From b91d287fa7a1ba0727eed5823c6ee4924ee5fa31 Mon Sep 17 00:00:00 2001 From: Ricardo Neri Date: Sat, 13 Jun 2026 15:17:47 -0700 Subject: [PATCH 1/2] thermal: intel: Fix dangling resources on thermal_throttle_online() failure The function thermal_throttle_add_dev() may fail and abort a CPU hotplug online operation. Since the failure occurs within the online callback, thermal_throttle_online(), the CPU hotplug framework does not invoke the corresponding offline callback. As a result, the hardware and software resources set up during the failed operation are not torn down. Since only thermal_throttle_add_dev() can fail, call it before setting up the rest of the resources. Fixes: f6656208f04e ("x86/mce/therm_throt: Optimize notifications of thermal throttle") Signed-off-by: Ricardo Neri Link: https://patch.msgid.link/20260613-rneri-directed-therm-intr-v3-1-3a26d1e47fc8@linux.intel.com Signed-off-by: Rafael J. Wysocki --- drivers/thermal/intel/therm_throt.c | 7 ++++++- 1 file changed, 6 insertions(+), 1 deletion(-) diff --git a/drivers/thermal/intel/therm_throt.c b/drivers/thermal/intel/therm_throt.c index 44fa4dd15dd1..45a8ef4a608b 100644 --- a/drivers/thermal/intel/therm_throt.c +++ b/drivers/thermal/intel/therm_throt.c @@ -529,8 +529,13 @@ static int thermal_throttle_online(unsigned int cpu) { struct thermal_state *state = &per_cpu(thermal_state, cpu); struct device *dev = get_cpu_device(cpu); + int err; u32 l; + err = thermal_throttle_add_dev(dev, cpu); + if (err) + return err; + state->package_throttle.level = PACKAGE_LEVEL; state->core_throttle.level = CORE_LEVEL; @@ -548,7 +553,7 @@ static int thermal_throttle_online(unsigned int cpu) l = apic_read(APIC_LVTTHMR); apic_write(APIC_LVTTHMR, l & ~APIC_LVT_MASKED); - return thermal_throttle_add_dev(dev, cpu); + return err; } static int thermal_throttle_offline(unsigned int cpu) From fb1a5dfe86d3af1e1c3ce168cf0d8d43897e0f77 Mon Sep 17 00:00:00 2001 From: "Rafael J. Wysocki" Date: Fri, 19 Jun 2026 17:01:27 +0200 Subject: [PATCH 2/2] thermal: testing: zone: Flush work items during cleanup To prevent freed module code from being executed during the thermal testing module unload, make it add a dedicated workqueue for thermal testing work items and flush it in thermal_testing_exit(). Fixes: f6a034f2df42 ("thermal: Introduce a debugfs-based testing facility") Link: https://sashiko.dev/#/patchset/20260605185212.2491144-1-sam.moelius%40trailofbits.com Signed-off-by: Rafael J. Wysocki Link: https://patch.msgid.link/1959388.tdWV9SEqCh@rafael.j.wysocki [ rjw: Make variable d_command static ] Signed-off-by: Rafael J. Wysocki --- drivers/thermal/testing/command.c | 36 ++++++++++++++++++++--- drivers/thermal/testing/thermal_testing.h | 8 +++++ drivers/thermal/testing/zone.c | 7 ++--- 3 files changed, 43 insertions(+), 8 deletions(-) diff --git a/drivers/thermal/testing/command.c b/drivers/thermal/testing/command.c index fbf7ab9729b5..90e06c593327 100644 --- a/drivers/thermal/testing/command.c +++ b/drivers/thermal/testing/command.c @@ -86,7 +86,10 @@ #include "thermal_testing.h" -struct dentry *d_testing; +struct workqueue_struct *tt_wq __ro_after_init; + +struct dentry *d_testing __ro_after_init; +static struct dentry *d_command __ro_after_init; #define TT_COMMAND_SIZE 16 @@ -203,17 +206,42 @@ static const struct file_operations tt_command_fops = { static int __init thermal_testing_init(void) { + int error; + + tt_wq = alloc_workqueue("thermal_testing", WQ_UNBOUND, 0); + if (!tt_wq) + return -ENOMEM; + d_testing = debugfs_create_dir("thermal-testing", NULL); - if (!IS_ERR(d_testing)) - debugfs_create_file("command", 0200, d_testing, NULL, - &tt_command_fops); + if (IS_ERR(d_testing)) { + error = PTR_ERR(d_testing); + goto destroy_wq; + } + + d_command = debugfs_create_file("command", 0200, d_testing, NULL, &tt_command_fops); + if (IS_ERR(d_command)) { + error = PTR_ERR(d_command); + goto remove_d_testing; + } return 0; + +remove_d_testing: + debugfs_remove(d_testing); +destroy_wq: + destroy_workqueue(tt_wq); + return error; } module_init(thermal_testing_init); static void __exit thermal_testing_exit(void) { + /* First, prevent new commands from being entered. */ + debugfs_remove(d_command); + /* Flush commands in progress (if any). */ + flush_workqueue(tt_wq); + destroy_workqueue(tt_wq); + /* Remove the directory structure and clean up. */ debugfs_remove(d_testing); tt_zone_cleanup(); } diff --git a/drivers/thermal/testing/thermal_testing.h b/drivers/thermal/testing/thermal_testing.h index c790a32aae4e..5880c9a63dba 100644 --- a/drivers/thermal/testing/thermal_testing.h +++ b/drivers/thermal/testing/thermal_testing.h @@ -1,4 +1,12 @@ /* SPDX-License-Identifier: GPL-2.0 */ +#include + +extern struct workqueue_struct *tt_wq; + +static inline void tt_queue_work(struct work_struct *work) +{ + queue_work(tt_wq, work); +} extern struct dentry *d_testing; diff --git a/drivers/thermal/testing/zone.c b/drivers/thermal/testing/zone.c index f7f9ca2f1f2c..6db26276dd41 100644 --- a/drivers/thermal/testing/zone.c +++ b/drivers/thermal/testing/zone.c @@ -13,7 +13,6 @@ #include #include #include -#include #include "thermal_testing.h" @@ -207,7 +206,7 @@ int tt_add_tz(void) INIT_WORK(&tt_work->work, tt_add_tz_work_fn); tt_work->tt_zone = no_free_ptr(tt_zone); - schedule_work(&(no_free_ptr(tt_work)->work)); + tt_queue_work(&(no_free_ptr(tt_work)->work)); return 0; } @@ -269,7 +268,7 @@ int tt_del_tz(const char *arg) INIT_WORK(&tt_work->work, tt_del_tz_work_fn); tt_work->tt_zone = tt_zone; - schedule_work(&(no_free_ptr(tt_work)->work)); + tt_queue_work(&(no_free_ptr(tt_work)->work)); return 0; } @@ -358,7 +357,7 @@ int tt_zone_add_trip(const char *arg) INIT_WORK(&tt_work->work, tt_zone_add_trip_work_fn); tt_work->tt_zone = no_free_ptr(tt_zone); tt_work->tt_trip = no_free_ptr(tt_trip); - schedule_work(&(no_free_ptr(tt_work)->work)); + tt_queue_work(&(no_free_ptr(tt_work)->work)); return 0; }