From 2128f3cca5a2e7ab4d1ffb16c0e0431c3a0106a1 Mon Sep 17 00:00:00 2001 From: Ma Jun Date: Mon, 11 Dec 2023 18:06:22 +0800 Subject: [PATCH 01/92] Documentation/driver-api: Add document about WBRF mechanism Add documentation about AMD's Wifi band RFI mitigation (WBRF) mechanism explaining the theory and how it is used. Signed-off-by: Ma Jun Reviewed-by: Hans de Goede Reviewed-by: Mario Limonciello Signed-off-by: Hans de Goede --- Documentation/driver-api/index.rst | 1 + Documentation/driver-api/wbrf.rst | 78 ++++++++++++++++++++++++++++++ 2 files changed, 79 insertions(+) create mode 100644 Documentation/driver-api/wbrf.rst diff --git a/Documentation/driver-api/index.rst b/Documentation/driver-api/index.rst index f549a68951d7..8bc4ebe7a36f 100644 --- a/Documentation/driver-api/index.rst +++ b/Documentation/driver-api/index.rst @@ -115,6 +115,7 @@ available subsections can be seen below. hte/index wmi dpll + wbrf .. only:: subproject and html diff --git a/Documentation/driver-api/wbrf.rst b/Documentation/driver-api/wbrf.rst new file mode 100644 index 000000000000..f48bfa029813 --- /dev/null +++ b/Documentation/driver-api/wbrf.rst @@ -0,0 +1,78 @@ +.. SPDX-License-Identifier: GPL-2.0-or-later + +================================= +WBRF - Wifi Band RFI Mitigations +================================= + +Due to electrical and mechanical constraints in certain platform designs +there may be likely interference of relatively high-powered harmonics of +the GPU memory clocks with local radio module frequency bands used by +certain Wifi bands. + +To mitigate possible RFI interference producers can advertise the +frequencies in use and consumers can use this information to avoid using +these frequencies for sensitive features. + +When a platform is known to have this issue with any contained devices, +the platform designer will advertise the availability of this feature via +ACPI devices with a device specific method (_DSM). +* Producers with this _DSM will be able to advertise the frequencies in use. +* Consumers with this _DSM will be able to register for notifications of +frequencies in use. + +Some general terms +================== + +Producer: such component who can produce high-powered radio frequency +Consumer: such component who can adjust its in-use frequency in +response to the radio frequencies of other components to mitigate the +possible RFI. + +To make the mechanism function, those producers should notify active use +of their particular frequencies so that other consumers can make relative +internal adjustments as necessary to avoid this resonance. + +ACPI interface +============== + +Although initially used by for wifi + dGPU use cases, the ACPI interface +can be scaled to any type of device that a platform designer discovers +can cause interference. + +The GUID used for the _DSM is 7B7656CF-DC3D-4C1C-83E9-66E721DE3070. + +3 functions are available in this _DSM: + +* 0: discover # of functions available +* 1: record RF bands in use +* 2: retrieve RF bands in use + +Driver programming interface +============================ + +.. kernel-doc:: drivers/platform/x86/amd/wbrf.c + +Sample Usage +============= + +The expected flow for the producers: +1. During probe, call `acpi_amd_wbrf_supported_producer` to check if WBRF +can be enabled for the device. +2. On using some frequency band, call `acpi_amd_wbrf_add_remove` with 'add' +param to get other consumers properly notified. +3. Or on stopping using some frequency band, call +`acpi_amd_wbrf_add_remove` with 'remove' param to get other consumers notified. + +The expected flow for the consumers: +1. During probe, call `acpi_amd_wbrf_supported_consumer` to check if WBRF +can be enabled for the device. +2. Call `amd_wbrf_register_notifier` to register for notification +of frequency band change(add or remove) from other producers. +3. Call the `amd_wbrf_retrieve_freq_band` initally to retrieve +current active frequency bands considering some producers may broadcast +such information before the consumer is up. +4. On receiving a notification for frequency band change, run +`amd_wbrf_retrieve_freq_band` again to retrieve the latest +active frequency bands. +5. During driver cleanup, call `amd_wbrf_unregister_notifier` to +unregister the notifier. From 58e82a62669da52e688f4a8b89922c1839bf1001 Mon Sep 17 00:00:00 2001 From: Ma Jun Date: Mon, 11 Dec 2023 18:06:23 +0800 Subject: [PATCH 02/92] platform/x86/amd: Add support for AMD ACPI based Wifi band RFI mitigation feature Due to electrical and mechanical constraints in certain platform designs there may be likely interference of relatively high-powered harmonics of the (G-)DDR memory clocks with local radio module frequency bands used by Wifi 6/6e/7. To mitigate this, AMD has introduced a mechanism that devices can use to notify active use of particular frequencies so that other devices can make relative internal adjustments as necessary to avoid this resonance. Co-developed-by: Evan Quan Signed-off-by: Evan Quan Signed-off-by: Ma Jun Reviewed-by: Mario Limonciello Reviewed-by: Hans de Goede Signed-off-by: Hans de Goede --- drivers/platform/x86/amd/Kconfig | 14 ++ drivers/platform/x86/amd/Makefile | 1 + drivers/platform/x86/amd/wbrf.c | 317 ++++++++++++++++++++++++++++++ include/linux/acpi_amd_wbrf.h | 91 +++++++++ 4 files changed, 423 insertions(+) create mode 100644 drivers/platform/x86/amd/wbrf.c create mode 100644 include/linux/acpi_amd_wbrf.h diff --git a/drivers/platform/x86/amd/Kconfig b/drivers/platform/x86/amd/Kconfig index 55f3a2fc6aec..54753213cc61 100644 --- a/drivers/platform/x86/amd/Kconfig +++ b/drivers/platform/x86/amd/Kconfig @@ -18,3 +18,17 @@ config AMD_HSMP If you choose to compile this driver as a module the module will be called amd_hsmp. + +config AMD_WBRF + bool "AMD Wifi RF Band mitigations (WBRF)" + depends on ACPI + help + WBRF(Wifi Band RFI mitigation) mechanism allows Wifi drivers + to notify the frequencies they are using so that other hardware + can be reconfigured to avoid harmonic conflicts. + + AMD provides an ACPI based mechanism to support WBRF on platform with + appropriate underlying support. + + This mechanism will only be activated on platforms that advertise a + need for it. diff --git a/drivers/platform/x86/amd/Makefile b/drivers/platform/x86/amd/Makefile index f04932b7a7d1..dcec0a46f8af 100644 --- a/drivers/platform/x86/amd/Makefile +++ b/drivers/platform/x86/amd/Makefile @@ -8,3 +8,4 @@ obj-$(CONFIG_AMD_PMC) += pmc/ amd_hsmp-y := hsmp.o obj-$(CONFIG_AMD_HSMP) += amd_hsmp.o obj-$(CONFIG_AMD_PMF) += pmf/ +obj-$(CONFIG_AMD_WBRF) += wbrf.o diff --git a/drivers/platform/x86/amd/wbrf.c b/drivers/platform/x86/amd/wbrf.c new file mode 100644 index 000000000000..dd197b3aebe0 --- /dev/null +++ b/drivers/platform/x86/amd/wbrf.c @@ -0,0 +1,317 @@ +// SPDX-License-Identifier: GPL-2.0 +/* + * Wifi Frequency Band Manage Interface + * Copyright (C) 2023 Advanced Micro Devices + */ + +#include +#include + +/* + * Functions bit vector for WBRF method + * + * Bit 0: WBRF supported. + * Bit 1: Function 1 (Add / Remove frequency) is supported. + * Bit 2: Function 2 (Get frequency list) is supported. + */ +#define WBRF_ENABLED 0x0 +#define WBRF_RECORD 0x1 +#define WBRF_RETRIEVE 0x2 + +#define WBRF_REVISION 0x1 + +/* + * The data structure used for WBRF_RETRIEVE is not naturally aligned. + * And unfortunately the design has been settled down. + */ +struct amd_wbrf_ranges_out { + u32 num_of_ranges; + struct freq_band_range band_list[MAX_NUM_OF_WBRF_RANGES]; +} __packed; + +static const guid_t wifi_acpi_dsm_guid = + GUID_INIT(0x7b7656cf, 0xdc3d, 0x4c1c, + 0x83, 0xe9, 0x66, 0xe7, 0x21, 0xde, 0x30, 0x70); + +/* + * Used to notify consumer (amdgpu driver currently) about + * the wifi frequency is change. + */ +static BLOCKING_NOTIFIER_HEAD(wbrf_chain_head); + +static int wbrf_record(struct acpi_device *adev, uint8_t action, struct wbrf_ranges_in_out *in) +{ + union acpi_object argv4; + union acpi_object *tmp; + union acpi_object *obj; + u32 num_of_ranges = 0; + u32 num_of_elements; + u32 arg_idx = 0; + int ret; + u32 i; + + if (!in) + return -EINVAL; + + for (i = 0; i < ARRAY_SIZE(in->band_list); i++) { + if (in->band_list[i].start && in->band_list[i].end) + num_of_ranges++; + } + + /* + * The num_of_ranges value in the "in" object supplied by + * the caller is required to be equal to the number of + * entries in the band_list array in there. + */ + if (num_of_ranges != in->num_of_ranges) + return -EINVAL; + + /* + * Every input frequency band comes with two end points(start/end) + * and each is accounted as an element. Meanwhile the range count + * and action type are accounted as an element each. + * So, the total element count = 2 * num_of_ranges + 1 + 1. + */ + num_of_elements = 2 * num_of_ranges + 2; + + tmp = kcalloc(num_of_elements, sizeof(*tmp), GFP_KERNEL); + if (!tmp) + return -ENOMEM; + + argv4.package.type = ACPI_TYPE_PACKAGE; + argv4.package.count = num_of_elements; + argv4.package.elements = tmp; + + /* save the number of ranges*/ + tmp[0].integer.type = ACPI_TYPE_INTEGER; + tmp[0].integer.value = num_of_ranges; + + /* save the action(WBRF_RECORD_ADD/REMOVE/RETRIEVE) */ + tmp[1].integer.type = ACPI_TYPE_INTEGER; + tmp[1].integer.value = action; + + arg_idx = 2; + for (i = 0; i < ARRAY_SIZE(in->band_list); i++) { + if (!in->band_list[i].start || !in->band_list[i].end) + continue; + + tmp[arg_idx].integer.type = ACPI_TYPE_INTEGER; + tmp[arg_idx++].integer.value = in->band_list[i].start; + tmp[arg_idx].integer.type = ACPI_TYPE_INTEGER; + tmp[arg_idx++].integer.value = in->band_list[i].end; + } + + obj = acpi_evaluate_dsm(adev->handle, &wifi_acpi_dsm_guid, + WBRF_REVISION, WBRF_RECORD, &argv4); + + if (!obj) + return -EINVAL; + + if (obj->type != ACPI_TYPE_INTEGER) { + ret = -EINVAL; + goto out; + } + + ret = obj->integer.value; + if (ret) + ret = -EINVAL; + +out: + ACPI_FREE(obj); + kfree(tmp); + + return ret; +} + +/** + * acpi_amd_wbrf_add_remove - add or remove the frequency band the device is using + * + * @dev: device pointer + * @action: remove or add the frequency band into bios + * @in: input structure containing the frequency band the device is using + * + * Broadcast to other consumers the frequency band the device starts + * to use. Underneath the surface the information is cached into an + * internal buffer first. Then a notification is sent to all those + * registered consumers. So then they can retrieve that buffer to + * know the latest active frequency bands. Consumers that haven't + * yet been registered can retrieve the information from the cache + * when they register. + * + * Return: + * 0 for success add/remove wifi frequency band. + * Returns a negative error code for failure. + */ +int acpi_amd_wbrf_add_remove(struct device *dev, uint8_t action, struct wbrf_ranges_in_out *in) +{ + struct acpi_device *adev; + int ret; + + adev = ACPI_COMPANION(dev); + if (!adev) + return -ENODEV; + + ret = wbrf_record(adev, action, in); + if (ret) + return ret; + + blocking_notifier_call_chain(&wbrf_chain_head, WBRF_CHANGED, NULL); + + return 0; +} +EXPORT_SYMBOL_GPL(acpi_amd_wbrf_add_remove); + +/** + * acpi_amd_wbrf_supported_producer - determine if the WBRF can be enabled + * for the device as a producer + * + * @dev: device pointer + * + * Check if the platform equipped with necessary implementations to + * support WBRF for the device as a producer. + * + * Return: + * true if WBRF is supported, otherwise returns false + */ +bool acpi_amd_wbrf_supported_producer(struct device *dev) +{ + struct acpi_device *adev; + + adev = ACPI_COMPANION(dev); + if (!adev) + return false; + + return acpi_check_dsm(adev->handle, &wifi_acpi_dsm_guid, + WBRF_REVISION, BIT(WBRF_RECORD)); +} +EXPORT_SYMBOL_GPL(acpi_amd_wbrf_supported_producer); + +/** + * acpi_amd_wbrf_supported_consumer - determine if the WBRF can be enabled + * for the device as a consumer + * + * @dev: device pointer + * + * Determine if the platform equipped with necessary implementations to + * support WBRF for the device as a consumer. + * + * Return: + * true if WBRF is supported, otherwise returns false. + */ +bool acpi_amd_wbrf_supported_consumer(struct device *dev) +{ + struct acpi_device *adev; + + adev = ACPI_COMPANION(dev); + if (!adev) + return false; + + return acpi_check_dsm(adev->handle, &wifi_acpi_dsm_guid, + WBRF_REVISION, BIT(WBRF_RETRIEVE)); +} +EXPORT_SYMBOL_GPL(acpi_amd_wbrf_supported_consumer); + +/** + * amd_wbrf_retrieve_freq_band - retrieve current active frequency bands + * + * @dev: device pointer + * @out: output structure containing all the active frequency bands + * + * Retrieve the current active frequency bands which were broadcasted + * by other producers. The consumer who calls this API should take + * proper actions if any of the frequency band may cause RFI with its + * own frequency band used. + * + * Return: + * 0 for getting wifi freq band successfully. + * Returns a negative error code for failure. + */ +int amd_wbrf_retrieve_freq_band(struct device *dev, struct wbrf_ranges_in_out *out) +{ + struct amd_wbrf_ranges_out acpi_out = {0}; + struct acpi_device *adev; + union acpi_object *obj; + union acpi_object param; + int ret = 0; + + adev = ACPI_COMPANION(dev); + if (!adev) + return -ENODEV; + + param.type = ACPI_TYPE_STRING; + param.string.length = 0; + param.string.pointer = NULL; + + obj = acpi_evaluate_dsm(adev->handle, &wifi_acpi_dsm_guid, + WBRF_REVISION, WBRF_RETRIEVE, ¶m); + if (!obj) + return -EINVAL; + + /* + * The return buffer is with variable length and the format below: + * number_of_entries(1 DWORD): Number of entries + * start_freq of 1st entry(1 QWORD): Start frequency of the 1st entry + * end_freq of 1st entry(1 QWORD): End frequency of the 1st entry + * ... + * ... + * start_freq of the last entry(1 QWORD) + * end_freq of the last entry(1 QWORD) + * + * Thus the buffer length is determined by the number of entries. + * - For zero entry scenario, the buffer length will be 4 bytes. + * - For one entry scenario, the buffer length will be 20 bytes. + */ + if (obj->buffer.length > sizeof(acpi_out) || obj->buffer.length < 4) { + dev_err(dev, "Wrong sized WBRT information"); + ret = -EINVAL; + goto out; + } + memcpy(&acpi_out, obj->buffer.pointer, obj->buffer.length); + + out->num_of_ranges = acpi_out.num_of_ranges; + memcpy(out->band_list, acpi_out.band_list, sizeof(acpi_out.band_list)); + +out: + ACPI_FREE(obj); + return ret; +} +EXPORT_SYMBOL_GPL(amd_wbrf_retrieve_freq_band); + +/** + * amd_wbrf_register_notifier - register for notifications of frequency + * band update + * + * @nb: driver notifier block + * + * The consumer should register itself via this API so that it can get + * notified on the frequency band updates from other producers. + * + * Return: + * 0 for registering a consumer driver successfully. + * Returns a negative error code for failure. + */ +int amd_wbrf_register_notifier(struct notifier_block *nb) +{ + return blocking_notifier_chain_register(&wbrf_chain_head, nb); +} +EXPORT_SYMBOL_GPL(amd_wbrf_register_notifier); + +/** + * amd_wbrf_unregister_notifier - unregister for notifications of + * frequency band update + * + * @nb: driver notifier block + * + * The consumer should call this API when it is longer interested with + * the frequency band updates from other producers. Usually, this should + * be performed during driver cleanup. + * + * Return: + * 0 for unregistering a consumer driver. + * Returns a negative error code for failure. + */ +int amd_wbrf_unregister_notifier(struct notifier_block *nb) +{ + return blocking_notifier_chain_unregister(&wbrf_chain_head, nb); +} +EXPORT_SYMBOL_GPL(amd_wbrf_unregister_notifier); diff --git a/include/linux/acpi_amd_wbrf.h b/include/linux/acpi_amd_wbrf.h new file mode 100644 index 000000000000..898f31d536d4 --- /dev/null +++ b/include/linux/acpi_amd_wbrf.h @@ -0,0 +1,91 @@ +/* SPDX-License-Identifier: GPL-2.0 */ +/* + * Wifi Band Exclusion Interface (AMD ACPI Implementation) + * Copyright (C) 2023 Advanced Micro Devices + */ + +#ifndef _ACPI_AMD_WBRF_H +#define _ACPI_AMD_WBRF_H + +#include +#include + +/* The maximum number of frequency band ranges */ +#define MAX_NUM_OF_WBRF_RANGES 11 + +/* Record actions */ +#define WBRF_RECORD_ADD 0x0 +#define WBRF_RECORD_REMOVE 0x1 + +/** + * struct freq_band_range - Wifi frequency band range definition + * @start: start frequency point (in Hz) + * @end: end frequency point (in Hz) + */ +struct freq_band_range { + u64 start; + u64 end; +}; + +/** + * struct wbrf_ranges_in_out - wbrf ranges info + * @num_of_ranges: total number of band ranges in this struct + * @band_list: array of Wifi band ranges + */ +struct wbrf_ranges_in_out { + u64 num_of_ranges; + struct freq_band_range band_list[MAX_NUM_OF_WBRF_RANGES]; +}; + +/** + * enum wbrf_notifier_actions - wbrf notifier actions index + * @WBRF_CHANGED: there was some frequency band updates. The consumers + * should retrieve the latest active frequency bands. + */ +enum wbrf_notifier_actions { + WBRF_CHANGED, +}; + +#if IS_ENABLED(CONFIG_AMD_WBRF) +bool acpi_amd_wbrf_supported_producer(struct device *dev); +int acpi_amd_wbrf_add_remove(struct device *dev, uint8_t action, struct wbrf_ranges_in_out *in); +bool acpi_amd_wbrf_supported_consumer(struct device *dev); +int amd_wbrf_retrieve_freq_band(struct device *dev, struct wbrf_ranges_in_out *out); +int amd_wbrf_register_notifier(struct notifier_block *nb); +int amd_wbrf_unregister_notifier(struct notifier_block *nb); +#else +static inline +bool acpi_amd_wbrf_supported_consumer(struct device *dev) +{ + return false; +} + +static inline +int acpi_amd_wbrf_add_remove(struct device *dev, uint8_t action, struct wbrf_ranges_in_out *in) +{ + return -ENODEV; +} + +static inline +bool acpi_amd_wbrf_supported_producer(struct device *dev) +{ + return false; +} +static inline +int amd_wbrf_retrieve_freq_band(struct device *dev, struct wbrf_ranges_in_out *out) +{ + return -ENODEV; +} +static inline +int amd_wbrf_register_notifier(struct notifier_block *nb) +{ + return -ENODEV; +} +static inline +int amd_wbrf_unregister_notifier(struct notifier_block *nb) +{ + return -ENODEV; +} +#endif /* CONFIG_AMD_WBRF */ + +#endif /* _ACPI_AMD_WBRF_H */ From 51ea405c47f833e55d19401b35b71100197e6d5d Mon Sep 17 00:00:00 2001 From: Alex Deucher Date: Mon, 11 Dec 2023 11:28:30 -0500 Subject: [PATCH 03/92] drm/amdgpu: fix buffer funcs setting order on suspend harder Part of commit c03581986234 ("drm/amdgpu: fix buffer funcs setting order on suspend") got dropped accidently. Add it back. Fixes: c03581986234 ("drm/amdgpu: fix buffer funcs setting order on suspend") Signed-off-by: Alex Deucher --- drivers/gpu/drm/amd/amdgpu/amdgpu_device.c | 2 -- 1 file changed, 2 deletions(-) diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_device.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_device.c index 138c7b37af69..f58fb719292d 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_device.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_device.c @@ -4589,8 +4589,6 @@ int amdgpu_device_suspend(struct drm_device *dev, bool fbcon) amdgpu_ras_suspend(adev); - amdgpu_ttm_set_buffer_funcs_status(adev, false); - amdgpu_device_ip_suspend_phase1(adev); if (!adev->in_s0ix) From 5d1ff65f80fd8c11476bd10d10aa2b2b639de432 Mon Sep 17 00:00:00 2001 From: Colin Ian King Date: Thu, 7 Dec 2023 10:54:54 +0000 Subject: [PATCH 04/92] drm/amd/display: Fix spelling mistake "SMC_MSG_AllowZstatesEntr" -> "SMC_MSG_AllowZstatesEntry" There is a spelling mistake in a smu_print message. Fix it. Signed-off-by: Colin Ian King Signed-off-by: Alex Deucher --- drivers/gpu/drm/amd/display/dc/clk_mgr/dcn35/dcn35_smu.c | 8 ++++---- 1 file changed, 4 insertions(+), 4 deletions(-) diff --git a/drivers/gpu/drm/amd/display/dc/clk_mgr/dcn35/dcn35_smu.c b/drivers/gpu/drm/amd/display/dc/clk_mgr/dcn35/dcn35_smu.c index d6db9d7fced2..6d4a1ffab5ed 100644 --- a/drivers/gpu/drm/amd/display/dc/clk_mgr/dcn35/dcn35_smu.c +++ b/drivers/gpu/drm/amd/display/dc/clk_mgr/dcn35/dcn35_smu.c @@ -361,26 +361,26 @@ void dcn35_smu_set_zstate_support(struct clk_mgr_internal *clk_mgr, enum dcn_zst case DCN_ZSTATE_SUPPORT_ALLOW: msg_id = VBIOSSMC_MSG_AllowZstatesEntry; param = (1 << 10) | (1 << 9) | (1 << 8); - smu_print("%s: SMC_MSG_AllowZstatesEntr msg = ALLOW, param = %d\n", __func__, param); + smu_print("%s: SMC_MSG_AllowZstatesEntry msg = ALLOW, param = %d\n", __func__, param); break; case DCN_ZSTATE_SUPPORT_DISALLOW: msg_id = VBIOSSMC_MSG_AllowZstatesEntry; param = 0; - smu_print("%s: SMC_MSG_AllowZstatesEntr msg_id = DISALLOW, param = %d\n", __func__, param); + smu_print("%s: SMC_MSG_AllowZstatesEntry msg_id = DISALLOW, param = %d\n", __func__, param); break; case DCN_ZSTATE_SUPPORT_ALLOW_Z10_ONLY: msg_id = VBIOSSMC_MSG_AllowZstatesEntry; param = (1 << 10); - smu_print("%s: SMC_MSG_AllowZstatesEntr msg = ALLOW_Z10_ONLY, param = %d\n", __func__, param); + smu_print("%s: SMC_MSG_AllowZstatesEntry msg = ALLOW_Z10_ONLY, param = %d\n", __func__, param); break; case DCN_ZSTATE_SUPPORT_ALLOW_Z8_Z10_ONLY: msg_id = VBIOSSMC_MSG_AllowZstatesEntry; param = (1 << 10) | (1 << 8); - smu_print("%s: SMC_MSG_AllowZstatesEntr msg = ALLOW_Z8_Z10_ONLY, param = %d\n", __func__, param); + smu_print("%s: SMC_MSG_AllowZstatesEntry msg = ALLOW_Z8_Z10_ONLY, param = %d\n", __func__, param); break; case DCN_ZSTATE_SUPPORT_ALLOW_Z8_ONLY: From 1e13c5644c443dee727ac1330bc118c909a1cf07 Mon Sep 17 00:00:00 2001 From: Melissa Wen Date: Thu, 16 Nov 2023 18:57:41 -0100 Subject: [PATCH 05/92] drm/drm_mode_object: increase max objects to accommodate new color props DRM_OBJECT_MAX_PROPERTY limits the number of properties to be attached and we are increasing that value all time we add a new property (generic or driver-specific). In this series, we are adding 13 new KMS driver-specific properties for AMD color manage: - CRTC Gamma enumerated Transfer Function - Plane: Degamma LUT+size+TF, HDR multiplier, shaper LUT+size+TF, 3D LUT+size, blend LUT+size+TF (12) Therefore, just increase DRM_OBJECT_MAX_PROPERTY to a number (64) that accomodates these new properties and gives some room for others, avoiding change this number everytime we add a new KMS property. Reviewed-by: Harry Wentland Reviewed-by: Simon Ser Signed-off-by: Melissa Wen Reviewed-by: Abhinav Kumar Acked-by: Maxime Ripard Signed-off-by: Alex Deucher --- include/drm/drm_mode_object.h | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/include/drm/drm_mode_object.h b/include/drm/drm_mode_object.h index 912f1e415685..08d7a7f0188f 100644 --- a/include/drm/drm_mode_object.h +++ b/include/drm/drm_mode_object.h @@ -60,7 +60,7 @@ struct drm_mode_object { void (*free_cb)(struct kref *kref); }; -#define DRM_OBJECT_MAX_PROPERTY 24 +#define DRM_OBJECT_MAX_PROPERTY 64 /** * struct drm_object_properties - property tracking for &drm_mode_object */ From 601603105325ad4ec62db95c9bc428202ece2c8f Mon Sep 17 00:00:00 2001 From: Melissa Wen Date: Thu, 16 Nov 2023 18:57:42 -0100 Subject: [PATCH 06/92] drm/drm_property: make replace_property_blob_from_id a DRM helper Place it in drm_property where drm_property_replace_blob and drm_property_lookup_blob live. Then we can use the DRM helper for driver-specific KMS properties too. Reviewed-by: Harry Wentland Reviewed-by: Liviu Dudau Signed-off-by: Melissa Wen Acked-by: Maxime Ripard Signed-off-by: Alex Deucher --- drivers/gpu/drm/arm/malidp_crtc.c | 2 +- drivers/gpu/drm/drm_atomic_uapi.c | 52 +++------------------------ drivers/gpu/drm/drm_property.c | 59 +++++++++++++++++++++++++++++++ include/drm/drm_property.h | 6 ++++ 4 files changed, 71 insertions(+), 48 deletions(-) diff --git a/drivers/gpu/drm/arm/malidp_crtc.c b/drivers/gpu/drm/arm/malidp_crtc.c index dc01c43f6193..d72c22dcf685 100644 --- a/drivers/gpu/drm/arm/malidp_crtc.c +++ b/drivers/gpu/drm/arm/malidp_crtc.c @@ -221,7 +221,7 @@ static int malidp_crtc_atomic_check_ctm(struct drm_crtc *crtc, /* * The size of the ctm is checked in - * drm_atomic_replace_property_blob_from_id. + * drm_property_replace_blob_from_id. */ ctm = (struct drm_color_ctm *)state->ctm->data; for (i = 0; i < ARRAY_SIZE(ctm->matrix); ++i) { diff --git a/drivers/gpu/drm/drm_atomic_uapi.c b/drivers/gpu/drm/drm_atomic_uapi.c index aee4a65d4959..29d4940188d4 100644 --- a/drivers/gpu/drm/drm_atomic_uapi.c +++ b/drivers/gpu/drm/drm_atomic_uapi.c @@ -362,48 +362,6 @@ static s32 __user *get_out_fence_for_connector(struct drm_atomic_state *state, return fence_ptr; } -static int -drm_atomic_replace_property_blob_from_id(struct drm_device *dev, - struct drm_property_blob **blob, - uint64_t blob_id, - ssize_t expected_size, - ssize_t expected_elem_size, - bool *replaced) -{ - struct drm_property_blob *new_blob = NULL; - - if (blob_id != 0) { - new_blob = drm_property_lookup_blob(dev, blob_id); - if (new_blob == NULL) { - drm_dbg_atomic(dev, - "cannot find blob ID %llu\n", blob_id); - return -EINVAL; - } - - if (expected_size > 0 && - new_blob->length != expected_size) { - drm_dbg_atomic(dev, - "[BLOB:%d] length %zu different from expected %zu\n", - new_blob->base.id, new_blob->length, expected_size); - drm_property_blob_put(new_blob); - return -EINVAL; - } - if (expected_elem_size > 0 && - new_blob->length % expected_elem_size != 0) { - drm_dbg_atomic(dev, - "[BLOB:%d] length %zu not divisible by element size %zu\n", - new_blob->base.id, new_blob->length, expected_elem_size); - drm_property_blob_put(new_blob); - return -EINVAL; - } - } - - *replaced |= drm_property_replace_blob(blob, new_blob); - drm_property_blob_put(new_blob); - - return 0; -} - static int drm_atomic_crtc_set_property(struct drm_crtc *crtc, struct drm_crtc_state *state, struct drm_property *property, uint64_t val) @@ -424,7 +382,7 @@ static int drm_atomic_crtc_set_property(struct drm_crtc *crtc, } else if (property == config->prop_vrr_enabled) { state->vrr_enabled = val; } else if (property == config->degamma_lut_property) { - ret = drm_atomic_replace_property_blob_from_id(dev, + ret = drm_property_replace_blob_from_id(dev, &state->degamma_lut, val, -1, sizeof(struct drm_color_lut), @@ -432,7 +390,7 @@ static int drm_atomic_crtc_set_property(struct drm_crtc *crtc, state->color_mgmt_changed |= replaced; return ret; } else if (property == config->ctm_property) { - ret = drm_atomic_replace_property_blob_from_id(dev, + ret = drm_property_replace_blob_from_id(dev, &state->ctm, val, sizeof(struct drm_color_ctm), -1, @@ -440,7 +398,7 @@ static int drm_atomic_crtc_set_property(struct drm_crtc *crtc, state->color_mgmt_changed |= replaced; return ret; } else if (property == config->gamma_lut_property) { - ret = drm_atomic_replace_property_blob_from_id(dev, + ret = drm_property_replace_blob_from_id(dev, &state->gamma_lut, val, -1, sizeof(struct drm_color_lut), @@ -581,7 +539,7 @@ static int drm_atomic_plane_set_property(struct drm_plane *plane, } else if (property == plane->color_range_property) { state->color_range = val; } else if (property == config->prop_fb_damage_clips) { - ret = drm_atomic_replace_property_blob_from_id(dev, + ret = drm_property_replace_blob_from_id(dev, &state->fb_damage_clips, val, -1, @@ -778,7 +736,7 @@ static int drm_atomic_connector_set_property(struct drm_connector *connector, if (state->link_status != DRM_LINK_STATUS_GOOD) state->link_status = val; } else if (property == config->hdr_output_metadata_property) { - ret = drm_atomic_replace_property_blob_from_id(dev, + ret = drm_property_replace_blob_from_id(dev, &state->hdr_output_metadata, val, sizeof(struct hdr_output_metadata), -1, diff --git a/drivers/gpu/drm/drm_property.c b/drivers/gpu/drm/drm_property.c index dfec479830e4..596272149a35 100644 --- a/drivers/gpu/drm/drm_property.c +++ b/drivers/gpu/drm/drm_property.c @@ -27,6 +27,7 @@ #include #include #include +#include #include #include "drm_crtc_internal.h" @@ -751,6 +752,64 @@ bool drm_property_replace_blob(struct drm_property_blob **blob, } EXPORT_SYMBOL(drm_property_replace_blob); +/** + * drm_property_replace_blob_from_id - replace a blob property taking a reference + * @dev: DRM device + * @blob: a pointer to the member blob to be replaced + * @blob_id: the id of the new blob to replace with + * @expected_size: expected size of the blob property + * @expected_elem_size: expected size of an element in the blob property + * @replaced: if the blob was in fact replaced + * + * Look up the new blob from id, take its reference, check expected sizes of + * the blob and its element and replace the old blob by the new one. Advertise + * if the replacement operation was successful. + * + * Return: true if the blob was in fact replaced. -EINVAL if the new blob was + * not found or sizes don't match. + */ +int drm_property_replace_blob_from_id(struct drm_device *dev, + struct drm_property_blob **blob, + uint64_t blob_id, + ssize_t expected_size, + ssize_t expected_elem_size, + bool *replaced) +{ + struct drm_property_blob *new_blob = NULL; + + if (blob_id != 0) { + new_blob = drm_property_lookup_blob(dev, blob_id); + if (new_blob == NULL) { + drm_dbg_atomic(dev, + "cannot find blob ID %llu\n", blob_id); + return -EINVAL; + } + + if (expected_size > 0 && + new_blob->length != expected_size) { + drm_dbg_atomic(dev, + "[BLOB:%d] length %zu different from expected %zu\n", + new_blob->base.id, new_blob->length, expected_size); + drm_property_blob_put(new_blob); + return -EINVAL; + } + if (expected_elem_size > 0 && + new_blob->length % expected_elem_size != 0) { + drm_dbg_atomic(dev, + "[BLOB:%d] length %zu not divisible by element size %zu\n", + new_blob->base.id, new_blob->length, expected_elem_size); + drm_property_blob_put(new_blob); + return -EINVAL; + } + } + + *replaced |= drm_property_replace_blob(blob, new_blob); + drm_property_blob_put(new_blob); + + return 0; +} +EXPORT_SYMBOL(drm_property_replace_blob_from_id); + int drm_mode_getblob_ioctl(struct drm_device *dev, void *data, struct drm_file *file_priv) { diff --git a/include/drm/drm_property.h b/include/drm/drm_property.h index 65bc9710a470..082f29156b3e 100644 --- a/include/drm/drm_property.h +++ b/include/drm/drm_property.h @@ -279,6 +279,12 @@ struct drm_property_blob *drm_property_create_blob(struct drm_device *dev, const void *data); struct drm_property_blob *drm_property_lookup_blob(struct drm_device *dev, uint32_t id); +int drm_property_replace_blob_from_id(struct drm_device *dev, + struct drm_property_blob **blob, + uint64_t blob_id, + ssize_t expected_size, + ssize_t expected_elem_size, + bool *replaced); int drm_property_replace_global_blob(struct drm_device *dev, struct drm_property_blob **replace, size_t length, From 24013b9301349881c9fcd27e7edacc672e0bf6d3 Mon Sep 17 00:00:00 2001 From: Melissa Wen Date: Thu, 16 Nov 2023 18:57:43 -0100 Subject: [PATCH 07/92] drm/drm_plane: track color mgmt changes per plane We will add color mgmt properties to DRM planes in the next patches and we want to track when one of this properties change to define atomic commit behaviors. Using a similar approach from CRTC color props, we set a color_mgmt_changed boolean whenever a plane color prop changes. Reviewed-by: Harry Wentland Signed-off-by: Melissa Wen Acked-by: Maxime Ripard Signed-off-by: Alex Deucher --- drivers/gpu/drm/drm_atomic.c | 1 + drivers/gpu/drm/drm_atomic_state_helper.c | 1 + include/drm/drm_plane.h | 7 +++++++ 3 files changed, 9 insertions(+) diff --git a/drivers/gpu/drm/drm_atomic.c b/drivers/gpu/drm/drm_atomic.c index c31fc0b48c31..a91737adf8e7 100644 --- a/drivers/gpu/drm/drm_atomic.c +++ b/drivers/gpu/drm/drm_atomic.c @@ -733,6 +733,7 @@ static void drm_atomic_plane_print_state(struct drm_printer *p, drm_get_color_encoding_name(state->color_encoding)); drm_printf(p, "\tcolor-range=%s\n", drm_get_color_range_name(state->color_range)); + drm_printf(p, "\tcolor_mgmt_changed=%d\n", state->color_mgmt_changed); if (plane->funcs->atomic_print_state) plane->funcs->atomic_print_state(p, state); diff --git a/drivers/gpu/drm/drm_atomic_state_helper.c b/drivers/gpu/drm/drm_atomic_state_helper.c index 54975de44a0e..519228eb1095 100644 --- a/drivers/gpu/drm/drm_atomic_state_helper.c +++ b/drivers/gpu/drm/drm_atomic_state_helper.c @@ -352,6 +352,7 @@ void __drm_atomic_helper_plane_duplicate_state(struct drm_plane *plane, state->fence = NULL; state->commit = NULL; state->fb_damage_clips = NULL; + state->color_mgmt_changed = false; } EXPORT_SYMBOL(__drm_atomic_helper_plane_duplicate_state); diff --git a/include/drm/drm_plane.h b/include/drm/drm_plane.h index c6565a6f9324..641fe298052d 100644 --- a/include/drm/drm_plane.h +++ b/include/drm/drm_plane.h @@ -251,6 +251,13 @@ struct drm_plane_state { /** @state: backpointer to global drm_atomic_state */ struct drm_atomic_state *state; + + /** + * @color_mgmt_changed: Color management properties have changed. Used + * by the atomic helpers and drivers to steer the atomic commit control + * flow. + */ + bool color_mgmt_changed : 1; }; static inline struct drm_rect From 9342a9ae54ef299ffe5e4ce3d0be6a4da5edba0e Mon Sep 17 00:00:00 2001 From: Melissa Wen Date: Thu, 16 Nov 2023 18:57:44 -0100 Subject: [PATCH 08/92] drm/amd/display: add driver-specific property for plane degamma LUT Hook up driver-specific atomic operations for managing AMD color properties. Create AMD driver-specific color management properties and attach them according to HW capabilities defined by `struct dc_color_caps`. First add plane degamma LUT properties that means user-blob and its size. We will add more plane color properties in the next patches. In addition, we define AMD_PRIVATE_COLOR to guard these driver-specific plane properties. Plane degamma can be used to linearize input space for arithmetical operations that are more accurate when applied in linear color. v2: - update degamma LUT prop description - move private color operations from amdgpu_display to amdgpu_dm_color v5: - get degamma blob correctly (Joshua) Reviewed-by: Harry Wentland Co-developed-by: Joshua Ashton Signed-off-by: Joshua Ashton Signed-off-by: Melissa Wen Signed-off-by: Alex Deucher --- drivers/gpu/drm/amd/amdgpu/amdgpu_mode.h | 12 +++ .../gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.c | 5 ++ .../gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.h | 11 +++ .../amd/display/amdgpu_dm/amdgpu_dm_color.c | 25 ++++++ .../amd/display/amdgpu_dm/amdgpu_dm_plane.c | 85 +++++++++++++++++++ 5 files changed, 138 insertions(+) diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_mode.h b/drivers/gpu/drm/amd/amdgpu/amdgpu_mode.h index d8083972e393..a63b01aa230d 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_mode.h +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_mode.h @@ -343,6 +343,18 @@ struct amdgpu_mode_info { int disp_priority; const struct amdgpu_display_funcs *funcs; const enum drm_plane_type *plane_type; + + /* Driver-private color mgmt props */ + + /* @plane_degamma_lut_property: Plane property to set a degamma LUT to + * convert encoded values to light linear values before sampling or + * blending. + */ + struct drm_property *plane_degamma_lut_property; + /* @plane_degamma_lut_size_property: Plane property to define the max + * size of degamma LUT as supported by the driver (read-only). + */ + struct drm_property *plane_degamma_lut_size_property; }; #define AMDGPU_MAX_BL_LEVEL 0xFF 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 8e6fe9860cd6..80346bebb7fc 100644 --- a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.c +++ b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.c @@ -4067,6 +4067,11 @@ static int amdgpu_dm_mode_config_init(struct amdgpu_device *adev) return r; } +#ifdef AMD_PRIVATE_COLOR + if (amdgpu_dm_create_color_properties(adev)) + return -ENOMEM; +#endif + r = amdgpu_dm_audio_init(adev); if (r) { dc_release_state(state->context); 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 0a9974017eba..446a3962d9c9 100644 --- a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.h +++ b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.h @@ -727,6 +727,16 @@ extern const struct amdgpu_ip_block_version dm_ip_block; struct dm_plane_state { struct drm_plane_state base; struct dc_plane_state *dc_state; + + /* Plane color mgmt */ + /** + * @degamma_lut: + * + * 1D LUT for mapping framebuffer/plane pixel data before sampling or + * blending operations. It's usually applied to linearize input space. + * The blob (if not NULL) is an array of &struct drm_color_lut. + */ + struct drm_property_blob *degamma_lut; }; struct dm_crtc_state { @@ -817,6 +827,7 @@ void amdgpu_dm_trigger_timing_sync(struct drm_device *dev); #define MAX_COLOR_LEGACY_LUT_ENTRIES 256 void amdgpu_dm_init_color_mod(void); +int amdgpu_dm_create_color_properties(struct amdgpu_device *adev); int amdgpu_dm_verify_lut_sizes(const struct drm_crtc_state *crtc_state); int amdgpu_dm_update_crtc_color_mgmt(struct dm_crtc_state *crtc); int amdgpu_dm_update_plane_color_mgmt(struct dm_crtc_state *crtc, diff --git a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_color.c b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_color.c index a4cb23d059bd..421af919387f 100644 --- a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_color.c +++ b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_color.c @@ -84,6 +84,31 @@ void amdgpu_dm_init_color_mod(void) setup_x_points_distribution(); } +#ifdef AMD_PRIVATE_COLOR +int +amdgpu_dm_create_color_properties(struct amdgpu_device *adev) +{ + struct drm_property *prop; + + prop = drm_property_create(adev_to_drm(adev), + DRM_MODE_PROP_BLOB, + "AMD_PLANE_DEGAMMA_LUT", 0); + if (!prop) + return -ENOMEM; + adev->mode_info.plane_degamma_lut_property = prop; + + prop = drm_property_create_range(adev_to_drm(adev), + DRM_MODE_PROP_IMMUTABLE, + "AMD_PLANE_DEGAMMA_LUT_SIZE", + 0, UINT_MAX); + if (!prop) + return -ENOMEM; + adev->mode_info.plane_degamma_lut_size_property = prop; + + return 0; +} +#endif + /** * __extract_blob_lut - Extracts the DRM lut and lut size from a blob. * @blob: DRM color mgmt property blob diff --git a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_plane.c b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_plane.c index 116121e647ca..b1e50cfa9f83 100644 --- a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_plane.c +++ b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_plane.c @@ -1357,6 +1357,10 @@ static struct drm_plane_state *amdgpu_dm_plane_drm_plane_duplicate_state(struct dc_plane_state_retain(dm_plane_state->dc_state); } + if (old_dm_plane_state->degamma_lut) + dm_plane_state->degamma_lut = + drm_property_blob_get(old_dm_plane_state->degamma_lut); + return &dm_plane_state->base; } @@ -1424,12 +1428,86 @@ static void amdgpu_dm_plane_drm_plane_destroy_state(struct drm_plane *plane, { struct dm_plane_state *dm_plane_state = to_dm_plane_state(state); + if (dm_plane_state->degamma_lut) + drm_property_blob_put(dm_plane_state->degamma_lut); + if (dm_plane_state->dc_state) dc_plane_state_release(dm_plane_state->dc_state); drm_atomic_helper_plane_destroy_state(plane, state); } +#ifdef AMD_PRIVATE_COLOR +static void +dm_atomic_plane_attach_color_mgmt_properties(struct amdgpu_display_manager *dm, + struct drm_plane *plane) +{ + struct amdgpu_mode_info mode_info = dm->adev->mode_info; + struct dpp_color_caps dpp_color_caps = dm->dc->caps.color.dpp; + + /* Check HW color pipeline capabilities on DPP block (pre-blending) + * before exposing related properties. + */ + if (dpp_color_caps.dgam_ram || dpp_color_caps.gamma_corr) { + drm_object_attach_property(&plane->base, + mode_info.plane_degamma_lut_property, + 0); + drm_object_attach_property(&plane->base, + mode_info.plane_degamma_lut_size_property, + MAX_COLOR_LUT_ENTRIES); + } +} + +static int +dm_atomic_plane_set_property(struct drm_plane *plane, + struct drm_plane_state *state, + struct drm_property *property, + uint64_t val) +{ + struct dm_plane_state *dm_plane_state = to_dm_plane_state(state); + struct amdgpu_device *adev = drm_to_adev(plane->dev); + bool replaced = false; + int ret; + + if (property == adev->mode_info.plane_degamma_lut_property) { + ret = drm_property_replace_blob_from_id(plane->dev, + &dm_plane_state->degamma_lut, + val, -1, + sizeof(struct drm_color_lut), + &replaced); + dm_plane_state->base.color_mgmt_changed |= replaced; + return ret; + } else { + drm_dbg_atomic(plane->dev, + "[PLANE:%d:%s] unknown property [PROP:%d:%s]]\n", + plane->base.id, plane->name, + property->base.id, property->name); + return -EINVAL; + } + + return 0; +} + +static int +dm_atomic_plane_get_property(struct drm_plane *plane, + const struct drm_plane_state *state, + struct drm_property *property, + uint64_t *val) +{ + struct dm_plane_state *dm_plane_state = to_dm_plane_state(state); + struct amdgpu_device *adev = drm_to_adev(plane->dev); + + if (property == adev->mode_info.plane_degamma_lut_property) { + *val = (dm_plane_state->degamma_lut) ? + dm_plane_state->degamma_lut->base.id : 0; + } else { + return -EINVAL; + } + + return 0; +} +#endif + static const struct drm_plane_funcs dm_plane_funcs = { .update_plane = drm_atomic_helper_update_plane, .disable_plane = drm_atomic_helper_disable_plane, @@ -1438,6 +1516,10 @@ static const struct drm_plane_funcs dm_plane_funcs = { .atomic_duplicate_state = amdgpu_dm_plane_drm_plane_duplicate_state, .atomic_destroy_state = amdgpu_dm_plane_drm_plane_destroy_state, .format_mod_supported = amdgpu_dm_plane_format_mod_supported, +#ifdef AMD_PRIVATE_COLOR + .atomic_set_property = dm_atomic_plane_set_property, + .atomic_get_property = dm_atomic_plane_get_property, +#endif }; int amdgpu_dm_plane_init(struct amdgpu_display_manager *dm, @@ -1517,6 +1599,9 @@ int amdgpu_dm_plane_init(struct amdgpu_display_manager *dm, drm_plane_helper_add(plane, &dm_plane_helper_funcs); +#ifdef AMD_PRIVATE_COLOR + dm_atomic_plane_attach_color_mgmt_properties(dm, plane); +#endif /* Create (reset) the plane state */ if (plane->funcs->reset) plane->funcs->reset(plane); From ed342a2e78c4e4a8d82c2d19c95e8a3eb092c0d0 Mon Sep 17 00:00:00 2001 From: Lijo Lazar Date: Fri, 1 Dec 2023 17:13:46 +0530 Subject: [PATCH 09/92] drm/amdgpu: Use the right method to get IP version Replace direct usage of adev->ip_versions with amdgpu_ip_version. Signed-off-by: Lijo Lazar Reviewed-by: Hawking Zhang Reviewed-by: Alex Deucher Signed-off-by: Alex Deucher --- drivers/gpu/drm/amd/amdgpu/amdgpu_device.c | 2 +- drivers/gpu/drm/amd/amdgpu/amdgpu_virt.c | 3 ++- drivers/gpu/drm/amd/amdgpu/gmc_v9_0.c | 2 +- drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.c | 7 ++++--- drivers/gpu/drm/amd/pm/swsmu/smu14/smu_v14_0.c | 6 +++--- 5 files changed, 11 insertions(+), 9 deletions(-) diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_device.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_device.c index f58fb719292d..85ed0d66a029 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_device.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_device.c @@ -1599,7 +1599,7 @@ bool amdgpu_device_seamless_boot_supported(struct amdgpu_device *adev) if (adev->mman.keep_stolen_vga_memory) return false; - return adev->ip_versions[DCE_HWIP][0] >= IP_VERSION(3, 0, 0); + return amdgpu_ip_version(adev, DCE_HWIP, 0) >= IP_VERSION(3, 0, 0); } /* diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_virt.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_virt.c index 3a632c3b1a2c..0dcff2889e25 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_virt.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_virt.c @@ -1099,7 +1099,8 @@ bool amdgpu_sriov_xnack_support(struct amdgpu_device *adev) { bool xnack_mode = true; - if (amdgpu_sriov_vf(adev) && adev->ip_versions[GC_HWIP][0] == IP_VERSION(9, 4, 2)) + if (amdgpu_sriov_vf(adev) && + amdgpu_ip_version(adev, GC_HWIP, 0) == IP_VERSION(9, 4, 2)) xnack_mode = false; return xnack_mode; diff --git a/drivers/gpu/drm/amd/amdgpu/gmc_v9_0.c b/drivers/gpu/drm/amd/amdgpu/gmc_v9_0.c index 2ac5820e9c92..473a774294ce 100644 --- a/drivers/gpu/drm/amd/amdgpu/gmc_v9_0.c +++ b/drivers/gpu/drm/amd/amdgpu/gmc_v9_0.c @@ -883,7 +883,7 @@ static void gmc_v9_0_flush_gpu_tlb(struct amdgpu_device *adev, uint32_t vmid, * GRBM interface. */ if ((vmhub == AMDGPU_GFXHUB(0)) && - (adev->ip_versions[GC_HWIP][0] < IP_VERSION(9, 4, 2))) + (amdgpu_ip_version(adev, GC_HWIP, 0) < IP_VERSION(9, 4, 2))) RREG32_NO_KIQ(req); for (j = 0; j < adev->usec_timeout; j++) { 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 80346bebb7fc..e819d05755b3 100644 --- a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.c +++ b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.c @@ -1712,7 +1712,7 @@ static int amdgpu_dm_init(struct amdgpu_device *adev) init_data.clk_reg_offsets = adev->reg_offset[CLK_HWIP][0]; /* Enable DWB for tested platforms only */ - if (adev->ip_versions[DCE_HWIP][0] >= IP_VERSION(3, 0, 0)) + if (amdgpu_ip_version(adev, DCE_HWIP, 0) >= IP_VERSION(3, 0, 0)) init_data.num_virtual_links = 1; INIT_LIST_HEAD(&adev->dm.da_list); @@ -8939,7 +8939,7 @@ static void dm_set_writeback(struct amdgpu_display_manager *dm, } wb_info->mcif_buf_params.p_vmid = 1; - if (adev->ip_versions[DCE_HWIP][0] >= IP_VERSION(3, 0, 0)) { + if (amdgpu_ip_version(adev, DCE_HWIP, 0) >= IP_VERSION(3, 0, 0)) { wb_info->mcif_warmup_params.start_address.quad_part = afb->address; wb_info->mcif_warmup_params.region_size = wb_info->mcif_buf_params.luma_pitch * wb_info->dwb_params.dest_height; @@ -9891,7 +9891,8 @@ static bool should_reset_plane(struct drm_atomic_state *state, * TODO: Remove this hack for all asics once it proves that the * fast updates works fine on DCN3.2+. */ - if (adev->ip_versions[DCE_HWIP][0] < IP_VERSION(3, 2, 0) && state->allow_modeset) + if (amdgpu_ip_version(adev, DCE_HWIP, 0) < IP_VERSION(3, 2, 0) && + state->allow_modeset) return true; /* Exit early if we know that we're adding or removing the plane. */ diff --git a/drivers/gpu/drm/amd/pm/swsmu/smu14/smu_v14_0.c b/drivers/gpu/drm/amd/pm/swsmu/smu14/smu_v14_0.c index d8f8ad0e7137..4894f7ee737b 100644 --- a/drivers/gpu/drm/amd/pm/swsmu/smu14/smu_v14_0.c +++ b/drivers/gpu/drm/amd/pm/swsmu/smu14/smu_v14_0.c @@ -224,7 +224,7 @@ int smu_v14_0_check_fw_version(struct smu_context *smu) if (smu->is_apu) adev->pm.fw_version = smu_version; - switch (adev->ip_versions[MP1_HWIP][0]) { + switch (amdgpu_ip_version(adev, MP1_HWIP, 0)) { case IP_VERSION(14, 0, 2): smu->smc_driver_if_version = SMU14_DRIVER_IF_VERSION_SMU_V14_0_2; break; @@ -235,7 +235,7 @@ int smu_v14_0_check_fw_version(struct smu_context *smu) break; default: dev_err(adev->dev, "smu unsupported IP version: 0x%x.\n", - adev->ip_versions[MP1_HWIP][0]); + amdgpu_ip_version(adev, MP1_HWIP, 0)); smu->smc_driver_if_version = SMU14_DRIVER_IF_VERSION_INV; break; } @@ -733,7 +733,7 @@ int smu_v14_0_gfx_off_control(struct smu_context *smu, bool enable) int ret = 0; struct amdgpu_device *adev = smu->adev; - switch (adev->ip_versions[MP1_HWIP][0]) { + switch (amdgpu_ip_version(adev, MP1_HWIP, 0)) { case IP_VERSION(14, 0, 2): case IP_VERSION(14, 0, 0): if (!(adev->pm.pp_feature & PP_GFXOFF_MASK)) From b70aed8f5d7686c4343f9ae618287404fa5a703e Mon Sep 17 00:00:00 2001 From: Saleemkhan Jamadar Date: Tue, 28 Nov 2023 17:02:06 +0530 Subject: [PATCH 10/92] drm/amdgpu/jpeg: configure doorbell for each playback Doorbell is configured during start of each playback. v1 - add comment for the doorbell programming change Signed-off-by: Saleemkhan Jamadar Acked-by: Leo Liu Reviewed-by: Veerabadhran Gopalakrishnan Signed-off-by: Alex Deucher --- drivers/gpu/drm/amd/amdgpu/jpeg_v4_0_5.c | 15 ++++++++------- 1 file changed, 8 insertions(+), 7 deletions(-) diff --git a/drivers/gpu/drm/amd/amdgpu/jpeg_v4_0_5.c b/drivers/gpu/drm/amd/amdgpu/jpeg_v4_0_5.c index 9df011323d4b..6ede85b28cc8 100644 --- a/drivers/gpu/drm/amd/amdgpu/jpeg_v4_0_5.c +++ b/drivers/gpu/drm/amd/amdgpu/jpeg_v4_0_5.c @@ -155,13 +155,6 @@ static int jpeg_v4_0_5_hw_init(void *handle) struct amdgpu_ring *ring = adev->jpeg.inst->ring_dec; int r; - adev->nbio.funcs->vcn_doorbell_range(adev, ring->use_doorbell, - (adev->doorbell_index.vcn.vcn_ring0_1 << 1), 0); - - WREG32_SOC15(VCN, 0, regVCN_JPEG_DB_CTRL, - ring->doorbell_index << VCN_JPEG_DB_CTRL__OFFSET__SHIFT | - VCN_JPEG_DB_CTRL__EN_MASK); - r = amdgpu_ring_test_helper(ring); if (r) return r; @@ -336,6 +329,14 @@ static int jpeg_v4_0_5_start(struct amdgpu_device *adev) if (adev->pm.dpm_enabled) amdgpu_dpm_enable_jpeg(adev, true); + /* doorbell programming is done for every playback */ + adev->nbio.funcs->vcn_doorbell_range(adev, ring->use_doorbell, + (adev->doorbell_index.vcn.vcn_ring0_1 << 1), 0); + + WREG32_SOC15(VCN, 0, regVCN_JPEG_DB_CTRL, + ring->doorbell_index << VCN_JPEG_DB_CTRL__OFFSET__SHIFT | + VCN_JPEG_DB_CTRL__EN_MASK); + /* disable power gating */ r = jpeg_v4_0_5_disable_static_power_gating(adev); if (r) From 31e6af1ff77533df2e8e006974a9b57adece0488 Mon Sep 17 00:00:00 2001 From: Ma Jun Date: Fri, 3 Nov 2023 13:41:42 +0800 Subject: [PATCH 11/92] drm/amd/pm: Remove redundant function members of pptable_funcs Remove redundant functions members of pptable_funcs and change the function type as static because they are not called by other files. Signed-off-by: Ma Jun Reviewed-by: Yang Wang Signed-off-by: Alex Deucher --- drivers/gpu/drm/amd/pm/swsmu/inc/smu_v13_0.h | 4 -- .../gpu/drm/amd/pm/swsmu/smu11/arcturus_ppt.c | 2 - .../gpu/drm/amd/pm/swsmu/smu11/navi10_ppt.c | 2 - .../amd/pm/swsmu/smu11/sienna_cichlid_ppt.c | 2 - .../gpu/drm/amd/pm/swsmu/smu13/smu_v13_0.c | 41 +++++++++---------- .../drm/amd/pm/swsmu/smu13/smu_v13_0_0_ppt.c | 2 - .../drm/amd/pm/swsmu/smu13/smu_v13_0_7_ppt.c | 2 - 7 files changed, 20 insertions(+), 35 deletions(-) diff --git a/drivers/gpu/drm/amd/pm/swsmu/inc/smu_v13_0.h b/drivers/gpu/drm/amd/pm/swsmu/inc/smu_v13_0.h index 48f5926d8153..5e7b8f29fecc 100644 --- a/drivers/gpu/drm/amd/pm/swsmu/inc/smu_v13_0.h +++ b/drivers/gpu/drm/amd/pm/swsmu/inc/smu_v13_0.h @@ -212,10 +212,6 @@ int smu_v13_0_get_max_sustainable_clocks_by_dc(struct smu_context *smu, bool smu_v13_0_baco_is_support(struct smu_context *smu); -enum smu_baco_state smu_v13_0_baco_get_state(struct smu_context *smu); - -int smu_v13_0_baco_set_state(struct smu_context *smu, enum smu_baco_state state); - int smu_v13_0_baco_enter(struct smu_context *smu); int smu_v13_0_baco_exit(struct smu_context *smu); diff --git a/drivers/gpu/drm/amd/pm/swsmu/smu11/arcturus_ppt.c b/drivers/gpu/drm/amd/pm/swsmu/smu11/arcturus_ppt.c index 2cb6b68222ba..4cd43bbec910 100644 --- a/drivers/gpu/drm/amd/pm/swsmu/smu11/arcturus_ppt.c +++ b/drivers/gpu/drm/amd/pm/swsmu/smu11/arcturus_ppt.c @@ -2407,8 +2407,6 @@ static const struct pptable_funcs arcturus_ppt_funcs = { .set_azalia_d3_pme = smu_v11_0_set_azalia_d3_pme, .get_max_sustainable_clocks_by_dc = smu_v11_0_get_max_sustainable_clocks_by_dc, .baco_is_support = smu_v11_0_baco_is_support, - .baco_get_state = smu_v11_0_baco_get_state, - .baco_set_state = smu_v11_0_baco_set_state, .baco_enter = smu_v11_0_baco_enter, .baco_exit = smu_v11_0_baco_exit, .get_dpm_ultimate_freq = smu_v11_0_get_dpm_ultimate_freq, diff --git a/drivers/gpu/drm/amd/pm/swsmu/smu11/navi10_ppt.c b/drivers/gpu/drm/amd/pm/swsmu/smu11/navi10_ppt.c index a38233cc5b7f..8d1d29ffb0f1 100644 --- a/drivers/gpu/drm/amd/pm/swsmu/smu11/navi10_ppt.c +++ b/drivers/gpu/drm/amd/pm/swsmu/smu11/navi10_ppt.c @@ -3537,8 +3537,6 @@ static const struct pptable_funcs navi10_ppt_funcs = { .set_azalia_d3_pme = smu_v11_0_set_azalia_d3_pme, .get_max_sustainable_clocks_by_dc = smu_v11_0_get_max_sustainable_clocks_by_dc, .baco_is_support = smu_v11_0_baco_is_support, - .baco_get_state = smu_v11_0_baco_get_state, - .baco_set_state = smu_v11_0_baco_set_state, .baco_enter = navi10_baco_enter, .baco_exit = navi10_baco_exit, .get_dpm_ultimate_freq = smu_v11_0_get_dpm_ultimate_freq, diff --git a/drivers/gpu/drm/amd/pm/swsmu/smu11/sienna_cichlid_ppt.c b/drivers/gpu/drm/amd/pm/swsmu/smu11/sienna_cichlid_ppt.c index 1de9f8b5cc5f..21fc033528fa 100644 --- a/drivers/gpu/drm/amd/pm/swsmu/smu11/sienna_cichlid_ppt.c +++ b/drivers/gpu/drm/amd/pm/swsmu/smu11/sienna_cichlid_ppt.c @@ -4428,8 +4428,6 @@ static const struct pptable_funcs sienna_cichlid_ppt_funcs = { .set_azalia_d3_pme = smu_v11_0_set_azalia_d3_pme, .get_max_sustainable_clocks_by_dc = smu_v11_0_get_max_sustainable_clocks_by_dc, .baco_is_support = smu_v11_0_baco_is_support, - .baco_get_state = smu_v11_0_baco_get_state, - .baco_set_state = smu_v11_0_baco_set_state, .baco_enter = sienna_cichlid_baco_enter, .baco_exit = sienna_cichlid_baco_exit, .mode1_reset_is_support = sienna_cichlid_is_mode1_reset_supported, diff --git a/drivers/gpu/drm/amd/pm/swsmu/smu13/smu_v13_0.c b/drivers/gpu/drm/amd/pm/swsmu/smu13/smu_v13_0.c index 86fc7273d588..2877e5da5b5a 100644 --- a/drivers/gpu/drm/amd/pm/swsmu/smu13/smu_v13_0.c +++ b/drivers/gpu/drm/amd/pm/swsmu/smu13/smu_v13_0.c @@ -2221,33 +2221,14 @@ static int smu_v13_0_baco_set_armd3_sequence(struct smu_context *smu, return 0; } -bool smu_v13_0_baco_is_support(struct smu_context *smu) -{ - struct smu_baco_context *smu_baco = &smu->smu_baco; - - if (amdgpu_sriov_vf(smu->adev) || - !smu_baco->platform_support) - return false; - - /* return true if ASIC is in BACO state already */ - if (smu_v13_0_baco_get_state(smu) == SMU_BACO_STATE_ENTER) - return true; - - if (smu_cmn_feature_is_supported(smu, SMU_FEATURE_BACO_BIT) && - !smu_cmn_feature_is_enabled(smu, SMU_FEATURE_BACO_BIT)) - return false; - - return true; -} - -enum smu_baco_state smu_v13_0_baco_get_state(struct smu_context *smu) +static enum smu_baco_state smu_v13_0_baco_get_state(struct smu_context *smu) { struct smu_baco_context *smu_baco = &smu->smu_baco; return smu_baco->state; } -int smu_v13_0_baco_set_state(struct smu_context *smu, +static int smu_v13_0_baco_set_state(struct smu_context *smu, enum smu_baco_state state) { struct smu_baco_context *smu_baco = &smu->smu_baco; @@ -2281,6 +2262,24 @@ int smu_v13_0_baco_set_state(struct smu_context *smu, return ret; } +bool smu_v13_0_baco_is_support(struct smu_context *smu) +{ + struct smu_baco_context *smu_baco = &smu->smu_baco; + + if (amdgpu_sriov_vf(smu->adev) || !smu_baco->platform_support) + return false; + + /* return true if ASIC is in BACO state already */ + if (smu_v13_0_baco_get_state(smu) == SMU_BACO_STATE_ENTER) + return true; + + if (smu_cmn_feature_is_supported(smu, SMU_FEATURE_BACO_BIT) && + !smu_cmn_feature_is_enabled(smu, SMU_FEATURE_BACO_BIT)) + return false; + + return true; +} + int smu_v13_0_baco_enter(struct smu_context *smu) { struct smu_baco_context *smu_baco = &smu->smu_baco; diff --git a/drivers/gpu/drm/amd/pm/swsmu/smu13/smu_v13_0_0_ppt.c b/drivers/gpu/drm/amd/pm/swsmu/smu13/smu_v13_0_0_ppt.c index a4debd7ab8bd..2d25a9e47651 100644 --- a/drivers/gpu/drm/amd/pm/swsmu/smu13/smu_v13_0_0_ppt.c +++ b/drivers/gpu/drm/amd/pm/swsmu/smu13/smu_v13_0_0_ppt.c @@ -3003,8 +3003,6 @@ static const struct pptable_funcs smu_v13_0_0_ppt_funcs = { .deep_sleep_control = smu_v13_0_deep_sleep_control, .gfx_ulv_control = smu_v13_0_gfx_ulv_control, .baco_is_support = smu_v13_0_baco_is_support, - .baco_get_state = smu_v13_0_baco_get_state, - .baco_set_state = smu_v13_0_baco_set_state, .baco_enter = smu_v13_0_baco_enter, .baco_exit = smu_v13_0_baco_exit, .mode1_reset_is_support = smu_v13_0_0_is_mode1_reset_supported, diff --git a/drivers/gpu/drm/amd/pm/swsmu/smu13/smu_v13_0_7_ppt.c b/drivers/gpu/drm/amd/pm/swsmu/smu13/smu_v13_0_7_ppt.c index f5596f031d00..2909ec06b1cb 100644 --- a/drivers/gpu/drm/amd/pm/swsmu/smu13/smu_v13_0_7_ppt.c +++ b/drivers/gpu/drm/amd/pm/swsmu/smu13/smu_v13_0_7_ppt.c @@ -2594,8 +2594,6 @@ static const struct pptable_funcs smu_v13_0_7_ppt_funcs = { .get_pp_feature_mask = smu_cmn_get_pp_feature_mask, .set_pp_feature_mask = smu_cmn_set_pp_feature_mask, .baco_is_support = smu_v13_0_baco_is_support, - .baco_get_state = smu_v13_0_baco_get_state, - .baco_set_state = smu_v13_0_baco_set_state, .baco_enter = smu_v13_0_baco_enter, .baco_exit = smu_v13_0_baco_exit, .mode1_reset_is_support = smu_v13_0_7_is_mode1_reset_supported, From 9a10bd0df618f500ca526cf99f42504900020c2c Mon Sep 17 00:00:00 2001 From: Joshua Aberback Date: Wed, 6 Dec 2023 14:52:22 -0500 Subject: [PATCH 12/92] drm/amd/display: Remove minor revision 5 until proper parser is ready Reviewed-by: Dillon Varone Acked-by: Aurabindo Pillai Signed-off-by: Joshua Aberback Signed-off-by: Alex Deucher --- drivers/gpu/drm/amd/display/dc/bios/bios_parser2.c | 2 -- 1 file changed, 2 deletions(-) diff --git a/drivers/gpu/drm/amd/display/dc/bios/bios_parser2.c b/drivers/gpu/drm/amd/display/dc/bios/bios_parser2.c index aef964f1bcbe..875a064bb9a5 100644 --- a/drivers/gpu/drm/amd/display/dc/bios/bios_parser2.c +++ b/drivers/gpu/drm/amd/display/dc/bios/bios_parser2.c @@ -1747,7 +1747,6 @@ static enum bp_result bios_parser_get_firmware_info( result = get_firmware_info_v3_2(bp, info); break; case 4: - case 5: result = get_firmware_info_v3_4(bp, info); break; default: @@ -2398,7 +2397,6 @@ static enum bp_result get_vram_info_v30( return result; } - /* * get_integrated_info_v11 * From 7f9b4fb450a65a46df3d454a53836cad7e1c79c6 Mon Sep 17 00:00:00 2001 From: Aurabindo Pillai Date: Wed, 6 Dec 2023 14:52:23 -0500 Subject: [PATCH 13/92] drm/amd/display: Use explicit size for types in DCCG's struct dp_dto_params Reviewed-by: Alvin Lee Acked-by: Aurabindo Pillai Signed-off-by: Aurabindo Pillai Signed-off-by: Alex Deucher --- drivers/gpu/drm/amd/display/dc/inc/hw/dccg.h | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/drivers/gpu/drm/amd/display/dc/inc/hw/dccg.h b/drivers/gpu/drm/amd/display/dc/inc/hw/dccg.h index 6b44557fcb1a..b9a06bf84cc9 100644 --- a/drivers/gpu/drm/amd/display/dc/inc/hw/dccg.h +++ b/drivers/gpu/drm/amd/display/dc/inc/hw/dccg.h @@ -59,8 +59,8 @@ enum dentist_dispclk_change_mode { struct dp_dto_params { int otg_inst; enum signal_type signal; - long long pixclk_hz; - long long refclk_hz; + uint64_t pixclk_hz; + uint64_t refclk_hz; }; enum pixel_rate_div { From af68153ffe8c4f778ba9cbe1d1725a939ab94576 Mon Sep 17 00:00:00 2001 From: Ran Shi Date: Wed, 6 Dec 2023 14:52:24 -0500 Subject: [PATCH 14/92] drm/amd/display: allow DP40 cables to do UHBR13.5 why: With DP2.1a expansion we are allowing DP40 cables to do UHBR13.5 how: Assume UHBR10 means UHBR13.5 also for unknown cable type and passive cable type. Reviewed-by: George Shen Acked-by: Aurabindo Pillai Signed-off-by: Ran Shi Signed-off-by: Alex Deucher --- .../display/dc/link/protocols/link_dp_capability.c | 14 ++++++++++---- 1 file changed, 10 insertions(+), 4 deletions(-) diff --git a/drivers/gpu/drm/amd/display/dc/link/protocols/link_dp_capability.c b/drivers/gpu/drm/amd/display/dc/link/protocols/link_dp_capability.c index db87aa7b5c90..3c5334cdb3fb 100644 --- a/drivers/gpu/drm/amd/display/dc/link/protocols/link_dp_capability.c +++ b/drivers/gpu/drm/amd/display/dc/link/protocols/link_dp_capability.c @@ -412,12 +412,18 @@ static enum dc_link_rate get_cable_max_link_rate(struct dc_link *link) { enum dc_link_rate cable_max_link_rate = LINK_RATE_UNKNOWN; - if (link->dpcd_caps.cable_id.bits.UHBR10_20_CAPABILITY & DP_UHBR20) + if (link->dpcd_caps.cable_id.bits.UHBR10_20_CAPABILITY & DP_UHBR20) { cable_max_link_rate = LINK_RATE_UHBR20; - else if (link->dpcd_caps.cable_id.bits.UHBR13_5_CAPABILITY) + } else if (link->dpcd_caps.cable_id.bits.UHBR13_5_CAPABILITY) { cable_max_link_rate = LINK_RATE_UHBR13_5; - else if (link->dpcd_caps.cable_id.bits.UHBR10_20_CAPABILITY & DP_UHBR10) - cable_max_link_rate = LINK_RATE_UHBR10; + } else if (link->dpcd_caps.cable_id.bits.UHBR10_20_CAPABILITY & DP_UHBR10) { + // allow DP40 cables to do UHBR13.5 for passive or unknown cable type + if (link->dpcd_caps.cable_id.bits.CABLE_TYPE < 2) { + cable_max_link_rate = LINK_RATE_UHBR13_5; + } else { + cable_max_link_rate = LINK_RATE_UHBR10; + } + } return cable_max_link_rate; } From d0f639c5869399bf6dde4d694d5f8c0ab8c0ec46 Mon Sep 17 00:00:00 2001 From: Taimur Hassan Date: Wed, 6 Dec 2023 14:52:25 -0500 Subject: [PATCH 15/92] drm/amd/display: Revert "Fix conversions between bytes and KB" [Why & How] HostVMMinPageSize is expected to be in KB according to spec, the checks later down the line reflect this as well. Reviewed-by: Nicholas Kazlauskas Acked-by: Aurabindo Pillai Signed-off-by: Taimur Hassan Signed-off-by: Alex Deucher --- .../drm/amd/display/dc/dml2/display_mode_core.c | 16 ++++++++-------- 1 file changed, 8 insertions(+), 8 deletions(-) diff --git a/drivers/gpu/drm/amd/display/dc/dml2/display_mode_core.c b/drivers/gpu/drm/amd/display/dc/dml2/display_mode_core.c index 4d1336e5afc2..180f8a98a361 100644 --- a/drivers/gpu/drm/amd/display/dc/dml2/display_mode_core.c +++ b/drivers/gpu/drm/amd/display/dc/dml2/display_mode_core.c @@ -6329,7 +6329,7 @@ static void dml_prefetch_check(struct display_mode_lib_st *mode_lib) mode_lib->ms.NoOfDPPThisState, mode_lib->ms.dpte_group_bytes, s->HostVMInefficiencyFactor, - mode_lib->ms.soc.hostvm_min_page_size_kbytes * 1024, + mode_lib->ms.soc.hostvm_min_page_size_kbytes, mode_lib->ms.cache_display_cfg.plane.HostVMMaxPageTableLevels); s->NextMaxVStartup = s->MaxVStartupAllPlanes[j]; @@ -6542,7 +6542,7 @@ static void dml_prefetch_check(struct display_mode_lib_st *mode_lib) mode_lib->ms.cache_display_cfg.plane.HostVMEnable, mode_lib->ms.cache_display_cfg.plane.HostVMMaxPageTableLevels, mode_lib->ms.cache_display_cfg.plane.GPUVMEnable, - mode_lib->ms.soc.hostvm_min_page_size_kbytes * 1024, + mode_lib->ms.soc.hostvm_min_page_size_kbytes, mode_lib->ms.PDEAndMetaPTEBytesPerFrame[j][k], mode_lib->ms.MetaRowBytes[j][k], mode_lib->ms.DPTEBytesPerRow[j][k], @@ -7687,7 +7687,7 @@ dml_bool_t dml_core_mode_support(struct display_mode_lib_st *mode_lib) CalculateVMRowAndSwath_params->HostVMMaxNonCachedPageTableLevels = mode_lib->ms.cache_display_cfg.plane.HostVMMaxPageTableLevels; CalculateVMRowAndSwath_params->GPUVMMaxPageTableLevels = mode_lib->ms.cache_display_cfg.plane.GPUVMMaxPageTableLevels; CalculateVMRowAndSwath_params->GPUVMMinPageSizeKBytes = mode_lib->ms.cache_display_cfg.plane.GPUVMMinPageSizeKBytes; - CalculateVMRowAndSwath_params->HostVMMinPageSize = mode_lib->ms.soc.hostvm_min_page_size_kbytes * 1024; + CalculateVMRowAndSwath_params->HostVMMinPageSize = mode_lib->ms.soc.hostvm_min_page_size_kbytes; CalculateVMRowAndSwath_params->PTEBufferModeOverrideEn = mode_lib->ms.cache_display_cfg.plane.PTEBufferModeOverrideEn; CalculateVMRowAndSwath_params->PTEBufferModeOverrideVal = mode_lib->ms.cache_display_cfg.plane.PTEBufferMode; CalculateVMRowAndSwath_params->PTEBufferSizeNotExceeded = mode_lib->ms.PTEBufferSizeNotExceededPerState; @@ -7957,7 +7957,7 @@ dml_bool_t dml_core_mode_support(struct display_mode_lib_st *mode_lib) UseMinimumDCFCLK_params->GPUVMMaxPageTableLevels = mode_lib->ms.cache_display_cfg.plane.GPUVMMaxPageTableLevels; UseMinimumDCFCLK_params->HostVMEnable = mode_lib->ms.cache_display_cfg.plane.HostVMEnable; UseMinimumDCFCLK_params->NumberOfActiveSurfaces = mode_lib->ms.num_active_planes; - UseMinimumDCFCLK_params->HostVMMinPageSize = mode_lib->ms.soc.hostvm_min_page_size_kbytes * 1024; + UseMinimumDCFCLK_params->HostVMMinPageSize = mode_lib->ms.soc.hostvm_min_page_size_kbytes; UseMinimumDCFCLK_params->HostVMMaxNonCachedPageTableLevels = mode_lib->ms.cache_display_cfg.plane.HostVMMaxPageTableLevels; UseMinimumDCFCLK_params->DynamicMetadataVMEnabled = mode_lib->ms.ip.dynamic_metadata_vm_enabled; UseMinimumDCFCLK_params->ImmediateFlipRequirement = s->ImmediateFlipRequiredFinal; @@ -8699,7 +8699,7 @@ void dml_core_mode_programming(struct display_mode_lib_st *mode_lib, const struc CalculateVMRowAndSwath_params->HostVMMaxNonCachedPageTableLevels = mode_lib->ms.cache_display_cfg.plane.HostVMMaxPageTableLevels; CalculateVMRowAndSwath_params->GPUVMMaxPageTableLevels = mode_lib->ms.cache_display_cfg.plane.GPUVMMaxPageTableLevels; CalculateVMRowAndSwath_params->GPUVMMinPageSizeKBytes = mode_lib->ms.cache_display_cfg.plane.GPUVMMinPageSizeKBytes; - CalculateVMRowAndSwath_params->HostVMMinPageSize = mode_lib->ms.soc.hostvm_min_page_size_kbytes * 1024; + CalculateVMRowAndSwath_params->HostVMMinPageSize = mode_lib->ms.soc.hostvm_min_page_size_kbytes; CalculateVMRowAndSwath_params->PTEBufferModeOverrideEn = mode_lib->ms.cache_display_cfg.plane.PTEBufferModeOverrideEn; CalculateVMRowAndSwath_params->PTEBufferModeOverrideVal = mode_lib->ms.cache_display_cfg.plane.PTEBufferMode; CalculateVMRowAndSwath_params->PTEBufferSizeNotExceeded = s->dummy_boolean_array[0]; @@ -8805,7 +8805,7 @@ void dml_core_mode_programming(struct display_mode_lib_st *mode_lib, const struc mode_lib->ms.cache_display_cfg.hw.DPPPerSurface, locals->dpte_group_bytes, s->HostVMInefficiencyFactor, - mode_lib->ms.soc.hostvm_min_page_size_kbytes * 1024, + mode_lib->ms.soc.hostvm_min_page_size_kbytes, mode_lib->ms.cache_display_cfg.plane.HostVMMaxPageTableLevels); locals->TCalc = 24.0 / locals->DCFCLKDeepSleep; @@ -8995,7 +8995,7 @@ void dml_core_mode_programming(struct display_mode_lib_st *mode_lib, const struc CalculatePrefetchSchedule_params->GPUVMEnable = mode_lib->ms.cache_display_cfg.plane.GPUVMEnable; CalculatePrefetchSchedule_params->HostVMEnable = mode_lib->ms.cache_display_cfg.plane.HostVMEnable; CalculatePrefetchSchedule_params->HostVMMaxNonCachedPageTableLevels = mode_lib->ms.cache_display_cfg.plane.HostVMMaxPageTableLevels; - CalculatePrefetchSchedule_params->HostVMMinPageSize = mode_lib->ms.soc.hostvm_min_page_size_kbytes * 1024; + CalculatePrefetchSchedule_params->HostVMMinPageSize = mode_lib->ms.soc.hostvm_min_page_size_kbytes; CalculatePrefetchSchedule_params->DynamicMetadataEnable = mode_lib->ms.cache_display_cfg.plane.DynamicMetadataEnable[k]; CalculatePrefetchSchedule_params->DynamicMetadataVMEnabled = mode_lib->ms.ip.dynamic_metadata_vm_enabled; CalculatePrefetchSchedule_params->DynamicMetadataLinesBeforeActiveRequired = mode_lib->ms.cache_display_cfg.plane.DynamicMetadataLinesBeforeActiveRequired[k]; @@ -9240,7 +9240,7 @@ void dml_core_mode_programming(struct display_mode_lib_st *mode_lib, const struc mode_lib->ms.cache_display_cfg.plane.HostVMEnable, mode_lib->ms.cache_display_cfg.plane.HostVMMaxPageTableLevels, mode_lib->ms.cache_display_cfg.plane.GPUVMEnable, - mode_lib->ms.soc.hostvm_min_page_size_kbytes * 1024, + mode_lib->ms.soc.hostvm_min_page_size_kbytes, locals->PDEAndMetaPTEBytesFrame[k], locals->MetaRowByte[k], locals->PixelPTEBytesPerRow[k], From 11edbb4497504540f5e73a8aabf1254b31cf0a82 Mon Sep 17 00:00:00 2001 From: Aurabindo Pillai Date: Wed, 6 Dec 2023 14:52:26 -0500 Subject: [PATCH 16/92] drm/amd/display: trivial comment change FP guard is valid for all recent asics, not just RV, so fix the comment. Reviewed-by: Chaitanya Dhere Acked-by: Aurabindo Pillai Signed-off-by: Aurabindo Pillai Signed-off-by: Alex Deucher --- drivers/gpu/drm/amd/display/dc/clk_mgr/clk_mgr.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/drivers/gpu/drm/amd/display/dc/clk_mgr/clk_mgr.c b/drivers/gpu/drm/amd/display/dc/clk_mgr/clk_mgr.c index 3e73c4e59d40..7dacb0f82d29 100644 --- a/drivers/gpu/drm/amd/display/dc/clk_mgr/clk_mgr.c +++ b/drivers/gpu/drm/amd/display/dc/clk_mgr/clk_mgr.c @@ -368,7 +368,7 @@ struct clk_mgr *dc_clk_mgr_create(struct dc_context *ctx, struct pp_smu_funcs *p } break; -#endif /* CONFIG_DRM_AMD_DC_FP - Family RV */ +#endif /* CONFIG_DRM_AMD_DC_FP */ default: ASSERT(0); /* Unknown Asic */ break; From 2170fb03be28ad7807ea460101a60689c3f383e4 Mon Sep 17 00:00:00 2001 From: Michael Strauss Date: Wed, 6 Dec 2023 14:52:27 -0500 Subject: [PATCH 17/92] drm/amd/display: Revert DP2 MST hub triple display fix [WHY] Introduces regression with DP2 native displays [HOW] Revert the change Reviewed-by: Charlene Liu Acked-by: Aurabindo Pillai Signed-off-by: Michael Strauss Signed-off-by: Alex Deucher --- drivers/gpu/drm/amd/display/dc/dml2/dml2_utils.c | 7 ------- 1 file changed, 7 deletions(-) diff --git a/drivers/gpu/drm/amd/display/dc/dml2/dml2_utils.c b/drivers/gpu/drm/amd/display/dc/dml2/dml2_utils.c index 38fa7441df51..814dbdcf9a78 100644 --- a/drivers/gpu/drm/amd/display/dc/dml2/dml2_utils.c +++ b/drivers/gpu/drm/amd/display/dc/dml2/dml2_utils.c @@ -160,13 +160,6 @@ bool is_dp2p0_output_encoder(const struct pipe_ctx *pipe_ctx) if (pipe_ctx->stream == NULL) return false; - /* Count MST hubs once by treating only 1st remote sink in topology as an encoder */ - if (pipe_ctx->stream->link && pipe_ctx->stream->link->remote_sinks[0]) { - return (pipe_ctx->stream_res.hpo_dp_stream_enc && - pipe_ctx->link_res.hpo_dp_link_enc && - dc_is_dp_signal(pipe_ctx->stream->signal) && - (pipe_ctx->stream->link->remote_sinks[0] == pipe_ctx->stream->sink)); - } return (pipe_ctx->stream_res.hpo_dp_stream_enc && pipe_ctx->link_res.hpo_dp_link_enc && From c1afbb715e33a2b208c27a989c5f929029ffe7d3 Mon Sep 17 00:00:00 2001 From: Fangzhi Zuo Date: Wed, 6 Dec 2023 14:52:28 -0500 Subject: [PATCH 18/92] drm/amd/display: Populate dtbclk from bounding box dtbclk is unavaliable from pmfw. Try to grab the value from bounding box Reviewed-by: Charlene Liu Acked-by: Aurabindo Pillai Signed-off-by: Fangzhi Zuo Signed-off-by: Alex Deucher --- .../gpu/drm/amd/display/dc/dml/dcn35/dcn35_fpu.c | 14 +++++++++----- .../amd/display/dc/dml2/dml2_translation_helper.c | 5 +++-- 2 files changed, 12 insertions(+), 7 deletions(-) diff --git a/drivers/gpu/drm/amd/display/dc/dml/dcn35/dcn35_fpu.c b/drivers/gpu/drm/amd/display/dc/dml/dcn35/dcn35_fpu.c index e9d88f52717b..3d12dabd39e4 100644 --- a/drivers/gpu/drm/amd/display/dc/dml/dcn35/dcn35_fpu.c +++ b/drivers/gpu/drm/amd/display/dc/dml/dcn35/dcn35_fpu.c @@ -124,7 +124,7 @@ struct _vcs_dpi_soc_bounding_box_st dcn3_5_soc = { .phyclk_mhz = 600.0, .phyclk_d18_mhz = 667.0, .dscclk_mhz = 186.0, - .dtbclk_mhz = 625.0, + .dtbclk_mhz = 600.0, }, { .state = 1, @@ -133,7 +133,7 @@ struct _vcs_dpi_soc_bounding_box_st dcn3_5_soc = { .phyclk_mhz = 810.0, .phyclk_d18_mhz = 667.0, .dscclk_mhz = 209.0, - .dtbclk_mhz = 625.0, + .dtbclk_mhz = 600.0, }, { .state = 2, @@ -142,7 +142,7 @@ struct _vcs_dpi_soc_bounding_box_st dcn3_5_soc = { .phyclk_mhz = 810.0, .phyclk_d18_mhz = 667.0, .dscclk_mhz = 209.0, - .dtbclk_mhz = 625.0, + .dtbclk_mhz = 600.0, }, { .state = 3, @@ -151,7 +151,7 @@ struct _vcs_dpi_soc_bounding_box_st dcn3_5_soc = { .phyclk_mhz = 810.0, .phyclk_d18_mhz = 667.0, .dscclk_mhz = 371.0, - .dtbclk_mhz = 625.0, + .dtbclk_mhz = 600.0, }, { .state = 4, @@ -160,7 +160,7 @@ struct _vcs_dpi_soc_bounding_box_st dcn3_5_soc = { .phyclk_mhz = 810.0, .phyclk_d18_mhz = 667.0, .dscclk_mhz = 417.0, - .dtbclk_mhz = 625.0, + .dtbclk_mhz = 600.0, }, }, .num_states = 5, @@ -367,6 +367,8 @@ void dcn35_update_bw_bounding_box_fpu(struct dc *dc, clock_limits[i].socclk_mhz; dc->dml2_options.bbox_overrides.clks_table.clk_entries[i].memclk_mhz = clk_table->entries[i].memclk_mhz * clk_table->entries[i].wck_ratio; + dc->dml2_options.bbox_overrides.clks_table.clk_entries[i].dtbclk_mhz = + clock_limits[i].dtbclk_mhz; dc->dml2_options.bbox_overrides.clks_table.num_entries_per_clk.num_dcfclk_levels = clk_table->num_entries; dc->dml2_options.bbox_overrides.clks_table.num_entries_per_clk.num_fclk_levels = @@ -379,6 +381,8 @@ void dcn35_update_bw_bounding_box_fpu(struct dc *dc, clk_table->num_entries; dc->dml2_options.bbox_overrides.clks_table.num_entries_per_clk.num_memclk_levels = clk_table->num_entries; + dc->dml2_options.bbox_overrides.clks_table.num_entries_per_clk.num_dtbclk_levels = + clk_table->num_entries; } } diff --git a/drivers/gpu/drm/amd/display/dc/dml2/dml2_translation_helper.c b/drivers/gpu/drm/amd/display/dc/dml2/dml2_translation_helper.c index 255af7875c08..279e7605a0a2 100644 --- a/drivers/gpu/drm/amd/display/dc/dml2/dml2_translation_helper.c +++ b/drivers/gpu/drm/amd/display/dc/dml2/dml2_translation_helper.c @@ -425,8 +425,9 @@ void dml2_init_soc_states(struct dml2_context *dml2, const struct dc *in_dc, } for (i = 0; i < dml2->config.bbox_overrides.clks_table.num_entries_per_clk.num_dtbclk_levels; i++) { - p->in_states->state_array[i].dtbclk_mhz = - dml2->config.bbox_overrides.clks_table.clk_entries[i].dtbclk_mhz; + if (dml2->config.bbox_overrides.clks_table.clk_entries[i].dtbclk_mhz > 0) + p->in_states->state_array[i].dtbclk_mhz = + dml2->config.bbox_overrides.clks_table.clk_entries[i].dtbclk_mhz; } for (i = 0; i < dml2->config.bbox_overrides.clks_table.num_entries_per_clk.num_dispclk_levels; i++) { From bbc42960f9b68e548403e57b2cfd6e93e684864f Mon Sep 17 00:00:00 2001 From: Allen Date: Wed, 6 Dec 2023 14:52:29 -0500 Subject: [PATCH 19/92] drm/amd/display: Disable OPTC pg to match DC Hubp/dpp pg [Why] To match the hardware sequence Reviewed-by: Charlene Liu Acked-by: Aurabindo Pillai Signed-off-by: Allen Signed-off-by: Alex Deucher --- drivers/gpu/drm/amd/display/dc/resource/dcn35/dcn35_resource.c | 1 + 1 file changed, 1 insertion(+) diff --git a/drivers/gpu/drm/amd/display/dc/resource/dcn35/dcn35_resource.c b/drivers/gpu/drm/amd/display/dc/resource/dcn35/dcn35_resource.c index 13324422ff50..4eb744f1bc9f 100644 --- a/drivers/gpu/drm/amd/display/dc/resource/dcn35/dcn35_resource.c +++ b/drivers/gpu/drm/amd/display/dc/resource/dcn35/dcn35_resource.c @@ -717,6 +717,7 @@ static const struct dc_debug_options debug_defaults_drv = { .disable_dcc = DCC_ENABLE, .disable_dpp_power_gate = true, .disable_hubp_power_gate = true, + .disable_optc_power_gate = true, /*should the same as above two*/ .disable_clock_gate = false, .disable_dsc_power_gate = true, .vsr_support = true, From fdb0ad2ff7c84bda30bfe3b8f90abd1f8d8788a0 Mon Sep 17 00:00:00 2001 From: Sung Joon Kim Date: Wed, 6 Dec 2023 14:52:30 -0500 Subject: [PATCH 20/92] drm/amd/display: Exit from idle state before accessing HW data [why & how] User interface cannot guarantee system is in idle state, so need to ensure we exit idle state before accessing any HW data. Reviewed-by: Nicholas Kazlauskas Acked-by: Aurabindo Pillai Signed-off-by: Sung Joon Kim Signed-off-by: Alex Deucher --- drivers/gpu/drm/amd/display/dc/dce/dmub_abm.c | 4 ++++ 1 file changed, 4 insertions(+) diff --git a/drivers/gpu/drm/amd/display/dc/dce/dmub_abm.c b/drivers/gpu/drm/amd/display/dc/dce/dmub_abm.c index 930fd929e93a..8c5e7f858be3 100644 --- a/drivers/gpu/drm/amd/display/dc/dce/dmub_abm.c +++ b/drivers/gpu/drm/amd/display/dc/dce/dmub_abm.c @@ -64,11 +64,15 @@ static void dmub_abm_init_ex(struct abm *abm, uint32_t backlight) static unsigned int dmub_abm_get_current_backlight_ex(struct abm *abm) { + dc_allow_idle_optimizations(abm->ctx->dc, false); + return dmub_abm_get_current_backlight(abm); } static unsigned int dmub_abm_get_target_backlight_ex(struct abm *abm) { + dc_allow_idle_optimizations(abm->ctx->dc, false); + return dmub_abm_get_target_backlight(abm); } From dd4e4bb28843393065eed279e869fac248d03f0f Mon Sep 17 00:00:00 2001 From: Alvin Lee Date: Wed, 6 Dec 2023 14:52:31 -0500 Subject: [PATCH 21/92] drm/amd/display: For prefetch mode > 0, extend prefetch if possible [Description] For mode programming we want to extend the prefetch as much as possible (up to oto, or as long as we can for equ) if we're not already applying the 60us prefetch requirement. This is to avoid intermittent underflow issues during prefetch. The prefetch extension is applied under the following scenarios: 1. We're in prefetch mode 1 (i.e. we don't support MCLK switch in blank) 2. We're using subvp or drr methods of p-state switch, in which case we we don't care if prefetch takes up more of the blanking time Mode programming typically chooses the smallest prefetch time possible (i.e. highest bandwidth during prefetch) presumably to create margin between p-states / c-states that happen in vblank and prefetch. Therefore we only apply this prefetch extension when p-state in vblank is not required (UCLK p-states take up the most vblank time). Reviewed-by: Jun Lei Acked-by: Aurabindo Pillai Signed-off-by: Alvin Lee Signed-off-by: Alex Deucher --- .../dc/dml/dcn32/display_mode_vba_32.c | 3 ++ .../dc/dml/dcn32/display_mode_vba_util_32.c | 33 +++++++++++++++---- .../dc/dml/dcn32/display_mode_vba_util_32.h | 1 + 3 files changed, 31 insertions(+), 6 deletions(-) diff --git a/drivers/gpu/drm/amd/display/dc/dml/dcn32/display_mode_vba_32.c b/drivers/gpu/drm/amd/display/dc/dml/dcn32/display_mode_vba_32.c index cbdfb762c10c..6c84b0fa40f4 100644 --- a/drivers/gpu/drm/amd/display/dc/dml/dcn32/display_mode_vba_32.c +++ b/drivers/gpu/drm/amd/display/dc/dml/dcn32/display_mode_vba_32.c @@ -813,6 +813,8 @@ static void DISPCLKDPPCLKDCFCLKDeepSleepPrefetchParametersWatermarksAndPerforman (v->DRAMSpeedPerState[mode_lib->vba.VoltageLevel] <= MEM_STROBE_FREQ_MHZ || v->DCFCLKPerState[mode_lib->vba.VoltageLevel] <= DCFCLK_FREQ_EXTRA_PREFETCH_REQ_MHZ) ? mode_lib->vba.ip.min_prefetch_in_strobe_us : 0, + mode_lib->vba.PrefetchModePerState[mode_lib->vba.VoltageLevel][mode_lib->vba.maxMpcComb] > 0 || mode_lib->vba.DRAMClockChangeRequirementFinal == false, + /* Output */ &v->DSTXAfterScaler[k], &v->DSTYAfterScaler[k], @@ -3317,6 +3319,7 @@ void dml32_ModeSupportAndSystemConfigurationFull(struct display_mode_lib *mode_l v->SwathHeightCThisState[k], v->TWait, (v->DRAMSpeedPerState[i] <= MEM_STROBE_FREQ_MHZ || v->DCFCLKState[i][j] <= DCFCLK_FREQ_EXTRA_PREFETCH_REQ_MHZ) ? mode_lib->vba.ip.min_prefetch_in_strobe_us : 0, + mode_lib->vba.PrefetchModePerState[i][j] > 0 || mode_lib->vba.DRAMClockChangeRequirementFinal == false, /* Output */ &v->dummy_vars.dml32_ModeSupportAndSystemConfigurationFull.DSTXAfterScaler[k], diff --git a/drivers/gpu/drm/amd/display/dc/dml/dcn32/display_mode_vba_util_32.c b/drivers/gpu/drm/amd/display/dc/dml/dcn32/display_mode_vba_util_32.c index d940dfa5ae43..80fccd4999a5 100644 --- a/drivers/gpu/drm/amd/display/dc/dml/dcn32/display_mode_vba_util_32.c +++ b/drivers/gpu/drm/amd/display/dc/dml/dcn32/display_mode_vba_util_32.c @@ -3423,6 +3423,7 @@ bool dml32_CalculatePrefetchSchedule( unsigned int SwathHeightC, double TWait, double TPreReq, + bool ExtendPrefetchIfPossible, /* Output */ double *DSTXAfterScaler, double *DSTYAfterScaler, @@ -3892,12 +3893,32 @@ bool dml32_CalculatePrefetchSchedule( /* Clamp to oto for bandwidth calculation */ LinesForPrefetchBandwidth = dst_y_prefetch_oto; } else { - *DestinationLinesForPrefetch = dst_y_prefetch_equ; - TimeForFetchingMetaPTE = Tvm_equ; - TimeForFetchingRowInVBlank = Tr0_equ; - *PrefetchBandwidth = prefetch_bw_equ; - /* Clamp to equ for bandwidth calculation */ - LinesForPrefetchBandwidth = dst_y_prefetch_equ; + /* For mode programming we want to extend the prefetch as much as possible + * (up to oto, or as long as we can for equ) if we're not already applying + * the 60us prefetch requirement. This is to avoid intermittent underflow + * issues during prefetch. + * + * The prefetch extension is applied under the following scenarios: + * 1. We're in prefetch mode > 0 (i.e. we don't support MCLK switch in blank) + * 2. We're using subvp or drr methods of p-state switch, in which case we + * we don't care if prefetch takes up more of the blanking time + * + * Mode programming typically chooses the smallest prefetch time possible + * (i.e. highest bandwidth during prefetch) presumably to create margin between + * p-states / c-states that happen in vblank and prefetch. Therefore we only + * apply this prefetch extension when p-state in vblank is not required (UCLK + * p-states take up the most vblank time). + */ + if (ExtendPrefetchIfPossible && TPreReq == 0 && VStartup < MaxVStartup) { + MyError = true; + } else { + *DestinationLinesForPrefetch = dst_y_prefetch_equ; + TimeForFetchingMetaPTE = Tvm_equ; + TimeForFetchingRowInVBlank = Tr0_equ; + *PrefetchBandwidth = prefetch_bw_equ; + /* Clamp to equ for bandwidth calculation */ + LinesForPrefetchBandwidth = dst_y_prefetch_equ; + } } *DestinationLinesToRequestVMInVBlank = dml_ceil(4.0 * TimeForFetchingMetaPTE / LineTime, 1.0) / 4.0; diff --git a/drivers/gpu/drm/amd/display/dc/dml/dcn32/display_mode_vba_util_32.h b/drivers/gpu/drm/amd/display/dc/dml/dcn32/display_mode_vba_util_32.h index 592d174df6c6..5d34735df83d 100644 --- a/drivers/gpu/drm/amd/display/dc/dml/dcn32/display_mode_vba_util_32.h +++ b/drivers/gpu/drm/amd/display/dc/dml/dcn32/display_mode_vba_util_32.h @@ -747,6 +747,7 @@ bool dml32_CalculatePrefetchSchedule( unsigned int SwathHeightC, double TWait, double TPreReq, + bool ExtendPrefetchIfPossible, /* Output */ double *DSTXAfterScaler, double *DSTYAfterScaler, From 9a902a9073c287353e25913c0761bfed49d75a88 Mon Sep 17 00:00:00 2001 From: Alvin Lee Date: Wed, 6 Dec 2023 14:52:32 -0500 Subject: [PATCH 22/92] drm/amd/display: Force p-state disallow if leaving no plane config [Description] - When we're in a no plane config, DCN is always asserting P-State allow - This creates a scenario where the P-State blackout can start just as VUPDATE takes place and transitions the DCN config to a one where one or more HUBP's are active which can result in underflow - To fix this issue, force p-state disallow and unforce after the transition from no planes case -> one or more planes active Reviewed-by: Samson Tam Acked-by: Aurabindo Pillai Signed-off-by: Alvin Lee Signed-off-by: Alex Deucher --- .../amd/display/dc/hwss/dcn20/dcn20_hwseq.c | 20 +++++++++++++++++++ 1 file changed, 20 insertions(+) diff --git a/drivers/gpu/drm/amd/display/dc/hwss/dcn20/dcn20_hwseq.c b/drivers/gpu/drm/amd/display/dc/hwss/dcn20/dcn20_hwseq.c index 608221b0dd5d..c3c83178eb1e 100644 --- a/drivers/gpu/drm/amd/display/dc/hwss/dcn20/dcn20_hwseq.c +++ b/drivers/gpu/drm/amd/display/dc/hwss/dcn20/dcn20_hwseq.c @@ -1877,6 +1877,8 @@ void dcn20_program_front_end_for_ctx( int i; struct dce_hwseq *hws = dc->hwseq; DC_LOGGER_INIT(dc->ctx->logger); + unsigned int prev_hubp_count = 0; + unsigned int hubp_count = 0; if (resource_is_pipe_topology_changed(dc->current_state, context)) resource_log_pipe_topology_update(dc, context); @@ -1894,6 +1896,20 @@ void dcn20_program_front_end_for_ctx( } } + for (i = 0; i < dc->res_pool->pipe_count; i++) { + if (dc->current_state->res_ctx.pipe_ctx[i].plane_state) + prev_hubp_count++; + if (context->res_ctx.pipe_ctx[i].plane_state) + hubp_count++; + } + + if (prev_hubp_count == 0 && hubp_count > 0) { + if (dc->res_pool->hubbub->funcs->force_pstate_change_control) + dc->res_pool->hubbub->funcs->force_pstate_change_control( + dc->res_pool->hubbub, true, false); + udelay(500); + } + /* Set pipe update flags and lock pipes */ for (i = 0; i < dc->res_pool->pipe_count; i++) dcn20_detect_pipe_changes(&dc->current_state->res_ctx.pipe_ctx[i], @@ -2039,6 +2055,10 @@ void dcn20_post_unlock_program_front_end( } } + if (dc->res_pool->hubbub->funcs->force_pstate_change_control) + dc->res_pool->hubbub->funcs->force_pstate_change_control( + dc->res_pool->hubbub, false, false); + for (i = 0; i < dc->res_pool->pipe_count; i++) { struct pipe_ctx *pipe = &context->res_ctx.pipe_ctx[i]; From 7253c36b1febe7e76be3da26fbf875978b37e92c Mon Sep 17 00:00:00 2001 From: Charlene Liu Date: Wed, 6 Dec 2023 14:52:33 -0500 Subject: [PATCH 23/92] drm/amd/display: fix HW block PG sequence [why] Power up and power down has reverted programming order. also make sure disable root clock last. Reviewed-by: Muhammad Ahmed Acked-by: Aurabindo Pillai Signed-off-by: Charlene Liu Signed-off-by: Alex Deucher --- .../gpu/drm/amd/display/dc/dcn35/dcn35_init.c | 3 +- .../amd/display/dc/hwss/dcn35/dcn35_hwseq.c | 138 ++++++++++++------ .../amd/display/dc/hwss/dcn35/dcn35_hwseq.h | 6 +- .../drm/amd/display/dc/hwss/hw_sequencer.h | 6 +- 4 files changed, 107 insertions(+), 46 deletions(-) diff --git a/drivers/gpu/drm/amd/display/dc/dcn35/dcn35_init.c b/drivers/gpu/drm/amd/display/dc/dcn35/dcn35_init.c index 296bf3a38cb9..d594905eb246 100644 --- a/drivers/gpu/drm/amd/display/dc/dcn35/dcn35_init.c +++ b/drivers/gpu/drm/amd/display/dc/dcn35/dcn35_init.c @@ -118,7 +118,8 @@ static const struct hw_sequencer_funcs dcn35_funcs = { .update_dsc_pg = dcn32_update_dsc_pg, .calc_blocks_to_gate = dcn35_calc_blocks_to_gate, .calc_blocks_to_ungate = dcn35_calc_blocks_to_ungate, - .block_power_control = dcn35_block_power_control, + .hw_block_power_up = dcn35_hw_block_power_up, + .hw_block_power_down = dcn35_hw_block_power_down, .root_clock_control = dcn35_root_clock_control, .set_idle_state = dcn35_set_idle_state, .get_idle_state = dcn35_get_idle_state diff --git a/drivers/gpu/drm/amd/display/dc/hwss/dcn35/dcn35_hwseq.c b/drivers/gpu/drm/amd/display/dc/hwss/dcn35/dcn35_hwseq.c index 09498aa92096..9262d3336182 100644 --- a/drivers/gpu/drm/amd/display/dc/hwss/dcn35/dcn35_hwseq.c +++ b/drivers/gpu/drm/amd/display/dc/hwss/dcn35/dcn35_hwseq.c @@ -1123,9 +1123,23 @@ void dcn35_calc_blocks_to_ungate(struct dc *dc, struct dc_state *context, update_state->pg_res_update[PG_HPO] = true; } - -void dcn35_block_power_control(struct dc *dc, - struct pg_block_update *update_state, bool power_on) +/** + * power down sequence + * ONO Region 3, DCPG 25: hpo - SKIPPED + * ONO Region 4, DCPG 0: dchubp0, dpp0 + * ONO Region 6, DCPG 1: dchubp1, dpp1 + * ONO Region 8, DCPG 2: dchubp2, dpp2 + * ONO Region 10, DCPG 3: dchubp3, dpp3 + * ONO Region 1, DCPG 23: dchubbub dchvm dchubbubmem - SKIPPED. PMFW will pwr dwn at IPS2 entry + * ONO Region 5, DCPG 16: dsc0 + * ONO Region 7, DCPG 17: dsc1 + * ONO Region 9, DCPG 18: dsc2 + * ONO Region 11, DCPG 19: dsc3 + * ONO Region 2, DCPG 24: mpc opp optc dwb + * ONO Region 0, DCPG 22: dccg dio dcio - SKIPPED. will be pwr dwn after lono timer is armed +*/ +void dcn35_hw_block_power_down(struct dc *dc, + struct pg_block_update *update_state) { int i = 0; struct pg_cntl *pg_cntl = dc->res_pool->pg_cntl; @@ -1134,50 +1148,81 @@ void dcn35_block_power_control(struct dc *dc, return; if (dc->debug.ignore_pg) return; + if (update_state->pg_res_update[PG_HPO]) { if (pg_cntl->funcs->hpo_pg_control) - pg_cntl->funcs->hpo_pg_control(pg_cntl, power_on); + pg_cntl->funcs->hpo_pg_control(pg_cntl, false); } for (i = 0; i < dc->res_pool->pipe_count; i++) { if (update_state->pg_pipe_res_update[PG_HUBP][i] && update_state->pg_pipe_res_update[PG_DPP][i]) { if (pg_cntl->funcs->hubp_dpp_pg_control) - pg_cntl->funcs->hubp_dpp_pg_control(pg_cntl, i, power_on); + pg_cntl->funcs->hubp_dpp_pg_control(pg_cntl, i, false); } - + } + for (i = 0; i < dc->res_pool->res_cap->num_dsc; i++) if (update_state->pg_pipe_res_update[PG_DSC][i]) { if (pg_cntl->funcs->dsc_pg_control) - pg_cntl->funcs->dsc_pg_control(pg_cntl, i, power_on); + pg_cntl->funcs->dsc_pg_control(pg_cntl, i, false); } - if (update_state->pg_pipe_res_update[PG_MPCC][i]) { - if (pg_cntl->funcs->mpcc_pg_control) - pg_cntl->funcs->mpcc_pg_control(pg_cntl, i, power_on); - } - - if (update_state->pg_pipe_res_update[PG_OPP][i]) { - if (pg_cntl->funcs->opp_pg_control) - pg_cntl->funcs->opp_pg_control(pg_cntl, i, power_on); - } - - if (update_state->pg_pipe_res_update[PG_OPTC][i]) { - if (pg_cntl->funcs->optc_pg_control) - pg_cntl->funcs->optc_pg_control(pg_cntl, i, power_on); - } - } - - if (update_state->pg_res_update[PG_DWB]) { - if (pg_cntl->funcs->dwb_pg_control) - pg_cntl->funcs->dwb_pg_control(pg_cntl, power_on); - } /*this will need all the clients to unregister optc interruts let dmubfw handle this*/ if (pg_cntl->funcs->plane_otg_pg_control) - pg_cntl->funcs->plane_otg_pg_control(pg_cntl, power_on); + pg_cntl->funcs->plane_otg_pg_control(pg_cntl, false); + + //domain22, 23, 25 currently always on. } +/** + * power up sequence + * ONO Region 0, DCPG 22: dccg dio dcio - SKIPPED + * ONO Region 2, DCPG 24: mpc opp optc dwb + * ONO Region 5, DCPG 16: dsc0 + * ONO Region 7, DCPG 17: dsc1 + * ONO Region 9, DCPG 18: dsc2 + * ONO Region 11, DCPG 19: dsc3 + * ONO Region 1, DCPG 23: dchubbub dchvm dchubbubmem - SKIPPED. PMFW will power up at IPS2 exit + * ONO Region 4, DCPG 0: dchubp0, dpp0 + * ONO Region 6, DCPG 1: dchubp1, dpp1 + * ONO Region 8, DCPG 2: dchubp2, dpp2 + * ONO Region 10, DCPG 3: dchubp3, dpp3 + * ONO Region 3, DCPG 25: hpo - SKIPPED + */ +void dcn35_hw_block_power_up(struct dc *dc, + struct pg_block_update *update_state) +{ + int i = 0; + struct pg_cntl *pg_cntl = dc->res_pool->pg_cntl; + if (!pg_cntl) + return; + if (dc->debug.ignore_pg) + return; + //domain22, 23, 25 currently always on. + /*this will need all the clients to unregister optc interruts let dmubfw handle this*/ + if (pg_cntl->funcs->plane_otg_pg_control) + pg_cntl->funcs->plane_otg_pg_control(pg_cntl, true); + + for (i = 0; i < dc->res_pool->res_cap->num_dsc; i++) + if (update_state->pg_pipe_res_update[PG_DSC][i]) { + if (pg_cntl->funcs->dsc_pg_control) + pg_cntl->funcs->dsc_pg_control(pg_cntl, i, true); + } + + for (i = 0; i < dc->res_pool->pipe_count; i++) { + if (update_state->pg_pipe_res_update[PG_HUBP][i] && + update_state->pg_pipe_res_update[PG_DPP][i]) { + if (pg_cntl->funcs->hubp_dpp_pg_control) + pg_cntl->funcs->hubp_dpp_pg_control(pg_cntl, i, true); + } + } + if (update_state->pg_res_update[PG_HPO]) { + if (pg_cntl->funcs->hpo_pg_control) + pg_cntl->funcs->hpo_pg_control(pg_cntl, true); + } +} void dcn35_root_clock_control(struct dc *dc, struct pg_block_update *update_state, bool power_on) { @@ -1186,14 +1231,16 @@ void dcn35_root_clock_control(struct dc *dc, if (!pg_cntl) return; - - for (i = 0; i < dc->res_pool->pipe_count; i++) { - if (update_state->pg_pipe_res_update[PG_HUBP][i] && - update_state->pg_pipe_res_update[PG_DPP][i]) { - if (dc->hwseq->funcs.dpp_root_clock_control) - dc->hwseq->funcs.dpp_root_clock_control(dc->hwseq, i, power_on); + /*enable root clock first when power up*/ + if (power_on) + for (i = 0; i < dc->res_pool->pipe_count; i++) { + if (update_state->pg_pipe_res_update[PG_HUBP][i] && + update_state->pg_pipe_res_update[PG_DPP][i]) { + if (dc->hwseq->funcs.dpp_root_clock_control) + dc->hwseq->funcs.dpp_root_clock_control(dc->hwseq, i, power_on); + } } - + for (i = 0; i < dc->res_pool->res_cap->num_dsc; i++) { if (update_state->pg_pipe_res_update[PG_DSC][i]) { if (power_on) { if (dc->res_pool->dccg->funcs->enable_dsc) @@ -1204,6 +1251,15 @@ void dcn35_root_clock_control(struct dc *dc, } } } + /*disable root clock first when power down*/ + if (!power_on) + for (i = 0; i < dc->res_pool->pipe_count; i++) { + if (update_state->pg_pipe_res_update[PG_HUBP][i] && + update_state->pg_pipe_res_update[PG_DPP][i]) { + if (dc->hwseq->funcs.dpp_root_clock_control) + dc->hwseq->funcs.dpp_root_clock_control(dc->hwseq, i, power_on); + } + } } void dcn35_prepare_bandwidth( @@ -1217,9 +1273,9 @@ void dcn35_prepare_bandwidth( if (dc->hwss.root_clock_control) dc->hwss.root_clock_control(dc, &pg_update_state, true); - - if (dc->hwss.block_power_control) - dc->hwss.block_power_control(dc, &pg_update_state, true); + /*power up required HW block*/ + if (dc->hwss.hw_block_power_up) + dc->hwss.hw_block_power_up(dc, &pg_update_state); } dcn20_prepare_bandwidth(dc, context); @@ -1235,9 +1291,9 @@ void dcn35_optimize_bandwidth( if (dc->hwss.calc_blocks_to_gate) { dc->hwss.calc_blocks_to_gate(dc, context, &pg_update_state); - - if (dc->hwss.block_power_control) - dc->hwss.block_power_control(dc, &pg_update_state, false); + /*try to power down unused block*/ + if (dc->hwss.hw_block_power_down) + dc->hwss.hw_block_power_down(dc, &pg_update_state); if (dc->hwss.root_clock_control) dc->hwss.root_clock_control(dc, &pg_update_state, false); diff --git a/drivers/gpu/drm/amd/display/dc/hwss/dcn35/dcn35_hwseq.h b/drivers/gpu/drm/amd/display/dc/hwss/dcn35/dcn35_hwseq.h index 0dff10d179b8..3837038dc4a8 100644 --- a/drivers/gpu/drm/amd/display/dc/hwss/dcn35/dcn35_hwseq.h +++ b/drivers/gpu/drm/amd/display/dc/hwss/dcn35/dcn35_hwseq.h @@ -63,8 +63,10 @@ void dcn35_calc_blocks_to_gate(struct dc *dc, struct dc_state *context, struct pg_block_update *update_state); void dcn35_calc_blocks_to_ungate(struct dc *dc, struct dc_state *context, struct pg_block_update *update_state); -void dcn35_block_power_control(struct dc *dc, - struct pg_block_update *update_state, bool power_on); +void dcn35_hw_block_power_up(struct dc *dc, + struct pg_block_update *update_state); +void dcn35_hw_block_power_down(struct dc *dc, + struct pg_block_update *update_state); void dcn35_root_clock_control(struct dc *dc, struct pg_block_update *update_state, bool power_on); diff --git a/drivers/gpu/drm/amd/display/dc/hwss/hw_sequencer.h b/drivers/gpu/drm/amd/display/dc/hwss/hw_sequencer.h index 452680fe9aab..45dc6d4e9562 100644 --- a/drivers/gpu/drm/amd/display/dc/hwss/hw_sequencer.h +++ b/drivers/gpu/drm/amd/display/dc/hwss/hw_sequencer.h @@ -414,8 +414,10 @@ struct hw_sequencer_funcs { struct pg_block_update *update_state); void (*calc_blocks_to_ungate)(struct dc *dc, struct dc_state *context, struct pg_block_update *update_state); - void (*block_power_control)(struct dc *dc, - struct pg_block_update *update_state, bool power_on); + void (*hw_block_power_up)(struct dc *dc, + struct pg_block_update *update_state); + void (*hw_block_power_down)(struct dc *dc, + struct pg_block_update *update_state); void (*root_clock_control)(struct dc *dc, struct pg_block_update *update_state, bool power_on); void (*set_idle_state)(const struct dc *dc, bool allow_idle); From bcbd0787f8be31b17125d05cfaf71724774b9964 Mon Sep 17 00:00:00 2001 From: Aric Cyr Date: Wed, 6 Dec 2023 14:52:34 -0500 Subject: [PATCH 24/92] drm/amd/display: 3.2.264 Summary: Bug fixes for: * DCN35 power gating * P-state change, & prefetch logic * ABM * DP 2.1 Acked-by: Aurabindo Pillai Signed-off-by: Aric Cyr Signed-off-by: Alex Deucher --- drivers/gpu/drm/amd/display/dc/dc.h | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/drivers/gpu/drm/amd/display/dc/dc.h b/drivers/gpu/drm/amd/display/dc/dc.h index 61d08349a0d7..2c85f8ee682f 100644 --- a/drivers/gpu/drm/amd/display/dc/dc.h +++ b/drivers/gpu/drm/amd/display/dc/dc.h @@ -49,7 +49,7 @@ struct aux_payload; struct set_config_cmd_payload; struct dmub_notification; -#define DC_VER "3.2.263" +#define DC_VER "3.2.264" #define MAX_SURFACES 3 #define MAX_PLANES 6 From 571c2fa26aa654946447c282a09d40a56c7ff128 Mon Sep 17 00:00:00 2001 From: Mario Limonciello Date: Mon, 19 Jun 2023 15:04:24 -0500 Subject: [PATCH 25/92] drm/amd/display: Disable PSR-SU on Parade 0803 TCON again When screen brightness is rapidly changed and PSR-SU is enabled the display hangs on panels with this TCON even on the latest DCN 3.1.4 microcode (0x8002a81 at this time). This was disabled previously as commit 072030b17830 ("drm/amd: Disable PSR-SU on Parade 0803 TCON") but reverted as commit 1e66a17ce546 ("Revert "drm/amd: Disable PSR-SU on Parade 0803 TCON"") in favor of testing for a new enough microcode (commit cd2e31a9ab93 ("drm/amd/display: Set minimum requirement for using PSR-SU on Phoenix")). As hangs are still happening specifically with this TCON, disable PSR-SU again for it until it can be root caused. Cc: stable@vger.kernel.org Cc: aaron.ma@canonical.com Cc: binli@gnome.org Cc: Marc Rossi Cc: Hamza Mahfooz Signed-off-by: Mario Limonciello Link: https://bugs.launchpad.net/ubuntu/+source/linux/+bug/2046131 Acked-by: Alex Deucher Reviewed-by: Harry Wentland Signed-off-by: Alex Deucher --- drivers/gpu/drm/amd/display/modules/power/power_helpers.c | 2 ++ 1 file changed, 2 insertions(+) diff --git a/drivers/gpu/drm/amd/display/modules/power/power_helpers.c b/drivers/gpu/drm/amd/display/modules/power/power_helpers.c index a522a7c02911..1675314a3ff2 100644 --- a/drivers/gpu/drm/amd/display/modules/power/power_helpers.c +++ b/drivers/gpu/drm/amd/display/modules/power/power_helpers.c @@ -839,6 +839,8 @@ bool is_psr_su_specific_panel(struct dc_link *link) ((dpcd_caps->sink_dev_id_str[1] == 0x08 && dpcd_caps->sink_dev_id_str[0] == 0x08) || (dpcd_caps->sink_dev_id_str[1] == 0x08 && dpcd_caps->sink_dev_id_str[0] == 0x07))) isPSRSUSupported = false; + else if (dpcd_caps->sink_dev_id_str[1] == 0x08 && dpcd_caps->sink_dev_id_str[0] == 0x03) + isPSRSUSupported = false; else if (dpcd_caps->psr_info.force_psrsu_cap == 0x1) isPSRSUSupported = true; } From d5a348d96e4e2b924fa83e729f8791c03a4f8e24 Mon Sep 17 00:00:00 2001 From: Joshua Ashton Date: Thu, 16 Nov 2023 18:57:45 -0100 Subject: [PATCH 26/92] drm/amd/display: add plane degamma TF driver-specific property Allow userspace to tell the kernel driver the input space and, therefore, uses correct predefined transfer function (TF) to go from encoded values to linear values. v2: - rename TF enum prefix from DRM_ to AMDGPU_ (Harry) - remove HLG TF Reviewed-by: Harry Wentland Signed-off-by: Joshua Ashton Co-developed-by: Melissa Wen Signed-off-by: Melissa Wen Signed-off-by: Alex Deucher --- drivers/gpu/drm/amd/amdgpu/amdgpu_mode.h | 5 +++++ .../gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.h | 19 +++++++++++++++++ .../amd/display/amdgpu_dm/amdgpu_dm_color.c | 21 +++++++++++++++++++ .../amd/display/amdgpu_dm/amdgpu_dm_plane.c | 19 +++++++++++++++-- 4 files changed, 62 insertions(+), 2 deletions(-) diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_mode.h b/drivers/gpu/drm/amd/amdgpu/amdgpu_mode.h index a63b01aa230d..a1b7ae5c2149 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_mode.h +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_mode.h @@ -355,6 +355,11 @@ struct amdgpu_mode_info { * size of degamma LUT as supported by the driver (read-only). */ struct drm_property *plane_degamma_lut_size_property; + /** + * @plane_degamma_tf_property: Plane pre-defined transfer function to + * to go from scanout/encoded values to linear values. + */ + struct drm_property *plane_degamma_tf_property; }; #define AMDGPU_MAX_BL_LEVEL 0xFF 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 446a3962d9c9..41bf0cf56433 100644 --- a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.h +++ b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.h @@ -724,6 +724,18 @@ struct amdgpu_dm_wb_connector { extern const struct amdgpu_ip_block_version dm_ip_block; +enum amdgpu_transfer_function { + AMDGPU_TRANSFER_FUNCTION_DEFAULT, + AMDGPU_TRANSFER_FUNCTION_SRGB, + AMDGPU_TRANSFER_FUNCTION_BT709, + AMDGPU_TRANSFER_FUNCTION_PQ, + AMDGPU_TRANSFER_FUNCTION_LINEAR, + AMDGPU_TRANSFER_FUNCTION_UNITY, + AMDGPU_TRANSFER_FUNCTION_GAMMA22, + AMDGPU_TRANSFER_FUNCTION_GAMMA24, + AMDGPU_TRANSFER_FUNCTION_GAMMA26, +}; + struct dm_plane_state { struct drm_plane_state base; struct dc_plane_state *dc_state; @@ -737,6 +749,13 @@ struct dm_plane_state { * The blob (if not NULL) is an array of &struct drm_color_lut. */ struct drm_property_blob *degamma_lut; + /** + * @degamma_tf: + * + * Predefined transfer function to tell DC driver the input space to + * linearize. + */ + enum amdgpu_transfer_function degamma_tf; }; struct dm_crtc_state { diff --git a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_color.c b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_color.c index 421af919387f..8e02d2b1249d 100644 --- a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_color.c +++ b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_color.c @@ -85,6 +85,18 @@ void amdgpu_dm_init_color_mod(void) } #ifdef AMD_PRIVATE_COLOR +static const struct drm_prop_enum_list amdgpu_transfer_function_enum_list[] = { + { AMDGPU_TRANSFER_FUNCTION_DEFAULT, "Default" }, + { AMDGPU_TRANSFER_FUNCTION_SRGB, "sRGB" }, + { AMDGPU_TRANSFER_FUNCTION_BT709, "BT.709" }, + { AMDGPU_TRANSFER_FUNCTION_PQ, "PQ (Perceptual Quantizer)" }, + { AMDGPU_TRANSFER_FUNCTION_LINEAR, "Linear" }, + { AMDGPU_TRANSFER_FUNCTION_UNITY, "Unity" }, + { AMDGPU_TRANSFER_FUNCTION_GAMMA22, "Gamma 2.2" }, + { AMDGPU_TRANSFER_FUNCTION_GAMMA24, "Gamma 2.4" }, + { AMDGPU_TRANSFER_FUNCTION_GAMMA26, "Gamma 2.6" }, +}; + int amdgpu_dm_create_color_properties(struct amdgpu_device *adev) { @@ -105,6 +117,15 @@ amdgpu_dm_create_color_properties(struct amdgpu_device *adev) return -ENOMEM; adev->mode_info.plane_degamma_lut_size_property = prop; + prop = drm_property_create_enum(adev_to_drm(adev), + DRM_MODE_PROP_ENUM, + "AMD_PLANE_DEGAMMA_TF", + amdgpu_transfer_function_enum_list, + ARRAY_SIZE(amdgpu_transfer_function_enum_list)); + if (!prop) + return -ENOMEM; + adev->mode_info.plane_degamma_tf_property = prop; + return 0; } #endif diff --git a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_plane.c b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_plane.c index b1e50cfa9f83..d5e12e05d161 100644 --- a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_plane.c +++ b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_plane.c @@ -1337,8 +1337,11 @@ static void amdgpu_dm_plane_drm_plane_reset(struct drm_plane *plane) amdgpu_state = kzalloc(sizeof(*amdgpu_state), GFP_KERNEL); WARN_ON(amdgpu_state == NULL); - if (amdgpu_state) - __drm_atomic_helper_plane_reset(plane, &amdgpu_state->base); + if (!amdgpu_state) + return; + + __drm_atomic_helper_plane_reset(plane, &amdgpu_state->base); + amdgpu_state->degamma_tf = AMDGPU_TRANSFER_FUNCTION_DEFAULT; } static struct drm_plane_state *amdgpu_dm_plane_drm_plane_duplicate_state(struct drm_plane *plane) @@ -1361,6 +1364,8 @@ static struct drm_plane_state *amdgpu_dm_plane_drm_plane_duplicate_state(struct dm_plane_state->degamma_lut = drm_property_blob_get(old_dm_plane_state->degamma_lut); + dm_plane_state->degamma_tf = old_dm_plane_state->degamma_tf; + return &dm_plane_state->base; } @@ -1455,6 +1460,9 @@ dm_atomic_plane_attach_color_mgmt_properties(struct amdgpu_display_manager *dm, drm_object_attach_property(&plane->base, mode_info.plane_degamma_lut_size_property, MAX_COLOR_LUT_ENTRIES); + drm_object_attach_property(&plane->base, + dm->adev->mode_info.plane_degamma_tf_property, + AMDGPU_TRANSFER_FUNCTION_DEFAULT); } } @@ -1477,6 +1485,11 @@ dm_atomic_plane_set_property(struct drm_plane *plane, &replaced); dm_plane_state->base.color_mgmt_changed |= replaced; return ret; + } else if (property == adev->mode_info.plane_degamma_tf_property) { + if (dm_plane_state->degamma_tf != val) { + dm_plane_state->degamma_tf = val; + dm_plane_state->base.color_mgmt_changed = 1; + } } else { drm_dbg_atomic(plane->dev, "[PLANE:%d:%s] unknown property [PROP:%d:%s]]\n", @@ -1500,6 +1513,8 @@ dm_atomic_plane_get_property(struct drm_plane *plane, if (property == adev->mode_info.plane_degamma_lut_property) { *val = (dm_plane_state->degamma_lut) ? dm_plane_state->degamma_lut->base.id : 0; + } else if (property == adev->mode_info.plane_degamma_tf_property) { + *val = dm_plane_state->degamma_tf; } else { return -EINVAL; } From 5a3b965b5810bd602d2c7d8ea79ffe8c6e81268d Mon Sep 17 00:00:00 2001 From: Melissa Wen Date: Thu, 16 Nov 2023 18:57:46 -0100 Subject: [PATCH 27/92] drm/amd/display: explicitly define EOTF and inverse EOTF Instead of relying on color block names to get the transfer function intention regarding encoding pixel's luminance, define supported Electro-Optical Transfer Functions (EOTFs) and inverse EOTFs, that includes pure gamma or standardized transfer functions. v3: - squash linear and unity TFs to identity (Pekka) - define the right TFs for BT.709 (Pekka and Harry) - add comment about AMD TF coefficients Suggested-by: Harry Wentland Reviewed-by: Harry Wentland Signed-off-by: Melissa Wen Signed-off-by: Alex Deucher --- .../gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.h | 27 +++++--- .../amd/display/amdgpu_dm/amdgpu_dm_color.c | 67 ++++++++++++++----- 2 files changed, 71 insertions(+), 23 deletions(-) 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 41bf0cf56433..baa73a6f47a9 100644 --- a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.h +++ b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.h @@ -724,16 +724,27 @@ struct amdgpu_dm_wb_connector { extern const struct amdgpu_ip_block_version dm_ip_block; +/* enum amdgpu_transfer_function: pre-defined transfer function supported by AMD. + * + * It includes standardized transfer functions and pure power functions. The + * transfer function coefficients are available at modules/color/color_gamma.c + */ enum amdgpu_transfer_function { AMDGPU_TRANSFER_FUNCTION_DEFAULT, - AMDGPU_TRANSFER_FUNCTION_SRGB, - AMDGPU_TRANSFER_FUNCTION_BT709, - AMDGPU_TRANSFER_FUNCTION_PQ, - AMDGPU_TRANSFER_FUNCTION_LINEAR, - AMDGPU_TRANSFER_FUNCTION_UNITY, - AMDGPU_TRANSFER_FUNCTION_GAMMA22, - AMDGPU_TRANSFER_FUNCTION_GAMMA24, - AMDGPU_TRANSFER_FUNCTION_GAMMA26, + AMDGPU_TRANSFER_FUNCTION_SRGB_EOTF, + AMDGPU_TRANSFER_FUNCTION_BT709_INV_OETF, + AMDGPU_TRANSFER_FUNCTION_PQ_EOTF, + AMDGPU_TRANSFER_FUNCTION_IDENTITY, + AMDGPU_TRANSFER_FUNCTION_GAMMA22_EOTF, + AMDGPU_TRANSFER_FUNCTION_GAMMA24_EOTF, + AMDGPU_TRANSFER_FUNCTION_GAMMA26_EOTF, + AMDGPU_TRANSFER_FUNCTION_SRGB_INV_EOTF, + AMDGPU_TRANSFER_FUNCTION_BT709_OETF, + AMDGPU_TRANSFER_FUNCTION_PQ_INV_EOTF, + AMDGPU_TRANSFER_FUNCTION_GAMMA22_INV_EOTF, + AMDGPU_TRANSFER_FUNCTION_GAMMA24_INV_EOTF, + AMDGPU_TRANSFER_FUNCTION_GAMMA26_INV_EOTF, + AMDGPU_TRANSFER_FUNCTION_COUNT }; struct dm_plane_state { diff --git a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_color.c b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_color.c index 8e02d2b1249d..a32abbd93a61 100644 --- a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_color.c +++ b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_color.c @@ -85,18 +85,57 @@ void amdgpu_dm_init_color_mod(void) } #ifdef AMD_PRIVATE_COLOR -static const struct drm_prop_enum_list amdgpu_transfer_function_enum_list[] = { - { AMDGPU_TRANSFER_FUNCTION_DEFAULT, "Default" }, - { AMDGPU_TRANSFER_FUNCTION_SRGB, "sRGB" }, - { AMDGPU_TRANSFER_FUNCTION_BT709, "BT.709" }, - { AMDGPU_TRANSFER_FUNCTION_PQ, "PQ (Perceptual Quantizer)" }, - { AMDGPU_TRANSFER_FUNCTION_LINEAR, "Linear" }, - { AMDGPU_TRANSFER_FUNCTION_UNITY, "Unity" }, - { AMDGPU_TRANSFER_FUNCTION_GAMMA22, "Gamma 2.2" }, - { AMDGPU_TRANSFER_FUNCTION_GAMMA24, "Gamma 2.4" }, - { AMDGPU_TRANSFER_FUNCTION_GAMMA26, "Gamma 2.6" }, +static const char * const +amdgpu_transfer_function_names[] = { + [AMDGPU_TRANSFER_FUNCTION_DEFAULT] = "Default", + [AMDGPU_TRANSFER_FUNCTION_IDENTITY] = "Identity", + [AMDGPU_TRANSFER_FUNCTION_SRGB_EOTF] = "sRGB EOTF", + [AMDGPU_TRANSFER_FUNCTION_BT709_INV_OETF] = "BT.709 inv_OETF", + [AMDGPU_TRANSFER_FUNCTION_PQ_EOTF] = "PQ EOTF", + [AMDGPU_TRANSFER_FUNCTION_GAMMA22_EOTF] = "Gamma 2.2 EOTF", + [AMDGPU_TRANSFER_FUNCTION_GAMMA24_EOTF] = "Gamma 2.4 EOTF", + [AMDGPU_TRANSFER_FUNCTION_GAMMA26_EOTF] = "Gamma 2.6 EOTF", + [AMDGPU_TRANSFER_FUNCTION_SRGB_INV_EOTF] = "sRGB inv_EOTF", + [AMDGPU_TRANSFER_FUNCTION_BT709_OETF] = "BT.709 OETF", + [AMDGPU_TRANSFER_FUNCTION_PQ_INV_EOTF] = "PQ inv_EOTF", + [AMDGPU_TRANSFER_FUNCTION_GAMMA22_INV_EOTF] = "Gamma 2.2 inv_EOTF", + [AMDGPU_TRANSFER_FUNCTION_GAMMA24_INV_EOTF] = "Gamma 2.4 inv_EOTF", + [AMDGPU_TRANSFER_FUNCTION_GAMMA26_INV_EOTF] = "Gamma 2.6 inv_EOTF", }; +static const u32 amdgpu_eotf = + BIT(AMDGPU_TRANSFER_FUNCTION_SRGB_EOTF) | + BIT(AMDGPU_TRANSFER_FUNCTION_BT709_INV_OETF) | + BIT(AMDGPU_TRANSFER_FUNCTION_PQ_EOTF) | + BIT(AMDGPU_TRANSFER_FUNCTION_GAMMA22_EOTF) | + BIT(AMDGPU_TRANSFER_FUNCTION_GAMMA24_EOTF) | + BIT(AMDGPU_TRANSFER_FUNCTION_GAMMA26_EOTF); + +static struct drm_property * +amdgpu_create_tf_property(struct drm_device *dev, + const char *name, + u32 supported_tf) +{ + u32 transfer_functions = supported_tf | + BIT(AMDGPU_TRANSFER_FUNCTION_DEFAULT) | + BIT(AMDGPU_TRANSFER_FUNCTION_IDENTITY); + struct drm_prop_enum_list enum_list[AMDGPU_TRANSFER_FUNCTION_COUNT]; + int i, len; + + len = 0; + for (i = 0; i < AMDGPU_TRANSFER_FUNCTION_COUNT; i++) { + if ((transfer_functions & BIT(i)) == 0) + continue; + + enum_list[len].type = i; + enum_list[len].name = amdgpu_transfer_function_names[i]; + len++; + } + + return drm_property_create_enum(dev, DRM_MODE_PROP_ENUM, + name, enum_list, len); +} + int amdgpu_dm_create_color_properties(struct amdgpu_device *adev) { @@ -117,11 +156,9 @@ amdgpu_dm_create_color_properties(struct amdgpu_device *adev) return -ENOMEM; adev->mode_info.plane_degamma_lut_size_property = prop; - prop = drm_property_create_enum(adev_to_drm(adev), - DRM_MODE_PROP_ENUM, - "AMD_PLANE_DEGAMMA_TF", - amdgpu_transfer_function_enum_list, - ARRAY_SIZE(amdgpu_transfer_function_enum_list)); + prop = amdgpu_create_tf_property(adev_to_drm(adev), + "AMD_PLANE_DEGAMMA_TF", + amdgpu_eotf); if (!prop) return -ENOMEM; adev->mode_info.plane_degamma_tf_property = prop; From e4cddd51bfab2a40529a4af35bd2c912b5a0c239 Mon Sep 17 00:00:00 2001 From: Melissa Wen Date: Thu, 16 Nov 2023 18:57:47 -0100 Subject: [PATCH 28/92] drm/amd/display: document AMDGPU pre-defined transfer functions Brief documentation about pre-defined transfer function usage on AMD display driver and standardized EOTFs and inverse EOTFs. v3: - Document BT709 OETF (Pekka) - Fix description of sRGB and pure power funcs (Pekka) v4: - Add description of linear and non-linear forms (Harry) Reviewed-by: Harry Wentland Co-developed-by: Harry Wentland Signed-off-by: Harry Wentland Signed-off-by: Melissa Wen Signed-off-by: Alex Deucher --- .../amd/display/amdgpu_dm/amdgpu_dm_color.c | 62 +++++++++++++++++++ 1 file changed, 62 insertions(+) diff --git a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_color.c b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_color.c index a32abbd93a61..49faaf606cfe 100644 --- a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_color.c +++ b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_color.c @@ -85,6 +85,68 @@ void amdgpu_dm_init_color_mod(void) } #ifdef AMD_PRIVATE_COLOR +/* Pre-defined Transfer Functions (TF) + * + * AMD driver supports pre-defined mathematical functions for transferring + * between encoded values and optical/linear space. Depending on HW color caps, + * ROMs and curves built by the AMD color module support these transforms. + * + * The driver-specific color implementation exposes properties for pre-blending + * degamma TF, shaper TF (before 3D LUT), and blend(dpp.ogam) TF and + * post-blending regamma (mpc.ogam) TF. However, only pre-blending degamma + * supports ROM curves. AMD color module uses pre-defined coefficients to build + * curves for the other blocks. What can be done by each color block is + * described by struct dpp_color_capsand struct mpc_color_caps. + * + * AMD driver-specific color API exposes the following pre-defined transfer + * functions: + * + * - Identity: linear/identity relationship between pixel value and + * luminance value; + * - Gamma 2.2, Gamma 2.4, Gamma 2.6: pure power functions; + * - sRGB: 2.4: The piece-wise transfer function from IEC 61966-2-1:1999; + * - BT.709: has a linear segment in the bottom part and then a power function + * with a 0.45 (~1/2.22) gamma for the rest of the range; standardized by + * ITU-R BT.709-6; + * - PQ (Perceptual Quantizer): used for HDR display, allows luminance range + * capability of 0 to 10,000 nits; standardized by SMPTE ST 2084. + * + * The AMD color model is designed with an assumption that SDR (sRGB, BT.709, + * Gamma 2.2, etc.) peak white maps (normalized to 1.0 FP) to 80 nits in the PQ + * system. This has the implication that PQ EOTF (non-linear to linear) maps to + * [0.0..125.0] where 125.0 = 10,000 nits / 80 nits. + * + * Non-linear and linear forms are described in the table below: + * + * ┌───────────┬─────────────────────┬──────────────────────┐ + * │ │ Non-linear │ Linear │ + * ├───────────┼─────────────────────┼──────────────────────┤ + * │ sRGB │ UNORM or [0.0, 1.0] │ [0.0, 1.0] │ + * ├───────────┼─────────────────────┼──────────────────────┤ + * │ BT709 │ UNORM or [0.0, 1.0] │ [0.0, 1.0] │ + * ├───────────┼─────────────────────┼──────────────────────┤ + * │ Gamma 2.x │ UNORM or [0.0, 1.0] │ [0.0, 1.0] │ + * ├───────────┼─────────────────────┼──────────────────────┤ + * │ PQ │ UNORM or FP16 CCCS* │ [0.0, 125.0] │ + * ├───────────┼─────────────────────┼──────────────────────┤ + * │ Identity │ UNORM or FP16 CCCS* │ [0.0, 1.0] or CCCS** │ + * └───────────┴─────────────────────┴──────────────────────┘ + * * CCCS: Windows canonical composition color space + * ** Respectively + * + * In the driver-specific API, color block names attached to TF properties + * suggest the intention regarding non-linear encoding pixel's luminance + * values. As some newer encodings don't use gamma curve, we make encoding and + * decoding explicit by defining an enum list of transfer functions supported + * in terms of EOTF and inverse EOTF, where: + * + * - EOTF (electro-optical transfer function): is the transfer function to go + * from the encoded value to an optical (linear) value. De-gamma functions + * traditionally do this. + * - Inverse EOTF (simply the inverse of the EOTF): is usually intended to go + * from an optical/linear space (which might have been used for blending) + * back to the encoded values. Gamma functions traditionally do this. + */ static const char * const amdgpu_transfer_function_names[] = { [AMDGPU_TRANSFER_FUNCTION_DEFAULT] = "Default", From ec7b2a55463ea50401a8146793b61ee590255a45 Mon Sep 17 00:00:00 2001 From: Joshua Ashton Date: Thu, 16 Nov 2023 18:57:48 -0100 Subject: [PATCH 29/92] drm/amd/display: add plane HDR multiplier driver-specific property Multiplier to 'gain' the plane. When PQ is decoded using the fixed func transfer function to the internal FP16 fb, 1.0 -> 80 nits (on AMD at least) When sRGB is decoded, 1.0 -> 1.0. Therefore, 1.0 multiplier = 80 nits for SDR content. So if you want, 203 nits for SDR content, pass in (203.0 / 80.0). v4: - comment about the PQ TF need for L-to-NL (from Harry's review) Reviewed-by: Harry Wentland Signed-off-by: Joshua Ashton Co-developed-by: Melissa Wen Signed-off-by: Melissa Wen Signed-off-by: Alex Deucher --- drivers/gpu/drm/amd/amdgpu/amdgpu_mode.h | 4 ++++ .../gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.h | 17 +++++++++++++++++ .../drm/amd/display/amdgpu_dm/amdgpu_dm_color.c | 6 ++++++ .../drm/amd/display/amdgpu_dm/amdgpu_dm_plane.c | 13 +++++++++++++ 4 files changed, 40 insertions(+) diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_mode.h b/drivers/gpu/drm/amd/amdgpu/amdgpu_mode.h index a1b7ae5c2149..218623e1b9aa 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_mode.h +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_mode.h @@ -360,6 +360,10 @@ struct amdgpu_mode_info { * to go from scanout/encoded values to linear values. */ struct drm_property *plane_degamma_tf_property; + /** + * @plane_hdr_mult_property: + */ + struct drm_property *plane_hdr_mult_property; }; #define AMDGPU_MAX_BL_LEVEL 0xFF 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 baa73a6f47a9..d36d68d517a8 100644 --- a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.h +++ b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.h @@ -55,6 +55,9 @@ #define HDMI_AMD_VENDOR_SPECIFIC_DATA_BLOCK_IEEE_REGISTRATION_ID 0x00001A #define AMD_VSDB_VERSION_3_FEATURECAP_REPLAYMODE 0x40 #define HDMI_AMD_VENDOR_SPECIFIC_DATA_BLOCK_VERSION_3 0x3 + +#define AMDGPU_HDR_MULT_DEFAULT (0x100000000LL) + /* #include "include/amdgpu_dal_power_if.h" #include "amdgpu_dm_irq.h" @@ -767,6 +770,20 @@ struct dm_plane_state { * linearize. */ enum amdgpu_transfer_function degamma_tf; + /** + * @hdr_mult: + * + * Multiplier to 'gain' the plane. When PQ is decoded using the fixed + * func transfer function to the internal FP16 fb, 1.0 -> 80 nits (on + * AMD at least). When sRGB is decoded, 1.0 -> 1.0, obviously. + * Therefore, 1.0 multiplier = 80 nits for SDR content. So if you + * want, 203 nits for SDR content, pass in (203.0 / 80.0). Format is + * S31.32 sign-magnitude. + * + * HDR multiplier can wide range beyond [0.0, 1.0]. This means that PQ + * TF is needed for any subsequent linear-to-non-linear transforms. + */ + __u64 hdr_mult; }; struct dm_crtc_state { diff --git a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_color.c b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_color.c index 49faaf606cfe..b5b34a9209e4 100644 --- a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_color.c +++ b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_color.c @@ -225,6 +225,12 @@ amdgpu_dm_create_color_properties(struct amdgpu_device *adev) return -ENOMEM; adev->mode_info.plane_degamma_tf_property = prop; + prop = drm_property_create_range(adev_to_drm(adev), + 0, "AMD_PLANE_HDR_MULT", 0, U64_MAX); + if (!prop) + return -ENOMEM; + adev->mode_info.plane_hdr_mult_property = prop; + return 0; } #endif diff --git a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_plane.c b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_plane.c index d5e12e05d161..45a2c9b36630 100644 --- a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_plane.c +++ b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_plane.c @@ -1342,6 +1342,7 @@ static void amdgpu_dm_plane_drm_plane_reset(struct drm_plane *plane) __drm_atomic_helper_plane_reset(plane, &amdgpu_state->base); amdgpu_state->degamma_tf = AMDGPU_TRANSFER_FUNCTION_DEFAULT; + amdgpu_state->hdr_mult = AMDGPU_HDR_MULT_DEFAULT; } static struct drm_plane_state *amdgpu_dm_plane_drm_plane_duplicate_state(struct drm_plane *plane) @@ -1365,6 +1366,7 @@ static struct drm_plane_state *amdgpu_dm_plane_drm_plane_duplicate_state(struct drm_property_blob_get(old_dm_plane_state->degamma_lut); dm_plane_state->degamma_tf = old_dm_plane_state->degamma_tf; + dm_plane_state->hdr_mult = old_dm_plane_state->hdr_mult; return &dm_plane_state->base; } @@ -1464,6 +1466,10 @@ dm_atomic_plane_attach_color_mgmt_properties(struct amdgpu_display_manager *dm, dm->adev->mode_info.plane_degamma_tf_property, AMDGPU_TRANSFER_FUNCTION_DEFAULT); } + /* HDR MULT is always available */ + drm_object_attach_property(&plane->base, + dm->adev->mode_info.plane_hdr_mult_property, + AMDGPU_HDR_MULT_DEFAULT); } static int @@ -1490,6 +1496,11 @@ dm_atomic_plane_set_property(struct drm_plane *plane, dm_plane_state->degamma_tf = val; dm_plane_state->base.color_mgmt_changed = 1; } + } else if (property == adev->mode_info.plane_hdr_mult_property) { + if (dm_plane_state->hdr_mult != val) { + dm_plane_state->hdr_mult = val; + dm_plane_state->base.color_mgmt_changed = 1; + } } else { drm_dbg_atomic(plane->dev, "[PLANE:%d:%s] unknown property [PROP:%d:%s]]\n", @@ -1515,6 +1526,8 @@ dm_atomic_plane_get_property(struct drm_plane *plane, dm_plane_state->degamma_lut->base.id : 0; } else if (property == adev->mode_info.plane_degamma_tf_property) { *val = dm_plane_state->degamma_tf; + } else if (property == adev->mode_info.plane_hdr_mult_property) { + *val = dm_plane_state->hdr_mult; } else { return -EINVAL; } From a9210714d23190b44eed32f8bcadbe3b18d51a1d Mon Sep 17 00:00:00 2001 From: Harshit Mogalapalli Date: Fri, 8 Dec 2023 01:58:24 -0800 Subject: [PATCH 30/92] drm/amd/display: Fix memory leak in dm_set_writeback() 'wb_info' needs to be freed on error paths or it would leak the memory. Smatch pointed this out. Fixes: c81e13b929df ("drm/amd/display: Hande writeback request from userspace") Signed-off-by: Harshit Mogalapalli Reviewed-by: Alex Hung Signed-off-by: Alex Deucher --- drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.c | 2 ++ 1 file changed, 2 insertions(+) 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 e819d05755b3..6efa0c1bba9d 100644 --- a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.c +++ b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.c @@ -8879,12 +8879,14 @@ static void dm_set_writeback(struct amdgpu_display_manager *dm, acrtc = to_amdgpu_crtc(wb_conn->encoder.crtc); if (!acrtc) { DRM_ERROR("no amdgpu_crtc found\n"); + kfree(wb_info); return; } afb = to_amdgpu_framebuffer(new_con_state->writeback_job->fb); if (!afb) { DRM_ERROR("No amdgpu_framebuffer found\n"); + kfree(wb_info); return; } From 4e95669ecb03d797355bc23871c5c43b9475d3dc Mon Sep 17 00:00:00 2001 From: Vignesh Chander Date: Fri, 8 Dec 2023 12:00:26 -0600 Subject: [PATCH 31/92] drm/amdgpu: xgmi_fill_topology_info 1. Use the mirrored topology info to fill links for VF. The new solution is required to simplify and optimize host driver logic. Only use the new solution for VFs that support full duplex and extended_peer_link_info otherwise the info would be incomplete. 2. avoid calling extended_link_info on VF as its not supported Signed-off-by: Vignesh Chander Reviewed-by: Zhigang Luo Reviewed-by: Jonathan Kim Signed-off-by: Alex Deucher --- drivers/gpu/drm/amd/amdgpu/amdgpu_psp.c | 4 +- drivers/gpu/drm/amd/amdgpu/amdgpu_xgmi.c | 58 ++++++++++++++++++++---- 2 files changed, 52 insertions(+), 10 deletions(-) diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_psp.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_psp.c index a21045d018f2..1bf975b8d083 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_psp.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_psp.c @@ -1433,8 +1433,8 @@ int psp_xgmi_get_topology_info(struct psp_context *psp, get_extended_data) || amdgpu_ip_version(psp->adev, MP0_HWIP, 0) == IP_VERSION(13, 0, 6); - bool ta_port_num_support = psp->xgmi_context.xgmi_ta_caps & - EXTEND_PEER_LINK_INFO_CMD_FLAG; + bool ta_port_num_support = amdgpu_sriov_vf(psp->adev) ? 0 : + psp->xgmi_context.xgmi_ta_caps & EXTEND_PEER_LINK_INFO_CMD_FLAG; /* popluate the shared output buffer rather than the cmd input buffer * with node_ids as the input for GET_PEER_LINKS command execution. diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_xgmi.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_xgmi.c index 44d8c1a11e1b..9a95b9f226b8 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_xgmi.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_xgmi.c @@ -823,6 +823,28 @@ static int amdgpu_xgmi_initialize_hive_get_data_partition(struct amdgpu_hive_inf return 0; } +static void amdgpu_xgmi_fill_topology_info(struct amdgpu_device *adev, + struct amdgpu_device *peer_adev) +{ + struct psp_xgmi_topology_info *top_info = &adev->psp.xgmi_context.top_info; + struct psp_xgmi_topology_info *peer_info = &peer_adev->psp.xgmi_context.top_info; + + for (int i = 0; i < peer_info->num_nodes; i++) { + if (peer_info->nodes[i].node_id == adev->gmc.xgmi.node_id) { + for (int j = 0; j < top_info->num_nodes; j++) { + if (top_info->nodes[j].node_id == peer_adev->gmc.xgmi.node_id) { + peer_info->nodes[i].num_hops = top_info->nodes[j].num_hops; + peer_info->nodes[i].is_sharing_enabled = + top_info->nodes[j].is_sharing_enabled; + peer_info->nodes[i].num_links = + top_info->nodes[j].num_links; + return; + } + } + } + } +} + int amdgpu_xgmi_add_device(struct amdgpu_device *adev) { struct psp_xgmi_topology_info *top_info; @@ -897,18 +919,38 @@ int amdgpu_xgmi_add_device(struct amdgpu_device *adev) goto exit_unlock; } - /* get latest topology info for each device from psp */ - list_for_each_entry(tmp_adev, &hive->device_list, gmc.xgmi.head) { - ret = psp_xgmi_get_topology_info(&tmp_adev->psp, count, - &tmp_adev->psp.xgmi_context.top_info, false); + if (amdgpu_sriov_vf(adev) && + adev->psp.xgmi_context.xgmi_ta_caps & EXTEND_PEER_LINK_INFO_CMD_FLAG) { + /* only get topology for VF being init if it can support full duplex */ + ret = psp_xgmi_get_topology_info(&adev->psp, count, + &adev->psp.xgmi_context.top_info, false); if (ret) { - dev_err(tmp_adev->dev, + dev_err(adev->dev, "XGMI: Get topology failure on device %llx, hive %llx, ret %d", - tmp_adev->gmc.xgmi.node_id, - tmp_adev->gmc.xgmi.hive_id, ret); - /* To do : continue with some node failed or disable the whole hive */ + adev->gmc.xgmi.node_id, + adev->gmc.xgmi.hive_id, ret); + /* To do: continue with some node failed or disable the whole hive*/ goto exit_unlock; } + + /* fill the topology info for peers instead of getting from PSP */ + list_for_each_entry(tmp_adev, &hive->device_list, gmc.xgmi.head) { + amdgpu_xgmi_fill_topology_info(adev, tmp_adev); + } + } else { + /* get latest topology info for each device from psp */ + list_for_each_entry(tmp_adev, &hive->device_list, gmc.xgmi.head) { + ret = psp_xgmi_get_topology_info(&tmp_adev->psp, count, + &tmp_adev->psp.xgmi_context.top_info, false); + if (ret) { + dev_err(tmp_adev->dev, + "XGMI: Get topology failure on device %llx, hive %llx, ret %d", + tmp_adev->gmc.xgmi.node_id, + tmp_adev->gmc.xgmi.hive_id, ret); + /* To do : continue with some node failed or disable the whole hive */ + goto exit_unlock; + } + } } /* get topology again for hives that support extended data */ From 1819200166ce511ac298dc96b9b17eb655a9edc4 Mon Sep 17 00:00:00 2001 From: Felix Kuehling Date: Thu, 24 Aug 2023 14:41:30 -0400 Subject: [PATCH 32/92] drm/amdkfd: Export DMABufs from KFD using GEM handles Create GEM handles for exporting DMABufs using GEM-Prime APIs. The GEM handles are created in a drm_client_dev context to avoid exposing them in user mode contexts through a DMABuf import. Signed-off-by: Felix Kuehling Reviewed-by: Ramesh Errabolu Signed-off-by: Alex Deucher --- drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd.c | 11 +++++++ drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd.h | 5 +++ .../gpu/drm/amd/amdgpu/amdgpu_amdkfd_gpuvm.c | 33 +++++++++++++++---- drivers/gpu/drm/amd/amdkfd/kfd_chardev.c | 4 +-- 4 files changed, 44 insertions(+), 9 deletions(-) diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd.c index 2d22f7d45512..067690ba7bff 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd.c @@ -142,6 +142,7 @@ void amdgpu_amdkfd_device_init(struct amdgpu_device *adev) { int i; int last_valid_bit; + int ret; amdgpu_amdkfd_gpuvm_init_mem_limits(); @@ -160,6 +161,12 @@ void amdgpu_amdkfd_device_init(struct amdgpu_device *adev) .enable_mes = adev->enable_mes, }; + ret = drm_client_init(&adev->ddev, &adev->kfd.client, "kfd", NULL); + if (ret) { + dev_err(adev->dev, "Failed to init DRM client: %d\n", ret); + return; + } + /* this is going to have a few of the MSBs set that we need to * clear */ @@ -198,6 +205,10 @@ void amdgpu_amdkfd_device_init(struct amdgpu_device *adev) adev->kfd.init_complete = kgd2kfd_device_init(adev->kfd.dev, &gpu_resources); + if (adev->kfd.init_complete) + drm_client_register(&adev->kfd.client); + else + drm_client_release(&adev->kfd.client); amdgpu_amdkfd_total_mem_size += adev->gmc.real_vram_size; diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd.h b/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd.h index 16794c2eea35..02973f5c8caf 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd.h +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd.h @@ -33,6 +33,7 @@ #include #include #include +#include #include "amdgpu_sync.h" #include "amdgpu_vm.h" #include "amdgpu_xcp.h" @@ -83,6 +84,7 @@ struct kgd_mem { struct amdgpu_sync sync; + uint32_t gem_handle; bool aql_queue; bool is_imported; }; @@ -105,6 +107,9 @@ struct amdgpu_kfd_dev { /* HMM page migration MEMORY_DEVICE_PRIVATE mapping */ struct dev_pagemap pgmap; + + /* Client for KFD BO GEM handle allocations */ + struct drm_client_dev client; }; enum kgd_engine_type { diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd_gpuvm.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd_gpuvm.c index 73288f9ccaf8..ae7dfaf59159 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd_gpuvm.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd_gpuvm.c @@ -25,6 +25,7 @@ #include #include #include +#include #include #include @@ -806,13 +807,22 @@ kfd_mem_dmaunmap_attachment(struct kgd_mem *mem, static int kfd_mem_export_dmabuf(struct kgd_mem *mem) { if (!mem->dmabuf) { - struct dma_buf *ret = amdgpu_gem_prime_export( - &mem->bo->tbo.base, + struct amdgpu_device *bo_adev; + struct dma_buf *dmabuf; + int r, fd; + + bo_adev = amdgpu_ttm_adev(mem->bo->tbo.bdev); + r = drm_gem_prime_handle_to_fd(&bo_adev->ddev, bo_adev->kfd.client.file, + mem->gem_handle, mem->alloc_flags & KFD_IOC_ALLOC_MEM_FLAGS_WRITABLE ? - DRM_RDWR : 0); - if (IS_ERR(ret)) - return PTR_ERR(ret); - mem->dmabuf = ret; + DRM_RDWR : 0, &fd); + if (r) + return r; + dmabuf = dma_buf_get(fd); + close_fd(fd); + if (WARN_ON_ONCE(IS_ERR(dmabuf))) + return PTR_ERR(dmabuf); + mem->dmabuf = dmabuf; } return 0; @@ -1778,6 +1788,9 @@ int amdgpu_amdkfd_gpuvm_alloc_memory_of_gpu( pr_debug("Failed to allow vma node access. ret %d\n", ret); goto err_node_allow; } + ret = drm_gem_handle_create(adev->kfd.client.file, gobj, &(*mem)->gem_handle); + if (ret) + goto err_gem_handle_create; bo = gem_to_amdgpu_bo(gobj); if (bo_type == ttm_bo_type_sg) { bo->tbo.sg = sg; @@ -1829,6 +1842,8 @@ int amdgpu_amdkfd_gpuvm_alloc_memory_of_gpu( err_pin_bo: err_validate_bo: remove_kgd_mem_from_kfd_bo_list(*mem, avm->process_info); + drm_gem_handle_delete(adev->kfd.client.file, (*mem)->gem_handle); +err_gem_handle_create: drm_vma_node_revoke(&gobj->vma_node, drm_priv); err_node_allow: /* Don't unreserve system mem limit twice */ @@ -1941,8 +1956,12 @@ int amdgpu_amdkfd_gpuvm_free_memory_of_gpu( /* Free the BO*/ drm_vma_node_revoke(&mem->bo->tbo.base.vma_node, drm_priv); - if (mem->dmabuf) + if (!mem->is_imported) + drm_gem_handle_delete(adev->kfd.client.file, mem->gem_handle); + if (mem->dmabuf) { dma_buf_put(mem->dmabuf); + mem->dmabuf = NULL; + } mutex_destroy(&mem->lock); /* If this releases the last reference, it will end up calling diff --git a/drivers/gpu/drm/amd/amdkfd/kfd_chardev.c b/drivers/gpu/drm/amd/amdkfd/kfd_chardev.c index f6d4748c1980..b4cb907f80c6 100644 --- a/drivers/gpu/drm/amd/amdkfd/kfd_chardev.c +++ b/drivers/gpu/drm/amd/amdkfd/kfd_chardev.c @@ -1855,8 +1855,8 @@ static uint32_t get_process_num_bos(struct kfd_process *p) return num_of_bos; } -static int criu_get_prime_handle(struct kgd_mem *mem, int flags, - u32 *shared_fd) +static int criu_get_prime_handle(struct kgd_mem *mem, + int flags, u32 *shared_fd) { struct dma_buf *dmabuf; int ret; From 0188006d7c797a37c04471a2b4a34a7dfb21f363 Mon Sep 17 00:00:00 2001 From: Felix Kuehling Date: Tue, 19 Sep 2023 14:57:02 -0400 Subject: [PATCH 33/92] drm/amdkfd: Import DMABufs for interop through DRM MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit Use drm_gem_prime_fd_to_handle to import DMABufs for interop. This ensures that a GEM handle is created on import and that obj->dma_buf will be set and remain set as long as the object is imported into KFD. Signed-off-by: Felix Kuehling Reviewed-by: Ramesh Errabolu Reviewed-by: Xiaogang.Chen Acked-by: Christian König Signed-off-by: Alex Deucher --- drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd.h | 9 ++- .../gpu/drm/amd/amdgpu/amdgpu_amdkfd_gpuvm.c | 64 +++++++++++++------ drivers/gpu/drm/amd/amdkfd/kfd_chardev.c | 15 ++--- 3 files changed, 52 insertions(+), 36 deletions(-) diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd.h b/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd.h index 02973f5c8caf..cf6ed5fce291 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd.h +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd.h @@ -314,11 +314,10 @@ int amdgpu_amdkfd_gpuvm_restore_process_bos(void *process_info, struct dma_fence **ef); int amdgpu_amdkfd_gpuvm_get_vm_fault_info(struct amdgpu_device *adev, struct kfd_vm_fault_info *info); -int amdgpu_amdkfd_gpuvm_import_dmabuf(struct amdgpu_device *adev, - struct dma_buf *dmabuf, - uint64_t va, void *drm_priv, - struct kgd_mem **mem, uint64_t *size, - uint64_t *mmap_offset); +int amdgpu_amdkfd_gpuvm_import_dmabuf_fd(struct amdgpu_device *adev, int fd, + uint64_t va, void *drm_priv, + struct kgd_mem **mem, uint64_t *size, + uint64_t *mmap_offset); int amdgpu_amdkfd_gpuvm_export_dmabuf(struct kgd_mem *mem, struct dma_buf **dmabuf); void amdgpu_amdkfd_debug_mem_fence(struct amdgpu_device *adev); diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd_gpuvm.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd_gpuvm.c index ae7dfaf59159..48697b789342 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd_gpuvm.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd_gpuvm.c @@ -1956,8 +1956,7 @@ int amdgpu_amdkfd_gpuvm_free_memory_of_gpu( /* Free the BO*/ drm_vma_node_revoke(&mem->bo->tbo.base.vma_node, drm_priv); - if (!mem->is_imported) - drm_gem_handle_delete(adev->kfd.client.file, mem->gem_handle); + drm_gem_handle_delete(adev->kfd.client.file, mem->gem_handle); if (mem->dmabuf) { dma_buf_put(mem->dmabuf); mem->dmabuf = NULL; @@ -2313,34 +2312,26 @@ int amdgpu_amdkfd_gpuvm_get_vm_fault_info(struct amdgpu_device *adev, return 0; } -int amdgpu_amdkfd_gpuvm_import_dmabuf(struct amdgpu_device *adev, - struct dma_buf *dma_buf, - uint64_t va, void *drm_priv, - struct kgd_mem **mem, uint64_t *size, - uint64_t *mmap_offset) +static int import_obj_create(struct amdgpu_device *adev, + struct dma_buf *dma_buf, + struct drm_gem_object *obj, + uint64_t va, void *drm_priv, + struct kgd_mem **mem, uint64_t *size, + uint64_t *mmap_offset) { struct amdgpu_vm *avm = drm_priv_to_vm(drm_priv); - struct drm_gem_object *obj; struct amdgpu_bo *bo; int ret; - obj = amdgpu_gem_prime_import(adev_to_drm(adev), dma_buf); - if (IS_ERR(obj)) - return PTR_ERR(obj); - bo = gem_to_amdgpu_bo(obj); if (!(bo->preferred_domains & (AMDGPU_GEM_DOMAIN_VRAM | - AMDGPU_GEM_DOMAIN_GTT))) { + AMDGPU_GEM_DOMAIN_GTT))) /* Only VRAM and GTT BOs are supported */ - ret = -EINVAL; - goto err_put_obj; - } + return -EINVAL; *mem = kzalloc(sizeof(struct kgd_mem), GFP_KERNEL); - if (!*mem) { - ret = -ENOMEM; - goto err_put_obj; - } + if (!*mem) + return -ENOMEM; ret = drm_vma_node_allow(&obj->vma_node, drm_priv); if (ret) @@ -2390,8 +2381,41 @@ int amdgpu_amdkfd_gpuvm_import_dmabuf(struct amdgpu_device *adev, drm_vma_node_revoke(&obj->vma_node, drm_priv); err_free_mem: kfree(*mem); + return ret; +} + +int amdgpu_amdkfd_gpuvm_import_dmabuf_fd(struct amdgpu_device *adev, int fd, + uint64_t va, void *drm_priv, + struct kgd_mem **mem, uint64_t *size, + uint64_t *mmap_offset) +{ + struct drm_gem_object *obj; + uint32_t handle; + int ret; + + ret = drm_gem_prime_fd_to_handle(&adev->ddev, adev->kfd.client.file, fd, + &handle); + if (ret) + return ret; + obj = drm_gem_object_lookup(adev->kfd.client.file, handle); + if (!obj) { + ret = -EINVAL; + goto err_release_handle; + } + + ret = import_obj_create(adev, obj->dma_buf, obj, va, drm_priv, mem, size, + mmap_offset); + if (ret) + goto err_put_obj; + + (*mem)->gem_handle = handle; + + return 0; + err_put_obj: drm_gem_object_put(obj); +err_release_handle: + drm_gem_handle_delete(adev->kfd.client.file, handle); return ret; } diff --git a/drivers/gpu/drm/amd/amdkfd/kfd_chardev.c b/drivers/gpu/drm/amd/amdkfd/kfd_chardev.c index b4cb907f80c6..ce4c52ec34d8 100644 --- a/drivers/gpu/drm/amd/amdkfd/kfd_chardev.c +++ b/drivers/gpu/drm/amd/amdkfd/kfd_chardev.c @@ -1564,16 +1564,11 @@ static int kfd_ioctl_import_dmabuf(struct file *filep, { struct kfd_ioctl_import_dmabuf_args *args = data; struct kfd_process_device *pdd; - struct dma_buf *dmabuf; int idr_handle; uint64_t size; void *mem; int r; - dmabuf = dma_buf_get(args->dmabuf_fd); - if (IS_ERR(dmabuf)) - return PTR_ERR(dmabuf); - mutex_lock(&p->mutex); pdd = kfd_process_device_data_by_id(p, args->gpu_id); if (!pdd) { @@ -1587,10 +1582,10 @@ static int kfd_ioctl_import_dmabuf(struct file *filep, goto err_unlock; } - r = amdgpu_amdkfd_gpuvm_import_dmabuf(pdd->dev->adev, dmabuf, - args->va_addr, pdd->drm_priv, - (struct kgd_mem **)&mem, &size, - NULL); + r = amdgpu_amdkfd_gpuvm_import_dmabuf_fd(pdd->dev->adev, args->dmabuf_fd, + args->va_addr, pdd->drm_priv, + (struct kgd_mem **)&mem, &size, + NULL); if (r) goto err_unlock; @@ -1601,7 +1596,6 @@ static int kfd_ioctl_import_dmabuf(struct file *filep, } mutex_unlock(&p->mutex); - dma_buf_put(dmabuf); args->handle = MAKE_HANDLE(args->gpu_id, idr_handle); @@ -1612,7 +1606,6 @@ static int kfd_ioctl_import_dmabuf(struct file *filep, pdd->drm_priv, NULL); err_unlock: mutex_unlock(&p->mutex); - dma_buf_put(dmabuf); return r; } From 296b29ce8acb5dbb3ca1937f1b537b3f6be0460a Mon Sep 17 00:00:00 2001 From: Evan Quan Date: Mon, 11 Dec 2023 18:06:26 +0800 Subject: [PATCH 34/92] drm/amd/pm: update driver_if and ppsmc headers for coming wbrf feature Add those data structures to support Wifi RFI mitigation feature. Signed-off-by: Evan Quan Reviewed-by: Mario Limonciello Signed-off-by: Ma Jun Signed-off-by: Alex Deucher --- drivers/gpu/drm/amd/pm/swsmu/inc/amdgpu_smu.h | 11 +++++++++++ .../pm/swsmu/inc/pmfw_if/smu13_driver_if_v13_0_0.h | 3 ++- .../pm/swsmu/inc/pmfw_if/smu13_driver_if_v13_0_7.h | 3 ++- .../drm/amd/pm/swsmu/inc/pmfw_if/smu_v13_0_0_ppsmc.h | 5 ++--- .../drm/amd/pm/swsmu/inc/pmfw_if/smu_v13_0_7_ppsmc.h | 3 ++- 5 files changed, 19 insertions(+), 6 deletions(-) diff --git a/drivers/gpu/drm/amd/pm/swsmu/inc/amdgpu_smu.h b/drivers/gpu/drm/amd/pm/swsmu/inc/amdgpu_smu.h index c2265e027ca8..ab36402f85b1 100644 --- a/drivers/gpu/drm/amd/pm/swsmu/inc/amdgpu_smu.h +++ b/drivers/gpu/drm/amd/pm/swsmu/inc/amdgpu_smu.h @@ -1501,6 +1501,17 @@ enum smu_baco_seq { __dst_size); \ }) +typedef struct { + uint16_t LowFreq; + uint16_t HighFreq; +} WifiOneBand_t; + +typedef struct { + uint32_t WifiBandEntryNum; + WifiOneBand_t WifiBandEntry[11]; + uint32_t MmHubPadding[8]; +} WifiBandEntryTable_t; + #if !defined(SWSMU_CODE_LAYER_L2) && !defined(SWSMU_CODE_LAYER_L3) && !defined(SWSMU_CODE_LAYER_L4) int smu_get_power_limit(void *handle, uint32_t *limit, diff --git a/drivers/gpu/drm/amd/pm/swsmu/inc/pmfw_if/smu13_driver_if_v13_0_0.h b/drivers/gpu/drm/amd/pm/swsmu/inc/pmfw_if/smu13_driver_if_v13_0_0.h index 9dd1ed5b8940..b114d14fc053 100644 --- a/drivers/gpu/drm/amd/pm/swsmu/inc/pmfw_if/smu13_driver_if_v13_0_0.h +++ b/drivers/gpu/drm/amd/pm/swsmu/inc/pmfw_if/smu13_driver_if_v13_0_0.h @@ -1615,7 +1615,8 @@ typedef struct { #define TABLE_I2C_COMMANDS 9 #define TABLE_DRIVER_INFO 10 #define TABLE_ECCINFO 11 -#define TABLE_COUNT 12 +#define TABLE_WIFIBAND 12 +#define TABLE_COUNT 13 //IH Interupt ID #define IH_INTERRUPT_ID_TO_DRIVER 0xFE diff --git a/drivers/gpu/drm/amd/pm/swsmu/inc/pmfw_if/smu13_driver_if_v13_0_7.h b/drivers/gpu/drm/amd/pm/swsmu/inc/pmfw_if/smu13_driver_if_v13_0_7.h index 62b7c0daff68..8b1496f8ce58 100644 --- a/drivers/gpu/drm/amd/pm/swsmu/inc/pmfw_if/smu13_driver_if_v13_0_7.h +++ b/drivers/gpu/drm/amd/pm/swsmu/inc/pmfw_if/smu13_driver_if_v13_0_7.h @@ -1605,7 +1605,8 @@ typedef struct { #define TABLE_I2C_COMMANDS 9 #define TABLE_DRIVER_INFO 10 #define TABLE_ECCINFO 11 -#define TABLE_COUNT 12 +#define TABLE_WIFIBAND 12 +#define TABLE_COUNT 13 //IH Interupt ID #define IH_INTERRUPT_ID_TO_DRIVER 0xFE diff --git a/drivers/gpu/drm/amd/pm/swsmu/inc/pmfw_if/smu_v13_0_0_ppsmc.h b/drivers/gpu/drm/amd/pm/swsmu/inc/pmfw_if/smu_v13_0_0_ppsmc.h index e2ee855c7748..e862d323caab 100644 --- a/drivers/gpu/drm/amd/pm/swsmu/inc/pmfw_if/smu_v13_0_0_ppsmc.h +++ b/drivers/gpu/drm/amd/pm/swsmu/inc/pmfw_if/smu_v13_0_0_ppsmc.h @@ -138,10 +138,9 @@ #define PPSMC_MSG_SetBadMemoryPagesRetiredFlagsPerChannel 0x4A #define PPSMC_MSG_SetPriorityDeltaGain 0x4B #define PPSMC_MSG_AllowIHHostInterrupt 0x4C - #define PPSMC_MSG_DALNotPresent 0x4E - -#define PPSMC_Message_Count 0x4F +#define PPSMC_MSG_EnableUCLKShadow 0x51 +#define PPSMC_Message_Count 0x52 //Debug Dump Message #define DEBUGSMC_MSG_TestMessage 0x1 diff --git a/drivers/gpu/drm/amd/pm/swsmu/inc/pmfw_if/smu_v13_0_7_ppsmc.h b/drivers/gpu/drm/amd/pm/swsmu/inc/pmfw_if/smu_v13_0_7_ppsmc.h index 6aaefca9b595..a6bf9cdd130e 100644 --- a/drivers/gpu/drm/amd/pm/swsmu/inc/pmfw_if/smu_v13_0_7_ppsmc.h +++ b/drivers/gpu/drm/amd/pm/swsmu/inc/pmfw_if/smu_v13_0_7_ppsmc.h @@ -134,6 +134,7 @@ #define PPSMC_MSG_SetBadMemoryPagesRetiredFlagsPerChannel 0x4A #define PPSMC_MSG_SetPriorityDeltaGain 0x4B #define PPSMC_MSG_AllowIHHostInterrupt 0x4C -#define PPSMC_Message_Count 0x4D +#define PPSMC_MSG_EnableUCLKShadow 0x51 +#define PPSMC_Message_Count 0x52 #endif From b8b39de646274366d17a3614fdaf65fa0716ab32 Mon Sep 17 00:00:00 2001 From: Evan Quan Date: Mon, 11 Dec 2023 18:06:27 +0800 Subject: [PATCH 35/92] drm/amd/pm: setup the framework to support Wifi RFI mitigation feature With WBRF feature supported, as a driver responding to the frequencies, amdgpu driver is able to do shadow pstate switching to mitigate possible interference(between its (G-)DDR memory clocks and local radio module frequency bands used by Wifi 6/6e/7). -- v1->v2: - update the prompt for feature support(Lijo) v8->v9: - update parameter document for smu_wbrf_event_handler(Simon) v9->v10: v10->v11: - correct the logics for wbrf range sorting(Lijo) v13: - Fix the format issue (IIpo Jarvinen) Signed-off-by: Evan Quan Reviewed-by: Mario Limonciello Signed-off-by: Ma Jun Signed-off-by: Ma Jun Signed-off-by: Alex Deucher --- drivers/gpu/drm/amd/amdgpu/amdgpu.h | 2 + drivers/gpu/drm/amd/amdgpu/amdgpu_drv.c | 17 ++ drivers/gpu/drm/amd/pm/swsmu/amdgpu_smu.c | 191 ++++++++++++++++++ drivers/gpu/drm/amd/pm/swsmu/inc/amdgpu_smu.h | 22 ++ drivers/gpu/drm/amd/pm/swsmu/smu_internal.h | 3 + 5 files changed, 235 insertions(+) diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu.h b/drivers/gpu/drm/amd/amdgpu/amdgpu.h index 0af8ac81facd..82d294069276 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu.h +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu.h @@ -252,6 +252,8 @@ extern int amdgpu_seamless; extern int amdgpu_user_partt_mode; extern int amdgpu_agp; +extern int amdgpu_wbrf; + #define AMDGPU_VM_MAX_NUM_CTX 4096 #define AMDGPU_SG_THRESHOLD (256*1024*1024) #define AMDGPU_WAIT_IDLE_TIMEOUT_IN_MS 3000 diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_drv.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_drv.c index 8b33b130ea36..a194db09cde6 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_drv.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_drv.c @@ -208,6 +208,7 @@ int amdgpu_umsch_mm; int amdgpu_seamless = -1; /* auto */ uint amdgpu_debug_mask; int amdgpu_agp = -1; /* auto */ +int amdgpu_wbrf = -1; static void amdgpu_drv_delayed_reset_work_handler(struct work_struct *work); @@ -971,6 +972,22 @@ module_param_named(debug_mask, amdgpu_debug_mask, uint, 0444); MODULE_PARM_DESC(agp, "AGP (-1 = auto (default), 0 = disable, 1 = enable)"); module_param_named(agp, amdgpu_agp, int, 0444); +/** + * DOC: wbrf (int) + * Enable Wifi RFI interference mitigation feature. + * Due to electrical and mechanical constraints there may be likely interference of + * relatively high-powered harmonics of the (G-)DDR memory clocks with local radio + * module frequency bands used by Wifi 6/6e/7. To mitigate the possible RFI interference, + * with this feature enabled, PMFW will use either “shadowed P-State” or “P-State” based + * on active list of frequencies in-use (to be avoided) as part of initial setting or + * P-state transition. However, there may be potential performance impact with this + * feature enabled. + * (0 = disabled, 1 = enabled, -1 = auto (default setting, will be enabled if supported)) + */ +MODULE_PARM_DESC(wbrf, + "Enable Wifi RFI interference mitigation (0 = disabled, 1 = enabled, -1 = auto(default)"); +module_param_named(wbrf, amdgpu_wbrf, int, 0444); + /* These devices are not supported by amdgpu. * They are supported by the mach64, r128, radeon drivers */ diff --git a/drivers/gpu/drm/amd/pm/swsmu/amdgpu_smu.c b/drivers/gpu/drm/amd/pm/swsmu/amdgpu_smu.c index f464610a959f..a3c54a053bd5 100644 --- a/drivers/gpu/drm/amd/pm/swsmu/amdgpu_smu.c +++ b/drivers/gpu/drm/amd/pm/swsmu/amdgpu_smu.c @@ -1322,6 +1322,170 @@ static int smu_get_thermal_temperature_range(struct smu_context *smu) return ret; } +/** + * smu_wbrf_handle_exclusion_ranges - consume the wbrf exclusion ranges + * + * @smu: smu_context pointer + * + * Retrieve the wbrf exclusion ranges and send them to PMFW for proper handling. + * Returns 0 on success, error on failure. + */ +static int smu_wbrf_handle_exclusion_ranges(struct smu_context *smu) +{ + struct wbrf_ranges_in_out wbrf_exclusion = {0}; + struct freq_band_range *wifi_bands = wbrf_exclusion.band_list; + struct amdgpu_device *adev = smu->adev; + uint32_t num_of_wbrf_ranges = MAX_NUM_OF_WBRF_RANGES; + uint64_t start, end; + int ret, i, j; + + ret = amd_wbrf_retrieve_freq_band(adev->dev, &wbrf_exclusion); + if (ret) { + dev_err(adev->dev, "Failed to retrieve exclusion ranges!\n"); + return ret; + } + + /* + * The exclusion ranges array we got might be filled with holes and duplicate + * entries. For example: + * {(2400, 2500), (0, 0), (6882, 6962), (2400, 2500), (0, 0), (6117, 6189), (0, 0)...} + * We need to do some sortups to eliminate those holes and duplicate entries. + * Expected output: {(2400, 2500), (6117, 6189), (6882, 6962), (0, 0)...} + */ + for (i = 0; i < num_of_wbrf_ranges; i++) { + start = wifi_bands[i].start; + end = wifi_bands[i].end; + + /* get the last valid entry to fill the intermediate hole */ + if (!start && !end) { + for (j = num_of_wbrf_ranges - 1; j > i; j--) + if (wifi_bands[j].start && wifi_bands[j].end) + break; + + /* no valid entry left */ + if (j <= i) + break; + + start = wifi_bands[i].start = wifi_bands[j].start; + end = wifi_bands[i].end = wifi_bands[j].end; + wifi_bands[j].start = 0; + wifi_bands[j].end = 0; + num_of_wbrf_ranges = j; + } + + /* eliminate duplicate entries */ + for (j = i + 1; j < num_of_wbrf_ranges; j++) { + if ((wifi_bands[j].start == start) && (wifi_bands[j].end == end)) { + wifi_bands[j].start = 0; + wifi_bands[j].end = 0; + } + } + } + + /* Send the sorted wifi_bands to PMFW */ + ret = smu_set_wbrf_exclusion_ranges(smu, wifi_bands); + /* Try to set the wifi_bands again */ + if (unlikely(ret == -EBUSY)) { + mdelay(5); + ret = smu_set_wbrf_exclusion_ranges(smu, wifi_bands); + } + + return ret; +} + +/** + * smu_wbrf_event_handler - handle notify events + * + * @nb: notifier block + * @action: event type + * @_arg: event data + * + * Calls relevant amdgpu function in response to wbrf event + * notification from kernel. + */ +static int smu_wbrf_event_handler(struct notifier_block *nb, + unsigned long action, void *_arg) +{ + struct smu_context *smu = container_of(nb, struct smu_context, wbrf_notifier); + + switch (action) { + case WBRF_CHANGED: + smu_wbrf_handle_exclusion_ranges(smu); + break; + default: + return NOTIFY_DONE; + }; + + return NOTIFY_OK; +} + +/** + * smu_wbrf_support_check - check wbrf support + * + * @smu: smu_context pointer + * + * Verifies the ACPI interface whether wbrf is supported. + */ +static void smu_wbrf_support_check(struct smu_context *smu) +{ + struct amdgpu_device *adev = smu->adev; + + smu->wbrf_supported = smu_is_asic_wbrf_supported(smu) && amdgpu_wbrf && + acpi_amd_wbrf_supported_consumer(adev->dev); + + if (smu->wbrf_supported) + dev_info(adev->dev, "RF interference mitigation is supported\n"); +} + +/** + * smu_wbrf_init - init driver wbrf support + * + * @smu: smu_context pointer + * + * Verifies the AMD ACPI interfaces and registers with the wbrf + * notifier chain if wbrf feature is supported. + * Returns 0 on success, error on failure. + */ +static int smu_wbrf_init(struct smu_context *smu) +{ + struct amdgpu_device *adev = smu->adev; + int ret; + + if (!smu->wbrf_supported) + return 0; + + smu->wbrf_notifier.notifier_call = smu_wbrf_event_handler; + ret = amd_wbrf_register_notifier(&smu->wbrf_notifier); + if (ret) + return ret; + + /* + * Some wifiband exclusion ranges may be already there + * before our driver loaded. To make sure our driver + * is awared of those exclusion ranges. + */ + ret = smu_wbrf_handle_exclusion_ranges(smu); + if (ret) + dev_err(adev->dev, "Failed to handle wbrf exclusion ranges\n"); + + return ret; +} + +/** + * smu_wbrf_fini - tear down driver wbrf support + * + * @smu: smu_context pointer + * + * Unregisters with the wbrf notifier chain. + */ +static void smu_wbrf_fini(struct smu_context *smu) +{ + if (!smu->wbrf_supported) + return; + + amd_wbrf_unregister_notifier(&smu->wbrf_notifier); +} + static int smu_smc_hw_setup(struct smu_context *smu) { struct smu_feature *feature = &smu->smu_feature; @@ -1414,6 +1578,15 @@ static int smu_smc_hw_setup(struct smu_context *smu) if (ret) return ret; + /* Enable UclkShadow on wbrf supported */ + if (smu->wbrf_supported) { + ret = smu_enable_uclk_shadow(smu, true); + if (ret) { + dev_err(adev->dev, "Failed to enable UclkShadow feature to support wbrf!\n"); + return ret; + } + } + /* * With SCPM enabled, these actions(and relevant messages) are * not needed and permitted. @@ -1512,6 +1685,15 @@ static int smu_smc_hw_setup(struct smu_context *smu) */ ret = smu_set_min_dcef_deep_sleep(smu, smu->smu_table.boot_values.dcefclk / 100); + if (ret) { + dev_err(adev->dev, "Error setting min deepsleep dcefclk\n"); + return ret; + } + + /* Init wbrf support. Properly setup the notifier */ + ret = smu_wbrf_init(smu); + if (ret) + dev_err(adev->dev, "Error during wbrf init call\n"); return ret; } @@ -1567,6 +1749,13 @@ static int smu_hw_init(void *handle) return ret; } + /* + * Check whether wbrf is supported. This needs to be done + * before SMU setup starts since part of SMU configuration + * relies on this. + */ + smu_wbrf_support_check(smu); + if (smu->is_apu) { ret = smu_set_gfx_imu_enable(smu); if (ret) @@ -1733,6 +1922,8 @@ static int smu_smc_hw_cleanup(struct smu_context *smu) struct amdgpu_device *adev = smu->adev; int ret = 0; + smu_wbrf_fini(smu); + cancel_work_sync(&smu->throttling_logging_work); cancel_work_sync(&smu->interrupt_work); diff --git a/drivers/gpu/drm/amd/pm/swsmu/inc/amdgpu_smu.h b/drivers/gpu/drm/amd/pm/swsmu/inc/amdgpu_smu.h index ab36402f85b1..f2a89a783c58 100644 --- a/drivers/gpu/drm/amd/pm/swsmu/inc/amdgpu_smu.h +++ b/drivers/gpu/drm/amd/pm/swsmu/inc/amdgpu_smu.h @@ -22,6 +22,8 @@ #ifndef __AMDGPU_SMU_H__ #define __AMDGPU_SMU_H__ +#include + #include "amdgpu.h" #include "kgd_pp_interface.h" #include "dm_pp_interface.h" @@ -570,6 +572,10 @@ struct smu_context { struct delayed_work swctf_delayed_work; enum pp_xgmi_plpd_mode plpd_mode; + + /* data structures for wbrf feature support */ + bool wbrf_supported; + struct notifier_block wbrf_notifier; }; struct i2c_adapter; @@ -1375,6 +1381,22 @@ struct pptable_funcs { * @notify_rlc_state: Notify RLC power state to SMU. */ int (*notify_rlc_state)(struct smu_context *smu, bool en); + + /** + * @is_asic_wbrf_supported: check whether PMFW supports the wbrf feature + */ + bool (*is_asic_wbrf_supported)(struct smu_context *smu); + + /** + * @enable_uclk_shadow: Enable the uclk shadow feature on wbrf supported + */ + int (*enable_uclk_shadow)(struct smu_context *smu, bool enable); + + /** + * @set_wbrf_exclusion_ranges: notify SMU the wifi bands occupied + */ + int (*set_wbrf_exclusion_ranges)(struct smu_context *smu, + struct freq_band_range *exclusion_ranges); }; typedef enum { diff --git a/drivers/gpu/drm/amd/pm/swsmu/smu_internal.h b/drivers/gpu/drm/amd/pm/swsmu/smu_internal.h index 64766ac69c53..6f4d212607d7 100644 --- a/drivers/gpu/drm/amd/pm/swsmu/smu_internal.h +++ b/drivers/gpu/drm/amd/pm/swsmu/smu_internal.h @@ -98,6 +98,9 @@ #define smu_set_config_table(smu, config_table) smu_ppt_funcs(set_config_table, -EOPNOTSUPP, smu, config_table) #define smu_init_pptable_microcode(smu) smu_ppt_funcs(init_pptable_microcode, 0, smu) #define smu_notify_rlc_state(smu, en) smu_ppt_funcs(notify_rlc_state, 0, smu, en) +#define smu_is_asic_wbrf_supported(smu) smu_ppt_funcs(is_asic_wbrf_supported, false, smu) +#define smu_enable_uclk_shadow(smu, enable) smu_ppt_funcs(enable_uclk_shadow, 0, smu, enable) +#define smu_set_wbrf_exclusion_ranges(smu, freq_band_range) smu_ppt_funcs(set_wbrf_exclusion_ranges, -EOPNOTSUPP, smu, freq_band_range) #endif #endif From 71f69557cb12a4674a05b4c5fb730880f13366b1 Mon Sep 17 00:00:00 2001 From: Evan Quan Date: Mon, 11 Dec 2023 18:06:28 +0800 Subject: [PATCH 36/92] drm/amd/pm: add flood detection for wbrf events To protect PMFW from being overloaded. Signed-off-by: Evan Quan Reviewed-by: Mario Limonciello Signed-off-by: Ma Jun Signed-off-by: Alex Deucher --- drivers/gpu/drm/amd/pm/swsmu/amdgpu_smu.c | 29 +++++++++++++++---- drivers/gpu/drm/amd/pm/swsmu/inc/amdgpu_smu.h | 7 +++++ 2 files changed, 30 insertions(+), 6 deletions(-) diff --git a/drivers/gpu/drm/amd/pm/swsmu/amdgpu_smu.c b/drivers/gpu/drm/amd/pm/swsmu/amdgpu_smu.c index a3c54a053bd5..d409857fd622 100644 --- a/drivers/gpu/drm/amd/pm/swsmu/amdgpu_smu.c +++ b/drivers/gpu/drm/amd/pm/swsmu/amdgpu_smu.c @@ -1410,7 +1410,8 @@ static int smu_wbrf_event_handler(struct notifier_block *nb, switch (action) { case WBRF_CHANGED: - smu_wbrf_handle_exclusion_ranges(smu); + schedule_delayed_work(&smu->wbrf_delayed_work, + msecs_to_jiffies(SMU_WBRF_EVENT_HANDLING_PACE)); break; default: return NOTIFY_DONE; @@ -1419,6 +1420,20 @@ static int smu_wbrf_event_handler(struct notifier_block *nb, return NOTIFY_OK; } +/** + * smu_wbrf_delayed_work_handler - callback on delayed work timer expired + * + * @work: struct work_struct pointer + * + * Flood is over and driver will consume the latest exclusion ranges. + */ +static void smu_wbrf_delayed_work_handler(struct work_struct *work) +{ + struct smu_context *smu = container_of(work, struct smu_context, wbrf_delayed_work.work); + + smu_wbrf_handle_exclusion_ranges(smu); +} + /** * smu_wbrf_support_check - check wbrf support * @@ -1448,12 +1463,13 @@ static void smu_wbrf_support_check(struct smu_context *smu) */ static int smu_wbrf_init(struct smu_context *smu) { - struct amdgpu_device *adev = smu->adev; int ret; if (!smu->wbrf_supported) return 0; + INIT_DELAYED_WORK(&smu->wbrf_delayed_work, smu_wbrf_delayed_work_handler); + smu->wbrf_notifier.notifier_call = smu_wbrf_event_handler; ret = amd_wbrf_register_notifier(&smu->wbrf_notifier); if (ret) @@ -1464,11 +1480,10 @@ static int smu_wbrf_init(struct smu_context *smu) * before our driver loaded. To make sure our driver * is awared of those exclusion ranges. */ - ret = smu_wbrf_handle_exclusion_ranges(smu); - if (ret) - dev_err(adev->dev, "Failed to handle wbrf exclusion ranges\n"); + schedule_delayed_work(&smu->wbrf_delayed_work, + msecs_to_jiffies(SMU_WBRF_EVENT_HANDLING_PACE)); - return ret; + return 0; } /** @@ -1484,6 +1499,8 @@ static void smu_wbrf_fini(struct smu_context *smu) return; amd_wbrf_unregister_notifier(&smu->wbrf_notifier); + + cancel_delayed_work_sync(&smu->wbrf_delayed_work); } static int smu_smc_hw_setup(struct smu_context *smu) diff --git a/drivers/gpu/drm/amd/pm/swsmu/inc/amdgpu_smu.h b/drivers/gpu/drm/amd/pm/swsmu/inc/amdgpu_smu.h index f2a89a783c58..f97178a602bc 100644 --- a/drivers/gpu/drm/amd/pm/swsmu/inc/amdgpu_smu.h +++ b/drivers/gpu/drm/amd/pm/swsmu/inc/amdgpu_smu.h @@ -473,6 +473,12 @@ struct stb_context { #define WORKLOAD_POLICY_MAX 7 +/* + * Configure wbrf event handling pace as there can be only one + * event processed every SMU_WBRF_EVENT_HANDLING_PACE ms. + */ +#define SMU_WBRF_EVENT_HANDLING_PACE 10 + struct smu_context { struct amdgpu_device *adev; struct amdgpu_irq_src irq_source; @@ -576,6 +582,7 @@ struct smu_context { /* data structures for wbrf feature support */ bool wbrf_supported; struct notifier_block wbrf_notifier; + struct delayed_work wbrf_delayed_work; }; struct i2c_adapter; From 18df969b44a0bdc1f24f6ca6b10595dad6f57398 Mon Sep 17 00:00:00 2001 From: Ma Jun Date: Mon, 11 Dec 2023 18:06:29 +0800 Subject: [PATCH 37/92] drm/amd/pm: enable Wifi RFI mitigation feature support for SMU13.0.0 Fulfill the SMU13.0.0 support for Wifi RFI mitigation feature. -- v10->v11: - downgrade the prompt level on message failure(Lijo) v13: - Fix the format issue (IIpo Jarvinen) - Move function smu_v13_0_0_set_wbrf_exclusion_ranges to smu_v13_0.c as a generic code for later use (IIpo Jarvinen) Co-developed-by: Evan Quan Signed-off-by: Evan Quan Reviewed-by: Mario Limonciello Signed-off-by: Ma Jun Signed-off-by: Alex Deucher --- drivers/gpu/drm/amd/pm/swsmu/inc/amdgpu_smu.h | 2 + drivers/gpu/drm/amd/pm/swsmu/inc/smu_types.h | 3 +- drivers/gpu/drm/amd/pm/swsmu/inc/smu_v13_0.h | 4 ++ .../gpu/drm/amd/pm/swsmu/smu13/smu_v13_0.c | 48 +++++++++++++++++++ .../drm/amd/pm/swsmu/smu13/smu_v13_0_0_ppt.c | 22 +++++++++ 5 files changed, 78 insertions(+), 1 deletion(-) diff --git a/drivers/gpu/drm/amd/pm/swsmu/inc/amdgpu_smu.h b/drivers/gpu/drm/amd/pm/swsmu/inc/amdgpu_smu.h index f97178a602bc..2aa4fea87314 100644 --- a/drivers/gpu/drm/amd/pm/swsmu/inc/amdgpu_smu.h +++ b/drivers/gpu/drm/amd/pm/swsmu/inc/amdgpu_smu.h @@ -23,6 +23,7 @@ #define __AMDGPU_SMU_H__ #include +#include #include "amdgpu.h" #include "kgd_pp_interface.h" @@ -320,6 +321,7 @@ enum smu_table_id { SMU_TABLE_PACE, SMU_TABLE_ECCINFO, SMU_TABLE_COMBO_PPTABLE, + SMU_TABLE_WIFIBAND, SMU_TABLE_COUNT, }; diff --git a/drivers/gpu/drm/amd/pm/swsmu/inc/smu_types.h b/drivers/gpu/drm/amd/pm/swsmu/inc/smu_types.h index ea8ea99b7436..953a767613b1 100644 --- a/drivers/gpu/drm/amd/pm/swsmu/inc/smu_types.h +++ b/drivers/gpu/drm/amd/pm/swsmu/inc/smu_types.h @@ -260,7 +260,8 @@ __SMU_DUMMY_MAP(PowerDownUmsch), \ __SMU_DUMMY_MAP(SetSoftMaxVpe), \ __SMU_DUMMY_MAP(SetSoftMinVpe), \ - __SMU_DUMMY_MAP(GetMetricsVersion), + __SMU_DUMMY_MAP(GetMetricsVersion), \ + __SMU_DUMMY_MAP(EnableUCLKShadow), #undef __SMU_DUMMY_MAP #define __SMU_DUMMY_MAP(type) SMU_MSG_##type diff --git a/drivers/gpu/drm/amd/pm/swsmu/inc/smu_v13_0.h b/drivers/gpu/drm/amd/pm/swsmu/inc/smu_v13_0.h index 5e7b8f29fecc..fbd57fa1a004 100644 --- a/drivers/gpu/drm/amd/pm/swsmu/inc/smu_v13_0.h +++ b/drivers/gpu/drm/amd/pm/swsmu/inc/smu_v13_0.h @@ -294,5 +294,9 @@ int smu_v13_0_update_pcie_parameters(struct smu_context *smu, int smu_v13_0_disable_pmfw_state(struct smu_context *smu); +int smu_v13_0_enable_uclk_shadow(struct smu_context *smu, bool enable); + +int smu_v13_0_set_wbrf_exclusion_ranges(struct smu_context *smu, + struct freq_band_range *exclusion_ranges); #endif #endif diff --git a/drivers/gpu/drm/amd/pm/swsmu/smu13/smu_v13_0.c b/drivers/gpu/drm/amd/pm/swsmu/smu13/smu_v13_0.c index 2877e5da5b5a..771a3d457c33 100644 --- a/drivers/gpu/drm/amd/pm/swsmu/smu13/smu_v13_0.c +++ b/drivers/gpu/drm/amd/pm/swsmu/smu13/smu_v13_0.c @@ -2507,3 +2507,51 @@ int smu_v13_0_disable_pmfw_state(struct smu_context *smu) return ret == 0 ? 0 : -EINVAL; } + +int smu_v13_0_enable_uclk_shadow(struct smu_context *smu, bool enable) +{ + return smu_cmn_send_smc_msg_with_param(smu, SMU_MSG_EnableUCLKShadow, enable, NULL); +} + +int smu_v13_0_set_wbrf_exclusion_ranges(struct smu_context *smu, + struct freq_band_range *exclusion_ranges) +{ + WifiBandEntryTable_t wifi_bands; + int valid_entries = 0; + int ret, i; + + memset(&wifi_bands, 0, sizeof(wifi_bands)); + for (i = 0; i < ARRAY_SIZE(wifi_bands.WifiBandEntry); i++) { + if (!exclusion_ranges[i].start && !exclusion_ranges[i].end) + break; + + /* PMFW expects the inputs to be in Mhz unit */ + wifi_bands.WifiBandEntry[valid_entries].LowFreq = + DIV_ROUND_DOWN_ULL(exclusion_ranges[i].start, HZ_PER_MHZ); + wifi_bands.WifiBandEntry[valid_entries++].HighFreq = + DIV_ROUND_UP_ULL(exclusion_ranges[i].end, HZ_PER_MHZ); + } + wifi_bands.WifiBandEntryNum = valid_entries; + + /* + * Per confirm with PMFW team, WifiBandEntryNum = 0 + * is a valid setting. + * + * Considering the scenarios below: + * - At first the wifi device adds an exclusion range e.g. (2400,2500) to + * BIOS and our driver gets notified. We will set WifiBandEntryNum = 1 + * and pass the WifiBandEntry (2400, 2500) to PMFW. + * + * - Later the wifi device removes the wifiband list added above and + * our driver gets notified again. At this time, driver will set + * WifiBandEntryNum = 0 and pass an empty WifiBandEntry list to PMFW. + * + * - PMFW may still need to do some uclk shadow update(e.g. switching + * from shadow clock back to primary clock) on receiving this. + */ + ret = smu_cmn_update_table(smu, SMU_TABLE_WIFIBAND, 0, &wifi_bands, true); + if (ret) + dev_warn(smu->adev->dev, "Failed to set wifiband!"); + + return ret; +} diff --git a/drivers/gpu/drm/amd/pm/swsmu/smu13/smu_v13_0_0_ppt.c b/drivers/gpu/drm/amd/pm/swsmu/smu13/smu_v13_0_0_ppt.c index 2d25a9e47651..a24aa886c636 100644 --- a/drivers/gpu/drm/amd/pm/swsmu/smu13/smu_v13_0_0_ppt.c +++ b/drivers/gpu/drm/amd/pm/swsmu/smu13/smu_v13_0_0_ppt.c @@ -169,6 +169,7 @@ static struct cmn2asic_msg_mapping smu_v13_0_0_message_map[SMU_MSG_MAX_COUNT] = MSG_MAP(AllowIHHostInterrupt, PPSMC_MSG_AllowIHHostInterrupt, 0), MSG_MAP(ReenableAcDcInterrupt, PPSMC_MSG_ReenableAcDcInterrupt, 0), MSG_MAP(DALNotPresent, PPSMC_MSG_DALNotPresent, 0), + MSG_MAP(EnableUCLKShadow, PPSMC_MSG_EnableUCLKShadow, 0), }; static struct cmn2asic_mapping smu_v13_0_0_clk_map[SMU_CLK_COUNT] = { @@ -253,6 +254,7 @@ static struct cmn2asic_mapping smu_v13_0_0_table_map[SMU_TABLE_COUNT] = { TAB_MAP(I2C_COMMANDS), TAB_MAP(ECCINFO), TAB_MAP(OVERDRIVE), + TAB_MAP(WIFIBAND), }; static struct cmn2asic_mapping smu_v13_0_0_pwr_src_map[SMU_POWER_SOURCE_COUNT] = { @@ -498,6 +500,9 @@ static int smu_v13_0_0_tables_init(struct smu_context *smu) PAGE_SIZE, AMDGPU_GEM_DOMAIN_VRAM); SMU_TABLE_INIT(tables, SMU_TABLE_ECCINFO, sizeof(EccInfoTable_t), PAGE_SIZE, AMDGPU_GEM_DOMAIN_VRAM); + SMU_TABLE_INIT(tables, SMU_TABLE_WIFIBAND, + sizeof(WifiBandEntryTable_t), PAGE_SIZE, + AMDGPU_GEM_DOMAIN_VRAM); smu_table->metrics_table = kzalloc(sizeof(SmuMetricsExternal_t), GFP_KERNEL); if (!smu_table->metrics_table) @@ -2938,6 +2943,20 @@ static ssize_t smu_v13_0_0_get_ecc_info(struct smu_context *smu, return ret; } +static bool smu_v13_0_0_wbrf_support_check(struct smu_context *smu) +{ + struct amdgpu_device *adev = smu->adev; + + switch (adev->ip_versions[MP1_HWIP][0]) { + case IP_VERSION(13, 0, 0): + return smu->smc_fw_version >= 0x004e6300; + case IP_VERSION(13, 0, 10): + return smu->smc_fw_version >= 0x00503300; + default: + return false; + } +} + static const struct pptable_funcs smu_v13_0_0_ppt_funcs = { .get_allowed_feature_mask = smu_v13_0_0_get_allowed_feature_mask, .set_default_dpm_table = smu_v13_0_0_set_default_dpm_table, @@ -3016,6 +3035,9 @@ static const struct pptable_funcs smu_v13_0_0_ppt_funcs = { .gpo_control = smu_v13_0_gpo_control, .get_ecc_info = smu_v13_0_0_get_ecc_info, .notify_display_change = smu_v13_0_notify_display_change, + .is_asic_wbrf_supported = smu_v13_0_0_wbrf_support_check, + .enable_uclk_shadow = smu_v13_0_enable_uclk_shadow, + .set_wbrf_exclusion_ranges = smu_v13_0_set_wbrf_exclusion_ranges, }; void smu_v13_0_0_set_ppt_funcs(struct smu_context *smu) From cca850267d33f1153e16e07dc7c32ce5bc3df1fe Mon Sep 17 00:00:00 2001 From: Evan Quan Date: Mon, 11 Dec 2023 18:06:30 +0800 Subject: [PATCH 38/92] drm/amd/pm: enable Wifi RFI mitigation feature support for SMU13.0.7 Fulfill the SMU13.0.7 support for Wifi RFI mitigation feature. -- v10->v11: - downgrade the prompt level on message failure(Lijo) v13: - Fix the format issue (IIpo Jarvinen) - Remove duplicate code (IIpo Jarvinen) Signed-off-by: Evan Quan Reviewed-by: Mario Limonciello Signed-off-by: Ma Jun Signed-off-by: Alex Deucher --- .../gpu/drm/amd/pm/swsmu/smu13/smu_v13_0_7_ppt.c | 13 +++++++++++++ 1 file changed, 13 insertions(+) diff --git a/drivers/gpu/drm/amd/pm/swsmu/smu13/smu_v13_0_7_ppt.c b/drivers/gpu/drm/amd/pm/swsmu/smu13/smu_v13_0_7_ppt.c index 2909ec06b1cb..59606a19e3d2 100644 --- a/drivers/gpu/drm/amd/pm/swsmu/smu13/smu_v13_0_7_ppt.c +++ b/drivers/gpu/drm/amd/pm/swsmu/smu13/smu_v13_0_7_ppt.c @@ -140,6 +140,7 @@ static struct cmn2asic_msg_mapping smu_v13_0_7_message_map[SMU_MSG_MAX_COUNT] = MSG_MAP(AllowGpo, PPSMC_MSG_SetGpoAllow, 0), MSG_MAP(GetPptLimit, PPSMC_MSG_GetPptLimit, 0), MSG_MAP(NotifyPowerSource, PPSMC_MSG_NotifyPowerSource, 0), + MSG_MAP(EnableUCLKShadow, PPSMC_MSG_EnableUCLKShadow, 0), }; static struct cmn2asic_mapping smu_v13_0_7_clk_map[SMU_CLK_COUNT] = { @@ -222,6 +223,7 @@ static struct cmn2asic_mapping smu_v13_0_7_table_map[SMU_TABLE_COUNT] = { TAB_MAP(ACTIVITY_MONITOR_COEFF), [SMU_TABLE_COMBO_PPTABLE] = {1, TABLE_COMBO_PPTABLE}, TAB_MAP(OVERDRIVE), + TAB_MAP(WIFIBAND), }; static struct cmn2asic_mapping smu_v13_0_7_pwr_src_map[SMU_POWER_SOURCE_COUNT] = { @@ -512,6 +514,9 @@ static int smu_v13_0_7_tables_init(struct smu_context *smu) AMDGPU_GEM_DOMAIN_VRAM); SMU_TABLE_INIT(tables, SMU_TABLE_COMBO_PPTABLE, MP0_MP1_DATA_REGION_SIZE_COMBOPPTABLE, PAGE_SIZE, AMDGPU_GEM_DOMAIN_VRAM); + SMU_TABLE_INIT(tables, SMU_TABLE_WIFIBAND, + sizeof(WifiBandEntryTable_t), PAGE_SIZE, + AMDGPU_GEM_DOMAIN_VRAM); smu_table->metrics_table = kzalloc(sizeof(SmuMetricsExternal_t), GFP_KERNEL); if (!smu_table->metrics_table) @@ -2535,6 +2540,11 @@ static int smu_v13_0_7_set_df_cstate(struct smu_context *smu, NULL); } +static bool smu_v13_0_7_wbrf_support_check(struct smu_context *smu) +{ + return smu->smc_fw_version > 0x00524600; +} + static const struct pptable_funcs smu_v13_0_7_ppt_funcs = { .get_allowed_feature_mask = smu_v13_0_7_get_allowed_feature_mask, .set_default_dpm_table = smu_v13_0_7_set_default_dpm_table, @@ -2601,6 +2611,9 @@ static const struct pptable_funcs smu_v13_0_7_ppt_funcs = { .set_mp1_state = smu_v13_0_7_set_mp1_state, .set_df_cstate = smu_v13_0_7_set_df_cstate, .gpo_control = smu_v13_0_gpo_control, + .is_asic_wbrf_supported = smu_v13_0_7_wbrf_support_check, + .enable_uclk_shadow = smu_v13_0_enable_uclk_shadow, + .set_wbrf_exclusion_ranges = smu_v13_0_set_wbrf_exclusion_ranges, }; void smu_v13_0_7_set_ppt_funcs(struct smu_context *smu) From 94b1e028e15c94362420f9f3f711fafbf9d52996 Mon Sep 17 00:00:00 2001 From: Alex Deucher Date: Thu, 7 Dec 2023 10:14:41 -0500 Subject: [PATCH 39/92] drm/amdgpu/sdma5.2: add begin/end_use ring callbacks MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit Add begin/end_use ring callbacks to disallow GFXOFF when SDMA work is submitted and allow it again afterward. This should avoid corner cases where GFXOFF is erroneously entered when SDMA is still active. For now just allow/disallow GFXOFF in the begin and end helpers until we root cause the issue. This should not impact power as SDMA usage is pretty minimal and GFXOSS should not be active when SDMA is active anyway, this just makes it explicit. v2: move everything into sdma5.2 code. No reason for this to be generic at this point. v3: Add comments in new code Link: https://gitlab.freedesktop.org/drm/amd/-/issues/2220 Reviewed-by: Mario Limonciello (v1) Tested-by: Mario Limonciello (v1) Reviewed-by: Christian König Signed-off-by: Alex Deucher Cc: stable@vger.kernel.org # 5.15+ --- drivers/gpu/drm/amd/amdgpu/sdma_v5_2.c | 28 ++++++++++++++++++++++++++ 1 file changed, 28 insertions(+) diff --git a/drivers/gpu/drm/amd/amdgpu/sdma_v5_2.c b/drivers/gpu/drm/amd/amdgpu/sdma_v5_2.c index 83c240f741b5..0058f3f7cf6e 100644 --- a/drivers/gpu/drm/amd/amdgpu/sdma_v5_2.c +++ b/drivers/gpu/drm/amd/amdgpu/sdma_v5_2.c @@ -1643,6 +1643,32 @@ static void sdma_v5_2_get_clockgating_state(void *handle, u64 *flags) *flags |= AMD_CG_SUPPORT_SDMA_LS; } +static void sdma_v5_2_ring_begin_use(struct amdgpu_ring *ring) +{ + struct amdgpu_device *adev = ring->adev; + + /* SDMA 5.2.3 (RMB) FW doesn't seem to properly + * disallow GFXOFF in some cases leading to + * hangs in SDMA. Disallow GFXOFF while SDMA is active. + * We can probably just limit this to 5.2.3, + * but it shouldn't hurt for other parts since + * this GFXOFF will be disallowed anyway when SDMA is + * active, this just makes it explicit. + */ + amdgpu_gfx_off_ctrl(adev, false); +} + +static void sdma_v5_2_ring_end_use(struct amdgpu_ring *ring) +{ + struct amdgpu_device *adev = ring->adev; + + /* SDMA 5.2.3 (RMB) FW doesn't seem to properly + * disallow GFXOFF in some cases leading to + * hangs in SDMA. Allow GFXOFF when SDMA is complete. + */ + amdgpu_gfx_off_ctrl(adev, true); +} + const struct amd_ip_funcs sdma_v5_2_ip_funcs = { .name = "sdma_v5_2", .early_init = sdma_v5_2_early_init, @@ -1690,6 +1716,8 @@ static const struct amdgpu_ring_funcs sdma_v5_2_ring_funcs = { .test_ib = sdma_v5_2_ring_test_ib, .insert_nop = sdma_v5_2_ring_insert_nop, .pad_ib = sdma_v5_2_ring_pad_ib, + .begin_use = sdma_v5_2_ring_begin_use, + .end_use = sdma_v5_2_ring_end_use, .emit_wreg = sdma_v5_2_ring_emit_wreg, .emit_reg_wait = sdma_v5_2_ring_emit_reg_wait, .emit_reg_write_reg_wait = sdma_v5_2_ring_emit_reg_write_reg_wait, From 91963397c49aa2907aeafa52d929555dcbc9cd07 Mon Sep 17 00:00:00 2001 From: Friedrich Vock Date: Sat, 2 Dec 2023 01:17:40 +0100 Subject: [PATCH 40/92] drm/amdgpu: Enable tunneling on high-priority compute queues MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit This improves latency if the GPU is already busy with other work. This is useful for VR compositors that submit highly latency-sensitive compositing work on high-priority compute queues while the GPU is busy rendering the next frame. Userspace merge request: https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/26462 v2: bump driver version (Alex) Reviewed-by: Marek Olšák Signed-off-by: Friedrich Vock Signed-off-by: Alex Deucher --- drivers/gpu/drm/amd/amdgpu/amdgpu.h | 1 + drivers/gpu/drm/amd/amdgpu/amdgpu_drv.c | 3 ++- drivers/gpu/drm/amd/amdgpu/amdgpu_ring.c | 10 ++++++---- drivers/gpu/drm/amd/amdgpu/gfx_v10_0.c | 3 ++- drivers/gpu/drm/amd/amdgpu/gfx_v11_0.c | 3 ++- 5 files changed, 13 insertions(+), 7 deletions(-) diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu.h b/drivers/gpu/drm/amd/amdgpu/amdgpu.h index 82d294069276..616b6c911767 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu.h +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu.h @@ -791,6 +791,7 @@ struct amdgpu_mqd_prop { uint64_t eop_gpu_addr; uint32_t hqd_pipe_priority; uint32_t hqd_queue_priority; + bool allow_tunneling; bool hqd_active; }; diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_drv.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_drv.c index a194db09cde6..880137774b4e 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_drv.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_drv.c @@ -115,9 +115,10 @@ * 3.54.0 - Add AMDGPU_CTX_QUERY2_FLAGS_RESET_IN_PROGRESS support * - 3.55.0 - Add AMDGPU_INFO_GPUVM_FAULT query * - 3.56.0 - Update IB start address and size alignment for decode and encode + * - 3.57.0 - Compute tunneling on GFX10+ */ #define KMS_DRIVER_MAJOR 3 -#define KMS_DRIVER_MINOR 56 +#define KMS_DRIVER_MINOR 57 #define KMS_DRIVER_PATCHLEVEL 0 /* diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_ring.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_ring.c index 6a80d3ec887e..45424ebf9681 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_ring.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_ring.c @@ -642,6 +642,10 @@ static void amdgpu_ring_to_mqd_prop(struct amdgpu_ring *ring, struct amdgpu_mqd_prop *prop) { struct amdgpu_device *adev = ring->adev; + bool is_high_prio_compute = ring->funcs->type == AMDGPU_RING_TYPE_COMPUTE && + amdgpu_gfx_is_high_priority_compute_queue(adev, ring); + bool is_high_prio_gfx = ring->funcs->type == AMDGPU_RING_TYPE_GFX && + amdgpu_gfx_is_high_priority_graphics_queue(adev, ring); memset(prop, 0, sizeof(*prop)); @@ -659,10 +663,8 @@ static void amdgpu_ring_to_mqd_prop(struct amdgpu_ring *ring, */ prop->hqd_active = ring->funcs->type == AMDGPU_RING_TYPE_KIQ; - if ((ring->funcs->type == AMDGPU_RING_TYPE_COMPUTE && - amdgpu_gfx_is_high_priority_compute_queue(adev, ring)) || - (ring->funcs->type == AMDGPU_RING_TYPE_GFX && - amdgpu_gfx_is_high_priority_graphics_queue(adev, ring))) { + prop->allow_tunneling = is_high_prio_compute; + if (is_high_prio_compute || is_high_prio_gfx) { prop->hqd_pipe_priority = AMDGPU_GFX_PIPE_PRIO_HIGH; prop->hqd_queue_priority = AMDGPU_GFX_QUEUE_PRIORITY_MAXIMUM; } diff --git a/drivers/gpu/drm/amd/amdgpu/gfx_v10_0.c b/drivers/gpu/drm/amd/amdgpu/gfx_v10_0.c index c8a3bf01743f..73f6d7e72c73 100644 --- a/drivers/gpu/drm/amd/amdgpu/gfx_v10_0.c +++ b/drivers/gpu/drm/amd/amdgpu/gfx_v10_0.c @@ -6593,7 +6593,8 @@ static int gfx_v10_0_compute_mqd_init(struct amdgpu_device *adev, void *m, tmp = REG_SET_FIELD(tmp, CP_HQD_PQ_CONTROL, ENDIAN_SWAP, 1); #endif tmp = REG_SET_FIELD(tmp, CP_HQD_PQ_CONTROL, UNORD_DISPATCH, 0); - tmp = REG_SET_FIELD(tmp, CP_HQD_PQ_CONTROL, TUNNEL_DISPATCH, 0); + tmp = REG_SET_FIELD(tmp, CP_HQD_PQ_CONTROL, TUNNEL_DISPATCH, + prop->allow_tunneling); tmp = REG_SET_FIELD(tmp, CP_HQD_PQ_CONTROL, PRIV_STATE, 1); tmp = REG_SET_FIELD(tmp, CP_HQD_PQ_CONTROL, KMD_QUEUE, 1); mqd->cp_hqd_pq_control = tmp; diff --git a/drivers/gpu/drm/amd/amdgpu/gfx_v11_0.c b/drivers/gpu/drm/amd/amdgpu/gfx_v11_0.c index c659ef0f47ce..bdcf96df69e6 100644 --- a/drivers/gpu/drm/amd/amdgpu/gfx_v11_0.c +++ b/drivers/gpu/drm/amd/amdgpu/gfx_v11_0.c @@ -3847,7 +3847,8 @@ static int gfx_v11_0_compute_mqd_init(struct amdgpu_device *adev, void *m, tmp = REG_SET_FIELD(tmp, CP_HQD_PQ_CONTROL, RPTR_BLOCK_SIZE, (order_base_2(AMDGPU_GPU_PAGE_SIZE / 4) - 1)); tmp = REG_SET_FIELD(tmp, CP_HQD_PQ_CONTROL, UNORD_DISPATCH, 0); - tmp = REG_SET_FIELD(tmp, CP_HQD_PQ_CONTROL, TUNNEL_DISPATCH, 0); + tmp = REG_SET_FIELD(tmp, CP_HQD_PQ_CONTROL, TUNNEL_DISPATCH, + prop->allow_tunneling); tmp = REG_SET_FIELD(tmp, CP_HQD_PQ_CONTROL, PRIV_STATE, 1); tmp = REG_SET_FIELD(tmp, CP_HQD_PQ_CONTROL, KMD_QUEUE, 1); mqd->cp_hqd_pq_control = tmp; From e747235ef3c253298157b6cd634b9b2695f33d20 Mon Sep 17 00:00:00 2001 From: Jani Nikula Date: Tue, 12 Dec 2023 15:53:38 +0200 Subject: [PATCH 41/92] drm/radeon: include drm/drm_edid.h only where needed Including drm_edid.h from radeon_mode.h causes the rebuild of more than a hundred files when drm_edid.h is modified, while there are only a handful of files that actually need to include drm_edid.h. Signed-off-by: Jani Nikula Signed-off-by: Alex Deucher --- drivers/gpu/drm/radeon/atombios_encoders.c | 1 + drivers/gpu/drm/radeon/dce3_1_afmt.c | 1 + drivers/gpu/drm/radeon/dce6_afmt.c | 1 + drivers/gpu/drm/radeon/evergreen.c | 1 + drivers/gpu/drm/radeon/evergreen_hdmi.c | 1 + drivers/gpu/drm/radeon/radeon_atombios.c | 1 + drivers/gpu/drm/radeon/radeon_audio.c | 1 + drivers/gpu/drm/radeon/radeon_audio.h | 4 +++- drivers/gpu/drm/radeon/radeon_combios.c | 1 + drivers/gpu/drm/radeon/radeon_encoders.c | 1 + drivers/gpu/drm/radeon/radeon_mode.h | 2 +- 11 files changed, 13 insertions(+), 2 deletions(-) diff --git a/drivers/gpu/drm/radeon/atombios_encoders.c b/drivers/gpu/drm/radeon/atombios_encoders.c index 4aca09cab4b8..6e537c5bd295 100644 --- a/drivers/gpu/drm/radeon/atombios_encoders.c +++ b/drivers/gpu/drm/radeon/atombios_encoders.c @@ -29,6 +29,7 @@ #include #include +#include #include #include #include diff --git a/drivers/gpu/drm/radeon/dce3_1_afmt.c b/drivers/gpu/drm/radeon/dce3_1_afmt.c index e8fe239b9d79..324e9b765098 100644 --- a/drivers/gpu/drm/radeon/dce3_1_afmt.c +++ b/drivers/gpu/drm/radeon/dce3_1_afmt.c @@ -21,6 +21,7 @@ * OTHER DEALINGS IN THE SOFTWARE. */ #include +#include #include "radeon.h" #include "radeon_asic.h" diff --git a/drivers/gpu/drm/radeon/dce6_afmt.c b/drivers/gpu/drm/radeon/dce6_afmt.c index 4a1d5447eac1..4c06f47453fd 100644 --- a/drivers/gpu/drm/radeon/dce6_afmt.c +++ b/drivers/gpu/drm/radeon/dce6_afmt.c @@ -21,6 +21,7 @@ * */ #include +#include #include "dce6_afmt.h" #include "radeon.h" diff --git a/drivers/gpu/drm/radeon/evergreen.c b/drivers/gpu/drm/radeon/evergreen.c index f0ae087be914..a424b86008b8 100644 --- a/drivers/gpu/drm/radeon/evergreen.c +++ b/drivers/gpu/drm/radeon/evergreen.c @@ -26,6 +26,7 @@ #include #include +#include #include #include #include diff --git a/drivers/gpu/drm/radeon/evergreen_hdmi.c b/drivers/gpu/drm/radeon/evergreen_hdmi.c index 5f3078f8ab95..681119c91d94 100644 --- a/drivers/gpu/drm/radeon/evergreen_hdmi.c +++ b/drivers/gpu/drm/radeon/evergreen_hdmi.c @@ -26,6 +26,7 @@ */ #include +#include #include #include "evergreen_hdmi.h" #include "radeon.h" diff --git a/drivers/gpu/drm/radeon/radeon_atombios.c b/drivers/gpu/drm/radeon/radeon_atombios.c index 85c4bb186203..3596ea4a8b60 100644 --- a/drivers/gpu/drm/radeon/radeon_atombios.c +++ b/drivers/gpu/drm/radeon/radeon_atombios.c @@ -27,6 +27,7 @@ #include #include +#include #include #include "radeon.h" diff --git a/drivers/gpu/drm/radeon/radeon_audio.c b/drivers/gpu/drm/radeon/radeon_audio.c index 279bf130a18c..91b58fbc2be7 100644 --- a/drivers/gpu/drm/radeon/radeon_audio.c +++ b/drivers/gpu/drm/radeon/radeon_audio.c @@ -27,6 +27,7 @@ #include #include +#include #include "dce6_afmt.h" #include "evergreen_hdmi.h" #include "radeon.h" diff --git a/drivers/gpu/drm/radeon/radeon_audio.h b/drivers/gpu/drm/radeon/radeon_audio.h index 05e67867469b..dacaaa007051 100644 --- a/drivers/gpu/drm/radeon/radeon_audio.h +++ b/drivers/gpu/drm/radeon/radeon_audio.h @@ -27,7 +27,9 @@ #include -#define RREG32_ENDPOINT(block, reg) \ +struct cea_sad; + +#define RREG32_ENDPOINT(block, reg) \ radeon_audio_endpoint_rreg(rdev, (block), (reg)) #define WREG32_ENDPOINT(block, reg, v) \ radeon_audio_endpoint_wreg(rdev, (block), (reg), (v)) diff --git a/drivers/gpu/drm/radeon/radeon_combios.c b/drivers/gpu/drm/radeon/radeon_combios.c index 2620efc7c675..6952b1273b0f 100644 --- a/drivers/gpu/drm/radeon/radeon_combios.c +++ b/drivers/gpu/drm/radeon/radeon_combios.c @@ -28,6 +28,7 @@ #include #include +#include #include #include "radeon.h" diff --git a/drivers/gpu/drm/radeon/radeon_encoders.c b/drivers/gpu/drm/radeon/radeon_encoders.c index 9cb6401fe97e..3de3dce9e89d 100644 --- a/drivers/gpu/drm/radeon/radeon_encoders.c +++ b/drivers/gpu/drm/radeon/radeon_encoders.c @@ -26,6 +26,7 @@ #include +#include #include #include diff --git a/drivers/gpu/drm/radeon/radeon_mode.h b/drivers/gpu/drm/radeon/radeon_mode.h index 1decdcec0264..59c4db13d90a 100644 --- a/drivers/gpu/drm/radeon/radeon_mode.h +++ b/drivers/gpu/drm/radeon/radeon_mode.h @@ -32,13 +32,13 @@ #include #include -#include #include #include #include #include #include +struct edid; struct radeon_bo; struct radeon_device; From 671994e3bf33a414dc6a8c147969dae3a15ba9de Mon Sep 17 00:00:00 2001 From: Melissa Wen Date: Thu, 16 Nov 2023 18:57:49 -0100 Subject: [PATCH 42/92] drm/amd/display: add plane 3D LUT driver-specific properties Add 3D LUT property for plane color transformations using a 3D lookup table. 3D LUT allows for highly accurate and complex color transformations and is suitable to adjust the balance between color channels. It's also more complex to manage and require more computational resources. Since a 3D LUT has a limited number of entries in each dimension we want to use them in an optimal fashion. This means using the 3D LUT in a colorspace that is optimized for human vision, such as sRGB, PQ, or another non-linear space. Therefore, userpace may need one 1D LUT (shaper) before it to delinearize content and another 1D LUT after 3D LUT (blend) to linearize content again for blending. The next patches add these 1D LUTs to the plane color mgmt pipeline. v3: - improve commit message about 3D LUT - describe the 3D LUT entries and size (Harry) v4: - advertise 3D LUT max size as the size of a single-dimension v5: - get lut3d blob correctly (Joshua) - fix doc about 3d-lut dimension size (Sebastian) Signed-off-by: Melissa Wen Reviewed-by: Harry Wentland Signed-off-by: Alex Deucher --- drivers/gpu/drm/amd/amdgpu/amdgpu_mode.h | 18 ++++++++++++++ .../gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.h | 9 +++++++ .../amd/display/amdgpu_dm/amdgpu_dm_color.c | 14 +++++++++++ .../amd/display/amdgpu_dm/amdgpu_dm_plane.c | 24 +++++++++++++++++++ 4 files changed, 65 insertions(+) diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_mode.h b/drivers/gpu/drm/amd/amdgpu/amdgpu_mode.h index 218623e1b9aa..7350094f3db6 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_mode.h +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_mode.h @@ -364,6 +364,24 @@ struct amdgpu_mode_info { * @plane_hdr_mult_property: */ struct drm_property *plane_hdr_mult_property; + /** + * @plane_lut3d_property: Plane property for color transformation using + * a 3D LUT (pre-blending), a three-dimensional array where each + * element is an RGB triplet. Each dimension has the size of + * lut3d_size. The array contains samples from the approximated + * function. On AMD, values between samples are estimated by + * tetrahedral interpolation. The array is accessed with three indices, + * one for each input dimension (color channel), blue being the + * outermost dimension, red the innermost. + */ + struct drm_property *plane_lut3d_property; + /** + * @plane_degamma_lut_size_property: Plane property to define the max + * size of 3D LUT as supported by the driver (read-only). The max size + * is the max size of one dimension and, therefore, the max number of + * entries for 3D LUT array is the 3D LUT size cubed; + */ + struct drm_property *plane_lut3d_size_property; }; #define AMDGPU_MAX_BL_LEVEL 0xFF 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 d36d68d517a8..0f78024ac0e5 100644 --- a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.h +++ b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.h @@ -784,6 +784,11 @@ struct dm_plane_state { * TF is needed for any subsequent linear-to-non-linear transforms. */ __u64 hdr_mult; + /** + * @lut3d: 3D lookup table blob. The blob (if not NULL) is an array of + * &struct drm_color_lut. + */ + struct drm_property_blob *lut3d; }; struct dm_crtc_state { @@ -869,6 +874,10 @@ void amdgpu_dm_update_freesync_caps(struct drm_connector *connector, void amdgpu_dm_trigger_timing_sync(struct drm_device *dev); +/* 3D LUT max size is 17x17x17 (4913 entries) */ +#define MAX_COLOR_3DLUT_SIZE 17 +#define MAX_COLOR_3DLUT_BITDEPTH 12 +/* 1D LUT size */ #define MAX_COLOR_LUT_ENTRIES 4096 /* Legacy gamm LUT users such as X doesn't like large LUT sizes */ #define MAX_COLOR_LEGACY_LUT_ENTRIES 256 diff --git a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_color.c b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_color.c index b5b34a9209e4..a3b4f6a4c4a8 100644 --- a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_color.c +++ b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_color.c @@ -231,6 +231,20 @@ amdgpu_dm_create_color_properties(struct amdgpu_device *adev) return -ENOMEM; adev->mode_info.plane_hdr_mult_property = prop; + prop = drm_property_create(adev_to_drm(adev), + DRM_MODE_PROP_BLOB, + "AMD_PLANE_LUT3D", 0); + if (!prop) + return -ENOMEM; + adev->mode_info.plane_lut3d_property = prop; + + prop = drm_property_create_range(adev_to_drm(adev), + DRM_MODE_PROP_IMMUTABLE, + "AMD_PLANE_LUT3D_SIZE", 0, UINT_MAX); + if (!prop) + return -ENOMEM; + adev->mode_info.plane_lut3d_size_property = prop; + return 0; } #endif diff --git a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_plane.c b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_plane.c index 45a2c9b36630..e588cff7d6b0 100644 --- a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_plane.c +++ b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_plane.c @@ -1364,6 +1364,9 @@ static struct drm_plane_state *amdgpu_dm_plane_drm_plane_duplicate_state(struct if (old_dm_plane_state->degamma_lut) dm_plane_state->degamma_lut = drm_property_blob_get(old_dm_plane_state->degamma_lut); + if (old_dm_plane_state->lut3d) + dm_plane_state->lut3d = + drm_property_blob_get(old_dm_plane_state->lut3d); dm_plane_state->degamma_tf = old_dm_plane_state->degamma_tf; dm_plane_state->hdr_mult = old_dm_plane_state->hdr_mult; @@ -1437,6 +1440,8 @@ static void amdgpu_dm_plane_drm_plane_destroy_state(struct drm_plane *plane, if (dm_plane_state->degamma_lut) drm_property_blob_put(dm_plane_state->degamma_lut); + if (dm_plane_state->lut3d) + drm_property_blob_put(dm_plane_state->lut3d); if (dm_plane_state->dc_state) dc_plane_state_release(dm_plane_state->dc_state); @@ -1470,6 +1475,14 @@ dm_atomic_plane_attach_color_mgmt_properties(struct amdgpu_display_manager *dm, drm_object_attach_property(&plane->base, dm->adev->mode_info.plane_hdr_mult_property, AMDGPU_HDR_MULT_DEFAULT); + + if (dpp_color_caps.hw_3d_lut) { + drm_object_attach_property(&plane->base, + mode_info.plane_lut3d_property, 0); + drm_object_attach_property(&plane->base, + mode_info.plane_lut3d_size_property, + MAX_COLOR_3DLUT_SIZE); + } } static int @@ -1501,6 +1514,14 @@ dm_atomic_plane_set_property(struct drm_plane *plane, dm_plane_state->hdr_mult = val; dm_plane_state->base.color_mgmt_changed = 1; } + } else if (property == adev->mode_info.plane_lut3d_property) { + ret = drm_property_replace_blob_from_id(plane->dev, + &dm_plane_state->lut3d, + val, -1, + sizeof(struct drm_color_lut), + &replaced); + dm_plane_state->base.color_mgmt_changed |= replaced; + return ret; } else { drm_dbg_atomic(plane->dev, "[PLANE:%d:%s] unknown property [PROP:%d:%s]]\n", @@ -1528,6 +1549,9 @@ dm_atomic_plane_get_property(struct drm_plane *plane, *val = dm_plane_state->degamma_tf; } else if (property == adev->mode_info.plane_hdr_mult_property) { *val = dm_plane_state->hdr_mult; + } else if (property == adev->mode_info.plane_lut3d_property) { + *val = (dm_plane_state->lut3d) ? + dm_plane_state->lut3d->base.id : 0; } else { return -EINVAL; } From 058eb51912ca3a5fb121668b30e8e94d976afb27 Mon Sep 17 00:00:00 2001 From: Hawking Zhang Date: Tue, 12 Dec 2023 16:46:30 +0800 Subject: [PATCH 43/92] drm/amdgpu: Switch to aca bank for xgmi pcs err cnt Instead of software managed counters. Signed-off-by: Hawking Zhang Reviewed-by: Yang Wang Reviewed-by: Stanley.Yang Signed-off-by: Alex Deucher --- drivers/gpu/drm/amd/amdgpu/amdgpu_mca.h | 2 ++ drivers/gpu/drm/amd/pm/swsmu/smu13/smu_v13_0_6_ppt.c | 6 ++++-- 2 files changed, 6 insertions(+), 2 deletions(-) diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_mca.h b/drivers/gpu/drm/amd/amdgpu/amdgpu_mca.h index e51e8918e667..b399f1b62887 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_mca.h +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_mca.h @@ -46,6 +46,8 @@ #define MCA_REG__STATUS__ERRORCODEEXT(x) MCA_REG_FIELD(x, 21, 16) #define MCA_REG__STATUS__ERRORCODE(x) MCA_REG_FIELD(x, 15, 0) +#define MCA_REG__MISC0__ERRCNT(x) MCA_REG_FIELD(x, 43, 32) + #define MCA_REG__SYND__ERRORINFORMATION(x) MCA_REG_FIELD(x, 17, 0) enum amdgpu_mca_ip { diff --git a/drivers/gpu/drm/amd/pm/swsmu/smu13/smu_v13_0_6_ppt.c b/drivers/gpu/drm/amd/pm/swsmu/smu13/smu_v13_0_6_ppt.c index ddd782fbee7a..3998c9b31d07 100644 --- a/drivers/gpu/drm/amd/pm/swsmu/smu13/smu_v13_0_6_ppt.c +++ b/drivers/gpu/drm/amd/pm/swsmu/smu13/smu_v13_0_6_ppt.c @@ -2537,13 +2537,15 @@ static int mca_pcs_xgmi_mca_get_err_count(const struct mca_ras_info *mca_ras, st uint32_t *count) { u32 ext_error_code; + u32 err_cnt; ext_error_code = MCA_REG__STATUS__ERRORCODEEXT(entry->regs[MCA_REG_IDX_STATUS]); + err_cnt = MCA_REG__MISC0__ERRCNT(entry->regs[MCA_REG_IDX_MISC0]); if (type == AMDGPU_MCA_ERROR_TYPE_UE && ext_error_code == 0) - *count = 1; + *count = err_cnt; else if (type == AMDGPU_MCA_ERROR_TYPE_CE && ext_error_code == 6) - *count = 1; + *count = err_cnt; return 0; } From c01b9be7b20999bfeaacc1e574be0db93c14c478 Mon Sep 17 00:00:00 2001 From: Mario Limonciello Date: Tue, 12 Dec 2023 01:09:16 -0600 Subject: [PATCH 44/92] drm/amd: Fix a probing order problem on SDMA 2.4 commit 751e293f2c99 ("drm/amd: Move microcode init from sw_init to early_init for SDMA v2.4") made a fateful mistake in `adev->sdma.num_instances` wasn't declared when sdma_v2_4_init_microcode() was run. This caused probing to fail. Move the declaration to right before sdma_v2_4_init_microcode(). Closes: https://gitlab.freedesktop.org/drm/amd/-/issues/3043 Fixes: 751e293f2c99 ("drm/amd: Move microcode init from sw_init to early_init for SDMA v2.4") Reviewed-by: Alex Deucher Signed-off-by: Mario Limonciello Signed-off-by: Alex Deucher --- drivers/gpu/drm/amd/amdgpu/sdma_v2_4.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/drivers/gpu/drm/amd/amdgpu/sdma_v2_4.c b/drivers/gpu/drm/amd/amdgpu/sdma_v2_4.c index 45377a175250..8d5d86675a7f 100644 --- a/drivers/gpu/drm/amd/amdgpu/sdma_v2_4.c +++ b/drivers/gpu/drm/amd/amdgpu/sdma_v2_4.c @@ -813,12 +813,12 @@ static int sdma_v2_4_early_init(void *handle) struct amdgpu_device *adev = (struct amdgpu_device *)handle; int r; + adev->sdma.num_instances = SDMA_MAX_INSTANCE; + r = sdma_v2_4_init_microcode(adev); if (r) return r; - adev->sdma.num_instances = SDMA_MAX_INSTANCE; - sdma_v2_4_set_ring_funcs(adev); sdma_v2_4_set_buffer_funcs(adev); sdma_v2_4_set_vm_pte_funcs(adev); From bd33bb1409b494558a2935f7bbc7842def957fcd Mon Sep 17 00:00:00 2001 From: Jonathan Kim Date: Tue, 5 Dec 2023 12:22:07 -0500 Subject: [PATCH 45/92] drm/amdkfd: fix mes set shader debugger process management MES provides the driver a call to explicitly flush stale process memory within the MES to avoid a race condition that results in a fatal memory violation. When SET_SHADER_DEBUGGER is called, the driver passes a memory address that represents a process context address MES uses to keep track of future per-process calls. Normally, MES will purge its process context list when the last queue has been removed. The driver, however, can call SET_SHADER_DEBUGGER regardless of whether a queue has been added or not. If SET_SHADER_DEBUGGER has been called with no queues as the last call prior to process termination, the passed process context address will still reside within MES. On a new process call to SET_SHADER_DEBUGGER, the driver may end up passing an identical process context address value (based on per-process gpu memory address) to MES but is now pointing to a new allocated buffer object during KFD process creation. Since the MES is unaware of this, access of the passed address points to the stale object within MES and triggers a fatal memory violation. The solution is for KFD to explicitly flush the process context address from MES on process termination. Note that the flush call and the MES debugger calls use the same MES interface but are separated as KFD calls to avoid conflicting with each other. Signed-off-by: Jonathan Kim Tested-by: Alice Wong Reviewed-by: Eric Huang Signed-off-by: Alex Deucher --- drivers/gpu/drm/amd/amdgpu/amdgpu_mes.c | 31 +++++++++++++++++++ drivers/gpu/drm/amd/amdgpu/amdgpu_mes.h | 10 +++--- .../amd/amdkfd/kfd_process_queue_manager.c | 1 + drivers/gpu/drm/amd/include/mes_v11_api_def.h | 3 +- 4 files changed, 40 insertions(+), 5 deletions(-) diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_mes.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_mes.c index e544b823abf6..e98de23250dc 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_mes.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_mes.c @@ -916,6 +916,11 @@ int amdgpu_mes_set_shader_debugger(struct amdgpu_device *adev, op_input.op = MES_MISC_OP_SET_SHADER_DEBUGGER; op_input.set_shader_debugger.process_context_addr = process_context_addr; op_input.set_shader_debugger.flags.u32all = flags; + + /* use amdgpu mes_flush_shader_debugger instead */ + if (op_input.set_shader_debugger.flags.process_ctx_flush) + return -EINVAL; + op_input.set_shader_debugger.spi_gdbg_per_vmid_cntl = spi_gdbg_per_vmid_cntl; memcpy(op_input.set_shader_debugger.tcp_watch_cntl, tcp_watch_cntl, sizeof(op_input.set_shader_debugger.tcp_watch_cntl)); @@ -935,6 +940,32 @@ int amdgpu_mes_set_shader_debugger(struct amdgpu_device *adev, return r; } +int amdgpu_mes_flush_shader_debugger(struct amdgpu_device *adev, + uint64_t process_context_addr) +{ + struct mes_misc_op_input op_input = {0}; + int r; + + if (!adev->mes.funcs->misc_op) { + DRM_ERROR("mes flush shader debugger is not supported!\n"); + return -EINVAL; + } + + op_input.op = MES_MISC_OP_SET_SHADER_DEBUGGER; + op_input.set_shader_debugger.process_context_addr = process_context_addr; + op_input.set_shader_debugger.flags.process_ctx_flush = true; + + amdgpu_mes_lock(&adev->mes); + + r = adev->mes.funcs->misc_op(&adev->mes, &op_input); + if (r) + DRM_ERROR("failed to set_shader_debugger\n"); + + amdgpu_mes_unlock(&adev->mes); + + return r; +} + static void amdgpu_mes_ring_to_queue_props(struct amdgpu_device *adev, struct amdgpu_ring *ring, diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_mes.h b/drivers/gpu/drm/amd/amdgpu/amdgpu_mes.h index 894b9b133000..7d4f93fea937 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_mes.h +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_mes.h @@ -296,9 +296,10 @@ struct mes_misc_op_input { uint64_t process_context_addr; union { struct { - uint64_t single_memop : 1; - uint64_t single_alu_op : 1; - uint64_t reserved: 30; + uint32_t single_memop : 1; + uint32_t single_alu_op : 1; + uint32_t reserved: 29; + uint32_t process_ctx_flush: 1; }; uint32_t u32all; } flags; @@ -374,7 +375,8 @@ int amdgpu_mes_set_shader_debugger(struct amdgpu_device *adev, const uint32_t *tcp_watch_cntl, uint32_t flags, bool trap_en); - +int amdgpu_mes_flush_shader_debugger(struct amdgpu_device *adev, + uint64_t process_context_addr); int amdgpu_mes_add_ring(struct amdgpu_device *adev, int gang_id, int queue_type, int idx, struct amdgpu_mes_ctx_data *ctx_data, diff --git a/drivers/gpu/drm/amd/amdkfd/kfd_process_queue_manager.c b/drivers/gpu/drm/amd/amdkfd/kfd_process_queue_manager.c index 77f493262e05..8e55e78fce4e 100644 --- a/drivers/gpu/drm/amd/amdkfd/kfd_process_queue_manager.c +++ b/drivers/gpu/drm/amd/amdkfd/kfd_process_queue_manager.c @@ -87,6 +87,7 @@ void kfd_process_dequeue_from_device(struct kfd_process_device *pdd) return; dev->dqm->ops.process_termination(dev->dqm, &pdd->qpd); + amdgpu_mes_flush_shader_debugger(dev->adev, pdd->proc_ctx_gpu_addr); pdd->already_dequeued = true; } diff --git a/drivers/gpu/drm/amd/include/mes_v11_api_def.h b/drivers/gpu/drm/amd/include/mes_v11_api_def.h index 1fbfd1aa987e..ec5b9ab67c5e 100644 --- a/drivers/gpu/drm/amd/include/mes_v11_api_def.h +++ b/drivers/gpu/drm/amd/include/mes_v11_api_def.h @@ -572,7 +572,8 @@ struct SET_SHADER_DEBUGGER { struct { uint32_t single_memop : 1; /* SQ_DEBUG.single_memop */ uint32_t single_alu_op : 1; /* SQ_DEBUG.single_alu_op */ - uint32_t reserved : 30; + uint32_t reserved : 29; + uint32_t process_ctx_flush : 1; }; uint32_t u32all; } flags; From f545d82479b46368bf00d0bfecf33fa914bd5f8f Mon Sep 17 00:00:00 2001 From: Melissa Wen Date: Thu, 16 Nov 2023 18:57:50 -0100 Subject: [PATCH 46/92] drm/amd/display: add plane shaper LUT and TF driver-specific properties On AMD HW, 3D LUT always assumes a preceding shaper 1D LUT used for delinearizing and/or normalizing the color space before applying a 3D LUT. Add pre-defined transfer function to enable delinearizing content with or without shaper LUT, where AMD color module calculates the resulted shaper curve. We apply an inverse EOTF to go from linear values to encoded values. If we are already in a non-linear space and/or don't need to normalize values, we can bypass shaper LUT with a linear transfer function that is also the default TF value. There is no shaper ROM. When setting shaper TF (!= Identity) and LUT at the same time, the color module will combine the pre-defined TF and the custom LUT values into the LUT that's actually programmed. v2: - squash commits for shaper LUT and shaper TF - define inverse EOTF as supported shaper TFs v3: - spell out TF+LUT behavior in the commit and comments (Harry) - replace BT709 EOTF by inv OETF v5: - get shaper blob correctly (Joshua) Reviewed-by: Harry Wentland Signed-off-by: Melissa Wen Signed-off-by: Alex Deucher --- drivers/gpu/drm/amd/amdgpu/amdgpu_mode.h | 21 ++++++++++++ .../gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.h | 11 +++++++ .../amd/display/amdgpu_dm/amdgpu_dm_color.c | 29 ++++++++++++++++ .../amd/display/amdgpu_dm/amdgpu_dm_plane.c | 33 +++++++++++++++++++ 4 files changed, 94 insertions(+) diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_mode.h b/drivers/gpu/drm/amd/amdgpu/amdgpu_mode.h index 7350094f3db6..981a2eb25257 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_mode.h +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_mode.h @@ -364,6 +364,27 @@ struct amdgpu_mode_info { * @plane_hdr_mult_property: */ struct drm_property *plane_hdr_mult_property; + /** + * @shaper_lut_property: Plane property to set pre-blending shaper LUT + * that converts color content before 3D LUT. If + * plane_shaper_tf_property != Identity TF, AMD color module will + * combine the user LUT values with pre-defined TF into the LUT + * parameters to be programmed. + */ + struct drm_property *plane_shaper_lut_property; + /** + * @shaper_lut_size_property: Plane property for the size of + * pre-blending shaper LUT as supported by the driver (read-only). + */ + struct drm_property *plane_shaper_lut_size_property; + /** + * @plane_shaper_tf_property: Plane property to set a predefined + * transfer function for pre-blending shaper (before applying 3D LUT) + * with or without LUT. There is no shaper ROM, but we can use AMD + * color modules to program LUT parameters from predefined TF (or + * from a combination of pre-defined TF and the custom 1D LUT). + */ + struct drm_property *plane_shaper_tf_property; /** * @plane_lut3d_property: Plane property for color transformation using * a 3D LUT (pre-blending), a three-dimensional array where each 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 0f78024ac0e5..9bee60cbc2b5 100644 --- a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.h +++ b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.h @@ -784,6 +784,17 @@ struct dm_plane_state { * TF is needed for any subsequent linear-to-non-linear transforms. */ __u64 hdr_mult; + /** + * @shaper_lut: shaper lookup table blob. The blob (if not NULL) is an + * array of &struct drm_color_lut. + */ + struct drm_property_blob *shaper_lut; + /** + * @shaper_tf: + * + * Predefined transfer function to delinearize color space. + */ + enum amdgpu_transfer_function shaper_tf; /** * @lut3d: 3D lookup table blob. The blob (if not NULL) is an array of * &struct drm_color_lut. diff --git a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_color.c b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_color.c index a3b4f6a4c4a8..866c6f1d3d99 100644 --- a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_color.c +++ b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_color.c @@ -173,6 +173,14 @@ static const u32 amdgpu_eotf = BIT(AMDGPU_TRANSFER_FUNCTION_GAMMA24_EOTF) | BIT(AMDGPU_TRANSFER_FUNCTION_GAMMA26_EOTF); +static const u32 amdgpu_inv_eotf = + BIT(AMDGPU_TRANSFER_FUNCTION_SRGB_INV_EOTF) | + BIT(AMDGPU_TRANSFER_FUNCTION_BT709_OETF) | + BIT(AMDGPU_TRANSFER_FUNCTION_PQ_INV_EOTF) | + BIT(AMDGPU_TRANSFER_FUNCTION_GAMMA22_INV_EOTF) | + BIT(AMDGPU_TRANSFER_FUNCTION_GAMMA24_INV_EOTF) | + BIT(AMDGPU_TRANSFER_FUNCTION_GAMMA26_INV_EOTF); + static struct drm_property * amdgpu_create_tf_property(struct drm_device *dev, const char *name, @@ -231,6 +239,27 @@ amdgpu_dm_create_color_properties(struct amdgpu_device *adev) return -ENOMEM; adev->mode_info.plane_hdr_mult_property = prop; + prop = drm_property_create(adev_to_drm(adev), + DRM_MODE_PROP_BLOB, + "AMD_PLANE_SHAPER_LUT", 0); + if (!prop) + return -ENOMEM; + adev->mode_info.plane_shaper_lut_property = prop; + + prop = drm_property_create_range(adev_to_drm(adev), + DRM_MODE_PROP_IMMUTABLE, + "AMD_PLANE_SHAPER_LUT_SIZE", 0, UINT_MAX); + if (!prop) + return -ENOMEM; + adev->mode_info.plane_shaper_lut_size_property = prop; + + prop = amdgpu_create_tf_property(adev_to_drm(adev), + "AMD_PLANE_SHAPER_TF", + amdgpu_inv_eotf); + if (!prop) + return -ENOMEM; + adev->mode_info.plane_shaper_tf_property = prop; + prop = drm_property_create(adev_to_drm(adev), DRM_MODE_PROP_BLOB, "AMD_PLANE_LUT3D", 0); diff --git a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_plane.c b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_plane.c index e588cff7d6b0..7f0de29889f0 100644 --- a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_plane.c +++ b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_plane.c @@ -1343,6 +1343,7 @@ static void amdgpu_dm_plane_drm_plane_reset(struct drm_plane *plane) __drm_atomic_helper_plane_reset(plane, &amdgpu_state->base); amdgpu_state->degamma_tf = AMDGPU_TRANSFER_FUNCTION_DEFAULT; amdgpu_state->hdr_mult = AMDGPU_HDR_MULT_DEFAULT; + amdgpu_state->shaper_tf = AMDGPU_TRANSFER_FUNCTION_DEFAULT; } static struct drm_plane_state *amdgpu_dm_plane_drm_plane_duplicate_state(struct drm_plane *plane) @@ -1364,12 +1365,16 @@ static struct drm_plane_state *amdgpu_dm_plane_drm_plane_duplicate_state(struct if (old_dm_plane_state->degamma_lut) dm_plane_state->degamma_lut = drm_property_blob_get(old_dm_plane_state->degamma_lut); + if (old_dm_plane_state->shaper_lut) + dm_plane_state->shaper_lut = + drm_property_blob_get(old_dm_plane_state->shaper_lut); if (old_dm_plane_state->lut3d) dm_plane_state->lut3d = drm_property_blob_get(old_dm_plane_state->lut3d); dm_plane_state->degamma_tf = old_dm_plane_state->degamma_tf; dm_plane_state->hdr_mult = old_dm_plane_state->hdr_mult; + dm_plane_state->shaper_tf = old_dm_plane_state->shaper_tf; return &dm_plane_state->base; } @@ -1442,6 +1447,8 @@ static void amdgpu_dm_plane_drm_plane_destroy_state(struct drm_plane *plane, drm_property_blob_put(dm_plane_state->degamma_lut); if (dm_plane_state->lut3d) drm_property_blob_put(dm_plane_state->lut3d); + if (dm_plane_state->shaper_lut) + drm_property_blob_put(dm_plane_state->shaper_lut); if (dm_plane_state->dc_state) dc_plane_state_release(dm_plane_state->dc_state); @@ -1477,6 +1484,14 @@ dm_atomic_plane_attach_color_mgmt_properties(struct amdgpu_display_manager *dm, AMDGPU_HDR_MULT_DEFAULT); if (dpp_color_caps.hw_3d_lut) { + drm_object_attach_property(&plane->base, + mode_info.plane_shaper_lut_property, 0); + drm_object_attach_property(&plane->base, + mode_info.plane_shaper_lut_size_property, + MAX_COLOR_LUT_ENTRIES); + drm_object_attach_property(&plane->base, + mode_info.plane_shaper_tf_property, + AMDGPU_TRANSFER_FUNCTION_DEFAULT); drm_object_attach_property(&plane->base, mode_info.plane_lut3d_property, 0); drm_object_attach_property(&plane->base, @@ -1514,6 +1529,19 @@ dm_atomic_plane_set_property(struct drm_plane *plane, dm_plane_state->hdr_mult = val; dm_plane_state->base.color_mgmt_changed = 1; } + } else if (property == adev->mode_info.plane_shaper_lut_property) { + ret = drm_property_replace_blob_from_id(plane->dev, + &dm_plane_state->shaper_lut, + val, -1, + sizeof(struct drm_color_lut), + &replaced); + dm_plane_state->base.color_mgmt_changed |= replaced; + return ret; + } else if (property == adev->mode_info.plane_shaper_tf_property) { + if (dm_plane_state->shaper_tf != val) { + dm_plane_state->shaper_tf = val; + dm_plane_state->base.color_mgmt_changed = 1; + } } else if (property == adev->mode_info.plane_lut3d_property) { ret = drm_property_replace_blob_from_id(plane->dev, &dm_plane_state->lut3d, @@ -1549,6 +1577,11 @@ dm_atomic_plane_get_property(struct drm_plane *plane, *val = dm_plane_state->degamma_tf; } else if (property == adev->mode_info.plane_hdr_mult_property) { *val = dm_plane_state->hdr_mult; + } else if (property == adev->mode_info.plane_shaper_lut_property) { + *val = (dm_plane_state->shaper_lut) ? + dm_plane_state->shaper_lut->base.id : 0; + } else if (property == adev->mode_info.plane_shaper_tf_property) { + *val = dm_plane_state->shaper_tf; } else if (property == adev->mode_info.plane_lut3d_property) { *val = (dm_plane_state->lut3d) ? dm_plane_state->lut3d->base.id : 0; From 0ef47454dc82358b62a424b37c7520a84f307edb Mon Sep 17 00:00:00 2001 From: Joshua Ashton Date: Thu, 16 Nov 2023 18:57:51 -0100 Subject: [PATCH 47/92] drm/amd/display: add plane blend LUT and TF driver-specific properties Blend 1D LUT or a pre-defined transfer function (TF) can be set to linearize content before blending, so that it's positioned just before blending planes in the AMD color mgmt pipeline, and after 3D LUT (non-linear space). Shaper and Blend LUTs are 1D LUTs that sandwich 3D LUT. Drivers should advertize blend properties according to HW caps. There is no blend ROM for pre-defined TF. When setting blend TF (!= Identity) and LUT at the same time, the color module will combine the pre-defined TF and the custom LUT values into the LUT that's actually programmed. v3: - spell out TF+LUT behavior in the commit and comments (Harry) v5: - get blend blob correctly Reviewed-by: Harry Wentland Signed-off-by: Joshua Ashton Signed-off-by: Melissa Wen Signed-off-by: Alex Deucher --- drivers/gpu/drm/amd/amdgpu/amdgpu_mode.h | 22 +++++++++++ .../gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.h | 12 ++++++ .../amd/display/amdgpu_dm/amdgpu_dm_color.c | 21 +++++++++++ .../amd/display/amdgpu_dm/amdgpu_dm_plane.c | 37 +++++++++++++++++++ 4 files changed, 92 insertions(+) diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_mode.h b/drivers/gpu/drm/amd/amdgpu/amdgpu_mode.h index 981a2eb25257..6ff37272ccad 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_mode.h +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_mode.h @@ -403,6 +403,28 @@ struct amdgpu_mode_info { * entries for 3D LUT array is the 3D LUT size cubed; */ struct drm_property *plane_lut3d_size_property; + /** + * @plane_blend_lut_property: Plane property for output gamma before + * blending. Userspace set a blend LUT to convert colors after 3D LUT + * conversion. It works as a post-3DLUT 1D LUT. With shaper LUT, they + * are sandwiching 3D LUT with two 1D LUT. If plane_blend_tf_property + * != Identity TF, AMD color module will combine the user LUT values + * with pre-defined TF into the LUT parameters to be programmed. + */ + struct drm_property *plane_blend_lut_property; + /** + * @plane_blend_lut_size_property: Plane property to define the max + * size of blend LUT as supported by the driver (read-only). + */ + struct drm_property *plane_blend_lut_size_property; + /** + * @plane_blend_tf_property: Plane property to set a predefined + * transfer function for pre-blending blend/out_gamma (after applying + * 3D LUT) with or without LUT. There is no blend ROM, but we can use + * AMD color modules to program LUT parameters from predefined TF (or + * from a combination of pre-defined TF and the custom 1D LUT). + */ + struct drm_property *plane_blend_tf_property; }; #define AMDGPU_MAX_BL_LEVEL 0xFF 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 9bee60cbc2b5..a9a0bc875b78 100644 --- a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.h +++ b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.h @@ -800,6 +800,18 @@ struct dm_plane_state { * &struct drm_color_lut. */ struct drm_property_blob *lut3d; + /** + * @blend_lut: blend lut lookup table blob. The blob (if not NULL) is an + * array of &struct drm_color_lut. + */ + struct drm_property_blob *blend_lut; + /** + * @blend_tf: + * + * Pre-defined transfer function for converting plane pixel data before + * applying blend LUT. + */ + enum amdgpu_transfer_function blend_tf; }; struct dm_crtc_state { diff --git a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_color.c b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_color.c index 866c6f1d3d99..425676f5afb7 100644 --- a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_color.c +++ b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_color.c @@ -274,6 +274,27 @@ amdgpu_dm_create_color_properties(struct amdgpu_device *adev) return -ENOMEM; adev->mode_info.plane_lut3d_size_property = prop; + prop = drm_property_create(adev_to_drm(adev), + DRM_MODE_PROP_BLOB, + "AMD_PLANE_BLEND_LUT", 0); + if (!prop) + return -ENOMEM; + adev->mode_info.plane_blend_lut_property = prop; + + prop = drm_property_create_range(adev_to_drm(adev), + DRM_MODE_PROP_IMMUTABLE, + "AMD_PLANE_BLEND_LUT_SIZE", 0, UINT_MAX); + if (!prop) + return -ENOMEM; + adev->mode_info.plane_blend_lut_size_property = prop; + + prop = amdgpu_create_tf_property(adev_to_drm(adev), + "AMD_PLANE_BLEND_TF", + amdgpu_eotf); + if (!prop) + return -ENOMEM; + adev->mode_info.plane_blend_tf_property = prop; + return 0; } #endif diff --git a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_plane.c b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_plane.c index 7f0de29889f0..35a4732483ba 100644 --- a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_plane.c +++ b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_plane.c @@ -1344,6 +1344,7 @@ static void amdgpu_dm_plane_drm_plane_reset(struct drm_plane *plane) amdgpu_state->degamma_tf = AMDGPU_TRANSFER_FUNCTION_DEFAULT; amdgpu_state->hdr_mult = AMDGPU_HDR_MULT_DEFAULT; amdgpu_state->shaper_tf = AMDGPU_TRANSFER_FUNCTION_DEFAULT; + amdgpu_state->blend_tf = AMDGPU_TRANSFER_FUNCTION_DEFAULT; } static struct drm_plane_state *amdgpu_dm_plane_drm_plane_duplicate_state(struct drm_plane *plane) @@ -1371,10 +1372,14 @@ static struct drm_plane_state *amdgpu_dm_plane_drm_plane_duplicate_state(struct if (old_dm_plane_state->lut3d) dm_plane_state->lut3d = drm_property_blob_get(old_dm_plane_state->lut3d); + if (old_dm_plane_state->blend_lut) + dm_plane_state->blend_lut = + drm_property_blob_get(old_dm_plane_state->blend_lut); dm_plane_state->degamma_tf = old_dm_plane_state->degamma_tf; dm_plane_state->hdr_mult = old_dm_plane_state->hdr_mult; dm_plane_state->shaper_tf = old_dm_plane_state->shaper_tf; + dm_plane_state->blend_tf = old_dm_plane_state->blend_tf; return &dm_plane_state->base; } @@ -1449,6 +1454,8 @@ static void amdgpu_dm_plane_drm_plane_destroy_state(struct drm_plane *plane, drm_property_blob_put(dm_plane_state->lut3d); if (dm_plane_state->shaper_lut) drm_property_blob_put(dm_plane_state->shaper_lut); + if (dm_plane_state->blend_lut) + drm_property_blob_put(dm_plane_state->blend_lut); if (dm_plane_state->dc_state) dc_plane_state_release(dm_plane_state->dc_state); @@ -1498,6 +1505,17 @@ dm_atomic_plane_attach_color_mgmt_properties(struct amdgpu_display_manager *dm, mode_info.plane_lut3d_size_property, MAX_COLOR_3DLUT_SIZE); } + + if (dpp_color_caps.ogam_ram) { + drm_object_attach_property(&plane->base, + mode_info.plane_blend_lut_property, 0); + drm_object_attach_property(&plane->base, + mode_info.plane_blend_lut_size_property, + MAX_COLOR_LUT_ENTRIES); + drm_object_attach_property(&plane->base, + mode_info.plane_blend_tf_property, + AMDGPU_TRANSFER_FUNCTION_DEFAULT); + } } static int @@ -1550,6 +1568,19 @@ dm_atomic_plane_set_property(struct drm_plane *plane, &replaced); dm_plane_state->base.color_mgmt_changed |= replaced; return ret; + } else if (property == adev->mode_info.plane_blend_lut_property) { + ret = drm_property_replace_blob_from_id(plane->dev, + &dm_plane_state->blend_lut, + val, -1, + sizeof(struct drm_color_lut), + &replaced); + dm_plane_state->base.color_mgmt_changed |= replaced; + return ret; + } else if (property == adev->mode_info.plane_blend_tf_property) { + if (dm_plane_state->blend_tf != val) { + dm_plane_state->blend_tf = val; + dm_plane_state->base.color_mgmt_changed = 1; + } } else { drm_dbg_atomic(plane->dev, "[PLANE:%d:%s] unknown property [PROP:%d:%s]]\n", @@ -1585,6 +1616,12 @@ dm_atomic_plane_get_property(struct drm_plane *plane, } else if (property == adev->mode_info.plane_lut3d_property) { *val = (dm_plane_state->lut3d) ? dm_plane_state->lut3d->base.id : 0; + } else if (property == adev->mode_info.plane_blend_lut_property) { + *val = (dm_plane_state->blend_lut) ? + dm_plane_state->blend_lut->base.id : 0; + } else if (property == adev->mode_info.plane_blend_tf_property) { + *val = dm_plane_state->blend_tf; + } else { return -EINVAL; } From 0f5afa190b890052cae187496f660699f00067ef Mon Sep 17 00:00:00 2001 From: Melissa Wen Date: Thu, 16 Nov 2023 18:57:52 -0100 Subject: [PATCH 48/92] drm/amd/display: add CRTC gamma TF driver-specific property Add AMD pre-defined transfer function property to default DRM CRTC gamma to convert to wire encoding with or without a user gamma LUT. There is no post-blending regamma ROM for pre-defined TF. When setting Gamma TF (!= Identity) and LUT at the same time, the color module will combine the pre-defined TF and the custom LUT values into the LUT that's actually programmed. v2: - enable CRTC prop in the end of driver-specific prop sequence - define inverse EOTFs as supported regamma TFs - reword driver-specific function doc to remove shaper/3D LUT v3: - spell out TF+LUT behavior in the commit and comments (Harry) Reviewed-by: Harry Wentland Co-developed-by: Joshua Ashton Signed-off-by: Joshua Ashton Signed-off-by: Melissa Wen Signed-off-by: Alex Deucher --- drivers/gpu/drm/amd/amdgpu/amdgpu_mode.h | 7 ++ .../gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.h | 8 +++ .../amd/display/amdgpu_dm/amdgpu_dm_color.c | 7 ++ .../amd/display/amdgpu_dm/amdgpu_dm_crtc.c | 72 +++++++++++++++++++ 4 files changed, 94 insertions(+) diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_mode.h b/drivers/gpu/drm/amd/amdgpu/amdgpu_mode.h index 6ff37272ccad..9fbe09743dde 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_mode.h +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_mode.h @@ -425,6 +425,13 @@ struct amdgpu_mode_info { * from a combination of pre-defined TF and the custom 1D LUT). */ struct drm_property *plane_blend_tf_property; + /* @regamma_tf_property: Transfer function for CRTC regamma + * (post-blending). Possible values are defined by `enum + * amdgpu_transfer_function`. There is no regamma ROM, but we can use + * AMD color modules to program LUT parameters from predefined TF (or + * from a combination of pre-defined TF and the custom 1D LUT). + */ + struct drm_property *regamma_tf_property; }; #define AMDGPU_MAX_BL_LEVEL 0xFF 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 a9a0bc875b78..d9923c9f39b0 100644 --- a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.h +++ b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.h @@ -836,6 +836,14 @@ struct dm_crtc_state { struct dc_info_packet vrr_infopacket; int abm_level; + + /** + * @regamma_tf: + * + * Pre-defined transfer function for converting internal FB -> wire + * encoding. + */ + enum amdgpu_transfer_function regamma_tf; }; #define to_dm_crtc_state(x) container_of(x, struct dm_crtc_state, base) diff --git a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_color.c b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_color.c index 425676f5afb7..4150e9370daf 100644 --- a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_color.c +++ b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_color.c @@ -295,6 +295,13 @@ amdgpu_dm_create_color_properties(struct amdgpu_device *adev) return -ENOMEM; adev->mode_info.plane_blend_tf_property = prop; + prop = amdgpu_create_tf_property(adev_to_drm(adev), + "AMD_CRTC_REGAMMA_TF", + amdgpu_inv_eotf); + if (!prop) + return -ENOMEM; + adev->mode_info.regamma_tf_property = prop; + return 0; } #endif 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 cb0b48bb2a7d..9224e2d7f32b 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 @@ -260,6 +260,7 @@ static struct drm_crtc_state *amdgpu_dm_crtc_duplicate_state(struct drm_crtc *cr state->freesync_config = cur->freesync_config; state->cm_has_degamma = cur->cm_has_degamma; state->cm_is_degamma_srgb = cur->cm_is_degamma_srgb; + state->regamma_tf = cur->regamma_tf; state->crc_skip_count = cur->crc_skip_count; state->mpo_requested = cur->mpo_requested; /* TODO Duplicate dc_stream after objects are stream object is flattened */ @@ -296,6 +297,70 @@ static int amdgpu_dm_crtc_late_register(struct drm_crtc *crtc) } #endif +#ifdef AMD_PRIVATE_COLOR +/** + * drm_crtc_additional_color_mgmt - enable additional color properties + * @crtc: DRM CRTC + * + * This function lets the driver enable post-blending CRTC regamma transfer + * function property in addition to DRM CRTC gamma LUT. Default value means + * linear transfer function, which is the default CRTC gamma LUT behaviour + * without this property. + */ +static void +dm_crtc_additional_color_mgmt(struct drm_crtc *crtc) +{ + struct amdgpu_device *adev = drm_to_adev(crtc->dev); + + if(adev->dm.dc->caps.color.mpc.ogam_ram) + drm_object_attach_property(&crtc->base, + adev->mode_info.regamma_tf_property, + AMDGPU_TRANSFER_FUNCTION_DEFAULT); +} + +static int +amdgpu_dm_atomic_crtc_set_property(struct drm_crtc *crtc, + struct drm_crtc_state *state, + struct drm_property *property, + uint64_t val) +{ + struct amdgpu_device *adev = drm_to_adev(crtc->dev); + struct dm_crtc_state *acrtc_state = to_dm_crtc_state(state); + + if (property == adev->mode_info.regamma_tf_property) { + if (acrtc_state->regamma_tf != val) { + acrtc_state->regamma_tf = val; + acrtc_state->base.color_mgmt_changed |= 1; + } + } else { + drm_dbg_atomic(crtc->dev, + "[CRTC:%d:%s] unknown property [PROP:%d:%s]]\n", + crtc->base.id, crtc->name, + property->base.id, property->name); + return -EINVAL; + } + + return 0; +} + +static int +amdgpu_dm_atomic_crtc_get_property(struct drm_crtc *crtc, + const struct drm_crtc_state *state, + struct drm_property *property, + uint64_t *val) +{ + struct amdgpu_device *adev = drm_to_adev(crtc->dev); + struct dm_crtc_state *acrtc_state = to_dm_crtc_state(state); + + if (property == adev->mode_info.regamma_tf_property) + *val = acrtc_state->regamma_tf; + else + return -EINVAL; + + return 0; +} +#endif + /* Implemented only the options currently available for the driver */ static const struct drm_crtc_funcs amdgpu_dm_crtc_funcs = { .reset = amdgpu_dm_crtc_reset_state, @@ -314,6 +379,10 @@ static const struct drm_crtc_funcs amdgpu_dm_crtc_funcs = { #if defined(CONFIG_DEBUG_FS) .late_register = amdgpu_dm_crtc_late_register, #endif +#ifdef AMD_PRIVATE_COLOR + .atomic_set_property = amdgpu_dm_atomic_crtc_set_property, + .atomic_get_property = amdgpu_dm_atomic_crtc_get_property, +#endif }; static void amdgpu_dm_crtc_helper_disable(struct drm_crtc *crtc) @@ -489,6 +558,9 @@ int amdgpu_dm_crtc_init(struct amdgpu_display_manager *dm, drm_mode_crtc_set_gamma_size(&acrtc->base, MAX_COLOR_LEGACY_LUT_ENTRIES); +#ifdef AMD_PRIVATE_COLOR + dm_crtc_additional_color_mgmt(&acrtc->base); +#endif return 0; fail: From 98fbb52772063ad2547d6d1b80ff99bc26761e79 Mon Sep 17 00:00:00 2001 From: Melissa Wen Date: Thu, 16 Nov 2023 18:57:53 -0100 Subject: [PATCH 49/92] drm/amd/display: add comments to describe DM crtc color mgmt behavior Describe some expected behavior of the AMD DM color mgmt programming. Reviewed-by: Harry Wentland Signed-off-by: Melissa Wen Signed-off-by: Alex Deucher --- .../drm/amd/display/amdgpu_dm/amdgpu_dm_color.c | 16 ++++++++++++++-- 1 file changed, 14 insertions(+), 2 deletions(-) diff --git a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_color.c b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_color.c index 4150e9370daf..0b92513eab98 100644 --- a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_color.c +++ b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_color.c @@ -661,13 +661,25 @@ int amdgpu_dm_update_crtc_color_mgmt(struct dm_crtc_state *crtc) crtc->cm_is_degamma_srgb = true; stream->out_transfer_func->type = TF_TYPE_DISTRIBUTED_POINTS; stream->out_transfer_func->tf = TRANSFER_FUNCTION_SRGB; - + /* + * Note: although we pass has_rom as parameter here, we never + * actually use ROM because the color module only takes the ROM + * path if transfer_func->type == PREDEFINED. + * + * See more in mod_color_calculate_regamma_params() + */ r = __set_legacy_tf(stream->out_transfer_func, regamma_lut, regamma_size, has_rom); if (r) return r; } else if (has_regamma) { - /* If atomic regamma, CRTC RGM goes into RGM LUT. */ + /* + * CRTC RGM goes into RGM LUT. + * + * Note: there is no implicit sRGB regamma here. We are using + * degamma calculation from color module to calculate the curve + * from a linear base. + */ stream->out_transfer_func->type = TF_TYPE_DISTRIBUTED_POINTS; stream->out_transfer_func->tf = TRANSFER_FUNCTION_LINEAR; From 8b6b3f668f31a24b5406661388b9a69202e83e9d Mon Sep 17 00:00:00 2001 From: Melissa Wen Date: Thu, 16 Nov 2023 18:57:54 -0100 Subject: [PATCH 50/92] drm/amd/display: encapsulate atomic regamma operation We will wire up MPC 3D LUT to DM CRTC color pipeline in the next patch, but so far, only for atomic interface. By checking set_output_transfer_func in DC drivers with MPC 3D LUT support, we can verify that regamma is only programmed when 3D LUT programming fails. As a groundwork to introduce 3D LUT programming and better understand each step, detach atomic regamma programming from the crtc colocr updating code. Reviewed-by: Harry Wentland Signed-off-by: Melissa Wen Signed-off-by: Alex Deucher --- .../amd/display/amdgpu_dm/amdgpu_dm_color.c | 55 ++++++++++++------- 1 file changed, 35 insertions(+), 20 deletions(-) diff --git a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_color.c b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_color.c index 0b92513eab98..dec734f03832 100644 --- a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_color.c +++ b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_color.c @@ -525,6 +525,37 @@ static int __set_output_tf(struct dc_transfer_func *func, return res ? 0 : -ENOMEM; } +static int amdgpu_dm_set_atomic_regamma(struct dc_stream_state *stream, + const struct drm_color_lut *regamma_lut, + uint32_t regamma_size, bool has_rom) +{ + struct dc_transfer_func *out_tf = stream->out_transfer_func; + int ret = 0; + + if (regamma_size) { + /* + * CRTC RGM goes into RGM LUT. + * + * Note: there is no implicit sRGB regamma here. We are using + * degamma calculation from color module to calculate the curve + * from a linear base. + */ + out_tf->type = TF_TYPE_DISTRIBUTED_POINTS; + out_tf->tf = TRANSFER_FUNCTION_LINEAR; + + ret = __set_output_tf(out_tf, regamma_lut, regamma_size, has_rom); + } else { + /* + * No CRTC RGM means we can just put the block into bypass + * since we don't have any plane level adjustments using it. + */ + out_tf->type = TF_TYPE_BYPASS; + out_tf->tf = TRANSFER_FUNCTION_LINEAR; + } + + return ret; +} + /** * __set_input_tf - calculates the input transfer function based on expected * input space. @@ -672,28 +703,12 @@ int amdgpu_dm_update_crtc_color_mgmt(struct dm_crtc_state *crtc) regamma_size, has_rom); if (r) return r; - } else if (has_regamma) { - /* - * CRTC RGM goes into RGM LUT. - * - * Note: there is no implicit sRGB regamma here. We are using - * degamma calculation from color module to calculate the curve - * from a linear base. - */ - stream->out_transfer_func->type = TF_TYPE_DISTRIBUTED_POINTS; - stream->out_transfer_func->tf = TRANSFER_FUNCTION_LINEAR; - - r = __set_output_tf(stream->out_transfer_func, regamma_lut, - regamma_size, has_rom); + } else { + regamma_size = has_regamma ? regamma_size : 0; + r = amdgpu_dm_set_atomic_regamma(stream, regamma_lut, + regamma_size, has_rom); if (r) return r; - } else { - /* - * No CRTC RGM means we can just put the block into bypass - * since we don't have any plane level adjustments using it. - */ - stream->out_transfer_func->type = TF_TYPE_BYPASS; - stream->out_transfer_func->tf = TRANSFER_FUNCTION_LINEAR; } /* From 6bd20f0f165f444c1d8184ebd238dd92966c9dca Mon Sep 17 00:00:00 2001 From: Joshua Ashton Date: Thu, 16 Nov 2023 18:57:55 -0100 Subject: [PATCH 51/92] drm/amd/display: add CRTC gamma TF support Add predefined transfer function programming. There is no post-blending out gamma ROM for hardcoded curves, but we can use AMD color modules to program LUT parameters from pre-defined coefficients and an empty regamma LUT (or bump up LUT parameters with pre-defined TF values). v2: - update crtc color mgmt if regamma TF differs between states (Joshua) - map inverse EOTF to DC transfer function (Melissa) v3: - update AMDGPU TF list v4: - update comment regarding regamma behavior Reviewed-by: Harry Wentland Signed-off-by: Joshua Ashton Co-developed-by: Melissa Wen Signed-off-by: Melissa Wen Signed-off-by: Alex Deucher --- .../gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.c | 1 + .../amd/display/amdgpu_dm/amdgpu_dm_color.c | 77 +++++++++++++++---- 2 files changed, 61 insertions(+), 17 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 6efa0c1bba9d..22c46be291d1 100644 --- a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.c +++ b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.c @@ -9860,6 +9860,7 @@ static int dm_update_crtc_state(struct amdgpu_display_manager *dm, * when a modeset is needed, to ensure it gets reprogrammed. */ if (dm_new_crtc_state->base.color_mgmt_changed || + dm_old_crtc_state->regamma_tf != dm_new_crtc_state->regamma_tf || drm_atomic_crtc_needs_modeset(new_crtc_state)) { ret = amdgpu_dm_update_crtc_color_mgmt(dm_new_crtc_state); if (ret) diff --git a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_color.c b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_color.c index dec734f03832..9b930e3eb79d 100644 --- a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_color.c +++ b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_color.c @@ -490,16 +490,18 @@ static int __set_output_tf(struct dc_transfer_func *func, struct calculate_buffer cal_buffer = {0}; bool res; - ASSERT(lut && lut_size == MAX_COLOR_LUT_ENTRIES); - cal_buffer.buffer_index = -1; - gamma = dc_create_gamma(); - if (!gamma) - return -ENOMEM; + if (lut_size) { + ASSERT(lut && lut_size == MAX_COLOR_LUT_ENTRIES); - gamma->num_entries = lut_size; - __drm_lut_to_dc_gamma(lut, gamma, false); + gamma = dc_create_gamma(); + if (!gamma) + return -ENOMEM; + + gamma->num_entries = lut_size; + __drm_lut_to_dc_gamma(lut, gamma, false); + } if (func->tf == TRANSFER_FUNCTION_LINEAR) { /* @@ -507,41 +509,49 @@ static int __set_output_tf(struct dc_transfer_func *func, * on top of a linear input. But degamma params can be used * instead to simulate this. */ - gamma->type = GAMMA_CUSTOM; + if (gamma) + gamma->type = GAMMA_CUSTOM; res = mod_color_calculate_degamma_params(NULL, func, - gamma, true); + gamma, gamma != NULL); } else { /* * Assume sRGB. The actual mapping will depend on whether the * input was legacy or not. */ - gamma->type = GAMMA_CS_TFM_1D; - res = mod_color_calculate_regamma_params(func, gamma, false, + if (gamma) + gamma->type = GAMMA_CS_TFM_1D; + res = mod_color_calculate_regamma_params(func, gamma, gamma != NULL, has_rom, NULL, &cal_buffer); } - dc_gamma_release(&gamma); + if (gamma) + dc_gamma_release(&gamma); return res ? 0 : -ENOMEM; } static int amdgpu_dm_set_atomic_regamma(struct dc_stream_state *stream, const struct drm_color_lut *regamma_lut, - uint32_t regamma_size, bool has_rom) + uint32_t regamma_size, bool has_rom, + enum dc_transfer_func_predefined tf) { struct dc_transfer_func *out_tf = stream->out_transfer_func; int ret = 0; - if (regamma_size) { + if (regamma_size || tf != TRANSFER_FUNCTION_LINEAR) { /* * CRTC RGM goes into RGM LUT. * * Note: there is no implicit sRGB regamma here. We are using * degamma calculation from color module to calculate the curve - * from a linear base. + * from a linear base if gamma TF is not set. However, if gamma + * TF (!= Linear) and LUT are set at the same time, we will use + * regamma calculation, and the color module will combine the + * pre-defined TF and the custom LUT values into the LUT that's + * actually programmed. */ out_tf->type = TF_TYPE_DISTRIBUTED_POINTS; - out_tf->tf = TRANSFER_FUNCTION_LINEAR; + out_tf->tf = tf; ret = __set_output_tf(out_tf, regamma_lut, regamma_size, has_rom); } else { @@ -587,6 +597,36 @@ static int __set_input_tf(struct dc_transfer_func *func, return res ? 0 : -ENOMEM; } +static enum dc_transfer_func_predefined +amdgpu_tf_to_dc_tf(enum amdgpu_transfer_function tf) +{ + switch (tf) + { + default: + case AMDGPU_TRANSFER_FUNCTION_DEFAULT: + case AMDGPU_TRANSFER_FUNCTION_IDENTITY: + return TRANSFER_FUNCTION_LINEAR; + case AMDGPU_TRANSFER_FUNCTION_SRGB_EOTF: + case AMDGPU_TRANSFER_FUNCTION_SRGB_INV_EOTF: + return TRANSFER_FUNCTION_SRGB; + case AMDGPU_TRANSFER_FUNCTION_BT709_OETF: + case AMDGPU_TRANSFER_FUNCTION_BT709_INV_OETF: + return TRANSFER_FUNCTION_BT709; + case AMDGPU_TRANSFER_FUNCTION_PQ_EOTF: + case AMDGPU_TRANSFER_FUNCTION_PQ_INV_EOTF: + return TRANSFER_FUNCTION_PQ; + case AMDGPU_TRANSFER_FUNCTION_GAMMA22_EOTF: + case AMDGPU_TRANSFER_FUNCTION_GAMMA22_INV_EOTF: + return TRANSFER_FUNCTION_GAMMA22; + case AMDGPU_TRANSFER_FUNCTION_GAMMA24_EOTF: + case AMDGPU_TRANSFER_FUNCTION_GAMMA24_INV_EOTF: + return TRANSFER_FUNCTION_GAMMA24; + case AMDGPU_TRANSFER_FUNCTION_GAMMA26_EOTF: + case AMDGPU_TRANSFER_FUNCTION_GAMMA26_INV_EOTF: + return TRANSFER_FUNCTION_GAMMA26; + } +} + /** * amdgpu_dm_verify_lut_sizes - verifies if DRM luts match the hw supported sizes * @crtc_state: the DRM CRTC state @@ -654,9 +694,12 @@ int amdgpu_dm_update_crtc_color_mgmt(struct dm_crtc_state *crtc) const struct drm_color_lut *degamma_lut, *regamma_lut; uint32_t degamma_size, regamma_size; bool has_regamma, has_degamma; + enum dc_transfer_func_predefined tf = TRANSFER_FUNCTION_LINEAR; bool is_legacy; int r; + tf = amdgpu_tf_to_dc_tf(crtc->regamma_tf); + r = amdgpu_dm_verify_lut_sizes(&crtc->base); if (r) return r; @@ -706,7 +749,7 @@ int amdgpu_dm_update_crtc_color_mgmt(struct dm_crtc_state *crtc) } else { regamma_size = has_regamma ? regamma_size : 0; r = amdgpu_dm_set_atomic_regamma(stream, regamma_lut, - regamma_size, has_rom); + regamma_size, has_rom, tf); if (r) return r; } From d9501844d53897ca7ac04697b8504940c6dfdbb3 Mon Sep 17 00:00:00 2001 From: Jani Nikula Date: Tue, 12 Dec 2023 15:53:09 +0200 Subject: [PATCH 52/92] drm/amd: include drm/drm_edid.h only where needed Including drm_edid.h from amdgpu_mode.h causes the rebuild of literally hundreds of files when drm_edid.h is modified, while there are only a handful of files that actually need to include drm_edid.h. Signed-off-by: Jani Nikula Signed-off-by: Alex Deucher --- drivers/gpu/drm/amd/amdgpu/amdgpu_mode.h | 2 +- drivers/gpu/drm/amd/amdgpu/amdgpu_vkms.c | 1 + drivers/gpu/drm/amd/amdgpu/atombios_encoders.c | 1 + drivers/gpu/drm/amd/amdgpu/dce_v10_0.c | 1 + drivers/gpu/drm/amd/amdgpu/dce_v11_0.c | 1 + drivers/gpu/drm/amd/amdgpu/dce_v6_0.c | 1 + drivers/gpu/drm/amd/amdgpu/dce_v8_0.c | 1 + drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_mst_types.c | 1 + drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_wb.c | 1 + 9 files changed, 9 insertions(+), 1 deletion(-) diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_mode.h b/drivers/gpu/drm/amd/amdgpu/amdgpu_mode.h index 9fbe09743dde..790c08b8262d 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_mode.h +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_mode.h @@ -32,7 +32,6 @@ #include #include -#include #include #include #include @@ -51,6 +50,7 @@ struct amdgpu_device; struct amdgpu_encoder; struct amdgpu_router; struct amdgpu_hpd; +struct edid; #define to_amdgpu_crtc(x) container_of(x, struct amdgpu_crtc, base) #define to_amdgpu_connector(x) container_of(x, struct amdgpu_connector, base) diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_vkms.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_vkms.c index db6fc0cb18eb..453a4b786cfc 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_vkms.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_vkms.c @@ -1,6 +1,7 @@ // SPDX-License-Identifier: GPL-2.0+ #include +#include #include #include diff --git a/drivers/gpu/drm/amd/amdgpu/atombios_encoders.c b/drivers/gpu/drm/amd/amdgpu/atombios_encoders.c index 3ee219aa2891..7672abe6c140 100644 --- a/drivers/gpu/drm/amd/amdgpu/atombios_encoders.c +++ b/drivers/gpu/drm/amd/amdgpu/atombios_encoders.c @@ -28,6 +28,7 @@ #include +#include #include #include "amdgpu.h" #include "amdgpu_connectors.h" diff --git a/drivers/gpu/drm/amd/amdgpu/dce_v10_0.c b/drivers/gpu/drm/amd/amdgpu/dce_v10_0.c index bb666cb7522e..587ee632a3b8 100644 --- a/drivers/gpu/drm/amd/amdgpu/dce_v10_0.c +++ b/drivers/gpu/drm/amd/amdgpu/dce_v10_0.c @@ -21,6 +21,7 @@ * */ +#include #include #include #include diff --git a/drivers/gpu/drm/amd/amdgpu/dce_v11_0.c b/drivers/gpu/drm/amd/amdgpu/dce_v11_0.c index 7af277f61cca..f22ec27365bd 100644 --- a/drivers/gpu/drm/amd/amdgpu/dce_v11_0.c +++ b/drivers/gpu/drm/amd/amdgpu/dce_v11_0.c @@ -21,6 +21,7 @@ * */ +#include #include #include #include diff --git a/drivers/gpu/drm/amd/amdgpu/dce_v6_0.c b/drivers/gpu/drm/amd/amdgpu/dce_v6_0.c index 143efc37a17f..4dbe9b3259b5 100644 --- a/drivers/gpu/drm/amd/amdgpu/dce_v6_0.c +++ b/drivers/gpu/drm/amd/amdgpu/dce_v6_0.c @@ -23,6 +23,7 @@ #include +#include #include #include #include diff --git a/drivers/gpu/drm/amd/amdgpu/dce_v8_0.c b/drivers/gpu/drm/amd/amdgpu/dce_v8_0.c index adeddfb7ff12..05bcce23385e 100644 --- a/drivers/gpu/drm/amd/amdgpu/dce_v8_0.c +++ b/drivers/gpu/drm/amd/amdgpu/dce_v8_0.c @@ -21,6 +21,7 @@ * */ +#include #include #include #include diff --git a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_mst_types.c b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_mst_types.c index 19ae5587dd5a..941e96f100f4 100644 --- a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_mst_types.c +++ b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_mst_types.c @@ -28,6 +28,7 @@ #include #include #include +#include #include "dm_services.h" #include "amdgpu.h" #include "amdgpu_dm.h" diff --git a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_wb.c b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_wb.c index b3e634b0f712..16e72d623630 100644 --- a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_wb.c +++ b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_wb.c @@ -32,6 +32,7 @@ #include "amdgpu_display.h" #include "dc.h" +#include #include #include From 88d26ea639a8e9d314e6bffef5f382167e7203e2 Mon Sep 17 00:00:00 2001 From: Joshua Ashton Date: Thu, 16 Nov 2023 18:57:56 -0100 Subject: [PATCH 53/92] drm/amd/display: set sdr_ref_white_level to 80 for out_transfer_func Otherwise this is just initialized to 0. This needs to actually have a value so that compute_curve can work for PQ EOTF. Reviewed-by: Harry Wentland Signed-off-by: Joshua Ashton Co-developed-by: Melissa Wen Signed-off-by: Melissa Wen Signed-off-by: Alex Deucher --- drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_color.c | 2 ++ 1 file changed, 2 insertions(+) diff --git a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_color.c b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_color.c index 9b930e3eb79d..83c14fb57721 100644 --- a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_color.c +++ b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_color.c @@ -72,6 +72,7 @@ */ #define MAX_DRM_LUT_VALUE 0xFFFF +#define SDR_WHITE_LEVEL_INIT_VALUE 80 /** * amdgpu_dm_init_color_mod - Initialize the color module. @@ -552,6 +553,7 @@ static int amdgpu_dm_set_atomic_regamma(struct dc_stream_state *stream, */ out_tf->type = TF_TYPE_DISTRIBUTED_POINTS; out_tf->tf = tf; + out_tf->sdr_ref_white_level = SDR_WHITE_LEVEL_INIT_VALUE; ret = __set_output_tf(out_tf, regamma_lut, regamma_size, has_rom); } else { From 6bed9d550e51534415a56f8de33f5b9d4e728e53 Mon Sep 17 00:00:00 2001 From: Joshua Ashton Date: Thu, 16 Nov 2023 18:57:57 -0100 Subject: [PATCH 54/92] drm/amd/display: mark plane as needing reset if color props change We should reset a plane state if at least one of the color management properties differs from old and new state. Reviewed-by: Harry Wentland Signed-off-by: Joshua Ashton Co-developed-by: Melissa Wen Signed-off-by: Melissa Wen Signed-off-by: Alex Deucher --- drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.c | 15 +++++++++++++++ 1 file changed, 15 insertions(+) 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 22c46be291d1..0af5d908e169 100644 --- a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.c +++ b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.c @@ -9929,6 +9929,10 @@ static bool should_reset_plane(struct drm_atomic_state *state, */ for_each_oldnew_plane_in_state(state, other, old_other_state, new_other_state, i) { struct amdgpu_framebuffer *old_afb, *new_afb; + struct dm_plane_state *dm_new_other_state, *dm_old_other_state; + + dm_new_other_state = to_dm_plane_state(new_other_state); + dm_old_other_state = to_dm_plane_state(old_other_state); if (other->type == DRM_PLANE_TYPE_CURSOR) continue; @@ -9965,6 +9969,17 @@ static bool should_reset_plane(struct drm_atomic_state *state, old_other_state->color_encoding != new_other_state->color_encoding) return true; + /* HDR/Transfer Function changes. */ + if (dm_old_other_state->degamma_tf != dm_new_other_state->degamma_tf || + dm_old_other_state->degamma_lut != dm_new_other_state->degamma_lut || + dm_old_other_state->hdr_mult != dm_new_other_state->hdr_mult || + dm_old_other_state->shaper_lut != dm_new_other_state->shaper_lut || + dm_old_other_state->shaper_tf != dm_new_other_state->shaper_tf || + dm_old_other_state->lut3d != dm_new_other_state->lut3d || + dm_old_other_state->blend_lut != dm_new_other_state->blend_lut || + dm_old_other_state->blend_tf != dm_new_other_state->blend_tf) + return true; + /* Framebuffer checks fall at the end. */ if (!old_other_state->fb || !new_other_state->fb) continue; From 73e5ea616a9f8c261d07e63b421947949ad6cbce Mon Sep 17 00:00:00 2001 From: Melissa Wen Date: Thu, 16 Nov 2023 18:57:58 -0100 Subject: [PATCH 55/92] drm/amd/display: decouple steps for mapping CRTC degamma to DC plane The next patch adds pre-blending degamma to AMD color mgmt pipeline, but pre-blending degamma caps (DPP) is currently in use to provide DRM CRTC atomic degamma or implict degamma on legacy gamma. Detach degamma usage regarging CRTC color properties to manage plane and CRTC color correction combinations. Reviewed-by: Harry Wentland Signed-off-by: Melissa Wen Signed-off-by: Alex Deucher --- .../amd/display/amdgpu_dm/amdgpu_dm_color.c | 60 +++++++++++++------ 1 file changed, 42 insertions(+), 18 deletions(-) diff --git a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_color.c b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_color.c index 83c14fb57721..f13d22851fe4 100644 --- a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_color.c +++ b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_color.c @@ -789,20 +789,9 @@ int amdgpu_dm_update_crtc_color_mgmt(struct dm_crtc_state *crtc) return 0; } -/** - * amdgpu_dm_update_plane_color_mgmt: Maps DRM color management to DC plane. - * @crtc: amdgpu_dm crtc state - * @dc_plane_state: target DC surface - * - * Update the underlying dc_stream_state's input transfer function (ITF) in - * preparation for hardware commit. The transfer function used depends on - * the preparation done on the stream for color management. - * - * Returns: - * 0 on success. -ENOMEM if mem allocation fails. - */ -int amdgpu_dm_update_plane_color_mgmt(struct dm_crtc_state *crtc, - struct dc_plane_state *dc_plane_state) +static int +map_crtc_degamma_to_dc_plane(struct dm_crtc_state *crtc, + struct dc_plane_state *dc_plane_state) { const struct drm_color_lut *degamma_lut; enum dc_transfer_func_predefined tf = TRANSFER_FUNCTION_SRGB; @@ -825,8 +814,7 @@ int amdgpu_dm_update_plane_color_mgmt(struct dm_crtc_state *crtc, °amma_size); ASSERT(degamma_size == MAX_COLOR_LUT_ENTRIES); - dc_plane_state->in_transfer_func->type = - TF_TYPE_DISTRIBUTED_POINTS; + dc_plane_state->in_transfer_func->type = TF_TYPE_DISTRIBUTED_POINTS; /* * This case isn't fully correct, but also fairly @@ -862,7 +850,7 @@ int amdgpu_dm_update_plane_color_mgmt(struct dm_crtc_state *crtc, degamma_lut, degamma_size); if (r) return r; - } else if (crtc->cm_is_degamma_srgb) { + } else { /* * For legacy gamma support we need the regamma input * in linear space. Assume that the input is sRGB. @@ -872,8 +860,44 @@ int amdgpu_dm_update_plane_color_mgmt(struct dm_crtc_state *crtc, if (tf != TRANSFER_FUNCTION_SRGB && !mod_color_calculate_degamma_params(NULL, - dc_plane_state->in_transfer_func, NULL, false)) + dc_plane_state->in_transfer_func, + NULL, false)) return -ENOMEM; + } + + return 0; +} + +/** + * amdgpu_dm_update_plane_color_mgmt: Maps DRM color management to DC plane. + * @crtc: amdgpu_dm crtc state + * @dc_plane_state: target DC surface + * + * Update the underlying dc_stream_state's input transfer function (ITF) in + * preparation for hardware commit. The transfer function used depends on + * the preparation done on the stream for color management. + * + * Returns: + * 0 on success. -ENOMEM if mem allocation fails. + */ +int amdgpu_dm_update_plane_color_mgmt(struct dm_crtc_state *crtc, + struct dc_plane_state *dc_plane_state) +{ + bool has_crtc_cm_degamma; + int ret; + + has_crtc_cm_degamma = (crtc->cm_has_degamma || crtc->cm_is_degamma_srgb); + if (has_crtc_cm_degamma){ + /* + * AMD HW doesn't have post-blending degamma caps. When DRM + * CRTC atomic degamma is set, we maps it to DPP degamma block + * (pre-blending) or, on legacy gamma, we use DPP degamma to + * linearize (implicit degamma) from sRGB/BT709 according to + * the input space. + */ + ret = map_crtc_degamma_to_dc_plane(crtc, dc_plane_state); + if (ret) + return ret; } else { /* ...Otherwise we can just bypass the DGM block. */ dc_plane_state->in_transfer_func->type = TF_TYPE_BYPASS; From 683b8c7e7a94fb7445b8d300c7404322ad040bab Mon Sep 17 00:00:00 2001 From: =?UTF-8?q?Christian=20K=C3=B6nig?= Date: Fri, 8 Dec 2023 13:43:09 +0100 Subject: [PATCH 56/92] drm/amdgpu: fix tear down order in amdgpu_vm_pt_free MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit When freeing PD/PT with shadows it can happen that the shadow destruction races with detaching the PD/PT from the VM causing a NULL pointer dereference in the invalidation code. Fix this by detaching the the PD/PT from the VM first and then freeing the shadow instead. Signed-off-by: Christian König Fixes: https://gitlab.freedesktop.org/drm/amd/-/issues/2867 Cc: Reviewed-by: Alex Deucher Signed-off-by: Alex Deucher --- drivers/gpu/drm/amd/amdgpu/amdgpu_vm_pt.c | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-) diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_vm_pt.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_vm_pt.c index a2287bb25223..a160265ddc07 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_vm_pt.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_vm_pt.c @@ -642,13 +642,14 @@ static void amdgpu_vm_pt_free(struct amdgpu_vm_bo_base *entry) if (!entry->bo) return; + + entry->bo->vm_bo = NULL; shadow = amdgpu_bo_shadowed(entry->bo); if (shadow) { ttm_bo_set_bulk_move(&shadow->tbo, NULL); amdgpu_bo_unref(&shadow); } ttm_bo_set_bulk_move(&entry->bo->tbo, NULL); - entry->bo->vm_bo = NULL; spin_lock(&entry->vm->status_lock); list_del(&entry->vm_status); From 980f8710075acaeb226a94cde6dda8ffad30123c Mon Sep 17 00:00:00 2001 From: Joshua Ashton Date: Thu, 16 Nov 2023 18:57:59 -0100 Subject: [PATCH 57/92] drm/amd/display: add plane degamma TF and LUT support Set DC plane with user degamma LUT or predefined TF from driver-specific plane color properties. If plane and CRTC degamma are set in the same time, plane degamma has priority. That means, we only set CRTC degamma if we don't have plane degamma LUT or TF to configure. We return -EINVAL if we don't have plane degamma settings, so we can continue and check CRTC degamma. Reviewed-by: Harry Wentland Signed-off-by: Joshua Ashton Signed-off-by: Melissa Wen Signed-off-by: Alex Deucher --- .../gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.c | 4 +- .../gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.h | 1 + .../amd/display/amdgpu_dm/amdgpu_dm_color.c | 70 +++++++++++++++++-- 3 files changed, 69 insertions(+), 6 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 0af5d908e169..911db58d0e1c 100644 --- a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.c +++ b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.c @@ -5169,7 +5169,9 @@ static int fill_dc_plane_attributes(struct amdgpu_device *adev, * Always set input transfer function, since plane state is refreshed * every time. */ - ret = amdgpu_dm_update_plane_color_mgmt(dm_crtc_state, dc_plane_state); + ret = amdgpu_dm_update_plane_color_mgmt(dm_crtc_state, + plane_state, + dc_plane_state); if (ret) return ret; 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 d9923c9f39b0..16af549d0b42 100644 --- a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.h +++ b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.h @@ -918,6 +918,7 @@ int amdgpu_dm_create_color_properties(struct amdgpu_device *adev); int amdgpu_dm_verify_lut_sizes(const struct drm_crtc_state *crtc_state); int amdgpu_dm_update_crtc_color_mgmt(struct dm_crtc_state *crtc); int amdgpu_dm_update_plane_color_mgmt(struct dm_crtc_state *crtc, + struct drm_plane_state *plane_state, struct dc_plane_state *dc_plane_state); void amdgpu_dm_update_connector_after_detect( diff --git a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_color.c b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_color.c index f13d22851fe4..13117516f2a4 100644 --- a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_color.c +++ b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_color.c @@ -868,9 +868,58 @@ map_crtc_degamma_to_dc_plane(struct dm_crtc_state *crtc, return 0; } +static int +__set_dm_plane_degamma(struct drm_plane_state *plane_state, + struct dc_plane_state *dc_plane_state) +{ + struct dm_plane_state *dm_plane_state = to_dm_plane_state(plane_state); + const struct drm_color_lut *degamma_lut; + enum amdgpu_transfer_function tf = AMDGPU_TRANSFER_FUNCTION_DEFAULT; + uint32_t degamma_size; + bool has_degamma_lut; + int ret; + + degamma_lut = __extract_blob_lut(dm_plane_state->degamma_lut, + °amma_size); + + has_degamma_lut = degamma_lut && + !__is_lut_linear(degamma_lut, degamma_size); + + tf = dm_plane_state->degamma_tf; + + /* If we don't have plane degamma LUT nor TF to set on DC, we have + * nothing to do here, return. + */ + if (!has_degamma_lut && tf == AMDGPU_TRANSFER_FUNCTION_DEFAULT) + return -EINVAL; + + dc_plane_state->in_transfer_func->tf = amdgpu_tf_to_dc_tf(tf); + + if (has_degamma_lut) { + ASSERT(degamma_size == MAX_COLOR_LUT_ENTRIES); + + dc_plane_state->in_transfer_func->type = + TF_TYPE_DISTRIBUTED_POINTS; + + ret = __set_input_tf(dc_plane_state->in_transfer_func, + degamma_lut, degamma_size); + if (ret) + return ret; + } else { + dc_plane_state->in_transfer_func->type = + TF_TYPE_PREDEFINED; + + if (!mod_color_calculate_degamma_params(NULL, + dc_plane_state->in_transfer_func, NULL, false)) + return -ENOMEM; + } + return 0; +} + /** * amdgpu_dm_update_plane_color_mgmt: Maps DRM color management to DC plane. * @crtc: amdgpu_dm crtc state + * @plane_state: DRM plane state * @dc_plane_state: target DC surface * * Update the underlying dc_stream_state's input transfer function (ITF) in @@ -881,13 +930,28 @@ map_crtc_degamma_to_dc_plane(struct dm_crtc_state *crtc, * 0 on success. -ENOMEM if mem allocation fails. */ int amdgpu_dm_update_plane_color_mgmt(struct dm_crtc_state *crtc, + struct drm_plane_state *plane_state, struct dc_plane_state *dc_plane_state) { bool has_crtc_cm_degamma; int ret; + /* Initially, we can just bypass the DGM block. */ + dc_plane_state->in_transfer_func->type = TF_TYPE_BYPASS; + dc_plane_state->in_transfer_func->tf = TRANSFER_FUNCTION_LINEAR; + + /* After, we start to update values according to color props */ has_crtc_cm_degamma = (crtc->cm_has_degamma || crtc->cm_is_degamma_srgb); - if (has_crtc_cm_degamma){ + + ret = __set_dm_plane_degamma(plane_state, dc_plane_state); + if (ret != -EINVAL) + return ret; + + /* If we are here, it means we don't have plane degamma settings, check + * if we have CRTC degamma waiting for mapping to pre-blending degamma + * block + */ + if (has_crtc_cm_degamma) { /* * AMD HW doesn't have post-blending degamma caps. When DRM * CRTC atomic degamma is set, we maps it to DPP degamma block @@ -898,10 +962,6 @@ int amdgpu_dm_update_plane_color_mgmt(struct dm_crtc_state *crtc, ret = map_crtc_degamma_to_dc_plane(crtc, dc_plane_state); if (ret) return ret; - } else { - /* ...Otherwise we can just bypass the DGM block. */ - dc_plane_state->in_transfer_func->type = TF_TYPE_BYPASS; - dc_plane_state->in_transfer_func->tf = TRANSFER_FUNCTION_LINEAR; } return 0; From ef113a3b1964b40dd87287806865b947d70f7df5 Mon Sep 17 00:00:00 2001 From: Melissa Wen Date: Thu, 16 Nov 2023 18:58:00 -0100 Subject: [PATCH 58/92] drm/amd/display: reject atomic commit if setting both plane and CRTC degamma DC only has pre-blending degamma caps (plane/DPP) that is currently in use for CRTC/post-blending degamma, so that we don't have HW caps to perform plane and CRTC degamma at the same time. Reject atomic updates when serspace sets both plane and CRTC degamma properties. Reviewed-by: Harry Wentland Signed-off-by: Melissa Wen Signed-off-by: Alex Deucher --- .../gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_color.c | 13 ++++++++++++- 1 file changed, 12 insertions(+), 1 deletion(-) diff --git a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_color.c b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_color.c index 13117516f2a4..8fc715fa426c 100644 --- a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_color.c +++ b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_color.c @@ -944,9 +944,20 @@ int amdgpu_dm_update_plane_color_mgmt(struct dm_crtc_state *crtc, has_crtc_cm_degamma = (crtc->cm_has_degamma || crtc->cm_is_degamma_srgb); ret = __set_dm_plane_degamma(plane_state, dc_plane_state); - if (ret != -EINVAL) + if (ret == -ENOMEM) return ret; + /* We only have one degamma block available (pre-blending) for the + * whole color correction pipeline, so that we can't actually perform + * plane and CRTC degamma at the same time. Explicitly reject atomic + * updates when userspace sets both plane and CRTC degamma properties. + */ + if (has_crtc_cm_degamma && ret != -EINVAL){ + drm_dbg_kms(crtc->base.crtc->dev, + "doesn't support plane and CRTC degamma at the same time\n"); + return -EINVAL; + } + /* If we are here, it means we don't have plane degamma settings, check * if we have CRTC degamma waiting for mapping to pre-blending degamma * block From 889044f9e04f0829dd92640c551941bbe77bc0ea Mon Sep 17 00:00:00 2001 From: Joshua Ashton Date: Thu, 16 Nov 2023 18:58:01 -0100 Subject: [PATCH 59/92] drm/amd/display: add dc_fixpt_from_s3132 helper Detach value translation from CTM to reuse it for programming HDR multiplier property. Reviewed-by: Harry Wentland Signed-off-by: Joshua Ashton Signed-off-by: Melissa Wen Signed-off-by: Alex Deucher --- .../gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_color.c | 8 +------- drivers/gpu/drm/amd/display/include/fixed31_32.h | 12 ++++++++++++ 2 files changed, 13 insertions(+), 7 deletions(-) diff --git a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_color.c b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_color.c index 8fc715fa426c..c160dbef5224 100644 --- a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_color.c +++ b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_color.c @@ -405,7 +405,6 @@ static void __drm_lut_to_dc_gamma(const struct drm_color_lut *lut, static void __drm_ctm_to_dc_matrix(const struct drm_color_ctm *ctm, struct fixed31_32 *matrix) { - int64_t val; int i; /* @@ -424,12 +423,7 @@ static void __drm_ctm_to_dc_matrix(const struct drm_color_ctm *ctm, } /* gamut_remap_matrix[i] = ctm[i - floor(i/4)] */ - val = ctm->matrix[i - (i / 4)]; - /* If negative, convert to 2's complement. */ - if (val & (1ULL << 63)) - val = -(val & ~(1ULL << 63)); - - matrix[i].value = val; + matrix[i] = dc_fixpt_from_s3132(ctm->matrix[i - (i / 4)]); } } diff --git a/drivers/gpu/drm/amd/display/include/fixed31_32.h b/drivers/gpu/drm/amd/display/include/fixed31_32.h index d4cf7ead1d87..84da1dd34efd 100644 --- a/drivers/gpu/drm/amd/display/include/fixed31_32.h +++ b/drivers/gpu/drm/amd/display/include/fixed31_32.h @@ -69,6 +69,18 @@ static const struct fixed31_32 dc_fixpt_epsilon = { 1LL }; static const struct fixed31_32 dc_fixpt_half = { 0x80000000LL }; static const struct fixed31_32 dc_fixpt_one = { 0x100000000LL }; +static inline struct fixed31_32 dc_fixpt_from_s3132(__u64 x) +{ + struct fixed31_32 val; + + /* If negative, convert to 2's complement. */ + if (x & (1ULL << 63)) + x = -(x & ~(1ULL << 63)); + + val.value = x; + return val; +} + /* * @brief * Initialization routines From 4bc59ddf57c1f68ea035c4f242108f29d91797fd Mon Sep 17 00:00:00 2001 From: Joshua Ashton Date: Thu, 16 Nov 2023 18:58:02 -0100 Subject: [PATCH 60/92] drm/amd/display: add HDR multiplier support With `dc_fixpt_from_s3132()` translation, we can just use it to set hdr_mult. Reviewed-by: Harry Wentland Signed-off-by: Joshua Ashton Signed-off-by: Melissa Wen Signed-off-by: Alex Deucher --- drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.c | 1 + drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_color.c | 3 +++ 2 files changed, 4 insertions(+) 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 911db58d0e1c..4a1139b3cbc5 100644 --- a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.c +++ b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.c @@ -8268,6 +8268,7 @@ static void amdgpu_dm_commit_planes(struct drm_atomic_state *state, bundle->surface_updates[planes_count].gamma = dc_plane->gamma_correction; bundle->surface_updates[planes_count].in_transfer_func = dc_plane->in_transfer_func; bundle->surface_updates[planes_count].gamut_remap_matrix = &dc_plane->gamut_remap_matrix; + bundle->surface_updates[planes_count].hdr_mult = dc_plane->hdr_mult; } amdgpu_dm_plane_fill_dc_scaling_info(dm->adev, new_plane_state, diff --git a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_color.c b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_color.c index c160dbef5224..35f148524827 100644 --- a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_color.c +++ b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_color.c @@ -927,6 +927,7 @@ int amdgpu_dm_update_plane_color_mgmt(struct dm_crtc_state *crtc, struct drm_plane_state *plane_state, struct dc_plane_state *dc_plane_state) { + struct dm_plane_state *dm_plane_state = to_dm_plane_state(plane_state); bool has_crtc_cm_degamma; int ret; @@ -937,6 +938,8 @@ int amdgpu_dm_update_plane_color_mgmt(struct dm_crtc_state *crtc, /* After, we start to update values according to color props */ has_crtc_cm_degamma = (crtc->cm_has_degamma || crtc->cm_is_degamma_srgb); + dc_plane_state->hdr_mult = dc_fixpt_from_s3132(dm_plane_state->hdr_mult); + ret = __set_dm_plane_degamma(plane_state, dc_plane_state); if (ret == -ENOMEM) return ret; From aba8b76baabde681ab4ff686452005d80d949345 Mon Sep 17 00:00:00 2001 From: Melissa Wen Date: Thu, 16 Nov 2023 18:58:03 -0100 Subject: [PATCH 61/92] drm/amd/display: add plane shaper LUT support Map DC shaper LUT to DM plane color management. Shaper LUT can be used to delinearize and/or normalize the color space for computational efficiency and achiving specific visual styles. If a plane degamma is apply to linearize the color space, a custom shaper 1D LUT can be used just before applying 3D LUT. v2: - use DPP color caps to verify plane 3D LUT support - add debug message if shaper LUT programming fails v4: - remove helper to check 3D LUT color caps (Harry) - update desc of lut3d-setup helper from MPC to DPP v5: - remove color_mgmt_changed check that prevents color updates (Joshua) Reviewed-by: Harry Wentland Signed-off-by: Melissa Wen Signed-off-by: Alex Deucher --- .../gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.c | 1 + .../gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.h | 2 + .../amd/display/amdgpu_dm/amdgpu_dm_color.c | 93 ++++++++++++++++++- 3 files changed, 92 insertions(+), 4 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 4a1139b3cbc5..73b12ca2fe60 100644 --- a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.c +++ b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.c @@ -8269,6 +8269,7 @@ static void amdgpu_dm_commit_planes(struct drm_atomic_state *state, bundle->surface_updates[planes_count].in_transfer_func = dc_plane->in_transfer_func; bundle->surface_updates[planes_count].gamut_remap_matrix = &dc_plane->gamut_remap_matrix; bundle->surface_updates[planes_count].hdr_mult = dc_plane->hdr_mult; + bundle->surface_updates[planes_count].func_shaper = dc_plane->in_shaper_func; } amdgpu_dm_plane_fill_dc_scaling_info(dm->adev, new_plane_state, 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 16af549d0b42..7706d412d206 100644 --- a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.h +++ b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.h @@ -908,6 +908,8 @@ void amdgpu_dm_trigger_timing_sync(struct drm_device *dev); /* 3D LUT max size is 17x17x17 (4913 entries) */ #define MAX_COLOR_3DLUT_SIZE 17 #define MAX_COLOR_3DLUT_BITDEPTH 12 +int amdgpu_dm_verify_lut3d_size(struct amdgpu_device *adev, + struct drm_plane_state *plane_state); /* 1D LUT size */ #define MAX_COLOR_LUT_ENTRIES 4096 /* Legacy gamm LUT users such as X doesn't like large LUT sizes */ diff --git a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_color.c b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_color.c index 35f148524827..daa500b04cb5 100644 --- a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_color.c +++ b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_color.c @@ -623,6 +623,63 @@ amdgpu_tf_to_dc_tf(enum amdgpu_transfer_function tf) } } +static int amdgpu_dm_atomic_shaper_lut(const struct drm_color_lut *shaper_lut, + uint32_t shaper_size, + struct dc_transfer_func *func_shaper) +{ + int ret = 0; + + if (shaper_size) { + /* + * If user shaper LUT is set, we assume a linear color space + * (linearized by degamma 1D LUT or not). + */ + func_shaper->type = TF_TYPE_DISTRIBUTED_POINTS; + func_shaper->tf = TRANSFER_FUNCTION_LINEAR; + + ret = __set_output_tf(func_shaper, shaper_lut, shaper_size, false); + } else { + func_shaper->type = TF_TYPE_BYPASS; + func_shaper->tf = TRANSFER_FUNCTION_LINEAR; + } + + return ret; +} + +/** + * amdgpu_dm_verify_lut3d_size - verifies if 3D LUT is supported and if user + * shaper and 3D LUTs match the hw supported size + * @adev: amdgpu device + * @crtc_state: the DRM CRTC state + * + * Verifies if pre-blending (DPP) 3D LUT is supported by the HW (DCN 2.0 or + * newer) and if the user shaper and 3D LUTs match the supported size. + * + * Returns: + * 0 on success. -EINVAL if lut size are invalid. + */ +int amdgpu_dm_verify_lut3d_size(struct amdgpu_device *adev, + struct drm_plane_state *plane_state) +{ + struct dm_plane_state *dm_plane_state = to_dm_plane_state(plane_state); + const struct drm_color_lut *shaper = NULL; + uint32_t exp_size, size; + bool has_3dlut = adev->dm.dc->caps.color.dpp.hw_3d_lut; + + /* shaper LUT is only available if 3D LUT color caps */ + exp_size = has_3dlut ? MAX_COLOR_LUT_ENTRIES : 0; + shaper = __extract_blob_lut(dm_plane_state->shaper_lut, &size); + + if (shaper && size != exp_size) { + drm_dbg(&adev->ddev, + "Invalid Shaper LUT size. Should be %u but got %u.\n", + exp_size, size); + return -EINVAL; + } + + return 0; +} + /** * amdgpu_dm_verify_lut_sizes - verifies if DRM luts match the hw supported sizes * @crtc_state: the DRM CRTC state @@ -910,6 +967,30 @@ __set_dm_plane_degamma(struct drm_plane_state *plane_state, return 0; } +static int +amdgpu_dm_plane_set_color_properties(struct drm_plane_state *plane_state, + struct dc_plane_state *dc_plane_state) +{ + struct dm_plane_state *dm_plane_state = to_dm_plane_state(plane_state); + const struct drm_color_lut *shaper_lut; + uint32_t shaper_size; + int ret; + + dc_plane_state->hdr_mult = dc_fixpt_from_s3132(dm_plane_state->hdr_mult); + + shaper_lut = __extract_blob_lut(dm_plane_state->shaper_lut, &shaper_size); + shaper_size = shaper_lut != NULL ? shaper_size : 0; + + ret = amdgpu_dm_atomic_shaper_lut(shaper_lut, shaper_size, + dc_plane_state->in_shaper_func); + if (ret) + drm_dbg_kms(plane_state->plane->dev, + "setting plane %d shaper LUT failed.\n", + plane_state->plane->index); + + return ret; +} + /** * amdgpu_dm_update_plane_color_mgmt: Maps DRM color management to DC plane. * @crtc: amdgpu_dm crtc state @@ -927,10 +1008,16 @@ int amdgpu_dm_update_plane_color_mgmt(struct dm_crtc_state *crtc, struct drm_plane_state *plane_state, struct dc_plane_state *dc_plane_state) { - struct dm_plane_state *dm_plane_state = to_dm_plane_state(plane_state); + struct amdgpu_device *adev = drm_to_adev(crtc->base.state->dev); bool has_crtc_cm_degamma; int ret; + ret = amdgpu_dm_verify_lut3d_size(adev, plane_state); + if (ret) { + drm_dbg_driver(&adev->ddev, "amdgpu_dm_verify_lut3d_size() failed\n"); + return ret; + } + /* Initially, we can just bypass the DGM block. */ dc_plane_state->in_transfer_func->type = TF_TYPE_BYPASS; dc_plane_state->in_transfer_func->tf = TRANSFER_FUNCTION_LINEAR; @@ -938,8 +1025,6 @@ int amdgpu_dm_update_plane_color_mgmt(struct dm_crtc_state *crtc, /* After, we start to update values according to color props */ has_crtc_cm_degamma = (crtc->cm_has_degamma || crtc->cm_is_degamma_srgb); - dc_plane_state->hdr_mult = dc_fixpt_from_s3132(dm_plane_state->hdr_mult); - ret = __set_dm_plane_degamma(plane_state, dc_plane_state); if (ret == -ENOMEM) return ret; @@ -972,5 +1057,5 @@ int amdgpu_dm_update_plane_color_mgmt(struct dm_crtc_state *crtc, return ret; } - return 0; + return amdgpu_dm_plane_set_color_properties(plane_state, dc_plane_state); } From 99de686115b00e765a5e9345e10c9d7312e4c7ea Mon Sep 17 00:00:00 2001 From: Melissa Wen Date: Thu, 16 Nov 2023 18:58:04 -0100 Subject: [PATCH 62/92] drm/amd/display: add plane shaper TF support Enable usage of predefined transfer func in addition to shaper 1D LUT. That means we can save some complexity by just setting a predefined curve, instead of programming a custom curve when preparing color space for applying 3D LUT. Reviewed-by: Harry Wentland Signed-off-by: Melissa Wen Signed-off-by: Alex Deucher --- .../drm/amd/display/amdgpu_dm/amdgpu_dm_color.c | 15 +++++++++++---- 1 file changed, 11 insertions(+), 4 deletions(-) diff --git a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_color.c b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_color.c index daa500b04cb5..8239904f9d90 100644 --- a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_color.c +++ b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_color.c @@ -624,20 +624,23 @@ amdgpu_tf_to_dc_tf(enum amdgpu_transfer_function tf) } static int amdgpu_dm_atomic_shaper_lut(const struct drm_color_lut *shaper_lut, + bool has_rom, + enum dc_transfer_func_predefined tf, uint32_t shaper_size, struct dc_transfer_func *func_shaper) { int ret = 0; - if (shaper_size) { + if (shaper_size || tf != TRANSFER_FUNCTION_LINEAR) { /* * If user shaper LUT is set, we assume a linear color space * (linearized by degamma 1D LUT or not). */ func_shaper->type = TF_TYPE_DISTRIBUTED_POINTS; - func_shaper->tf = TRANSFER_FUNCTION_LINEAR; + func_shaper->tf = tf; + func_shaper->sdr_ref_white_level = SDR_WHITE_LEVEL_INIT_VALUE; - ret = __set_output_tf(func_shaper, shaper_lut, shaper_size, false); + ret = __set_output_tf(func_shaper, shaper_lut, shaper_size, has_rom); } else { func_shaper->type = TF_TYPE_BYPASS; func_shaper->tf = TRANSFER_FUNCTION_LINEAR; @@ -972,6 +975,7 @@ amdgpu_dm_plane_set_color_properties(struct drm_plane_state *plane_state, struct dc_plane_state *dc_plane_state) { struct dm_plane_state *dm_plane_state = to_dm_plane_state(plane_state); + enum amdgpu_transfer_function shaper_tf = AMDGPU_TRANSFER_FUNCTION_DEFAULT; const struct drm_color_lut *shaper_lut; uint32_t shaper_size; int ret; @@ -980,8 +984,11 @@ amdgpu_dm_plane_set_color_properties(struct drm_plane_state *plane_state, shaper_lut = __extract_blob_lut(dm_plane_state->shaper_lut, &shaper_size); shaper_size = shaper_lut != NULL ? shaper_size : 0; + shaper_tf = dm_plane_state->shaper_tf; - ret = amdgpu_dm_atomic_shaper_lut(shaper_lut, shaper_size, + ret = amdgpu_dm_atomic_shaper_lut(shaper_lut, false, + amdgpu_tf_to_dc_tf(shaper_tf), + shaper_size, dc_plane_state->in_shaper_func); if (ret) drm_dbg_kms(plane_state->plane->dev, From 69a83fd3f0a86374b2fcfab1c02363495704e652 Mon Sep 17 00:00:00 2001 From: Woody Suwalski Date: Tue, 12 Dec 2023 18:31:34 -0500 Subject: [PATCH 63/92] drm/radeon: Prevent multiple debug error lines on suspend MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit Fix to avoid multiple debug error lines printed on every suspend by Radeon driver's debugfs. radeon_debugfs_init() calls debugfs_create_file() for every ring. This results in printing multiple error lines to the screen and dmesg similar to this: [   92.378726] debugfs: File 'radeon_ring_gfx' in directory '0000:00:01.0' already present! [   92.378732] debugfs: File 'radeon_ring_cp1' in directory '0000:00:01.0' already present! [   92.378734] debugfs: File 'radeon_ring_cp2' in directory '0000:00:01.0' already present! [   92.378737] debugfs: File 'radeon_ring_dma1' in directory '0000:00:01.0' already present! [   92.378739] debugfs: File 'radeon_ring_dma2' in directory '0000:00:01.0' already present! [   92.380775] debugfs: File 'radeon_ring_uvd' in directory '0000:00:01.0' already present! [   92.406620] debugfs: File 'radeon_ring_vce1' in directory '0000:00:01.0' already present! [   92.406624] debugfs: File 'radeon_ring_vce2' in directory '0000:00:01.0' already present! Patch v1: The fix was to run lookup() for the file before trying to (re)create that debug file. Patch v2: Call the radeon_debugfs_init() only once when radeon ring is initialized (as suggested by Christian K. - thanks) Reviewed-by: Christian König Signed-off-by: Woody Suwalski Signed-off-by: Alex Deucher --- drivers/gpu/drm/radeon/radeon_ring.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/drivers/gpu/drm/radeon/radeon_ring.c b/drivers/gpu/drm/radeon/radeon_ring.c index e6534fa9f1fb..38048593bb4a 100644 --- a/drivers/gpu/drm/radeon/radeon_ring.c +++ b/drivers/gpu/drm/radeon/radeon_ring.c @@ -413,6 +413,7 @@ int radeon_ring_init(struct radeon_device *rdev, struct radeon_ring *ring, unsig dev_err(rdev->dev, "(%d) ring map failed\n", r); return r; } + radeon_debugfs_ring_init(rdev, ring); } ring->ptr_mask = (ring->ring_size / 4) - 1; ring->ring_free_dw = ring->ring_size / 4; @@ -421,7 +422,6 @@ int radeon_ring_init(struct radeon_device *rdev, struct radeon_ring *ring, unsig ring->next_rptr_gpu_addr = rdev->wb.gpu_addr + index; ring->next_rptr_cpu_addr = &rdev->wb.wb[index/4]; } - radeon_debugfs_ring_init(rdev, ring); radeon_ring_lockup_update(rdev, ring); return 0; } From 65d2765d6291a49d5cdfc0fd88ba5689ed27dbe2 Mon Sep 17 00:00:00 2001 From: =?UTF-8?q?Christian=20K=C3=B6nig?= Date: Mon, 4 Dec 2023 15:51:50 +0100 Subject: [PATCH 64/92] drm/amdgpu: warn when there are still mappings when a BO is destroyed v2 MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit This can only happen when there is a reference counting bug. v2: fix typo Signed-off-by: Christian König Reviewed-by: Alex Deucher Signed-off-by: Alex Deucher --- drivers/gpu/drm/amd/amdgpu/amdgpu_object.c | 2 ++ 1 file changed, 2 insertions(+) diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_object.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_object.c index d79b4ca1ecfc..5ad03f2afdb4 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_object.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_object.c @@ -1343,6 +1343,8 @@ void amdgpu_bo_release_notify(struct ttm_buffer_object *bo) abo = ttm_to_amdgpu_bo(bo); + WARN_ON(abo->vm_bo); + if (abo->kfd_bo) amdgpu_amdkfd_release_notify(abo); From 486c95af5d76047d5cb50727270b1961dacb9380 Mon Sep 17 00:00:00 2001 From: Melissa Wen Date: Thu, 16 Nov 2023 18:58:05 -0100 Subject: [PATCH 65/92] drm/amd/display: add plane 3D LUT support Wire up DC 3D LUT to DM plane color management (pre-blending). On AMD display HW, 3D LUT comes after a shaper curve and we always have to program a shaper curve to delinearize or normalize the color space before applying a 3D LUT (since we have a reduced number of LUT entries). In this version, the default values of 3D LUT for size and bit_depth are 17x17x17 and 12-bit, but we already provide here a more generic mechanisms to program other supported values (9x9x9 size and 10-bit). v2: - started with plane 3D LUT instead of CRTC 3D LUT support v4: - lut3d_size is the max dimension size instead of # of entries Reviewed-by: Harry Wentland Signed-off-by: Melissa Wen Signed-off-by: Alex Deucher --- .../gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.c | 1 + .../amd/display/amdgpu_dm/amdgpu_dm_color.c | 102 +++++++++++++++++- 2 files changed, 99 insertions(+), 4 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 73b12ca2fe60..131174c17cf0 100644 --- a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.c +++ b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.c @@ -8270,6 +8270,7 @@ static void amdgpu_dm_commit_planes(struct drm_atomic_state *state, bundle->surface_updates[planes_count].gamut_remap_matrix = &dc_plane->gamut_remap_matrix; bundle->surface_updates[planes_count].hdr_mult = dc_plane->hdr_mult; bundle->surface_updates[planes_count].func_shaper = dc_plane->in_shaper_func; + bundle->surface_updates[planes_count].lut3d_func = dc_plane->lut3d_func; } amdgpu_dm_plane_fill_dc_scaling_info(dm->adev, new_plane_state, diff --git a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_color.c b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_color.c index 8239904f9d90..f55e0a69dd32 100644 --- a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_color.c +++ b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_color.c @@ -623,6 +623,86 @@ amdgpu_tf_to_dc_tf(enum amdgpu_transfer_function tf) } } +static void __to_dc_lut3d_color(struct dc_rgb *rgb, + const struct drm_color_lut lut, + int bit_precision) +{ + rgb->red = drm_color_lut_extract(lut.red, bit_precision); + rgb->green = drm_color_lut_extract(lut.green, bit_precision); + rgb->blue = drm_color_lut_extract(lut.blue, bit_precision); +} + +static void __drm_3dlut_to_dc_3dlut(const struct drm_color_lut *lut, + uint32_t lut3d_size, + struct tetrahedral_params *params, + bool use_tetrahedral_9, + int bit_depth) +{ + struct dc_rgb *lut0; + struct dc_rgb *lut1; + struct dc_rgb *lut2; + struct dc_rgb *lut3; + int lut_i, i; + + + if (use_tetrahedral_9) { + lut0 = params->tetrahedral_9.lut0; + lut1 = params->tetrahedral_9.lut1; + lut2 = params->tetrahedral_9.lut2; + lut3 = params->tetrahedral_9.lut3; + } else { + lut0 = params->tetrahedral_17.lut0; + lut1 = params->tetrahedral_17.lut1; + lut2 = params->tetrahedral_17.lut2; + lut3 = params->tetrahedral_17.lut3; + } + + for (lut_i = 0, i = 0; i < lut3d_size - 4; lut_i++, i += 4) { + /* + * We should consider the 3D LUT RGB values are distributed + * along four arrays lut0-3 where the first sizes 1229 and the + * other 1228. The bit depth supported for 3dlut channel is + * 12-bit, but DC also supports 10-bit. + * + * TODO: improve color pipeline API to enable the userspace set + * bit depth and 3D LUT size/stride, as specified by VA-API. + */ + __to_dc_lut3d_color(&lut0[lut_i], lut[i], bit_depth); + __to_dc_lut3d_color(&lut1[lut_i], lut[i + 1], bit_depth); + __to_dc_lut3d_color(&lut2[lut_i], lut[i + 2], bit_depth); + __to_dc_lut3d_color(&lut3[lut_i], lut[i + 3], bit_depth); + } + /* lut0 has 1229 points (lut_size/4 + 1) */ + __to_dc_lut3d_color(&lut0[lut_i], lut[i], bit_depth); +} + +/* amdgpu_dm_atomic_lut3d - set DRM 3D LUT to DC stream + * @drm_lut3d: user 3D LUT + * @drm_lut3d_size: size of 3D LUT + * @lut3d: DC 3D LUT + * + * Map user 3D LUT data to DC 3D LUT and all necessary bits to program it + * on DCN accordingly. + */ +static void amdgpu_dm_atomic_lut3d(const struct drm_color_lut *drm_lut3d, + uint32_t drm_lut3d_size, + struct dc_3dlut *lut) +{ + if (!drm_lut3d_size) { + lut->state.bits.initialized = 0; + } else { + /* Stride and bit depth are not programmable by API yet. + * Therefore, only supports 17x17x17 3D LUT (12-bit). + */ + lut->lut_3d.use_tetrahedral_9 = false; + lut->lut_3d.use_12bits = true; + lut->state.bits.initialized = 1; + __drm_3dlut_to_dc_3dlut(drm_lut3d, drm_lut3d_size, &lut->lut_3d, + lut->lut_3d.use_tetrahedral_9, + MAX_COLOR_3DLUT_BITDEPTH); + } +} + static int amdgpu_dm_atomic_shaper_lut(const struct drm_color_lut *shaper_lut, bool has_rom, enum dc_transfer_func_predefined tf, @@ -665,8 +745,8 @@ int amdgpu_dm_verify_lut3d_size(struct amdgpu_device *adev, struct drm_plane_state *plane_state) { struct dm_plane_state *dm_plane_state = to_dm_plane_state(plane_state); - const struct drm_color_lut *shaper = NULL; - uint32_t exp_size, size; + const struct drm_color_lut *shaper = NULL, *lut3d = NULL; + uint32_t exp_size, size, dim_size = MAX_COLOR_3DLUT_SIZE; bool has_3dlut = adev->dm.dc->caps.color.dpp.hw_3d_lut; /* shaper LUT is only available if 3D LUT color caps */ @@ -680,6 +760,17 @@ int amdgpu_dm_verify_lut3d_size(struct amdgpu_device *adev, return -EINVAL; } + /* The number of 3D LUT entries is the dimension size cubed */ + exp_size = has_3dlut ? dim_size * dim_size * dim_size : 0; + lut3d = __extract_blob_lut(dm_plane_state->lut3d, &size); + + if (lut3d && size != exp_size) { + drm_dbg(&adev->ddev, + "Invalid 3D LUT size. Should be %u but got %u.\n", + exp_size, size); + return -EINVAL; + } + return 0; } @@ -976,8 +1067,8 @@ amdgpu_dm_plane_set_color_properties(struct drm_plane_state *plane_state, { struct dm_plane_state *dm_plane_state = to_dm_plane_state(plane_state); enum amdgpu_transfer_function shaper_tf = AMDGPU_TRANSFER_FUNCTION_DEFAULT; - const struct drm_color_lut *shaper_lut; - uint32_t shaper_size; + const struct drm_color_lut *shaper_lut, *lut3d; + uint32_t shaper_size, lut3d_size; int ret; dc_plane_state->hdr_mult = dc_fixpt_from_s3132(dm_plane_state->hdr_mult); @@ -985,7 +1076,10 @@ amdgpu_dm_plane_set_color_properties(struct drm_plane_state *plane_state, shaper_lut = __extract_blob_lut(dm_plane_state->shaper_lut, &shaper_size); shaper_size = shaper_lut != NULL ? shaper_size : 0; shaper_tf = dm_plane_state->shaper_tf; + lut3d = __extract_blob_lut(dm_plane_state->lut3d, &lut3d_size); + lut3d_size = lut3d != NULL ? lut3d_size : 0; + amdgpu_dm_atomic_lut3d(lut3d, lut3d_size, dc_plane_state->lut3d_func); ret = amdgpu_dm_atomic_shaper_lut(shaper_lut, false, amdgpu_tf_to_dc_tf(shaper_tf), shaper_size, From 8d26795ae61a5f64ba7db4f3240dc9ab2138d361 Mon Sep 17 00:00:00 2001 From: Joshua Ashton Date: Thu, 16 Nov 2023 18:58:06 -0100 Subject: [PATCH 66/92] drm/amd/display: handle empty LUTs in __set_input_tf Unlike degamma, blend gamma doesn't support hardcoded curve (predefined/ROM), but we can use AMD color module to fill blend gamma parameters when we have non-linear plane gamma TF without plane gamma LUT. The regular degamma path doesn't hit this. Reviewed-by: Harry Wentland Signed-off-by: Joshua Ashton Signed-off-by: Melissa Wen Signed-off-by: Alex Deucher --- .../amd/display/amdgpu_dm/amdgpu_dm_color.c | 20 +++++++++++-------- 1 file changed, 12 insertions(+), 8 deletions(-) diff --git a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_color.c b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_color.c index f55e0a69dd32..738399538432 100644 --- a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_color.c +++ b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_color.c @@ -578,17 +578,21 @@ static int __set_input_tf(struct dc_transfer_func *func, struct dc_gamma *gamma = NULL; bool res; - gamma = dc_create_gamma(); - if (!gamma) - return -ENOMEM; + if (lut_size) { + gamma = dc_create_gamma(); + if (!gamma) + return -ENOMEM; - gamma->type = GAMMA_CUSTOM; - gamma->num_entries = lut_size; + gamma->type = GAMMA_CUSTOM; + gamma->num_entries = lut_size; - __drm_lut_to_dc_gamma(lut, gamma, false); + __drm_lut_to_dc_gamma(lut, gamma, false); + } - res = mod_color_calculate_degamma_params(NULL, func, gamma, true); - dc_gamma_release(&gamma); + res = mod_color_calculate_degamma_params(NULL, func, gamma, gamma != NULL); + + if (gamma) + dc_gamma_release(&gamma); return res ? 0 : -ENOMEM; } From 783ed4460fe55b01ff32a7c6ad8239974874a16a Mon Sep 17 00:00:00 2001 From: Joshua Ashton Date: Thu, 16 Nov 2023 18:58:07 -0100 Subject: [PATCH 67/92] drm/amd/display: add plane blend LUT and TF support Map plane blend properties to DPP blend gamma. Plane blend is a post-3D LUT curve that linearizes color space for blending. It may be defined by a user-blob LUT and/or predefined transfer function. As hardcoded curve (ROM) is not supported on blend gamma, we use AMD color module to fill parameters when setting non-linear TF with empty LUT. v2: - rename DRM TFs to AMDGPU TFs Reviewed-by: Harry Wentland Signed-off-by: Joshua Ashton Signed-off-by: Melissa Wen Signed-off-by: Alex Deucher --- .../gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.c | 1 + .../amd/display/amdgpu_dm/amdgpu_dm_color.c | 56 +++++++++++++++++-- 2 files changed, 53 insertions(+), 4 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 131174c17cf0..7fd85ec7f811 100644 --- a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.c +++ b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.c @@ -8271,6 +8271,7 @@ static void amdgpu_dm_commit_planes(struct drm_atomic_state *state, bundle->surface_updates[planes_count].hdr_mult = dc_plane->hdr_mult; bundle->surface_updates[planes_count].func_shaper = dc_plane->in_shaper_func; bundle->surface_updates[planes_count].lut3d_func = dc_plane->lut3d_func; + bundle->surface_updates[planes_count].blend_tf = dc_plane->blend_tf; } amdgpu_dm_plane_fill_dc_scaling_info(dm->adev, new_plane_state, diff --git a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_color.c b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_color.c index 738399538432..27ddc6576719 100644 --- a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_color.c +++ b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_color.c @@ -733,6 +733,35 @@ static int amdgpu_dm_atomic_shaper_lut(const struct drm_color_lut *shaper_lut, return ret; } +static int amdgpu_dm_atomic_blend_lut(const struct drm_color_lut *blend_lut, + bool has_rom, + enum dc_transfer_func_predefined tf, + uint32_t blend_size, + struct dc_transfer_func *func_blend) +{ + int ret = 0; + + if (blend_size || tf != TRANSFER_FUNCTION_LINEAR) { + /* + * DRM plane gamma LUT or TF means we are linearizing color + * space before blending (similar to degamma programming). As + * we don't have hardcoded curve support, or we use AMD color + * module to fill the parameters that will be translated to HW + * points. + */ + func_blend->type = TF_TYPE_DISTRIBUTED_POINTS; + func_blend->tf = tf; + func_blend->sdr_ref_white_level = SDR_WHITE_LEVEL_INIT_VALUE; + + ret = __set_input_tf(func_blend, blend_lut, blend_size); + } else { + func_blend->type = TF_TYPE_BYPASS; + func_blend->tf = TRANSFER_FUNCTION_LINEAR; + } + + return ret; +} + /** * amdgpu_dm_verify_lut3d_size - verifies if 3D LUT is supported and if user * shaper and 3D LUTs match the hw supported size @@ -1071,8 +1100,9 @@ amdgpu_dm_plane_set_color_properties(struct drm_plane_state *plane_state, { struct dm_plane_state *dm_plane_state = to_dm_plane_state(plane_state); enum amdgpu_transfer_function shaper_tf = AMDGPU_TRANSFER_FUNCTION_DEFAULT; - const struct drm_color_lut *shaper_lut, *lut3d; - uint32_t shaper_size, lut3d_size; + enum amdgpu_transfer_function blend_tf = AMDGPU_TRANSFER_FUNCTION_DEFAULT; + const struct drm_color_lut *shaper_lut, *lut3d, *blend_lut; + uint32_t shaper_size, lut3d_size, blend_size; int ret; dc_plane_state->hdr_mult = dc_fixpt_from_s3132(dm_plane_state->hdr_mult); @@ -1088,12 +1118,30 @@ amdgpu_dm_plane_set_color_properties(struct drm_plane_state *plane_state, amdgpu_tf_to_dc_tf(shaper_tf), shaper_size, dc_plane_state->in_shaper_func); - if (ret) + if (ret) { drm_dbg_kms(plane_state->plane->dev, "setting plane %d shaper LUT failed.\n", plane_state->plane->index); - return ret; + return ret; + } + + blend_tf = dm_plane_state->blend_tf; + blend_lut = __extract_blob_lut(dm_plane_state->blend_lut, &blend_size); + blend_size = blend_lut != NULL ? blend_size : 0; + + ret = amdgpu_dm_atomic_blend_lut(blend_lut, false, + amdgpu_tf_to_dc_tf(blend_tf), + blend_size, dc_plane_state->blend_tf); + if (ret) { + drm_dbg_kms(plane_state->plane->dev, + "setting plane %d gamma lut failed.\n", + plane_state->plane->index); + + return ret; + } + + return 0; } /** From f81996637000a050477d597ef99e832079f99bd2 Mon Sep 17 00:00:00 2001 From: Joshua Ashton Date: Thu, 16 Nov 2023 18:58:08 -0100 Subject: [PATCH 68/92] drm/amd/display: allow newer DC hardware to use degamma ROM for PQ/HLG Need to funnel the color caps through to these functions so it can check that the hardware is capable. v2: - remove redundant color caps assignment on plane degamma map (Harry) - pass color caps to degamma params v3: - remove unused color_caps parameter from set_color_properties (Harry) Reviewed-by: Harry Wentland Signed-off-by: Joshua Ashton Signed-off-by: Melissa Wen Signed-off-by: Alex Deucher --- .../amd/display/amdgpu_dm/amdgpu_dm_color.c | 29 ++++++++++++------- 1 file changed, 18 insertions(+), 11 deletions(-) diff --git a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_color.c b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_color.c index 27ddc6576719..3a1ca13eaee4 100644 --- a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_color.c +++ b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_color.c @@ -565,6 +565,7 @@ static int amdgpu_dm_set_atomic_regamma(struct dc_stream_state *stream, /** * __set_input_tf - calculates the input transfer function based on expected * input space. + * @caps: dc color capabilities * @func: transfer function * @lut: lookup table that defines the color space * @lut_size: size of respective lut. @@ -572,7 +573,7 @@ static int amdgpu_dm_set_atomic_regamma(struct dc_stream_state *stream, * Returns: * 0 in case of success. -ENOMEM if fails. */ -static int __set_input_tf(struct dc_transfer_func *func, +static int __set_input_tf(struct dc_color_caps *caps, struct dc_transfer_func *func, const struct drm_color_lut *lut, uint32_t lut_size) { struct dc_gamma *gamma = NULL; @@ -589,7 +590,7 @@ static int __set_input_tf(struct dc_transfer_func *func, __drm_lut_to_dc_gamma(lut, gamma, false); } - res = mod_color_calculate_degamma_params(NULL, func, gamma, gamma != NULL); + res = mod_color_calculate_degamma_params(caps, func, gamma, gamma != NULL); if (gamma) dc_gamma_release(&gamma); @@ -753,7 +754,7 @@ static int amdgpu_dm_atomic_blend_lut(const struct drm_color_lut *blend_lut, func_blend->tf = tf; func_blend->sdr_ref_white_level = SDR_WHITE_LEVEL_INIT_VALUE; - ret = __set_input_tf(func_blend, blend_lut, blend_size); + ret = __set_input_tf(NULL, func_blend, blend_lut, blend_size); } else { func_blend->type = TF_TYPE_BYPASS; func_blend->tf = TRANSFER_FUNCTION_LINEAR; @@ -969,7 +970,8 @@ int amdgpu_dm_update_crtc_color_mgmt(struct dm_crtc_state *crtc) static int map_crtc_degamma_to_dc_plane(struct dm_crtc_state *crtc, - struct dc_plane_state *dc_plane_state) + struct dc_plane_state *dc_plane_state, + struct dc_color_caps *caps) { const struct drm_color_lut *degamma_lut; enum dc_transfer_func_predefined tf = TRANSFER_FUNCTION_SRGB; @@ -1024,7 +1026,7 @@ map_crtc_degamma_to_dc_plane(struct dm_crtc_state *crtc, dc_plane_state->in_transfer_func->tf = TRANSFER_FUNCTION_LINEAR; - r = __set_input_tf(dc_plane_state->in_transfer_func, + r = __set_input_tf(caps, dc_plane_state->in_transfer_func, degamma_lut, degamma_size); if (r) return r; @@ -1037,7 +1039,7 @@ map_crtc_degamma_to_dc_plane(struct dm_crtc_state *crtc, dc_plane_state->in_transfer_func->tf = tf; if (tf != TRANSFER_FUNCTION_SRGB && - !mod_color_calculate_degamma_params(NULL, + !mod_color_calculate_degamma_params(caps, dc_plane_state->in_transfer_func, NULL, false)) return -ENOMEM; @@ -1048,7 +1050,8 @@ map_crtc_degamma_to_dc_plane(struct dm_crtc_state *crtc, static int __set_dm_plane_degamma(struct drm_plane_state *plane_state, - struct dc_plane_state *dc_plane_state) + struct dc_plane_state *dc_plane_state, + struct dc_color_caps *color_caps) { struct dm_plane_state *dm_plane_state = to_dm_plane_state(plane_state); const struct drm_color_lut *degamma_lut; @@ -1079,7 +1082,7 @@ __set_dm_plane_degamma(struct drm_plane_state *plane_state, dc_plane_state->in_transfer_func->type = TF_TYPE_DISTRIBUTED_POINTS; - ret = __set_input_tf(dc_plane_state->in_transfer_func, + ret = __set_input_tf(color_caps, dc_plane_state->in_transfer_func, degamma_lut, degamma_size); if (ret) return ret; @@ -1087,7 +1090,7 @@ __set_dm_plane_degamma(struct drm_plane_state *plane_state, dc_plane_state->in_transfer_func->type = TF_TYPE_PREDEFINED; - if (!mod_color_calculate_degamma_params(NULL, + if (!mod_color_calculate_degamma_params(color_caps, dc_plane_state->in_transfer_func, NULL, false)) return -ENOMEM; } @@ -1162,6 +1165,7 @@ int amdgpu_dm_update_plane_color_mgmt(struct dm_crtc_state *crtc, struct dc_plane_state *dc_plane_state) { struct amdgpu_device *adev = drm_to_adev(crtc->base.state->dev); + struct dc_color_caps *color_caps = NULL; bool has_crtc_cm_degamma; int ret; @@ -1171,6 +1175,9 @@ int amdgpu_dm_update_plane_color_mgmt(struct dm_crtc_state *crtc, return ret; } + if (dc_plane_state->ctx && dc_plane_state->ctx->dc) + color_caps = &dc_plane_state->ctx->dc->caps.color; + /* Initially, we can just bypass the DGM block. */ dc_plane_state->in_transfer_func->type = TF_TYPE_BYPASS; dc_plane_state->in_transfer_func->tf = TRANSFER_FUNCTION_LINEAR; @@ -1178,7 +1185,7 @@ int amdgpu_dm_update_plane_color_mgmt(struct dm_crtc_state *crtc, /* After, we start to update values according to color props */ has_crtc_cm_degamma = (crtc->cm_has_degamma || crtc->cm_is_degamma_srgb); - ret = __set_dm_plane_degamma(plane_state, dc_plane_state); + ret = __set_dm_plane_degamma(plane_state, dc_plane_state, color_caps); if (ret == -ENOMEM) return ret; @@ -1205,7 +1212,7 @@ int amdgpu_dm_update_plane_color_mgmt(struct dm_crtc_state *crtc, * linearize (implicit degamma) from sRGB/BT709 according to * the input space. */ - ret = map_crtc_degamma_to_dc_plane(crtc, dc_plane_state); + ret = map_crtc_degamma_to_dc_plane(crtc, dc_plane_state, color_caps); if (ret) return ret; } From 94aeb4117343d072e3a35b9595bcbfc0058ee724 Mon Sep 17 00:00:00 2001 From: "Wang, Beyond" Date: Tue, 12 Dec 2023 21:03:04 +0800 Subject: [PATCH 69/92] drm/amdgpu: fix ftrace event amdgpu_bo_move always move on same heap MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit Issue: during evict or validate happened on amdgpu_bo, the 'from' and 'to' is always same in ftrace event of amdgpu_bo_move where calling the 'trace_amdgpu_bo_move', the comment says move_notify is called before move happens, but actually it is called after move happens, here the new_mem is same as bo->resource Fix: move trace_amdgpu_bo_move from move_notify to amdgpu_bo_move Signed-off-by: Wang, Beyond Reviewed-by: Christian König Signed-off-by: Alex Deucher --- drivers/gpu/drm/amd/amdgpu/amdgpu_object.c | 13 +------------ drivers/gpu/drm/amd/amdgpu/amdgpu_object.h | 4 +--- drivers/gpu/drm/amd/amdgpu/amdgpu_ttm.c | 5 +++-- 3 files changed, 5 insertions(+), 17 deletions(-) diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_object.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_object.c index 5ad03f2afdb4..425cebcc5cbf 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_object.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_object.c @@ -1245,19 +1245,15 @@ int amdgpu_bo_get_metadata(struct amdgpu_bo *bo, void *buffer, * amdgpu_bo_move_notify - notification about a memory move * @bo: pointer to a buffer object * @evict: if this move is evicting the buffer from the graphics address space - * @new_mem: new information of the bufer object * * Marks the corresponding &amdgpu_bo buffer object as invalid, also performs * bookkeeping. * TTM driver callback which is called when ttm moves a buffer. */ -void amdgpu_bo_move_notify(struct ttm_buffer_object *bo, - bool evict, - struct ttm_resource *new_mem) +void amdgpu_bo_move_notify(struct ttm_buffer_object *bo, bool evict) { struct amdgpu_device *adev = amdgpu_ttm_adev(bo->bdev); struct amdgpu_bo *abo; - struct ttm_resource *old_mem = bo->resource; if (!amdgpu_bo_is_amdgpu_bo(bo)) return; @@ -1274,13 +1270,6 @@ void amdgpu_bo_move_notify(struct ttm_buffer_object *bo, /* remember the eviction */ if (evict) atomic64_inc(&adev->num_evictions); - - /* update statistics */ - if (!new_mem) - return; - - /* move_notify is called before move happens */ - trace_amdgpu_bo_move(abo, new_mem->mem_type, old_mem->mem_type); } void amdgpu_bo_get_memory(struct amdgpu_bo *bo, diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_object.h b/drivers/gpu/drm/amd/amdgpu/amdgpu_object.h index d28e21baef16..a3ea8a82db23 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_object.h +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_object.h @@ -344,9 +344,7 @@ int amdgpu_bo_set_metadata (struct amdgpu_bo *bo, void *metadata, int amdgpu_bo_get_metadata(struct amdgpu_bo *bo, void *buffer, size_t buffer_size, uint32_t *metadata_size, uint64_t *flags); -void amdgpu_bo_move_notify(struct ttm_buffer_object *bo, - bool evict, - struct ttm_resource *new_mem); +void amdgpu_bo_move_notify(struct ttm_buffer_object *bo, bool evict); void amdgpu_bo_release_notify(struct ttm_buffer_object *bo); vm_fault_t amdgpu_bo_fault_reserve_notify(struct ttm_buffer_object *bo); void amdgpu_bo_fence(struct amdgpu_bo *bo, struct dma_fence *fence, diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_ttm.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_ttm.c index ab4a762aed5b..75c9fd2c6c2a 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_ttm.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_ttm.c @@ -545,10 +545,11 @@ static int amdgpu_bo_move(struct ttm_buffer_object *bo, bool evict, return r; } + trace_amdgpu_bo_move(abo, new_mem->mem_type, old_mem->mem_type); out: /* update statistics */ atomic64_add(bo->base.size, &adev->num_bytes_moved); - amdgpu_bo_move_notify(bo, evict, new_mem); + amdgpu_bo_move_notify(bo, evict); return 0; } @@ -1553,7 +1554,7 @@ static int amdgpu_ttm_access_memory(struct ttm_buffer_object *bo, static void amdgpu_bo_delete_mem_notify(struct ttm_buffer_object *bo) { - amdgpu_bo_move_notify(bo, false, NULL); + amdgpu_bo_move_notify(bo, false); } static struct ttm_device_funcs amdgpu_bo_driver = { From a2f2f43f74cd050146cd2660bbc3c7e1e7c0da0b Mon Sep 17 00:00:00 2001 From: Peyton Lee Date: Tue, 12 Dec 2023 10:12:33 +0800 Subject: [PATCH 70/92] drm/amd/pm: support return vpe clock table pm supports return vpe clock table and soc clock table Signed-off-by: Peyton Lee Reviewed-by: Li Ma Signed-off-by: Alex Deucher --- drivers/gpu/drm/amd/display/dc/dm_pp_smu.h | 2 ++ drivers/gpu/drm/amd/pm/amdgpu_dpm.c | 10 ++++++++++ drivers/gpu/drm/amd/pm/inc/amdgpu_dpm.h | 1 + .../drm/amd/pm/swsmu/smu14/smu_v14_0_0_ppt.c | 20 +++++++++++++++++++ 4 files changed, 33 insertions(+) diff --git a/drivers/gpu/drm/amd/display/dc/dm_pp_smu.h b/drivers/gpu/drm/amd/display/dc/dm_pp_smu.h index 4440d08743aa..bd7ba0a25198 100644 --- a/drivers/gpu/drm/amd/display/dc/dm_pp_smu.h +++ b/drivers/gpu/drm/amd/display/dc/dm_pp_smu.h @@ -247,6 +247,7 @@ struct pp_smu_funcs_nv { #define PP_SMU_NUM_MEMCLK_DPM_LEVELS 4 #define PP_SMU_NUM_DCLK_DPM_LEVELS 8 #define PP_SMU_NUM_VCLK_DPM_LEVELS 8 +#define PP_SMU_NUM_VPECLK_DPM_LEVELS 8 struct dpm_clock { uint32_t Freq; // In MHz @@ -262,6 +263,7 @@ struct dpm_clocks { struct dpm_clock MemClocks[PP_SMU_NUM_MEMCLK_DPM_LEVELS]; struct dpm_clock VClocks[PP_SMU_NUM_VCLK_DPM_LEVELS]; struct dpm_clock DClocks[PP_SMU_NUM_DCLK_DPM_LEVELS]; + struct dpm_clock VPEClocks[PP_SMU_NUM_VPECLK_DPM_LEVELS]; }; diff --git a/drivers/gpu/drm/amd/pm/amdgpu_dpm.c b/drivers/gpu/drm/amd/pm/amdgpu_dpm.c index 97b40c5fa1ff..6627ee07d52d 100644 --- a/drivers/gpu/drm/amd/pm/amdgpu_dpm.c +++ b/drivers/gpu/drm/amd/pm/amdgpu_dpm.c @@ -616,6 +616,16 @@ void amdgpu_dpm_enable_jpeg(struct amdgpu_device *adev, bool enable) enable ? "enable" : "disable", ret); } +void amdgpu_dpm_enable_vpe(struct amdgpu_device *adev, bool enable) +{ + int ret = 0; + + ret = amdgpu_dpm_set_powergating_by_smu(adev, AMD_IP_BLOCK_TYPE_VPE, !enable); + if (ret) + DRM_ERROR("Dpm %s vpe failed, ret = %d.\n", + enable ? "enable" : "disable", ret); +} + int amdgpu_pm_load_smu_firmware(struct amdgpu_device *adev, uint32_t *smu_version) { const struct amd_pm_funcs *pp_funcs = adev->powerplay.pp_funcs; diff --git a/drivers/gpu/drm/amd/pm/inc/amdgpu_dpm.h b/drivers/gpu/drm/amd/pm/inc/amdgpu_dpm.h index d76b0a60db44..3047ffe7f244 100644 --- a/drivers/gpu/drm/amd/pm/inc/amdgpu_dpm.h +++ b/drivers/gpu/drm/amd/pm/inc/amdgpu_dpm.h @@ -445,6 +445,7 @@ void amdgpu_dpm_compute_clocks(struct amdgpu_device *adev); void amdgpu_dpm_enable_uvd(struct amdgpu_device *adev, bool enable); void amdgpu_dpm_enable_vce(struct amdgpu_device *adev, bool enable); void amdgpu_dpm_enable_jpeg(struct amdgpu_device *adev, bool enable); +void amdgpu_dpm_enable_vpe(struct amdgpu_device *adev, bool enable); int amdgpu_pm_load_smu_firmware(struct amdgpu_device *adev, uint32_t *smu_version); int amdgpu_dpm_handle_passthrough_sbr(struct amdgpu_device *adev, bool enable); int amdgpu_dpm_send_hbm_bad_pages_num(struct amdgpu_device *adev, uint32_t size); diff --git a/drivers/gpu/drm/amd/pm/swsmu/smu14/smu_v14_0_0_ppt.c b/drivers/gpu/drm/amd/pm/swsmu/smu14/smu_v14_0_0_ppt.c index 94ccdbfd7090..47fdbae4adfc 100644 --- a/drivers/gpu/drm/amd/pm/swsmu/smu14/smu_v14_0_0_ppt.c +++ b/drivers/gpu/drm/amd/pm/swsmu/smu14/smu_v14_0_0_ppt.c @@ -1085,6 +1085,25 @@ static int smu_v14_0_0_set_umsch_mm_enable(struct smu_context *smu, 0, NULL); } +static int smu_14_0_0_get_dpm_table(struct smu_context *smu, struct dpm_clocks *clock_table) +{ + DpmClocks_t *clk_table = smu->smu_table.clocks_table; + uint8_t idx; + + /* Only the Clock information of SOC and VPE is copied to provide VPE DPM settings for use. */ + for (idx = 0; idx < NUM_SOCCLK_DPM_LEVELS; idx++) { + clock_table->SocClocks[idx].Freq = (idx < clk_table->NumSocClkLevelsEnabled) ? clk_table->SocClocks[idx]:0; + clock_table->SocClocks[idx].Vol = 0; + } + + for (idx = 0; idx < NUM_VPE_DPM_LEVELS; idx++) { + clock_table->VPEClocks[idx].Freq = (idx < clk_table->VpeClkLevelsEnabled) ? clk_table->VPEClocks[idx]:0; + clock_table->VPEClocks[idx].Vol = 0; + } + + return 0; +} + static const struct pptable_funcs smu_v14_0_0_ppt_funcs = { .check_fw_status = smu_v14_0_check_fw_status, .check_fw_version = smu_v14_0_check_fw_version, @@ -1115,6 +1134,7 @@ static const struct pptable_funcs smu_v14_0_0_ppt_funcs = { .set_gfx_power_up_by_imu = smu_v14_0_set_gfx_power_up_by_imu, .dpm_set_vpe_enable = smu_v14_0_0_set_vpe_enable, .dpm_set_umsch_mm_enable = smu_v14_0_0_set_umsch_mm_enable, + .get_dpm_clock_table = smu_14_0_0_get_dpm_table, }; static void smu_v14_0_0_set_smu_mailbox_registers(struct smu_context *smu) From cb19dc4a64598ffbfd4354083f809fae082fa4c3 Mon Sep 17 00:00:00 2001 From: Joshua Ashton Date: Thu, 16 Nov 2023 18:58:09 -0100 Subject: [PATCH 71/92] drm/amd/display: copy 3D LUT settings from crtc state to stream_update When commiting planes, we copy color mgmt resources to the stream state. Do the same for shaper and 3D LUTs. Reviewed-by: Harry Wentland Signed-off-by: Joshua Ashton Co-developed-by: Melissa Wen Signed-off-by: Melissa Wen Signed-off-by: Alex Deucher --- drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.c | 4 ++++ 1 file changed, 4 insertions(+) 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 7fd85ec7f811..25ec9129758b 100644 --- a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.c +++ b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.c @@ -8483,6 +8483,10 @@ static void amdgpu_dm_commit_planes(struct drm_atomic_state *state, &acrtc_state->stream->csc_color_matrix; bundle->stream_update.out_transfer_func = acrtc_state->stream->out_transfer_func; + bundle->stream_update.lut3d_func = + (struct dc_3dlut *) acrtc_state->stream->lut3d_func; + bundle->stream_update.func_shaper = + (struct dc_transfer_func *) acrtc_state->stream->func_shaper; } acrtc_state->stream->abm_level = acrtc_state->abm_level; From b8b92c1bd7788b1f13d547ee2ce8a93baf55b814 Mon Sep 17 00:00:00 2001 From: Melissa Wen Date: Thu, 16 Nov 2023 18:58:10 -0100 Subject: [PATCH 72/92] drm/amd/display: add plane CTM driver-specific property Plane CTM for pre-blending color space conversion. Only enable driver-specific plane CTM property on drivers that support both pre- and post-blending gamut remap matrix, i.e., DCN3+ family. Otherwise it conflits with DRM CRTC CTM property. Reviewed-by: Harry Wentland Signed-off-by: Melissa Wen Signed-off-by: Alex Deucher --- drivers/gpu/drm/amd/amdgpu/amdgpu_mode.h | 2 ++ .../gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.h | 7 +++++++ .../amd/display/amdgpu_dm/amdgpu_dm_color.c | 7 +++++++ .../amd/display/amdgpu_dm/amdgpu_dm_plane.c | 21 +++++++++++++++++++ 4 files changed, 37 insertions(+) diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_mode.h b/drivers/gpu/drm/amd/amdgpu/amdgpu_mode.h index 790c08b8262d..2e4911050cc5 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_mode.h +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_mode.h @@ -364,6 +364,8 @@ struct amdgpu_mode_info { * @plane_hdr_mult_property: */ struct drm_property *plane_hdr_mult_property; + + struct drm_property *plane_ctm_property; /** * @shaper_lut_property: Plane property to set pre-blending shaper LUT * that converts color content before 3D LUT. If 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 7706d412d206..2d5af83d40b5 100644 --- a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.h +++ b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.h @@ -784,6 +784,13 @@ struct dm_plane_state { * TF is needed for any subsequent linear-to-non-linear transforms. */ __u64 hdr_mult; + /** + * @ctm: + * + * Color transformation matrix. The blob (if not NULL) is a &struct + * drm_color_ctm_3x4. + */ + struct drm_property_blob *ctm; /** * @shaper_lut: shaper lookup table blob. The blob (if not NULL) is an * array of &struct drm_color_lut. diff --git a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_color.c b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_color.c index 3a1ca13eaee4..3eed47736b26 100644 --- a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_color.c +++ b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_color.c @@ -240,6 +240,13 @@ amdgpu_dm_create_color_properties(struct amdgpu_device *adev) return -ENOMEM; adev->mode_info.plane_hdr_mult_property = prop; + prop = drm_property_create(adev_to_drm(adev), + DRM_MODE_PROP_BLOB, + "AMD_PLANE_CTM", 0); + if (!prop) + return -ENOMEM; + adev->mode_info.plane_ctm_property = prop; + prop = drm_property_create(adev_to_drm(adev), DRM_MODE_PROP_BLOB, "AMD_PLANE_SHAPER_LUT", 0); diff --git a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_plane.c b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_plane.c index 35a4732483ba..f10c5154d06a 100644 --- a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_plane.c +++ b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_plane.c @@ -1366,6 +1366,9 @@ static struct drm_plane_state *amdgpu_dm_plane_drm_plane_duplicate_state(struct if (old_dm_plane_state->degamma_lut) dm_plane_state->degamma_lut = drm_property_blob_get(old_dm_plane_state->degamma_lut); + if (old_dm_plane_state->ctm) + dm_plane_state->ctm = + drm_property_blob_get(old_dm_plane_state->ctm); if (old_dm_plane_state->shaper_lut) dm_plane_state->shaper_lut = drm_property_blob_get(old_dm_plane_state->shaper_lut); @@ -1450,6 +1453,8 @@ static void amdgpu_dm_plane_drm_plane_destroy_state(struct drm_plane *plane, if (dm_plane_state->degamma_lut) drm_property_blob_put(dm_plane_state->degamma_lut); + if (dm_plane_state->ctm) + drm_property_blob_put(dm_plane_state->ctm); if (dm_plane_state->lut3d) drm_property_blob_put(dm_plane_state->lut3d); if (dm_plane_state->shaper_lut) @@ -1490,6 +1495,11 @@ dm_atomic_plane_attach_color_mgmt_properties(struct amdgpu_display_manager *dm, dm->adev->mode_info.plane_hdr_mult_property, AMDGPU_HDR_MULT_DEFAULT); + /* Only enable plane CTM if both DPP and MPC gamut remap is available. */ + if (dm->dc->caps.color.mpc.gamut_remap) + drm_object_attach_property(&plane->base, + dm->adev->mode_info.plane_ctm_property, 0); + if (dpp_color_caps.hw_3d_lut) { drm_object_attach_property(&plane->base, mode_info.plane_shaper_lut_property, 0); @@ -1547,6 +1557,14 @@ dm_atomic_plane_set_property(struct drm_plane *plane, dm_plane_state->hdr_mult = val; dm_plane_state->base.color_mgmt_changed = 1; } + } else if (property == adev->mode_info.plane_ctm_property) { + ret = drm_property_replace_blob_from_id(plane->dev, + &dm_plane_state->ctm, + val, + sizeof(struct drm_color_ctm), -1, + &replaced); + dm_plane_state->base.color_mgmt_changed |= replaced; + return ret; } else if (property == adev->mode_info.plane_shaper_lut_property) { ret = drm_property_replace_blob_from_id(plane->dev, &dm_plane_state->shaper_lut, @@ -1608,6 +1626,9 @@ dm_atomic_plane_get_property(struct drm_plane *plane, *val = dm_plane_state->degamma_tf; } else if (property == adev->mode_info.plane_hdr_mult_property) { *val = dm_plane_state->hdr_mult; + } else if (property == adev->mode_info.plane_ctm_property) { + *val = (dm_plane_state->ctm) ? + dm_plane_state->ctm->base.id : 0; } else if (property == adev->mode_info.plane_shaper_lut_property) { *val = (dm_plane_state->shaper_lut) ? dm_plane_state->shaper_lut->base.id : 0; From 5f82a0c90ccaf0d1390b5c1b83a83d38bca526da Mon Sep 17 00:00:00 2001 From: Peyton Lee Date: Tue, 12 Dec 2023 11:11:24 +0800 Subject: [PATCH 73/92] drm/amdgpu/vpe: enable vpe dpm enable vpe dpm Signed-off-by: Peyton Lee Reviewed-by: Lang Yu Signed-off-by: Alex Deucher --- drivers/gpu/drm/amd/amdgpu/amdgpu_vpe.c | 249 ++++++++++++++++++++++++ drivers/gpu/drm/amd/amdgpu/amdgpu_vpe.h | 12 ++ drivers/gpu/drm/amd/amdgpu/vpe_v6_1.c | 15 ++ 3 files changed, 276 insertions(+) diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_vpe.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_vpe.c index e81579708e96..6f149b54d4d3 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_vpe.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_vpe.c @@ -26,6 +26,7 @@ #include "amdgpu.h" #include "amdgpu_ucode.h" #include "amdgpu_vpe.h" +#include "amdgpu_smu.h" #include "soc15_common.h" #include "vpe_v6_1.h" @@ -33,8 +34,186 @@ /* VPE CSA resides in the 4th page of CSA */ #define AMDGPU_CSA_VPE_OFFSET (4096 * 3) +/* 1 second timeout */ +#define VPE_IDLE_TIMEOUT msecs_to_jiffies(1000) + +#define VPE_MAX_DPM_LEVEL 4 +#define FIXED1_8_BITS_PER_FRACTIONAL_PART 8 +#define GET_PRATIO_INTEGER_PART(x) ((x) >> FIXED1_8_BITS_PER_FRACTIONAL_PART) + static void vpe_set_ring_funcs(struct amdgpu_device *adev); +static inline uint16_t div16_u16_rem(uint16_t dividend, uint16_t divisor, uint16_t *remainder) +{ + *remainder = dividend % divisor; + return dividend / divisor; +} + +static inline uint16_t complete_integer_division_u16( + uint16_t dividend, + uint16_t divisor, + uint16_t *remainder) +{ + return div16_u16_rem(dividend, divisor, (uint16_t *)remainder); +} + +static uint16_t vpe_u1_8_from_fraction(uint16_t numerator, uint16_t denominator) +{ + bool arg1_negative = numerator < 0; + bool arg2_negative = denominator < 0; + + uint16_t arg1_value = (uint16_t)(arg1_negative ? -numerator : numerator); + uint16_t arg2_value = (uint16_t)(arg2_negative ? -denominator : denominator); + + uint16_t remainder; + + /* determine integer part */ + uint16_t res_value = complete_integer_division_u16( + arg1_value, arg2_value, &remainder); + + if (res_value > 127 /* CHAR_MAX */) + return 0; + + /* determine fractional part */ + { + unsigned int i = FIXED1_8_BITS_PER_FRACTIONAL_PART; + + do { + remainder <<= 1; + + res_value <<= 1; + + if (remainder >= arg2_value) { + res_value |= 1; + remainder -= arg2_value; + } + } while (--i != 0); + } + + /* round up LSB */ + { + uint16_t summand = (remainder << 1) >= arg2_value; + + if ((res_value + summand) > 32767 /* SHRT_MAX */) + return 0; + + res_value += summand; + } + + if (arg1_negative ^ arg2_negative) + res_value = -res_value; + + return res_value; +} + +static uint16_t vpe_internal_get_pratio(uint16_t from_frequency, uint16_t to_frequency) +{ + uint16_t pratio = vpe_u1_8_from_fraction(from_frequency, to_frequency); + + if (GET_PRATIO_INTEGER_PART(pratio) > 1) + pratio = 0; + + return pratio; +} + +/* + * VPE has 4 DPM levels from level 0 (lowerest) to 3 (highest), + * VPE FW will dynamically decide which level should be used according to current loading. + * + * Get VPE and SOC clocks from PM, and select the appropriate four clock values, + * calculate the ratios of adjusting from one clock to another. + * The VPE FW can then request the appropriate frequency from the PMFW. + */ +int amdgpu_vpe_configure_dpm(struct amdgpu_vpe *vpe) +{ + struct amdgpu_device *adev = vpe->ring.adev; + uint32_t dpm_ctl; + + if (adev->pm.dpm_enabled) { + struct dpm_clocks clock_table = { 0 }; + struct dpm_clock *VPEClks; + struct dpm_clock *SOCClks; + uint32_t idx; + uint32_t pratio_vmax_vnorm = 0, pratio_vnorm_vmid = 0, pratio_vmid_vmin = 0; + uint16_t pratio_vmin_freq = 0, pratio_vmid_freq = 0, pratio_vnorm_freq = 0, pratio_vmax_freq = 0; + + dpm_ctl = RREG32(vpe_get_reg_offset(vpe, 0, vpe->regs.dpm_enable)); + dpm_ctl |= 1; /* DPM enablement */ + WREG32(vpe_get_reg_offset(vpe, 0, vpe->regs.dpm_enable), dpm_ctl); + + /* Get VPECLK and SOCCLK */ + if (amdgpu_dpm_get_dpm_clock_table(adev, &clock_table)) { + dev_dbg(adev->dev, "%s: get clock failed!\n", __func__); + goto disable_dpm; + } + + SOCClks = clock_table.SocClocks; + VPEClks = clock_table.VPEClocks; + + /* vpe dpm only cares 4 levels. */ + for (idx = 0; idx < VPE_MAX_DPM_LEVEL; idx++) { + uint32_t soc_dpm_level; + uint32_t min_freq; + + if (idx == 0) + soc_dpm_level = 0; + else + soc_dpm_level = (idx * 2) + 1; + + /* clamp the max level */ + if (soc_dpm_level > PP_SMU_NUM_VPECLK_DPM_LEVELS - 1) + soc_dpm_level = PP_SMU_NUM_VPECLK_DPM_LEVELS - 1; + + min_freq = (SOCClks[soc_dpm_level].Freq < VPEClks[soc_dpm_level].Freq) ? + SOCClks[soc_dpm_level].Freq : VPEClks[soc_dpm_level].Freq; + + switch (idx) { + case 0: + pratio_vmin_freq = min_freq; + break; + case 1: + pratio_vmid_freq = min_freq; + break; + case 2: + pratio_vnorm_freq = min_freq; + break; + case 3: + pratio_vmax_freq = min_freq; + break; + default: + break; + } + } + + if (pratio_vmin_freq && pratio_vmid_freq && pratio_vnorm_freq && pratio_vmax_freq) { + uint32_t pratio_ctl; + + pratio_vmax_vnorm = (uint32_t)vpe_internal_get_pratio(pratio_vmax_freq, pratio_vnorm_freq); + pratio_vnorm_vmid = (uint32_t)vpe_internal_get_pratio(pratio_vnorm_freq, pratio_vmid_freq); + pratio_vmid_vmin = (uint32_t)vpe_internal_get_pratio(pratio_vmid_freq, pratio_vmin_freq); + + pratio_ctl = pratio_vmax_vnorm | (pratio_vnorm_vmid << 9) | (pratio_vmid_vmin << 18); + WREG32(vpe_get_reg_offset(vpe, 0, vpe->regs.dpm_pratio), pratio_ctl); /* PRatio */ + WREG32(vpe_get_reg_offset(vpe, 0, vpe->regs.dpm_request_interval), 24000); /* 1ms, unit=1/24MHz */ + WREG32(vpe_get_reg_offset(vpe, 0, vpe->regs.dpm_decision_threshold), 1200000); /* 50ms */ + WREG32(vpe_get_reg_offset(vpe, 0, vpe->regs.dpm_busy_clamp_threshold), 1200000);/* 50ms */ + WREG32(vpe_get_reg_offset(vpe, 0, vpe->regs.dpm_idle_clamp_threshold), 1200000);/* 50ms */ + dev_dbg(adev->dev, "%s: configure vpe dpm pratio done!\n", __func__); + } else { + dev_dbg(adev->dev, "%s: invalid pratio parameters!\n", __func__); + goto disable_dpm; + } + } + return 0; + +disable_dpm: + dpm_ctl = RREG32(vpe_get_reg_offset(vpe, 0, vpe->regs.dpm_enable)); + dpm_ctl &= 0xfffffffe; /* Disable DPM */ + WREG32(vpe_get_reg_offset(vpe, 0, vpe->regs.dpm_enable), dpm_ctl); + dev_dbg(adev->dev, "%s: disable vpe dpm\n", __func__); + return 0; +} + int amdgpu_vpe_psp_update_sram(struct amdgpu_device *adev) { struct amdgpu_firmware_info ucode = { @@ -134,6 +313,19 @@ static int vpe_early_init(void *handle) return 0; } +static void vpe_idle_work_handler(struct work_struct *work) +{ + struct amdgpu_device *adev = + container_of(work, struct amdgpu_device, vpe.idle_work.work); + unsigned int fences = 0; + + fences += amdgpu_fence_count_emitted(&adev->vpe.ring); + + if (fences == 0) + amdgpu_device_ip_set_powergating_state(adev, AMD_IP_BLOCK_TYPE_VPE, AMD_PG_STATE_GATE); + else + schedule_delayed_work(&adev->vpe.idle_work, VPE_IDLE_TIMEOUT); +} static int vpe_common_init(struct amdgpu_vpe *vpe) { @@ -150,6 +342,9 @@ static int vpe_common_init(struct amdgpu_vpe *vpe) return r; } + vpe->context_started = false; + INIT_DELAYED_WORK(&adev->vpe.idle_work, vpe_idle_work_handler); + return 0; } @@ -219,6 +414,9 @@ static int vpe_hw_fini(void *handle) vpe_ring_stop(vpe); + /* Power off VPE */ + amdgpu_device_ip_set_powergating_state(adev, AMD_IP_BLOCK_TYPE_VPE, AMD_PG_STATE_GATE); + return 0; } @@ -226,6 +424,8 @@ static int vpe_suspend(void *handle) { struct amdgpu_device *adev = (struct amdgpu_device *)handle; + cancel_delayed_work_sync(&adev->vpe.idle_work); + return vpe_hw_fini(adev); } @@ -430,6 +630,21 @@ static int vpe_set_clockgating_state(void *handle, static int vpe_set_powergating_state(void *handle, enum amd_powergating_state state) { + struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_vpe *vpe = &adev->vpe; + + if (!adev->pm.dpm_enabled) + dev_err(adev->dev, "Without PM, cannot support powergating\n"); + + dev_dbg(adev->dev, "%s: %s!\n", __func__, (state == AMD_PG_STATE_GATE) ? "GATE":"UNGATE"); + + if (state == AMD_PG_STATE_GATE) { + amdgpu_dpm_enable_vpe(adev, false); + vpe->context_started = false; + } else { + amdgpu_dpm_enable_vpe(adev, true); + } + return 0; } @@ -595,6 +810,38 @@ static int vpe_ring_test_ib(struct amdgpu_ring *ring, long timeout) return ret; } +static void vpe_ring_begin_use(struct amdgpu_ring *ring) +{ + struct amdgpu_device *adev = ring->adev; + struct amdgpu_vpe *vpe = &adev->vpe; + + cancel_delayed_work_sync(&adev->vpe.idle_work); + + /* Power on VPE and notify VPE of new context */ + if (!vpe->context_started) { + uint32_t context_notify; + + /* Power on VPE */ + amdgpu_device_ip_set_powergating_state(adev, AMD_IP_BLOCK_TYPE_VPE, AMD_PG_STATE_UNGATE); + + /* Indicates that a job from a new context has been submitted. */ + context_notify = RREG32(vpe_get_reg_offset(vpe, 0, vpe->regs.context_indicator)); + if ((context_notify & 0x1) == 0) + context_notify |= 0x1; + else + context_notify &= ~(0x1); + WREG32(vpe_get_reg_offset(vpe, 0, vpe->regs.context_indicator), context_notify); + vpe->context_started = true; + } +} + +static void vpe_ring_end_use(struct amdgpu_ring *ring) +{ + struct amdgpu_device *adev = ring->adev; + + schedule_delayed_work(&adev->vpe.idle_work, VPE_IDLE_TIMEOUT); +} + static const struct amdgpu_ring_funcs vpe_ring_funcs = { .type = AMDGPU_RING_TYPE_VPE, .align_mask = 0xf, @@ -625,6 +872,8 @@ static const struct amdgpu_ring_funcs vpe_ring_funcs = { .init_cond_exec = vpe_ring_init_cond_exec, .patch_cond_exec = vpe_ring_patch_cond_exec, .preempt_ib = vpe_ring_preempt_ib, + .begin_use = vpe_ring_begin_use, + .end_use = vpe_ring_end_use, }; static void vpe_set_ring_funcs(struct amdgpu_device *adev) diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_vpe.h b/drivers/gpu/drm/amd/amdgpu/amdgpu_vpe.h index 29d56f7ae4a9..1153ddaea64d 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_vpe.h +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_vpe.h @@ -47,6 +47,15 @@ struct vpe_regs { uint32_t queue0_rb_wptr_lo; uint32_t queue0_rb_wptr_hi; uint32_t queue0_preempt; + + uint32_t dpm_enable; + uint32_t dpm_pratio; + uint32_t dpm_request_interval; + uint32_t dpm_decision_threshold; + uint32_t dpm_busy_clamp_threshold; + uint32_t dpm_idle_clamp_threshold; + uint32_t dpm_request_lv; + uint32_t context_indicator; }; struct amdgpu_vpe { @@ -63,12 +72,15 @@ struct amdgpu_vpe { struct amdgpu_bo *cmdbuf_obj; uint64_t cmdbuf_gpu_addr; uint32_t *cmdbuf_cpu_addr; + struct delayed_work idle_work; + bool context_started; }; int amdgpu_vpe_psp_update_sram(struct amdgpu_device *adev); int amdgpu_vpe_init_microcode(struct amdgpu_vpe *vpe); int amdgpu_vpe_ring_init(struct amdgpu_vpe *vpe); int amdgpu_vpe_ring_fini(struct amdgpu_vpe *vpe); +int amdgpu_vpe_configure_dpm(struct amdgpu_vpe *vpe); #define vpe_ring_init(vpe) ((vpe)->funcs->ring_init ? (vpe)->funcs->ring_init((vpe)) : 0) #define vpe_ring_start(vpe) ((vpe)->funcs->ring_start ? (vpe)->funcs->ring_start((vpe)) : 0) diff --git a/drivers/gpu/drm/amd/amdgpu/vpe_v6_1.c b/drivers/gpu/drm/amd/amdgpu/vpe_v6_1.c index 174f13eff575..d20060a51e05 100644 --- a/drivers/gpu/drm/amd/amdgpu/vpe_v6_1.c +++ b/drivers/gpu/drm/amd/amdgpu/vpe_v6_1.c @@ -96,6 +96,10 @@ static int vpe_v6_1_load_microcode(struct amdgpu_vpe *vpe) adev->vpe.cmdbuf_cpu_addr[1] = f32_cntl; amdgpu_vpe_psp_update_sram(adev); + + /* Config DPM */ + amdgpu_vpe_configure_dpm(vpe); + return 0; } @@ -128,6 +132,8 @@ static int vpe_v6_1_load_microcode(struct amdgpu_vpe *vpe) } vpe_v6_1_halt(vpe, false); + /* Config DPM */ + amdgpu_vpe_configure_dpm(vpe); return 0; } @@ -264,6 +270,15 @@ static int vpe_v6_1_set_regs(struct amdgpu_vpe *vpe) vpe->regs.queue0_rb_wptr_hi = regVPEC_QUEUE0_RB_WPTR_HI; vpe->regs.queue0_preempt = regVPEC_QUEUE0_PREEMPT; + vpe->regs.dpm_enable = regVPEC_PUB_DUMMY2; + vpe->regs.dpm_pratio = regVPEC_QUEUE6_DUMMY4; + vpe->regs.dpm_request_interval = regVPEC_QUEUE5_DUMMY3; + vpe->regs.dpm_decision_threshold = regVPEC_QUEUE5_DUMMY4; + vpe->regs.dpm_busy_clamp_threshold = regVPEC_QUEUE7_DUMMY2; + vpe->regs.dpm_idle_clamp_threshold = regVPEC_QUEUE7_DUMMY3; + vpe->regs.dpm_request_lv = regVPEC_QUEUE7_DUMMY1; + vpe->regs.context_indicator = regVPEC_QUEUE6_DUMMY3; + return 0; } From 34dc227bf2f34085313be39d76b12f08bfe8efc0 Mon Sep 17 00:00:00 2001 From: Kenneth Feng Date: Wed, 13 Dec 2023 14:58:34 +0800 Subject: [PATCH 74/92] drm/amd/pm: add power save mode workload for smu 13.0.10 add power save mode workload for smu 13.0.10, so that in compute mode, pmfw will add margin since some applications requres higher margin. Signed-off-by: Kenneth Feng Reviewed-by: Likun Gao Signed-off-by: Alex Deucher --- .../drm/amd/pm/swsmu/smu13/smu_v13_0_0_ppt.c | 23 +++++++++++-------- 1 file changed, 13 insertions(+), 10 deletions(-) diff --git a/drivers/gpu/drm/amd/pm/swsmu/smu13/smu_v13_0_0_ppt.c b/drivers/gpu/drm/amd/pm/swsmu/smu13/smu_v13_0_0_ppt.c index a24aa886c636..231122622a9c 100644 --- a/drivers/gpu/drm/amd/pm/swsmu/smu13/smu_v13_0_0_ppt.c +++ b/drivers/gpu/drm/amd/pm/swsmu/smu13/smu_v13_0_0_ppt.c @@ -2545,16 +2545,19 @@ static int smu_v13_0_0_set_power_profile_mode(struct smu_context *smu, workload_mask = 1 << workload_type; - /* Add optimizations for SMU13.0.0. Reuse the power saving profile */ - if (smu->power_profile_mode == PP_SMC_POWER_PROFILE_COMPUTE && - (amdgpu_ip_version(smu->adev, MP1_HWIP, 0) == IP_VERSION(13, 0, 0)) && - ((smu->adev->pm.fw_version == 0x004e6601) || - (smu->adev->pm.fw_version >= 0x004e7400))) { - workload_type = smu_cmn_to_asic_specific_index(smu, - CMN2ASIC_MAPPING_WORKLOAD, - PP_SMC_POWER_PROFILE_POWERSAVING); - if (workload_type >= 0) - workload_mask |= 1 << workload_type; + /* Add optimizations for SMU13.0.0/10. Reuse the power saving profile */ + if (smu->power_profile_mode == PP_SMC_POWER_PROFILE_COMPUTE) { + if ((amdgpu_ip_version(smu->adev, MP1_HWIP, 0) == IP_VERSION(13, 0, 0) && + ((smu->adev->pm.fw_version == 0x004e6601) || + (smu->adev->pm.fw_version >= 0x004e7300))) || + (amdgpu_ip_version(smu->adev, MP1_HWIP, 0) == IP_VERSION(13, 0, 10) && + smu->adev->pm.fw_version >= 0x00504500)) { + workload_type = smu_cmn_to_asic_specific_index(smu, + CMN2ASIC_MAPPING_WORKLOAD, + PP_SMC_POWER_PROFILE_POWERSAVING); + if (workload_type >= 0) + workload_mask |= 1 << workload_type; + } } return smu_cmn_send_smc_msg_with_param(smu, From 3dad69090743c5f4642aeb628b8542a1e335dded Mon Sep 17 00:00:00 2001 From: Melissa Wen Date: Thu, 16 Nov 2023 18:58:11 -0100 Subject: [PATCH 75/92] drm/amd/display: add plane CTM support Map the plane CTM driver-specific property to DC plane, instead of DC stream. The remaining steps to program DPP block are already implemented on DC shared-code. v3: - fix comment about plane and CRTC CTMs priorities (Harry) Reviewed-by: Harry Wentland Signed-off-by: Melissa Wen Signed-off-by: Alex Deucher --- .../gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.c | 1 + .../amd/display/amdgpu_dm/amdgpu_dm_color.c | 26 +++++++++++++++++++ 2 files changed, 27 insertions(+) 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 25ec9129758b..d7ac020bd8af 100644 --- a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.c +++ b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.c @@ -9983,6 +9983,7 @@ static bool should_reset_plane(struct drm_atomic_state *state, if (dm_old_other_state->degamma_tf != dm_new_other_state->degamma_tf || dm_old_other_state->degamma_lut != dm_new_other_state->degamma_lut || dm_old_other_state->hdr_mult != dm_new_other_state->hdr_mult || + dm_old_other_state->ctm != dm_new_other_state->ctm || dm_old_other_state->shaper_lut != dm_new_other_state->shaper_lut || dm_old_other_state->shaper_tf != dm_new_other_state->shaper_tf || dm_old_other_state->lut3d != dm_new_other_state->lut3d || diff --git a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_color.c b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_color.c index 3eed47736b26..d52c3333ea13 100644 --- a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_color.c +++ b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_color.c @@ -1172,6 +1172,8 @@ int amdgpu_dm_update_plane_color_mgmt(struct dm_crtc_state *crtc, struct dc_plane_state *dc_plane_state) { struct amdgpu_device *adev = drm_to_adev(crtc->base.state->dev); + struct dm_plane_state *dm_plane_state = to_dm_plane_state(plane_state); + struct drm_color_ctm *ctm = NULL; struct dc_color_caps *color_caps = NULL; bool has_crtc_cm_degamma; int ret; @@ -1224,5 +1226,29 @@ int amdgpu_dm_update_plane_color_mgmt(struct dm_crtc_state *crtc, return ret; } + /* Setup CRTC CTM. */ + if (dm_plane_state->ctm) { + ctm = (struct drm_color_ctm *)dm_plane_state->ctm->data; + /* + * DCN2 and older don't support both pre-blending and + * post-blending gamut remap. For this HW family, if we have + * the plane and CRTC CTMs simultaneously, CRTC CTM takes + * priority, and we discard plane CTM, as implemented in + * dcn10_program_gamut_remap(). However, DCN3+ has DPP + * (pre-blending) and MPC (post-blending) `gamut remap` blocks; + * therefore, we can program plane and CRTC CTMs together by + * mapping CRTC CTM to MPC and keeping plane CTM setup at DPP, + * as it's done by dcn30_program_gamut_remap(). + */ + __drm_ctm_to_dc_matrix(ctm, dc_plane_state->gamut_remap_matrix.matrix); + + dc_plane_state->gamut_remap_matrix.enable_remap = true; + dc_plane_state->input_csc_color_matrix.enable_adjustment = false; + } else { + /* Bypass CTM. */ + dc_plane_state->gamut_remap_matrix.enable_remap = false; + dc_plane_state->input_csc_color_matrix.enable_adjustment = false; + } + return amdgpu_dm_plane_set_color_properties(plane_state, dc_plane_state); } From 6872a189be508b9383bc081d462a5d99cbb8319d Mon Sep 17 00:00:00 2001 From: Joshua Ashton Date: Thu, 16 Nov 2023 18:58:12 -0100 Subject: [PATCH 76/92] drm/amd/display: Add 3x4 CTM support for plane CTM Create drm_color_ctm_3x4 to support 3x4-dimension plane CTM matrix and convert DRM CTM to DC CSC float matrix. v3: - rename ctm2 to ctm_3x4 (Harry) Reviewed-by: Harry Wentland Signed-off-by: Joshua Ashton Signed-off-by: Alex Deucher --- .../amd/display/amdgpu_dm/amdgpu_dm_color.c | 28 +++++++++++++++++-- .../amd/display/amdgpu_dm/amdgpu_dm_plane.c | 2 +- include/uapi/drm/drm_mode.h | 8 ++++++ 3 files changed, 34 insertions(+), 4 deletions(-) diff --git a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_color.c b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_color.c index d52c3333ea13..96aecc1a71a3 100644 --- a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_color.c +++ b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_color.c @@ -434,6 +434,28 @@ static void __drm_ctm_to_dc_matrix(const struct drm_color_ctm *ctm, } } +/** + * __drm_ctm_3x4_to_dc_matrix - converts a DRM CTM 3x4 to a DC CSC float matrix + * @ctm: DRM color transformation matrix with 3x4 dimensions + * @matrix: DC CSC float matrix + * + * The matrix needs to be a 3x4 (12 entry) matrix. + */ +static void __drm_ctm_3x4_to_dc_matrix(const struct drm_color_ctm_3x4 *ctm, + struct fixed31_32 *matrix) +{ + int i; + + /* The format provided is S31.32, using signed-magnitude representation. + * Our fixed31_32 is also S31.32, but is using 2's complement. We have + * to convert from signed-magnitude to 2's complement. + */ + for (i = 0; i < 12; i++) { + /* gamut_remap_matrix[i] = ctm[i - floor(i/4)] */ + matrix[i] = dc_fixpt_from_s3132(ctm->matrix[i]); + } +} + /** * __set_legacy_tf - Calculates the legacy transfer function * @func: transfer function @@ -1173,7 +1195,7 @@ int amdgpu_dm_update_plane_color_mgmt(struct dm_crtc_state *crtc, { struct amdgpu_device *adev = drm_to_adev(crtc->base.state->dev); struct dm_plane_state *dm_plane_state = to_dm_plane_state(plane_state); - struct drm_color_ctm *ctm = NULL; + struct drm_color_ctm_3x4 *ctm = NULL; struct dc_color_caps *color_caps = NULL; bool has_crtc_cm_degamma; int ret; @@ -1228,7 +1250,7 @@ int amdgpu_dm_update_plane_color_mgmt(struct dm_crtc_state *crtc, /* Setup CRTC CTM. */ if (dm_plane_state->ctm) { - ctm = (struct drm_color_ctm *)dm_plane_state->ctm->data; + ctm = (struct drm_color_ctm_3x4 *)dm_plane_state->ctm->data; /* * DCN2 and older don't support both pre-blending and * post-blending gamut remap. For this HW family, if we have @@ -1240,7 +1262,7 @@ int amdgpu_dm_update_plane_color_mgmt(struct dm_crtc_state *crtc, * mapping CRTC CTM to MPC and keeping plane CTM setup at DPP, * as it's done by dcn30_program_gamut_remap(). */ - __drm_ctm_to_dc_matrix(ctm, dc_plane_state->gamut_remap_matrix.matrix); + __drm_ctm_3x4_to_dc_matrix(ctm, dc_plane_state->gamut_remap_matrix.matrix); dc_plane_state->gamut_remap_matrix.enable_remap = true; dc_plane_state->input_csc_color_matrix.enable_adjustment = false; diff --git a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_plane.c b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_plane.c index f10c5154d06a..8a4c40b4c27e 100644 --- a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_plane.c +++ b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_plane.c @@ -1561,7 +1561,7 @@ dm_atomic_plane_set_property(struct drm_plane *plane, ret = drm_property_replace_blob_from_id(plane->dev, &dm_plane_state->ctm, val, - sizeof(struct drm_color_ctm), -1, + sizeof(struct drm_color_ctm_3x4), -1, &replaced); dm_plane_state->base.color_mgmt_changed |= replaced; return ret; diff --git a/include/uapi/drm/drm_mode.h b/include/uapi/drm/drm_mode.h index 95630f170110..39d9ac0c0a80 100644 --- a/include/uapi/drm/drm_mode.h +++ b/include/uapi/drm/drm_mode.h @@ -846,6 +846,14 @@ struct drm_color_ctm { __u64 matrix[9]; }; +struct drm_color_ctm_3x4 { + /* + * Conversion matrix with 3x4 dimensions in S31.32 sign-magnitude + * (not two's complement!) format. + */ + __u64 matrix[12]; +}; + struct drm_color_lut { /* * Values are mapped linearly to 0.0 - 1.0 range, with 0x0 == 0.0 and From afe58346d5d3887b3e49ff623d2f2e471f232a8d Mon Sep 17 00:00:00 2001 From: Alex Deucher Date: Mon, 27 Nov 2023 17:26:29 -0500 Subject: [PATCH 77/92] drm/amdgpu/debugfs: fix error code when smc register accessors are NULL MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit Should be -EOPNOTSUPP. Fixes: 5104fdf50d32 ("drm/amdgpu: Fix a null pointer access when the smc_rreg pointer is NULL") Reviewed-by: Christian König Signed-off-by: Alex Deucher --- drivers/gpu/drm/amd/amdgpu/amdgpu_debugfs.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_debugfs.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_debugfs.c index 424bed738296..2cebf2145d9a 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_debugfs.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_debugfs.c @@ -755,7 +755,7 @@ static ssize_t amdgpu_debugfs_regs_smc_read(struct file *f, char __user *buf, int r; if (!adev->smc_rreg) - return -EPERM; + return -EOPNOTSUPP; if (size & 0x3 || *pos & 0x3) return -EINVAL; @@ -814,7 +814,7 @@ static ssize_t amdgpu_debugfs_regs_smc_write(struct file *f, const char __user * int r; if (!adev->smc_wreg) - return -EPERM; + return -EOPNOTSUPP; if (size & 0x3 || *pos & 0x3) return -EINVAL; From 804c49ef30735d70c1df0c58ebec313149a3933c Mon Sep 17 00:00:00 2001 From: Yang Li Date: Thu, 14 Dec 2023 09:01:54 +0800 Subject: [PATCH 78/92] drm/amd/pm: Remove unneeded semicolon ./drivers/gpu/drm/amd/pm/swsmu/amdgpu_smu.c:1418:2-3: Unneeded semicolon Reported-by: Abaci Robot Closes: https://bugzilla.openanolis.cn/show_bug.cgi?id=7743 Signed-off-by: Yang Li Signed-off-by: Alex Deucher --- drivers/gpu/drm/amd/pm/swsmu/amdgpu_smu.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/drivers/gpu/drm/amd/pm/swsmu/amdgpu_smu.c b/drivers/gpu/drm/amd/pm/swsmu/amdgpu_smu.c index d409857fd622..c16703868e5c 100644 --- a/drivers/gpu/drm/amd/pm/swsmu/amdgpu_smu.c +++ b/drivers/gpu/drm/amd/pm/swsmu/amdgpu_smu.c @@ -1415,7 +1415,7 @@ static int smu_wbrf_event_handler(struct notifier_block *nb, break; default: return NOTIFY_DONE; - }; + } return NOTIFY_OK; } From ac16667237a82e2597e329eb9bc520d1cf9dff30 Mon Sep 17 00:00:00 2001 From: Zhipeng Lu Date: Thu, 14 Dec 2023 23:24:11 +0800 Subject: [PATCH 79/92] drm/amd/pm: fix a double-free in si_dpm_init When the allocation of adev->pm.dpm.dyn_state.vddc_dependency_on_dispclk.entries fails, amdgpu_free_extended_power_table is called to free some fields of adev. However, when the control flow returns to si_dpm_sw_init, it goes to label dpm_failed and calls si_dpm_fini, which calls amdgpu_free_extended_power_table again and free those fields again. Thus a double-free is triggered. Fixes: 841686df9f7d ("drm/amdgpu: add SI DPM support (v4)") Signed-off-by: Zhipeng Lu Signed-off-by: Alex Deucher --- drivers/gpu/drm/amd/pm/legacy-dpm/si_dpm.c | 5 ++--- 1 file changed, 2 insertions(+), 3 deletions(-) diff --git a/drivers/gpu/drm/amd/pm/legacy-dpm/si_dpm.c b/drivers/gpu/drm/amd/pm/legacy-dpm/si_dpm.c index fc8e4ac6c8e7..df4f20293c16 100644 --- a/drivers/gpu/drm/amd/pm/legacy-dpm/si_dpm.c +++ b/drivers/gpu/drm/amd/pm/legacy-dpm/si_dpm.c @@ -7379,10 +7379,9 @@ static int si_dpm_init(struct amdgpu_device *adev) kcalloc(4, sizeof(struct amdgpu_clock_voltage_dependency_entry), GFP_KERNEL); - if (!adev->pm.dpm.dyn_state.vddc_dependency_on_dispclk.entries) { - amdgpu_free_extended_power_table(adev); + if (!adev->pm.dpm.dyn_state.vddc_dependency_on_dispclk.entries) return -ENOMEM; - } + adev->pm.dpm.dyn_state.vddc_dependency_on_dispclk.count = 4; adev->pm.dpm.dyn_state.vddc_dependency_on_dispclk.entries[0].clk = 0; adev->pm.dpm.dyn_state.vddc_dependency_on_dispclk.entries[0].v = 0; From 28dd788382c43b330480f57cd34cde0840896743 Mon Sep 17 00:00:00 2001 From: Zhipeng Lu Date: Fri, 15 Dec 2023 00:24:58 +0800 Subject: [PATCH 80/92] drivers/amd/pm: fix a use-after-free in kv_parse_power_table When ps allocated by kzalloc equals to NULL, kv_parse_power_table frees adev->pm.dpm.ps that allocated before. However, after the control flow goes through the following call chains: kv_parse_power_table |-> kv_dpm_init |-> kv_dpm_sw_init |-> kv_dpm_fini The adev->pm.dpm.ps is used in the for loop of kv_dpm_fini after its first free in kv_parse_power_table and causes a use-after-free bug. Fixes: a2e73f56fa62 ("drm/amdgpu: Add support for CIK parts") Signed-off-by: Zhipeng Lu Signed-off-by: Alex Deucher --- drivers/gpu/drm/amd/pm/legacy-dpm/kv_dpm.c | 4 +--- 1 file changed, 1 insertion(+), 3 deletions(-) diff --git a/drivers/gpu/drm/amd/pm/legacy-dpm/kv_dpm.c b/drivers/gpu/drm/amd/pm/legacy-dpm/kv_dpm.c index 5d28c951a319..5cb4725c773f 100644 --- a/drivers/gpu/drm/amd/pm/legacy-dpm/kv_dpm.c +++ b/drivers/gpu/drm/amd/pm/legacy-dpm/kv_dpm.c @@ -2735,10 +2735,8 @@ static int kv_parse_power_table(struct amdgpu_device *adev) non_clock_info = (struct _ATOM_PPLIB_NONCLOCK_INFO *) &non_clock_info_array->nonClockInfo[non_clock_array_index]; ps = kzalloc(sizeof(struct kv_ps), GFP_KERNEL); - if (ps == NULL) { - kfree(adev->pm.dpm.ps); + if (ps == NULL) return -ENOMEM; - } adev->pm.dpm.ps[i].ps_priv = ps; k = 0; idx = (u8 *)&power_state->v2.clockInfoIndex[0]; From c2709b2d6a537ca0fa0f1da36fdaf07e48ef447d Mon Sep 17 00:00:00 2001 From: Zhipeng Lu Date: Fri, 15 Dec 2023 00:58:42 +0800 Subject: [PATCH 81/92] gpu/drm/radeon: fix two memleaks in radeon_vm_init When radeon_bo_create and radeon_vm_clear_bo fail, the vm->page_tables allocated before need to be freed. However, neither radeon_vm_init itself nor its caller have done such deallocation. Fixes: 6d2f2944e95e ("drm/radeon: use normal BOs for the page tables v4") Signed-off-by: Zhipeng Lu Signed-off-by: Alex Deucher --- drivers/gpu/drm/radeon/radeon_vm.c | 8 ++++++-- 1 file changed, 6 insertions(+), 2 deletions(-) diff --git a/drivers/gpu/drm/radeon/radeon_vm.c b/drivers/gpu/drm/radeon/radeon_vm.c index 987cabbf1318..c38b4d5d6a14 100644 --- a/drivers/gpu/drm/radeon/radeon_vm.c +++ b/drivers/gpu/drm/radeon/radeon_vm.c @@ -1204,13 +1204,17 @@ int radeon_vm_init(struct radeon_device *rdev, struct radeon_vm *vm) r = radeon_bo_create(rdev, pd_size, align, true, RADEON_GEM_DOMAIN_VRAM, 0, NULL, NULL, &vm->page_directory); - if (r) + if (r) { + kfree(vm->page_tables); + vm->page_tables = NULL; return r; - + } r = radeon_vm_clear_bo(rdev, vm->page_directory); if (r) { radeon_bo_unref(&vm->page_directory); vm->page_directory = NULL; + kfree(vm->page_tables); + vm->page_tables = NULL; return r; } From a6582701178a47c4d0cb2188c965c59c0c0647c8 Mon Sep 17 00:00:00 2001 From: Zhipeng Lu Date: Fri, 15 Dec 2023 00:59:38 +0800 Subject: [PATCH 82/92] drm/amd/pm: fix a double-free in amdgpu_parse_extended_power_table The amdgpu_free_extended_power_table is called in every error-handling paths of amdgpu_parse_extended_power_table. However, after the following call chain of returning: amdgpu_parse_extended_power_table |-> kv_dpm_init / si_dpm_init (the only two caller of amdgpu_parse_extended_power_table) |-> kv_dpm_sw_init / si_dpm_sw_init (the only caller of kv_dpm_init / si_dpm_init, accordingly) |-> kv_dpm_fini / si_dpm_fini (goto dpm_failed in xx_dpm_sw_init) |-> amdgpu_free_extended_power_table As above, the amdgpu_free_extended_power_table is called twice in this returning chain and thus a double-free is triggered. Similarily, the last kfree in amdgpu_parse_extended_power_table also cause a double free with amdgpu_free_extended_power_table in kv_dpm_fini. Fixes: 84176663e70d ("drm/amd/pm: create a new holder for those APIs used only by legacy ASICs(si/kv)") Signed-off-by: Zhipeng Lu Signed-off-by: Alex Deucher --- .../gpu/drm/amd/pm/legacy-dpm/legacy_dpm.c | 52 +++++-------------- 1 file changed, 13 insertions(+), 39 deletions(-) diff --git a/drivers/gpu/drm/amd/pm/legacy-dpm/legacy_dpm.c b/drivers/gpu/drm/amd/pm/legacy-dpm/legacy_dpm.c index 81fb4e5dd804..60377747bab4 100644 --- a/drivers/gpu/drm/amd/pm/legacy-dpm/legacy_dpm.c +++ b/drivers/gpu/drm/amd/pm/legacy-dpm/legacy_dpm.c @@ -272,10 +272,8 @@ int amdgpu_parse_extended_power_table(struct amdgpu_device *adev) le16_to_cpu(power_info->pplib4.usVddcDependencyOnSCLKOffset)); ret = amdgpu_parse_clk_voltage_dep_table(&adev->pm.dpm.dyn_state.vddc_dependency_on_sclk, dep_table); - if (ret) { - amdgpu_free_extended_power_table(adev); + if (ret) return ret; - } } if (power_info->pplib4.usVddciDependencyOnMCLKOffset) { dep_table = (ATOM_PPLIB_Clock_Voltage_Dependency_Table *) @@ -283,10 +281,8 @@ int amdgpu_parse_extended_power_table(struct amdgpu_device *adev) le16_to_cpu(power_info->pplib4.usVddciDependencyOnMCLKOffset)); ret = amdgpu_parse_clk_voltage_dep_table(&adev->pm.dpm.dyn_state.vddci_dependency_on_mclk, dep_table); - if (ret) { - amdgpu_free_extended_power_table(adev); + if (ret) return ret; - } } if (power_info->pplib4.usVddcDependencyOnMCLKOffset) { dep_table = (ATOM_PPLIB_Clock_Voltage_Dependency_Table *) @@ -294,10 +290,8 @@ int amdgpu_parse_extended_power_table(struct amdgpu_device *adev) le16_to_cpu(power_info->pplib4.usVddcDependencyOnMCLKOffset)); ret = amdgpu_parse_clk_voltage_dep_table(&adev->pm.dpm.dyn_state.vddc_dependency_on_mclk, dep_table); - if (ret) { - amdgpu_free_extended_power_table(adev); + if (ret) return ret; - } } if (power_info->pplib4.usMvddDependencyOnMCLKOffset) { dep_table = (ATOM_PPLIB_Clock_Voltage_Dependency_Table *) @@ -305,10 +299,8 @@ int amdgpu_parse_extended_power_table(struct amdgpu_device *adev) le16_to_cpu(power_info->pplib4.usMvddDependencyOnMCLKOffset)); ret = amdgpu_parse_clk_voltage_dep_table(&adev->pm.dpm.dyn_state.mvdd_dependency_on_mclk, dep_table); - if (ret) { - amdgpu_free_extended_power_table(adev); + if (ret) return ret; - } } if (power_info->pplib4.usMaxClockVoltageOnDCOffset) { ATOM_PPLIB_Clock_Voltage_Limit_Table *clk_v = @@ -339,10 +331,8 @@ int amdgpu_parse_extended_power_table(struct amdgpu_device *adev) kcalloc(psl->ucNumEntries, sizeof(struct amdgpu_phase_shedding_limits_entry), GFP_KERNEL); - if (!adev->pm.dpm.dyn_state.phase_shedding_limits_table.entries) { - amdgpu_free_extended_power_table(adev); + if (!adev->pm.dpm.dyn_state.phase_shedding_limits_table.entries) return -ENOMEM; - } entry = &psl->entries[0]; for (i = 0; i < psl->ucNumEntries; i++) { @@ -383,10 +373,8 @@ int amdgpu_parse_extended_power_table(struct amdgpu_device *adev) ATOM_PPLIB_CAC_Leakage_Record *entry; u32 size = cac_table->ucNumEntries * sizeof(struct amdgpu_cac_leakage_table); adev->pm.dpm.dyn_state.cac_leakage_table.entries = kzalloc(size, GFP_KERNEL); - if (!adev->pm.dpm.dyn_state.cac_leakage_table.entries) { - amdgpu_free_extended_power_table(adev); + if (!adev->pm.dpm.dyn_state.cac_leakage_table.entries) return -ENOMEM; - } entry = &cac_table->entries[0]; for (i = 0; i < cac_table->ucNumEntries; i++) { if (adev->pm.dpm.platform_caps & ATOM_PP_PLATFORM_CAP_EVV) { @@ -438,10 +426,8 @@ int amdgpu_parse_extended_power_table(struct amdgpu_device *adev) sizeof(struct amdgpu_vce_clock_voltage_dependency_entry); adev->pm.dpm.dyn_state.vce_clock_voltage_dependency_table.entries = kzalloc(size, GFP_KERNEL); - if (!adev->pm.dpm.dyn_state.vce_clock_voltage_dependency_table.entries) { - amdgpu_free_extended_power_table(adev); + if (!adev->pm.dpm.dyn_state.vce_clock_voltage_dependency_table.entries) return -ENOMEM; - } adev->pm.dpm.dyn_state.vce_clock_voltage_dependency_table.count = limits->numEntries; entry = &limits->entries[0]; @@ -493,10 +479,8 @@ int amdgpu_parse_extended_power_table(struct amdgpu_device *adev) sizeof(struct amdgpu_uvd_clock_voltage_dependency_entry); adev->pm.dpm.dyn_state.uvd_clock_voltage_dependency_table.entries = kzalloc(size, GFP_KERNEL); - if (!adev->pm.dpm.dyn_state.uvd_clock_voltage_dependency_table.entries) { - amdgpu_free_extended_power_table(adev); + if (!adev->pm.dpm.dyn_state.uvd_clock_voltage_dependency_table.entries) return -ENOMEM; - } adev->pm.dpm.dyn_state.uvd_clock_voltage_dependency_table.count = limits->numEntries; entry = &limits->entries[0]; @@ -525,10 +509,8 @@ int amdgpu_parse_extended_power_table(struct amdgpu_device *adev) sizeof(struct amdgpu_clock_voltage_dependency_entry); adev->pm.dpm.dyn_state.samu_clock_voltage_dependency_table.entries = kzalloc(size, GFP_KERNEL); - if (!adev->pm.dpm.dyn_state.samu_clock_voltage_dependency_table.entries) { - amdgpu_free_extended_power_table(adev); + if (!adev->pm.dpm.dyn_state.samu_clock_voltage_dependency_table.entries) return -ENOMEM; - } adev->pm.dpm.dyn_state.samu_clock_voltage_dependency_table.count = limits->numEntries; entry = &limits->entries[0]; @@ -548,10 +530,8 @@ int amdgpu_parse_extended_power_table(struct amdgpu_device *adev) le16_to_cpu(ext_hdr->usPPMTableOffset)); adev->pm.dpm.dyn_state.ppm_table = kzalloc(sizeof(struct amdgpu_ppm_table), GFP_KERNEL); - if (!adev->pm.dpm.dyn_state.ppm_table) { - amdgpu_free_extended_power_table(adev); + if (!adev->pm.dpm.dyn_state.ppm_table) return -ENOMEM; - } adev->pm.dpm.dyn_state.ppm_table->ppm_design = ppm->ucPpmDesign; adev->pm.dpm.dyn_state.ppm_table->cpu_core_number = le16_to_cpu(ppm->usCpuCoreNumber); @@ -583,10 +563,8 @@ int amdgpu_parse_extended_power_table(struct amdgpu_device *adev) sizeof(struct amdgpu_clock_voltage_dependency_entry); adev->pm.dpm.dyn_state.acp_clock_voltage_dependency_table.entries = kzalloc(size, GFP_KERNEL); - if (!adev->pm.dpm.dyn_state.acp_clock_voltage_dependency_table.entries) { - amdgpu_free_extended_power_table(adev); + if (!adev->pm.dpm.dyn_state.acp_clock_voltage_dependency_table.entries) return -ENOMEM; - } adev->pm.dpm.dyn_state.acp_clock_voltage_dependency_table.count = limits->numEntries; entry = &limits->entries[0]; @@ -606,10 +584,8 @@ int amdgpu_parse_extended_power_table(struct amdgpu_device *adev) ATOM_PowerTune_Table *pt; adev->pm.dpm.dyn_state.cac_tdp_table = kzalloc(sizeof(struct amdgpu_cac_tdp_table), GFP_KERNEL); - if (!adev->pm.dpm.dyn_state.cac_tdp_table) { - amdgpu_free_extended_power_table(adev); + if (!adev->pm.dpm.dyn_state.cac_tdp_table) return -ENOMEM; - } if (rev > 0) { ATOM_PPLIB_POWERTUNE_Table_V1 *ppt = (ATOM_PPLIB_POWERTUNE_Table_V1 *) (mode_info->atom_context->bios + data_offset + @@ -645,10 +621,8 @@ int amdgpu_parse_extended_power_table(struct amdgpu_device *adev) ret = amdgpu_parse_clk_voltage_dep_table( &adev->pm.dpm.dyn_state.vddgfx_dependency_on_sclk, dep_table); - if (ret) { - kfree(adev->pm.dpm.dyn_state.vddgfx_dependency_on_sclk.entries); + if (ret) return ret; - } } } From 8b881b5d6fe9ebb7736097f37103c9b07ea45642 Mon Sep 17 00:00:00 2001 From: Alex Deucher Date: Thu, 14 Dec 2023 10:41:41 -0500 Subject: [PATCH 83/92] drm/amd/display: fix documentation for amdgpu_dm_verify_lut3d_size() It takes the plane state rather than the crtc state. Fixes: aba8b76baabd ("drm/amd/display: add plane shaper LUT support") Reported-by: Stephen Rothwell Reviewed-by: Melissa Wen Signed-off-by: Alex Deucher Cc: Melissa Wen Cc: Harry.Wentland@amd.com --- drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_color.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_color.c b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_color.c index 96aecc1a71a3..c6ed0d854b01 100644 --- a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_color.c +++ b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_color.c @@ -796,7 +796,7 @@ static int amdgpu_dm_atomic_blend_lut(const struct drm_color_lut *blend_lut, * amdgpu_dm_verify_lut3d_size - verifies if 3D LUT is supported and if user * shaper and 3D LUTs match the hw supported size * @adev: amdgpu device - * @crtc_state: the DRM CRTC state + * @plane_state: the DRM plane state * * Verifies if pre-blending (DPP) 3D LUT is supported by the HW (DCN 2.0 or * newer) and if the user shaper and 3D LUTs match the supported size. From c6ef0a2265c518aa6699b64d10a7e5a9049ac96a Mon Sep 17 00:00:00 2001 From: Melissa Wen Date: Thu, 14 Dec 2023 18:45:16 -0100 Subject: [PATCH 84/92] drm/amd/display: fix documentation for dm_crtc_additional_color_mgmt() warning: expecting prototype for drm_crtc_additional_color_mgmt(). Prototype was for dm_crtc_additional_color_mgmt() instead Reported-by: kernel test robot Closes: https://lore.kernel.org/oe-kbuild-all/202312141801.o9eBCxt9-lkp@intel.com/ Signed-off-by: Melissa Wen Signed-off-by: Alex Deucher --- drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_crtc.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) 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 9224e2d7f32b..7545a184e43a 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 @@ -299,7 +299,7 @@ static int amdgpu_dm_crtc_late_register(struct drm_crtc *crtc) #ifdef AMD_PRIVATE_COLOR /** - * drm_crtc_additional_color_mgmt - enable additional color properties + * dm_crtc_additional_color_mgmt - enable additional color properties * @crtc: DRM CRTC * * This function lets the driver enable post-blending CRTC regamma transfer From 24149412dfc71f7f4a54868702e9145e396263d3 Mon Sep 17 00:00:00 2001 From: Jonathan Kim Date: Wed, 13 Dec 2023 22:08:03 -0500 Subject: [PATCH 85/92] drm/amdkfd: only flush mes process context if mes support is there Fix up on mes process context flush to prevent non-mes devices from spamming error messages or running into undefined behaviour during process termination. Fixes: bd33bb1409b4 ("drm/amdkfd: fix mes set shader debugger process management") Signed-off-by: Jonathan Kim Reviewed-by: Eric Huang Signed-off-by: Alex Deucher --- drivers/gpu/drm/amd/amdkfd/kfd_process_queue_manager.c | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-) diff --git a/drivers/gpu/drm/amd/amdkfd/kfd_process_queue_manager.c b/drivers/gpu/drm/amd/amdkfd/kfd_process_queue_manager.c index 8e55e78fce4e..43eff221eae5 100644 --- a/drivers/gpu/drm/amd/amdkfd/kfd_process_queue_manager.c +++ b/drivers/gpu/drm/amd/amdkfd/kfd_process_queue_manager.c @@ -87,7 +87,8 @@ void kfd_process_dequeue_from_device(struct kfd_process_device *pdd) return; dev->dqm->ops.process_termination(dev->dqm, &pdd->qpd); - amdgpu_mes_flush_shader_debugger(dev->adev, pdd->proc_ctx_gpu_addr); + if (dev->kfd->shared_resources.enable_mes) + amdgpu_mes_flush_shader_debugger(dev->adev, pdd->proc_ctx_gpu_addr); pdd->already_dequeued = true; } From 65a618dd73216e111baab144a837f842dbb6a738 Mon Sep 17 00:00:00 2001 From: Philip Yang Date: Thu, 14 Dec 2023 09:42:03 -0500 Subject: [PATCH 86/92] drm/amdkfd: svm range always mapped flag not working on APU On gfx943 APU there is no VRAM and page migration, queue CWSR area, svm range with always mapped flag, is not mapped to GPU correctly. This works fine if retry fault on CWSR area can be recovered, but could cause deadlock if there is another retry fault recover waiting for CWSR to finish. Fix this by mapping svm range with always mapped flag to GPU with ACCESS attribute if XNACK ON. There is side effect, because all GPUs have ACCESS attribute by default on new svm range with XNACK on, the CWSR area will be mapped to all GPUs after this change. This side effect will be fixed with Thunk change to set CWSR svm range with ACCESS_IN_PLACE attribute on the GPU that user queue is created. Signed-off-by: Philip Yang Reviewed-by: Felix Kuehling Signed-off-by: Alex Deucher --- drivers/gpu/drm/amd/amdkfd/kfd_svm.c | 18 ++++++++++++------ 1 file changed, 12 insertions(+), 6 deletions(-) diff --git a/drivers/gpu/drm/amd/amdkfd/kfd_svm.c b/drivers/gpu/drm/amd/amdkfd/kfd_svm.c index 617d20f0380f..fb0011f80277 100644 --- a/drivers/gpu/drm/amd/amdkfd/kfd_svm.c +++ b/drivers/gpu/drm/amd/amdkfd/kfd_svm.c @@ -1607,18 +1607,24 @@ static int svm_range_validate_and_map(struct mm_struct *mm, if (test_bit(gpuidx, prange->bitmap_access)) bitmap_set(ctx->bitmap, gpuidx, 1); } + + /* + * If prange is already mapped or with always mapped flag, + * update mapping on GPUs with ACCESS attribute + */ + if (bitmap_empty(ctx->bitmap, MAX_GPU_INSTANCE)) { + if (prange->mapped_to_gpu || + prange->flags & KFD_IOCTL_SVM_FLAG_GPU_ALWAYS_MAPPED) + bitmap_copy(ctx->bitmap, prange->bitmap_access, MAX_GPU_INSTANCE); + } } else { bitmap_or(ctx->bitmap, prange->bitmap_access, prange->bitmap_aip, MAX_GPU_INSTANCE); } if (bitmap_empty(ctx->bitmap, MAX_GPU_INSTANCE)) { - bitmap_copy(ctx->bitmap, prange->bitmap_access, MAX_GPU_INSTANCE); - if (!prange->mapped_to_gpu || - bitmap_empty(ctx->bitmap, MAX_GPU_INSTANCE)) { - r = 0; - goto free_ctx; - } + r = 0; + goto free_ctx; } if (prange->actual_loc && !prange->ttm_res) { From 78b4dfd35999e22b4f589a3e070c4aa5f07ce3a2 Mon Sep 17 00:00:00 2001 From: James Zhu Date: Fri, 8 Dec 2023 17:41:25 -0500 Subject: [PATCH 87/92] drm/amdgpu: increase hmm range get pages timeout When application tries to allocate all system memory and cause memory to swap out. Needs more time for hmm_range_fault to validate the remaining page for allocation. To be safe, increase timeout value to 1 second for 64MB range. Signed-off-by: James Zhu Acked-by: Felix Kuehling Signed-off-by: Alex Deucher --- drivers/gpu/drm/amd/amdgpu/amdgpu_hmm.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_hmm.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_hmm.c index 081267161d40..b24eb5821fd1 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_hmm.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_hmm.c @@ -190,8 +190,8 @@ int amdgpu_hmm_range_get_pages(struct mmu_interval_notifier *notifier, pr_debug("hmm range: start = 0x%lx, end = 0x%lx", hmm_range->start, hmm_range->end); - /* Assuming 128MB takes maximum 1 second to fault page address */ - timeout = max((hmm_range->end - hmm_range->start) >> 27, 1UL); + /* Assuming 64MB takes maximum 1 second to fault page address */ + timeout = max((hmm_range->end - hmm_range->start) >> 26, 1UL); timeout *= HMM_RANGE_DEFAULT_TIMEOUT; timeout = jiffies + msecs_to_jiffies(timeout); From 0c8c0e7a9eebc2de03d161de4376e0d9158b6817 Mon Sep 17 00:00:00 2001 From: James Zhu Date: Fri, 8 Dec 2023 17:49:54 -0500 Subject: [PATCH 88/92] drm/amdgpu: make an improvement on amdgpu_hmm_range_get_pages Only schedule when hmm_range_fault returns error. Signed-off-by: James Zhu Acked-by: Felix Kuehling Signed-off-by: Alex Deucher --- drivers/gpu/drm/amd/amdgpu/amdgpu_hmm.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_hmm.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_hmm.c index b24eb5821fd1..55b65fc04b65 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_hmm.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_hmm.c @@ -199,6 +199,7 @@ int amdgpu_hmm_range_get_pages(struct mmu_interval_notifier *notifier, hmm_range->notifier_seq = mmu_interval_read_begin(notifier); r = hmm_range_fault(hmm_range); if (unlikely(r)) { + schedule(); /* * FIXME: This timeout should encompass the retry from * mmu_interval_read_retry() as well. @@ -212,7 +213,6 @@ int amdgpu_hmm_range_get_pages(struct mmu_interval_notifier *notifier, break; hmm_range->hmm_pfns += MAX_WALK_BYTE >> PAGE_SHIFT; hmm_range->start = hmm_range->end; - schedule(); } while (hmm_range->end < end); hmm_range->start = start; From 7046ca9c1ba64938f1b498026419d47b0993c69f Mon Sep 17 00:00:00 2001 From: Li Ma Date: Fri, 15 Dec 2023 11:37:20 +0800 Subject: [PATCH 89/92] drm/amd/swsmu: remove duplicate definition of smu v14_0_0 driver if version There is a repeated define of smu v14_0_0 driver if version, so delete one in driver if header. Signed-off-by: Li Ma Reviewed-by: Kenneth Feng Reviewed-by: Yifan Zhang Signed-off-by: Alex Deucher --- .../drm/amd/pm/swsmu/inc/pmfw_if/smu14_driver_if_v14_0_0.h | 5 ----- drivers/gpu/drm/amd/pm/swsmu/inc/smu_v14_0.h | 2 +- 2 files changed, 1 insertion(+), 6 deletions(-) diff --git a/drivers/gpu/drm/amd/pm/swsmu/inc/pmfw_if/smu14_driver_if_v14_0_0.h b/drivers/gpu/drm/amd/pm/swsmu/inc/pmfw_if/smu14_driver_if_v14_0_0.h index 8f42771e1f0a..5bb7a63c0602 100644 --- a/drivers/gpu/drm/amd/pm/swsmu/inc/pmfw_if/smu14_driver_if_v14_0_0.h +++ b/drivers/gpu/drm/amd/pm/swsmu/inc/pmfw_if/smu14_driver_if_v14_0_0.h @@ -24,11 +24,6 @@ #ifndef SMU14_DRIVER_IF_V14_0_0_H #define SMU14_DRIVER_IF_V14_0_0_H -// *** IMPORTANT *** -// SMU TEAM: Always increment the interface version if -// any structure is changed in this file -#define PMFW_DRIVER_IF_VERSION 7 - typedef struct { int32_t value; uint32_t numFractionalBits; diff --git a/drivers/gpu/drm/amd/pm/swsmu/inc/smu_v14_0.h b/drivers/gpu/drm/amd/pm/swsmu/inc/smu_v14_0.h index a5b569976f19..3f7463c1c1a9 100644 --- a/drivers/gpu/drm/amd/pm/swsmu/inc/smu_v14_0.h +++ b/drivers/gpu/drm/amd/pm/swsmu/inc/smu_v14_0.h @@ -26,8 +26,8 @@ #include "amdgpu_smu.h" #define SMU14_DRIVER_IF_VERSION_INV 0xFFFFFFFF +#define SMU14_DRIVER_IF_VERSION_SMU_V14_0_0 0x7 #define SMU14_DRIVER_IF_VERSION_SMU_V14_0_2 0x1 -#define SMU14_DRIVER_IF_VERSION_SMU_V14_0_0 0x6 #define FEATURE_MASK(feature) (1ULL << feature) From 0f657938e4345a77be871d906f3e0de3c58a7a49 Mon Sep 17 00:00:00 2001 From: Samson Tam Date: Tue, 28 Nov 2023 16:53:12 -0500 Subject: [PATCH 90/92] drm/amd/display: do not send commands to DMUB if DMUB is inactive from S3 [Why] On resume from S3, may get apply_idle_optimizations call while DMUB is inactive which will just time out. [How] Set and track power state in dmub_srv and check power state before sending commands to DMUB. Add interface in both dmub_srv and dc_dmub_srv Reviewed-by: Nicholas Kazlauskas Acked-by: Wayne Lin Signed-off-by: Samson Tam Tested-by: Daniel Wheeler Signed-off-by: Alex Deucher --- .../gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.c | 3 +++ drivers/gpu/drm/amd/display/dc/dc_dmub_srv.c | 14 +++++++++++++ drivers/gpu/drm/amd/display/dc/dc_dmub_srv.h | 2 ++ drivers/gpu/drm/amd/display/dmub/dmub_srv.h | 21 +++++++++++++++++++ .../gpu/drm/amd/display/dmub/src/dmub_srv.c | 15 +++++++++++++ 5 files changed, 55 insertions(+) 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 d7ac020bd8af..54861136dafd 100644 --- a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.c +++ b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.c @@ -2687,6 +2687,7 @@ static int dm_suspend(void *handle) hpd_rx_irq_work_suspend(dm); dc_set_power_state(dm->dc, DC_ACPI_CM_POWER_STATE_D3); + dc_dmub_srv_set_power_state(dm->dc->ctx->dmub_srv, DC_ACPI_CM_POWER_STATE_D3); return 0; } @@ -2882,6 +2883,7 @@ static int dm_resume(void *handle) if (r) DRM_ERROR("DMUB interface failed to initialize: status=%d\n", r); + dc_dmub_srv_set_power_state(dm->dc->ctx->dmub_srv, DC_ACPI_CM_POWER_STATE_D0); dc_set_power_state(dm->dc, DC_ACPI_CM_POWER_STATE_D0); dc_resume(dm->dc); @@ -2932,6 +2934,7 @@ static int dm_resume(void *handle) } /* power on hardware */ + dc_dmub_srv_set_power_state(dm->dc->ctx->dmub_srv, DC_ACPI_CM_POWER_STATE_D0); dc_set_power_state(dm->dc, DC_ACPI_CM_POWER_STATE_D0); /* program HPD filter */ diff --git a/drivers/gpu/drm/amd/display/dc/dc_dmub_srv.c b/drivers/gpu/drm/amd/display/dc/dc_dmub_srv.c index 53400cc05b5b..1a4d615ccdec 100644 --- a/drivers/gpu/drm/amd/display/dc/dc_dmub_srv.c +++ b/drivers/gpu/drm/amd/display/dc/dc_dmub_srv.c @@ -1268,3 +1268,17 @@ void dc_dmub_srv_exit_low_power_state(const struct dc *dc) ASSERT(0); } +void dc_dmub_srv_set_power_state(struct dc_dmub_srv *dc_dmub_srv, enum dc_acpi_cm_power_state powerState) +{ + struct dmub_srv *dmub; + + if (!dc_dmub_srv) + return; + + dmub = dc_dmub_srv->dmub; + + if (powerState == DC_ACPI_CM_POWER_STATE_D0) + dmub_srv_set_power_state(dmub, DMUB_POWER_STATE_D0); + else + dmub_srv_set_power_state(dmub, DMUB_POWER_STATE_D3); +} diff --git a/drivers/gpu/drm/amd/display/dc/dc_dmub_srv.h b/drivers/gpu/drm/amd/display/dc/dc_dmub_srv.h index d4a60f53faab..c25ce7546f71 100644 --- a/drivers/gpu/drm/amd/display/dc/dc_dmub_srv.h +++ b/drivers/gpu/drm/amd/display/dc/dc_dmub_srv.h @@ -102,4 +102,6 @@ void dc_dmub_srv_subvp_save_surf_addr(const struct dc_dmub_srv *dc_dmub_srv, con bool dc_dmub_srv_is_hw_pwr_up(struct dc_dmub_srv *dc_dmub_srv, bool wait); void dc_dmub_srv_notify_idle(const struct dc *dc, bool allow_idle); void dc_dmub_srv_exit_low_power_state(const struct dc *dc); + +void dc_dmub_srv_set_power_state(struct dc_dmub_srv *dc_dmub_srv, enum dc_acpi_cm_power_state powerState); #endif /* _DMUB_DC_SRV_H_ */ diff --git a/drivers/gpu/drm/amd/display/dmub/dmub_srv.h b/drivers/gpu/drm/amd/display/dmub/dmub_srv.h index df63aa8f01e9..d1a4ed6f5916 100644 --- a/drivers/gpu/drm/amd/display/dmub/dmub_srv.h +++ b/drivers/gpu/drm/amd/display/dmub/dmub_srv.h @@ -150,6 +150,13 @@ enum dmub_memory_access_type { DMUB_MEMORY_ACCESS_DMA }; +/* enum dmub_power_state type - to track DC power state in dmub_srv */ +enum dmub_srv_power_state_type { + DMUB_POWER_STATE_UNDEFINED = 0, + DMUB_POWER_STATE_D0 = 1, + DMUB_POWER_STATE_D3 = 8 +}; + /** * struct dmub_region - dmub hw memory region * @base: base address for region, must be 256 byte aligned @@ -485,6 +492,8 @@ struct dmub_srv { /* Feature capabilities reported by fw */ struct dmub_feature_caps feature_caps; struct dmub_visual_confirm_color visual_confirm_color; + + enum dmub_srv_power_state_type power_state; }; /** @@ -889,6 +898,18 @@ enum dmub_status dmub_srv_clear_inbox0_ack(struct dmub_srv *dmub); */ void dmub_srv_subvp_save_surf_addr(struct dmub_srv *dmub, const struct dc_plane_address *addr, uint8_t subvp_index); +/** + * dmub_srv_set_power_state() - Track DC power state in dmub_srv + * @dmub: The dmub service + * @power_state: DC power state setting + * + * Store DC power state in dmub_srv. If dmub_srv is in D3, then don't send messages to DMUB + * + * Return: + * void + */ +void dmub_srv_set_power_state(struct dmub_srv *dmub, enum dmub_srv_power_state_type dmub_srv_power_state); + #if defined(__cplusplus) } #endif diff --git a/drivers/gpu/drm/amd/display/dmub/src/dmub_srv.c b/drivers/gpu/drm/amd/display/dmub/src/dmub_srv.c index a329a6ecf4dd..53ac1c66dd86 100644 --- a/drivers/gpu/drm/amd/display/dmub/src/dmub_srv.c +++ b/drivers/gpu/drm/amd/display/dmub/src/dmub_srv.c @@ -713,6 +713,7 @@ enum dmub_status dmub_srv_hw_init(struct dmub_srv *dmub, dmub->hw_funcs.reset_release(dmub); dmub->hw_init = true; + dmub->power_state = DMUB_POWER_STATE_D0; return DMUB_STATUS_OK; } @@ -766,6 +767,9 @@ enum dmub_status dmub_srv_cmd_queue(struct dmub_srv *dmub, if (!dmub->hw_init) return DMUB_STATUS_INVALID; + if (dmub->power_state != DMUB_POWER_STATE_D0) + return DMUB_STATUS_INVALID; + if (dmub->inbox1_rb.rptr > dmub->inbox1_rb.capacity || dmub->inbox1_rb.wrpt > dmub->inbox1_rb.capacity) { return DMUB_STATUS_HW_FAILURE; @@ -784,6 +788,9 @@ enum dmub_status dmub_srv_cmd_execute(struct dmub_srv *dmub) if (!dmub->hw_init) return DMUB_STATUS_INVALID; + if (dmub->power_state != DMUB_POWER_STATE_D0) + return DMUB_STATUS_INVALID; + /** * Read back all the queued commands to ensure that they've * been flushed to framebuffer memory. Otherwise DMCUB might @@ -1100,3 +1107,11 @@ void dmub_srv_subvp_save_surf_addr(struct dmub_srv *dmub, const struct dc_plane_ subvp_index); } } + +void dmub_srv_set_power_state(struct dmub_srv *dmub, enum dmub_srv_power_state_type dmub_srv_power_state) +{ + if (!dmub || !dmub->hw_init) + return; + + dmub->power_state = dmub_srv_power_state; +} From c57a0f50c060b7c58f974306fe103eabb881ccbc Mon Sep 17 00:00:00 2001 From: Muhammad Ahmed Date: Fri, 1 Dec 2023 19:11:50 -0500 Subject: [PATCH 91/92] drm/amd/display: remove HPO PG in driver side [why & how] Add config to make HPO PG handled in dmubfw ips entry/exit Reviewed-by: Charlene Liu Acked-by: Wayne Lin Signed-off-by: Muhammad Ahmed Tested-by: Daniel Wheeler Signed-off-by: Alex Deucher --- drivers/gpu/drm/amd/display/dc/resource/dcn35/dcn35_resource.c | 1 + 1 file changed, 1 insertion(+) diff --git a/drivers/gpu/drm/amd/display/dc/resource/dcn35/dcn35_resource.c b/drivers/gpu/drm/amd/display/dc/resource/dcn35/dcn35_resource.c index 4eb744f1bc9f..4e1db842b98c 100644 --- a/drivers/gpu/drm/amd/display/dc/resource/dcn35/dcn35_resource.c +++ b/drivers/gpu/drm/amd/display/dc/resource/dcn35/dcn35_resource.c @@ -718,6 +718,7 @@ static const struct dc_debug_options debug_defaults_drv = { .disable_dpp_power_gate = true, .disable_hubp_power_gate = true, .disable_optc_power_gate = true, /*should the same as above two*/ + .disable_hpo_power_gate = true, /*dmubfw force domain25 on*/ .disable_clock_gate = false, .disable_dsc_power_gate = true, .vsr_support = true, From 669080888691c312cc926322a7b24600121c90fb Mon Sep 17 00:00:00 2001 From: Alvin Lee Date: Mon, 4 Dec 2023 15:19:34 -0500 Subject: [PATCH 92/92] drm/amd/display: Revert " drm/amd/display: Use channel_width = 2 for vram table 3.0" [Description] Revert commit fec05adc40c2 ("drm/amd/display: Use channel_width = 2 for vram table 3.0") Because the issue is being fixed from VBIOS side. Reviewed-by: Samson Tam Acked-by: Wayne Lin Signed-off-by: Alvin Lee Tested-by: Daniel Wheeler Signed-off-by: Alex Deucher --- drivers/gpu/drm/amd/display/dc/bios/bios_parser2.c | 8 +------- 1 file changed, 1 insertion(+), 7 deletions(-) diff --git a/drivers/gpu/drm/amd/display/dc/bios/bios_parser2.c b/drivers/gpu/drm/amd/display/dc/bios/bios_parser2.c index 875a064bb9a5..fcd65a2057ad 100644 --- a/drivers/gpu/drm/amd/display/dc/bios/bios_parser2.c +++ b/drivers/gpu/drm/amd/display/dc/bios/bios_parser2.c @@ -2386,13 +2386,7 @@ static enum bp_result get_vram_info_v30( return BP_RESULT_BADBIOSTABLE; info->num_chans = info_v30->channel_num; - /* As suggested by VBIOS we should always use - * dram_channel_width_bytes = 2 when using VRAM - * table version 3.0. This is because the channel_width - * param in the VRAM info table is changed in 7000 series and - * no longer represents the memory channel width. - */ - info->dram_channel_width_bytes = 2; + info->dram_channel_width_bytes = (1 << info_v30->channel_width) / 8; return result; }