowtable: Add pending bit for offload work") Assisted-by: Codex:gpt-5 Signed-off-by: Jérémy Jean Signed-off-by: Pablo Neira Ayuso --- net/netfilter/nf_flow_table_offload.c | 7 ++++++- 1 file changed, 6 insertions(+), 1 deletion(-) diff --git a/net/netfilter/nf_flow_table_offload.c b/net/netfilter/nf_flow_table_offload.c index 801a3dd9ceea..6757fd89c1f1 100644 --- a/net/netfilter/nf_flow_table_offload.c +++ b/net/netfilter/nf_flow_table_offload.c @@ -995,7 +995,6 @@ static void flow_offload_work_del(struct flow_offload_work *offload) flow_offload_tuple_del(offload, FLOW_OFFLOAD_DIR_ORIGINAL); if (test_bit(NF_FLOW_HW_BIDIRECTIONAL, &offload->flow->flags)) flow_offload_tuple_del(offload, FLOW_OFFLOAD_DIR_REPLY); - set_bit(NF_FLOW_HW_DEAD, &offload->flow->flags); } static void flow_offload_tuple_stats(struct flow_offload_work *offload, @@ -1059,6 +1058,12 @@ static void flow_offload_work_handler(struct work_struct *work) } clear_bit(NF_FLOW_HW_PENDING, &offload->flow->flags); + if (offload->cmd == FLOW_CLS_DESTROY) { + /* Publish after the worker's last flow access. */ + smp_mb__before_atomic(); + set_bit(NF_FLOW_HW_DEAD, &offload->flow->flags); + } + kfree(offload); } From 9461613afc59acef44a0071b0dd5075f6e993ffe Mon Sep 17 00:00:00 2001 From: Florian Westphal Date: Thu, 3 Sep 2026 02:41:46 +0200 Subject: [PATCH 354/592] netfilter: nfnetlink_queue: hold nfnl mutex in event notifier We must serialize the release notifier and the config netlink function. A concurrent thread can issue close() which can call the release function while unrelated socket processes UNBIND request for same portid: Oops: general protection fault, [..] RIP: 0010:__instance_destroy+0x60/0x210 [nfnetlink_queue] Call Trace: nfqnl_recv_config+0x9b0/0xdc0 [nfnetlink_queue] nfnetlink_rcv_msg+0x7c2/0xeb0 ? __pfx_nfnetlink_rcv_msg+0x10/0x10 After this, parallel UNBIND and URELEASE events are impossible. This change isn't nice, but its the shortest fix given instances are not refcounted and the nfnetlink config callback drops the rcu read lock early due to need for sleeping allocations. Fixes: 7af4cc3fa158 ("[NETFILTER]: Add "nfnetlink_queue" netfilter queue handler over nfnetlink") Signed-off-by: Florian Westphal Signed-off-by: Pablo Neira Ayuso --- net/netfilter/nfnetlink_queue.c | 8 +++++--- 1 file changed, 5 insertions(+), 3 deletions(-) diff --git a/net/netfilter/nfnetlink_queue.c b/net/netfilter/nfnetlink_queue.c index c727668b0c5b..a3bc00280051 100644 --- a/net/netfilter/nfnetlink_queue.c +++ b/net/netfilter/nfnetlink_queue.c @@ -1593,6 +1593,7 @@ nfqnl_rcv_nl_event(struct notifier_block *this, if (event == NETLINK_URELEASE && n->protocol == NETLINK_NETFILTER) { int i; + nfnl_lock(NFNL_SUBSYS_QUEUE); /* destroy all instances for this portid */ spin_lock(&q->instances_lock); for (i = 0; i < INSTANCE_BUCKETS; i++) { @@ -1606,6 +1607,7 @@ nfqnl_rcv_nl_event(struct notifier_block *this, } } spin_unlock(&q->instances_lock); + nfnl_unlock(NFNL_SUBSYS_QUEUE); } return NOTIFY_DONE; } @@ -1925,9 +1927,9 @@ static int nfqnl_recv_config(struct sk_buff *skb, const struct nfnl_info *info, /* Lookup queue under RCU. After peer_portid check (or for new queue * in BIND case), the queue is owned by the socket sending this message. - * A socket cannot simultaneously send a message and close, so while - * processing this CONFIG message, nfqnl_rcv_nl_event() (triggered by - * socket close) cannot destroy this queue. Safe to use without RCU. + * nfqnl_rcv_nl_event() will block on the nfnl subsys mutex that is + * held by the caller, so the queue cannot be destroyed in parallel, + * even after we drop the RCU read lock. */ rcu_read_lock(); queue = instance_lookup(q, queue_num); From 1b9b5323725e458906c7620a3bc10398b51ad954 Mon Sep 17 00:00:00 2001 From: Weiming Shi Date: Sun, 6 Sep 2026 16:44:10 +0800 Subject: [PATCH 355/592] netfilter: ip6t_rpfilter: reject routes without inet6_dev ip6_route_lookup() can return an error-free route whose rt6i_idev is NULL. Lowering an external nexthop device's MTU below IPV6_MIN_MTU tears down its inet6_dev while fib6_ifdown() leaves routes using nexthop objects in the FIB. An unprivileged user can construct this state with rtnetlink in a private user and network namespace, then trigger a NULL dereference through an IPv6 rpfilter lookup: Oops: general protection fault, probably for non-canonical address 0xdffffc0000000000 KASAN: null-ptr-deref in range [0x0000000000000000-0x0000000000000007] RIP: rpfilter_mt (net/ipv6/netfilter/ip6t_rpfilter.c:75) Call Trace: ip6t_do_table (net/ipv6/netfilter/ip6_tables.c:316) nf_hook_slow (net/netfilter/core.c:619) ipv6_rcv (net/ipv6/ip6_input.c:351) __netif_receive_skb_one_core (net/core/dev.c:6216) process_backlog (net/core/dev.c:6680) __napi_poll (net/core/dev.c:7739) net_rx_action (net/core/dev.c:7959) handle_softirqs (kernel/softirq.c:622) do_softirq.part.0 (kernel/softirq.c:523) __local_bh_enable_ip (kernel/softirq.c:450) __dev_queue_xmit (net/core/dev.c:4913) packet_sendmsg (net/packet/af_packet.c:3139) __sys_sendto (net/socket.c:2252) __x64_sys_sendto (net/socket.c:2259) do_syscall_64 (arch/x86/entry/syscall_64.c:94) entry_SYSCALL_64_after_hwframe (arch/x86/entry/entry_64.S:121) Kernel panic - not syncing: Fatal exception in interrupt Reject routes without an inet6_dev immediately after lookup. Such routes are not eligible for reverse-path filtering, and the check protects all later rt6i_idev dereferences. Fixes: e26f9a480fb6 ("netfilter: add ipv6 reverse path filter match") Reported-by: co+459f67f4d8af8ce6@bugs.sh Closes: https://lore.kernel.org/all/VtWUkE8QzJt5CroTj2V2v3ZQ0gwbXZ7nq7I3@bugs.sh/ Suggested-by: Florian Westphal Assisted-by: Claude:gpt-5 Cc: stable@vger.kernel.org Signed-off-by: Weiming Shi Signed-off-by: Pablo Neira Ayuso --- net/ipv6/netfilter/ip6t_rpfilter.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/net/ipv6/netfilter/ip6t_rpfilter.c b/net/ipv6/netfilter/ip6t_rpfilter.c index 67c87a88cde4..b5def30c3127 100644 --- a/net/ipv6/netfilter/ip6t_rpfilter.c +++ b/net/ipv6/netfilter/ip6t_rpfilter.c @@ -61,7 +61,7 @@ static bool rpfilter_lookup_reverse6(struct net *net, const struct sk_buff *skb, fl6.flowi6_oif = dev->ifindex; rt = (void *)ip6_route_lookup(net, &fl6, skb, lookup_flags); - if (rt->dst.error) + if (rt->dst.error || !rt->rt6i_idev) goto out; if (rt->rt6i_flags & (RTF_REJECT|RTF_ANYCAST)) From 82313c169eddc02b1bf5ba6b427803e272d3ec42 Mon Sep 17 00:00:00 2001 From: Luxiao Xu Date: Sun, 6 Sep 2026 21:29:55 +0800 Subject: [PATCH 356/592] netfilter: ip6t_rt: fix zero-address non-strict match out-of-bounds read rt_mt6_check() permits rules to be configured with rtinfo->addrnr == 0 even when address matching (IP6T_RT_FST_MASK) is requested. In the IP6T_RT_FST_NSTRICT path, rt_mt6() evaluates packet routing addresses against rtinfo->addrs[i] and terminates backwards at the bottom of the loop: if (ipv6_addr_equal(ap, &rtinfo->addrs[i])) { i++; } if (i == rtinfo->addrnr) break; When addrnr is 0, if the first packet address matches rtinfo->addrs[0], i is incremented to 1. Because i is now strictly greater than addrnr (0), the loop termination condition (i == rtinfo->addrnr) is bypassed and will never be satisfied. If a crafted IPv6 packet contains matching routing addresses, i will advance past IP6T_RT_HOPS (16). The subsequent call to ipv6_addr_equal() reads beyond struct ip6t_rt, triggering UBSAN/KASAN out-of-bounds warnings or kernel panics. Fix this by: 1. Rejecting rules in rt_mt6_check() where IP6T_RT_FST_MASK is set but rtinfo->addrnr is zero. 2. In rt_mt6(), moving the termination condition (i < rtinfo->addrnr) into the for-loop header condition and removing the backwards break at the end of the loop body. Fixes: 1da177e4c3f4 ("Linux-2.6.12-rc2") Cc: stable@vger.kernel.org Reported-by: Vega Suggested-by: Florian Westphal Assisted-by: LLM Signed-off-by: Luxiao Xu Signed-off-by: Ren Wei Signed-off-by: Pablo Neira Ayuso --- net/ipv6/netfilter/ip6t_rt.c | 11 ++++++++--- 1 file changed, 8 insertions(+), 3 deletions(-) diff --git a/net/ipv6/netfilter/ip6t_rt.c b/net/ipv6/netfilter/ip6t_rt.c index 8051425213dd..9880faf3cc7d 100644 --- a/net/ipv6/netfilter/ip6t_rt.c +++ b/net/ipv6/netfilter/ip6t_rt.c @@ -96,7 +96,8 @@ static bool rt_mt6(const struct sk_buff *skb, struct xt_action_param *par) unsigned int i = 0; for (temp = 0; - temp < (unsigned int)((hdrlen - 8) / 16); + temp < (unsigned int)((hdrlen - 8) / 16) && + i < rtinfo->addrnr; temp++) { ap = skb_header_pointer(skb, ptr @@ -112,8 +113,6 @@ static bool rt_mt6(const struct sk_buff *skb, struct xt_action_param *par) if (ipv6_addr_equal(ap, &rtinfo->addrs[i])) i++; - if (i == rtinfo->addrnr) - break; } if (i == rtinfo->addrnr) return ret; @@ -162,6 +161,12 @@ static int rt_mt6_check(const struct xt_mtchk_param *par) pr_info_ratelimited("too many addresses specified\n"); return -EINVAL; } + + if ((rtinfo->flags & IP6T_RT_FST_MASK) && !rtinfo->addrnr) { + pr_info_ratelimited("address list match requested but addrnr is 0\n"); + return -EINVAL; + } + if ((rtinfo->flags & (IP6T_RT_RES | IP6T_RT_FST_MASK)) && (!(rtinfo->flags & IP6T_RT_TYP) || (rtinfo->rt_type != 0) || From a311a898172743558b82f6035ef2aa8c310a4223 Mon Sep 17 00:00:00 2001 From: Karl Mehltretter Date: Thu, 10 Sep 2026 22:02:28 +0200 Subject: [PATCH 357/592] netfilter: nft_synproxy: use the family-aware checksum helper nft_synproxy_do_eval() verifies the TCP checksum before it switches on skb->protocol. It uses nf_ip_checksum(), which constructs an IPv4 pseudo header and relies on the IPv4 header checksum when folding the whole skb. Neither operation is valid for an IPv6 packet. A correctly checksummed IPv6 segment can therefore fail verification when it reaches the hook as CHECKSUM_NONE or, at NF_INET_LOCAL_IN, CHECKSUM_COMPLETE. nft_synproxy_do_eval() returns NF_DROP before nft_synproxy_eval_v6() can send a SYN-ACK. nft_synproxy_validate() deliberately admits NFPROTO_IPV6 and NFPROTO_INET, and the xtables counterpart ip6t_SYNPROXY.c already calls nf_ip6_checksum(). Use nf_checksum() with nft_pf() so the checksum helper dispatches to the packet family's implementation. Fixes: ad49d86e07a4 ("netfilter: nf_tables: Add synproxy support") Assisted-by: LLM Signed-off-by: Karl Mehltretter Signed-off-by: Pablo Neira Ayuso --- net/netfilter/nft_synproxy.c | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-) diff --git a/net/netfilter/nft_synproxy.c b/net/netfilter/nft_synproxy.c index 9ed288c9d168..554a96a000f4 100644 --- a/net/netfilter/nft_synproxy.c +++ b/net/netfilter/nft_synproxy.c @@ -118,7 +118,8 @@ static void nft_synproxy_do_eval(const struct nft_synproxy *priv, return; } - if (nf_ip_checksum(skb, nft_hook(pkt), thoff, IPPROTO_TCP)) { + if (nf_checksum(skb, nft_hook(pkt), thoff, IPPROTO_TCP, + nft_pf(pkt))) { regs->verdict.code = NF_DROP; return; } From e290145564886d6a3038810c621f738c1fe9fa51 Mon Sep 17 00:00:00 2001 From: Julian Anastasov Date: Fri, 11 Sep 2026 14:43:15 +0300 Subject: [PATCH 358/592] ipvs: revalidate ihl before icmp_send While the outer IP header is already pulled into the skb head, we must be careful and revalidate the embedded headers after reading them from the skb frags to prevent possible out-of-bounds access. One such place reported by Sashiko is ip_vs_in_icmp() where local process can change the ihl field and after pskb_may_pull() we can see larger value. Even if icmp_send() has checks to prevent out-of-bounds access, play safe and add check to drop the packet if the ihl field is changed. As the outer headers are pulled, make sure the transport header is updated too, it was used before commit 7fcc2fe39fed ("net: icmp: avoid invalid transport header access in icmp_send tracepoint") Fixes: f2edb9f7706d ("ipvs: implement passive PMTUD for IPIP packets") Link: https://sashiko.dev/#/patchset/20260806105211.34622-1-ja%40ssi.bg Signed-off-by: Julian Anastasov Signed-off-by: Pablo Neira Ayuso --- net/netfilter/ipvs/ip_vs_core.c | 6 ++++++ 1 file changed, 6 insertions(+) diff --git a/net/netfilter/ipvs/ip_vs_core.c b/net/netfilter/ipvs/ip_vs_core.c index ba0957798bad..fd503f0efb57 100644 --- a/net/netfilter/ipvs/ip_vs_core.c +++ b/net/netfilter/ipvs/ip_vs_core.c @@ -1960,6 +1960,12 @@ ip_vs_in_icmp(struct netns_ipvs *ipvs, struct sk_buff *skb, int *related, /* Ensure the IP header is present in headroom */ if (!pskb_may_pull(skb, hlen_orig)) goto ignore_tunnel; + skb_set_transport_header(skb, hlen_orig); + /* Before now we may used ihl from skb frag, revalidate it after + * copying it into skb head to prevent out-of-bounds access + */ + if (ip_hdr(skb)->ihl * 4 != hlen_orig) + goto ignore_tunnel; IP_VS_DBG(12, "Sending ICMP for %pI4->%pI4: t=%u, c=%u, i=%u\n", &ip_hdr(skb)->saddr, &ip_hdr(skb)->daddr, type, code, ntohl(info)); From 207d591c353201f3bd3e0c89bb7d44a849c8fd59 Mon Sep 17 00:00:00 2001 From: Naman Gulati Date: Sat, 12 Sep 2026 01:10:51 +0000 Subject: [PATCH 359/592] netfilter: ctnetlink: fix suspicious RCU usage in expect_iter_name expect_iter_name() is invoked by nf_ct_expect_iterate_net() under spin_lock_bh(&nf_conntrack_expect_lock). It does not hold rcu_read_lock(). When accessing exp->helper with rcu_dereference() in syzbot's report, lockdep warns: ============================= WARNING: suspicious RCU usage syzkaller #0 Not tainted ----------------------------- net/netfilter/nf_conntrack_netlink.c:3393 suspicious rcu_dereference_check() usage! locks held by syz-executor381/5628: 2, last CPU#1: #0: ffffffff9aee42a0 (nfnl_subsys_ctnetlink_exp){+.+.}-{4:4}, at: nfnetlink_rcv_msg+0xa69/0x12b0 #1: ffffffff8ea74d58 (nf_conntrack_expect_lock){+...}-{3:3}, at: nf_ct_expect_iterate_net+0x38/0x180 Call Trace: dump_stack_lvl+0xe8/0x150 lockdep_rcu_suspicious+0x140/0x1d0 expect_iter_name+0xfb/0x100 nf_ct_expect_iterate_net+0xf2/0x180 ctnetlink_del_expect+0x45d/0x640 nfnetlink_rcv_msg+0xcc2/0x12b0 netlink_rcv_skb+0x226/0x4a0 nfnetlink_rcv+0x2b9/0x28c0 netlink_unicast+0x7bd/0x940 netlink_sendmsg+0x813/0xb40 ____sys_sendmsg+0x54e/0x850 ___sys_sendmsg+0x2a5/0x360 __sys_sendmsg+0x2a5/0x360 do_syscall_64+0x166/0x520 entry_SYSCALL_64_after_hwframe+0x77/0x7f Use rcu_dereference_protected() with lockdep_is_held() on nf_conntrack_expect_lock instead, similar to expect_iter_me() in nf_conntrack_helper.c. Fixes: f01794106042 ("netfilter: nf_conntrack_expect: use expect->helper") Reported-by: syzbot+4bd730aede2791e40bdf@syzkaller.appspotmail.com Closes: https://lore.kernel.org/netdev/6aa4a377.f81106d8.2ab401.0024.GAE@google.com/T/#u Signed-off-by: Naman Gulati Signed-off-by: Pablo Neira Ayuso --- net/netfilter/nf_conntrack_netlink.c | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-) diff --git a/net/netfilter/nf_conntrack_netlink.c b/net/netfilter/nf_conntrack_netlink.c index 579ada063b1b..4e5d7c701436 100644 --- a/net/netfilter/nf_conntrack_netlink.c +++ b/net/netfilter/nf_conntrack_netlink.c @@ -3392,7 +3392,8 @@ static bool expect_iter_name(struct nf_conntrack_expect *exp, void *data) struct nf_conntrack_helper *helper; const char *name = data; - helper = rcu_dereference(exp->helper); + helper = rcu_dereference_protected(exp->helper, + lockdep_is_held(&nf_conntrack_expect_lock)); if (!helper) return false; From 70194dc37670bd08e44b471389861cc01bd3a3c9 Mon Sep 17 00:00:00 2001 From: Aohan Mei Date: Mon, 14 Sep 2026 19:51:47 +0800 Subject: [PATCH 360/592] netfilter: nf_tables: skip expired catchall elements on insert and delete nft_setelem_catchall_insert() looks up duplicates with nft_set_elem_active() only, while nft_set_catchall_lookup() and the dump path additionally skip expired elements. Once a catchall element with a timeout expires, this predicate drift makes it invisible to userspace dumps, yet it still blocks re-insertion: with NLM_F_EXCL the request fails with -EEXIST, and without it the request reports success but silently inserts nothing. The stale entry only goes away when the (user-tunable) gc interval elapses, so the catchall rule may silently stop matching for an arbitrarily long time after its first expiration. The delete path shows the same drift: nft_setelem_catchall_deactivate() picks the first active-next entry in the catchall list, so with an expired entry still pending GC it retires the stale entry instead of the fresh one, and it deactivates an element that userspace no longer sees instead of failing with -ENOENT. Align both walks with the lookup and dump predicates: only an element that is active and not expired counts as a duplicate or delete candidate, using the per-netns timestamp taken at transaction start, in line with the set backend .insert/.deactivate and catchall GC sync paths. Reported-by: TencentOS Corvus AI Cc: stable@vger.kernel.org Fixes: aaa31047a6d2 ("netfilter: nftables: add catch-all set element support") Assisted-by: CodeBuddy:Kimi-K3 Signed-off-by: Aohan Mei Signed-off-by: Pablo Neira Ayuso --- net/netfilter/nf_tables_api.c | 10 ++++++++-- 1 file changed, 8 insertions(+), 2 deletions(-) diff --git a/net/netfilter/nf_tables_api.c b/net/netfilter/nf_tables_api.c index c0b754a2d45b..b59628e6240c 100644 --- a/net/netfilter/nf_tables_api.c +++ b/net/netfilter/nf_tables_api.c @@ -6995,11 +6995,14 @@ static int nft_setelem_catchall_insert(const struct net *net, { struct nft_set_elem_catchall *catchall; u8 genmask = nft_genmask_next(net); + u64 tstamp = nft_net_tstamp(net); struct nft_set_ext *ext; list_for_each_entry(catchall, &set->catchall_list, list) { ext = nft_set_elem_ext(set, catchall->elem); - if (nft_set_elem_active(ext, genmask)) { + if (nft_set_elem_active(ext, genmask) && + !__nft_set_elem_expired(ext, tstamp) && + !nft_set_elem_is_dead(ext)) { *priv = catchall->elem; return -EEXIST; } @@ -7092,11 +7095,14 @@ static int nft_setelem_catchall_deactivate(const struct net *net, struct nft_set_elem *elem) { struct nft_set_elem_catchall *catchall; + u64 tstamp = nft_net_tstamp(net); struct nft_set_ext *ext; list_for_each_entry(catchall, &set->catchall_list, list) { ext = nft_set_elem_ext(set, catchall->elem); - if (!nft_is_active_next(net, ext)) + if (!nft_is_active_next(net, ext) || + __nft_set_elem_expired(ext, tstamp) || + nft_set_elem_is_dead(ext)) continue; kfree(elem->priv); From 88aed0422f39b22406f35f1e758cea25e7bbcfb5 Mon Sep 17 00:00:00 2001 From: Vinay Belgaumkar Date: Fri, 4 Sep 2026 11:16:24 -0700 Subject: [PATCH 361/592] perf: Fix null pointer access in is_include_guest_event() A typical module unload occurring event when there is an active perf connection leads to freeing of the pmu pointer. The call log is something like: .. __pmu_detach_event pmu_detach_event pmu_detach_events perf_pmu_unregister .. __pmu_detach_event() sets event->pmu to null. When the perf connection finally is closed, the following stack trace is observed: Oops: general protection fault, kernel NULL pointer dereference ... RIP: 0010:_free_event+0x3e/0x370 ... Call Trace: ... perf_event_release_kernel+0x260/0x2d0 perf_release+0x12/0x20 A call to mediated_pmu_unaccount_event() inside _free_event() is the root cause of this crash. Adding a check inside is_include_guest_event() ensures we don't accidentally access a null pmu ptr. In addition to this, we will now call mediated_pmu_unaccount_event() before clearing the pmu ptr so that nr_include_guest_events counts are maintained correctly. Fixes: eff95e170275 ("perf: Add APIs to create/release mediated guest vPMUs") Assisted-by: Claude:Claude-Sonnet-5 Signed-off-by: Vinay Belgaumkar Signed-off-by: Peter Zijlstra (Intel) Reviewed-by: Dapeng Mi Link: https://patch.msgid.link/20260904181625.1394082-1-vinay.belgaumkar@intel.com --- kernel/events/core.c | 4 ++++ 1 file changed, 4 insertions(+) diff --git a/kernel/events/core.c b/kernel/events/core.c index fe33fe15689d..db7b76d6b68a 100644 --- a/kernel/events/core.c +++ b/kernel/events/core.c @@ -6350,6 +6350,9 @@ static DEFINE_MUTEX(perf_mediated_pmu_mutex); /* !exclude_guest event of PMU with PERF_PMU_CAP_MEDIATED_VPMU */ static inline bool is_include_guest_event(struct perf_event *event) { + if (!event->pmu) + return false; + if ((event->pmu->capabilities & PERF_PMU_CAP_MEDIATED_VPMU) && !event->attr.exclude_guest) return true; @@ -13002,6 +13005,7 @@ static void __pmu_detach_event(struct pmu *pmu, struct perf_event *event, exclusive_event_destroy(event); module_put(pmu->module); + mediated_pmu_unaccount_event(event); event->pmu = NULL; /* force fault instead of UAF */ } From fe3c73d7bc769e7afc252f867a3421fe168b898d Mon Sep 17 00:00:00 2001 From: Andrea Righi Date: Tue, 15 Sep 2026 20:41:01 +0200 Subject: [PATCH 362/592] sched/core: Avoid false migration warning for proxy donors Proxy execution can move a blocked donor's scheduling context to the lock owner's CPU even when the donor is migration-disabled. The donor does not execute there, and its original execution CPU remains recorded in wake_cpu. set_task_cpu() warns unconditionally for migration-disabled tasks, so a subsequent proxy migration or the wakeup path returning the donor home triggers a false positive: moving a blocked scheduling context does not violate the migration-disabled execution context. For example, creating a mutex owner on CPU1 and a migration-disabled waiter on CPU0 can trigger the following warning: proxy_migrate_repro: donor blocking on CPU0 with migration disabled proxy_migrate_repro: donor moved from CPU0 to CPU1 WARNING: kernel/sched/core.c:3389 at set_task_cpu+0x1d3/0x280 ... Call Trace: try_to_wake_up+0x43f/0x780 __mutex_unlock_slowpath+0x330/0x540 owner_fn+0x9f/0xc0 [proxy_migrate_repro] ... proxy_migrate_repro: donor woke on CPU0, task_cpu=0 proxy_migrate_repro: completed Exclude blocked proxy donors from the warning. The proxy wakeup path restores an executable placement before clearing the blocked state. Fixes: b049b81bdff6 ("sched: Handle blocked-waiter migration (and return migration)") Signed-off-by: Andrea Righi Signed-off-by: Peter Zijlstra (Intel) Acked-by: John Stultz Link: https://patch.msgid.link/20260915184101.2621252-1-arighi@nvidia.com --- kernel/sched/core.c | 9 ++++++++- 1 file changed, 8 insertions(+), 1 deletion(-) diff --git a/kernel/sched/core.c b/kernel/sched/core.c index 7885ff76e69f..0b846a13c628 100644 --- a/kernel/sched/core.c +++ b/kernel/sched/core.c @@ -3351,6 +3351,8 @@ void relax_compatible_cpus_allowed_ptr(struct task_struct *p) void set_task_cpu(struct task_struct *p, unsigned int new_cpu) { unsigned int state = READ_ONCE(p->__state); + bool proxy_migrated = sched_proxy_exec() && p->is_blocked && + task_cpu(p) != p->wake_cpu; /* * We should never call set_task_cpu() on a blocked task, @@ -3386,7 +3388,12 @@ void set_task_cpu(struct task_struct *p, unsigned int new_cpu) */ WARN_ON_ONCE(!cpu_online(new_cpu)); - WARN_ON_ONCE(is_migration_disabled(p)); + /* + * Proxy execution can move a blocked task's scheduling context to any + * CPU without moving its migration-disabled execution context. The + * wakeup path will return the task to a CPU where it can execute. + */ + WARN_ON_ONCE(is_migration_disabled(p) && !proxy_migrated); trace_sched_migrate_task(p, new_cpu); From 93f53499d0b945e8ae447f497faf743d60069f61 Mon Sep 17 00:00:00 2001 From: Matthew Schwartz Date: Thu, 17 Sep 2026 16:09:06 -0700 Subject: [PATCH 363/592] x86/fred: Reconstruct the #GP context for rejected INT instructions FRED event delivery does not use the IDT, so the gate DPL check that rejects a user INT n falls to software (Intel FRED specification [1], section 8.3). fred_intx() rejects the same vectors as IDT delivery, but reports a zero error code and the IP after the INT. This breaks the signal ABI. Wine uses the error code to recognize INT 0x2d, so the changed context turns a handled breakpoint into an access violation in Elden Ring. Rewind IP using the instruction length in the augmented SS and synthesize the IDT selector error code, (vector << 3) | 2. Set RF in the saved flags, as the CPU does for a #GP fault. Section 5.2.1 defines the saved vector, instruction length and RF state. The supplied length handles prefixes without reading user memory. Limit the changes to already-rejected software interrupts, preserving the accepted INT3, INT4 and enabled INT80 paths and hardware exceptions. With IA32 emulation disabled, INT 0x80 now reports the same #GP as the DPL 0 gate IDT installs there. The rewound IP also stops fixup_iopl_exception() from inspecting the byte after the INT. Also clear the software event flag. Section 6.2.3 specifies that ERETU with this flag and TF set traps before executing any user instruction. A tracer that suppresses SIGSEGV and resumes with TF set expects the next instruction to run first, as after IRET. The sigreturn path clears the same flag for this reason in prevent_single_step_upon_eretu(). [1] Intel Flexible Return and Event Delivery (FRED) Specification, revision 9.0 (346446-009US), sections 5.2.1, 6.2.3 and 8.3. Fixes: 14619d912b65 ("x86/fred: FRED entry/exit and dispatch code") Closes: https://gitlab.freedesktop.org/mesa/mesa/-/work_items/15745 Closes: https://gitlab.freedesktop.org/mesa/mesa/-/work_items/16132 Reported-by: Paul Gofman Signed-off-by: Matthew Schwartz Signed-off-by: Peter Zijlstra (Intel) Reviewed-by: H. Peter Anvin Link: https://cdrdv2.intel.com/v1/dl/getContent/678938 # [1] Link: https://patch.msgid.link/20260917230907.2080792-2-matthew.schwartz@linux.dev --- arch/x86/entry/entry_fred.c | 11 ++++++++++- 1 file changed, 10 insertions(+), 1 deletion(-) diff --git a/arch/x86/entry/entry_fred.c b/arch/x86/entry/entry_fred.c index fb3594ddf731..854899bfec5d 100644 --- a/arch/x86/entry/entry_fred.c +++ b/arch/x86/entry/entry_fred.c @@ -10,6 +10,7 @@ #include #include #include +#include #include #include #include @@ -71,7 +72,15 @@ static noinstr void fred_intx(struct pt_regs *regs) #endif default: - return exc_general_protection(regs, 0); + /* + * Reconstruct the #GP fault state that IDT delivery would produce. + * Clear the software event flag so ERETU with TF set does not trap + * before the resumed instruction. See prevent_single_step_upon_eretu(). + */ + regs->ip -= regs->fred_ss.insnlen; + regs->flags |= X86_EFLAGS_RF; + regs->fred_ss.swevent = 0; + return exc_general_protection(regs, (regs->fred_ss.vector << 3) | 2); } } From 96443a53bc3ef4b67dab0c497878fc8d56f799f3 Mon Sep 17 00:00:00 2001 From: Matthew Schwartz Date: Thu, 17 Sep 2026 16:09:07 -0700 Subject: [PATCH 364/592] selftests/x86: Check signal state for rejected software interrupts Add a test of the signal ABI for INT instructions in both 32-bit and 64-bit processes. Check the signal number, trap number, error code, si_code, si_addr, instruction pointer and RF/TF state against legacy IDT behavior. Include a 15-byte prefixed INT to check that IP uses the hardware instruction length. Exercise both INT3 encodings, INT4, UD2 and HLT to cover the unchanged trap and fault paths. Run each instruction with TF clear and set. Resume at a known NOP after handling the signal and check that single-stepping traps after the NOP. Also drive INT 0x2d under ptrace, which resumes through the fault frame rather than sigreturn and so exposes a stale FRED software event flag. Start from an INT3 stop, whose FRED frame has no software event flag, instead of the syscall frame of raise(SIGSTOP). Single-step into the INT and check that the fault reports its address. Then suppress SIGSEGV and resume at the NOP, once with PTRACE_SINGLESTEP and once with PTRACE_CONT and TF set. Section 6.2.3 of the Intel FRED specification [1] specifies the immediate single-step trap caused by returning with both that flag and TF set. Check that each trap occurs after the NOP, rather than at its address. Report whether the CPU supports FRED, since a pass looks the same on either entry path. INT 0x80 with IA32 emulation disabled and a 64-bit tracer of a 32-bit tracee are not covered. Both variants pass all 29 checks on a non-FRED AMD host and on Panther Lake with FRED enabled and the fix applied. With the same binaries on unpatched Panther Lake, 16 signal-context checks fail and the first ptrace check reports the IP after the INT. The two dependent ptrace resume checks are not reached. [1] Intel Flexible Return and Event Delivery (FRED) Specification, revision 9.0 (346446-009US), section 6.2.3. Signed-off-by: Matthew Schwartz Signed-off-by: Peter Zijlstra (Intel) Link: https://cdrdv2.intel.com/v1/dl/getContent/678938 # [1] Link: https://patch.msgid.link/20260917230907.2080792-3-matthew.schwartz@linux.dev --- tools/testing/selftests/x86/Makefile | 2 +- tools/testing/selftests/x86/int_signal.c | 311 +++++++++++++++++++++++ 2 files changed, 312 insertions(+), 1 deletion(-) create mode 100644 tools/testing/selftests/x86/int_signal.c diff --git a/tools/testing/selftests/x86/Makefile b/tools/testing/selftests/x86/Makefile index 434065215d12..d478b13cc8d5 100644 --- a/tools/testing/selftests/x86/Makefile +++ b/tools/testing/selftests/x86/Makefile @@ -13,7 +13,7 @@ CAN_BUILD_WITH_NOPIE := $(shell ./check_cc.sh "$(CC)" trivial_program.c -no-pie) TARGETS_C_BOTHBITS := single_step_syscall sysret_ss_attrs syscall_nt test_mremap_vdso \ check_initial_reg_state sigreturn iopl ioperm \ test_vsyscall mov_ss_trap sigtrap_loop \ - syscall_arg_fault fsgsbase_restore sigaltstack + syscall_arg_fault fsgsbase_restore sigaltstack int_signal TARGETS_C_BOTHBITS += nx_stack TARGETS_C_32BIT_ONLY := entry_from_vm86 test_syscall_vdso unwind_vdso \ test_FCMOV test_FCOMI test_FISTTP \ diff --git a/tools/testing/selftests/x86/int_signal.c b/tools/testing/selftests/x86/int_signal.c new file mode 100644 index 000000000000..22676dac72b5 --- /dev/null +++ b/tools/testing/selftests/x86/int_signal.c @@ -0,0 +1,311 @@ +// SPDX-License-Identifier: GPL-2.0-only +/* Check the signal context for INT instructions with IDT and FRED entry. */ +#define _GNU_SOURCE + +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include + +#include "helpers.h" + +#ifdef __x86_64__ +#define REG_IP REG_RIP +#define USER_IP rip +#define STACK_PTR "%rsp" +#else +#define REG_IP REG_EIP +#define USER_IP eip +#define STACK_PTR "%esp" +#endif + +/* + * Each instruction has normal and single-step entry points. Resume at the + * NOP after handling its signal, then expect a trace trap after that NOP + * when TF is set. Explicit labels avoid assuming the kernel's saved IP. + */ +#define PROBE(name, insn) \ + extern void name(void); \ + extern void name##_tf(void); \ + extern const char name##_end[], name##_step[]; \ + asm(".pushsection .text\n" \ + ".globl " #name "_tf\n" \ + ".type " #name "_tf, @function\n" \ + #name "_tf:\n" \ + "pushf\n" \ + "orl $0x100, (" STACK_PTR ")\n" \ + "popf\n" \ + ".globl " #name "\n" \ + ".type " #name ", @function\n" \ + #name ":\n" insn "\n" \ + ".globl " #name "_end\n" \ + #name "_end:\nnop\n" \ + ".globl " #name "_step\n" \ + #name "_step:\nret\n" \ + ".size " #name ", .-" #name "\n" \ + ".size " #name "_tf, .-" #name "_tf\n" \ + ".popsection\n") + +PROBE(int1, ".byte 0xcd, 0x01"); +PROBE(int29, ".byte 0xcd, 0x29"); +PROBE(int2c, ".byte 0xcd, 0x2c"); +PROBE(int2d, ".byte 0xcd, 0x2d"); +PROBE(prefixed_int2d, ".byte 0x66, 0xcd, 0x2d"); +PROBE(long_int2d, ".fill 13, 1, 0x2e\n.byte 0xcd, 0x2d"); +PROBE(int81, ".byte 0xcd, 0x81"); +PROBE(intff, ".byte 0xcd, 0xff"); +PROBE(short_int3, ".byte 0xcc"); +PROBE(long_int3, ".byte 0xcd, 0x03"); +PROBE(int4, ".byte 0xcd, 0x04"); +PROBE(ud2, ".byte 0x0f, 0x0b"); +PROBE(hlt, ".byte 0xf4"); + +struct test { + const char *name; + void (*run)(void); + void (*run_tf)(void); + const char *end, *step; + int signo, trap, error, ip_offset, flags, code; +}; + +#define TEST(name, sig, trap, error, offset, flags, code) \ + { #name, name, name##_tf, name##_end, name##_step, \ + sig, trap, error, offset, flags, code } + +#define GP(name, error) \ + TEST(name, SIGSEGV, 13, error, 0, X86_EFLAGS_RF, SI_KERNEL) + +static const struct test tests[] = { + GP(int1, 0x00a), + GP(int29, 0x14a), + GP(int2c, 0x162), + GP(int2d, 0x16a), + GP(prefixed_int2d, 0x16a), + GP(long_int2d, 0x16a), + GP(int81, 0x40a), + GP(intff, 0x7fa), + GP(hlt, 0), + TEST(short_int3, SIGTRAP, 3, 0, 1, 0, SI_KERNEL), + TEST(long_int3, SIGTRAP, 3, 0, 2, 0, SI_KERNEL), + TEST(int4, SIGSEGV, 4, 0, 2, 0, SI_KERNEL), + TEST(ud2, SIGILL, 6, 0, 0, X86_EFLAGS_RF, ILL_ILLOPN), +}; + +static const struct test *active; +static volatile sig_atomic_t seen, signo, trap, error, ip_offset, flags; +static volatile sig_atomic_t code, addr_ok, single_step, stepped, step_ok; + +static void handler(int sig, siginfo_t *info, void *context) +{ + ucontext_t *uc = context; + uintptr_t ip = uc->uc_mcontext.gregs[REG_IP]; + uintptr_t start = (uintptr_t)active->run; + uintptr_t end = (uintptr_t)active->end; + + if (seen && single_step && sig == SIGTRAP) { + if (stepped++) { + ksft_print_msg("%s: second trace trap at %#lx\n", + active->name, (unsigned long)ip); + _exit(KSFT_FAIL); + } + step_ok = ip == (uintptr_t)active->step && + uc->uc_mcontext.gregs[REG_TRAPNO] == 1 && + info->si_code == TRAP_TRACE; + uc->uc_mcontext.gregs[REG_EFL] &= ~X86_EFLAGS_TF; + return; + } + + if (seen || ip < start || ip > end) { + ksft_print_msg("%s: unexpected signal %d at %#lx\n", + active->name, sig, (unsigned long)ip); + _exit(KSFT_FAIL); + } + + signo = sig; + trap = uc->uc_mcontext.gregs[REG_TRAPNO]; + error = uc->uc_mcontext.gregs[REG_ERR]; + ip_offset = ip - start; + flags = uc->uc_mcontext.gregs[REG_EFL] & (X86_EFLAGS_RF | X86_EFLAGS_TF); + code = info->si_code; + /* force_sig() reports no address, force_sig_fault() reports the IP. */ + addr_ok = info->si_addr == (code == SI_KERNEL ? NULL : (void *)ip); + seen = 1; + uc->uc_mcontext.gregs[REG_IP] = end; +} + +static void wait_for_child(pid_t child, int *status) +{ + pid_t ret; + + do { + ret = waitpid(child, status, 0); + } while (ret < 0 && errno == EINTR); + if (ret != child) + ksft_exit_fail_perror("waitpid"); +} + +/* Resume the tracee and check where the next stop lands. */ +static bool resume_to(pid_t child, int *status, int request, int sig, + const void *ip, const char *what) +{ + struct user_regs_struct regs; + + if (ptrace(request, child, 0, 0)) + return false; + wait_for_child(child, status); + if (!WIFSTOPPED(*status)) { + ksft_print_msg("%s: tracee did not stop\n", what); + return false; + } + if (WSTOPSIG(*status) != sig) { + ksft_print_msg("%s: stopped with signal %d, expected %d\n", + what, WSTOPSIG(*status), sig); + return false; + } + if (ptrace(PTRACE_GETREGS, child, 0, ®s)) + return false; + if ((unsigned long)regs.USER_IP != (unsigned long)ip) { + ksft_print_msg("%s: stopped at %#lx, expected %#lx\n", what, + (unsigned long)regs.USER_IP, (unsigned long)ip); + return false; + } + return true; +} + +static bool set_ip(pid_t child, const void *ip, bool tf) +{ + struct user_regs_struct regs; + + if (ptrace(PTRACE_GETREGS, child, 0, ®s)) + return false; + regs.USER_IP = (unsigned long)ip; + if (tf) + regs.eflags |= X86_EFLAGS_TF; + return !ptrace(PTRACE_SETREGS, child, 0, ®s); +} + +/* + * Exercise the tracer paths that resume through the fault frame rather than + * sigreturn. A stale FRED software event flag on that frame traps before the + * NOP executes instead of after it. + */ +static void test_ptrace(void) +{ + bool into = false, step = false, cont = false; + pid_t child; + int status; + + child = fork(); + if (child < 0) + ksft_exit_fail_perror("fork"); + if (!child) { + if (ptrace(PTRACE_TRACEME, 0, 0, 0)) + _exit(KSFT_FAIL); + /* Start from a breakpoint frame, not the syscall frame of raise(). */ + asm volatile("int3"); + _exit(KSFT_FAIL); + } + + wait_for_child(child, &status); + if (!WIFSTOPPED(status) || WSTOPSIG(status) != SIGTRAP) + goto out; + if (ptrace(PTRACE_SETOPTIONS, child, 0, PTRACE_O_EXITKILL)) + goto out; + + /* Single-step into the INT. The fault must report the INT's address. */ + if (!set_ip(child, int2d, false)) + goto out; + into = resume_to(child, &status, PTRACE_SINGLESTEP, SIGSEGV, int2d, + "single-step into INT"); + if (!into) + goto out; + + /* Suppress SIGSEGV and single-step the NOP. */ + if (!set_ip(child, int2d_end, false)) + goto out; + step = resume_to(child, &status, PTRACE_SINGLESTEP, SIGTRAP, int2d_step, + "single-step after INT"); + if (!step) + goto out; + + /* Fault again, then suppress SIGSEGV and continue with TF set. */ + if (!set_ip(child, int2d, false)) + goto out; + if (!resume_to(child, &status, PTRACE_CONT, SIGSEGV, int2d, + "continue to INT")) + goto out; + if (!set_ip(child, int2d_end, true)) + goto out; + cont = resume_to(child, &status, PTRACE_CONT, SIGTRAP, int2d_step, + "continue with TF after INT"); +out: + if (WIFSTOPPED(status)) { + kill(child, SIGKILL); + wait_for_child(child, &status); + } + ksft_test_result(into, "ptrace single-step into INT faults at the INT\n"); + ksft_test_result(step, "ptrace single-step after suppressing SIGSEGV\n"); + ksft_test_result(cont, "ptrace continue with TF after suppressing SIGSEGV\n"); +} + +static bool cpu_has_fred(void) +{ + unsigned int eax, ebx, ecx, edx; + + if (__get_cpuid_max(0, NULL) < 7) + return false; + __cpuid_count(7, 1, eax, ebx, ecx, edx); + return eax & (1 << 17); +} + +int main(void) +{ + unsigned int i, tf; + int expected_flags, ok; + + ksft_print_header(); + ksft_set_plan(2 * ARRAY_SIZE(tests) + 3); + ksft_print_msg("CPU %s FRED\n", cpu_has_fred() ? "supports" : "lacks"); + sethandler(SIGSEGV, handler, 0); + sethandler(SIGTRAP, handler, 0); + sethandler(SIGILL, handler, 0); + + for (tf = 0; tf < 2; tf++) { + for (i = 0; i < ARRAY_SIZE(tests); i++) { + active = &tests[i]; + single_step = tf; + seen = signo = trap = error = ip_offset = flags = 0; + code = addr_ok = stepped = step_ok = 0; + expected_flags = active->flags | (tf ? X86_EFLAGS_TF : 0); + if (tf) + active->run_tf(); + else + active->run(); + + ok = seen && signo == active->signo && trap == active->trap && + error == active->error && ip_offset == active->ip_offset && + flags == expected_flags && code == active->code && addr_ok && + (!tf || (stepped && step_ok)); + ksft_test_result(ok, "%s%s\n", active->name, tf ? " with TF" : ""); + if (!ok) { + ksft_print_msg("got signal=%d trap=%d error=%#x ip=%d\n", + signo, trap, error, ip_offset); + ksft_print_msg("got flags=%#x code=%d addr_ok=%d step_ok=%d\n", + flags, code, addr_ok, step_ok); + ksft_print_msg("expected signal=%d trap=%d error=%#x ip=%d\n", + active->signo, active->trap, active->error, + active->ip_offset); + ksft_print_msg("expected flags=%#x code=%d\n", + expected_flags, active->code); + } + } + } + test_ptrace(); + ksft_finished(); +} From 406aa2b186d3f13a35bc1ad6aff4274917851bc4 Mon Sep 17 00:00:00 2001 From: Robin Murphy Date: Tue, 15 Sep 2026 13:24:22 +0100 Subject: [PATCH 365/592] perf/arm-cmn: Fix multi-filter encoding The current special-case for EVICT_STATE_SEL filtering effectively assigns the "filter" and "filter2" controls in the opposite order from how the CMN S3 r2 TRM states "Filtering is programmed in pmu_hbt_lbt_sel and pmu_evict_state_sel". On reflection, not only does this seem unnecessarily non-obvious to users, but it's also likely to be a problem for scaling to a full multi-filter abstraction in future. There is a logical order to filters based on their bitfield positions in the pmu_event_sel register, which the TRM descriptions allude to, and the cmn_filter_select enum already (almost) follows, so let's fix the UABI to follow suit while it's still unreleased. Fixes: 09178f536bb9 ("perf/arm-cmn: Plumb in new filter types") Signed-off-by: Robin Murphy Reviewed-by: Ilkka Koskinen Reviewed-by: Leo Yan Signed-off-by: Will Deacon --- drivers/perf/arm-cmn.c | 58 +++++++++++++++++++++++++----------------- 1 file changed, 34 insertions(+), 24 deletions(-) diff --git a/drivers/perf/arm-cmn.c b/drivers/perf/arm-cmn.c index 33ee2be9b386..5378fba916cf 100644 --- a/drivers/perf/arm-cmn.c +++ b/drivers/perf/arm-cmn.c @@ -305,9 +305,9 @@ enum cmn_filter_select { SEL_NONE, SEL_OCCUP1_ID, SEL_CLASS_OCCUP_ID, - SEL_CBUSY_SNTHROTTLE_SEL, SEL_HBT_LBT_SEL, SEL_SN_HOME_SEL, + SEL_CBUSY_SNTHROTTLE_SEL, SEL_SNP_VC_SEL, SEL_ENHANCED_HBT_LBT_SEL, SEL_EVICT_STATE_SEL, @@ -978,10 +978,14 @@ static umode_t arm_cmn_event_attr_is_visible(struct kobject *kobj, _CMN_EVENT_HNS(_model, _name##_all, _event, _sel, 0), \ _CMN_EVENT_HNS(_model, _name##_hbt, _event, _sel, 1), \ _CMN_EVENT_HNS(_model, _name##_lbt, _event, _sel, 2) -#define _CMN_EVENT_HNS_HBT2(_model, _name, _event, _fsel1, f1) \ - _CMN_EVENT_HNS(_model, _name##_all, _event, _fsel1, f1, SEL_HBT_LBT_SEL, 0), \ - _CMN_EVENT_HNS(_model, _name##_hbt, _event, _fsel1, f1, SEL_HBT_LBT_SEL, 1), \ - _CMN_EVENT_HNS(_model, _name##_lbt, _event, _fsel1, f1, SEL_HBT_LBT_SEL, 2) +#define _CMN_EVENT_HNS_EVICT(_model, _name, _event, _fsel1, f1) \ + _CMN_EVENT_HNS(_model, _name##_all, _event, _fsel1, f1, SEL_EVICT_STATE_SEL, 0), \ + _CMN_EVENT_HNS(_model, _name##_eu, _event, _fsel1, f1, SEL_EVICT_STATE_SEL, 1), \ + _CMN_EVENT_HNS(_model, _name##_en, _event, _fsel1, f1, SEL_EVICT_STATE_SEL, 2), \ + _CMN_EVENT_HNS(_model, _name##_su, _event, _fsel1, f1, SEL_EVICT_STATE_SEL, 3), \ + _CMN_EVENT_HNS(_model, _name##_sn, _event, _fsel1, f1, SEL_EVICT_STATE_SEL, 4), \ + _CMN_EVENT_HNS(_model, _name##_mu, _event, _fsel1, f1, SEL_EVICT_STATE_SEL, 5), \ + _CMN_EVENT_HNS(_model, _name##_mn, _event, _fsel1, f1, SEL_EVICT_STATE_SEL, 6) #define CMN_EVENT_HNS_OCC(_model, _name, _event) \ CMN_EVENT_HN_OCC(_model, hns_##_name, CMN_TYPE_HNS, _event), \ @@ -1014,13 +1018,9 @@ static umode_t arm_cmn_event_attr_is_visible(struct kobject *kobj, _CMN_EVENT_HNS(CMNS3R2, _name##_ccglcn, _event, SEL_ENHANCED_HBT_LBT_SEL, 5), \ _CMN_EVENT_HNS(CMNS3R2, _name##_ccgrn, _event, SEL_ENHANCED_HBT_LBT_SEL, 6) #define CMN_EVENT_HNS_EVICT(_model, _name, _event) \ - _CMN_EVENT_HNS_HBT2(_model, _name##_all, _event, SEL_EVICT_STATE_SEL, 0), \ - _CMN_EVENT_HNS_HBT2(_model, _name##_eu, _event, SEL_EVICT_STATE_SEL, 1), \ - _CMN_EVENT_HNS_HBT2(_model, _name##_en, _event, SEL_EVICT_STATE_SEL, 2), \ - _CMN_EVENT_HNS_HBT2(_model, _name##_su, _event, SEL_EVICT_STATE_SEL, 3), \ - _CMN_EVENT_HNS_HBT2(_model, _name##_sn, _event, SEL_EVICT_STATE_SEL, 4), \ - _CMN_EVENT_HNS_HBT2(_model, _name##_mu, _event, SEL_EVICT_STATE_SEL, 5), \ - _CMN_EVENT_HNS_HBT2(_model, _name##_mn, _event, SEL_EVICT_STATE_SEL, 6) + _CMN_EVENT_HNS_EVICT(_model, _name##_all, _event, SEL_HBT_LBT_SEL, 0), \ + _CMN_EVENT_HNS_EVICT(_model, _name##_hbt, _event, SEL_HBT_LBT_SEL, 1), \ + _CMN_EVENT_HNS_EVICT(_model, _name##_lbt, _event, SEL_HBT_LBT_SEL, 2) #define CMN_EVENT_HNSR0_HBT(_name, _event) \ _CMN_EVENT_HNS_HBT(CMN700 | CMNS3R01, _name, _event, SEL_HBT_LBT_SEL) @@ -1860,10 +1860,12 @@ static void arm_cmn_val_add_event(struct arm_cmn *cmn, struct arm_cmn_val *val, val->dtm_count[dtm]++; + if (sel == SEL_EVICT_STATE_SEL) { + val->filter[dtm][sel] = CMN_EVENT_FILTER2(event) + 1; + sel = SEL_HBT_LBT_SEL; + } if (sel) val->filter[dtm][sel] = CMN_EVENT_FILTER(event) + 1; - if (sel == SEL_EVICT_STATE_SEL) - val->filter[dtm][SEL_HBT_LBT_SEL] = CMN_EVENT_FILTER2(event) + 1; if (type != CMN_TYPE_WP) continue; @@ -1914,14 +1916,17 @@ static int arm_cmn_validate_group(struct arm_cmn *cmn, struct perf_event *event) if (val->dtm_count[dtm] == CMN_DTM_NUM_COUNTERS) goto done; + if (sel == SEL_EVICT_STATE_SEL) { + if (val->filter[dtm][sel] && + val->filter[dtm][sel] != CMN_EVENT_FILTER2(event) + 1) + goto done; + sel = SEL_HBT_LBT_SEL; + } + if (sel && val->filter[dtm][sel] && val->filter[dtm][sel] != CMN_EVENT_FILTER(event) + 1) goto done; - if (sel == SEL_EVICT_STATE_SEL && val->filter[dtm][SEL_HBT_LBT_SEL] && - val->filter[dtm][SEL_HBT_LBT_SEL] != CMN_EVENT_FILTER2(event) + 1) - goto done; - if (type != CMN_TYPE_WP) continue; @@ -1950,7 +1955,7 @@ static enum cmn_filter_select arm_cmn_event_filter(const struct arm_cmn *cmn, for (int i = 0; i < ARRAY_SIZE(arm_cmn_event_attrs) - 1; i++) { e = container_of(arm_cmn_event_attrs[i], typeof(*e), attr.attr); if (e->model & model && e->type == type && e->eventid == eventid) - return e->filter[0].sel; + return e->filter[1].sel ?: e->filter[0].sel; } return SEL_NONE; } @@ -2081,16 +2086,21 @@ static void arm_cmn_event_clear(struct arm_cmn *cmn, struct perf_event *event, static int arm_cmn_set_event_filter(struct arm_cmn_node *dn, struct perf_event *event) { enum cmn_filter_select fsel = to_cmn_hw(event)->filter_sel; + bool evict_state = fsel == SEL_EVICT_STATE_SEL; int ret = 0; + if (evict_state) { + ret = arm_cmn_set_event_sel_hi(dn, fsel, CMN_EVENT_FILTER2(event)); + if (ret) + return ret; + fsel = SEL_HBT_LBT_SEL; + } if (fsel) ret = arm_cmn_set_event_sel_hi(dn, fsel, CMN_EVENT_FILTER(event)); - if (fsel == SEL_EVICT_STATE_SEL && !ret) { - ret = arm_cmn_set_event_sel_hi(dn, SEL_HBT_LBT_SEL, CMN_EVENT_FILTER2(event)); - if (ret) - dn->filter[fsel].count--; - } + if (ret && evict_state) + dn->filter[SEL_EVICT_STATE_SEL].count--; + return ret; } From bb756b11ad63832ebee58caf9e8f9381eaecff9f Mon Sep 17 00:00:00 2001 From: Zeng Heng Date: Fri, 11 Sep 2026 09:58:59 +0800 Subject: [PATCH 366/592] arm64: io: Reject non-user protection in ioremap_prot() Mapping a stack-top page via /dev/mem with PROT_NONE and then reading that process's /proc//cmdline triggers a spurious WARN in ioremap_prot() through generic_access_phys(): WARNING: ./arch/arm64/include/asm/io.h:275 at generic_access_phys Call trace: generic_access_phys+0x1c8/0x228 (P) __access_remote_vm+0x2b4/0x398 access_remote_vm+0x14/0x30 get_mm_cmdline+0xf8/0x2a0 proc_pid_cmdline_read+0x68/0x120 generic_access_phys() passes the protection derived from the user PTE to ioremap_prot(). On arm64, a PROT_NONE mapping is represented by a present-invalid PTE, so pte_present() still returns true and the protection reaches ioremap_prot(). A PROT_NONE mapping does not have PTE_USER, causing the existing WARN_ON_ONCE() in ioremap_prot() to fire even though this is a valid user mapping. Execute-only mappings have the same issue and must not be readable through this path either. ioremap_prot() should therefore reject protection values without PTE_USER without warning. This makes the access fail cleanly for PROT_NONE and execute-only mappings while retaining the existing user-protection contract. Fixes: 8f098037139b ("arm64: io: Extract user memory type in ioremap_prot()") Signed-off-by: Zeng Heng Reviewed-by: Catalin Marinas Signed-off-by: Will Deacon --- arch/arm64/include/asm/io.h | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-) diff --git a/arch/arm64/include/asm/io.h b/arch/arm64/include/asm/io.h index 49a7002661a9..9aa0bb08ab60 100644 --- a/arch/arm64/include/asm/io.h +++ b/arch/arm64/include/asm/io.h @@ -280,7 +280,8 @@ static inline void __iomem *ioremap_prot(phys_addr_t phys, size_t size, pgprot_t prot; ptval_t user_prot_val = pgprot_val(user_prot); - if (WARN_ON_ONCE(!(user_prot_val & PTE_USER))) + /* Reject PROT_NONE and exec-only */ + if (!(user_prot_val & PTE_USER)) return NULL; prot = __pgprot_modify(PAGE_KERNEL, PTE_ATTRINDX_MASK, From baaa9b126b875b40ffd9356da52c7c871917e5bb Mon Sep 17 00:00:00 2001 From: Yureka Lilian Date: Fri, 11 Sep 2026 19:02:10 +0200 Subject: [PATCH 367/592] arm64: Add override for WFxT Add an override for WFxT support within ID_AA64ISAR2_EL1 to allow it to be disabled using the new arm64.nowfxt command line parameter. This accompanies the idle=nop param introduced in a previous patch series [1] in dealing with misbehaving WFI and WFIT instructions on Apple Silicon SoCs, and eases debugging of other quirky WFxT implementations. Link[1]: https://lore.kernel.org/all/20260804-arm64-idle-param-v3-1-d10f8159062b@cyberchaos.dev/ Suggested-by: Will Deacon Signed-off-by: Yureka Lilian Signed-off-by: Will Deacon --- Documentation/admin-guide/kernel-parameters.txt | 3 +++ arch/arm64/kernel/pi/idreg-override.c | 2 ++ 2 files changed, 5 insertions(+) diff --git a/Documentation/admin-guide/kernel-parameters.txt b/Documentation/admin-guide/kernel-parameters.txt index 68647ff4bdd2..0fe9e36ce61a 100644 --- a/Documentation/admin-guide/kernel-parameters.txt +++ b/Documentation/admin-guide/kernel-parameters.txt @@ -589,6 +589,9 @@ Kernel parameters arm64.nosve [ARM64] Unconditionally disable Scalable Vector Extension support + arm64.nowfxt [ARM64] Unconditionally disable Wait For Event with + Timeout and Wait For Interrupt with Timeout support + ataflop= [HW,M68k] atarimouse= [HW,MOUSE] Atari Mouse diff --git a/arch/arm64/kernel/pi/idreg-override.c b/arch/arm64/kernel/pi/idreg-override.c index 274bc72ba98d..acaf56234973 100644 --- a/arch/arm64/kernel/pi/idreg-override.c +++ b/arch/arm64/kernel/pi/idreg-override.c @@ -186,6 +186,7 @@ static const struct ftr_set_desc isar2 __prel64_initconst = { .name = "id_aa64isar2", .override = &id_aa64isar2_override, .fields = { + FIELD("wfxt", ID_AA64ISAR2_EL1_WFxT_SHIFT, NULL), FIELD("gpa3", ID_AA64ISAR2_EL1_GPA3_SHIFT, NULL), FIELD("apa3", ID_AA64ISAR2_EL1_APA3_SHIFT, NULL), FIELD("mops", ID_AA64ISAR2_EL1_MOPS_SHIFT, NULL), @@ -254,6 +255,7 @@ static const struct { "id_aa64isar1.api=0 id_aa64isar1.apa=0 " "id_aa64isar2.gpa3=0 id_aa64isar2.apa3=0" }, { "arm64.nomops", "id_aa64isar2.mops=0" }, + { "arm64.nowfxt", "id_aa64isar2.wfxt=0" }, { "arm64.nomte", "id_aa64pfr1.mte=0" }, { "nokaslr", "arm64_sw.nokaslr=1" }, { "rodata=off", "arm64_sw.rodataoff=1" }, From b7403afb7a5f85073243df10238b3483958ad69e Mon Sep 17 00:00:00 2001 From: David Carlier Date: Sun, 6 Sep 2026 13:14:16 +0100 Subject: [PATCH 368/592] arm64: errata: match the target implementation CPU's own MIDR __is_affected_midr_range() is handed the MIDR and REVIDR of one target implementation CPU, but tests the erratum's range with is_midr_in_range(), which re-scans all of target_impl_cpus[] and ignores the @midr argument. The range test is thus constant across the per-CPU loop in is_affected_midr_range() and only answers "is any target CPU in range". Since just the fixed_revs REVIDR check uses the iteration's own registers, an out-of-range target CPU can decide whether a MIDR_FIXED() exemption applies. A VM then enables a workaround whose only in-range CPU is fixed silicon, e.g. erratum 2658417 on a Cortex-A510 r1p1 with REVIDR_EL1[25] set. Factor the range test into __is_midr_in_range(), which takes an explicit MIDR, and use it in __is_affected_midr_range(). Fixes: 86edf6bdcf05 ("smccc/kvm_guest: Enable errata based on implementation CPUs") Cc: stable@vger.kernel.org Assisted-by: Claude:claude-opus-5 Signed-off-by: David Carlier Signed-off-by: Will Deacon --- arch/arm64/kernel/cpu_errata.c | 15 +++++++++------ 1 file changed, 9 insertions(+), 6 deletions(-) diff --git a/arch/arm64/kernel/cpu_errata.c b/arch/arm64/kernel/cpu_errata.c index b33dccfafaf8..8ec47d89b45b 100644 --- a/arch/arm64/kernel/cpu_errata.c +++ b/arch/arm64/kernel/cpu_errata.c @@ -28,18 +28,21 @@ bool cpu_errata_set_target_impl(u64 num, void *impl_cpus) return true; } +static inline bool __is_midr_in_range(u32 midr, struct midr_range const *range) +{ + return midr_is_cpu_model_range(midr, range->model, + range->rv_min, range->rv_max); +} + static inline bool is_midr_in_range(struct midr_range const *range) { int i; if (!target_impl_cpu_num) - return midr_is_cpu_model_range(read_cpuid_id(), range->model, - range->rv_min, range->rv_max); + return __is_midr_in_range(read_cpuid_id(), range); for (i = 0; i < target_impl_cpu_num; i++) { - if (midr_is_cpu_model_range(target_impl_cpus[i].midr, - range->model, - range->rv_min, range->rv_max)) + if (__is_midr_in_range(target_impl_cpus[i].midr, range)) return true; } return false; @@ -59,7 +62,7 @@ __is_affected_midr_range(const struct arm64_cpu_capabilities *entry, u32 midr, u32 revidr) { const struct arm64_midr_revidr *fix; - if (!is_midr_in_range(&entry->midr_range)) + if (!__is_midr_in_range(midr, &entry->midr_range)) return false; midr &= MIDR_REVISION_MASK | MIDR_VARIANT_MASK; From 4485a01f4df1c9683d8ffe3e4ade6c33c9572d3c Mon Sep 17 00:00:00 2001 From: Sean Young Date: Sun, 13 Sep 2026 15:12:27 +0100 Subject: [PATCH 369/592] parisc: unwind: Replace open-coded binary search with bsearch() There is a bug in the binary search where hi can underflow. If addr is less than the first entry, then "hi = mid - 1" will underflow to ULONG_MAX. Then we have an out-of-bounds read. Replace the open-coded binary search with bsearch(). Issue found by an LLM. Signed-off-by: Sean Young Signed-off-by: Helge Deller --- arch/parisc/kernel/unwind.c | 33 +++++++++++++++------------------ 1 file changed, 15 insertions(+), 18 deletions(-) diff --git a/arch/parisc/kernel/unwind.c b/arch/parisc/kernel/unwind.c index 32103a270a8e..fab9ae22191a 100644 --- a/arch/parisc/kernel/unwind.c +++ b/arch/parisc/kernel/unwind.c @@ -14,6 +14,7 @@ #include #include #include +#include #include #include @@ -49,27 +50,23 @@ static DEFINE_SPINLOCK(unwind_lock); static struct unwind_table kernel_unwind_table __ro_after_init; static LIST_HEAD(unwind_tables); +static int cmp_unwind_entry(const void *key, const void *elt) +{ + unsigned long addr = (unsigned long)key; + const struct unwind_table_entry *e = elt; + + if (addr < e->region_start) + return -1; + if (addr > e->region_end) + return 1; + return 0; +} + static inline const struct unwind_table_entry * find_unwind_entry_in_table(const struct unwind_table *table, unsigned long addr) { - const struct unwind_table_entry *e = NULL; - unsigned long lo, hi, mid; - - lo = 0; - hi = table->length - 1; - - while (lo <= hi) { - mid = (hi - lo) / 2 + lo; - e = &table->table[mid]; - if (addr < e->region_start) - hi = mid - 1; - else if (addr > e->region_end) - lo = mid + 1; - else - return e; - } - - return NULL; + return bsearch((void *)addr, table->table, table->length, + sizeof(*table->table), cmp_unwind_entry); } static const struct unwind_table_entry * From cb917b1e1c23f6f9b73802cd576b48bd606458c0 Mon Sep 17 00:00:00 2001 From: Ethan Nelson-Moore Date: Thu, 17 Sep 2026 20:04:14 -0700 Subject: [PATCH 370/592] parisc: remove unused header The last use of arch/parisc/include/asm/compat_ucontext.h was removed in commit d313d4e72df3 ("parisc: remove unused compat_rt_sigframe.h header") but it was left behind. Remove it. Signed-off-by: Ethan Nelson-Moore Signed-off-by: Helge Deller --- arch/parisc/include/asm/compat_ucontext.h | 18 ------------------ 1 file changed, 18 deletions(-) delete mode 100644 arch/parisc/include/asm/compat_ucontext.h diff --git a/arch/parisc/include/asm/compat_ucontext.h b/arch/parisc/include/asm/compat_ucontext.h deleted file mode 100644 index c606f1bc891d..000000000000 --- a/arch/parisc/include/asm/compat_ucontext.h +++ /dev/null @@ -1,18 +0,0 @@ -/* SPDX-License-Identifier: GPL-2.0 */ -#ifndef _ASM_PARISC_COMPAT_UCONTEXT_H -#define _ASM_PARISC_COMPAT_UCONTEXT_H - -#include - -/* 32-bit ucontext as seen from an 64-bit kernel */ -struct compat_ucontext { - compat_uint_t uc_flags; - compat_uptr_t uc_link; - compat_stack_t uc_stack; /* struct compat_sigaltstack (12 bytes)*/ - /* FIXME: Pad out to get uc_mcontext to start at an 8-byte aligned boundary */ - compat_uint_t pad[1]; - struct compat_sigcontext uc_mcontext; - compat_sigset_t uc_sigmask; /* mask last for extensibility */ -}; - -#endif /* !_ASM_PARISC_COMPAT_UCONTEXT_H */ From 289e99e7a263c6fd6a5d07d6d8f2156b70f3a9d5 Mon Sep 17 00:00:00 2001 From: Zhenghui Hao Date: Fri, 18 Sep 2026 13:46:49 +0800 Subject: [PATCH 371/592] parisc: parse early parameters in setup_arch() parisc is one of the few architectures that does not call parse_early_param() from setup_arch(). That was mostly harmless until commit d49004c5f0c1 ("arch, mm: consolidate initialization of nodes, zones and memory map") moved the consumer of several hugetlb command line parameters into mm_core_init_early(), which runs before the generic parse_early_param() call in start_kernel(). As a result hugepages=, hugepagesz=, default_hugepagesz=, hugetlb_cma= and hugetlb_free_vmemmap= are recorded after they have already been consumed and are silently dropped on parisc. Call parse_early_param() from setup_arch(), after the command line has been set up and the memory inventory has been taken. jump_label_init() must be called first because early parameter handlers may enable or disable static keys. Both functions are safe to call more than once: the generic calls in start_kernel() remain in place and turn into no-ops. Suggested-by: Mike Rapoport (Microsoft) Fixes: d49004c5f0c1 ("arch, mm: consolidate initialization of nodes, zones and memory map") Cc: Signed-off-by: Zhenghui Hao Tested-by: Helge Deller Signed-off-by: Helge Deller --- arch/parisc/kernel/setup.c | 12 ++++++++++++ 1 file changed, 12 insertions(+) diff --git a/arch/parisc/kernel/setup.c b/arch/parisc/kernel/setup.c index d3e17a7a8901..4d3015a411d6 100644 --- a/arch/parisc/kernel/setup.c +++ b/arch/parisc/kernel/setup.c @@ -132,6 +132,18 @@ void __init setup_arch(char **cmdline_p) parisc_cache_init(); paging_init(); + /* + * Parse early parameters before mm_core_init_early() runs. + * Several early_param() handlers only record data that is consumed + * from there - for example hugepages=, hugepagesz=, + * default_hugepagesz=, hugetlb_cma= and hugetlb_free_vmemmap= - so + * the generic parse_early_param() call in start_kernel() is too late + * for them. jump_label_init() must come first, since early param + * handlers may enable or disable static keys. + */ + jump_label_init(); + parse_early_param(); + #ifdef CONFIG_PA11 dma_ops_init(); #endif From 4dd1999783d7d12434006289338373e49492dc96 Mon Sep 17 00:00:00 2001 From: Alexey Klimov Date: Thu, 17 Sep 2026 10:16:42 +0200 Subject: [PATCH 372/592] soc: samsung: exynos-pmu: fix use-after-free of interrupt generator node The setup_cpuhp_and_cpuidle() parses the device tree node for the interrupt generation block via of_parse_phandle() and decrements its reference count using of_node_put() immediately after fetching the resource address. However, later the intr_gen_node pointer is passed into of_syscon_register_regmap(). Fix this by declaring intr_gen_node with __free() and removing of_node_put(). Reported-by: Sashiko Closes: https://sashiko.dev/#/patchset/20260513-exynos850-cpuhotplug-v4-0-54fec5f65362@linaro.org?part=3 Fixes: 78b72897a5c8 ("soc: samsung: exynos-pmu: Enable CPU Idle for gs101") Cc: stable@vger.kernel.org Signed-off-by: Alexey Klimov Link: https://patch.msgid.link/20260828-exynos-pmu-cpuhp-idle-fixes-v2-1-06bce6107bd6@linaro.org Signed-off-by: Krzysztof Kozlowski Link: https://lore.kernel.org/r/20260917081641.72291-2-krzk@kernel.org Signed-off-by: Arnd Bergmann --- drivers/soc/samsung/exynos-pmu.c | 7 ++----- 1 file changed, 2 insertions(+), 5 deletions(-) diff --git a/drivers/soc/samsung/exynos-pmu.c b/drivers/soc/samsung/exynos-pmu.c index f5fcdde9750e..efccdd63e40e 100644 --- a/drivers/soc/samsung/exynos-pmu.c +++ b/drivers/soc/samsung/exynos-pmu.c @@ -409,13 +409,12 @@ static struct notifier_block exynos_cpupm_reboot_nb = { static int setup_cpuhp_and_cpuidle(struct device *dev) { - struct device_node *intr_gen_node; + struct device_node *intr_gen_node __free(device_node) = + of_parse_phandle(dev->of_node, "google,pmu-intr-gen-syscon", 0); struct resource intrgen_res; void __iomem *virt_addr; int ret, cpu; - intr_gen_node = of_parse_phandle(dev->of_node, - "google,pmu-intr-gen-syscon", 0); if (!intr_gen_node) { /* * To maintain support for older DTs that didn't specify syscon @@ -431,8 +430,6 @@ static int setup_cpuhp_and_cpuidle(struct device *dev) * syscon provided regmap. */ ret = of_address_to_resource(intr_gen_node, 0, &intrgen_res); - of_node_put(intr_gen_node); - virt_addr = devm_ioremap(dev, intrgen_res.start, resource_size(&intrgen_res)); if (!virt_addr) From 7cb575b71ab98194d2e040bded3a7281e089c5ed Mon Sep 17 00:00:00 2001 From: Li Jun Date: Thu, 17 Sep 2026 09:37:10 +0800 Subject: [PATCH 373/592] watchdog: da9063: fix suspend/resume handling of HW_RUNNING watchdog MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit da9063_wdt_suspend() and da9063_wdt_resume() only check watchdog_active(), when the watchdog is left running by the driver sets WDOG_HW_RUNNING in da9063_wdt_probe() but userspace never opens the device, so WDOG_ACTIVE remains cleared, the wdt_disable() will not be executed in da9063_wdt_suspend. In this case, the suspend callback is a no-op and the watchdog keeps counting during system suspend, leading to an unexpected system reset. Check WDOG_HW_RUNNING and wdd，can fix this issue. Fixes: a7ceca4398bc8 ("watchdog: da9063: optionally disable watchdog during suspend") Cc: stable@vger.kernel.org Signed-off-by: Li Jun Link: https://patch.msgid.link/20260917013710.2754679-1-lijun01@kylinos.cn Signed-off-by: Guenter Roeck --- drivers/watchdog/da9063_wdt.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/drivers/watchdog/da9063_wdt.c b/drivers/watchdog/da9063_wdt.c index 92e1b78ff481..3703110e82fc 100644 --- a/drivers/watchdog/da9063_wdt.c +++ b/drivers/watchdog/da9063_wdt.c @@ -271,7 +271,7 @@ static int da9063_wdt_suspend(struct device *dev) if (!da9063->use_sw_pm) return 0; - if (watchdog_active(wdd)) + if (watchdog_active(wdd) || watchdog_hw_running(wdd)) return da9063_wdt_stop(wdd); return 0; @@ -285,7 +285,7 @@ static int da9063_wdt_resume(struct device *dev) if (!da9063->use_sw_pm) return 0; - if (watchdog_active(wdd)) + if (watchdog_active(wdd) || watchdog_hw_running(wdd)) return da9063_wdt_start(wdd); return 0; From 8c7fdc0b4c64d6583fff3e8f7696237a16ff7c29 Mon Sep 17 00:00:00 2001 From: Joshua Hahn Date: Wed, 2 Sep 2026 12:45:20 -0700 Subject: [PATCH 374/592] selftests/cgroup: account for zswap shrinker writeback The test_no_invasive_cgroup_shrink selftest checks that when a cgroup has zswapped out more memory than memory.zswap.max, it does not trigger writeback for other cgroups. To do this, it compares the writeback count in a control cgroup and makes sure that it is 0, and then checks the writeback count in an aggressor cgroup who does expect to see writeback. However, when the zswap shrinker is enabled, the victim cgroup can see legitimate writebacks not triggered by the aggressor. In some Meta CI tests, we have seen this failure mode happen. Instead of checking that the victim cgroup has 0 writeback, compare the writeback values before and after the aggressor runs and check that the victim cgroup did not perform any additional writeback. Note that this can still lead to probabilistic failures if writebacks take longer than 5 seconds, but this should fix the systematic failure case and make "not ok test_no_invasive_cgroup_shrink" less likely. Link: https://lore.kernel.org/20260902194521.3652178-1-joshua.hahnjy@gmail.com Fixes: b5ba474f3f51 ("zswap: shrink zswap pool based on memory pressure") Signed-off-by: Joshua Hahn Signed-off-by: Andrew Morton Reported-by: Krush Chavan Suggested-by: Nhat Pham Cc: --- tools/testing/selftests/cgroup/test_zswap.c | 28 +++++++++++++++++++-- 1 file changed, 26 insertions(+), 2 deletions(-) diff --git a/tools/testing/selftests/cgroup/test_zswap.c b/tools/testing/selftests/cgroup/test_zswap.c index 609c48f38524..8df54b59513a 100644 --- a/tools/testing/selftests/cgroup/test_zswap.c +++ b/tools/testing/selftests/cgroup/test_zswap.c @@ -20,6 +20,7 @@ static int page_size; #define PATH_ZSWAP "/sys/module/zswap" #define PATH_ZSWAP_ENABLED "/sys/module/zswap/parameters/enabled" +#define PATH_ZSWAP_SHRINKER_ENABLED "/sys/module/zswap/parameters/shrinker_enabled" #define PATH_ZSWAP_STORED_PAGES "/sys/kernel/debug/zswap/stored_pages" static int read_int(const char *path, size_t *value) @@ -444,6 +445,16 @@ static int test_zswap_writeback_disabled(const char *root) return test_zswap_writeback(root, false); } +static bool zswap_shrinker_enabled(void) +{ + char value[2]; + + if (read_text(PATH_ZSWAP_SHRINKER_ENABLED, value, sizeof(value)) <= 0) + return 0; + + return value[0] == 'Y'; +} + /* * When trying to store a memcg page in zswap, if the memcg hits its memory * limit in zswap, writeback should affect only the zswapped pages of that @@ -453,6 +464,7 @@ static int test_no_invasive_cgroup_shrink(const char *root) { int ret = KSFT_FAIL; unsigned int off; + long zswpwb_before, zswpwb_after, zswpwb_target; size_t allocation_size = page_size * 1024; unsigned int nr_pages = allocation_size / page_size; char zswap_max_buf[32], mem_max_buf[32]; @@ -488,6 +500,14 @@ static int test_no_invasive_cgroup_shrink(const char *root) if (cg_read_key_long(zw_group, "memory.stat", "zswapped") < 1) goto out; + /* If the shrinker is enabled, try to let the writebacks finish first */ + if (zswap_shrinker_enabled()) + sleep(5); + + zswpwb_before = get_cg_wb_count(zw_group); + if (zswpwb_before < 0) + goto out; + /* Push wb_group memory into zswap with hard-to-compress data to trigger wb */ if (cg_enter_current(wb_group)) goto out; @@ -500,9 +520,13 @@ static int test_no_invasive_cgroup_shrink(const char *root) getrandom(&wb_allocation[off], page_size/4, 0); } - /* Verify that only zswapped memory from gwb_group has been written back */ - if (wait_for_writeback(wb_group, 5000) > 0 && get_cg_wb_count(zw_group) == 0) + /* Verify that only zswapped memory from wb_group has been written back */ + zswpwb_target = wait_for_writeback(wb_group, 5000); + zswpwb_after = get_cg_wb_count(zw_group); + + if (zswpwb_target > 0 && zswpwb_before == zswpwb_after) ret = KSFT_PASS; + out: cg_enter_current(root); if (zw_group) { From 8d50c2f37bcc766cd5ecde9bbb14c9c27c609f33 Mon Sep 17 00:00:00 2001 From: Haowen Bai Date: Thu, 3 Sep 2026 21:28:54 +0800 Subject: [PATCH 375/592] mailmap: update Haowen Bai's email address Map Haowen Bai's former Meizu and current Ugreen addresses to baihaowen88@gmail.com as the canonical public address. Link: https://lore.kernel.org/20260903132854.1930923-1-calvin.bai@ugreen.com Signed-off-by: Haowen Bai Signed-off-by: Andrew Morton --- .mailmap | 2 ++ 1 file changed, 2 insertions(+) diff --git a/.mailmap b/.mailmap index 90ff4831f592..59518b814b4f 100644 --- a/.mailmap +++ b/.mailmap @@ -355,6 +355,8 @@ Hans Verkuil Hans Verkuil Hans Verkuil Hao Ge +Haowen Bai +Haowen Bai Harry Yoo <42.hyeyoo@gmail.com> Harry Yoo Heiko Carstens From 525c0edc032b3297d0c1056cf1fa20cf1f9e6184 Mon Sep 17 00:00:00 2001 From: Joseph Qi Date: Fri, 4 Sep 2026 10:37:51 +0800 Subject: [PATCH 376/592] ocfs2: make ocfs2_calc_xattr_init() return void ocfs2_calc_xattr_init() used to read the default ACL off the parent inode itself, so it could return an error from ocfs2_xattr_get_nolock(). Commit bd7c05fb4a47 ("ocfs2: fix circular locking dependency in ocfs2_init_acl()") moved that lookup before the transaction starts and deleted the error path, but left the now vestigial 'int ret = 0' declaration and both 'return ret' statements behind, along with an unreachable error branch in ocfs2_mknod(). Drop the leftover variable and convert the return type to void, so the callee states that it always succeeds and the caller no longer carries a check that can never trigger. No functional change. Link: https://lore.kernel.org/20260904023751.3703334-1-joseph.qi@linux.alibaba.com Fixes: bd7c05fb4a47 ("ocfs2: fix circular locking dependency in ocfs2_init_acl()") Signed-off-by: Joseph Qi Signed-off-by: Andrew Morton Reported-by: kernel test robot Closes: https://lore.kernel.org/oe-kbuild-all/202609040247.8B3lmoqX-lkp@intel.com/ Cc: Mark Fasheh Cc: Joel Becker Cc: Junxiao Bi Cc: Changwei Ge Cc: Jun Piao Cc: Heming Zhao --- fs/ocfs2/namei.c | 9 ++------- fs/ocfs2/xattr.c | 13 +++++-------- fs/ocfs2/xattr.h | 8 ++++---- 3 files changed, 11 insertions(+), 19 deletions(-) diff --git a/fs/ocfs2/namei.c b/fs/ocfs2/namei.c index e9c7774ccf91..58c6061ed983 100644 --- a/fs/ocfs2/namei.c +++ b/fs/ocfs2/namei.c @@ -336,13 +336,8 @@ static int ocfs2_mknod(struct mnt_idmap *idmap, goto leave; /* calculate meta data/clusters for setting security and acl xattr */ - status = ocfs2_calc_xattr_init(dir, mode, &si, &want_clusters, - &xattr_credits, &want_meta, - &acl_state); - if (status < 0) { - mlog_errno(status); - goto leave; - } + ocfs2_calc_xattr_init(dir, mode, &si, &want_clusters, &xattr_credits, + &want_meta, &acl_state); /* Reserve a cluster if creating an extent based directory. */ if (S_ISDIR(mode) && !ocfs2_supports_inline_data(osb)) { diff --git a/fs/ocfs2/xattr.c b/fs/ocfs2/xattr.c index 35bcbb0ff607..bfafe059bedf 100644 --- a/fs/ocfs2/xattr.c +++ b/fs/ocfs2/xattr.c @@ -635,12 +635,11 @@ int ocfs2_calc_security_init(struct inode *dir, return ret; } -int ocfs2_calc_xattr_init(struct inode *dir, umode_t mode, - struct ocfs2_security_xattr_info *si, - int *want_clusters, int *xattr_credits, - int *want_meta, struct ocfs2_acl_state *acl_state) +void ocfs2_calc_xattr_init(struct inode *dir, umode_t mode, + struct ocfs2_security_xattr_info *si, + int *want_clusters, int *xattr_credits, + int *want_meta, struct ocfs2_acl_state *acl_state) { - int ret = 0; struct ocfs2_super *osb = OCFS2_SB(dir->i_sb); int s_size = 0, a_size = 0, acl_len = 0, new_clusters; @@ -662,7 +661,7 @@ int ocfs2_calc_xattr_init(struct inode *dir, umode_t mode, } if (!(s_size + a_size)) - return ret; + return; /* * The max space of security xattr taken inline is @@ -728,8 +727,6 @@ int ocfs2_calc_xattr_init(struct inode *dir, umode_t mode, } } } - - return ret; } static int ocfs2_xattr_extend_allocation(struct inode *inode, diff --git a/fs/ocfs2/xattr.h b/fs/ocfs2/xattr.h index 5e18513277f1..887cc1a18b1a 100644 --- a/fs/ocfs2/xattr.h +++ b/fs/ocfs2/xattr.h @@ -59,10 +59,10 @@ int ocfs2_calc_security_init(struct inode *, int *, int *, struct ocfs2_alloc_context **); struct ocfs2_acl_state; -int ocfs2_calc_xattr_init(struct inode *dir, umode_t mode, - struct ocfs2_security_xattr_info *si, - int *want_clusters, int *xattr_credits, - int *want_meta, struct ocfs2_acl_state *acl_state); +void ocfs2_calc_xattr_init(struct inode *dir, umode_t mode, + struct ocfs2_security_xattr_info *si, + int *want_clusters, int *xattr_credits, + int *want_meta, struct ocfs2_acl_state *acl_state); /* * xattrs can live inside an inode, as part of an external xattr block, From f166586f74dd5d9cbadaabf86ef81c8ddf6cafa7 Mon Sep 17 00:00:00 2001 From: Nathan Gao Date: Thu, 3 Sep 2026 17:28:27 -0700 Subject: [PATCH 377/592] mm/damon/ops-common: use a page-aligned address in damon_ptep_mkold() __damon_va_prepare_access_check() picks a random byte address within the region and stores it in r->sampling_addr. damon_va_mkold() passes it into a page table walk, which hands it to damon_ptep_mkold() as the address of the page to sample: damon_va_mkold(mm, r->sampling_addr) damon_va_walk_page_range(mm, addr, addr + 1) damon_mkold_pmd_entry() damon_ptep_mkold(pte, vma, addr) ptep_test_and_clear_young(vma, addr, pte) mmu_notifier_clear_young(mm, addr, addr + PAGE_SIZE) For arm64, before commit 6f0e1142173a ("arm64: mm: support batch clearing of the young flag for large folios"), the contpte helper walked exactly CONT_PTES entries from the aligned-down page table pointer and used @addr only to pass down to each entry, so an unaligned value was harmless: ptep = contpte_align_down(ptep); addr = ALIGN_DOWN(addr, CONT_PTE_SIZE); for (i = 0; i < CONT_PTES; i++, ptep++, addr += PAGE_SIZE) Now the range to walk is derived from @addr instead: end = addr + nr * PAGE_SIZE, rounded up to CONT_PTE_SIZE. For a sample in the last page of a contpte block, the sub-page offset puts end just past the block boundary, so the round-up lands a whole block further and the walk clears PTE_AF in CONT_PTES entries beyond the sampled block. For the last block in a page table page, those entries are past the end of that page, so the walk writes into the page that follows. Triggered by the full 7.1/7.2 kernel selftest suite on arm64 (EC2 c/m6g.4xlarge). The kernel sometimes crashes at or shortly after the DAMON test. What the overrun does depends on the page that happens to follow the page table, so there is no single signature. If that page is read-only, the write faults in the sampling path itself: Unable to handle kernel write to read-only memory at virtual address ffff0003c5d2d000 FSC = 0x0f: level 3 permission fault CM = 0, WnR = 1, TnD = 0, TagAccess = 0 CPU: 10 UID: 0 PID: 3487 Comm: kdamond.2 pc : contpte_test_and_clear_young_ptes+0x70/0xc0 lr : damon_ptep_mkold+0x1e8/0x1f8 Call trace: contpte_test_and_clear_young_ptes+0x70/0xc0 (P) damon_mkold_pmd_entry+0x150/0x170 walk_pmd_range+0x110/0x2b0 walk_pud_range+0x10c/0x208 walk_pgd_range+0x134/0x258 __walk_page_range+0x98/0x1b0 walk_page_range_vma_unsafe+0x90/0x148 walk_page_range_vma+0x28/0x40 damon_va_walk_page_range+0x114/0x2b8 damon_va_prepare_access_checks+0xec/0x1a8 kdamond_fn+0x534/0x770 kthread+0x128/0x138 ret_from_fork+0x10/0x20 Otherwise the page is writable, the PTE_AF clearing succeeds silently and the damage only surfaces later, in whatever happened to own the page, so the backtrace is unrelated to DAMON and differs between runs. Pass a page-aligned address to the ptep_test_and_clear_young() call in damon_ptep_mkold(), which is the only place DAMON can reach contpte_test_and_clear_young_ptes() from. Nothing else sees the aligned address, and r->sampling_addr itself is left as is, so the sampling and region bookkeeping semantics are unchanged. Link: https://lore.kernel.org/20260904002829.116381-1-sj@kernel.org Fixes: 6f0e1142173a ("arm64: mm: support batch clearing of the young flag for large folios") Signed-off-by: Nathan Gao Signed-off-by: SJ Park Signed-off-by: Andrew Morton Reviewed-by: SJ Park Reviewed-by: Baolin Wang Cc: Baolin Wang Cc: David Hildenbrand (Arm) Cc: Ryan Roberts Cc: --- mm/damon/ops-common.c | 7 ++++++- 1 file changed, 6 insertions(+), 1 deletion(-) diff --git a/mm/damon/ops-common.c b/mm/damon/ops-common.c index fbda70d8ea4d..8fc61d06d358 100644 --- a/mm/damon/ops-common.c +++ b/mm/damon/ops-common.c @@ -61,7 +61,12 @@ void damon_ptep_mkold(pte_t *pte, struct vm_area_struct *vma, unsigned long addr * device aspects. */ if (likely(pte_present(pteval))) - young |= ptep_test_and_clear_young(vma, addr, pte); + /* + * Arch implementation of ptep_test_and_clear_young() may + * require aligned @addr + */ + young |= ptep_test_and_clear_young(vma, PAGE_ALIGN_DOWN(addr), + pte); young |= mmu_notifier_clear_young(vma->vm_mm, addr, addr + PAGE_SIZE); if (young) folio_set_young(folio); From 90179da203ba8b708c84a12a07cd44be0f346334 Mon Sep 17 00:00:00 2001 From: Liew Rui Yan Date: Tue, 8 Sep 2026 06:54:11 -0700 Subject: [PATCH 378/592] mm/damon/core: allow esz to be set to zero When the temporal quota goal tuner determines that the goal has been achieved (score >= 10000), it sets esz_bp to zero so that the esz becomes zero. However, damos_set_effective_quota() clamps the esz to min_region_sz when quota->ms is set. This is a minor issue, the main problem is that it doesn't match the description in the documentation, which state that if the goal has already been [over-]achieved, the quota will be set to zero. Fix this by set quota (esz) as minimum as possible. Link: https://lore.kernel.org/20260908135413.97570-1-sj@kernel.org Fixes: 8bbde987c2b8 ("mm/damon/core: disallow time-quota setting zero esz") Signed-off-by: SJ Park Signed-off-by: Liew Rui Yan Signed-off-by: Andrew Morton Reviewed-by: SJ Park Cc: # v7.1.x --- mm/damon/core.c | 5 +++-- 1 file changed, 3 insertions(+), 2 deletions(-) diff --git a/mm/damon/core.c b/mm/damon/core.c index 644daf5a1656..2b294fb46648 100644 --- a/mm/damon/core.c +++ b/mm/damon/core.c @@ -3091,6 +3091,7 @@ static void damos_set_effective_quota(struct damon_ctx *ctx, struct damos *s) struct damos_quota *quota = &s->quota; unsigned long throughput; unsigned long esz = ULONG_MAX; + unsigned long esz_time; if (!quota->ms && list_empty("a->goals)) { quota->esz = quota->sz; @@ -3111,8 +3112,8 @@ static void damos_set_effective_quota(struct damon_ctx *ctx, struct damos *s) 1000000, quota->total_charged_ns); else throughput = PAGE_SIZE * 1024; - esz = min(throughput * quota->ms, esz); - esz = max(ctx->min_region_sz, esz); + esz_time = max(throughput * quota->ms, ctx->min_region_sz); + esz = min(esz_time, esz); } if (quota->sz && quota->sz < esz) From 39c0ceedd54557bdc1542de08d22b2ed33e534e4 Mon Sep 17 00:00:00 2001 From: SJ Park Date: Mon, 7 Sep 2026 10:03:56 -0700 Subject: [PATCH 379/592] mm/damon/vaddr: avoid hw-driven pte updates during damon_hugetlb_mkold() damon_hugetlb_mkold() reads the page table entry into a local variable, unsets the accessed bit in the variable, and updates the page table entry with the updated variable value. If hardware updates the same page table entry in parallel, the hw updates could be lost. For example, hardware-updated dirty bits might be lost. Avoid the parallel updates by clearing the page table entry when reading it together, using huge_ptep_get_and_clear(). If a parallel write to the memory is made after the clearing, the hw will see the page table entry is cleared, trigger page fault and wait until it is handled. The page fault handling will wait for damon_hugetlb_mkold() due to the page table lock. Because hugetlbfs is an in-memory file system and hugetlb pages cannot be reclaimed, no critical issue is expected to my best knowledge. But definitely this is a nasty bug that should be fixed sooner rather than later. The issue was discovered [1] by Sashiko. Link: https://lore.kernel.org/20260907170358.100168-1-sj@kernel.org Link: https://lore.kernel.org/20260830160545.98969-1-sj@kernel.org [1] Fixes: 49f4203aae06 ("mm/damon: add access checking for hugetlb pages") Signed-off-by: SJ Park Signed-off-by: Andrew Morton Cc: Baolin Wang Cc: # 5.17.x --- mm/damon/vaddr.c | 21 ++++++++++++++------- 1 file changed, 14 insertions(+), 7 deletions(-) diff --git a/mm/damon/vaddr.c b/mm/damon/vaddr.c index 0648400b2d65..04ee2a2c6a4d 100644 --- a/mm/damon/vaddr.c +++ b/mm/damon/vaddr.c @@ -293,22 +293,29 @@ static int damon_mkold_pmd_entry(pmd_t *pmd, unsigned long addr, } #ifdef CONFIG_HUGETLB_PAGE +static bool damon_hugetlb_ptep_mkold(pte_t *pte, struct mm_struct *mm, + struct vm_area_struct *vma, unsigned long addr, pte_t *entry) +{ + unsigned long psize = huge_page_size(hstate_vma(vma)); + + if (!pte_young(*entry)) + return false; + *entry = huge_ptep_get_and_clear(mm, addr, pte, psize); + *entry = pte_mkold(*entry); + set_huge_pte_at(mm, addr, pte, *entry, psize); + return true; +} + static void damon_hugetlb_mkold(pte_t *pte, struct mm_struct *mm, struct vm_area_struct *vma, unsigned long addr) { bool referenced = false; pte_t entry = huge_ptep_get(mm, addr, pte); struct folio *folio = pfn_folio(pte_pfn(entry)); - unsigned long psize = huge_page_size(hstate_vma(vma)); folio_get(folio); - if (pte_young(entry)) { - referenced = true; - entry = pte_mkold(entry); - set_huge_pte_at(mm, addr, pte, entry, psize); - } - + referenced = damon_hugetlb_ptep_mkold(pte, mm, vma, addr, &entry); if (mmu_notifier_clear_young(mm, addr, addr + huge_page_size(hstate_vma(vma)))) referenced = true; From b3723b596b548c837a766aae3553c14a7b15af2b Mon Sep 17 00:00:00 2001 From: Liew Rui Yan Date: Tue, 8 Sep 2026 06:47:38 -0700 Subject: [PATCH 380/592] mm/damon/core: fix unconditionally skip last region Once quota set, the charge_{target,addr}_from unconditionally skips and resets at the last region of the tracked target, so the last region can be skipped even when it has not been processed. Example: 1. Target has 2 regions: R1 (0-100 bytes) and R2 (100-200 bytes). 2. Quota is configured to process only 100 bytes per window. 3. Window 1: Processes R1 (0-100). Quota is full. charge_{target, addr}_from is saved at (Target, 100). 4. Window 2: The loop reaches R2. Because R2 is damon_last_region(t), the old code unconditionally returns true, skipping R2 entirely and resetting the charge_{target,addr}_from. Result: R2 is permanently skipped even though it has never been processed. However, it is important to note that this is a very minor issue. This is because it is triggered only when the previous window saved/kept charge_{target,addr}_from, and in the next window, all regions except the last region were skipped by damos_skip_charged_region(). Fix this by only resetting the charge_{target,addr}_from when last region is reached, only skipping when it is applied or cannot split. Link: https://lore.kernel.org/20260908134739.96919-1-sj@kernel.org Fixes: 50585192bc2e ("mm/damon/schemes: skip already charged targets and regions") Signed-off-by: Liew Rui Yan Reviewed-by: SJ Park Signed-off-by: SJ Park Signed-off-by: Andrew Morton Cc: # v5.16.x --- mm/damon/core.c | 25 ++++++++++++++----------- 1 file changed, 14 insertions(+), 11 deletions(-) diff --git a/mm/damon/core.c b/mm/damon/core.c index 2b294fb46648..06a253df8d4b 100644 --- a/mm/damon/core.c +++ b/mm/damon/core.c @@ -2342,36 +2342,39 @@ static bool damos_skip_charged_region(struct damon_target *t, { struct damos_quota *quota = &s->quota; unsigned long sz_to_skip; + bool skip = false; /* Skip previously charged regions */ if (quota->charge_target_from) { if (t != quota->charge_target_from) return true; - if (r == damon_last_region(t)) { - quota->charge_target_from = NULL; - quota->charge_addr_from = 0; - return true; - } if (quota->charge_addr_from && - r->ar.end <= quota->charge_addr_from) - return true; + r->ar.end <= quota->charge_addr_from) { + skip = true; + goto out; + } if (quota->charge_addr_from && r->ar.start < quota->charge_addr_from) { sz_to_skip = ALIGN_DOWN(quota->charge_addr_from - r->ar.start, min_region_sz); if (!sz_to_skip) { - if (damon_sz_region(r) <= min_region_sz) - return true; + if (damon_sz_region(r) <= min_region_sz) { + skip = true; + goto out; + } sz_to_skip = min_region_sz; } damon_split_region_at(t, r, sz_to_skip); - return true; + skip = true; } + } +out: + if (r == damon_last_region(t)) { quota->charge_target_from = NULL; quota->charge_addr_from = 0; } - return false; + return skip; } static void damos_update_stat(struct damos *s, From 9bdad082d44bdcf93716973dcba6be77e8a06e7b Mon Sep 17 00:00:00 2001 From: Jaewook You Date: Mon, 14 Sep 2026 22:23:52 +0900 Subject: [PATCH 381/592] mm/hugetlb: preserve mremap address delta when skipping page tables move_hugetlb_page_tables() optimizes mremap() by advancing to the last entry in the page table when the source page table does not exist, either initially or after unsharing a PMD table. The common loop increment then steps to the first entry in the next page table. However, the code advances both the source and destination addresses to the last entries in their respective page tables, which is wrong. The destination address must be advanced only by the same amount as the source address. If the source and destination offsets within their page tables differ, the destination address can be advanced too far, causing follow-up issues. Fix this by advancing the destination address by the source advance distance. With a reproducer, we were able to trigger a kernel panic on x86-64. With this fix in place, we can no longer reproduce the issue. Link: https://lore.kernel.org/20260914132352.472-1-jaewook376@gmail.com Fixes: e95a9851787b ("hugetlb: skip to end of PT page mapping when pte not present") Fixes: 4ddb4d91b82f ("hugetlb: do not update address in huge_pmd_unshare") Signed-off-by: Jaewook You Signed-off-by: Andrew Morton Acked-by: David Hildenbrand (Arm) Cc: Johan Hovold Cc: Muchun Song Cc: Oscar Salvador Cc: Assisted-by: LLM --- mm/hugetlb.c | 11 +++++++---- 1 file changed, 7 insertions(+), 4 deletions(-) diff --git a/mm/hugetlb.c b/mm/hugetlb.c index d28972cd33f5..cea25773a6c9 100644 --- a/mm/hugetlb.c +++ b/mm/hugetlb.c @@ -5170,18 +5170,21 @@ int move_hugetlb_page_tables(struct vm_area_struct *vma, hugetlb_vma_lock_write(vma); i_mmap_lock_write(mapping); for (; old_addr < old_end; old_addr += sz, new_addr += sz) { + const unsigned long offset_to_last_entry = + (old_addr | last_addr_mask) - old_addr; + src_pte = hugetlb_walk(vma, old_addr, sz); if (!src_pte) { - old_addr |= last_addr_mask; - new_addr |= last_addr_mask; + old_addr += offset_to_last_entry; + new_addr += offset_to_last_entry; continue; } if (huge_pte_none(huge_ptep_get(mm, old_addr, src_pte))) continue; if (huge_pmd_unshare(&tlb, vma, old_addr, src_pte)) { - old_addr |= last_addr_mask; - new_addr |= last_addr_mask; + old_addr += offset_to_last_entry; + new_addr += offset_to_last_entry; continue; } From b6ac0b3f6013c168f22cad97e79967accacb08e1 Mon Sep 17 00:00:00 2001 From: Jinjiang Tu Date: Tue, 8 Sep 2026 20:29:24 +0800 Subject: [PATCH 382/592] mm/rmap: fix missing barrier between anon_vma init and vma->anon_vma publish On arm64 server, we find that a task trying to grab the anon_vma lock triggers hungtask. INFO: task main:2354726 blocked for more than 120 seconds. Tainted: G E 5.10.0-0021.aarch64 #1 "echo 0 > /proc/sys/kernel/hung_task_timeout_secs" disables this message. task:main state:D stack: 0 pid:2354726 ppid:2350673 flags:0x00000a01 Call trace: __switch_to+0x7c/0xbc __schedule+0x3b4/0x8a0 schedule+0x50/0xe0 rwsem_down_write_slowpath+0x3cc/0x6cc down_write+0x60/0x260 __anon_vma_prepare+0x6c/0x210 do_anonymous_page+0x258/0x660 handle_pte_fault+0x188/0x214 __handle_mm_fault+0x1b0/0x380 handle_mm_fault+0xf4/0x284 do_page_fault+0x19c/0x494 do_translation_fault+0xcc/0xf8 do_mem_abort+0x48/0xac el0_da+0x44/0x80 el0_sync_handler+0x88/0xb4 el0_sync+0x160/0x180 After analyzing the vmcore, we found the anon_vma->root->rwsem.count is -1. There is another anon_vma whose anon_vma->root->rwsem.count is 1, the anon_vma->root->rwsem.owner shows the lock is held, but the stack of the task shows the task doesn't hold the anon_vma lock. After adding more debugging info, we found __anon_vma_prepare() reuses anon_vma and triggers the UAF of anon_vma->root due to missing memory barrier, leading to locking and unlocking two different anon_vma->root, thus leading to an anon_vma will never be unlocked, and another anon_vma couldn't be locked anymore. This race requires two adjacent VMAs that are not merged but are anon_vma-compatible (e.g., they differ in VMA_ACCESS_FLAGS that can be changed by mprotect()). Two threads fault on each VMA concurrently, both calling __anon_vma_prepare() with only mmap_lock held for reading. THREAD A THREAD B __anon_vma_prepare __anon_vma_prepare find_mergeable_anon_vma() -> NULL anon_vma = anon_vma_alloc(); anon_vma->root = anon_vma; // the two stores may be reordered vma->anon_vma = anon_vma; // finds A's anon_vma anon_vma = find_mergeable_anon_vma(vma); anon_vma_lock_write(anon_vma); // may still see the old root down_write(&anon_vma->root->rwsem); anon_vma_unlock_write(anon_vma); // see the new root, never unlock old up_write(&anon_vma->root->rwsem); thread A triggers page fault and calls __anon_vma_prepare() to prepare anon_vma for the faulting vma. __anon_vma_prepare() allocates and initializes a new anon_vma, and then publishes it to the vma with a plain store. anon_vma_prepare() only requires the mmap_lock to be held for reading, so two threads can fault on adjacent VMAs at the same time. While thread A publishes a new anon_vma, thread B could find the anon_vma via find_mergeable_anon_vma() and then locks anon_vma->root->rwsem. The store to anon_vma->root in anon_vma_alloc() and the store to vma->anon_vma can be reordered. The anon_vma_lock_write() and spin_lock() only provide acquire semantics, which do not prevent prior stores from being reordered after them. The release semantics of the corresponding spin_unlock() and anon_vma_unlock_write() come too late, the store to vma->anon_vma is already published before they take effect. As a result, thread B can observe the following order: vma->anon_vma = anon_vma; anon_vma->root = anon_vma; The anon_vma slab is SLAB_TYPESAFE_BY_RCU, so a newly allocated anon_vma may reuse memory from a previously freed one. The constructor (anon_vma_ctor) does not reset anon_vma->root, and __put_anon_vma() doesn't clear it either, so the old root value persists until anon_vma_alloc() overwrites it. If that store isn't visible, thread B reads a root that points to the old anon_vma and locks it. As a result, thread B can call anon_vma_lock_write() with the old root, and call anon_vma_unlock_write() with the new root, leading to an anon_vma will never be unlocked, and another anon_vma couldn't be locked anymore (its count is dropped from 0 to -1 due to wrong unlock). To fix it, change the plain store `vma->anon_vma = anon_vma` to store release, so that the fields of anon_vma are visible before anon_vma is published to vma->anon_vma. At read side, the load of anon_vma and anon_vma->root have address dependency. According to Documentation/memory-barriers.txt and some investigations, only Alpha needs address-dependency barriers and it has been handled by READ_ONCE() in reusable_anon_vma(). We reproduced this issue in v5.10 with KSM enabled. The kernel doesn't merge commit cf7e7a3503df ("mm: prevent KSM from breaking VMA merging for new VMAs"), so there are many adjacent VMAs that aren't merged but are compatible for anon_vma. Without this fix, our production environment could reproduce this issue about 2-5 times each month. After adding a smp_mb() before anon_vma_lock_write(anon_vma) in __anon_vma_prepare(), which is different to this patch, this issue hasn't been reproduced for one month. Link: https://lore.kernel.org/20260908122924.554373-1-tujinjiang@huawei.com Fixes: 5c341ee1dfc8 ("mm: track the root (oldest) anon_vma") Signed-off-by: Jinjiang Tu Signed-off-by: Andrew Morton Reviewed-by: Lance Yang Reviewed-by: Lorenzo Stoakes (ARM) Acked-by: David Hildenbrand (Arm) Acked-by: Vlastimil Babka (SUSE) Cc: Minchan Kim Cc: Harry Yoo Cc: Hiroyouki Kamezawa Cc: Jann Horn Cc: Jinjiang Tu Cc: Kefeng Wang Cc: Larry Woodman Cc: Liam R. Howlett Cc: Nanyong Sun Cc: Rik van Riel Cc: --- mm/rmap.c | 6 +++++- mm/vma.c | 8 ++++++++ 2 files changed, 13 insertions(+), 1 deletion(-) diff --git a/mm/rmap.c b/mm/rmap.c index d1819fd69938..f3b21aaa34ee 100644 --- a/mm/rmap.c +++ b/mm/rmap.c @@ -209,7 +209,11 @@ int __anon_vma_prepare(struct vm_area_struct *vma) /* page_table_lock to protect against threads */ spin_lock(&mm->page_table_lock); if (likely(!vma->anon_vma)) { - vma->anon_vma = anon_vma; + /* + * Make anon_vma fields visible before anon_vma is published. + * Paired with an address dependency in reusable_anon_vma(). + */ + smp_store_release(&vma->anon_vma, anon_vma); anon_vma_chain_assign(vma, avc, anon_vma); anon_rmap_tree_insert(avc, anon_vma); anon_vma->num_active_vmas++; diff --git a/mm/vma.c b/mm/vma.c index f29abb30956b..9f0a0acf694a 100644 --- a/mm/vma.c +++ b/mm/vma.c @@ -2094,6 +2094,13 @@ static int anon_vma_compatible(struct vm_area_struct *a, struct vm_area_struct * * acceptable for merging, so we can do all of this optimistically. But * we do that READ_ONCE() to make sure that we never re-load the pointer. * + * The READ_ONCE() establishes an address dependency between anon_vma and + * any access to its fields, which pairs with the assignment to + * vma->anon_vma performed with release semantics in __anon_vma_prepare(). + * + * This is especially important as anon_vma's are SLAB_TYPESAFE_BY_RCU so + * accessing an uninitialised anon_vma's fields may result in a UAF. + * * IOW: that the "list_is_singular()" test on the anon_vma_chain only * matters for the 'stable anon_vma' case (ie the thing we want to avoid * is to return an anon_vma that is "complex" due to having gone through @@ -2108,6 +2115,7 @@ static struct anon_vma *reusable_anon_vma(struct vm_area_struct *old, struct vm_area_struct *b) { if (anon_vma_compatible(a, b)) { + /* Paired with a memory barrier in __anon_vma_prepare(). */ struct anon_vma *anon_vma = READ_ONCE(old->anon_vma); if (anon_vma && list_is_singular(&old->anon_vma_chain)) From 44fcc0bfb0874a95cec2c1672f14d426f86dc68f Mon Sep 17 00:00:00 2001 From: Baolin Wang Date: Tue, 8 Sep 2026 09:42:12 +0800 Subject: [PATCH 383/592] MAINTAINERS: add Baoquan and Baolin as MGLRU reviewers Baoquan and I have been contributing MGLRU patches and helping review MGLRU related patches for some time. We will continue to follow MGLRU changes, so we'd like to be CC'd on MGLRU related patches. Link: https://lore.kernel.org/06e20ef4f603a4ffeafcdbf623ce2936281457bf.1788831480.git.baolin.wang@linux.alibaba.com Signed-off-by: Baolin Wang Signed-off-by: Andrew Morton Acked-by: Barry Song Acked-by: Qi Zheng Acked-by: Baoquan He Acked-by: Kairui Song Cc: Axel Rasmussen Cc: Shakeel Butt Cc: Wei Xu Cc: Yuanchu Xie --- MAINTAINERS | 2 ++ 1 file changed, 2 insertions(+) diff --git a/MAINTAINERS b/MAINTAINERS index c9f866debdf0..a6c0e4bcbc78 100644 --- a/MAINTAINERS +++ b/MAINTAINERS @@ -17240,6 +17240,8 @@ R: Barry Song R: Axel Rasmussen R: Yuanchu Xie R: Wei Xu +R: Baoquan He +R: Baolin Wang L: linux-mm@kvack.org S: Maintained W: http://www.linux-mm.org From eb64948249781bda35de04feab5a0acc36aa9051 Mon Sep 17 00:00:00 2001 From: SJ Park Date: Thu, 10 Sep 2026 07:28:45 -0700 Subject: [PATCH 384/592] mm/damon/core: reset invalid quota->charge_target_from DAMOS can suddenly stop working if a target process that the quota is just fully charged on is terminated. Fix by catching and processing the corner case. When DAMOS quota is fully charged, the target and the region to continue applying the action in the next round is saved in damos_quota->charge_{target,addr}_from. In the next round, DAMOS iterates targets and regions from the beginning. It skips applying the action to the regions until it visits and skips the saved target/region. Virtual address space targets become invalid if the process is terminated. Trying to apply the scheme to invalid target is just a waste of time. Hence commit 6e4930e33329 ("mm/damon/core: fix wasteful CPU calls by skipping non-existent targets") made the logic to skip invalid targets. However, it does skip before the charged target/region skipping/updating. Let's suppose the user runs DAMOS for multiple virtual address spaces with a quota. The quota exceeded in the middle of a virtual address space. And the process of the address space is terminated. Then the charge_target_from points to the invalid target. The pointer update logic is skipped for the invalid target, so the charge_target_from is never updated. DAMOS action to every target/region is skipped. From the user's perspective, it would look like suddenly DAMOS has stopped working. No critical leak or crash can happen. The user could reinstall the scheme. But this makes use of DAMOS under certain setups quite unreliable. When the invalid target is found, further check the corner case and reset the pointer. This issue was discovered [1] by Sashiko. Link: https://lore.kernel.org/20260910142846.172957-1-sj@kernel.org Link: https://lore.kernel.org/20260830064708.40CA61F000E9@smtp.kernel.org [1] Fixes: 6e4930e33329 ("mm/damon/core: fix wasteful CPU calls by skipping non-existent targets") Signed-off-by: SJ Park Signed-off-by: Andrew Morton Cc: Enze Li Cc: # 7.0.x --- mm/damon/core.c | 9 ++++++++- 1 file changed, 8 insertions(+), 1 deletion(-) diff --git a/mm/damon/core.c b/mm/damon/core.c index 06a253df8d4b..1764620903f2 100644 --- a/mm/damon/core.c +++ b/mm/damon/core.c @@ -3245,8 +3245,15 @@ static void kdamond_apply_schemes(struct damon_ctx *c) max_region_sz = damon_region_sz_limit(c); mutex_lock(&c->walk_control_lock); damon_for_each_target(t, c) { - if (c->ops.target_valid && c->ops.target_valid(t) == false) + if (c->ops.target_valid && c->ops.target_valid(t) == false) { + damon_for_each_scheme(s, c) { + if (s->quota.charge_target_from != t) + continue; + s->quota.charge_target_from = NULL; + s->quota.charge_addr_from = 0; + } continue; + } damos_apply_target(c, t, max_region_sz); } From 407a5d205179a4ab186571b0e16ec42725dc77bc Mon Sep 17 00:00:00 2001 From: Josef Bacik Date: Wed, 9 Sep 2026 18:01:07 +0000 Subject: [PATCH 385/592] writeback: report a Tasks-RCU quiescent state per cgwb drain pass cleanup_offline_cgwbs_workfn() drains a dying cgwb by calling cleanup_offline_cgwb() until it returns false, with a cond_resched() between passes. On a CONFIG_PREEMPTION kernel that cond_resched() does nothing: _cond_resched() is a plain "return 0", and under PREEMPT_DYNAMIC the full and lazy modes disable it. Since commit 7dadeaa6e851 ("sched: Further restrict the preemption modes") those are the only two models on the architectures with PREEMPT_LAZY support, arm64 and x86 among them, so the drain loop never reports a Tasks-RCU quiescent state. A worker draining a cgwb with millions of attached inodes runs for minutes. On a 6.18 arm64 host in lazy mode the cgwb worker drained one dying cgroup's writeback domain for over 11 minutes. A BPF program unlink (bpf_trampoline_unlink_prog -> bpf_trampoline_update -> unregister_ftrace_direct -> ftrace_shutdown -> synchronize_rcu_tasks()) waited on that grace period while holding the trampoline mutex, 42 tasks queued behind it in D state, and the hung task detector fired at 614 s and panicked the host. Any BPF or ftrace detach during a long drain inherits the drain's length. Fix this by calling cond_resched_tasks_rcu_qs() so we do not stall out anybody who calls sycnrhonize_rcu_tasks(). We put this in a do { } while loop because if we have many small cgroups cleanup_offline_cgwb() will return false and we will never call cond_resched_tasks_rcu_qs(), creating the same problem. Link: https://lore.kernel.org/20260909-cgwb-tasks-rcu-qs-v1-1-967a7754771f@toxicpanda.com Fixes: c22d70a162d3 ("writeback, cgroup: release dying cgwbs by switching attached inodes") Signed-off-by: Josef Bacik Signed-off-by: Andrew Morton Link: https://lore.kernel.org/bpf/9d444098-7c03-4163-af12-bd0a79a51443@paulmck-laptop/ Assisted-by: LLM Acked-by: Tejun Heo Reviewed-by: Roman Gushchin Reviewed-by: Jan Kara Acked-by: Lorenzo Stoakes (ARM) Cc: David Hildenbrand Cc: Dennis Zhou Cc: Liam R. Howlett Cc: Matthew Wilcox (Oracle) Cc: Michal Hocko Cc: Mike Rapoport Cc: "Paul E . McKenney" Cc: Suren Baghdasaryan Cc: Vlastimil Babka Cc: --- mm/backing-dev.c | 5 +++-- 1 file changed, 3 insertions(+), 2 deletions(-) diff --git a/mm/backing-dev.c b/mm/backing-dev.c index cecbcf9060a6..18e999053bae 100644 --- a/mm/backing-dev.c +++ b/mm/backing-dev.c @@ -910,8 +910,9 @@ static void cleanup_offline_cgwbs_workfn(struct work_struct *work) continue; spin_unlock_irq(&cgwb_lock); - while (cleanup_offline_cgwb(wb)) - cond_resched(); + do { + cond_resched_tasks_rcu_qs(); + } while (cleanup_offline_cgwb(wb)); spin_lock_irq(&cgwb_lock); wb_put(wb); From 692dd08e03f4a5523650e055f033f846bee43a0c Mon Sep 17 00:00:00 2001 From: Xu Xin Date: Mon, 7 Sep 2026 14:53:42 +0800 Subject: [PATCH 386/592] MAINTAINERS: update Xu Xin's email Now I'm moving to the @linux.dev account, so map my old email addresses and update them to my new address. Link: https://lore.kernel.org/20260907145342529uOGtNYWTAzrSCFnghfgCb@zte.com.cn Signed-off-by: Xu Xin Signed-off-by: Andrew Morton --- .mailmap | 2 ++ MAINTAINERS | 2 +- 2 files changed, 3 insertions(+), 1 deletion(-) diff --git a/.mailmap b/.mailmap index 59518b814b4f..2447811f773a 100644 --- a/.mailmap +++ b/.mailmap @@ -980,6 +980,8 @@ Wesley Cheng Will Deacon Wolfram Sang Wolfram Sang +Xu Xin +Xu Xin xu xin Yakir Yang Yanteng Si Ying Huang diff --git a/MAINTAINERS b/MAINTAINERS index a6c0e4bcbc78..7d12dfac0149 100644 --- a/MAINTAINERS +++ b/MAINTAINERS @@ -17192,7 +17192,7 @@ F: tools/testing/selftests/mm/gup_test.c MEMORY MANAGEMENT - KSM (Kernel Samepage Merging) M: Andrew Morton M: David Hildenbrand -R: Xu Xin +R: Xu Xin R: Chengming Zhou L: linux-mm@kvack.org S: Maintained From b0be01f133aa36d2fd12d71c916cb8a47826b4ed Mon Sep 17 00:00:00 2001 From: Shawn Guo Date: Sun, 30 Aug 2026 11:02:01 +0800 Subject: [PATCH 387/592] pinctrl: qcom: nord: Split QUP1 SE2/SE3 into lane-pair functions QUP1 SE2 and SE3 pack all four of their lanes pair-wise onto only two pins each: lanes 0/1 (I2C SDA/SCL) at mux value 2 and lanes 2/3 (UART TX/RX) at mux value 1, on gpio127/gpio128 and gpio129/gpio130 respectively. Both mux values were named "qup1_se2" (respectively "qup1_se3"), so the two distinct lane pairs became indistinguishable. msm_pinmux_set_mux() stops at the first entry matching the requested function, which means mux value 1 was always selected and the I2C lanes could never be muxed out. In practice i2c9 and i2c10 got the UART lanes and did not work, while uart9 and uart10 happened to be muxed correctly. Give each lane pair its own function, following the _01/_23 naming already used for the same hardware arrangement by the shikra, eliza, hawi and maili TLMM drivers. Both functions still cover the full pin pair, so a single pinctrl state per protocol remains sufficient. Drop gpio129/gpio130 from the SE2 group list, since those pins belong to SE3 and were never reachable through the SE2 function. Also rename QUP1 SE2/SE3 functions in the binding doc accordingly. While at it, add missing "gpio", "qup3_se0_mira" and "qup3_se0_mirb" to the binding function enum to get the list complete. Fixes: c24dd0826f06 ("pinctrl: qcom: add the TLMM driver for the Nord platforms") Cc: stable@vger.kernel.org Assisted-by: LLM Signed-off-by: Shawn Guo Reviewed-by: Konrad Dybcio Link: https://patch.msgid.link/20260830030201.135637-1-shengchao.guo@oss.qualcomm.com Signed-off-by: Bartosz Golaszewski Signed-off-by: Linus Walleij --- .../bindings/pinctrl/qcom,nord-tlmm.yaml | 7 ++-- drivers/pinctrl/qcom/pinctrl-nord.c | 34 +++++++++++++------ 2 files changed, 27 insertions(+), 14 deletions(-) diff --git a/Documentation/devicetree/bindings/pinctrl/qcom,nord-tlmm.yaml b/Documentation/devicetree/bindings/pinctrl/qcom,nord-tlmm.yaml index 4bb511719f31..56758a85ead8 100644 --- a/Documentation/devicetree/bindings/pinctrl/qcom,nord-tlmm.yaml +++ b/Documentation/devicetree/bindings/pinctrl/qcom,nord-tlmm.yaml @@ -67,7 +67,7 @@ $defs: Specify the alternative function to be configured for the specified pins. - enum: [ aoss_cti, atest_char, atest_usb20, atest_usb21, + enum: [ gpio, aoss_cti, atest_char, atest_usb20, atest_usb21, aud_intfc0_clk, aud_intfc0_data, aud_intfc0_ws, aud_intfc10_clk, aud_intfc10_data, aud_intfc10_ws, aud_intfc1_clk, aud_intfc1_data, aud_intfc1_ws, @@ -98,9 +98,10 @@ $defs: pcie3_clk_req_n, phase_flag, pll_bist_sync, pll_clk_aux, prng_rosc0, prng_rosc1, pwrbrk_i_n, qdss, qdss_cti, qspi, qup0_se0, qup0_se1, qup0_se2, qup0_se3, qup0_se4, qup0_se5, - qup1_se0, qup1_se1, qup1_se3, qup1_se2, qup1_se4, qup1_se5, + qup1_se0, qup1_se1, qup1_se2_01, qup1_se2_23, qup1_se3_01, + qup1_se3_23, qup1_se4, qup1_se5, qup1_se6, qup2_se0, qup2_se1, qup2_se2, qup2_se3, qup2_se4, - qup2_se5, qup2_se6, + qup2_se5, qup2_se6, qup3_se0_mira, qup3_se0_mirb, sailss_ospi, sdc4_clk, sdc4_cmd, sdc4_data, smb_alert, smb_alert_n, smb_clk, smb_dat, tb_trig_sdc4, tmess_prng0, tmess_prng1, tsc_timer, tsense_pwm, usb0_hs, diff --git a/drivers/pinctrl/qcom/pinctrl-nord.c b/drivers/pinctrl/qcom/pinctrl-nord.c index 7c21306e77ff..7f37f8e819ba 100644 --- a/drivers/pinctrl/qcom/pinctrl-nord.c +++ b/drivers/pinctrl/qcom/pinctrl-nord.c @@ -570,8 +570,10 @@ enum nord_functions { msm_mux_qup0_se5, msm_mux_qup1_se0, msm_mux_qup1_se1, - msm_mux_qup1_se2, - msm_mux_qup1_se3, + msm_mux_qup1_se2_01, + msm_mux_qup1_se2_23, + msm_mux_qup1_se3_01, + msm_mux_qup1_se3_23, msm_mux_qup1_se4, msm_mux_qup1_se5, msm_mux_qup1_se6, @@ -1152,11 +1154,19 @@ static const char *const qup1_se1_groups[] = { "gpio123", "gpio124", "gpio125", "gpio126", }; -static const char *const qup1_se2_groups[] = { - "gpio127", "gpio128", "gpio129", "gpio130", +static const char *const qup1_se2_01_groups[] = { + "gpio127", "gpio128", }; -static const char *const qup1_se3_groups[] = { +static const char *const qup1_se2_23_groups[] = { + "gpio127", "gpio128", +}; + +static const char *const qup1_se3_01_groups[] = { + "gpio129", "gpio130", +}; + +static const char *const qup1_se3_23_groups[] = { "gpio129", "gpio130", }; @@ -1428,8 +1438,10 @@ static const struct pinfunction nord_functions[] = { MSM_PIN_FUNCTION(qup0_se5), MSM_PIN_FUNCTION(qup1_se0), MSM_PIN_FUNCTION(qup1_se1), - MSM_PIN_FUNCTION(qup1_se2), - MSM_PIN_FUNCTION(qup1_se3), + MSM_PIN_FUNCTION(qup1_se2_01), + MSM_PIN_FUNCTION(qup1_se2_23), + MSM_PIN_FUNCTION(qup1_se3_01), + MSM_PIN_FUNCTION(qup1_se3_23), MSM_PIN_FUNCTION(qup1_se4), MSM_PIN_FUNCTION(qup1_se5), MSM_PIN_FUNCTION(qup1_se6), @@ -1633,13 +1645,13 @@ static const struct msm_pingroup nord_groups[] = { _, _, _, _, _, _, _), [126] = PINGROUP(126, qup1_se1, qup1_se0, ccu_i2c_scl, mdp1_vsync_out, _, atest_usb20, ddr_pxi, _, _, _, _), - [127] = PINGROUP(127, qup1_se2, qup1_se2, _, atest_usb21, ddr_pxi, + [127] = PINGROUP(127, qup1_se2_23, qup1_se2_01, _, atest_usb21, ddr_pxi, _, _, _, _, _, _), - [128] = PINGROUP(128, qup1_se2, qup1_se2, _, atest_usb20, ddr_pxi, + [128] = PINGROUP(128, qup1_se2_23, qup1_se2_01, _, atest_usb20, ddr_pxi, _, _, _, _, _, _), - [129] = PINGROUP(129, qup1_se3, qup1_se3, ccu_i2c_sda, mdp1_vsync_out, + [129] = PINGROUP(129, qup1_se3_23, qup1_se3_01, ccu_i2c_sda, mdp1_vsync_out, _, atest_usb21, ddr_pxi, _, _, _, _), - [130] = PINGROUP(130, qup1_se3, qup1_se3, ccu_i2c_scl, mdp1_vsync_out, + [130] = PINGROUP(130, qup1_se3_23, qup1_se3_01, ccu_i2c_scl, mdp1_vsync_out, _, atest_usb20, ddr_pxi, _, _, _, _), [131] = PINGROUP(131, qup1_se4, qup1_se6, ccu_i2c_sda, mdp1_vsync_out, _, atest_usb21, ddr_pxi, _, _, _, _), From 447dcff90557748a5ac384aaf5d70b2c7e3b961d Mon Sep 17 00:00:00 2001 From: "hpp.iscas" Date: Sat, 5 Sep 2026 21:43:40 +0800 Subject: [PATCH 388/592] pinctrl: qcom: ipq5210: Publish the OF module alias The IPQ5210 TLMM platform driver can be a module and matches through ipq5210_tlmm_of_match. This table is not published for OF modalias matching. Publish the existing table, preserving arch_initcall ordering and the shared MSM pinctrl probe. Fixes: a549fe22376f ("pinctrl: qcom: Introduce IPQ5210 TLMM driver") Signed-off-by: hpp.iscas Reviewed-by: Konrad Dybcio Link: https://patch.msgid.link/20260905134340.67477-1-hppiscas@163.com Signed-off-by: Bartosz Golaszewski Signed-off-by: Linus Walleij --- drivers/pinctrl/qcom/pinctrl-ipq5210.c | 1 + 1 file changed, 1 insertion(+) diff --git a/drivers/pinctrl/qcom/pinctrl-ipq5210.c b/drivers/pinctrl/qcom/pinctrl-ipq5210.c index 827a4ad07a6b..d5b4eb3e7343 100644 --- a/drivers/pinctrl/qcom/pinctrl-ipq5210.c +++ b/drivers/pinctrl/qcom/pinctrl-ipq5210.c @@ -867,6 +867,7 @@ static const struct of_device_id ipq5210_tlmm_of_match[] = { { .compatible = "qcom,ipq5210-tlmm", }, { }, }; +MODULE_DEVICE_TABLE(of, ipq5210_tlmm_of_match); static int ipq5210_tlmm_probe(struct platform_device *pdev) { From 99cc2a62e07a44a22254d7beca9ef1f8ad886d0d Mon Sep 17 00:00:00 2001 From: Eric Dumazet Date: Sun, 13 Sep 2026 04:42:33 +0000 Subject: [PATCH 389/592] tipc: reject invalid and unexpected GRP_ACK_MSG to prevent bc_ackers underflow Commit 48a5fe38772b ("tipc: fix bc_ackers underflow on duplicate GRP_ACK_MSG") rejected duplicate/stale ACKs in tipc_group_proto_rcv() by returning early when less_eq(acked, m->bc_acked). However, that check remains incomplete in two ways: 1. When grp->bc_ackers is zero (e.g. on a quiet group, when replicast ACKs were not requested, or after all expected members have already acknowledged), an unexpected GRP_ACK_MSG with acked > m->bc_acked passes less_eq() and unconditionally decrements grp->bc_ackers. Because bc_ackers is a u16, this wraps to 65535, causing tipc_group_bc_cong() to permanently report congestion and blocking all future group broadcasts on the socket. 2. During an active broadcast round (grp->bc_ackers > 0), the sender transmits packet S and advances grp->bc_snd_nxt to S + 1. Receivers increment their expected counter to S + 1 upon consuming packet S, so the only valid ACK value for the current round is strictly acked == grp->bc_snd_nxt. However, tipc_group_update_bc_members() initializes each member's m->bc_acked to prev = grp->bc_snd_nxt - 1 (S - 1 before increment). This leaves a 2-sequence gap (S - 1 to S + 1) in sequence space. An incoming ACK is therefore neither rejected as duplicate nor prevented from decrementing grp->bc_ackers if an unexpected or stale value (such as S) is received. A member sending acked = S followed by acked = S + 1 could decrement grp->bc_ackers twice in the same round, prematurely clearing bc_ackers or underflowing it. Fix this by: - Dropping GRP_ACK_MSG immediately if grp->bc_ackers is zero. - Requiring acked == grp->bc_snd_nxt and rejecting duplicates where m->bc_acked == acked. Because replicast broadcast rounds are strictly sequential, only grp->bc_snd_nxt can be acknowledged, and each member can acknowledge at most once per round. Note that a related pre-existing issue in tipc_group_delete_member() (where grp->bc_ackers decrementing to zero upon member departure does not restore *grp->open or trigger a socket wakeup) will be addressed in a separate patch. Fixes: 48a5fe38772b ("tipc: fix bc_ackers underflow on duplicate GRP_ACK_MSG") Fixes: 2f487712b893 ("tipc: guarantee that group broadcast doesn't bypass group unicast") Reported-by: James Burton Cc: stable@vger.kernel.org Signed-off-by: Eric Dumazet Link: https://patch.msgid.link/20260913044233.193927-1-edumazet@google.com Signed-off-by: Jakub Kicinski --- net/tipc/group.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/net/tipc/group.c b/net/tipc/group.c index 14e6732624e2..74f6d3dac078 100644 --- a/net/tipc/group.c +++ b/net/tipc/group.c @@ -797,10 +797,10 @@ void tipc_group_proto_rcv(struct tipc_group *grp, bool *usr_wakeup, tipc_group_open(m, usr_wakeup); return; case GRP_ACK_MSG: - if (!m) + if (!m || !grp->bc_ackers) return; acked = msg_grp_bc_acked(hdr); - if (less_eq(acked, m->bc_acked)) + if (acked != grp->bc_snd_nxt || m->bc_acked == acked) return; m->bc_acked = acked; if (--grp->bc_ackers) From 47abe7a5c4eb53269aca3506446f851572a059a3 Mon Sep 17 00:00:00 2001 From: Nguyen Ngoc Thang Date: Tue, 15 Sep 2026 22:08:16 +0700 Subject: [PATCH 390/592] net/sched: act_ct: don't WARN on benign flow_offload_alloc() failure flow_offload_alloc() returns NULL when the conntrack entry is dying (e.g. raced with a conntrack flush) or when the GFP_ATOMIC allocation fails; both are expected under load and neither is a kernel bug. This path runs from softirq on every committed packet, so with panic_on_warn=1 an unprivileged user can panic the box just by racing a conntrack flush against a `tc ... action ct commit` classifier. Reproduced with a custom repro under QEMU: a small, fixed set of UDP flows through `tc filter ... action ct commit` on lo, raced against threads flooding bare ctnetlink CT_DELETE (flush) requests. Hits WARNING: net/sched/act_ct.c:437 (tcf_ct_flow_table_add(), inlined into tcf_ct_act() in this build) within ~15s on the unpatched kernel; same setup is clean on the patched kernel. The fix itself is behavior-preserving: both branches already did `goto err_alloc` before and after, only the WARN is removed. Fixes: 64ff70b80fd4 ("net/sched: act_ct: Offload established connections to flow table") Reported-by: syzbot+6cc37aba98dac721c415@syzkaller.appspotmail.com Closes: https://syzkaller.appspot.com/bug?extid=6cc37aba98dac721c415 Signed-off-by: Nguyen Ngoc Thang Reviewed-by: Simon Horman Link: https://patch.msgid.link/20260915150816.36487-1-ngocthang2710.1999@gmail.com Signed-off-by: Jakub Kicinski --- net/sched/act_ct.c | 5 ++--- 1 file changed, 2 insertions(+), 3 deletions(-) diff --git a/net/sched/act_ct.c b/net/sched/act_ct.c index 9080cb386c16..55f3521edb4c 100644 --- a/net/sched/act_ct.c +++ b/net/sched/act_ct.c @@ -432,11 +432,10 @@ static void tcf_ct_flow_table_add(struct tcf_ct_flow_table *ct_ft, if (test_and_set_bit(IPS_OFFLOAD_BIT, &ct->status)) return; + /* NULL if ct is dying (raced flush) or the atomic alloc failed. */ entry = flow_offload_alloc(ct); - if (!entry) { - WARN_ON_ONCE(1); + if (!entry) goto err_alloc; - } if (tcp) { ct->proto.tcp.seen[0].flags |= IP_CT_TCP_FLAG_BE_LIBERAL; From 2566866fc30965d915d0b52b5c3323b362619f0e Mon Sep 17 00:00:00 2001 From: =?UTF-8?q?J=C3=A9r=C3=A9my=20Jean?= Date: Tue, 15 Sep 2026 12:48:07 +0000 Subject: [PATCH 391/592] net: gue: reject invalid REMCSUM offsets MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit The REMCSUM option carries an absolute checksum start and checksum field offset. gue_remcsum() passes them to skb_remcsum_process(), whose partial path stores offset - start in the u16 skb->csum_offset variable. If offset is less than start, this underflows. A forwarded packet can retain CHECKSUM_PARTIAL and reach a NETIF_F_HW_CSUM driver which trusts the metadata, leading skb_copy_and_csum_dev() to write two bytes about 64 KiB beyond the destination buffer. Reject reversed tuples in validate_gue_flags(), after the existing length validation, so all GUE parsers enforce the ordering in one place. Fixes: fe881ef11cf0 ("gue: Use checksum partial with remote checksum offload") Signed-off-by: Jérémy Jean Link: https://patch.msgid.link/20260915124806.2852293-2-Jeremy.Jean@oss.cyber.gouv.fr Signed-off-by: Jakub Kicinski --- include/net/gue.h | 19 +++++++++++++++---- 1 file changed, 15 insertions(+), 4 deletions(-) diff --git a/include/net/gue.h b/include/net/gue.h index caefd6da8693..d377155fd0b3 100644 --- a/include/net/gue.h +++ b/include/net/gue.h @@ -84,8 +84,9 @@ static inline size_t guehdr_priv_flags_len(__be32 flags) } /* Validate standard and private flags. Returns non-zero (meaning invalid) - * if there is an unknown standard or private flags, or the options length for - * the flags exceeds the options length specific in hlen of the GUE header. + * if there is an unknown standard or private flags, if the options length for + * the flags exceeds the options length specified in hlen of the GUE header, or + * if a private option contains invalid data. */ static inline int validate_gue_flags(struct guehdr *guehdr, size_t optlen) { @@ -103,8 +104,8 @@ static inline int validate_gue_flags(struct guehdr *guehdr, size_t optlen) /* Private flags are last four bytes accounted in * guehdr_flags_len */ - __be32 pflags = *(__be32 *)((void *)&guehdr[1] + - len - GUE_LEN_PRIV); + void *data = (void *)&guehdr[1] + len; + __be32 pflags = *(__be32 *)(data - GUE_LEN_PRIV); if (pflags & ~GUE_PFLAGS_ALL) return 1; @@ -112,6 +113,16 @@ static inline int validate_gue_flags(struct guehdr *guehdr, size_t optlen) len += guehdr_priv_flags_len(pflags); if (len > optlen) return 1; + + if (pflags & GUE_PFLAG_REMCSUM) { + __be16 *pd = data; + + /* The field offset pd[1] must not be less + * than the start pd[0]. + */ + if (ntohs(pd[1]) < ntohs(pd[0])) + return 1; + } } return 0; From 310d1ac61a4d5a2ca8356a3a48d263acf54503ce Mon Sep 17 00:00:00 2001 From: Lorenzo Bianconi Date: Wed, 16 Sep 2026 15:30:13 +0200 Subject: [PATCH 392/592] net: ethernet: mtk_eth_soc: unregister net_devices in case of probe failure If register_netdev() fails for one of the MTK_MAX_DEVS devices in mtk_probe(), the error path jumps to err_deinit_ppe, skipping mtk_unreg_dev(). The previously registered net_devices are then freed by mtk_free_dev() while still in NETREG_REGISTERED state, hitting the BUG_ON(dev->reg_state != NETREG_UNREGISTERED). Route the register_netdev() failure to err_unreg_netdev so the net_devices registered so far are properly unregistered before being freed. Fixes: 8a8a9e89f801 ("net: ethernet: mediatek: cleanup error path inside mtk_hw_init") Signed-off-by: Lorenzo Bianconi Link: https://patch.msgid.link/20260916-mtk_eth_soc-netdev-fix-v1-1-5dac50eb65b1@oss.qualcomm.com Signed-off-by: Jakub Kicinski --- drivers/net/ethernet/mediatek/mtk_eth_soc.c | 6 +++++- 1 file changed, 5 insertions(+), 1 deletion(-) diff --git a/drivers/net/ethernet/mediatek/mtk_eth_soc.c b/drivers/net/ethernet/mediatek/mtk_eth_soc.c index fd7a49ae88d0..2ea5dfe85539 100644 --- a/drivers/net/ethernet/mediatek/mtk_eth_soc.c +++ b/drivers/net/ethernet/mediatek/mtk_eth_soc.c @@ -4509,6 +4509,10 @@ static int mtk_unreg_dev(struct mtk_eth *eth) mac = netdev_priv(eth->netdev[i]); if (MTK_HAS_CAPS(eth->soc->caps, MTK_QDMA)) unregister_netdevice_notifier(&mac->device_notifier); + + if (eth->netdev[i]->reg_state != NETREG_REGISTERED) + continue; + unregister_netdev(eth->netdev[i]); } @@ -5344,7 +5348,7 @@ static int mtk_probe(struct platform_device *pdev) err = register_netdev(eth->netdev[i]); if (err) { dev_err(eth->dev, "error bringing up device\n"); - goto err_deinit_ppe; + goto err_unreg_netdev; } else netif_info(eth, probe, eth->netdev[i], "mediatek frame engine at 0x%08lx, irq %d\n", From 24fedc7a569bce181728f0dd616504bef9f1513b Mon Sep 17 00:00:00 2001 From: Andy Moreton Date: Wed, 16 Sep 2026 13:56:41 +0100 Subject: [PATCH 393/592] sfc: add X4D PF support X4D is an X4 controller instance as an IP block in an SoC. It has the same feature set as X4. Signed-off-by: Andy Moreton Reviewed-by: Pieter Jansen van Vuuren Reviewed-by: Alejandro Lucero Reviewed-by: Simon Horman Link: https://patch.msgid.link/20260916125641.12238-1-alucerop@amd.com Signed-off-by: Jakub Kicinski --- drivers/net/ethernet/sfc/efx.c | 4 ++++ 1 file changed, 4 insertions(+) diff --git a/drivers/net/ethernet/sfc/efx.c b/drivers/net/ethernet/sfc/efx.c index 3806cd3dd7f4..953f1d90b9b1 100644 --- a/drivers/net/ethernet/sfc/efx.c +++ b/drivers/net/ethernet/sfc/efx.c @@ -904,6 +904,10 @@ static const struct pci_device_id efx_pci_table[] = { .driver_data = (unsigned long)&efx_x4_nic_type}, {PCI_DEVICE(PCI_VENDOR_ID_SOLARFLARE, 0x2c03), /* X4 PF (FF only) */ .driver_data = (unsigned long)&efx_x4_nic_type}, + {PCI_DEVICE(PCI_VENDOR_ID_SOLARFLARE, 0x8c03), /* X4D PF (FF/LL) */ + .driver_data = (unsigned long)&efx_x4_nic_type}, + {PCI_DEVICE(PCI_VENDOR_ID_SOLARFLARE, 0xac03), /* X4D PF (FF only) */ + .driver_data = (unsigned long)&efx_x4_nic_type}, {0} /* end of list */ }; From ee319bd3a0e976af5087cbe59ebc50a66f31d202 Mon Sep 17 00:00:00 2001 From: Norbert Szetei Date: Wed, 16 Sep 2026 21:57:53 +0200 Subject: [PATCH 394/592] ipv6: do not let ipv6_find_hdr() return an offset past the packet end ipv6_find_hdr() walks the extension header chain, skipping each header by the length that header itself declares. ipv6_optlen() returns up to 2048, and the skip is never checked against skb->len, so the offset stored in *offset can point past the end of the packet. openvswitch installs that offset as the transport header, and update_ipv6_checksum() then reads and writes the transport checksum field out of bounds: BUG: KASAN: slab-use-after-free in inet_proto_csum_replace16+0x445/0x470 Read of size 2 at addr ffff88810b754b06 by task ovs_ipv6_oob/629 CPU: 4 UID: 1000 PID: 629 Comm: ovs_ipv6_oob Tainted: G N 7.3.0-rc3+ #348 Call Trace: inet_proto_csum_replace16+0x445/0x470 set_ipv6_addr+0x3dd/0x460 do_execute_actions+0x6a3d/0x7c40 ovs_execute_actions+0xfd/0x480 ovs_packet_cmd_execute+0xc38/0xf20 genl_rcv_msg+0x59e/0x870 netlink_rcv_skb+0x18b/0x450 genl_rcv+0x2d/0x40 netlink_unicast+0x6bc/0xa20 The buggy address belongs to the object at ffff88810b754980 which belongs to the cache skbuff_small_head of size 704 The buggy address is located 390 bytes inside of freed 704-byte region [ffff88810b754980, ffff88810b754c40) Other callers use that offset too, so bound it here rather than in one caller. Reject a header whose declared length does not fit in the packet. ipv6_find_hdr() already fails with -EBADMSG on a malformed chain, so this adds no new failure mode. Fixes: f8f626754ebe ("ipv6: Move ipv6_find_hdr() out of Netfilter code.") Suggested-by: Ilya Maximets Suggested-by: Eric Dumazet Cc: stable@vger.kernel.org Signed-off-by: Norbert Szetei Reviewed-by: Ido Schimmel Reviewed-by: Ilya Maximets Link: https://patch.msgid.link/8F80BA1A-DDFD-432D-9075-242A3435FEB5@doyensec.com Signed-off-by: Jakub Kicinski --- net/ipv6/exthdrs_core.c | 3 +++ 1 file changed, 3 insertions(+) diff --git a/net/ipv6/exthdrs_core.c b/net/ipv6/exthdrs_core.c index 9d06d487e8b1..4a9748338cf4 100644 --- a/net/ipv6/exthdrs_core.c +++ b/net/ipv6/exthdrs_core.c @@ -278,6 +278,9 @@ int ipv6_find_hdr(const struct sk_buff *skb, unsigned int *offset, hdrlen = ipv6_optlen(hp); if (!found) { + if (skb->len - start < hdrlen) + return -EBADMSG; + nexthdr = hp->nexthdr; start += hdrlen; } From dd47bcf279f1083f09bf5266890b26263361022b Mon Sep 17 00:00:00 2001 From: Kuniyuki Iwashima Date: Wed, 16 Sep 2026 23:09:24 +0000 Subject: [PATCH 395/592] ip6_gre: Call ip6erspan_tunnel_unlink_md() in ip6erspan_changelink(). The cited commit accidentally added ip6gre_tunnel_unlink_md() in ip6erspan_changelink(). Let's correct it to ip6erspan_tunnel_unlink_md(). Fixes: b80d0b93b991 ("net: ip6_gre: fix tunnel metadata device sharing.") Signed-off-by: Kuniyuki Iwashima Reviewed-by: Xuanqiang Luo Reviewed-by: Ido Schimmel Link: https://patch.msgid.link/20260916230927.378957-1-kuniyu@google.com Signed-off-by: Jakub Kicinski --- net/ipv6/ip6_gre.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/net/ipv6/ip6_gre.c b/net/ipv6/ip6_gre.c index 8ebda0b6a78b..e61cb10b50dc 100644 --- a/net/ipv6/ip6_gre.c +++ b/net/ipv6/ip6_gre.c @@ -2279,7 +2279,7 @@ static int ip6erspan_changelink(struct net_device *dev, struct nlattr *tb[], return PTR_ERR(t); ip6erspan_set_version(data, &p); - ip6gre_tunnel_unlink_md(ign, t); + ip6erspan_tunnel_unlink_md(ign, t); ip6gre_tunnel_unlink(ign, t); ip6erspan_tnl_change(t, &p, !tb[IFLA_MTU]); ip6erspan_tunnel_link_md(ign, t); From 95c4d54ed02283e9a09e8cd7360e384daa67a741 Mon Sep 17 00:00:00 2001 From: Abhishek Ojha Date: Wed, 16 Sep 2026 19:19:28 -0400 Subject: [PATCH 396/592] net: phy: micrel: Advance register data pointer in write loop lanphy_write_reg_data() does not advance the data pointer while iterating over the register table. As a result, it writes the first entry num times and leaves the remaining errata registers unconfigured. Single-entry tables are unaffected, but tables with multiple entries leave every entry after the first unapplied. Advance the data pointer after each successful write so every table entry is applied in order. Fixes: c8732e933925 ("net: phy: micrel: lan8842 errata") Cc: stable@vger.kernel.org Signed-off-by: Abhishek Ojha Reviewed-by: Andrew Lunn Link: https://patch.msgid.link/20260916231928.1336305-1-abhishek.ojha@savoirfairelinux.com Signed-off-by: Jakub Kicinski --- drivers/net/phy/micrel.c | 1 + 1 file changed, 1 insertion(+) diff --git a/drivers/net/phy/micrel.c b/drivers/net/phy/micrel.c index ae830781824b..5c8461db7b4b 100644 --- a/drivers/net/phy/micrel.c +++ b/drivers/net/phy/micrel.c @@ -6344,6 +6344,7 @@ static int lanphy_write_reg_data(struct phy_device *phydev, data->val); if (ret) break; + data++; } return ret; From 1b82958f3f035df5ccaab5430a2302f08a5d5351 Mon Sep 17 00:00:00 2001 From: Alexander Duyck Date: Mon, 14 Sep 2026 14:09:57 -0700 Subject: [PATCH 397/592] net: ethtool: keep rtnl_lock for the ioctl self test An offline self test that brings the interface down and back up with netif_close() / netif_open() requires rtnl_lock for both. Since the ethtool IOCTL path became rtnl-optional for ops-locked drivers, the ETHTOOL_TEST ioctl runs holding only the netdev instance lock, so on an ops-locked driver the self test now tears the device down without rtnl_lock. With lockdep this reproduces deterministically on every offline self test on such a driver; note the sole lock held is the instance lock, not rtnl: WARNING: suspicious RCU usage net/core/netpoll.c:207 suspicious rcu_dereference_protected() usage! 1 lock held by ethtool/107: #0: (&dev->lock){+.+.}, at: dev_ethtool Call Trace: netpoll_poll_disable __dev_close_many netif_close_many netif_close fbnic_self_test dev_ethtool_locked dev_ethtool dev_ioctl sock_ioctl __x64_sys_ioctl Without lockdep the same condition trips ASSERT_RTNL() in __dev_close_many() / __dev_open(); that check only samples the global rtnl state, so it can be masked by a concurrent rtnl holder, but the device is still being reconfigured without the lock it requires. The ethtool self_test is a legacy ioctl-only command, so an ETHTOOL_TEST case is only needed on the ioctl path. Add an opt-in bit for drivers whose self test needs rtnl_lock and set it on the ops-locked drivers whose offline self test tears the interface down and up: - fbnic (ops-locked via queue_mgmt_ops): fbnic_self_test() offline path uses netif_close() / netif_open(). - bnxt (ops-locked via queue_mgmt_ops): bnxt_self_test() offline path goes through bnxt_close_nic() / bnxt_half_open_nic() / bnxt_half_close_nic() / bnxt_open_nic(), which close and reopen the device. Fixes: f994752b1127 ("net: ethtool: optionally skip rtnl_lock on IOCTL path") Signed-off-by: Alexander Duyck Reviewed-by: Simon Horman Link: https://patch.msgid.link/178942019771.7700.338431553546884773.stgit@ahduyck-xeon-server.home.arpa Signed-off-by: Jakub Kicinski --- drivers/net/ethernet/broadcom/bnxt/bnxt_ethtool.c | 3 ++- drivers/net/ethernet/meta/fbnic/fbnic_ethtool.c | 3 ++- include/linux/ethtool.h | 2 ++ net/ethtool/common.h | 2 ++ 4 files changed, 8 insertions(+), 2 deletions(-) diff --git a/drivers/net/ethernet/broadcom/bnxt/bnxt_ethtool.c b/drivers/net/ethernet/broadcom/bnxt/bnxt_ethtool.c index 62bc9cae613c..622e89587e5d 100644 --- a/drivers/net/ethernet/broadcom/bnxt/bnxt_ethtool.c +++ b/drivers/net/ethernet/broadcom/bnxt/bnxt_ethtool.c @@ -5733,7 +5733,8 @@ const struct ethtool_ops bnxt_ethtool_ops = { .op_needs_rtnl = ETHTOOL_OP_NEEDS_RTNL_SCHANNELS | ETHTOOL_OP_NEEDS_RTNL_SRINGPARAM | ETHTOOL_OP_NEEDS_RTNL_SCOALESCE | - ETHTOOL_OP_NEEDS_RTNL_RSS, + ETHTOOL_OP_NEEDS_RTNL_RSS | + ETHTOOL_OP_NEEDS_RTNL_TEST, .supported_coalesce_params = ETHTOOL_COALESCE_USECS | ETHTOOL_COALESCE_MAX_FRAMES | ETHTOOL_COALESCE_USECS_IRQ | diff --git a/drivers/net/ethernet/meta/fbnic/fbnic_ethtool.c b/drivers/net/ethernet/meta/fbnic/fbnic_ethtool.c index 0e47088ec44b..423f179c9d47 100644 --- a/drivers/net/ethernet/meta/fbnic/fbnic_ethtool.c +++ b/drivers/net/ethernet/meta/fbnic/fbnic_ethtool.c @@ -2025,7 +2025,8 @@ static const struct ethtool_ops fbnic_ethtool_ops = { ETHTOOL_OP_NEEDS_RTNL_SPAUSEPARAM | ETHTOOL_OP_NEEDS_RTNL_SCHANNELS | ETHTOOL_OP_NEEDS_RTNL_SRINGPARAM | - ETHTOOL_OP_NEEDS_RTNL_GLINK, + ETHTOOL_OP_NEEDS_RTNL_GLINK | + ETHTOOL_OP_NEEDS_RTNL_TEST, .get_drvinfo = fbnic_get_drvinfo, .get_regs_len = fbnic_get_regs_len, .get_regs = fbnic_get_regs, diff --git a/include/linux/ethtool.h b/include/linux/ethtool.h index 253600c0eccd..c4c9ce038611 100644 --- a/include/linux/ethtool.h +++ b/include/linux/ethtool.h @@ -944,6 +944,7 @@ struct kernel_ethtool_ts_info { #define ETHTOOL_OP_NEEDS_RTNL_SPAUSEPARAM BIT(6) #define ETHTOOL_OP_NEEDS_RTNL_RSS BIT(7) #define ETHTOOL_OP_NEEDS_RTNL_GLINK BIT(8) +#define ETHTOOL_OP_NEEDS_RTNL_TEST BIT(9) /** * struct ethtool_ops - optional netdev operations @@ -981,6 +982,7 @@ struct kernel_ethtool_ts_info { * - netdev_update_features() * - netif_set_real_num_tx_queues() * - ethtool_op_get_link() (syncs link watch under rtnl_lock) + * - netif_open() / netif_close() (used by @self_test) * * @get_drvinfo: Report driver/device information. Modern drivers no * longer have to implement this callback. Most fields are diff --git a/net/ethtool/common.h b/net/ethtool/common.h index 4e5356e26f40..ae32e7fdb563 100644 --- a/net/ethtool/common.h +++ b/net/ethtool/common.h @@ -163,6 +163,8 @@ ethtool_ioctl_needs_rtnl(const struct net_device *dev, u32 ethcmd) return ops->op_needs_rtnl & ETHTOOL_OP_NEEDS_RTNL_RSS; case ETHTOOL_GLINK: return ops->op_needs_rtnl & ETHTOOL_OP_NEEDS_RTNL_GLINK; + case ETHTOOL_TEST: + return ops->op_needs_rtnl & ETHTOOL_OP_NEEDS_RTNL_TEST; } return false; } From 1f4c73064a50f53d596c6f1d06d2d700f43c4b32 Mon Sep 17 00:00:00 2001 From: =?UTF-8?q?Bj=C3=B6rn=20T=C3=B6pel?= Date: Mon, 14 Sep 2026 14:10:04 -0700 Subject: [PATCH 398/592] eth: fbnic: Handle maximum standalone channels MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit Standalone channels use one NAPI vector for each Tx and Rx queue. fbnic's allocation path excludes FBNIC_MAX_TXQS from that layout. A 64-Tx/64-Rx configuration therefore records 128 vectors but allocates only 64, leaving NULL entries that resource setup dereferences. Include the maximum vector count in standalone allocation. Fixes: bc6107771bb4 ("eth: fbnic: Allocate a netdevice and napi vectors with queues") Signed-off-by: Björn Töpel Reviewed-by: Simon Horman Link: https://patch.msgid.link/178942020457.7700.13129750616387075931.stgit@ahduyck-xeon-server.home.arpa Signed-off-by: Jakub Kicinski --- drivers/net/ethernet/meta/fbnic/fbnic_txrx.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/drivers/net/ethernet/meta/fbnic/fbnic_txrx.c b/drivers/net/ethernet/meta/fbnic/fbnic_txrx.c index 661dee1661af..a30aa4450848 100644 --- a/drivers/net/ethernet/meta/fbnic/fbnic_txrx.c +++ b/drivers/net/ethernet/meta/fbnic/fbnic_txrx.c @@ -1791,7 +1791,7 @@ int fbnic_alloc_napi_vectors(struct fbnic_net *fbn) int err; /* Allocate 1 Tx queue per napi vector */ - if (num_napi < FBNIC_MAX_TXQS && num_napi == num_tx + num_rx) { + if (num_napi <= FBNIC_MAX_TXQS && num_napi == num_tx + num_rx) { while (num_tx) { err = fbnic_alloc_napi_vector(fbd, fbn, num_napi, v_idx, From b5d9e9d4d0c13bc8b60d8d97e7a07fb25fea639e Mon Sep 17 00:00:00 2001 From: Alexander Duyck Date: Mon, 14 Sep 2026 14:10:11 -0700 Subject: [PATCH 399/592] eth: fbnic: use the Rx queue napi pointer to find the napi vector The queue management ndos pick the napi vector for an Rx queue with: nv = fbn->napi[idx % fbn->num_napi]; The issue is this is only correct in the cases where there are no standalone Tx vectors. In those cases we were allocating the Tx vectors first and then the Rx so the queues would be pointing to Tx NAPI vectors instead of the Rx ones. The mapping the ndos want is already recorded. fbnic_set_netif_napi() publishes it with netif_queue_set_napi(), which stores the napi pointer in netdev_rx_queue.napi, and fbnic_reset_netif_napi() clears it again. Both run under the netdev instance lock that the queue management ndos also hold, so the pointer can be read directly. Use it and drop the divide. The pointer is NULL exactly while the datapath is down, so fbnic_queue_mem_alloc() can reject that case rather than reaching into freed state: netdev_rx_queue_restart() calls it before it tests netif_running(), and fbnic_pm_suspend() leaves netif_running() true across a PCIe recovery that never completes, so a queue restart can arrive after fbnic_stop() has freed the rings and the vectors. fbnic_stop() clears the association in fbnic_reset_netif_queues() before fbnic_free_napi_vectors(), so the NULL is always published first. fbnic_queue_start() and fbnic_queue_stop() need no check of their own, as netdev_rx_queue_reconfig() only reaches them once fbnic_queue_mem_alloc() has succeeded under the same instance lock. Fixes: da43127a8edc ("eth: fbnic: support queue ops / zero-copy Rx") Signed-off-by: Alexander Duyck Reviewed-by: Simon Horman Link: https://patch.msgid.link/178942021136.7700.4391219358260544104.stgit@ahduyck-xeon-server.home.arpa Signed-off-by: Jakub Kicinski --- drivers/net/ethernet/meta/fbnic/fbnic_txrx.c | 26 +++++++++++++++++--- 1 file changed, 23 insertions(+), 3 deletions(-) diff --git a/drivers/net/ethernet/meta/fbnic/fbnic_txrx.c b/drivers/net/ethernet/meta/fbnic/fbnic_txrx.c index a30aa4450848..10caacffee0f 100644 --- a/drivers/net/ethernet/meta/fbnic/fbnic_txrx.c +++ b/drivers/net/ethernet/meta/fbnic/fbnic_txrx.c @@ -7,6 +7,7 @@ #include #include #include +#include #include #include #include @@ -2853,6 +2854,17 @@ void fbnic_napi_depletion_check(struct net_device *netdev) fbnic_wrfl(fbd); } +/* Returns the napi vector servicing an Rx queue, or NULL if the datapath + * is torn down. The association is published by fbnic_set_netif_napi() + * and cleared by fbnic_reset_netif_napi(), both under the instance lock. + */ +static struct fbnic_napi_vector *fbnic_rxq_nv(struct net_device *dev, int idx) +{ + struct napi_struct *napi = __netif_get_rx_queue(dev, idx)->napi; + + return napi ? container_of(napi, struct fbnic_napi_vector, napi) : NULL; +} + static int fbnic_queue_mem_alloc(struct net_device *dev, struct netdev_queue_config *qcfg, void *qmem, int idx) @@ -2865,8 +2877,16 @@ static int fbnic_queue_mem_alloc(struct net_device *dev, if (!netif_running(dev)) return fbnic_alloc_qt_page_pools(fbn, qt, idx); + /* A failed PCIe recovery or resume can leave the datapath torn down + * while netif_running() is still true. This ndo runs before + * netdev_rx_queue_restart() checks netif_running(), so bail out + * rather than touching rings and vectors that are already freed. + */ + nv = fbnic_rxq_nv(dev, idx); + if (!nv) + return -ENETDOWN; + real = container_of(fbn->rx[idx], struct fbnic_q_triad, cmpl); - nv = fbn->napi[idx % fbn->num_napi]; fbnic_ring_init(&qt->sub0, real->sub0.doorbell, real->sub0.q_idx, real->sub0.flags); @@ -2917,7 +2937,7 @@ static int fbnic_queue_start(struct net_device *dev, struct fbnic_q_triad *real; real = container_of(fbn->rx[idx], struct fbnic_q_triad, cmpl); - nv = fbn->napi[idx % fbn->num_napi]; + nv = fbnic_rxq_nv(dev, idx); fbnic_aggregate_ring_bdq_counters(fbn, &real->sub0); fbnic_aggregate_ring_bdq_counters(fbn, &real->sub1); @@ -2939,7 +2959,7 @@ static int fbnic_queue_stop(struct net_device *dev, void *qmem, int idx) int err; real = container_of(fbn->rx[idx], struct fbnic_q_triad, cmpl); - nv = fbn->napi[idx % fbn->num_napi]; + nv = fbnic_rxq_nv(dev, idx); fbnic_dbg_nv_exit(nv); napi_disable_locked(&nv->napi); From 4bcc4a92c603fe7f062cea22e20da2e0ad6b12c3 Mon Sep 17 00:00:00 2001 From: Alexander Duyck Date: Mon, 14 Sep 2026 14:10:18 -0700 Subject: [PATCH 400/592] eth: fbnic: reset num_napi when the napi vectors are freed fbn->num_napi is the count of live napi vectors, each of which owns an IRQ. The PM path had freed them without clearing the count. fbnic_pm_suspend() tears the datapath down via ndo_stop() and frees the IRQs, but leaves netif_running() true so resume knows to re-open. Resume rebuilds the datapath in __fbnic_pm_resume() and fbnic_reset_queues() sets num_napi and __fbnic_open() re-allocates the vectors. When the datapath is torn down but never rebuilt, num_napi is left pointing at freed vectors under 2 different scenarios: - a PCIe error recovery that fails (fbnic_err_slot_reset() -> __fbnic_pm_resume() returns an error -> PCI_ERS_RESULT_DISCONNECT), so .resume never runs; or - an __fbnic_open() that fails partway on resume and unwinds, freeing the vectors after fbnic_reset_queues() has already set num_napi. The netdev is then running with num_napi > 0 but napi[] freed, and the eventual remove/unbind close re-enters fbnic_down() -> fbnic_dbg_down() and dereferences the freed vectors: BUG: kernel NULL pointer dereference, address: 0000000000000210 RIP: fbnic_dbg_down+0x28 Clear num_napi when the vectors are freed: in the suspend teardown (a good resume re-establishes it before __fbnic_open()) and on the resume open failure. A redundant ndo_stop() then walks an empty napi[]. The normal ndo_stop() down/up cycle is untouched and keeps num_napi for the next ndo_open(). Fixes: bc6107771bb4 ("eth: fbnic: Allocate a netdevice and napi vectors with queues") Signed-off-by: Alexander Duyck Reviewed-by: Simon Horman Link: https://patch.msgid.link/178942021809.7700.10804028989308077839.stgit@ahduyck-xeon-server.home.arpa Signed-off-by: Jakub Kicinski --- drivers/net/ethernet/meta/fbnic/fbnic_pci.c | 18 ++++++++++++++---- 1 file changed, 14 insertions(+), 4 deletions(-) diff --git a/drivers/net/ethernet/meta/fbnic/fbnic_pci.c b/drivers/net/ethernet/meta/fbnic/fbnic_pci.c index 8b9bc9e8ea56..c6698e3002a1 100644 --- a/drivers/net/ethernet/meta/fbnic/fbnic_pci.c +++ b/drivers/net/ethernet/meta/fbnic/fbnic_pci.c @@ -434,6 +434,7 @@ static int fbnic_pm_suspend(struct device *dev) { struct fbnic_dev *fbd = dev_get_drvdata(dev); struct net_device *netdev = fbd->netdev; + struct fbnic_net *fbn; if (fbnic_init_failure(fbd)) goto null_uc_addr; @@ -441,11 +442,16 @@ static int fbnic_pm_suspend(struct device *dev) rtnl_lock(); netdev_lock(netdev); + fbn = netdev_priv(netdev); + netif_device_detach(netdev); if (netif_running(netdev)) netdev->netdev_ops->ndo_stop(netdev); + /* The IRQs are about to be freed, so drop the napi vector count */ + fbn->num_napi = 0; + netdev_unlock(netdev); rtnl_unlock(); @@ -508,16 +514,20 @@ static int __fbnic_pm_resume(struct device *dev) if (fbnic_init_failure(fbd)) return 0; + rtnl_lock(); + netdev_lock(netdev); + fbn = netdev_priv(netdev); /* Reset the queues if needed */ fbnic_reset_queues(fbn, fbn->num_tx_queues, fbn->num_rx_queues); - rtnl_lock(); - netdev_lock(netdev); - - if (netif_running(netdev)) + if (netif_running(netdev)) { err = __fbnic_open(fbn); + /* On failure the vectors are freed, so drop the count */ + if (err) + fbn->num_napi = 0; + } netdev_unlock(netdev); rtnl_unlock(); From 8947f13e436a4ff5eed9f8f019b2865a07af4bb2 Mon Sep 17 00:00:00 2001 From: Alexander Duyck Date: Mon, 14 Sep 2026 14:10:25 -0700 Subject: [PATCH 401/592] eth: fbnic: Set AW_FLUSH_MODE alongside AW_FLUSH when flushing the mailbox When tearing down the FW mailbox Rx ring, fbnic_mbx_reset_desc_ring() writes AW_CFG with FLUSH set and everything else, BME included, cleared. Clearing BME halts the device's writes to the host but leaves the staged requests parked in the PUL write pipeline rather than draining them, so on the write path FLUSH alone never terminates the outstanding requests and the flush the firmware waits on never completes. Add the FLUSH_MODE definition and set both bits so the staged writes drain out of the pipeline on their own. BME stays cleared, so nothing lands on the host; it is restored later in fbnic_mbx_init_desc_ring() when the ring is rebuilt, once the outstanding writes are gone. The read path is unaffected. AR_CFG has no equivalent mode bit and AR_FLUSH terminates the outstanding reads by itself, so it is left as is. Both writes remain plain stores rather than read-modify-writes. That is deliberate: the matching write in fbnic_mbx_init_desc_ring() restores BME and the TLP attributes, and clears both flush bits as a side effect. Fixes: 3b12f00ddd08 ("fbnic: Gate AXI read/write enabling on FW mailbox") Signed-off-by: Alexander Duyck Reviewed-by: Simon Horman Link: https://patch.msgid.link/178942022583.7700.11050671998277309744.stgit@ahduyck-xeon-server.home.arpa Signed-off-by: Jakub Kicinski --- drivers/net/ethernet/meta/fbnic/fbnic_csr.h | 1 + drivers/net/ethernet/meta/fbnic/fbnic_fw.c | 9 ++++++++- 2 files changed, 9 insertions(+), 1 deletion(-) diff --git a/drivers/net/ethernet/meta/fbnic/fbnic_csr.h b/drivers/net/ethernet/meta/fbnic/fbnic_csr.h index 64b958df7774..14af30e189d6 100644 --- a/drivers/net/ethernet/meta/fbnic/fbnic_csr.h +++ b/drivers/net/ethernet/meta/fbnic/fbnic_csr.h @@ -974,6 +974,7 @@ enum { /* PUL User Registers */ #define FBNIC_CSR_START_PUL_USER 0x31000 /* CSR section delimiter */ #define FBNIC_PUL_OB_TLP_HDR_AW_CFG 0x3103d /* 0xc40f4 */ +#define FBNIC_PUL_OB_TLP_HDR_AW_CFG_FLUSH_MODE CSR_BIT(20) #define FBNIC_PUL_OB_TLP_HDR_AW_CFG_FLUSH CSR_BIT(19) #define FBNIC_PUL_OB_TLP_HDR_AW_CFG_BME CSR_BIT(18) #define FBNIC_PUL_OB_TLP_HDR_AW_CFG_RDE_ATTR CSR_GENMASK(17, 15) diff --git a/drivers/net/ethernet/meta/fbnic/fbnic_fw.c b/drivers/net/ethernet/meta/fbnic/fbnic_fw.c index 283d25fae79e..59aa879798b9 100644 --- a/drivers/net/ethernet/meta/fbnic/fbnic_fw.c +++ b/drivers/net/ethernet/meta/fbnic/fbnic_fw.c @@ -60,8 +60,15 @@ static void fbnic_mbx_reset_desc_ring(struct fbnic_dev *fbd, int mbx_idx) */ switch (mbx_idx) { case FBNIC_IPC_MBX_RX_IDX: + /* Clearing BME blocks the device from writing to the host + * but leaves the requests parked in the write pipeline. The + * write path only clears outstanding requests when both FLUSH + * and FLUSH_MODE are set; FLUSH_MODE lets them drain without + * landing on the host. + */ wr32(fbd, FBNIC_PUL_OB_TLP_HDR_AW_CFG, - FBNIC_PUL_OB_TLP_HDR_AW_CFG_FLUSH); + FBNIC_PUL_OB_TLP_HDR_AW_CFG_FLUSH | + FBNIC_PUL_OB_TLP_HDR_AW_CFG_FLUSH_MODE); break; case FBNIC_IPC_MBX_TX_IDX: wr32(fbd, FBNIC_PUL_OB_TLP_HDR_AR_CFG, From 1b97a269a5bdde20d4e69511f27649c9cb82b7c7 Mon Sep 17 00:00:00 2001 From: Alexander Duyck Date: Mon, 14 Sep 2026 14:10:33 -0700 Subject: [PATCH 402/592] eth: fbnic: Handle FW mailbox completions flagged with an error The firmware can complete a mailbox descriptor while also setting FW_ERR to indicate it could not process the request, for example on a mailbox DMA error. The completion carries no valid data. The driver did not check FW_ERR. On the Rx mailbox it would sync and parse the stale page as a normal message, and on the Tx mailbox it silently freed the request. If the initial capabilities exchange in fbnic_mbx_poll_tx_ready() hit FW_ERR -- on the Tx request or on the Rx response descriptor -- no response was parsed and the poll spun until it timed out even though the ring was healthy. Check FW_ERR on both mailboxes. Count it per-mailbox in fbnic_fw_mbx.resp_error, which is also shown in debugfs, warn (rate limited, since the bit is firmware controlled), and drop the Rx page instead of parsing it. In fbnic_mbx_poll_tx_ready() re-issue the capabilities request when either the Tx or the Rx resp_error counter advances, so a FW_ERR on the request or on its response triggers a retry rather than a timeout. A valid capabilities response is honored before the retry check, so a response parsed in the same poll as an unrelated FW_ERR is not discarded. The counters are mailbox-wide rather than keyed to the capabilities request; that is sufficient here because the exchange runs during bring-up before any other mailbox traffic, and any spurious retry is bounded by the existing 10s timeout. Fixes: da3cde08209e ("eth: fbnic: Add FW communication mechanism") Signed-off-by: Alexander Duyck Reviewed-by: Simon Horman Link: https://patch.msgid.link/178942023343.7700.9423398932961964439.stgit@ahduyck-xeon-server.home.arpa Signed-off-by: Jakub Kicinski --- drivers/net/ethernet/meta/fbnic/fbnic_csr.h | 4 ++ .../net/ethernet/meta/fbnic/fbnic_debugfs.c | 4 +- drivers/net/ethernet/meta/fbnic/fbnic_fw.c | 38 ++++++++++++++++++- drivers/net/ethernet/meta/fbnic/fbnic_fw.h | 1 + 4 files changed, 44 insertions(+), 3 deletions(-) diff --git a/drivers/net/ethernet/meta/fbnic/fbnic_csr.h b/drivers/net/ethernet/meta/fbnic/fbnic_csr.h index 14af30e189d6..baba3471bf5a 100644 --- a/drivers/net/ethernet/meta/fbnic/fbnic_csr.h +++ b/drivers/net/ethernet/meta/fbnic/fbnic_csr.h @@ -1216,6 +1216,10 @@ enum { #define FBNIC_IPC_MBX_DESC_LEN_MASK DESC_GENMASK(63, 48) #define FBNIC_IPC_MBX_DESC_EOM DESC_BIT(46) #define FBNIC_IPC_MBX_DESC_ADDR_MASK DESC_GENMASK(45, 3) +/* Set with FW_CMPL when the FW completed a descriptor without successfully + * processing it (e.g. a mailbox DMA error); the completion has no valid data. + */ +#define FBNIC_IPC_MBX_DESC_FW_ERR DESC_BIT(2) #define FBNIC_IPC_MBX_DESC_FW_CMPL DESC_BIT(1) #define FBNIC_IPC_MBX_DESC_HOST_CMPL DESC_BIT(0) diff --git a/drivers/net/ethernet/meta/fbnic/fbnic_debugfs.c b/drivers/net/ethernet/meta/fbnic/fbnic_debugfs.c index 3c4563c8f403..6edfa0aa69f1 100644 --- a/drivers/net/ethernet/meta/fbnic/fbnic_debugfs.c +++ b/drivers/net/ethernet/meta/fbnic/fbnic_debugfs.c @@ -539,8 +539,8 @@ static void fbnic_dbg_fw_mbx_display(struct seq_file *s, /* Generate header */ seq_puts(s, mbx_idx == FBNIC_IPC_MBX_RX_IDX ? "Rx\n" : "Tx\n"); - seq_printf(s, "Rdy: %d Head: %d Tail: %d\n", - mbx->ready, mbx->head, mbx->tail); + seq_printf(s, "Rdy: %d Head: %d Tail: %d resp_error: %llu\n", + mbx->ready, mbx->head, mbx->tail, mbx->resp_error); snprintf(hdr, sizeof(hdr), "%3s %-4s %s %-12s %s %-3s %-16s\n", "Idx", "Len", "E", "Addr", "F", "H", "Raw"); diff --git a/drivers/net/ethernet/meta/fbnic/fbnic_fw.c b/drivers/net/ethernet/meta/fbnic/fbnic_fw.c index 59aa879798b9..6d7eb8479edf 100644 --- a/drivers/net/ethernet/meta/fbnic/fbnic_fw.c +++ b/drivers/net/ethernet/meta/fbnic/fbnic_fw.c @@ -292,6 +292,12 @@ static void fbnic_mbx_process_tx_msgs(struct fbnic_dev *fbd) if (!(desc & FBNIC_IPC_MBX_DESC_FW_CMPL)) break; + if (desc & FBNIC_IPC_MBX_DESC_FW_ERR) { + tx_mbx->resp_error++; + dev_warn_ratelimited(fbd->dev, + "FW completed a Tx mailbox request with an error\n"); + } + fbnic_mbx_unmap_and_free_msg(fbd, FBNIC_IPC_MBX_TX_IDX, head); head++; @@ -1673,6 +1679,13 @@ static void fbnic_mbx_process_rx_msgs(struct fbnic_dev *fbd) if (!(desc & FBNIC_IPC_MBX_DESC_FW_CMPL)) break; + if (desc & FBNIC_IPC_MBX_DESC_FW_ERR) { + rx_mbx->resp_error++; + dev_warn_ratelimited(fbd->dev, + "FW reported an error on an Rx mailbox message; dropping\n"); + goto next_page; + } + dma_sync_single_for_cpu(fbd->dev, rx_mbx->buf_info[head].addr, FBNIC_RX_PAGE_SIZE, DMA_FROM_DEVICE); @@ -1740,7 +1753,9 @@ void fbnic_mbx_poll(struct fbnic_dev *fbd) int fbnic_mbx_poll_tx_ready(struct fbnic_dev *fbd) { struct fbnic_fw_mbx *tx_mbx = &fbd->mbx[FBNIC_IPC_MBX_TX_IDX]; + struct fbnic_fw_mbx *rx_mbx = &fbd->mbx[FBNIC_IPC_MBX_RX_IDX]; unsigned long timeout = jiffies + 10 * HZ + 1; + u64 tx_resp_error, rx_resp_error; int err, i; do { @@ -1771,6 +1786,9 @@ int fbnic_mbx_poll_tx_ready(struct fbnic_dev *fbd) * mgmt.version once we get the actual version from the firmware * in the capabilities request message. */ +send_cap_req: + tx_resp_error = tx_mbx->resp_error; + rx_resp_error = rx_mbx->resp_error; err = fbnic_fw_xmit_simple_msg(fbd, FBNIC_TLV_MSG_ID_HOST_CAP_REQ); if (err) goto clean_mbx; @@ -1788,9 +1806,27 @@ int fbnic_mbx_poll_tx_ready(struct fbnic_dev *fbd) msleep(20); fbnic_mbx_poll(fbd); + /* A valid capabilities response ends the poll. Check it + * before the FW_ERR retry below so a response parsed in the + * same poll as an unrelated FW_ERR is not discarded. + */ + if (fbd->fw_cap.running.mgmt.version >= MIN_FW_VER_CODE) + break; + /* set err, but wait till mgmt.version check to report it */ - if (!time_is_after_jiffies(timeout)) + if (!time_is_after_jiffies(timeout)) { err = -ETIMEDOUT; + continue; + } + + /* The FW can flag our capabilities request (Tx) or its + * response (Rx) with FW_ERR, in which case it produced no + * usable response. The ring is not wedged, so re-issue the + * request instead of spinning until the timeout. + */ + if (tx_mbx->resp_error != tx_resp_error || + rx_mbx->resp_error != rx_resp_error) + goto send_cap_req; } return 0; diff --git a/drivers/net/ethernet/meta/fbnic/fbnic_fw.h b/drivers/net/ethernet/meta/fbnic/fbnic_fw.h index d84723e4cfa3..5f9969247e30 100644 --- a/drivers/net/ethernet/meta/fbnic/fbnic_fw.h +++ b/drivers/net/ethernet/meta/fbnic/fbnic_fw.h @@ -13,6 +13,7 @@ struct fbnic_tlv_msg; struct fbnic_fw_mbx { u8 ready, head, tail; + u64 resp_error; struct { struct fbnic_tlv_msg *msg; dma_addr_t addr; From d2c31b837406395e576afeb25958c98e9938f3f6 Mon Sep 17 00:00:00 2001 From: Yiqi Sun Date: Tue, 15 Sep 2026 17:50:17 +0800 Subject: [PATCH 403/592] sctp: avoid livelock while updating retransmit path sctp_assoc_update_retran_path() can loop forever when every remaining transport, including retran_path, is SCTP_UNCONFIRMED: the state check runs before the wraparound test, so the loop cannot observe that it has completed a full pass. Fix this by considering a transport only when it is not UNCONFIRMED, then checking whether the walk has returned to retran_path. This makes the full-pass termination independent of the transport state while preserving the existing fallback selection semantics. Also restore the NULL guard around the retran_path assignment. In the all-UNCONFIRMED case there is no eligible replacement transport, and installing NULL would leave later retransmit-path users and the debug print with a NULL path. Fixes: 4c47af4d5eb2 ("net: sctp: rework multihoming retransmission path selection to rfc4960") Signed-off-by: Yiqi Sun Acked-by: Xin Long Link: https://patch.msgid.link/20260915095017.942213-1-sunyiqixm@gmail.com Signed-off-by: Jakub Kicinski --- net/sctp/associola.c | 15 ++++++++------- 1 file changed, 8 insertions(+), 7 deletions(-) diff --git a/net/sctp/associola.c b/net/sctp/associola.c index c0512c827d0f..4521be3bd85a 100644 --- a/net/sctp/associola.c +++ b/net/sctp/associola.c @@ -1289,18 +1289,19 @@ void sctp_assoc_update_retran_path(struct sctp_association *asoc) /* Manually skip the head element. */ if (&trans->transports == &asoc->peer.transport_addr_list) continue; - if (trans->state == SCTP_UNCONFIRMED) - continue; - trans_next = sctp_trans_elect_best(trans, trans_next); - /* Active is good enough for immediate return. */ - if (trans_next->state == SCTP_ACTIVE) - break; + if (trans->state != SCTP_UNCONFIRMED) { + trans_next = sctp_trans_elect_best(trans, trans_next); + /* Active is good enough for immediate return. */ + if (trans_next->state == SCTP_ACTIVE) + break; + } /* We've reached the end, time to update path. */ if (trans == asoc->peer.retran_path) break; } - asoc->peer.retran_path = trans_next; + if (trans_next) + asoc->peer.retran_path = trans_next; pr_debug("%s: association:%p updated new path to addr:%pISpc\n", __func__, asoc, &asoc->peer.retran_path->ipaddr.sa); From 4581c3d2adc3c73a019bc38db64ca11f28bbd7fd Mon Sep 17 00:00:00 2001 From: Ralf Lici Date: Thu, 17 Sep 2026 14:27:23 +0200 Subject: [PATCH 404/592] net/mlx5e: advertise MACsec offload only when supported Commit 339ccec8d43d ("net/mlx5: Enable MACsec offload feature for VLAN interface") added NETIF_F_HW_MACSEC unconditionally to vlan_features so that VLAN devices could inherit MACsec offload support. mlx5e_build_nic_netdev subsequently copies vlan_features into hw_features and features. As a result, all mlx5e NIC netdevices advertise MACsec hardware offload, even when the firmware does not support it and the driver does not install macsec_ops. Set the MACsec feature bits in mlx5e_macsec_build_netdev, after device capabilities have been validated. This preserves MACsec-over-VLAN support and the ethtool feature control on capable devices, without advertising either on unsupported hardware. Fixes: 339ccec8d43d ("net/mlx5: Enable MACsec offload feature for VLAN interface") Cc: stable@vger.kernel.org Reviewed-by: Tariq Toukan Signed-off-by: Ralf Lici Link: https://patch.msgid.link/20260917122724.654639-1-ralf@mandelbit.com Signed-off-by: Jakub Kicinski --- drivers/net/ethernet/mellanox/mlx5/core/en_accel/macsec.c | 2 ++ drivers/net/ethernet/mellanox/mlx5/core/en_main.c | 1 - 2 files changed, 2 insertions(+), 1 deletion(-) diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en_accel/macsec.c b/drivers/net/ethernet/mellanox/mlx5/core/en_accel/macsec.c index daff53ba7d09..38a3415acf7a 100644 --- a/drivers/net/ethernet/mellanox/mlx5/core/en_accel/macsec.c +++ b/drivers/net/ethernet/mellanox/mlx5/core/en_accel/macsec.c @@ -1724,6 +1724,8 @@ void mlx5e_macsec_build_netdev(struct mlx5e_priv *priv) mlx5_core_dbg(priv->mdev, "mlx5e: MACsec acceleration enabled\n"); netdev->macsec_ops = &macsec_offload_ops; netdev->features |= NETIF_F_HW_MACSEC; + netdev->hw_features |= NETIF_F_HW_MACSEC; + netdev->vlan_features |= NETIF_F_HW_MACSEC; netif_keep_dst(netdev); } diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en_main.c b/drivers/net/ethernet/mellanox/mlx5/core/en_main.c index fc110a7d16e8..4c6060d54fbe 100644 --- a/drivers/net/ethernet/mellanox/mlx5/core/en_main.c +++ b/drivers/net/ethernet/mellanox/mlx5/core/en_main.c @@ -5851,7 +5851,6 @@ static void mlx5e_build_nic_netdev(struct net_device *netdev) netdev->vlan_features |= NETIF_F_SG; netdev->vlan_features |= NETIF_F_HW_CSUM; - netdev->vlan_features |= NETIF_F_HW_MACSEC; netdev->vlan_features |= NETIF_F_GRO; netdev->vlan_features |= NETIF_F_TSO; netdev->vlan_features |= NETIF_F_TSO6; From 6c096bb08de97cdca051fecddad22cac6a1fd275 Mon Sep 17 00:00:00 2001 From: Quentin Armitage Date: Tue, 15 Sep 2026 22:33:21 +0100 Subject: [PATCH 405/592] net: allow IFLA_INET_CONF messages when NLA_F_NESTED unset Commit fa8fca88714c ("ipv4: validate IPV4_DEVCONF attributes properly") added validation of IFLA_INET_CONF attributes, and in the process changed the call of nla_for_each_nested() to nla_parse_nested(). A side effect of this change is that the IFLA_INET_CONF option is now tested for NLA_F_NESTED being set, and fails if it is not. Prior to the commit there was no check of NLA_F_NESTED. Change nla_parse_nested() to nla_parse(). This restores the previous functionality of not checking NLA_F_NESTED, thereby allowing code that (incorrectly) doesn't set NLA_F_NESTED to continue to work. This issue was identified because keepalived started logging errors when it was configuring macvlans that it created. Fixes: fa8fca88714c ("ipv4: validate IPV4_DEVCONF attributes properly") Signed-off-by: Quentin Armitage Reviewed-by: Ido Schimmel Link: https://patch.msgid.link/20260915213320.1527029-2-quentin@armitage.org.uk Signed-off-by: Jakub Kicinski --- net/ipv4/devinet.c | 7 ++++--- 1 file changed, 4 insertions(+), 3 deletions(-) diff --git a/net/ipv4/devinet.c b/net/ipv4/devinet.c index a90be57c63be..5b6b11c943e4 100644 --- a/net/ipv4/devinet.c +++ b/net/ipv4/devinet.c @@ -2117,9 +2117,10 @@ static int inet_validate_link_af(const struct net_device *dev, return err; if (tb[IFLA_INET_CONF]) { - err = nla_parse_nested(nested_tb, IPV4_DEVCONF_MAX, - tb[IFLA_INET_CONF], inet_devconf_policy, - extack); + err = nla_parse(nested_tb, IPV4_DEVCONF_MAX, + nla_data(tb[IFLA_INET_CONF]), + nla_len(tb[IFLA_INET_CONF]), + inet_devconf_policy, extack); if (err < 0) return err; From 10396a2d6d41d594975b6ece712278570c3c970c Mon Sep 17 00:00:00 2001 From: Holger Dengler Date: Fri, 14 Aug 2026 16:20:10 +0200 Subject: [PATCH 406/592] crypto: s390/hmac - Generate intermediate CV for API partial block handling The API partial block handling requires a intermediate chaining value (CV). The internal function hash_data() sets the function code correctly, so also call cpacf_kimd() instruction for intermediate CV generation, as cpacf_klmd() always generate the final hash value. Cc: stable@vger.kernel.org # 6.15+ Fixes: 08811169ac01 ("crypto: s390/hmac - Use API partial block handling") Signed-off-by: Holger Dengler Reviewed-by: Harald Freudenberger Signed-off-by: Herbert Xu --- arch/s390/crypto/hmac_s390.c | 5 ++++- 1 file changed, 4 insertions(+), 1 deletion(-) diff --git a/arch/s390/crypto/hmac_s390.c b/arch/s390/crypto/hmac_s390.c index f8cd09f341d4..445fa7bbd958 100644 --- a/arch/s390/crypto/hmac_s390.c +++ b/arch/s390/crypto/hmac_s390.c @@ -150,7 +150,10 @@ static int hash_data(const u8 *in, unsigned int inlen, #undef PARAM_INIT - cpacf_klmd(func, ¶m, in, inlen); + if (final) + cpacf_klmd(func, ¶m, in, inlen); + else + cpacf_kimd(func, ¶m, in, inlen); memcpy(digest, ¶m, digestsize); From 63edf5a009ae366369a1b484cd9ae4ee7c51946c Mon Sep 17 00:00:00 2001 From: "Chang S. Bae" Date: Wed, 16 Sep 2026 23:00:03 +0000 Subject: [PATCH 407/592] x86/build/64: Prevent native builds from generating EGPR use Omar reports that CONFIG_X86_NATIVE_CPU=y allows builds to opportunistically emit instructions using %r16-%r31 (EGPRs) when the build host supports APX since the commit: ea1dcca1de12 ("x86/kbuild/64: Add the CONFIG_X86_NATIVE_CPU option to locally optimize the kernel with '-march=native'") But the kernel is not yet prepared to use new registers internally. For example, there is no context-switch support for general in-kernel use. Explicitly disable EGPR use when building with -march=native. For C, since GCC 14 and Clang 18, both compilers support suppressing EGPR use with -mno-apx-features=egpr, whose availability can be detected via cc-option. For Rust, pass features=-apxf through the generated JSON to avoid unstable-feature warnings, see https://github.com/rust-lang/rust/issues/139284 Note Rust only accepts the option to disable APX instructions entirely or not. Support for this gating also depends on the Rust/LLVM combination. Rust 1.88 introduced the `apxf` feature option, but versions prior to 1.93 may emit an `apxf` attribute to the backend that only LLVM 23 or later can interpret. Restrict native Rust builds accordingly. Fixes: ea1dcca1de12 ("x86/kbuild/64: Add the CONFIG_X86_NATIVE_CPU option to locally optimize the kernel with '-march=native'") Reported-by: Omar Avelar Signed-off-by: Chang S. Bae Signed-off-by: Borislav Petkov (AMD) Reviewed-by: Nathan Chancellor Acked-by: Miguel Ojeda Cc: stable@vger.kernel.org Link: https://patch.msgid.link/20260916230003.1144622-1-chang.seok.bae@intel.com --- arch/x86/Kconfig.cpu | 11 +++++++++++ arch/x86/Makefile | 5 +++++ scripts/generate_rust_target.rs | 5 +++++ 3 files changed, 21 insertions(+) diff --git a/arch/x86/Kconfig.cpu b/arch/x86/Kconfig.cpu index e4654388d794..6e7a366f0798 100644 --- a/arch/x86/Kconfig.cpu +++ b/arch/x86/Kconfig.cpu @@ -204,10 +204,21 @@ config CC_HAS_MARCH_NATIVE # usage warnings that only appear wth '-march=native'. depends on CC_IS_GCC || CLANG_VERSION >= 190100 +config RUSTC_HAS_APXF + # The kernel isn't ready for in-kernel APX instructions. Without + # explicit frontend gating of APX, the backend may emit those + # instructions in native builds. + # + # Rust 1.88 added the `apxf` feature option, but versions before 1.93 + # emit an `apxf` target attribute that only LLVM 23+ can interpret. + def_bool (RUSTC_VERSION >= 108800 && RUSTC_LLVM_MAJOR_VERSION >= 23) || \ + RUSTC_VERSION >= 109300 + config X86_NATIVE_CPU bool "Build and optimize for local/native CPU" depends on X86_64 depends on CC_HAS_MARCH_NATIVE + depends on !RUST || RUSTC_HAS_APXF help Optimize for the current CPU used to compile the kernel. Use this option if you intend to build the kernel for your diff --git a/arch/x86/Makefile b/arch/x86/Makefile index 598f178102ee..8af6b80cffdd 100644 --- a/arch/x86/Makefile +++ b/arch/x86/Makefile @@ -161,6 +161,11 @@ else ifdef CONFIG_X86_NATIVE_CPU KBUILD_CFLAGS += -march=native + # Prevent the compiler from generating EGPR use. The kernel is + # not yet prepared for general in-kernel use. + KBUILD_CFLAGS += $(call cc-option,-mno-apx-features=egpr) + + # generate_rust_target.rs handles Rust APX gating. KBUILD_RUSTFLAGS += -Ctarget-cpu=native else KBUILD_CFLAGS += -march=x86-64 -mtune=generic diff --git a/scripts/generate_rust_target.rs b/scripts/generate_rust_target.rs index 3bf296581a88..7687b0dd5474 100644 --- a/scripts/generate_rust_target.rs +++ b/scripts/generate_rust_target.rs @@ -224,6 +224,11 @@ fn main() { features += ",+harden-sls-ijmp"; features += ",+harden-sls-ret"; } + if cfg.has("X86_NATIVE_CPU") { + // Prevent the backend from generating APX instructions. The kernel is not yet prepared + // for general in-kernel EGPR use. + features += ",-apxf"; + } ts.push("features", features); ts.push("llvm-target", "x86_64-linux-gnu"); ts.push("supported-sanitizers", ["kcfi", "kernel-address"]); From 8901cee9316d53a5a97398f026c2d59a3043159d Mon Sep 17 00:00:00 2001 From: Kumar Kartikeya Dwivedi Date: Sat, 19 Sep 2026 03:42:10 +0200 Subject: [PATCH 408/592] bpf: Compare stack frames in regs_exact() regs_exact() compares the register state up to id, followed by the ID mappings, but does not compare frameno. The PTR_TO_STACK case in regsafe() checks frameno separately, which is bypassed when exact comparison is requested. Consequently, infinite-loop detection can treat pointers to different stack frames as the same pointer and reject a finite loop. For example, initialize fp-8 to zero in the caller and to one in the callee, then pass the caller's fp-8 to the callee as r1: loop: r0 = *(u64 *)(r1 + 0); if r0 != 0 goto done; r1 = r10; r1 += -8; goto loop; done: exit; The loop terminates after reading the callee's slot on its second iteration. At the loop header, however, the only relevant difference is r1's frameno, so exact comparison incorrectly reports an infinite loop. The same problem occurs when the pointer is spilled to the stack. Move frameno into the type-specific metadata union, ahead of id, so the existing prefix comparison in regs_exact() covers it. Ordinary stack pointers do not use another union member. Iterator and IRQ stack-slot states use their dedicated union views and do not need a frame lookup. This also keeps bpf_reg_state at 80 bytes. Since frameno now shares storage with other pointer metadata, it is only meaningful for PTR_TO_STACK registers. Return NULL from bpf_func() for other register types. process_iter_arg(), get_constant_map_key() and is_dynptr_reg_valid_init() look up the frame before checking the register type and would otherwise index frame[] with a byte of the register's map or BTF pointer. They dereference the frame only after their type check. Move the states_maybe_looping() boundary from frameno to precise after the field relocation. Its prefix comparison continues to cover the complete value state and now includes frameno. Continue to ignore precise. Precision marks control whether pruning may ignore scalar ranges; they do not change the represented values, and exact comparison already compares those ranges unconditionally. Marks can also change through backtracking while an ancestor state is still being explored. Fixes: d5b892fd607a ("bpf: make infinite loop detection in is_state_visited() exact") Reported-by: Eduard Zingerman Signed-off-by: Kumar Kartikeya Dwivedi Signed-off-by: Alexei Starovoitov Acked-by: Eduard Zingerman Link: https://patch.msgid.link/20260919014213.1840880-2-memxor@gmail.com --- include/linux/bpf_verifier.h | 23 +++++++++++++++-------- kernel/bpf/states.c | 7 ++----- 2 files changed, 17 insertions(+), 13 deletions(-) diff --git a/include/linux/bpf_verifier.h b/include/linux/bpf_verifier.h index a7202b44ab10..6fe8e5dc57aa 100644 --- a/include/linux/bpf_verifier.h +++ b/include/linux/bpf_verifier.h @@ -45,6 +45,14 @@ struct bpf_reg_state { union { /* valid when type == PTR_TO_PACKET */ int range; + /* + * Valid when type == PTR_TO_STACK. Inside the callee two registers + * can be both PTR_TO_STACK like R1=fp-8 and R2=fp-8, but one of them + * points to this function stack while another to the caller's stack. + * To differentiate them 'frameno' is used which is an index in + * bpf_verifier_state->frame[] array pointing to bpf_func_state. + */ + u8 frameno; /* * For CONST_PTR_TO_MAP, PTR_TO_MAP_KEY, PTR_TO_MAP_VALUE and @@ -155,14 +163,6 @@ struct bpf_reg_state { * during state comparisons. */ u32 map_uid; - /* - * Inside the callee two registers can be both PTR_TO_STACK like - * R1=fp-8 and R2=fp-8, but one of them points to this function stack - * while another to the caller's stack. To differentiate them 'frameno' - * is used which is an index in bpf_verifier_state->frame[] array - * pointing to bpf_func_state. - */ - u8 frameno; /* if (!precise && SCALAR_VALUE) min/max/tnum don't affect safety */ bool precise; }; @@ -1239,11 +1239,18 @@ static inline int bpf_get_spi(s32 off) return (-off - 1) / BPF_REG_SIZE; } +/* + * Return the function state a stack pointer register refers to. frameno + * shares storage with other pointer metadata, so return NULL for any + * other register type instead of indexing frame[] with aliased bytes. + */ static inline struct bpf_func_state *bpf_func(struct bpf_verifier_env *env, const struct bpf_reg_state *reg) { struct bpf_verifier_state *cur = env->cur_state; + if (reg->type != PTR_TO_STACK) + return NULL; return cur->frame[reg->frameno]; } diff --git a/kernel/bpf/states.c b/kernel/bpf/states.c index e4ec007f7fa6..012b82513a3b 100644 --- a/kernel/bpf/states.c +++ b/kernel/bpf/states.c @@ -644,10 +644,7 @@ static bool regsafe(struct bpf_verifier_env *env, struct bpf_reg_state *rold, return range_within(rold, rcur) && tnum_in(rold->var_off, rcur->var_off); case PTR_TO_STACK: - /* two stack pointers are equal only if they're pointing to - * the same stack frame, since fp-8 in foo != fp-8 in bar - */ - return regs_exact(rold, rcur, idmap) && rold->frameno == rcur->frameno; + return regs_exact(rold, rcur, idmap); case PTR_TO_ARENA: return true; case PTR_TO_INSN: @@ -1126,7 +1123,7 @@ static bool states_maybe_looping(struct bpf_verifier_state *old, fcur = cur->frame[fr]; for (i = 0; i < MAX_BPF_REG; i++) if (memcmp(&fold->regs[i], &fcur->regs[i], - offsetof(struct bpf_reg_state, frameno))) + offsetof(struct bpf_reg_state, precise))) return false; return true; } From 070587848985efcfcd45104c5266ba76a1165f55 Mon Sep 17 00:00:00 2001 From: Kumar Kartikeya Dwivedi Date: Sat, 19 Sep 2026 03:42:11 +0200 Subject: [PATCH 409/592] selftests/bpf: Cover frame changes in bounded loops Add a finite loop whose progress is represented only by changing the frame number of a stack pointer. The loop first reads zero from the caller's stack, switches to the same offset in the callee's stack, and exits after reading one on its next iteration. Force frequent checkpoints so the test exercises infinite-loop detection, and check that the program returns one when run. Without the frameno comparison in regs_exact(), the program is rejected with an "infinite loop detected" diagnostic instead of loading successfully. Signed-off-by: Kumar Kartikeya Dwivedi Signed-off-by: Alexei Starovoitov Tested-by: Eduard Zingerman Link: https://patch.msgid.link/20260919014213.1840880-3-memxor@gmail.com --- .../selftests/bpf/progs/verifier_loops1.c | 36 +++++++++++++++++++ 1 file changed, 36 insertions(+) diff --git a/tools/testing/selftests/bpf/progs/verifier_loops1.c b/tools/testing/selftests/bpf/progs/verifier_loops1.c index d248ce877f14..48a966cda199 100644 --- a/tools/testing/selftests/bpf/progs/verifier_loops1.c +++ b/tools/testing/selftests/bpf/progs/verifier_loops1.c @@ -303,4 +303,40 @@ __naked void maybe_exit_scc_bug1(void) ::: __clobber_all); } +/* + * The loop reads zero from the caller's stack on its first iteration and + * one from the callee's stack on its second iteration. At the loop header, + * only the frame number of the pointer in r1 changes. + */ +static __naked __noinline __used +void loop_stack_frames_reg(void) +{ + asm volatile ( + "*(u64 *)(r10 - 8) = 1;" +"1:" + "r0 = *(u64 *)(r1 + 0);" + "if r0 != 0 goto 2f;" + "r1 = r10;" + "r1 += -8;" + "goto 1b;" +"2:" + "exit;" + ::: __clobber_all); +} + +SEC("xdp") +__description("bounded loop changing stack frame in a register") +__success __retval(1) +__flag(BPF_F_TEST_STATE_FREQ) +__naked void bounded_loop_stack_frames_reg(void) +{ + asm volatile ( + "*(u64 *)(r10 - 8) = 0;" + "r1 = r10;" + "r1 += -8;" + "call loop_stack_frames_reg;" + "exit;" + ::: __clobber_all); +} + char _license[] SEC("license") = "GPL"; From bfc888f04588f591851e95c974954cfca58e6c19 Mon Sep 17 00:00:00 2001 From: Kumar Kartikeya Dwivedi Date: Mon, 14 Sep 2026 15:24:42 +0200 Subject: [PATCH 410/592] bpf: Bound ownership depth through local kptrs and graph roots Program-allocated objects can own other local objects through referenced kptrs. bpf_obj_free_fields() follows those pointers through __bpf_obj_drop_impl() synchronously, before the object storage is freed through RCU. A self-referential local kptr type therefore permits arbitrarily deep object chains, and dropping the head can exhaust the kernel stack. Long acyclic type chains have the same problem. btf_check_and_fixup_fields() still assumes referenced kptrs only point to kernel types and checks ownership through list and rbtree roots only. Its existing rule is sufficient for graph-only cycles: the target of each graph edge must contain a node, so every type in a cycle has both a root and a node. The rule rejects such a type owning another root, breaking every cycle. It also limits graph-only chains to three types, or two if the first type contains a node, and conservatively rejects longer acyclic chains. The missing local-kptr edges, rather than a missed graph-only cycle, are the bug introduced by support for bpf_kptr_xchg() into local kptrs. Replace that restriction with one bounded ownership walk covering graph roots and local referenced kptrs. Run it after all BTF records have been fixed up, reject cycles and paths deeper than eight record-bearing types, and cache each type's suffix depth while checking it against the remaining budget. This also permits the longer acyclic graph-only layouts rejected by the old rule; update their existing BTF tests accordingly. Keep the bound independent of MAX_CALL_FRAMES because recursive destruction can run below a BPF call chain. A plain local pointee without special-field metadata adds only a final non-recursing drop. Non-owning kptrs and kernel-BTF kptrs do not recurse through local records and remain outside the walk. Include local percpu-kptr edges too, although allocation of percpu objects with special fields is currently forbidden, so that relaxing that restriction cannot bypass the ownership bound. btf_check_and_fixup_fields() continues to initialize graph_root.value_rec, including for separately allocated map records. The ownership relationships belong to immutable program BTF and only need validation at BTF load time. Fixes: b0966c724584 ("bpf: Support bpf_kptr_xchg into local kptr") Reported-by: Nicholas Carlini Suggested-by: Nicholas Carlini Signed-off-by: Kumar Kartikeya Dwivedi Signed-off-by: Alexei Starovoitov Link: https://patch.msgid.link/20260914132444.2564218-2-memxor@gmail.com --- kernel/bpf/btf.c | 138 +++++++++++------- .../selftests/bpf/prog_tests/linked_list.c | 4 +- 2 files changed, 90 insertions(+), 52 deletions(-) diff --git a/kernel/bpf/btf.c b/kernel/bpf/btf.c index d67a169ba2bb..d870bc5e50bc 100644 --- a/kernel/bpf/btf.c +++ b/kernel/bpf/btf.c @@ -4268,13 +4268,10 @@ int btf_check_and_fixup_fields(const struct btf *btf, struct btf_record *rec) { int i; - /* There are three types that signify ownership of some other type: - * kptr_ref, bpf_list_head, bpf_rb_root. - * kptr_ref only supports storing kernel types, which can't store - * references to program allocated local types. - * - * Hence we only need to ensure that bpf_{list_head,rb_root} ownership - * does not form cycles. + /* + * Check fields which require the complete BTF and initialize runtime + * metadata. Ownership relationships are validated after every record has + * been fixed up. */ if (IS_ERR_OR_NULL(rec) || !(rec->field_mask & (BPF_GRAPH_ROOT | BPF_UPTR))) return 0; @@ -4305,53 +4302,90 @@ int btf_check_and_fixup_fields(const struct btf *btf, struct btf_record *rec) if (!meta) return -EFAULT; rec->fields[i].graph_root.value_rec = meta->record; - - /* We need to set value_rec for all root types, but no need - * to check ownership cycle for a type unless it's also a - * node type. - */ - if (!(rec->field_mask & BPF_GRAPH_NODE)) - continue; - - /* We need to ensure ownership acyclicity among all types. The - * proper way to do it would be to topologically sort all BTF - * IDs based on the ownership edges, since there can be multiple - * bpf_{list_head,rb_node} in a type. Instead, we use the - * following resaoning: - * - * - A type can only be owned by another type in user BTF if it - * has a bpf_{list,rb}_node. Let's call these node types. - * - A type can only _own_ another type in user BTF if it has a - * bpf_{list_head,rb_root}. Let's call these root types. - * - * We ensure that if a type is both a root and node, its - * element types cannot be root types. - * - * To ensure acyclicity: - * - * When A is an root type but not a node, its ownership - * chain can be: - * A -> B -> C - * Where: - * - A is an root, e.g. has bpf_rb_root. - * - B is both a root and node, e.g. has bpf_rb_node and - * bpf_list_head. - * - C is only an root, e.g. has bpf_list_node - * - * When A is both a root and node, some other type already - * owns it in the BTF domain, hence it can not own - * another root type through any of the ownership edges. - * A -> B - * Where: - * - A is both an root and node. - * - B is only an node. - */ - if (meta->record->field_mask & BPF_GRAPH_ROOT) - return -ELOOP; } return 0; } +static int btf_owned_type_idx(const struct btf *btf, struct btf_struct_metas *tab, + const struct btf_field *field) +{ + struct btf_struct_meta *meta; + u32 btf_id; + + if (field->type & BPF_GRAPH_ROOT) { + btf_id = field->graph_root.value_btf_id; + } else if (field->type == BPF_KPTR_REF || field->type == BPF_KPTR_PERCPU) { + if (btf_is_kernel(field->kptr.btf)) + return -ENOENT; + btf_id = field->kptr.btf_id; + } else { + return -ENOENT; + } + + meta = btf_find_struct_meta(btf, btf_id); + if (!meta) + return field->type & BPF_GRAPH_ROOT ? -EFAULT : -ENOENT; + return meta - tab->types; +} + +/* + * Each ownership edge adds kernel frames through bpf_obj_free_fields() and + * __bpf_obj_drop_impl(). Keep the bound deliberately small because object + * destruction can itself run below a BPF call chain. A final pointee without + * special fields is not present in the struct metadata table and adds only a + * non-recursing drop. + */ +#define BTF_MAX_OWNERSHIP_DEPTH 8 + +static int btf_ownership_depth(const struct btf *btf, + struct btf_struct_metas *tab, u8 *depth, + int idx, int depth_left) +{ + const struct btf_record *rec = tab->types[idx].record; + int i, ret, max_depth = 0; + + if (!depth_left) + return -ELOOP; + if (depth[idx]) + goto done; + + for (i = 0; i < rec->cnt; i++) { + ret = btf_owned_type_idx(btf, tab, &rec->fields[i]); + if (ret == -ENOENT) + continue; + if (ret < 0) + return ret; + ret = btf_ownership_depth(btf, tab, depth, ret, depth_left - 1); + if (ret < 0) + return ret; + max_depth = max(max_depth, ret); + } + depth[idx] = max_depth + 1; +done: + return depth[idx] > depth_left ? -ELOOP : depth[idx]; +} + +static int btf_check_ownership_depth(const struct btf *btf, + struct btf_struct_metas *tab) +{ + u8 *depth; + int i, ret = 0; + + depth = kvcalloc(tab->cnt, sizeof(*depth), GFP_KERNEL | __GFP_NOWARN); + if (!depth) + return -ENOMEM; + + for (i = 0; i < tab->cnt; i++) { + ret = btf_ownership_depth(btf, tab, depth, i, + BTF_MAX_OWNERSHIP_DEPTH); + if (ret < 0) + break; + ret = 0; + } + kvfree(depth); + return ret; +} + static void __btf_struct_show(const struct btf *btf, const struct btf_type *t, u32 type_id, void *data, u8 bits_offset, struct btf_show *show) @@ -6044,6 +6078,10 @@ static struct btf *btf_parse(const union bpf_attr *attr, bpfptr_t uattr, if (err < 0) goto errout_meta; } + + err = btf_check_ownership_depth(btf, struct_meta_tab); + if (err < 0) + goto errout_meta; } err = bpf_log_attr_finalize(attr_log, &env->log); diff --git a/tools/testing/selftests/bpf/prog_tests/linked_list.c b/tools/testing/selftests/bpf/prog_tests/linked_list.c index c3d133c6a00d..52fabbee3dd5 100644 --- a/tools/testing/selftests/bpf/prog_tests/linked_list.c +++ b/tools/testing/selftests/bpf/prog_tests/linked_list.c @@ -714,7 +714,7 @@ static void test_btf(void) break; err = btf__load_into_kernel(btf); - ASSERT_EQ(err, -ELOOP, "check btf"); + ASSERT_EQ(err, 0, "check btf"); btf__free(btf); break; } @@ -773,7 +773,7 @@ static void test_btf(void) break; err = btf__load_into_kernel(btf); - ASSERT_EQ(err, -ELOOP, "check btf"); + ASSERT_EQ(err, 0, "check btf"); btf__free(btf); break; } From 0288ed67482b6e370eb3fa6b07720df435907970 Mon Sep 17 00:00:00 2001 From: Kumar Kartikeya Dwivedi Date: Mon, 14 Sep 2026 15:24:43 +0200 Subject: [PATCH 411/592] selftests/bpf: Check local object ownership depth Build raw program BTF records to exercise local object ownership without constructing a runtime chain deep enough to threaten the kernel stack. Cover referenced-kptr and percpu-kptr self-cycles, a two-type cycle, and a cycle mixing a graph root with a referenced kptr. The existing graph-only check accepts these local-kptr cycles and over-limit kptr chains; the fix rejects them with -ELOOP. Pin the eight-record depth boundary with a terminal plain object. Exercise both parent-first and child-first BTF orders, and a shared suffix reached first through a shorter path. These cases require cached suffix depths to be checked against the remaining depth budget on each path. Check list and rbtree chains of three, four, eight, and nine types. This covers the old graph-only depth boundary and the new explicit bound. The existing linked-list BTF tests still reject pure graph cycles and now accept the longer acyclic layouts previously rejected by the conservative rule. Although bpf_percpu_obj_new() currently rejects types with special fields, require the percpu cycle to fail at BTF load so future support cannot bypass the ownership bound. Keep a positive control for a non-owning kptr, which remains outside the ownership graph. Signed-off-by: Kumar Kartikeya Dwivedi Signed-off-by: Alexei Starovoitov Link: https://patch.msgid.link/20260914132444.2564218-3-memxor@gmail.com --- .../bpf/prog_tests/local_kptr_ownership.c | 296 ++++++++++++++++++ 1 file changed, 296 insertions(+) create mode 100644 tools/testing/selftests/bpf/prog_tests/local_kptr_ownership.c diff --git a/tools/testing/selftests/bpf/prog_tests/local_kptr_ownership.c b/tools/testing/selftests/bpf/prog_tests/local_kptr_ownership.c new file mode 100644 index 000000000000..a487aa68f2ee --- /dev/null +++ b/tools/testing/selftests/bpf/prog_tests/local_kptr_ownership.c @@ -0,0 +1,296 @@ +// SPDX-License-Identifier: GPL-2.0 +/* Copyright (c) 2026 Meta Platforms, Inc. and affiliates. */ + +#include +#include +#include + +#define SPIN_LOCK 2 +#define LIST_HEAD 3 +#define LIST_NODE 4 +/* Keep in sync with BTF_MAX_OWNERSHIP_DEPTH. */ +#define MAX_OWNERSHIP_DEPTH 8 + +static struct btf *init_btf(void) +{ + struct btf *btf; + int id; + + btf = btf__new_empty(); + if (!ASSERT_OK_PTR(btf, "btf__new_empty")) + return NULL; + id = btf__add_int(btf, "int", 4, BTF_INT_SIGNED); + if (!ASSERT_EQ(id, 1, "btf__add_int")) + goto err_out; + id = btf__add_struct(btf, "bpf_spin_lock", 4); + if (!ASSERT_EQ(id, SPIN_LOCK, "btf__add_struct bpf_spin_lock")) + goto err_out; + id = btf__add_struct(btf, "bpf_list_head", 16); + if (!ASSERT_EQ(id, LIST_HEAD, "btf__add_struct bpf_list_head")) + goto err_out; + id = btf__add_struct(btf, "bpf_list_node", 24); + if (!ASSERT_EQ(id, LIST_NODE, "btf__add_struct bpf_list_node")) + goto err_out; + return btf; + +err_out: + btf__free(btf); + return NULL; +} + +static int add_local_kptr(struct btf *btf, int pointee_id, const char *tag) +{ + int id; + + id = btf__add_type_tag(btf, tag, pointee_id); + if (!ASSERT_GT(id, 0, "btf__add_type_tag")) + return id; + id = btf__add_ptr(btf, id); + ASSERT_GT(id, 0, "btf__add_ptr"); + return id; +} + +static void test_self_cycle(const char *tag, int expected_err) +{ + struct btf *btf; + int id, err; + + btf = init_btf(); + if (!ASSERT_OK_PTR(btf, "init_btf")) + return; + id = add_local_kptr(btf, 7, tag); + if (id <= 0) + goto out; + id = btf__add_struct(btf, "self_cycle", 8); + if (!ASSERT_EQ(id, 7, "btf__add_struct self_cycle")) + goto out; + err = btf__add_field(btf, "next", 6, 0, 0); + if (!ASSERT_OK(err, "btf__add_field self_cycle::next")) + goto out; + + err = btf__load_into_kernel(btf); + ASSERT_EQ(err, expected_err, "check btf"); +out: + btf__free(btf); +} + +static void test_aba_cycle(void) +{ + struct btf *btf; + int id, err; + + btf = init_btf(); + if (!ASSERT_OK_PTR(btf, "init_btf")) + return; + id = add_local_kptr(btf, 10, "kptr"); + if (id <= 0) + goto out; + id = add_local_kptr(btf, 9, "kptr"); + if (id <= 0) + goto out; + id = btf__add_struct(btf, "cycle_a", 8); + if (!ASSERT_EQ(id, 9, "btf__add_struct cycle_a")) + goto out; + err = btf__add_field(btf, "b", 6, 0, 0); + if (!ASSERT_OK(err, "btf__add_field cycle_a::b")) + goto out; + id = btf__add_struct(btf, "cycle_b", 8); + if (!ASSERT_EQ(id, 10, "btf__add_struct cycle_b")) + goto out; + err = btf__add_field(btf, "a", 8, 0, 0); + if (!ASSERT_OK(err, "btf__add_field cycle_b::a")) + goto out; + + err = btf__load_into_kernel(btf); + ASSERT_EQ(err, -ELOOP, "check btf"); +out: + btf__free(btf); +} + +static void test_mixed_cycle(void) +{ + struct btf *btf; + int id, err; + + btf = init_btf(); + if (!ASSERT_OK_PTR(btf, "init_btf")) + return; + id = add_local_kptr(btf, 7, "kptr"); + if (id <= 0) + goto out; + id = btf__add_struct(btf, "mixed_owner", 20); + if (!ASSERT_EQ(id, 7, "btf__add_struct mixed_owner")) + goto out; + err = btf__add_field(btf, "root", LIST_HEAD, 0, 0); + if (!ASSERT_OK(err, "btf__add_field mixed_owner::root")) + goto out; + err = btf__add_field(btf, "lock", SPIN_LOCK, 128, 0); + if (!ASSERT_OK(err, "btf__add_field mixed_owner::lock")) + goto out; + id = btf__add_decl_tag(btf, "contains:mixed_node:node", 7, 0); + if (!ASSERT_EQ(id, 8, "btf__add_decl_tag mixed_owner")) + goto out; + id = btf__add_struct(btf, "mixed_node", 32); + if (!ASSERT_EQ(id, 9, "btf__add_struct mixed_node")) + goto out; + err = btf__add_field(btf, "node", LIST_NODE, 0, 0); + if (!ASSERT_OK(err, "btf__add_field mixed_node::node")) + goto out; + err = btf__add_field(btf, "owner", 6, 192, 0); + if (!ASSERT_OK(err, "btf__add_field mixed_node::owner")) + goto out; + + err = btf__load_into_kernel(btf); + ASSERT_EQ(err, -ELOOP, "check btf"); +out: + btf__free(btf); +} + +static void test_acyclic_depth(int depth, bool child_first, bool shared_suffix, int expected_err) +{ + int ptr_id[MAX_OWNERSHIP_DEPTH + 1]; + int first_struct_id; + struct btf *btf; + int id, err, i, n, pointee_id; + + btf = init_btf(); + if (!ASSERT_OK_PTR(btf, "init_btf")) + return; + first_struct_id = 5 + 2 * depth; + for (i = 0; i < depth; i++) { + if (i == depth - 1) + pointee_id = first_struct_id + depth; + else + pointee_id = first_struct_id + (child_first ? depth - 2 - i : i + 1); + ptr_id[i] = add_local_kptr(btf, pointee_id, "kptr"); + if (ptr_id[i] <= 0) + goto out; + } + for (n = 0; n < depth; n++) { + char name[32]; + int offset = 0; + + i = child_first ? depth - 1 - n : n; + snprintf(name, sizeof(name), "owner_%d", i); + id = btf__add_struct(btf, name, shared_suffix && !i ? 16 : 8); + if (!ASSERT_EQ(id, first_struct_id + n, "btf__add_struct owner")) + goto out; + if (shared_suffix && !i) { + /* + * Visit the shared suffix through the shorter path before + * reaching it again with less remaining depth. + */ + err = btf__add_field(btf, "suffix", ptr_id[1], 0, 0); + if (!ASSERT_OK(err, "btf__add_field owner::suffix")) + goto out; + offset = 64; + } + err = btf__add_field(btf, "next", ptr_id[i], offset, 0); + if (!ASSERT_OK(err, "btf__add_field owner::next")) + goto out; + } + id = btf__add_struct(btf, "plain_leaf", 4); + if (!ASSERT_EQ(id, first_struct_id + depth, "btf__add_struct plain_leaf")) + goto out; + + err = btf__load_into_kernel(btf); + ASSERT_EQ(err, expected_err, "check btf"); +out: + btf__free(btf); +} + +static void test_graph_depth(bool rbtree, int depth, int expected_err) +{ + int root_type = LIST_HEAD, node_type = LIST_NODE, node_size = 24; + int id, err, i, lock_off, root_off, size; + struct btf *btf; + + btf = init_btf(); + if (!ASSERT_OK_PTR(btf, "init_btf")) + return; + if (rbtree) { + root_type = btf__add_struct(btf, "bpf_rb_root", 16); + if (!ASSERT_GT(root_type, 0, "btf__add_struct bpf_rb_root")) + goto out; + node_type = btf__add_struct(btf, "bpf_rb_node", 32); + if (!ASSERT_GT(node_type, 0, "btf__add_struct bpf_rb_node")) + goto out; + node_size = 32; + } + + for (i = 0; i < depth; i++) { + char name[32], tag[64]; + + lock_off = i ? node_size : 0; + root_off = lock_off + 8; + size = i == depth - 1 ? node_size : root_off + 16; + snprintf(name, sizeof(name), "graph_owner_%d", i); + id = btf__add_struct(btf, name, size); + if (!ASSERT_GT(id, 0, "btf__add_struct graph_owner")) + goto out; + if (i) { + err = btf__add_field(btf, "node", node_type, 0, 0); + if (!ASSERT_OK(err, "btf__add_field graph_owner::node")) + goto out; + } + if (i == depth - 1) + continue; + err = btf__add_field(btf, "lock", SPIN_LOCK, lock_off * 8, 0); + if (!ASSERT_OK(err, "btf__add_field graph_owner::lock")) + goto out; + err = btf__add_field(btf, "root", root_type, root_off * 8, 0); + if (!ASSERT_OK(err, "btf__add_field graph_owner::root")) + goto out; + snprintf(tag, sizeof(tag), "contains:graph_owner_%d:node", i + 1); + err = btf__add_decl_tag(btf, tag, id, i ? 2 : 1); + if (!ASSERT_GT(err, 0, "btf__add_decl_tag graph_owner")) + goto out; + } + + err = btf__load_into_kernel(btf); + ASSERT_EQ(err, expected_err, "check btf"); +out: + btf__free(btf); +} + +void test_local_kptr_ownership(void) +{ + if (test__start_subtest("self_cycle")) + test_self_cycle("kptr", -ELOOP); + if (test__start_subtest("untrusted_self_cycle")) + test_self_cycle("kptr_untrusted", 0); + if (test__start_subtest("percpu_self_cycle")) + test_self_cycle("percpu_kptr", -ELOOP); + if (test__start_subtest("ABA_cycle")) + test_aba_cycle(); + if (test__start_subtest("mixed_graph_root_cycle")) + test_mixed_cycle(); + if (test__start_subtest("max_acyclic")) + test_acyclic_depth(MAX_OWNERSHIP_DEPTH, false, false, 0); + if (test__start_subtest("too_deep_acyclic")) + test_acyclic_depth(MAX_OWNERSHIP_DEPTH + 1, false, false, -ELOOP); + if (test__start_subtest("max_acyclic_child_first")) + test_acyclic_depth(MAX_OWNERSHIP_DEPTH, true, false, 0); + if (test__start_subtest("too_deep_acyclic_child_first")) + test_acyclic_depth(MAX_OWNERSHIP_DEPTH + 1, true, false, -ELOOP); + if (test__start_subtest("max_acyclic_shared_suffix")) + test_acyclic_depth(MAX_OWNERSHIP_DEPTH, false, true, 0); + if (test__start_subtest("too_deep_acyclic_shared_suffix")) + test_acyclic_depth(MAX_OWNERSHIP_DEPTH + 1, false, true, -ELOOP); + if (test__start_subtest("list_three_types")) + test_graph_depth(false, 3, 0); + if (test__start_subtest("list_four_types")) + test_graph_depth(false, 4, 0); + if (test__start_subtest("list_max_depth")) + test_graph_depth(false, MAX_OWNERSHIP_DEPTH, 0); + if (test__start_subtest("list_too_deep")) + test_graph_depth(false, MAX_OWNERSHIP_DEPTH + 1, -ELOOP); + if (test__start_subtest("rbtree_three_types")) + test_graph_depth(true, 3, 0); + if (test__start_subtest("rbtree_four_types")) + test_graph_depth(true, 4, 0); + if (test__start_subtest("rbtree_max_depth")) + test_graph_depth(true, MAX_OWNERSHIP_DEPTH, 0); + if (test__start_subtest("rbtree_too_deep")) + test_graph_depth(true, MAX_OWNERSHIP_DEPTH + 1, -ELOOP); +} From c21eaa72f02fc6e85621cbe09d303d8fb8bd39cd Mon Sep 17 00:00:00 2001 From: Thomas Gleixner Date: Wed, 16 Sep 2026 20:48:30 +0200 Subject: [PATCH 412/592] posix-cpu-timers: Prevent freeing a timer which is queued on the expiry list Kijo analyzed another race in the POSIX CPU timer code: Commit bf635681c906 converted cpu_timer::firing from a tristate value to a boolean. This lost the distinction between "not owned by the firing list" and "still owned, but delivery was canceled". The resulting race is: expiry handler timer_settime() timer_delete() -------------- --------------- -------------- collect timer onto private firing list firing = true observes firing = true firing = false return TIMER_RETRY wait for handler observes firing = false finish deletion unhash and free timer resume list traversal read freed elist.next -> UAF The firing bit is clearly the wrong indicator since that commit. Check whether the timer is queued on the expiry list or not instead. If it is queued clear the firing bit to prevent signal delivery as before and return TIMER_RETRY so the caller unlocks the timer which allows the expiry code to make progress and remove it from the list. Fixes: bf635681c906 ("posix-cpu-timers: Cleanup the firing logic") Reported-by: Kijo Park Debugged-by: Kijo Park Signed-off-by: Thomas Gleixner Tested-by: Kijo Park Reviewed-by: Frederic Weisbecker Cc: stable@vger.kernel.org --- kernel/time/posix-cpu-timers.c | 40 +++++++++++++++++++--------------- 1 file changed, 22 insertions(+), 18 deletions(-) diff --git a/kernel/time/posix-cpu-timers.c b/kernel/time/posix-cpu-timers.c index d73d31c7994f..0bf4fcd969c8 100644 --- a/kernel/time/posix-cpu-timers.c +++ b/kernel/time/posix-cpu-timers.c @@ -408,6 +408,7 @@ static int posix_cpu_timer_create(struct k_itimer *new_timer) new_timer->kclock = &clock_posix_cpu; timerqueue_init(&new_timer->it.cpu.node); + INIT_LIST_HEAD(&new_timer->it.cpu.elist); new_timer->it.cpu.pid = get_pid(pid); rcu_read_unlock(); return 0; @@ -565,6 +566,24 @@ static struct task_struct *timer_lock_sighand(struct k_itimer *timer, unsigned l return NULL; } +/* + * If the timer is queued on the expiry list, then it cannot be dequeued because + * the firing list is not protected by sighand->lock. The delivery path is + * waiting for the timer lock. So go back, unlock and retry. + */ +static bool posix_cpu_timer_on_expiry_list(struct k_itimer *timer) +{ + if (list_empty(&timer->it.cpu.elist)) + return false; + + /* + * Prevent signal delivery as there is no point in delivering a signal + * which is made obsolete right away. + */ + timer->it.cpu.firing = false; + return true; +} + /* * Clean up a CPU-clock timer that is about to be destroyed. * This is called from timer deletion with the timer already locked. @@ -580,18 +599,10 @@ static int posix_cpu_timer_del(struct k_itimer *timer) p = timer_lock_sighand(timer, &flags); if (likely(p)) { - if (timer->it.cpu.firing) { - /* - * Prevent signal delivery. The timer cannot be dequeued - * because it is on the firing list which is not protected - * by sighand->lock. The delivery path is waiting for - * the timer lock. So go back, unlock and retry. - */ - timer->it.cpu.firing = false; + if (posix_cpu_timer_on_expiry_list(timer)) ret = TIMER_RETRY; - } else { + else disarm_timer(timer, p); - } unlock_task_sighand(p, &flags); } @@ -731,14 +742,7 @@ static int posix_cpu_timer_set(struct k_itimer *timer, int timer_flags, /* Retrieve the current expiry time before disarming the timer */ old_expires = cpu_timer_getexpires(ctmr); - if (unlikely(timer->it.cpu.firing)) { - /* - * Prevent signal delivery. The timer cannot be dequeued - * because it is on the firing list which is not protected - * by sighand->lock. The delivery path is waiting for - * the timer lock. So go back, unlock and retry. - */ - timer->it.cpu.firing = false; + if (posix_cpu_timer_on_expiry_list(timer)) { ret = TIMER_RETRY; } else { cpu_timer_dequeue(ctmr); From c82b797abe668d0b668601a93ba2c0b071a63574 Mon Sep 17 00:00:00 2001 From: Weiming Shi Date: Mon, 14 Sep 2026 14:51:23 +0800 Subject: [PATCH 413/592] net/sched: reject IDR error pointers when deleting actions tcf_action_delete() drops the reference held by its lookup before calling tcf_idr_delete_index() with the saved action index. An unlocked classifier can remove that action and reserve the same IDR slot with ERR_PTR(-EBUSY) in between. tcf_idr_delete_index() only checks the lookup result for NULL. It therefore treats the reservation as a tc_action and dereferences tcfa_bindcnt. A hardware execution breakpoint was used to schedule the interleaving without changing the kernel source. KASAN reported this decoded trace: BUG: KASAN: null-ptr-deref in tca_action_gd+0x5b9/0x1010 Read of size 4 at addr 0000000000000010 by task poc/150 Oops: general protection fault, probably for non-canonical address 0xdffffc0000000002 RIP: tca_action_gd+0x5c0/0x1010: arch_atomic_read at arch/x86/include/asm/atomic.h:23 raw_atomic_read at include/linux/atomic/atomic-arch-fallback.h:457 atomic_read at include/linux/atomic/atomic-instrumented.h:33 tcf_idr_delete_index at net/sched/act_api.c:766 tcf_action_delete at net/sched/act_api.c:1859 tcf_del_notify at net/sched/act_api.c:2014 tca_action_gd at net/sched/act_api.c:2064 R13: 0000000000000010 R15: fffffffffffffff0 Kernel panic - not syncing: Fatal exception R15 contains ERR_PTR(-EBUSY), and adding the tcfa_bindcnt offset produces the address in R13. With the guard applied, the same reproducer returned -ENOENT without a KASAN report or panic. Treat error pointers as absent and return -ENOENT. Fixes: 0190c1d452a9 ("net: sched: atomically check-allocate action") Cc: stable@vger.kernel.org Reported-by: Xiang Mei Signed-off-by: Weiming Shi Link: https://patch.msgid.link/20260914065123.4109709-2-bestswngs@gmail.com Signed-off-by: Jakub Kicinski --- net/sched/act_api.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/net/sched/act_api.c b/net/sched/act_api.c index 3f653721c45f..e45a63be397c 100644 --- a/net/sched/act_api.c +++ b/net/sched/act_api.c @@ -758,7 +758,7 @@ static int tcf_idr_delete_index(struct tcf_idrinfo *idrinfo, u32 index) mutex_lock(&idrinfo->lock); p = idr_find(&idrinfo->action_idr, index); - if (!p) { + if (IS_ERR_OR_NULL(p)) { mutex_unlock(&idrinfo->lock); return -ENOENT; } From 9d565b6b72fe3f41fd43636e143072848105189f Mon Sep 17 00:00:00 2001 From: Aamir Ahmed Date: Tue, 15 Sep 2026 00:06:58 +0100 Subject: [PATCH 414/592] net: usb: catc: bound the RX packet length in catc_rx_done() catc_rx_done() walks a multi-packet URB, reading a two-byte length from each packet header. Its bound, pkt_len > urb->actual_length, ignores the header offset and compares against the whole transfer rather than the bytes left from pkt_start, so a crafted packet header makes skb_copy_to_linear_data() read past the buffer. A length below ETH_HLEN is also accepted, including zero, and eth_type_trans() then reads a MAC header from the uninitialised tailroom of a shorter skb. The is_f5u011 branch takes its length straight from the transfer, so a zero-length URB reaches the same path. Track the bytes remaining from the current packet, and reject a header that does not fit, a length past what is left, and a length below an Ethernet header. A transfer shorter than an Ethernet header, including a zero-length one, previously became a runt skb passed to netif_rx() and counted as received; it is now counted in rx_length_errors and ends the walk. Fixes: 1da177e4c3f4 ("Linux-2.6.12-rc2") Signed-off-by: Aamir Ahmed Reviewed-by: Simon Horman Link: https://patch.msgid.link/AS8P251MB00015FD7716F38C345619B56C8BB2@AS8P251MB0001.EURP251.PROD.OUTLOOK.COM Signed-off-by: Jakub Kicinski --- drivers/net/usb/catc.c | 15 ++++++++++++--- 1 file changed, 12 insertions(+), 3 deletions(-) diff --git a/drivers/net/usb/catc.c b/drivers/net/usb/catc.c index 96e82f94edcf..39b678f175dd 100644 --- a/drivers/net/usb/catc.c +++ b/drivers/net/usb/catc.c @@ -233,17 +233,26 @@ static void catc_rx_done(struct urb *urb) } do { - if(!catc->is_f5u011) { - pkt_len = le16_to_cpup((__le16*)pkt_start); - if (pkt_len > urb->actual_length) { + int remaining = urb->actual_length - + (pkt_start - (u8 *)urb->transfer_buffer); + + if (!catc->is_f5u011) { + if (remaining < pkt_offset) { catc->netdev->stats.rx_length_errors++; catc->netdev->stats.rx_errors++; break; } + pkt_len = le16_to_cpup((__le16 *)pkt_start); } else { pkt_len = urb->actual_length; } + if (pkt_len < ETH_HLEN || pkt_len + pkt_offset > remaining) { + catc->netdev->stats.rx_length_errors++; + catc->netdev->stats.rx_errors++; + break; + } + if (!(skb = dev_alloc_skb(pkt_len))) return; From ab888242fce4f16f6c4d4c6ec53939ad36aa3b3a Mon Sep 17 00:00:00 2001 From: Xiang Mei Date: Tue, 15 Sep 2026 01:31:52 -0700 Subject: [PATCH 415/592] vlan: require the MAC header to be present in __vlan_insert_inner_tag() __vlan_insert_inner_tag() only guarantees head room via skb_cow_head(), never that mac_len bytes of MAC header are present. Its ETH_HLEN wrappers - __vlan_insert_tag() under skb_vlan_push(), and vlan_insert_tag() under validate_xmit_vlan() on the generic transmit path - therefore rewrite the first 16 bytes at skb->data: a 12-byte memmove plus two 2-byte stores at +12 and +14. No caller supplies the bound, while the pop helpers use skb_ensure_writable()/pskb_may_pull(). An IFF_TUN device has hard_header_len == 0, so packet_snd() accepts a one-byte AF_PACKET/SOCK_RAW frame. The first vlan push only sets a hwaccel tag; the next - clsact "action vlan push" or bpf_skb_vlan_push() - enters the helper with skb->len still 1. The head comes from skbuff_small_head without __GFP_ZERO, so each push drags bytes from beyond skb->tail into the frame. After three the one-byte send leaves as 13 bytes carrying 11 bytes of uninitialised slab: 0000: 5a b3 62 12 80 88 ff ff 00 b3 62 12 81 `------------------------------' only 0x5a was sent; the rest is slab, here the top 56 bits of a linear-map address Require the MAC header the helper rewrites to be present, so such a frame is dropped rather than transmitted. Fixes: 1da177e4c3f4 ("Linux-2.6.12-rc2") Cc: stable@vger.kernel.org Reported-by: co+0ea1ac045375cf05@bugs.sh Signed-off-by: Xiang Mei Reviewed-by: Simon Horman Link: https://patch.msgid.link/20260915083152.705309-1-xmei5@asu.edu Signed-off-by: Jakub Kicinski --- include/linux/if_vlan.h | 3 +++ 1 file changed, 3 insertions(+) diff --git a/include/linux/if_vlan.h b/include/linux/if_vlan.h index 20cc16ea4e5a..4846032bf4ff 100644 --- a/include/linux/if_vlan.h +++ b/include/linux/if_vlan.h @@ -365,6 +365,9 @@ static inline int __vlan_insert_inner_tag(struct sk_buff *skb, const u8 meta_len = mac_len > ETH_TLEN ? skb_metadata_len(skb) : 0; struct vlan_ethhdr *veth; + if (unlikely(!pskb_may_pull(skb, mac_len))) + return -EINVAL; + if (skb_cow_head(skb, meta_len + VLAN_HLEN) < 0) return -ENOMEM; From a11212910cf09b2fe8db9afa41ef60c4f81879c5 Mon Sep 17 00:00:00 2001 From: Yuqi Xu Date: Sat, 19 Sep 2026 16:45:02 +0800 Subject: [PATCH 416/592] bpf: Check params size before reading reserved fields bpf_crypto_ctx_create() is a kfunc whose second argument is declared with the __sz annotation, so the verifier only guarantees that params__sz bytes of params are valid. The function nevertheless reads params->reserved[0] and params->reserved[1] (offsets 14 and 15) before comparing params__sz against the size of struct bpf_crypto_params, so a BPF program can pass a shorter buffer and have the kernel read past the region that was validated for it. Move the size check in front of the reserved field reads. Fixes: 3e1c6f35409f ("bpf: make common crypto API for TC/XDP programs") Reported-by: Vega Signed-off-by: Yuqi Xu Signed-off-by: Alexei Starovoitov Reviewed-by: Ren Wei Link: https://patch.msgid.link/4f3ab4b03e79017e215521743996555439bf0bb3.1789802413.git.xuyuqiabc@gmail.com --- kernel/bpf/crypto.c | 5 +++-- 1 file changed, 3 insertions(+), 2 deletions(-) diff --git a/kernel/bpf/crypto.c b/kernel/bpf/crypto.c index 51f89cecefb4..3f3fe2450fc6 100644 --- a/kernel/bpf/crypto.c +++ b/kernel/bpf/crypto.c @@ -149,8 +149,9 @@ bpf_crypto_ctx_create(const struct bpf_crypto_params *params, u32 params__sz, const struct bpf_crypto_type *type; struct bpf_crypto_ctx *ctx; - if (!params || params->reserved[0] || params->reserved[1] || - params__sz != sizeof(struct bpf_crypto_params)) { + if (!params || + params__sz != sizeof(struct bpf_crypto_params) || + params->reserved[0] || params->reserved[1]) { *err = -EINVAL; return NULL; } From 8a60ade2277e1f0e0d0578d565354e52292fa46d Mon Sep 17 00:00:00 2001 From: Jamal Hadi Salim Date: Thu, 17 Sep 2026 06:57:32 -0400 Subject: [PATCH 417/592] net/sched: sch_hfsc: bound the classify inner-filter walk with a drift budget MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit hfsc_classify() applies the "filter may only point downwards" level check only when the filter result carries no bound class. A filter created with a flowid gets res.class set once at bind time, so the check never runs for it during classification. hfsc_adjust_levels() can later raise a class's level without revalidating existing bindings, leaving two binds that were each legal at bind time pointing at each other; the classify walk then bounces between two interior classes forever with the qdisc lock held and BH disabled — a soft lockup from a single packet. The stuck walk trips the watchdog: watchdog: BUG: soft lockup - CPU#3 stuck for 13s! [ping:444] RIP: 0010:u32_classify+0x542/0x17f0 ... tcf_classify+0x66/0xa0 hfsc_enqueue+0x166/0xdf0 Bound the traversal with a budget of non-descending hops, the only way a configured walk can move without descending the class tree once levels drift after bind time. The budget is cumulative over the whole walk and is deliberately not reset on a descending hop: a chain that alternates a descent with a lateral hop would return the budget every lap and never trip. Descending hops never decrement it, so legitimately deep trees are unaffected and a terminating lateral chain still classifies normally. Drop the packet with a rate-limited warning once the budget is exhausted, mirroring the merged HTB fix. This is a follow-up to commit 729c4896ab82 ("net/sched: sch_htb: limit htb_classify inner-class filter hops"), which bounded the same classify loop on the HTB side but left the HFSC walk unbounded. Conditions to recreate the bug: - CONFIG_NET_SCHED, CONFIG_NET_SCH_HFSC, CONFIG_NET_CLS_U32, CONFIG_LOCKUP_DETECTOR. - Build a cycle with two legal-at-bind-time flowid binds and a level drift: class X 1:1 (child of root) with leaf child 1:10; class Y 1:2 (sibling of X) with children 1:20 and 1:200; root u32 filter flowid 1:1; filter on X flowid 1:2 (legal when Y is a leaf); after Y's level rises to 2, filter on Y flowid 1:1 (legal then). Send one packet (ping on the device). Unfixed kernel: classify spins with the qdisc lock held; with softlockup_panic=1 it panics. - Reachable from unprivileged user via unshare -Urn (CAP_NET_ADMIN). Fixes: a2f79227138c ("net_sched: sch_hfsc: fix classification loops") Reported-by: Sashiko (gemini + nipa) Closes: https://lore.kernel.org/netdev/QDISC-CTUU.v2.20260913192614@mojatatu.com/ Link: https://sashiko.dev/#/patchset/QDISC-CTUU.v2.20260913192614@mojatatu.com Link: https://netdev-ai.bots.linux.dev/sashiko/#/patchset/QDISC-CTUU.v2.20260913192614%40mojatatu.com Reviewed-by: Victor Nogueira Tested-by: hybris Signed-off-by: Jamal Hadi Salim Link: https://patch.msgid.link/QDISC-CTUU.v3.20260916184908@mojatatu.com Signed-off-by: Jakub Kicinski --- net/sched/sch_hfsc.c | 22 ++++++++++++++++++++++ 1 file changed, 22 insertions(+) diff --git a/net/sched/sch_hfsc.c b/net/sched/sch_hfsc.c index e87f5021a199..284490fd6ca9 100644 --- a/net/sched/sch_hfsc.c +++ b/net/sched/sch_hfsc.c @@ -386,6 +386,15 @@ cftree_update(struct hfsc_class *cl) #define SM_MASK ((1ULL << SM_SHIFT) - 1) #define ISM_MASK ((1ULL << ISM_SHIFT) - 1) +/* + * Cap on the non-descending hops a classify walk may take before its + * filter chain is treated as misconfigured. A flowid binding that was + * legal at bind time can become lateral once hfsc_adjust_levels() + * raises a class level; a few such hops are legitimate, an unbounded + * run means the chain cycles. + */ +#define HFSC_CLASSIFY_MAX_DRIFT 8 + static inline u64 seg_x2y(u64 x, u64 sm) { @@ -1133,6 +1142,7 @@ hfsc_classify(struct sk_buff *skb, struct Qdisc *sch, int *qerr) struct hfsc_class *head, *cl; struct tcf_result res; struct tcf_proto *tcf; + unsigned int drift; int result; if (TC_H_MAJ(skb->priority ^ sch->handle) == 0 && @@ -1142,6 +1152,7 @@ hfsc_classify(struct sk_buff *skb, struct Qdisc *sch, int *qerr) *qerr = NET_XMIT_SUCCESS | __NET_XMIT_BYPASS; head = &q->root; + drift = HFSC_CLASSIFY_MAX_DRIFT; tcf = rcu_dereference_bh(q->root.filter_list); while (tcf && (result = tcf_classify_qdisc(skb, tcf, &res, false)) >= 0) { #ifdef CONFIG_NET_CLS_ACT @@ -1167,6 +1178,17 @@ hfsc_classify(struct sk_buff *skb, struct Qdisc *sch, int *qerr) if (cl->level == 0) return cl; /* hit leaf class */ + /* + * flowid binds skip the level check above (res.class is set + * at bind time and levels drift after), so a walk can follow + * lateral hops without descending; a bounded number of them + * is legal, more means the chain cycles. + */ + if (cl->level >= head->level && drift-- == 0) { + pr_warn_ratelimited("hfsc: classify hop budget exhausted, dropping packet\n"); + return NULL; + } + /* apply inner filter chain */ tcf = rcu_dereference_bh(cl->filter_list); head = cl; From 1e24c4f2ee44be0eee94092b5d13cbdb4bdf0d60 Mon Sep 17 00:00:00 2001 From: Jamal Hadi Salim Date: Thu, 17 Sep 2026 06:57:33 -0400 Subject: [PATCH 418/592] selftests: tc-testing: add a lateral-drift hfsc classify-walk test The classify-loop fix bounds a walk's non-descending hops, so the guard must not misfire on a legal walk that reaches its leaf through a level-drift lateral chain. Add a case that builds exactly that chain and asserts traffic still reaches the chain's own leaf. A lateral hop can only exist because a bind was legal when it was made and a later class add raised the target's level, so the setup binds each hop while the target is still a leaf and only then deepens it: bind 1:1 -> 1:2 while 1:2 is a leaf, add 1:20 under 1:2, add 1:3 and bind 1:2 -> 1:3 while 1:3 is a leaf, then add 1:30 and 1:31 under 1:3 and bind 1:3 -> 1:31. The walk root -> 1:1 -> 1:2 -> 1:3 -> 1:31 then takes two lateral hops and must reach leaf 1:31. The default class is 1:30, distinct from the asserted leaf, and the verify pattern is anchored to the 1:31 stats line, so neither a fall-through to the default nor a nonzero count on another class can satisfy the check. On the patched kernel the test passes; with the bound forced to zero the walk falls to the default and 1:31 stays idle, so the test fails. Reviewed-by: Victor Nogueira Tested-by: hybris Signed-off-by: Jamal Hadi Salim Link: https://patch.msgid.link/QDISC-CTUU.v3.20260916184908@mojatatu.com.2 Signed-off-by: Jakub Kicinski --- .../tc-testing/tc-tests/qdiscs/hfsc.json | 34 +++++++++++++++++++ 1 file changed, 34 insertions(+) diff --git a/tools/testing/selftests/tc-testing/tc-tests/qdiscs/hfsc.json b/tools/testing/selftests/tc-testing/tc-tests/qdiscs/hfsc.json index c98c339424d4..4f6bbb8b57f9 100644 --- a/tools/testing/selftests/tc-testing/tc-tests/qdiscs/hfsc.json +++ b/tools/testing/selftests/tc-testing/tc-tests/qdiscs/hfsc.json @@ -169,5 +169,39 @@ "teardown": [ "$TC qdisc del dev $DUMMY handle 1: root" ] + }, + { + "id": "8c39", + "name": "HFSC classify walk still reaches leaf after lateral drift", + "category": [ + "qdisc", + "hfsc" + ], + "plugins": { + "requires": "nsPlugin" + }, + "setup": [ + "ip link set lo up", + "$TC qdisc add dev lo handle 1: root hfsc default 30", + "$TC class add dev lo parent 1: classid 1:1 hfsc rt m2 100kbit", + "$TC class add dev lo parent 1:1 classid 1:10 hfsc rt m2 50kbit", + "$TC class add dev lo parent 1: classid 1:2 hfsc rt m2 100kbit", + "$TC filter add dev lo parent 1: protocol ip prio 1 u32 match u8 0 0 at 0 flowid 1:1", + "$TC filter add dev lo parent 1:1 protocol ip prio 1 u32 match u8 0 0 at 0 flowid 1:2", + "$TC class add dev lo parent 1:2 classid 1:20 hfsc rt m2 10kbit", + "$TC class add dev lo parent 1: classid 1:3 hfsc rt m2 100kbit", + "$TC filter add dev lo parent 1:2 protocol ip prio 1 u32 match u8 0 0 at 0 flowid 1:3", + "$TC class add dev lo parent 1:3 classid 1:30 hfsc rt m2 10kbit", + "$TC class add dev lo parent 1:3 classid 1:31 hfsc rt m2 100kbit", + "$TC filter add dev lo parent 1:3 protocol ip prio 1 u32 match u8 0 0 at 0 flowid 1:31" + ], + "cmdUnderTest": "ping -n -c 10 -W 1 127.0.0.1", + "expExitCode": "0", + "verifyCmd": "$TC -s class show dev lo", + "matchPattern": "class hfsc 1:31 parent 1:3 rt[^\\n]*\\n Sent [0-9]+ bytes [1-9][0-9]* pkt", + "matchCount": "1", + "teardown": [ + "$TC qdisc del dev lo handle 1: root" + ] } ] From f7eeb1af8537b05953fb1c88ab8b59d94059a381 Mon Sep 17 00:00:00 2001 From: Shengzhuo Wei Date: Thu, 27 Aug 2026 23:43:01 +0800 Subject: [PATCH 419/592] i2c: at91: release DMA channels on remove and probe error at91_twi_configure_dma() requests exclusive tx/rx DMA channels, but nothing ever releases them on driver detach, and the probe error path after the channels are acquired (i2c_add_numbered_adapter() failure) returns without releasing them either, because the remove callback is not invoked after a failed probe. Move the release into a helper, call it from the existing configure-failure path, the adapter-registration failure path, and at91_twi_remove(). Fixes: 60937b2cdbf9 ("i2c: at91: add dma support") Assisted-by: GLM:5.3 Signed-off-by: Shengzhuo Wei Cc: # v3.8+ Acked-by: Mukesh Kumar Savaliya Signed-off-by: Andi Shyti Link: https://patch.msgid.link/20260827-i2c-dma-channel-leak-v1-1-271d4adc03a0@cherr.cc --- drivers/i2c/busses/i2c-at91-core.c | 3 +++ drivers/i2c/busses/i2c-at91-master.c | 12 +++++++++++- drivers/i2c/busses/i2c-at91.h | 1 + 3 files changed, 15 insertions(+), 1 deletion(-) diff --git a/drivers/i2c/busses/i2c-at91-core.c b/drivers/i2c/busses/i2c-at91-core.c index b64adef778d4..8ca4556d9664 100644 --- a/drivers/i2c/busses/i2c-at91-core.c +++ b/drivers/i2c/busses/i2c-at91-core.c @@ -255,6 +255,7 @@ static int at91_twi_probe(struct platform_device *pdev) if (rc) { pm_runtime_disable(dev->dev); pm_runtime_set_suspended(dev->dev); + at91_twi_dma_release(dev); return rc; } @@ -270,6 +271,8 @@ static void at91_twi_remove(struct platform_device *pdev) i2c_del_adapter(&dev->adapter); + at91_twi_dma_release(dev); + pm_runtime_disable(dev->dev); pm_runtime_set_suspended(dev->dev); } diff --git a/drivers/i2c/busses/i2c-at91-master.c b/drivers/i2c/busses/i2c-at91-master.c index 894cedbca99f..68238cc8aee0 100644 --- a/drivers/i2c/busses/i2c-at91-master.c +++ b/drivers/i2c/busses/i2c-at91-master.c @@ -817,11 +817,21 @@ static int at91_twi_configure_dma(struct at91_twi_dev *dev, u32 phy_addr) error: if (ret != -EPROBE_DEFER) dev_info(dev->dev, "can't get DMA channel, continue without DMA support\n"); + at91_twi_dma_release(dev); + return ret; +} + +void at91_twi_dma_release(struct at91_twi_dev *dev) +{ + struct at91_twi_dma *dma = &dev->dma; + if (dma->chan_rx) dma_release_channel(dma->chan_rx); if (dma->chan_tx) dma_release_channel(dma->chan_tx); - return ret; + dma->chan_rx = NULL; + dma->chan_tx = NULL; + dev->use_dma = false; } static int at91_init_twi_recovery_gpio(struct platform_device *pdev, diff --git a/drivers/i2c/busses/i2c-at91.h b/drivers/i2c/busses/i2c-at91.h index 942e9c3973bb..d68fcbbc3e0f 100644 --- a/drivers/i2c/busses/i2c-at91.h +++ b/drivers/i2c/busses/i2c-at91.h @@ -172,6 +172,7 @@ void at91_twi_irq_restore(struct at91_twi_dev *dev); void at91_init_twi_bus(struct at91_twi_dev *dev); void at91_init_twi_bus_master(struct at91_twi_dev *dev); +void at91_twi_dma_release(struct at91_twi_dev *dev); int at91_twi_probe_master(struct platform_device *pdev, u32 phy_addr, struct at91_twi_dev *dev); From e9f03b9625e2eeaca357b065c92d5b14064a1583 Mon Sep 17 00:00:00 2001 From: Shengzhuo Wei Date: Thu, 27 Aug 2026 23:43:02 +0800 Subject: [PATCH 420/592] i2c: imx: release DMA channels on probe error i2c_imx_dma_request() acquires exclusive tx/rx DMA channels and is optional: on errors other than -EPROBE_DEFER the driver falls back to PIO mode and probe continues. If i2c_add_numbered_adapter() then fails, probe returns through clk_notifier_unregister without releasing the channels, because the remove callback is not invoked after a failed probe. Release the channels on the probe error path, mirroring i2c_imx_remove(). Fixes: ce1a78840ff7 ("i2c: imx: add DMA support for freescale i2c driver") Assisted-by: GLM:5.3 Signed-off-by: Shengzhuo Wei Cc: # v3.19+ Reviewed-by: Frank Li Signed-off-by: Andi Shyti Link: https://patch.msgid.link/20260827-i2c-dma-channel-leak-v1-2-271d4adc03a0@cherr.cc --- drivers/i2c/busses/i2c-imx.c | 2 ++ 1 file changed, 2 insertions(+) diff --git a/drivers/i2c/busses/i2c-imx.c b/drivers/i2c/busses/i2c-imx.c index 4dc3df6a41a4..ff5a91158df9 100644 --- a/drivers/i2c/busses/i2c-imx.c +++ b/drivers/i2c/busses/i2c-imx.c @@ -1880,6 +1880,8 @@ static int i2c_imx_probe(struct platform_device *pdev) clk_notifier_unregister: clk_notifier_unregister(i2c_imx->clk, &i2c_imx->clk_change_nb); + if (i2c_imx->dma) + i2c_imx_dma_free(i2c_imx); free_irq(irq, i2c_imx); rpm_disable: pm_runtime_put_noidle(&pdev->dev); From 268aacb2e2a5c94d08e23194961234d0710c8407 Mon Sep 17 00:00:00 2001 From: Shengzhuo Wei Date: Thu, 27 Aug 2026 23:43:03 +0800 Subject: [PATCH 421/592] i2c: qcom-geni: release DMA channels on probe error geni_i2c_init() grabs exclusive GPI tx/rx DMA channels when the serial engine runs in GPI mode. If i2c_add_adapter() subsequently fails, probe returns without releasing the channels, because the remove callback is not invoked after a failed probe. The adapter-registration failure path used to release the channels via its err_dma label; that release was dropped when the probe tail was restructured into geni_i2c_init(). Release the channels on the adapter-registration failure path, mirroring geni_i2c_remove(). Fixes: d8d3bb127ad1 ("i2c: qcom-geni: Isolate serial engine setup") Assisted-by: GLM:5.3 Signed-off-by: Shengzhuo Wei Reviewed-by: Konrad Dybcio Reviewed-by: Mukesh Kumar Savaliya Signed-off-by: Andi Shyti Link: https://patch.msgid.link/20260827-i2c-dma-channel-leak-v1-3-271d4adc03a0@cherr.cc --- drivers/i2c/busses/i2c-qcom-geni.c | 4 +++- 1 file changed, 3 insertions(+), 1 deletion(-) diff --git a/drivers/i2c/busses/i2c-qcom-geni.c b/drivers/i2c/busses/i2c-qcom-geni.c index 00013b41a6f5..e9e41a174c20 100644 --- a/drivers/i2c/busses/i2c-qcom-geni.c +++ b/drivers/i2c/busses/i2c-qcom-geni.c @@ -1189,8 +1189,10 @@ static int geni_i2c_probe(struct platform_device *pdev) return ret; ret = i2c_add_adapter(&gi2c->adap); - if (ret) + if (ret) { + release_gpi_dma(gi2c); return dev_err_probe(dev, ret, "Error adding i2c adapter\n"); + } dev_dbg(dev, "Geni-I2C adaptor successfully added\n"); From 7362a1553eb09a8cdf8be7e509bd5309a8342486 Mon Sep 17 00:00:00 2001 From: Liu Zhenlong Date: Wed, 19 Aug 2026 01:57:50 +0800 Subject: [PATCH 422/592] i2c: qcom-cci: fix device_node refcount leak in cci_probe()/cci_remove() The of_node_put() matching of_node_get() runs after i2c_del_adapter(), whose trailing memset() zeroes adap->dev and thus adap->dev.of_node, making the put a no-op and leaking the node on every adapter removal and error cleanup. Use a devm action: the pointer is captured at registration, out of reach of that memset(), and devres runs the put once on probe failure and detach, replacing the three manual of_node_put() calls. The setup loop uses the scoped iterator form so the child node is released automatically if devm_add_action_or_reset() fails mid-loop. Suggested-by: Konrad Dybcio Fixes: 02a4a69667a2 ("i2c: qcom-cci: don't put a device tree node before i2c_add_adapter()") Assisted-by: Claude:claude-opus-5 Signed-off-by: Liu Zhenlong Cc: # v5.17+ Reviewed-by: Vladimir Zapolskiy Reviewed-by: Konrad Dybcio Signed-off-by: Andi Shyti Link: https://patch.msgid.link/20260818175750.4205-1-dragonliu2018@gmail.com --- drivers/i2c/busses/i2c-qcom-cci.c | 20 +++++++++++--------- 1 file changed, 11 insertions(+), 9 deletions(-) diff --git a/drivers/i2c/busses/i2c-qcom-cci.c b/drivers/i2c/busses/i2c-qcom-cci.c index 25b6e4e9e3fa..873e901a23d7 100644 --- a/drivers/i2c/busses/i2c-qcom-cci.c +++ b/drivers/i2c/busses/i2c-qcom-cci.c @@ -497,10 +497,14 @@ static const struct dev_pm_ops qcom_cci_pm = { SET_RUNTIME_PM_OPS(cci_suspend_runtime, cci_resume_runtime, NULL) }; +static void cci_put_of_node(void *data) +{ + of_node_put(data); +} + static int cci_probe(struct platform_device *pdev) { struct device *dev = &pdev->dev; - struct device_node *child; struct resource *r; struct cci *cci; int ret, i; @@ -516,7 +520,7 @@ static int cci_probe(struct platform_device *pdev) if (!cci->data) return -ENOENT; - for_each_available_child_of_node(dev->of_node, child) { + for_each_available_child_of_node_scoped(dev->of_node, child) { struct cci_master *master; u32 idx; @@ -537,6 +541,9 @@ static int cci_probe(struct platform_device *pdev) master->adap.algo = &cci_algo; master->adap.dev.parent = dev; master->adap.dev.of_node = of_node_get(child); + ret = devm_add_action_or_reset(dev, cci_put_of_node, child); + if (ret) + return ret; master->master = idx; master->cci = cci; @@ -606,10 +613,8 @@ static int cci_probe(struct platform_device *pdev) continue; ret = i2c_add_adapter(&cci->master[i].adap); - if (ret < 0) { - of_node_put(cci->master[i].adap.dev.of_node); + if (ret < 0) goto error_i2c; - } } return 0; @@ -617,10 +622,8 @@ static int cci_probe(struct platform_device *pdev) error_i2c: for (--i ; i >= 0; i--) { - if (cci->master[i].cci) { + if (cci->master[i].cci) i2c_del_adapter(&cci->master[i].adap); - of_node_put(cci->master[i].adap.dev.of_node); - } } disable_clocks: cci_disable_clocks(cci); @@ -636,7 +639,6 @@ static void cci_remove(struct platform_device *pdev) for (i = 0; i < cci->data->num_masters; i++) { if (cci->master[i].cci) { i2c_del_adapter(&cci->master[i].adap); - of_node_put(cci->master[i].adap.dev.of_node); cci_halt(cci, i); } } From c4e941bb7654bcbdfb0b6f3341dc2acfdf235c8d Mon Sep 17 00:00:00 2001 From: Arnd Bergmann Date: Tue, 15 Sep 2026 22:10:30 +0200 Subject: [PATCH 423/592] landlock: Work around gcc-16 -Wuninitialized warning MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit gcc has a bug with -ftrivial-auto-var-init=pattern that produces a warning for correct code that uses sparse bitfields: security/landlock/fs.c: In function 'is_access_to_paths_allowed.isra': security/landlock/fs.c:767:28: error: '_layer_masks_child1' is used uninitialized [-Werror=uninitialized] 767 | struct layer_masks _layer_masks_child1, _layer_masks_child2; | ^~~~~~~~~~~~~~~~~~~ security/landlock/fs.c:767:28: note: '_layer_masks_child1' declared here 767 | struct layer_masks _layer_masks_child1, _layer_masks_child2; | ^~~~~~~~~~~~~~~~~~~ security/landlock/fs.c: In function 'hook_unix_find': security/landlock/fs.c:1649:28: error: 'layer_masks' is used uninitialized [-Werror=uninitialized] 1649 | struct layer_masks layer_masks; | ^~~~~~~~~~~ security/landlock/fs.c:1649:28: note: 'layer_masks' declared here 1649 | struct layer_masks layer_masks; | ^~~~~~~~~~~ To work around this, change the definition of struct layer_mask to use an explictit padding field. Link: https://gcc.gnu.org/bugzilla/show_bug.cgi?id=110743 Link: https://lore.kernel.org/all/20260619082133.3504146-1-arnd@kernel.org/ Fixes: a260c0055665 ("landlock: Add a place for flags to layer rules") Signed-off-by: Arnd Bergmann Link: https://patch.msgid.link/20260915201036.3527935-1-arnd@kernel.org [mic: Use BITS_PER_TYPE(), fix kdoc warnings, fix commit message according to v2 changes] Cc: stable@vger.kernel.org Signed-off-by: Mickaël Salaün --- security/landlock/access.h | 8 ++++++++ 1 file changed, 8 insertions(+) diff --git a/security/landlock/access.h b/security/landlock/access.h index bbbb41f41147..f843835851d0 100644 --- a/security/landlock/access.h +++ b/security/landlock/access.h @@ -61,6 +61,10 @@ union access_masks_all { static_assert(sizeof(typeof_member(union access_masks_all, masks)) == sizeof(typeof_member(union access_masks_all, all))); +#define _LANDLOCK_LAYER_MASK_PADDING \ + (BITS_PER_TYPE(access_mask_t) - LANDLOCK_NUM_ACCESS_MAX - \ + IS_ENABLED(CONFIG_SECURITY_LANDLOCK_LOG)) + /** * struct layer_mask - The access rights and rule flags for a layer. * @@ -81,6 +85,10 @@ struct layer_mask { */ access_mask_t quiet : 1; #endif /* CONFIG_SECURITY_LANDLOCK_LOG */ + /** + * @__pad: Padding for the compiler's bitfield initialization. + */ + access_mask_t __pad : _LANDLOCK_LAYER_MASK_PADDING; } __packed __aligned(sizeof(access_mask_t)); /* From f71ecaece401cef287cdca12aad785fec809cb9e Mon Sep 17 00:00:00 2001 From: =?UTF-8?q?Micka=C3=ABl=20Sala=C3=BCn?= Date: Fri, 18 Sep 2026 20:50:24 +0200 Subject: [PATCH 424/592] landlock: Fix tracepoint fixed-width type names MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit The new Landlock tracepoints use UAPI-prefixed __u32 and __u64 names for callback arguments and record fields, including internal IDs that are not Landlock UAPI values. Typed BPF consumers see callback typedef names through BTF. Use the kernel u32 and u64 aliases before release so the tracepoint contract does not present internal values as Landlock UAPI types. This changes BTF-visible typedef spelling but not integer widths, calling conventions, tracefs formats, or record layouts. Cc: Günther Noack Cc: Steven Rostedt Link: https://patch.msgid.link/20260918185036.608651-2-mic@digikod.net Signed-off-by: Mickaël Salaün --- include/trace/events/landlock.h | 68 ++++++++++++++++----------------- 1 file changed, 34 insertions(+), 34 deletions(-) diff --git a/include/trace/events/landlock.h b/include/trace/events/landlock.h index d05253afaf59..4984f80923ed 100644 --- a/include/trace/events/landlock.h +++ b/include/trace/events/landlock.h @@ -243,8 +243,8 @@ static inline const char *__trace_landlock_print_layers( * Field encoding * ~~~~~~~~~~~~~~ * - * Fields that mirror the Landlock UAPI use the same C types and endianness - * (e.g. network ports are __u64 in host endianness, like + * Fields that mirror the Landlock UAPI preserve their widths and endianness + * (e.g. network ports are u64 in host endianness, like * landlock_net_port_attr.port). Per-event details, such as where a value * is byte-swapped, live in the field's own kdoc. * @@ -319,8 +319,8 @@ TRACE_EVENT(landlock_create_ruleset, TP_ARGS(ruleset), TP_STRUCT__entry( - __field( __u64, ruleset_id ) - __field( __u32, ruleset_version ) + __field( u64, ruleset_id ) + __field( u32, ruleset_version ) __field( access_mask_t, handled_fs ) __field( access_mask_t, handled_net ) __field( access_mask_t, scoped ) @@ -359,8 +359,8 @@ TRACE_EVENT(landlock_free_ruleset, TP_ARGS(ruleset), TP_STRUCT__entry( - __field( __u64, ruleset_id ) - __field( __u32, ruleset_version ) + __field( u64, ruleset_id ) + __field( u32, ruleset_version ) ), TP_fast_assign( @@ -396,8 +396,8 @@ TRACE_EVENT(landlock_add_rule_fs, TP_ARGS(ruleset, access_rights, path, pathname), TP_STRUCT__entry( - __field( __u64, ruleset_id ) - __field( __u32, ruleset_version ) + __field( u64, ruleset_id ) + __field( u32, ruleset_version ) __field( access_mask_t, access_rights ) __field( dev_t, dev ) __field( ino_t, ino ) @@ -443,15 +443,15 @@ TRACE_EVENT(landlock_add_rule_fs, TRACE_EVENT(landlock_add_rule_net, TP_PROTO(const struct landlock_ruleset *ruleset, - access_mask_t access_rights, __u64 port), + access_mask_t access_rights, u64 port), TP_ARGS(ruleset, access_rights, port), TP_STRUCT__entry( - __field( __u64, ruleset_id ) - __field( __u32, ruleset_version ) + __field( u64, ruleset_id ) + __field( u32, ruleset_version ) __field( access_mask_t, access_rights ) - __field( __u64, port ) + __field( u64, port ) ), TP_fast_assign( @@ -495,10 +495,10 @@ TRACE_EVENT(landlock_create_domain, TP_ARGS(domain, ruleset), TP_STRUCT__entry( - __field( __u64, domain_id ) - __field( __u64, parent_id ) - __field( __u64, ruleset_id ) - __field( __u32, ruleset_version ) + __field( u64, domain_id ) + __field( u64, parent_id ) + __field( u64, ruleset_id ) + __field( u32, ruleset_version ) ), TP_fast_assign( @@ -557,7 +557,7 @@ TRACE_EVENT(landlock_enforce_domain, TP_ARGS(domain, complete, process_wide, no_new_privs), TP_STRUCT__entry( - __field( __u64, domain_id ) + __field( u64, domain_id ) __field( bool, complete ) __field( bool, process_wide ) __field( bool, no_new_privs ) @@ -595,8 +595,8 @@ TRACE_EVENT(landlock_free_domain, TP_ARGS(hierarchy), TP_STRUCT__entry( - __field( __u64, domain_id ) - __field( __u64, denials ) + __field( u64, domain_id ) + __field( u64, denials ) ), TP_fast_assign( @@ -631,7 +631,7 @@ TRACE_EVENT(landlock_check_rule_fs, TP_ARGS(domain, rule, access_request, dentry), TP_STRUCT__entry( - __field( __u64, domain_id ) + __field( u64, domain_id ) __field( access_mask_t, access_request ) __field( dev_t, dev ) __field( ino_t, ino ) @@ -675,14 +675,14 @@ TRACE_EVENT(landlock_check_rule_net, TP_PROTO(const struct landlock_domain *domain, const struct landlock_rule *rule, - access_mask_t access_request, __u64 port), + access_mask_t access_request, u64 port), TP_ARGS(domain, rule, access_request, port), TP_STRUCT__entry( - __field( __u64, domain_id ) + __field( u64, domain_id ) __field( access_mask_t, access_request ) - __field( __u64, port ) + __field( u64, port ) __dynamic_array(access_mask_t, grants, domain->num_layers) ), @@ -729,7 +729,7 @@ TRACE_EVENT(landlock_deny_access_fs, TP_ARGS(hierarchy, same_exec, logged, blockers, path, pathname), TP_STRUCT__entry( - __field( __u64, domain_id ) + __field( u64, domain_id ) __field( bool, same_exec ) __field( bool, logged ) __field( access_mask_t, blockers ) @@ -791,17 +791,17 @@ TRACE_EVENT(landlock_deny_access_net, TP_PROTO(const struct landlock_hierarchy *hierarchy, bool same_exec, bool logged, access_mask_t blockers, const struct sock *sk, - __u64 sport, __u64 dport), + u64 sport, u64 dport), TP_ARGS(hierarchy, same_exec, logged, blockers, sk, sport, dport), TP_STRUCT__entry( - __field( __u64, domain_id ) + __field( u64, domain_id ) __field( bool, same_exec ) __field( bool, logged ) __field( access_mask_t, blockers ) - __field( __u64, sport ) - __field( __u64, dport ) + __field( u64, sport ) + __field( u64, dport ) ), TP_fast_assign( @@ -842,10 +842,10 @@ TRACE_EVENT(landlock_deny_ptrace, TP_ARGS(hierarchy, same_exec, logged, tracee_domain_id, tracee), TP_STRUCT__entry( - __field( __u64, domain_id ) + __field( u64, domain_id ) __field( bool, same_exec ) __field( bool, logged ) - __field( __u64, tracee_domain_id) + __field( u64, tracee_domain_id) __field( pid_t, tracee_pid ) __string( tracee_comm, tracee->comm ) ), @@ -891,10 +891,10 @@ TRACE_EVENT(landlock_deny_scope_signal, TP_ARGS(hierarchy, same_exec, logged, target_domain_id, target), TP_STRUCT__entry( - __field( __u64, domain_id ) + __field( u64, domain_id ) __field( bool, same_exec ) __field( bool, logged ) - __field( __u64, target_domain_id) + __field( u64, target_domain_id) __field( pid_t, target_pid ) __string( target_comm, target->comm ) ), @@ -940,10 +940,10 @@ TRACE_EVENT(landlock_deny_scope_abstract_unix_socket, TP_ARGS(hierarchy, same_exec, logged, peer_domain_id, peer), TP_STRUCT__entry( - __field( __u64, domain_id ) + __field( u64, domain_id ) __field( bool, same_exec ) __field( bool, logged ) - __field( __u64, peer_domain_id ) + __field( u64, peer_domain_id ) __field( pid_t, peer_pid ) /* * Abstract socket names are untrusted binary data from From 0de33ca344fbf983d380d78db6eeb6fe312d5a5e Mon Sep 17 00:00:00 2001 From: =?UTF-8?q?Micka=C3=ABl=20Sala=C3=BCn?= Date: Fri, 18 Sep 2026 20:50:25 +0200 Subject: [PATCH 425/592] landlock: Fix filesystem denial blocker reporting MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit Filesystem topology denials are rendered with an empty blockers value because their blocker is identified by the request type instead of an access mask. Introduce the private struct landlock_blockers to carry the request type and final missing access mask to filesystem and network denial tracepoints. Copy both members into named trace-record fields, then use the type to print change_topology for topology denials while preserving symbolic access masks for ordinary denials. The request type lets typed BPF consumers distinguish topology denials from access denials. Keeping the native access mask in a pointer-reached field also lets CO-RE adjust existing programs' load width if access_mask_t grows. Cc: Günther Noack Cc: Steven Rostedt Fixes: 01ce260f5ccf ("landlock: Add landlock_deny_access_fs and landlock_deny_access_net") Link: https://patch.msgid.link/20260918185036.608651-3-mic@digikod.net Signed-off-by: Mickaël Salaün --- include/trace/events/landlock.h | 51 ++++++++++++++++++++++++--------- security/landlock/log.h | 5 ++++ security/landlock/trace.c | 19 +++++++++--- 3 files changed, 58 insertions(+), 17 deletions(-) diff --git a/include/trace/events/landlock.h b/include/trace/events/landlock.h index 4984f80923ed..4e304cab1dab 100644 --- a/include/trace/events/landlock.h +++ b/include/trace/events/landlock.h @@ -17,7 +17,9 @@ #include #include +enum landlock_request_type; struct dentry; +struct landlock_blockers; struct landlock_domain; struct landlock_hierarchy; struct landlock_rule; @@ -26,6 +28,10 @@ struct path; struct sock; struct task_struct; +TRACE_DEFINE_ENUM(LANDLOCK_REQUEST_FS_CHANGE_TOPOLOGY); +TRACE_DEFINE_ENUM(LANDLOCK_REQUEST_FS_ACCESS); +TRACE_DEFINE_ENUM(LANDLOCK_REQUEST_NET_ACCESS); + #ifdef CREATE_TRACE_POINTS /* About 6 KiB, leaving about 2 KiB for sibling helpers and fixed fields. */ @@ -182,6 +188,9 @@ static inline const char *__trace_landlock_print_layers( /* Maps a shared _LANDLOCK_*_NAMES entry to a __print_flags() pair. */ #define _LANDLOCK_NAME_ENTRY(mask, name) { mask, name } +#define _LANDLOCK_FS_BLOCKER_TYPE_NAMES \ + { LANDLOCK_REQUEST_FS_CHANGE_TOPOLOGY, "change_topology" } + /** * DOC: Landlock trace events * @@ -281,6 +290,13 @@ static inline const char *__trace_landlock_print_layers( * the two parties without kernel-internal state. The ID is a scalar * snapshot, not a live domain pointer that could dangle: an optional * relational referent is a scalar (0 sentinel), not a nullable pointer. + * + * Blocker fields + * ~~~~~~~~~~~~~~ + * + * The filesystem and network blocker arguments identify the request type + * and carry its final missing access subset when applicable. The type + * determines how to interpret the access value. */ /* @@ -712,8 +728,7 @@ TRACE_EVENT(landlock_check_rule_net, * domain field. * @same_exec: Whether the current task entered the denying domain itself. * @logged: The domain's audit-logging decision for this denial. - * @blockers: Access mask that was blocked (zero for a mount-topology - * change, whose only blocker is the operation itself). + * @blockers: Request type and final missing access subset (never NULL). * @path: Filesystem path that was denied (never NULL). * @pathname: Resolved path string (never NULL; an error placeholder on * resolution failure). @@ -723,8 +738,8 @@ TRACE_EVENT(landlock_check_rule_net, TRACE_EVENT(landlock_deny_access_fs, TP_PROTO(const struct landlock_hierarchy *hierarchy, bool same_exec, - bool logged, access_mask_t blockers, const struct path *path, - const char *pathname), + bool logged, const struct landlock_blockers *blockers, + const struct path *path, const char *pathname), TP_ARGS(hierarchy, same_exec, logged, blockers, path, pathname), @@ -732,7 +747,8 @@ TRACE_EVENT(landlock_deny_access_fs, __field( u64, domain_id ) __field( bool, same_exec ) __field( bool, logged ) - __field( access_mask_t, blockers ) + __field( enum landlock_request_type, blockers_type ) + __field( access_mask_t, blockers_access ) __field( dev_t, dev ) __field( ino_t, ino ) __string( pathname, pathname ) @@ -744,7 +760,8 @@ TRACE_EVENT(landlock_deny_access_fs, __entry->domain_id = hierarchy->id; __entry->same_exec = same_exec; __entry->logged = logged; - __entry->blockers = blockers; + __entry->blockers_type = blockers->type; + __entry->blockers_access = blockers->access; __entry->dev = path->dentry->d_sb->s_dev; /* * A negative dentry has no backing inode, so mirror the @@ -756,7 +773,10 @@ TRACE_EVENT(landlock_deny_access_fs, TP_printk("domain=%llx same_exec=%d logged=%d blockers=%s dev=%u:%u ino=%lu path=%s", __entry->domain_id, __entry->same_exec, __entry->logged, - __print_flags(__entry->blockers, "|", _LANDLOCK_ACCESS_FS_NAMES), + __entry->blockers_type == LANDLOCK_REQUEST_FS_ACCESS ? + __print_flags(__entry->blockers_access, "|", _LANDLOCK_ACCESS_FS_NAMES) : + __print_symbolic(__entry->blockers_type, + _LANDLOCK_FS_BLOCKER_TYPE_NAMES), MAJOR(__entry->dev), MINOR(__entry->dev), __entry->ino, __trace_print_untrusted_str(p, __get_str(pathname), __get_dynamic_array_len(pathname) - 1)) @@ -769,7 +789,7 @@ TRACE_EVENT(landlock_deny_access_fs, * domain field. * @same_exec: Whether the current task entered the denying domain itself. * @logged: The domain's audit-logging decision for this denial. - * @blockers: Access mask that was blocked. + * @blockers: Request type and final missing access subset (never NULL). * @sk: Socket object (never NULL), read without a socket lock, so its * fields are a best-effort snapshot. The denied endpoint is not * available: the hook runs before :manpage:`bind(2)` / @@ -790,8 +810,8 @@ TRACE_EVENT(landlock_deny_access_fs, TRACE_EVENT(landlock_deny_access_net, TP_PROTO(const struct landlock_hierarchy *hierarchy, bool same_exec, - bool logged, access_mask_t blockers, const struct sock *sk, - u64 sport, u64 dport), + bool logged, const struct landlock_blockers *blockers, + const struct sock *sk, u64 sport, u64 dport), TP_ARGS(hierarchy, same_exec, logged, blockers, sk, sport, dport), @@ -799,7 +819,8 @@ TRACE_EVENT(landlock_deny_access_net, __field( u64, domain_id ) __field( bool, same_exec ) __field( bool, logged ) - __field( access_mask_t, blockers ) + __field( enum landlock_request_type, blockers_type ) + __field( access_mask_t, blockers_access ) __field( u64, sport ) __field( u64, dport ) ), @@ -808,14 +829,17 @@ TRACE_EVENT(landlock_deny_access_net, __entry->domain_id = hierarchy->id; __entry->same_exec = same_exec; __entry->logged = logged; - __entry->blockers = blockers; + __entry->blockers_type = blockers->type; + __entry->blockers_access = blockers->access; __entry->sport = sport; __entry->dport = dport; ), TP_printk("domain=%llx same_exec=%d logged=%d blockers=%s sport=%llu dport=%llu", __entry->domain_id, __entry->same_exec, __entry->logged, - __print_flags(__entry->blockers, "|", _LANDLOCK_ACCESS_NET_NAMES), + __entry->blockers_type == LANDLOCK_REQUEST_NET_ACCESS ? + __print_flags(__entry->blockers_access, "|", _LANDLOCK_ACCESS_NET_NAMES) : + "unknown", __entry->sport, __entry->dport) ); @@ -991,6 +1015,7 @@ TRACE_EVENT(landlock_deny_scope_abstract_unix_socket, __get_dynamic_array_len(sun_path) - 1)) ); +#undef _LANDLOCK_FS_BLOCKER_TYPE_NAMES #undef _LANDLOCK_NAME_ENTRY #endif /* _TRACE_LANDLOCK_H */ diff --git a/security/landlock/log.h b/security/landlock/log.h index e0a6e44f3ddd..04f3e241e765 100644 --- a/security/landlock/log.h +++ b/security/landlock/log.h @@ -25,6 +25,11 @@ enum landlock_request_type { LANDLOCK_REQUEST_SCOPE_SIGNAL, }; +struct landlock_blockers { + access_mask_t access; + enum landlock_request_type type; +}; + /* * We should be careful to only use a variable of this type for * landlock_log_denial(). This way, the compiler can remove it entirely if diff --git a/security/landlock/trace.c b/security/landlock/trace.c index 8c21e5de6f0d..58276cc32d3f 100644 --- a/security/landlock/trace.c +++ b/security/landlock/trace.c @@ -61,7 +61,7 @@ void landlock_trace_free_domain(const struct landlock_hierarchy *const hierarchy * * @request: Detail of the user space request. * @youngest_denied: The youngest hierarchy node that denied the access. - * @missing: The set of denied access rights. + * @missing: The final missing access subset, when applicable. * @same_exec: Whether the current task is the same executable that called * landlock_restrict_self() for the denying domain, as computed * by landlock_log_denial(). @@ -83,6 +83,10 @@ void landlock_trace_denial( case LANDLOCK_REQUEST_FS_ACCESS: case LANDLOCK_REQUEST_FS_CHANGE_TOPOLOGY: if (trace_landlock_deny_access_fs_enabled()) { + const struct landlock_blockers blockers = { + .access = missing, + .type = request->type, + }; char *buf __free(__putname) = __getname(); struct path dentry_path; const char *pathname; @@ -147,16 +151,23 @@ void landlock_trace_denial( trace_landlock_deny_access_fs(youngest_denied, same_exec, logged, - missing, path, pathname); + &blockers, path, + pathname); } break; case LANDLOCK_REQUEST_NET_ACCESS: - if (trace_landlock_deny_access_net_enabled()) + if (trace_landlock_deny_access_net_enabled()) { + const struct landlock_blockers blockers = { + .access = missing, + .type = request->type, + }; + trace_landlock_deny_access_net( - youngest_denied, same_exec, logged, missing, + youngest_denied, same_exec, logged, &blockers, request->audit.u.net->sk, ntohs(request->audit.u.net->sport), ntohs(request->audit.u.net->dport)); + } break; case LANDLOCK_REQUEST_PTRACE: if (trace_landlock_deny_ptrace_enabled()) From 1a985d3890ed8caa428390a5682e957503f14060 Mon Sep 17 00:00:00 2001 From: =?UTF-8?q?Micka=C3=ABl=20Sala=C3=BCn?= Date: Fri, 18 Sep 2026 20:50:26 +0200 Subject: [PATCH 426/592] landlock: Fix rule tracepoint context MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit Name each event after the identity it reports. Add-rule events describe UAPI rule insertion, so rename them after LANDLOCK_RULE_PATH_BENEATH and LANDLOCK_RULE_NET_PORT. Check-rule events describe matches in internal rule trees, so rename them after LANDLOCK_KEY_INODE and LANDLOCK_KEY_NET_PORT. This remains accurate if multiple UAPI rule types share one lookup and stored rule. Keep denial event names based on filesystem and network families because they describe final access decisions. Use u64 for growable access masks passed by value to add-rule and check-rule typed BTF callbacks. CO-RE can relocate pointer-reached fields, but it cannot widen a scalar callback slot declared by a BPF program. Keep native access_mask_t for internal state and trace records. For add-rule callbacks, report the normalized per-call contribution passed to landlock_insert_rule() and expose the complete validated flags value. Put the ruleset and flags first as a common invocation prefix. This distinguishes duplicate and effective-zero additions without recovering arguments from saved syscall registers. Cc: Günther Noack Cc: Steven Rostedt Fixes: 63747c94774d ("landlock: Add landlock_add_rule_fs and landlock_add_rule_net tracepoints") Fixes: 3f1f106e4c14 ("landlock: Add tracepoints for rule checking") Link: https://patch.msgid.link/20260918185036.608651-4-mic@digikod.net Signed-off-by: Mickaël Salaün --- Documentation/trace/events-landlock.rst | 22 +++---- include/trace/events/landlock.h | 62 +++++++++++-------- security/landlock/fs.c | 10 +-- security/landlock/net.c | 5 +- tools/testing/selftests/landlock/fs_test.c | 28 ++++----- tools/testing/selftests/landlock/net_test.c | 11 ++-- tools/testing/selftests/landlock/trace.h | 56 ++++++++--------- .../selftests/landlock/trace_fs_test.c | 54 +++++++++------- tools/testing/selftests/landlock/trace_test.c | 51 ++++++++------- 9 files changed, 162 insertions(+), 137 deletions(-) diff --git a/Documentation/trace/events-landlock.rst b/Documentation/trace/events-landlock.rst index af9267cca47d..9bb81a5c676e 100644 --- a/Documentation/trace/events-landlock.rst +++ b/Documentation/trace/events-landlock.rst @@ -6,7 +6,7 @@ Landlock Trace Events ===================== :Author: Mickaël Salaün -:Date: August 2026 +:Date: September 2026 Landlock emits trace events for sandbox lifecycle operations and access denials. These events can be consumed by ftrace (for human-readable @@ -31,8 +31,8 @@ Landlock trace events are organized in four categories: **Syscall events** are emitted during Landlock system calls: - ``landlock_create_ruleset``: a new ruleset is created -- ``landlock_add_rule_fs``: a filesystem rule is added to a ruleset -- ``landlock_add_rule_net``: a network port rule is added to a ruleset +- ``landlock_add_rule_path_beneath``: a filesystem rule is added to a ruleset +- ``landlock_add_rule_net_port``: a network port rule is added to a ruleset - ``landlock_create_domain``: a new domain is created from a ruleset - ``landlock_enforce_domain``: a domain is enforced on a thread @@ -47,8 +47,8 @@ Landlock trace events are organized in four categories: **Rule evaluation events** are emitted during rule matching: -- ``landlock_check_rule_fs``: a filesystem rule is evaluated -- ``landlock_check_rule_net``: a network port rule is evaluated +- ``landlock_check_rule_inode``: an inode-keyed rule is evaluated +- ``landlock_check_rule_net_port``: a network-port-keyed rule is evaluated **Lifecycle events**: @@ -189,7 +189,7 @@ rather than the caller's, so correlate those to the syscall by domain ID. Interpreting check_rule events ============================== -The ``check_rule_fs`` and ``check_rule_net`` events expose the per-layer +The ``check_rule_inode`` and ``check_rule_net_port`` events expose the per-layer rule evaluation, which is useful for understanding *why* a specific access is allowed or denied. @@ -244,22 +244,22 @@ check. For example, a program sandboxed with read and execute access to the whole filesystem reads ``/etc/passwd``; both the ``execve()`` and the -read match the rule covering ``/`` (inode 2), so ``check_rule_fs`` fires +read match the rule covering ``/`` (inode 2), so ``check_rule_inode`` fires with the requested rights intersected against what that rule grants. The ``access_request=`` mask includes ``truncate`` because the file-open hook evaluates that optional right alongside the required access, but the rule does not grant it, so ``truncate`` never appears in ``grants=``:: - cat-127 [...] landlock_check_rule_fs: domain=1e40cb56f access_request=execute|read_file|truncate dev=0:17 ino=2 grants={execute|read_file} - cat-127 [...] landlock_check_rule_fs: domain=1e40cb56f access_request=read_file|truncate dev=0:17 ino=2 grants={read_file} + cat-127 [...] landlock_check_rule_inode: domain=1e40cb56f access_request=execute|read_file|truncate dev=0:17 ino=2 grants={execute|read_file} + cat-127 [...] landlock_check_rule_inode: domain=1e40cb56f access_request=read_file|truncate dev=0:17 ino=2 grants={read_file} The ``[...]`` replaces the ftrace CPU, flags, and timestamp columns. A single ``grants=`` group means the enforcing domain has one layer. With two nested sandboxes that each grant the same rights, the rule spans both layers, so ``grants=`` has one group per layer:: - cat-128 [...] landlock_check_rule_fs: domain=184788b52 access_request=execute|read_file|truncate dev=0:17 ino=2 grants={execute|read_file,execute|read_file} - cat-128 [...] landlock_check_rule_fs: domain=184788b52 access_request=read_file|truncate dev=0:17 ino=2 grants={read_file,read_file} + cat-128 [...] landlock_check_rule_inode: domain=184788b52 access_request=execute|read_file|truncate dev=0:17 ino=2 grants={execute|read_file,execute|read_file} + cat-128 [...] landlock_check_rule_inode: domain=184788b52 access_request=read_file|truncate dev=0:17 ino=2 grants={read_file,read_file} eBPF access =========== diff --git a/include/trace/events/landlock.h b/include/trace/events/landlock.h index 4e304cab1dab..b7e6a6638822 100644 --- a/include/trace/events/landlock.h +++ b/include/trace/events/landlock.h @@ -28,6 +28,8 @@ struct path; struct sock; struct task_struct; +static_assert(sizeof(access_mask_t) <= sizeof(u64)); + TRACE_DEFINE_ENUM(LANDLOCK_REQUEST_FS_CHANGE_TOPOLOGY); TRACE_DEFINE_ENUM(LANDLOCK_REQUEST_FS_ACCESS); TRACE_DEFINE_ENUM(LANDLOCK_REQUEST_NET_ACCESS); @@ -114,7 +116,7 @@ __trace_print_untrusted_str(struct trace_seq *p, const char *src, size_t len) * Fills the dense per-domain-layer array layers (one access mask per layer, * indexed by level - 1) from rule's sparse layer stack, keeping only the * requested rights (access_request). Layers with no matching rule entry get - * a zero mask. Shared by the check_rule_fs and check_rule_net events. + * a zero mask. Shared by the check_rule_inode and check_rule_net_port events. * * rule->layers is sorted by ascending level, with levels in the domain's * [1, num_layers] range (see landlock_merge_ruleset()), so every entry maps @@ -389,12 +391,14 @@ TRACE_EVENT(landlock_free_ruleset, ); /** - * landlock_add_rule_fs - Filesystem rule added to a ruleset + * landlock_add_rule_path_beneath - Path-beneath rule added to a ruleset * * @ruleset: Source ruleset (never NULL). - * @access_rights: Effective access mask stored in the rule, not the raw - * sys_landlock_add_rule() argument (unhandled rights - * added). + * @flags: Complete validated landlock_add_rule_flags value supplied by this + * successful call, not the rule's accumulated quiet state. + * @access_rights: Canonical per-call access mask passed to + * landlock_insert_rule() after normalization, not the raw + * sys_landlock_add_rule() argument or accumulated rule. * @path: Filesystem path for the rule (never NULL). * @pathname: Resolved absolute path string (never NULL; error placeholder * on resolution failure). @@ -403,13 +407,13 @@ TRACE_EVENT(landlock_free_ruleset, * the reported ruleset is a stable snapshot that no concurrent writer can * change. */ -TRACE_EVENT(landlock_add_rule_fs, +TRACE_EVENT(landlock_add_rule_path_beneath, - TP_PROTO(const struct landlock_ruleset *ruleset, - access_mask_t access_rights, const struct path *path, + TP_PROTO(const struct landlock_ruleset *ruleset, u32 flags, + u64 access_rights, const struct path *path, const char *pathname), - TP_ARGS(ruleset, access_rights, path, pathname), + TP_ARGS(ruleset, flags, access_rights, path, pathname), TP_STRUCT__entry( __field( u64, ruleset_id ) @@ -443,25 +447,27 @@ TRACE_EVENT(landlock_add_rule_fs, ); /** - * landlock_add_rule_net - Network port rule added to a ruleset + * landlock_add_rule_net_port - Network-port rule added to a ruleset * * @ruleset: Source ruleset (never NULL). - * @access_rights: Effective access mask stored in the rule, not the raw - * sys_landlock_add_rule() argument (unhandled rights - * added). - * @port: Network port, the landlock_net_port_attr.port UAPI value - * forwarded directly. + * @flags: Complete validated landlock_add_rule_flags value supplied by this + * successful call, not the rule's accumulated quiet state. + * @access_rights: Canonical per-call access mask passed to + * landlock_insert_rule() after normalization, not the raw + * sys_landlock_add_rule() argument or accumulated rule. + * @port: Network port in host endianness, forwarded directly from + * &landlock_net_port_attr.port. * * Emitted by sys_landlock_add_rule() under the modified ruleset's lock, so * the reported ruleset is a stable snapshot that no concurrent writer can * change. */ -TRACE_EVENT(landlock_add_rule_net, +TRACE_EVENT(landlock_add_rule_net_port, - TP_PROTO(const struct landlock_ruleset *ruleset, - access_mask_t access_rights, u64 port), + TP_PROTO(const struct landlock_ruleset *ruleset, u32 flags, + u64 access_rights, u64 port), - TP_ARGS(ruleset, access_rights, port), + TP_ARGS(ruleset, flags, access_rights, port), TP_STRUCT__entry( __field( u64, ruleset_id ) @@ -625,7 +631,7 @@ TRACE_EVENT(landlock_free_domain, ); /** - * landlock_check_rule_fs - Filesystem rule evaluated during access check + * landlock_check_rule_inode - Inode rule evaluated during access check * * @domain: Enforcing domain (never NULL). * @rule: Matching rule with per-layer access masks (never NULL). @@ -638,11 +644,11 @@ TRACE_EVENT(landlock_free_domain, * domain layer. See Documentation/trace/events-landlock.rst for how to * interpret it. */ -TRACE_EVENT(landlock_check_rule_fs, +TRACE_EVENT(landlock_check_rule_inode, TP_PROTO(const struct landlock_domain *domain, const struct landlock_rule *rule, - access_mask_t access_request, const struct dentry *dentry), + u64 access_request, const struct dentry *dentry), TP_ARGS(domain, rule, access_request, dentry), @@ -664,7 +670,8 @@ TRACE_EVENT(landlock_check_rule_fs, __trace_landlock_fill_layers(__get_dynamic_array(grants), __get_dynamic_array_len(grants) / sizeof(access_mask_t), - rule, access_request); + rule, + (access_mask_t)access_request); ), TP_printk("domain=%llx access_request=%s dev=%u:%u ino=%lu grants=%s", @@ -675,7 +682,7 @@ TRACE_EVENT(landlock_check_rule_fs, ); /** - * landlock_check_rule_net - Network port rule evaluated during access check + * landlock_check_rule_net_port - Network port rule evaluated * * @domain: Enforcing domain (never NULL). * @rule: Matching rule with per-layer access masks (never NULL). @@ -687,11 +694,11 @@ TRACE_EVENT(landlock_check_rule_fs, * layer. See Documentation/trace/events-landlock.rst for how to * interpret it. */ -TRACE_EVENT(landlock_check_rule_net, +TRACE_EVENT(landlock_check_rule_net_port, TP_PROTO(const struct landlock_domain *domain, const struct landlock_rule *rule, - access_mask_t access_request, u64 port), + u64 access_request, u64 port), TP_ARGS(domain, rule, access_request, port), @@ -711,7 +718,8 @@ TRACE_EVENT(landlock_check_rule_net, __trace_landlock_fill_layers(__get_dynamic_array(grants), __get_dynamic_array_len(grants) / sizeof(access_mask_t), - rule, access_request); + rule, + (access_mask_t)access_request); ), TP_printk("domain=%llx access_request=%s port=%llu grants=%s", diff --git a/security/landlock/fs.c b/security/landlock/fs.c index 330a1871bf94..cab43892ec2f 100644 --- a/security/landlock/fs.c +++ b/security/landlock/fs.c @@ -356,14 +356,14 @@ int landlock_append_fs_rule(struct landlock_ruleset *const ruleset, * held for BTF consistency (enforced by lockdep_assert_held in * TP_fast_assign). */ - if (!err && trace_landlock_add_rule_fs_enabled()) { + if (!err && trace_landlock_add_rule_path_beneath_enabled()) { char *buffer __free(__putname) = __getname(); const char *pathname = buffer ? resolve_path_for_trace(path, buffer) : ""; - trace_landlock_add_rule_fs(ruleset, access_rights, path, - pathname); + trace_landlock_add_rule_path_beneath( + ruleset, flags, access_rights, path, pathname); } mutex_unlock(&ruleset->lock); @@ -423,8 +423,8 @@ static bool unmask_layers_fs(const struct landlock_domain *const domain, ret = landlock_unmask_layers(domain, id, masks, &rule); if (rule) - trace_landlock_check_rule_fs(domain, rule, access_request, - dentry); + trace_landlock_check_rule_inode(domain, rule, access_request, + dentry); return ret; } diff --git a/security/landlock/net.c b/security/landlock/net.c index 8f2aaac54b33..5552c60388f8 100644 --- a/security/landlock/net.c +++ b/security/landlock/net.c @@ -47,7 +47,8 @@ int landlock_append_net_rule(struct landlock_ruleset *const ruleset, * TP_fast_assign). */ if (!err) - trace_landlock_add_rule_net(ruleset, access_rights, port); + trace_landlock_add_rule_net_port(ruleset, flags, access_rights, + port); mutex_unlock(&ruleset->lock); return err; @@ -63,7 +64,7 @@ static bool unmask_layers_net(const struct landlock_domain *const domain, ret = landlock_unmask_layers(domain, id, masks, &rule); if (rule) - trace_landlock_check_rule_net( + trace_landlock_check_rule_net_port( domain, rule, access_request, ntohs((__force __be16)id.key.data)); return ret; diff --git a/tools/testing/selftests/landlock/fs_test.c b/tools/testing/selftests/landlock/fs_test.c index 18dbdb99aeba..6e979cef884d 100644 --- a/tools/testing/selftests/landlock/fs_test.c +++ b/tools/testing/selftests/landlock/fs_test.c @@ -10493,9 +10493,9 @@ FIXTURE_TEARDOWN_PARENT(trace_layout1) } /* - * Verifies that check_rule_fs events include correct field values: domain, dev, - * ino, access_request, and grants. All values are verified against stat() of - * the rule path on a deterministic tmpfs layout. + * Verifies that check_rule_inode events include correct field values: domain, + * dev, ino, access_request, and grants. All values are verified against stat() + * of the rule path on a deterministic tmpfs layout. */ TEST_F(trace_layout1, check_rule_fs_fields) { @@ -10529,7 +10529,7 @@ TEST_F(trace_layout1, check_rule_fs_fields) EXPECT_EQ(1, tracefs_count_matches(buf, REGEX_CHECK_RULE_FS(TRACE_TASK))) { - TH_LOG("Expected 1 check_rule_fs event\n%s", buf); + TH_LOG("Expected 1 check_rule_inode event\n%s", buf); } ASSERT_EQ(0, tracefs_extract_field(buf, REGEX_CHECK_RULE_FS(TRACE_TASK), @@ -10570,8 +10570,8 @@ TEST_F(trace_layout1, check_rule_fs_fields) } /* - * Verifies check_rule_fs behavior with multiple rules. With rules at s1d1 and - * s1d2 (a child of s1d1), accessing s1d2 produces only 1 event because the + * Verifies check_rule_inode behavior with multiple rules. With rules at s1d1 + * and s1d2 (a child of s1d1), accessing s1d2 produces only 1 event because the * pathwalk short-circuits after the first rule fully unmasks the single layer. */ TEST_F(trace_layout1, check_rule_fs_multiple_rules) @@ -10643,14 +10643,14 @@ TEST_F(trace_layout1, check_rule_fs_multiple_rules) ASSERT_NE(NULL, buf); /* - * Only 1 check_rule_fs event: the rule on dir_s1d2 fully unmasked the - * single layer, so the pathwalk short-circuits before reaching the + * Only one check_rule_inode event: the rule on dir_s1d2 fully unmasks + * the single layer, so the pathwalk short-circuits before reaching the * dir_s1d1 rule. */ count = tracefs_count_matches(buf, REGEX_CHECK_RULE_FS(TRACE_TASK)); EXPECT_EQ(1, count) { - TH_LOG("Expected 1 check_rule_fs event, got %d\n%s", count, + TH_LOG("Expected 1 check_rule_inode event, got %d\n%s", count, buf); } @@ -10777,7 +10777,7 @@ TEST_F(trace_layout1, check_rule_fs_optional_access) count = tracefs_count_matches(buf, REGEX_CHECK_RULE_FS(TRACE_TASK)); EXPECT_EQ(1, count) { - TH_LOG("Expected 1 check_rule_fs event, got %d\n%s", count, + TH_LOG("Expected 1 check_rule_inode event, got %d\n%s", count, buf); } @@ -10796,7 +10796,7 @@ TEST_F(trace_layout1, check_rule_fs_optional_access) } /* - * Verifies that check_rule_fs fires for a rule that matches the inode even when + * Verifies that check_rule_inode fires for a rule matching the inode even when * it grants none of the requested rights, so the grants set is empty. Landlock * cannot know a rule ignores the request before reading it, so the event is * still emitted (grants={}), which lets a tracer see that the rule matched. @@ -10884,7 +10884,7 @@ TEST_F(trace_layout1, check_rule_fs_empty_grant) count = tracefs_count_matches(buf, REGEX_CHECK_RULE_FS(TRACE_TASK)); EXPECT_EQ(2, count) { - TH_LOG("Expected 2 check_rule_fs events, got %d\n%s", count, + TH_LOG("Expected 2 check_rule_inode events, got %d\n%s", count, buf); } @@ -10894,7 +10894,7 @@ TEST_F(trace_layout1, check_rule_fs_empty_grant) tracefs_count_matches( buf, TRACE_PREFIX( - TRACE_TASK) "landlock_check_rule_fs: domain=[0-9a-f]\\+ " + TRACE_TASK) "landlock_check_rule_inode: domain=[0-9a-f]\\+ " "access_request=read_dir " "dev=[0-9]\\+:[0-9]\\+ ino=[0-9]\\+ " "grants={}$")) @@ -10908,7 +10908,7 @@ TEST_F(trace_layout1, check_rule_fs_empty_grant) tracefs_count_matches( buf, TRACE_PREFIX( - TRACE_TASK) "landlock_check_rule_fs: domain=[0-9a-f]\\+ " + TRACE_TASK) "landlock_check_rule_inode: domain=[0-9a-f]\\+ " "access_request=read_dir " "dev=[0-9]\\+:[0-9]\\+ ino=[0-9]\\+ " "grants={read_dir}$")) diff --git a/tools/testing/selftests/landlock/net_test.c b/tools/testing/selftests/landlock/net_test.c index a18761e0fd82..4fb705e1596d 100644 --- a/tools/testing/selftests/landlock/net_test.c +++ b/tools/testing/selftests/landlock/net_test.c @@ -3712,7 +3712,7 @@ TEST_F(trace_net_connect, deny_access_net) free(buf); } -/* Field verification for the check_rule_net event on an allowed access. */ +/* Field verification for the check_rule_net_port event on an allowed access. */ /* clang-format off */ FIXTURE(trace_net_check_rule) { @@ -3757,10 +3757,11 @@ FIXTURE_TEARDOWN(trace_net_check_rule) /* * Verifies that an allowed bind matching a net-port rule emits exactly one - * landlock_check_rule_net event with the enforcing domain, the requested + * landlock_check_rule_net_port event with the enforcing domain, the requested * access, the checked port (host endianness), and the per-layer grants. The - * whole event is anchored to exact values so a revert of the check_rule_net - * emit (or a byte-order or field-plumbing regression) fails the test. + * whole event is anchored to exact values so removing the check_rule_net_port + * emission or introducing a byte-order or field-plumbing regression fails the + * test. */ TEST_F(trace_net_check_rule, check_rule_net_fields) { @@ -3832,7 +3833,7 @@ TEST_F(trace_net_check_rule, check_rule_net_fields) EXPECT_EQ(1, tracefs_count_matches(buf, REGEX_CHECK_RULE_NET(TRACE_TASK))) { - TH_LOG("Expected 1 check_rule_net event\n%s", buf); + TH_LOG("Expected 1 check_rule_net_port event\n%s", buf); } ASSERT_EQ(0, diff --git a/tools/testing/selftests/landlock/trace.h b/tools/testing/selftests/landlock/trace.h index ba0c5e92001f..e6873853376c 100644 --- a/tools/testing/selftests/landlock/trace.h +++ b/tools/testing/selftests/landlock/trace.h @@ -27,14 +27,14 @@ TRACEFS_LANDLOCK_DIR "/landlock_create_domain/enable" #define TRACEFS_ENFORCE_DOMAIN_ENABLE \ TRACEFS_LANDLOCK_DIR "/landlock_enforce_domain/enable" -#define TRACEFS_ADD_RULE_FS_ENABLE \ - TRACEFS_LANDLOCK_DIR "/landlock_add_rule_fs/enable" -#define TRACEFS_ADD_RULE_NET_ENABLE \ - TRACEFS_LANDLOCK_DIR "/landlock_add_rule_net/enable" +#define TRACEFS_ADD_RULE_PATH_BENEATH_ENABLE \ + TRACEFS_LANDLOCK_DIR "/landlock_add_rule_path_beneath/enable" +#define TRACEFS_ADD_RULE_NET_PORT_ENABLE \ + TRACEFS_LANDLOCK_DIR "/landlock_add_rule_net_port/enable" #define TRACEFS_CHECK_RULE_FS_ENABLE \ - TRACEFS_LANDLOCK_DIR "/landlock_check_rule_fs/enable" + TRACEFS_LANDLOCK_DIR "/landlock_check_rule_inode/enable" #define TRACEFS_CHECK_RULE_NET_ENABLE \ - TRACEFS_LANDLOCK_DIR "/landlock_check_rule_net/enable" + TRACEFS_LANDLOCK_DIR "/landlock_check_rule_net_port/enable" #define TRACEFS_DENY_ACCESS_FS_ENABLE \ TRACEFS_LANDLOCK_DIR "/landlock_deny_access_fs/enable" #define TRACEFS_DENY_ACCESS_NET_ENABLE \ @@ -79,18 +79,18 @@ */ #define KWORKER_TASK "kworker/[0-9]\\+:[0-9]\\+" -#define REGEX_ADD_RULE_FS(task) \ - TRACE_PREFIX(task) \ - "landlock_add_rule_fs: " \ - "ruleset=[0-9a-f]\\+\\.[0-9]\\+ " \ - "access_rights=[a-z_|]* " \ - "dev=[0-9]\\+:[0-9]\\+ " \ - "ino=[0-9]\\+ " \ +#define REGEX_ADD_RULE_PATH_BENEATH(task) \ + TRACE_PREFIX(task) \ + "landlock_add_rule_path_beneath: " \ + "ruleset=[0-9a-f]\\+\\.[0-9]\\+ " \ + "access_rights=[a-z_|]* " \ + "dev=[0-9]\\+:[0-9]\\+ " \ + "ino=[0-9]\\+ " \ "path=[^ ]\\+$" -#define REGEX_ADD_RULE_NET(task) \ +#define REGEX_ADD_RULE_NET_PORT(task) \ TRACE_PREFIX(task) \ - "landlock_add_rule_net: " \ + "landlock_add_rule_net_port: " \ "ruleset=[0-9a-f]\\+\\.[0-9]\\+ " \ "access_rights=[a-z_|]* " \ "port=[0-9]\\+$" @@ -110,21 +110,21 @@ "parent=[0-9a-f]\\+ " \ "ruleset=[0-9a-f]\\+\\.[0-9]\\+$" -#define REGEX_CHECK_RULE_FS(task) \ - TRACE_PREFIX(task) \ - "landlock_check_rule_fs: " \ - "domain=[0-9a-f]\\+ " \ - "access_request=[a-z_|]* " \ - "dev=[0-9]\\+:[0-9]\\+ " \ - "ino=[0-9]\\+ " \ +#define REGEX_CHECK_RULE_FS(task) \ + TRACE_PREFIX(task) \ + "landlock_check_rule_inode: " \ + "domain=[0-9a-f]\\+ " \ + "access_request=[a-z_|]* " \ + "dev=[0-9]\\+:[0-9]\\+ " \ + "ino=[0-9]\\+ " \ "grants={[a-z_|,]*}$" -#define REGEX_CHECK_RULE_NET(task) \ - TRACE_PREFIX(task) \ - "landlock_check_rule_net: " \ - "domain=[0-9a-f]\\+ " \ - "access_request=[a-z_|]* " \ - "port=[0-9]\\+ " \ +#define REGEX_CHECK_RULE_NET(task) \ + TRACE_PREFIX(task) \ + "landlock_check_rule_net_port: " \ + "domain=[0-9a-f]\\+ " \ + "access_request=[a-z_|]* " \ + "port=[0-9]\\+ " \ "grants={[a-z_|,]*}$" #define REGEX_DENY_ACCESS_FS(task) \ diff --git a/tools/testing/selftests/landlock/trace_fs_test.c b/tools/testing/selftests/landlock/trace_fs_test.c index 4543a25c1f55..6666d4746cb1 100644 --- a/tools/testing/selftests/landlock/trace_fs_test.c +++ b/tools/testing/selftests/landlock/trace_fs_test.c @@ -121,7 +121,8 @@ FIXTURE_SETUP(trace_fs) } self->tracefs_ok = 1; - ASSERT_EQ(0, tracefs_enable_event(TRACEFS_ADD_RULE_FS_ENABLE, true)); + ASSERT_EQ(0, tracefs_enable_event(TRACEFS_ADD_RULE_PATH_BENEATH_ENABLE, + true)); ASSERT_EQ(0, tracefs_enable_event(TRACEFS_CHECK_RULE_FS_ENABLE, true)); ASSERT_EQ(0, tracefs_enable_event(TRACEFS_DENY_ACCESS_FS_ENABLE, true)); ASSERT_EQ(0, tracefs_clear()); @@ -134,7 +135,7 @@ FIXTURE_TEARDOWN(trace_fs) return; set_cap(_metadata, CAP_SYS_ADMIN); - tracefs_enable_event(TRACEFS_ADD_RULE_FS_ENABLE, false); + tracefs_enable_event(TRACEFS_ADD_RULE_PATH_BENEATH_ENABLE, false); tracefs_enable_event(TRACEFS_CHECK_RULE_FS_ENABLE, false); tracefs_enable_event(TRACEFS_DENY_ACCESS_FS_ENABLE, false); tracefs_fixture_teardown(); @@ -183,11 +184,11 @@ TEST_F(trace_fs, unsandboxed) } /* - * Verifies that adding a filesystem rule emits a landlock_add_rule_fs trace - * event with the expected path and field values: ruleset ID is non-zero, - * access_rights is non-zero, and path matches. + * Verifies that adding a filesystem rule emits a landlock_add_rule_path_beneath + * event with the expected path and field values: the ruleset ID and + * access_rights are non-zero, and the path matches. */ -TEST_F(trace_fs, add_rule_fs) +TEST_F(trace_fs, add_rule_path_beneath) { struct landlock_ruleset_attr ruleset_attr = { .handled_access_fs = LANDLOCK_ACCESS_FS_READ_FILE | @@ -215,28 +216,30 @@ TEST_F(trace_fs, add_rule_fs) buf = tracefs_read_buf(); ASSERT_NE(NULL, buf); - count = tracefs_count_matches(buf, REGEX_ADD_RULE_FS(TRACE_TASK)); + count = tracefs_count_matches(buf, + REGEX_ADD_RULE_PATH_BENEATH(TRACE_TASK)); EXPECT_EQ(1, count) { - TH_LOG("Expected 1 add_rule_fs event, got %d\n%s", count, buf); + TH_LOG("Expected 1 add_rule_path_beneath event, got %d\n%s", + count, buf); } /* Ruleset ID should be non-zero. */ - ASSERT_EQ(0, tracefs_extract_field(buf, REGEX_ADD_RULE_FS(TRACE_TASK), - "ruleset", field_buf, - sizeof(field_buf))); + ASSERT_EQ(0, tracefs_extract_field( + buf, REGEX_ADD_RULE_PATH_BENEATH(TRACE_TASK), + "ruleset", field_buf, sizeof(field_buf))); EXPECT_STRNE("0", field_buf); /* Access rights should be non-zero. */ - ASSERT_EQ(0, tracefs_extract_field(buf, REGEX_ADD_RULE_FS(TRACE_TASK), - "access_rights", field_buf, - sizeof(field_buf))); + ASSERT_EQ(0, tracefs_extract_field( + buf, REGEX_ADD_RULE_PATH_BENEATH(TRACE_TASK), + "access_rights", field_buf, sizeof(field_buf))); EXPECT_STRNE("", field_buf); /* Path should be /usr. */ - ASSERT_EQ(0, - tracefs_extract_field(buf, REGEX_ADD_RULE_FS(TRACE_TASK), - "path", field_buf, sizeof(field_buf))); + ASSERT_EQ(0, tracefs_extract_field( + buf, REGEX_ADD_RULE_PATH_BENEATH(TRACE_TASK), + "path", field_buf, sizeof(field_buf))); EXPECT_STREQ("/usr", field_buf); free(buf); @@ -246,7 +249,7 @@ TEST_F(trace_fs, add_rule_fs) * Verifies that a path whose escaping exceeds the trace scratch sequence does * not corrupt a sibling symbolic field. */ -TEST_F(trace_fs, add_rule_fs_escaped_path_overflow) +TEST_F(trace_fs, add_rule_path_beneath_escaped_path_overflow) { static const char access_prefix[] = "execute|write_file|read_file|"; static const char access_suffix[] = "|ioctl_dev|resolve_unix"; @@ -277,10 +280,12 @@ TEST_F(trace_fs, add_rule_fs_escaped_path_overflow) buf = tracefs_read_buf(); ASSERT_NE(NULL, buf); - count = tracefs_count_matches(buf, REGEX_ADD_RULE_FS(TRACE_TASK)); + count = tracefs_count_matches(buf, + REGEX_ADD_RULE_PATH_BENEATH(TRACE_TASK)); EXPECT_EQ(1, count) { - TH_LOG("Expected 1 add_rule_fs event, got %d\n%s", count, buf); + TH_LOG("Expected 1 add_rule_path_beneath event, got %d\n%s", + count, buf); } /* @@ -288,9 +293,9 @@ TEST_F(trace_fs, add_rule_fs_escaped_path_overflow) * field also catches scratch-sequence poisoning when the compiler * evaluates the overflowing path first, as GCC currently does. */ - ASSERT_EQ(0, tracefs_extract_field(buf, REGEX_ADD_RULE_FS(TRACE_TASK), - "access_rights", field_buf, - sizeof(field_buf))); + ASSERT_EQ(0, tracefs_extract_field( + buf, REGEX_ADD_RULE_PATH_BENEATH(TRACE_TASK), + "access_rights", field_buf, sizeof(field_buf))); EXPECT_EQ(0, strncmp(field_buf, access_prefix, sizeof(access_prefix) - 1)); EXPECT_EQ(NULL, strstr(field_buf, "|refer|")); @@ -298,7 +303,8 @@ TEST_F(trace_fs, add_rule_fs_escaped_path_overflow) ASSERT_LE(sizeof(access_suffix) - 1, field_len); EXPECT_STREQ(access_suffix, field_buf + field_len - (sizeof(access_suffix) - 1)); - expect_truncated_path(_metadata, buf, REGEX_ADD_RULE_FS(TRACE_TASK)); + expect_truncated_path(_metadata, buf, + REGEX_ADD_RULE_PATH_BENEATH(TRACE_TASK)); free(buf); } diff --git a/tools/testing/selftests/landlock/trace_test.c b/tools/testing/selftests/landlock/trace_test.c index afdaf8511b3a..f9b293a9dd56 100644 --- a/tools/testing/selftests/landlock/trace_test.c +++ b/tools/testing/selftests/landlock/trace_test.c @@ -49,8 +49,10 @@ FIXTURE_SETUP(trace) ASSERT_EQ(0, tracefs_enable_event(TRACEFS_CREATE_RULESET_ENABLE, true)); ASSERT_EQ(0, tracefs_enable_event(TRACEFS_CREATE_DOMAIN_ENABLE, true)); ASSERT_EQ(0, tracefs_enable_event(TRACEFS_ENFORCE_DOMAIN_ENABLE, true)); - ASSERT_EQ(0, tracefs_enable_event(TRACEFS_ADD_RULE_FS_ENABLE, true)); - ASSERT_EQ(0, tracefs_enable_event(TRACEFS_ADD_RULE_NET_ENABLE, true)); + ASSERT_EQ(0, tracefs_enable_event(TRACEFS_ADD_RULE_PATH_BENEATH_ENABLE, + true)); + ASSERT_EQ(0, + tracefs_enable_event(TRACEFS_ADD_RULE_NET_PORT_ENABLE, true)); ASSERT_EQ(0, tracefs_enable_event(TRACEFS_CHECK_RULE_FS_ENABLE, true)); ASSERT_EQ(0, tracefs_enable_event(TRACEFS_CHECK_RULE_NET_ENABLE, true)); ASSERT_EQ(0, tracefs_enable_event(TRACEFS_DENY_ACCESS_FS_ENABLE, true)); @@ -72,8 +74,8 @@ FIXTURE_TEARDOWN(trace) tracefs_enable_event(TRACEFS_CREATE_RULESET_ENABLE, false); tracefs_enable_event(TRACEFS_CREATE_DOMAIN_ENABLE, false); tracefs_enable_event(TRACEFS_ENFORCE_DOMAIN_ENABLE, false); - tracefs_enable_event(TRACEFS_ADD_RULE_FS_ENABLE, false); - tracefs_enable_event(TRACEFS_ADD_RULE_NET_ENABLE, false); + tracefs_enable_event(TRACEFS_ADD_RULE_PATH_BENEATH_ENABLE, false); + tracefs_enable_event(TRACEFS_ADD_RULE_NET_PORT_ENABLE, false); tracefs_enable_event(TRACEFS_CHECK_RULE_FS_ENABLE, false); tracefs_enable_event(TRACEFS_CHECK_RULE_NET_ENABLE, false); tracefs_enable_event(TRACEFS_DENY_ACCESS_FS_ENABLE, false); @@ -103,8 +105,10 @@ TEST_F(trace, no_trace_when_disabled) ASSERT_EQ(0, tracefs_enable_event(TRACEFS_CREATE_DOMAIN_ENABLE, false)); ASSERT_EQ(0, tracefs_enable_event(TRACEFS_ENFORCE_DOMAIN_ENABLE, false)); - ASSERT_EQ(0, tracefs_enable_event(TRACEFS_ADD_RULE_FS_ENABLE, false)); - ASSERT_EQ(0, tracefs_enable_event(TRACEFS_ADD_RULE_NET_ENABLE, false)); + ASSERT_EQ(0, tracefs_enable_event(TRACEFS_ADD_RULE_PATH_BENEATH_ENABLE, + false)); + ASSERT_EQ(0, tracefs_enable_event(TRACEFS_ADD_RULE_NET_PORT_ENABLE, + false)); ASSERT_EQ(0, tracefs_enable_event(TRACEFS_CHECK_RULE_FS_ENABLE, false)); ASSERT_EQ(0, tracefs_enable_event(TRACEFS_CHECK_RULE_NET_ENABLE, false)); @@ -265,10 +269,11 @@ TEST_F(trace, ruleset_version) ASSERT_NE(0, !!dot); EXPECT_STREQ("0", dot + 1); - /* Verify 2 add_rule_fs events were emitted. */ - EXPECT_EQ(2, tracefs_count_matches(buf, REGEX_ADD_RULE_FS(TRACE_TASK))) + /* Verify two add_rule_path_beneath events were emitted. */ + EXPECT_EQ(2, tracefs_count_matches( + buf, REGEX_ADD_RULE_PATH_BENEATH(TRACE_TASK))) { - TH_LOG("Expected 2 add_rule_fs events\n%s", buf); + TH_LOG("Expected 2 add_rule_path_beneath events\n%s", buf); } /* @@ -373,7 +378,7 @@ TEST_F(trace, create_domain) tracefs_count_matches(buf, REGEX_CHECK_RULE_FS(TRACE_TASK)); ASSERT_LE(1, check_count) { - TH_LOG("Expected check_rule_fs events\n%s", buf); + TH_LOG("Expected check_rule_inode events\n%s", buf); } EXPECT_EQ(0, tracefs_extract_field(buf, REGEX_CHECK_RULE_FS(TRACE_TASK), @@ -508,9 +513,11 @@ TEST_F(trace, add_rule_invalid_fd) buf = tracefs_read_buf(); ASSERT_NE(NULL, buf); - EXPECT_EQ(0, tracefs_count_matches(buf, REGEX_ADD_RULE_FS(TRACE_TASK))) + EXPECT_EQ(0, tracefs_count_matches( + buf, REGEX_ADD_RULE_PATH_BENEATH(TRACE_TASK))) { - TH_LOG("No add_rule_fs event expected on invalid fd\n%s", buf); + TH_LOG("No add_rule_path_beneath event expected on invalid fd\n%s", + buf); } free(buf); @@ -902,10 +909,10 @@ TEST_F(trace, non_audit_visible_denial_counting) } /* - * Verifies that landlock_add_rule_net emits a trace event with the correct port - * and allowed access mask fields. + * Verifies that landlock_add_rule_net_port emits a trace event with the correct + * port and allowed access mask fields. */ -TEST_F(trace, add_rule_net_fields) +TEST_F(trace, add_rule_net_port_fields) { struct landlock_ruleset_attr ruleset_attr = { .handled_access_net = LANDLOCK_ACCESS_NET_BIND_TCP, @@ -931,9 +938,10 @@ TEST_F(trace, add_rule_net_fields) buf = tracefs_read_buf(); ASSERT_NE(NULL, buf); - EXPECT_EQ(1, tracefs_count_matches(buf, REGEX_ADD_RULE_NET(TRACE_TASK))) + EXPECT_EQ(1, tracefs_count_matches(buf, + REGEX_ADD_RULE_NET_PORT(TRACE_TASK))) { - TH_LOG("Expected 1 add_rule_net event\n%s", buf); + TH_LOG("Expected 1 add_rule_net_port event\n%s", buf); } /* @@ -941,7 +949,8 @@ TEST_F(trace, add_rule_net_fields) * (landlock_net_port_attr.port). On little-endian, htons(8080) is * 36895, so this comparison catches byte-order bugs. */ - EXPECT_EQ(0, tracefs_extract_field(buf, REGEX_ADD_RULE_NET(TRACE_TASK), + EXPECT_EQ(0, tracefs_extract_field(buf, + REGEX_ADD_RULE_NET_PORT(TRACE_TASK), "port", field, sizeof(field))); EXPECT_STREQ("8080", field); /* @@ -950,9 +959,9 @@ TEST_F(trace, add_rule_net_fields) * net access bits are unhandled because the ruleset only handles * BIND_TCP). */ - EXPECT_EQ(0, - tracefs_extract_field(buf, REGEX_ADD_RULE_NET(TRACE_TASK), - "access_rights", field, sizeof(field))); + EXPECT_EQ(0, tracefs_extract_field( + buf, REGEX_ADD_RULE_NET_PORT(TRACE_TASK), + "access_rights", field, sizeof(field))); EXPECT_STREQ("bind_tcp|connect_tcp|bind_udp|connect_send_udp", field); free(buf); From 98b04ab00f0e738d69bd718228db55483a911b7a Mon Sep 17 00:00:00 2001 From: =?UTF-8?q?Micka=C3=ABl=20Sala=C3=BCn?= Date: Fri, 18 Sep 2026 20:50:27 +0200 Subject: [PATCH 427/592] landlock: Fix network denial trace context MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit Network denial events report source and destination ports reconstructed from audit data. Their zero values are ambiguous, and neither identifies the complete endpoint that Landlock checked. Carry the checked sockaddr and its signed length in a private trace-only context. For an enabled event, validate the length and copy only the initialized prefix into zeroed local storage. This prevents a typed BPF program from reading uninitialized bytes while exposing the socket family, socket, address, and length. Replace the source and destination trace-record fields with one signed port derived from the checked address. A value of -1 means that no port was checked, zero is a valid port, and positive values use host endianness. Bind blockers select the bind address; connect and send blockers select the destination. Cc: Günther Noack Cc: Steven Rostedt Fixes: 01ce260f5ccf ("landlock: Add landlock_deny_access_fs and landlock_deny_access_net") Link: https://patch.msgid.link/20260918185036.608651-5-mic@digikod.net Signed-off-by: Mickaël Salaün --- include/trace/events/landlock.h | 73 +++++++++++++-------- security/landlock/log.h | 32 +++++++-- security/landlock/net.c | 37 ++++++++--- security/landlock/trace.c | 20 +++++- tools/testing/selftests/landlock/net_test.c | 49 +++++--------- tools/testing/selftests/landlock/trace.h | 3 +- 6 files changed, 135 insertions(+), 79 deletions(-) diff --git a/include/trace/events/landlock.h b/include/trace/events/landlock.h index b7e6a6638822..a982a7cfa881 100644 --- a/include/trace/events/landlock.h +++ b/include/trace/events/landlock.h @@ -10,7 +10,10 @@ #if !defined(_TRACE_LANDLOCK_H) || defined(TRACE_HEADER_MULTI_READ) #define _TRACE_LANDLOCK_H +#include +#include #include +#include #include #include #include @@ -790,6 +793,11 @@ TRACE_EVENT(landlock_deny_access_fs, __get_dynamic_array_len(pathname) - 1)) ); +static_assert(offsetof(struct sockaddr_in, sin_port) == + offsetof(struct sockaddr_in6, sin6_port)); +static_assert(sizeof_field(struct sockaddr_in, sin_port) == + sizeof_field(struct sockaddr_in6, sin6_port)); + /** * landlock_deny_access_net - Network access denied * @@ -798,30 +806,31 @@ TRACE_EVENT(landlock_deny_access_fs, * @same_exec: Whether the current task entered the denying domain itself. * @logged: The domain's audit-logging decision for this denial. * @blockers: Request type and final missing access subset (never NULL). - * @sk: Socket object (never NULL), read without a socket lock, so its - * fields are a best-effort snapshot. The denied endpoint is not - * available: the hook runs before :manpage:`bind(2)` / - * :manpage:`connect(2)` sets the socket addresses. - * @sport: Source port in host endianness, set for bind denials (zero for - * an autobind/ephemeral port); zero for connect and send denials. - * @dport: Destination port in host endianness, set for connect and send - * denials; zero for bind denials, and also zero for a UDP send to - * an AF_UNSPEC address on an IPv6 socket (indistinguishable from a - * real destination port 0). The bind-vs-connect direction is - * given by @blockers, not by which port is set. + * @sk: Socket object (never NULL), read without a socket lock, so its fields + * are a best-effort snapshot. + * @socket_family: Socket-family snapshot used by the verdict. + * @address: Authoritative address checked by the verdict (never NULL). + * The producer copies @addrlen bytes from the checked address and + * zeroes the remaining storage before emission. The + * &sockaddr_in.sin_port or &sockaddr_in6.sin6_port member, when + * present, remains in network endianness. + * @addrlen: Validated signed length of @address. * - * Emitted when a Landlock domain denies a network operation. - * - * The port fields are converted from the socket's network byte order to - * host endianness before emitting. + * Emitted when a Landlock domain denies a network operation. The blocker + * identifies whether the address is a bind or connect/send policy object. + * The flattened port field is converted from the checked address to host + * endianness, or is -1 when no port was checked. Zero is a valid checked + * port. */ TRACE_EVENT(landlock_deny_access_net, TP_PROTO(const struct landlock_hierarchy *hierarchy, bool same_exec, bool logged, const struct landlock_blockers *blockers, - const struct sock *sk, u64 sport, u64 dport), + const struct sock *sk, u16 socket_family, + const struct sockaddr_storage *address, int addrlen), - TP_ARGS(hierarchy, same_exec, logged, blockers, sk, sport, dport), + TP_ARGS(hierarchy, same_exec, logged, blockers, sk, socket_family, + address, addrlen), TP_STRUCT__entry( __field( u64, domain_id ) @@ -829,26 +838,36 @@ TRACE_EVENT(landlock_deny_access_net, __field( bool, logged ) __field( enum landlock_request_type, blockers_type ) __field( access_mask_t, blockers_access ) - __field( u64, sport ) - __field( u64, dport ) + __field( s64, port ) ), TP_fast_assign( + const struct sockaddr *const addr = + (const struct sockaddr *)address; + const bool has_port = + addrlen >= (int)offsetofend(struct sockaddr_in, sin_port) && + (addr->sa_family == AF_INET || + addr->sa_family == AF_INET6 || + (addr->sa_family == AF_UNSPEC && + socket_family == AF_INET)); + __entry->domain_id = hierarchy->id; __entry->same_exec = same_exec; __entry->logged = logged; __entry->blockers_type = blockers->type; __entry->blockers_access = blockers->access; - __entry->sport = sport; - __entry->dport = dport; + __entry->port = + has_port ? + ntohs(((const struct sockaddr_in *)addr)->sin_port) : + -1; ), - TP_printk("domain=%llx same_exec=%d logged=%d blockers=%s sport=%llu dport=%llu", - __entry->domain_id, __entry->same_exec, __entry->logged, - __entry->blockers_type == LANDLOCK_REQUEST_NET_ACCESS ? - __print_flags(__entry->blockers_access, "|", _LANDLOCK_ACCESS_NET_NAMES) : - "unknown", - __entry->sport, __entry->dport) + TP_printk("domain=%llx same_exec=%d logged=%d blockers=%s port=%lld", + __entry->domain_id, __entry->same_exec, __entry->logged, + __entry->blockers_type == LANDLOCK_REQUEST_NET_ACCESS ? + __print_flags(__entry->blockers_access, "|", _LANDLOCK_ACCESS_NET_NAMES) : + "unknown", + __entry->port) ); /** diff --git a/security/landlock/log.h b/security/landlock/log.h index 04f3e241e765..4587c2b1566d 100644 --- a/security/landlock/log.h +++ b/security/landlock/log.h @@ -15,6 +15,7 @@ struct landlock_cred_security; struct landlock_hierarchy; +struct sockaddr; enum landlock_request_type { LANDLOCK_REQUEST_PTRACE = 1, @@ -30,6 +31,16 @@ struct landlock_blockers { enum landlock_request_type type; }; +#ifdef CONFIG_TRACEPOINTS + +struct landlock_net_trace { + const struct sockaddr *address; + int addrlen; + u16 socket_family; +}; + +#endif /* CONFIG_TRACEPOINTS */ + /* * We should be careful to only use a variable of this type for * landlock_log_denial(). This way, the compiler can remove it entirely if @@ -57,13 +68,20 @@ struct landlock_request { deny_masks_t deny_masks; optional_access_t quiet_optional_accesses; - /* - * Other-party domain ID for a relational (scope/ptrace) denial, or 0 if - * that party is unsandboxed. An ID, not a pointer: the other task can - * replace its credential and free the domain it referenced. Trace path - * only; audit ignores it. - */ - u64 other_domain_id; + union { + /* + * Other-party domain ID for a relational (scope/ptrace) denial, + * or 0 if that party is unsandboxed. Store an ID, not a + * pointer: the other task can replace its credential and free + * the domain it referenced. Trace-only; audit ignores it. + */ + u64 other_domain_id; + +#ifdef CONFIG_TRACEPOINTS + /* Synchronous context for a network denial. */ + const struct landlock_net_trace *trace_net; +#endif /* CONFIG_TRACEPOINTS */ + }; }; #ifdef CONFIG_SECURITY_LANDLOCK_LOG diff --git a/security/landlock/net.c b/security/landlock/net.c index 5552c60388f8..6fe0dbde3b78 100644 --- a/security/landlock/net.c +++ b/security/landlock/net.c @@ -93,7 +93,7 @@ static int current_check_access_socket(struct socket *const sock, return 0; /* Checks for minimal header length to safely read sa_family. */ - if (addrlen < offsetofend(typeof(*address), sa_family)) + if (addrlen < (int)offsetofend(typeof(*address), sa_family)) return -EINVAL; /* @@ -145,6 +145,17 @@ static int current_check_access_socket(struct socket *const sock, .audit.u.net = &audit_net, .access = access_request, .layer_masks = &layer_masks, +#ifdef CONFIG_TRACEPOINTS + .trace_net = + &(struct landlock_net_trace){ + .address = + address, + .addrlen = + addrlen, + .socket_family = + sock_family, + }, +#endif /* CONFIG_TRACEPOINTS */ }); return -EACCES; } @@ -276,14 +287,22 @@ static int current_check_access_socket(struct socket *const sock, audit_net.family = address->sa_family; audit_net.sk = sock->sk; - landlock_log_denial(subject, - &(struct landlock_request){ - .type = LANDLOCK_REQUEST_NET_ACCESS, - .audit.type = LSM_AUDIT_DATA_NET, - .audit.u.net = &audit_net, - .access = access_request, - .layer_masks = &layer_masks, - }); + landlock_log_denial( + subject, &(struct landlock_request){ + .type = LANDLOCK_REQUEST_NET_ACCESS, + .audit.type = LSM_AUDIT_DATA_NET, + .audit.u.net = &audit_net, + .access = access_request, + .layer_masks = &layer_masks, +#ifdef CONFIG_TRACEPOINTS + .trace_net = + &(struct landlock_net_trace){ + .address = address, + .addrlen = addrlen, + .socket_family = sock_family, + }, +#endif /* CONFIG_TRACEPOINTS */ + }); return -EACCES; } diff --git a/security/landlock/trace.c b/security/landlock/trace.c index 58276cc32d3f..9be86638f905 100644 --- a/security/landlock/trace.c +++ b/security/landlock/trace.c @@ -12,6 +12,7 @@ #include #include #include +#include #include #include "access.h" @@ -157,16 +158,31 @@ void landlock_trace_denial( break; case LANDLOCK_REQUEST_NET_ACCESS: if (trace_landlock_deny_access_net_enabled()) { + const struct landlock_net_trace *const trace_net = + request->trace_net; const struct landlock_blockers blockers = { .access = missing, .type = request->type, }; + struct sockaddr_storage address = {}; + if (WARN_ON_ONCE(!trace_net || !trace_net->address)) + return; + + if (WARN_ON_ONCE( + trace_net->addrlen < + (int)offsetofend(struct sockaddr, + sa_family) || + trace_net->addrlen > (int)sizeof(address))) + return; + + memcpy(&address, trace_net->address, + trace_net->addrlen); trace_landlock_deny_access_net( youngest_denied, same_exec, logged, &blockers, request->audit.u.net->sk, - ntohs(request->audit.u.net->sport), - ntohs(request->audit.u.net->dport)); + trace_net->socket_family, &address, + trace_net->addrlen); } break; case LANDLOCK_REQUEST_PTRACE: diff --git a/tools/testing/selftests/landlock/net_test.c b/tools/testing/selftests/landlock/net_test.c index 4fb705e1596d..28942438e154 100644 --- a/tools/testing/selftests/landlock/net_test.c +++ b/tools/testing/selftests/landlock/net_test.c @@ -3482,8 +3482,8 @@ TEST_F(trace_net, deny_access_net_bind) /* * Anchors the denial fields shared by every deny_access_net event so a field - * test proves more than sport/dport: the denying domain, the same-exec bit, the - * audit-logging verdict, and the blocked access all stay populated. + * test proves more than the checked endpoint: the denying domain, the same-exec + * bit, the audit-logging verdict, and the blocked access all stay populated. */ static void expect_net_deny_common_fields(struct __test_metadata *const _metadata, @@ -3569,38 +3569,38 @@ FIXTURE_VARIANT(trace_net_connect) { bool deny_connect; }; +/* Denied connect() to the next IPv4 port. */ /* clang-format off */ - -/* Denied connect(): sport=0, dport=. */ FIXTURE_VARIANT_ADD(trace_net_connect, connect_denied) { + /* clang-format on */ .handled = LANDLOCK_ACCESS_NET_CONNECT_TCP, .bind_base_first = false, .deny_connect = true, }; -/* Denied bind(): sport=, dport=0. */ +/* Denied bind() to the next IPv4 port. */ +/* clang-format off */ FIXTURE_VARIANT_ADD(trace_net_connect, bind_fields) { + /* clang-format on */ .handled = LANDLOCK_ACCESS_NET_BIND_TCP, .bind_base_first = false, .deny_connect = false, }; -/* Denied connect() after an allowed bind(): the connect fields (sport=0). */ +/* Denied connect() after an allowed bind() uses the checked destination. */ +/* clang-format off */ FIXTURE_VARIANT_ADD(trace_net_connect, connect_after_bind) { - .handled = LANDLOCK_ACCESS_NET_BIND_TCP | LANDLOCK_ACCESS_NET_CONNECT_TCP, + /* clang-format on */ + .handled = LANDLOCK_ACCESS_NET_BIND_TCP | + LANDLOCK_ACCESS_NET_CONNECT_TCP, .bind_base_first = true, .deny_connect = true, }; -/* clang-format on */ - /* - * A denied TCP bind(2) or connect(2) emits one deny_access_net event. The port - * is reported in the field matching the denied operation, in host endianness - * (the UAPI landlock_net_port_attr.port convention): a connect denial reports - * sport=0 dport=, a bind denial reports sport= dport=0, so a - * byte-order or field-swap bug is caught. A prior allowed bind - * (connect_after_bind) does not change the connect denial's fields. + * A denied TCP bind(2) or connect(2) emits one deny_access_net event with the + * checked IPv4 port in host endianness (the UAPI landlock_net_port_attr.port + * convention). A prior allowed bind does not change a connect denial's port. */ TEST_F(trace_net_connect, deny_access_net) { @@ -3693,21 +3693,13 @@ TEST_F(trace_net_connect, deny_access_net) expect_net_deny_common_fields(_metadata, buf); - /* - * The denied operation's port field carries the port; the other is 0. - */ snprintf(expected, sizeof(expected), "%llu", (unsigned long long)(sock_port_start + 1)); ASSERT_EQ(0, tracefs_extract_field(buf, REGEX_DENY_ACCESS_NET(TRACE_TASK), - "sport", field, sizeof(field))); - EXPECT_STREQ(variant->deny_connect ? "0" : expected, field); - - ASSERT_EQ(0, - tracefs_extract_field(buf, REGEX_DENY_ACCESS_NET(TRACE_TASK), - "dport", field, sizeof(field))); - EXPECT_STREQ(variant->deny_connect ? expected : "0", field); + "port", field, sizeof(field))); + EXPECT_STREQ(expected, field); free(buf); } @@ -3867,11 +3859,4 @@ TEST_F(trace_net_check_rule, check_rule_net_fields) free(buf); } -/* - * IPv6 network trace tests are intentionally elided. IPv6 hook dispatch uses - * the same current_check_access_socket() code path as IPv4, validated by the - * audit tests in this file. The trace events use the same blockers/sport/dport - * fields regardless of address family. - */ - TEST_HARNESS_MAIN diff --git a/tools/testing/selftests/landlock/trace.h b/tools/testing/selftests/landlock/trace.h index e6873853376c..2ec863362173 100644 --- a/tools/testing/selftests/landlock/trace.h +++ b/tools/testing/selftests/landlock/trace.h @@ -145,8 +145,7 @@ "same_exec=[01] " \ "logged=[01] " \ "blockers=[a-z_|]* " \ - "sport=[0-9]\\+ " \ - "dport=[0-9]\\+$" + "port=-\\?[0-9]\\+$" #define REGEX_DENY_PTRACE(task) \ TRACE_PREFIX(task) \ From 7ad69ac63315506e2091bf46d5c96c49f6ce439e Mon Sep 17 00:00:00 2001 From: =?UTF-8?q?Micka=C3=ABl=20Sala=C3=BCn?= Date: Fri, 18 Sep 2026 20:50:28 +0200 Subject: [PATCH 428/592] landlock: Report the actual ptrace tracer MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit The ptrace denial callback identifies only the tracee. Current is the tracer during hook_ptrace_access_check(), but it is the tracee during PTRACE_TRACEME, where the parent is the actual tracer. A consumer therefore cannot infer both parties from the existing arguments. Append the actual tracer task to the typed-BPF callback: current for hook_ptrace_access_check() and parent for hook_ptrace_traceme(). Carry it with the tracee domain ID in a private ptrace context. Both hooks keep the selected tasks alive through synchronous dispatch, so no extra task reference is needed. Keep the tracefs record unchanged. The new context is available only to typed BPF, while same_exec continues to describe the tracer that owns the denying policy. Cc: Günther Noack Cc: Steven Rostedt Fixes: bb91730f16c0 ("landlock: Add tracepoints for ptrace and scope denials") Link: https://patch.msgid.link/20260918185036.608651-6-mic@digikod.net Signed-off-by: Mickaël Salaün --- include/trace/events/landlock.h | 11 ++++++++--- security/landlock/log.h | 16 ++++++++++++---- security/landlock/task.c | 28 +++++++++++++++++----------- security/landlock/trace.c | 20 ++++++++++++++------ 4 files changed, 51 insertions(+), 24 deletions(-) diff --git a/include/trace/events/landlock.h b/include/trace/events/landlock.h index a982a7cfa881..8c6ebf958d66 100644 --- a/include/trace/events/landlock.h +++ b/include/trace/events/landlock.h @@ -295,6 +295,8 @@ static inline const char *__trace_landlock_print_layers( * the two parties without kernel-internal state. The ID is a scalar * snapshot, not a live domain pointer that could dangle: an optional * relational referent is a scalar (0 sentinel), not a nullable pointer. + * For ptrace, same_exec instead describes the tracer, even for + * PTRACE_TRACEME, and may differ from the current task. * * Blocker fields * ~~~~~~~~~~~~~~ @@ -875,12 +877,14 @@ TRACE_EVENT(landlock_deny_access_net, * * @hierarchy: Denying domain's hierarchy node (never NULL); its id is the * domain field. - * @same_exec: Whether the current task entered the denying domain itself. + * @same_exec: Whether the tracer entered the denying domain itself. * @logged: The domain's audit-logging decision for this denial. * @tracee_domain_id: The tracee's Landlock domain ID, or 0 if the tracee * is unsandboxed. * @tracee: The target task ptrace acted on (never NULL). tracee_pid is * the init-namespace TGID (like audit's opid). + * @tracer: The tracer or proposed tracer (never NULL); for PTRACE_TRACEME + * this is the parent, not the syscall caller. * * Emitted when a Landlock domain denies a ptrace operation. */ @@ -888,9 +892,10 @@ TRACE_EVENT(landlock_deny_ptrace, TP_PROTO(const struct landlock_hierarchy *hierarchy, bool same_exec, bool logged, u64 tracee_domain_id, - const struct task_struct *tracee), + const struct task_struct *tracee, + const struct task_struct *tracer), - TP_ARGS(hierarchy, same_exec, logged, tracee_domain_id, tracee), + TP_ARGS(hierarchy, same_exec, logged, tracee_domain_id, tracee, tracer), TP_STRUCT__entry( __field( u64, domain_id ) diff --git a/security/landlock/log.h b/security/landlock/log.h index 4587c2b1566d..821df6f711c6 100644 --- a/security/landlock/log.h +++ b/security/landlock/log.h @@ -16,6 +16,7 @@ struct landlock_cred_security; struct landlock_hierarchy; struct sockaddr; +struct task_struct; enum landlock_request_type { LANDLOCK_REQUEST_PTRACE = 1, @@ -39,6 +40,11 @@ struct landlock_net_trace { u16 socket_family; }; +struct landlock_ptrace_trace { + u64 tracee_domain_id; + const struct task_struct *tracer; +}; + #endif /* CONFIG_TRACEPOINTS */ /* @@ -70,16 +76,18 @@ struct landlock_request { union { /* - * Other-party domain ID for a relational (scope/ptrace) denial, - * or 0 if that party is unsandboxed. Store an ID, not a - * pointer: the other task can replace its credential and free - * the domain it referenced. Trace-only; audit ignores it. + * Other-party domain ID for a scope denial, or 0 if that party + * is unsandboxed. Store an ID, not a pointer: the other task + * can replace its credential and free the domain it referenced. + * Audit ignores this trace-only field. */ u64 other_domain_id; #ifdef CONFIG_TRACEPOINTS /* Synchronous context for a network denial. */ const struct landlock_net_trace *trace_net; + /* Consumed only by the synchronous trace dispatcher. */ + const struct landlock_ptrace_trace *trace_ptrace; #endif /* CONFIG_TRACEPOINTS */ }; }; diff --git a/security/landlock/task.c b/security/landlock/task.c index 4491ce31ae04..445f6b921a87 100644 --- a/security/landlock/task.c +++ b/security/landlock/task.c @@ -88,7 +88,9 @@ static int hook_ptrace_access_check(struct task_struct *const child, const unsigned int mode) { const struct landlock_cred_security *parent_subject; +#ifdef CONFIG_TRACEPOINTS u64 tracee_domain_id = 0; +#endif /* CONFIG_TRACEPOINTS */ int err; /* Quick return for non-landlocked tasks. */ @@ -100,10 +102,10 @@ static int hook_ptrace_access_check(struct task_struct *const child, const struct landlock_domain *const child_dom = landlock_get_task_domain(child); err = domain_ptrace(parent_subject->domain, child_dom); -#ifdef CONFIG_SECURITY_LANDLOCK_LOG +#ifdef CONFIG_TRACEPOINTS if (child_dom) tracee_domain_id = child_dom->hierarchy->id; -#endif /* CONFIG_SECURITY_LANDLOCK_LOG */ +#endif /* CONFIG_TRACEPOINTS */ } if (!err) @@ -121,7 +123,12 @@ static int hook_ptrace_access_check(struct task_struct *const child, .u.tsk = child, }, .layer_plus_one = parent_subject->domain->num_layers, - .other_domain_id = tracee_domain_id, +#ifdef CONFIG_TRACEPOINTS + .trace_ptrace = &(struct landlock_ptrace_trace) { + .tracee_domain_id = tracee_domain_id, + .tracer = current, + }, +#endif /* CONFIG_TRACEPOINTS */ }); return err; @@ -142,7 +149,6 @@ static int hook_ptrace_traceme(struct task_struct *const parent) { const struct landlock_cred_security *parent_subject; const struct landlock_domain *child_dom; - u64 tracee_domain_id = 0; int err; child_dom = landlock_get_current_domain(); @@ -154,12 +160,6 @@ static int hook_ptrace_traceme(struct task_struct *const parent) if (!err) return 0; -#ifdef CONFIG_SECURITY_LANDLOCK_LOG - /* The tracee is the current task; its domain is stable here. */ - if (child_dom) - tracee_domain_id = child_dom->hierarchy->id; -#endif /* CONFIG_SECURITY_LANDLOCK_LOG */ - /* * For the ptrace_traceme case, we log the domain which is the cause of * the denial, which means the parent domain instead of the current @@ -174,7 +174,13 @@ static int hook_ptrace_traceme(struct task_struct *const parent) .u.tsk = current, }, .layer_plus_one = parent_subject->domain->num_layers, - .other_domain_id = tracee_domain_id, +#ifdef CONFIG_TRACEPOINTS + .trace_ptrace = &(struct landlock_ptrace_trace) { + /* The current task's domain is stable here. */ + .tracee_domain_id = child_dom ? child_dom->hierarchy->id : 0, + .tracer = parent, + }, +#endif /* CONFIG_TRACEPOINTS */ }); return err; } diff --git a/security/landlock/trace.c b/security/landlock/trace.c index 9be86638f905..43091c052f77 100644 --- a/security/landlock/trace.c +++ b/security/landlock/trace.c @@ -63,7 +63,7 @@ void landlock_trace_free_domain(const struct landlock_hierarchy *const hierarchy * @request: Detail of the user space request. * @youngest_denied: The youngest hierarchy node that denied the access. * @missing: The final missing access subset, when applicable. - * @same_exec: Whether the current task is the same executable that called + * @same_exec: Whether the policy subject is the same executable that called * landlock_restrict_self() for the denying domain, as computed * by landlock_log_denial(). * @logged: Whether the domain's policy selects this denial for logging, as @@ -186,11 +186,19 @@ void landlock_trace_denial( } break; case LANDLOCK_REQUEST_PTRACE: - if (trace_landlock_deny_ptrace_enabled()) - trace_landlock_deny_ptrace(youngest_denied, same_exec, - logged, - request->other_domain_id, - request->audit.u.tsk); + if (trace_landlock_deny_ptrace_enabled()) { + const struct landlock_ptrace_trace *const trace_ptrace = + request->trace_ptrace; + + if (WARN_ON_ONCE(!trace_ptrace || + !trace_ptrace->tracer)) + return; + + trace_landlock_deny_ptrace( + youngest_denied, same_exec, logged, + trace_ptrace->tracee_domain_id, + request->audit.u.tsk, trace_ptrace->tracer); + } break; case LANDLOCK_REQUEST_SCOPE_SIGNAL: if (trace_landlock_deny_scope_signal_enabled()) From 0889db596a25ecde210e66fa0db70bc6a6d91f6c Mon Sep 17 00:00:00 2001 From: =?UTF-8?q?Micka=C3=ABl=20Sala=C3=BCn?= Date: Fri, 18 Sep 2026 20:50:29 +0200 Subject: [PATCH 429/592] landlock: Report the effective signal number MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit The signal-scope denial callback identifies its target but not the effective signal. This loses permission-probe signal zero and makes the file-owner hook's zero sentinel ambiguous. Append an int signal argument to the typed-BPF callback. Preserve sig, including zero, in hook_task_kill(). In hook_file_send_sigiotask(), translate signum zero to SIGIO at the producer, where its meaning is known. Carry the effective signal and target domain ID in a private, stack-backed context consumed synchronously. This requires no allocation or task reference in the interrupt-capable file-owner path. Gate this context and the remaining scope-only domain IDs with CONFIG_TRACEPOINTS. Keep the tracefs record and audit output unchanged. Cc: Günther Noack Cc: Steven Rostedt Fixes: bb91730f16c0 ("landlock: Add tracepoints for ptrace and scope denials") Link: https://patch.msgid.link/20260918185036.608651-7-mic@digikod.net Signed-off-by: Mickaël Salaün --- include/trace/events/landlock.h | 8 +++-- security/landlock/log.h | 21 ++++++++---- security/landlock/task.c | 59 +++++++++++++++++++++++---------- security/landlock/trace.c | 12 +++++-- 4 files changed, 71 insertions(+), 29 deletions(-) diff --git a/include/trace/events/landlock.h b/include/trace/events/landlock.h index 8c6ebf958d66..5da0f12ed2c3 100644 --- a/include/trace/events/landlock.h +++ b/include/trace/events/landlock.h @@ -928,12 +928,14 @@ TRACE_EVENT(landlock_deny_ptrace, * * @hierarchy: Denying domain's hierarchy node (never NULL); its id is the * domain field. - * @same_exec: Whether the current task entered the denying domain itself. + * @same_exec: Whether the policy subject entered the denying domain itself. * @logged: The domain's audit-logging decision for this denial. * @target_domain_id: The target's Landlock domain ID, or 0 if the target * is unsandboxed. * @target: The task the signal was aimed at (never NULL). target_pid is * the init-namespace TGID (like audit's opid). + * @signal: The signal selected by the denied check. Zero is a permission + * probe, not an absent value. * * Emitted when a Landlock domain denies signal delivery to a scoped-out * target. @@ -942,9 +944,9 @@ TRACE_EVENT(landlock_deny_scope_signal, TP_PROTO(const struct landlock_hierarchy *hierarchy, bool same_exec, bool logged, u64 target_domain_id, - const struct task_struct *target), + const struct task_struct *target, int signal), - TP_ARGS(hierarchy, same_exec, logged, target_domain_id, target), + TP_ARGS(hierarchy, same_exec, logged, target_domain_id, target, signal), TP_STRUCT__entry( __field( u64, domain_id ) diff --git a/security/landlock/log.h b/security/landlock/log.h index 821df6f711c6..faa30e26e42a 100644 --- a/security/landlock/log.h +++ b/security/landlock/log.h @@ -45,6 +45,11 @@ struct landlock_ptrace_trace { const struct task_struct *tracer; }; +struct landlock_signal_trace { + u64 target_domain_id; + int signal; +}; + #endif /* CONFIG_TRACEPOINTS */ /* @@ -74,22 +79,24 @@ struct landlock_request { deny_masks_t deny_masks; optional_access_t quiet_optional_accesses; +#ifdef CONFIG_TRACEPOINTS union { /* - * Other-party domain ID for a scope denial, or 0 if that party - * is unsandboxed. Store an ID, not a pointer: the other task - * can replace its credential and free the domain it referenced. - * Audit ignores this trace-only field. + * Other-party domain ID for an abstract UNIX socket scope + * denial, or 0 if that party is unsandboxed. Store an ID, not + * a pointer: the other task can replace its credential and free + * the domain it referenced. */ u64 other_domain_id; -#ifdef CONFIG_TRACEPOINTS /* Synchronous context for a network denial. */ const struct landlock_net_trace *trace_net; - /* Consumed only by the synchronous trace dispatcher. */ + /* Synchronous context for a ptrace denial. */ const struct landlock_ptrace_trace *trace_ptrace; -#endif /* CONFIG_TRACEPOINTS */ + /* Synchronous context for a signal denial. */ + const struct landlock_signal_trace *trace_signal; }; +#endif /* CONFIG_TRACEPOINTS */ }; #ifdef CONFIG_SECURITY_LANDLOCK_LOG diff --git a/security/landlock/task.c b/security/landlock/task.c index 445f6b921a87..d9eae86fc552 100644 --- a/security/landlock/task.c +++ b/security/landlock/task.c @@ -256,8 +256,7 @@ static bool domain_is_scoped(const struct landlock_domain *const client, } static bool sock_is_scoped(struct sock *const other, - const struct landlock_domain *const domain, - u64 *const peer_domain_id) + const struct landlock_domain *const domain) { const struct landlock_domain *dom_other; @@ -275,13 +274,23 @@ static bool sock_is_scoped(struct sock *const other, return false; dom_other = landlock_cred(other->sk_socket->file->f_cred)->domain; -#ifdef CONFIG_SECURITY_LANDLOCK_LOG - *peer_domain_id = dom_other ? dom_other->hierarchy->id : 0; -#endif /* CONFIG_SECURITY_LANDLOCK_LOG */ return domain_is_scoped(domain, dom_other, LANDLOCK_SCOPE_ABSTRACT_UNIX_SOCKET); } +#ifdef CONFIG_TRACEPOINTS + +static u64 get_socket_domain_id(const struct sock *const other) +{ + const struct landlock_domain *domain; + + lockdep_assert_held(&unix_sk(other)->lock); + domain = landlock_cred(other->sk_socket->file->f_cred)->domain; + return domain ? domain->hierarchy->id : 0; +} + +#endif /* CONFIG_TRACEPOINTS */ + static bool is_abstract_socket(struct sock *const sock) { struct unix_address *addr = unix_sk(sock)->addr; @@ -305,7 +314,6 @@ static int hook_unix_stream_connect(struct sock *const sock, struct sock *const newsk) { size_t handle_layer; - u64 peer_domain_id = 0; const struct landlock_cred_security *const subject = landlock_get_applicable_subject(current_cred(), unix_scope, &handle_layer); @@ -317,7 +325,7 @@ static int hook_unix_stream_connect(struct sock *const sock, if (!is_abstract_socket(other)) return 0; - if (!sock_is_scoped(other, subject->domain, &peer_domain_id)) + if (!sock_is_scoped(other, subject->domain)) return 0; landlock_log_denial(subject, &(struct landlock_request) { @@ -329,7 +337,9 @@ static int hook_unix_stream_connect(struct sock *const sock, }, }, .layer_plus_one = handle_layer + 1, - .other_domain_id = peer_domain_id, +#ifdef CONFIG_TRACEPOINTS + .other_domain_id = get_socket_domain_id(other), +#endif /* CONFIG_TRACEPOINTS */ }); return -EPERM; } @@ -338,7 +348,6 @@ static int hook_unix_may_send(struct socket *const sock, struct socket *const other) { size_t handle_layer; - u64 peer_domain_id = 0; const struct landlock_cred_security *const subject = landlock_get_applicable_subject(current_cred(), unix_scope, &handle_layer); @@ -356,7 +365,7 @@ static int hook_unix_may_send(struct socket *const sock, if (!is_abstract_socket(other->sk)) return 0; - if (!sock_is_scoped(other->sk, subject->domain, &peer_domain_id)) + if (!sock_is_scoped(other->sk, subject->domain)) return 0; landlock_log_denial(subject, &(struct landlock_request) { @@ -368,7 +377,9 @@ static int hook_unix_may_send(struct socket *const sock, }, }, .layer_plus_one = handle_layer + 1, - .other_domain_id = peer_domain_id, +#ifdef CONFIG_TRACEPOINTS + .other_domain_id = get_socket_domain_id(other->sk), +#endif /* CONFIG_TRACEPOINTS */ }); return -EPERM; } @@ -383,7 +394,9 @@ static int hook_task_kill(struct task_struct *const p, { bool is_scoped; size_t handle_layer; +#ifdef CONFIG_TRACEPOINTS u64 target_domain_id = 0; +#endif /* CONFIG_TRACEPOINTS */ const struct landlock_cred_security *subject; if (!cred) { @@ -415,10 +428,10 @@ static int hook_task_kill(struct task_struct *const p, is_scoped = domain_is_scoped(subject->domain, other, signal_scope.scope); -#ifdef CONFIG_SECURITY_LANDLOCK_LOG +#ifdef CONFIG_TRACEPOINTS if (other) target_domain_id = other->hierarchy->id; -#endif /* CONFIG_SECURITY_LANDLOCK_LOG */ +#endif /* CONFIG_TRACEPOINTS */ } if (!is_scoped) @@ -431,7 +444,12 @@ static int hook_task_kill(struct task_struct *const p, .u.tsk = p, }, .layer_plus_one = handle_layer + 1, - .other_domain_id = target_domain_id, +#ifdef CONFIG_TRACEPOINTS + .trace_signal = &(struct landlock_signal_trace) { + .target_domain_id = target_domain_id, + .signal = sig, + }, +#endif /* CONFIG_TRACEPOINTS */ }); return -EPERM; } @@ -441,7 +459,9 @@ static int hook_file_send_sigiotask(struct task_struct *tsk, { const struct landlock_cred_security *subject; bool is_scoped = false; +#ifdef CONFIG_TRACEPOINTS u64 target_domain_id = 0; +#endif /* CONFIG_TRACEPOINTS */ /* Lock already held by send_sigio() and send_sigurg(). */ lockdep_assert_held(&fown->lock); @@ -474,10 +494,10 @@ static int hook_file_send_sigiotask(struct task_struct *tsk, is_scoped = domain_is_scoped(subject->domain, other, signal_scope.scope); -#ifdef CONFIG_SECURITY_LANDLOCK_LOG +#ifdef CONFIG_TRACEPOINTS if (other) target_domain_id = other->hierarchy->id; -#endif /* CONFIG_SECURITY_LANDLOCK_LOG */ +#endif /* CONFIG_TRACEPOINTS */ } if (!is_scoped) @@ -492,7 +512,12 @@ static int hook_file_send_sigiotask(struct task_struct *tsk, #ifdef CONFIG_SECURITY_LANDLOCK_LOG .layer_plus_one = landlock_file(fown->file)->fown_layer + 1, #endif /* CONFIG_SECURITY_LANDLOCK_LOG */ - .other_domain_id = target_domain_id, +#ifdef CONFIG_TRACEPOINTS + .trace_signal = &(struct landlock_signal_trace) { + .target_domain_id = target_domain_id, + .signal = signum ? signum : SIGIO, + }, +#endif /* CONFIG_TRACEPOINTS */ }); return -EPERM; } diff --git a/security/landlock/trace.c b/security/landlock/trace.c index 43091c052f77..225dbf37bab0 100644 --- a/security/landlock/trace.c +++ b/security/landlock/trace.c @@ -201,10 +201,18 @@ void landlock_trace_denial( } break; case LANDLOCK_REQUEST_SCOPE_SIGNAL: - if (trace_landlock_deny_scope_signal_enabled()) + if (trace_landlock_deny_scope_signal_enabled()) { + const struct landlock_signal_trace *const trace_signal = + request->trace_signal; + + if (WARN_ON_ONCE(!trace_signal)) + return; + trace_landlock_deny_scope_signal( youngest_denied, same_exec, logged, - request->other_domain_id, request->audit.u.tsk); + trace_signal->target_domain_id, + request->audit.u.tsk, trace_signal->signal); + } break; case LANDLOCK_REQUEST_SCOPE_ABSTRACT_UNIX_SOCKET: if (trace_landlock_deny_scope_abstract_unix_socket_enabled()) From c6dea91d846f192eb6ddbd44f5fd873035b8b285 Mon Sep 17 00:00:00 2001 From: =?UTF-8?q?Micka=C3=ABl=20Sala=C3=BCn?= Date: Fri, 18 Sep 2026 20:50:30 +0200 Subject: [PATCH 430/592] selftests/landlock: Test filesystem denial blockers MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit Filesystem denial traces identify the policy change needed to allow a request, so require exact blocker values rather than merely nonempty output. Pin a READ_DIR denial to exactly one event with blockers=read_dir. Pin a REFER-only mount denial to EPERM and exactly one event with blockers=change_topology. The mount child retains CAP_SYS_ADMIN so Landlock is the only expected source of EPERM. This prevents a later capability failure from masking a Landlock regression; the trace-collecting parent remains unsandboxed. Cc: Günther Noack Cc: Steven Rostedt Link: https://patch.msgid.link/20260918185036.608651-8-mic@digikod.net Signed-off-by: Mickaël Salaün --- .../selftests/landlock/trace_fs_test.c | 76 ++++++++++++++++++- 1 file changed, 73 insertions(+), 3 deletions(-) diff --git a/tools/testing/selftests/landlock/trace_fs_test.c b/tools/testing/selftests/landlock/trace_fs_test.c index 6666d4746cb1..64014ade3a0e 100644 --- a/tools/testing/selftests/landlock/trace_fs_test.c +++ b/tools/testing/selftests/landlock/trace_fs_test.c @@ -548,7 +548,8 @@ TEST_F(trace_fs, check_rule_nested) */ TEST_F(trace_fs, deny_access_fs_denied) { - char *buf; + const char *const event_regex = REGEX_DENY_ACCESS_FS(TRACE_TASK); + char *buf, blockers[64]; int count; ASSERT_EQ(0, tracefs_clear_buf()); @@ -564,8 +565,77 @@ TEST_F(trace_fs, deny_access_fs_denied) buf = tracefs_read_buf(); ASSERT_NE(NULL, buf); - count = tracefs_count_matches(buf, REGEX_DENY_ACCESS_FS(TRACE_TASK)); - EXPECT_LE(1, count); + count = tracefs_count_matches(buf, event_regex); + EXPECT_EQ(1, count) + { + TH_LOG("Expected 1 access denial, got %d\n%s", count, buf); + } + ASSERT_EQ(0, tracefs_extract_field(buf, event_regex, "blockers", + blockers, sizeof(blockers))); + EXPECT_STREQ("read_dir", blockers); + + free(buf); +} + +/* + * Verifies that a denied mount reports the singleton topology blocker rather + * than an empty access mask. + */ +TEST_F(trace_fs, deny_change_topology) +{ + const char *const event_regex = REGEX_DENY_ACCESS_FS(TRACE_TASK); + const struct landlock_ruleset_attr ruleset_attr = { + .handled_access_fs = LANDLOCK_ACCESS_FS_REFER, + }; + char *buf, blockers[64]; + int count, ruleset_fd, status; + pid_t pid; + + ruleset_fd = + landlock_create_ruleset(&ruleset_attr, sizeof(ruleset_attr), 0); + ASSERT_LE(0, ruleset_fd); + ASSERT_EQ(0, tracefs_clear_buf()); + + /* Ensure that Landlock is the only expected mount denial. */ + set_cap(_metadata, CAP_SYS_ADMIN); + pid = fork(); + ASSERT_LE(0, pid); + if (pid == 0) { + if (prctl(PR_SET_NO_NEW_PRIVS, 1, 0, 0, 0)) { + close(ruleset_fd); + _exit(1); + } + if (landlock_restrict_self(ruleset_fd, 0)) { + close(ruleset_fd); + _exit(2); + } + close(ruleset_fd); + + if (mount(NULL, "/", NULL, MS_PRIVATE | MS_REC, NULL) != -1) + _exit(3); + + if (errno != EPERM) + _exit(4); + + _exit(0); + } + close(ruleset_fd); + clear_cap(_metadata, CAP_SYS_ADMIN); + + ASSERT_EQ(pid, waitpid(pid, &status, 0)); + ASSERT_TRUE(WIFEXITED(status)); + EXPECT_EQ(0, WEXITSTATUS(status)); + + buf = tracefs_read_buf(); + ASSERT_NE(NULL, buf); + count = tracefs_count_matches(buf, event_regex); + EXPECT_EQ(1, count) + { + TH_LOG("Expected 1 topology denial, got %d\n%s", count, buf); + } + ASSERT_EQ(0, tracefs_extract_field(buf, event_regex, "blockers", + blockers, sizeof(blockers))); + EXPECT_STREQ("change_topology", blockers); free(buf); } From c8dcb17205a689e96e2b8e46f22575a39b5b6320 Mon Sep 17 00:00:00 2001 From: =?UTF-8?q?Micka=C3=ABl=20Sala=C3=BCn?= Date: Fri, 18 Sep 2026 20:50:31 +0200 Subject: [PATCH 431/592] selftests/landlock: Test network denial context MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit Network denial events now report the port from their one authoritative validated address through one signed field. Verify this value directly without inferring a source or destination role. Exercise a nonzero IPv4 TCP bind, an explicit-zero IPv4 UDP bind, a synthetic-zero IPv6 UDP autobind, and a family-only AF_UNSPEC UDP send. Require exactly one event with the exact policy blocker for each shape. The value -1 distinguishes a family-only address with no validated port from the two valid port-zero cases. Test coverage for security/landlock is 91.6% of 2625 lines according to LLVM 22. Cc: Günther Noack Cc: Steven Rostedt Link: https://patch.msgid.link/20260918185036.608651-9-mic@digikod.net [mic: Add test coverage] Signed-off-by: Mickaël Salaün --- tools/testing/selftests/landlock/net_test.c | 257 +++++++++++--------- 1 file changed, 146 insertions(+), 111 deletions(-) diff --git a/tools/testing/selftests/landlock/net_test.c b/tools/testing/selftests/landlock/net_test.c index 28942438e154..16afbfdf06bb 100644 --- a/tools/testing/selftests/landlock/net_test.c +++ b/tools/testing/selftests/landlock/net_test.c @@ -3481,13 +3481,14 @@ TEST_F(trace_net, deny_access_net_bind) } /* - * Anchors the denial fields shared by every deny_access_net event so a field - * test proves more than the checked endpoint: the denying domain, the same-exec - * bit, the audit-logging verdict, and the blocked access all stay populated. + * Anchors the denial fields shared by every deny_access_net event so a port + * test also proves the denying domain, execution status, logging verdict, and + * exact blocked access. */ static void expect_net_deny_common_fields(struct __test_metadata *const _metadata, - const char *const buf) + const char *const buf, + const char *const expected_blockers) { char field[64]; @@ -3511,18 +3512,21 @@ expect_net_deny_common_fields(struct __test_metadata *const _metadata, ASSERT_EQ(0, tracefs_extract_field(buf, REGEX_DENY_ACCESS_NET(TRACE_TASK), "blockers", field, sizeof(field))); - EXPECT_STRNE("", field); + EXPECT_STREQ(expected_blockers, field); } -/* Connect and field-check tests use a separate fixture without variants. */ +enum trace_net_operation { + TRACE_NET_BIND, + TRACE_NET_SEND, +}; /* clang-format off */ -FIXTURE(trace_net_connect) { +FIXTURE(trace_net_address) { /* clang-format on */ int tracefs_ok; }; -FIXTURE_SETUP(trace_net_connect) +FIXTURE_SETUP(trace_net_address) { int ret; @@ -3547,7 +3551,7 @@ FIXTURE_SETUP(trace_net_connect) clear_cap(_metadata, CAP_SYS_ADMIN); } -FIXTURE_TEARDOWN(trace_net_connect) +FIXTURE_TEARDOWN(trace_net_address) { if (!self->tracefs_ok) return; @@ -3559,146 +3563,177 @@ FIXTURE_TEARDOWN(trace_net_connect) } /* clang-format off */ -FIXTURE_VARIANT(trace_net_connect) { +FIXTURE_VARIANT(trace_net_address) { /* clang-format on */ - /* handled_access_net, also the access allowed on the base port. */ - __u64 handled; - /* Bind the allowed base port before the denied operation. */ - bool bind_base_first; - /* Denied operation on the next port: connect (true) or bind (false). */ - bool deny_connect; + int socket_family; + int socket_type; + enum trace_net_operation operation; + int address_family; + socklen_t addrlen; + __u64 handled_access; + const char *expected_blockers; + bool address_port_zero; + bool expected_address_port; + int expected_port; }; -/* Denied connect() to the next IPv4 port. */ /* clang-format off */ -FIXTURE_VARIANT_ADD(trace_net_connect, connect_denied) { +FIXTURE_VARIANT_ADD(trace_net_address, ipv4_tcp_bind) { /* clang-format on */ - .handled = LANDLOCK_ACCESS_NET_CONNECT_TCP, - .bind_base_first = false, - .deny_connect = true, + .socket_family = AF_INET, + .socket_type = SOCK_STREAM, + .operation = TRACE_NET_BIND, + .address_family = AF_INET, + .addrlen = sizeof(struct sockaddr_in), + .handled_access = LANDLOCK_ACCESS_NET_BIND_TCP, + .expected_blockers = "bind_tcp", + .expected_address_port = true, }; -/* Denied bind() to the next IPv4 port. */ +/* Explicit bind(0) has a checked zero port. */ /* clang-format off */ -FIXTURE_VARIANT_ADD(trace_net_connect, bind_fields) { +FIXTURE_VARIANT_ADD(trace_net_address, ipv4_udp_bind_zero) { /* clang-format on */ - .handled = LANDLOCK_ACCESS_NET_BIND_TCP, - .bind_base_first = false, - .deny_connect = false, + .socket_family = AF_INET, + .socket_type = SOCK_DGRAM, + .operation = TRACE_NET_BIND, + .address_family = AF_INET, + .addrlen = sizeof(struct sockaddr_in), + .handled_access = LANDLOCK_ACCESS_NET_BIND_UDP, + .expected_blockers = "bind_udp", + .address_port_zero = true, + .expected_port = 0, }; -/* Denied connect() after an allowed bind() uses the checked destination. */ +/* A UDP send can deny its synthetic unspecified bind endpoint. */ /* clang-format off */ -FIXTURE_VARIANT_ADD(trace_net_connect, connect_after_bind) { +FIXTURE_VARIANT_ADD(trace_net_address, ipv6_udp_autobind) { /* clang-format on */ - .handled = LANDLOCK_ACCESS_NET_BIND_TCP | - LANDLOCK_ACCESS_NET_CONNECT_TCP, - .bind_base_first = true, - .deny_connect = true, + .socket_family = AF_INET6, + .socket_type = SOCK_DGRAM, + .operation = TRACE_NET_SEND, + .address_family = AF_INET6, + .addrlen = sizeof(struct sockaddr_in6), + .handled_access = LANDLOCK_ACCESS_NET_BIND_UDP, + .expected_blockers = "bind_udp", + .expected_port = 0, }; -/* - * A denied TCP bind(2) or connect(2) emits one deny_access_net event with the - * checked IPv4 port in host endianness (the UAPI landlock_net_port_attr.port - * convention). A prior allowed bind does not change a connect denial's port. - */ -TEST_F(trace_net_connect, deny_access_net) +/* A family-only address has no checked port. */ +/* clang-format off */ +FIXTURE_VARIANT_ADD(trace_net_address, ipv6_unspec_udp_send_min) { + /* clang-format on */ + .socket_family = AF_INET6, + .socket_type = SOCK_DGRAM, + .operation = TRACE_NET_SEND, + .address_family = AF_UNSPEC, + .addrlen = sizeof(sa_family_t), + .handled_access = LANDLOCK_ACCESS_NET_CONNECT_SEND_UDP, + .expected_blockers = "connect_send_udp", + .expected_port = -1, +}; + +static void set_trace_net_address(struct sockaddr_storage *const storage, + const int socket_family, + const int address_family, + const unsigned short port) { - pid_t child; - int status; - char *buf; + memset(storage, 0, sizeof(*storage)); + + if (socket_family == AF_INET) { + struct sockaddr_in *const addr4 = (struct sockaddr_in *)storage; + + addr4->sin_family = address_family; + addr4->sin_port = htons(port); + addr4->sin_addr.s_addr = htonl(INADDR_LOOPBACK); + } else { + struct sockaddr_in6 *const addr6 = + (struct sockaddr_in6 *)storage; + + addr6->sin6_family = address_family; + addr6->sin6_port = htons(port); + addr6->sin6_addr = in6addr_loopback; + } +} + +/* Verifies the actionable signed port for representative checked shapes. */ +TEST_F(trace_net_address, deny_access_net) +{ + const char *const event_regex = REGEX_DENY_ACCESS_NET(TRACE_TASK); + const unsigned short address_port = + variant->address_port_zero ? 0 : sock_port_start + 1; + const int expected_port = variant->expected_address_port ? + address_port : + variant->expected_port; + const struct landlock_ruleset_attr ruleset_attr = { + .handled_access_net = variant->handled_access, + }; + struct sockaddr_storage address; char field[64], expected[16]; + char *buf; + int count, ret, ruleset_fd, socket_fd, status; + pid_t child; if (!self->tracefs_ok) SKIP(return, "tracefs not available"); + set_trace_net_address(&address, variant->socket_family, + variant->address_family, address_port); + socket_fd = socket(variant->socket_family, + variant->socket_type | SOCK_CLOEXEC, 0); + ASSERT_LE(0, socket_fd); + ruleset_fd = + landlock_create_ruleset(&ruleset_attr, sizeof(ruleset_attr), 0); + ASSERT_LE(0, ruleset_fd); + ASSERT_EQ(0, tracefs_clear_buf()); + child = fork(); ASSERT_LE(0, child); - if (child == 0) { - struct landlock_ruleset_attr ruleset_attr = { - .handled_access_net = variant->handled, - }; - struct landlock_net_port_attr port_attr = { - .allowed_access = variant->handled, - .port = sock_port_start, - }; - struct sockaddr_in addr = { - .sin_family = AF_INET, - .sin_addr.s_addr = htonl(INADDR_LOOPBACK), - }; - int ruleset_fd, sock_fd, optval = 1, ret; - - ruleset_fd = landlock_create_ruleset(&ruleset_attr, - sizeof(ruleset_attr), 0); - if (ruleset_fd < 0) + if (prctl(PR_SET_NO_NEW_PRIVS, 1, 0, 0, 0)) _exit(1); - if (landlock_add_rule(ruleset_fd, LANDLOCK_RULE_NET_PORT, - &port_attr, 0)) { - close(ruleset_fd); - _exit(1); - } - prctl(PR_SET_NO_NEW_PRIVS, 1, 0, 0, 0); - if (landlock_restrict_self(ruleset_fd, 0)) { - close(ruleset_fd); - _exit(1); - } + if (landlock_restrict_self(ruleset_fd, 0)) + _exit(2); close(ruleset_fd); - sock_fd = socket(AF_INET, SOCK_STREAM | SOCK_CLOEXEC, 0); - if (sock_fd < 0) - _exit(1); - - /* Bind the allowed base port first (succeeds, no event). */ - if (variant->bind_base_first) { - setsockopt(sock_fd, SOL_SOCKET, SO_REUSEADDR, &optval, - sizeof(optval)); - addr.sin_port = htons(sock_port_start); - if (bind(sock_fd, (struct sockaddr *)&addr, - sizeof(addr))) { - close(sock_fd); - _exit(1); - } - } - - /* Denied operation on the next port. */ - addr.sin_port = htons(sock_port_start + 1); - if (variant->deny_connect) - ret = connect(sock_fd, (struct sockaddr *)&addr, - sizeof(addr)); - else - ret = bind(sock_fd, (struct sockaddr *)&addr, - sizeof(addr)); - if (ret == 0) { - close(sock_fd); - _exit(2); - } - if (errno != EACCES) { - close(sock_fd); + switch (variant->operation) { + case TRACE_NET_BIND: + ret = bind(socket_fd, (const struct sockaddr *)&address, + variant->addrlen); + break; + case TRACE_NET_SEND: + ret = sendto(socket_fd, "A", 1, MSG_NOSIGNAL, + (const struct sockaddr *)&address, + variant->addrlen); + break; + default: _exit(3); } - close(sock_fd); + if (ret >= 0 || errno != EACCES) + _exit(4); + close(socket_fd); + _exit(0); } + close(ruleset_fd); + close(socket_fd); ASSERT_EQ(child, waitpid(child, &status, 0)); ASSERT_TRUE(WIFEXITED(status)); - EXPECT_EQ(0, WEXITSTATUS(status)); + ASSERT_EQ(0, WEXITSTATUS(status)); buf = tracefs_read_buf(); ASSERT_NE(NULL, buf); + count = tracefs_count_matches(buf, event_regex); + if (count != 1) + TH_LOG("Expected 1 denial event, got %d\n%s", count, buf); + ASSERT_EQ(1, count); + expect_net_deny_common_fields(_metadata, buf, + variant->expected_blockers); - EXPECT_EQ(1, tracefs_count_matches(buf, - REGEX_DENY_ACCESS_NET(TRACE_TASK))); - - expect_net_deny_common_fields(_metadata, buf); - - snprintf(expected, sizeof(expected), "%llu", - (unsigned long long)(sock_port_start + 1)); - - ASSERT_EQ(0, - tracefs_extract_field(buf, REGEX_DENY_ACCESS_NET(TRACE_TASK), - "port", field, sizeof(field))); + ASSERT_EQ(0, tracefs_extract_field(buf, event_regex, "port", field, + sizeof(field))); + snprintf(expected, sizeof(expected), "%d", expected_port); EXPECT_STREQ(expected, field); free(buf); From 3fa5aa398edf94653926ad5e68b89f69490481b5 Mon Sep 17 00:00:00 2001 From: =?UTF-8?q?Micka=C3=ABl=20Sala=C3=BCn?= Date: Fri, 18 Sep 2026 20:50:32 +0200 Subject: [PATCH 432/592] landlock: Fix tracepoint contract documentation MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit The tracepoint documentation claims that denial and lifecycle events expose every input needed to reproduce a verdict. Instead document how denial, ruleset, and domain events identify the denying policy, checked operation and object, and reason for denial. Direct consumers to generic tracepoints for additional operational context. State the reconstruction limits: IDs are boot-local, rule checks have no request ID, and exported records may be lost or cross-CPU reordered. Also replace the incorrect BPF_RAW_TRACEPOINT guidance with libbpf SEC("tp_btf/...") attachment and refer consumers to the event prototypes for callback argument layouts. Cc: Günther Noack Cc: Steven Rostedt Link: https://patch.msgid.link/20260918185036.608651-10-mic@digikod.net Signed-off-by: Mickaël Salaün --- Documentation/trace/events-landlock.rst | 30 ++++++++++++--------- include/trace/events/landlock.h | 36 ++++++++++++------------- 2 files changed, 36 insertions(+), 30 deletions(-) diff --git a/Documentation/trace/events-landlock.rst b/Documentation/trace/events-landlock.rst index 9bb81a5c676e..304a8d06273e 100644 --- a/Documentation/trace/events-landlock.rst +++ b/Documentation/trace/events-landlock.rst @@ -141,10 +141,10 @@ in some field formats: treated as untrusted input and escaped in the trace text output so it cannot inject field separators or control characters. -- **Other party's domain**: A scope or ptrace denial compares the - subject's denying domain (``domain=``, always the enforcing domain and - never the current task) with the other party's domain, so these - tracepoints also report the other party's domain as a scalar ID: +- **Other party's domain**: A scope or ptrace denial compares the subject's + denying domain (``domain=``), which is the enforcing domain and not + necessarily the current task's domain, with the other party's domain. + These tracepoints also report the other party's domain as a scalar ID: ``tracee_domain=`` (ptrace), ``target_domain=`` (signal), and ``peer_domain=`` (abstract unix socket). It is ``0`` when the other party is unsandboxed, and otherwise a domain ID that a consumer resolves @@ -231,9 +231,10 @@ contribution, not the final decision: through another matching rule, or the right is denied and appears in the ``blockers=`` field of the corresponding ``deny_access`` event. -To reconstruct the decision for an object, aggregate the ``grants=`` -groups of all ``check_rule`` events emitted for that object during the -check. +For an access check that a consumer can delimit, aggregate the ``grants=`` +groups of all matching ``check_rule`` events. These events do not carry a +request ID; use their execution context and generic tracepoints to separate +concurrent or successive checks of the same object. .. note:: @@ -264,9 +265,11 @@ layers, so ``grants=`` has one group per layer:: eBPF access =========== -eBPF programs attached via ``BPF_RAW_TRACEPOINT`` can access the -tracepoint arguments directly through BTF. The arguments include both -standard kernel objects and Landlock-internal objects: +BTF-enabled raw tracepoint programs attached through libbpf +``SEC("tp_btf/...")`` sections receive typed callback arguments. The +event prototypes in `Event reference`_ document their argument layouts. +The arguments include both standard kernel objects and Landlock-internal +objects: - Standard kernel objects (``struct task_struct``, ``struct sock``, ``struct path``, ``struct dentry``) can be used with existing BPF @@ -277,7 +280,7 @@ standard kernel objects and Landlock-internal objects: Internal struct layouts may change between kernel versions; use CO-RE for field relocation. -A stateful eBPF program observes the full event stream and maintains +A stateful eBPF program attached before sandbox construction can maintain per-domain state in BPF maps: 1. On ``landlock_create_domain``: record the domain ID and parent (the @@ -293,7 +296,10 @@ per-domain state in BPF maps: final statistics. This approach requires no kernel modification and no Landlock-specific -BPF helpers. The Landlock IDs serve as correlation keys across events. +BPF helpers. Landlock IDs serve as correlation keys within one boot. +Records exported through tracing or BPF buffers can be lost, and records +from different CPUs are not globally ordered, so consumers must detect and +reconcile incomplete state. Audit filtering equivalence =========================== diff --git a/include/trace/events/landlock.h b/include/trace/events/landlock.h index 5da0f12ed2c3..523ba5ea9870 100644 --- a/include/trace/events/landlock.h +++ b/include/trace/events/landlock.h @@ -206,22 +206,22 @@ static inline const char *__trace_landlock_print_layers( * Decision context * ~~~~~~~~~~~~~~~~ * - * A denial event, together with the lifecycle events, exposes the full - * set of inputs the verdict consumed, so a consumer that tracked domain - * creation (landlock_create_ruleset, landlock_create_domain) can verify - * or reproduce the Landlock decision rather than merely observe it - * happened. In who/what/why terms: who is the denying domain (the domain - * field, always the subject that enforced the policy, never the current - * task), what is the operation and its object, and why is every other - * input the verdict weighed. + * A denial event identifies the domain whose policy denied the request, the + * Landlock operation and policy object that were checked, and the blocker or + * domain relationship responsible for the denial. When tracing starts with + * sandbox construction, ruleset and domain events provide the policy history + * needed to interpret these identifiers. The denying domain is the subject + * that enforced the policy, not necessarily current. Generic tracepoints can + * provide additional operational context. * * Lifecycle consistency * ~~~~~~~~~~~~~~~~~~~~~~ * - * Lifecycle events are balanced: a creation event always has a matching - * deallocation event and vice versa, so an eBPF program can model object - * lifetimes from the trace stream without reconciliation logic. A creation - * event fires while the object is still private to the calling thread + * Lifecycle emission is balanced: a creation event always has a matching + * deallocation event and vice versa. A consumer that observes an object's + * complete lifetime can model it from this pair; one that attaches late or + * loses exported records must reconcile incomplete state. A creation event + * fires while the object is still private to the calling thread * (landlock_create_ruleset fires before the ruleset's file descriptor is * installed, so it cannot race a concurrent :manpage:`close(2)`); if fd * installation later fails and the ruleset is freed, free_ruleset still @@ -295,7 +295,8 @@ static inline const char *__trace_landlock_print_layers( * the two parties without kernel-internal state. The ID is a scalar * snapshot, not a live domain pointer that could dangle: an optional * relational referent is a scalar (0 sentinel), not a nullable pointer. - * For ptrace, same_exec instead describes the tracer, even for + * Nonzero IDs are unique within one boot. For ptrace, same_exec instead + * describes the tracer, even for * PTRACE_TRACEME, and may differ from the current task. * * Blocker fields @@ -1031,11 +1032,10 @@ TRACE_EVENT(landlock_deny_scope_abstract_unix_socket, __entry->logged = logged; __entry->peer_domain_id = peer_domain_id; /* - * Best-effort (0 for a datagram peer). sk_peer_pid is - * canonically guarded by sk->sk_peer_lock, but the target - * peer's peercred is set once and not updated concurrently in - * these hooks, so this READ_ONCE() is safe; sun_path is the - * reliable identifier. + * Best-effort (0 for a datagram peer). The caller holds the + * peer's AF_UNIX state lock, serializing published peercred + * updates. The peer socket keeps a reference to sk_peer_pid + * through pid_nr(); sun_path is the reliable identifier. */ peer_pid = READ_ONCE(peer->sk_peer_pid); __entry->peer_pid = peer_pid ? pid_nr(peer_pid) : 0; From 6a5719cc3ef2e4d9857cc4ae18e6db09d59a8cc9 Mon Sep 17 00:00:00 2001 From: Daniel J Blueman Date: Sun, 20 Sep 2026 16:16:08 +0200 Subject: [PATCH 433/592] net: qrtr: resend HELLO on MHI resume Since the MHI HELLO exchange was relocated, it is sent only at device registration. During a suspend-resume cycle, the firmware in WiFi cards such as WCN7850 indefinitely waits for another HELLO, triggering: ath12k_wifi7_pci 0004:01:00.0: timeout while waiting for restart complete ath12k_wifi7_pci 0004:01:00.0: failed to resume core: -110 Fix this by triggering the handshake from resume_early in the MHI transport. Validated on Qualcomm X1E-801800 on Lenovo Slim 7x across 10 suspend-resume cycles. Fixes: 544d85de4dc2 ("net: qrtr: Send HELLO message on endpoint register") Signed-off-by: Daniel J Blueman Reviewed-by: Manivannan Sadhasivam Reported-by: Jeff Johnson Link: https://lore.kernel.org/all/6257c447-788d-4362-851e-0d552bcf7c56@oss.qualcomm.com/ Tested-by: Jeff Johnson Reported-by: Vlastimil Babka (SUSE) Link: https://lore.kernel.org/all/ab1491bb-cca5-4145-ac7d-31c966abf7b4@suse.com/ Tested-by: Vlastimil Babka (SUSE) Reported-by: Takashi Iwai Link: https://lore.kernel.org/all/87a4plsg4w.wl-tiwai@suse.de/ Tested-by: Takashi Iwai Signed-off-by: Thorsten Leemhuis Signed-off-by: Linus Torvalds --- net/qrtr/af_qrtr.c | 13 +++++++++++++ net/qrtr/mhi.c | 9 ++++++++- net/qrtr/qrtr.h | 2 ++ 3 files changed, 23 insertions(+), 1 deletion(-) diff --git a/net/qrtr/af_qrtr.c b/net/qrtr/af_qrtr.c index 78347c937af7..e7b3647424b8 100644 --- a/net/qrtr/af_qrtr.c +++ b/net/qrtr/af_qrtr.c @@ -623,6 +623,19 @@ static void qrtr_hello_work(struct work_struct *work) qrtr_port_put(ctrl); } +/* Trigger the HELLO handshake after the remote has been reset, eg on resume */ +void qrtr_endpoint_hello(struct qrtr_endpoint *ep) +{ + struct qrtr_node *node = ep->node; + + mutex_lock(&node->ep_lock); + node->hello_sent = false; + mutex_unlock(&node->ep_lock); + + schedule_delayed_work(&node->say_hello, 0); +} +EXPORT_SYMBOL_GPL(qrtr_endpoint_hello); + /** * qrtr_endpoint_register() - register a new endpoint * @ep: endpoint to register diff --git a/net/qrtr/mhi.c b/net/qrtr/mhi.c index 3990da1a65dc..e9a4bb92ce76 100644 --- a/net/qrtr/mhi.c +++ b/net/qrtr/mhi.c @@ -183,6 +183,7 @@ static int __maybe_unused qcom_mhi_qrtr_pm_suspend_late(struct device *dev) static int __maybe_unused qcom_mhi_qrtr_pm_resume_early(struct device *dev) { struct mhi_device *mhi_dev = container_of(dev, struct mhi_device, dev); + struct qrtr_mhi_dev *qdev = dev_get_drvdata(dev); enum mhi_state state; int rc; @@ -201,7 +202,13 @@ static int __maybe_unused qcom_mhi_qrtr_pm_resume_early(struct device *dev) return rc; } - return qcom_mhi_qrtr_queue_dl_buffers(mhi_dev); + rc = qcom_mhi_qrtr_queue_dl_buffers(mhi_dev); + if (rc) + return rc; + + qrtr_endpoint_hello(&qdev->ep); + + return 0; } static const struct dev_pm_ops qcom_mhi_qrtr_pm_ops = { diff --git a/net/qrtr/qrtr.h b/net/qrtr/qrtr.h index 3f2d28696062..de2de69a6199 100644 --- a/net/qrtr/qrtr.h +++ b/net/qrtr/qrtr.h @@ -27,6 +27,8 @@ int qrtr_endpoint_register(struct qrtr_endpoint *ep, unsigned int nid); void qrtr_endpoint_unregister(struct qrtr_endpoint *ep); +void qrtr_endpoint_hello(struct qrtr_endpoint *ep); + int qrtr_endpoint_post(struct qrtr_endpoint *ep, const void *data, size_t len); int qrtr_ns_init(void); From 93f51579e7df248780214094418f205253383cc5 Mon Sep 17 00:00:00 2001 From: Linus Torvalds Date: Sun, 20 Sep 2026 13:48:15 -0700 Subject: [PATCH 434/592] Linux 7.3-rc4 --- Makefile | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/Makefile b/Makefile index 0f1b80100b47..751a08643bf8 100644 --- a/Makefile +++ b/Makefile @@ -2,7 +2,7 @@ VERSION = 7 PATCHLEVEL = 3 SUBLEVEL = 0 -EXTRAVERSION = -rc3 +EXTRAVERSION = -rc4 NAME = Baby Opossum Posse # *DOCUMENTATION* From ae2c5bf969573708cd5b6bb6393631255d83c3fe Mon Sep 17 00:00:00 2001 From: Prathamesh Shete Date: Wed, 16 Sep 2026 06:51:32 +0000 Subject: [PATCH 435/592] pinctrl: tegra238: Fix register bank for AON pin groups The AON pin controller has a single register region and therefore, the bank defined in the tegra238_functions[] and tegra238_aon_groups[] for the AON pin groups must be 0. However, commit 25cac7292d49 ("pinctrl: tegra: Add Tegra238 pinmux driver") incorrectly specified the bank for these pins as 1 and not 0. This means that in the tegra_pinctrl_probe() function we use an invalid index when accessing the pmx->regs[] array which causes an incorrect address to be used for accessing the pinmux registers. Fix this by correcting the bank for the AON pin groups. Fixes: 25cac7292d49 ("pinctrl: tegra: Add Tegra238 pinmux driver") Signed-off-by: Prathamesh Shete Signed-off-by: Linus Walleij --- drivers/pinctrl/tegra/pinctrl-tegra238.c | 204 +++++++++++------------ 1 file changed, 102 insertions(+), 102 deletions(-) diff --git a/drivers/pinctrl/tegra/pinctrl-tegra238.c b/drivers/pinctrl/tegra/pinctrl-tegra238.c index ec482365f14f..40aba285944e 100644 --- a/drivers/pinctrl/tegra/pinctrl-tegra238.c +++ b/drivers/pinctrl/tegra/pinctrl-tegra238.c @@ -1744,57 +1744,57 @@ static const char * const tegra238_functions[] = { #define drive_sdmmc1_dat0_pu2 DRV_PINGROUP_ENTRY_Y(0x8034, 28, 2, 30, 2, -1, -1, -1, -1, 0) #define drive_ufs0_rst_n_pv1 DRV_PINGROUP_ENTRY_Y(0x11004, 12, 5, 24, 5, -1, -1, -1, -1, 0) #define drive_ufs0_ref_clk_pv0 DRV_PINGROUP_ENTRY_Y(0x1100c, 12, 5, 24, 5, -1, -1, -1, -1, 0) -#define drive_batt_oc_paa4 DRV_PINGROUP_ENTRY_Y(0x1024, 12, 5, 20, 5, -1, -1, -1, -1, 1) -#define drive_bootv_ctl_n_paa0 DRV_PINGROUP_ENTRY_Y(0x102c, 12, 5, 20, 5, -1, -1, -1, -1, 1) -#define drive_vcomp_alert_paa2 DRV_PINGROUP_ENTRY_Y(0x105c, 12, 5, 20, 5, -1, -1, -1, -1, 1) -#define drive_hdmi_cec_pbb0 DRV_PINGROUP_ENTRY_Y(0x1064, 12, 5, 20, 5, -1, -1, -1, -1, 1) -#define drive_touch_clk_pdd3 DRV_PINGROUP_ENTRY_Y(0x106c, 12, 5, 20, 5, -1, -1, -1, -1, 1) -#define drive_uart3_rx_pcc6 DRV_PINGROUP_ENTRY_Y(0x1074, 12, 5, 20, 5, -1, -1, -1, -1, 1) -#define drive_uart3_tx_pcc5 DRV_PINGROUP_ENTRY_Y(0x107c, 12, 5, 20, 5, -1, -1, -1, -1, 1) -#define drive_gen8_i2c_sda_pdd2 DRV_PINGROUP_ENTRY_Y(0x1084, 12, 5, 20, 5, -1, -1, -1, -1, 1) -#define drive_gen8_i2c_scl_pdd1 DRV_PINGROUP_ENTRY_Y(0x108c, 12, 5, 20, 5, -1, -1, -1, -1, 1) -#define drive_spi2_mosi_pcc2 DRV_PINGROUP_ENTRY_Y(0x1094, 12, 5, 20, 5, -1, -1, -1, -1, 1) -#define drive_gen2_i2c_scl_pcc7 DRV_PINGROUP_ENTRY_Y(0x109c, 12, 5, 20, 5, -1, -1, -1, -1, 1) -#define drive_spi2_cs0_pcc3 DRV_PINGROUP_ENTRY_Y(0x10a4, 12, 5, 20, 5, -1, -1, -1, -1, 1) -#define drive_gen2_i2c_sda_pdd0 DRV_PINGROUP_ENTRY_Y(0x10ac, 12, 5, 20, 5, -1, -1, -1, -1, 1) -#define drive_spi2_sck_pcc0 DRV_PINGROUP_ENTRY_Y(0x10b4, 12, 5, 20, 5, -1, -1, -1, -1, 1) -#define drive_spi2_miso_pcc1 DRV_PINGROUP_ENTRY_Y(0x10bc, 12, 5, 20, 5, -1, -1, -1, -1, 1) -#define drive_soc_gpio49_pee2 DRV_PINGROUP_ENTRY_Y(0x10c4, 12, 5, 20, 5, -1, -1, -1, -1, 1) -#define drive_soc_gpio50_pee4 DRV_PINGROUP_ENTRY_Y(0x10cc, 12, 5, 20, 5, -1, -1, -1, -1, 1) -#define drive_soc_gpio82_pee3 DRV_PINGROUP_ENTRY_Y(0x10d4, 12, 5, 20, 5, -1, -1, -1, -1, 1) -#define drive_soc_gpio71_pff2 DRV_PINGROUP_ENTRY_Y(0x10dc, 12, 5, 20, 5, -1, -1, -1, -1, 1) -#define drive_soc_gpio76_pff7 DRV_PINGROUP_ENTRY_Y(0x10e4, 12, 5, 20, 5, -1, -1, -1, -1, 1) -#define drive_soc_gpio74_pff5 DRV_PINGROUP_ENTRY_Y(0x10ec, 12, 5, 20, 5, -1, -1, -1, -1, 1) -#define drive_soc_gpio00_paa1 DRV_PINGROUP_ENTRY_Y(0x10f4, 12, 5, 20, 5, -1, -1, -1, -1, 1) -#define drive_soc_gpio19_pdd6 DRV_PINGROUP_ENTRY_Y(0x10fc, 12, 5, 20, 5, -1, -1, -1, -1, 1) -#define drive_soc_gpio86_phh3 DRV_PINGROUP_ENTRY_Y(0x1104, 12, 5, 20, 5, -1, -1, -1, -1, 1) -#define drive_soc_gpio72_pff3 DRV_PINGROUP_ENTRY_Y(0x110c, 12, 5, 20, 5, -1, -1, -1, -1, 1) -#define drive_soc_gpio77_pgg0 DRV_PINGROUP_ENTRY_Y(0x1114, 12, 5, 20, 5, -1, -1, -1, -1, 1) -#define drive_soc_gpio80_pff6 DRV_PINGROUP_ENTRY_Y(0x111c, 12, 5, 20, 5, -1, -1, -1, -1, 1) -#define drive_soc_gpio84_pgg1 DRV_PINGROUP_ENTRY_Y(0x1124, 12, 5, 20, 5, -1, -1, -1, -1, 1) -#define drive_soc_gpio83_pee5 DRV_PINGROUP_ENTRY_Y(0x112c, 12, 5, 20, 5, -1, -1, -1, -1, 1) -#define drive_soc_gpio73_pff4 DRV_PINGROUP_ENTRY_Y(0x1134, 12, 5, 20, 5, -1, -1, -1, -1, 1) -#define drive_soc_gpio70_pff1 DRV_PINGROUP_ENTRY_Y(0x113c, 12, 5, 20, 5, -1, -1, -1, -1, 1) -#define drive_soc_gpio04_paa5 DRV_PINGROUP_ENTRY_Y(0x1144, 12, 5, 20, 5, -1, -1, -1, -1, 1) -#define drive_soc_gpio85_pgg6 DRV_PINGROUP_ENTRY_Y(0x114c, 12, 5, 20, 5, -1, -1, -1, -1, 1) -#define drive_soc_gpio69_pff0 DRV_PINGROUP_ENTRY_Y(0x1154, 12, 5, 20, 5, -1, -1, -1, -1, 1) -#define drive_soc_gpio25_paa6 DRV_PINGROUP_ENTRY_Y(0x115c, 12, 5, 20, 5, -1, -1, -1, -1, 1) -#define drive_soc_gpio26_paa7 DRV_PINGROUP_ENTRY_Y(0x1164, 12, 5, 20, 5, -1, -1, -1, -1, 1) -#define drive_uart5_tx_pgg7 DRV_PINGROUP_ENTRY_Y(0x116c, 12, 5, 20, 5, -1, -1, -1, -1, 1) -#define drive_uart5_rx_phh0 DRV_PINGROUP_ENTRY_Y(0x1174, 12, 5, 20, 5, -1, -1, -1, -1, 1) -#define drive_uart2_tx_pgg2 DRV_PINGROUP_ENTRY_Y(0x117c, 12, 5, 20, 5, -1, -1, -1, -1, 1) -#define drive_uart2_rx_pgg3 DRV_PINGROUP_ENTRY_Y(0x1184, 12, 5, 20, 5, -1, -1, -1, -1, 1) -#define drive_uart2_cts_pgg5 DRV_PINGROUP_ENTRY_Y(0x118c, 12, 5, 20, 5, -1, -1, -1, -1, 1) -#define drive_uart2_rts_pgg4 DRV_PINGROUP_ENTRY_Y(0x1194, 12, 5, 20, 5, -1, -1, -1, -1, 1) -#define drive_uart5_cts_phh2 DRV_PINGROUP_ENTRY_Y(0x119c, 12, 5, 20, 5, -1, -1, -1, -1, 1) -#define drive_uart5_rts_phh1 DRV_PINGROUP_ENTRY_Y(0x11a4, 12, 5, 20, 5, -1, -1, -1, -1, 1) -#define drive_pwm7_pee1 DRV_PINGROUP_ENTRY_Y(0x11ac, 12, 5, 20, 5, -1, -1, -1, -1, 1) -#define drive_pwm2_pdd7 DRV_PINGROUP_ENTRY_Y(0x11b4, 12, 5, 20, 5, -1, -1, -1, -1, 1) -#define drive_pwm3_pee0 DRV_PINGROUP_ENTRY_Y(0x11bc, 12, 5, 20, 5, -1, -1, -1, -1, 1) -#define drive_pwm1_paa3 DRV_PINGROUP_ENTRY_Y(0x11c4, 12, 5, 20, 5, -1, -1, -1, -1, 1) -#define drive_spi2_cs1_pcc4 DRV_PINGROUP_ENTRY_Y(0x11cc, 12, 5, 20, 5, -1, -1, -1, -1, 1) -#define drive_dmic1_clk_pdd4 DRV_PINGROUP_ENTRY_Y(0x11d4, 12, 5, 20, 5, -1, -1, -1, -1, 1) -#define drive_dmic1_dat_pdd5 DRV_PINGROUP_ENTRY_Y(0x11dc, 12, 5, 20, 5, -1, -1, -1, -1, 1) +#define drive_batt_oc_paa4 DRV_PINGROUP_ENTRY_Y(0x1024, 12, 5, 20, 5, -1, -1, -1, -1, 0) +#define drive_bootv_ctl_n_paa0 DRV_PINGROUP_ENTRY_Y(0x102c, 12, 5, 20, 5, -1, -1, -1, -1, 0) +#define drive_vcomp_alert_paa2 DRV_PINGROUP_ENTRY_Y(0x105c, 12, 5, 20, 5, -1, -1, -1, -1, 0) +#define drive_hdmi_cec_pbb0 DRV_PINGROUP_ENTRY_Y(0x1064, 12, 5, 20, 5, -1, -1, -1, -1, 0) +#define drive_touch_clk_pdd3 DRV_PINGROUP_ENTRY_Y(0x106c, 12, 5, 20, 5, -1, -1, -1, -1, 0) +#define drive_uart3_rx_pcc6 DRV_PINGROUP_ENTRY_Y(0x1074, 12, 5, 20, 5, -1, -1, -1, -1, 0) +#define drive_uart3_tx_pcc5 DRV_PINGROUP_ENTRY_Y(0x107c, 12, 5, 20, 5, -1, -1, -1, -1, 0) +#define drive_gen8_i2c_sda_pdd2 DRV_PINGROUP_ENTRY_Y(0x1084, 12, 5, 20, 5, -1, -1, -1, -1, 0) +#define drive_gen8_i2c_scl_pdd1 DRV_PINGROUP_ENTRY_Y(0x108c, 12, 5, 20, 5, -1, -1, -1, -1, 0) +#define drive_spi2_mosi_pcc2 DRV_PINGROUP_ENTRY_Y(0x1094, 12, 5, 20, 5, -1, -1, -1, -1, 0) +#define drive_gen2_i2c_scl_pcc7 DRV_PINGROUP_ENTRY_Y(0x109c, 12, 5, 20, 5, -1, -1, -1, -1, 0) +#define drive_spi2_cs0_pcc3 DRV_PINGROUP_ENTRY_Y(0x10a4, 12, 5, 20, 5, -1, -1, -1, -1, 0) +#define drive_gen2_i2c_sda_pdd0 DRV_PINGROUP_ENTRY_Y(0x10ac, 12, 5, 20, 5, -1, -1, -1, -1, 0) +#define drive_spi2_sck_pcc0 DRV_PINGROUP_ENTRY_Y(0x10b4, 12, 5, 20, 5, -1, -1, -1, -1, 0) +#define drive_spi2_miso_pcc1 DRV_PINGROUP_ENTRY_Y(0x10bc, 12, 5, 20, 5, -1, -1, -1, -1, 0) +#define drive_soc_gpio49_pee2 DRV_PINGROUP_ENTRY_Y(0x10c4, 12, 5, 20, 5, -1, -1, -1, -1, 0) +#define drive_soc_gpio50_pee4 DRV_PINGROUP_ENTRY_Y(0x10cc, 12, 5, 20, 5, -1, -1, -1, -1, 0) +#define drive_soc_gpio82_pee3 DRV_PINGROUP_ENTRY_Y(0x10d4, 12, 5, 20, 5, -1, -1, -1, -1, 0) +#define drive_soc_gpio71_pff2 DRV_PINGROUP_ENTRY_Y(0x10dc, 12, 5, 20, 5, -1, -1, -1, -1, 0) +#define drive_soc_gpio76_pff7 DRV_PINGROUP_ENTRY_Y(0x10e4, 12, 5, 20, 5, -1, -1, -1, -1, 0) +#define drive_soc_gpio74_pff5 DRV_PINGROUP_ENTRY_Y(0x10ec, 12, 5, 20, 5, -1, -1, -1, -1, 0) +#define drive_soc_gpio00_paa1 DRV_PINGROUP_ENTRY_Y(0x10f4, 12, 5, 20, 5, -1, -1, -1, -1, 0) +#define drive_soc_gpio19_pdd6 DRV_PINGROUP_ENTRY_Y(0x10fc, 12, 5, 20, 5, -1, -1, -1, -1, 0) +#define drive_soc_gpio86_phh3 DRV_PINGROUP_ENTRY_Y(0x1104, 12, 5, 20, 5, -1, -1, -1, -1, 0) +#define drive_soc_gpio72_pff3 DRV_PINGROUP_ENTRY_Y(0x110c, 12, 5, 20, 5, -1, -1, -1, -1, 0) +#define drive_soc_gpio77_pgg0 DRV_PINGROUP_ENTRY_Y(0x1114, 12, 5, 20, 5, -1, -1, -1, -1, 0) +#define drive_soc_gpio80_pff6 DRV_PINGROUP_ENTRY_Y(0x111c, 12, 5, 20, 5, -1, -1, -1, -1, 0) +#define drive_soc_gpio84_pgg1 DRV_PINGROUP_ENTRY_Y(0x1124, 12, 5, 20, 5, -1, -1, -1, -1, 0) +#define drive_soc_gpio83_pee5 DRV_PINGROUP_ENTRY_Y(0x112c, 12, 5, 20, 5, -1, -1, -1, -1, 0) +#define drive_soc_gpio73_pff4 DRV_PINGROUP_ENTRY_Y(0x1134, 12, 5, 20, 5, -1, -1, -1, -1, 0) +#define drive_soc_gpio70_pff1 DRV_PINGROUP_ENTRY_Y(0x113c, 12, 5, 20, 5, -1, -1, -1, -1, 0) +#define drive_soc_gpio04_paa5 DRV_PINGROUP_ENTRY_Y(0x1144, 12, 5, 20, 5, -1, -1, -1, -1, 0) +#define drive_soc_gpio85_pgg6 DRV_PINGROUP_ENTRY_Y(0x114c, 12, 5, 20, 5, -1, -1, -1, -1, 0) +#define drive_soc_gpio69_pff0 DRV_PINGROUP_ENTRY_Y(0x1154, 12, 5, 20, 5, -1, -1, -1, -1, 0) +#define drive_soc_gpio25_paa6 DRV_PINGROUP_ENTRY_Y(0x115c, 12, 5, 20, 5, -1, -1, -1, -1, 0) +#define drive_soc_gpio26_paa7 DRV_PINGROUP_ENTRY_Y(0x1164, 12, 5, 20, 5, -1, -1, -1, -1, 0) +#define drive_uart5_tx_pgg7 DRV_PINGROUP_ENTRY_Y(0x116c, 12, 5, 20, 5, -1, -1, -1, -1, 0) +#define drive_uart5_rx_phh0 DRV_PINGROUP_ENTRY_Y(0x1174, 12, 5, 20, 5, -1, -1, -1, -1, 0) +#define drive_uart2_tx_pgg2 DRV_PINGROUP_ENTRY_Y(0x117c, 12, 5, 20, 5, -1, -1, -1, -1, 0) +#define drive_uart2_rx_pgg3 DRV_PINGROUP_ENTRY_Y(0x1184, 12, 5, 20, 5, -1, -1, -1, -1, 0) +#define drive_uart2_cts_pgg5 DRV_PINGROUP_ENTRY_Y(0x118c, 12, 5, 20, 5, -1, -1, -1, -1, 0) +#define drive_uart2_rts_pgg4 DRV_PINGROUP_ENTRY_Y(0x1194, 12, 5, 20, 5, -1, -1, -1, -1, 0) +#define drive_uart5_cts_phh2 DRV_PINGROUP_ENTRY_Y(0x119c, 12, 5, 20, 5, -1, -1, -1, -1, 0) +#define drive_uart5_rts_phh1 DRV_PINGROUP_ENTRY_Y(0x11a4, 12, 5, 20, 5, -1, -1, -1, -1, 0) +#define drive_pwm7_pee1 DRV_PINGROUP_ENTRY_Y(0x11ac, 12, 5, 20, 5, -1, -1, -1, -1, 0) +#define drive_pwm2_pdd7 DRV_PINGROUP_ENTRY_Y(0x11b4, 12, 5, 20, 5, -1, -1, -1, -1, 0) +#define drive_pwm3_pee0 DRV_PINGROUP_ENTRY_Y(0x11bc, 12, 5, 20, 5, -1, -1, -1, -1, 0) +#define drive_pwm1_paa3 DRV_PINGROUP_ENTRY_Y(0x11c4, 12, 5, 20, 5, -1, -1, -1, -1, 0) +#define drive_spi2_cs1_pcc4 DRV_PINGROUP_ENTRY_Y(0x11cc, 12, 5, 20, 5, -1, -1, -1, -1, 0) +#define drive_dmic1_clk_pdd4 DRV_PINGROUP_ENTRY_Y(0x11d4, 12, 5, 20, 5, -1, -1, -1, -1, 0) +#define drive_dmic1_dat_pdd5 DRV_PINGROUP_ENTRY_Y(0x11dc, 12, 5, 20, 5, -1, -1, -1, -1, 0) #define drive_sdmmc1_comp DRV_PINGROUP_ENTRY_N @@ -1961,57 +1961,57 @@ static const struct tegra_pingroup tegra238_groups[] = { }; static const struct tegra_pingroup tegra238_aon_groups[] = { - PINGROUP(bootv_ctl_n_paa0, RSVD0, RSVD1, RSVD2, RSVD3, 0x1028, 1, Y, -1, 7, 6, 8, -1, 10, 12), - PINGROUP(soc_gpio00_paa1, RSVD0, RSVD1, RSVD2, RSVD3, 0x10f0, 1, Y, 5, 7, 6, 8, -1, 10, 12), - PINGROUP(vcomp_alert_paa2, SOC_THERM_OC1, RSVD1, RSVD2, RSVD3, 0x1058, 1, Y, 5, 7, 6, 8, -1, 10, 12), - PINGROUP(pwm1_paa3, GP_PWM1, RSVD1, RSVD2, RSVD3, 0x11c0, 1, Y, 5, 7, 6, 8, -1, 10, 12), - PINGROUP(batt_oc_paa4, SOC_THERM_OC2, RSVD1, RSVD2, RSVD3, 0x1020, 1, Y, 5, 7, 6, 8, -1, 10, 12), - PINGROUP(soc_gpio04_paa5, RSVD0, RSVD1, RSVD2, RSVD3, 0x1140, 1, Y, 5, 7, 6, 8, -1, 10, 12), - PINGROUP(soc_gpio25_paa6, RSVD0, RSVD1, RSVD2, RSVD3, 0x1158, 1, Y, -1, 7, 6, 8, -1, 10, 12), - PINGROUP(soc_gpio26_paa7, RSVD0, SOC_THERM_OC3, RSVD2, RSVD3, 0x1160, 1, Y, -1, 7, 6, 8, -1, 10, 12), - PINGROUP(hdmi_cec_pbb0, HDMI_CEC, RSVD1, RSVD2, RSVD3, 0x1060, 1, Y, 5, 7, 6, 8, -1, 10, 12), - PINGROUP(spi2_sck_pcc0, SPI2_SCK, RSVD1, RSVD2, RSVD3, 0x10b0, 1, Y, -1, 7, 6, 8, -1, 10, 12), - PINGROUP(spi2_miso_pcc1, SPI2_DIN, RSVD1, RSVD2, RSVD3, 0x10b8, 1, Y, -1, 7, 6, 8, -1, 10, 12), - PINGROUP(spi2_mosi_pcc2, SPI2_DOUT, RSVD1, RSVD2, RSVD3, 0x1090, 1, Y, -1, 7, 6, 8, -1, 10, 12), - PINGROUP(spi2_cs0_pcc3, SPI2_CS0, RSVD1, RSVD2, RSVD3, 0x10a0, 1, Y, -1, 7, 6, 8, -1, 10, 12), - PINGROUP(spi2_cs1_pcc4, SPI2_CS1, RSVD1, RSVD2, RSVD3, 0x11c8, 1, Y, -1, 7, 6, 8, -1, 10, 12), - PINGROUP(uart3_tx_pcc5, UARTC_TXD, RSVD1, RSVD2, RSVD3, 0x1078, 1, Y, 5, 7, 6, 8, -1, 10, 12), - PINGROUP(uart3_rx_pcc6, UARTC_RXD, RSVD1, RSVD2, RSVD3, 0x1070, 1, Y, 5, 7, 6, 8, -1, 10, 12), - PINGROUP(gen2_i2c_scl_pcc7, I2C2_CLK, RSVD1, RSVD2, RSVD3, 0x1098, 1, Y, 5, 7, 6, 8, -1, 10, 12), - PINGROUP(gen2_i2c_sda_pdd0, I2C2_DAT, RSVD1, RSVD2, RSVD3, 0x10a8, 1, Y, 5, 7, 6, 8, -1, 10, 12), - PINGROUP(gen8_i2c_scl_pdd1, I2C8_CLK, RSVD1, RSVD2, RSVD3, 0x1088, 1, Y, 5, 7, 6, 8, -1, 10, 12), - PINGROUP(gen8_i2c_sda_pdd2, I2C8_DAT, RSVD1, RSVD2, RSVD3, 0x1080, 1, Y, 5, 7, 6, 8, -1, 10, 12), - PINGROUP(touch_clk_pdd3, GP_PWM4, TOUCH_CLK, RSVD2, RSVD3, 0x1068, 1, Y, 5, 7, 6, 8, -1, 10, 12), - PINGROUP(dmic1_clk_pdd4, DMIC1_CLK, RSVD1, DMIC5_CLK, RSVD3, 0x11d0, 1, Y, -1, 7, 6, 8, -1, 10, 12), - PINGROUP(dmic1_dat_pdd5, DMIC1_DAT, RSVD1, DMIC5_DAT, RSVD3, 0x11d8, 1, Y, -1, 7, 6, 8, -1, 10, 12), - PINGROUP(soc_gpio19_pdd6, RSVD0, WDT_RESET_OUTB, RSVD2, RSVD3, 0x10f8, 1, Y, -1, 7, 6, 8, -1, 10, 12), - PINGROUP(soc_gpio49_pee2, RSVD0, RSVD1, RSVD2, RSVD3, 0x10c0, 1, Y, -1, 7, 6, 8, -1, 10, 12), - PINGROUP(soc_gpio50_pee4, RSVD0, RSVD1, RSVD2, RSVD3, 0x10c8, 1, Y, -1, 7, 6, 8, -1, 10, 12), - PINGROUP(soc_gpio82_pee3, RSVD0, RSVD1, RSVD2, RSVD3, 0x10d0, 1, Y, -1, 7, 6, 8, -1, 10, 12), - PINGROUP(soc_gpio71_pff2, PPC_MODE_1, RSVD1, RSVD2, RSVD3, 0x10d8, 1, Y, 5, 7, 6, 8, -1, 10, 12), - PINGROUP(soc_gpio76_pff7, RSVD0, RSVD1, TSC_EDGE_OUT0, TSC_EDGE_OUT0A, 0x10e0, 1, Y, -1, 7, 6, 8, -1, 10, 12), - PINGROUP(soc_gpio74_pff5, PPC_READY, PPC_I2C_DAT, RSVD2, RSVD3, 0x10e8, 1, Y, 5, 7, 6, 8, -1, 10, 12), - PINGROUP(soc_gpio86_phh3, RSVD0, SPI5_CS1, TSC_EDGE_OUT3, TSC_EDGE_OUT0D, 0x1100, 1, Y, 5, 7, 6, 8, -1, 10, 12), - PINGROUP(soc_gpio72_pff3, PPC_MODE_2, RSVD1, RSVD2, RSVD3, 0x1108, 1, Y, 5, 7, 6, 8, -1, 10, 12), - PINGROUP(soc_gpio77_pgg0, RSVD0, RSVD1, TSC_EDGE_OUT1, TSC_EDGE_OUT0B, 0x1110, 1, Y, -1, 7, 6, 8, -1, 10, 12), - PINGROUP(soc_gpio80_pff6, RSVD0, PPC_RST_N, RSVD2, RSVD3, 0x1118, 1, Y, 5, 7, 6, 8, -1, 10, 12), - PINGROUP(soc_gpio84_pgg1, RSVD0, RSVD1, TSC_EDGE_OUT2, TSC_EDGE_OUT0C, 0x1120, 1, Y, -1, 7, 6, 8, -1, 10, 12), - PINGROUP(soc_gpio83_pee5, RSVD0, RSVD1, RSVD2, RSVD3, 0x1128, 1, Y, -1, 7, 6, 8, -1, 10, 12), - PINGROUP(soc_gpio73_pff4, PPC_CC, PPC_I2C_CLK, RSVD2, RSVD3, 0x1130, 1, Y, 5, 7, 6, 8, -1, 10, 12), - PINGROUP(soc_gpio70_pff1, PPC_MODE_0, RSVD1, RSVD2, RSVD3, 0x1138, 1, Y, 5, 7, 6, 8, -1, 10, 12), - PINGROUP(soc_gpio85_pgg6, RSVD0, SPI4_CS1, RSVD2, RSVD3, 0x1148, 1, Y, 5, 7, 6, 8, -1, 10, 12), - PINGROUP(soc_gpio69_pff0, PPC_INT_N, RSVD1, RSVD2, RSVD3, 0x1150, 1, Y, 5, 7, 6, 8, -1, 10, 12), - PINGROUP(uart5_tx_pgg7, UARTE_TXD, SPI5_SCK, RSVD2, RSVD3, 0x1168, 1, Y, 5, 7, 6, 8, -1, 10, 12), - PINGROUP(uart5_rx_phh0, UARTE_RXD, SPI5_MISO, RSVD2, RSVD3, 0x1170, 1, Y, 5, 7, 6, 8, -1, 10, 12), - PINGROUP(uart2_tx_pgg2, UARTB_TXD, SPI4_SCK, RSVD2, RSVD3, 0x1178, 1, Y, 5, 7, 6, 8, -1, 10, 12), - PINGROUP(uart2_rx_pgg3, UARTB_RXD, SPI4_MISO, RSVD2, RSVD3, 0x1180, 1, Y, 5, 7, 6, 8, -1, 10, 12), - PINGROUP(uart2_cts_pgg5, UARTB_CTS, SPI4_CS0, RSVD2, RSVD3, 0x1188, 1, Y, 5, 7, 6, 8, -1, 10, 12), - PINGROUP(uart2_rts_pgg4, UARTB_RTS, SPI4_MOSI, RSVD2, RSVD3, 0x1190, 1, Y, 5, 7, 6, 8, -1, 10, 12), - PINGROUP(uart5_cts_phh2, UARTE_CTS, SPI5_CS0, RSVD2, RSVD3, 0x1198, 1, Y, 5, 7, 6, 8, -1, 10, 12), - PINGROUP(uart5_rts_phh1, UARTE_RTS, SPI5_MOSI, RSVD2, RSVD3, 0x11a0, 1, Y, 5, 7, 6, 8, -1, 10, 12), - PINGROUP(pwm2_pdd7, GP_PWM2, LED_BLINK, RSVD2, RSVD3, 0x11b0, 1, Y, 5, 7, 6, 8, -1, 10, 12), - PINGROUP(pwm3_pee0, GP_PWM3, RSVD1, RSVD2, RSVD3, 0x11b8, 1, Y, 5, 7, 6, 8, -1, 10, 12), - PINGROUP(pwm7_pee1, GP_PWM7, RSVD1, RSVD2, RSVD3, 0x11a8, 1, Y, 5, 7, 6, 8, -1, 10, 12), + PINGROUP(bootv_ctl_n_paa0, RSVD0, RSVD1, RSVD2, RSVD3, 0x1028, 0, Y, -1, 7, 6, 8, -1, 10, 12), + PINGROUP(soc_gpio00_paa1, RSVD0, RSVD1, RSVD2, RSVD3, 0x10f0, 0, Y, 5, 7, 6, 8, -1, 10, 12), + PINGROUP(vcomp_alert_paa2, SOC_THERM_OC1, RSVD1, RSVD2, RSVD3, 0x1058, 0, Y, 5, 7, 6, 8, -1, 10, 12), + PINGROUP(pwm1_paa3, GP_PWM1, RSVD1, RSVD2, RSVD3, 0x11c0, 0, Y, 5, 7, 6, 8, -1, 10, 12), + PINGROUP(batt_oc_paa4, SOC_THERM_OC2, RSVD1, RSVD2, RSVD3, 0x1020, 0, Y, 5, 7, 6, 8, -1, 10, 12), + PINGROUP(soc_gpio04_paa5, RSVD0, RSVD1, RSVD2, RSVD3, 0x1140, 0, Y, 5, 7, 6, 8, -1, 10, 12), + PINGROUP(soc_gpio25_paa6, RSVD0, RSVD1, RSVD2, RSVD3, 0x1158, 0, Y, -1, 7, 6, 8, -1, 10, 12), + PINGROUP(soc_gpio26_paa7, RSVD0, SOC_THERM_OC3, RSVD2, RSVD3, 0x1160, 0, Y, -1, 7, 6, 8, -1, 10, 12), + PINGROUP(hdmi_cec_pbb0, HDMI_CEC, RSVD1, RSVD2, RSVD3, 0x1060, 0, Y, 5, 7, 6, 8, -1, 10, 12), + PINGROUP(spi2_sck_pcc0, SPI2_SCK, RSVD1, RSVD2, RSVD3, 0x10b0, 0, Y, -1, 7, 6, 8, -1, 10, 12), + PINGROUP(spi2_miso_pcc1, SPI2_DIN, RSVD1, RSVD2, RSVD3, 0x10b8, 0, Y, -1, 7, 6, 8, -1, 10, 12), + PINGROUP(spi2_mosi_pcc2, SPI2_DOUT, RSVD1, RSVD2, RSVD3, 0x1090, 0, Y, -1, 7, 6, 8, -1, 10, 12), + PINGROUP(spi2_cs0_pcc3, SPI2_CS0, RSVD1, RSVD2, RSVD3, 0x10a0, 0, Y, -1, 7, 6, 8, -1, 10, 12), + PINGROUP(spi2_cs1_pcc4, SPI2_CS1, RSVD1, RSVD2, RSVD3, 0x11c8, 0, Y, -1, 7, 6, 8, -1, 10, 12), + PINGROUP(uart3_tx_pcc5, UARTC_TXD, RSVD1, RSVD2, RSVD3, 0x1078, 0, Y, 5, 7, 6, 8, -1, 10, 12), + PINGROUP(uart3_rx_pcc6, UARTC_RXD, RSVD1, RSVD2, RSVD3, 0x1070, 0, Y, 5, 7, 6, 8, -1, 10, 12), + PINGROUP(gen2_i2c_scl_pcc7, I2C2_CLK, RSVD1, RSVD2, RSVD3, 0x1098, 0, Y, 5, 7, 6, 8, -1, 10, 12), + PINGROUP(gen2_i2c_sda_pdd0, I2C2_DAT, RSVD1, RSVD2, RSVD3, 0x10a8, 0, Y, 5, 7, 6, 8, -1, 10, 12), + PINGROUP(gen8_i2c_scl_pdd1, I2C8_CLK, RSVD1, RSVD2, RSVD3, 0x1088, 0, Y, 5, 7, 6, 8, -1, 10, 12), + PINGROUP(gen8_i2c_sda_pdd2, I2C8_DAT, RSVD1, RSVD2, RSVD3, 0x1080, 0, Y, 5, 7, 6, 8, -1, 10, 12), + PINGROUP(touch_clk_pdd3, GP_PWM4, TOUCH_CLK, RSVD2, RSVD3, 0x1068, 0, Y, 5, 7, 6, 8, -1, 10, 12), + PINGROUP(dmic1_clk_pdd4, DMIC1_CLK, RSVD1, DMIC5_CLK, RSVD3, 0x11d0, 0, Y, -1, 7, 6, 8, -1, 10, 12), + PINGROUP(dmic1_dat_pdd5, DMIC1_DAT, RSVD1, DMIC5_DAT, RSVD3, 0x11d8, 0, Y, -1, 7, 6, 8, -1, 10, 12), + PINGROUP(soc_gpio19_pdd6, RSVD0, WDT_RESET_OUTB, RSVD2, RSVD3, 0x10f8, 0, Y, -1, 7, 6, 8, -1, 10, 12), + PINGROUP(soc_gpio49_pee2, RSVD0, RSVD1, RSVD2, RSVD3, 0x10c0, 0, Y, -1, 7, 6, 8, -1, 10, 12), + PINGROUP(soc_gpio50_pee4, RSVD0, RSVD1, RSVD2, RSVD3, 0x10c8, 0, Y, -1, 7, 6, 8, -1, 10, 12), + PINGROUP(soc_gpio82_pee3, RSVD0, RSVD1, RSVD2, RSVD3, 0x10d0, 0, Y, -1, 7, 6, 8, -1, 10, 12), + PINGROUP(soc_gpio71_pff2, PPC_MODE_1, RSVD1, RSVD2, RSVD3, 0x10d8, 0, Y, 5, 7, 6, 8, -1, 10, 12), + PINGROUP(soc_gpio76_pff7, RSVD0, RSVD1, TSC_EDGE_OUT0, TSC_EDGE_OUT0A, 0x10e0, 0, Y, -1, 7, 6, 8, -1, 10, 12), + PINGROUP(soc_gpio74_pff5, PPC_READY, PPC_I2C_DAT, RSVD2, RSVD3, 0x10e8, 0, Y, 5, 7, 6, 8, -1, 10, 12), + PINGROUP(soc_gpio86_phh3, RSVD0, SPI5_CS1, TSC_EDGE_OUT3, TSC_EDGE_OUT0D, 0x1100, 0, Y, 5, 7, 6, 8, -1, 10, 12), + PINGROUP(soc_gpio72_pff3, PPC_MODE_2, RSVD1, RSVD2, RSVD3, 0x1108, 0, Y, 5, 7, 6, 8, -1, 10, 12), + PINGROUP(soc_gpio77_pgg0, RSVD0, RSVD1, TSC_EDGE_OUT1, TSC_EDGE_OUT0B, 0x1110, 0, Y, -1, 7, 6, 8, -1, 10, 12), + PINGROUP(soc_gpio80_pff6, RSVD0, PPC_RST_N, RSVD2, RSVD3, 0x1118, 0, Y, 5, 7, 6, 8, -1, 10, 12), + PINGROUP(soc_gpio84_pgg1, RSVD0, RSVD1, TSC_EDGE_OUT2, TSC_EDGE_OUT0C, 0x1120, 0, Y, -1, 7, 6, 8, -1, 10, 12), + PINGROUP(soc_gpio83_pee5, RSVD0, RSVD1, RSVD2, RSVD3, 0x1128, 0, Y, -1, 7, 6, 8, -1, 10, 12), + PINGROUP(soc_gpio73_pff4, PPC_CC, PPC_I2C_CLK, RSVD2, RSVD3, 0x1130, 0, Y, 5, 7, 6, 8, -1, 10, 12), + PINGROUP(soc_gpio70_pff1, PPC_MODE_0, RSVD1, RSVD2, RSVD3, 0x1138, 0, Y, 5, 7, 6, 8, -1, 10, 12), + PINGROUP(soc_gpio85_pgg6, RSVD0, SPI4_CS1, RSVD2, RSVD3, 0x1148, 0, Y, 5, 7, 6, 8, -1, 10, 12), + PINGROUP(soc_gpio69_pff0, PPC_INT_N, RSVD1, RSVD2, RSVD3, 0x1150, 0, Y, 5, 7, 6, 8, -1, 10, 12), + PINGROUP(uart5_tx_pgg7, UARTE_TXD, SPI5_SCK, RSVD2, RSVD3, 0x1168, 0, Y, 5, 7, 6, 8, -1, 10, 12), + PINGROUP(uart5_rx_phh0, UARTE_RXD, SPI5_MISO, RSVD2, RSVD3, 0x1170, 0, Y, 5, 7, 6, 8, -1, 10, 12), + PINGROUP(uart2_tx_pgg2, UARTB_TXD, SPI4_SCK, RSVD2, RSVD3, 0x1178, 0, Y, 5, 7, 6, 8, -1, 10, 12), + PINGROUP(uart2_rx_pgg3, UARTB_RXD, SPI4_MISO, RSVD2, RSVD3, 0x1180, 0, Y, 5, 7, 6, 8, -1, 10, 12), + PINGROUP(uart2_cts_pgg5, UARTB_CTS, SPI4_CS0, RSVD2, RSVD3, 0x1188, 0, Y, 5, 7, 6, 8, -1, 10, 12), + PINGROUP(uart2_rts_pgg4, UARTB_RTS, SPI4_MOSI, RSVD2, RSVD3, 0x1190, 0, Y, 5, 7, 6, 8, -1, 10, 12), + PINGROUP(uart5_cts_phh2, UARTE_CTS, SPI5_CS0, RSVD2, RSVD3, 0x1198, 0, Y, 5, 7, 6, 8, -1, 10, 12), + PINGROUP(uart5_rts_phh1, UARTE_RTS, SPI5_MOSI, RSVD2, RSVD3, 0x11a0, 0, Y, 5, 7, 6, 8, -1, 10, 12), + PINGROUP(pwm2_pdd7, GP_PWM2, LED_BLINK, RSVD2, RSVD3, 0x11b0, 0, Y, 5, 7, 6, 8, -1, 10, 12), + PINGROUP(pwm3_pee0, GP_PWM3, RSVD1, RSVD2, RSVD3, 0x11b8, 0, Y, 5, 7, 6, 8, -1, 10, 12), + PINGROUP(pwm7_pee1, GP_PWM7, RSVD1, RSVD2, RSVD3, 0x11a8, 0, Y, 5, 7, 6, 8, -1, 10, 12), }; static const struct tegra_pinctrl_soc_data tegra238_pinctrl_aon = { From 7a6d08ee0f0e30023d18779bb314db8fd9a3b6d4 Mon Sep 17 00:00:00 2001 From: Ralf Lici Date: Fri, 28 Aug 2026 16:50:22 +0200 Subject: [PATCH 436/592] ovpn: preserve IPv6 scope id for netlink peer endpoints ovpn accepts OVPN_A_PEER_REMOTE_IPV6_SCOPE_ID and reports bind->remote.in6.sin6_scope_id in peer dumps, but the netlink endpoint parser never copied the attribute into the sockaddr_in6 used to create or update the peer bind. As a result, an IPv6 link-local remote endpoint configured through netlink loses its interface scope, unlike on the peer float path where ipv6_iface_scope_id populates the field. The UDPv6 output path then builds a flow with flowi6_oif set to zero and route lookup can fail or select the wrong interface. Copy the scope id when parsing non-v4-mapped IPv6 remote endpoints. The existing precheck already rejects the scope-id attribute for IPv4 and v4-mapped IPv6 remotes. Fixes: 1d36a36f6d53 ("ovpn: implement peer add/get/dump/delete via netlink") Signed-off-by: Ralf Lici Signed-off-by: Antonio Quartulli --- drivers/net/ovpn/netlink.c | 6 ++++++ 1 file changed, 6 insertions(+) diff --git a/drivers/net/ovpn/netlink.c b/drivers/net/ovpn/netlink.c index 4dad85294198..2ba762082acc 100644 --- a/drivers/net/ovpn/netlink.c +++ b/drivers/net/ovpn/netlink.c @@ -100,6 +100,8 @@ static bool ovpn_nl_attr_sockaddr_remote(struct nlattr **attrs, struct sockaddr_in6 *sin6; struct sockaddr_in *sin; struct in6_addr *in6; + struct nlattr *scope; + u32 scope_id = 0; __be16 port = 0; __be32 *in; @@ -114,6 +116,9 @@ static bool ovpn_nl_attr_sockaddr_remote(struct nlattr **attrs, } else if (attrs[OVPN_A_PEER_REMOTE_IPV6]) { ss->ss_family = AF_INET6; in6 = nla_data(attrs[OVPN_A_PEER_REMOTE_IPV6]); + scope = attrs[OVPN_A_PEER_REMOTE_IPV6_SCOPE_ID]; + if (scope) + scope_id = nla_get_u32(scope); } else { return false; } @@ -126,6 +131,7 @@ static bool ovpn_nl_attr_sockaddr_remote(struct nlattr **attrs, if (!ipv6_addr_v4mapped(in6)) { sin6 = (struct sockaddr_in6 *)ss; sin6->sin6_port = port; + sin6->sin6_scope_id = scope_id; memcpy(&sin6->sin6_addr, in6, sizeof(*in6)); break; } From 77393b4d72dfeb764b2af2b848acc659f6fcfd0a Mon Sep 17 00:00:00 2001 From: Ralf Lici Date: Fri, 28 Aug 2026 16:50:23 +0200 Subject: [PATCH 437/592] ovpn: skip UDP source validation for unspecified addresses ovpn validates the cached local UDP source address before reusing or refreshing a peer dst cache. This is only meaningful when a concrete source address is selected. For IPv6, calling ipv6_chk_addr with :: checks whether the unspecified address itself is configured on the host. A peer may legitimately have bind->local.ipv6 set to :: when no local endpoint was configured or after a stale learned address was cleared. In that case the source should be left unspecified and selected by ip6_dst_lookup_flow(). For IPv4, inet_confirm_addr(..., local = 0, ...) asks for local address autoselection rather than validating a chosen source. Skip the precheck there as well and let ip_route_output_flow select or reject the source. Only validate non-zero/non-any source addresses. Fixes: 08857b5ec5d9 ("ovpn: implement basic TX path (UDP)") Signed-off-by: Ralf Lici Signed-off-by: Antonio Quartulli --- drivers/net/ovpn/udp.c | 7 ++++--- 1 file changed, 4 insertions(+), 3 deletions(-) diff --git a/drivers/net/ovpn/udp.c b/drivers/net/ovpn/udp.c index 7f69e8890b5b..df4750dabd1e 100644 --- a/drivers/net/ovpn/udp.c +++ b/drivers/net/ovpn/udp.c @@ -161,8 +161,8 @@ static int ovpn_udp4_output(struct ovpn_peer *peer, struct ovpn_bind *bind, if (rt) goto transmit; - if (unlikely(!inet_confirm_addr(sock_net(sk), NULL, 0, fl.saddr, - RT_SCOPE_HOST))) { + if (fl.saddr && unlikely(!inet_confirm_addr(sock_net(sk), NULL, 0, + fl.saddr, RT_SCOPE_HOST))) { /* we may end up here when the cached address is not usable * anymore. In this case we reset address/cache and perform a * new look up @@ -238,7 +238,8 @@ static int ovpn_udp6_output(struct ovpn_peer *peer, struct ovpn_bind *bind, if (dst) goto transmit; - if (unlikely(!ipv6_chk_addr(sock_net(sk), &fl.saddr, NULL, 0))) { + if (!ipv6_addr_any(&fl.saddr) && + unlikely(!ipv6_chk_addr(sock_net(sk), &fl.saddr, NULL, 0))) { /* we may end up here when the cached address is not usable * anymore. In this case we reset address/cache and perform a * new look up From 7c66b7a4ae80a9309e6dc1d24b7b6b897e6348eb Mon Sep 17 00:00:00 2001 From: Ralf Lici Date: Fri, 28 Aug 2026 16:50:24 +0200 Subject: [PATCH 438/592] ovpn: track UDP socket route key for peer dst cache ovpn stores the route used to transmit UDP packets in a per-peer dst cache. A cached dst is only valid for the route lookup inputs used when it was resolved. Some of those inputs are mutable while userspace still owns the UDP socket. In particular, changes to the socket mark or UDP source port do not invalidate ovpn's peer dst cache, so ovpn can keep using a route selected with an old socket route key. Replace the cached mark with a route key containing the socket-owned lookup inputs currently used by ovpn, and reset the peer dst cache when the key changes. Before storing a newly looked-up dst, recheck the route key under the peer lock so a dst resolved for stale socket state is not published. Fixes: 08857b5ec5d9 ("ovpn: implement basic TX path (UDP)") Signed-off-by: Ralf Lici Signed-off-by: Antonio Quartulli --- drivers/net/ovpn/peer.c | 1 + drivers/net/ovpn/peer.h | 19 ++++++++- drivers/net/ovpn/udp.c | 90 +++++++++++++++++++++++++++++++++++------ 3 files changed, 95 insertions(+), 15 deletions(-) diff --git a/drivers/net/ovpn/peer.c b/drivers/net/ovpn/peer.c index c95656ca7c35..b400783c2efa 100644 --- a/drivers/net/ovpn/peer.c +++ b/drivers/net/ovpn/peer.c @@ -113,6 +113,7 @@ struct ovpn_peer *ovpn_peer_new(struct ovpn_priv *ovpn, u32 id) RCU_INIT_POINTER(peer->bind, NULL); ovpn_crypto_state_init(&peer->crypto); spin_lock_init(&peer->lock); + seqcount_spinlock_init(&peer->route_key_seq, &peer->lock); kref_init(&peer->refcount); ovpn_peer_stats_init(&peer->vpn_stats); ovpn_peer_stats_init(&peer->link_stats); diff --git a/drivers/net/ovpn/peer.h b/drivers/net/ovpn/peer.h index dfa5c0037e02..063535699ecd 100644 --- a/drivers/net/ovpn/peer.h +++ b/drivers/net/ovpn/peer.h @@ -10,6 +10,7 @@ #ifndef _NET_OVPN_OVPNPEER_H_ #define _NET_OVPN_OVPNPEER_H_ +#include #include #include @@ -17,6 +18,16 @@ #include "socket.h" #include "stats.h" +/** + * struct ovpn_route_key - route key used for the peer dst cache + * @mark: fwmark used for route lookup + * @sport: UDP source port used for route lookup + */ +struct ovpn_route_key { + u32 mark; + __be16 sport; +}; + /** * struct ovpn_peer - the main remote peer object * @ovpn: main openvpn instance this peer belongs to @@ -45,6 +56,8 @@ * @tcp.sk_cb.ops: pointer to the original prot_ops object (TCP only) * @crypto: the crypto configuration (ciphers, keys, etc..) * @dst_cache: cache for dst_entry used to send to peer + * @route_key: route key matching the current dst cache contents + * @route_key_seq: seqcount protecting lockless route_key reads * @bind: remote peer binding * @keepalive_interval: seconds after which a new keepalive should be sent * @keepalive_xmit_exp: future timestamp when next keepalive should be sent @@ -55,7 +68,7 @@ * @vpn_stats: per-peer in-VPN TX/RX stats * @link_stats: per-peer link/transport TX/RX stats * @delete_reason: why peer was deleted (i.e. timeout, transport error, ..) - * @lock: protects binding to peer (bind) and keepalive* fields + * @lock: protects binding to peer (bind), route_key and keepalive* fields * @refcount: reference counter * @rcu: used to free peer in an RCU safe way * @release_entry: entry for the socket release list @@ -99,6 +112,8 @@ struct ovpn_peer { } tcp; struct ovpn_crypto_state crypto; struct dst_cache dst_cache; + struct ovpn_route_key route_key; + seqcount_spinlock_t route_key_seq; struct ovpn_bind __rcu *bind; unsigned long keepalive_interval; unsigned long keepalive_xmit_exp; @@ -109,7 +124,7 @@ struct ovpn_peer { struct ovpn_peer_stats vpn_stats; struct ovpn_peer_stats link_stats; enum ovpn_del_peer_reason delete_reason; - spinlock_t lock; /* protects bind and keepalive* */ + spinlock_t lock; /* protects bind, route_key and keepalive* */ struct kref refcount; struct rcu_head rcu; struct llist_node release_entry; diff --git a/drivers/net/ovpn/udp.c b/drivers/net/ovpn/udp.c index df4750dabd1e..c6d591cb7ff4 100644 --- a/drivers/net/ovpn/udp.c +++ b/drivers/net/ovpn/udp.c @@ -131,6 +131,48 @@ static int ovpn_udp_encap_recv(struct sock *sk, struct sk_buff *skb) return 0; } +static bool ovpn_route_key_equal(const struct ovpn_route_key *a, + const struct ovpn_route_key *b) +{ + return a->mark == b->mark && a->sport == b->sport; +} + +/** + * ovpn_dst_cache_check_key - reset peer dst cache after key changes + * @peer: the peer owning the dst cache + * @cache: the cache that might need to be reset + * @key: the route key for the packet being transmitted + * + * Reset the peer dst cache if it was populated for a different route key. + */ +static void ovpn_dst_cache_check_key(struct ovpn_peer *peer, + struct dst_cache *cache, + const struct ovpn_route_key *key) +{ + struct ovpn_route_key old_key; + unsigned int seq; + + /* snapshot the saved key before deciding whether the cache matches */ + do { + seq = read_seqcount_begin(&peer->route_key_seq); + old_key = peer->route_key; + } while (read_seqcount_retry(&peer->route_key_seq, seq)); + + /* nothing changed: the current cache can be reused */ + if (likely(ovpn_route_key_equal(&old_key, key))) + return; + + /* recheck under lock because another path may have updated the key */ + spin_lock_bh(&peer->lock); + if (!ovpn_route_key_equal(&peer->route_key, key)) { + write_seqcount_begin(&peer->route_key_seq); + peer->route_key = *key; + dst_cache_reset(cache); + write_seqcount_end(&peer->route_key_seq); + } + spin_unlock_bh(&peer->lock); +} + /** * ovpn_udp4_output - send IPv4 packet over udp socket * @peer: the destination peer @@ -138,21 +180,23 @@ static int ovpn_udp_encap_recv(struct sock *sk, struct sk_buff *skb) * @cache: dst cache * @sk: the socket to send the packet over * @skb: the packet to send + * @key: the route key snapshot used for cache validation and flow lookup * * Return: 0 on success or a negative error code otherwise */ static int ovpn_udp4_output(struct ovpn_peer *peer, struct ovpn_bind *bind, struct dst_cache *cache, struct sock *sk, - struct sk_buff *skb) + struct sk_buff *skb, + const struct ovpn_route_key *key) { struct rtable *rt; struct flowi4 fl = { .saddr = bind->local.ipv4.s_addr, .daddr = bind->remote.in4.sin_addr.s_addr, - .fl4_sport = inet_sk(sk)->inet_sport, + .fl4_sport = key->sport, .fl4_dport = bind->remote.in4.sin_port, .flowi4_proto = sk->sk_protocol, - .flowi4_mark = sk->sk_mark, + .flowi4_mark = key->mark, }; int ret; @@ -193,7 +237,12 @@ static int ovpn_udp4_output(struct ovpn_peer *peer, struct ovpn_bind *bind, ret); goto err; } - dst_cache_set_ip4(cache, &rt->dst, fl.saddr); + + /* avoid storing a stale cache */ + spin_lock_bh(&peer->lock); + if (likely(ovpn_route_key_equal(key, &peer->route_key))) + dst_cache_set_ip4(cache, &rt->dst, fl.saddr); + spin_unlock_bh(&peer->lock); transmit: udp_tunnel_xmit_skb(rt, sk, skb, fl.saddr, fl.daddr, 0, @@ -213,12 +262,14 @@ static int ovpn_udp4_output(struct ovpn_peer *peer, struct ovpn_bind *bind, * @cache: dst cache * @sk: the socket to send the packet over * @skb: the packet to send + * @key: the route key snapshot used for cache validation and flow lookup * * Return: 0 on success or a negative error code otherwise */ static int ovpn_udp6_output(struct ovpn_peer *peer, struct ovpn_bind *bind, struct dst_cache *cache, struct sock *sk, - struct sk_buff *skb) + struct sk_buff *skb, + const struct ovpn_route_key *key) { struct dst_entry *dst; int ret; @@ -226,10 +277,10 @@ static int ovpn_udp6_output(struct ovpn_peer *peer, struct ovpn_bind *bind, struct flowi6 fl = { .saddr = bind->local.ipv6, .daddr = bind->remote.in6.sin6_addr, - .fl6_sport = inet_sk(sk)->inet_sport, + .fl6_sport = key->sport, .fl6_dport = bind->remote.in6.sin6_port, .flowi6_proto = sk->sk_protocol, - .flowi6_mark = sk->sk_mark, + .flowi6_mark = key->mark, .flowi6_oif = bind->remote.in6.sin6_scope_id, }; @@ -259,7 +310,12 @@ static int ovpn_udp6_output(struct ovpn_peer *peer, struct ovpn_bind *bind, &bind->remote.in6, ret); goto err; } - dst_cache_set_ip6(cache, dst, &fl.saddr); + + /* avoid storing a stale cache */ + spin_lock_bh(&peer->lock); + if (likely(ovpn_route_key_equal(key, &peer->route_key))) + dst_cache_set_ip6(cache, dst, &fl.saddr); + spin_unlock_bh(&peer->lock); transmit: /* user IPv6 packets may be larger than the transport interface @@ -288,6 +344,7 @@ static int ovpn_udp6_output(struct ovpn_peer *peer, struct ovpn_bind *bind, * @cache: dst cache * @sk: the socket to send the packet over * @skb: the packet to send + * @key: route key snapshot used for cache validation and flow lookup * * rcu_read_lock should be held on entry. * On return, the skb is consumed. @@ -295,7 +352,8 @@ static int ovpn_udp6_output(struct ovpn_peer *peer, struct ovpn_bind *bind, * Return: 0 on success or a negative error code otherwise */ static int ovpn_udp_output(struct ovpn_peer *peer, struct dst_cache *cache, - struct sock *sk, struct sk_buff *skb) + struct sock *sk, struct sk_buff *skb, + struct ovpn_route_key *key) { struct ovpn_bind *bind; int ret; @@ -315,11 +373,11 @@ static int ovpn_udp_output(struct ovpn_peer *peer, struct dst_cache *cache, switch (bind->remote.in4.sin_family) { case AF_INET: - ret = ovpn_udp4_output(peer, bind, cache, sk, skb); + ret = ovpn_udp4_output(peer, bind, cache, sk, skb, key); break; #if IS_ENABLED(CONFIG_IPV6) case AF_INET6: - ret = ovpn_udp6_output(peer, bind, cache, sk, skb); + ret = ovpn_udp6_output(peer, bind, cache, sk, skb, key); break; #endif default: @@ -341,15 +399,21 @@ static int ovpn_udp_output(struct ovpn_peer *peer, struct dst_cache *cache, void ovpn_udp_send_skb(struct ovpn_peer *peer, struct sock *sk, struct sk_buff *skb) { + struct ovpn_route_key key = { + .mark = READ_ONCE(sk->sk_mark), + .sport = READ_ONCE(inet_sk(sk)->inet_sport), + }; int ret; skb->dev = peer->ovpn->dev; - skb->mark = READ_ONCE(sk->sk_mark); + skb->mark = key.mark; /* no checksum performed at this layer */ skb->ip_summed = CHECKSUM_NONE; + ovpn_dst_cache_check_key(peer, &peer->dst_cache, &key); + /* crypto layer -> transport (UDP) */ - ret = ovpn_udp_output(peer, &peer->dst_cache, sk, skb); + ret = ovpn_udp_output(peer, &peer->dst_cache, sk, skb, &key); if (unlikely(ret < 0)) kfree_skb(skb); } From fa603710bdb9aea33c0d9cc2c05ed24d84f58753 Mon Sep 17 00:00:00 2001 From: Ralf Lici Date: Fri, 28 Aug 2026 16:50:25 +0200 Subject: [PATCH 439/592] ovpn: validate peer state before caching UDP dst UDP route lookup runs without peer->lock while the bind is protected by RCU. The route key is snapshotted separately. Either can change while the lookup is in progress. The TX path currently checks only the route key before publishing the looked-up dst. If the bind changes but the route key does not, a dst resolved from the old endpoint can be installed in the cache after the bind replacement. Compare both the bind pointer and the route key under peer->lock before updating the cache. The RCU read-side critical section keeps the old bind alive throughout the lookup, so pointer identity is sufficient to detect a replacement. Fixes: f0281c1d3732 ("ovpn: add support for updating local or remote UDP endpoint") Signed-off-by: Ralf Lici Signed-off-by: Antonio Quartulli --- drivers/net/ovpn/udp.c | 33 +++++++++++++++++++++++++++++++-- 1 file changed, 31 insertions(+), 2 deletions(-) diff --git a/drivers/net/ovpn/udp.c b/drivers/net/ovpn/udp.c index c6d591cb7ff4..eeef4a7229f5 100644 --- a/drivers/net/ovpn/udp.c +++ b/drivers/net/ovpn/udp.c @@ -173,6 +173,35 @@ static void ovpn_dst_cache_check_key(struct ovpn_peer *peer, spin_unlock_bh(&peer->lock); } +/** + * ovpn_dst_cache_current - check whether a route lookup matches peer state + * @peer: the peer owning the bind and dst cache + * @bind: the RCU bind used for the route lookup + * @key: the route key used for the route lookup + * + * Check that @bind is still the current peer bind and that @key still matches + * the peer route key. The caller must hold @peer->lock. The TX path keeps + * @bind inside an RCU read-side critical section, so pointer identity is enough + * to detect whether the bind was replaced while the route lookup was running. + * + * Return: true if the lookup result still matches the current peer state and + * may update the dst cache. + */ +static bool ovpn_dst_cache_current(const struct ovpn_peer *peer, + const struct ovpn_bind *bind, + const struct ovpn_route_key *key) +{ + const struct ovpn_bind *curr_bind; + + lockdep_assert_held(&peer->lock); + + curr_bind = rcu_dereference_protected(peer->bind, + lockdep_is_held(&peer->lock)); + + return curr_bind == bind && + ovpn_route_key_equal(key, &peer->route_key); +} + /** * ovpn_udp4_output - send IPv4 packet over udp socket * @peer: the destination peer @@ -240,7 +269,7 @@ static int ovpn_udp4_output(struct ovpn_peer *peer, struct ovpn_bind *bind, /* avoid storing a stale cache */ spin_lock_bh(&peer->lock); - if (likely(ovpn_route_key_equal(key, &peer->route_key))) + if (likely(ovpn_dst_cache_current(peer, bind, key))) dst_cache_set_ip4(cache, &rt->dst, fl.saddr); spin_unlock_bh(&peer->lock); @@ -313,7 +342,7 @@ static int ovpn_udp6_output(struct ovpn_peer *peer, struct ovpn_bind *bind, /* avoid storing a stale cache */ spin_lock_bh(&peer->lock); - if (likely(ovpn_route_key_equal(key, &peer->route_key))) + if (likely(ovpn_dst_cache_current(peer, bind, key))) dst_cache_set_ip6(cache, dst, &fl.saddr); spin_unlock_bh(&peer->lock); From aea934a221ec6a867221e5b765f65f1857befd53 Mon Sep 17 00:00:00 2001 From: Ralf Lici Date: Fri, 28 Aug 2026 16:50:26 +0200 Subject: [PATCH 440/592] ovpn: replace bind when learning local endpoint struct ovpn_bind is published through peer->bind with RCU, but local endpoint learning updates bind->local in place under peer->lock. UDP TX reads the field without that lock. In particular, a concurrent IPv6 update can therefore result in a torn address read. Use ovpn_peer_reset_sockaddr to publish a replacement bind when learning a new local endpoint, just as a remote endpoint change does. Preserve the current remote address and reset the dst cache only after the new bind has been published successfully. Track remote endpoint changes separately so that float notification and transport-address rehashing remain limited to actual peer floats. Fixes: f0281c1d3732 ("ovpn: add support for updating local or remote UDP endpoint") Signed-off-by: Ralf Lici Signed-off-by: Antonio Quartulli --- drivers/net/ovpn/peer.c | 43 ++++++++++++++++++++++------------------- 1 file changed, 23 insertions(+), 20 deletions(-) diff --git a/drivers/net/ovpn/peer.c b/drivers/net/ovpn/peer.c index b400783c2efa..430c6cd48db8 100644 --- a/drivers/net/ovpn/peer.c +++ b/drivers/net/ovpn/peer.c @@ -200,13 +200,12 @@ static void __ovpn_peer_hash_transp_addr(struct ovpn_peer *peer, */ void ovpn_peer_endpoints_update(struct ovpn_peer *peer, struct sk_buff *skb) { + const void *local_ip = NULL; struct sockaddr_storage ss; struct sockaddr_in6 *sa6; - bool reset_cache = false; struct sockaddr_in *sa; struct ovpn_bind *bind; - const void *local_ip; - size_t salen = 0; + bool floated = false; spin_lock_bh(&peer->lock); bind = rcu_dereference_protected(peer->bind, @@ -233,8 +232,7 @@ void ovpn_peer_endpoints_update(struct ovpn_peer *peer, struct sk_buff *skb) .sin_addr.s_addr = ip_hdr(skb)->saddr, .sin_port = udp_hdr(skb)->source, }; - salen = sizeof(*sa); - reset_cache = true; + floated = true; break; } @@ -246,10 +244,12 @@ void ovpn_peer_endpoints_update(struct ovpn_peer *peer, struct sk_buff *skb) netdev_name(peer->ovpn->dev), peer->id, &bind->local.ipv4.s_addr, &ip_hdr(skb)->daddr); - bind->local.ipv4.s_addr = ip_hdr(skb)->daddr; - reset_cache = true; + local_ip = &ip_hdr(skb)->daddr; + memcpy(&ss, &bind->remote, sizeof(struct sockaddr_in)); + break; } - break; + /* nothing changed */ + goto unlock; case htons(ETH_P_IPV6): /* float check */ if (unlikely(!ovpn_bind_skb_src_match(bind, skb))) { @@ -271,8 +271,7 @@ void ovpn_peer_endpoints_update(struct ovpn_peer *peer, struct sk_buff *skb) ipv6_iface_scope_id(&ipv6_hdr(skb)->saddr, skb->skb_iif), }; - salen = sizeof(*sa6); - reset_cache = true; + floated = true; break; } @@ -285,26 +284,30 @@ void ovpn_peer_endpoints_update(struct ovpn_peer *peer, struct sk_buff *skb) netdev_name(peer->ovpn->dev), peer->id, &bind->local.ipv6, &ipv6_hdr(skb)->daddr); - bind->local.ipv6 = ipv6_hdr(skb)->daddr; - reset_cache = true; + local_ip = &ipv6_hdr(skb)->daddr; + memcpy(&ss, &bind->remote, sizeof(struct sockaddr_in6)); + break; } - break; + /* nothing changed */ + goto unlock; default: goto unlock; } - if (unlikely(reset_cache)) - dst_cache_reset(&peer->dst_cache); - - /* if the peer did not float, we can bail out now */ - if (likely(!salen)) - goto unlock; - if (unlikely(ovpn_peer_reset_sockaddr(peer, (struct sockaddr_storage *)&ss, local_ip) < 0)) goto unlock; + /* reset the cache only after a successful bind update to avoid useless + * cache misses on concurrent TX + */ + dst_cache_reset(&peer->dst_cache); + + /* if only the local address changed, bail out now */ + if (!floated) + goto unlock; + net_dbg_ratelimited("%s: peer %d floated to %pIScp", netdev_name(peer->ovpn->dev), peer->id, &ss); From 7d8104988f423572df1f3347ce578037b1043f34 Mon Sep 17 00:00:00 2001 From: Ralf Lici Date: Fri, 28 Aug 2026 16:50:27 +0200 Subject: [PATCH 441/592] ovpn: replace bind when clearing stale local source The UDP output fallback clears bind->local in place when the remembered source address is no longer usable. The bind is RCU-published and read locklessly by concurrent TX, so an IPv6 reader can observe a torn address. Retry the route lookup with source address autoselection without modifying the bind. After a successful lookup, revalidate the bind and route key under peer->lock, reset the dst cache, and best-effort publish a replacement bind with a wildcard local address. Do not cache the resolved dst when clearing the local source. Replacing the source invalidates all per-CPU cache entries, while dst_cache_set_ip4 and dst_cache_set_ip6 update only the current CPU slot. The current packet can still use the resolved route; if bind allocation fails, a later cache miss retries the repair. Fixes: 08857b5ec5d9 ("ovpn: implement basic TX path (UDP)") Signed-off-by: Ralf Lici Signed-off-by: Antonio Quartulli --- drivers/net/ovpn/udp.c | 81 +++++++++++++++++++++++++++++------------- 1 file changed, 56 insertions(+), 25 deletions(-) diff --git a/drivers/net/ovpn/udp.c b/drivers/net/ovpn/udp.c index eeef4a7229f5..055cdb1bee13 100644 --- a/drivers/net/ovpn/udp.c +++ b/drivers/net/ovpn/udp.c @@ -185,7 +185,7 @@ static void ovpn_dst_cache_check_key(struct ovpn_peer *peer, * to detect whether the bind was replaced while the route lookup was running. * * Return: true if the lookup result still matches the current peer state and - * may update the dst cache. + * may update the dst cache or replace the bind. */ static bool ovpn_dst_cache_current(const struct ovpn_peer *peer, const struct ovpn_bind *bind, @@ -218,6 +218,9 @@ static int ovpn_udp4_output(struct ovpn_peer *peer, struct ovpn_bind *bind, struct sk_buff *skb, const struct ovpn_route_key *key) { + struct sockaddr_storage remote; + struct in_addr local = {}; + bool reset_local = false; struct rtable *rt; struct flowi4 fl = { .saddr = bind->local.ipv4.s_addr, @@ -236,24 +239,17 @@ static int ovpn_udp4_output(struct ovpn_peer *peer, struct ovpn_bind *bind, if (fl.saddr && unlikely(!inet_confirm_addr(sock_net(sk), NULL, 0, fl.saddr, RT_SCOPE_HOST))) { - /* we may end up here when the cached address is not usable - * anymore. In this case we reset address/cache and perform a - * new look up + /* The learned local address is not usable anymore. + * Retry with source address autoselection. */ fl.saddr = 0; - spin_lock_bh(&peer->lock); - bind->local.ipv4.s_addr = 0; - spin_unlock_bh(&peer->lock); - dst_cache_reset(cache); + reset_local = true; } rt = ip_route_output_flow(sock_net(sk), &fl, sk); if (IS_ERR(rt) && PTR_ERR(rt) == -EINVAL) { fl.saddr = 0; - spin_lock_bh(&peer->lock); - bind->local.ipv4.s_addr = 0; - spin_unlock_bh(&peer->lock); - dst_cache_reset(cache); + reset_local = true; rt = ip_route_output_flow(sock_net(sk), &fl, sk); } @@ -267,10 +263,28 @@ static int ovpn_udp4_output(struct ovpn_peer *peer, struct ovpn_bind *bind, goto err; } - /* avoid storing a stale cache */ + /* avoid storing a stale cache or local address */ spin_lock_bh(&peer->lock); - if (likely(ovpn_dst_cache_current(peer, bind, key))) - dst_cache_set_ip4(cache, &rt->dst, fl.saddr); + if (likely(ovpn_dst_cache_current(peer, bind, key))) { + if (!reset_local) { + dst_cache_set_ip4(cache, &rt->dst, fl.saddr); + spin_unlock_bh(&peer->lock); + goto transmit; + } + + /* invalidate per-CPU dst entries that may still carry + * the stale source + */ + dst_cache_reset(cache); + + /* preserve the current remote */ + memcpy(&remote, &bind->remote, sizeof(struct sockaddr_in)); + /* The current packet already has a valid wildcard-source route. + * If replacing the bind fails, leave the stale local in place; + * a later cache miss will retry the repair. + */ + ovpn_peer_reset_sockaddr(peer, &remote, &local); + } spin_unlock_bh(&peer->lock); transmit: @@ -300,6 +314,9 @@ static int ovpn_udp6_output(struct ovpn_peer *peer, struct ovpn_bind *bind, struct sk_buff *skb, const struct ovpn_route_key *key) { + struct in6_addr local = in6addr_any; + struct sockaddr_storage remote; + bool reset_local = false; struct dst_entry *dst; int ret; @@ -320,15 +337,11 @@ static int ovpn_udp6_output(struct ovpn_peer *peer, struct ovpn_bind *bind, if (!ipv6_addr_any(&fl.saddr) && unlikely(!ipv6_chk_addr(sock_net(sk), &fl.saddr, NULL, 0))) { - /* we may end up here when the cached address is not usable - * anymore. In this case we reset address/cache and perform a - * new look up + /* The learned local address is not usable anymore. + * Retry with source address autoselection. */ fl.saddr = in6addr_any; - spin_lock_bh(&peer->lock); - bind->local.ipv6 = in6addr_any; - spin_unlock_bh(&peer->lock); - dst_cache_reset(cache); + reset_local = true; } dst = ip6_dst_lookup_flow(sock_net(sk), sk, &fl, NULL); @@ -340,10 +353,28 @@ static int ovpn_udp6_output(struct ovpn_peer *peer, struct ovpn_bind *bind, goto err; } - /* avoid storing a stale cache */ + /* avoid storing a stale cache or local address */ spin_lock_bh(&peer->lock); - if (likely(ovpn_dst_cache_current(peer, bind, key))) - dst_cache_set_ip6(cache, dst, &fl.saddr); + if (likely(ovpn_dst_cache_current(peer, bind, key))) { + if (!reset_local) { + dst_cache_set_ip6(cache, dst, &fl.saddr); + spin_unlock_bh(&peer->lock); + goto transmit; + } + + /* invalidate per-CPU dst entries that may still carry + * the stale source + */ + dst_cache_reset(cache); + + /* preserve the current remote */ + memcpy(&remote, &bind->remote, sizeof(struct sockaddr_in6)); + /* The current packet already has a valid wildcard-source route. + * If replacing the bind fails, leave the stale local in place; + * a later cache miss will retry the repair. + */ + ovpn_peer_reset_sockaddr(peer, &remote, &local); + } spin_unlock_bh(&peer->lock); transmit: From b43beccb3713fafada57814b0a652f4a876eb75f Mon Sep 17 00:00:00 2001 From: Ralf Lici Date: Fri, 28 Aug 2026 15:00:06 +0200 Subject: [PATCH 442/592] ovpn: always unhash old VPN addresses before rehashing ovpn_peer_hash_vpn_ip updates the per-peer VPN address hash entries after userspace changes a peer VPN address. The current code removes an old hash entry only when the new address for that family is not the unspecified address. When an address is cleared to 0.0.0.0 or ::, its hash node therefore remains linked in the bucket selected by the old address. The address comparison performed during lookup prevents the old address from matching, but the table retains a stale entry until the peer is removed or another address is configured for that family. Always remove both old VPN address hash entries before conditionally adding the currently configured addresses back. This ensures that a cleared address leaves its hash node unhashed. Fixes: 1d36a36f6d53 ("ovpn: implement peer add/get/dump/delete via netlink") Signed-off-by: Ralf Lici Signed-off-by: Antonio Quartulli --- drivers/net/ovpn/peer.c | 10 ++++------ 1 file changed, 4 insertions(+), 6 deletions(-) diff --git a/drivers/net/ovpn/peer.c b/drivers/net/ovpn/peer.c index 430c6cd48db8..bbd9e17fb0bf 100644 --- a/drivers/net/ovpn/peer.c +++ b/drivers/net/ovpn/peer.c @@ -994,10 +994,11 @@ void ovpn_peer_hash_vpn_ip(struct ovpn_peer *peer) if (hlist_unhashed(&peer->hash_entry_id)) return; - if (peer->vpn_addrs.ipv4.s_addr != htonl(INADDR_ANY)) { - /* remove potential old hashing */ - hlist_nulls_del_init_rcu(&peer->hash_entry_addr4); + /* remove potential old hashing */ + hlist_nulls_del_init_rcu(&peer->hash_entry_addr4); + hlist_nulls_del_init_rcu(&peer->hash_entry_addr6); + if (peer->vpn_addrs.ipv4.s_addr != htonl(INADDR_ANY)) { nhead = ovpn_get_hash_head(peer->ovpn->peers->by_vpn_addr4, &peer->vpn_addrs.ipv4, sizeof(peer->vpn_addrs.ipv4)); @@ -1005,9 +1006,6 @@ void ovpn_peer_hash_vpn_ip(struct ovpn_peer *peer) } if (!ipv6_addr_any(&peer->vpn_addrs.ipv6)) { - /* remove potential old hashing */ - hlist_nulls_del_init_rcu(&peer->hash_entry_addr6); - nhead = ovpn_get_hash_head(peer->ovpn->peers->by_vpn_addr6, &peer->vpn_addrs.ipv6, sizeof(peer->vpn_addrs.ipv6)); From d25e885b31a0f2808d936f95c9a558a8a792669b Mon Sep 17 00:00:00 2001 From: Ralf Lici Date: Fri, 28 Aug 2026 15:00:07 +0200 Subject: [PATCH 443/592] ovpn: reject duplicate peer VPN addresses In MP mode, ovpn uses the peer VPN addresses as lookup keys for selecting the peer that should receive an outgoing tunnel packet. However, the netlink peer configuration path does not currently reject duplicate VPN addresses. If two peers are configured with the same VPN address, both can be inserted in the VPN address hash table and lookups return whichever peer is found first. This makes peer selection ambiguous and dependent on hash insertion order. Reject peer creation or update when the resulting VPN address is already assigned to another peer. Ignore unspecified addresses because those are not inserted in the VPN address hash tables. This changes such configurations from being accepted to being rejected, but they have never worked reliably because peer selection is ambiguous. Fixes: 1d36a36f6d53 ("ovpn: implement peer add/get/dump/delete via netlink") Signed-off-by: Ralf Lici Signed-off-by: Antonio Quartulli --- drivers/net/ovpn/netlink.c | 37 ++++++++++++++++----- drivers/net/ovpn/peer.c | 67 +++++++++++++++++++++++++++++++++++++- drivers/net/ovpn/peer.h | 6 ++++ 3 files changed, 101 insertions(+), 9 deletions(-) diff --git a/drivers/net/ovpn/netlink.c b/drivers/net/ovpn/netlink.c index 2ba762082acc..e23f7d1f49e0 100644 --- a/drivers/net/ovpn/netlink.c +++ b/drivers/net/ovpn/netlink.c @@ -480,8 +480,10 @@ int ovpn_nl_peer_new_doit(struct sk_buff *skb, struct genl_info *info) int ovpn_nl_peer_set_doit(struct sk_buff *skb, struct genl_info *info) { - struct nlattr *attrs[OVPN_A_PEER_MAX + 1]; struct ovpn_priv *ovpn = info->user_ptr[0]; + struct nlattr *attrs[OVPN_A_PEER_MAX + 1]; + struct in6_addr vpn_addr6; + struct in_addr vpn_addr4; struct ovpn_socket *sock; struct ovpn_peer *peer; u32 peer_id; @@ -528,28 +530,47 @@ int ovpn_nl_peer_set_doit(struct sk_buff *skb, struct genl_info *info) rcu_read_unlock(); spin_lock_bh(&ovpn->lock); - ret = ovpn_nl_peer_modify(peer, info, attrs); - if (ret < 0) { - spin_unlock_bh(&ovpn->lock); - ovpn_peer_put(peer); - return ret; + + /* reject peer with conflicting VPN address */ + if (attrs[OVPN_A_PEER_VPN_IPV4]) { + vpn_addr4.s_addr = nla_get_in_addr(attrs[OVPN_A_PEER_VPN_IPV4]); + if (ovpn_peer_vpn_addr_conflict4(ovpn, peer, &vpn_addr4)) + goto addr_conflict; } + if (attrs[OVPN_A_PEER_VPN_IPV6]) { + vpn_addr6 = nla_get_in6_addr(attrs[OVPN_A_PEER_VPN_IPV6]); + if (ovpn_peer_vpn_addr_conflict6(ovpn, peer, &vpn_addr6)) + goto addr_conflict; + } + + ret = ovpn_nl_peer_modify(peer, info, attrs); + if (ret < 0) + goto unlock; /* ret == 1 means that VPN IPv4/6 has been modified and rehashing * is required */ - if (ret > 0) + if (ret > 0) { ovpn_peer_hash_vpn_ip(peer); + ret = 0; + } /* if the remote endpoint was updated, the by_transp_addr hash bucket * also needs to be refreshed, otherwise incoming packets from the new * remote address would fail the lockless lookup */ if (attrs[OVPN_A_PEER_REMOTE_IPV4] || attrs[OVPN_A_PEER_REMOTE_IPV6]) ovpn_peer_hash_transp_addr(peer); + +unlock: spin_unlock_bh(&ovpn->lock); ovpn_peer_put(peer); - return 0; + return ret; +addr_conflict: + NL_SET_ERR_MSG_FMT_MOD(info->extack, + "VPN IP is already assigned to another peer"); + ret = -EADDRINUSE; + goto unlock; } static int ovpn_nl_send_peer(struct sk_buff *skb, const struct genl_info *info, diff --git a/drivers/net/ovpn/peer.c b/drivers/net/ovpn/peer.c index bbd9e17fb0bf..2067825bb5b6 100644 --- a/drivers/net/ovpn/peer.c +++ b/drivers/net/ovpn/peer.c @@ -488,7 +488,7 @@ static struct ovpn_peer *ovpn_peer_get_by_vpn_addr4(struct ovpn_priv *ovpn, * Return: the peer if found or NULL otherwise */ static struct ovpn_peer *ovpn_peer_get_by_vpn_addr6(struct ovpn_priv *ovpn, - struct in6_addr *addr) + const struct in6_addr *addr) { struct hlist_nulls_head *nhead; struct hlist_nulls_node *ntmp; @@ -513,6 +513,64 @@ static struct ovpn_peer *ovpn_peer_get_by_vpn_addr6(struct ovpn_priv *ovpn, return NULL; } +/** + * ovpn_peer_vpn_addr_conflict4 - check if the VPN v4 address is already in use + * @ovpn: the openvpn instance to search + * @peer: peer being added or updated, or NULL + * @addr: VPN IPv4 address to check + * + * Check whether @addr is already assigned to another peer. @peer is ignored + * when found, allowing peer updates that keep an existing address. + * Unspecified addresses are ignored. + * + * Note: the caller must hold @ovpn->lock. + * + * Return: true on conflict, false otherwise. + */ +bool ovpn_peer_vpn_addr_conflict4(struct ovpn_priv *ovpn, + const struct ovpn_peer *peer, + const struct in_addr *addr) +{ + struct ovpn_peer *tmp = NULL; + + lockdep_assert_held(&ovpn->lock); + + /* we don't hash INADDR_ANY, no conflict in that case */ + if (addr->s_addr != htonl(INADDR_ANY)) + tmp = ovpn_peer_get_by_vpn_addr4(ovpn, addr->s_addr); + + return tmp && tmp != peer; +} + +/** + * ovpn_peer_vpn_addr_conflict6 - check if the VPN v6 address is already in use + * @ovpn: the openvpn instance to search + * @peer: peer being added or updated, or NULL + * @addr: VPN IPv6 address to check + * + * Check whether @addr is already assigned to another peer. @peer is ignored + * when found, allowing peer updates that keep an existing address. + * Unspecified addresses are ignored. + * + * Note: the caller must hold @ovpn->lock. + * + * Return: true on conflict, false otherwise. + */ +bool ovpn_peer_vpn_addr_conflict6(struct ovpn_priv *ovpn, + const struct ovpn_peer *peer, + const struct in6_addr *addr) +{ + struct ovpn_peer *tmp = NULL; + + lockdep_assert_held(&ovpn->lock); + + /* we don't hash ::, no conflict in that case */ + if (!ipv6_addr_any(addr)) + tmp = ovpn_peer_get_by_vpn_addr6(ovpn, addr); + + return tmp && tmp != peer; +} + /** * ovpn_peer_transp_match - check if sockaddr and peer binding match * @peer: the peer to get the binding from @@ -1040,6 +1098,13 @@ static int ovpn_peer_add_mp(struct ovpn_priv *ovpn, struct ovpn_peer *peer) goto out; } + /* reject peer with conflicting VPN address */ + if (ovpn_peer_vpn_addr_conflict4(ovpn, NULL, &peer->vpn_addrs.ipv4) || + ovpn_peer_vpn_addr_conflict6(ovpn, NULL, &peer->vpn_addrs.ipv6)) { + ret = -EADDRINUSE; + goto out; + } + bind = rcu_dereference_protected(peer->bind, true); /* peers connected via TCP have bind == NULL */ if (bind) { diff --git a/drivers/net/ovpn/peer.h b/drivers/net/ovpn/peer.h index 063535699ecd..1879bfb76992 100644 --- a/drivers/net/ovpn/peer.h +++ b/drivers/net/ovpn/peer.h @@ -164,6 +164,12 @@ struct ovpn_peer *ovpn_peer_get_by_transp_addr(struct ovpn_priv *ovpn, struct ovpn_peer *ovpn_peer_get_by_id(struct ovpn_priv *ovpn, u32 peer_id); struct ovpn_peer *ovpn_peer_get_by_dst(struct ovpn_priv *ovpn, struct sk_buff *skb); +bool ovpn_peer_vpn_addr_conflict4(struct ovpn_priv *ovpn, + const struct ovpn_peer *peer, + const struct in_addr *addr); +bool ovpn_peer_vpn_addr_conflict6(struct ovpn_priv *ovpn, + const struct ovpn_peer *peer, + const struct in6_addr *addr); void ovpn_peer_hash_vpn_ip(struct ovpn_peer *peer); void ovpn_peer_hash_transp_addr(struct ovpn_peer *peer); bool ovpn_peer_check_by_src(struct ovpn_priv *ovpn, struct sk_buff *skb, From 025af3a0a892514f9f27f186338ba3d44365547a Mon Sep 17 00:00:00 2001 From: Ralf Lici Date: Fri, 28 Aug 2026 15:00:08 +0200 Subject: [PATCH 444/592] ovpn: reject multipeer peers without VPN addresses In MP mode, ovpn uses the peer VPN addresses to select the peer for outgoing tunnel packets. Peer creation currently requires a VPN IPv4 or IPv6 attribute, but it only checks for the presence of the attribute and not for a usable address value. This allows userspace to create an MP peer with only unspecified VPN addresses, or to update an existing peer so that both VPN address families become unspecified. Such a peer cannot be selected through the VPN address hash tables. Reject MP peer creation or update when the resulting peer would not have at least one VPN address configured. This changes such configurations from being accepted to being rejected, but they have never been usable because the peer cannot be selected through the VPN address hash tables. Fixes: 1d36a36f6d53 ("ovpn: implement peer add/get/dump/delete via netlink") Signed-off-by: Ralf Lici Signed-off-by: Antonio Quartulli --- drivers/net/ovpn/netlink.c | 36 ++++++++++++++++++++++++++++++------ 1 file changed, 30 insertions(+), 6 deletions(-) diff --git a/drivers/net/ovpn/netlink.c b/drivers/net/ovpn/netlink.c index e23f7d1f49e0..e9e0f75e0443 100644 --- a/drivers/net/ovpn/netlink.c +++ b/drivers/net/ovpn/netlink.c @@ -352,8 +352,10 @@ static int ovpn_nl_peer_modify(struct ovpn_peer *peer, struct genl_info *info, int ovpn_nl_peer_new_doit(struct sk_buff *skb, struct genl_info *info) { - struct nlattr *attrs[OVPN_A_PEER_MAX + 1]; + struct in_addr vpn_addr4 = { .s_addr = htonl(INADDR_ANY) }; + struct in6_addr vpn_addr6 = IN6ADDR_ANY_INIT; struct ovpn_priv *ovpn = info->user_ptr[0]; + struct nlattr *attrs[OVPN_A_PEER_MAX + 1]; struct ovpn_socket *ovpn_sock; struct socket *sock = NULL; struct ovpn_peer *peer; @@ -377,11 +379,20 @@ int ovpn_nl_peer_new_doit(struct sk_buff *skb, struct genl_info *info) return -EINVAL; /* in MP mode VPN IPs are required for selecting the right peer */ - if (ovpn->mode == OVPN_MODE_MP && !attrs[OVPN_A_PEER_VPN_IPV4] && - !attrs[OVPN_A_PEER_VPN_IPV6]) { - NL_SET_ERR_MSG_FMT_MOD(info->extack, - "VPN IP must be provided in MP mode"); - return -EINVAL; + if (ovpn->mode == OVPN_MODE_MP) { + if (attrs[OVPN_A_PEER_VPN_IPV4]) + vpn_addr4.s_addr = + nla_get_in_addr(attrs[OVPN_A_PEER_VPN_IPV4]); + if (attrs[OVPN_A_PEER_VPN_IPV6]) + vpn_addr6 = + nla_get_in6_addr(attrs[OVPN_A_PEER_VPN_IPV6]); + + if (vpn_addr4.s_addr == htonl(INADDR_ANY) && + ipv6_addr_any(&vpn_addr6)) { + NL_SET_ERR_MSG_FMT_MOD(info->extack, + "at least one VPN IP must be configured in MP mode"); + return -EINVAL; + } } peer_id = nla_get_u32(attrs[OVPN_A_PEER_ID]); @@ -531,6 +542,9 @@ int ovpn_nl_peer_set_doit(struct sk_buff *skb, struct genl_info *info) spin_lock_bh(&ovpn->lock); + vpn_addr4 = peer->vpn_addrs.ipv4; + vpn_addr6 = peer->vpn_addrs.ipv6; + /* reject peer with conflicting VPN address */ if (attrs[OVPN_A_PEER_VPN_IPV4]) { vpn_addr4.s_addr = nla_get_in_addr(attrs[OVPN_A_PEER_VPN_IPV4]); @@ -543,6 +557,16 @@ int ovpn_nl_peer_set_doit(struct sk_buff *skb, struct genl_info *info) goto addr_conflict; } + /* in MP mode VPN IPs are required for selecting the right peer */ + if (ovpn->mode == OVPN_MODE_MP && + vpn_addr4.s_addr == htonl(INADDR_ANY) && + ipv6_addr_any(&vpn_addr6)) { + NL_SET_ERR_MSG_FMT_MOD(info->extack, + "at least one VPN IP must be configured in MP mode"); + ret = -EINVAL; + goto unlock; + } + ret = ovpn_nl_peer_modify(peer, info, attrs); if (ret < 0) goto unlock; From 5940f3407b78062442cb01f541ef6eed709fc380 Mon Sep 17 00:00:00 2001 From: Ralf Lici Date: Fri, 28 Aug 2026 15:00:09 +0200 Subject: [PATCH 445/592] ovpn: reject invalid peer VPN addresses In MP mode, ovpn uses peer VPN addresses as lookup keys for selecting the peer that should receive outgoing tunnel packets. The netlink configuration path currently accepts address values that cannot sensibly identify a VPN peer, such as multicast, broadcast or loopback addresses. Reject invalid peer VPN addresses when creating or updating an MP peer. Keep accepting the unspecified address as the internal unset value, provided that at least one VPN address family remains configured. Fixes: 1d36a36f6d53 ("ovpn: implement peer add/get/dump/delete via netlink") Signed-off-by: Ralf Lici Signed-off-by: Antonio Quartulli --- drivers/net/ovpn/netlink.c | 55 +++++++++++++++++++++++++++++--------- 1 file changed, 42 insertions(+), 13 deletions(-) diff --git a/drivers/net/ovpn/netlink.c b/drivers/net/ovpn/netlink.c index e9e0f75e0443..5432bc2eb8e8 100644 --- a/drivers/net/ovpn/netlink.c +++ b/drivers/net/ovpn/netlink.c @@ -185,6 +185,39 @@ static sa_family_t ovpn_nl_family_get(struct nlattr *addr4, return AF_UNSPEC; } +static int ovpn_nl_peer_check_vpn_addrs(const struct in_addr *addr4, + const struct in6_addr *addr6, + struct genl_info *info) +{ + int addr6_type; + + if (addr4->s_addr == htonl(INADDR_ANY) && ipv6_addr_any(addr6)) { + NL_SET_ERR_MSG_MOD(info->extack, + "at least one VPN IP must be configured in MP mode"); + return -EINVAL; + } + + if (ipv4_is_multicast(addr4->s_addr) || ipv4_is_lbcast(addr4->s_addr) || + ipv4_is_loopback(addr4->s_addr)) { + NL_SET_ERR_MSG_MOD(info->extack, + "VPN IPv4 address must be valid unicast or any"); + return -EADDRNOTAVAIL; + } + + if (!ipv6_addr_any(addr6)) { + addr6_type = ipv6_addr_type(addr6); + + if (!(addr6_type & IPV6_ADDR_UNICAST) || + (addr6_type & (IPV6_ADDR_LOOPBACK | IPV6_ADDR_COMPATv4))) { + NL_SET_ERR_MSG_MOD(info->extack, + "VPN IPv6 address must be valid unicast or any"); + return -EADDRNOTAVAIL; + } + } + + return 0; +} + static int ovpn_nl_peer_precheck(struct ovpn_priv *ovpn, struct genl_info *info, struct nlattr **attrs) @@ -387,12 +420,10 @@ int ovpn_nl_peer_new_doit(struct sk_buff *skb, struct genl_info *info) vpn_addr6 = nla_get_in6_addr(attrs[OVPN_A_PEER_VPN_IPV6]); - if (vpn_addr4.s_addr == htonl(INADDR_ANY) && - ipv6_addr_any(&vpn_addr6)) { - NL_SET_ERR_MSG_FMT_MOD(info->extack, - "at least one VPN IP must be configured in MP mode"); - return -EINVAL; - } + ret = ovpn_nl_peer_check_vpn_addrs(&vpn_addr4, &vpn_addr6, + info); + if (ret < 0) + return ret; } peer_id = nla_get_u32(attrs[OVPN_A_PEER_ID]); @@ -558,13 +589,11 @@ int ovpn_nl_peer_set_doit(struct sk_buff *skb, struct genl_info *info) } /* in MP mode VPN IPs are required for selecting the right peer */ - if (ovpn->mode == OVPN_MODE_MP && - vpn_addr4.s_addr == htonl(INADDR_ANY) && - ipv6_addr_any(&vpn_addr6)) { - NL_SET_ERR_MSG_FMT_MOD(info->extack, - "at least one VPN IP must be configured in MP mode"); - ret = -EINVAL; - goto unlock; + if (ovpn->mode == OVPN_MODE_MP) { + ret = ovpn_nl_peer_check_vpn_addrs(&vpn_addr4, &vpn_addr6, + info); + if (ret < 0) + goto unlock; } ret = ovpn_nl_peer_modify(peer, info, attrs); From 006208026819d5e9e5ec07b3e73d95960a059327 Mon Sep 17 00:00:00 2001 From: Ralf Lici Date: Fri, 28 Aug 2026 15:00:10 +0200 Subject: [PATCH 446/592] selftests: ovpn: validate peer VPN addresses Exercise peer VPN address validation through both peer creation and update. Check missing, unspecified, duplicate, multicast, broadcast, loopback, IPv4-compatible and IPv4-mapped addresses. Temporarily configure a peer with both address families to verify that either family can be cleared while the other remains configured, then restore the original addresses before running the existing traffic tests. Extend ovpn-cli peer updates with an optional VPN address and preserve peer creation errors so the negative tests can observe rejected netlink requests. Signed-off-by: Ralf Lici Signed-off-by: Antonio Quartulli --- tools/testing/selftests/net/ovpn/common.sh | 13 ++++ tools/testing/selftests/net/ovpn/ovpn-cli.c | 54 ++++++++++----- tools/testing/selftests/net/ovpn/test.sh | 75 ++++++++++++++++++++- 3 files changed, 123 insertions(+), 19 deletions(-) diff --git a/tools/testing/selftests/net/ovpn/common.sh b/tools/testing/selftests/net/ovpn/common.sh index 2d844eb3aa6e..5e9c81e885e6 100644 --- a/tools/testing/selftests/net/ovpn/common.sh +++ b/tools/testing/selftests/net/ovpn/common.sh @@ -136,6 +136,19 @@ ovpn_create_ns() { ip netns add "ovpn_peer${1}" } +ovpn_peer_vpn_addr() { + local peer="$1" + local file + + if [ "${OVPN_PROTO}" == "UDP" ]; then + file="${OVPN_UDP_PEERS_FILE}" + else + file="${OVPN_TCP_PEERS_FILE}" + fi + + awk -v peer="${peer}" '$1 == peer {print $NF; exit}' "${file}" +} + ovpn_setup_ns() { local peer="ovpn_peer${1}" local server_ns="ovpn_peer0" diff --git a/tools/testing/selftests/net/ovpn/ovpn-cli.c b/tools/testing/selftests/net/ovpn/ovpn-cli.c index f4effa7580c0..3b612a8a18fe 100644 --- a/tools/testing/selftests/net/ovpn/ovpn-cli.c +++ b/tools/testing/selftests/net/ovpn/ovpn-cli.c @@ -650,6 +650,26 @@ static int ovpn_connect(struct ovpn_ctx *ovpn) return ret; } +static int ovpn_nl_put_vpn_addr(struct nl_msg *msg, + const struct ovpn_ctx *ovpn) +{ + if (!ovpn->peer_ip_set) + return 0; + + switch (ovpn->peer_ip.in4.sin_family) { + case AF_INET: + return nla_put_u32(msg, OVPN_A_PEER_VPN_IPV4, + ovpn->peer_ip.in4.sin_addr.s_addr); + case AF_INET6: + return nla_put(msg, OVPN_A_PEER_VPN_IPV6, + sizeof(struct in6_addr), + &ovpn->peer_ip.in6.sin6_addr); + default: + fprintf(stderr, "Invalid family for peer address\n"); + return -EAFNOSUPPORT; + } +} + static int ovpn_new_peer(struct ovpn_ctx *ovpn, bool is_tcp) { struct nlattr *attr; @@ -691,22 +711,9 @@ static int ovpn_new_peer(struct ovpn_ctx *ovpn, bool is_tcp) } } - if (ovpn->peer_ip_set) { - switch (ovpn->peer_ip.in4.sin_family) { - case AF_INET: - NLA_PUT_U32(ctx->nl_msg, OVPN_A_PEER_VPN_IPV4, - ovpn->peer_ip.in4.sin_addr.s_addr); - break; - case AF_INET6: - NLA_PUT(ctx->nl_msg, OVPN_A_PEER_VPN_IPV6, - sizeof(struct in6_addr), - &ovpn->peer_ip.in6.sin6_addr); - break; - default: - fprintf(stderr, "Invalid family for peer address\n"); - goto nla_put_failure; - } - } + ret = ovpn_nl_put_vpn_addr(ctx->nl_msg, ovpn); + if (ret) + goto nla_put_failure; nla_nest_end(ctx->nl_msg, attr); @@ -732,6 +739,10 @@ static int ovpn_set_peer(struct ovpn_ctx *ovpn) ovpn->keepalive_interval); NLA_PUT_U32(ctx->nl_msg, OVPN_A_PEER_KEEPALIVE_TIMEOUT, ovpn->keepalive_timeout); + + ret = ovpn_nl_put_vpn_addr(ctx->nl_msg, ovpn); + if (ret) + goto nla_put_failure; nla_nest_end(ctx->nl_msg, attr); ret = ovpn_nl_msg_send(ctx, NULL); @@ -1730,13 +1741,14 @@ static void usage(const char *cmd) fprintf(stderr, "\tmark: socket FW mark value\n"); fprintf(stderr, - "* set_peer : set peer attributes\n"); + "* set_peer [vpnaddr]: set peer attributes\n"); fprintf(stderr, "\tiface: ovpn interface name\n"); fprintf(stderr, "\tpeer_id: peer ID of the peer to modify\n"); fprintf(stderr, "\tkeepalive_interval: interval for sending ping messages\n"); fprintf(stderr, "\tkeepalive_timeout: time after which a peer is timed out\n"); + fprintf(stderr, "\tvpnaddr: peer VPN IP\n"); fprintf(stderr, "* del_peer : delete peer\n"); fprintf(stderr, "\tiface: ovpn interface name\n"); @@ -2090,6 +2102,8 @@ static int ovpn_run_cmd(struct ovpn_ctx *ovpn) return ret; ret = ovpn_new_peer(ovpn, false); + if (ret < 0) + return ret; ovpn_waitbg(); break; case CMD_NEW_MULTI_PEER: @@ -2331,6 +2345,12 @@ static int ovpn_parse_cmd_args(struct ovpn_ctx *ovpn, int argc, char *argv[]) "keepalive interval value out of range\n"); return -1; } + + if (argc > 6) { + ret = ovpn_parse_remote(ovpn, NULL, NULL, argv[6]); + if (ret < 0) + return -1; + } break; case CMD_DEL_PEER: if (argc < 4) diff --git a/tools/testing/selftests/net/ovpn/test.sh b/tools/testing/selftests/net/ovpn/test.sh index 9b5610837032..392109d5e14e 100755 --- a/tools/testing/selftests/net/ovpn/test.sh +++ b/tools/testing/selftests/net/ovpn/test.sh @@ -56,6 +56,76 @@ ovpn_prepare_network() { done } +ovpn_new_test_peer() { + local peer_id="$1" + + shift + ip netns exec ovpn_peer0 "${OVPN_CLI}" new_peer tun0 \ + "${peer_id}" none 65000 10.10.1.2 1 "$@" +} + +ovpn_set_peer_vpn_addr() { + ip netns exec ovpn_peer0 "${OVPN_CLI}" set_peer tun0 \ + "$1" 60 120 "$2" +} + +ovpn_run_vpn_addr_validation() { + local addr + local peer1_addr4 + local test_peer_id=$((OVPN_NUM_PEERS + 1)) + local test_peer_addr6="2001:db8::2" + # Do not include 0.0.0.0 or :: here. They are invalid on creation, but + # clear one address family on update and are valid if the other remains. + local -a invalid_addrs=( + "127.0.0.1" + "224.0.0.1" + "255.255.255.255" + "::1" + "::192.0.2.1" + "::ffff:192.0.2.1" + "ff02::1" + ) + + peer1_addr4=$(ovpn_peer_vpn_addr 1) + + ovpn_cmd_fail "reject peer without VPN address" \ + ovpn_new_test_peer "${test_peer_id}" + + for addr in "0.0.0.0" "::" "${invalid_addrs[@]}"; do + ovpn_cmd_fail "reject new peer VPN address ${addr}" \ + ovpn_new_test_peer "${test_peer_id}" "${addr}" + done + + ovpn_cmd_fail "reject duplicate IPv4 address on peer creation" \ + ovpn_new_test_peer "${test_peer_id}" "${peer1_addr4}" + ovpn_cmd_fail "reject clearing the last peer VPN address" \ + ovpn_set_peer_vpn_addr 1 0.0.0.0 + + for addr in "${invalid_addrs[@]}"; do + ovpn_cmd_fail "reject updated peer VPN address ${addr}" \ + ovpn_set_peer_vpn_addr 1 "${addr}" + done + + ovpn_cmd_fail "reject duplicate IPv4 address on peer update" \ + ovpn_set_peer_vpn_addr 2 "${peer1_addr4}" + + ovpn_cmd_ok "add peer IPv6 address" \ + ovpn_set_peer_vpn_addr 1 "${test_peer_addr6}" + ovpn_cmd_fail "reject duplicate IPv6 address on peer creation" \ + ovpn_new_test_peer "${test_peer_id}" "${test_peer_addr6}" + ovpn_cmd_fail "reject duplicate IPv6 address on peer update" \ + ovpn_set_peer_vpn_addr 2 "${test_peer_addr6}" + + ovpn_cmd_ok "clear peer IPv4 address" \ + ovpn_set_peer_vpn_addr 1 0.0.0.0 + ovpn_cmd_fail "reject clearing the remaining peer IPv6 address" \ + ovpn_set_peer_vpn_addr 1 :: + ovpn_cmd_ok "restore peer IPv4 address" \ + ovpn_set_peer_vpn_addr 1 "${peer1_addr4}" + ovpn_cmd_ok "clear peer IPv6 address" \ + ovpn_set_peer_vpn_addr 1 :: +} + ovpn_run_basic_traffic() { local p local header1 @@ -293,15 +363,16 @@ trap ovpn_stage_err ERR ktap_print_header if [ "${OVPN_FLOAT}" == "1" ]; then - ktap_set_plan 13 + ktap_set_plan 14 else - ktap_set_plan 12 + ktap_set_plan 13 fi ovpn_cleanup modprobe -q ovpn || true ovpn_run_stage "setup network topology" ovpn_prepare_network +ovpn_run_stage "validate peer VPN addresses" ovpn_run_vpn_addr_validation ovpn_run_stage "run baseline data traffic" ovpn_run_basic_traffic ovpn_run_stage "run LAN traffic behind peer1" ovpn_run_lan_traffic [ "${OVPN_FLOAT}" == "1" ] && ovpn_run_stage "run floating peer checks" \ From f0ca020cbb9bb7f3f4ea8ba1dfcf30a282aec91e Mon Sep 17 00:00:00 2001 From: Hui Peng Date: Sat, 19 Sep 2026 22:17:38 +0000 Subject: [PATCH 447/592] Bluetooth: bnep: fix out-of-bounds reads on short RX/TX frames and control fallthrough Fix multiple out-of-bounds reads in Bluetooth BNEP frame processing: 1. In bnep_rx_frame() and bnep_ctrl_frame() (net/bluetooth/bnep/core.c), use pskb_may_pull() to verify the BNEP header, control type byte, filter count, and extension headers exist before reading them, and return 0 after handling BNEP_CONTROL instead of falling through to Ethernet frame submission when no extension headers follow. 2. In bnep_net_xmit() (net/bluetooth/bnep/netdev.c), verify skb->len >= ETH_HLEN with pskb_may_pull() before reading the 14-byte Ethernet header to prevent an out-of-bounds heap read and infoleak on short AF_PACKET TX frames. Fixes: 1da177e4c3f4 ("Linux-2.6.12-rc2") Assisted-by: LLM Signed-off-by: Hui Peng Signed-off-by: Luiz Augusto von Dentz --- net/bluetooth/bnep/core.c | 17 ++++++++++++++++- net/bluetooth/bnep/netdev.c | 8 +++++++- 2 files changed, 23 insertions(+), 2 deletions(-) diff --git a/net/bluetooth/bnep/core.c b/net/bluetooth/bnep/core.c index f7d88c33e23e..ad24d2486665 100644 --- a/net/bluetooth/bnep/core.c +++ b/net/bluetooth/bnep/core.c @@ -270,9 +270,14 @@ static int bnep_rx_extension(struct bnep_session *s, struct sk_buff *skb) BT_DBG("type 0x%x len %u", h->type, h->len); + if (skb->len < h->len) { + err = -EILSEQ; + break; + } + switch (h->type & BNEP_TYPE_MASK) { case BNEP_EXT_CONTROL: - bnep_rx_control(s, skb->data, skb->len); + bnep_rx_control(s, skb->data, h->len); break; default: @@ -373,6 +378,11 @@ static int bnep_rx_frame(struct bnep_session *s, struct sk_buff *skb) goto badframe; } + if ((type & BNEP_TYPE_MASK) == BNEP_CONTROL) { + kfree_skb(skb); + return 0; + } + /* Strip 802.1p header */ if (ntohs(s->eh.h_proto) == ETH_P_8021Q) { if (!skb_pull(skb, 4)) @@ -451,6 +461,11 @@ static int bnep_tx_frame(struct bnep_session *s, struct sk_buff *skb) goto send; } + if (skb->len < ETH_HLEN) { + kfree_skb(skb); + return 0; + } + iv[il++] = (struct kvec) { &type, 1 }; len++; diff --git a/net/bluetooth/bnep/netdev.c b/net/bluetooth/bnep/netdev.c index ee1e39a3daff..b451ef457741 100644 --- a/net/bluetooth/bnep/netdev.c +++ b/net/bluetooth/bnep/netdev.c @@ -166,6 +166,12 @@ static netdev_tx_t bnep_net_xmit(struct sk_buff *skb, BT_DBG("skb %p, dev %p", skb, dev); + if (!pskb_may_pull(skb, ETH_HLEN)) { + dev->stats.tx_dropped++; + kfree_skb(skb); + return NETDEV_TX_OK; + } + #ifdef CONFIG_BT_BNEP_MC_FILTER if (bnep_net_mc_filter(skb, s)) { kfree_skb(skb); @@ -218,7 +224,7 @@ void bnep_net_setup(struct net_device *dev) dev->addr_len = ETH_ALEN; ether_setup(dev); - dev->min_mtu = 0; + dev->min_mtu = ETH_MIN_MTU; dev->max_mtu = ETH_MAX_MTU; dev->priv_flags &= ~IFF_TX_SKB_SHARING; dev->netdev_ops = &bnep_netdev_ops; From 37a11129345337efd6eef8e62b03b6348cd0dd8b Mon Sep 17 00:00:00 2001 From: Ravindra Date: Tue, 15 Sep 2026 10:42:15 +0530 Subject: [PATCH 448/592] Bluetooth: btintel_pcie: validate device-supplied DMA indices In btintel_pcie_msix_rx_handle(), the driver processes RX completion descriptors (urbd1) written by the PCIe device into DMA-coherent memory. urbd1->frbd_tag (a 16-bit field fully controlled by the device firmware via DMA) is used directly as an array index into rxq->bufs[] without any bounds check. rxq->bufs[] has only BTINTEL_PCIE_RX_DESCS_COUNT (64) entries, while frbd_tag can be any value 0-65535. A malicious or malfunctioning device can write an out-of-range frbd_tag, causing the driver to dereference an out-of-bounds data_buf pointer. Additionally, cr_hia is read from a DMA-shared index array also writable by the device; if the device sets cr_hia >= rxq->count, the while-loop never terminates because cr_tia is wrapped via modulo rxq->count and can never equal an out-of-range cr_hia. Add bounds validation for cr_hia and frbd_tag in the RX path, and cr_hia in the TX path. Log invalid values with bt_dev_err before returning. Fixes: c2b636b3f788 ("Bluetooth: btintel_pcie: Add support for PCIe transport") Signed-off-by: Ravindra Signed-off-by: Luiz Augusto von Dentz --- drivers/bluetooth/btintel_pcie.c | 16 ++++++++++++++++ 1 file changed, 16 insertions(+) diff --git a/drivers/bluetooth/btintel_pcie.c b/drivers/bluetooth/btintel_pcie.c index 6e6e2b19815c..2819e001797b 100644 --- a/drivers/bluetooth/btintel_pcie.c +++ b/drivers/bluetooth/btintel_pcie.c @@ -1099,6 +1099,11 @@ static void btintel_pcie_msix_tx_handle(struct btintel_pcie_data *data) txq = &data->txq; + if (cr_hia >= txq->count) { + bt_dev_err(data->hdev, "TXQ: invalid cr_hia %u", cr_hia); + return; + } + while (cr_tia != cr_hia) { data->tx_wait_done = true; wake_up(&data->tx_wait_q); @@ -1650,6 +1655,11 @@ static void btintel_pcie_msix_rx_handle(struct btintel_pcie_data *data) rxq = &data->rxq; + if (cr_hia >= rxq->count) { + bt_dev_err(hdev, "RXQ: invalid cr_hia %u", cr_hia); + return; + } + /* The firmware sends multiple CD in a single MSI-X and it needs to * process all received CDs in this interrupt. */ @@ -1657,6 +1667,12 @@ static void btintel_pcie_msix_rx_handle(struct btintel_pcie_data *data) urbd1 = &rxq->urbd1s[cr_tia]; ipc_print_urbd1(data->hdev, urbd1, cr_tia); + if (urbd1->frbd_tag >= rxq->count) { + bt_dev_err(hdev, "RXQ: invalid frbd_tag %u", + urbd1->frbd_tag); + return; + } + buf = &rxq->bufs[urbd1->frbd_tag]; if (!buf) { bt_dev_err(hdev, "RXQ: failed to get the DMA buffer for %d", From 46f8ffd0a1f1eb6cbc94946a92c11ef601e228a1 Mon Sep 17 00:00:00 2001 From: Hui Peng Date: Sat, 19 Sep 2026 11:25:18 +0000 Subject: [PATCH 449/592] Bluetooth: RFCOMM: fix NULL dereference of dlc->session in RFCOMM_CONNINFO The RFCOMM_CONNINFO getsockopt handler accepts a socket that is not connected as long as deferred setup is enabled: if (sk->sk_state != BT_CONNECTED && !rfcomm_pi(sk)->dlc->defer_setup) { err = -ENOTCONN; break; } l2cap_sk = rfcomm_pi(sk)->dlc->session->sock->sk; dlc->defer_setup is set in rfcomm_sock_init() when rfcomm_connect_ind() creates a child socket for an incoming connection on a listening socket that has BT_DEFER_SETUP enabled. It is never cleared afterwards. The session, however, can go away underneath it. rfcomm_recv_disc() forces the dlc state before tearing it down: d->state = BT_CLOSED; __rfcomm_dlc_close(d, err); The RFCOMM_DEFER_SETUP early return in __rfcomm_dlc_close() only covers BT_CONNECT, BT_CONFIG, BT_OPEN and BT_CONNECT2, so with the state already BT_CLOSED that switch does not match and the function falls through to rfcomm_dlc_unlink(), which sets d->session = NULL, while d->defer_setup stays 1. A getsockopt(SOL_RFCOMM, RFCOMM_CONNINFO) on the accepted socket after that point therefore skips the -ENOTCONN path -- sk->sk_state is BT_CLOSED, but dlc->defer_setup is still set -- and dereferences the NULL session. No race is needed: once the DISC has been processed, the dereference is unconditional. Reproduced on a KASAN kernel under QEMU with a BR/EDR peer emulated over /dev/vhci: the peer brings up an ACL link, opens L2CAP on the RFCOMM PSM, starts a session and sends SABM for a channel bound with BT_DEFER_SETUP, and sends DISC for that dlci after the socket has been accepted. getsockopt(SOL_RFCOMM, RFCOMM_CONNINFO) on the accepted socket then hits: Oops: general protection fault, probably for non-canonical address 0xdffffc0000000002: 0000 [#1] SMP KASAN PTI KASAN: null-ptr-deref in range [0x0000000000000010-0x0000000000000017] CPU: 1 UID: 0 PID: 150 Comm: init Tainted: G B 7.3.0-rc3-g5dd1818b15d9 Hardware name: QEMU Standard PC (i440FX + PIIX, 1996) RIP: 0010:rfcomm_sock_getsockopt+0x529/0x780 Call Trace: do_sock_getsockopt+0x3ad/0x7d0 __sys_getsockopt+0x10e/0x1b0 __x64_sys_getsockopt+0xc2/0x160 do_syscall_64+0xda/0x4b0 entry_SYSCALL_64_after_hwframe+0x77/0x7f 0x10 is the offset of sock in struct rfcomm_session; rfcomm_sock_getsockopt_old() is inlined into rfcomm_sock_getsockopt(). Commit 43a556b2fd43 ("Bluetooth: RFCOMM: take rfcomm_mutex for the deferred setup accept") fixed the same "a remote DISC clears the session while deferred setup is still flagged" problem in rfcomm_dlc_accept(); this is the remaining instance of it, in the getsockopt path. Deferred setup only leaves a socket usable here once it has reached BT_CONNECT2, so restrict the exception to that state and check that a session is actually present before following it. Fixes: bb23c0ab8246 ("Bluetooth: Add support for deferring RFCOMM connection setup") Cc: stable@vger.kernel.org Assisted-by: LLM Signed-off-by: Hui Peng Signed-off-by: Luiz Augusto von Dentz --- net/bluetooth/rfcomm/sock.c | 6 ++++-- 1 file changed, 4 insertions(+), 2 deletions(-) diff --git a/net/bluetooth/rfcomm/sock.c b/net/bluetooth/rfcomm/sock.c index e2486bc11cbc..fb924d0e34ec 100644 --- a/net/bluetooth/rfcomm/sock.c +++ b/net/bluetooth/rfcomm/sock.c @@ -786,8 +786,10 @@ static int rfcomm_sock_getsockopt_old(struct socket *sock, int optname, break; case RFCOMM_CONNINFO: - if (sk->sk_state != BT_CONNECTED && - !rfcomm_pi(sk)->dlc->defer_setup) { + if ((sk->sk_state != BT_CONNECTED && + !(sk->sk_state == BT_CONNECT2 && + rfcomm_pi(sk)->dlc->defer_setup)) || + !rfcomm_pi(sk)->dlc->session) { err = -ENOTCONN; break; } From 6d91041bb38b97e2feb625123cc0529d7b83a0e1 Mon Sep 17 00:00:00 2001 From: Hui Peng Date: Sat, 19 Sep 2026 11:25:14 +0000 Subject: [PATCH 450/592] Bluetooth: RFCOMM: Reject short EA=0 frames in rfcomm_recv_frame() While rfcomm_recv_frame() verifies that skb->len is at least sizeof(*hdr) + 1 (4 bytes: 3-byte header + 1-byte FCS), an RFCOMM frame with an extended 2-byte length field (!__test_ea(hdr->len)) has a 4-byte header plus a 1-byte FCS (5 bytes minimum, sizeof(*hdr) + 2). When a 4-byte RFCOMM frame with EA == 0 arrives: 1. The initial skb->len < sizeof(*hdr) + 1 check passes (4 < 4 is false). 2. Trimming the FCS byte decrements skb->len to 3. 3. If __check_fcs() succeeds, skb_pull(skb, 4) fails (4 > 3) and returns NULL without advancing skb->data. 4. Because the return value of skb_pull() is ignored, the un-pulled 3-byte struct rfcomm_hdr remains at skb->data and is either queued as application payload via rfcomm_recv_data() or parsed as a multiplexer control command via rfcomm_recv_mcc() on DLCI 0. Fix this by extending the length check in rfcomm_recv_frame() to also require skb->len >= sizeof(*hdr) + 2 when !__test_ea(hdr->len). Fixes: b230e5bf501c ("Bluetooth: RFCOMM: validate skb length in rfcomm_recv_frame") Assisted-by: LLM Signed-off-by: Hui Peng Signed-off-by: Luiz Augusto von Dentz --- net/bluetooth/rfcomm/core.c | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-) diff --git a/net/bluetooth/rfcomm/core.c b/net/bluetooth/rfcomm/core.c index f7463f092283..d91e2a6ee26c 100644 --- a/net/bluetooth/rfcomm/core.c +++ b/net/bluetooth/rfcomm/core.c @@ -1817,7 +1817,8 @@ static struct rfcomm_session *rfcomm_recv_frame(struct rfcomm_session *s, return s; } - if (skb->len < sizeof(*hdr) + 1) { + if (skb->len < sizeof(*hdr) + 1 || + (!__test_ea(hdr->len) && skb->len < sizeof(*hdr) + 2)) { kfree_skb(skb); return s; } From 94b7e3a7e871ae27d4935c76959dfc61829f27f9 Mon Sep 17 00:00:00 2001 From: Helge Deller Date: Sat, 19 Sep 2026 22:09:20 +0200 Subject: [PATCH 451/592] parisc: Increase kernel stack size to 32kb For 64-bit Linux kernels, increase the default kernel stack size (THREAD_SIZE_ORDER) to 32 kB, in order to avoid kernel crashes which have been triggered recently when building the debian vtk9 package with gcc 17: stackcheck: kworker/u128:0 will most likely overflow kernel stack (sp:179a83af0, stk bottom-top:179a80000-179a84000) Kernel panic - not syncing: low stack detected by irq handler - check messages CPU: 2 UID: 0 PID: 30760 Comm: kworker/u128:0 Tainted: G W 6.18.46-dirty #1 NONE Tainted: [W]=WARN Hardware name: 9000/800/rp3440 Workqueue: writeback wb_workfn (flush-259:0) Backtrace: [<000000004022f050>] show_stack+0x70/0x90 [<000000004022378c>] dump_stack_lvl+0x124/0x190 [<000000004022382c>] dump_stack+0x34/0x48 [<000000004020212c>] vpanic+0x204/0x648 [<00000000402025c4>] panic+0x54/0x58 [<0000000040232230>] do_cpu_irq_mask+0x3f8/0x440 [<0000000040227070>] intr_return+0x0/0xc Signed-off-by: Helge Deller Reported-by: John David Anglin Cc: stable@vger.kernel.org # v6.18+ --- arch/parisc/include/asm/thread_info.h | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/arch/parisc/include/asm/thread_info.h b/arch/parisc/include/asm/thread_info.h index b283738bb6da..249370706242 100644 --- a/arch/parisc/include/asm/thread_info.h +++ b/arch/parisc/include/asm/thread_info.h @@ -24,7 +24,7 @@ struct thread_info { /* thread information allocation */ -#ifdef CONFIG_IRQSTACKS +#if defined(CONFIG_IRQSTACKS) && !defined(CONFIG_64BIT) #define THREAD_SIZE_ORDER 2 /* PA-RISC requires at least 16k stack */ #else #define THREAD_SIZE_ORDER 3 /* PA-RISC requires at least 32k stack */ From 79a9172f3ab4ad8392c5e5c8944b9b7710ade620 Mon Sep 17 00:00:00 2001 From: Xu Yunxiang Date: Mon, 21 Sep 2026 05:04:21 +0800 Subject: [PATCH 452/592] bpf: Reject non-negative offsets in stack_slot_obj_get_spi() bpf_get_spi() computes (-off - 1) / BPF_REG_SIZE using C division, which truncates toward zero. For off == 0, this produces spi 0, the same index used by the valid stack slot at fp-8. stack_slot_obj_get_spi() currently checks alignment and the resulting spi bounds, but does not reject the non-negative offset itself. It can therefore validate a PTR_TO_STACK register holding fp+0 against an iterator stored at fp-8 even though the runtime receives the actual fp+0 pointer. An effectful iterator kfunc can then interpret memory outside the BPF stack as iterator state. Reject non-negative offsets before converting the offset to an spi. All valid stack objects begin at a negative offset from the frame pointer. Fixes: 06accc8779c1 ("bpf: add support for open-coded iterator loops") Signed-off-by: Xu Yunxiang Signed-off-by: Andrii Nakryiko Reviewed-by: Sun Jian Link: https://lore.kernel.org/bpf/20260920210423.345636-2-xyx2021@mail.ustc.edu.cn --- kernel/bpf/verifier.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/kernel/bpf/verifier.c b/kernel/bpf/verifier.c index 02be326f235c..dd8bb179d39e 100644 --- a/kernel/bpf/verifier.c +++ b/kernel/bpf/verifier.c @@ -567,7 +567,7 @@ static int stack_slot_obj_get_spi(struct bpf_verifier_env *env, struct bpf_reg_s } off = reg->var_off.value; - if (off % BPF_REG_SIZE) { + if (off >= 0 || off % BPF_REG_SIZE) { verbose(env, "cannot pass in %s at an offset=%d\n", obj_kind, off); return -EINVAL; } From 8244668cbbffc8e242b2c5204a2dea20bfca096c Mon Sep 17 00:00:00 2001 From: Xu Yunxiang Date: Mon, 21 Sep 2026 05:04:22 +0800 Subject: [PATCH 453/592] selftests/bpf: Reject iterator destruction through fp+0 Add a verifier regression test that initializes a numeric iterator at fp-8 and attempts to destroy it through fp+0. The verifier must reject the non-negative offset instead of treating it as the initialized stack slot. Check the offset diagnostic to ensure rejection happens at the stack object address check. The numeric iterator destroy operation is a no-op; this test checks verifier rejection and does not run the program. Signed-off-by: Xu Yunxiang Signed-off-by: Andrii Nakryiko Reviewed-by: Sun Jian Link: https://lore.kernel.org/bpf/20260920210423.345636-3-xyx2021@mail.ustc.edu.cn --- .../selftests/bpf/progs/iters_state_safety.c | 22 +++++++++++++++++++ 1 file changed, 22 insertions(+) diff --git a/tools/testing/selftests/bpf/progs/iters_state_safety.c b/tools/testing/selftests/bpf/progs/iters_state_safety.c index 646026430e9b..e5bb9fe6d5e5 100644 --- a/tools/testing/selftests/bpf/progs/iters_state_safety.c +++ b/tools/testing/selftests/bpf/progs/iters_state_safety.c @@ -52,6 +52,28 @@ int create_and_destroy(void *ctx) return 0; } +/* fp+0 is not a stack slot. bpf_get_spi(0) used to alias spi 0 (fp-8). */ +SEC("?raw_tp") +__failure __msg("cannot pass in iter at an offset=0") +int destroy_fp0_fail(void *ctx) +{ + struct bpf_iter_num iter; + + asm volatile ("r1 = %[iter];" + "r2 = 0;" + "r3 = 1000;" + "call %[bpf_iter_num_new];" + /* r10 is fp+0, one byte above the top of the BPF stack */ + "r1 = r10;" + "call %[bpf_iter_num_destroy];" + : + : __imm_ptr(iter), ITER_HELPERS + : __clobber_common + ); + + return 0; +} + SEC("?raw_tp") __failure __msg("Unreleased reference id=1") int create_and_forget_to_destroy_fail(void *ctx) From d06f2ebf67ff2962fe00d687e4f0d4703eb41a12 Mon Sep 17 00:00:00 2001 From: Ratheesh Kannoth Date: Wed, 16 Sep 2026 07:51:11 +0530 Subject: [PATCH 454/592] octeontx2-af: Fix memory scaling limitation in SR-IOV mode The original code used DMA_ATTR_FORCE_CONTIGUOUS, which could exhaust the CMA pool when a large number of VFs were requested. Fix this by switching to the DMA streaming API. This is equivalent on Octeon platforms, which provide full I/O coherency via the SMMU. Cc: Leon Romanovsky Fixes: 73d33dbc0723 ("octeontx2-af: Use DMA_ATTR_FORCE_CONTIGUOUS attribute in DMA alloc") Signed-off-by: Ratheesh Kannoth Reviewed-by: Leon Romanovsky Link: https://patch.msgid.link/20260916022111.1083017-1-rkannoth@marvell.com Signed-off-by: Jakub Kicinski --- .../ethernet/marvell/octeontx2/af/common.h | 45 ++++++++++++++++--- 1 file changed, 39 insertions(+), 6 deletions(-) diff --git a/drivers/net/ethernet/marvell/octeontx2/af/common.h b/drivers/net/ethernet/marvell/octeontx2/af/common.h index 779413a383b7..78e42549d990 100644 --- a/drivers/net/ethernet/marvell/octeontx2/af/common.h +++ b/drivers/net/ethernet/marvell/octeontx2/af/common.h @@ -7,6 +7,10 @@ #ifndef COMMON_H #define COMMON_H +#include +#include +#include + #include "rvu_struct.h" #define OTX2_ALIGN 128 /* Align to cacheline */ @@ -44,6 +48,33 @@ struct qmem { u32 qsize; }; +static inline void *otx2_dma_alloc_coherent(struct device *dev, size_t size, + dma_addr_t *dma_handle) +{ + dma_addr_t dma_addr; + void *vaddr; + + vaddr = kzalloc(size, GFP_KERNEL); + if (!vaddr) + return NULL; + + dma_addr = dma_map_single(dev, vaddr, size, DMA_BIDIRECTIONAL); + if (dma_mapping_error(dev, dma_addr)) { + kfree(vaddr); + return NULL; + } + + *dma_handle = dma_addr; + return vaddr; +} + +static inline void otx2_dma_free_coherent(struct device *dev, size_t size, + void *vaddr, dma_addr_t dma_handle) +{ + dma_unmap_single(dev, dma_handle, size, DMA_BIDIRECTIONAL); + kfree(vaddr); +} + static inline int qmem_alloc(struct device *dev, struct qmem **q, int qsize, int entry_sz) { @@ -60,8 +91,11 @@ static inline int qmem_alloc(struct device *dev, struct qmem **q, qmem->entry_sz = entry_sz; qmem->alloc_sz = (qsize * entry_sz) + OTX2_ALIGN; - qmem->base = dma_alloc_attrs(dev, qmem->alloc_sz, &qmem->iova, - GFP_KERNEL, DMA_ATTR_FORCE_CONTIGUOUS); + + if (get_order(PAGE_ALIGN(qmem->alloc_sz)) > MAX_PAGE_ORDER) + return -ENOMEM; + + qmem->base = otx2_dma_alloc_coherent(dev, qmem->alloc_sz, &qmem->iova); if (!qmem->base) return -ENOMEM; @@ -80,10 +114,9 @@ static inline void qmem_free(struct device *dev, struct qmem *qmem) return; if (qmem->base) - dma_free_attrs(dev, qmem->alloc_sz, - qmem->base - qmem->align, - qmem->iova - qmem->align, - DMA_ATTR_FORCE_CONTIGUOUS); + otx2_dma_free_coherent(dev, qmem->alloc_sz, + qmem->base - qmem->align, + qmem->iova - qmem->align); devm_kfree(dev, qmem); } From 0346ec2f080b40d95ed05b853bb9226289e75212 Mon Sep 17 00:00:00 2001 From: Kuniyuki Iwashima Date: Fri, 18 Sep 2026 08:22:05 +0000 Subject: [PATCH 455/592] ipv6: Prevent rt6_insert_exception() for dying fib6_info. Before the cited commit, fib6_nh_flush_exceptions() always set from->exception_bucket_flushed = 1 under rt6_exception_lock to prevent rt6_insert_exception() from inserting a new exception for a dying fib6_info. The flag was replaced with the FIB6_EXCEPTION_BUCKET_FLUSHED bit stored in nh->rt6i_exception_bucket. The problem is that now the bit is only set when the bucket is not NULL and fib6_nh_flush_exceptions() is called from fib6_nh_release() after fib6_ref has already reached zero. If rt6_insert_exception() is called while the target fib6_info is being removed via fib6_purge_rt(), a new exception could be created successfully because rt6_flush_exceptions() no longer sets the bit. This creates a reference cycle between the fib6_info and the exception route, leaking the fib6_info, its nexthop device, and all per-CPU routes in fib6_nh->rt6i_pcpu, which stalls netdev unregistration. [ 34.680602] unregister_netdevice: waiting for gre6 to become free. Usage count = 68 [ 44.920675] unregister_netdevice: waiting for gre6 to become free. Usage count = 68 [ 55.176582] unregister_netdevice: waiting for gre6 to become free. Usage count = 68 Let's call fib6_drop_pcpu_from() before rt6_flush_exceptions(), to set fib6_destroying before rt6_exception_lock, and check f6i->fib6_destroying in rt6_insert_exception(). Note that FIB6_EXCEPTION_BUCKET_FLUSHED logic is dead and we can clean it up in net-next. Fixes: cc5c073a693f ("ipv6: Move exception bucket to fib6_nh") Signed-off-by: Kuniyuki Iwashima Reviewed-by: Ido Schimmel Link: https://patch.msgid.link/20260918082209.2853582-1-kuniyu@google.com Signed-off-by: Jakub Kicinski --- net/ipv6/ip6_fib.c | 2 +- net/ipv6/route.c | 5 +++++ 2 files changed, 6 insertions(+), 1 deletion(-) diff --git a/net/ipv6/ip6_fib.c b/net/ipv6/ip6_fib.c index 9ea75703b38d..9ff761962b45 100644 --- a/net/ipv6/ip6_fib.c +++ b/net/ipv6/ip6_fib.c @@ -1043,8 +1043,8 @@ static void fib6_purge_rt(struct fib6_info *rt, struct fib6_node *fn, struct fib6_table *table = rt->fib6_table; /* Flush all cached dst in exception table */ - rt6_flush_exceptions(rt); fib6_drop_pcpu_from(rt); + rt6_flush_exceptions(rt); if (rt->nh) { spin_lock(&rt->nh->lock); diff --git a/net/ipv6/route.c b/net/ipv6/route.c index 08bd68f1b5bb..884d9ab0d50d 100644 --- a/net/ipv6/route.c +++ b/net/ipv6/route.c @@ -1729,6 +1729,11 @@ static int rt6_insert_exception(struct rt6_info *nrt, spin_lock_bh(&rt6_exception_lock); + if (f6i->fib6_destroying) { + err = -ENOENT; + goto out; + } + bucket = rcu_dereference_protected(nh->rt6i_exception_bucket, lockdep_is_held(&rt6_exception_lock)); if (!bucket) { From 31995571219c8ac30913d9c0dccad033fbb0b3da Mon Sep 17 00:00:00 2001 From: Nicolai Buchwitz Date: Fri, 18 Sep 2026 11:55:40 +0200 Subject: [PATCH 456/592] net: don't require the hwtstamp NDOs when a PHY provides timestamping Removing the legacy ioctl fallback made both hwtstamp NDOs mandatory. A device that only timestamps in its PHY implements neither, so SIOCSHWTSTAMP fails with EOPNOTSUPP before anything looks at the PHY and PTP stops working there. The check only ever picked the legacy path. That path is gone, so drop it and test where the NDOs are actually called. SIOCGHWTSTAMP is new here, not restored. The old path went through phy_mii_ioctl(), which only handled SIOCSHWTSTAMP. Such a device now returns -ENODEV while absent instead of -EOPNOTSUPP, like the ones that do implement the NDOs. Fixes: 5062245a5a7f ("net: remove legacy way to get/set HW timestamp config") Signed-off-by: Nicolai Buchwitz Reviewed-by: Kory Maincent Link: https://patch.msgid.link/20260918095540.34286-1-nb@tipi-net.de Signed-off-by: Jakub Kicinski --- net/core/dev_ioctl.c | 25 +++++++++---------------- 1 file changed, 9 insertions(+), 16 deletions(-) diff --git a/net/core/dev_ioctl.c b/net/core/dev_ioctl.c index a320e264eaaf..164643140a52 100644 --- a/net/core/dev_ioctl.c +++ b/net/core/dev_ioctl.c @@ -276,19 +276,18 @@ int dev_get_hwtstamp_phylib(struct net_device *dev, if (phy_is_default_hwtstamp(dev->phydev)) return phy_hwtstamp_get(dev->phydev, cfg); + if (!dev->netdev_ops->ndo_hwtstamp_get) + return -EOPNOTSUPP; + return dev->netdev_ops->ndo_hwtstamp_get(dev, cfg); } static int dev_get_hwtstamp(struct net_device *dev, struct ifreq *ifr) { - const struct net_device_ops *ops = dev->netdev_ops; struct kernel_hwtstamp_config kernel_cfg = {}; struct hwtstamp_config cfg; int err; - if (!ops->ndo_hwtstamp_get) - return -EOPNOTSUPP; - if (!netif_device_present(dev)) return -ENODEV; @@ -359,12 +358,18 @@ int dev_set_hwtstamp_phylib(struct net_device *dev, cfg->source = phy_ts ? HWTSTAMP_SOURCE_PHYLIB : HWTSTAMP_SOURCE_NETDEV; if (phy_ts && dev->see_all_hwtstamp_requests) { + if (!ops->ndo_hwtstamp_get) + return -EOPNOTSUPP; + err = ops->ndo_hwtstamp_get(dev, &old_cfg); if (err) return err; } if (!phy_ts || dev->see_all_hwtstamp_requests) { + if (!ops->ndo_hwtstamp_set) + return -EOPNOTSUPP; + err = ops->ndo_hwtstamp_set(dev, cfg, extack); if (err) { if (extack->_msg) @@ -390,7 +395,6 @@ int dev_set_hwtstamp_phylib(struct net_device *dev, static int dev_set_hwtstamp(struct net_device *dev, struct ifreq *ifr) { - const struct net_device_ops *ops = dev->netdev_ops; struct kernel_hwtstamp_config kernel_cfg = {}; struct netlink_ext_ack extack = {}; struct hwtstamp_config cfg; @@ -413,9 +417,6 @@ static int dev_set_hwtstamp(struct net_device *dev, struct ifreq *ifr) return err; } - if (!ops->ndo_hwtstamp_set) - return -EOPNOTSUPP; - if (!netif_device_present(dev)) return -ENODEV; @@ -441,15 +442,11 @@ static int dev_set_hwtstamp(struct net_device *dev, struct ifreq *ifr) int generic_hwtstamp_get_lower(struct net_device *dev, struct kernel_hwtstamp_config *kernel_cfg) { - const struct net_device_ops *ops = dev->netdev_ops; int err; if (!netif_device_present(dev)) return -ENODEV; - if (!ops->ndo_hwtstamp_get) - return -EOPNOTSUPP; - netdev_lock_ops(dev); err = dev_get_hwtstamp_phylib(dev, kernel_cfg); netdev_unlock_ops(dev); @@ -462,15 +459,11 @@ int generic_hwtstamp_set_lower(struct net_device *dev, struct kernel_hwtstamp_config *kernel_cfg, struct netlink_ext_ack *extack) { - const struct net_device_ops *ops = dev->netdev_ops; int err; if (!netif_device_present(dev)) return -ENODEV; - if (!ops->ndo_hwtstamp_set) - return -EOPNOTSUPP; - netdev_lock_ops(dev); err = dev_set_hwtstamp_phylib(dev, kernel_cfg, extack); netdev_unlock_ops(dev); From 7cce782d8327b7291334c4a304cf3fd909a74d9d Mon Sep 17 00:00:00 2001 From: Ivan Vecera Date: Thu, 17 Sep 2026 16:37:36 +0200 Subject: [PATCH 457/592] dpll: use exact lookup for reference sync pin id dpll_pin_ref_sync_state_set() looks up the reference sync pin in the pin->ref_sync_pins xarray, which is keyed by the sync pin's id (see dpll_pin_ref_sync_pair_add() using xa_insert() with ref_sync_pin->id). The pin id to operate on is supplied by userspace via DPLL_A_PIN_ID. The lookup however used xa_find() with a ULONG_MAX limit, which returns the first present entry with an index greater than or equal to the requested id, not the entry stored exactly at that id. If userspace passes an id that is not paired as a reference sync pin, but another pin with a higher id is present in the xarray, xa_find() silently returns that wrong pin and the subsequent ref_sync_set() operates on it. The request only fails when the given id is larger than every present key. Use xa_load() for an exact-key lookup instead, mirroring the deletion path in dpll_pin_ref_sync_pair_del(). Fixes: 58256a26bfb3 ("dpll: add reference sync get/set") Signed-off-by: Ivan Vecera Link: https://patch.msgid.link/20260917143736.526221-1-ivecera@redhat.com Signed-off-by: Jakub Kicinski --- drivers/dpll/dpll_netlink.c | 3 +-- 1 file changed, 1 insertion(+), 2 deletions(-) diff --git a/drivers/dpll/dpll_netlink.c b/drivers/dpll/dpll_netlink.c index 45365214fbef..fb24fd53f2e1 100644 --- a/drivers/dpll/dpll_netlink.c +++ b/drivers/dpll/dpll_netlink.c @@ -1210,8 +1210,7 @@ dpll_pin_ref_sync_state_set(struct dpll_pin *pin, struct dpll_device *dpll; int ret; - ref_sync_pin = xa_find(&pin->ref_sync_pins, &ref_sync_pin_idx, - ULONG_MAX, XA_PRESENT); + ref_sync_pin = xa_load(&pin->ref_sync_pins, ref_sync_pin_idx); if (!ref_sync_pin) { NL_SET_ERR_MSG(extack, "reference sync pin not found"); return -EINVAL; From c06bde80ae7a7b595732f7cabcb92cf08db9d56a Mon Sep 17 00:00:00 2001 From: Pengpeng Hou Date: Sun, 20 Sep 2026 11:47:45 +0800 Subject: [PATCH 458/592] net: usb: sr9700: include receive overhead in the length check The receive fixup subtracts the Ethernet CRC from the reported packet length, but compares that payload length against the whole remaining receive buffer. The following copy starts after the three-byte header, and the cursor advance consumes both that header and the four-byte CRC. Require the payload to fit after SR_RX_OVERHEAD before copying it or advancing to the next packet. The loop already ensures that the remaining buffer is larger than the overhead, so the subtraction is safe. The issue was found by our static-analysis tool. Fixes: c9b37458e956 ("USB2NET : SR9700 : One chip USB 1.1 USB2NET SR9700Device Driver Support") Reviewed-by: Ethan Nelson-Moore Tested-by: Ethan Nelson-Moore Signed-off-by: Pengpeng Hou Link: https://patch.msgid.link/20260920034745.18468-1-hppiscas@163.com Signed-off-by: Jakub Kicinski --- drivers/net/usb/sr9700.c | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-) diff --git a/drivers/net/usb/sr9700.c b/drivers/net/usb/sr9700.c index 937e6fef3ac6..50981a28376a 100644 --- a/drivers/net/usb/sr9700.c +++ b/drivers/net/usb/sr9700.c @@ -355,7 +355,8 @@ static int sr9700_rx_fixup(struct usbnet *dev, struct sk_buff *skb) /* ignore the CRC length */ len = (skb->data[1] | (skb->data[2] << 8)) - 4; - if (len > ETH_FRAME_LEN || len > skb->len || len < 0) + if (len > ETH_FRAME_LEN || len < 0 || + len > skb->len - SR_RX_OVERHEAD) return 0; /* the last packet of current skb */ From be581d6635579489eadff9cea4caee142ec6d8bc Mon Sep 17 00:00:00 2001 From: Yuya Kusakabe Date: Fri, 18 Sep 2026 22:49:30 +0900 Subject: [PATCH 459/592] selftests: net: fix CONFIG_SYSCTL sort order in configs Commit 8d75c338f0bc ("sysctl: remove CONFIG_PROC_SYSCTL, it just mirrors CONFIG_SYSCTL") renamed CONFIG_PROC_SYSCTL to CONFIG_SYSCTL in place, which left the entry out of alphabetical order in the net and packetdrill configs. The netdev CI check for sorted selftest configs now fails for every patch that touches either file. Fixes: 8d75c338f0bc ("sysctl: remove CONFIG_PROC_SYSCTL, it just mirrors CONFIG_SYSCTL") Signed-off-by: Yuya Kusakabe Reviewed-by: Joel Granados Link: https://patch.msgid.link/20260918-selftests-net-config-sort-v1-1-968ea6e8c1b7@gmail.com Signed-off-by: Jakub Kicinski --- tools/testing/selftests/net/config | 2 +- tools/testing/selftests/net/packetdrill/config | 2 +- 2 files changed, 2 insertions(+), 2 deletions(-) diff --git a/tools/testing/selftests/net/config b/tools/testing/selftests/net/config index 30d5fcb09a83..737e7e6327b3 100644 --- a/tools/testing/selftests/net/config +++ b/tools/testing/selftests/net/config @@ -118,10 +118,10 @@ CONFIG_NFT_NAT=m CONFIG_NUMA=y CONFIG_OPENVSWITCH=m CONFIG_PAGE_POOL_STATS=y -CONFIG_SYSCTL=y CONFIG_PSAMPLE=m CONFIG_RPS=y CONFIG_SYN_COOKIES=y +CONFIG_SYSCTL=y CONFIG_SYSFS=y CONFIG_TAP=m CONFIG_TCP_CONG_DCTCP=y diff --git a/tools/testing/selftests/net/packetdrill/config b/tools/testing/selftests/net/packetdrill/config index 83dde525c53c..b7df8bf30920 100644 --- a/tools/testing/selftests/net/packetdrill/config +++ b/tools/testing/selftests/net/packetdrill/config @@ -4,8 +4,8 @@ CONFIG_IPV6=y CONFIG_NET_NS=y CONFIG_NET_SCH_FIFO=y CONFIG_NET_SCH_FQ=y -CONFIG_SYSCTL=y CONFIG_SYN_COOKIES=y +CONFIG_SYSCTL=y CONFIG_TCP_CONG_CUBIC=y CONFIG_TCP_MD5SIG=y CONFIG_TUN=y From 10de7ed8ef4840da9ca21de4c29578657ac367db Mon Sep 17 00:00:00 2001 From: Andrea Parri Date: Thu, 17 Sep 2026 13:55:42 +0200 Subject: [PATCH 460/592] net/mlx5e: fix swapped IPv6 IPsec policy masks IPv6 XFRM policies may use different source and destination prefix lengths. mlx5e_ipsec_policy_mask() builds the corresponding masks independently, but setup_fte_addr6() installs each mask in the opposite address field. When the prefix lengths differ, this makes the source match use the destination prefix and the destination match use the source prefix. The resulting hardware rule can both miss traffic covered by the policy and match traffic outside it. Install each mask in its corresponding match field. Fixes: ca7992f52c2c ("net/mlx5e: Properly match IPsec subnet addresses") Cc: stable@vger.kernel.org Signed-off-by: Andrea Parri Reviewed-by: Tariq Toukan Link: https://patch.msgid.link/20260917115542.177675-1-parri.andrea@gmail.com Signed-off-by: Jakub Kicinski --- drivers/net/ethernet/mellanox/mlx5/core/en_accel/ipsec_fs.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en_accel/ipsec_fs.c b/drivers/net/ethernet/mellanox/mlx5/core/en_accel/ipsec_fs.c index 329608c59313..8ffa8068e90a 100644 --- a/drivers/net/ethernet/mellanox/mlx5/core/en_accel/ipsec_fs.c +++ b/drivers/net/ethernet/mellanox/mlx5/core/en_accel/ipsec_fs.c @@ -1564,14 +1564,14 @@ static void setup_fte_addr6(struct mlx5_flow_spec *spec, memcpy(MLX5_ADDR_OF(fte_match_param, spec->match_value, outer_headers.src_ipv4_src_ipv6.ipv6_layout.ipv6), saddr, 16); memcpy(MLX5_ADDR_OF(fte_match_param, spec->match_criteria, - outer_headers.src_ipv4_src_ipv6.ipv6_layout.ipv6), dmask, 16); + outer_headers.src_ipv4_src_ipv6.ipv6_layout.ipv6), smask, 16); } if (!addr6_all_zero(daddr)) { memcpy(MLX5_ADDR_OF(fte_match_param, spec->match_value, outer_headers.dst_ipv4_dst_ipv6.ipv6_layout.ipv6), daddr, 16); memcpy(MLX5_ADDR_OF(fte_match_param, spec->match_criteria, - outer_headers.dst_ipv4_dst_ipv6.ipv6_layout.ipv6), smask, 16); + outer_headers.dst_ipv4_dst_ipv6.ipv6_layout.ipv6), dmask, 16); } } From be31fe6333f534155e6b408f1ef6d77974bb41aa Mon Sep 17 00:00:00 2001 From: Kuniyuki Iwashima Date: Sun, 20 Sep 2026 19:14:32 +0000 Subject: [PATCH 461/592] ipv6: Fix dst leak for uncached routes. ip6_route_output_flags(), ip6_rt_put_flags(), and ip6_dst_check() detect an uncached route by list_empty(&rt->dst.rt_uncached), which replaced the static DST_NOCACHE flag check in commit a4c2fd7f7891 ("net: remove DST_NOCACHE flag"). When a device is unregistered, rt6_uncached_list_flush_dev() unlinks uncached routes tied to the device from rt6_uncached_list. Previously, they were moved to another list with list_move() (__list_del_entry() + list_add()), and since commit 98aa546af5e4 ("inet: remove (struct uncached_list)->quarantine"), the routes are just unlinked with list_del_init(). If list_del_init() runs concurrently, list_empty() evaluates to true; ip6_route_output_flags() calls dst_hold_safe() incorrectly and ip6_rt_put_flags() skips ip6_rt_put(), leaking dst, and thus dev tied via rt->from as well. The same race is partially fixed by commit 9a6f0c4d5796 ("dst: fix races in rt6_uncached_list_del() and rt_del_uncached_list()"). Let's check rt6->dst.rt_uncached_list instead. Note that IPv4 does not have the same issue. Fixes: 98aa546af5e4 ("inet: remove (struct uncached_list)->quarantine") Signed-off-by: Kuniyuki Iwashima Reviewed-by: Hangbin Liu Reviewed-by: Xuanqiang Luo Reviewed-by: Ido Schimmel Reviewed-by: Eric Dumazet Link: https://patch.msgid.link/20260920191558.2990636-1-kuniyu@google.com Signed-off-by: Jakub Kicinski --- include/net/ip6_route.h | 4 ++-- net/ipv6/route.c | 7 ++++--- 2 files changed, 6 insertions(+), 5 deletions(-) diff --git a/include/net/ip6_route.h b/include/net/ip6_route.h index b9e8d2b759e9..0f9b7a260d25 100644 --- a/include/net/ip6_route.h +++ b/include/net/ip6_route.h @@ -101,12 +101,12 @@ static inline struct dst_entry *ip6_route_output(struct net *net, } /* Only conditionally release dst if flags indicates - * !RT6_LOOKUP_F_DST_NOREF or dst is in uncached_list. + * !RT6_LOOKUP_F_DST_NOREF or dst is uncached. */ static inline void ip6_rt_put_flags(struct rt6_info *rt, int flags) { if (!(flags & RT6_LOOKUP_F_DST_NOREF) || - !list_empty(&rt->dst.rt_uncached)) + rt->dst.rt_uncached_list) ip6_rt_put(rt); } diff --git a/net/ipv6/route.c b/net/ipv6/route.c index 884d9ab0d50d..153ce16628c1 100644 --- a/net/ipv6/route.c +++ b/net/ipv6/route.c @@ -139,6 +139,7 @@ void rt6_uncached_list_add(struct rt6_info *rt) { struct uncached_list *ul = raw_cpu_ptr(&rt6_uncached_list); + /* Set once and never cleared: non-NULL marks an uncached route. */ rt->dst.rt_uncached_list = ul; spin_lock_bh(&ul->lock); @@ -2726,8 +2727,8 @@ struct dst_entry *ip6_route_output_flags(struct net *net, rcu_read_lock(); dst = ip6_route_output_flags_noref(net, sk, fl6, flags); rt6 = dst_rt6_info(dst); - /* For dst cached in uncached_list, refcnt is already taken. */ - if (list_empty(&rt6->dst.rt_uncached) && !dst_hold_safe(dst)) { + /* For an uncached dst, refcnt is already taken. */ + if (!rt6->dst.rt_uncached_list && !dst_hold_safe(dst)) { dst = &net->ipv6.ip6_null_entry->dst; dst_hold(dst); } @@ -2836,7 +2837,7 @@ INDIRECT_CALLABLE_SCOPE struct dst_entry *ip6_dst_check(struct dst_entry *dst, from = rcu_dereference(rt->from); if (from && (rt->rt6i_flags & RTF_PCPU || - unlikely(!list_empty(&rt->dst.rt_uncached)))) + unlikely(rt->dst.rt_uncached_list))) dst_ret = rt6_dst_from_check(rt, from, cookie); else dst_ret = rt6_check(rt, from, cookie); From 5fd0783b99d4af98f65cd58b56ec203d1d426104 Mon Sep 17 00:00:00 2001 From: Shardul Bankar Date: Thu, 17 Sep 2026 14:55:32 +0530 Subject: [PATCH 462/592] udp: relocate a connected socket in the 4-tuple hash table on re-connect A connected UDP socket that connects again to a different peer is not re-filed in the 4-tuple hash table: sk binds to 127.0.0.1:21001 sk connects to 127.0.0.2:20001 // filed under hash(sk, peer1) sk connects to 127.0.0.3:20002 // still filed under hash(sk, peer1) packet from 127.0.0.3:20002 // hash(sk, peer2) misses, so the // lookup falls back to scoring the // hash2 chain for this address // and port udp_lib_hash4() returns early when the socket is already hashed, assuming ->rehash() relocates it. ->rehash() runs from __ip{4,6}_datagram_connect() only while the receive address is unset, which a second connect never is: the first connect assigns it, whether the socket was bound to a specific address or to the wildcard. commit 644f9108f3a5 ("udp: Make rehash4 independent in udp_lib_rehash()") added that early return and named connect(AF_UNSPEC) as the way around it. That workaround does not help a socket with both SOCK_BINDADDR_LOCK and SOCK_BINDPORT_LOCK set, because __udp_disconnect() skips ->rehash() for the first and ->unhash() for the second. Delivery is correct either way. Relocate the socket when the hash it is filed under differs from the one requested, which is what commit 78c91ae2c6de ("ipv4/udp: Add 4-tuple hash for connected socket") did before the early return became unconditional. It is done here under hslot->lock, which that version did not take, to match udp_lib_rehash() and udp_lib_unhash(). hslot2 is unchanged, so hash4_cnt needs no adjustment, as in udp_lib_rehash(). A first connect is unaffected, and IPv6 shares the code. With 500 sockets on the port, a re-connected socket measured 522,553 pps without this change and 2,055,078 with it. The UDP side was noted as remaining work in [1]. Link: https://lore.kernel.org/netdev/apnHqmYZQ4yzOP4N@v4bel/ [1] Fixes: 644f9108f3a5 ("udp: Make rehash4 independent in udp_lib_rehash()") Assisted-by: LLM Signed-off-by: Shardul Bankar Reviewed-by: Kuniyuki Iwashima Link: https://patch.msgid.link/20260917-udp_hash4_fix_v1-v1-1-718891af0d7a@mpiricsoftware.com Signed-off-by: Paolo Abeni --- net/ipv4/udp.c | 21 +++++++++++++++------ 1 file changed, 15 insertions(+), 6 deletions(-) diff --git a/net/ipv4/udp.c b/net/ipv4/udp.c index bb8cfc62cb00..0fa3cdbdcc21 100644 --- a/net/ipv4/udp.c +++ b/net/ipv4/udp.c @@ -617,14 +617,23 @@ void udp_lib_hash4(struct sock *sk, u16 hash) struct net *net = sock_net(sk); struct udp_table *udptable; - /* Connected udp socket can re-connect to another remote address, which - * will be handled by rehash. Thus no need to redo hash4 here. - */ - if (udp_hashed4(sk)) - return; - udptable = net->ipv4.udp_table; hslot = udp_hashslot(udptable, net, udp_sk(sk)->udp_port_hash); + + /* A connected socket can re-connect to another address. rehash() + * relocates it, but only runs when the local address changes, so a + * socket bound to a specific address would stay filed under the + * previous peer's hash. Move it here. + */ + if (udp_hashed4(sk)) { + if (udp_sk(sk)->udp_lrpa_hash != hash) { + spin_lock_bh(&hslot->lock); + udp_rehash4(udptable, sk, hash); + spin_unlock_bh(&hslot->lock); + } + return; + } + hslot2 = udp_hashslot2(udptable, udp_sk(sk)->udp_portaddr_hash); hslot4 = udp_hashslot4(udptable, hash); udp_sk(sk)->udp_lrpa_hash = hash; From 9e95b1a94c9c49b4ba722251bbca2759b9c51737 Mon Sep 17 00:00:00 2001 From: Shardul Bankar Date: Thu, 17 Sep 2026 14:55:33 +0530 Subject: [PATCH 463/592] udp: remove a disconnected socket from the 4-tuple hash table A UDP socket bound to a specific address and port keeps its entry in the 4-tuple hash table after it is disconnected: sk binds to 127.0.0.1:21001 sk connects to 127.0.0.2:20001 // filed in the 4-tuple table sk disconnects, connect(AF_UNSPEC) // still filed, peer now 0.0.0.0:0 __udp_disconnect() takes a socket out of that table only as a side effect of ->rehash() or ->unhash(), and it skips ->rehash() when SOCK_BINDADDR_LOCK is set and ->unhash() when SOCK_BINDPORT_LOCK is set. commit 6996a2d2d0a6 ("udp: Unhash auto-bound connected sk from 4-tuple hash table when disconnected.") fixed the same end state for a wildcard-bound socket, by a path this one does not take. The entry is counted whether or not anything hits it. hash4_cnt on the hash2 slot stays raised for as long as the socket lives, so udp_has_hash4() keeps sending every packet for that address and port through the 4-tuple lookup first. On IPv6 it can also be hit. __udp_disconnect() does not clear sk_v6_daddr, so udp_v6_rehash() files the entry under the peer the socket was connected to with a zero dport, and inet6_match() compares that same field: a datagram from the former peer with a zero source port matches, and source port zero is accepted on receive. On IPv4 the peer is cleared, so a match would need a zero source address as well, which the routing layer rejects as martian. The stale sk_v6_daddr is a separate defect, not addressed here; removing the entry closes this path either way. The entry can also be relocated. __udp_disconnect() clears sk_bound_dev_if, so a subsequent SO_BINDTODEVICE calls ->rehash(), and because the receive address is still specific udp_lib_rehash() moves the entry instead of removing it, into the bucket that (rcv_saddr, num, 0, 0) hashes to -- a pure function of the address and port, so every socket reaching this state on one address and port collects in one bucket. The bucket cannot be chosen from outside, as udp_ehashfn() is seeded with a per-boot secret. This last one became reachable only with commit 644f9108f3a5 ("udp: Make rehash4 independent in udp_lib_rehash()"), which moved the hash4 handling out of a branch a disconnected socket does not take; the stale entry itself dates from the commit in Fixes. Take the socket out of the table before __udp_disconnect() runs, while it still matches how it was filed. This also reaches the wildcard case ahead of udp_lib_rehash()'s udp_unhash4() branch, leaving that branch unreachable from udp_disconnect(); removing it belongs in net-next. udp_disconnect() and udp_abort() are the only UDP entries into __udp_disconnect(), which is shared with raw, ping and l2tp sockets that are not struct udp_sock: ping_prot.obj_size is sizeof(struct inet_sock), so udp_hashed4() on one would read past the allocation. Fixes: 78c91ae2c6de ("ipv4/udp: Add 4-tuple hash for connected socket") Assisted-by: LLM Signed-off-by: Shardul Bankar Reviewed-by: Kuniyuki Iwashima Link: https://patch.msgid.link/20260917-udp_hash4_fix_v1-v1-2-718891af0d7a@mpiricsoftware.com Signed-off-by: Paolo Abeni --- net/ipv4/udp.c | 23 +++++++++++++++++++++++ 1 file changed, 23 insertions(+) diff --git a/net/ipv4/udp.c b/net/ipv4/udp.c index 0fa3cdbdcc21..b090bd1f59e8 100644 --- a/net/ipv4/udp.c +++ b/net/ipv4/udp.c @@ -2206,9 +2206,31 @@ int __udp_disconnect(struct sock *sk, int flags) } EXPORT_SYMBOL(__udp_disconnect); +/* __udp_disconnect() takes a socket out of the 4-tuple hash table only via + * ->rehash() or ->unhash(), and neither runs for a socket bound to a + * specific address and port. Remove it here, before its peer is cleared. + */ +static void udp_unhash4_on_disconnect(struct sock *sk) +{ + struct net *net = sock_net(sk); + struct udp_table *udptable; + struct udp_hslot *hslot; + + if (!udp_hashed4(sk)) + return; + + udptable = net->ipv4.udp_table; + hslot = udp_hashslot(udptable, net, udp_sk(sk)->udp_port_hash); + + spin_lock_bh(&hslot->lock); + udp_unhash4(udptable, sk); + spin_unlock_bh(&hslot->lock); +} + int udp_disconnect(struct sock *sk, int flags) { lock_sock(sk); + udp_unhash4_on_disconnect(sk); __udp_disconnect(sk, flags); release_sock(sk); return 0; @@ -3140,6 +3162,7 @@ int udp_abort(struct sock *sk, int err) sk->sk_err = err; sk_error_report(sk); + udp_unhash4_on_disconnect(sk); __udp_disconnect(sk, 0); out: From a92e1a412c53dc0d9ad639e7abf8b3fc70a5b6ad Mon Sep 17 00:00:00 2001 From: Myeonghun Pak Date: Thu, 17 Sep 2026 14:33:36 -0400 Subject: [PATCH 464/592] tg3: clean up PHYLIB resources on probe failure tg3_get_invariants() can register an MDIO bus and connect a PHY for USE_PHYLIB devices. If tg3_init_one() later fails, its common error path releases the mappings and netdev without undoing those PHYLIB resources. Disconnect the PHY and unregister the MDIO bus before the remaining teardown. Guard PHY cleanup with USE_PHYLIB to match tg3_phy_init(), and call tg3_mdio_fini() unconditionally to match tg3_mdio_init(). The existing IS_CONNECTED and MDIOBUS_INITED flags make both helpers safe when initialization only completed partially. This issue was identified during our ongoing static-analysis research while reviewing kernel code. Fixes: 158d7abdae85 ("tg3: Add mdio bus registration") Assisted-by: OpenAI:GPT-5.6 Co-developed-by: Ijae Kim Signed-off-by: Ijae Kim Signed-off-by: Myeonghun Pak Link: https://patch.msgid.link/20260917183336.36239-1-mhun512@gmail.com Signed-off-by: Paolo Abeni --- drivers/net/ethernet/broadcom/tg3.c | 4 ++++ 1 file changed, 4 insertions(+) diff --git a/drivers/net/ethernet/broadcom/tg3.c b/drivers/net/ethernet/broadcom/tg3.c index 73a4b569b03e..caa7a6caa6e2 100644 --- a/drivers/net/ethernet/broadcom/tg3.c +++ b/drivers/net/ethernet/broadcom/tg3.c @@ -18047,6 +18047,10 @@ static int tg3_init_one(struct pci_dev *pdev, return 0; err_out_apeunmap: + if (tg3_flag(tp, USE_PHYLIB)) + tg3_phy_fini(tp); + tg3_mdio_fini(tp); + if (tp->aperegs) { iounmap(tp->aperegs); tp->aperegs = NULL; From a644f09b2090ad22a13fbcf9d141084f573108ef Mon Sep 17 00:00:00 2001 From: Wentao Liang Date: Thu, 17 Sep 2026 11:01:35 +0000 Subject: [PATCH 465/592] fsl/fman: Fix clk reference leak in read_dts_node() of_clk_get() returns a clock with its reference count incremented, but read_dts_node() only uses it to read the rate and never calls clk_put(). The clock is not stored anywhere, so the reference cannot be released later either. Release the clock once its rate has been read, which also covers the error path taken when the rate is zero. Fixes: 414fd46e7762 ("fsl/fman: Add FMan support") Cc: stable@vger.kernel.org Signed-off-by: Wentao Liang Reviewed-by: Simon Horman Link: https://patch.msgid.link/20260917110135.2148068-1-vulab@iscas.ac.cn Signed-off-by: Paolo Abeni --- drivers/net/ethernet/freescale/fman/fman.c | 1 + 1 file changed, 1 insertion(+) diff --git a/drivers/net/ethernet/freescale/fman/fman.c b/drivers/net/ethernet/freescale/fman/fman.c index 299bab043175..46cc28895e56 100644 --- a/drivers/net/ethernet/freescale/fman/fman.c +++ b/drivers/net/ethernet/freescale/fman/fman.c @@ -2734,6 +2734,7 @@ static struct fman *read_dts_node(struct platform_device *of_dev) } clk_rate = clk_get_rate(clk); + clk_put(clk); if (!clk_rate) { err = -EINVAL; dev_err(&of_dev->dev, "%s: Failed to determine FM%d clock rate\n", From 2d14720beb58870b15a52b236c3ab0be0e06e915 Mon Sep 17 00:00:00 2001 From: Muhammad Bilal Date: Sun, 20 Sep 2026 00:19:37 +0500 Subject: [PATCH 466/592] net: spacemit: clear TX descriptor on fragment mapping failure emac_tx_mem_map() writes TX_DESC_0_OWN into the ring descriptor for every slot beyond old_head as soon as that slot's memset()'d local copy is committed with "*tx_desc_addr = tx_desc", i.e. before the buffers for that slot have necessarily all been mapped successfully. If emac_tx_map_frag() then fails on a later fragment, the err_free_skb path calls emac_free_tx_buf() to unmap and drop the skb, but leaves the already-written descriptor memory untouched, and tx_ring->head is never advanced past old_head (the "tx_ring->head = head" store is skipped by the goto). So a slot between old_head and the rolled-back head can be left with TX_DESC_0_OWN set and buffer_addr_{1,2} pointing at DMA mappings that emac_free_tx_buf() just tore down, while software considers that slot free again. The next successful emac_tx_mem_map() call only rebuilds old_head itself; if the DMA engine auto-advances into the following descriptor once it finishes old_head's packet, it will fetch that stale, already-unmapped address. emac_tx_clean_desc() already treats emac_free_tx_buf() and clearing the descriptor as a pair when reclaiming completed descriptors; do the same in the mapping failure path. Fixes: bfec6d7f2001 ("net: spacemit: Add K1 Ethernet MAC") Signed-off-by: Muhammad Bilal Reviewed-by: Vivian Wang Reviewed-by: Troy Mitchell Link: https://patch.msgid.link/20260919191937.271202-1-meatuni001@gmail.com Signed-off-by: Paolo Abeni --- drivers/net/ethernet/spacemit/k1_emac.c | 3 +++ 1 file changed, 3 insertions(+) diff --git a/drivers/net/ethernet/spacemit/k1_emac.c b/drivers/net/ethernet/spacemit/k1_emac.c index f7f16397a2c2..d641ac26a1e8 100644 --- a/drivers/net/ethernet/spacemit/k1_emac.c +++ b/drivers/net/ethernet/spacemit/k1_emac.c @@ -803,6 +803,9 @@ static void emac_tx_mem_map(struct emac_priv *priv, struct sk_buff *skb) while (i != head) { emac_free_tx_buf(priv, i); + tx_desc_addr = &((struct emac_desc *)tx_ring->desc_addr)[i]; + memset(tx_desc_addr, 0, sizeof(*tx_desc_addr)); + if (++i == tx_ring->total_cnt) i = 0; } From ac4334522e4ba4a3b6710dd5d4cc98092824b8ca Mon Sep 17 00:00:00 2001 From: bui duc phuc Date: Fri, 18 Sep 2026 11:28:04 +0700 Subject: [PATCH 467/592] net: ethernet: ti: netcp: fix pm_runtime usage counter leak on error pm_runtime_get_sync() leaves the runtime PM usage counter incremented even when it fails, but the error path in netcp_probe() does not call pm_runtime_put_noidle() to balance it, leaking a reference each time resume fails. Use pm_runtime_resume_and_get() instead, which automatically drops the usage counter on failure, fixing the leak. Fixes: 84640e27f230 ("net: netcp: Add Keystone NetCP core ethernet driver") Signed-off-by: bui duc phuc Reviewed-by: Simon Horman Link: https://patch.msgid.link/20260918042804.13101-1-phucduc.bui@gmail.com Signed-off-by: Paolo Abeni --- drivers/net/ethernet/ti/netcp_core.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/drivers/net/ethernet/ti/netcp_core.c b/drivers/net/ethernet/ti/netcp_core.c index eb8fc2ed05f4..4f9e20468bbb 100644 --- a/drivers/net/ethernet/ti/netcp_core.c +++ b/drivers/net/ethernet/ti/netcp_core.c @@ -2225,7 +2225,7 @@ static int netcp_probe(struct platform_device *pdev) return -ENOMEM; pm_runtime_enable(&pdev->dev); - ret = pm_runtime_get_sync(&pdev->dev); + ret = pm_runtime_resume_and_get(&pdev->dev); if (ret < 0) { dev_err(dev, "Failed to enable NETCP power-domain\n"); pm_runtime_disable(&pdev->dev); From 999e8295bc41d6ce45b8e54f88150efa96f3f01e Mon Sep 17 00:00:00 2001 From: Wentao Liang Date: Thu, 17 Sep 2026 11:08:28 +0000 Subject: [PATCH 468/592] net: hisilicon: hns_dsaf_mac: fix mdio device leak in hns_mac_register_phy() hns_dsaf_find_platform_device() returns the mdio platform device with its reference count incremented. hns_mac_register_phy() never drops that reference, so the mdio device can not be released. Release the reference on both the deferred probe and the normal path. Fixes: 1d1afa2ebf82 ("net: hns: register phy device in each mac initial sequence") Cc: stable@vger.kernel.org Signed-off-by: Wentao Liang Reviewed-by: Simon Horman Link: https://patch.msgid.link/20260917110828.2148390-1-vulab@iscas.ac.cn Signed-off-by: Paolo Abeni --- drivers/net/ethernet/hisilicon/hns/hns_dsaf_mac.c | 3 +++ 1 file changed, 3 insertions(+) diff --git a/drivers/net/ethernet/hisilicon/hns/hns_dsaf_mac.c b/drivers/net/ethernet/hisilicon/hns/hns_dsaf_mac.c index bc6b269be299..f1cb6d56e4b9 100644 --- a/drivers/net/ethernet/hisilicon/hns/hns_dsaf_mac.c +++ b/drivers/net/ethernet/hisilicon/hns/hns_dsaf_mac.c @@ -793,6 +793,7 @@ static int hns_mac_register_phy(struct hns_mac_cb *mac_cb) dev_err(mac_cb->dev, "mac%d mdio is NULL, dsaf will probe again later\n", mac_cb->mac_id); + put_device(&pdev->dev); return -EPROBE_DEFER; } @@ -801,6 +802,8 @@ static int hns_mac_register_phy(struct hns_mac_cb *mac_cb) dev_dbg(mac_cb->dev, "mac%d register phy addr:%d\n", mac_cb->mac_id, addr); + put_device(&pdev->dev); + return rc; } From 23d42b9a3bcd55b17d3b371544fedc708c2397e9 Mon Sep 17 00:00:00 2001 From: =?UTF-8?q?Th=C3=A9o=20Lebrun?= Date: Fri, 18 Sep 2026 21:53:52 +0200 Subject: [PATCH 469/592] net: macb: fix dma_alloc_coherent() leak on macb_alloc() error paths MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit Fix 3 leaks in macb_alloc() error paths: - Tx buffer allocated but crossing a 4G boundary: Tx leaked. - Rx buffer allocation fails: Tx leaked. - Rx buffer allocated but crossing a 4G boundary: Tx & Rx leaked. This is because our error handling calls macb_free(bp) which in turn frees the buffers stored in bp->queues[0], but nothing has been stored in there. Fix by storing allocated buffers into bp->queues[0] ASAP. Fixes: 78d901897b3c ("net: macb: single dma_alloc_coherent() for DMA descriptors") Cc: stable@vger.kernel.org Signed-off-by: Théo Lebrun Reviewed-by: Nicolai Buchwitz Link: https://patch.msgid.link/20260918-macb-alloc-leak-v1-1-aba9a3d4f6e3@bootlin.com Signed-off-by: Paolo Abeni --- drivers/net/ethernet/cadence/macb_main.c | 14 ++++++++++++-- 1 file changed, 12 insertions(+), 2 deletions(-) diff --git a/drivers/net/ethernet/cadence/macb_main.c b/drivers/net/ethernet/cadence/macb_main.c index b8234ac4b602..8e5c034dc3a4 100644 --- a/drivers/net/ethernet/cadence/macb_main.c +++ b/drivers/net/ethernet/cadence/macb_main.c @@ -2749,14 +2749,24 @@ static int macb_alloc(struct macb *bp) size = bp->num_queues * macb_tx_ring_size_per_queue(bp); tx = dma_alloc_coherent(dev, size, &tx_dma, GFP_KERNEL); - if (!tx || upper_32_bits(tx_dma) != upper_32_bits(tx_dma + size - 1)) + if (!tx) + goto out_err; + /* Record the buffer so that the error path frees it. */ + bp->queues[0].tx_ring = tx; + bp->queues[0].tx_ring_dma = tx_dma; + if (upper_32_bits(tx_dma) != upper_32_bits(tx_dma + size - 1)) goto out_err; netdev_dbg(bp->netdev, "Allocated %zu bytes for %u TX rings at %08lx (mapped %p)\n", size, bp->num_queues, (unsigned long)tx_dma, tx); size = bp->num_queues * macb_rx_ring_size_per_queue(bp); rx = dma_alloc_coherent(dev, size, &rx_dma, GFP_KERNEL); - if (!rx || upper_32_bits(rx_dma) != upper_32_bits(rx_dma + size - 1)) + if (!rx) + goto out_err; + /* Record the buffer so that the error path frees it. */ + bp->queues[0].rx_ring = rx; + bp->queues[0].rx_ring_dma = rx_dma; + if (upper_32_bits(rx_dma) != upper_32_bits(rx_dma + size - 1)) goto out_err; netdev_dbg(bp->netdev, "Allocated %zu bytes for %u RX rings at %08lx (mapped %p)\n", size, bp->num_queues, (unsigned long)rx_dma, rx); From 261e8a37ecbaf462cdf9c336d2b2f5056088401a Mon Sep 17 00:00:00 2001 From: Jakub Kicinski Date: Fri, 18 Sep 2026 15:29:48 -0700 Subject: [PATCH 470/592] genetlink: report the real command id for dump-only ops in policy dumps The op-to-policy map a CTRL_CMD_GETPOLICY dump returns is the only way for userspace to find out which policy index belongs to which command. ctrl_dumppolicy_put_op() tags the nest with doit->cmd, but an op which only has a dumpit has no doit and every path which fills the split ops in zeroes it out, so those entries all claim to be command 0. nlctrl's own CTRL_CMD_GETPOLICY and NETDEV_CMD_QSTATS_GET are both in that group: [{'family-id': 16, 'op-policy': {'do': 0, 'dump': 0, 'op-id': 3}}, {'family-id': 16, 'op-policy': {'dump': 1, 'op-id': 0}}, ctrl_fill_info() gets this right - it uses the iterator's cmd for CTRL_ATTR_OP_ID - so the two introspection interfaces of the same family contradict each other today. Pass the command in rather than reconstructing it from doit->cmd | dumpit->cmd inside the helper, both callers already have it. Fixes: 26588edbef60 ("genetlink: support split policies in ctrl_dumppolicy_put_op()") Signed-off-by: Jakub Kicinski Link: https://patch.msgid.link/20260918222949.4190284-1-kuba@kernel.org Signed-off-by: Paolo Abeni --- net/netlink/genetlink.c | 8 +++++--- 1 file changed, 5 insertions(+), 3 deletions(-) diff --git a/net/netlink/genetlink.c b/net/netlink/genetlink.c index 41d37442f186..5cc1037d4917 100644 --- a/net/netlink/genetlink.c +++ b/net/netlink/genetlink.c @@ -1656,7 +1656,7 @@ static void *ctrl_dumppolicy_prep(struct sk_buff *skb, } static int ctrl_dumppolicy_put_op(struct sk_buff *skb, - struct netlink_callback *cb, + struct netlink_callback *cb, u32 cmd, struct genl_split_ops *doit, struct genl_split_ops *dumpit) { @@ -1677,7 +1677,7 @@ static int ctrl_dumppolicy_put_op(struct sk_buff *skb, if (!nest_pol) goto err; - nest_op = nla_nest_start(skb, doit->cmd); + nest_op = nla_nest_start(skb, cmd); if (!nest_op) goto err; @@ -1721,7 +1721,8 @@ static int ctrl_dumppolicy(struct sk_buff *skb, struct netlink_callback *cb) &doit, &dumpit))) return -ENOENT; - if (ctrl_dumppolicy_put_op(skb, cb, &doit, &dumpit)) + if (ctrl_dumppolicy_put_op(skb, cb, ctx->op, + &doit, &dumpit)) return skb->len; /* done with the per-op policy index list */ @@ -1730,6 +1731,7 @@ static int ctrl_dumppolicy(struct sk_buff *skb, struct netlink_callback *cb) while (ctx->dump_map) { if (ctrl_dumppolicy_put_op(skb, cb, + ctx->op_iter->cmd, &ctx->op_iter->doit, &ctx->op_iter->dumpit)) return skb->len; From a87529034b9c3c3f3bb71c33e2d6d94658e06383 Mon Sep 17 00:00:00 2001 From: Jakub Kicinski Date: Fri, 18 Sep 2026 15:29:49 -0700 Subject: [PATCH 471/592] selftests: net: nl_nlctrl: check the op ids in the policy map Validate that the op map in the policy dump is correct. Signed-off-by: Jakub Kicinski Link: https://patch.msgid.link/20260918222949.4190284-2-kuba@kernel.org Signed-off-by: Paolo Abeni --- tools/testing/selftests/net/nl_nlctrl.py | 116 +++++++++++++++++++---- 1 file changed, 99 insertions(+), 17 deletions(-) diff --git a/tools/testing/selftests/net/nl_nlctrl.py b/tools/testing/selftests/net/nl_nlctrl.py index fe1f66dc9435..237b3d273260 100755 --- a/tools/testing/selftests/net/nl_nlctrl.py +++ b/tools/testing/selftests/net/nl_nlctrl.py @@ -9,40 +9,86 @@ from lib.py import ksft_run, ksft_exit from lib.py import ksft_eq, ksft_ge, ksft_true, ksft_in, ksft_not_in from lib.py import NetdevFamily, EthtoolFamily, NlctrlFamily +# Families we can expect to always be around, and which between them +# cover ops with a do, with a dump, and with both. +FAMILIES = ('nlctrl', 'netdev') -def getfamily_do(ctrl) -> None: - """Query a single family by name and validate its ops.""" - fam = ctrl.getfamily({'family-name': 'netdev'}) - ksft_eq(fam['family-name'], 'netdev') + +def _get_ops(ctrl, name): + """Get the ops of a family, keyed by command id.""" + fam = ctrl.getfamily({'family-name': name}) + ksft_eq(fam['family-name'], name) ksft_true(fam['family-id'] > 0) # The format of ops is quite odd, [{$idx: {"id"...}}, {$idx: {"id"...}}] # Discard the indices and re-key by command id. ops_by_id = {v['id']: v for op in fam['ops'] for v in op.values()} - ksft_eq(len(ops_by_id), len(fam['ops'])) + ksft_eq(len(ops_by_id), len(fam['ops']), + comment=f"{name} lists a command twice") + return ops_by_id - # All ops should have a policy (either do or dump has one) - for op in ops_by_id.values(): - ksft_in('cmd-cap-haspol', op['flags'], - comment=f"op {op['id']} missing haspol") + +def _get_policy_map(ctrl, req): + """ + The policy map in the Netlink replies looks like this: + + [{'family-id': 16, 'op-policy': {'do': 0, 'dump': 0, 'op-id': 3}}, + {'family-id': 16, 'op-policy': {'dump': 1, 'op-id': 4}}, ...] + + Return the mapping: + + {3:{'do','dump'}, 4:{'dump'}} + + The policy itself is discarded here, only return which command has policy. + """ + pol_map = {} + for msg in ctrl.getpolicy(req, dump=True): + if 'op-policy' not in msg: + continue + modes = dict(msg['op-policy']) + cmd = modes.pop('op-id') + ksft_not_in(cmd, pol_map, comment=f"command {cmd} reported twice") + pol_map[cmd] = set(modes.keys()) + return pol_map + + +def getfamily_do(ctrl) -> None: + """Query single families by name and validate their ops.""" + ops = {name: _get_ops(ctrl, name) for name in FAMILIES} + + for name, ops_by_id in ops.items(): + for op in ops_by_id.values(): + # All ops in nlctrl and netdev have a policy + ksft_in('cmd-cap-haspol', op['flags'], + comment=f"{name} op {op['id']} missing haspol") + ksft_true(op['flags'] & {'cmd-cap-do', 'cmd-cap-dump'}, + comment=f"{name} op {op['id']} has no handler") + + # nlctrl getfamily (id 3) does both, getpolicy (id 10) is dump-only + ksft_in('cmd-cap-do', ops['nlctrl'][3]['flags']) + ksft_in('cmd-cap-dump', ops['nlctrl'][3]['flags']) + ksft_not_in('cmd-cap-do', ops['nlctrl'][10]['flags']) + ksft_in('cmd-cap-dump', ops['nlctrl'][10]['flags']) + + netdev = ops['netdev'] # dev-get (id 1) should support both do and dump - ksft_in('cmd-cap-do', ops_by_id[1]['flags']) - ksft_in('cmd-cap-dump', ops_by_id[1]['flags']) + ksft_in('cmd-cap-do', netdev[1]['flags']) + ksft_in('cmd-cap-dump', netdev[1]['flags']) # qstats-get (id 12) is dump-only - ksft_not_in('cmd-cap-do', ops_by_id[12]['flags']) - ksft_in('cmd-cap-dump', ops_by_id[12]['flags']) + ksft_not_in('cmd-cap-do', netdev[12]['flags']) + ksft_in('cmd-cap-dump', netdev[12]['flags']) # napi-set (id 14) is do-only and requires admin - ksft_in('cmd-cap-do', ops_by_id[14]['flags']) - ksft_not_in('cmd-cap-dump', ops_by_id[14]['flags']) - ksft_in('admin-perm', ops_by_id[14]['flags']) + ksft_in('cmd-cap-do', netdev[14]['flags']) + ksft_not_in('cmd-cap-dump', netdev[14]['flags']) + ksft_in('admin-perm', netdev[14]['flags']) # Notification-only commands (dev-add/del/change-ntf etc.) must # not appear in the ops list since they have no do/dump handlers. for ntf_id in [2, 3, 4, 6, 7, 8]: - ksft_not_in(ntf_id, ops_by_id, + ksft_not_in(ntf_id, netdev, comment=f"ntf-only cmd {ntf_id} should not be in ops") @@ -103,6 +149,41 @@ def getpolicy_dump(_ctrl) -> None: comment="linkinfo-set should not have a dump policy") +def getpolicy_op_map(ctrl) -> None: + """Check the op-to-policy map consistency. Each op with 'haspol' flag + has to have a policy. The policy back-references must name only + real ops that exist, have given modes (do vs dump) and have 'haspol'. + """ + for name in FAMILIES: + ops_by_id = _get_ops(ctrl, name) + haspol = {cmd for cmd, op in ops_by_id.items() + if 'cmd-cap-haspol' in op['flags']} + + pol_map = _get_policy_map(ctrl, {'family-name': name}) + ksft_eq(set(pol_map), haspol, + comment=f"{name} policy map does not match the op list") + + # Walk the op list rather than the map, the map may be missing + # the very op we are after. Asking for a command the family does + # not have is an error, so it must not come from the map either. + for cmd in sorted(haspol): + modes = pol_map.get(cmd, set()) + + # The kernel only reports a mode the op actually has. + if 'do' in modes: + ksft_in('cmd-cap-do', ops_by_id[cmd]['flags'], + comment=f"{name} cmd {cmd} has no do") + if 'dump' in modes: + ksft_in('cmd-cap-dump', ops_by_id[cmd]['flags'], + comment=f"{name} cmd {cmd} has no dump") + + # Asking for one op builds the map in a different place in + # the kernel, it has to report what the full dump did. + single = _get_policy_map(ctrl, {'family-name': name, 'op': cmd}) + ksft_eq(single, {cmd: modes}, + comment=f"{name} cmd {cmd} policy differs from the dump") + + def getpolicy_by_op(_ctrl) -> None: """Query policy for specific ops, check attr names are resolved.""" ndev = NetdevFamily() @@ -122,6 +203,7 @@ def main() -> None: ksft_run([getfamily_do, getfamily_dump, getpolicy_dump, + getpolicy_op_map, getpolicy_by_op], args=(ctrl, )) ksft_exit() From 9892d71cf0ce3ff3d4fed2d9a3968fd4feb1c918 Mon Sep 17 00:00:00 2001 From: Coia Prant Date: Sun, 20 Sep 2026 01:20:21 +0800 Subject: [PATCH 472/592] net: pcs: xpcs: fix clock reference leak on xpcs_init_clks failure xpcs_init_clks() takes references with clk_bulk_get_optional() and then enables them with clk_bulk_prepare_enable(). If the enable step fails, the function returns without dropping the references. xpcs_create() handles the failure through out_free_data, which calls xpcs_free_data() but never xpcs_clear_clks(), so the clk references are leaked. Add the missing clk_bulk_put() on the enable failure path. The prepare/enable side is already rolled back by clk_bulk_prepare_enable() itself. Fixes: f6bb3e9d98c2 ("net: pcs: xpcs: Add Synopsys DW xPCS platform device driver") Signed-off-by: Coia Prant Reviewed-by: Simon Horman Link: https://patch.msgid.link/20260919172021.2336748-1-coiaprant@gmail.com Signed-off-by: Paolo Abeni --- drivers/net/pcs/pcs-xpcs.c | 4 +++- 1 file changed, 3 insertions(+), 1 deletion(-) diff --git a/drivers/net/pcs/pcs-xpcs.c b/drivers/net/pcs/pcs-xpcs.c index 0337e2bcc012..b415b93d77c1 100644 --- a/drivers/net/pcs/pcs-xpcs.c +++ b/drivers/net/pcs/pcs-xpcs.c @@ -1545,8 +1545,10 @@ static int xpcs_init_clks(struct dw_xpcs *xpcs) return dev_err_probe(dev, ret, "Failed to get clocks\n"); ret = clk_bulk_prepare_enable(DW_XPCS_NUM_CLKS, xpcs->clks); - if (ret) + if (ret) { + clk_bulk_put(DW_XPCS_NUM_CLKS, xpcs->clks); return dev_err_probe(dev, ret, "Failed to enable clocks\n"); + } return 0; } From 17741334d00bf5ebd37f8c1c36bc9c146a351deb Mon Sep 17 00:00:00 2001 From: Wentao Liang Date: Thu, 17 Sep 2026 11:58:11 +0000 Subject: [PATCH 473/592] net: usb: lan78xx: Fix URB reference leak in lan78xx_submit_deferred_urbs() usb_get_from_anchor() hands over a reference to the URB, which the caller must release. lan78xx_submit_deferred_urbs() never does, so every deferred Tx URB keeps an extra reference: the counter grows on each suspend/resume cycle and the URBs are never freed when the buffers are released. Drop the reference after submitting, and on the path that drops the packet instead of submitting it. Fixes: 5f4cc6e25148 ("lan78xx: Fix race conditions in suspend/resume handling") Cc: stable@vger.kernel.org Signed-off-by: Wentao Liang Link: https://patch.msgid.link/20260917115811.2150119-1-vulab@iscas.ac.cn Signed-off-by: Paolo Abeni --- drivers/net/usb/lan78xx.c | 2 ++ 1 file changed, 2 insertions(+) diff --git a/drivers/net/usb/lan78xx.c b/drivers/net/usb/lan78xx.c index cb782d81d84f..5655941f1478 100644 --- a/drivers/net/usb/lan78xx.c +++ b/drivers/net/usb/lan78xx.c @@ -5239,10 +5239,12 @@ static bool lan78xx_submit_deferred_urbs(struct lan78xx_net *dev) !netif_carrier_ok(dev->net) || pipe_halted) { lan78xx_release_tx_buf(dev, skb); + usb_put_urb(urb); continue; } ret = usb_submit_urb(urb, GFP_ATOMIC); + usb_put_urb(urb); if (ret == 0) { netif_trans_update(dev->net); From bd3a19800dd1ba1ba9a4c307d9dfe50eac443c01 Mon Sep 17 00:00:00 2001 From: Tingmao Wang Date: Sun, 8 Feb 2026 23:54:48 +0000 Subject: [PATCH 474/592] landlock: Add counted_by in landlock_domain MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit For a domain, this array stores the access masks for each layer (of which there are num_layers of them). This annotation serves as useful documentation. Signed-off-by: Tingmao Wang Reviewed-by: Günther Noack Link: https://patch.msgid.link/20260208235449.1124354-1-m@maowtm.org [mic: Rebase on the ruleset/domain split, and update commit message] Signed-off-by: Mickaël Salaün --- security/landlock/domain.h | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-) diff --git a/security/landlock/domain.h b/security/landlock/domain.h index 5ce2f91488d5..caa3d19d2c43 100644 --- a/security/landlock/domain.h +++ b/security/landlock/domain.h @@ -243,7 +243,8 @@ struct landlock_domain { * overlapping access rights. These layers are set once * and never changed for the lifetime of the domain. */ - struct access_masks handled_masks[]; + struct access_masks + handled_masks[] __counted_by(num_layers); }; }; }; From e7e0a54300a896e731dffd3e2e8dae5631d6243d Mon Sep 17 00:00:00 2001 From: =?UTF-8?q?Micka=C3=ABl=20Sala=C3=BCn?= Date: Tue, 22 Sep 2026 15:26:14 +0200 Subject: [PATCH 475/592] landlock: Widen ruleset versions to 64 bits MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit Tracepoint consumers use a ruleset ID and version to identify the successful landlock_add_rule(2) call prefix used to create a domain. LANDLOCK_MAX_NUM_RULES bounds distinct stored rules, not successful calls: re-adding already-present rights for an object or port succeeds without increasing num_rules. Because every successful call increments the version, these calls can wrap the 32-bit counter and give different prefixes the same trace identity. Widen the counter and its trace fields to 64 bits so the counter cannot wrap in practice, while preserving the successful-call semantics. Saturating would alias all subsequent histories, while rejecting a call at the limit would change otherwise valid syscall behavior solely for trace metadata. Cc: Günther Noack Cc: Steven Rostedt Fixes: 63747c94774d ("landlock: Add landlock_add_rule_fs and landlock_add_rule_net tracepoints") Reviewed-by: Günther Noack Link: https://patch.msgid.link/20260922132615.1025945-1-mic@digikod.net Signed-off-by: Mickaël Salaün --- include/trace/events/landlock.h | 20 ++++++++++---------- security/landlock/ruleset.h | 5 +++-- 2 files changed, 13 insertions(+), 12 deletions(-) diff --git a/include/trace/events/landlock.h b/include/trace/events/landlock.h index 523ba5ea9870..3a43638c9bc2 100644 --- a/include/trace/events/landlock.h +++ b/include/trace/events/landlock.h @@ -344,7 +344,7 @@ TRACE_EVENT(landlock_create_ruleset, TP_STRUCT__entry( __field( u64, ruleset_id ) - __field( u32, ruleset_version ) + __field( u64, ruleset_version ) __field( access_mask_t, handled_fs ) __field( access_mask_t, handled_net ) __field( access_mask_t, scoped ) @@ -358,7 +358,7 @@ TRACE_EVENT(landlock_create_ruleset, __entry->scoped = ruleset->handled_masks.scope; ), - TP_printk("ruleset=%llx.%u handled_fs=%s handled_net=%s scoped=%s", + TP_printk("ruleset=%llx.%llu handled_fs=%s handled_net=%s scoped=%s", __entry->ruleset_id, __entry->ruleset_version, __print_flags(__entry->handled_fs, "|", _LANDLOCK_ACCESS_FS_NAMES), __print_flags(__entry->handled_net, "|", _LANDLOCK_ACCESS_NET_NAMES), @@ -384,7 +384,7 @@ TRACE_EVENT(landlock_free_ruleset, TP_STRUCT__entry( __field( u64, ruleset_id ) - __field( u32, ruleset_version ) + __field( u64, ruleset_version ) ), TP_fast_assign( @@ -392,7 +392,7 @@ TRACE_EVENT(landlock_free_ruleset, __entry->ruleset_version = ruleset->version; ), - TP_printk("ruleset=%llx.%u", + TP_printk("ruleset=%llx.%llu", __entry->ruleset_id, __entry->ruleset_version) ); @@ -423,7 +423,7 @@ TRACE_EVENT(landlock_add_rule_path_beneath, TP_STRUCT__entry( __field( u64, ruleset_id ) - __field( u32, ruleset_version ) + __field( u64, ruleset_version ) __field( access_mask_t, access_rights ) __field( dev_t, dev ) __field( ino_t, ino ) @@ -444,7 +444,7 @@ TRACE_EVENT(landlock_add_rule_path_beneath, __assign_str(pathname); ), - TP_printk("ruleset=%llx.%u access_rights=%s dev=%u:%u ino=%lu path=%s", + TP_printk("ruleset=%llx.%llu access_rights=%s dev=%u:%u ino=%lu path=%s", __entry->ruleset_id, __entry->ruleset_version, __print_flags(__entry->access_rights, "|", _LANDLOCK_ACCESS_FS_NAMES), MAJOR(__entry->dev), MINOR(__entry->dev), __entry->ino, @@ -477,7 +477,7 @@ TRACE_EVENT(landlock_add_rule_net_port, TP_STRUCT__entry( __field( u64, ruleset_id ) - __field( u32, ruleset_version ) + __field( u64, ruleset_version ) __field( access_mask_t, access_rights ) __field( u64, port ) ), @@ -490,7 +490,7 @@ TRACE_EVENT(landlock_add_rule_net_port, __entry->port = port; ), - TP_printk("ruleset=%llx.%u access_rights=%s port=%llu", + TP_printk("ruleset=%llx.%llu access_rights=%s port=%llu", __entry->ruleset_id, __entry->ruleset_version, __print_flags(__entry->access_rights, "|", _LANDLOCK_ACCESS_NET_NAMES), __entry->port) @@ -526,7 +526,7 @@ TRACE_EVENT(landlock_create_domain, __field( u64, domain_id ) __field( u64, parent_id ) __field( u64, ruleset_id ) - __field( u32, ruleset_version ) + __field( u64, ruleset_version ) ), TP_fast_assign( @@ -538,7 +538,7 @@ TRACE_EVENT(landlock_create_domain, __entry->ruleset_version = ruleset->version; ), - TP_printk("domain=%llx parent=%llx ruleset=%llx.%u", + TP_printk("domain=%llx parent=%llx ruleset=%llx.%llu", __entry->domain_id, __entry->parent_id, __entry->ruleset_id, __entry->ruleset_version) ); diff --git a/security/landlock/ruleset.h b/security/landlock/ruleset.h index b536fa0425b7..cf77f1806a95 100644 --- a/security/landlock/ruleset.h +++ b/security/landlock/ruleset.h @@ -171,9 +171,10 @@ struct landlock_ruleset { * @version: Counter incremented on each successful * landlock_add_rule(2), including when it only extends an existing * rule's access rights. Used by tracepoints to correlate a domain with - * the exact ruleset state it was created from. Protected by @lock. + * the exact successful rule history it was created from. Protected by + * @lock. */ - u32 version; + u64 version; /** * @id: Unique identifier for this ruleset, used for tracing. */ From ed6eec97b534979dcf28b40c389cee57bd6561d4 Mon Sep 17 00:00:00 2001 From: Emil Tsalapatis Date: Tue, 22 Sep 2026 17:20:18 +0000 Subject: [PATCH 476/592] bpf: Fix bounds check for skb-backed dynptrs The skb_pointer_if_linear() function checks whether a memory region of length len starting at offset off into the skb is in the linear area, and returns a pointer to the region if so. The check currently subtracts between skb_headlen and offset of the check, and since skb_headlen is unsigned the subtraction can underflow. This causes the bounds check to spuriously pass and generate an arbitrary pointer of the form *(skb->data + off). The only user of this helper is currently skb-backed BPF dynptr code. Returning the wrong pointer leads to the dynptr erroneously being backed with invalid memory. Ensure the subtraction cannot underflow, and fail the check if it would. Use u64 arithmetic to also prevent overflow when calculating (skb_headlen(skb) - off) since off is unsigned. Fixes: 6f5a630d7c57 ("bpf, net: Introduce skb_pointer_if_linear().") Reported-by: Nicholas Carlini Signed-off-by: Emil Tsalapatis Signed-off-by: Alexei Starovoitov Reviewed-by: Jiayuan Chen Link: https://patch.msgid.link/20260922172028.6269-2-emil@etsalapatis.com --- include/linux/skbuff.h | 5 ++++- 1 file changed, 4 insertions(+), 1 deletion(-) diff --git a/include/linux/skbuff.h b/include/linux/skbuff.h index 421f6fc45451..c8e21903074c 100644 --- a/include/linux/skbuff.h +++ b/include/linux/skbuff.h @@ -4372,7 +4372,10 @@ skb_header_pointer_careful(const struct sk_buff *skb, int offset, static inline void * __must_check skb_pointer_if_linear(const struct sk_buff *skb, int offset, int len) { - if (likely(skb_headlen(skb) - offset >= len)) + unsigned int uoffset = (unsigned int)offset; + + if (likely(uoffset <= skb_headlen(skb) && + (unsigned int)len <= skb_headlen(skb) - uoffset)) return skb->data + offset; return NULL; } From 4fd72eb9f1c939ce33bb714c6f745a125ac5f40c Mon Sep 17 00:00:00 2001 From: Emil Tsalapatis Date: Tue, 22 Sep 2026 17:20:19 +0000 Subject: [PATCH 477/592] selftests/bpf: Test dynptr slices past end of skb Add a selftest to ensure dynptr slices cannot include past the end of the linear area of an skb. Signed-off-by: Emil Tsalapatis Signed-off-by: Alexei Starovoitov Link: https://patch.msgid.link/20260922172028.6269-3-emil@etsalapatis.com --- .../testing/selftests/bpf/prog_tests/dynptr.c | 10 ++++++++++ .../selftests/bpf/progs/dynptr_success.c | 20 +++++++++++++++++++ 2 files changed, 30 insertions(+) diff --git a/tools/testing/selftests/bpf/prog_tests/dynptr.c b/tools/testing/selftests/bpf/prog_tests/dynptr.c index 5fda11590708..4396560365e8 100644 --- a/tools/testing/selftests/bpf/prog_tests/dynptr.c +++ b/tools/testing/selftests/bpf/prog_tests/dynptr.c @@ -9,6 +9,7 @@ enum test_setup_type { SETUP_SYSCALL_SLEEP, SETUP_SKB_PROG, + SETUP_SKB_PROG_NONLINEAR, SETUP_SKB_PROG_TP, SETUP_XDP_PROG, }; @@ -32,6 +33,7 @@ static struct { {"test_ringbuf", SETUP_SYSCALL_SLEEP}, {"test_skb_readonly", SETUP_SKB_PROG}, {"test_dynptr_skb_data", SETUP_SKB_PROG}, + {"test_dynptr_skb_slice_non_linear", SETUP_SKB_PROG_NONLINEAR}, {"test_dynptr_skb_meta_data", SETUP_SKB_PROG}, {"test_dynptr_skb_meta_flags", SETUP_SKB_PROG}, {"test_adjust", SETUP_SYSCALL_SLEEP}, @@ -94,7 +96,9 @@ static void verify_success(const char *prog_name, enum test_setup_type setup_typ bpf_link__destroy(link); break; case SETUP_SKB_PROG: + case SETUP_SKB_PROG_NONLINEAR: { + struct __sk_buff ctx = {}; int prog_fd; char buf[64]; @@ -106,6 +110,12 @@ static void verify_success(const char *prog_name, enum test_setup_type setup_typ .repeat = 1, ); + if (setup_type == SETUP_SKB_PROG_NONLINEAR) { + ctx.data_end = ETH_HLEN + sizeof(struct iphdr); + topts.ctx_in = &ctx; + topts.ctx_size_in = sizeof(ctx); + } + prog_fd = bpf_program__fd(prog); if (!ASSERT_GE(prog_fd, 0, "prog_fd")) goto cleanup; diff --git a/tools/testing/selftests/bpf/progs/dynptr_success.c b/tools/testing/selftests/bpf/progs/dynptr_success.c index e0745b6e467e..b668ebd61fc7 100644 --- a/tools/testing/selftests/bpf/progs/dynptr_success.c +++ b/tools/testing/selftests/bpf/progs/dynptr_success.c @@ -10,6 +10,7 @@ #include "errno.h" #define PAGE_SIZE_64K 65536 +#define TEST_SKB_LINEAR_SIZE (sizeof(struct ethhdr) + sizeof(struct iphdr)) char _license[] SEC("license") = "GPL"; @@ -211,6 +212,25 @@ int test_dynptr_skb_data(struct __sk_buff *skb) return 1; } +SEC("?tc") +int test_dynptr_skb_slice_non_linear(struct __sk_buff *skb) +{ + struct bpf_dynptr ptr; + void *data; + + if (bpf_dynptr_from_skb(skb, 0, &ptr)) { + err = 1; + return 1; + } + + /* Ensure we cannot read past the end of the buffer. */ + data = bpf_dynptr_slice(&ptr, TEST_SKB_LINEAR_SIZE + 1, NULL, 1); + if (data) + err = 2; + + return 1; +} + SEC("?tc") int test_dynptr_skb_meta_data(struct __sk_buff *skb) { From 4a4852376e3a2727ea40e61143d6d7c22bb6dfad Mon Sep 17 00:00:00 2001 From: Emil Tsalapatis Date: Tue, 22 Sep 2026 17:20:20 +0000 Subject: [PATCH 478/592] bpf: Fix bpf_sock context code generation Currently, the ctx access code reads the rx_queue_mapping field with either a 4-byte or 2-byte load. The rest of the bits in the register are marked known zero by the verifier. However, the emitted ctx access code places in the register on certain the special value (-1) using BPF_MOV_IMM64, which gets sign-extended to turn on all the bits in the register. By shifting this value right, the program ends up with a value at runtime above what the verifier assumes is possible. Fix this by ensuring the read value is as wide as the assumed size. Use MOV32 instructions instead of MOV64 instructions to keep the upper bits zero as assumed by the verifier. Also properly report the size of the destination variable (the bpf_sock field, 4 bytes) instead of the source (the socket field, 2 bytes). Fixes: c3c16f2ea6d2 ("bpf: Add rx_queue_mapping to bpf_sock") Reported-by: Nicholas Carlini Suggested-by: Nicholas Carlini Signed-off-by: Emil Tsalapatis Signed-off-by: Alexei Starovoitov Reviewed-by: Jiayuan Chen Link: https://patch.msgid.link/20260922172028.6269-4-emil@etsalapatis.com --- net/core/filter.c | 7 ++++--- 1 file changed, 4 insertions(+), 3 deletions(-) diff --git a/net/core/filter.c b/net/core/filter.c index 532405988fd9..70dc621672f2 100644 --- a/net/core/filter.c +++ b/net/core/filter.c @@ -10566,11 +10566,12 @@ u32 bpf_sock_convert_ctx_access(enum bpf_access_type type, target_size)); *insn++ = BPF_JMP_IMM(BPF_JNE, si->dst_reg, NO_QUEUE_MAPPING, 1); - *insn++ = BPF_MOV64_IMM(si->dst_reg, -1); + *insn++ = BPF_MOV32_IMM(si->dst_reg, -1); #else - *insn++ = BPF_MOV64_IMM(si->dst_reg, -1); - *target_size = 2; + *insn++ = BPF_MOV32_IMM(si->dst_reg, -1); #endif + *target_size = sizeof_field(struct bpf_sock, rx_queue_mapping); + break; } From dec0c209a6808fe6de2e4787538e02786a16a4ef Mon Sep 17 00:00:00 2001 From: Emil Tsalapatis Date: Tue, 22 Sep 2026 17:20:21 +0000 Subject: [PATCH 479/592] selftests/bpf: Add selftests for rx_queue_mapping context access Add tests to ensure the verifier properly tracks the 0 bit state and width of the rx_queue_mapping field read from struct sock. Signed-off-by: Emil Tsalapatis Signed-off-by: Alexei Starovoitov Link: https://patch.msgid.link/20260922172028.6269-5-emil@etsalapatis.com --- .../selftests/bpf/progs/verifier_sock.c | 38 +++++++++++++++++++ 1 file changed, 38 insertions(+) diff --git a/tools/testing/selftests/bpf/progs/verifier_sock.c b/tools/testing/selftests/bpf/progs/verifier_sock.c index 4f2f3209eec8..bf9f6fb6582c 100644 --- a/tools/testing/selftests/bpf/progs/verifier_sock.c +++ b/tools/testing/selftests/bpf/progs/verifier_sock.c @@ -88,6 +88,44 @@ l0_%=: r0 = *(u32*)(r1 + %[bpf_sock_family]); \ : __clobber_all); } +SEC("socket") +__description("skb->sk: sk->rx_queue_mapping [no sign extension]") +__success __success_unpriv __retval(0) +__naked void sk_rx_queue_mapping_no_sign_ext(void) +{ + asm volatile (" \ + r1 = *(u64*)(r1 + %[__sk_buff_sk]); \ + if r1 != 0 goto l0_%=; \ + r0 = 0xdead; \ + exit; \ +l0_%=: r0 = *(u32*)(r1 + %[bpf_sock_rx_queue_mapping]); \ + r0 >>= 32; \ + exit; \ +" : + : __imm_const(__sk_buff_sk, offsetof(struct __sk_buff, sk)), + __imm_const(bpf_sock_rx_queue_mapping, offsetof(struct bpf_sock, rx_queue_mapping)) + : __clobber_all); +} + +SEC("socket") +__description("skb->sk: sk->rx_queue_mapping [narrow load mask]") +__success __success_unpriv __retval(0) +__naked void sk_rx_queue_mapping_narrow_load_mask(void) +{ + asm volatile (" \ + r1 = *(u64*)(r1 + %[__sk_buff_sk]); \ + if r1 != 0 goto l0_%=; \ + r0 = 0xdead; \ + exit; \ +l0_%=: r0 = *(u16*)(r1 + %[bpf_sock_rx_queue_mapping]); \ + r0 >>= 16; \ + exit; \ +" : + : __imm_const(__sk_buff_sk, offsetof(struct __sk_buff, sk)), + __imm_const(bpf_sock_rx_queue_mapping, offsetof(struct bpf_sock, rx_queue_mapping)) + : __clobber_all); +} + SEC("cgroup/skb") __description("skb->sk: sk->type [fullsock field]") __failure __msg("invalid sock_common access") From a6c1edfbe240e4377038a0e1d233ad81fde9a21b Mon Sep 17 00:00:00 2001 From: Emil Tsalapatis Date: Tue, 22 Sep 2026 17:20:22 +0000 Subject: [PATCH 480/592] bpf: Reject pkt arguments in mutating subprogs The verifier tracks changes in how PTR_TO_PACKET registers' bounds are modified across subprog boundaries. PTR_TO_PACKET registers are actually passed as PTR_TO_MEM, which is assumed valid for the entire call. This is not the case with packet memory, where a pskb_* call may invalidate its memory region. Reject BPF code that passes PTR_TO_PACKET pointers to subprogs that may mutate a packet. We cannot pass the pointer as a true PTR_TO_PACKET because we would also need to somehow pass the PTR_TO_PACKET_META or PTR_TO_PACKET_END to the subprog. Since we cannot avoid representing the pointer in the subprog as PTR_TO_MEM, only permit it if the subprog is guaranteed not to mutate the packet. Fixes: 80f281664f5a ("bpf: Support pointers in global func args") Reported-by: Nicholas Carlini Suggested-by: Nicholas Carlini Signed-off-by: Emil Tsalapatis Signed-off-by: Alexei Starovoitov Link: https://patch.msgid.link/20260922172028.6269-6-emil@etsalapatis.com --- kernel/bpf/verifier.c | 10 ++++++++++ 1 file changed, 10 insertions(+) diff --git a/kernel/bpf/verifier.c b/kernel/bpf/verifier.c index dd8bb179d39e..7ffbb804184f 100644 --- a/kernel/bpf/verifier.c +++ b/kernel/bpf/verifier.c @@ -9750,6 +9750,16 @@ static int btf_check_func_arg_match(struct bpf_verifier_env *env, int subprog, if (check_mem_reg(env, reg, argno, arg->mem_size, BPF_READ | BPF_WRITE, NULL, NULL)) return -EINVAL; + /* + * PTR_TO_PACKET get passed as PTR_TO_MEM, preventing + * us from adjusting bounds tracking info. + */ + if ((reg_is_pkt_pointer_any(reg) || reg_is_dynptr_slice_pkt(reg)) && + sub->changes_pkt_data) { + bpf_log(log, "%s is a packet pointer, but func#%d may change packet data\n", + reg_arg_name(env, argno), subprog); + return -EINVAL; + } if (!(arg->arg_type & PTR_MAYBE_NULL) && (type_may_be_null(reg->type) || bpf_register_is_null(reg))) { bpf_log(log, "%s is expected to be non-NULL\n", From 1ed69a54d31848618131aaf3311a78adbe78ced0 Mon Sep 17 00:00:00 2001 From: Emil Tsalapatis Date: Tue, 22 Sep 2026 17:20:23 +0000 Subject: [PATCH 481/592] selftests/bpf: Test rejection of pkt args to mutating subprogs Add a selftests that ensures that PTR_TO_PACKET arguments can only be passed to subprogs that will never adjust the underlying packet memory, and are rejected otherwise. Signed-off-by: Emil Tsalapatis Signed-off-by: Alexei Starovoitov Link: https://patch.msgid.link/20260922172028.6269-7-emil@etsalapatis.com --- .../bpf/progs/verifier_global_ptr_args.c | 53 +++++++++++++++++++ 1 file changed, 53 insertions(+) diff --git a/tools/testing/selftests/bpf/progs/verifier_global_ptr_args.c b/tools/testing/selftests/bpf/progs/verifier_global_ptr_args.c index a3d2af8dc839..dcc2dd46751a 100644 --- a/tools/testing/selftests/bpf/progs/verifier_global_ptr_args.c +++ b/tools/testing/selftests/bpf/progs/verifier_global_ptr_args.c @@ -350,4 +350,57 @@ int anything_to_untrusted_mem(void *ctx) return 0; } +struct pkt_arg { + __u64 x; + __u8 pad[56]; +}; + +__weak int subprog_pkt_ptr_no_change(struct pkt_arg *p) +{ + if (!p) + return 0; + + return p->x; +} + +SEC("?tc") +__success +int pkt_ptr_to_global_mem_arg_no_change(struct __sk_buff *skb) +{ + void *data = (void *)(long)skb->data; + void *data_end = (void *)(long)skb->data_end; + struct pkt_arg *p = data; + + if ((void *)(p + 1) > data_end) + return 0; + + return subprog_pkt_ptr_no_change(p); +} + +__weak int subprog_pkt_ptr_changes_data(struct __sk_buff *skb __arg_ctx, + struct pkt_arg *p) +{ + if (!p) + return 0; + + bpf_skb_pull_data(skb, 0); + return p->x; +} + +SEC("?tc") +__failure __log_level(2) +__msg("R2 is a packet pointer, but func#{{[0-9]+}} may change packet data") +__msg("Caller passes invalid args into func#{{[0-9]+}} ('subprog_pkt_ptr_changes_data')") +int pkt_ptr_to_global_mem_arg_changes_data(struct __sk_buff *skb) +{ + void *data = (void *)(long)skb->data; + void *data_end = (void *)(long)skb->data_end; + struct pkt_arg *p = data; + + if ((void *)(p + 1) > data_end) + return 0; + + return subprog_pkt_ptr_changes_data(skb, p); +} + char _license[] SEC("license") = "GPL"; From f85f5917aa2fbd861c72ea71ecb14a2fa915c3f6 Mon Sep 17 00:00:00 2001 From: Emil Tsalapatis Date: Tue, 22 Sep 2026 17:20:24 +0000 Subject: [PATCH 482/592] bpf: Prevent variable arena/non-arena register contents The verifier marks ALU instructions that include at least one arena operand with needs_zext: These instructions are fixed up after verification to be ALU32 instructions to ensure that the result is a valid offset into an arena. However, different code paths may provide two non-arena 64-bit arguments to the same instruction. The result of the operation in that code path is wrong, since it is now unexpectedly truncated to 32 bits and zero-extended. Add logic to the verifier to ensure every instruction either always has at least one PTR_TO_ARENA argument, or never does. Since needs_zext already tracks the first scenario, add a prevent_zext field in bpf_insn_aux to track the latter. Reject instructions that use arena arguments and have prevent_zext set, or do not have arena arguments and have needs_zext set. Fixes: 6082b6c328b5 ("bpf: Recognize addr_space_cast instruction in the verifier.") Reported-by: Nicholas Carlini Suggested-by: Nicholas Carlini Signed-off-by: Emil Tsalapatis Signed-off-by: Alexei Starovoitov Link: https://patch.msgid.link/20260922172028.6269-8-emil@etsalapatis.com --- include/linux/bpf_verifier.h | 1 + kernel/bpf/verifier.c | 24 +++++++++++++++++++++++- 2 files changed, 24 insertions(+), 1 deletion(-) diff --git a/include/linux/bpf_verifier.h b/include/linux/bpf_verifier.h index 6fe8e5dc57aa..b83ec99f1a13 100644 --- a/include/linux/bpf_verifier.h +++ b/include/linux/bpf_verifier.h @@ -680,6 +680,7 @@ struct bpf_insn_aux_data { bool nospec_result; /* result is unsafe under speculation, nospec must follow */ bool zext_dst; /* this insn zero extends dst reg */ bool needs_zext; /* alu op needs to clear upper bits */ + bool prevent_zext; /* alu op cannot be zext (already used with 64-bit scalars) */ bool non_sleepable; /* helper/kfunc may be called from non-sleepable context */ bool is_iter_next; /* bpf_iter__next() kfunc call */ bool call_with_percpu_alloc_ptr; /* {this,per}_cpu_ptr() with prog percpu alloc */ diff --git a/kernel/bpf/verifier.c b/kernel/bpf/verifier.c index 7ffbb804184f..41b49c56e123 100644 --- a/kernel/bpf/verifier.c +++ b/kernel/bpf/verifier.c @@ -15741,6 +15741,7 @@ static int adjust_reg_min_max_vals(struct bpf_verifier_env *env, struct bpf_reg_state *regs = state->regs, *dst_reg, *src_reg; struct bpf_reg_state *ptr_reg = NULL, off_reg = {0}; bool alu32 = (BPF_CLASS(insn->code) != BPF_ALU64); + struct bpf_insn_aux_data *aux = cur_aux(env); u8 opcode = BPF_OP(insn->code); int err; @@ -15752,12 +15753,23 @@ static int adjust_reg_min_max_vals(struct bpf_verifier_env *env, /* Case where at least one operand is an arena. */ if (dst_reg->type == PTR_TO_ARENA || (src_reg && src_reg->type == PTR_TO_ARENA)) { - struct bpf_insn_aux_data *aux = cur_aux(env); if (dst_reg->type != PTR_TO_ARENA) *dst_reg = *src_reg; if (BPF_CLASS(insn->code) == BPF_ALU64) { + /* + * Only arena pointers set needs_zext, but doing so + * modifies the instruction at fixup time to an ALU32 + * and makes it unsuitable for 64-bit scalar args. We + * prevent zext from being set if the instruction has + * been previously called with non-arena registers. + */ + if (aux->prevent_zext) { + verbose(env, "same insn cannot be used with and without arena pointer\n"); + return -EINVAL; + } + /* * 32-bit operations zero upper bits automatically. * 64-bit operations need to be converted to 32. @@ -15770,6 +15782,16 @@ static int adjust_reg_min_max_vals(struct bpf_verifier_env *env, return 0; } + /* Prevent the instruction from being used with arena pointers (see above). */ + if (env->prog->aux->arena && BPF_CLASS(insn->code) == BPF_ALU64) { + if (aux->needs_zext) { + verbose(env, "same insn cannot be used with and without arena pointer\n"); + return -EINVAL; + } + + aux->prevent_zext = true; + } + if (dst_reg->type != SCALAR_VALUE) ptr_reg = dst_reg; From a9e86dd9de4f933b69f5cca6293fd4c8919224ec Mon Sep 17 00:00:00 2001 From: Emil Tsalapatis Date: Tue, 22 Sep 2026 17:20:25 +0000 Subject: [PATCH 483/592] selftests/bpf: Test for mixed arena/nonarena code paths Add a selftest to confirm the verifier rejects ALU operations that return arena or non-arena results depending on code path. Signed-off-by: Emil Tsalapatis Signed-off-by: Alexei Starovoitov Link: https://patch.msgid.link/20260922172028.6269-9-emil@etsalapatis.com --- .../selftests/bpf/progs/verifier_arena.c | 49 +++++++++++++++++++ 1 file changed, 49 insertions(+) diff --git a/tools/testing/selftests/bpf/progs/verifier_arena.c b/tools/testing/selftests/bpf/progs/verifier_arena.c index 815f342eb4b0..3e33766547c0 100644 --- a/tools/testing/selftests/bpf/progs/verifier_arena.c +++ b/tools/testing/selftests/bpf/progs/verifier_arena.c @@ -561,6 +561,55 @@ int arena_ptr_add_arena_ptr(void *ctx) return 0; } +SEC("syscall") +__failure __msg("same insn cannot be used with and without arena pointer") +int mixed_arena_scalar_alu64_scalar_first(void *ctx) +{ + volatile register __u64 reg asm("r3"); + __u32 pick_arena = bpf_get_prandom_u32(); + + reg = 1ULL << 32; + + if (pick_arena) { + asm volatile ( + "r9 = %[arena] ll;" + "%[reg] = 0;" + "%[reg] = addr_space_cast(%[reg], 0x0, 0x1);" + : [reg] "=r"(reg) + : __imm_addr(arena) + : "r9" + ); + } + + reg += 1; + + return 0; +} + +SEC("syscall") +__failure __msg("same insn cannot be used with and without arena pointer") +int mixed_arena_scalar_alu64_arena_first(void *ctx) +{ + volatile register __u64 reg asm("r3"); + __u32 pick_scalar = bpf_get_prandom_u32(); + + asm volatile ( + "r9 = %[arena] ll;" + "%[reg] = 0;" + "%[reg] = addr_space_cast(%[reg], 0x0, 0x1);" + : [reg] "=r"(reg) + : __imm_addr(arena) + : "r9" + ); + + if (pick_scalar) + reg = 1ULL << 32; + + reg += 1; + + return 0; +} + SEC("syscall") __success __retval(0) int scalar_xor_arena_ptr(void *ctx) From 814a81c842bd88f6bd8a4ce550d560df071a5d03 Mon Sep 17 00:00:00 2001 From: Zhao Gongyi Date: Thu, 17 Sep 2026 20:10:16 +0800 Subject: [PATCH 484/592] bpf, sockmap: Reject max_entries > INT_MAX in sock_map_alloc sock_map_alloc() only rejects max_entries == 0 and otherwise allows any u32 value. sock_map_free() then walks the sks[] array with a signed int iterator: int i; for (i = 0; i < stab->map.max_entries; i++) struct sock **psk = &stab->sks[i]; When a SOCKMAP is created with max_entries = 0xffffffff (UINT_MAX), the allocation of 32 GiB can succeed on large-memory hosts. During free the counter reaches 0x80000000, wraps to INT_MIN, is sign-extended by movslq and turned into a ~16 GiB negative offset from stab->sks, pointing far below the allocation. The faulting access is an xchg() write in sock_map_free(). Without KASAN, the same out-of-bounds write can fault on an unmapped vmalloc page or corrupt an unrelated allocation if that vmalloc address is populated. On a KASAN kernel with CONFIG_KASAN_VMALLOC=y, the shadow check for that address hits an unmapped shadow page and oopses first: BUG: unable to handle page fault for address: fffff521b59c5a00 RIP: 0010:kasan_check_range+0x107/0x190 Call Trace: sock_map_free+0x93/0x190 map_create+0x68d/0xb30 __sys_bpf+0x21e/0x2e70 Vmcore confirmed stab->map.max_entries == 0xffffffff, stab->sks == 0xffffc911ace2d000, and the faulting address sks + (s64)INT_MIN * 8 exactly at 0xffffc90dace2d000. The same buggy path is reached on the normal close()/bpf_map_free_deferred() path whenever such a map is destroyed. sock_map_alloc() used to bound its allocation size through bpf_map_charge_init(), but the bound was dropped when rlimit-based memory accounting was removed. Reject max_entries > INT_MAX at creation time so the signed iterator in sock_map_free() never sees a value that would overflow. Triggered by syzkaller and reproduced on both a 6.6-based KASAN kernel and the upstream v7.3-rc2 kernel. Fixes: 0d2c4f964050 ("bpf: Eliminate rlimit-based memory accounting for sockmap and sockhash maps") Signed-off-by: Zhao Gongyi Signed-off-by: Alexei Starovoitov Link: https://patch.msgid.link/20260917121016.48171-1-zhaogongyi@bytedance.com --- net/core/sock_map.c | 1 + 1 file changed, 1 insertion(+) diff --git a/net/core/sock_map.c b/net/core/sock_map.c index ca49bc7f8687..38df84284328 100644 --- a/net/core/sock_map.c +++ b/net/core/sock_map.c @@ -41,6 +41,7 @@ static struct bpf_map *sock_map_alloc(union bpf_attr *attr) struct bpf_stab *stab; if (attr->max_entries == 0 || + attr->max_entries > INT_MAX || attr->key_size != 4 || (attr->value_size != sizeof(u32) && attr->value_size != sizeof(u64)) || From 35e6f970f553954d92ba20afa885139a8e7dd0d7 Mon Sep 17 00:00:00 2001 From: Bernardo Soares Date: Fri, 18 Sep 2026 10:59:30 +0100 Subject: [PATCH 485/592] net/mlx5: Bridge, don't fail switchdev events of sibling eswitch ports mlx5 registers the bridge offload switchdev notifiers once per eswitch instance, but the notifier chains are global, so every instance sees every event and must filter out the ones that aren't its own. The existing filter, mlx5_esw_bridge_dev_same_hw(), only checks that the event netdevice sits on the same HCA - intentional for merged eswitch, where one bridge can span representors of several eswitches on one HCA - but same-HCA doesn't mean the instance actually has that port: peer ports are only created reactively from NETDEV_CHANGEUPPER, so an instance brought up after a sibling PF's port was already enslaved has none. The port object and attribute handlers claim the event anyway once same-HW passes, then fail the port lookup and return -EINVAL, which gets reported to user space even though the owning instance already handled it (e.g. "bridge vlan add ... RTNETLINK answers: Invalid argument"). Fix by filtering on the tracked port instead. The same gap exists in the generic recursive lower-device walk used by attribute changes on a bridge with more than one representor enslaved directly: mlx5_esw_bridge_lower_rep_vport_num_vhca_id_get() is entered with the bridge master netdevice, falls through to its generic netdev_for_each_lower_dev() loop, and returns as soon as the recursion into any one lower device yields a non-NULL rep - the underlying base case, mlx5_esw_bridge_rep_vport_num_vhca_id_get(), only checks mlx5_esw_bridge_dev_same_hw(), not ownership by the calling instance's br_offloads. mlx5_esw_bridge_lag_rep_get(), used for the LAG-master case, already filters on mlx5_esw_bridge_dev_same_esw() per candidate and so cannot select a sibling's rep; it is not the source of this bug. On a merged-eswitch HCA with a bridge spanning representors of more than one eswitch instance directly, the walk can return a sibling's rep instead of continuing to the one the calling instance actually owns, so the attribute change fails the same way as above. Fix by checking mlx5_esw_bridge_port_exists() at the point each rep is picked, same as the previous fix did for the notifier filter. Fixes: c358ea1741bc ("net/mlx5: Bridge, allow merged eswitch connectivity") Signed-off-by: Bernardo Soares Cc: Vlad Buslov Cc: Saeed Mahameed Reviewed-by: Mark Bloch Link: https://patch.msgid.link/20260918095931.29792-2-bsoares.it@gmail.com Signed-off-by: Jakub Kicinski --- .../mellanox/mlx5/core/en/rep/bridge.c | 45 +++++++++++++++---- .../ethernet/mellanox/mlx5/core/esw/bridge.c | 6 +++ .../ethernet/mellanox/mlx5/core/esw/bridge.h | 2 + 3 files changed, 44 insertions(+), 9 deletions(-) diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en/rep/bridge.c b/drivers/net/ethernet/mellanox/mlx5/core/en/rep/bridge.c index baac38bece14..4b7b0a0fc2b2 100644 --- a/drivers/net/ethernet/mellanox/mlx5/core/en/rep/bridge.c +++ b/drivers/net/ethernet/mellanox/mlx5/core/en/rep/bridge.c @@ -85,9 +85,16 @@ mlx5_esw_bridge_lower_rep_vport_num_vhca_id_get(struct net_device *dev, struct m struct net_device *lower_dev; struct list_head *iter; - if (netif_is_lag_master(dev) || mlx5e_eswitch_rep(dev)) - return mlx5_esw_bridge_rep_vport_num_vhca_id_get(dev, esw, vport_num, - esw_owner_vhca_id); + if (netif_is_lag_master(dev) || mlx5e_eswitch_rep(dev)) { + struct net_device *rep; + + rep = mlx5_esw_bridge_rep_vport_num_vhca_id_get(dev, esw, vport_num, + esw_owner_vhca_id); + if (rep && !mlx5_esw_bridge_port_exists(*vport_num, *esw_owner_vhca_id, + esw->br_offloads)) + return NULL; + return rep; + } netdev_for_each_lower_dev(dev, lower_dev, iter) { struct net_device *rep; @@ -104,6 +111,28 @@ mlx5_esw_bridge_lower_rep_vport_num_vhca_id_get(struct net_device *dev, struct m return NULL; } +static bool mlx5_esw_bridge_rep_port_lookup(struct net_device *dev, + struct mlx5_esw_bridge_offloads *br_offloads, + u16 *vport_num, u16 *esw_owner_vhca_id) +{ + if (!mlx5_esw_bridge_rep_vport_num_vhca_id_get(dev, br_offloads->esw, vport_num, + esw_owner_vhca_id)) + return false; + + return mlx5_esw_bridge_port_exists(*vport_num, *esw_owner_vhca_id, br_offloads); +} + +static bool mlx5_esw_bridge_lower_rep_port_lookup(struct net_device *dev, + struct mlx5_esw_bridge_offloads *br_offloads, + u16 *vport_num, u16 *esw_owner_vhca_id) +{ + if (!mlx5_esw_bridge_lower_rep_vport_num_vhca_id_get(dev, br_offloads->esw, vport_num, + esw_owner_vhca_id)) + return false; + + return mlx5_esw_bridge_port_exists(*vport_num, *esw_owner_vhca_id, br_offloads); +} + static bool mlx5_esw_bridge_is_local(struct net_device *dev, struct net_device *rep, struct mlx5_eswitch *esw) { @@ -218,8 +247,7 @@ mlx5_esw_bridge_port_obj_add(struct net_device *dev, u16 vport_num, esw_owner_vhca_id; int err; - if (!mlx5_esw_bridge_rep_vport_num_vhca_id_get(dev, br_offloads->esw, &vport_num, - &esw_owner_vhca_id)) + if (!mlx5_esw_bridge_rep_port_lookup(dev, br_offloads, &vport_num, &esw_owner_vhca_id)) return 0; port_obj_info->handled = true; @@ -251,8 +279,7 @@ mlx5_esw_bridge_port_obj_del(struct net_device *dev, const struct switchdev_obj_port_mdb *mdb; u16 vport_num, esw_owner_vhca_id; - if (!mlx5_esw_bridge_rep_vport_num_vhca_id_get(dev, br_offloads->esw, &vport_num, - &esw_owner_vhca_id)) + if (!mlx5_esw_bridge_rep_port_lookup(dev, br_offloads, &vport_num, &esw_owner_vhca_id)) return 0; port_obj_info->handled = true; @@ -283,8 +310,8 @@ mlx5_esw_bridge_port_obj_attr_set(struct net_device *dev, u16 vport_num, esw_owner_vhca_id; int err = 0; - if (!mlx5_esw_bridge_lower_rep_vport_num_vhca_id_get(dev, br_offloads->esw, &vport_num, - &esw_owner_vhca_id)) + if (!mlx5_esw_bridge_lower_rep_port_lookup(dev, br_offloads, &vport_num, + &esw_owner_vhca_id)) return 0; port_attr_info->handled = true; diff --git a/drivers/net/ethernet/mellanox/mlx5/core/esw/bridge.c b/drivers/net/ethernet/mellanox/mlx5/core/esw/bridge.c index 87b5fd349594..ac90ccda1272 100644 --- a/drivers/net/ethernet/mellanox/mlx5/core/esw/bridge.c +++ b/drivers/net/ethernet/mellanox/mlx5/core/esw/bridge.c @@ -1686,6 +1686,12 @@ int mlx5_esw_bridge_vport_peer_unlink(struct net_device *br_netdev, u16 vport_nu extack); } +bool mlx5_esw_bridge_port_exists(u16 vport_num, u16 esw_owner_vhca_id, + struct mlx5_esw_bridge_offloads *br_offloads) +{ + return mlx5_esw_bridge_port_lookup(vport_num, esw_owner_vhca_id, br_offloads); +} + int mlx5_esw_bridge_port_vlan_add(u16 vport_num, u16 esw_owner_vhca_id, u16 vid, u16 flags, struct mlx5_esw_bridge_offloads *br_offloads, struct netlink_ext_ack *extack) diff --git a/drivers/net/ethernet/mellanox/mlx5/core/esw/bridge.h b/drivers/net/ethernet/mellanox/mlx5/core/esw/bridge.h index d6f539161993..a4e59cc21089 100644 --- a/drivers/net/ethernet/mellanox/mlx5/core/esw/bridge.h +++ b/drivers/net/ethernet/mellanox/mlx5/core/esw/bridge.h @@ -80,6 +80,8 @@ int mlx5_esw_bridge_vlan_proto_set(u16 vport_num, u16 esw_owner_vhca_id, u16 pro struct mlx5_esw_bridge_offloads *br_offloads); int mlx5_esw_bridge_mcast_set(u16 vport_num, u16 esw_owner_vhca_id, bool enable, struct mlx5_esw_bridge_offloads *br_offloads); +bool mlx5_esw_bridge_port_exists(u16 vport_num, u16 esw_owner_vhca_id, + struct mlx5_esw_bridge_offloads *br_offloads); int mlx5_esw_bridge_port_vlan_add(u16 vport_num, u16 esw_owner_vhca_id, u16 vid, u16 flags, struct mlx5_esw_bridge_offloads *br_offloads, struct netlink_ext_ack *extack); From 2e51097c982b7b22382fda3202ba29f0ea33e8c0 Mon Sep 17 00:00:00 2001 From: Bernardo Soares Date: Fri, 18 Sep 2026 10:59:31 +0100 Subject: [PATCH 486/592] net/mlx5: Bridge, don't fail unlink of untracked/unsupported peer ports mlx5_esw_bridge_vport_unlink() returns -EINVAL when the port isn't tracked by this instance's br_offloads. This is reachable on a sibling instance that registered its notifier after the port was already enslaved: it never saw the NETDEV_CHANGEUPPER link event, so peer_link() never created a peer port for it, but it does see the later unlink event and fails. Return 0 instead, and give mlx5_esw_bridge_vport_peer_unlink() the same merged_eswitch capability guard peer_link() already has, since without it peer_link() likewise never creates a port to unlink. This also matters beyond the -EINVAL itself: mlx5_esw_bridge_switchdev_port_event() runs on the per-netns netdev_chain, and notifier_from_errno(-EINVAL) sets NOTIFY_STOP_MASK, which call_netdevice_notifiers_info() checks to stop calling further listeners on that chain - so the old -EINVAL silently dropped the event for any listener registered later on the same chain, even though none of it was visible to user space since __netdev_upper_dev_unlink() discards the return value. Fixes: c358ea1741bc ("net/mlx5: Bridge, allow merged eswitch connectivity") Signed-off-by: Bernardo Soares Reviewed-by: Mark Bloch Link: https://patch.msgid.link/20260918095931.29792-3-bsoares.it@gmail.com Signed-off-by: Jakub Kicinski --- drivers/net/ethernet/mellanox/mlx5/core/esw/bridge.c | 9 +++++---- 1 file changed, 5 insertions(+), 4 deletions(-) diff --git a/drivers/net/ethernet/mellanox/mlx5/core/esw/bridge.c b/drivers/net/ethernet/mellanox/mlx5/core/esw/bridge.c index ac90ccda1272..b4cf3c5ac0dd 100644 --- a/drivers/net/ethernet/mellanox/mlx5/core/esw/bridge.c +++ b/drivers/net/ethernet/mellanox/mlx5/core/esw/bridge.c @@ -1649,10 +1649,8 @@ int mlx5_esw_bridge_vport_unlink(struct net_device *br_netdev, u16 vport_num, int err; port = mlx5_esw_bridge_port_lookup(vport_num, esw_owner_vhca_id, br_offloads); - if (!port) { - NL_SET_ERR_MSG_MOD(extack, "Port is not attached to any bridge"); - return -EINVAL; - } + if (!port) + return 0; if (port->bridge->ifindex != br_netdev->ifindex) { NL_SET_ERR_MSG_MOD(extack, "Port is attached to another bridge"); return -EINVAL; @@ -1682,6 +1680,9 @@ int mlx5_esw_bridge_vport_peer_unlink(struct net_device *br_netdev, u16 vport_nu struct mlx5_esw_bridge_offloads *br_offloads, struct netlink_ext_ack *extack) { + if (!MLX5_CAP_ESW(br_offloads->esw->dev, merged_eswitch)) + return 0; + return mlx5_esw_bridge_vport_unlink(br_netdev, vport_num, esw_owner_vhca_id, br_offloads, extack); } From 0bf6bb567f0edaa771e7dd208ef98da50e6a4485 Mon Sep 17 00:00:00 2001 From: Wentao Liang Date: Thu, 17 Sep 2026 11:31:31 +0000 Subject: [PATCH 487/592] net/mlx5: Fix rev_entry reference leak in mlx5_tc_ct_shared_counter_get() When the reverse entry is found but its counter is already being released, refcount_inc_not_zero() fails and the reference taken by mlx5_tc_ct_entry_get() is never dropped before falling through to create_counter. Drop it so the reverse entry is not kept alive forever by a shared counter lookup that did not use it. Fixes: 1edae2335adf ("net/mlx5e: CT: Use the same counter for both directions") Cc: stable@vger.kernel.org Signed-off-by: Wentao Liang Reviewed-by: Tariq Toukan Link: https://patch.msgid.link/20260917113131.2149024-1-vulab@iscas.ac.cn Signed-off-by: Jakub Kicinski --- drivers/net/ethernet/mellanox/mlx5/core/en/tc_ct.c | 3 +++ 1 file changed, 3 insertions(+) diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en/tc_ct.c b/drivers/net/ethernet/mellanox/mlx5/core/en/tc_ct.c index 6c87a1c7db09..c1841ce74d9a 100644 --- a/drivers/net/ethernet/mellanox/mlx5/core/en/tc_ct.c +++ b/drivers/net/ethernet/mellanox/mlx5/core/en/tc_ct.c @@ -1082,6 +1082,9 @@ mlx5_tc_ct_shared_counter_get(struct mlx5_tc_ct_priv *ct_priv, spin_unlock_bh(&ct_priv->ht_lock); + if (rev_entry) + mlx5_tc_ct_entry_put(rev_entry); + create_counter: shared_counter = mlx5_tc_ct_counter_create(ct_priv); From 4498467a8af06cfa3d71cb04bd7c4170dec8f449 Mon Sep 17 00:00:00 2001 From: Aohan Mei Date: Mon, 21 Sep 2026 17:37:04 +0800 Subject: [PATCH 488/592] sctp: discard the rest of the packet on a stale-cookie error When an association is in COOKIE-ECHOED state and the peer sends a bundled [ERROR(Stale Cookie)][DATA] packet from one of its non-primary addresses, processing the ERROR chunk takes the non-fatal stale-cookie retry path sctp_sf_do_5_2_6_stale(), which queues SCTP_CMD_DEL_NON_PRIMARY while keeping the association alive. sctp_cmd_del_non_primary() removes every non-primary transport - including the very transport this packet arrived on, which is still referenced by the receive lookup and shared by all chunks of the packet via chunk->transport. sctp_assoc_rm_peer() does redirect asoc->peer.last_data_from away from the removed transport, but right afterwards the bundled DATA chunk makes sctp_assoc_bh_rcv() re-register asoc->peer.last_data_from = chunk->transport unconditionally, undoing the redirection with the just-removed transport. Once the packet is done, the receive reference is dropped and the transport is RCU-freed, while the surviving association keeps the dangling last_data_from. A later FWD-TSN (or the delayed SACK timer) makes sctp_gen_sack() dereference it (->param_flags and friends), and sctp_make_sack()/sctp_outq_select_transport() may write to the freed object and link it into the live transport list. This is a use-after-free triggerable by any malicious SCTP peer (or a local unprivileged user acting as one) with no capabilities required: BUG: KASAN: slab-use-after-free in sctp_do_sm+0x498a/0x5660 Read of size 4 at addr ffff88800e1e356c by task poc/115 Call Trace: sctp_do_sm <- sctp_assoc_bh_rcv <- sctp_inq_push <- sctp_rcv <- ip_protocol_deliver_rcu <- ip_rcv Allocated: sctp_transport_new <- sctp_assoc_add_peer <- sctp_process_init (INIT-ACK processing) Freed: kfree <- sctp_transport_destroy_rcu <- rcu_core (call_rcu queued by sctp_transport_put at end of sctp_rcv) The buggy address is located 364 bytes inside of freed 1024-byte region [ffff88800e1e3400, ffff88800e1e3800), cache kmalloc-1k Note that commit 03a9d10ecf71 ("sctp: drop a chunk if its transport was removed") only covers the window between the receive lookup and the chunk processing (e.g. an ASCONF DEL-IP racing the socket backlog); here the transport is removed *while* the packet is being processed, by an earlier chunk of the same packet, so the drop in sctp_inq_push() does not reach this path. Verified with the bundled [ERROR(Stale Cookie)][DATA] + FWD-TSN reproducer: the KASAN report above still fires with that commit applied, and is gone with this patch on top. Fix it by discarding the rest of the packet on this path, as suggested by Xin. After the stale-cookie ERROR has sent the association back to COOKIE-WAIT and removed the non-primary transports, the remaining chunks of the packet can only run against the restarted handshake while referencing the removed arrival transport through chunk->transport: besides the last_data_from registration above, sctp_cmd_setup_t2() and the sctp_make_*() reply builders would also copy that pointer into association-lifetime state that sctp_assoc_rm_peer() has already sanitized. Let the peer retransmit them, in line with what sctp_inq_push() does for chunks whose transport was removed before processing. Fixes: 1da177e4c3f4 ("Linux-2.6.12-rc2") Suggested-by: Xin Long Reported-by: TencentOS Corvus AI Cc: stable@vger.kernel.org Signed-off-by: Aohan Mei Acked-by: Xin Long Link: https://patch.msgid.link/20260921093707.1432184-1-ljp1205831794@gmail.com Signed-off-by: Jakub Kicinski --- net/sctp/sm_statefuns.c | 2 ++ 1 file changed, 2 insertions(+) diff --git a/net/sctp/sm_statefuns.c b/net/sctp/sm_statefuns.c index c701cff6ea67..8dc65498763c 100644 --- a/net/sctp/sm_statefuns.c +++ b/net/sctp/sm_statefuns.c @@ -2654,6 +2654,8 @@ static enum sctp_disposition sctp_sf_do_5_2_6_stale( sctp_add_cmd_sf(commands, SCTP_CMD_REPLY, SCTP_CHUNK(reply)); + sctp_add_cmd_sf(commands, SCTP_CMD_DISCARD_PACKET, SCTP_NULL()); + return SCTP_DISPOSITION_CONSUME; nomem: From 07e1a9408b6c2f9d0cfb757b67dabb52da7a32b2 Mon Sep 17 00:00:00 2001 From: Willem de Bruijn Date: Fri, 18 Sep 2026 20:47:28 -0400 Subject: [PATCH 489/592] virtio_net: copy zerocopy frags in start_xmit without NAPI Virtio-net without NAPI frees completed skbs lazily on the next start_xmit. Senders waiting for in-flight zerocopy buffers can deadlock if they cannot transmit more packets, as then no completed packets will be freed. When !use_napi, virtio-net already calls skb_orphan to avoid waiting up for transmitted skbs to be freed. For zerocopy packets that require deep copying on orphan (i.e. those that do not set SKBFL_DONT_ORPHAN, such as PACKET_TX_RING), call skb_orphan_frags before orphaning to release the buffers. This fixes the tpacket_snd slot reuse bug on skb_orphan for virtio-net, and prevents PACKET_TX_RING from running out of slots. This fix also touches vhost_net zerocopy packets, which also do not set SKBFL_DONT_ORPHAN. This is fine: vhost_net packets only encounter virtio-net in nested virtualization, and only if napi_tx is explicitly disabled (it has been default-enabled since Linux 4.12). In that rare case, copying the frags is desirable anyway to prevent holding guest descriptors pinned across unbounded intervals. This is a prerequisite for the next patch, which converts PACKET_TX_RING to standard zerocopy completion. Without this patch first, a bounded ring sender can stall indefinitely behind a virtio-net virtqueue that cannot reclaim. Fixes: 5cd8d46ea156 ("packet: copy user buffers before orphan or clone") Cc: stable@vger.kernel.org Cc: mst@redhat.com Cc: jasowangio@gmail.com Signed-off-by: Willem de Bruijn Link: https://patch.msgid.link/20260919004748.1463985-2-willemdebruijn.kernel@gmail.com Signed-off-by: Jakub Kicinski --- drivers/net/virtio_net.c | 9 +++++++++ 1 file changed, 9 insertions(+) diff --git a/drivers/net/virtio_net.c b/drivers/net/virtio_net.c index e34c52d059d3..bf82ef9874ab 100644 --- a/drivers/net/virtio_net.c +++ b/drivers/net/virtio_net.c @@ -3349,6 +3349,14 @@ static netdev_tx_t start_xmit(struct sk_buff *skb, struct net_device *dev) else virtqueue_disable_cb(sq->vq); + if (!use_napi && + unlikely(skb_orphan_frags(skb, GFP_ATOMIC))) { + DEV_STATS_INC(dev, tx_dropped); + dev_kfree_skb_any(skb); + kick = !xmit_more || netif_xmit_stopped(txq); + goto kick_vq; + } + /* timestamp packet in software */ skb_tx_timestamp(skb); @@ -3381,6 +3389,7 @@ static netdev_tx_t start_xmit(struct sk_buff *skb, struct net_device *dev) kick = use_napi ? __netdev_tx_sent_queue(txq, skb->len, xmit_more) : !xmit_more || netif_xmit_stopped(txq); +kick_vq: if (kick) { if (virtqueue_kick_prepare(sq->vq) && virtqueue_notify(sq->vq)) { u64_stats_update_begin(&sq->stats.syncp); From 9518405613863d0bf0700927a21367f5942cf058 Mon Sep 17 00:00:00 2001 From: Willem de Bruijn Date: Fri, 18 Sep 2026 20:47:29 -0400 Subject: [PATCH 490/592] packet: use ubuf_info completion for TX_RING packets tpacket_snd sends skbs with frags pointing into its ring slots. Slots are released when skb->destructor is called. A call to skb_orphan calls skb->destructor before the skb is freed. This can cause the slot to be reused while still linked into the skb. Switch to standard zerocopy completion (ubuf_info) so the slot is only released once all references to the payload are freed or copied. Restore skb->destructor to standard sock_wfree. The ubuf_info completion callback can be called with a NULL skb, but only from net_zcopy_put and related API, used by zerocopy implementations that hold their own reference on the uarg, such as MSG_ZEROCOPY. This uarg is only ever completed from skb_zcopy_clear, so skb is always set. To prevent userspace from aliasing in-flight state on shared ring slots, allocate tpacket_uarg per packet, rather than per slot. This adds a small allocation to the transmit path. Use standard kmalloc to allow backporting to stable kernels. The uarg holds an sk_wmem_alloc reference, rather than an sk_refcnt reference. packet_free_tx_ring waits on sk_wmem_alloc before freeing the ring pages. Always allocate vec->deferred for tx_ring so page-backed rings also wait on sk_wmem_alloc when skb_copy_ubufs drops page refs before calling tpacket_ubuf_complete. Drop the tx_ring.pg_vec test that tpacket_destruct_skb performed before accessing the slot. The sk_wmem_alloc reference now guarantees that the slot is valid. The test is also not sufficient by itself, as it reads pg_vec without pg_vec_lock, so it can race with packet_set_ring. As a result a slot is released when its payload is copied, which can be before transmission (e.g., in skb_orphan_frags_rx). If copied before skb_tx_timestamp() is called, no slot timestamp is recorded, similar to when skb_orphan() was called early in the datapath before this patch. Revert the now unused previous skb_zcopy_.._nouarg infra. Depends on commit 992cc9f94ca9 ("net/packet: defer vmalloc TX_RING free until skbs finish"). Reported-by: Katherine Leaver Reported-by: Bjoern Doebel Closes: https://lore.kernel.org/netdev/20260909085542.3370986-1-doebel@amazon.de/ Fixes: 5cd8d46ea156 ("packet: copy user buffers before orphan or clone") Cc: stable@vger.kernel.org Signed-off-by: Willem de Bruijn Link: https://patch.msgid.link/20260919004748.1463985-3-willemdebruijn.kernel@gmail.com Signed-off-by: Jakub Kicinski --- include/linux/skbuff.h | 19 +------- net/packet/af_packet.c | 102 ++++++++++++++++++++++++++--------------- 2 files changed, 65 insertions(+), 56 deletions(-) diff --git a/include/linux/skbuff.h b/include/linux/skbuff.h index 421f6fc45451..b14d6be7370b 100644 --- a/include/linux/skbuff.h +++ b/include/linux/skbuff.h @@ -1834,22 +1834,6 @@ static inline void skb_zcopy_set(struct sk_buff *skb, struct ubuf_info *uarg, } } -static inline void skb_zcopy_set_nouarg(struct sk_buff *skb, void *val) -{ - skb_shinfo(skb)->destructor_arg = (void *)((uintptr_t) val | 0x1UL); - skb_shinfo(skb)->flags |= SKBFL_ZEROCOPY_FRAG; -} - -static inline bool skb_zcopy_is_nouarg(struct sk_buff *skb) -{ - return (uintptr_t) skb_shinfo(skb)->destructor_arg & 0x1UL; -} - -static inline void *skb_zcopy_get_nouarg(struct sk_buff *skb) -{ - return (void *)((uintptr_t) skb_shinfo(skb)->destructor_arg & ~0x1UL); -} - static inline void net_zcopy_put(struct ubuf_info *uarg) { if (uarg) @@ -1872,8 +1856,7 @@ static inline void skb_zcopy_clear(struct sk_buff *skb, bool zerocopy_success) struct ubuf_info *uarg = skb_zcopy(skb); if (uarg) { - if (!skb_zcopy_is_nouarg(skb)) - uarg->ops->complete(skb, uarg, zerocopy_success); + uarg->ops->complete(skb, uarg, zerocopy_success); skb_shinfo(skb)->flags &= ~SKBFL_ALL_ZEROCOPY; } diff --git a/net/packet/af_packet.c b/net/packet/af_packet.c index 50cae32ae269..64b501db660a 100644 --- a/net/packet/af_packet.c +++ b/net/packet/af_packet.c @@ -2530,26 +2530,6 @@ static int tpacket_rcv(struct sk_buff *skb, struct net_device *dev, goto drop_n_restore; } -static void tpacket_destruct_skb(struct sk_buff *skb) -{ - struct packet_sock *po = pkt_sk(skb->sk); - - if (likely(po->tx_ring.pg_vec)) { - void *ph; - __u32 ts; - - ph = skb_zcopy_get_nouarg(skb); - - ts = __packet_set_timestamp(po, ph, skb); - __packet_set_status(po, ph, TP_STATUS_AVAILABLE | ts); - - packet_dec_pending(&po->tx_ring); - complete(&po->skb_completion); - } - - sock_wfree(skb); -} - static int __packet_snd_vnet_parse(struct virtio_net_hdr *vnet_hdr, size_t len) { if ((vnet_hdr->flags & VIRTIO_NET_HDR_F_NEEDS_CSUM) && @@ -2589,27 +2569,56 @@ static int packet_snd_vnet_parse(struct msghdr *msg, size_t *len, return 0; } +struct tpacket_uarg { + struct ubuf_info ubuf; + struct packet_sock *po; + void *ph; +}; + +static void tpacket_ubuf_complete(struct sk_buff *skb, struct ubuf_info *uarg, + bool success) +{ + struct tpacket_uarg *tu = container_of(uarg, struct tpacket_uarg, ubuf); + struct packet_sock *po = tu->po; + void *ph = tu->ph; + __u32 ts; + + DEBUG_NET_WARN_ON_ONCE(!skb); + + if (!refcount_dec_and_test(&uarg->refcnt)) + return; + + ts = __packet_set_timestamp(po, ph, skb); + __packet_set_status(po, ph, TP_STATUS_AVAILABLE | ts); + + packet_dec_pending(&po->tx_ring); + complete(&po->skb_completion); + + kfree(tu); + sk_free(&po->sk); +} + +static const struct ubuf_info_ops tpacket_ubuf_ops = { + .complete = tpacket_ubuf_complete, +}; + static int tpacket_fill_skb(struct packet_sock *po, struct sk_buff *skb, - void *frame, struct net_device *dev, void *data, int tp_len, + struct net_device *dev, void *data, int tp_len, __be16 proto, unsigned char *addr, int hlen, int copylen, int hard_header_len, const struct sockcm_cookie *sockc) { - union tpacket_uhdr ph; int to_write, offset, len, nr_frags, len_max; struct socket *sock = po->sk.sk_socket; struct page *page; int err; - ph.raw = frame; - skb->protocol = proto; skb->dev = dev; skb->priority = sockc->priority; skb->mark = sockc->mark; skb_set_delivery_type_by_clockid(skb, sockc->transmit_time, po->sk.sk_clockid); skb_setup_tx_timestamp(skb, sockc); - skb_zcopy_set_nouarg(skb, ph.raw); skb_reserve(skb, hlen); skb_reset_network_header(skb); @@ -2749,6 +2758,7 @@ static int tpacket_snd(struct packet_sock *po, struct msghdr *msg) struct virtio_net_hdr vnet_hdr; bool has_vnet_hdr = false; struct sockcm_cookie sockc; + struct tpacket_uarg *uarg; __be16 proto; int err, reserve = 0; void *ph; @@ -2876,7 +2886,7 @@ static int tpacket_snd(struct packet_sock *po, struct msghdr *msg) err = len_sum; goto out_status; } - tp_len = tpacket_fill_skb(po, skb, ph, dev, data, tp_len, proto, + tp_len = tpacket_fill_skb(po, skb, dev, data, tp_len, proto, addr, hlen, copylen, hard_header_len, &sockc); if (likely(tp_len >= 0) && @@ -2908,7 +2918,24 @@ static int tpacket_snd(struct packet_sock *po, struct msghdr *msg) virtio_net_hdr_set_proto(skb, &vnet_hdr); } - skb->destructor = tpacket_destruct_skb; + uarg = kmalloc(sizeof(*uarg), GFP_KERNEL); + if (unlikely(!uarg)) { + if (likely(len_sum > 0)) + err = len_sum; + else + err = -ENOMEM; + goto out_status; + } + uarg->po = po; + uarg->ph = ph; + uarg->ubuf.ops = &tpacket_ubuf_ops; + uarg->ubuf.flags = SKBFL_ZEROCOPY_FRAG; + refcount_set(&uarg->ubuf.refcnt, 1); + + /* Hold a sk_wmem_alloc reference until completion */ + refcount_inc(&po->sk.sk_wmem_alloc); + skb_zcopy_init(skb, &uarg->ubuf); + __packet_set_status(po, ph, TP_STATUS_SENDING); packet_inc_pending(&po->tx_ring); @@ -4486,21 +4513,20 @@ static struct pgv *alloc_pg_vec(struct tpacket_req *req, int order, bool tx_ring vec->len = block_nr; pg_vec = vec->pg_vec; + if (tx_ring) { + vec->deferred = kzalloc_obj(*vec->deferred, + GFP_KERNEL | __GFP_NOWARN); + if (!vec->deferred) + goto out_free_pgvec; + vec->deferred->vec = vec; + INIT_DELAYED_WORK(&vec->deferred->work, + packet_free_pg_vec_work); + } + for (i = 0; i < block_nr; i++) { pg_vec[i].buffer = alloc_one_pg_vec_page(order); if (unlikely(!pg_vec[i].buffer)) goto out_free_pgvec; - - if (tx_ring && !vec->deferred && - is_vmalloc_addr(pg_vec[i].buffer)) { - vec->deferred = kzalloc_obj(*vec->deferred, - GFP_KERNEL | __GFP_NOWARN); - if (!vec->deferred) - goto out_free_pgvec; - vec->deferred->vec = vec; - INIT_DELAYED_WORK(&vec->deferred->work, - packet_free_pg_vec_work); - } } out: From cae23ae3f7887a1cf8a75da38edcebeef695040f Mon Sep 17 00:00:00 2001 From: Xin Long Date: Mon, 21 Sep 2026 14:03:45 -0400 Subject: [PATCH 491/592] sctp: hold asoc or transport before mod_timer() in timer handlers Take the association or transport reference before rearming a timer in the timer handlers. The existing code calls mod_timer() before taking the reference needed by the rearmed timer without holding the sock lock. This creates a race with timer cleanup: if the timer is deleted after mod_timer() returns but before the reference is taken, the cleanup path can drop the timer's reference and destroy the transport or association. The timer handler then takes a reference on the already freed object and eventually drops it, causing a refcount underflow. Hold the object before mod_timer() and drop the reference if mod_timer() reports that the timer was already pending in timer handlers. Apply the same ordering to the proto-unreachable path, which can rearm a transport timer outside the timer handlers without holding the sock lock. Fixes: 1da177e4c3f4 ("Linux-2.6.12-rc2") Reported-by: Tangxin Xie Signed-off-by: Xin Long Link: https://patch.msgid.link/c31b5e3ee2b7274e804f5eba2f21e2412e7eef7a.1790013825.git.lucien.xin@gmail.com Signed-off-by: Jakub Kicinski --- net/sctp/input.c | 7 ++++--- net/sctp/sm_sideeffect.c | 37 ++++++++++++++++++++++--------------- 2 files changed, 26 insertions(+), 18 deletions(-) diff --git a/net/sctp/input.c b/net/sctp/input.c index 864741fae418..9494cfa51106 100644 --- a/net/sctp/input.c +++ b/net/sctp/input.c @@ -436,9 +436,10 @@ void sctp_icmp_proto_unreachable(struct sock *sk, if (timer_pending(&t->proto_unreach_timer)) return; else { - if (!mod_timer(&t->proto_unreach_timer, - jiffies + (HZ/20))) - sctp_transport_hold(t); + sctp_transport_hold(t); + if (mod_timer(&t->proto_unreach_timer, + jiffies + (HZ / 20))) + sctp_transport_put(t); } } else { struct net *net = sock_net(sk); diff --git a/net/sctp/sm_sideeffect.c b/net/sctp/sm_sideeffect.c index 0d99b7e8c082..35f540fb15fc 100644 --- a/net/sctp/sm_sideeffect.c +++ b/net/sctp/sm_sideeffect.c @@ -244,8 +244,9 @@ void sctp_generate_t3_rtx_event(struct timer_list *t) pr_debug("%s: sock is busy\n", __func__); /* Try again later. */ - if (!mod_timer(&transport->T3_rtx_timer, jiffies + (HZ/20))) - sctp_transport_hold(transport); + sctp_transport_hold(transport); + if (mod_timer(&transport->T3_rtx_timer, jiffies + (HZ / 20))) + sctp_transport_put(transport); goto out_unlock; } @@ -280,8 +281,9 @@ static void sctp_generate_timeout_event(struct sctp_association *asoc, timeout_type); /* Try again later. */ - if (!mod_timer(&asoc->timers[timeout_type], jiffies + (HZ/20))) - sctp_association_hold(asoc); + sctp_association_hold(asoc); + if (mod_timer(&asoc->timers[timeout_type], jiffies + (HZ / 20))) + sctp_association_put(asoc); goto out_unlock; } @@ -378,8 +380,9 @@ void sctp_generate_heartbeat_event(struct timer_list *t) pr_debug("%s: sock is busy\n", __func__); /* Try again later. */ - if (!mod_timer(&transport->hb_timer, jiffies + (HZ/20))) - sctp_transport_hold(transport); + sctp_transport_hold(transport); + if (mod_timer(&transport->hb_timer, jiffies + (HZ / 20))) + sctp_transport_put(transport); goto out_unlock; } @@ -388,8 +391,9 @@ void sctp_generate_heartbeat_event(struct timer_list *t) timeout = sctp_transport_timeout(transport); if (elapsed < timeout) { elapsed = timeout - elapsed; - if (!mod_timer(&transport->hb_timer, jiffies + elapsed)) - sctp_transport_hold(transport); + sctp_transport_hold(transport); + if (mod_timer(&transport->hb_timer, jiffies + elapsed)) + sctp_transport_put(transport); goto out_unlock; } @@ -422,9 +426,10 @@ void sctp_generate_proto_unreach_event(struct timer_list *t) pr_debug("%s: sock is busy\n", __func__); /* Try again later. */ - if (!mod_timer(&transport->proto_unreach_timer, - jiffies + (HZ/20))) - sctp_transport_hold(transport); + sctp_transport_hold(transport); + if (mod_timer(&transport->proto_unreach_timer, + jiffies + (HZ / 20))) + sctp_transport_put(transport); goto out_unlock; } @@ -458,8 +463,9 @@ void sctp_generate_reconf_event(struct timer_list *t) pr_debug("%s: sock is busy\n", __func__); /* Try again later. */ - if (!mod_timer(&transport->reconf_timer, jiffies + (HZ / 20))) - sctp_transport_hold(transport); + sctp_transport_hold(transport); + if (mod_timer(&transport->reconf_timer, jiffies + (HZ / 20))) + sctp_transport_put(transport); goto out_unlock; } @@ -495,8 +501,9 @@ void sctp_generate_probe_event(struct timer_list *t) pr_debug("%s: sock is busy\n", __func__); /* Try again later. */ - if (!mod_timer(&transport->probe_timer, jiffies + (HZ / 20))) - sctp_transport_hold(transport); + sctp_transport_hold(transport); + if (mod_timer(&transport->probe_timer, jiffies + (HZ / 20))) + sctp_transport_put(transport); goto out_unlock; } From 73b3fc67a493e9b9a8e94a38839f6638d12fe7a6 Mon Sep 17 00:00:00 2001 From: Mina Almasry Date: Mon, 21 Sep 2026 19:54:12 +0000 Subject: [PATCH 492/592] net: devmem: document that bind-tx is unprivileged by design Unlike bind-rx, which configures shared NIC RX queues to steer incoming traffic into the caller's dmabuf and requires CAP_NET_ADMIN (uns-admin-perm), bind-tx only DMA-maps the caller's dmabuf so the caller can transmit from it on their own sockets without affecting other traffic or device configuration. Add a comment in netdev.yaml and above netdev_nl_bind_tx_doit() to make it explicit that NETDEV_CMD_BIND_TX is unprivileged by design. Signed-off-by: Mina Almasry Acked-by: Stanislav Fomichev Acked-by: Daniel Borkmann Link: https://patch.msgid.link/20260921195545.493253-1-almasrymina@google.com Signed-off-by: Jakub Kicinski --- Documentation/netlink/specs/netdev.yaml | 2 ++ net/core/netdev-genl.c | 6 ++++++ 2 files changed, 8 insertions(+) diff --git a/Documentation/netlink/specs/netdev.yaml b/Documentation/netlink/specs/netdev.yaml index 3e3f03bd5c29..0562ba1f567a 100644 --- a/Documentation/netlink/specs/netdev.yaml +++ b/Documentation/netlink/specs/netdev.yaml @@ -851,6 +851,8 @@ operations: name: bind-tx doc: Bind dmabuf to netdev for TX attribute-set: dmabuf + # Intentionally unprivileged (no admin-perm / uns-admin-perm); see + # comment above netdev_nl_bind_tx_doit(). do: request: attributes: diff --git a/net/core/netdev-genl.c b/net/core/netdev-genl.c index cb18db681640..33b9f4eb9565 100644 --- a/net/core/netdev-genl.c +++ b/net/core/netdev-genl.c @@ -1168,6 +1168,12 @@ netdev_find_netmem_tx_dev(struct net_device *dev) return NULL; } +/* Note: NETDEV_CMD_BIND_TX is intentionally unprivileged (no + * GENL_ADMIN_PERM / GENL_UNS_ADMIN_PERM). Unlike bind-rx, which configures + * shared NIC RX queues, bind-tx only DMA-maps the caller's dmabuf so they can + * transmit from it on their own sockets without affecting other traffic or + * device state. + */ int netdev_nl_bind_tx_doit(struct sk_buff *skb, struct genl_info *info) { struct net_devmem_dmabuf_binding *binding; From d68acbf93531abdb5b02b21994cd4c15a3c95b42 Mon Sep 17 00:00:00 2001 From: Maxime Chevallier Date: Thu, 17 Sep 2026 23:53:32 +0200 Subject: [PATCH 493/592] net: stmmac: selftests: Support running selftests on DSA conduits Most stmmac selftests rely on dev_add_pack() to add custom handlers, that validate the packets sent to ourselves through MAC loopback. However, when the stmmac-driven interface is a DSA CPU conduit, all frames that are received have ETH_P_XDSA as a protocol, even though they don't actually contain any tag as they come from the loopback and not the switch. This will prevent any incoming packet to match our packet handlers. Let's register a ETH_P_ALL packet handler when we detect that we're a DSA conduit, and use a proxy packet handler to filter the h_proto. As this allows external frames to be received through our .func(), the packet handler is added after the dev->addr field is populated in our selftest attributes. Note that we may still receive incoming packets from the switch, but these frames shouldn't interfere with the very specific frames used for selftests, and stmmac selftests in general aren't safe against external traffic interferences. This was validated on a WPQ864 devkit for IPQ8064, that has the SoC connected to a QCA8k switch. The ARP offload's packet handler is left alone, this feature is just not implemented in stmmac and due for removal. Fixes: 091810dbded9 ("net: stmmac: Introduce selftests support") Reviewed-by: Nicolai Buchwitz Signed-off-by: Maxime Chevallier Link: https://patch.msgid.link/20260917215339.2022523-2-maxime.chevallier@bootlin.com Signed-off-by: Jakub Kicinski --- .../stmicro/stmmac/stmmac_selftests.c | 89 +++++++++++++++---- 1 file changed, 71 insertions(+), 18 deletions(-) diff --git a/drivers/net/ethernet/stmicro/stmmac/stmmac_selftests.c b/drivers/net/ethernet/stmicro/stmmac/stmmac_selftests.c index 6372ec7c3f31..614b5995dec5 100644 --- a/drivers/net/ethernet/stmicro/stmmac/stmmac_selftests.c +++ b/drivers/net/ethernet/stmicro/stmmac/stmmac_selftests.c @@ -12,6 +12,7 @@ #include #include #include +#include #include #include #include @@ -237,6 +238,9 @@ struct stmmac_test_priv { struct stmmac_packet_attrs *packet; struct packet_type pt; struct completion comp; + __be16 packet_type; + int (*func)(struct sk_buff *skb, struct net_device *ndev, + struct packet_type *pt, struct net_device *orig_ndev); int double_vlan; int vlan_id; int ok; @@ -316,6 +320,50 @@ static int stmmac_test_loopback_validate(struct sk_buff *skb, return 0; } +static int stmmac_sft_filter(struct sk_buff *skb, struct net_device *ndev, + struct packet_type *pt, + struct net_device *orig_ndev) +{ + struct stmmac_test_priv *tpriv = pt->af_packet_priv; + struct ethhdr *hdr = eth_hdr(skb); + int ret = 0; + + if (hdr->h_proto == tpriv->packet_type) { + struct sk_buff *nskb = skb_clone(skb, GFP_ATOMIC); + + if (nskb) + ret = tpriv->func(nskb, ndev, pt, orig_ndev); + } + + kfree_skb(skb); + return ret; +} + +static void stmmac_sft_add_pack(struct packet_type *pt) +{ + struct stmmac_test_priv *tpriv = pt->af_packet_priv; + + if (netdev_uses_dsa(tpriv->pt.dev)) { + tpriv->packet_type = tpriv->pt.type; + tpriv->func = tpriv->pt.func; + + /* DSA conduit will report ETH_P_XDSA, so our packet handler + * won't match. Let's register a ETH_P_ALL match and filter + * manually in stmmac_sft_filter. + */ + tpriv->pt.type = htons(ETH_P_ALL); + tpriv->pt.func = stmmac_sft_filter; + tpriv->pt.ignore_outgoing = true; + } + + dev_add_pack(pt); +} + +static void stmmac_sft_remove_pack(struct packet_type *pt) +{ + dev_remove_pack(pt); +} + static int __stmmac_test_loopback(struct stmmac_priv *priv, struct stmmac_packet_attrs *attr) { @@ -337,7 +385,7 @@ static int __stmmac_test_loopback(struct stmmac_priv *priv, tpriv->packet = attr; if (!attr->dont_wait) - dev_add_pack(&tpriv->pt); + stmmac_sft_add_pack(&tpriv->pt); skb = stmmac_test_get_udp_skb(priv, attr); if (!skb) { @@ -360,7 +408,7 @@ static int __stmmac_test_loopback(struct stmmac_priv *priv, cleanup: if (!attr->dont_wait) - dev_remove_pack(&tpriv->pt); + stmmac_sft_remove_pack(&tpriv->pt); kfree(tpriv); return ret; } @@ -767,7 +815,7 @@ static int stmmac_test_flowctrl(struct stmmac_priv *priv) tpriv->pt.func = stmmac_test_flowctrl_validate; tpriv->pt.dev = priv->dev; tpriv->pt.af_packet_priv = tpriv; - dev_add_pack(&tpriv->pt); + stmmac_sft_add_pack(&tpriv->pt); /* Compute minimum number of packets to make FIFO full */ pkt_count = rx_fifo_size; @@ -823,7 +871,7 @@ static int stmmac_test_flowctrl(struct stmmac_priv *priv) cleanup: dev_mc_del(priv->dev, paddr); dev_set_promiscuity(priv->dev, -1); - dev_remove_pack(&tpriv->pt); + stmmac_sft_remove_pack(&tpriv->pt); kfree(tpriv); return ret; } @@ -928,18 +976,20 @@ static int __stmmac_test_vlanfilt(struct stmmac_priv *priv) * HASH values. */ tpriv->vlan_id = 0x123; - dev_add_pack(&tpriv->pt); ret = vlan_vid_add(priv->dev, htons(ETH_P_8021Q), tpriv->vlan_id); if (ret) goto cleanup; + attr.vlan = 1; + attr.dst = priv->dev->dev_addr; + attr.sport = 9; + attr.dport = 9; + + stmmac_sft_add_pack(&tpriv->pt); + for (i = 0; i < 4; i++) { - attr.vlan = 1; attr.vlan_id_out = tpriv->vlan_id + i; - attr.dst = priv->dev->dev_addr; - attr.sport = 9; - attr.dport = 9; skb = stmmac_test_get_udp_skb(priv, &attr); if (!skb) { @@ -966,9 +1016,9 @@ static int __stmmac_test_vlanfilt(struct stmmac_priv *priv) } vlan_del: + stmmac_sft_remove_pack(&tpriv->pt); vlan_vid_del(priv->dev, htons(ETH_P_8021Q), tpriv->vlan_id); cleanup: - dev_remove_pack(&tpriv->pt); kfree(tpriv); return ret; } @@ -1022,18 +1072,20 @@ static int __stmmac_test_dvlanfilt(struct stmmac_priv *priv) * HASH values. */ tpriv->vlan_id = 0x123; - dev_add_pack(&tpriv->pt); ret = vlan_vid_add(priv->dev, htons(ETH_P_8021AD), tpriv->vlan_id); if (ret) goto cleanup; + attr.vlan = 2; + attr.dst = priv->dev->dev_addr; + attr.sport = 9; + attr.dport = 9; + + stmmac_sft_add_pack(&tpriv->pt); + for (i = 0; i < 4; i++) { - attr.vlan = 2; attr.vlan_id_out = tpriv->vlan_id + i; - attr.dst = priv->dev->dev_addr; - attr.sport = 9; - attr.dport = 9; skb = stmmac_test_get_udp_skb(priv, &attr); if (!skb) { @@ -1060,9 +1112,9 @@ static int __stmmac_test_dvlanfilt(struct stmmac_priv *priv) } vlan_del: + stmmac_sft_remove_pack(&tpriv->pt); vlan_vid_del(priv->dev, htons(ETH_P_8021AD), tpriv->vlan_id); cleanup: - dev_remove_pack(&tpriv->pt); kfree(tpriv); return ret; } @@ -1293,7 +1345,6 @@ static int stmmac_test_vlanoff_common(struct stmmac_priv *priv, bool svlan) tpriv->pt.af_packet_priv = tpriv; tpriv->packet = &attr; tpriv->vlan_id = 0x123; - dev_add_pack(&tpriv->pt); ret = vlan_vid_add(priv->dev, htons(proto), tpriv->vlan_id); if (ret) @@ -1301,6 +1352,8 @@ static int stmmac_test_vlanoff_common(struct stmmac_priv *priv, bool svlan) attr.dst = priv->dev->dev_addr; + stmmac_sft_add_pack(&tpriv->pt); + skb = stmmac_test_get_udp_skb(priv, &attr); if (!skb) { ret = -ENOMEM; @@ -1318,9 +1371,9 @@ static int stmmac_test_vlanoff_common(struct stmmac_priv *priv, bool svlan) ret = tpriv->ok ? 0 : -ETIMEDOUT; vlan_del: + stmmac_sft_remove_pack(&tpriv->pt); vlan_vid_del(priv->dev, htons(proto), tpriv->vlan_id); cleanup: - dev_remove_pack(&tpriv->pt); kfree(tpriv); return ret; } From c8c1795aa8106293020a836d01e127e98f442925 Mon Sep 17 00:00:00 2001 From: Maxime Chevallier Date: Thu, 17 Sep 2026 23:53:33 +0200 Subject: [PATCH 494/592] net: stmmac: selftests: Validate EEE based on the actual LPI timer value The EEE selftest is a 2-step test : - It validates that we enter in LPI mode with the irq_tx_path_in_lpi_mode_n counter - It then validates that we exit LPI when sending a frame, with the irq_tx_path_exit_lpi_mode_n counter. The current state of the test lacks 2 main things : - We don't know exactly when was the previous frame sent (it's from the previous selftest) - The timeout is hardcoded, while the LPI is entered after a user-configurable delay. On top of that, the timeout loop uses a pre-decrement iterator (--retries) that actually only iterate nine times, so 900ms while the default LPI value is 1 second. Let's therefore make it more deterministic : - Send a frame at the beginning of the test - Wait for more than the lpi timer value, we timeout after about twice the value, - Then send another frame, and verify that we do go out of LPI, also with a timeout. As LPI timer can get pretty high, bail out if LPI timer is over 5 seconds. Note that the test's goal isn't to validate the LPI timer value itself, only that we enter/leave LPI mode. Fixes: 091810dbded9 ("net: stmmac: Introduce selftests support") Reviewed-by: Nicolai Buchwitz Signed-off-by: Maxime Chevallier Link: https://patch.msgid.link/20260917215339.2022523-3-maxime.chevallier@bootlin.com Signed-off-by: Jakub Kicinski --- .../stmicro/stmmac/stmmac_selftests.c | 44 ++++++++++++++++--- 1 file changed, 37 insertions(+), 7 deletions(-) diff --git a/drivers/net/ethernet/stmicro/stmmac/stmmac_selftests.c b/drivers/net/ethernet/stmicro/stmmac/stmmac_selftests.c index 614b5995dec5..2f9f7746c40a 100644 --- a/drivers/net/ethernet/stmicro/stmmac/stmmac_selftests.c +++ b/drivers/net/ethernet/stmicro/stmmac/stmmac_selftests.c @@ -30,6 +30,7 @@ struct stmmachdr { sizeof(struct stmmachdr)) #define STMMAC_TEST_PKT_MAGIC 0xdeadcafecafedeadULL #define STMMAC_LB_TIMEOUT msecs_to_jiffies(200) +#define STMMAC_SFT_MAX_LPI (5 * USEC_PER_SEC) struct stmmac_packet_attrs { int vlan; @@ -462,12 +463,16 @@ static int stmmac_test_mmc(struct stmmac_priv *priv) static int stmmac_test_eee(struct stmmac_priv *priv) { struct stmmac_extra_stats *initial, *final; - int retries = 10; + unsigned long timeout, max_duration; int ret; if (!priv->dma_cap.eee || !priv->eee_active) return -EOPNOTSUPP; + /* Bail out if the configured LPI timer is too long */ + if (priv->tx_lpi_timer > STMMAC_SFT_MAX_LPI) + return -EOPNOTSUPP; + initial = kzalloc_obj(*initial); if (!initial) return -ENOMEM; @@ -478,14 +483,21 @@ static int stmmac_test_eee(struct stmmac_priv *priv) goto out_free_initial; } + /* Snapshot stats, we want to count the in_lpi events. We may enter + * LPI just after the packet was sent. + */ memcpy(initial, &priv->xstats, sizeof(*initial)); + /* Send a frame, then wait to enter LPI */ ret = stmmac_test_mac_loopback(priv); if (ret) goto out_free_final; + max_duration = usecs_to_jiffies(2 * priv->tx_lpi_timer); + /* We have no traffic in the line so, sooner or later it will go LPI */ - while (--retries) { + timeout = jiffies + max_duration; + while (!time_after(jiffies, timeout)) { memcpy(final, &priv->xstats, sizeof(*final)); if (final->irq_tx_path_in_lpi_mode_n > @@ -494,20 +506,38 @@ static int stmmac_test_eee(struct stmmac_priv *priv) msleep(100); } - if (!retries) { + memcpy(final, &priv->xstats, sizeof(*final)); + if (final->irq_tx_path_in_lpi_mode_n <= + initial->irq_tx_path_in_lpi_mode_n) { ret = -ETIMEDOUT; goto out_free_final; } - if (final->irq_tx_path_in_lpi_mode_n <= - initial->irq_tx_path_in_lpi_mode_n) { - ret = -EINVAL; + /* Re-snapshot, as we want to measure exit_lpi events. We should be + * in LPI right now. + */ + memcpy(initial, &priv->xstats, sizeof(*initial)); + + /* TX something so we go out of LPI */ + ret = stmmac_test_mac_loopback(priv); + if (ret) goto out_free_final; + + /* Wait for the exit LPI interrupt */ + timeout = jiffies + max_duration; + while (!time_after(jiffies, timeout)) { + memcpy(final, &priv->xstats, sizeof(*final)); + + if (final->irq_tx_path_exit_lpi_mode_n > + initial->irq_tx_path_exit_lpi_mode_n) + break; + msleep(100); } + memcpy(final, &priv->xstats, sizeof(*final)); if (final->irq_tx_path_exit_lpi_mode_n <= initial->irq_tx_path_exit_lpi_mode_n) { - ret = -EINVAL; + ret = -ETIMEDOUT; goto out_free_final; } From ba804b23d76d278ee475b8427fa7c5623ce5e270 Mon Sep 17 00:00:00 2001 From: Maxime Chevallier Date: Thu, 17 Sep 2026 23:53:34 +0200 Subject: [PATCH 495/592] net: stmmac: selftests: Check the dev->features for S-TAG offload testing The S-TAG offload insertion incorrectly checks the dvlan (double vlan) DMA cap, which is different than S-TAG support. Use NETIF_F_HW_VLAN_STAG_TX to check if the feature is supported instead. Note that this flag isn't set in stmmac yet, but contrary to ARP offload, this is a feature that has a chance to get there eventually so let's leave the selftest here for now. It'll report -EOPNOTSUPP in the meantime. Fixes: 091810dbded9 ("net: stmmac: Introduce selftests support") Reviewed-by: Nicolai Buchwitz Signed-off-by: Maxime Chevallier Link: https://patch.msgid.link/20260917215339.2022523-4-maxime.chevallier@bootlin.com Signed-off-by: Jakub Kicinski --- drivers/net/ethernet/stmicro/stmmac/stmmac_selftests.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/drivers/net/ethernet/stmicro/stmmac/stmmac_selftests.c b/drivers/net/ethernet/stmicro/stmmac/stmmac_selftests.c index 2f9f7746c40a..de02c0da56dc 100644 --- a/drivers/net/ethernet/stmicro/stmmac/stmmac_selftests.c +++ b/drivers/net/ethernet/stmicro/stmmac/stmmac_selftests.c @@ -1415,7 +1415,7 @@ static int stmmac_test_vlanoff(struct stmmac_priv *priv) static int stmmac_test_svlanoff(struct stmmac_priv *priv) { - if (!priv->dma_cap.dvlan) + if (!(priv->dev->features & NETIF_F_HW_VLAN_STAG_TX)) return -EOPNOTSUPP; return stmmac_test_vlanoff_common(priv, true); } From 960db6f65788c21249ea04a947d5c01e38d19294 Mon Sep 17 00:00:00 2001 From: Maxime Chevallier Date: Thu, 17 Sep 2026 23:53:35 +0200 Subject: [PATCH 496/592] net: stmmac: selftests: Capture all packets for vlan checks While we use vlan_vid_add to trigger the tag filtering machinery in the driver, there's no netdev associated to the VLAN. This causes the skb to arrive with empty skb->vlan_tci fields, as the packet is marked OTHERHOST in __netif_receive_skb_core(), and we fail our validation. Let's use the proxy mechanism introduced for DSA, that registers a ETH_P_ALL packet handler that runs earlier, before the vlan netdev lookup, then filters for the correct ethertype before passing an skb clone to our validation function. As we may receive external frames with the right tag from the outside, let's move the address check in the vlan validation function earlier. Fixes: 091810dbded9 ("net: stmmac: Introduce selftests support") Reviewed-by: Nicolai Buchwitz Signed-off-by: Maxime Chevallier Link: https://patch.msgid.link/20260917215339.2022523-5-maxime.chevallier@bootlin.com Signed-off-by: Jakub Kicinski --- .../stmicro/stmmac/stmmac_selftests.c | 19 +++++++++++++------ 1 file changed, 13 insertions(+), 6 deletions(-) diff --git a/drivers/net/ethernet/stmicro/stmmac/stmmac_selftests.c b/drivers/net/ethernet/stmicro/stmmac/stmmac_selftests.c index de02c0da56dc..43b8411c5112 100644 --- a/drivers/net/ethernet/stmicro/stmmac/stmmac_selftests.c +++ b/drivers/net/ethernet/stmicro/stmmac/stmmac_selftests.c @@ -242,6 +242,7 @@ struct stmmac_test_priv { __be16 packet_type; int (*func)(struct sk_buff *skb, struct net_device *ndev, struct packet_type *pt, struct net_device *orig_ndev); + bool capture_all; int double_vlan; int vlan_id; int ok; @@ -344,13 +345,15 @@ static void stmmac_sft_add_pack(struct packet_type *pt) { struct stmmac_test_priv *tpriv = pt->af_packet_priv; - if (netdev_uses_dsa(tpriv->pt.dev)) { + if (netdev_uses_dsa(tpriv->pt.dev) || tpriv->capture_all) { tpriv->packet_type = tpriv->pt.type; tpriv->func = tpriv->pt.func; /* DSA conduit will report ETH_P_XDSA, so our packet handler * won't match. Let's register a ETH_P_ALL match and filter - * manually in stmmac_sft_filter. + * manually in stmmac_sft_filter. This is also useful for + * VLAN tests, to capture packets otherwise marked as + * OTHERHOST. */ tpriv->pt.type = htons(ETH_P_ALL); tpriv->pt.func = stmmac_sft_filter; @@ -943,6 +946,11 @@ static int stmmac_test_vlan_validate(struct sk_buff *skb, goto out; if (skb_headlen(skb) < (STMMAC_TEST_PKT_SIZE - ETH_HLEN)) goto out; + + ehdr = (struct ethhdr *)skb_mac_header(skb); + if (!ether_addr_equal_unaligned(ehdr->h_dest, tpriv->packet->dst)) + goto out; + if (tpriv->vlan_id) { if (skb->vlan_proto != htons(proto)) goto out; @@ -954,10 +962,6 @@ static int stmmac_test_vlan_validate(struct sk_buff *skb, } } - ehdr = (struct ethhdr *)skb_mac_header(skb); - if (!ether_addr_equal_unaligned(ehdr->h_dest, tpriv->packet->dst)) - goto out; - ihdr = ip_hdr(skb); if (tpriv->double_vlan) ihdr = (struct iphdr *)(skb_network_header(skb) + 4); @@ -999,6 +1003,7 @@ static int __stmmac_test_vlanfilt(struct stmmac_priv *priv) tpriv->pt.dev = priv->dev; tpriv->pt.af_packet_priv = tpriv; tpriv->packet = &attr; + tpriv->capture_all = true; /* * As we use HASH filtering, false positives may appear. This is a @@ -1095,6 +1100,7 @@ static int __stmmac_test_dvlanfilt(struct stmmac_priv *priv) tpriv->pt.dev = priv->dev; tpriv->pt.af_packet_priv = tpriv; tpriv->packet = &attr; + tpriv->capture_all = true; /* * As we use HASH filtering, false positives may appear. This is a @@ -1375,6 +1381,7 @@ static int stmmac_test_vlanoff_common(struct stmmac_priv *priv, bool svlan) tpriv->pt.af_packet_priv = tpriv; tpriv->packet = &attr; tpriv->vlan_id = 0x123; + tpriv->capture_all = true; ret = vlan_vid_add(priv->dev, htons(proto), tpriv->vlan_id); if (ret) From b42e7012773a0e81e97a2dda6ef907f5147a6658 Mon Sep 17 00:00:00 2001 From: Maxime Chevallier Date: Thu, 17 Sep 2026 23:53:36 +0200 Subject: [PATCH 497/592] net: stmmac: dwmac4: Use the correct bufzise when the len is exactly 8K DMA bufsize selection isn't made on the MTU but the actual frame length, so including the L2 header. On DWMAC4, if the len is exactly BUF_SIZE_8KiB, the next larger size is incorrectly selected. Lets fix the comparison and while at it, rename the parameter from len to mtu. Fixes: c3efed5ad1b0 ("net: stmmac: Enable dwmac4 jumbo frame more than 8KiB"). Signed-off-by: Maxime Chevallier Reviewed-by: Nicolai Buchwitz Link: https://patch.msgid.link/20260917215339.2022523-6-maxime.chevallier@bootlin.com Signed-off-by: Jakub Kicinski --- drivers/net/ethernet/stmicro/stmmac/dwmac4_descs.c | 4 ++-- drivers/net/ethernet/stmicro/stmmac/hwif.h | 2 +- drivers/net/ethernet/stmicro/stmmac/ring_mode.c | 4 ++-- 3 files changed, 5 insertions(+), 5 deletions(-) diff --git a/drivers/net/ethernet/stmicro/stmmac/dwmac4_descs.c b/drivers/net/ethernet/stmicro/stmmac/dwmac4_descs.c index 2994df41ec2c..c6a8f8d73501 100644 --- a/drivers/net/ethernet/stmicro/stmmac/dwmac4_descs.c +++ b/drivers/net/ethernet/stmicro/stmmac/dwmac4_descs.c @@ -474,11 +474,11 @@ static void dwmac4_set_sarc(struct dma_desc *p, u32 sarc_type) sarc_type)); } -static int set_16kib_bfsize(int mtu) +static int set_16kib_bfsize(int len) { int ret = 0; - if (unlikely(mtu >= BUF_SIZE_8KiB)) + if (unlikely(len > BUF_SIZE_8KiB)) ret = BUF_SIZE_16KiB; return ret; } diff --git a/drivers/net/ethernet/stmicro/stmmac/hwif.h b/drivers/net/ethernet/stmicro/stmmac/hwif.h index 9314bcb85c22..857f7562c6c6 100644 --- a/drivers/net/ethernet/stmicro/stmmac/hwif.h +++ b/drivers/net/ethernet/stmicro/stmmac/hwif.h @@ -540,7 +540,7 @@ struct stmmac_mode_ops { bool (*is_jumbo_frm)(unsigned int len, bool enh_desc); int (*jumbo_frm)(struct stmmac_tx_queue *tx_q, struct sk_buff *skb, int csum); - int (*set_16kib_bfsize)(int mtu); + int (*set_16kib_bfsize)(int len); void (*init_desc3)(struct dma_desc *p); void (*refill_desc3)(struct stmmac_rx_queue *rx_q, struct dma_desc *p); void (*clean_desc3)(struct stmmac_tx_queue *tx_q, struct dma_desc *p); diff --git a/drivers/net/ethernet/stmicro/stmmac/ring_mode.c b/drivers/net/ethernet/stmicro/stmmac/ring_mode.c index f7949419eb9f..d2f0c321661d 100644 --- a/drivers/net/ethernet/stmicro/stmmac/ring_mode.c +++ b/drivers/net/ethernet/stmicro/stmmac/ring_mode.c @@ -124,10 +124,10 @@ static void clean_desc3(struct stmmac_tx_queue *tx_q, struct dma_desc *p) p->des3 = 0; } -static int set_16kib_bfsize(int mtu) +static int set_16kib_bfsize(int len) { int ret = 0; - if (unlikely(mtu > BUF_SIZE_8KiB)) + if (unlikely(len > BUF_SIZE_8KiB)) ret = BUF_SIZE_16KiB; return ret; } From b8a26d46c0a4254f8bfe143681adb9d18d020299 Mon Sep 17 00:00:00 2001 From: Maxime Chevallier Date: Thu, 17 Sep 2026 23:53:37 +0200 Subject: [PATCH 498/592] net: stmmac: size the RX buffers from the frame length, not the MTU When picking the buffsize to use based on the MTU, we shouldn't check only the MTU value, but also : - ETH_HLEN for the L2 header, - up to 2 VLAN tags, - the FCS, The default bufsize is 1536 bytes, which is enough to contain all the above so this hasn't surfaced before, but the addition of NET_IP_ALIGN to the start of buffer address tripped the Jumbo selftest, leading to this discovery. With that, we don't need the '>=' checks on the buffer len, we can use more consistent comparison operators in stmmac_set_bfsize. Fixes: 286a83721720 ("stmmac: add CHAINED descriptor mode support (V4)") Reviewed-by: Nicolai Buchwitz Signed-off-by: Maxime Chevallier Link: https://patch.msgid.link/20260917215339.2022523-7-maxime.chevallier@bootlin.com Signed-off-by: Jakub Kicinski --- .../net/ethernet/stmicro/stmmac/stmmac_main.c | 20 ++++++++++--------- 1 file changed, 11 insertions(+), 9 deletions(-) diff --git a/drivers/net/ethernet/stmicro/stmmac/stmmac_main.c b/drivers/net/ethernet/stmicro/stmmac/stmmac_main.c index 1fb5f804ea23..d5a984ad864f 100644 --- a/drivers/net/ethernet/stmicro/stmmac/stmmac_main.c +++ b/drivers/net/ethernet/stmicro/stmmac/stmmac_main.c @@ -1536,17 +1536,17 @@ static unsigned int stmmac_rx_offset(struct stmmac_priv *priv) return NET_SKB_PAD + NET_IP_ALIGN; } -static int stmmac_set_bfsize(int mtu) +static int stmmac_set_bfsize(int len) { int ret; - if (mtu >= BUF_SIZE_8KiB) + if (len > BUF_SIZE_8KiB) ret = BUF_SIZE_16KiB; - else if (mtu >= BUF_SIZE_4KiB) + else if (len > BUF_SIZE_4KiB) ret = BUF_SIZE_8KiB; - else if (mtu >= BUF_SIZE_2KiB) + else if (len > BUF_SIZE_2KiB) ret = BUF_SIZE_4KiB; - else if (mtu > DEFAULT_BUFSIZE) + else if (len > DEFAULT_BUFSIZE) ret = BUF_SIZE_2KiB; else ret = DEFAULT_BUFSIZE; @@ -4063,7 +4063,7 @@ static struct stmmac_dma_conf * stmmac_setup_dma_desc(struct stmmac_priv *priv, unsigned int mtu) { struct stmmac_dma_conf *dma_conf; - int bfsize, ret; + int bfsize, len, ret; u8 chan; dma_conf = kzalloc_obj(*dma_conf); @@ -4073,13 +4073,15 @@ stmmac_setup_dma_desc(struct stmmac_priv *priv, unsigned int mtu) return ERR_PTR(-ENOMEM); } - /* Returns 0 or BUF_SIZE_16KiB if mtu > 8KiB and dwmac4 or ring mode */ - bfsize = stmmac_set_16kib_bfsize(priv, mtu); + len = mtu + ETH_HLEN + 2 * VLAN_HLEN + ETH_FCS_LEN; + + /* Returns 0 or BUF_SIZE_16KiB if len > 8KiB and dwmac4 or ring mode */ + bfsize = stmmac_set_16kib_bfsize(priv, len); if (bfsize < 0) bfsize = 0; if (bfsize < BUF_SIZE_16KiB) - bfsize = stmmac_set_bfsize(mtu); + bfsize = stmmac_set_bfsize(len); dma_conf->dma_buf_sz = bfsize; /* Chose the tx/rx size from the already defined one in the From c4ac6e94eb9423126bda907a7f2933284f7450ff Mon Sep 17 00:00:00 2001 From: Maxime Chevallier Date: Thu, 17 Sep 2026 23:53:38 +0200 Subject: [PATCH 499/592] net: stmmac: selftests: Account for alignment shift on dwmac1000 for Jumbo test On dwmac1000, we currently only support single-descriptor frames. The Jumbo test started failing when NET_IP_ALIGN was added to align the IP header, as this tests tries to send the biggest possible frame. On dwmac1000 the DMA transfer is aligned on 4-bytes, so adding a 2-byte shift at the start-of-buffer address means it takes a whole extra 4-byte DMA burst to receive the Jumbo packet, causing it to spill over the next descriptor. This doesn't seem to happen on dwmac4 and xgmac that appear to correctly handle unaligned xfers (only tested on dwmac4) Let's account for that in the Jumbo test, reduce the size of our big packet by the align size. Fixes: 23680bf5f8c6 ("net: stmmac: restore NET_IP_ALIGN in the RX DMA offset") Reviewed-by: Nicolai Buchwitz Signed-off-by: Maxime Chevallier Link: https://patch.msgid.link/20260917215339.2022523-8-maxime.chevallier@bootlin.com Signed-off-by: Jakub Kicinski --- drivers/net/ethernet/stmicro/stmmac/stmmac_selftests.c | 3 +++ 1 file changed, 3 insertions(+) diff --git a/drivers/net/ethernet/stmicro/stmmac/stmmac_selftests.c b/drivers/net/ethernet/stmicro/stmmac/stmmac_selftests.c index 43b8411c5112..c25dc9f89270 100644 --- a/drivers/net/ethernet/stmicro/stmmac/stmmac_selftests.c +++ b/drivers/net/ethernet/stmicro/stmmac/stmmac_selftests.c @@ -1789,6 +1789,9 @@ static int __stmmac_test_jumbo(struct stmmac_priv *priv, u16 queue) struct stmmac_packet_attrs attr = { }; int size = priv->dma_conf.dma_buf_sz; + if (!dwmac_is_xmac(priv->plat->core_type)) + size -= NET_IP_ALIGN; + attr.dst = priv->dev->dev_addr; attr.max_size = size - ETH_FCS_LEN; attr.queue_mapping = queue; From 0160953d8eec75c3c55562158c46442ff1fd410b Mon Sep 17 00:00:00 2001 From: =?UTF-8?q?Bj=C3=B6rn=20T=C3=B6pel?= Date: Fri, 18 Sep 2026 13:46:40 +0200 Subject: [PATCH 500/592] eth: fbnic: Avoid rounding zero ring sizes MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit roundup_pow_of_two() is undefined for zero. ethtool permits a zero ring size to reach the driver, where the minimum-size check should reject it. Leave zero unchanged while rounding nonzero ring sizes. The minimum-size check then rejects zero deterministically without changing the established behavior for other values. Fixes: 6cbf18a05c06 ("eth: fbnic: support ring size configuration") Reported-by: Sashiko Link: https://lore.kernel.org/netdev/178971206933.22033.236948278674126701@kernel.org/ Suggested-by: Alexander Duyck Signed-off-by: Björn Töpel Reviewed-by: Joe Damato Link: https://patch.msgid.link/20260918114641.1281172-1-bjorn@kernel.org Signed-off-by: Jakub Kicinski --- drivers/net/ethernet/meta/fbnic/fbnic_ethtool.c | 13 +++++++++---- 1 file changed, 9 insertions(+), 4 deletions(-) diff --git a/drivers/net/ethernet/meta/fbnic/fbnic_ethtool.c b/drivers/net/ethernet/meta/fbnic/fbnic_ethtool.c index 423f179c9d47..76e9a545bb16 100644 --- a/drivers/net/ethernet/meta/fbnic/fbnic_ethtool.c +++ b/drivers/net/ethernet/meta/fbnic/fbnic_ethtool.c @@ -313,6 +313,11 @@ fbnic_get_ringparam(struct net_device *netdev, struct ethtool_ringparam *ring, kernel_ring->hds_thresh = fbn->hds_thresh; } +static u32 fbnic_ring_size_pow2(u32 size) +{ + return size ? roundup_pow_of_two(size) : 0; +} + static void fbnic_set_rings(struct fbnic_net *fbn, struct ethtool_ringparam *ring, struct kernel_ethtool_ringparam *kernel_ring) @@ -334,10 +339,10 @@ fbnic_set_ringparam(struct net_device *netdev, struct ethtool_ringparam *ring, struct fbnic_net *clone; int err; - ring->rx_pending = roundup_pow_of_two(ring->rx_pending); - ring->rx_mini_pending = roundup_pow_of_two(ring->rx_mini_pending); - ring->rx_jumbo_pending = roundup_pow_of_two(ring->rx_jumbo_pending); - ring->tx_pending = roundup_pow_of_two(ring->tx_pending); + ring->rx_pending = fbnic_ring_size_pow2(ring->rx_pending); + ring->rx_mini_pending = fbnic_ring_size_pow2(ring->rx_mini_pending); + ring->rx_jumbo_pending = fbnic_ring_size_pow2(ring->rx_jumbo_pending); + ring->tx_pending = fbnic_ring_size_pow2(ring->tx_pending); /* These are absolute minimums allowing the device and driver to operate * but not necessarily guarantee reasonable performance. Settings below From 9c572a83037a7dcd653ba3a9cc468c16b857d0c9 Mon Sep 17 00:00:00 2001 From: Bernard Ladenthin Date: Fri, 18 Sep 2026 15:39:37 +0200 Subject: [PATCH 501/592] net/sched: fix potential stack infoleak in em_text_dump() em_text_dump() allocates struct tcf_em_text on the stack without zeroing it. strscpy() writes the algorithm name and a NUL terminator into conf.algo[], leaving the remaining bytes uninitialised. nla_put_nohdr() then copies the full struct to the netlink response. KMSAN on Linux 7.2-rc6 reports two kernel-infoleak splats from this path, one triggered via "tc filter show" and one via a raw RTM_GETTFILTER dump: BUG: KMSAN: kernel-infoleak in _copy_to_iter+0x1c9/0x2620 nla_put_nohdr+0x83/0x130 em_text_dump+0x291/0x550 Local variable conf created at: em_text_dump+0x5d/0x550 Bytes 168-179 of 199 are uninitialized I am not certain whether this constitutes a real security problem in practice: the test was conducted in a controlled KMSAN environment and the leaked stack bytes may or may not carry sensitive data on actual production kernels. I am reporting it because KMSAN flagged it as a kernel-infoleak and the fix is straightforward. I can provide a userspace reproducer on request. The original code used strncpy() which zero-pads to the destination size. Commit b04202d6065c ("net/sched: replace strncpy with strscpy") replaced it with strscpy(), which does not pad, creating this condition. Zero-initialising the struct closes it. Fixes: b04202d6065c ("net/sched: replace strncpy with strscpy") Link: https://lore.kernel.org/netdev/20250327143733.187438-1-richard120310@gmail.com/ Assisted-by: Claude:claude-sonnet-4-6 [KMSAN] Signed-off-by: Bernard Ladenthin Acked-by: Jamal Hadi Salim Link: https://patch.msgid.link/20260918133953.12494-1-bernard.ladenthin@gmail.com Signed-off-by: Jakub Kicinski --- net/sched/em_text.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/net/sched/em_text.c b/net/sched/em_text.c index 343f1aebeec2..4132f8c3c5fc 100644 --- a/net/sched/em_text.c +++ b/net/sched/em_text.c @@ -113,7 +113,7 @@ static void em_text_destroy(struct tcf_ematch *m) static int em_text_dump(struct sk_buff *skb, struct tcf_ematch *m) { struct text_match *tm = EM_TEXT_PRIV(m); - struct tcf_em_text conf; + struct tcf_em_text conf = {}; strscpy(conf.algo, tm->config->ops->name); conf.from_offset = tm->from_offset; From 6db1ce73e9853f533eb7f413f14ba00f8ec6f80d Mon Sep 17 00:00:00 2001 From: Weiming Shi Date: Sun, 20 Sep 2026 21:23:04 +0800 Subject: [PATCH 502/592] bpf: Reject dev-bound-only programs on other devices __bpf_offload_dev_match() falls back to comparing offdev pointers after an exact netdev mismatch. Bound-only programs normally have NULL offdevs, so unrelated netdevs compare equal. A bound-only program on an offload-registered netdev can instead inherit a real offdev and match a sibling port. With CAP_BPF and CAP_NET_ADMIN, a caller can use bpf(BPF_LINK_CREATE) with a different target ifindex to run metadata kfuncs specialized for the bound driver on the target driver's xdp_buff. Running a veth-bound program on tun reads beyond tun's bare stack xdp_buff as a veth_xdp_buff. Oops: general protection fault, probably for non-canonical address KASAN: null-ptr-deref in range [0x0000000000000010-0x0000000000000017] RIP: 0010:veth_xdp_rx_timestamp (drivers/net/veth.c:1673) Call Trace: ... tun_build_skb (drivers/net/tun.c:1739) tun_get_user (drivers/net/tun.c:1856) tun_chr_write_iter (drivers/net/tun.c:2091) vfs_write (fs/read_write.c:595 fs/read_write.c:687) ksys_write (fs/read_write.c:739) do_syscall_64 (arch/x86/entry/syscall_64.c:84) entry_SYSCALL_64_after_hwframe (arch/x86/entry/entry_64.S:121) Kernel panic - not syncing: Fatal exception in interrupt Restrict non-offloaded programs to exact netdev matches and retain the shared-offdev fallback only for genuinely offloaded multi-port programs. Fixes: 2b3486bc2d23 ("bpf: Introduce device-bound XDP programs") Reported-by: Signed-off-by: Weiming Shi Signed-off-by: Alexei Starovoitov Link: https://lore.kernel.org/bpf/20260917161335.1020405-2-bestswngs@gmail.com/ Link: https://patch.msgid.link/20260920132303.4109240-3-bestswngs@gmail.com --- kernel/bpf/offload.c | 2 ++ 1 file changed, 2 insertions(+) diff --git a/kernel/bpf/offload.c b/kernel/bpf/offload.c index 0d6f5569588c..d855399812ee 100644 --- a/kernel/bpf/offload.c +++ b/kernel/bpf/offload.c @@ -698,6 +698,8 @@ static bool __bpf_offload_dev_match(struct bpf_prog *prog, return false; if (offload->netdev == netdev) return true; + if (!bpf_prog_is_offloaded(prog->aux)) + return false; ondev1 = bpf_offload_find_netdev(offload->netdev); ondev2 = bpf_offload_find_netdev(netdev); From 2bc6b218717b9d08f466f88209251d54bc09b207 Mon Sep 17 00:00:00 2001 From: Fuad Tabba Date: Tue, 22 Sep 2026 19:14:30 +0100 Subject: [PATCH 503/592] arm64/boot: Disable trapping of PMZR_EL0 writes to EL2 __init_el2_fgt2() writes one mask to both HDFGRTR2_EL2 and HDFGWTR2_EL2. PMZR_EL0 is write-only, so its trap bit, nPMZR_EL0, exists only in HDFGWTR2_EL2 and is therefore never set: a PMZR_EL0 write from the host traps to EL2, where the nVHE hypervisor has no handler and BUG()s. The kernel never writes PMZR_EL0, but kernel.perf_user_access=1 has the PMU driver set PMUSERENR_EL0.UEN for a task with a user-read event, so a write from EL0 reaches the trap and takes the host down without a panic message. Accumulate the HDFGWTR2_EL2 bits separately, as __init_el2_fgt() already does for HDFGWTR_EL2, and set nPMZR_EL0 with the other FEAT_PMUv3p9 bits. Fixes: 858c7bfcb35e1 ("arm64/boot: Enable EL2 requirements for FEAT_PMUv3p9") Cc: stable@vger.kernel.org Signed-off-by: Fuad Tabba Reviewed-by: Anshuman Khandual Reviewed-by: Oliver Upton Signed-off-by: Will Deacon --- Documentation/arch/arm64/booting.rst | 1 + arch/arm64/include/asm/el2_setup.h | 9 ++++++++- 2 files changed, 9 insertions(+), 1 deletion(-) diff --git a/Documentation/arch/arm64/booting.rst b/Documentation/arch/arm64/booting.rst index 13ef311dace8..3fea4b14ef7c 100644 --- a/Documentation/arch/arm64/booting.rst +++ b/Documentation/arch/arm64/booting.rst @@ -465,6 +465,7 @@ Before jumping into the kernel, the following conditions must be met: - HDFGWTR2_EL2.nPMICNTR_EL0 (bit 2) must be initialised to 0b1. - HDFGWTR2_EL2.nPMICFILTR_EL0 (bit 3) must be initialised to 0b1. - HDFGWTR2_EL2.nPMUACR_EL1 (bit 4) must be initialised to 0b1. + - HDFGWTR2_EL2.nPMZR_EL0 (bit 21) must be initialised to 0b1. For CPUs with SPE data source filtering (FEAT_SPE_FDS): diff --git a/arch/arm64/include/asm/el2_setup.h b/arch/arm64/include/asm/el2_setup.h index aa8ec9df8024..87560d8b254e 100644 --- a/arch/arm64/include/asm/el2_setup.h +++ b/arch/arm64/include/asm/el2_setup.h @@ -418,6 +418,7 @@ b.lt .Lskip_fgt2_\@ mov x0, xzr + mov x2, xzr mrs x1, id_aa64dfr0_el1 ubfx x1, x1, #ID_AA64DFR0_EL1_PMUVer_SHIFT, #4 cmp x1, #ID_AA64DFR0_EL1_PMUVer_V3P9 @@ -426,6 +427,11 @@ orr x0, x0, #HDFGRTR2_EL2_nPMICNTR_EL0 orr x0, x0, #HDFGRTR2_EL2_nPMICFILTR_EL0 orr x0, x0, #HDFGRTR2_EL2_nPMUACR_EL1 + orr x2, x2, #HDFGWTR2_EL2_nPMICNTR_EL0 + orr x2, x2, #HDFGWTR2_EL2_nPMICFILTR_EL0 + orr x2, x2, #HDFGWTR2_EL2_nPMUACR_EL1 + /* PMZR_EL0 is write-only, so it has no read trap to disable */ + orr x2, x2, #HDFGWTR2_EL2_nPMZR_EL0 .Lskip_pmuv3p9_\@: /* If SPE is implemented, */ __spe_vers_imp .Lskip_spefds_\@, ID_AA64DFR0_EL1_PMSVer_IMP, x1 @@ -436,10 +442,11 @@ cbz x1, .Lskip_spefds_\@ /* disable traps of PMSDSFR to EL2. */ orr x0, x0, #HDFGRTR2_EL2_nPMSDSFR_EL1 + orr x2, x2, #HDFGWTR2_EL2_nPMSDSFR_EL1 .Lskip_spefds_\@: msr_s SYS_HDFGRTR2_EL2, x0 - msr_s SYS_HDFGWTR2_EL2, x0 + msr_s SYS_HDFGWTR2_EL2, x2 msr_s SYS_HFGRTR2_EL2, xzr msr_s SYS_HFGWTR2_EL2, xzr msr_s SYS_HFGITR2_EL2, xzr From 686f942332b1667f13f3b8d6a2f50bcfbf42e277 Mon Sep 17 00:00:00 2001 From: Pengpeng Hou Date: Wed, 15 Jul 2026 16:43:25 +0800 Subject: [PATCH 504/592] nfc: nfcmrvl: validate helper command length before pull The firmware download receive path removes the NCI data header and reads the helper command before validating the remaining packet length. A short frame can therefore reach the data access before the malformed packet is rejected. Validate the complete helper command length before stripping the NCI data header. Fixes: 3194c6870158 ("NFC: nfcmrvl: add firmware download support") Signed-off-by: Pengpeng Hou Link: https://patch.msgid.link/20260715084325.40276-1-pengpeng@iscas.ac.cn Signed-off-by: David Heidelberg --- drivers/nfc/nfcmrvl/fw_dnld.c | 11 ++++++++--- 1 file changed, 8 insertions(+), 3 deletions(-) diff --git a/drivers/nfc/nfcmrvl/fw_dnld.c b/drivers/nfc/nfcmrvl/fw_dnld.c index 2b8f401d8fd7..8b9d5257320d 100644 --- a/drivers/nfc/nfcmrvl/fw_dnld.c +++ b/drivers/nfc/nfcmrvl/fw_dnld.c @@ -263,9 +263,14 @@ static int process_state_fw_dnld(struct nfcmrvl_private *priv, * B8..N: payload */ - /* Remove NCI HDR */ - skb_pull(skb, 3); - if (skb->data[0] != HELPER_CMD_PACKET_FORMAT || skb->len != 5) { + if (skb->len != NCI_DATA_HDR_SIZE + 5) { + nfc_err(priv->dev, "bad command"); + return -EINVAL; + } + + /* Remove NCI header */ + skb_pull(skb, NCI_DATA_HDR_SIZE); + if (skb->data[0] != HELPER_CMD_PACKET_FORMAT) { nfc_err(priv->dev, "bad command"); return -EINVAL; } From a653c01ce447f10c36b901646888c0330363af4f Mon Sep 17 00:00:00 2001 From: Pengpeng Hou Date: Wed, 15 Jul 2026 16:44:05 +0800 Subject: [PATCH 505/592] nfc: st21nfca: validate received frame size st21nfca_hci_i2c_repack() trims a received frame at its EOF marker before removing byte stuffing. It then assumes the truncated frame contains the LLC header and two CRC bytes, and it unconditionally reads the byte after an escape marker. A malformed frame can place EOF immediately after the start marker or can end its data portion with an escape marker. The former leaves too few bytes for check_crc(), while the latter makes the unstuffing loop read past the current skb length. Require the minimum framing bytes both before and after unstuffing. Use separate input and output cursors while removing byte stuffing, and reject an escape marker without its encoded byte. This keeps malformed frames within the received frame boundary before CRC processing. Fixes: 3096e25a3e40 ("NFC: st21nfca: Fix incorrect byte stuffing revocation") Signed-off-by: Pengpeng Hou Link: https://patch.msgid.link/20260715084405.41546-1-pengpeng@iscas.ac.cn Signed-off-by: David Heidelberg --- drivers/nfc/st21nfca/i2c.c | 29 +++++++++++++++++++---------- 1 file changed, 19 insertions(+), 10 deletions(-) diff --git a/drivers/nfc/st21nfca/i2c.c b/drivers/nfc/st21nfca/i2c.c index a4c93ff7c5b0..0f44c783bd04 100644 --- a/drivers/nfc/st21nfca/i2c.c +++ b/drivers/nfc/st21nfca/i2c.c @@ -289,27 +289,36 @@ static int check_crc(u8 *buf, int buflen) */ static int st21nfca_hci_i2c_repack(struct sk_buff *skb) { - int i, j, r, size; + int read, write, r, size; - if (skb->len < 1 || (skb->len > 1 && skb->data[1] != 0)) + if (skb->len < ST21NFCA_FRAME_HEADROOM || + !IS_START_OF_FRAME(skb->data)) return -EBADMSG; size = get_frame_size(skb->data, skb->len); if (size > 0) { + if (size < ST21NFCA_FRAME_HEADROOM + 2) + return -EBADMSG; + skb_trim(skb, size); /* remove ST21NFCA byte stuffing for upper layer */ - for (i = 1, j = 0; i < skb->len; i++) { - if (skb->data[i + j] == + for (read = 1, write = 1; read < skb->len;) { + if (skb->data[read] == (u8) ST21NFCA_ESCAPE_BYTE_STUFFING) { - skb->data[i] = skb->data[i + j + 1] - | ST21NFCA_BYTE_STUFFING_MASK; - i++; - j++; + if (read + 1 == skb->len) + return -EBADMSG; + + skb->data[write++] = skb->data[read + 1] + | ST21NFCA_BYTE_STUFFING_MASK; + read += 2; + } else { + skb->data[write++] = skb->data[read++]; } - skb->data[i] = skb->data[i + j]; } /* remove byte stuffing useless byte */ - skb_trim(skb, i - j); + skb_trim(skb, write); + if (skb->len < ST21NFCA_FRAME_HEADROOM + 2) + return -EBADMSG; /* remove ST21NFCA_SOF_EOF from head */ skb_pull(skb, 1); From bf1460acdf8cf5a07c819f59785d40f20d113099 Mon Sep 17 00:00:00 2001 From: Aldo Ariel Panzardo Date: Thu, 16 Jul 2026 20:26:57 -0300 Subject: [PATCH 506/592] nfc: llcp: Fix list corruption / refcount desync in nfc_llcp_recv_dm() nfc_llcp_recv_dm() handles DM(NOBOUND)/DM(REJ) for a socket that is still linked on local->connecting_sockets: it looks the socket up with nfc_llcp_connecting_sock_get(), sets sk->sk_state = LLCP_CLOSED and returns, without taking the socket lock and without unlinking the socket from the connecting_sockets list. llcp_sock_release() selects the list to unlink from by sk_state: a socket in LLCP_CONNECTING is unlinked from connecting_sockets, otherwise from the sockets list. Because recv_dm left the socket physically on connecting_sockets but in the LLCP_CLOSED state, release() takes the else branch and calls nfc_llcp_sock_unlink(&local->sockets, sk). That runs sk_del_node_init() while holding sockets.lock, i.e. it removes the socket from the connecting_sockets hlist under the wrong lock. A concurrent connect() linking another socket onto connecting_sockets under connecting_sockets.lock then mutates the same hlist unserialized, which corrupts the list and desyncs the sk_add_node()/sk_del_node_init() sock_hold()/__sock_put() pairing. An unprivileged local process holding LLCP sockets, with the DM supplied by the remote peer over an established LLCP link, can drive this to leak kernel sockets without bound (the mis-decrement goes through the non-freeing __sock_put() path, so the object is never released), leading to memory exhaustion / DoS. This is the same class of bug that was fixed in the sibling handler nfc_llcp_recv_cc() by commit b493ea2765cc ("nfc: llcp: Fix use-after-free race in nfc_llcp_recv_cc()"); recv_dm did not receive the equivalent fix. Fix it the same way: take lock_sock(), re-check that the socket is still hashed (release() may have won the race), and for the NOBOUND/REJ case unlink it from connecting_sockets before moving it to LLCP_CLOSED. The unlink drops the connecting_sockets membership reference via sk_del_node_init(), leaving the socket unhashed, so the later nfc_llcp_sock_unlink() in llcp_sock_release() becomes a no-op and no double put occurs. Fixes: a69f32af86e3 ("NFC: Socket linked list") Signed-off-by: Aldo Ariel Panzardo Link: https://patch.msgid.link/20260716232657.203145-1-qwe.aldo@gmail.com Signed-off-by: David Heidelberg --- net/nfc/llcp_core.c | 25 +++++++++++++++++++++++++ 1 file changed, 25 insertions(+) diff --git a/net/nfc/llcp_core.c b/net/nfc/llcp_core.c index cac1b5487064..bd6361e2efa4 100644 --- a/net/nfc/llcp_core.c +++ b/net/nfc/llcp_core.c @@ -1251,6 +1251,7 @@ static void nfc_llcp_recv_dm(struct nfc_llcp_local *local, struct nfc_llcp_sock *llcp_sock; struct sock *sk; u8 dsap, ssap, reason; + bool connecting = false; dsap = nfc_llcp_dsap(skb); ssap = nfc_llcp_ssap(skb); @@ -1262,6 +1263,7 @@ static void nfc_llcp_recv_dm(struct nfc_llcp_local *local, case LLCP_DM_NOBOUND: case LLCP_DM_REJ: llcp_sock = nfc_llcp_connecting_sock_get(local, dsap); + connecting = true; break; default: @@ -1276,10 +1278,33 @@ static void nfc_llcp_recv_dm(struct nfc_llcp_local *local, sk = &llcp_sock->sk; + lock_sock(sk); + + /* Check if socket was destroyed whilst waiting for the lock */ + if (!sk_hashed(sk)) { + release_sock(sk); + nfc_llcp_sock_put(llcp_sock); + return; + } + + /* + * For DM(NOBOUND)/DM(REJ) the socket is still linked on the + * connecting_sockets list. Unlink it here, under the socket lock, + * before moving it to LLCP_CLOSED: llcp_sock_release() selects the + * list to unlink from by sk_state, so leaving a connecting socket + * in the CLOSED state would make it unlink from the wrong list and + * corrupt the connecting_sockets list / desync the socket refcount. + * This mirrors nfc_llcp_recv_cc(). + */ + if (connecting) + nfc_llcp_sock_unlink(&local->connecting_sockets, sk); + sk->sk_err = ENXIO; sk->sk_state = LLCP_CLOSED; sk->sk_state_change(sk); + release_sock(sk); + nfc_llcp_sock_put(llcp_sock); } From 092c6a605cbd6414ef499834c2e0da69c2c3388e Mon Sep 17 00:00:00 2001 From: Doruk Tan Ozturk Date: Sat, 11 Jul 2026 14:36:51 +0200 Subject: [PATCH 507/592] nfc: port100: reject frames whose declared length exceeds the received data port100_recv_response() passes the URB transfer buffer to port100_rx_frame_is_valid(), which checksums le16_to_cpu(frame->datalen) bytes of frame->data. datalen is a 16-bit field supplied by the device and is never checked against the number of bytes actually received (urb->actual_length), so a device reporting a datalen larger than the received frame makes port100_data_checksum() read out of bounds past the transfer buffer. Reject a response whose declared frame size does not fit the received length before validating it. Found by 0sec (https://0sec.ai) using automated source analysis; the missing bound is evident from source. Compile-tested. Fixes: 562d4d59b8a1 ("NFC: Sony Port-100 Series driver") Cc: stable@vger.kernel.org Assisted-by: 0sec:claude-opus-4-8 Signed-off-by: Doruk Tan Ozturk Reviewed-by: Simon Horman Link: https://patch.msgid.link/20260711123651.32595-1-doruk@0sec.ai Signed-off-by: David Heidelberg --- drivers/nfc/port100.c | 7 +++++++ 1 file changed, 7 insertions(+) diff --git a/drivers/nfc/port100.c b/drivers/nfc/port100.c index b613f5e2fd57..e769a30b8b5c 100644 --- a/drivers/nfc/port100.c +++ b/drivers/nfc/port100.c @@ -636,6 +636,13 @@ static void port100_recv_response(struct urb *urb) in_frame = dev->in_urb->transfer_buffer; + if (urb->actual_length < PORT100_FRAME_HEADER_LEN || + urb->actual_length < port100_rx_frame_size(in_frame)) { + nfc_err(&dev->interface->dev, "Received a truncated frame\n"); + cmd->status = -EIO; + goto sched_wq; + } + if (!port100_rx_frame_is_valid(in_frame)) { nfc_err(&dev->interface->dev, "Received an invalid frame\n"); cmd->status = -EIO; From 3d8afc5243ea2ee803d98e69eb4a01748167ac1b Mon Sep 17 00:00:00 2001 From: Lei Zhu Date: Wed, 29 Jul 2026 15:24:26 +0800 Subject: [PATCH 508/592] selftests: nci: Correct pthread_create return value check The pthread_create() functions returns 0 on success and a positive value on failure. Modify the return value check to correctly detect failure cases. Fixes: 72696bd8a09d ("selftests: nci: Extract the start/stop discovery function") Signed-off-by: Lei Zhu Link: https://patch.msgid.link/20260729072426.303484-1-zhulei_szu@163.com Signed-off-by: David Heidelberg --- tools/testing/selftests/nci/nci_dev.c | 12 +++++++----- 1 file changed, 7 insertions(+), 5 deletions(-) diff --git a/tools/testing/selftests/nci/nci_dev.c b/tools/testing/selftests/nci/nci_dev.c index 312f84ee0444..c053f5cf2745 100644 --- a/tools/testing/selftests/nci/nci_dev.c +++ b/tools/testing/selftests/nci/nci_dev.c @@ -438,7 +438,7 @@ FIXTURE_SETUP(NCI) else rc = pthread_create(&thread_t, NULL, virtual_dev_open, (void *)&self->virtual_nci_fd); - ASSERT_GT(rc, -1); + ASSERT_EQ(rc, 0); rc = send_cmd_with_idx(self->sd, self->fid, self->pid, NFC_CMD_DEV_UP, self->dev_idex); @@ -509,7 +509,7 @@ FIXTURE_TEARDOWN(NCI) rc = pthread_create(&thread_t, NULL, virtual_deinit, (void *)&self->virtual_nci_fd); - ASSERT_GT(rc, -1); + ASSERT_EQ(rc, 0); rc = send_cmd_with_idx(self->sd, self->fid, self->pid, NFC_CMD_DEV_DOWN, self->dev_idex); EXPECT_EQ(rc, 0); @@ -590,7 +590,7 @@ int start_polling(int dev_idx, int proto, int virtual_fd, int sd, int fid, int p rc = pthread_create(&thread_t, NULL, virtual_poll_start, (void *)&virtual_fd); - if (rc < 0) + if (rc) return rc; rc = send_cmd_mt_nla(sd, fid, pid, NFC_CMD_START_POLL, 2, nla_start_poll_type, @@ -610,7 +610,7 @@ int stop_polling(int dev_idx, int virtual_fd, int sd, int fid, int pid) rc = pthread_create(&thread_t, NULL, virtual_poll_stop, (void *)&virtual_fd); - if (rc < 0) + if (rc) return rc; rc = send_cmd_with_idx(sd, fid, pid, @@ -830,6 +830,8 @@ int disconnect_tag(int nfc_sock, int virtual_fd) status = pthread_create(&thread_t, NULL, virtual_deactivate_proc, (void *)&virtual_fd); + if (status) + return status; close(nfc_sock); pthread_join(thread_t, (void **)&status); @@ -874,7 +876,7 @@ TEST_F(NCI, deinit) else rc = pthread_create(&thread_t, NULL, virtual_deinit, (void *)&self->virtual_nci_fd); - ASSERT_GT(rc, -1); + ASSERT_EQ(rc, 0); rc = send_cmd_with_idx(self->sd, self->fid, self->pid, NFC_CMD_DEV_DOWN, self->dev_idex); From c3eef2f988a3db9690369d7cef9a3344dd9788d3 Mon Sep 17 00:00:00 2001 From: Lee Jones Date: Wed, 2 Sep 2026 12:30:31 +0000 Subject: [PATCH 509/592] nfc: llcp: Fix race condition in accept_queue lifecycle In nfc_llcp_socket_release(), sockets and listener accept queues are walked under the local sockets rwlock and bh_lock_sock(). However, bh_lock_sock() does not synchronise against process-context lock_sock() held by nfc_llcp_accept_dequeue() during accept(). Because socket_release() does not check sock_owned_by_user(), both paths can concurrently unlink and release the same child socket, resulting in use-after-free or a NULL pointer dereference of child->parent in nfc_llcp_accept_unlink(). Fix this synchronisation race by having nfc_llcp_socket_release() use process-context lock_sock() instead of bh_lock_sock(): 1. Pop sockets from the local sockets list under the write lock using nfc_llcp_sock_list_pop() so lock_sock() can be acquired without holding the rwlock. 2. Because lock_sock() can sleep, defer the final release of the nfc_llcp_local structure to a workqueue (release_work). This avoids a sleeping-in-atomic bug when the last local reference is dropped from softirq context. Additionally, hold a single device reference on local from registration until final destruction. 3. In nfc_llcp_local_get(), use kref_get_unless_zero() to prevent resurrecting a local object whose teardown has been scheduled. 4. In llcp_sock_accept(), verify that the listener socket state is still LLCP_LISTEN after waking from schedule_timeout() to prevent hangs if the listener is closed concurrently. 5. When unlinking unaccepted child sockets during listener release, unlink them from local->sockets, call sock_orphan(), and drop their initial sk_alloc creation reference via sock_put(). 6. Make nfc_llcp_accept_unlink() idempotent by guarding parent access with a NULL check. Fixes: 50b78b2a6500 ("NFC: Fix sleeping in atomic when releasing socket") Signed-off-by: Lee Jones Link: https://patch.msgid.link/20260902123033.1169067-1-lee@kernel.org Signed-off-by: David Heidelberg --- net/nfc/llcp.h | 1 + net/nfc/llcp_core.c | 123 +++++++++++++++++++++++++++----------------- net/nfc/llcp_sock.c | 51 +++++++++++++----- 3 files changed, 116 insertions(+), 59 deletions(-) diff --git a/net/nfc/llcp.h b/net/nfc/llcp.h index d8345ed57c95..23ae7a0112d3 100644 --- a/net/nfc/llcp.h +++ b/net/nfc/llcp.h @@ -91,6 +91,7 @@ struct nfc_llcp_local { struct hlist_head pending_sdreqs; struct timer_list sdreq_timer; struct work_struct sdreq_timeout_work; + struct work_struct release_work; u8 sdreq_next_tid; /* sockets array */ diff --git a/net/nfc/llcp_core.c b/net/nfc/llcp_core.c index bd6361e2efa4..23553e7426ec 100644 --- a/net/nfc/llcp_core.c +++ b/net/nfc/llcp_core.c @@ -20,6 +20,8 @@ static LIST_HEAD(llcp_devices); /* Protects llcp_devices list */ static DEFINE_SPINLOCK(llcp_devices_lock); +static struct workqueue_struct *llcp_wq; + static void nfc_llcp_rx_skb(struct nfc_llcp_local *local, struct sk_buff *skb); void nfc_llcp_sock_link(struct llcp_sock_list *l, struct sock *sk) @@ -63,21 +65,33 @@ static void nfc_llcp_socket_purge(struct nfc_llcp_sock *sock) } } +static struct sock *nfc_llcp_sock_list_pop(struct llcp_sock_list *l) +{ + struct sock *sk; + + write_lock(&l->lock); + sk = sk_head(&l->head); + if (sk) { + sock_hold(sk); + sk_del_node_init(sk); + } + write_unlock(&l->lock); + + return sk; +} + static void nfc_llcp_socket_release(struct nfc_llcp_local *local, bool device, int err) { struct sock *sk; - struct hlist_node *tmp; struct nfc_llcp_sock *llcp_sock; skb_queue_purge(&local->tx_queue); - write_lock(&local->sockets.lock); - - sk_for_each_safe(sk, tmp, &local->sockets.head) { + while ((sk = nfc_llcp_sock_list_pop(&local->sockets))) { llcp_sock = nfc_llcp_sock(sk); - bh_lock_sock(sk); + lock_sock(sk); nfc_llcp_socket_purge(llcp_sock); @@ -91,17 +105,27 @@ static void nfc_llcp_socket_release(struct nfc_llcp_local *local, bool device, list_for_each_entry_safe(lsk, n, &llcp_sock->accept_queue, accept_queue) { + bool put_creation = false; + accept_sk = &lsk->sk; - bh_lock_sock(accept_sk); + lock_sock_nested(accept_sk, + SINGLE_DEPTH_NESTING); - nfc_llcp_accept_unlink(accept_sk); + if (nfc_llcp_sock(accept_sk)->parent == sk) { + nfc_llcp_accept_unlink(accept_sk); + nfc_llcp_sock_unlink(&local->sockets, accept_sk); - if (err) - accept_sk->sk_err = err; - accept_sk->sk_state = LLCP_CLOSED; - accept_sk->sk_state_change(sk); + if (err) + accept_sk->sk_err = err; + accept_sk->sk_state = LLCP_CLOSED; + accept_sk->sk_state_change(accept_sk); + sock_orphan(accept_sk); + put_creation = true; + } - bh_unlock_sock(accept_sk); + release_sock(accept_sk); + if (put_creation) + sock_put(accept_sk); /* creation ref */ } } @@ -110,23 +134,18 @@ static void nfc_llcp_socket_release(struct nfc_llcp_local *local, bool device, sk->sk_state = LLCP_CLOSED; sk->sk_state_change(sk); - bh_unlock_sock(sk); - - sk_del_node_init(sk); + release_sock(sk); + sock_put(sk); } - write_unlock(&local->sockets.lock); - /* If we still have a device, we keep the RAW sockets alive */ if (device == true) return; - write_lock(&local->raw_sockets.lock); - - sk_for_each_safe(sk, tmp, &local->raw_sockets.head) { + while ((sk = nfc_llcp_sock_list_pop(&local->raw_sockets))) { llcp_sock = nfc_llcp_sock(sk); - bh_lock_sock(sk); + lock_sock(sk); nfc_llcp_socket_purge(llcp_sock); @@ -135,26 +154,20 @@ static void nfc_llcp_socket_release(struct nfc_llcp_local *local, bool device, sk->sk_state = LLCP_CLOSED; sk->sk_state_change(sk); - bh_unlock_sock(sk); - - sk_del_node_init(sk); + release_sock(sk); + sock_put(sk); } - - write_unlock(&local->raw_sockets.lock); } static struct nfc_llcp_local *nfc_llcp_local_get(struct nfc_llcp_local *local) { - /* Since using nfc_llcp_local may result in usage of nfc_dev, whenever - * we hold a reference to local, we also need to hold a reference to - * the device to avoid UAF. - */ - if (!nfc_get_device(local->dev->idx)) + if (!local) return NULL; - kref_get(&local->ref); + if (kref_get_unless_zero(&local->ref)) + return local; - return local; + return NULL; } static void local_cleanup(struct nfc_llcp_local *local) @@ -172,30 +185,34 @@ static void local_cleanup(struct nfc_llcp_local *local) nfc_llcp_free_sdp_tlv_list(&local->pending_sdreqs); } +static void local_release_work(struct work_struct *work) +{ + struct nfc_llcp_local *local; + struct nfc_dev *dev; + + local = container_of(work, struct nfc_llcp_local, release_work); + dev = local->dev; + + local_cleanup(local); + kfree(local); + nfc_put_device(dev); +} + static void local_release(struct kref *ref) { struct nfc_llcp_local *local; local = container_of(ref, struct nfc_llcp_local, ref); - local_cleanup(local); - kfree(local); + queue_work(llcp_wq, &local->release_work); } int nfc_llcp_local_put(struct nfc_llcp_local *local) { - struct nfc_dev *dev; - int ret; - - if (local == NULL) + if (!local) return 0; - dev = local->dev; - - ret = kref_put(&local->ref, local_release); - nfc_put_device(dev); - - return ret; + return kref_put(&local->ref, local_release); } static struct nfc_llcp_sock *nfc_llcp_sock_get(struct nfc_llcp_local *local, @@ -1705,6 +1722,7 @@ int nfc_llcp_register_device(struct nfc_dev *ndev) INIT_WORK(&local->rx_work, nfc_llcp_rx_work); INIT_WORK(&local->timeout_work, nfc_llcp_timeout_work); + INIT_WORK(&local->release_work, local_release_work); rwlock_init(&local->sockets.lock); rwlock_init(&local->connecting_sockets.lock); @@ -1748,10 +1766,23 @@ void nfc_llcp_unregister_device(struct nfc_dev *dev) int __init nfc_llcp_init(void) { - return nfc_llcp_sock_init(); + int ret; + + llcp_wq = alloc_workqueue("nfc_llcp_wq", WQ_UNBOUND, 0); + if (!llcp_wq) + return -ENOMEM; + + ret = nfc_llcp_sock_init(); + if (ret) { + destroy_workqueue(llcp_wq); + return ret; + } + + return 0; } void nfc_llcp_exit(void) { nfc_llcp_sock_exit(); + destroy_workqueue(llcp_wq); } diff --git a/net/nfc/llcp_sock.c b/net/nfc/llcp_sock.c index 5558d8a4d48b..ce6875eb58fb 100644 --- a/net/nfc/llcp_sock.c +++ b/net/nfc/llcp_sock.c @@ -392,11 +392,12 @@ void nfc_llcp_accept_unlink(struct sock *sk) pr_debug("state %d\n", sk->sk_state); - list_del_init(&llcp_sock->accept_queue); - sk_acceptq_removed(llcp_sock->parent); - llcp_sock->parent = NULL; - - sock_put(sk); + if (llcp_sock->parent) { + list_del_init(&llcp_sock->accept_queue); + sk_acceptq_removed(llcp_sock->parent); + llcp_sock->parent = NULL; + sock_put(sk); + } } void nfc_llcp_accept_enqueue(struct sock *parent, struct sock *sk) @@ -423,12 +424,20 @@ struct sock *nfc_llcp_accept_dequeue(struct sock *parent, list_for_each_entry_safe(lsk, n, &llcp_parent->accept_queue, accept_queue) { + struct nfc_llcp_local *local; + sk = &lsk->sk; - lock_sock(sk); + lock_sock_nested(sk, SINGLE_DEPTH_NESTING); if (sk->sk_state == LLCP_CLOSED) { - release_sock(sk); + local = nfc_llcp_sock(sk)->local; + nfc_llcp_accept_unlink(sk); + if (local) + nfc_llcp_sock_unlink(&local->sockets, sk); + sock_orphan(sk); + release_sock(sk); + sock_put(sk); continue; } @@ -464,7 +473,7 @@ static int llcp_sock_accept(struct socket *sock, struct socket *newsock, pr_debug("parent %p\n", sk); - lock_sock_nested(sk, SINGLE_DEPTH_NESTING); + lock_sock(sk); if (sk->sk_state != LLCP_LISTEN) { ret = -EBADFD; @@ -490,7 +499,12 @@ static int llcp_sock_accept(struct socket *sock, struct socket *newsock, release_sock(sk); timeo = schedule_timeout(timeo); - lock_sock_nested(sk, SINGLE_DEPTH_NESTING); + lock_sock(sk); + + if (sk->sk_state != LLCP_LISTEN) { + ret = -EBADFD; + break; + } } __set_current_state(TASK_RUNNING); remove_wait_queue(sk_sleep(sk), &wait); @@ -629,13 +643,24 @@ static int llcp_sock_release(struct socket *sock) list_for_each_entry_safe(lsk, n, &llcp_sock->accept_queue, accept_queue) { - accept_sk = &lsk->sk; - lock_sock(accept_sk); + bool put_creation = false; - nfc_llcp_send_disconnect(lsk); - nfc_llcp_accept_unlink(accept_sk); + accept_sk = &lsk->sk; + lock_sock_nested(accept_sk, SINGLE_DEPTH_NESTING); + + if (nfc_llcp_sock(accept_sk)->parent == sk) { + nfc_llcp_send_disconnect(lsk); + nfc_llcp_accept_unlink(accept_sk); + nfc_llcp_sock_unlink(&local->sockets, accept_sk); + + accept_sk->sk_state = LLCP_CLOSED; + sock_orphan(accept_sk); + put_creation = true; + } release_sock(accept_sk); + if (put_creation) + sock_put(accept_sk); /* creation ref */ } } From eda518d2cdb6074a0bcdfa06af291616bcb5c421 Mon Sep 17 00:00:00 2001 From: Chaithanya Lagisetty Date: Tue, 1 Sep 2026 07:06:18 +0000 Subject: [PATCH 510/592] selftests: nci: Fix uninitialized family ID on missing attribute get_family_id() walks the generic netlink CTRL_CMD_GETFAMILY reply looking for the CTRL_ATTR_FAMILY_ID attribute and returns the parsed value in the local variable "id". If the reply does not carry that attribute, the parsing loop never assigns "id" and the function returns an indeterminate stack value, which the caller stores in self->fid and uses for subsequent netlink requests. Initialize "id" to 0 so a missing attribute yields a deterministic (invalid) family ID instead of a garbage value. Fixes: f595cf1242f3 ("selftests: Add nci suite") Signed-off-by: Chaithanya Lagisetty Reviewed-by: Hangbin Liu Link: https://patch.msgid.link/20260901070618.3299012-1-nagachaithanya9911@gmail.com Signed-off-by: David Heidelberg --- tools/testing/selftests/nci/nci_dev.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/tools/testing/selftests/nci/nci_dev.c b/tools/testing/selftests/nci/nci_dev.c index c053f5cf2745..23fd38acfcf4 100644 --- a/tools/testing/selftests/nci/nci_dev.c +++ b/tools/testing/selftests/nci/nci_dev.c @@ -182,7 +182,7 @@ static int get_family_id(int sd, __u32 pid, __u32 *event_group) } ans; struct nlattr *na; int resp_len; - __u16 id; + __u16 id = 0; int len; int rc; From 6be581aeffc215bfc77939cd59902b0dbc4af23e Mon Sep 17 00:00:00 2001 From: Chris Gellermann Date: Fri, 4 Sep 2026 11:59:15 +0200 Subject: [PATCH 511/592] selftests/nci: Fix out-of-bounds store on thread join The NCI test collects the exit status of its helper threads by passing the address of an int to pthread_join(): int status; ... pthread_join(thread_t, (void **) &status); pthread_join() stores a void pointer to the memory location. On 64-bit systems, a void pointer is wider than an int, so the store overruns the 4 bytes of space allocated on the stack for the integer and corrupts the adjacent stack. On our CHERI system, this caused a fault due to a capability bounds violation. Fix this by introducing a helper that joins a thread through a void pointer and converts the result back to an integer, which is what the helper threads return. While here, also fix the logic in disconnect_tag() if the helper thread creation failed. Previously, it would have joined a thread that was never created when pthread_create() failed. Fixes: f595cf1242f3 ("selftests: Add nci suite") Signed-off-by: Chris Gellermann Reviewed-by: Simon Horman Link: https://patch.msgid.link/20260904095915.3372241-1-christian.gellermann@codasip.com Signed-off-by: David Heidelberg --- tools/testing/selftests/nci/nci_dev.c | 31 +++++++++++++++++---------- 1 file changed, 20 insertions(+), 11 deletions(-) diff --git a/tools/testing/selftests/nci/nci_dev.c b/tools/testing/selftests/nci/nci_dev.c index 23fd38acfcf4..07427fa42888 100644 --- a/tools/testing/selftests/nci/nci_dev.c +++ b/tools/testing/selftests/nci/nci_dev.c @@ -8,6 +8,7 @@ #include #include +#include #include #include #include @@ -87,6 +88,16 @@ struct msgtemplate { char buf[MAX_MSG_SIZE]; }; +static int join_thread_status(pthread_t thread) +{ + void *thread_ret = NULL; + + if (pthread_join(thread, &thread_ret)) + return -1; + + return (int)(intptr_t)thread_ret; +} + static int create_nl_socket(void) { int fd; @@ -444,7 +455,7 @@ FIXTURE_SETUP(NCI) NFC_CMD_DEV_UP, self->dev_idex); EXPECT_EQ(rc, 0); - pthread_join(thread_t, (void **)&status); + status = join_thread_status(thread_t); ASSERT_EQ(status, 0); self->open_state = true; } @@ -514,7 +525,7 @@ FIXTURE_TEARDOWN(NCI) NFC_CMD_DEV_DOWN, self->dev_idex); EXPECT_EQ(rc, 0); - pthread_join(thread_t, (void **)&status); + status = join_thread_status(thread_t); ASSERT_EQ(status, 0); } @@ -585,7 +596,6 @@ int start_polling(int dev_idx, int proto, int virtual_fd, int sd, int fid, int p void *nla_start_poll_data[2] = {&dev_idx, &proto}; int nla_start_poll_len[2] = {4, 4}; pthread_t thread_t; - int status; int rc; rc = pthread_create(&thread_t, NULL, virtual_poll_start, @@ -598,14 +608,12 @@ int start_polling(int dev_idx, int proto, int virtual_fd, int sd, int fid, int p if (rc != 0) return rc; - pthread_join(thread_t, (void **)&status); - return status; + return join_thread_status(thread_t); } int stop_polling(int dev_idx, int virtual_fd, int sd, int fid, int pid) { pthread_t thread_t; - int status; int rc; rc = pthread_create(&thread_t, NULL, virtual_poll_stop, @@ -618,8 +626,7 @@ int stop_polling(int dev_idx, int virtual_fd, int sd, int fid, int pid) if (rc != 0) return rc; - pthread_join(thread_t, (void **)&status); - return status; + return join_thread_status(thread_t); } TEST_F(NCI, start_poll) @@ -834,8 +841,10 @@ int disconnect_tag(int nfc_sock, int virtual_fd) return status; close(nfc_sock); - pthread_join(thread_t, (void **)&status); - return status; + if (status) + return -1; + + return join_thread_status(thread_t); } TEST_F(NCI, t4t_tag_read) @@ -882,7 +891,7 @@ TEST_F(NCI, deinit) NFC_CMD_DEV_DOWN, self->dev_idex); EXPECT_EQ(rc, 0); - pthread_join(thread_t, (void **)&status); + status = join_thread_status(thread_t); self->open_state = 0; ASSERT_EQ(status, 0); From 51814683e28fc64eceb415962376956c3cfc75a7 Mon Sep 17 00:00:00 2001 From: Chris Gellermann Date: Fri, 4 Sep 2026 18:42:52 +0200 Subject: [PATCH 512/592] nfc: virtual_ncidev: Add missing ioctl compat handler The compat handler for ioctls to the virtual nci device is missing. So, nci-specific ioctls of a compat task return with -1 and errno set to ENOTTY. Add a handler. The handling of an ioctl() call of a compat task to get the index of virtual nci device (IOCTL_GET_NCIDEV_IDX) lands in the default case of the ioctl compat handler (see fs/ioctl.c): COMPAT_SYSCALL_DEFINE3(ioctl, ...) { ... default: error = do_vfs_ioctl(fd_file(f), fd, cmd, ...); if (error != -ENOIOCTLCMD) break; if (fd_file(f)->f_op->compat_ioctl) error = fd_file(f)->f_op->compat_ioctl(fd_file(f), cmd, arg); if (error == -ENOIOCTLCMD) error = -ENOTTY; ... } There, do_vfs_ioctl() returns -ENOIOCTLCMD and compat_ioctl is not set for virtual_ncidev_fops, i.e. f_op->compat_ioctl == NULL. So, the ioctl() syscall returns with -1 and errno set to ENOTTY to the compat task. To fix this, use the compat_ptr_ioctl helper for compat handling here. It shall be used for ioctls that "either ignore the argument or pass a pointer to a compatible data type". The driver's sole ioctl takes a user void pointer and copies nfc_dev->idx to it, a 4-byte integer across all ABIs. This issue has been found by running the nci_dev kernel selftest as rv64 binary on top of a CHERI kernel, where the ioctl() ends up in the ioctl compat handler, similar to a 32-bit application on top of a 64-bit kernel. Fixes: e624e6c3e777 ("nfc: Add a virtual nci device driver") Signed-off-by: Chris Gellermann Reviewed-by: Simon Horman Link: https://patch.msgid.link/20260904164252.18351-1-christian.gellermann@codasip.com Signed-off-by: David Heidelberg --- drivers/nfc/virtual_ncidev.c | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-) diff --git a/drivers/nfc/virtual_ncidev.c b/drivers/nfc/virtual_ncidev.c index 8eeb447ac96e..e51c647b27eb 100644 --- a/drivers/nfc/virtual_ncidev.c +++ b/drivers/nfc/virtual_ncidev.c @@ -195,7 +195,8 @@ static const struct file_operations virtual_ncidev_fops = { .write = virtual_ncidev_write, .open = virtual_ncidev_open, .release = virtual_ncidev_close, - .unlocked_ioctl = virtual_ncidev_ioctl + .unlocked_ioctl = virtual_ncidev_ioctl, + .compat_ioctl = compat_ptr_ioctl, }; static struct miscdevice miscdev = { From 273f9d667cde649f8de9d72b1303cc2f4b658c50 Mon Sep 17 00:00:00 2001 From: Aamir Ahmed Date: Tue, 15 Sep 2026 19:54:27 +0100 Subject: [PATCH 513/592] nfc: llcp: drop truncated I/RR/RNR PDUs in nfc_llcp_recv_hdlc() nfc_llcp_recv_hdlc() reads the sequence byte skb->data[2], via nfc_llcp_ns()/nfc_llcp_nr(), before any length check. The receive path only guarantees the two-byte LLCP header -- __nfc_llcp_recv() checks it with pskb_may_pull() and nfc_llcp_recv_agf() admits two-byte inner PDUs -- so a two-byte I, RR or RNR PDU reads one byte of uninitialised skb tailroom. The byte becomes N(R)/N(S); a peer can already set those with a well-formed PDU, so this is acting on uninitialised memory, not new peer control. Guard the read with pskb_may_pull(), as commit 95674f506c63 ("nfc: llcp: reject PDUs shorter than the LLCP header") did for the two-byte header, so the sequence byte is present and linear before it is read. RR and RNR PDUs are LLCP_HEADER_SIZE + LLCP_SEQUENCE_SIZE bytes and an I PDU is longer, so no valid frame is rejected; a truncated PDU is malformed, so return without a DM reply. Fixes: d646960f7986 ("NFC: Initial LLCP support") Cc: stable@vger.kernel.org Assisted-by: LLM Signed-off-by: Aamir Ahmed Reviewed-by: Simon Horman Link: https://patch.msgid.link/AS8P251MB0001789BBF04B72745C7D96BC8BA2@AS8P251MB0001.EURP251.PROD.OUTLOOK.COM Signed-off-by: David Heidelberg --- net/nfc/llcp_core.c | 3 +++ 1 file changed, 3 insertions(+) diff --git a/net/nfc/llcp_core.c b/net/nfc/llcp_core.c index 23553e7426ec..5017f6aa57a0 100644 --- a/net/nfc/llcp_core.c +++ b/net/nfc/llcp_core.c @@ -1091,6 +1091,9 @@ static void nfc_llcp_recv_hdlc(struct nfc_llcp_local *local, struct sock *sk; u8 dsap, ssap, ptype, ns, nr; + if (!pskb_may_pull(skb, LLCP_HEADER_SIZE + LLCP_SEQUENCE_SIZE)) + return; + ptype = nfc_llcp_ptype(skb); dsap = nfc_llcp_dsap(skb); ssap = nfc_llcp_ssap(skb); From 66f4300206b82b0b143ef0d9be90cd8d29f23a47 Mon Sep 17 00:00:00 2001 From: Cong Nguyen Date: Mon, 14 Sep 2026 19:11:29 +0700 Subject: [PATCH 514/592] nfc: llcp: fix sdreq TLV list leak on parse/alloc/send failure nfc_genl_llc_sdreq() builds a list of TLV nodes while walking nested netlink attrs, but 3 error paths (nested-attr parse failure, TLV alloc ENOMEM, nfc_llcp_send_snl_sdreq() failure) all skip freeing what was already queued. Route them through a new free_list label, mirroring the SDRES path in the same file which already does this. Harmless on the success path too -- send_snl_sdreq() drains the list as it moves nodes, so it's already empty by the time free_list runs. Fixes: d9b8d8e19b07 ("NFC: llcp: Service Name Lookup netlink interface") Assisted-by: Claude:claude-opus-4 Signed-off-by: Cong Nguyen Reviewed-by: Simon Horman Link: https://patch.msgid.link/20260914121129.2098606-1-congnt264@gmail.com Signed-off-by: David Heidelberg --- net/nfc/netlink.c | 7 +++++-- 1 file changed, 5 insertions(+), 2 deletions(-) diff --git a/net/nfc/netlink.c b/net/nfc/netlink.c index 0c58824cb150..224bdfa2dd0d 100644 --- a/net/nfc/netlink.c +++ b/net/nfc/netlink.c @@ -1181,7 +1181,7 @@ static int nfc_genl_llc_sdreq(struct sk_buff *skb, struct genl_info *info) if (rc != 0) { rc = -EINVAL; - goto put_local; + goto free_list; } if (!sdp_attrs[NFC_SDP_ATTR_URI]) @@ -1200,7 +1200,7 @@ static int nfc_genl_llc_sdreq(struct sk_buff *skb, struct genl_info *info) sdreq = nfc_llcp_build_sdreq_tlv(tid, uri, uri_len); if (sdreq == NULL) { rc = -ENOMEM; - goto put_local; + goto free_list; } tlvs_len += sdreq->tlv_len; @@ -1215,6 +1215,9 @@ static int nfc_genl_llc_sdreq(struct sk_buff *skb, struct genl_info *info) rc = nfc_llcp_send_snl_sdreq(local, &sdreq_list, tlvs_len); +free_list: + nfc_llcp_free_sdp_tlv_list(&sdreq_list); + put_local: nfc_llcp_local_put(local); From d2acbde7e67df44efa8f0963462d1192e7694ffc Mon Sep 17 00:00:00 2001 From: Myeonghun Pak Date: Sun, 13 Sep 2026 00:26:25 -0400 Subject: [PATCH 515/592] nfc: trf7970a: power down on startup RX gain failure trf7970a_startup() powers up the device before applying the optional RX gain reduction. If the register read or write fails, it returns without undoing that power-up. Probe's unwind only drops the separate regulator references acquired by probe, leaving the additional VIN enable from startup unbalanced. The system resume caller also has no power-down on this error. Call trf7970a_power_down() before returning the RX gain error to deassert the enable GPIOs, release the startup VIN reference and restore the powered-off state. Runtime PM has not been enabled yet, so the full shutdown helper is not appropriate here. Preserve the original SPI error. This issue was identified during our ongoing static-analysis research while reviewing kernel code. Fixes: 5d69351820ea ("NFC: trf7970a: Create device-tree parameter for RX gain reduction") Cc: stable@vger.kernel.org Assisted-by: OpenAI:GPT-5.6 Co-developed-by: Ijae Kim Signed-off-by: Ijae Kim Signed-off-by: Myeonghun Pak Reviewed-by: Paul Geurts Link: https://patch.msgid.link/20260913042625.31296-1-mhun512@gmail.com Signed-off-by: David Heidelberg --- drivers/nfc/trf7970a.c | 4 +++- 1 file changed, 3 insertions(+), 1 deletion(-) diff --git a/drivers/nfc/trf7970a.c b/drivers/nfc/trf7970a.c index 60883001fa5d..ddfc58c29964 100644 --- a/drivers/nfc/trf7970a.c +++ b/drivers/nfc/trf7970a.c @@ -1997,8 +1997,10 @@ static int trf7970a_startup(struct trf7970a *trf) return ret; ret = trf7970a_update_rx_gain_reduction(trf); - if (ret) + if (ret) { + trf7970a_power_down(trf); return ret; + } pm_runtime_set_active(trf->dev); pm_runtime_enable(trf->dev); From dcab71a7011918f6fdba7adcec02d217dcb84b8d Mon Sep 17 00:00:00 2001 From: Luxiao Xu Date: Wed, 9 Sep 2026 13:19:24 +0800 Subject: [PATCH 516/592] nfc: fix use-after-free in nfc_get_local_general_bytes Commit 6709d4b7bc2e ("net: nfc: Fix use-after-free caused by nfc_llcp_find_local") attempted to fix a use-after-free (UAF) issue by invoking nfc_llcp_local_put(local) after accessing local->gb. However, if the reference count drops to zero, local is freed immediately, leading to a use-after-free when callers access the returned pointer. Alternative approaches using dynamic allocation (e.g. kmemdup) introduced memory leaks because callers consistently treat the returned pointer as borrowed memory. Fix this properly by refactoring nfc_llcp_general_bytes() and nfc_get_local_general_bytes() to accept a caller-provided output buffer (out_gb) and its maximum length (gb_max_len). The general bytes are safely copied into out_gb before calling nfc_llcp_local_put(local), ensuring safe lifetime management without ownership transfer complications. Update all callers across drivers (microread, pn533, pn544, st21nfca, digital_dep, and nci) to provide their own destination buffers and pass them to nfc_get_local_general_bytes(). Fixes: 6709d4b7bc2e ("net: nfc: Fix use-after-free caused by nfc_llcp_find_local") Cc: stable@vger.kernel.org Reported-by: Vega Assisted-by: LLM Signed-off-by: Luxiao Xu Signed-off-by: Ren Wei Reviewed-by: Simon Horman Link: https://patch.msgid.link/3cbaac3bee23f8ff3a3284ed32d347696eb1d208.1788841683.git.rakukuip@gmail.com Signed-off-by: David Heidelberg --- drivers/nfc/microread/microread.c | 6 +++--- drivers/nfc/pn533/pn533.c | 14 ++++++++------ drivers/nfc/pn533/pn533.h | 4 +++- drivers/nfc/pn544/pn544.c | 7 +++---- drivers/nfc/st21nfca/core.c | 8 ++++---- include/net/nfc/hci.h | 2 +- include/net/nfc/nfc.h | 3 ++- net/nfc/core.c | 15 +++++++-------- net/nfc/digital_dep.c | 8 ++++---- net/nfc/llcp_core.c | 17 +++++++++++++---- net/nfc/nci/core.c | 10 +++++----- net/nfc/nfc.h | 3 ++- 12 files changed, 55 insertions(+), 42 deletions(-) diff --git a/drivers/nfc/microread/microread.c b/drivers/nfc/microread/microread.c index dfa2490db545..2bfafa94e83d 100644 --- a/drivers/nfc/microread/microread.c +++ b/drivers/nfc/microread/microread.c @@ -251,9 +251,9 @@ static int microread_start_poll(struct nfc_hci_dev *hdev, param[1] |= (1 << 1); if ((im_protocols | tm_protocols) & NFC_PROTO_NFC_DEP_MASK) { - hdev->gb = nfc_get_local_general_bytes(hdev->ndev, - &hdev->gb_len); - if (hdev->gb == NULL || hdev->gb_len == 0) { + nfc_get_local_general_bytes(hdev->ndev, hdev->gb, + sizeof(hdev->gb), &hdev->gb_len); + if (hdev->gb_len == 0) { im_protocols &= ~NFC_PROTO_NFC_DEP_MASK; tm_protocols &= ~NFC_PROTO_NFC_DEP_MASK; } diff --git a/drivers/nfc/pn533/pn533.c b/drivers/nfc/pn533/pn533.c index f5a6a7c20d5a..b0133e51dce9 100644 --- a/drivers/nfc/pn533/pn533.c +++ b/drivers/nfc/pn533/pn533.c @@ -1357,10 +1357,11 @@ static int pn533_poll_dep(struct nfc_dev *nfc_dev) u8 *next, nfcid3[NFC_NFCID3_MAXSIZE]; u8 passive_data[PASSIVE_DATA_LEN] = {0x00, 0xff, 0xff, 0x00, 0x3}; - if (!dev->gb) { - dev->gb = nfc_get_local_general_bytes(nfc_dev, &dev->gb_len); - - if (!dev->gb || !dev->gb_len) { + if (!dev->gb_len) { + nfc_get_local_general_bytes(nfc_dev, dev->gb, + sizeof(dev->gb), + &dev->gb_len); + if (!dev->gb_len) { dev->poll_dep = 0; queue_work(dev->wq, &dev->rf_work); } @@ -1658,8 +1659,9 @@ static int pn533_start_poll(struct nfc_dev *nfc_dev, } if (tm_protocols) { - dev->gb = nfc_get_local_general_bytes(nfc_dev, &dev->gb_len); - if (dev->gb == NULL) + nfc_get_local_general_bytes(nfc_dev, dev->gb, + sizeof(dev->gb), &dev->gb_len); + if (dev->gb_len == 0) tm_protocols = 0; } diff --git a/drivers/nfc/pn533/pn533.h b/drivers/nfc/pn533/pn533.h index 09e35b8693f5..5ab668e05121 100644 --- a/drivers/nfc/pn533/pn533.h +++ b/drivers/nfc/pn533/pn533.h @@ -6,6 +6,8 @@ * Copyright (C) 2012-2013 Tieto Poland */ +#include + #define PN533_DEVICE_STD 0x1 #define PN533_DEVICE_PASORI 0x2 #define PN533_DEVICE_ACR122U 0x3 @@ -166,7 +168,7 @@ struct pn533 { struct timer_list listen_timer; int cancel_listen; - u8 *gb; + u8 gb[NFC_MAX_GT_LEN]; size_t gb_len; u8 tgt_available_prots; diff --git a/drivers/nfc/pn544/pn544.c b/drivers/nfc/pn544/pn544.c index 9d0a16ac465e..c4fa70e45c14 100644 --- a/drivers/nfc/pn544/pn544.c +++ b/drivers/nfc/pn544/pn544.c @@ -377,10 +377,9 @@ static int pn544_hci_start_poll(struct nfc_hci_dev *hdev, return r; if ((im_protocols | tm_protocols) & NFC_PROTO_NFC_DEP_MASK) { - hdev->gb = nfc_get_local_general_bytes(hdev->ndev, - &hdev->gb_len); - pr_debug("generate local bytes %p\n", hdev->gb); - if (hdev->gb == NULL || hdev->gb_len == 0) { + nfc_get_local_general_bytes(hdev->ndev, hdev->gb, + sizeof(hdev->gb), &hdev->gb_len); + if (hdev->gb_len == 0) { im_protocols &= ~NFC_PROTO_NFC_DEP_MASK; tm_protocols &= ~NFC_PROTO_NFC_DEP_MASK; } diff --git a/drivers/nfc/st21nfca/core.c b/drivers/nfc/st21nfca/core.c index fd39a05c9622..6bfeb8e7ed89 100644 --- a/drivers/nfc/st21nfca/core.c +++ b/drivers/nfc/st21nfca/core.c @@ -351,10 +351,10 @@ static int st21nfca_hci_start_poll(struct nfc_hci_dev *hdev, if (r < 0) return r; } else { - hdev->gb = nfc_get_local_general_bytes(hdev->ndev, - &hdev->gb_len); - - if (hdev->gb == NULL || hdev->gb_len == 0) { + nfc_get_local_general_bytes(hdev->ndev, hdev->gb, + sizeof(hdev->gb), + &hdev->gb_len); + if (hdev->gb_len == 0) { im_protocols &= ~NFC_PROTO_NFC_DEP_MASK; tm_protocols &= ~NFC_PROTO_NFC_DEP_MASK; } diff --git a/include/net/nfc/hci.h b/include/net/nfc/hci.h index 756c11084f65..86ed63e5d533 100644 --- a/include/net/nfc/hci.h +++ b/include/net/nfc/hci.h @@ -144,7 +144,7 @@ struct nfc_hci_dev { data_exchange_cb_t async_cb; void *async_cb_context; - u8 *gb; + u8 gb[NFC_MAX_GT_LEN]; size_t gb_len; unsigned long quirks; diff --git a/include/net/nfc/nfc.h b/include/net/nfc/nfc.h index c54df042db6b..bcafab5c53e5 100644 --- a/include/net/nfc/nfc.h +++ b/include/net/nfc/nfc.h @@ -273,7 +273,8 @@ struct sk_buff *nfc_alloc_recv_skb(unsigned int size, gfp_t gfp); int nfc_set_remote_general_bytes(struct nfc_dev *dev, const u8 *gt, u8 gt_len); -u8 *nfc_get_local_general_bytes(struct nfc_dev *dev, size_t *gb_len); +u8 *nfc_get_local_general_bytes(struct nfc_dev *dev, u8 *out_gb, + size_t gb_max_len, size_t *gb_len); int nfc_fw_download_done(struct nfc_dev *dev, const char *firmware_name, u32 result); diff --git a/net/nfc/core.c b/net/nfc/core.c index a92a6566e6a0..f521669293f0 100644 --- a/net/nfc/core.c +++ b/net/nfc/core.c @@ -279,10 +279,10 @@ static struct nfc_target *nfc_find_target(struct nfc_dev *dev, u32 target_idx) int nfc_dep_link_up(struct nfc_dev *dev, int target_index, u8 comm_mode) { - int rc = 0; - u8 *gb; - size_t gb_len; struct nfc_target *target; + u8 gb[NFC_MAX_GT_LEN]; + size_t gb_len = 0; + int rc = 0; pr_debug("dev_name=%s comm %d\n", dev_name(&dev->dev), comm_mode); @@ -301,7 +301,7 @@ int nfc_dep_link_up(struct nfc_dev *dev, int target_index, u8 comm_mode) goto error; } - gb = nfc_llcp_general_bytes(dev, &gb_len); + nfc_get_local_general_bytes(dev, gb, sizeof(gb), &gb_len); if (gb_len > NFC_MAX_GT_LEN) { rc = -EINVAL; goto error; @@ -644,11 +644,10 @@ int nfc_set_remote_general_bytes(struct nfc_dev *dev, const u8 *gb, u8 gb_len) } EXPORT_SYMBOL(nfc_set_remote_general_bytes); -u8 *nfc_get_local_general_bytes(struct nfc_dev *dev, size_t *gb_len) +u8 *nfc_get_local_general_bytes(struct nfc_dev *dev, u8 *out_gb, + size_t gb_max_len, size_t *gb_len) { - pr_debug("dev_name=%s\n", dev_name(&dev->dev)); - - return nfc_llcp_general_bytes(dev, gb_len); + return nfc_llcp_general_bytes(dev, out_gb, gb_max_len, gb_len); } EXPORT_SYMBOL(nfc_get_local_general_bytes); diff --git a/net/nfc/digital_dep.c b/net/nfc/digital_dep.c index 3982fa084737..968547c306a5 100644 --- a/net/nfc/digital_dep.c +++ b/net/nfc/digital_dep.c @@ -1490,14 +1490,14 @@ static int digital_tg_send_atr_res(struct nfc_digital_dev *ddev, struct digital_atr_req *atr_req) { struct digital_atr_res *atr_res; + u8 gb[NFC_MAX_GT_LEN]; struct sk_buff *skb; - u8 *gb, payload_bits; + u8 payload_bits; size_t gb_len; int rc; - gb = nfc_get_local_general_bytes(ddev->nfc_dev, &gb_len); - if (!gb) - gb_len = 0; + nfc_get_local_general_bytes(ddev->nfc_dev, gb, sizeof(gb), + &gb_len); skb = digital_skb_alloc(ddev, sizeof(struct digital_atr_res) + gb_len); if (!skb) diff --git a/net/nfc/llcp_core.c b/net/nfc/llcp_core.c index 5017f6aa57a0..5ce3ce64baf2 100644 --- a/net/nfc/llcp_core.c +++ b/net/nfc/llcp_core.c @@ -652,23 +652,32 @@ static int nfc_llcp_build_gb(struct nfc_llcp_local *local) return ret; } -u8 *nfc_llcp_general_bytes(struct nfc_dev *dev, size_t *general_bytes_len) +u8 *nfc_llcp_general_bytes(struct nfc_dev *dev, u8 *out_gb, size_t gb_max_len, + size_t *general_bytes_len) { struct nfc_llcp_local *local; + if (!out_gb || !general_bytes_len) + return NULL; + local = nfc_llcp_find_local(dev); - if (local == NULL) { + if (!local) { *general_bytes_len = 0; return NULL; } nfc_llcp_build_gb(local); - *general_bytes_len = local->gb_len; + if (local->gb_len) { + *general_bytes_len = min_t(size_t, local->gb_len, gb_max_len); + memcpy(out_gb, local->gb, *general_bytes_len); + } else { + *general_bytes_len = 0; + } nfc_llcp_local_put(local); - return local->gb; + return out_gb; } int nfc_llcp_set_remote_gb(struct nfc_dev *dev, const u8 *gb, u8 gb_len) diff --git a/net/nfc/nci/core.c b/net/nfc/nci/core.c index 5f46c4b5720f..73e3a96470ac 100644 --- a/net/nfc/nci/core.c +++ b/net/nfc/nci/core.c @@ -780,15 +780,15 @@ static int nci_set_local_general_bytes(struct nfc_dev *nfc_dev) { struct nci_dev *ndev = nfc_get_drvdata(nfc_dev); struct nci_set_config_param param; + u8 gb[NFC_MAX_GT_LEN]; int rc; - param.val = nfc_get_local_general_bytes(nfc_dev, ¶m.len); - if ((param.val == NULL) || (param.len == 0)) + nfc_get_local_general_bytes(nfc_dev, gb, sizeof(gb), + ¶m.len); + if (param.len == 0) return 0; - if (param.len > NFC_MAX_GT_LEN) - return -EINVAL; - + param.val = gb; param.id = NCI_PN_ATR_REQ_GEN_BYTES; rc = nci_request(ndev, nci_set_config_req, ¶m, diff --git a/net/nfc/nfc.h b/net/nfc/nfc.h index 0b1e6466f4fb..82c5dfdad10e 100644 --- a/net/nfc/nfc.h +++ b/net/nfc/nfc.h @@ -49,7 +49,8 @@ void nfc_llcp_mac_is_up(struct nfc_dev *dev, u32 target_idx, int nfc_llcp_register_device(struct nfc_dev *dev); void nfc_llcp_unregister_device(struct nfc_dev *dev); int nfc_llcp_set_remote_gb(struct nfc_dev *dev, const u8 *gb, u8 gb_len); -u8 *nfc_llcp_general_bytes(struct nfc_dev *dev, size_t *general_bytes_len); +u8 *nfc_llcp_general_bytes(struct nfc_dev *dev, u8 *out_gb, size_t gb_max_len, + size_t *general_bytes_len); int nfc_llcp_data_received(struct nfc_dev *dev, struct sk_buff *skb); struct nfc_llcp_local *nfc_llcp_find_local(struct nfc_dev *dev); int nfc_llcp_local_put(struct nfc_llcp_local *local); From 7f2ea5ed588c03d481f0301e6c3d4240132383fb Mon Sep 17 00:00:00 2001 From: Pengpeng Hou Date: Sun, 30 Aug 2026 21:29:58 +0800 Subject: [PATCH 517/592] nfc: st21nfca: validate ISO15693 inventory length The ISO15693 inventory helper removes a two-byte prefix without checking that it exists, then accepts a one-byte remainder before reading data[1] as the DSFID. Require the prefix and at least two remaining bytes before copying the UID data and reading the DSFID. Fixes: 7974728094d3 ("NFC: st21nfca: Add ISO15693 Reader/Writer support") Signed-off-by: Pengpeng Hou Link: https://patch.msgid.link/20260830132958.6397-1-pengpeng@iscas.ac.cn Signed-off-by: David Heidelberg --- drivers/nfc/st21nfca/core.c | 4 +--- 1 file changed, 1 insertion(+), 3 deletions(-) diff --git a/drivers/nfc/st21nfca/core.c b/drivers/nfc/st21nfca/core.c index 6bfeb8e7ed89..b5c1ca3acfbe 100644 --- a/drivers/nfc/st21nfca/core.c +++ b/drivers/nfc/st21nfca/core.c @@ -577,9 +577,7 @@ static int st21nfca_get_iso15693_inventory(struct nfc_hci_dev *hdev, if (r < 0) goto exit; - skb_pull(inventory_skb, 2); - - if (inventory_skb->len == 0 || + if (!skb_pull(inventory_skb, 2) || inventory_skb->len < 2 || inventory_skb->len > NFC_ISO15693_UID_MAXSIZE) { r = -EPROTO; goto exit; From c04981e42d94f39c1dba965cc462a046e946a6c5 Mon Sep 17 00:00:00 2001 From: =?UTF-8?q?=C3=96mer=20Mete=20Kaya?= Date: Wed, 9 Sep 2026 15:16:22 +0300 Subject: [PATCH 518/592] nfc: llcp: fix -ENOMEM on connect with zero-length service name MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit When service_name_len is 0, kmemdup() returns ZERO_SIZE_PTR which passes the NULL check, causing nfc_llcp_send_connect() to attempt building a zero-length service name TLV and fail with -ENOMEM. Fix by setting service_name to NULL directly when service_name_len is 0. Fixes: d646960f7986 ("NFC: Initial LLCP support") Signed-off-by: Ömer Mete Kaya Link: https://patch.msgid.link/20260909122029.34081-1-omermetekaya0@gmail.com Signed-off-by: David Heidelberg --- net/nfc/llcp_sock.c | 16 ++++++++++------ 1 file changed, 10 insertions(+), 6 deletions(-) diff --git a/net/nfc/llcp_sock.c b/net/nfc/llcp_sock.c index ce6875eb58fb..1e5ee4bcde68 100644 --- a/net/nfc/llcp_sock.c +++ b/net/nfc/llcp_sock.c @@ -759,12 +759,16 @@ static int llcp_sock_connect(struct socket *sock, struct sockaddr_unsized *_addr llcp_sock->service_name_len = min_t(unsigned int, addr->service_name_len, NFC_LLCP_MAX_SERVICE_NAME); - llcp_sock->service_name = kmemdup(addr->service_name, - llcp_sock->service_name_len, - GFP_KERNEL); - if (!llcp_sock->service_name) { - ret = -ENOMEM; - goto sock_llcp_release; + if (llcp_sock->service_name_len == 0) { + llcp_sock->service_name = NULL; + } else { + llcp_sock->service_name = kmemdup(addr->service_name, + llcp_sock->service_name_len, + GFP_KERNEL); + if (!llcp_sock->service_name) { + ret = -ENOMEM; + goto sock_llcp_release; + } } nfc_llcp_sock_link(&local->connecting_sockets, sk); From 408cff6bd60636df201274d320edfdfde9ed41db Mon Sep 17 00:00:00 2001 From: =?UTF-8?q?=C3=96mer=20Mete=20Kaya?= Date: Wed, 9 Sep 2026 15:14:31 +0300 Subject: [PATCH 519/592] nfc: llcp: fix WKS SAP hijacking via prefix match in nfc_llcp_wks_sap() MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit nfc_llcp_wks_sap() compares only service_name_len bytes, so a short service_name like "u" matches longer WKS strings like "urn:nfc:sn:snep". Fix by requiring exact length match before strncmp(). Fixes: d646960f7986 ("NFC: Initial LLCP support") Signed-off-by: Ömer Mete Kaya Link: https://patch.msgid.link/20260909121437.33744-1-omermetekaya0@gmail.com Signed-off-by: David Heidelberg --- net/nfc/llcp_core.c | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-) diff --git a/net/nfc/llcp_core.c b/net/nfc/llcp_core.c index 5ce3ce64baf2..def712e1b2b2 100644 --- a/net/nfc/llcp_core.c +++ b/net/nfc/llcp_core.c @@ -369,7 +369,8 @@ static int nfc_llcp_wks_sap(const char *service_name, size_t service_name_len) if (wks[sap] == NULL) continue; - if (strncmp(wks[sap], service_name, service_name_len) == 0) + if (strlen(wks[sap]) == service_name_len && + !strncmp(wks[sap], service_name, service_name_len)) return sap; } From 7dcf371a35632f035baf77bcf2c129165f772ce4 Mon Sep 17 00:00:00 2001 From: =?UTF-8?q?=C3=96mer=20Mete=20Kaya?= Date: Tue, 8 Sep 2026 19:18:01 +0300 Subject: [PATCH 520/592] nfc: llcp: fix slab-out-of-bounds reads when logging service names MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit nfc_llcp_wks_sap() and nfc_llcp_build_sdreq_tlv() pass non-null- terminated strings to pr_debug() using the %s format specifier. The buffers are allocated via kmemdup() or come from netlink attributes and are not guaranteed to be null-terminated, causing __dynamic_pr_debug() to read beyond the allocated region: KASAN: slab-out-of-bounds Read in __dynamic_pr_debug Fix both call sites by using %.*s with the explicit length to limit the output to the actual length of the string. Fixes: d9b8d8e19b07 ("NFC: llcp: Service Name Lookup netlink interface") Reported-by: syzbot+1e3df0852e82c21ca418@syzkaller.appspotmail.com Closes: https://syzkaller.appspot.com/bug?extid=1e3df0852e82c21ca418 Signed-off-by: Ömer Mete Kaya Link: https://patch.msgid.link/20260908161952.731468-1-omermetekaya0@gmail.com Signed-off-by: David Heidelberg --- net/nfc/llcp_commands.c | 2 +- net/nfc/llcp_core.c | 2 +- 2 files changed, 2 insertions(+), 2 deletions(-) diff --git a/net/nfc/llcp_commands.c b/net/nfc/llcp_commands.c index ca89fe967d6a..80a00938c869 100644 --- a/net/nfc/llcp_commands.c +++ b/net/nfc/llcp_commands.c @@ -135,7 +135,7 @@ struct nfc_llcp_sdp_tlv *nfc_llcp_build_sdreq_tlv(u8 tid, const char *uri, { struct nfc_llcp_sdp_tlv *sdreq; - pr_debug("uri: %s, len: %zu\n", uri, uri_len); + pr_debug("uri: %.*s, len: %zu\n", (int)uri_len, uri, uri_len); /* sdreq->tlv_len is u8, takes uri_len, + 3 for header, + 1 for NULL */ if (WARN_ON_ONCE(uri_len > U8_MAX - 4)) diff --git a/net/nfc/llcp_core.c b/net/nfc/llcp_core.c index def712e1b2b2..74bf817007cf 100644 --- a/net/nfc/llcp_core.c +++ b/net/nfc/llcp_core.c @@ -358,7 +358,7 @@ static int nfc_llcp_wks_sap(const char *service_name, size_t service_name_len) { int sap, num_wks; - pr_debug("%s\n", service_name); + pr_debug("%.*s\n", (int)service_name_len, service_name); if (service_name == NULL) return -EINVAL; From b61732f47316d45f27706db7812950145d3327b5 Mon Sep 17 00:00:00 2001 From: Deepanshu Kartikey Date: Wed, 23 Sep 2026 09:26:27 +0530 Subject: [PATCH 521/592] nfc: pn533: fix OOB read in pn533_acr122_is_rx_frame_valid() frame->ccid.datalen is read directly from the USB response frame and used, unchecked, as an index into frame->data[]. A malicious or malfunctioning device can set this field to an arbitrary value, causing the driver to read far outside the received buffer. Bound ccid.datalen against the maximum possible ACR122 frame size before using it. This replaces the existing datalen == 0 check, since datalen < 2 already covers that case and additionally rejects datalen == 1, which would still underflow the "datalen - 2" offset used below. Fixes: 9815c7cf22da ("NFC: pn533: Separate physical layer from the core implementation") Reported-by: syzbot+1853daab1a47603d4678@syzkaller.appspotmail.com Closes: https://syzkaller.appspot.com/bug?extid=1853daab1a47603d4678 Tested-by: syzbot+1853daab1a47603d4678@syzkaller.appspotmail.com Assisted-by: LLM Signed-off-by: Deepanshu Kartikey Link: https://patch.msgid.link/20260923035627.6210-1-kartikey406@gmail.com Signed-off-by: David Heidelberg --- drivers/nfc/pn533/usb.c | 4 +++- 1 file changed, 3 insertions(+), 1 deletion(-) diff --git a/drivers/nfc/pn533/usb.c b/drivers/nfc/pn533/usb.c index efb07f944fce..972eaac09e59 100644 --- a/drivers/nfc/pn533/usb.c +++ b/drivers/nfc/pn533/usb.c @@ -319,7 +319,9 @@ static bool pn533_acr122_is_rx_frame_valid(void *_frame, struct pn533 *dev) if (frame->ccid.type != 0x83) return false; - if (!frame->ccid.datalen) + if (frame->ccid.datalen < 2 || + frame->ccid.datalen > PN533_ACR122_FRAME_MAX_PAYLOAD_LEN + + PN533_ACR122_RX_FRAME_TAIL_LEN) return false; if (frame->data[frame->ccid.datalen - 2] == 0x63) From cfa165cbfbed9d0f4bbc22fef4309f595a3ab187 Mon Sep 17 00:00:00 2001 From: Victor Nogueira Date: Sun, 20 Sep 2026 14:07:01 -0300 Subject: [PATCH 522/592] net/sched: act_gate: budget the per-entry list in get_fill_size tcf_gate_get_fill_size returns only the TCA_GATE_PARMS size, but tcf_gate_dump also emits three 64-bit timestamps, the clock id, flags, priority and the variable-length TCA_GATE_ENTRY_LIST nest. The per-entry nest is unbounded: parse_gate_list places no cap on the number of sched-entries, so a gate with many entries can push the real dump well past the skb that tca_get_fill allocates from this size. RTM_NEWACTION then fails the add-notify with -EINVAL while the action is already committed to the IDR, and a subsequent RTM_GETACTION on the installed gate also returns -EINVAL because its dump no longer fits. Fix this by accounting for the missing fields in tcf_gate_get_fill_size along with all elements in the entries list. Note that sizing the reply from the action lets an oversized gate install cleanly for the first time: with the input unbounded by parse_gate_list, the sized skb can now grow well above NLMSG_GOODSIZE per netlink request (a transient GFP_KERNEL allocation reachable only with namespace-local CAP_NET_ADMIN). Overload from a malicious netns admin is hardening material, not net, per the discussion at https://lore.kernel.org/netdev/20260914191108.55a1a4f1@kernel.org/; a follow-up patch for net-next will cap the sched-entry count. Fixes: 4e76e75d6aba ("net sched actions: calculate add/delete event message size") Reported-by: Sashiko Link: https://netdev-ai.bots.linux.dev/sashiko/#/patchset/20260824153903.4143642-1-victor@mojatatu.com Tested-by: hybris Co-developed-by: Jamal Hadi Salim Signed-off-by: Jamal Hadi Salim Signed-off-by: Victor Nogueira Link: https://patch.msgid.link/QDISC-3BLH.v1.20260914203033@mojatatu.com Signed-off-by: Jakub Kicinski --- net/sched/act_gate.c | 30 +++++++++++++++++++++++++++++- 1 file changed, 29 insertions(+), 1 deletion(-) diff --git a/net/sched/act_gate.c b/net/sched/act_gate.c index 5d228a402204..6d6d45e03c07 100644 --- a/net/sched/act_gate.c +++ b/net/sched/act_gate.c @@ -681,7 +681,35 @@ static void tcf_gate_stats_update(struct tc_action *a, u64 bytes, u64 packets, static size_t tcf_gate_get_fill_size(const struct tc_action *act) { - return nla_total_size(sizeof(struct tc_gate)); + struct tcf_gate *gact = to_gate(act); + const struct tcf_gate_params *p; + struct tcfg_gate_entry *entry; + size_t size = nla_total_size(sizeof(struct tc_gate)) /* TCA_GATE_PARMS */ + + 3 * nla_total_size_64bit(sizeof(u64)) /* TCA_GATE_BASE_TIME + * TCA_GATE_CYCLE_TIME + * TCA_GATE_CYCLE_TIME_EXT + */ + + nla_total_size(sizeof(s32)) /* TCA_GATE_CLOCKID */ + + nla_total_size(sizeof(u32)) /* TCA_GATE_FLAGS */ + + nla_total_size(sizeof(s32)) /* TCA_GATE_PRIORITY */ + + nla_total_size(0); /* TCA_GATE_ENTRY_LIST */ + /* TCA_GATE_TM is budgeted by tcf_action_shared_attrs_size() */ + + rcu_read_lock(); + p = rcu_dereference(gact->param); + if (p) { + list_for_each_entry_rcu(entry, &p->entries, list) + /* TCA_GATE_ONE_ENTRY nest and its attributes */ + size += nla_total_size(0) + + nla_total_size(sizeof(u32)) /* TCA_GATE_ENTRY_INDEX */ + + nla_total_size(0) /* TCA_GATE_ENTRY_GATE */ + + nla_total_size(sizeof(u32)) /* TCA_GATE_ENTRY_INTERVAL */ + + nla_total_size(sizeof(s32)) /* TCA_GATE_ENTRY_MAX_OCTETS */ + + nla_total_size(sizeof(s32)); /* TCA_GATE_ENTRY_IPV */ + } + rcu_read_unlock(); + + return size; } static void tcf_gate_entry_destructor(void *priv) From ab1404ac81154a89fb61ac50ae9a04cd8d4834dc Mon Sep 17 00:00:00 2001 From: Fourie Zhang Date: Sun, 20 Sep 2026 19:08:43 +0800 Subject: [PATCH 523/592] net: bridge: mdb: restart port group walk after deletion br_mdb_flush_pgs() keeps a pointer-to-pointer cursor while walking mp->ports. br_multicast_del_pg() can re-enter the same MDB entry through br_multicast_sg_del_exclude_ports() and unlink other port groups. If the cursor points into one of those groups, the next iteration dereferences a stale cursor and can leave mp->ports pointing at freed memory. A following RTM_GETMDB exposes the dangling pointer: BUG: KASAN: slab-use-after-free in br_mdb_dump Read of size 8 br_mdb_dump rtnl_mdb_dump rtnl_dumpit netlink_dump Reset the cursor to mp->ports after every deletion. The deletion removes at least the selected group, so the restarted walk always makes progress. Fixes: a6acb535afb2 ("bridge: mdb: Add MDB bulk deletion support") Cc: stable@vger.kernel.org Signed-off-by: Fourie Zhang Acked-by: Nikolay Aleksandrov Link: https://patch.msgid.link/20260920110852.60293-1-fouriezhang@tencent.com Signed-off-by: Jakub Kicinski --- net/bridge/br_mdb.c | 2 ++ 1 file changed, 2 insertions(+) diff --git a/net/bridge/br_mdb.c b/net/bridge/br_mdb.c index e0c7020b12f5..a01bd280c722 100644 --- a/net/bridge/br_mdb.c +++ b/net/bridge/br_mdb.c @@ -1523,6 +1523,8 @@ static void br_mdb_flush_pgs(struct net_bridge *br, } br_multicast_del_pg(mp, p, pp); + /* br_multicast_del_pg() can remove other groups from this list. */ + pp = &mp->ports; } } From fdfec06ac1eb5cdbc18d556c70a7b26837208218 Mon Sep 17 00:00:00 2001 From: Nicolai Buchwitz Date: Tue, 22 Sep 2026 09:31:40 +0200 Subject: [PATCH 524/592] MAINTAINERS: add Nicolai Buchwitz as GENET maintainer I have been contributing to and reviewing the GENET driver for a while now. Florian asked if I would like to formalize this commitment, so add myself as a maintainer. Signed-off-by: Nicolai Buchwitz Acked-by: Florian Fainelli Acked-by: Justin Chen Link: https://patch.msgid.link/20260922073140.1471858-1-nb@tipi-net.de Signed-off-by: Jakub Kicinski --- MAINTAINERS | 1 + 1 file changed, 1 insertion(+) diff --git a/MAINTAINERS b/MAINTAINERS index 3b2eb2a7a89a..7200905501b8 100644 --- a/MAINTAINERS +++ b/MAINTAINERS @@ -5454,6 +5454,7 @@ F: include/linux/brcmphy.h BROADCOM GENET ETHERNET DRIVER M: Doug Berger M: Florian Fainelli +M: Nicolai Buchwitz R: Broadcom internal kernel review list L: netdev@vger.kernel.org S: Maintained From 7e87508b5c4d81210d0a736ed01962e52f5c4c56 Mon Sep 17 00:00:00 2001 From: Nicolai Buchwitz Date: Tue, 22 Sep 2026 15:06:38 +0200 Subject: [PATCH 525/592] net: bcmgenet: stop Tx NAPI before disabling the queues bcmgenet_netif_stop() and the Wake-on-LAN branch of bcmgenet_suspend() both disable the Tx queues first and stop Tx NAPI several steps later. A completion in flight calls netif_tx_wake_queue() in between, and nothing stops the queue again, so a transmit can reach the rings after they have been freed. Close is safe because dev_deactivate_many() stops the qdisc first. bcmgenet_suspend() does not, so stop Tx NAPI before the queues on both paths. KASAN on a Raspberry Pi CM4, driven from an MTU change because suspend freezes user space before the callback runs: BUG: KASAN: use-after-free in bcmgenet_xmit+0x17f8/0x2258 Write of size 8 at addr ffffff8055844a68 by task ksoftirqd/0/14 bcmgenet_xmit+0x17f8/0x2258 dev_hard_start_xmit+0x13c/0x588 sch_direct_xmit+0x108/0x340 __dev_queue_xmit+0x1190/0x3848 Fixes: 254f3239dd07 ("net: bcmgenet: revise suspend/resume") Signed-off-by: Nicolai Buchwitz Link: https://patch.msgid.link/20260922130639.1660797-1-nb@tipi-net.de Signed-off-by: Jakub Kicinski --- drivers/net/ethernet/broadcom/genet/bcmgenet.c | 6 ++++-- 1 file changed, 4 insertions(+), 2 deletions(-) diff --git a/drivers/net/ethernet/broadcom/genet/bcmgenet.c b/drivers/net/ethernet/broadcom/genet/bcmgenet.c index b916080f4ff1..ca62041efecd 100644 --- a/drivers/net/ethernet/broadcom/genet/bcmgenet.c +++ b/drivers/net/ethernet/broadcom/genet/bcmgenet.c @@ -3441,6 +3441,8 @@ static void bcmgenet_netif_stop(struct net_device *dev, bool stop_phy) { struct bcmgenet_priv *priv = netdev_priv(dev); + /* Stop completion polling before it can wake a stopped queue */ + bcmgenet_disable_tx_napi(priv); netif_tx_disable(dev); /* Disable MAC receive */ @@ -3455,7 +3457,6 @@ static void bcmgenet_netif_stop(struct net_device *dev, bool stop_phy) /* Disable MAC transmit. TX DMA disabled must be done before this */ umac_enable_set(priv, CMD_TX_EN, false); - bcmgenet_disable_tx_napi(priv); bcmgenet_disable_rx_napi(priv); bcmgenet_intr_disable(priv); @@ -4320,6 +4321,8 @@ static int bcmgenet_suspend(struct device *d) netif_device_detach(dev); if (device_may_wakeup(d) && priv->wolopts) { + /* Stop completion polling before it can wake a stopped queue */ + bcmgenet_disable_tx_napi(priv); netif_tx_disable(dev); /* Suspend non-wake Rx data flows */ @@ -4348,7 +4351,6 @@ static int bcmgenet_suspend(struct device *d) netdev_warn(priv->dev, "Timed out while disabling TX DMA\n"); - bcmgenet_disable_tx_napi(priv); bcmgenet_disable_rx_napi(priv); disable_irq(priv->irq1); bcmgenet_tx_reclaim_all(dev); From 7104a370714346b667712913dc16abf14bbc97ed Mon Sep 17 00:00:00 2001 From: Jakub Kicinski Date: Mon, 21 Sep 2026 16:18:56 -0700 Subject: [PATCH 526/592] veth: manage XDP program pointers during channel resize veth_set_channels() tears down XDP resources for removed RX queues without clearing rq->xdp_prog. If the program is then detached or replaced, those queues keep the old pointer after bpf_prog_put(). A later channel increase can re-enable NAPI and run the freed program. BUG: unable to handle page fault for address: ffffc90000256048 Oops: Oops: 0000 [#1] SMP KASAN NOPTI RIP: veth_xdp_rcv_skb (include/linux/filter.h:779 include/net/xdp.h:696 drivers/net/veth.c:820) Call Trace: veth_xdp_rcv (drivers/net/veth.c:941) veth_poll (drivers/net/veth.c:986) __napi_poll (net/core/dev.c:7787) net_rx_action (net/core/dev.c:7850 net/core/dev.c:8007) handle_softirqs (kernel/softirq.c:645) Kernel panic - not syncing: Fatal exception in interrupt Fixes: 4752eeb3d891 ("veth: implement support for set_channel ethtool op") Signed-off-by: Weiming Shi Acked-by: Stanislav Fomichev Reviewed-by: Jiayuan Chen Reviewed-by: Jason Xing Link: https://patch.msgid.link/20260921231856.1798630-1-kuba@kernel.org Signed-off-by: Jakub Kicinski --- drivers/net/veth.c | 2 ++ 1 file changed, 2 insertions(+) diff --git a/drivers/net/veth.c b/drivers/net/veth.c index 6ed3ee81153f..71227d0389aa 100644 --- a/drivers/net/veth.c +++ b/drivers/net/veth.c @@ -1054,6 +1054,7 @@ static int __veth_napi_enable_range(struct net_device *dev, int start, int end) for (i = start; i < end; i++) { struct veth_rq *rq = &priv->rq[i]; + rcu_assign_pointer(rq->xdp_prog, priv->_xdp_prog); napi_enable(&rq->xdp_napi); rcu_assign_pointer(priv->rq[i].napi, &priv->rq[i].xdp_napi); } @@ -1088,6 +1089,7 @@ static void veth_napi_del_range(struct net_device *dev, int start, int end) rcu_assign_pointer(priv->rq[i].napi, NULL); napi_disable(&rq->xdp_napi); + rcu_assign_pointer(rq->xdp_prog, NULL); __netif_napi_del(&rq->xdp_napi); } synchronize_net(); From 3b4e0b0c008a8c1b474730248cd5b873026c74bd Mon Sep 17 00:00:00 2001 From: Shihuang Liu Date: Sat, 19 Sep 2026 21:36:04 +0800 Subject: [PATCH 527/592] net: skbuff: fix pull-bound underflow in skb_checksum_setup_ipv6() skb_maybe_pull_tail() subtracts skb_headlen(skb) from the unsigned max argument and passes the result to __pskb_pull_tail() as a signed int. The function does not ensure that max is at least skb_headlen(skb). This can happen while parsing IPv6 extension headers when an skb already has a linear area larger than MAX_IPV6_HDR_LEN. Once the parser needs data beyond the linear area, max - skb_headlen(skb) wraps and is converted to a negative delta. __pskb_pull_tail() then passes that negative length to skb_copy_bits(), where it can become a very large copy length. Pass the requested length itself as the pull bound at the three extension-header call sites, so the delta can no longer go negative. Fixes: 1431fb31ecba ("xen-netback: fix fragment detection in checksum setup") Suggested-by: Eric Dumazet Signed-off-by: Shihuang Liu Link: https://patch.msgid.link/20260919133604.50948-1-shlomojune6@gmail.com Signed-off-by: Jakub Kicinski --- net/core/skbuff.c | 9 ++++++--- 1 file changed, 6 insertions(+), 3 deletions(-) diff --git a/net/core/skbuff.c b/net/core/skbuff.c index 609f2c7f4a47..b4edbd06655e 100644 --- a/net/core/skbuff.c +++ b/net/core/skbuff.c @@ -5977,7 +5977,8 @@ static int skb_checksum_setup_ipv6(struct sk_buff *skb, bool recalculate) err = skb_maybe_pull_tail(skb, off + sizeof(struct ipv6_opt_hdr), - MAX_IPV6_HDR_LEN); + off + + sizeof(struct ipv6_opt_hdr)); if (err < 0) goto out; @@ -5992,7 +5993,8 @@ static int skb_checksum_setup_ipv6(struct sk_buff *skb, bool recalculate) err = skb_maybe_pull_tail(skb, off + sizeof(struct ip_auth_hdr), - MAX_IPV6_HDR_LEN); + off + + sizeof(struct ip_auth_hdr)); if (err < 0) goto out; @@ -6007,7 +6009,8 @@ static int skb_checksum_setup_ipv6(struct sk_buff *skb, bool recalculate) err = skb_maybe_pull_tail(skb, off + sizeof(struct frag_hdr), - MAX_IPV6_HDR_LEN); + off + + sizeof(struct frag_hdr)); if (err < 0) goto out; From ab7aa05c06ae340e5c7530bb78fa8d23794e460b Mon Sep 17 00:00:00 2001 From: Ido Schimmel Date: Tue, 22 Sep 2026 16:12:39 +0300 Subject: [PATCH 528/592] vrf: Stop corrupting skb->csum when capturing CHECKSUM_COMPLETE packets The VRF device is an Ethernet device but it can have non-Ethernet ports such as IP tunnels. Before the cited commit, capturing packets from such ports on the VRF device resulted in these packets being detected as malformed since they lack an Ethernet header. The cited commit fixed it by pushing a dummy Ethernet header to such packets before the capture and pulling it afterwards. In the case of CHECKSUM_COMPLETE packets it also updated skb->csum with the checksum of the dummy Ethernet header. This is wrong as skb->csum should not include the checksum of the Ethernet header ("checksum of the _whole_ packet as seen by netif_rx()"). This also means that L4 protocols receive a corrupted skb->csum and potentially drop the packet, as is the case with UDP packets whose checksum was completed by software. Fix by removing the unnecessary call to skb_postpush_rcsum(). Fixes: 048939088220 ("vrf: add mac header for tunneled packets when sniffer is attached") Reported-by: Stefano Sasso Closes: https://lore.kernel.org/netdev/CALtE316UtL3x7LL6uxfXzx8rW6AbzYPeDOb478hqJCr_-dj=Wg@mail.gmail.com/ Signed-off-by: Ido Schimmel Reviewed-by: David Ahern Reviewed-by: Eric Dumazet Reviewed-by: Andrea Mayer Link: https://patch.msgid.link/20260922131239.2509494-1-idosch@nvidia.com Signed-off-by: Jakub Kicinski --- drivers/net/vrf.c | 2 -- 1 file changed, 2 deletions(-) diff --git a/drivers/net/vrf.c b/drivers/net/vrf.c index a0557a3a7026..d4dc6d690a75 100644 --- a/drivers/net/vrf.c +++ b/drivers/net/vrf.c @@ -1175,8 +1175,6 @@ static int vrf_prepare_mac_header(struct sk_buff *skb, skb->protocol = eth->h_proto; skb->pkt_type = PACKET_HOST; - skb_postpush_rcsum(skb, skb->data, ETH_HLEN); - skb_pull_inline(skb, ETH_HLEN); return 0; From 2d959c75c27f90e9ec489d18ce5ee6b852ad4741 Mon Sep 17 00:00:00 2001 From: Hui Peng Date: Mon, 21 Sep 2026 04:40:25 +0000 Subject: [PATCH 529/592] ipv6: sr: enforce exact attribute length for SEG6_ATTR_DST In seg6_genl_policy, SEG6_ATTR_DST is defined with .type = NLA_BINARY and .len = sizeof(struct in6_addr). For NLA_BINARY, .len only enforces the maximum payload length and permits shorter payloads (e.g., 0 bytes). When seg6_genl_set_tunsrc() copies sizeof(struct in6_addr) bytes via kmemdup(val, sizeof(*val), GFP_KERNEL), a short SEG6_ATTR_DST attribute triggers a 16-byte out-of-bounds read past skb->tail into uninitialized skb->head memory, which is stored in sdata->tun_src and leaked back to userspace via SEG6_CMD_GET_TUNSRC. Switch SEG6_ATTR_DST in seg6_genl_policy to NLA_POLICY_EXACT_LEN(sizeof(struct in6_addr)) so that generic netlink validation rejects any attribute whose length is not exactly sizeof(struct in6_addr) with -ERANGE. Tested in QEMU against Linux 7.3.0-rc3 by sending a SEG6_CMD_SET_TUNSRC Generic Netlink message with a 0-byte SEG6_ATTR_DST attribute followed by SEG6_CMD_GET_TUNSRC. On the unfixed kernel, SEG6_CMD_SET_TUNSRC succeeds (err = 0) and SEG6_CMD_GET_TUNSRC leaks 16 bytes of uninitialized kernel heap memory (tun_src = 836a61ecc4d25a1042a8d60411cfb378); with this patch applied, SEG6_CMD_SET_TUNSRC is rejected by netlink policy validation with -ERANGE (-34) and tun_src remains zeroed. Fixes: 915d7e5e5930 ("ipv6: sr: add code base for control plane support of SR-IPv6") Cc: stable@vger.kernel.org Signed-off-by: Hui Peng Reviewed-by: Hangbin Liu Reviewed-by: Justin Iurman Reviewed-by: Andrea Mayer Link: https://patch.msgid.link/20260921044025.1535982-1-benquike@gmail.com Signed-off-by: Jakub Kicinski --- net/ipv6/seg6.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/net/ipv6/seg6.c b/net/ipv6/seg6.c index 62a7eb779202..8c2b156c227a 100644 --- a/net/ipv6/seg6.c +++ b/net/ipv6/seg6.c @@ -138,8 +138,8 @@ void seg6_icmp_srh(struct sk_buff *skb, struct inet6_skb_parm *opt) static struct genl_family seg6_genl_family; static const struct nla_policy seg6_genl_policy[SEG6_ATTR_MAX + 1] = { - [SEG6_ATTR_DST] = { .type = NLA_BINARY, - .len = sizeof(struct in6_addr) }, + [SEG6_ATTR_DST] = + NLA_POLICY_EXACT_LEN(sizeof(struct in6_addr)), [SEG6_ATTR_DSTLEN] = { .type = NLA_S32, }, [SEG6_ATTR_HMACKEYID] = { .type = NLA_U32, }, [SEG6_ATTR_SECRET] = { .type = NLA_BINARY, }, From 6b491af01aa5c0633a580e3b11bc7277adc903b2 Mon Sep 17 00:00:00 2001 From: Nicolo Giuliani Date: Mon, 21 Sep 2026 05:56:29 +0200 Subject: [PATCH 530/592] net: dsa: mv88e6xxx: 88E6191X and 88E6193X have no PTP The 88E6191X and 88E6193X are 6393 family devices that share mv88e6393x_ops with the 88E6393X and are marked as ptp_support. Marvell's UMSD driver describes both as parts without AVB (88E6193X: "BGA package - No AVB, No Routing, No Cut-through"), and the register access confirms it on an 88E6193X: the whole indirect AVB register space behind Global 2 registers 0x16 and 0x17 reads zero, for every port, block and address, with the 6390 and with the 6352 command encoding. Writes to the TAI registers, including the clock period register and the TAI global configuration register, read back as zero. Since commit 7e3c18097a70 ("net: dsa: mv88e6xxx: read cycle counter period from hardware") the PTP setup reads the TAI clock period, so the switch fails to probe: mv88e6xxx ...: unexpected cycle counter period of 0 ps Add mv88e6191x_ops, a copy of mv88e6393x_ops without avb_ops and ptp_ops, use it for the 88E6191X and the 88E6193X and stop setting ptp_support for them. The 88E6393X is unchanged. Tested on an 88E6193X (Sophos XGS 107w): the switch probes and the ports work. I do not have an 88E6191X, it is changed because UMSD describes it the same way. Fixes: de776d0d316f ("net: dsa: mv88e6xxx: add support for mv88e6393x family") Suggested-by: Andrew Lunn Signed-off-by: Nicolo Giuliani Reviewed-by: Andrew Lunn Link: https://patch.msgid.link/20260921-send-net-v2-1-031ad720f140@studio.unibo.it Signed-off-by: Jakub Kicinski --- drivers/net/dsa/mv88e6xxx/chip.c | 68 ++++++++++++++++++++++++++++++-- 1 file changed, 64 insertions(+), 4 deletions(-) diff --git a/drivers/net/dsa/mv88e6xxx/chip.c b/drivers/net/dsa/mv88e6xxx/chip.c index 7f68a0c55802..a4a8c7e11bf4 100644 --- a/drivers/net/dsa/mv88e6xxx/chip.c +++ b/drivers/net/dsa/mv88e6xxx/chip.c @@ -5639,6 +5639,68 @@ static const struct mv88e6xxx_ops mv88e6390x_ops = { .pcs_ops = &mv88e6390_pcs_ops, }; +static const struct mv88e6xxx_ops mv88e6191x_ops = { + /* MV88E6XXX_FAMILY_6393 without AVB and PTP: 6191X and 6193X */ + .irl_init_all = mv88e6390_g2_irl_init_all, + .get_eeprom = mv88e6xxx_g2_get_eeprom8, + .set_eeprom = mv88e6xxx_g2_set_eeprom8, + .set_switch_mac = mv88e6xxx_g2_set_switch_mac, + .phy_read = mv88e6xxx_g2_smi_phy_read_c22, + .phy_write = mv88e6xxx_g2_smi_phy_write_c22, + .phy_read_c45 = mv88e6xxx_g2_smi_phy_read_c45, + .phy_write_c45 = mv88e6xxx_g2_smi_phy_write_c45, + .port_set_link = mv88e6xxx_port_set_link, + .port_sync_link = mv88e6xxx_port_sync_link, + .port_set_rgmii_delay = mv88e6390_port_set_rgmii_delay, + .port_set_speed_duplex = mv88e6393x_port_set_speed_duplex, + .port_tag_remap = mv88e6390_port_tag_remap, + .port_set_policy = mv88e6393x_port_set_policy, + .port_set_frame_mode = mv88e6351_port_set_frame_mode, + .port_set_ucast_flood = mv88e6352_port_set_ucast_flood, + .port_set_mcast_flood = mv88e6352_port_set_mcast_flood, + .port_set_ether_type = mv88e6393x_port_set_ether_type, + .port_set_jumbo_size = mv88e6165_port_set_jumbo_size, + .port_egress_rate_limiting = mv88e6097_port_egress_rate_limiting, + .port_pause_limit = mv88e6390_port_pause_limit, + .port_disable_learn_limit = mv88e6xxx_port_disable_learn_limit, + .port_disable_pri_override = mv88e6xxx_port_disable_pri_override, + .port_get_cmode = mv88e6352_port_get_cmode, + .port_set_cmode = mv88e6393x_port_set_cmode, + .port_setup_message_port = mv88e6xxx_setup_message_port, + .port_set_upstream_port = mv88e6393x_port_set_upstream_port, + .port_enable_tcam = mv88e6xxx_port_enable_tcam, + .stats_snapshot = mv88e6390_g1_stats_snapshot, + .stats_set_histogram = mv88e6390_g1_stats_set_histogram, + .stats_get_sset_count = mv88e6320_stats_get_sset_count, + .stats_get_strings = mv88e6320_stats_get_strings, + .stats_get_stat = mv88e6390_stats_get_stat, + /* .set_cpu_port is missing because this family does not support a global + * CPU port, only per port CPU port which is set via + * .port_set_upstream_port method. + */ + .set_egress_port = mv88e6393x_set_egress_port, + .watchdog_ops = &mv88e6393x_watchdog_ops, + .mgmt_rsvd2cpu = mv88e6393x_port_mgmt_rsvd2cpu, + .pot_clear = mv88e6xxx_g2_pot_clear, + .hardware_reset_pre = mv88e6xxx_g2_eeprom_wait, + .hardware_reset_post = mv88e6xxx_g2_eeprom_wait, + .reset = mv88e6352_g1_reset, + .rmu_disable = mv88e6390_g1_rmu_disable, + .atu_get_hash = mv88e6165_g1_atu_get_hash, + .atu_set_hash = mv88e6165_g1_atu_set_hash, + .vtu_getnext = mv88e6390_g1_vtu_getnext, + .vtu_loadpurge = mv88e6390_g1_vtu_loadpurge, + .stu_getnext = mv88e6390_g1_stu_getnext, + .stu_loadpurge = mv88e6390_g1_stu_loadpurge, + .serdes_get_lane = mv88e6393x_serdes_get_lane, + .serdes_irq_mapping = mv88e6390_serdes_irq_mapping, + /* TODO: serdes stats */ + .gpio_ops = &mv88e6352_gpio_ops, + .phylink_get_caps = mv88e6393x_phylink_get_caps, + .pcs_ops = &mv88e6393x_pcs_ops, + .tcam_ops = &mv88e6393_tcam_ops, +}; + static const struct mv88e6xxx_ops mv88e6393x_ops = { /* MV88E6XXX_FAMILY_6393 */ .irl_init_all = mv88e6390_g2_irl_init_all, @@ -6163,8 +6225,7 @@ static const struct mv88e6xxx_info mv88e6xxx_table[] = { .atu_move_port_mask = 0x1f, .pvt = true, .multi_chip = true, - .ptp_support = true, - .ops = &mv88e6393x_ops, + .ops = &mv88e6191x_ops, }, [MV88E6193X] = { @@ -6190,8 +6251,7 @@ static const struct mv88e6xxx_info mv88e6xxx_table[] = { .atu_move_port_mask = 0x1f, .pvt = true, .multi_chip = true, - .ptp_support = true, - .ops = &mv88e6393x_ops, + .ops = &mv88e6191x_ops, }, [MV88E6220] = { From d22609f3d13fc5baacd92c222731b03c593401db Mon Sep 17 00:00:00 2001 From: Hui Peng Date: Mon, 21 Sep 2026 04:59:20 +0000 Subject: [PATCH 531/592] fou: reject omitted FOU_ATTR_IPPROTO on FOU_ENCAP_DIRECT Commit 7a9bc9e3f423 ("fou: Don't allow 0 for FOU_ATTR_IPPROTO.") added NLA_POLICY_MIN(NLA_U8, 1) to fou_nl_policy[FOU_ATTR_IPPROTO], which rejects an explicitly supplied FOU_ATTR_IPPROTO == 0 attribute with -ERANGE. However, FOU_ATTR_IPPROTO is an optional netlink attribute. When a user sends FOU_CMD_ADD with FOU_ATTR_TYPE set to FOU_ENCAP_DIRECT and omits FOU_ATTR_IPPROTO entirely, nla_policy validation succeeds and parse_nl_config() leaves cfg->protocol as 0 (from memset(cfg, 0, sizeof(*cfg))). fou_create() then creates a FOU_ENCAP_DIRECT socket with fou->protocol == 0. In fou_udp_recv(), returning -fou->protocol to udp_queue_rcv_one_skb() triggers IP protocol resubmission when fou->protocol > 0, whereas returning 0 tells the UDP tunnel layer that the skb was consumed without freeing it. When fou->protocol == 0, every packet received on the socket returns 0 from fou_udp_recv() and leaks the sk_buff. Reject FOU_ENCAP_DIRECT when !cfg->protocol in fou_create() so that creating a direct encapsulation port without FOU_ATTR_IPPROTO fails with -EINVAL while leaving FOU_CMD_DEL and FOU_CMD_GET (which share parse_nl_config()) unaffected. Tested in QEMU against Linux 7.3.0-rc3 by sending a FOU_CMD_ADD Generic Netlink request with FOU_ATTR_PORT = 5555 and FOU_ATTR_TYPE = FOU_ENCAP_DIRECT while omitting FOU_ATTR_IPPROTO. On the unfixed kernel, FOU_CMD_ADD succeeds (err = 0), FOU_CMD_GET reports fou->type = 1 and fou->protocol = 0, and sending 4000 UDP packets to 127.0.0.1:5555 leaks all 4000 sk_buffs (SUnreclaim in /proc/meminfo grows from 41456 kB to 59008 kB, +17552 kB); with this patch applied, FOU_CMD_ADD is rejected with -EINVAL (-22). Fixes: 23461551c006 ("fou: Support for foo-over-udp RX path") Fixes: 7a9bc9e3f423 ("fou: Don't allow 0 for FOU_ATTR_IPPROTO.") Cc: stable@vger.kernel.org Signed-off-by: Hui Peng Reviewed-by: Hangbin Liu Link: https://patch.msgid.link/20260921045920.1613098-1-benquike@gmail.com Signed-off-by: Jakub Kicinski --- net/ipv4/fou_core.c | 4 ++++ 1 file changed, 4 insertions(+) diff --git a/net/ipv4/fou_core.c b/net/ipv4/fou_core.c index 5e867f1b5c1d..076fdca44f51 100644 --- a/net/ipv4/fou_core.c +++ b/net/ipv4/fou_core.c @@ -600,6 +600,10 @@ static int fou_create(struct net *net, struct fou_cfg *cfg, /* Initial for fou type */ switch (cfg->type) { case FOU_ENCAP_DIRECT: + if (!cfg->protocol) { + err = -EINVAL; + goto error; + } tunnel_cfg.encap_rcv = fou_udp_recv; tunnel_cfg.gro_receive = fou_gro_receive; tunnel_cfg.gro_complete = fou_gro_complete; From 89a8a1eef2d441b7825a6c0116ce817235a7ffe6 Mon Sep 17 00:00:00 2001 From: Jonas Jelonek Date: Fri, 18 Sep 2026 21:19:55 +0000 Subject: [PATCH 532/592] net: mdio: realtek-rtl9300: fix RTL931x C22 extended page selection The RTL931x indirect access engine has a separate nine-bit extended page field. The driver leaves it at zero, and otto_emdio_run_cmd() therefore programs extended page zero for every Clause 22 transaction. This overrides page selection made through PHY register 30, causing accesses to private PHY pages to hit extended page zero instead. Set the field to its 0x1ff "do not change" value for RTL931x Clause 22 reads and writes. This preserves extended page selection made through PHY register 30 and restores access to its private register pages. Fixes: 5ebdcac59aff ("net: mdio: realtek-rtl9300: Add support for RTL931x") Signed-off-by: Jonas Jelonek Acked-by: Markus Stockhausen Link: https://patch.msgid.link/20260918211955.3955777-1-jonas@jonasjelonek.de Signed-off-by: Jakub Kicinski --- drivers/net/mdio/mdio-realtek-rtl9300.c | 6 ++++++ 1 file changed, 6 insertions(+) diff --git a/drivers/net/mdio/mdio-realtek-rtl9300.c b/drivers/net/mdio/mdio-realtek-rtl9300.c index afd52a1cd7f8..9ce2b7807532 100644 --- a/drivers/net/mdio/mdio-realtek-rtl9300.c +++ b/drivers/net/mdio/mdio-realtek-rtl9300.c @@ -88,6 +88,8 @@ #define RTL9310_SMI_INDRT_ACCESS_BC_PHYID_CTRL 0x0c14 #define RTL9310_BC_PORT_ID GENMASK(10, 5) #define RTL9310_SMI_INDRT_ACCESS_CTRL_1 0x0c04 +#define RTL9310_SMI_INDRT_EXT_PAGE GENMASK(8, 0) +#define RTL9310_SMI_INDRT_EXT_PAGE_NO_CHANGE 0x1ff #define RTL9310_SMI_INDRT_ACCESS_CTRL_2_LOW 0x0c08 #define RTL9310_SMI_INDRT_ACCESS_CTRL_2_HIGH 0x0c0c #define RTL9310_SMI_INDRT_ACCESS_CTRL_3 0x0c10 /* I/O fields flipped */ @@ -325,6 +327,8 @@ static int otto_emdio_9310_read_c22(struct mii_bus *bus, int port, int regnum, u .broadcast = FIELD_PREP(RTL9310_BC_PORT_ID, port), .c22_data = FIELD_PREP(RTL9310_PHY_CTRL_REG_ADDR, regnum) | FIELD_PREP(RTL9310_PHY_CTRL_MAIN_PAGE, RAW_PAGE(priv)), + .ext_page = FIELD_PREP(RTL9310_SMI_INDRT_EXT_PAGE, + RTL9310_SMI_INDRT_EXT_PAGE_NO_CHANGE), }; return otto_emdio_read_cmd(bus, RTL9310_PHY_CTRL_TYPE_C22, &cmd_data, @@ -337,6 +341,8 @@ static int otto_emdio_9310_write_c22(struct mii_bus *bus, int port, int regnum, struct otto_emdio_cmd_regs cmd_data = { .c22_data = FIELD_PREP(RTL9310_PHY_CTRL_REG_ADDR, regnum) | FIELD_PREP(RTL9310_PHY_CTRL_MAIN_PAGE, RAW_PAGE(priv)), + .ext_page = FIELD_PREP(RTL9310_SMI_INDRT_EXT_PAGE, + RTL9310_SMI_INDRT_EXT_PAGE_NO_CHANGE), .io_data = FIELD_PREP(RTL9310_PHY_CTRL_INDATA, value), .port_mask_high = (u32)(BIT_ULL(port) >> 32), .port_mask_low = (u32)(BIT_ULL(port)), From db762fd96be225bd06161c9631160c755d891693 Mon Sep 17 00:00:00 2001 From: Johan Almbladh Date: Wed, 23 Sep 2026 12:51:57 +0200 Subject: [PATCH 533/592] bpf: Fix immediate JMP JEQ/JNE on MIPS32 An addu instruction was emitted instead of addiu, causing the immediate value 1 to be interpreted as register $at. This made the comparison result invalid when the immediate operand was negative. Note that $at is mapped to BPF_REG_AX, which is used for constant blinding. Fix the instruction to use the immediate form. Found with test_bpf on MIPS32r1 emulated by QEMU. Fixes: eb63cfcd2ee8 ("mips, bpf: Add eBPF JIT for 32-bit MIPS") Signed-off-by: Johan Almbladh Signed-off-by: Alexei Starovoitov Link: https://patch.msgid.link/20260923105158.3514342-1-johan.almbladh@anyfinetworks.com --- arch/mips/net/bpf_jit_comp32.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/arch/mips/net/bpf_jit_comp32.c b/arch/mips/net/bpf_jit_comp32.c index 40a878b672f5..15a2a153dc87 100644 --- a/arch/mips/net/bpf_jit_comp32.c +++ b/arch/mips/net/bpf_jit_comp32.c @@ -1111,7 +1111,7 @@ static void emit_jmp_i64(struct jit_context *ctx, emit(ctx, xor, tmp, lo(dst), tmp); } if (imm < 0) { /* Compare sign extension */ - emit(ctx, addu, MIPS_R_T9, hi(dst), 1); + emit(ctx, addiu, MIPS_R_T9, hi(dst), 1); emit(ctx, or, tmp, tmp, MIPS_R_T9); } else { /* Compare zero extension */ emit(ctx, or, tmp, tmp, hi(dst)); From 8110ba09777873443db286b3cbb89b0e6311c554 Mon Sep 17 00:00:00 2001 From: Johan Almbladh Date: Wed, 23 Sep 2026 12:51:58 +0200 Subject: [PATCH 534/592] bpf: Fix BSWAP 32 and 16 on MIPS64 The 16/32-bit byteswap implementations for MIPS64r1 and earlier do not have an explicit zero extension afterwards. The input is first sign-extended to 64 bits, and the byteswap sequence can then leave the result sign-extended depending on the value of the low bits. Add the missing zero-extension. Found with test_bpf on MIPS64r1 emulated by QEMU. Fixes: fbc802de6b10 ("mips, bpf: Add new eBPF JIT for 64-bit MIPS") Signed-off-by: Johan Almbladh Signed-off-by: Alexei Starovoitov Link: https://patch.msgid.link/20260923105158.3514342-2-johan.almbladh@anyfinetworks.com --- arch/mips/net/bpf_jit_comp64.c | 3 +-- 1 file changed, 1 insertion(+), 2 deletions(-) diff --git a/arch/mips/net/bpf_jit_comp64.c b/arch/mips/net/bpf_jit_comp64.c index fa7e9aa37f49..6681ccac9dd9 100644 --- a/arch/mips/net/bpf_jit_comp64.c +++ b/arch/mips/net/bpf_jit_comp64.c @@ -305,8 +305,7 @@ static void emit_bswap_r64(struct jit_context *ctx, u8 dst, u32 width) case 16: emit_sext(ctx, dst, dst); emit_bswap_r(ctx, dst, width); - if (cpu_has_mips64r2 || cpu_has_mips64r6) - emit_zext(ctx, dst); + emit_zext(ctx, dst); break; } clobber_reg(ctx, dst); From d8b6529e80bcb4fb8177121404cbb3377acaebd2 Mon Sep 17 00:00:00 2001 From: Zijie Huang Date: Mon, 21 Sep 2026 01:36:19 +0800 Subject: [PATCH 535/592] net: arp: terminate device name before lookup The ARP ioctl copies a user-provided struct arpreq into a stack object. Its arp_dev field may contain IFNAMSIZ bytes without a NUL terminator. Such input is passed to dev_get_by_name_rcu() or __dev_get_by_name(), where strcmp() can read past the end of the stack object when a matching alternative interface name exists. Terminate the field before the lookup to prevent the out-of-bounds read. Fixes: 36fbf1e52bd3 ("net: rtnetlink: add linkprop commands to add and delete alternative ifnames") Cc: stable@vger.kernel.org Reported-by: Vega Signed-off-by: Zijie Huang Signed-off-by: Ren Wei Reviewed-by: Ido Schimmel Link: https://patch.msgid.link/fabf02a70787d17299e4b3153eadffaf20d154b3.1789910973.git.milkory@outlook.com Signed-off-by: Jakub Kicinski --- net/ipv4/arp.c | 1 + 1 file changed, 1 insertion(+) diff --git a/net/ipv4/arp.c b/net/ipv4/arp.c index d409f606aec0..60009d92e071 100644 --- a/net/ipv4/arp.c +++ b/net/ipv4/arp.c @@ -1278,6 +1278,7 @@ int arp_ioctl(struct net *net, unsigned int cmd, void __user *arg) err = copy_from_user(&r, arg, sizeof(struct arpreq)); if (err) return -EFAULT; + r.arp_dev[IFNAMSIZ - 1] = '\0'; break; default: return -EINVAL; From 4eb3f195ef08c5acaed87958297e41cc49588dde Mon Sep 17 00:00:00 2001 From: Ivan Delalande Date: Fri, 18 Sep 2026 15:47:15 -0700 Subject: [PATCH 536/592] tg3: use random MAC address when tg3_get_device_address fails Some of the tg3 NICs we use (BCM57762) reset the SRAM MAC address to the placeholder address on link flaps, tg3_chip_reset, etc. We've typically fixed it from userspace, but since e4c00ba7274b ("tg3: replace placeholder MAC address with device property") was merged, tg3 just fails probe as we don't have a way to get it through the generic device_get_mac_address infrastructure as fallback on our systems. Make the driver assign a random address in this condition instead of being fatal for probe. Set deferred_probe_reason through dev_warn_probe if the address isn't yet available from the provider. Fixes: e4c00ba7274b ("tg3: replace placeholder MAC address with device property") Suggested-by: Jakub Kicinski Link: https://lore.kernel.org/netdev/20260909191751.651aa5c4@kernel.org/ Signed-off-by: Ivan Delalande Link: https://patch.msgid.link/20260918224715.GA654128@visor Signed-off-by: Jakub Kicinski --- drivers/net/ethernet/broadcom/tg3.c | 11 +++++++---- 1 file changed, 7 insertions(+), 4 deletions(-) diff --git a/drivers/net/ethernet/broadcom/tg3.c b/drivers/net/ethernet/broadcom/tg3.c index caa7a6caa6e2..8b6806a79edf 100644 --- a/drivers/net/ethernet/broadcom/tg3.c +++ b/drivers/net/ethernet/broadcom/tg3.c @@ -17915,11 +17915,14 @@ static int tg3_init_one(struct pci_dev *pdev, err = tg3_get_device_address(tp, addr); if (err) { - dev_err(&pdev->dev, - "Could not obtain valid ethernet address, aborting\n"); - goto err_out_apeunmap; + dev_warn_probe(&pdev->dev, err, + "Could not obtain a valid ethernet address\n"); + if (err == -EPROBE_DEFER) + goto err_out_apeunmap; + eth_hw_addr_random(dev); + } else { + eth_hw_addr_set(dev, addr); } - eth_hw_addr_set(dev, addr); intmbx = MAILBOX_INTERRUPT_0 + TG3_64BIT_REG_LOW; rcvmbx = MAILBOX_RCVRET_CON_IDX_0 + TG3_64BIT_REG_LOW; From 0a7822e34a0bfde31b194ac3da3253e5032b44cc Mon Sep 17 00:00:00 2001 From: Lorenzo Bianconi Date: Mon, 21 Sep 2026 16:46:18 +0200 Subject: [PATCH 537/592] net: stmmac: clear stale buf->page after recycling on skb build failure In stmmac_rx(), when napi_build_skb() fails the descriptor page is recycled back to the page pool with page_pool_recycle_direct(), but buf->page is left pointing at the recycled page, unlike every other consumption site in the function which clears the pointer after handing the page away. With the stale pointer stmmac_rx_refill() skips the replacement allocation and programs the already-recycled page back into the RX descriptor. Clear buf->page on the napi_build_skb() failure path to keep the buffer lifecycle consistent with the other consumption sites. Fixes: df542f669307 ("net: stmmac: Switch to zero-copy in non-XDP RX path") Signed-off-by: Lorenzo Bianconi Reviewed-by: Maxime Chevallier Link: https://patch.msgid.link/20260921-stmmac-fix-napi-build-skb-error-v1-1-3d54bf6d9bb6@oss.qualcomm.com Signed-off-by: Jakub Kicinski --- drivers/net/ethernet/stmicro/stmmac/stmmac_main.c | 1 + 1 file changed, 1 insertion(+) diff --git a/drivers/net/ethernet/stmicro/stmmac/stmmac_main.c b/drivers/net/ethernet/stmicro/stmmac/stmmac_main.c index d5a984ad864f..3f34d491c959 100644 --- a/drivers/net/ethernet/stmicro/stmmac/stmmac_main.c +++ b/drivers/net/ethernet/stmicro/stmmac/stmmac_main.c @@ -5889,6 +5889,7 @@ static int stmmac_rx(struct stmmac_priv *priv, int limit, u32 queue) if (!skb) { page_pool_recycle_direct(rx_q->page_pool, buf->page); + buf->page = NULL; rx_dropped++; count++; goto drain_data; From 87cd6b717e4069dca34e0866e57eaeb44d3b173e Mon Sep 17 00:00:00 2001 From: Yuya Kusakabe Date: Tue, 22 Sep 2026 05:49:56 +0900 Subject: [PATCH 538/592] net: ipv6: keep room for the mac header in dst_dev_overhead() The seg6, ioam6 and rpl lwtunnels size their skb_cow_head() request as the length they are about to push plus dst_dev_overhead(), then push the new headers and rebuild the mac header below them with skb_mac_header_rebuild(). That rebuild needs skb->mac_len of headroom, but dst_dev_overhead() leaves LL_RESERVED_SPACE() of the egress device, 16 bytes for plain Ethernet. Where the mac header is longer than that, as it is on ingress through a VLAN device with reorder_hdr off, the rebuild runs out of room: skb_set_mac_header(skb, -skb->mac_len) computes a negative offset, stores it unchecked in the u16 skb->mac_header, and the memmove that follows writes skb->mac_len bytes about 64 KB past skb->head. Forwarding plain ping6 traffic through such a device reproduces it on all five seg6 encapsulation modes and on the rpl and ioam6 inline paths; skb->mac_header comes back as 65534 on a 704-byte head. Return the larger of the two. The helper already returns skb->mac_len when it has no dst, so this only makes the other branch agree, and it covers every caller rather than each call site in turn. Fixes: 40475b63761a ("net: ipv6: seg6_iptunnel: mitigate 2-realloc issue") Fixes: dce525185bc9 ("net: ipv6: ioam6_iptunnel: mitigate 2-realloc issue") Fixes: 985ec6f5e623 ("net: ipv6: rpl_iptunnel: mitigate 2-realloc issue") Suggested-by: Andrea Mayer Signed-off-by: Yuya Kusakabe Reviewed-by: Justin Iurman Reviewed-by: Gabriel Goller Reviewed-by: Eric Dumazet Reviewed-by: Andrea Mayer Link: https://patch.msgid.link/20260922-seg6-maclen-headroom-v3-1-7b2f982ef79d@gmail.com Signed-off-by: Jakub Kicinski --- include/net/dst.h | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-) diff --git a/include/net/dst.h b/include/net/dst.h index 307073eae7f8..dbedfe72e1fd 100644 --- a/include/net/dst.h +++ b/include/net/dst.h @@ -455,7 +455,8 @@ static inline unsigned int dst_dev_overhead(struct dst_entry *dst, struct sk_buff *skb) { if (likely(dst)) - return LL_RESERVED_SPACE(dst->dev); + return max_t(unsigned int, skb->mac_len, + LL_RESERVED_SPACE(dst->dev)); return skb->mac_len; } From 0e2bec77ea62895416600c90588f593516572bca Mon Sep 17 00:00:00 2001 From: Florian Fainelli Date: Mon, 21 Sep 2026 15:00:17 -0700 Subject: [PATCH 539/592] net: bcmgenet: fix 64-bit RTNL stats reading in ethtool on 32-bit systems When bcmgenet was converted to 64-bit statistics, STAT_RTNL members were switched to point into struct rtnl_link_stats64, whose fields are 64-bit (__u64) regardless of architecture. However, bcmgenet_get_ethtool_stats() retained a legacy check: if (sizeof(unsigned long) != sizeof(u32) && s->stat_sizeof == sizeof(unsigned long)) On 32-bit systems, sizeof(unsigned long) == sizeof(u32), causing this condition to evaluate to false. As a result, 64-bit RTNL stats fields were read via *(u32 *)p. On 32-bit Big-Endian systems (such as MIPS BE), this reads the high 32 bits and returns 0 until the counter exceeds 4GB; on 32-bit Little-Endian systems (such as 32-bit ARM), the value is truncated to 32 bits. Fix this by checking if s->stat_sizeof == sizeof(u64) so 64-bit fields are always read as 64-bit values. Fixes: 59aa6e3072aa ("net: bcmgenet: switch to use 64bit statistics") Reviewed-by: Nicolai Buchwitz Signed-off-by: Florian Fainelli Link: https://patch.msgid.link/20260921220021.281418-2-florian.fainelli@broadcom.com Signed-off-by: Jakub Kicinski --- drivers/net/ethernet/broadcom/genet/bcmgenet.c | 5 ++--- 1 file changed, 2 insertions(+), 3 deletions(-) diff --git a/drivers/net/ethernet/broadcom/genet/bcmgenet.c b/drivers/net/ethernet/broadcom/genet/bcmgenet.c index ca62041efecd..47a6c073c8d9 100644 --- a/drivers/net/ethernet/broadcom/genet/bcmgenet.c +++ b/drivers/net/ethernet/broadcom/genet/bcmgenet.c @@ -1346,9 +1346,8 @@ static void bcmgenet_get_ethtool_stats(struct net_device *dev, p = (char *)&stats64; p += s->stat_offset; - if (sizeof(unsigned long) != sizeof(u32) && - s->stat_sizeof == sizeof(unsigned long)) - data[i] = *(unsigned long *)p; + if (s->stat_sizeof == sizeof(u64)) + data[i] = *(u64 *)p; else data[i] = *(u32 *)p; } From 3aeaa609fda19c09d5298c9fedaaa3b6229601b5 Mon Sep 17 00:00:00 2001 From: Florian Fainelli Date: Mon, 21 Sep 2026 15:00:18 -0700 Subject: [PATCH 540/592] net: bcmgenet: initialize u64 stats seq counter for all queues bcmgenet_gstrings_stats statically defines ethtool statistics for queues 0 through GENET_MAX_MQ_CNT (4). However, bcmgenet_probe() only initialized the u64_stats_sync seq counter up to priv->hw_params->rx_queues and priv->hw_params->tx_queues. Since priv->hw_params->rx_queues is 0 across all hardware versions (and priv->hw_params->tx_queues is 0 on GENET V1), rings 1..4 have uninitialized u64_stats_sync structures. When ethtool -S is run on 32-bit kernels, bcmgenet_get_ethtool_stats() reads stats from rx_rings[1..4], causing lockdep warnings due to the uninitialized sequence counters. Initialize the sequence counters for all GENET_MAX_MQ_CNT + 1 queues. Fixes: ffc2c8c4a714 ("net: bcmgenet: Initialize u64 stats seq counter") Reviewed-by: Nicolai Buchwitz Signed-off-by: Florian Fainelli Link: https://patch.msgid.link/20260921220021.281418-3-florian.fainelli@broadcom.com Signed-off-by: Jakub Kicinski --- drivers/net/ethernet/broadcom/genet/bcmgenet.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/drivers/net/ethernet/broadcom/genet/bcmgenet.c b/drivers/net/ethernet/broadcom/genet/bcmgenet.c index 47a6c073c8d9..2ab37bb031ed 100644 --- a/drivers/net/ethernet/broadcom/genet/bcmgenet.c +++ b/drivers/net/ethernet/broadcom/genet/bcmgenet.c @@ -4135,10 +4135,10 @@ static int bcmgenet_probe(struct platform_device *pdev) priv->rx_rings[i].rx_max_coalesced_frames = 1; /* Initialize u64 stats seq counter for 32bit machines */ - for (i = 0; i <= priv->hw_params->rx_queues; i++) + for (i = 0; i <= GENET_MAX_MQ_CNT; i++) { u64_stats_init(&priv->rx_rings[i].stats64.syncp); - for (i = 0; i <= priv->hw_params->tx_queues; i++) u64_stats_init(&priv->tx_rings[i].stats64.syncp); + } /* libphy will determine the link state */ netif_carrier_off(dev); From cbbc1aee7776c7fa1d89e6cb963a23e58c495dca Mon Sep 17 00:00:00 2001 From: Florian Fainelli Date: Mon, 21 Sep 2026 15:00:19 -0700 Subject: [PATCH 541/592] net: bcmgenet: do not skip WoL power up on GENET V1 bcmgenet_power_up() had an early check for bcmgenet_has_ext(priv) before dispatching by power mode. GENET V1 does not have the EXT block (unlike GENET V2+), which causes bcmgenet_power_up() to immediately return 0. As a consequence, when waking up from GENET_POWER_WOL_MAGIC on GENET V1, bcmgenet_wol_power_up_cfg() is never invoked to disable the WoL clock, clear wake event masks, and restore normal PHY and MAC operations. Move the bcmgenet_has_ext() checks to the GENET_POWER_PASSIVE and GENET_POWER_CABLE_SENSE cases where the EXT registers are actually accessed, allowing GENET_POWER_WOL_MAGIC cleanup to execute on all hardware versions. Fixes: c3ae64ae0c08 ("net: bcmgenet: handle GENET_POWER_WOL_MAGIC") Reviewed-by: Nicolai Buchwitz Signed-off-by: Florian Fainelli Link: https://patch.msgid.link/20260921220021.281418-4-florian.fainelli@broadcom.com Signed-off-by: Jakub Kicinski --- drivers/net/ethernet/broadcom/genet/bcmgenet.c | 13 ++++++++----- 1 file changed, 8 insertions(+), 5 deletions(-) diff --git a/drivers/net/ethernet/broadcom/genet/bcmgenet.c b/drivers/net/ethernet/broadcom/genet/bcmgenet.c index 2ab37bb031ed..07032e4193ea 100644 --- a/drivers/net/ethernet/broadcom/genet/bcmgenet.c +++ b/drivers/net/ethernet/broadcom/genet/bcmgenet.c @@ -1762,13 +1762,12 @@ static int bcmgenet_power_up(struct bcmgenet_priv *priv, int ret = 0; u32 reg; - if (!bcmgenet_has_ext(priv)) - return ret; - - reg = bcmgenet_ext_readl(priv, EXT_EXT_PWR_MGMT); - switch (mode) { case GENET_POWER_PASSIVE: + if (!bcmgenet_has_ext(priv)) + break; + + reg = bcmgenet_ext_readl(priv, EXT_EXT_PWR_MGMT); reg &= ~(EXT_PWR_DOWN_DLL | EXT_PWR_DOWN_BIAS | EXT_ENERGY_DET_MASK); if (GENET_IS_V5(priv) && !bcmgenet_has_ephy_16nm(priv)) { @@ -1792,8 +1791,12 @@ static int bcmgenet_power_up(struct bcmgenet_priv *priv, break; case GENET_POWER_CABLE_SENSE: + if (!bcmgenet_has_ext(priv)) + break; + /* enable APD */ if (!GENET_IS_V5(priv)) { + reg = bcmgenet_ext_readl(priv, EXT_EXT_PWR_MGMT); reg |= EXT_PWR_DN_EN_LD; bcmgenet_ext_writel(priv, reg, EXT_EXT_PWR_MGMT); } From 273941c85fc2632cd3e56ddff737b9245de7697d Mon Sep 17 00:00:00 2001 From: Florian Fainelli Date: Mon, 21 Sep 2026 15:00:20 -0700 Subject: [PATCH 542/592] net: bcmgenet: validate Ethernet address in bcmgenet_set_mac_addr bcmgenet_set_mac_addr() did not check whether the provided MAC address is a valid Ethernet address before applying it. Userspace could configure an invalid address (such as all zeroes or a multicast address) while the interface is down. Add a call to is_valid_ether_addr() and return -EADDRNOTAVAIL if the MAC address is not valid. Fixes: 1c1008c793fa ("net: bcmgenet: add main driver file") Reviewed-by: Nicolai Buchwitz Signed-off-by: Florian Fainelli Link: https://patch.msgid.link/20260921220021.281418-5-florian.fainelli@broadcom.com Signed-off-by: Jakub Kicinski --- drivers/net/ethernet/broadcom/genet/bcmgenet.c | 3 +++ 1 file changed, 3 insertions(+) diff --git a/drivers/net/ethernet/broadcom/genet/bcmgenet.c b/drivers/net/ethernet/broadcom/genet/bcmgenet.c index 07032e4193ea..011376a1678a 100644 --- a/drivers/net/ethernet/broadcom/genet/bcmgenet.c +++ b/drivers/net/ethernet/broadcom/genet/bcmgenet.c @@ -3635,6 +3635,9 @@ static int bcmgenet_set_mac_addr(struct net_device *dev, void *p) if (netif_running(dev)) return -EBUSY; + if (!is_valid_ether_addr(addr->sa_data)) + return -EADDRNOTAVAIL; + eth_hw_addr_set(dev, addr->sa_data); return 0; From d64e277b955be4506931802837499b62c8f3968a Mon Sep 17 00:00:00 2001 From: Florian Fainelli Date: Mon, 21 Sep 2026 15:00:21 -0700 Subject: [PATCH 543/592] net: bcmgenet: mask DMA_TIMEOUT_MASK when reading DMA_RING0_TIMEOUT bcmgenet_get_coalesce() reads DMA_RING0_TIMEOUT to calculate rx_coalesce_usecs without masking out bits outside DMA_TIMEOUT_MASK (16 bits). If upper bits are non-zero or contain status/flags, the computed value of rx_coalesce_usecs returned to userspace via ethtool becomes corrupted. Mask the register read with DMA_TIMEOUT_MASK before computing the timeout in microseconds. Fixes: 4a29645bfe6c ("net: bcmgenet: Implement RX coalescing control knobs") Reviewed-by: Nicolai Buchwitz Signed-off-by: Florian Fainelli Link: https://patch.msgid.link/20260921220021.281418-6-florian.fainelli@broadcom.com Signed-off-by: Jakub Kicinski --- drivers/net/ethernet/broadcom/genet/bcmgenet.c | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-) diff --git a/drivers/net/ethernet/broadcom/genet/bcmgenet.c b/drivers/net/ethernet/broadcom/genet/bcmgenet.c index 011376a1678a..21668e41b696 100644 --- a/drivers/net/ethernet/broadcom/genet/bcmgenet.c +++ b/drivers/net/ethernet/broadcom/genet/bcmgenet.c @@ -852,7 +852,8 @@ static int bcmgenet_get_coalesce(struct net_device *dev, ec->rx_max_coalesced_frames = bcmgenet_rdma_ring_readl(priv, 0, DMA_MBUF_DONE_THRESH); ec->rx_coalesce_usecs = - bcmgenet_rdma_readl(priv, DMA_RING0_TIMEOUT) * 8192 / 1000; + (bcmgenet_rdma_readl(priv, DMA_RING0_TIMEOUT) & + DMA_TIMEOUT_MASK) * 8192 / 1000; for (i = 0; i <= priv->hw_params->rx_queues; i++) { ring = &priv->rx_rings[i]; From 4bdee8060d1e4581624e68fbd369b1afb14df4bc Mon Sep 17 00:00:00 2001 From: Myeonghun Pak Date: Mon, 21 Sep 2026 20:09:14 -0400 Subject: [PATCH 544/592] net: airoha: npu: cancel wdt_work after releasing the WDT IRQ airoha_npu_remove() calls cancel_work_sync() on each core's wdt_work, but the watchdog IRQ that queues it is requested with devm_request_irq() and is freed only after .remove() returns. airoha_npu_wdt_handler() can therefore schedule_work() again once the cancel has returned. struct airoha_npu, which contains the work, is devm_kzalloc()'d and is freed in that same unwind, so the late work dereferences freed memory. Register the work with devm_work_autocancel() before devm_request_irq() and drop .remove(). Devres runs in reverse order, so the IRQ is freed before cancel_work_sync(), including when probe fails. A cancel left in .remove() cannot get that order. Initializing the work first also stops a pending watchdog interrupt from queuing an uninitialized work item. Probe currently calls INIT_WORK() only after devm_request_irq(). This issue was identified during our ongoing static-analysis research while reviewing kernel code. Fixes: 23290c7bc190 ("net: airoha: Introduce Airoha NPU support") Cc: stable@vger.kernel.org # 6.15+ Co-developed-by: Ijae Kim Signed-off-by: Ijae Kim Signed-off-by: Myeonghun Pak Acked-by: Lorenzo Bianconi Link: https://patch.msgid.link/20260922000914.542068-1-mhun512@gmail.com Signed-off-by: Jakub Kicinski --- drivers/net/ethernet/airoha/airoha_npu.c | 18 ++++++------------ 1 file changed, 6 insertions(+), 12 deletions(-) diff --git a/drivers/net/ethernet/airoha/airoha_npu.c b/drivers/net/ethernet/airoha/airoha_npu.c index 5bb4817a898d..4d3195eb00f7 100644 --- a/drivers/net/ethernet/airoha/airoha_npu.c +++ b/drivers/net/ethernet/airoha/airoha_npu.c @@ -5,6 +5,7 @@ */ #include +#include #include #include #include @@ -751,12 +752,15 @@ static int airoha_npu_probe(struct platform_device *pdev) if (irq < 0) return irq; + err = devm_work_autocancel(dev, &core->wdt_work, + airoha_npu_wdt_work); + if (err) + return err; + err = devm_request_irq(dev, irq, airoha_npu_wdt_handler, IRQF_SHARED, "airoha-npu-wdt", core); if (err) return err; - - INIT_WORK(&core->wdt_work, airoha_npu_wdt_work); } /* wlan IRQ lines */ @@ -803,18 +807,8 @@ static int airoha_npu_probe(struct platform_device *pdev) return 0; } -static void airoha_npu_remove(struct platform_device *pdev) -{ - struct airoha_npu *npu = platform_get_drvdata(pdev); - int i; - - for (i = 0; i < ARRAY_SIZE(npu->cores); i++) - cancel_work_sync(&npu->cores[i].wdt_work); -} - static struct platform_driver airoha_npu_driver = { .probe = airoha_npu_probe, - .remove = airoha_npu_remove, .driver = { .name = "airoha-npu", .of_match_table = of_airoha_npu_match, From a3f315be9d30eeb6938d11fa17fd4b32d52f7c42 Mon Sep 17 00:00:00 2001 From: Xuanqiang Luo Date: Mon, 21 Sep 2026 11:18:59 +0800 Subject: [PATCH 545/592] ip_gre: Reject enabling collect metadata through changelink ipgre_netlink_parms() can enable collect_md on an existing GRE, GRETAP or ERSPAN device. Unlike newlink, changelink does not enforce metadata tunnel uniqueness. Converting a non-metadata device can therefore replace the metadata receive entry for another device of the same type in the same netns. Deleting either device then clears the shared entry, breaking metadata receive lookup for the surviving device. If parameter validation fails after collect_md is set, deleting the modified device can also clear an entry it never owned. Reject enabling metadata mode in both changelink callbacks before any encapsulation or tunnel parameters are modified. Allow requests that repeat the metadata attribute on an existing metadata device. Fixes: 2e15ea390e6f ("ip_gre: Add support to collect tunnel metadata.") Signed-off-by: Xuanqiang Luo Reviewed-by: Ido Schimmel Reviewed-by: Hangbin Liu Link: https://patch.msgid.link/20260921031859.9283-1-xuanqiang.luo@linux.dev Signed-off-by: Jakub Kicinski --- net/ipv4/ip_gre.c | 12 ++++++++++++ 1 file changed, 12 insertions(+) diff --git a/net/ipv4/ip_gre.c b/net/ipv4/ip_gre.c index 82309efd417e..e4878e9aa636 100644 --- a/net/ipv4/ip_gre.c +++ b/net/ipv4/ip_gre.c @@ -1464,6 +1464,12 @@ static int ipgre_changelink(struct net_device *dev, struct nlattr *tb[], if (!rtnl_dev_link_net_capable(dev, t->net)) return -EPERM; + if (data && data[IFLA_GRE_COLLECT_METADATA] && !t->collect_md) { + NL_SET_ERR_MSG(extack, + "Enabling collect_md on an existing device is not supported"); + return -EOPNOTSUPP; + } + err = ipgre_newlink_encap_setup(dev, data); if (err) return err; @@ -1496,6 +1502,12 @@ static int erspan_changelink(struct net_device *dev, struct nlattr *tb[], if (!rtnl_dev_link_net_capable(dev, t->net)) return -EPERM; + if (data && data[IFLA_GRE_COLLECT_METADATA] && !t->collect_md) { + NL_SET_ERR_MSG(extack, + "Enabling collect_md on an existing device is not supported"); + return -EOPNOTSUPP; + } + err = ipgre_newlink_encap_setup(dev, data); if (err) return err; From 0f2fd31f63c65c409bd336af3a42d478a9207c1f Mon Sep 17 00:00:00 2001 From: Gilberto Conde Date: Mon, 21 Sep 2026 10:34:26 +0100 Subject: [PATCH 546/592] net: usb: qmi_wwan: add Quectel EG120K-EA Add support for the Quectel EG120K-EA LTE Cat.12 module (USB ID 2c7c:030b). Its QMI interface (interface 4) uses class/subclass/protocol ff/ff/ff like the other recent Quectel modules, so match it the same way. The product ID is shared with the EM060K, which the option driver already knows and uses to claim the serial interfaces. Without a qmi_wwan entry the data interface is left unbound. Tested on a GL.iNet GL-X2000, where the module is soldered down and enumerates at SuperSpeed: cdc-wdm0 and wwan0 appear and ModemManager brings up a data session. T: Bus=02 Lev=01 Prnt=01 Port=00 Cnt=01 Dev#= 2 Spd=5000 MxCh= 0 D: Ver= 3.10 Cls=00(>ifc ) Sub=00 Prot=00 MxPS= 9 #Cfgs= 1 P: Vendor=2c7c ProdID=030b Rev= 5.04 S: Manufacturer=Quectel S: Product=EG120K-EA S: SerialNumber=45546267 C:* #Ifs= 6 Cfg#= 1 Atr=a0 MxPwr=896mA I:* If#= 0 Alt= 0 #EPs= 2 Cls=ff(vend.) Sub=ff Prot=30 Driver=option E: Ad=01(O) Atr=02(Bulk) MxPS=1024 Ivl=0ms E: Ad=81(I) Atr=02(Bulk) MxPS=1024 Ivl=0ms I:* If#= 1 Alt= 0 #EPs= 3 Cls=ff(vend.) Sub=00 Prot=40 Driver=option E: Ad=83(I) Atr=03(Int.) MxPS= 10 Ivl=32ms E: Ad=82(I) Atr=02(Bulk) MxPS=1024 Ivl=0ms E: Ad=02(O) Atr=02(Bulk) MxPS=1024 Ivl=0ms I:* If#= 2 Alt= 0 #EPs= 3 Cls=ff(vend.) Sub=ff Prot=40 Driver=option E: Ad=85(I) Atr=03(Int.) MxPS= 10 Ivl=32ms E: Ad=84(I) Atr=02(Bulk) MxPS=1024 Ivl=0ms E: Ad=03(O) Atr=02(Bulk) MxPS=1024 Ivl=0ms I:* If#= 3 Alt= 0 #EPs= 3 Cls=ff(vend.) Sub=ff Prot=40 Driver=option E: Ad=87(I) Atr=03(Int.) MxPS= 10 Ivl=32ms E: Ad=86(I) Atr=02(Bulk) MxPS=1024 Ivl=0ms E: Ad=04(O) Atr=02(Bulk) MxPS=1024 Ivl=0ms I:* If#= 4 Alt= 0 #EPs= 3 Cls=ff(vend.) Sub=ff Prot=ff Driver=qmi_wwan E: Ad=88(I) Atr=03(Int.) MxPS= 8 Ivl=32ms E: Ad=8e(I) Atr=02(Bulk) MxPS=1024 Ivl=0ms E: Ad=0f(O) Atr=02(Bulk) MxPS=1024 Ivl=0ms I:* If#=12 Alt= 0 #EPs= 1 Cls=ff(vend.) Sub=ff Prot=70 Driver=(none) E: Ad=89(I) Atr=02(Bulk) MxPS=1024 Ivl=0ms Signed-off-by: Gilberto Conde Link: https://patch.msgid.link/20260921093426.2870266-1-gilbertorconde@gmail.com Signed-off-by: Jakub Kicinski --- drivers/net/usb/qmi_wwan.c | 1 + 1 file changed, 1 insertion(+) diff --git a/drivers/net/usb/qmi_wwan.c b/drivers/net/usb/qmi_wwan.c index f51cf9cb9421..0e2ab567dd40 100644 --- a/drivers/net/usb/qmi_wwan.c +++ b/drivers/net/usb/qmi_wwan.c @@ -1086,6 +1086,7 @@ static const struct usb_device_id products[] = { {QMI_MATCH_FF_FF_FF(0x2c7c, 0x0125)}, /* Quectel EC25, EC20 R2.0 Mini PCIe */ {QMI_MATCH_FF_FF_FF(0x2c7c, 0x013d)}, /* Quectel RG660QB */ {QMI_MATCH_FF_FF_FF(0x2c7c, 0x0306)}, /* Quectel EP06/EG06/EM06 */ + {QMI_MATCH_FF_FF_FF(0x2c7c, 0x030b)}, /* Quectel EM060K/EG120K-EA */ {QMI_MATCH_FF_FF_FF(0x2c7c, 0x0512)}, /* Quectel EG12/EM12 */ {QMI_MATCH_FF_FF_FF(0x2c7c, 0x0620)}, /* Quectel EM160R-GL */ {QMI_MATCH_FF_FF_FF(0x2c7c, 0x0800)}, /* Quectel RM500Q-GL */ From 481506a756dcd828ef42391cb08f38d8d96d38fc Mon Sep 17 00:00:00 2001 From: Sanghyun Park Date: Fri, 18 Sep 2026 12:26:58 +0900 Subject: [PATCH 547/592] vxlan: use one headroom snapshot for neighbour replies vxlan_na_create() samples LL_RESERVED_SPACE() to size the reply skb and then samples it again to reserve headroom. A concurrent vxlan_changelink() can update needed_headroom between the two reads, creating a TOCTOU race. The second value can exceed the allocation and make the Ethernet header write out of bounds. The race is reproducible on the unpatched kernel. It occurred when vxlan_na_create() generated a neighbour reply while vxlan_changelink() changed the link headroom. KASAN caught a four-byte write two bytes beyond a 704-byte skbuff_small_head allocation. Snapshot the headroom once and use that value for both allocation and reservation. Fixes: 4b29dba9c085 ("vxlan: fix nonfunctional neigh_reduce()") Signed-off-by: Sanghyun Park Link: https://patch.msgid.link/20260918032842.502409-2-sanghyun.park.cnu@gmail.com Signed-off-by: Jakub Kicinski --- drivers/net/vxlan/vxlan_core.c | 6 ++++-- 1 file changed, 4 insertions(+), 2 deletions(-) diff --git a/drivers/net/vxlan/vxlan_core.c b/drivers/net/vxlan/vxlan_core.c index c1d54339fa2b..3390e341d1e5 100644 --- a/drivers/net/vxlan/vxlan_core.c +++ b/drivers/net/vxlan/vxlan_core.c @@ -1958,13 +1958,15 @@ static struct sk_buff *vxlan_na_create(struct sk_buff *request, struct ipv6hdr *pip6; u8 *daddr; int na_olen = 8; /* opt hdr + ETH_ALEN for target */ + int headroom; int ns_olen; int i, len; if (dev == NULL || !pskb_may_pull(request, request->len)) return NULL; - len = LL_RESERVED_SPACE(dev) + sizeof(struct ipv6hdr) + + headroom = LL_RESERVED_SPACE(dev); + len = headroom + sizeof(struct ipv6hdr) + sizeof(*na) + na_olen + dev->needed_tailroom; reply = alloc_skb(len, GFP_ATOMIC); if (reply == NULL) @@ -1972,7 +1974,7 @@ static struct sk_buff *vxlan_na_create(struct sk_buff *request, reply->protocol = htons(ETH_P_IPV6); reply->dev = dev; - skb_reserve(reply, LL_RESERVED_SPACE(request->dev)); + skb_reserve(reply, headroom); skb_push(reply, sizeof(struct ethhdr)); skb_reset_mac_header(reply); From 4da3b7b8b50f3e2fde54a4c18a82a8e3f6223910 Mon Sep 17 00:00:00 2001 From: Norbert Szetei Date: Mon, 21 Sep 2026 17:03:57 +0200 Subject: [PATCH 548/592] net: xps: reject an out of range traffic class Only the entries below dev->num_tc are valid in dev->tc_to_txq[], and dev->prio_tc_map[] may only name classes below it. netdev_set_num_tc() lowers dev->num_tc without touching either array. netdev_txq_to_tc() walks all TC_MAX_QUEUE slots and netdev_get_prio_tc_map() returns the entry as it stands, so a leftover entry is handed out as a traffic class >= dev->num_tc. Taking that class from netdev_txq_to_tc(), __netif_set_xps_queue() rejects only a negative one and indexes an XPS map sized for dev->num_tc classes: tci = j * num_tc + tc; RCU_INIT_POINTER(new_dev_maps->attr_map[tci], map); attr_map[] holds nr_ids * num_tc entries and j runs over the ids named in the mask, so a class that is not below num_tc pushes tci past the end of the map for the last ids and the store overruns it. Any caller that lowers num_tc leaves such entries behind, and mqprio_destroy() tears down with netdev_set_num_tc(dev, 0) rather than netdev_reset_tc(). After mqprio with 8 classes then 1, tc_to_txq[1..7] still describe txq 1..7. The splat is from an XPS write to txq 2 on a veth with 8 rx queues: attr_map[] has 8 * 1 entries, tci = j + 2, and j == 6 stores one past the end of the 88-byte map: BUG: KASAN: slab-out-of-bounds in __netif_set_xps_queue (net/core/dev.c:2954) Write of size 8 at addr ffff88813016bc58 by task xps_oob/634 __netif_set_xps_queue (net/core/dev.c:2954) xps_rxqs_store (net/core/net-sysfs.c:1880) netdev_queue_attr_store (net/core/net-sysfs.c:1390) Allocated by task 634: __kmalloc_noprof (mm/slub.c:5439) __netif_set_xps_queue (net/core/dev.c:2937) The buggy address is located 0 bytes to the right of allocated 88-byte region [ffff88813016bc00, ffff88813016bc58) Reject a class the map has no room for. Fixes: 184c449f91fe ("net: Add support for XPS with QoS via traffic classes") Signed-off-by: Norbert Szetei Link: https://patch.msgid.link/162DD16F-54C6-444A-9E09-0B8CB3D591F2@doyensec.com Signed-off-by: Jakub Kicinski --- net/core/dev.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/net/core/dev.c b/net/core/dev.c index c67900354fa6..0292a16e16c2 100644 --- a/net/core/dev.c +++ b/net/core/dev.c @@ -2901,7 +2901,7 @@ int __netif_set_xps_queue(struct net_device *dev, const unsigned long *mask, dev = netdev_get_tx_queue(dev, index)->sb_dev ? : dev; tc = netdev_txq_to_tc(dev, index); - if (tc < 0) + if (tc < 0 || tc >= num_tc) return -EINVAL; } From e47a1958e12abc3a17b5231a4f21c8f1bf662e08 Mon Sep 17 00:00:00 2001 From: Yuqi Xu Date: Sat, 19 Sep 2026 16:45:27 +0800 Subject: [PATCH 549/592] net: ipconfig: bound DHCP option construction ic_dhcp_init_options() appends the hostname (option 12), vendor-class (option 60) and client-ID (option 61) options into the fixed 312-byte bootp_pkt.exten[] buffer. Only the client-ID branch checked the remaining space; the hostname and vendor-class writes were unbounded. A 64-byte hostname together with the maximum 252-byte dhcpclass= identifier needs 18 + (2 + 64) + (2 + 252) = 338 of the 312 available bytes even before the terminating END marker, so the vendor-class memcpy runs past the end of exten[]. With CONFIG_FORTIFY_SOURCE this is reported as a field-spanning write and, when the kernel is booted with panic_on_warn=1, aborts boot with a panic. Route the optional options through a common helper that makes sure the option, its 2-byte header and the END marker all fit and drops an option that would not. Configurations with short options keep sending exactly the same bytes as before. Fixes: 130c0f47fdf9 ("ipconfig: send host-name in DHCP requests") Cc: stable@vger.kernel.org Reported-by: Vega Assisted-by: LLM Signed-off-by: Yuqi Xu Reviewed-by: Ren Wei Reviewed-by: Simon Horman Link: https://patch.msgid.link/7808dfbfa2162dfd0b19f59aff5742d6e0db2abb.1789798023.git.xuyuqiabc@gmail.com Signed-off-by: Paolo Abeni --- net/ipv4/ipconfig.c | 45 ++++++++++++++++++++++++++------------------- 1 file changed, 26 insertions(+), 19 deletions(-) diff --git a/net/ipv4/ipconfig.c b/net/ipv4/ipconfig.c index 155db067eaec..1b8585404a41 100644 --- a/net/ipv4/ipconfig.c +++ b/net/ipv4/ipconfig.c @@ -676,6 +676,24 @@ static const u8 ic_bootp_cookie[4] = { 99, 130, 83, 99 }; #ifdef IPCONFIG_DHCP +static bool __init +ic_dhcp_add_option(u8 **options, const u8 *end, u8 type, const void *value, + int len) +{ + u8 *e = *options; + + /* leave room for the option header and the END marker */ + if (len > U8_MAX || end - e < len + 3) + return false; + + *e++ = type; + *e++ = len; + memcpy(e, value, len); + *options = e + len; + + return true; +} + static void __init ic_dhcp_init_options(u8 *options, struct ic_device *d) { @@ -691,6 +709,7 @@ ic_dhcp_init_options(u8 *options, struct ic_device *d) 42, /* NTP servers */ }; u8 mt = (ic_servaddr == NONE) ? DHCPDISCOVER : DHCPREQUEST; + u8 *end = options + sizeof(((struct bootp_pkt *)0)->exten); u8 *e = options; int len; @@ -721,31 +740,19 @@ ic_dhcp_init_options(u8 *options, struct ic_device *d) e += sizeof(ic_req_params); if (ic_host_name_set) { - *e++ = 12; /* host-name */ len = strlen(utsname()->nodename); - *e++ = len; - memcpy(e, utsname()->nodename, len); - e += len; + ic_dhcp_add_option(&e, end, 12, utsname()->nodename, len); } if (*vendor_class_identifier) { - pr_info("DHCP: sending class identifier \"%s\"\n", - vendor_class_identifier); - *e++ = 60; /* Class-identifier */ len = strlen(vendor_class_identifier); - *e++ = len; - memcpy(e, vendor_class_identifier, len); - e += len; + if (ic_dhcp_add_option(&e, end, 60, vendor_class_identifier, len)) + pr_info("DHCP: sending class identifier \"%s\"\n", + vendor_class_identifier); } len = strlen(dhcp_client_identifier + 1); - /* the minimum length of identifier is 2, include 1 byte type, - * and can not be larger than the length of options - */ - if (len >= 1 && len < 312 - (e - options) - 1) { - *e++ = 61; - *e++ = len + 1; - memcpy(e, dhcp_client_identifier, len + 1); - e += len + 1; - } + /* the minimum length of identifier is 2, include 1 byte type */ + if (len >= 1) + ic_dhcp_add_option(&e, end, 61, dhcp_client_identifier, len + 1); *e++ = 255; /* End of the list */ } From 00efbbd40bd5fd92c67b7cf1aab8904fa59a96f6 Mon Sep 17 00:00:00 2001 From: David Dai Date: Fri, 18 Sep 2026 16:11:55 -0500 Subject: [PATCH 550/592] bonding: crypto offload enabled, non-offload slave failover, rekey failed Create a bonding device (i.e. bond0) in active-backup mode, 2 slaves. Active slave: offload capable interface (i.e. eth1), primary interface. Backup slave: non-offload capable interface(i.e. eth2). Configure strongswan service swantl.conf child SA "hw_offload = crypto" Start strongswan service IPSec Crytpo Offload is enabled on top of bond0. i.e. ip xfrm state |grep offload crypto offload parameters: dev bond0 dir out mode crypto crypto offload parameters: dev bond0 dir in mode crypto Active slave eth1 takes adavantage of IPSec Crypto Offload capability. If active slave eth1 is down for any reason (i.e. eth1 link down): ip link set down dev eth1 non-offload capable interface eth2 failover to becomes active slave. The existing SAs can continue use software IPsec after failover. Traffic still keeps going properly. However if eth1 link had not recovered yet, strongswan service does new child SA rekey, or uses swanctl command to do new child SA rekey, it will fail because active slave eth2 doesn't support crypto offload. In bond_ipsec_add_sa routine, it returns -EINVAL now, which is treated as fatal error by xfrm_dev_state_add routine in kernel xfrm. To make the non-offload active slave survive the child SA rekey, need to make bond_ipsec_add_sa routine returns -EOPNOTSUPP instead when active slave doesn't support IPsec Crypto offload, the xfrm will gracefully fallback to create new SA using Software IPsec. Network traffic can keep going. After offload capable interface eth1 link is up, becomes active slave, next time strongswan child SA rekey will create a new SA which enables crypto offload again. Fixes: 18cb261afd7b ("bonding: support hardware encryption offload to slaves") Signed-off-by: David Dai Reviewed-by: Hangbin Liu Link: https://patch.msgid.link/20260918211155.1664493-1-zdai@linux.ibm.com Signed-off-by: Paolo Abeni --- drivers/net/bonding/bond_main.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/drivers/net/bonding/bond_main.c b/drivers/net/bonding/bond_main.c index a9bff7663eec..de2489c3d9bf 100644 --- a/drivers/net/bonding/bond_main.c +++ b/drivers/net/bonding/bond_main.c @@ -490,7 +490,7 @@ static int bond_ipsec_add_sa(struct net_device *bond_dev, !real_dev->xfrmdev_ops->xdo_dev_state_add || netif_is_bond_master(real_dev)) { NL_SET_ERR_MSG_MOD(extack, "Slave does not support ipsec offload"); - err = -EINVAL; + err = -EOPNOTSUPP; goto out; } From 36c2009d90f2210ef92e6f4f2850e8b57b09e754 Mon Sep 17 00:00:00 2001 From: =?UTF-8?q?Gajdos=20Tam=C3=A1s?= Date: Mon, 21 Sep 2026 11:13:32 +0200 Subject: [PATCH 551/592] net: atl1c: fix soft lockup on out-of-range tpd_cons read MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit The hardware can report an out-of-range tpd_cons (seen as 0xffff) while the PCIe link/MAC is resetting. An out-of-range value can never be reached and the loop below would spin forever. To avoid a soft lockup treat it as "nothing new to clean" instead. Reproduced on two machines, same NIC (Qualcomm Atheros AR8151 v2.0, 4-port), triggered by rebooting a Mikrotik CCR2004 PCIe card that the ports are directly linked to: - Ubuntu 26.04.1 LTS, kernel 7.0.0-31-generic. The link-flap precursor, before the lockup was captured with a full trace elsewhere: atl1c 0000:05:00.0 enp5s0f0: NETDEV WATCHDOG: CPU: 4: transmit queue 2 timed out 489984 ms atl1c 0000:05:00.0: MAC state machine can't be idle since disabled for 10ms second atl1c 0000:05:00.0: atl1c: enp5s0f0 NIC Link is Up<65535 Mbps Full Duplex> 65535 (0xffff) here is the same value tpd_cons reads back once the loop below gets stuck. - Proxmox VE, kernel 7.0.14-11-pve. Same NIC/trigger, this time caught by the soft lockup watchdog with a full stack trace: watchdog: BUG: soft lockup - CPU#12 stuck for 354s! [napi/eth%d-0:329] CPU: 12 UID: 0 PID: 329 Comm: napi/eth%d-0 Tainted: P O L 7.0.14-11-pve #1 PREEMPT(lazy) RIP: 0010:atl1c_clean_tx+0x142/0x2d0 [atl1c] Call Trace: __napi_poll+0x32/0x1e0 napi_threaded_poll_loop+0x286/0x2e0 napi_threaded_poll+0xfd/0x140 kthread+0xf7/0x130 ret_from_fork+0x2da/0x3a0 ret_from_fork_asm+0x1a/0x30 Fixes: 43250ddd75a35d ("atl1c: Atheros L1C Gigabit Ethernet driver") Cc: stable@vger.kernel.org Signed-off-by: Gajdos Tamás Link: https://patch.msgid.link/20260921091334.3571525-2-tamas@rimpianto.com Signed-off-by: Paolo Abeni --- drivers/net/ethernet/atheros/atl1c/atl1c_main.c | 3 +++ 1 file changed, 3 insertions(+) diff --git a/drivers/net/ethernet/atheros/atl1c/atl1c_main.c b/drivers/net/ethernet/atheros/atl1c/atl1c_main.c index 7efa3fc257b3..e58f1d2c26bd 100644 --- a/drivers/net/ethernet/atheros/atl1c/atl1c_main.c +++ b/drivers/net/ethernet/atheros/atl1c/atl1c_main.c @@ -1602,6 +1602,9 @@ static int atl1c_clean_tx(struct napi_struct *napi, int budget) AT_READ_REGW(&adapter->hw, atl1c_qregs[tpd_ring->num].tpd_cons, &hw_next_to_clean); + if (unlikely(hw_next_to_clean >= tpd_ring->count)) + hw_next_to_clean = next_to_clean; + while (next_to_clean != hw_next_to_clean) { buffer_info = &tpd_ring->buffer_info[next_to_clean]; if (buffer_info->skb) { From 374bf9e4b90f979e052332c4faca2d745c491a12 Mon Sep 17 00:00:00 2001 From: =?UTF-8?q?Gajdos=20Tam=C3=A1s?= Date: Mon, 21 Sep 2026 11:13:33 +0200 Subject: [PATCH 552/592] net: atl1e: fix soft lockup on out-of-range hw_next_to_clean read MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit Same issue as atl1c (see the first commit in this series, "net: atl1c: fix soft lockup on out-of-range tpd_cons read"): the hardware can report an out-of-range hw_next_to_clean (seen as 0xffff) while the PCIe link/MAC is resetting. An out-of-range value can never be reached and the loop below would spin forever. Treat it as "nothing new to clean" instead. Fixes: a6a5325239c202 ("atl1e: Atheros L1E Gigabit Ethernet driver") Cc: stable@vger.kernel.org Signed-off-by: Gajdos Tamás Link: https://patch.msgid.link/20260921091334.3571525-3-tamas@rimpianto.com Signed-off-by: Paolo Abeni --- drivers/net/ethernet/atheros/atl1e/atl1e_main.c | 3 +++ 1 file changed, 3 insertions(+) diff --git a/drivers/net/ethernet/atheros/atl1e/atl1e_main.c b/drivers/net/ethernet/atheros/atl1e/atl1e_main.c index 40290028580b..437989edb780 100644 --- a/drivers/net/ethernet/atheros/atl1e/atl1e_main.c +++ b/drivers/net/ethernet/atheros/atl1e/atl1e_main.c @@ -1234,6 +1234,9 @@ static bool atl1e_clean_tx_irq(struct atl1e_adapter *adapter) u16 hw_next_to_clean = AT_READ_REGW(&adapter->hw, REG_TPD_CONS_IDX); u16 next_to_clean = atomic_read(&tx_ring->next_to_clean); + if (unlikely(hw_next_to_clean >= tx_ring->count)) + hw_next_to_clean = next_to_clean; + while (next_to_clean != hw_next_to_clean) { tx_buffer = &tx_ring->tx_buffer[next_to_clean]; if (tx_buffer->dma) { From 43e746821f5f5afbbf68e388bf9fbe221e03bfca Mon Sep 17 00:00:00 2001 From: =?UTF-8?q?Gajdos=20Tam=C3=A1s?= Date: Mon, 21 Sep 2026 11:13:34 +0200 Subject: [PATCH 553/592] net: atl1: fix soft lockup on out-of-range cmb_tpd_next_to_clean read MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit Same issue as atl1c (see the first commit in this series, "net: atl1c: fix soft lockup on out-of-range tpd_cons read"): the hardware can report an out-of-range cmb_tpd_next_to_clean (seen as 0xffff) while the PCIe link/MAC is resetting. An out-of-range value can never be reached and the loop below would spin forever. Treat it as "nothing new to clean" instead. Fixes: f3cc28c797604f ("Add Attansic L1 ethernet driver.") Cc: stable@vger.kernel.org Signed-off-by: Gajdos Tamás Link: https://patch.msgid.link/20260921091334.3571525-4-tamas@rimpianto.com Signed-off-by: Paolo Abeni --- drivers/net/ethernet/atheros/atlx/atl1.c | 3 +++ 1 file changed, 3 insertions(+) diff --git a/drivers/net/ethernet/atheros/atlx/atl1.c b/drivers/net/ethernet/atheros/atlx/atl1.c index 98a4d089270e..957d5598dda5 100644 --- a/drivers/net/ethernet/atheros/atlx/atl1.c +++ b/drivers/net/ethernet/atheros/atlx/atl1.c @@ -2066,6 +2066,9 @@ static int atl1_intr_tx(struct atl1_adapter *adapter) sw_tpd_next_to_clean = atomic_read(&tpd_ring->next_to_clean); cmb_tpd_next_to_clean = le16_to_cpu(adapter->cmb.cmb->tpd_cons_idx); + if (unlikely(cmb_tpd_next_to_clean >= tpd_ring->count)) + cmb_tpd_next_to_clean = sw_tpd_next_to_clean; + while (cmb_tpd_next_to_clean != sw_tpd_next_to_clean) { buffer_info = &tpd_ring->buffer_info[sw_tpd_next_to_clean]; if (buffer_info->dma) { From 77b1718e39e5c9f6956fb60807326af20baf889d Mon Sep 17 00:00:00 2001 From: Myeonghun Pak Date: Mon, 21 Sep 2026 21:46:05 -0400 Subject: [PATCH 554/592] bna: prevent IOC timer rearm during teardown bna: prevent IOC timer rearm during teardown bnad_pci_remove() and the probe disable_ioceth path call timer_delete_sync() for ioc_timer, sem_timer and hb_timer, but not for iocpf_timer. bnad_iocpf_timeout() then takes bnad->bna_lock after free_netdev() has freed the struct bnad. Deleting iocpf_timer last does not fix this. sem_timer and iocpf_timer rearm each other: bnad_iocpf_sem_timeout() can arm iocpf_timer, and bnad_iocpf_timeout() arms sem_timer from bfa_ioc_hw_sem_get() when the semaphore is busy. timer_delete_sync() only waits out its own callback. bnad_ioceth_disable() can time out and leave that callback live. Shut all four IOC timers down with timer_shutdown_sync() on both paths, so a later mod_timer() is ignored. This issue was identified during our ongoing static-analysis research while reviewing kernel code. Fixes: 1d32f7696286 ("bna: IOC failure auto recovery fix") Cc: stable@vger.kernel.org Assisted-by: LLM Co-developed-by: Ijae Kim Signed-off-by: Ijae Kim Signed-off-by: Myeonghun Pak Link: https://patch.msgid.link/20260922014605.588040-1-mhun512@gmail.com Signed-off-by: Paolo Abeni --- drivers/net/ethernet/brocade/bna/bnad.c | 24 ++++++++++++++++++------ 1 file changed, 18 insertions(+), 6 deletions(-) diff --git a/drivers/net/ethernet/brocade/bna/bnad.c b/drivers/net/ethernet/brocade/bna/bnad.c index 8b75004ba7c9..55dfd4896785 100644 --- a/drivers/net/ethernet/brocade/bna/bnad.c +++ b/drivers/net/ethernet/brocade/bna/bnad.c @@ -2571,6 +2571,22 @@ bnad_ioceth_disable(struct bnad *bnad) return err; } +/* + * The IOC timers rearm one another, so deleting one cannot stop a + * sibling callback from arming it again. Shut them down so a later + * mod_timer() is ignored. + */ +static void +bnad_ioc_timers_shutdown(struct bnad *bnad) +{ + struct bfa_ioc *ioc = &bnad->bna.ioceth.ioc; + + timer_shutdown_sync(&ioc->ioc_timer); + timer_shutdown_sync(&ioc->sem_timer); + timer_shutdown_sync(&ioc->hb_timer); + timer_shutdown_sync(&ioc->iocpf_timer); +} + static int bnad_ioceth_enable(struct bnad *bnad) { @@ -3727,9 +3743,7 @@ bnad_pci_probe(struct pci_dev *pdev, bnad_res_free(bnad, &bnad->mod_res_info[0], BNA_MOD_RES_T_MAX); disable_ioceth: bnad_ioceth_disable(bnad); - timer_delete_sync(&bnad->bna.ioceth.ioc.ioc_timer); - timer_delete_sync(&bnad->bna.ioceth.ioc.sem_timer); - timer_delete_sync(&bnad->bna.ioceth.ioc.hb_timer); + bnad_ioc_timers_shutdown(bnad); spin_lock_irqsave(&bnad->bna_lock, flags); bna_uninit(bna); spin_unlock_irqrestore(&bnad->bna_lock, flags); @@ -3770,9 +3784,7 @@ bnad_pci_remove(struct pci_dev *pdev) mutex_lock(&bnad->conf_mutex); bnad_ioceth_disable(bnad); - timer_delete_sync(&bnad->bna.ioceth.ioc.ioc_timer); - timer_delete_sync(&bnad->bna.ioceth.ioc.sem_timer); - timer_delete_sync(&bnad->bna.ioceth.ioc.hb_timer); + bnad_ioc_timers_shutdown(bnad); spin_lock_irqsave(&bnad->bna_lock, flags); bna_uninit(bna); spin_unlock_irqrestore(&bnad->bna_lock, flags); From 58eb1b3325edac42dc6df72c80962bd53a3c8ca7 Mon Sep 17 00:00:00 2001 From: Dongliang Qin Date: Tue, 22 Sep 2026 11:15:45 +0800 Subject: [PATCH 555/592] rds: ib: Clear the sg list when mapping an MR fails rds_ib_map_frmr() stores the caller's scatterlist in the MR before DMA mapping and registration can fail. On failure, __rds_rdma_map() unpins the pages and frees the scatterlist, but rds_ib_free_frmr() can still return the MR to the pool with the stale pointer set. This leaves the pool with a dangling scatterlist and can lead to local privilege escalation. KASAN detects the resulting use-after-free when the MR is later torn down: BUG: KASAN: slab-use-after-free in __rds_ib_teardown_mr Read of size 8 Call Trace: __rds_ib_teardown_mr rds_ib_unreg_frmr rds_ib_flush_mr_pool rds_ib_flush_mrs rds_free_mr rds_setsockopt Store the scatterlist in the MR only after DMA mapping succeeds. If DMA mapping fails, return directly while the MR fields remain clear; the caller keeps ownership of the scatterlist and its pinned pages. If a later registration step fails, unmap the scatterlist and clear the MR fields before returning. Fixes: 1659185fb4d0 ("RDS: IB: Support Fastreg MR (FRMR) memory registration mode") Cc: stable@vger.kernel.org Signed-off-by: Dongliang Qin Reviewed-by: Allison Henderson Link: https://patch.msgid.link/20260922031546.3874605-1-cccccccccccc777777@gmail.com Signed-off-by: Paolo Abeni --- net/rds/ib_frmr.c | 11 +++++------ 1 file changed, 5 insertions(+), 6 deletions(-) diff --git a/net/rds/ib_frmr.c b/net/rds/ib_frmr.c index bd861191157b..8397aa4a17ac 100644 --- a/net/rds/ib_frmr.c +++ b/net/rds/ib_frmr.c @@ -204,19 +204,16 @@ static int rds_ib_map_frmr(struct rds_ib_device *rds_ibdev, */ rds_ib_teardown_mr(ibmr); - ibmr->sg = sg; - ibmr->sg_len = sg_len; - ibmr->sg_dma_len = 0; frmr->sg_byte_len = 0; - WARN_ON(ibmr->sg_dma_len); - ibmr->sg_dma_len = ib_dma_map_sg(dev, ibmr->sg, ibmr->sg_len, + ibmr->sg_dma_len = ib_dma_map_sg(dev, sg, sg_len, DMA_BIDIRECTIONAL); if (unlikely(!ibmr->sg_dma_len)) { pr_warn("RDS/IB: %s failed!\n", __func__); return -EBUSY; } - frmr->sg_byte_len = 0; + ibmr->sg = sg; + ibmr->sg_len = sg_len; frmr->dma_npages = 0; len = 0; @@ -264,6 +261,8 @@ static int rds_ib_map_frmr(struct rds_ib_device *rds_ibdev, ib_dma_unmap_sg(rds_ibdev->dev, ibmr->sg, ibmr->sg_len, DMA_BIDIRECTIONAL); ibmr->sg_dma_len = 0; + ibmr->sg = NULL; + ibmr->sg_len = 0; return ret; } From c2de369c5c5b8599ca10fd5ca8d11fcd845c1331 Mon Sep 17 00:00:00 2001 From: Haseeb Malik Date: Mon, 21 Sep 2026 16:40:30 -0400 Subject: [PATCH 556/592] macsec: initialize SecY before registering the netdevice Creating a MACsec device with MAC offload over an LRO-capable lower device triggers a warning in rtmsg_ifinfo_build_skb() when IPv4 forwarding is enabled by default. register_netdevice() invokes inetdev_init(), which disables LRO and emits a NETDEV_FEAT_CHANGE notification. This reaches macsec_fill_info() before macsec_add_dev() initializes the SecY. key_len is still zero, so macsec_fill_info() returns -EMSGSIZE and trips the WARN_ON in rtmsg_ifinfo_build_skb(), even though the skb has enough space. Even without the warning, notifications during registration can report uninitialized SecY attributes, including the SCI. This ordering has existed since the driver was introduced. Initialize the SecY and apply the new-link attributes before registration. Move MAC address inheritance into macsec_newlink() so the SCI can also be initialized before registration-time notifications report it. Move the per-CPU statistics and metadata destination allocation into ndo_init(), and release partial allocations on failure. Fixes: c09440f7dcb3 ("macsec: introduce IEEE 802.1AE driver") Reported-by: syzbot+f2f6312ad1b5a0bfe316@syzkaller.appspotmail.com Closes: https://syzkaller.appspot.com/bug?extid=f2f6312ad1b5a0bfe316 Suggested-by: Sabrina Dubroca Link: https://lists.openwall.net/linux-kernel/2026/08/19/552 Signed-off-by: Haseeb Malik Reviewed-by: Sabrina Dubroca Link: https://patch.msgid.link/20260921-fix-macsec-net-v3-1-accf94f93f5e@gmail.com Signed-off-by: Paolo Abeni --- drivers/net/macsec.c | 84 +++++++++++++++++++++++--------------------- 1 file changed, 43 insertions(+), 41 deletions(-) diff --git a/drivers/net/macsec.c b/drivers/net/macsec.c index 6f9f3aceffaa..78a19b134632 100644 --- a/drivers/net/macsec.c +++ b/drivers/net/macsec.c @@ -3539,6 +3539,22 @@ static int macsec_dev_init(struct net_device *dev) if (err) return err; + err = -ENOMEM; + macsec->stats = netdev_alloc_pcpu_stats(struct pcpu_secy_stats); + if (!macsec->stats) + goto destroy_gro_cells; + + macsec->secy.tx_sc.stats = + netdev_alloc_pcpu_stats(struct pcpu_tx_sc_stats); + if (!macsec->secy.tx_sc.stats) + goto free_secy_stats; + + macsec->secy.tx_sc.md_dst = metadata_dst_alloc(0, METADATA_MACSEC, + GFP_KERNEL); + if (!macsec->secy.tx_sc.md_dst) + goto free_tx_sc_stats; + macsec->secy.tx_sc.md_dst->u.macsec_info.sci = macsec->secy.sci; + macsec_inherit_tso_max(dev); dev->hw_features = real_dev->hw_features & MACSEC_OFFLOAD_FEATURES; @@ -3551,8 +3567,6 @@ static int macsec_dev_init(struct net_device *dev) macsec_set_head_tail_room(dev); - if (is_zero_ether_addr(dev->dev_addr)) - eth_hw_addr_inherit(dev, real_dev); if (is_zero_ether_addr(dev->broadcast)) memcpy(dev->broadcast, real_dev->broadcast, dev->addr_len); @@ -3560,6 +3574,14 @@ static int macsec_dev_init(struct net_device *dev) netdev_hold(real_dev, &macsec->dev_tracker, GFP_KERNEL); return 0; + +free_tx_sc_stats: + free_percpu(macsec->secy.tx_sc.stats); +free_secy_stats: + free_percpu(macsec->stats); +destroy_gro_cells: + gro_cells_destroy(&macsec->gro_cells); + return err; } static void macsec_dev_uninit(struct net_device *dev) @@ -4116,26 +4138,11 @@ static sci_t dev_to_sci(struct net_device *dev, __be16 port) return make_sci(dev->dev_addr, port); } -static int macsec_add_dev(struct net_device *dev, sci_t sci, u8 icv_len) +static void macsec_init_secy(struct net_device *dev, sci_t sci, u8 icv_len) { struct macsec_dev *macsec = macsec_priv(dev); struct macsec_secy *secy = &macsec->secy; - macsec->stats = netdev_alloc_pcpu_stats(struct pcpu_secy_stats); - if (!macsec->stats) - return -ENOMEM; - - secy->tx_sc.stats = netdev_alloc_pcpu_stats(struct pcpu_tx_sc_stats); - if (!secy->tx_sc.stats) - return -ENOMEM; - - secy->tx_sc.md_dst = metadata_dst_alloc(0, METADATA_MACSEC, GFP_KERNEL); - if (!secy->tx_sc.md_dst) - /* macsec and secy percpu stats will be freed when unregistering - * net_device in macsec_free_netdev() - */ - return -ENOMEM; - if (sci == MACSEC_UNDEF_SCI) sci = dev_to_sci(dev, MACSEC_PORT_ES); @@ -4149,15 +4156,12 @@ static int macsec_add_dev(struct net_device *dev, sci_t sci, u8 icv_len) secy->xpn = DEFAULT_XPN; secy->sci = sci; - secy->tx_sc.md_dst->u.macsec_info.sci = sci; secy->tx_sc.active = true; secy->tx_sc.encoding_sa = DEFAULT_ENCODING_SA; secy->tx_sc.encrypt = DEFAULT_ENCRYPT; secy->tx_sc.send_sci = DEFAULT_SEND_SCI; secy->tx_sc.end_station = false; secy->tx_sc.scb = false; - - return 0; } static struct lock_class_key macsec_netdev_addr_lock_key; @@ -4220,6 +4224,24 @@ static int macsec_newlink(struct net_device *dev, if (rx_handler && rx_handler != macsec_handle_frame) return -EBUSY; + if (is_zero_ether_addr(dev->dev_addr)) + eth_hw_addr_inherit(dev, real_dev); + + if (data && data[IFLA_MACSEC_SCI]) + sci = nla_get_sci(data[IFLA_MACSEC_SCI]); + else if (data && data[IFLA_MACSEC_PORT]) + sci = dev_to_sci(dev, nla_get_be16(data[IFLA_MACSEC_PORT])); + else + sci = dev_to_sci(dev, MACSEC_PORT_ES); + + /* Registration can notify listeners before returning. */ + macsec_init_secy(dev, sci, icv_len); + if (data) { + err = macsec_changelink_common(dev, data); + if (err) + return err; + } + err = register_netdevice(dev); if (err < 0) return err; @@ -4232,31 +4254,11 @@ static int macsec_newlink(struct net_device *dev, if (err < 0) goto unregister; - /* need to be already registered so that ->init has run and - * the MAC addr is set - */ - if (data && data[IFLA_MACSEC_SCI]) - sci = nla_get_sci(data[IFLA_MACSEC_SCI]); - else if (data && data[IFLA_MACSEC_PORT]) - sci = dev_to_sci(dev, nla_get_be16(data[IFLA_MACSEC_PORT])); - else - sci = dev_to_sci(dev, MACSEC_PORT_ES); - if (rx_handler && sci_exists(real_dev, sci)) { err = -EBUSY; goto unlink; } - err = macsec_add_dev(dev, sci, icv_len); - if (err) - goto unlink; - - if (data) { - err = macsec_changelink_common(dev, data); - if (err) - goto del_dev; - } - /* If h/w offloading is available, propagate to the device */ if (macsec_is_offloaded(macsec)) { const struct macsec_ops *ops; From d6ec384c87cc851cfd13bb18c99ce351ccee6192 Mon Sep 17 00:00:00 2001 From: Fang Xieyan Date: Mon, 21 Sep 2026 20:54:41 +0800 Subject: [PATCH 557/592] net/sched: act_ife: validate metadata length before decoding skbmark_decode(), skbprio_decode() and skbtcindex_decode() read fixed-size values from the TLV payload without validating its length. A malformed IFE frame can declare a shorter payload, causing the decoders to consume bytes beyond the declared metadata value: [TLV type=IFE_META_SKBMARK len=4] -> dlen == 0, but decode reads 4 bytes The decoder may therefore set skb metadata from unintended input. Validate the payload length before decoding and return -EINVAL for invalid lengths. Read the values with get_unaligned_be32() and get_unaligned_be16(), as TLV payloads are not guaranteed to be aligned. Teach tcf_ife_decode() to log a decoder error separately from an unknown metaid; both are counted as overlimits and decoding continues with the remaining metadata. The metadata length issue was found by an automated audit of the IFE decode path at v6.18-rc7 and reproduced with a userspace sanitizer model of the decode path. Compile-tested on x86_64 with defconfig and NET_ACT_IFE=y: act_ife.o and the three act_meta_*.o build warning-free. Fixes: 084e2f6566d2 ("Support to encoding decoding skb mark on IFE action") Fixes: 200e10f46936 ("Support to encoding decoding skb prio on IFE action") Fixes: 408fbc22ef1e ("net sched ife action: Introduce skb tcindex metadata encap decap") Assisted-by: Hawkeye:GLM-5.3-flash Assisted-by: Qoder:Qwen3.8-Max Signed-off-by: Fang Xieyan Link: https://patch.msgid.link/20260921125441.81459-1-fangxy@xiaopeng.com Signed-off-by: Paolo Abeni --- net/sched/act_ife.c | 17 ++++++++++++----- net/sched/act_meta_mark.c | 6 ++++-- net/sched/act_meta_skbprio.c | 6 ++++-- net/sched/act_meta_skbtcindex.c | 6 ++++-- 4 files changed, 24 insertions(+), 11 deletions(-) diff --git a/net/sched/act_ife.c b/net/sched/act_ife.c index 9cea71fc1db3..2afd68983ece 100644 --- a/net/sched/act_ife.c +++ b/net/sched/act_ife.c @@ -737,6 +737,7 @@ static int tcf_ife_decode(struct sk_buff *skb, const struct tc_action *a, u8 *curr_data; u16 mtype; u16 dlen; + int ret; curr_data = ife_tlv_meta_decode(tlv_data, ifehdr_end, &mtype, &dlen, NULL); @@ -745,13 +746,19 @@ static int tcf_ife_decode(struct sk_buff *skb, const struct tc_action *a, return TC_ACT_SHOT; } - if (find_decode_metaid(skb, p, mtype, dlen, curr_data)) { - /* abuse overlimits to count when we receive metadata - * but dont have an ops for it + ret = find_decode_metaid(skb, p, mtype, dlen, curr_data); + if (ret < 0) { + /* abuse overlimits to count metadata we cannot + * decode: no ops for it, or the decoder rejected it */ - pr_info_ratelimited("Unknown metaid %d dlen %d\n", - mtype, dlen); qstats_cpu_overlimit_inc(ife->common.cpu_qstats); + + if (ret == -ENOENT) + pr_info_ratelimited("Unknown metaid %d dlen %d\n", + mtype, dlen); + else + pr_info_ratelimited("Failed to decode metaid %d dlen %d err %d\n", + mtype, dlen, ret); } } diff --git a/net/sched/act_meta_mark.c b/net/sched/act_meta_mark.c index ea0573cb8b2d..e2f61b22bf0f 100644 --- a/net/sched/act_meta_mark.c +++ b/net/sched/act_meta_mark.c @@ -10,6 +10,7 @@ #include #include #include +#include #include #include #include @@ -28,9 +29,10 @@ static int skbmark_encode(struct sk_buff *skb, void *skbdata, static int skbmark_decode(struct sk_buff *skb, void *data, u16 len) { - u32 ifemark = *(u32 *)data; + if (len != sizeof(u32)) + return -EINVAL; - skb->mark = ntohl(ifemark); + skb->mark = get_unaligned_be32(data); return 0; } diff --git a/net/sched/act_meta_skbprio.c b/net/sched/act_meta_skbprio.c index 2df3133ce5ad..5cdb57931eab 100644 --- a/net/sched/act_meta_skbprio.c +++ b/net/sched/act_meta_skbprio.c @@ -10,6 +10,7 @@ #include #include #include +#include #include #include #include @@ -33,9 +34,10 @@ static int skbprio_encode(struct sk_buff *skb, void *skbdata, static int skbprio_decode(struct sk_buff *skb, void *data, u16 len) { - u32 ifeprio = *(u32 *)data; + if (len != sizeof(u32)) + return -EINVAL; - skb->priority = ntohl(ifeprio); + skb->priority = get_unaligned_be32(data); return 0; } diff --git a/net/sched/act_meta_skbtcindex.c b/net/sched/act_meta_skbtcindex.c index 44547caead46..8803710c0905 100644 --- a/net/sched/act_meta_skbtcindex.c +++ b/net/sched/act_meta_skbtcindex.c @@ -10,6 +10,7 @@ #include #include #include +#include #include #include #include @@ -28,9 +29,10 @@ static int skbtcindex_encode(struct sk_buff *skb, void *skbdata, static int skbtcindex_decode(struct sk_buff *skb, void *data, u16 len) { - u16 ifetc_index = *(u16 *)data; + if (len != sizeof(u16)) + return -EINVAL; - skb->tc_index = ntohs(ifetc_index); + skb->tc_index = get_unaligned_be16(data); return 0; } From 7c9f391ec89cb621d7af375ace2eb9a6248e5b9d Mon Sep 17 00:00:00 2001 From: Christian Lamparter Date: Mon, 21 Sep 2026 18:28:17 +0200 Subject: [PATCH 558/592] net: emac: move setting of netops to fix crash fixes the following crash on driver initialization: |BUG: Kernel NULL pointer dereference on read at 0x00000158 |Faulting instruction address: 0xc0566b40 |Oops: Kernel access of bad area, sig: 11 [#1] |BE PAGE_SIZE=4K PowerPC 44x Platform |Modules linked in: |CPU: 0 UID: 0 PID: 1 Comm: swapper/0 Tainted: GW 7.3.0-rc3+ #1 |Tainted: [W]=WARN |Hardware name: MyBook Live APM821XX 0x12c41c83 PowerPC 44x Platform |NIP: c0566b40 LR: c05648f8 CTR: c04c1e9c |REGS: c1053a20 TRAP: 0300 Tainted: GW (7.3.0-rc3+) |MSR: 0002b000 CR: 24008808 XER: 00000000 |DEAR: 00000158 ESR: 00000000 |GPR00: c05648f8 c1053b10 c1063600 c1030000 c5ab3000 00000000 [...] |GPR08: 00000002 00000000 00000000 c1053b40 84002808 00000000 [...] |GPR16: cfffd210 00000002 c0beafcc cfffc960 00000000 c1030644 [...] |GPR24: c0beafbc c1037000 00000000 0000000a 00000000 c1030000 [...] |NIP [c0566b40] phy_link_topo_add_phy+0x2c/0x1d0 |LR [c05648f8] phy_attach_direct+0x1a4/0x368 |Call Trace: |[c1053b10] [c0811e04] klist_put+0x54/0xb4 (unreliable) |[c1053b40] [c05648f8] phy_attach_direct+0x1a4/0x368 |[c1053b70] [c0564ae8] phy_connect_direct+0x2c/0x60 |[c1053b90] [c056c7a8] of_phy_connect+0x50/0x74 |[c1053bc0] [c0572a88] emac_probe+0xd50/0x119c |[c1053c90] [c04cb770] platform_probe+0x74/0xa4 |[c1053cb0] [c04c8f08] really_probe+0x120/0x2b0 |[c1053cd0] [c04c9254] __driver_probe_device+0x1bc/0x1fc |[c1053d00] [c04c9334] driver_probe_device+0x38/0xa8 |[c1053d30] [c04c9560] __driver_attach+0xf4/0x10c |[c1053d50] [c04c6af8] bus_for_each_dev+0x68/0xd0 |[c1053d90] [c04c7c34] bus_add_driver+0xcc/0x1ec |[c1053dc0] [c04c9f6c] driver_register+0xcc/0x110 |[c1053de0] [c0aa3f40] emac_init+0x1c4/0x200 This bug showed up starting with v7.3-rc1. At the NIP in phy_link_topo_add_phy() is a netdev_need_ops_lock() check. This was added by the following commit ded86da4bbb7 ("net: ethtool: relax ethnl_req_get_phydev() locking assertion") The bug shows up because at the time of_phy_connect() was called, the netdev_ops were *not yet* determined. My fix is to move the code that sets netdev_ops+commac.ops+ethtool_ops further up as emac_init_config() derives that by looking at the device-tree and sets the required dev->phy_mode accordingly. During review, the Sashiko bot's AI stated that the commac assignment became a dead store. Great catch! To keep the original behavior as-is, one mentioned option "set dev->commac.ops = &emac_commac_ops only in the non-gige case?" sounded like a great plan. So the dev->commac.ops assignment for the non-gige-case moves into the else block. This patch was tested on a WD MyBook Live (RGMII). the device now works again. Fixes: ded86da4bbb7 ("net: ethtool: relax ethnl_req_get_phydev() locking assertion") Signed-off-by: Christian Lamparter Reviewed-by: Nicolai Buchwitz Reviewed-by: Maxime Chevallier Link: https://patch.msgid.link/49cd7343bc0e507c022071f3e2b5662b053dca73.1790007431.git.chunkeey@gmail.com Signed-off-by: Paolo Abeni --- drivers/net/ethernet/ibm/emac/core.c | 16 +++++++++------- 1 file changed, 9 insertions(+), 7 deletions(-) diff --git a/drivers/net/ethernet/ibm/emac/core.c b/drivers/net/ethernet/ibm/emac/core.c index 1d46cf6c2c12..e7043523457c 100644 --- a/drivers/net/ethernet/ibm/emac/core.c +++ b/drivers/net/ethernet/ibm/emac/core.c @@ -3044,6 +3044,15 @@ static int emac_probe(struct platform_device *ofdev) if (err) goto err_gone; + if (emac_phy_supports_gige(dev->phy_mode)) { + ndev->netdev_ops = &emac_gige_netdev_ops; + dev->commac.ops = &emac_commac_sg_ops; + } else { + ndev->netdev_ops = &emac_netdev_ops; + dev->commac.ops = &emac_commac_ops; + } + ndev->ethtool_ops = &emac_ethtool_ops; + dev->emacp = devm_platform_ioremap_resource(ofdev, 0); if (IS_ERR(dev->emacp)) { err = PTR_ERR(dev->emacp); @@ -3076,7 +3085,6 @@ static int emac_probe(struct platform_device *ofdev) dev->mdio_instance = platform_get_drvdata(dev->mdio_dev); /* Register with MAL */ - dev->commac.ops = &emac_commac_ops; dev->commac.dev = dev; dev->commac.tx_chan_mask = MAL_CHAN_MASK(dev->mal_tx_chan); dev->commac.rx_chan_mask = MAL_CHAN_MASK(dev->mal_rx_chan); @@ -3144,12 +3152,6 @@ static int emac_probe(struct platform_device *ofdev) ndev->features |= ndev->hw_features | NETIF_F_RXCSUM; } ndev->watchdog_timeo = 5 * HZ; - if (emac_phy_supports_gige(dev->phy_mode)) { - ndev->netdev_ops = &emac_gige_netdev_ops; - dev->commac.ops = &emac_commac_sg_ops; - } else - ndev->netdev_ops = &emac_netdev_ops; - ndev->ethtool_ops = &emac_ethtool_ops; /* MTU range: 46 - 1500 or whatever is in OF */ ndev->min_mtu = EMAC_MIN_MTU; From c3a66e5f5bab3912e9f84223c5a982bf4333d5a1 Mon Sep 17 00:00:00 2001 From: Donggeun Yoo Date: Thu, 24 Sep 2026 19:23:20 +0900 Subject: [PATCH 559/592] bpf: Zero-fill other CPUs when BPF_F_CPU creates a per-cpu hash element pcpu_init_value() initializes the per-cpu area of a newly created [lru_]percpu_hash element. The area is recycled, so when the value comes from a BPF program (onallcpus == false) it writes the running CPU's slot and zeroes the rest. bpf_percpu_hash_update() passes onallcpus == true, which delegates to pcpu_copy_value(). pcpu_copy_value() writes only the CPU named in map_flags when BPF_F_CPU is set, so on the create path the other slots keep the recycled element's values: update(k1, 0xdeadc0de, BPF_F_ALL_CPUS) every CPU holds 0xdeadc0de delete(k1) element back on the freelist update(k2, 0xc0ffee, BPF_F_CPU | 0) creates, writes CPU 0 only lookup(k2) CPU 0 0xc0ffee, rest 0xdeadc0de Zero-fill the other CPUs on that arm too. Fixes: c6936161fd55 ("bpf: Add BPF_F_CPU and BPF_F_ALL_CPUS flags support for percpu_hash and lru_percpu_hash maps") Signed-off-by: Donggeun Yoo Signed-off-by: Alexei Starovoitov Link: https://patch.msgid.link/20260924102321.2120434-2-donggeunyoo.kernel@gmail.com --- kernel/bpf/hashtab.c | 11 +++++++---- 1 file changed, 7 insertions(+), 4 deletions(-) diff --git a/kernel/bpf/hashtab.c b/kernel/bpf/hashtab.c index 4f495dcbf670..f9464e566f10 100644 --- a/kernel/bpf/hashtab.c +++ b/kernel/bpf/hashtab.c @@ -1054,14 +1054,17 @@ static void pcpu_init_value(struct bpf_htab *htab, void __percpu *pptr, /* When not setting the initial value on all cpus, zero-fill element * values for other cpus. Otherwise, bpf program has no way to ensure * known initial values for cpus other than current one - * (onallcpus=false always when coming from bpf prog). + * (onallcpus=false always when coming from bpf prog, + * map_flags & BPF_F_CPU when coming from syscall but setting + * only one cpu). */ - if (!onallcpus) { - int current_cpu = raw_smp_processor_id(); + if (!onallcpus || (map_flags & BPF_F_CPU)) { + int init_cpu = (map_flags & BPF_F_CPU) ? map_flags >> 32 : + raw_smp_processor_id(); int cpu; for_each_possible_cpu(cpu) { - if (cpu == current_cpu) + if (cpu == init_cpu) copy_map_value(&htab->map, per_cpu_ptr(pptr, cpu), value); else /* Since elem is preallocated, we cannot touch special fields */ zero_map_value(&htab->map, per_cpu_ptr(pptr, cpu)); From 3422808f4e959266ea7790101ac90b8341f48226 Mon Sep 17 00:00:00 2001 From: Donggeun Yoo Date: Thu, 24 Sep 2026 19:23:21 +0900 Subject: [PATCH 560/592] selftests/bpf: Test per-cpu initialization of a BPF_F_CPU created element The existing cpu_flag subtests always prime a key with BPF_F_ALL_CPUS before any BPF_F_CPU write, so the create path is never covered. Add a subtest that creates the element with BPF_F_CPU on a map with max_entries 1, so the key can only reuse the element the previous key released, and check that the CPUs the update did not name read back zero. Run it for PERCPU_HASH preallocated and BPF_F_NO_PREALLOC, whose per-cpu areas come from different allocators, and for LRU_PERCPU_HASH. Under BPF_F_NO_PREALLOC the reuse is only guaranteed on the cpu that ran the delete, so pin the thread across the pair, and name a cpu other than that one in map_flags. Signed-off-by: Donggeun Yoo Signed-off-by: Alexei Starovoitov Link: https://patch.msgid.link/20260924102321.2120434-3-donggeunyoo.kernel@gmail.com --- .../selftests/bpf/prog_tests/percpu_alloc.c | 105 ++++++++++++++++++ 1 file changed, 105 insertions(+) diff --git a/tools/testing/selftests/bpf/prog_tests/percpu_alloc.c b/tools/testing/selftests/bpf/prog_tests/percpu_alloc.c index a72ae0b29f6e..7b4a1e24363b 100644 --- a/tools/testing/selftests/bpf/prog_tests/percpu_alloc.c +++ b/tools/testing/selftests/bpf/prog_tests/percpu_alloc.c @@ -1,4 +1,6 @@ // SPDX-License-Identifier: GPL-2.0 +#define _GNU_SOURCE +#include #include #include "cgroup_helpers.h" #include "percpu_alloc_array.skel.h" @@ -350,6 +352,103 @@ static void test_lru_percpu_hash_cpu_flag(void) test_percpu_map_cpu_flag(BPF_MAP_TYPE_LRU_PERCPU_HASH); } +/* + * A BPF_F_CPU update that creates an element must zero the value on the other + * cpus, rather than leave them holding whatever the recycled element last + * contained. max_entries is 1 so the second key can only reuse the element + * the first one released. + */ +static void test_percpu_map_cpu_flag_create(enum bpf_map_type map_type, __u32 map_flags) +{ + LIBBPF_OPTS(bpf_map_create_opts, opts, .map_flags = map_flags); + const u32 stale = 0xDEADC0DE, fresh = 0xC0FFEE; + int nr_cpus, cpu, map_fd, err, key; + int pinned_cpu, value_cpu; + cpu_set_t old_mask, new_mask; + bool restore_mask = false; + u32 value; + u64 flags; + + nr_cpus = libbpf_num_possible_cpus(); + if (!ASSERT_GT(nr_cpus, 0, "libbpf_num_possible_cpus")) + return; + + if (nr_cpus < 2) { + test__skip(); + return; + } + + map_fd = bpf_map_create(map_type, "cpu_flag_create", sizeof(key), sizeof(value), 1, &opts); + if (!ASSERT_GE(map_fd, 0, "bpf_map_create")) + return; + + /* NO_PREALLOC recycles per cpu, so keep the delete and the create on one cpu. */ + err = sched_getaffinity(0, sizeof(old_mask), &old_mask); + if (!ASSERT_OK(err, "sched_getaffinity")) + goto out; + + pinned_cpu = sched_getcpu(); + if (!ASSERT_GE(pinned_cpu, 0, "sched_getcpu")) + goto out; + + CPU_ZERO(&new_mask); + CPU_SET(pinned_cpu, &new_mask); + err = sched_setaffinity(0, sizeof(new_mask), &new_mask); + if (!ASSERT_OK(err, "sched_setaffinity")) + goto out; + restore_mask = true; + + value_cpu = pinned_cpu ? 0 : 1; + + key = 1; + value = stale; + err = bpf_map_update_elem(map_fd, &key, &value, BPF_F_ALL_CPUS); + if (!ASSERT_OK(err, "bpf_map_update_elem all_cpus")) + goto out; + + err = bpf_map_delete_elem(map_fd, &key); + if (!ASSERT_OK(err, "bpf_map_delete_elem")) + goto out; + + key = 2; + value = fresh; + flags = (u64)value_cpu << 32 | BPF_F_CPU; + err = bpf_map_update_elem(map_fd, &key, &value, flags); + if (!ASSERT_OK(err, "bpf_map_update_elem specified cpu")) + goto out; + + for (cpu = 0; cpu < nr_cpus; cpu++) { + value = 0; + flags = (u64)cpu << 32 | BPF_F_CPU; + err = bpf_map_lookup_elem_flags(map_fd, &key, &value, flags); + if (!ASSERT_OK(err, "bpf_map_lookup_elem_flags specified cpu")) + goto out; + if (!ASSERT_EQ(value, cpu == value_cpu ? fresh : 0, "value on specified cpu")) + goto out; + } + +out: + if (restore_mask) + sched_setaffinity(0, sizeof(old_mask), &old_mask); + close(map_fd); +} + +static void test_percpu_hash_cpu_flag_create(void) +{ + test_percpu_map_cpu_flag_create(BPF_MAP_TYPE_PERCPU_HASH, 0); +} + +static void test_percpu_hash_cpu_flag_create_malloc(void) +{ + test_percpu_map_cpu_flag_create(BPF_MAP_TYPE_PERCPU_HASH, BPF_F_NO_PREALLOC); +} + +static void test_lru_percpu_hash_cpu_flag_create(void) +{ + /* lru without prealloc is -ENOTSUPP, so there is no malloc variant */ + test_percpu_map_cpu_flag_create(BPF_MAP_TYPE_LRU_PERCPU_HASH, 0); +} + static void test_percpu_cgroup_storage_cpu_flag(void) { struct percpu_alloc_array *skel = NULL; @@ -454,6 +553,12 @@ void test_percpu_alloc(void) test_percpu_hash_cpu_flag(); if (test__start_subtest("cpu_flag_lru_percpu_hash")) test_lru_percpu_hash_cpu_flag(); + if (test__start_subtest("cpu_flag_create_percpu_hash")) + test_percpu_hash_cpu_flag_create(); + if (test__start_subtest("cpu_flag_create_percpu_hash_malloc")) + test_percpu_hash_cpu_flag_create_malloc(); + if (test__start_subtest("cpu_flag_create_lru_percpu_hash")) + test_lru_percpu_hash_cpu_flag_create(); if (test__start_subtest("cpu_flag_percpu_cgroup_storage")) test_percpu_cgroup_storage_cpu_flag(); if (test__start_subtest("cpu_flag_array")) From c2cdef41e0b4d8ed23a5b41e6ad4e64594e055e4 Mon Sep 17 00:00:00 2001 From: Aleksei Sviridkin Date: Fri, 18 Sep 2026 04:50:19 +0300 Subject: [PATCH 561/592] net: dsa: mt7530: fix NULL dereference on unbind of MT7531 and MT7621 The core and io supplies are only requested for ID_MT7530: both the devm_regulator_get() in probe and the regulator_enable() in mt7530_setup() are guarded by the switch id, but mt7530_remove() disables them unconditionally. On an MT7621 or an MT7531 both pointers are still NULL from devm_kzalloc(), so rmmod or a sysfs unbind calls regulator_disable() on NULL. Fixes: ddda1ac116c8 ("net: dsa: mt7530: support the 7530 switch on the Mediatek MT7621 SoC") Signed-off-by: Aleksei Sviridkin Link: https://patch.msgid.link/20260918015020.2518315-2-f@lex.la Signed-off-by: Jakub Kicinski --- drivers/net/dsa/mt7530-mdio.c | 18 ++++++++++-------- 1 file changed, 10 insertions(+), 8 deletions(-) diff --git a/drivers/net/dsa/mt7530-mdio.c b/drivers/net/dsa/mt7530-mdio.c index 784dd58a7158..de42f70afcfa 100644 --- a/drivers/net/dsa/mt7530-mdio.c +++ b/drivers/net/dsa/mt7530-mdio.c @@ -227,15 +227,17 @@ mt7530_remove(struct mdio_device *mdiodev) if (!priv) return; - ret = regulator_disable(priv->core_pwr); - if (ret < 0) - dev_err(priv->dev, - "Failed to disable core power: %d\n", ret); + if (priv->id == ID_MT7530) { + ret = regulator_disable(priv->core_pwr); + if (ret < 0) + dev_err(priv->dev, + "Failed to disable core power: %d\n", ret); - ret = regulator_disable(priv->io_pwr); - if (ret < 0) - dev_err(priv->dev, "Failed to disable io pwr: %d\n", - ret); + ret = regulator_disable(priv->io_pwr); + if (ret < 0) + dev_err(priv->dev, "Failed to disable io pwr: %d\n", + ret); + } mt7530_remove_common(priv); From 0d80ba0a204c6a16bd7778b50de578dff107c0fe Mon Sep 17 00:00:00 2001 From: Aleksei Sviridkin Date: Fri, 18 Sep 2026 04:50:20 +0300 Subject: [PATCH 562/592] net: dsa: mt7530: leave the MDIO IRQ mappings to regmap-irq mt7530_remove_common() disposes the per-PHY interrupt mappings from .remove, but the regmap-irq chip that owns the domain is devm-registered, so its parent interrupt is only freed once .remove has returned. The switch's own regmap-irq thread can therefore still dispatch on a mapping that is already gone: irq_find_mapping() returns 0, irq_to_desc() returns NULL and handle_nested_irq() locks desc->lock without checking it. The attached PHYs have not given those interrupts back yet either, which the kernel warns about a moment before the fault. regmap_del_irq_chip() disposes the same mappings itself, after freeing the parent interrupt and before removing the domain, so there is nothing left for the driver to do here. Until it runs the descriptors stay alive, and a late dispatch on one of them is harmless: dsa_unregister_switch() has freed the PHY handlers by then, so handle_nested_irq() finds no action and returns. Fixes: 254f6b272e3b ("dsa: mt7530: Utilize REGMAP_IRQ for interrupt handling") Signed-off-by: Aleksei Sviridkin Link: https://patch.msgid.link/20260918015020.2518315-3-f@lex.la Signed-off-by: Jakub Kicinski --- drivers/net/dsa/mt7530.c | 3 --- 1 file changed, 3 deletions(-) diff --git a/drivers/net/dsa/mt7530.c b/drivers/net/dsa/mt7530.c index 3e61eb3c2b1e..96832852c65a 100644 --- a/drivers/net/dsa/mt7530.c +++ b/drivers/net/dsa/mt7530.c @@ -3593,9 +3593,6 @@ EXPORT_SYMBOL_GPL(mt7530_probe_common); void mt7530_remove_common(struct mt7530_priv *priv) { - if (priv->irq_domain) - mt7530_free_mdio_irq(priv); - dsa_unregister_switch(priv->ds); mutex_destroy(&priv->reg_mutex); From 26cc0e69cce062cd3aa6fae33074684669c35a71 Mon Sep 17 00:00:00 2001 From: Hui Peng Date: Mon, 21 Sep 2026 05:10:01 +0000 Subject: [PATCH 563/592] mctp: route: iterate socket tag list in mctp_lookup_prealloc_tag() When a socket transmits a packet with MCTP_TAG_PREALLOC set, mctp_lookup_prealloc_tag() iterates over the per-netns &mns->keys list and matches netid, req_tag, peer_addr, and manual_alloc, without checking whether tmp->sk == &msk->sk. This allows any MCTP socket in the same network namespace to use and consume another socket's preallocated tag. Iterate the socket's own tag list (&msk->keys via sklist) instead of the namespace-wide &mns->keys list in mctp_lookup_prealloc_tag(), ensuring that only tags allocated by msk are matched. Tested in QEMU against Linux 7.3.0-rc3 by allocating a manual tag (0x18) on socket A via SIOCMCTPALLOCTAG for peer EID 9 and sending a 4-byte message with MCTP_TAG_PREALLOC from socket B in the same network namespace. On the unfixed kernel, sendto(sock_b) using socket A's preallocated tag succeeds (ret = 4); with this patch applied, sendto(sock_b) fails with -ENOENT (errno = 2) while sendto(sock_a) succeeds (ret = 4). Fixes: 63ed1aab3d40 ("mctp: Add SIOCMCTP{ALLOC,DROP}TAG ioctls for tag control") Suggested-by: Jeremy Kerr Cc: stable@vger.kernel.org Signed-off-by: Hui Peng Link: https://patch.msgid.link/20260921051002.1656692-1-benquike@gmail.com Signed-off-by: Jakub Kicinski --- net/mctp/route.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/net/mctp/route.c b/net/mctp/route.c index b19c63a5691a..e7c95eeacb48 100644 --- a/net/mctp/route.c +++ b/net/mctp/route.c @@ -825,7 +825,7 @@ static struct mctp_sk_key *mctp_lookup_prealloc_tag(struct mctp_sock *msk, spin_lock_irqsave(&mns->keys_lock, flags); - hlist_for_each_entry(tmp, &mns->keys, hlist) { + hlist_for_each_entry(tmp, &msk->keys, sklist) { if (tmp->net != netid) continue; From fe99bbeee5c5dbd3abc30721a8079ced59649d97 Mon Sep 17 00:00:00 2001 From: Yilin Zhang Date: Thu, 24 Sep 2026 12:49:00 +0800 Subject: [PATCH 564/592] tcp: fix use-after-free of retransmit_skb_hint in tcp_send_synack() When tcp_send_synack() replaces the cloned SYN skb at the head of the retransmit queue with a copy, it frees the original with tcp_rtx_queue_unlink_and_free() and only repairs tp->highest_sack. tp->retransmit_skb_hint keeps pointing at the freed skbuff_fclone_cache object. The dangling hint is read in tcp_verify_retransmit_hint() and used as the root of the rbtree walk in tcp_xmit_retransmit_queue(). An unprivileged TFO client (sendmsg(MSG_FASTOPEN)) can arm the hint with an attacker-supplied ICMP fragmentation-needed message, after which a simultaneous open frees the armed SYN skb: BUG: KASAN: slab-use-after-free in tcp_mark_skb_lost (net/ipv4/tcp_input.c:1316) Read of size 4 at addr ffff88800604d928 by task swapper/1/0 Call Trace: tcp_mark_skb_lost (net/ipv4/tcp_input.c:1316) tcp_simple_retransmit (net/ipv4/tcp_input.c:3158) tcp_v4_err (net/ipv4/tcp_ipv4.c:587) Sync the hint to the copy. Fixes: c31b70c9968f ("tcp: Add logic to check for SYN w/ data in tcp_simple_retransmit") Reported-by: Kimi Security Team Tested-by: Weiming Shi Signed-off-by: Yilin Zhang Reviewed-by: Eric Dumazet Link: https://patch.msgid.link/8a9dff4063a2745653b7e88ceb745d75efa16e68.1790224474.git.yilinzhang@moonshot.ai Signed-off-by: Jakub Kicinski --- net/ipv4/tcp_output.c | 3 +++ 1 file changed, 3 insertions(+) diff --git a/net/ipv4/tcp_output.c b/net/ipv4/tcp_output.c index d960e3de7d50..e0c392e29de5 100644 --- a/net/ipv4/tcp_output.c +++ b/net/ipv4/tcp_output.c @@ -3886,6 +3886,7 @@ void tcp_send_active_reset(struct sock *sk, enum sk_rst_reason reason) */ int tcp_send_synack(struct sock *sk) { + struct tcp_sock *tp = tcp_sk(sk); struct sk_buff *skb; skb = tcp_rtx_queue_head(sk); @@ -3903,6 +3904,8 @@ int tcp_send_synack(struct sock *sk) if (!nskb) return -ENOMEM; INIT_LIST_HEAD(&nskb->tcp_tsorted_anchor); + if (skb == tp->retransmit_skb_hint) + tp->retransmit_skb_hint = nskb; tcp_highest_sack_replace(sk, skb, nskb); tcp_rtx_queue_unlink_and_free(skb, sk); __skb_header_release(nskb); From f75f21ef36285e5f56ee0c428bd2909ee81165b9 Mon Sep 17 00:00:00 2001 From: Ming Wang Date: Sun, 20 Sep 2026 15:44:59 +0800 Subject: [PATCH 565/592] net: usb: cdc_mbim: add MeiG Smart SRM821 to ZLP whitelist The MeiG Smart SRM821 5G module (0x2dee:0x4d53) crashes and drops off the USB bus when it receives a Zero Length Packet (ZLP) after sending or receiving an NTB of exactly 16384 bytes (tx_max). According to the MBIM specification, devices do not require a ZLP if the NTB size is exactly dwNtbOutMaxSize. However, the cdc_mbim driver defaults to sending ZLPs for devices not explicitly whitelisted to accommodate non-conformant hardware. This default behavior breaks the strictly conformant MeiG SRM821 module. Add this device to the ZLP conformance whitelist (cdc_mbim_info) so the driver will pad the NTB to avoid sending ZLPs, preventing the device firmware from crashing. Cc: stable@vger.kernel.org Signed-off-by: Ming Wang Link: https://patch.msgid.link/20260920074500.826121-1-wangming01@loongson.cn Signed-off-by: Jakub Kicinski --- drivers/net/usb/cdc_mbim.c | 5 +++++ 1 file changed, 5 insertions(+) diff --git a/drivers/net/usb/cdc_mbim.c b/drivers/net/usb/cdc_mbim.c index 877fb0ed7d3d..a7010a0664c7 100644 --- a/drivers/net/usb/cdc_mbim.c +++ b/drivers/net/usb/cdc_mbim.c @@ -635,6 +635,11 @@ static const struct usb_device_id mbim_devs[] = { .driver_info = (unsigned long)&cdc_mbim_info, }, + /* MeiG Smart SRM821 ZLP conformance */ + { USB_DEVICE_AND_INTERFACE_INFO(0x2dee, 0x4d53, USB_CLASS_COMM, USB_CDC_SUBCLASS_MBIM, USB_CDC_PROTO_NONE), + .driver_info = (unsigned long)&cdc_mbim_info, + }, + /* Some Huawei devices, ME906s-158 (12d1:15c1) and E3372 * (12d1:157d), are known to fail unless the NDP is placed * after the IP packets. Applying the quirk to all Huawei From a940003f44e7e441c228151dd212642152700ec8 Mon Sep 17 00:00:00 2001 From: Aleksei Sviridkin Date: Mon, 21 Sep 2026 01:20:44 +0300 Subject: [PATCH 566/592] net: phylink: record the PHY only once bringup cannot fail phylink_bringup_phy() stores the PHY in pl->phydev before its last fallible step: on a MAC whose phylink ops implement LPI, phy_eee_rx_clock_stop() can fail with a real MDIO error. The callers unwind with phy_detach(), which knows nothing about pl->phydev, so a pointer to a PHY that is no longer attached outlives the failed connect. What that costs depends on how the caller got here. phylink_connect_phy() goes through phylink_attach_phy(), which refuses to attach while pl->phydev is set, turning a transient MDIO error into a permanent -EBUSY. The SFP path is worse than that: sfp_sm_probe_phy() answers the failure with phy_device_remove() and phy_device_free(), and it assigns sfp->mod_phy only past that error return, so nothing clears pl->phydev and it is left pointing at a freed phy_device that phylink_resolve() and the ethtool helpers go on reading. phylink_fwnode_phy_connect() has no such check, so a later connect overwrites the stale pointer and hides the problem. A disconnect does not: phylink_disconnect_phy() hands that pointer to phy_disconnect(), and the second phy_detach() on the same PHY drops references the first one already released. Found while making a DSA port survive a PHY whose driver arrives after the switch probes: keeping the port across a failed connect and retrying is what makes this window reachable. Publish the pointer after the last call that can fail instead of unwinding it afterwards. Nothing between the two points reads pl->phydev, and the registration that follows cannot fail: phy_request_interrupt() falls back to polling on its own. The PHY-side state keeps the order it had, so no MDIO operation moves relative to another. Fixes: 03abf2a7c654 ("net: phylink: add EEE management") Signed-off-by: Aleksei Sviridkin Link: https://patch.msgid.link/20260920222044.1752860-1-f@lex.la Signed-off-by: Jakub Kicinski --- drivers/net/phy/phylink.c | 20 +++++++++++++++++--- 1 file changed, 17 insertions(+), 3 deletions(-) diff --git a/drivers/net/phy/phylink.c b/drivers/net/phy/phylink.c index a1458da8111b..1bbcf46c8356 100644 --- a/drivers/net/phy/phylink.c +++ b/drivers/net/phy/phylink.c @@ -2129,7 +2129,6 @@ static int phylink_bringup_phy(struct phylink *pl, struct phy_device *phy, mutex_lock(&pl->phydev_mutex); mutex_lock(&phy->lock); mutex_lock(&pl->state_mutex); - pl->phydev = phy; pl->phy_state.interface = interface; pl->phy_state.pause = MLO_PAUSE_NONE; pl->phy_state.speed = SPEED_UNKNOWN; @@ -2196,10 +2195,25 @@ static int phylink_bringup_phy(struct phylink *pl, struct phy_device *phy, ret = 0; } - if (ret == 0 && phy_interrupt_is_valid(phy)) + if (ret) + return ret; + + /* Nothing below can fail, so the PHY can be recorded now. Doing it + * here rather than above keeps a failed bringup from leaving + * pl->phydev pointing at a PHY the caller is about to detach. + */ + mutex_lock(&pl->phydev_mutex); + mutex_lock(&phy->lock); + mutex_lock(&pl->state_mutex); + pl->phydev = phy; + mutex_unlock(&pl->state_mutex); + mutex_unlock(&phy->lock); + mutex_unlock(&pl->phydev_mutex); + + if (phy_interrupt_is_valid(phy)) phy_request_interrupt(phy); - return ret; + return 0; } static int phylink_attach_phy(struct phylink *pl, struct phy_device *phy, From 3173cba1170131816972ed2b6185cb970ba8b747 Mon Sep 17 00:00:00 2001 From: Jiawen Wu Date: Mon, 21 Sep 2026 15:15:49 +0800 Subject: [PATCH 567/592] net: libwx: fix races in Tx timestamp handling wx->ptp_tx_skb is shared between the Tx path, the PTP auxiliary worker and the timestamp cleanup paths. The WX_STATE_PTP_TX_IN_PROGRESS bit prevents multiple Tx paths from submitting timestamp requests, but does not serialize the worker against cleanup. As a result, wx_ptp_clear_tx_timestamp() can free an skb after wx_ptp_tx_hwtstamp_work() has obtained its pointer. The worker may then pass the freed skb to skb_tstamp_tx() and release the same reference again. The cleanup path may also clear the in-progress bit while the worker is still processing the old skb. This allows the Tx path to publish a new skb which the worker can subsequently overwrite with NULL, leaking its reference. Add a dedicated spinlock to protect publication and consumption of the Tx timestamp skb. Detach the skb and clear the in-progress bit while holding the lock, then deliver the timestamp and release the skb after dropping it. Use the same locked cleanup in the quiesce path, but keep the detach there free of register accesses: quiesce runs during PCIe error recovery, where MMIO is not reliable, and it deliberately did not touch the device before. The lock is taken with interrupts disabled, because netpoll can call ndo_start_xmit() with hard interrupts already off. When handling a Tx DMA mapping failure, keep the transmit path reference until after comparing the skb under the lock. This prevents skb address reuse from making the error path mistake a newer timestamp request for the failed one. Fixes: 06e75161b9d4 ("net: wangxun: Add support for PTP clock") Reported-by: Sashiko Link: https://netdev-ai.bots.linux.dev/sashiko/#/patchset/6C7EC12D69217315%2B20260818074721.45536-1-jiawenwu%40trustnetic.com Signed-off-by: Jiawen Wu Link: https://patch.msgid.link/77431AF9A0E369F3+20260921071549.1141804-1-jiawenwu@trustnetic.com Signed-off-by: Jakub Kicinski --- drivers/net/ethernet/wangxun/libwx/wx_hw.c | 1 + drivers/net/ethernet/wangxun/libwx/wx_lib.c | 47 ++++-- drivers/net/ethernet/wangxun/libwx/wx_ptp.c | 144 ++++++++++++------- drivers/net/ethernet/wangxun/libwx/wx_type.h | 2 + 4 files changed, 133 insertions(+), 61 deletions(-) diff --git a/drivers/net/ethernet/wangxun/libwx/wx_hw.c b/drivers/net/ethernet/wangxun/libwx/wx_hw.c index 122c4952d203..113552586be7 100644 --- a/drivers/net/ethernet/wangxun/libwx/wx_hw.c +++ b/drivers/net/ethernet/wangxun/libwx/wx_hw.c @@ -2518,6 +2518,7 @@ int wx_sw_init(struct wx *wx) } spin_lock_init(&wx->hw_stats_lock); + spin_lock_init(&wx->ptp_tx_lock); mutex_init(&wx->reset_lock); bitmap_zero(wx->state, WX_STATE_NBITS); bitmap_zero(wx->flags, WX_PF_FLAGS_NBITS); diff --git a/drivers/net/ethernet/wangxun/libwx/wx_lib.c b/drivers/net/ethernet/wangxun/libwx/wx_lib.c index ed5aad7857bd..dcbf5811046e 100644 --- a/drivers/net/ethernet/wangxun/libwx/wx_lib.c +++ b/drivers/net/ethernet/wangxun/libwx/wx_lib.c @@ -1200,9 +1200,11 @@ static int wx_tx_map(struct wx_ring *tx_ring, i--; } - dev_kfree_skb_any(first->skb); - first->skb = NULL; - + /* first->skb is released by the caller, which keeps a reference on it + * until the PTP cleanup has compared it against wx->ptp_tx_skb. That + * prevents the address from being reused by a newer request while the + * comparison is pending. + */ tx_ring->next_to_use = i; return -ENOMEM; @@ -1649,9 +1651,11 @@ static netdev_tx_t wx_xmit_frame_ring(struct sk_buff *skb, if (unlikely(skb_shinfo(skb)->tx_flags & SKBTX_HW_TSTAMP) && wx->ptp_clock) { + unsigned long flags; + + spin_lock_irqsave(&wx->ptp_tx_lock, flags); if (wx->tstamp_config.tx_type == HWTSTAMP_TX_ON && - !test_and_set_bit_lock(WX_STATE_PTP_TX_IN_PROGRESS, - wx->state)) { + !test_and_set_bit(WX_STATE_PTP_TX_IN_PROGRESS, wx->state)) { skb_shinfo(skb)->tx_flags |= SKBTX_IN_PROGRESS; tx_flags |= WX_TX_FLAGS_TSTAMP; wx->ptp_tx_skb = skb_get(skb); @@ -1659,6 +1663,7 @@ static netdev_tx_t wx_xmit_frame_ring(struct sk_buff *skb, } else { wx->tx_hwtstamp_skipped++; } + spin_unlock_irqrestore(&wx->ptp_tx_lock, flags); } /* record initial flags and protocol */ @@ -1677,19 +1682,35 @@ static netdev_tx_t wx_xmit_frame_ring(struct sk_buff *skb, wx->atr(tx_ring, first, ptype); if (wx_tx_map(tx_ring, first, hdr_len)) - goto cleanup_tx_tstamp; + goto out_drop; return NETDEV_TX_OK; out_drop: + /* The frame never reached the hardware, so no timestamp will ever be + * reported for it and the request has to be cancelled. The slot is + * shared, though: wx_ptp_clear_tx_timestamp() or wx_ptp_tx_hang() may + * have dropped our request already, and a transmit on another queue + * can have claimed the slot since. Only cancel it while it is still + * ours, otherwise we would free somebody else's skb and release their + * in-progress bit. + */ + if (unlikely(tx_flags & WX_TX_FLAGS_TSTAMP)) { + struct sk_buff *ptp_tx_skb = NULL; + unsigned long flags; + + spin_lock_irqsave(&wx->ptp_tx_lock, flags); + if (wx->ptp_tx_skb == skb) { + ptp_tx_skb = wx->ptp_tx_skb; + wx->ptp_tx_skb = NULL; + clear_bit(WX_STATE_PTP_TX_IN_PROGRESS, wx->state); + wx->tx_hwtstamp_errors++; + } + spin_unlock_irqrestore(&wx->ptp_tx_lock, flags); + + dev_kfree_skb_any(ptp_tx_skb); + } dev_kfree_skb_any(first->skb); first->skb = NULL; -cleanup_tx_tstamp: - if (unlikely(tx_flags & WX_TX_FLAGS_TSTAMP)) { - dev_kfree_skb_any(wx->ptp_tx_skb); - wx->ptp_tx_skb = NULL; - wx->tx_hwtstamp_errors++; - clear_bit_unlock(WX_STATE_PTP_TX_IN_PROGRESS, wx->state); - } return NETDEV_TX_OK; } diff --git a/drivers/net/ethernet/wangxun/libwx/wx_ptp.c b/drivers/net/ethernet/wangxun/libwx/wx_ptp.c index 4708e7f3958f..65b8937f6e94 100644 --- a/drivers/net/ethernet/wangxun/libwx/wx_ptp.c +++ b/drivers/net/ethernet/wangxun/libwx/wx_ptp.c @@ -129,6 +129,34 @@ static int wx_ptp_settime64(struct ptp_clock_info *ptp, return 0; } +/** + * __wx_ptp_detach_tx_skb - detach the skb tracking the Tx timestamp request + * @wx: the private board structure + * + * Detach the skb of the outstanding request and release the in-progress bit, + * so that a new request can be submitted. + * + * This performs no register access. Callers that need a timestamp the hardware + * may have left latched must unlatch it themselves, while the device is known + * to be alive. wx_ptp_quiesce() runs during PCIe error recovery, where MMIO is + * not reliable, and therefore deliberately skips the unlatch. + * + * Context: Expects wx->ptp_tx_lock to be held by the caller. + * Return: the detached skb, or NULL if no request was outstanding. The caller + * owns the returned reference and must release it once the lock is dropped. + */ +static struct sk_buff *__wx_ptp_detach_tx_skb(struct wx *wx) +{ + struct sk_buff *skb = wx->ptp_tx_skb; + + lockdep_assert_held(&wx->ptp_tx_lock); + + wx->ptp_tx_skb = NULL; + clear_bit(WX_STATE_PTP_TX_IN_PROGRESS, wx->state); + + return skb; +} + /** * wx_ptp_clear_tx_timestamp - utility function to clear Tx timestamp state * @wx: the private board structure @@ -139,12 +167,16 @@ static int wx_ptp_settime64(struct ptp_clock_info *ptp, */ static void wx_ptp_clear_tx_timestamp(struct wx *wx) { + struct sk_buff *skb; + unsigned long flags; + + spin_lock_irqsave(&wx->ptp_tx_lock, flags); + /* Unlatch a timestamp the hardware may have left pending. */ rd32ptp(wx, WX_TSC_1588_STMPH); - if (wx->ptp_tx_skb) { - dev_kfree_skb_any(wx->ptp_tx_skb); - wx->ptp_tx_skb = NULL; - } - clear_bit_unlock(WX_STATE_PTP_TX_IN_PROGRESS, wx->state); + skb = __wx_ptp_detach_tx_skb(wx); + spin_unlock_irqrestore(&wx->ptp_tx_lock, flags); + + dev_kfree_skb_any(skb); } /** @@ -175,49 +207,54 @@ static void wx_ptp_convert_to_hwtstamp(struct wx *wx, } /** - * wx_ptp_tx_hwtstamp - utility function which checks for TX time stamp + * wx_ptp_tx_hwtstamp_work - check for a pending Tx time stamp * @wx: the private board struct * - * if the timestamp is valid, we convert it into the timecounter ns - * value, then store that result into the shhwtstamps structure which - * is passed up the network stack + * If a Tx timestamp request is outstanding and the hardware has latched a + * valid value, we convert it into the timecounter ns value, then store that + * result into the shhwtstamps structure which is passed up the network stack. + * + * Return: 0 when there is nothing left to poll for, -1 when the timestamp is + * not available yet and the caller should poll again. */ -static void wx_ptp_tx_hwtstamp(struct wx *wx) -{ - struct skb_shared_hwtstamps shhwtstamps; - struct sk_buff *skb = wx->ptp_tx_skb; - u64 regval = 0; - - regval |= (u64)rd32ptp(wx, WX_TSC_1588_STMPL); - regval |= (u64)rd32ptp(wx, WX_TSC_1588_STMPH) << 32; - - wx_ptp_convert_to_hwtstamp(wx, &shhwtstamps, regval); - - wx->ptp_tx_skb = NULL; - clear_bit_unlock(WX_STATE_PTP_TX_IN_PROGRESS, wx->state); - skb_tstamp_tx(skb, &shhwtstamps); - dev_kfree_skb_any(skb); - wx->tx_hwtstamp_pkts++; -} - static int wx_ptp_tx_hwtstamp_work(struct wx *wx) { + struct skb_shared_hwtstamps shhwtstamps; + unsigned long flags; + struct sk_buff *skb; u32 tsynctxctl; + u64 regval = 0; + + spin_lock_irqsave(&wx->ptp_tx_lock, flags); /* we have to have a valid skb to poll for a timestamp */ if (!wx->ptp_tx_skb) { - wx_ptp_clear_tx_timestamp(wx); + rd32ptp(wx, WX_TSC_1588_STMPH); + __wx_ptp_detach_tx_skb(wx); + spin_unlock_irqrestore(&wx->ptp_tx_lock, flags); return 0; } /* stop polling once we have a valid timestamp */ tsynctxctl = rd32ptp(wx, WX_TSC_1588_CTL); - if (tsynctxctl & WX_TSC_1588_CTL_VALID) { - wx_ptp_tx_hwtstamp(wx); - return 0; + if (!(tsynctxctl & WX_TSC_1588_CTL_VALID)) { + spin_unlock_irqrestore(&wx->ptp_tx_lock, flags); + return -1; } - return -1; + regval |= (u64)rd32ptp(wx, WX_TSC_1588_STMPL); + regval |= (u64)rd32ptp(wx, WX_TSC_1588_STMPH) << 32; + skb = wx->ptp_tx_skb; + wx->ptp_tx_skb = NULL; + clear_bit(WX_STATE_PTP_TX_IN_PROGRESS, wx->state); + spin_unlock_irqrestore(&wx->ptp_tx_lock, flags); + + wx_ptp_convert_to_hwtstamp(wx, &shhwtstamps, regval); + skb_tstamp_tx(skb, &shhwtstamps); + dev_kfree_skb_any(skb); + wx->tx_hwtstamp_pkts++; + + return 0; } /** @@ -296,24 +333,29 @@ static void wx_ptp_rx_hang(struct wx *wx) */ static void wx_ptp_tx_hang(struct wx *wx) { - bool timeout = time_is_before_jiffies(wx->ptp_tx_start + - WX_PTP_TX_TIMEOUT); + struct sk_buff *skb = NULL; + unsigned long flags; - if (!wx->ptp_tx_skb) - return; - - if (!test_bit(WX_STATE_PTP_TX_IN_PROGRESS, wx->state)) - return; + spin_lock_irqsave(&wx->ptp_tx_lock, flags); /* If we haven't received a timestamp within the timeout, it is * reasonable to assume that it will never occur, so we can unlock the * timestamp bit when this occurs. */ - if (timeout) { - wx_ptp_clear_tx_timestamp(wx); - wx->tx_hwtstamp_timeouts++; - dev_warn(&wx->pdev->dev, "clearing Tx timestamp hang\n"); + if (wx->ptp_tx_skb && + test_bit(WX_STATE_PTP_TX_IN_PROGRESS, wx->state) && + time_is_before_jiffies(wx->ptp_tx_start + WX_PTP_TX_TIMEOUT)) { + rd32ptp(wx, WX_TSC_1588_STMPH); + skb = __wx_ptp_detach_tx_skb(wx); } + spin_unlock_irqrestore(&wx->ptp_tx_lock, flags); + + if (!skb) + return; + + dev_kfree_skb_any(skb); + wx->tx_hwtstamp_timeouts++; + dev_warn(&wx->pdev->dev, "clearing Tx timestamp hang\n"); } static long wx_ptp_do_aux_work(struct ptp_clock_info *ptp) @@ -841,6 +883,9 @@ EXPORT_SYMBOL(wx_ptp_stop); void wx_ptp_quiesce(struct wx *wx) { + struct sk_buff *skb; + unsigned long flags; + if (!test_and_clear_bit(WX_STATE_PTP_RUNNING, wx->state)) return; @@ -849,11 +894,14 @@ void wx_ptp_quiesce(struct wx *wx) if (wx->ptp_clock) ptp_cancel_worker_sync(wx->ptp_clock); - if (wx->ptp_tx_skb) { - dev_kfree_skb_any(wx->ptp_tx_skb); - wx->ptp_tx_skb = NULL; - } - clear_bit_unlock(WX_STATE_PTP_TX_IN_PROGRESS, wx->state); + /* Drop a pending Tx timestamp request. Do not touch the registers + * here: quiesce runs during PCIe error recovery, where the device may + * already be gone and MMIO is not reliable. + */ + spin_lock_irqsave(&wx->ptp_tx_lock, flags); + skb = __wx_ptp_detach_tx_skb(wx); + spin_unlock_irqrestore(&wx->ptp_tx_lock, flags); + dev_kfree_skb_any(skb); if (wx->ptp_clock) { ptp_clock_unregister(wx->ptp_clock); diff --git a/drivers/net/ethernet/wangxun/libwx/wx_type.h b/drivers/net/ethernet/wangxun/libwx/wx_type.h index 9454e90258d8..afd980dbb793 100644 --- a/drivers/net/ethernet/wangxun/libwx/wx_type.h +++ b/drivers/net/ethernet/wangxun/libwx/wx_type.h @@ -1430,6 +1430,8 @@ struct wx { unsigned long last_overflow_check; unsigned long last_rx_ptp_check; unsigned long ptp_tx_start; + /* protects ptp_tx_skb, ptp_tx_start and the in-progress state bit */ + spinlock_t ptp_tx_lock; seqlock_t hw_tc_lock; /* seqlock for ptp */ struct cyclecounter hw_cc; struct timecounter hw_tc; From 26b2bd70d22457556e2fa01cbf1192cb1a94d619 Mon Sep 17 00:00:00 2001 From: Ilya Maximets Date: Mon, 21 Sep 2026 16:55:43 +0200 Subject: [PATCH 568/592] net: openvswitch: conntrack: avoid modifying shared unconfirmed ct entry In a case where skb with an unconfirmed ct entry gets cloned, we may end up committing both but with different sets of extensions. The series of events: 1. The first clone wants to commit and runs the helpers wiring up the extension pointer into the expectation list. 2. Then it looses the confirmation keeping the entry unconfirmed. 3. Second clone now wants to commit labels and adds the new extension for that breaking the pointer in the expectation list causing UAF on the destruction path later. While this is possible to trigger, there should be no practical network pipeline where committing both clones without modifications into the same zone is needed. So, let's just reset the entry in case for some reason we got an skb with a shared one during commit. This doesn't affect any known use cases, but avoids any potential problems with sharing and modification of the unconfirmed ct entry. The fixes tag points to the introduction of helpers, since that's the main UAF trigger for the sharing. Fixes: cae3a2627520 ("openvswitch: Allow attaching helpers to ct action") Cc: stable@vger.kernel.org Reported-by: Axel Mierczuk Signed-off-by: Ilya Maximets Reviewed-by: Aaron Conole Link: https://patch.msgid.link/20260921145655.3167436-2-i.maximets@ovn.org Signed-off-by: Jakub Kicinski --- include/net/netfilter/nf_conntrack.h | 5 +++++ net/openvswitch/conntrack.c | 12 ++++++++++++ 2 files changed, 17 insertions(+) diff --git a/include/net/netfilter/nf_conntrack.h b/include/net/netfilter/nf_conntrack.h index bc42dd0e10e6..c39425e54d87 100644 --- a/include/net/netfilter/nf_conntrack.h +++ b/include/net/netfilter/nf_conntrack.h @@ -185,6 +185,11 @@ static inline void nf_ct_put(struct nf_conn *ct) nf_ct_destroy(&ct->ct_general); } +static inline bool nf_ct_shared(const struct nf_conn *ct) +{ + return refcount_read(&ct->ct_general.use) > 1; +} + /* load module; enable/disable conntrack in this namespace */ int nf_ct_netns_get(struct net *net, u8 nfproto); void nf_ct_netns_put(struct net *net, u8 nfproto); diff --git a/net/openvswitch/conntrack.c b/net/openvswitch/conntrack.c index 0f433688e17b..a733029c28dd 100644 --- a/net/openvswitch/conntrack.c +++ b/net/openvswitch/conntrack.c @@ -734,6 +734,18 @@ static int __ovs_ct_lookup(struct net *net, struct sw_flow_key *key, enum ip_conntrack_info ctinfo; struct nf_conn *ct; + /* If the ct entry is not confirmed and shared with some other skb, + * e.g., a cloned one, we can't just modify it with the commit as we + * must not modify the extension set. Reset. + */ + if (cached && info->commit) { + ct = nf_ct_get(skb, &ctinfo); + if (ct && !nf_ct_is_confirmed(ct) && nf_ct_shared(ct)) { + nf_reset_ct(skb); + cached = false; + } + } + if (!cached) { struct nf_hook_state state = { .hook = NF_INET_PRE_ROUTING, From 5e6c14dd42a1c1fe938e573dc6c9098145b2b0c4 Mon Sep 17 00:00:00 2001 From: Ilya Maximets Date: Mon, 21 Sep 2026 16:55:44 +0200 Subject: [PATCH 569/592] net: openvswitch: conntrack: remove 'add_helper' dead code This variable can only become 'true' when the connection is not confirmed, but it is only checked when it is confirmed. So, it can be treated as being always false and just removed. Fixes: 3c1860543fcc ("openvswitch: add nf_ct_is_confirmed check before assigning the helper") Cc: stable@vger.kernel.org Signed-off-by: Ilya Maximets Reviewed-by: Aaron Conole Link: https://patch.msgid.link/20260921145655.3167436-3-i.maximets@ovn.org Signed-off-by: Jakub Kicinski --- net/openvswitch/conntrack.c | 10 ++-------- 1 file changed, 2 insertions(+), 8 deletions(-) diff --git a/net/openvswitch/conntrack.c b/net/openvswitch/conntrack.c index a733029c28dd..c20f096eef40 100644 --- a/net/openvswitch/conntrack.c +++ b/net/openvswitch/conntrack.c @@ -778,8 +778,6 @@ static int __ovs_ct_lookup(struct net *net, struct sw_flow_key *key, ct = nf_ct_get(skb, &ctinfo); if (ct) { - bool add_helper = false; - /* Packets starting a new connection must be NATted before the * helper, so that the helper knows about the NAT. We enforce * this by delaying both NAT and helper calls for unconfirmed @@ -811,7 +809,6 @@ static int __ovs_ct_lookup(struct net *net, struct sw_flow_key *key, GFP_ATOMIC); if (err) return err; - add_helper = true; /* helper installed, add seqadj if NAT is required */ if (info->nat && !nfct_seqadj(ct)) { @@ -821,13 +818,10 @@ static int __ovs_ct_lookup(struct net *net, struct sw_flow_key *key, } /* Call the helper only if: - * - nf_conntrack_in() was executed above ("!cached") or a - * helper was just attached ("add_helper") for a confirmed - * connection, or + * - nf_conntrack_in() was executed above ("!cached"), or * - When committing an unconfirmed connection. */ - if ((nf_ct_is_confirmed(ct) ? !cached || add_helper : - info->commit)) { + if ((nf_ct_is_confirmed(ct) ? !cached : info->commit)) { int err = nf_ct_helper(skb, ct, ctinfo, info->family); err = verdict_to_errno(err); From 1a4151e6be57b098b7a5ebfbde58585e83200cdc Mon Sep 17 00:00:00 2001 From: Ilya Maximets Date: Mon, 21 Sep 2026 16:55:45 +0200 Subject: [PATCH 570/592] net: openvswitch: conntrack: fix helper UAF due to extensions realloc While calling the helpers, a raw pointer to the extensions area is wired into expectations list: -> nf_ct_helper() -> helper->help() -> nf_ct_expect_related_report() -> nf_ct_expect_insert() -> hlist_add_head_rcu(&exp->lnode, &master_help->expectations) In case the connection is not confirmed yet, more extensions can be added afterwards with *_ext_add() calls reallocating the extension space and leaving the now invalid pointer in the expectations list that is later accessed while removing the expectation. Make sure that helpers are called at the end after all the other extensions are already added. Note that the helper rejection now leaves the mark and labels set, but that's not different from how the NAT was handled before or how the mark and the labels were handled on confirmation failure. And there are no atomicity guarantees provided by the API anyway. Fixes: cae3a2627520 ("openvswitch: Allow attaching helpers to ct action") Cc: stable@vger.kernel.org Reported-by: Axel Mierczuk Signed-off-by: Ilya Maximets Reviewed-by: Aaron Conole Link: https://patch.msgid.link/20260921145655.3167436-4-i.maximets@ovn.org Signed-off-by: Jakub Kicinski --- net/openvswitch/conntrack.c | 19 +++++++++++++++---- 1 file changed, 15 insertions(+), 4 deletions(-) diff --git a/net/openvswitch/conntrack.c b/net/openvswitch/conntrack.c index c20f096eef40..d3326edcabf7 100644 --- a/net/openvswitch/conntrack.c +++ b/net/openvswitch/conntrack.c @@ -817,11 +817,14 @@ static int __ovs_ct_lookup(struct net *net, struct sw_flow_key *key, } } - /* Call the helper only if: - * - nf_conntrack_in() was executed above ("!cached"), or - * - When committing an unconfirmed connection. + /* Call the helper only if nf_conntrack_in() was executed + * above ("!cached"). + * + * For unconfirmed connections it will be called later during + * commit as we need to have all the other extensions allocated + * before the call. */ - if ((nf_ct_is_confirmed(ct) ? !cached : info->commit)) { + if (nf_ct_is_confirmed(ct) && !cached) { int err = nf_ct_helper(skb, ct, ctinfo, info->family); err = verdict_to_errno(err); @@ -1025,6 +1028,14 @@ static int ovs_ct_commit(struct net *net, struct sw_flow_key *key, return err; nf_conn_act_ct_ext_add(skb, ct, ctinfo); + + /* Call the helpers now. We couldn't do this before as + * all the extensions must be allocated before the call. + */ + err = nf_ct_helper(skb, ct, ctinfo, info->family); + err = verdict_to_errno(err); + if (err) + return err; } else if (IS_ENABLED(CONFIG_NF_CONNTRACK_LABELS) && labels_nonzero(&info->labels.mask)) { err = ovs_ct_set_labels(ct, key, &info->labels.value, From f85009dfcd65e5969526b0db7a49b5413746e630 Mon Sep 17 00:00:00 2001 From: Ilya Maximets Date: Mon, 21 Sep 2026 16:55:46 +0200 Subject: [PATCH 571/592] net/sched: act_ct: avoid modifying shared unconfirmed ct entry In a case where skb with an unconfirmed ct entry gets cloned, we may end up processing both again but with different sets of extensions. The series of events: 1. The first clone wants to commit and runs the helpers wiring up the extension pointer into the expectation list. 2. Then it looses the confirmation keeping the entry unconfirmed. 3. Second clone now wants to commit labels or run NAT and adds the new extension for that breaking the pointer in the expectation list causing UAF on the destruction path later. While this is possible to trigger, there should be no practical network pipeline where we need to process both clones without modifications in the same zone. So, let's just reset the entry in case for some reason we got an skb with a shared one. This doesn't affect any known use cases, but avoids any potential problems with sharing and modification of the unconfirmed ct entry. Unlike openvswitch module, act_ct allows for NAT without commit. Changing that would be a uAPI break. So, act_ct needs to reset on NAT regardless of the commit flag to avoid reallocation of the extension space. This, however, doesn't really change the picture for sensible networking cases as there should be no need to run the same packet twice (before and after the clone) through conntrack without packet header or zone changes and without commit. The fixes tag points to the introduction of helpers, since that's the main UAF trigger for the sharing. Fixes: a21b06e73191 ("net: sched: add helper support in act_ct") Cc: stable@vger.kernel.org Reported-by: Axel Mierczuk Signed-off-by: Ilya Maximets Reviewed-by: Aaron Conole Reviewed-by: Xin Long Reviewed-by: Jamal Hadi Salim Link: https://patch.msgid.link/20260921145655.3167436-5-i.maximets@ovn.org Signed-off-by: Jakub Kicinski --- net/sched/act_ct.c | 18 ++++++++++++++++-- 1 file changed, 16 insertions(+), 2 deletions(-) diff --git a/net/sched/act_ct.c b/net/sched/act_ct.c index 55f3521edb4c..e72143d36b11 100644 --- a/net/sched/act_ct.c +++ b/net/sched/act_ct.c @@ -979,11 +979,11 @@ TC_INDIRECT_SCOPE int tcf_ct_act(struct sk_buff *skb, const struct tc_action *a, struct tcf_result *res) { struct net *net = dev_net(skb->dev); + bool cached, commit, clear, nat; enum ip_conntrack_info ctinfo; struct tcf_ct *c = to_ct(a); struct nf_conn *tmpl = NULL; struct nf_hook_state state; - bool cached, commit, clear; int nh_ofs, err, retval; struct tcf_ct_params *p; bool add_helper = false; @@ -998,6 +998,7 @@ TC_INDIRECT_SCOPE int tcf_ct_act(struct sk_buff *skb, const struct tc_action *a, retval = p->action; commit = p->ct_action & TCA_CT_ACT_COMMIT; clear = p->ct_action & TCA_CT_ACT_CLEAR; + nat = p->ct_action & TCA_CT_ACT_NAT; tmpl = p->tmpl; tcf_lastuse_update(&c->tcf_tm); @@ -1046,6 +1047,19 @@ TC_INDIRECT_SCOPE int tcf_ct_act(struct sk_buff *skb, const struct tc_action *a, * different zone. */ cached = tcf_ct_skb_nfct_cached(net, skb, p); + + /* If the ct entry is not confirmed and shared with some other skb, + * e.g., a cloned one, we can't just modify it with a commit or nat + * as we must not modify the extension set. Reset. + */ + if (cached && (commit || nat)) { + ct = nf_ct_get(skb, &ctinfo); + if (ct && !nf_ct_is_confirmed(ct) && nf_ct_shared(ct)) { + nf_reset_ct(skb); + cached = false; + } + } + if (!cached) { if (tcf_ct_flow_table_lookup(p, skb, family)) { skip_add = true; @@ -1083,7 +1097,7 @@ TC_INDIRECT_SCOPE int tcf_ct_act(struct sk_buff *skb, const struct tc_action *a, if (err) goto drop; add_helper = true; - if (p->ct_action & TCA_CT_ACT_NAT && !nfct_seqadj(ct)) { + if (nat && !nfct_seqadj(ct)) { if (!nfct_seqadj_ext_add(ct)) goto drop; } From 00df72e39f306e2f7adb68528a5c92109da6a0a9 Mon Sep 17 00:00:00 2001 From: Ilya Maximets Date: Mon, 21 Sep 2026 16:55:47 +0200 Subject: [PATCH 572/592] net/sched: act_ct: remove 'add_helper' dead code This variable can only become 'true' when the connection is not confirmed, but it is only checked when it is confirmed. So, it can be treated as being always false and just removed. Fixes: a21b06e73191 ("net: sched: add helper support in act_ct") Cc: stable@vger.kernel.org Signed-off-by: Ilya Maximets Reviewed-by: Aaron Conole Reviewed-by: Xin Long Reviewed-by: Jamal Hadi Salim Link: https://patch.msgid.link/20260921145655.3167436-6-i.maximets@ovn.org Signed-off-by: Jakub Kicinski --- net/sched/act_ct.c | 5 ++--- 1 file changed, 2 insertions(+), 3 deletions(-) diff --git a/net/sched/act_ct.c b/net/sched/act_ct.c index e72143d36b11..f62051ec9d57 100644 --- a/net/sched/act_ct.c +++ b/net/sched/act_ct.c @@ -986,7 +986,6 @@ TC_INDIRECT_SCOPE int tcf_ct_act(struct sk_buff *skb, const struct tc_action *a, struct nf_hook_state state; int nh_ofs, err, retval; struct tcf_ct_params *p; - bool add_helper = false; bool skb_is_ours = false; bool skip_add = false; bool defrag = false; @@ -1096,14 +1095,14 @@ TC_INDIRECT_SCOPE int tcf_ct_act(struct sk_buff *skb, const struct tc_action *a, err = __nf_ct_try_assign_helper(ct, p->tmpl, GFP_ATOMIC); if (err) goto drop; - add_helper = true; + if (nat && !nfct_seqadj(ct)) { if (!nfct_seqadj_ext_add(ct)) goto drop; } } - if (nf_ct_is_confirmed(ct) ? ((!cached && !skip_add) || add_helper) : commit) { + if (nf_ct_is_confirmed(ct) ? (!cached && !skip_add) : commit) { err = nf_ct_helper(skb, ct, ctinfo, family); if (err != NF_ACCEPT) goto nf_error; From dad19b59da050cb60d3f7023dac2a042a84bf0bd Mon Sep 17 00:00:00 2001 From: Ilya Maximets Date: Mon, 21 Sep 2026 16:55:48 +0200 Subject: [PATCH 573/592] net/sched: act_ct: fix helper UAF due to extensions realloc While calling the helpers, a raw pointer to the extensions area is wired into expectations list: -> nf_ct_helper() -> helper->help() -> nf_ct_expect_related_report() -> nf_ct_expect_insert() -> hlist_add_head_rcu(&exp->lnode, &master_help->expectations) In case the connection is not confirmed yet, more extensions can be added afterwards with *_ext_add() calls reallocating the extension space and leaving the now invalid pointer in the expectations list that is later accessed while removing the expectation. Make sure that helpers are called at the end after all the other extensions are already added. Note that the helper rejection now leaves the mark and labels set, but that's not different from how the NAT was handled before or how the mark and the labels were handled on confirmation failure. And there are no atomicity guarantees provided by the API anyway. Fixes: a21b06e73191 ("net: sched: add helper support in act_ct") Cc: stable@vger.kernel.org Reported-by: Axel Mierczuk Signed-off-by: Ilya Maximets Reviewed-by: Xin Long Reviewed-by: Jamal Hadi Salim Reviewed-by: Aaron Conole Link: https://patch.msgid.link/20260921145655.3167436-7-i.maximets@ovn.org Signed-off-by: Jakub Kicinski --- net/sched/act_ct.c | 18 ++++++++++++------ 1 file changed, 12 insertions(+), 6 deletions(-) diff --git a/net/sched/act_ct.c b/net/sched/act_ct.c index f62051ec9d57..411e3dd92d07 100644 --- a/net/sched/act_ct.c +++ b/net/sched/act_ct.c @@ -1102,6 +1102,18 @@ TC_INDIRECT_SCOPE int tcf_ct_act(struct sk_buff *skb, const struct tc_action *a, } } + if (commit) { + tcf_ct_act_set_mark(ct, p->mark, p->mark_mask); + tcf_ct_act_set_labels(ct, p->labels, p->labels_mask); + + if (!nf_ct_is_confirmed(ct)) + nf_conn_act_ct_ext_add(skb, ct, ctinfo); + } + + /* Run helpers for the connection if nf_conntrack_in() was executed + * or if we're about to commit. This has to be done after all the + * extensions are already added. + */ if (nf_ct_is_confirmed(ct) ? (!cached && !skip_add) : commit) { err = nf_ct_helper(skb, ct, ctinfo, family); if (err != NF_ACCEPT) @@ -1109,12 +1121,6 @@ TC_INDIRECT_SCOPE int tcf_ct_act(struct sk_buff *skb, const struct tc_action *a, } if (commit) { - tcf_ct_act_set_mark(ct, p->mark, p->mark_mask); - tcf_ct_act_set_labels(ct, p->labels, p->labels_mask); - - if (!nf_ct_is_confirmed(ct)) - nf_conn_act_ct_ext_add(skb, ct, ctinfo); - /* This will take care of sending queued events * even if the connection is already confirmed. */ From 0958ea4355e2e9220ad4e13da3b7d94f365ed34f Mon Sep 17 00:00:00 2001 From: Guangshuo Li Date: Mon, 21 Sep 2026 23:42:01 +0800 Subject: [PATCH 574/592] net: ena: fix PHC cleanup on probe failure ena_probe() initializes the PHC as part of ena_device_init(), but the probe failure path does not destroy it before freeing the PHC private data. The normal removal path calls ena_phc_destroy() through ena_destroy_device() before ena_phc_free(). However, if probe fails after ena_device_init() succeeds, the error path reaches ena_phc_free() without unregistering the PTP clock or destroying the device PHC resources. Call ena_phc_destroy() in the probe error path before freeing the PHC private data. This issue was found by manual code inspection. Cc: stable@vger.kernel.org tags and describe this as a consistency cleanup Fixes: e0ea34158ee8 ("net: ena: Add PHC support in the ENA driver") Cc: stable@vger.kernel.org Signed-off-by: Guangshuo Li Cc: stable Link: https://patch.msgid.link/20260921154202.471662-2-lgs201920130244@gmail.com Signed-off-by: Jakub Kicinski --- drivers/net/ethernet/amazon/ena/ena_netdev.c | 1 + 1 file changed, 1 insertion(+) diff --git a/drivers/net/ethernet/amazon/ena/ena_netdev.c b/drivers/net/ethernet/amazon/ena/ena_netdev.c index ea89619039d8..5f0864d16dd3 100644 --- a/drivers/net/ethernet/amazon/ena/ena_netdev.c +++ b/drivers/net/ethernet/amazon/ena/ena_netdev.c @@ -4122,6 +4122,7 @@ static int ena_probe(struct pci_dev *pdev, const struct pci_device_id *ent) err_device_destroy: ena_com_delete_host_info(ena_dev); ena_com_admin_destroy(ena_dev); + ena_phc_destroy(adapter); ena_devlink_destroy: ena_devlink_free(devlink); err_metrics_destroy: From 9476b4468862927297c94c440863cd8ed1e7cc83 Mon Sep 17 00:00:00 2001 From: Guangshuo Li Date: Mon, 21 Sep 2026 23:42:02 +0800 Subject: [PATCH 575/592] net: ena: fix MMIO read buffer leak on probe failure ena_device_init() initializes the MMIO read mechanism with ena_com_mmio_reg_read_request_init(), which allocates a coherent DMA buffer for MMIO read responses. The normal removal path releases this buffer through ena_com_mmio_reg_read_request_destroy(). However, if ena_probe() fails after ena_device_init() succeeds, the error path destroys the admin resources and eventually frees ena_dev without destroying the MMIO read request, leaving the coherent DMA buffer allocated. Call ena_com_mmio_reg_read_request_destroy() in the probe error path before releasing the remaining device resources. This issue was found by manual code inspection. Fixes: 1738cd3ed342 ("net: ena: Add a driver for Amazon Elastic Network Adapters (ENA)") Cc: stable@vger.kernel.org Signed-off-by: Guangshuo Li Link: https://patch.msgid.link/20260921154202.471662-3-lgs201920130244@gmail.com Signed-off-by: Jakub Kicinski --- drivers/net/ethernet/amazon/ena/ena_netdev.c | 1 + 1 file changed, 1 insertion(+) diff --git a/drivers/net/ethernet/amazon/ena/ena_netdev.c b/drivers/net/ethernet/amazon/ena/ena_netdev.c index 5f0864d16dd3..7eb6456ed0d5 100644 --- a/drivers/net/ethernet/amazon/ena/ena_netdev.c +++ b/drivers/net/ethernet/amazon/ena/ena_netdev.c @@ -4123,6 +4123,7 @@ static int ena_probe(struct pci_dev *pdev, const struct pci_device_id *ent) ena_com_delete_host_info(ena_dev); ena_com_admin_destroy(ena_dev); ena_phc_destroy(adapter); + ena_com_mmio_reg_read_request_destroy(ena_dev); ena_devlink_destroy: ena_devlink_free(devlink); err_metrics_destroy: From 1a983a4e14c635c40354be110cd9a1a5c94e01e6 Mon Sep 17 00:00:00 2001 From: Sang-Hoon Choi Date: Tue, 22 Sep 2026 03:15:59 +0900 Subject: [PATCH 576/592] nfp: hold IPsec RX state under the XArray lock nfp_net_ipsec_rx() drops the XArray lock before taking a reference to the xfrm_state it found. The delete path can erase the entry and drop the last state reference in that interval. RX can then try to increment a zero refcount after the state has been queued for destruction. The driver queues firmware invalidation asynchronously; the delete path does not wait for the command to complete or drain pending RX processing. The XFRM garbage collector waits for an RCU grace period before freeing the state. That delays reclamation but does not make acquiring a reference from zero valid. Take the xfrm_state reference before releasing the XArray lock so xa_erase() cannot run between lookup and reference acquisition. Fixes: 57f273adbcd4 ("nfp: add framework to support ipsec offloading") Reported-by: Changyul Lee Signed-off-by: Sang-Hoon Choi Link: https://patch.msgid.link/179001455912.44752.17153022439349797877.idr-bug-92@gmail.com Signed-off-by: Jakub Kicinski --- drivers/net/ethernet/netronome/nfp/crypto/ipsec.c | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-) diff --git a/drivers/net/ethernet/netronome/nfp/crypto/ipsec.c b/drivers/net/ethernet/netronome/nfp/crypto/ipsec.c index 9e7c285eaa6b..960d7513aa8d 100644 --- a/drivers/net/ethernet/netronome/nfp/crypto/ipsec.c +++ b/drivers/net/ethernet/netronome/nfp/crypto/ipsec.c @@ -625,11 +625,12 @@ int nfp_net_ipsec_rx(struct nfp_meta_parsed *meta, struct sk_buff *skb) xa_lock(&nn->xa_ipsec); x = xa_load(&nn->xa_ipsec, saidx); + if (x) + xfrm_state_hold(x); xa_unlock(&nn->xa_ipsec); if (!x) return -EINVAL; - xfrm_state_hold(x); sp->xvec[sp->len++] = x; sp->olen++; xo = xfrm_offload(skb); From ba3d1f480c7a3fba963e7867ad6cc557c197acbc Mon Sep 17 00:00:00 2001 From: Allison Henderson Date: Mon, 21 Sep 2026 14:50:27 -0700 Subject: [PATCH 577/592] net/rds: size a connection's path set by the transport it ends up with __rds_conn_create() computes npaths from the caller's transport before it decides whether a connection to one of the host's own addresses is to be handled by the loopback transport instead. That substitution is what an RDS/TCP socket sending to a local address gets, and after it the path init loop still runs for the TCP transport's RDS_MPATH_WORKERS paths and allocates an ordered workqueue for each, while rds_loop_conn_alloc() only ever provides transport data for path 0. rds_conn_destroy() sizes its teardown from c_trans, by then the loopback transport, so it visits path 0 only - and rds_conn_path_destroy() would skip the other paths anyway, since it returns before destroy_workqueue() for a path without transport data. kfree(c_path) then drops the last pointers to seven workqueues. That repeats for every such connection, on every netns teardown or module unload, and every distinct local destination address is a separate connection. Recompute npaths once the transport is final, so that creation and destruction agree on the set of paths. The c_path array stays sized for the caller's transport; the unused entries are freed with it. Fixes: 4716af3897e9 ("net/rds: Give each connection path its own workqueue") Signed-off-by: Allison Henderson Link: https://patch.msgid.link/20260921215027.174657-1-achender@kernel.org Signed-off-by: Jakub Kicinski --- net/rds/connection.c | 6 ++++++ 1 file changed, 6 insertions(+) diff --git a/net/rds/connection.c b/net/rds/connection.c index b6c4beb50eaf..c752a8623cfc 100644 --- a/net/rds/connection.c +++ b/net/rds/connection.c @@ -276,6 +276,12 @@ static struct rds_connection *__rds_conn_create(struct net *net, conn->c_trans = trans; + /* The transport may just have been swapped for loopback; size the + * set of paths - which is also what rds_conn_destroy() tears down + * again - by the transport the connection actually uses. + */ + npaths = (trans->t_mp_capable ? RDS_MPATH_WORKERS : 1); + init_waitqueue_head(&conn->c_hs_waitq); for (i = 0; i < npaths; i++) { __rds_conn_path_init(conn, &conn->c_path[i], From 61cb282fe97b3b0ba32ca09417a693162bf4ae3f Mon Sep 17 00:00:00 2001 From: Sidraya Jayagond Date: Tue, 22 Sep 2026 09:31:49 +0200 Subject: [PATCH 578/592] net/smc: fix UAF on lgr list traversal in smcr_port_err() smcr_port_err() traverses smc_lgr_list.list without holding smc_lgr_list.lock, allowing a concurrent smc_lgr_terminate_sched() to free an lgr while it is still being dereferenced. Hold smc_lgr_list.lock across the traversal. Update smc_ib_gid_check() to call smcr_port_err() after releasing the lock. Fixes: 541afa10c126 ("net/smc: add smcr_port_err() and smcr_link_down() processing") Reviewed-by: Mahanta Jambigi Signed-off-by: Sidraya Jayagond Reviewed-by: Dust Li Link: https://patch.msgid.link/20260922073149.474762-1-sidraya@linux.ibm.com Signed-off-by: Jakub Kicinski --- net/smc/smc_core.c | 2 ++ net/smc/smc_ib.c | 10 ++++++++-- 2 files changed, 10 insertions(+), 2 deletions(-) diff --git a/net/smc/smc_core.c b/net/smc/smc_core.c index 04aedd957543..9974149659c2 100644 --- a/net/smc/smc_core.c +++ b/net/smc/smc_core.c @@ -1849,6 +1849,7 @@ void smcr_port_err(struct smc_ib_device *smcibdev, u8 ibport) struct smc_link_group *lgr, *n; int i; + spin_lock_bh(&smc_lgr_list.lock); list_for_each_entry_safe(lgr, n, &smc_lgr_list.list, list) { if (strncmp(smcibdev->pnetid[ibport - 1], lgr->pnet_id, SMC_MAX_PNETID_LEN)) @@ -1863,6 +1864,7 @@ void smcr_port_err(struct smc_ib_device *smcibdev, u8 ibport) smcr_link_down_cond_sched(lnk); } } + spin_unlock_bh(&smc_lgr_list.lock); } static void smc_link_down_work(struct work_struct *work) diff --git a/net/smc/smc_ib.c b/net/smc/smc_ib.c index 9bb495707445..daaa8a72da90 100644 --- a/net/smc/smc_ib.c +++ b/net/smc/smc_ib.c @@ -333,6 +333,7 @@ static bool smc_ib_check_link_gid(u8 gid[SMC_GID_SIZE], bool smcrv2, static void smc_ib_gid_check(struct smc_ib_device *smcibdev, u8 ibport) { struct smc_link_group *lgr; + bool stale_gid = false; int i; spin_lock_bh(&smc_lgr_list.lock); @@ -348,11 +349,16 @@ static void smc_ib_gid_check(struct smc_ib_device *smcibdev, u8 ibport) continue; if (!smc_ib_check_link_gid(lgr->lnk[i].gid, lgr->smc_version == SMC_V2, - smcibdev, ibport)) - smcr_port_err(smcibdev, ibport); + smcibdev, ibport)) { + stale_gid = true; + goto out; + } } } +out: spin_unlock_bh(&smc_lgr_list.lock); + if (stale_gid) + smcr_port_err(smcibdev, ibport); } static int smc_ib_remember_port_attr(struct smc_ib_device *smcibdev, u8 ibport) From b94773dc4df7026a6f29b2c65e96e88d29cdb576 Mon Sep 17 00:00:00 2001 From: Alexander Sverdlin Date: Tue, 22 Sep 2026 09:52:46 +0200 Subject: [PATCH 579/592] net: phy: intel-xway: workaround 100BASE-TX Link-Up issue MaxLinear GSW12x/GSW14x Ethernet Switch Errata Sheet states: "An issue has been sporadically observed after device power-on on the first link-up attempt in 100BASE-TX mode resulting in either the link-up taking a long time, or failing to link-up altogether... Workaround: After power-on, enable Cable Diagnostic Mode for all ports and disable it..." Implement the proposed workaround unconditionally in the Intel XWAY driver (MaxLinear GSW1xx switches incorporate Intel XWAY PHYs) because the diagnostic bits have the same meaning even in older integral PHYs such as GPY111/PEF7071/PHY11G. So it's not clear how to distinguish the affected newer integrated PHYs, but the workaround should not hurt the older PHYs. Cc: stable@vger.kernel.org Fixes: 22335939ec90 ("net: dsa: add driver for MaxLinear GSW1xx switch family") Signed-off-by: Alexander Sverdlin Reviewed-by: Andrew Lunn Link: https://patch.msgid.link/20260922075251.23386-1-alexander.sverdlin@siemens.com Signed-off-by: Jakub Kicinski --- drivers/net/phy/intel-xway.c | 29 ++++++++++++++++++++++++++++- 1 file changed, 28 insertions(+), 1 deletion(-) diff --git a/drivers/net/phy/intel-xway.c b/drivers/net/phy/intel-xway.c index afbcec711744..3cee31bb931f 100644 --- a/drivers/net/phy/intel-xway.c +++ b/drivers/net/phy/intel-xway.c @@ -16,6 +16,11 @@ #define XWAY_MDIO_ISTAT 0x1A /* interrupt status */ #define XWAY_MDIO_LED 0x1B /* led control */ +#define XWAY_MDIO_GCTRL_TM_MASK GENMASK(15, 13) +#define XWAY_MDIO_GCTRL_TM(mode) FIELD_PREP(XWAY_MDIO_GCTRL_TM_MASK, (mode)) +#define XWAY_MDIO_GCTRL_TM_NOP XWAY_MDIO_GCTRL_TM(0) /* Normal operation */ +#define XWAY_MDIO_GCTRL_TM_CDIAG XWAY_MDIO_GCTRL_TM(6) /* Cable diagnostics */ + #define XWAY_MDIO_ERRCNT_SEL GENMASK(11, 8) #define XWAY_MDIO_ERRCNT_COUNT GENMASK(7, 0) #define XWAY_MDIO_ERRCNT_SEL_RXERR 0 @@ -326,6 +331,28 @@ static int xway_gphy_probe(struct phy_device *phydev) return 0; } +static int xway_11g_int_config_init(struct phy_device *phydev) +{ + int err; + + /* An issue has been sporadically observed after device power-on on the + * first link-up attempt in 100BASE-TX mode resulting in either the + * link-up taking a long time, or failing to link-up altogether. + * + * Workaround: + * After power-on, enable Cable Diagnostic Mode for all ports and + * disable it. + */ + err = phy_modify(phydev, MII_CTRL1000, XWAY_MDIO_GCTRL_TM_MASK, XWAY_MDIO_GCTRL_TM_CDIAG); + if (err) + return err; + err = phy_modify(phydev, MII_CTRL1000, XWAY_MDIO_GCTRL_TM_MASK, XWAY_MDIO_GCTRL_TM_NOP); + if (err) + return err; + + return xway_gphy_config_init(phydev); +} + static int xway_gphy14_config_aneg(struct phy_device *phydev) { int reg, err; @@ -735,7 +762,7 @@ static struct phy_driver xway_gphy[] = { .phy_id_mask = 0xffffffff, .name = "Intel XWAY PHY11G (xRX v1.2 integrated)", /* PHY_GBIT_FEATURES */ - .config_init = xway_gphy_config_init, + .config_init = xway_11g_int_config_init, .probe = xway_gphy_probe, .handle_interrupt = xway_gphy_handle_interrupt, .config_intr = xway_gphy_config_intr, From 8e1937fed6738460554ec123c64839e2445e7d53 Mon Sep 17 00:00:00 2001 From: Ginger Li Date: Tue, 22 Sep 2026 16:09:09 +0800 Subject: [PATCH 580/592] tipc: Fix a data race on mon->peer_cnt in mon_timeout() mon_timeout() evaluates dom_size(mon->peer_cnt) before it takes mon->lock, while mon->peer_cnt is updated under that lock by tipc_mon_add_peer() and tipc_mon_remove_peer(). The value can therefore be stale, and the decision whether the local domain has to be recomputed can be based on an outdated member count. Read mon->peer_cnt inside the write_lock_bh(&mon->lock) protected region. Fixes: 35c55c9877f8 ("tipc: add neighbor monitoring framework") Signed-off-by: Ginger Li Reviewed-by: Tung Nguyen Link: https://patch.msgid.link/20260922080909.21123-1-ginger.jzllee@gmail.com Signed-off-by: Jakub Kicinski --- net/tipc/monitor.c | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-) diff --git a/net/tipc/monitor.c b/net/tipc/monitor.c index a94b9b36a700..1a438e312d66 100644 --- a/net/tipc/monitor.c +++ b/net/tipc/monitor.c @@ -632,9 +632,10 @@ static void mon_timeout(struct timer_list *t) { struct tipc_monitor *mon = timer_container_of(mon, t, timer); struct tipc_peer *self; - int best_member_cnt = dom_size(mon->peer_cnt) - 1; + int best_member_cnt; write_lock_bh(&mon->lock); + best_member_cnt = dom_size(mon->peer_cnt) - 1; self = mon->self; if (self && (best_member_cnt != self->applied)) { mon_update_local_domain(mon); From 56d82862a0a243ac14ba11b6d7b57ddc2d064b95 Mon Sep 17 00:00:00 2001 From: Dairui Zhang Date: Wed, 23 Sep 2026 13:01:01 +0800 Subject: [PATCH 581/592] af_packet: fix integer overflow in prb_calc_retire_blk_tmo() prb_calc_retire_blk_tmo() computes in 32-bit int arithmetic: mbits = (blk_size_in_bytes * 8) / (1024 * 1024); If I'm reading the validation right, tp_block_size is user controlled and packet_set_ring() only rejects values that are <= 0 as int or not page aligned, so a 256MiB block goes right through (and alloc_one_pg_vec_page() even has a vzalloc fallback for it). 0x10000000 * 8 wraps to INT_MIN, and on a NIC reporting 1 Gbps (div == 1) the function ends up returning -2047. The condition is actually (8 * size) mod 2^32 >= 2^31 && div == 1, so the trigger set is [256,512), [768,1024), [1280,1536) and [1792,2048) MiB. Other sizes wrap to non-negative values and faster links divide the unsigned value back below 2^31, which is why this doesn't blow up for everyone. What makes it fatal is what happens next in init_prb_bdqc(): p1->interval_ktime = ms_to_ktime(prb_calc_retire_blk_tmo(...)); hrtimer_start(&p1->retire_blk_timer, p1->interval_ktime, HRTIMER_MODE_REL_SOFT); A negative relative timeout expires immediately. The callback unconditionally returns HRTIMER_RESTART, and hrtimer_forward() turns the negative interval into hrtimer_resolution: if (interval < hrtimer_resolution) interval = hrtimer_resolution; So the SOFT timer re-fires at the maximum rate forever, holding sk_receive_queue.lock each pass. One CPU spins in softirq until the socket is closed. Repeat with more rings and the machine is gone. The overflow itself is ancient - it was introduced together with TPACKET_V3 in f6fb8f100b80 ("af-packet: TPACKET_V3 flexible buffer implementation."). Its effect prior to f7460d2989fa ("net: af_packet: Use hrtimer to do the retire operation", v6.18) was not as clear-cut, though: the return value was stored into an unsigned short retire_blk_tov, so a negative result was truncated, and a 0-jiffy delay loop could be programmed as well. Neither is nearly as detrimental as the immediate maximum-rate spin the hrtimer conversion turned it into. (Unrelated to CVE-2019-20812 - that one was the ethtool failure path returning 0, which now returns DEFAULT_PRB_RETIRE_TOV.) Reproducer, needs CAP_NET_RAW (a --network host container has it by default) and a 1 Gbps NIC (QEMU e1000 works): int fd = socket(AF_PACKET, SOCK_RAW, htons(ETH_P_ALL)); bind(fd, ...); int v = TPACKET_V3; setsockopt(fd, SOL_PACKET, PACKET_VERSION, &v, sizeof(v)); struct tpacket_req3 req = { .tp_block_size = 0x10000000, .tp_block_nr = 1, .tp_frame_size = 2048, .tp_frame_nr = 0x10000000 / 2048, .tp_retire_blk_tov = 0, }; setsockopt(fd, SOL_PACKET, PACKET_RX_RING, &req, sizeof(req)); Compute in 64 bits instead. The operands are already bounded by the existing validation, so nothing else changes. If you'd prefer a different fix, just say so and I'll respin. Fixes: f6fb8f100b80 ("af-packet: TPACKET_V3 flexible buffer implementation.") Cc: stable@vger.kernel.org Signed-off-by: Dairui Zhang Reviewed-by: Willem de Bruijn Link: https://patch.msgid.link/20260923050101.1510064-1-zhangdairui@gmail.com Signed-off-by: Jakub Kicinski --- net/packet/af_packet.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/net/packet/af_packet.c b/net/packet/af_packet.c index 64b501db660a..7c83e01526ed 100644 --- a/net/packet/af_packet.c +++ b/net/packet/af_packet.c @@ -617,7 +617,7 @@ static int prb_calc_retire_blk_tmo(struct packet_sock *po, return DEFAULT_PRB_RETIRE_TOV; div = ecmd.base.speed / 1000; - mbits = (blk_size_in_bytes * 8) / (1024 * 1024); + mbits = (u64)blk_size_in_bytes * 8 / (1024 * 1024); if (div) mbits /= div; From 8db67bb6a1fffa4df68fbbc22e39943aeeff9178 Mon Sep 17 00:00:00 2001 From: Coia Prant Date: Wed, 23 Sep 2026 20:37:13 +0800 Subject: [PATCH 582/592] net: ethernet: stmmac: dwmac-rk: fix bulk clock leak when the PHY clock fails gmac_clk_enable() enables the bulk clocks first and then the optional PHY clock. If clk_prepare_enable() on the PHY clock fails, the function returns without rolling back the bulk clocks, and bsp_priv->clk_enabled stays false, so the later gmac_clk_enable(bsp_priv, false) becomes a no-op and the bulk clock references are leaked. Add the missing clk_bulk_disable_unprepare() on that failure path. Fixes: ea449f7fa0bf ("net: ethernet: stmmac: dwmac-rk: rework optional clock handling") Reviewed-by: Maxime Chevallier Reviewed-by: Heiko Stuebner Acked-by: Lorenzo Bianconi Signed-off-by: Coia Prant Link: https://patch.msgid.link/20260923123713.3137146-1-coiaprant@gmail.com Signed-off-by: Jakub Kicinski --- drivers/net/ethernet/stmicro/stmmac/dwmac-rk.c | 5 ++++- 1 file changed, 4 insertions(+), 1 deletion(-) diff --git a/drivers/net/ethernet/stmicro/stmmac/dwmac-rk.c b/drivers/net/ethernet/stmicro/stmmac/dwmac-rk.c index 8d7042e68926..72bdbcb5e863 100644 --- a/drivers/net/ethernet/stmicro/stmmac/dwmac-rk.c +++ b/drivers/net/ethernet/stmicro/stmmac/dwmac-rk.c @@ -1162,8 +1162,11 @@ static int gmac_clk_enable(struct rk_priv_data *bsp_priv, bool enable) return ret; ret = clk_prepare_enable(bsp_priv->clk_phy); - if (ret) + if (ret) { + clk_bulk_disable_unprepare(bsp_priv->num_clks, + bsp_priv->clks); return ret; + } rk_configure_io_clksel(bsp_priv); rk_ungate_rmii_clock(bsp_priv); From 06e3f54e8b22040ada01a28343badf1990b4dd7d Mon Sep 17 00:00:00 2001 From: Eric Dumazet Date: Wed, 23 Sep 2026 13:03:18 +0000 Subject: [PATCH 583/592] net: flush skb_defer_nodes in dev_cpu_dead() When a CPU goes offline, dev_cpu_dead() drains its softnet queues (completion_queue, output_queue, poll_list, process_queue, and input_pkt_queue), but leaves net_hotdata.skb_defer_nodes untouched. If oldcpu goes offline while holding pending skbs in its skb_defer_nodes lists (e.g. below the sysctl_skb_defer_max >> 1 IPI threshold, or if the IPI races with CPU teardown), those skbs remain stranded until oldcpu is brought back online. If any of these skbs hold page_pool fragments, page_pool_destroy() will stall indefinitely waiting for inflight pages to be returned when a netdev or driver is torn down while oldcpu is offline. Additionally, if smp_call_function_single_async() fails in kick_defer_list_purge() because the target CPU went offline, reset defer_ipi_scheduled to 0 so future IPI kicks are not blocked when the CPU comes back online. Also, if oldcpu was the last online CPU on its NUMA node, drain that node's slot across all CPUs so no skbs deferred from that node remain stranded on idle remote CPUs (or if the node itself is subsequently offlined). Finally, in skb_attempt_defer_free(), re-check cpu_online(cpu) and whether the caller migrated CPUs after llist_add(), flushing the node list if so, to close the preemption TOCTOU race against CPU/node teardown. Fixes: 68822bdf76f1 ("net: generalize skb freeing deferral to per-cpu lists") Fixes: 5628f3fe3b16 ("net: add NUMA awareness to skb_attempt_defer_free()") Closes: https://lore.kernel.org/netdev/20260916003430.3612956-1-kris.pan@intel.com/ Signed-off-by: Eric Dumazet Cc: Kris Pan Link: https://patch.msgid.link/20260923130318.607255-1-edumazet@google.com Signed-off-by: Jakub Kicinski --- net/core/dev.c | 47 +++++++++++++++++++++++++++++++++++------------ net/core/dev.h | 2 ++ net/core/skbuff.c | 12 +++++++++--- 3 files changed, 46 insertions(+), 15 deletions(-) diff --git a/net/core/dev.c b/net/core/dev.c index 0292a16e16c2..f660fccfc0db 100644 --- a/net/core/dev.c +++ b/net/core/dev.c @@ -5376,7 +5376,8 @@ void kick_defer_list_purge(unsigned int cpu) backlog_unlock_irq_restore(sd, flags); } else if (!cmpxchg(&sd->defer_ipi_scheduled, 0, 1)) { - smp_call_function_single_async(cpu, &sd->defer_csd); + if (smp_call_function_single_async(cpu, &sd->defer_csd)) + WRITE_ONCE(sd->defer_ipi_scheduled, 0); } } @@ -6900,25 +6901,35 @@ bool napi_complete_done(struct napi_struct *n, int work_done) } EXPORT_SYMBOL(napi_complete_done); -static void skb_defer_free_flush(void) +static void __skb_defer_free_flush(struct skb_defer_node *sdn, int budget) { struct llist_node *free_list; struct sk_buff *skb, *next; + + if (llist_empty(&sdn->defer_list)) + return; + atomic_long_set(&sdn->defer_count, 0); + free_list = llist_del_all(&sdn->defer_list); + + llist_for_each_entry_safe(skb, next, free_list, ll_node) { + prefetch(next); + napi_consume_skb(skb, budget); + } +} + +void skb_defer_node_flush(struct skb_defer_node *sdn) +{ + __skb_defer_free_flush(sdn, 0); +} + +static void skb_defer_free_flush(void) +{ struct skb_defer_node *sdn; int node; for_each_node(node) { sdn = this_cpu_ptr(net_hotdata.skb_defer_nodes) + node; - - if (llist_empty(&sdn->defer_list)) - continue; - atomic_long_set(&sdn->defer_count, 0); - free_list = llist_del_all(&sdn->defer_list); - - llist_for_each_entry_safe(skb, next, free_list, ll_node) { - prefetch(next); - napi_consume_skb(skb, 1); - } + __skb_defer_free_flush(sdn, 1); } } @@ -12897,6 +12908,7 @@ static int dev_cpu_dead(unsigned int oldcpu) struct sk_buff **list_skb; struct sk_buff *skb; unsigned int cpu; + int node; struct softnet_data *sd, *oldsd, *remsd = NULL; local_irq_disable(); @@ -12957,6 +12969,17 @@ static int dev_cpu_dead(unsigned int oldcpu) rps_input_queue_head_incr(oldsd); } + for_each_node(node) + skb_defer_node_flush(per_cpu_ptr(net_hotdata.skb_defer_nodes, + oldcpu) + node); + node = cpu_to_node(oldcpu); + if (node_possible(node) && + !cpumask_intersects(cpumask_of_node(node), cpu_online_mask)) { + for_each_possible_cpu(cpu) + skb_defer_node_flush(per_cpu_ptr(net_hotdata.skb_defer_nodes, + cpu) + node); + } + return 0; } diff --git a/net/core/dev.h b/net/core/dev.h index b757faead4d1..04fb0e9a571e 100644 --- a/net/core/dev.h +++ b/net/core/dev.h @@ -399,6 +399,8 @@ static inline void napi_assert_will_not_race(const struct napi_struct *napi) WARN_ON(READ_ONCE(napi->list_owner) != -1); } +struct skb_defer_node; +void skb_defer_node_flush(struct skb_defer_node *sdn); void kick_defer_list_purge(unsigned int cpu); int dev_set_hwtstamp_phylib(struct net_device *dev, diff --git a/net/core/skbuff.c b/net/core/skbuff.c index b4edbd06655e..c3042d822afa 100644 --- a/net/core/skbuff.c +++ b/net/core/skbuff.c @@ -7360,8 +7360,8 @@ void skb_attempt_defer_free(struct sk_buff *skb) struct skb_defer_node *sdn; unsigned long defer_count; unsigned int defer_max; + int cpu, my_cpu; bool kick; - int cpu; if (static_branch_unlikely(&skb_defer_disable_key)) goto nodefer; @@ -7371,7 +7371,8 @@ void skb_attempt_defer_free(struct sk_buff *skb) goto nodefer; cpu = skb->alloc_cpu; - if (cpu == raw_smp_processor_id() || + my_cpu = raw_smp_processor_id(); + if (cpu == my_cpu || WARN_ON_ONCE(cpu >= nr_cpu_ids) || !cpu_online(cpu)) { nodefer: kfree_skb_napi_cache(skb); @@ -7382,7 +7383,7 @@ nodefer: kfree_skb_napi_cache(skb); DEBUG_NET_WARN_ON_ONCE(skb->destructor); DEBUG_NET_WARN_ON_ONCE(skb_nfct(skb)); - sdn = per_cpu_ptr(net_hotdata.skb_defer_nodes, cpu) + numa_node_id(); + sdn = per_cpu_ptr(net_hotdata.skb_defer_nodes, cpu) + cpu_to_node(my_cpu); defer_max = READ_ONCE(net_hotdata.sysctl_skb_defer_max); defer_count = atomic_long_inc_return(&sdn->defer_count); @@ -7392,6 +7393,11 @@ nodefer: kfree_skb_napi_cache(skb); llist_add(&skb->ll_node, &sdn->defer_list); + if (unlikely(!cpu_online(cpu) || my_cpu != raw_smp_processor_id())) { + skb_defer_node_flush(sdn); + return; + } + /* Send an IPI every time queue reaches half capacity. */ kick = (defer_count - 1) == (defer_max >> 1); From 83769c23fb1879edc916a526ba424285033baf2d Mon Sep 17 00:00:00 2001 From: Eric Dumazet Date: Wed, 23 Sep 2026 14:59:42 +0000 Subject: [PATCH 584/592] gve: DQO: fix header length used by gve_can_send_tso() for UDP GSO gve_can_send_tso() computes how many buffers each segment of a GSO packet would span, and for this it needs the length of the headers that the device replicates in front of every segment. It unconditionally uses skb_tcp_all_headers(), which reads the doff field of the TCP header. SKB_GSO_UDP_L4 packets have no TCP header: tcp_hdrlen() then reads one byte of the UDP payload, and header_len can be anything in [0, 60] instead of the transport offset plus the eight bytes of the UDP header that gve_prep_tso() programs into the TSO context descriptor. A wrong header length shifts all the segment boundaries computed in the loop, so the number of buffers per segment can be over or under estimated. In the first case, GSO is needlessly disabled for this packet by gve_features_check_dqo() and the stack has to segment it. In the second case, the driver hands the device a packet whose segments span more than GVE_TX_MAX_DATA_DESCS buffers. Use the UDP header length for SKB_GSO_UDP_L4 packets, matching what gve_prep_tso() does. Fixes: 014c607f86ab ("gve: add support for UDP GSO for DQO format") Closes: https://lore.kernel.org/netdev/CANn89i+MS4L60sFQ49=-f-mibeveUfcrpVkD5X+Qy6SOnEpd6w@mail.gmail.com/ Signed-off-by: Eric Dumazet Cc: Ankit Garg Cc: Harshitha Ramamurthy Cc: Joshua Washington Cc: Willem de Bruijn Reviewed-by: Ankit Garg Reviewed-by: Harshitha Ramamurthy Link: https://patch.msgid.link/20260923145942.731365-1-edumazet@google.com Signed-off-by: Jakub Kicinski --- drivers/net/ethernet/google/gve/gve_tx_dqo.c | 8 +++++++- 1 file changed, 7 insertions(+), 1 deletion(-) diff --git a/drivers/net/ethernet/google/gve/gve_tx_dqo.c b/drivers/net/ethernet/google/gve/gve_tx_dqo.c index 80ab0a449ff5..0f6f7c5dbb2e 100644 --- a/drivers/net/ethernet/google/gve/gve_tx_dqo.c +++ b/drivers/net/ethernet/google/gve/gve_tx_dqo.c @@ -918,13 +918,19 @@ static bool gve_can_send_tso(const struct sk_buff *skb) { const int max_bufs_per_seg = GVE_TX_MAX_DATA_DESCS - 1; const struct skb_shared_info *shinfo = skb_shinfo(skb); - const int header_len = skb_tcp_all_headers(skb); const int gso_size = shinfo->gso_size; int cur_seg_num_bufs; int prev_frag_size; int cur_seg_size; + int header_len; int i; + /* Must match the header length programmed by gve_prep_tso(). */ + if (skb_is_gso_tcp(skb)) + header_len = skb_tcp_all_headers(skb); + else + header_len = skb_transport_offset(skb) + sizeof(struct udphdr); + cur_seg_size = skb_headlen(skb) - header_len; prev_frag_size = skb_headlen(skb); cur_seg_num_bufs = cur_seg_size > 0; From 3b430ea6234087957b0d3cd181e3116722b59819 Mon Sep 17 00:00:00 2001 From: Eddie Phillips Date: Thu, 24 Sep 2026 00:42:51 +0000 Subject: [PATCH 585/592] gve: fix TX drop when GSO MSS is too small for hw The device has a strict requirement that the minimum MSS (gso_size) for TSO/GSO packets must be at least 88 bytes. If a packet below this threshold is pushed to the hardware, it can cause hardware to silently drop the packet, leading to increased latency and retransmissions. Currently, this is validated too late in the transmit pipeline (gve_prep_tso), leading to silent drops. Fix this by moving the validation into the .ndo_features_check callback (gve_features_check_dqo). If we detect a GSO packet with a gso_size smaller than GVE_TX_MIN_TSO_MSS_DQO, we clear the GSO feature flags for this packet. Fixes: a57e5de476be ("gve: DQO: Add TX path") Signed-off-by: Eddie Phillips Signed-off-by: Eric Dumazet Reviewed-by: Harshitha Ramamurthy Link: https://patch.msgid.link/20260924004252.1196328-2-edumazet@google.com Signed-off-by: Jakub Kicinski --- drivers/net/ethernet/google/gve/gve_tx_dqo.c | 14 +++----------- 1 file changed, 3 insertions(+), 11 deletions(-) diff --git a/drivers/net/ethernet/google/gve/gve_tx_dqo.c b/drivers/net/ethernet/google/gve/gve_tx_dqo.c index 0f6f7c5dbb2e..e5fe17b04798 100644 --- a/drivers/net/ethernet/google/gve/gve_tx_dqo.c +++ b/drivers/net/ethernet/google/gve/gve_tx_dqo.c @@ -577,17 +577,6 @@ static int gve_prep_tso(struct sk_buff *skb) int header_len; int err; - /* Note: HW requires MSS (gso_size) to be <= 9728 and the total length - * of the TSO to be <= 262143. - * - * However, we don't validate these because: - * - Hypervisor enforces a limit of 9K MTU - * - Kernel will not produce a TSO larger than 64k - */ - - if (unlikely(shinfo->gso_size < GVE_TX_MIN_TSO_MSS_DQO)) - return -1; - /* Needed because we will modify header. */ err = skb_cow_head(skb, 0); if (err < 0) @@ -925,6 +914,9 @@ static bool gve_can_send_tso(const struct sk_buff *skb) int header_len; int i; + if (unlikely(gso_size < GVE_TX_MIN_TSO_MSS_DQO)) + return false; + /* Must match the header length programmed by gve_prep_tso(). */ if (skb_is_gso_tcp(skb)) header_len = skb_tcp_all_headers(skb); From 296c83b5ccc808c080865eb20fd7a477b0355bb7 Mon Sep 17 00:00:00 2001 From: Eric Dumazet Date: Thu, 24 Sep 2026 00:42:52 +0000 Subject: [PATCH 586/592] gve: DQO: reject TSO packets with an out of range MSS gve_prep_tso() notes that the device requires the MSS to be <= 9728, but does not enforce it, assuming the 9K MTU enforced by the hypervisor and the 64KB limit on TSO sizes are enough. This does not hold for packets that were not generated locally. A guest behind a tap, or any packet socket user, can provide an arbitrary gso_size in virtio_net_hdr. Layer 2 forwarding does not check the MTU for GSO packets (is_skb_forwardable()), and gso_features_check() only bounds skb->len and gso_segs, never gso_size. Such a packet reaches gve_tx_fill_tso_ctx_desc(), which puts gso_size into the mss field of the TSO context descriptor. This field is 14 bits wide, so a gso_size of 16384 is silently turned into an MSS of zero. Drop these packets from gve_prep_tso(), and make sure that gve_features_check_dqo() leaves their GSO bits alone: skb_segment() splits at gso_size regardless of the MTU, so falling back to software segmentation would give the device non TSO packets bigger than the 9728 bytes it supports. Note that the device can still be given oversized non TSO packets when the stack segments in software for other reasons, for instance after TSO has been disabled with ethtool. This is a generic issue, because the MTU check is skipped for GSO packets in the forwarding path, and is addressed separately. Fixes: a57e5de476be ("gve: DQO: Add TX path") Signed-off-by: Eric Dumazet Reviewed-by: Harshitha Ramamurthy Link: https://patch.msgid.link/20260924004252.1196328-3-edumazet@google.com Signed-off-by: Jakub Kicinski --- .../net/ethernet/google/gve/gve_desc_dqo.h | 5 ++++ drivers/net/ethernet/google/gve/gve_tx_dqo.c | 26 ++++++++++++++++++- 2 files changed, 30 insertions(+), 1 deletion(-) diff --git a/drivers/net/ethernet/google/gve/gve_desc_dqo.h b/drivers/net/ethernet/google/gve/gve_desc_dqo.h index f7786b03c744..d2c86c8eeae2 100644 --- a/drivers/net/ethernet/google/gve/gve_desc_dqo.h +++ b/drivers/net/ethernet/google/gve/gve_desc_dqo.h @@ -14,6 +14,11 @@ #define GVE_TX_MAX_HDR_SIZE_DQO 255 #define GVE_TX_MIN_TSO_MSS_DQO 88 +/* HW limit. This also has to fit in the 14 bits of the mss field of + * struct gve_tx_tso_context_desc_dqo. + */ +#define GVE_TX_MAX_TSO_MSS_DQO 9728 + #ifndef __LITTLE_ENDIAN_BITFIELD #error "Only little endian supported" #endif diff --git a/drivers/net/ethernet/google/gve/gve_tx_dqo.c b/drivers/net/ethernet/google/gve/gve_tx_dqo.c index e5fe17b04798..616c1921aebe 100644 --- a/drivers/net/ethernet/google/gve/gve_tx_dqo.c +++ b/drivers/net/ethernet/google/gve/gve_tx_dqo.c @@ -577,6 +577,20 @@ static int gve_prep_tso(struct sk_buff *skb) int header_len; int err; + /* Note: HW requires the total length of the TSO to be <= 262143, + * this is enforced by netif_set_tso_max_size(). + * + * MSS (gso_size) can not be trusted: packets forwarded from a tap or + * injected by a packet socket can carry an arbitrary value, while the + * mss field of the TSO context descriptor is only 14 bits wide. + * + * A too big MSS is dropped here instead of being rejected from + * gve_features_check_dqo(), because software segmentation would + * produce packets larger than the device can send. + */ + if (unlikely(shinfo->gso_size > GVE_TX_MAX_TSO_MSS_DQO)) + return -1; + /* Needed because we will modify header. */ err = skb_cow_head(skb, 0); if (err < 0) @@ -964,7 +978,17 @@ netdev_features_t gve_features_check_dqo(struct sk_buff *skb, struct net_device *dev, netdev_features_t features) { - if (skb_is_gso(skb) && !gve_can_send_tso(skb)) + if (!skb_is_gso(skb)) + return features; + + /* Keep the GSO bits for a too big MSS, so that gve_prep_tso() drops + * the packet: software segmentation would give packets larger than + * the device can send. + */ + if (skb_shinfo(skb)->gso_size > GVE_TX_MAX_TSO_MSS_DQO) + return features; + + if (!gve_can_send_tso(skb)) return features & ~NETIF_F_GSO_MASK; return features; From 72b5b9a28b996e09b8b5b944370c79851bb68f52 Mon Sep 17 00:00:00 2001 From: Zixuan Chai Date: Thu, 24 Sep 2026 09:26:05 +0800 Subject: [PATCH 587/592] llc: reserve device headroom for allocated frames llc_alloc_frame() reserves link-layer headroom using the device type. This is insufficient for stacked Ethernet devices such as VLAN devices, where vlan_dev_hard_header() pushes a VLAN header before the lower device's Ethernet header. An LLC response on such a device can therefore underflow skb headroom in eth_header(). Use LL_RESERVED_SPACE() to account for the device's actual required headroom while preserving the existing LLC device-type check. Fixes: bf9ae5386bca ("llc: use dev_hard_header") Cc: stable@vger.kernel.org Reported-by: VEGA Signed-off-by: Zixuan Chai Signed-off-by: Ren Wei Reviewed-by: Eric Dumazet Link: https://patch.msgid.link/20260924012613.2533934-1-weir@nebusec.ai Signed-off-by: Jakub Kicinski --- net/llc/llc_sap.c | 8 ++++---- 1 file changed, 4 insertions(+), 4 deletions(-) diff --git a/net/llc/llc_sap.c b/net/llc/llc_sap.c index 1bd446a21092..3904a1b4ba84 100644 --- a/net/llc/llc_sap.c +++ b/net/llc/llc_sap.c @@ -19,12 +19,12 @@ #include #include -static int llc_mac_header_len(unsigned short devtype) +static int llc_mac_header_len(struct net_device *dev) { - switch (devtype) { + switch (dev->type) { case ARPHRD_ETHER: case ARPHRD_LOOPBACK: - return sizeof(struct ethhdr); + return LL_RESERVED_SPACE(dev); } return 0; } @@ -45,7 +45,7 @@ struct sk_buff *llc_alloc_frame(struct sock *sk, struct net_device *dev, int hlen = type == LLC_PDU_TYPE_U ? 3 : 4; struct sk_buff *skb; - hlen += llc_mac_header_len(dev->type); + hlen += llc_mac_header_len(dev); skb = alloc_skb(hlen + data_size, GFP_ATOMIC); if (skb) { From 72f9dd522f8d6c5a00be9695c7bb74631eb5069e Mon Sep 17 00:00:00 2001 From: Eric Dumazet Date: Thu, 24 Sep 2026 08:29:48 +0000 Subject: [PATCH 588/592] llc: fix skb UAF and leaks on llc_mac_hdr_init() failure In llc_conn_ac_resend_i_xxx_x_set_0_or_send_rr(), if llc_mac_hdr_init() fails, kfree_skb(skb) is called instead of kfree_skb(nskb). This leaks the newly allocated nskb, reads from the freed skb via LLC_I_GET_NR(pdu), and double-frees skb when llc_conn_state_process() drops its reference. In llc_sap_action_send_xid_r() and llc_sap_action_send_test_r(), nskb is leaked if llc_mac_hdr_init() returns an error. Free nskb in all three error paths. Fixes: 1da177e4c3f4 ("Linux-2.6.12-rc2") Closes: https://lore.kernel.org/netdev/179022851638.2160803.1808206741379444999@kernel.org/ Signed-off-by: Eric Dumazet Link: https://patch.msgid.link/20260924082951.1599377-2-edumazet@google.com Signed-off-by: Jakub Kicinski --- net/llc/llc_c_ac.c | 2 +- net/llc/llc_s_ac.c | 4 ++++ 2 files changed, 5 insertions(+), 1 deletion(-) diff --git a/net/llc/llc_c_ac.c b/net/llc/llc_c_ac.c index 724ecd741d4c..1aa7fe28acdd 100644 --- a/net/llc/llc_c_ac.c +++ b/net/llc/llc_c_ac.c @@ -437,7 +437,7 @@ int llc_conn_ac_resend_i_xxx_x_set_0_or_send_rr(struct sock *sk, if (likely(!rc)) llc_conn_send_pdu(sk, nskb); else - kfree_skb(skb); + kfree_skb(nskb); } if (rc) { nr = LLC_I_GET_NR(pdu); diff --git a/net/llc/llc_s_ac.c b/net/llc/llc_s_ac.c index 98deee560373..831998211b52 100644 --- a/net/llc/llc_s_ac.c +++ b/net/llc/llc_s_ac.c @@ -121,6 +121,8 @@ int llc_sap_action_send_xid_r(struct llc_sap *sap, struct sk_buff *skb) rc = llc_mac_hdr_init(nskb, mac_sa, mac_da); if (likely(!rc)) rc = dev_queue_xmit(nskb); + else + kfree_skb(nskb); out: return rc; } @@ -170,6 +172,8 @@ int llc_sap_action_send_test_r(struct llc_sap *sap, struct sk_buff *skb) rc = llc_mac_hdr_init(nskb, mac_sa, mac_da); if (likely(!rc)) rc = dev_queue_xmit(nskb); + else + kfree_skb(nskb); out: return rc; } From ac704ff08e511c87643799c385f55ecd69b85e03 Mon Sep 17 00:00:00 2001 From: Eric Dumazet Date: Thu, 24 Sep 2026 08:29:49 +0000 Subject: [PATCH 589/592] bridge: check llc_mac_hdr_init() return value in br_send_bpdu() If llc_mac_hdr_init() fails (for instance if the port device type does not support LLC or dev_hard_header() fails), br_send_bpdu() should drop the skb instead of resetting the mac header to the LLC payload and transmitting a malformed frame. Fixes: 1da177e4c3f4 ("Linux-2.6.12-rc2") Closes: https://lore.kernel.org/netdev/179022851638.2160803.1808206741379444999@kernel.org/ Cc: Nikolay Aleksandrov Cc: Ido Schimmel Cc: bridge@lists.linux.dev Signed-off-by: Eric Dumazet Acked-by: Nikolay Aleksandrov Link: https://patch.msgid.link/20260924082951.1599377-3-edumazet@google.com Signed-off-by: Jakub Kicinski --- net/bridge/br_stp_bpdu.c | 5 ++++- 1 file changed, 4 insertions(+), 1 deletion(-) diff --git a/net/bridge/br_stp_bpdu.c b/net/bridge/br_stp_bpdu.c index 74ec42ba1e7d..21d092f5acbb 100644 --- a/net/bridge/br_stp_bpdu.c +++ b/net/bridge/br_stp_bpdu.c @@ -52,7 +52,10 @@ static void br_send_bpdu(struct net_bridge_port *p, LLC_SAP_BSPAN, LLC_PDU_CMD); llc_pdu_init_as_ui_cmd(skb); - llc_mac_hdr_init(skb, p->dev->dev_addr, p->br->group_addr); + if (llc_mac_hdr_init(skb, p->dev->dev_addr, p->br->group_addr)) { + kfree_skb(skb); + return; + } skb_reset_mac_header(skb); From 907b978e82cb4c1c245fc2985bb27c5d5c88c8f6 Mon Sep 17 00:00:00 2001 From: Eric Dumazet Date: Thu, 24 Sep 2026 08:29:50 +0000 Subject: [PATCH 590/592] net/sched: sch_teql: fix shadowed err in __teql_resolve() __teql_resolve() declares an inner 'int err;' inside the 'if (neigh_event_send(n, skb_res) == 0)' block, shadowing the outer 'int err = 0;'. As a result, a negative return from dev_hard_header() is written to the inner variable and __teql_resolve() still returns 0. Remove the shadowed variable and set the outer err to -EINVAL when dev_hard_header() returns a negative error. Fixes: 1da177e4c3f4 ("Linux-2.6.12-rc2") Closes: https://lore.kernel.org/netdev/179022851638.2160803.1808206741379444999@kernel.org/ Cc: Jamal Hadi Salim Cc: Jiri Pirko Signed-off-by: Eric Dumazet Link: https://patch.msgid.link/20260924082951.1599377-4-edumazet@google.com Signed-off-by: Jakub Kicinski --- net/sched/sch_teql.c | 7 ++----- 1 file changed, 2 insertions(+), 5 deletions(-) diff --git a/net/sched/sch_teql.c b/net/sched/sch_teql.c index 9e52afc2d980..409ce50cc0db 100644 --- a/net/sched/sch_teql.c +++ b/net/sched/sch_teql.c @@ -265,14 +265,11 @@ __teql_resolve(struct sk_buff *skb, struct sk_buff *skb_res, } if (neigh_event_send(n, skb_res) == 0) { - int err; char haddr[MAX_ADDR_LEN]; neigh_ha_snapshot(haddr, n, dev); - err = dev_hard_header(skb, dev, ntohs(skb_protocol(skb, false)), - haddr, NULL, skb->len); - - if (err < 0) + if (dev_hard_header(skb, dev, ntohs(skb_protocol(skb, false)), + haddr, NULL, skb->len) < 0) err = -EINVAL; } else { err = (skb_res == NULL) ? -EAGAIN : 1; From cd5dd68267c4238795fadaf02b3575ca3f8a6500 Mon Sep 17 00:00:00 2001 From: Eric Dumazet Date: Thu, 24 Sep 2026 08:29:51 +0000 Subject: [PATCH 591/592] vlan: ensure sufficient headroom in vlan_dev_hard_header() Callers that only reserve ETH_HLEN or less (such as llc_alloc_frame()), or skbs allocated before dynamic device/headroom changes (e.g. toggling VLAN_FLAG_REORDER_HDR or bonding/team switching slaves), can reach vlan_dev_hard_header() with insufficient headroom and trigger skb_under_panic(). Use skb_cow_head() in vlan_dev_hard_header() when VLAN_FLAG_REORDER_HDR is not set to ensure sufficient headroom for the VLAN header(s) and the underlying device hard header. Use READ_ONCE() to read dev->hard_header_len and dev->needed_headroom as they can be updated concurrently under RTNL (e.g. in vlan_transfer_features()) while vlan_dev_hard_header() runs locklessly on the transmit path. Also avoid LL_RESERVED_SPACE(dev) here so that the extra HH_DATA_MOD alignment padding does not trigger unnecessary pskb_expand_head() reallocations on inner stacked VLAN devices after the outer VLAN header has been pushed. Fixes: 1da177e4c3f4 ("Linux-2.6.12-rc2") Reported-by: Zixuan Chai Closes: https://lore.kernel.org/netdev/cover.1789987105.git.petalzu987@gmail.com/ Link: https://lore.kernel.org/netdev/179022851638.2160803.1808206741379444999@kernel.org/ Cc: Hangbin Liu Signed-off-by: Eric Dumazet Link: https://patch.msgid.link/20260924082951.1599377-5-edumazet@google.com Signed-off-by: Jakub Kicinski --- net/8021q/vlan_dev.c | 5 +++++ 1 file changed, 5 insertions(+) diff --git a/net/8021q/vlan_dev.c b/net/8021q/vlan_dev.c index 2859cbac3f26..c949c6a82945 100644 --- a/net/8021q/vlan_dev.c +++ b/net/8021q/vlan_dev.c @@ -55,6 +55,11 @@ static int vlan_dev_hard_header(struct sk_buff *skb, struct net_device *dev, int rc; if (!(vlan->flags & VLAN_FLAG_REORDER_HDR)) { + unsigned int hlen = READ_ONCE(dev->hard_header_len) + + READ_ONCE(dev->needed_headroom); + + if (skb_cow_head(skb, hlen) < 0) + return -ENOMEM; vhdr = skb_push(skb, VLAN_HLEN); vlan_tci = vlan->vlan_id; From fc6d80eb504458d6416b75a94188b268c95c6533 Mon Sep 17 00:00:00 2001 From: Willem de Bruijn Date: Thu, 24 Sep 2026 11:44:12 -0400 Subject: [PATCH 592/592] tcp: prevent collapsing skbs across boundary in rtx queue tcp_write_collapse_fence() sets TCP_SKB_CB(skb)->eor = 1 on tcp_write_queue_tail(sk) to prevent skbs queued after a switch to device encryption from being collapsed into earlier skbs. The fence is a no-op if all earlier data has already been transmitted when the switch happens: sk->sk_write_queue is empty. The not yet acknowledged earlier skbs wait in sk->tcp_rtx_queue with eor 0. On a subsequent retransmit or SACK shift, tcp_retrans_try_collapse() or tcp_shift_skb_data() can then merge an skb queued after the switch into one queued before it. Both users of the fence are affected: - psp: devices only encrypt skbs with skb->decrypted set. The merged skb keeps decrypted = 0 from the earlier skb, so merged data sent after psp_sock_assoc_set_tx() is retransmitted in cleartext. - tls device offload: the merged skb straddles the start marker set in tls_set_device_offload(). The software fallback (fill_sg_in() returns -EINVAL) and the mlx5, nfp and funeth drivers cannot handle such an skb and drop it. Every retransmit rebuilds the same skb, so the connection stalls. Fix this in two places, for defense in depth: 1. Fall back to tcp_rtx_queue_tail(sk) in tcp_write_collapse_fence() when tcp_write_queue_tail(sk) is NULL. 2. Check !skb_cmp_decrypted(to, from) in tcp_skb_can_collapse(), as tcp_skb_can_collapse_rx() does on receive. skb_shift(), which both collapse paths call, already has a DEBUG_NET_WARN_ON_ONCE() for this condition. Fixes: e8f69799810c ("net/tls: Add generic NIC offload infrastructure") Cc: stable@vger.kernel.org Signed-off-by: Willem de Bruijn Reviewed-by: Eric Dumazet Reviewed-by: Daniel Zahka Link: https://patch.msgid.link/20260924154427.953800-1-willemdebruijn.kernel@gmail.com Signed-off-by: Jakub Kicinski --- include/net/tcp.h | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/include/net/tcp.h b/include/net/tcp.h index 436495ff2271..4416cdf9bf30 100644 --- a/include/net/tcp.h +++ b/include/net/tcp.h @@ -1232,9 +1232,9 @@ static inline bool tcp_skb_can_collapse_to(const struct sk_buff *skb) static inline bool tcp_skb_can_collapse(const struct sk_buff *to, const struct sk_buff *from) { - /* skb_cmp_decrypted() not needed, use tcp_write_collapse_fence() */ return likely(tcp_skb_can_collapse_to(to) && mptcp_skb_can_collapse(to, from) && + !skb_cmp_decrypted(to, from) && skb_pure_zcopy_same(to, from) && skb_frags_readable(to) == skb_frags_readable(from)); } @@ -2327,7 +2327,7 @@ static inline void tcp_rtx_queue_unlink_and_free(struct sk_buff *skb, struct soc static inline void tcp_write_collapse_fence(struct sock *sk) { - struct sk_buff *skb = tcp_write_queue_tail(sk); + struct sk_buff *skb = tcp_write_queue_tail(sk) ?: tcp_rtx_queue_tail(sk); if (skb) TCP_SKB_CB(skb)->eor = 1;