sched/cache: Keep nr_pref_llc_running in the runnable domain, to fix LLC mis-scheduling buggrafted

alb_break_llc() decides whether to break LLC preference during active
load balance. It does so by testing that every runnable fair task on the
source rq prefers its LLC:

	env->src_rq->nr_pref_llc_running == env->src_rq->cfs.h_nr_runnable

But the two counters cover different sets. nr_pref_llc_running is updated
in account_llc_enqueue()/account_llc_dequeue(), next to cfs_rq->nr_queued,
so it follows queued tasks. h_nr_runnable is updated in set_delayed()/
clear_delayed() and drops delay-dequeued tasks.

So under DELAY_DEQUEUE, a preferring task that goes to sleep stays counted
in nr_pref_llc_running while h_nr_runnable falls. The equality then breaks,
alb_break_llc() returns false, and active balance is free to pull a task
off its preferred LLC. Active balance only moves runnable tasks, and this
is the only LLC check it consults: once the stopper runs, LBF_ACTIVE_LB
skips the per-task test in can_migrate_task(). The runnable set is the one
we want.

Fix it on the counter side. A task should be counted in
nr_pref_llc_running exactly while it is both queued on its preferred LLC
(pref_llc_queued) and runnable (!sched_delayed). Define that membership
once in task_pref_llc_runnable(), and adjust the counter only through
pref_llc_running_inc()/pref_llc_running_dec() from the four sites that
change either input: account_llc_enqueue(), account_llc_dequeue(),
set_delayed() and clear_delayed(). Gating every update on the same
predicate keeps the delay, wake and dequeue paths from double-counting
or underflowing; see the comments at those sites for the ordering.

nr_llc_running and sd->llc_counts are not touched and stay on queued
semantics.

Fixes: 714059f79ff0 ("sched/cache: Handle moving single tasks to/from their preferred LLC")
Closes: https://lore.kernel.org/lkml/20260827135000.735138-1-zhanxusheng@xiaomi.com/
Reported-by: Zhan Xusheng <zhanxusheng@xiaomi.com>
Suggested-by: Chen Yu <yu.c.chen@intel.com>
Signed-off-by: Tim Chen <tim.c.chen@linux.intel.com>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Signed-off-by: Ingo Molnar <mingo@kernel.org>
Reviewed-by: Kayra Cizmeci <kayracizmeci@gmail.com>
Cc: <stable@kernel.org> # v7.2.x
Link: https://patch.msgid.link/06af61afedac32e6477f57feb4d658f6c411c3af.1790035273.git.tim.c.chen@linux.intel.com

1 files changed, 170 insertions, 0 deletions

diff --git a/Documentation/gpu/amdgpu/thermal.rst b/Documentation/gpu/amdgpu/thermal.rst
new file mode 100644
index 000000000..1768a106a
--- /dev/null
+++ b/Documentation/gpu/amdgpu/thermal.rst
@@ -0,0 +1,170 @@
+===========================================
+ GPU Power/Thermal Controls and Monitoring
+===========================================
+
+HWMON Interfaces
+================
+
+.. kernel-doc:: drivers/gpu/drm/amd/pm/amdgpu_pm.c
+   :doc: hwmon
+
+GPU sysfs Power State Interfaces
+================================
+
+GPU power controls are exposed via sysfs files.
+
+power_dpm_state
+---------------
+
+.. kernel-doc:: drivers/gpu/drm/amd/pm/amdgpu_pm.c
+   :doc: power_dpm_state
+
+power_dpm_force_performance_level
+---------------------------------
+
+.. kernel-doc:: drivers/gpu/drm/amd/pm/amdgpu_pm.c
+   :doc: power_dpm_force_performance_level
+
+pp_table
+--------
+
+.. kernel-doc:: drivers/gpu/drm/amd/pm/amdgpu_pm.c
+   :doc: pp_table
+
+pp_od_clk_voltage
+-----------------
+
+.. kernel-doc:: drivers/gpu/drm/amd/pm/amdgpu_pm.c
+   :doc: pp_od_clk_voltage
+
+pp_dpm_*
+--------
+
+.. kernel-doc:: drivers/gpu/drm/amd/pm/amdgpu_pm.c
+   :doc: pp_dpm_sclk pp_dpm_mclk pp_dpm_socclk pp_dpm_fclk pp_dpm_dcefclk pp_dpm_pcie
+
+pp_power_profile_mode
+---------------------
+
+.. kernel-doc:: drivers/gpu/drm/amd/pm/amdgpu_pm.c
+   :doc: pp_power_profile_mode
+
+pm_policy
+---------------------
+
+.. kernel-doc:: drivers/gpu/drm/amd/pm/amdgpu_pm.c
+   :doc: pm_policy
+
+\*_busy_percent
+---------------
+
+.. kernel-doc:: drivers/gpu/drm/amd/pm/amdgpu_pm.c
+   :doc: gpu_busy_percent
+
+.. kernel-doc:: drivers/gpu/drm/amd/pm/amdgpu_pm.c
+   :doc: mem_busy_percent
+
+gpu_metrics
+-----------
+
+.. kernel-doc:: drivers/gpu/drm/amd/pm/amdgpu_pm.c
+   :doc: gpu_metrics
+
+fan_curve
+---------
+
+.. kernel-doc:: drivers/gpu/drm/amd/pm/amdgpu_pm.c
+   :doc: fan_curve
+
+acoustic_limit_rpm_threshold
+----------------------------
+
+.. kernel-doc:: drivers/gpu/drm/amd/pm/amdgpu_pm.c
+   :doc: acoustic_limit_rpm_threshold
+
+acoustic_target_rpm_threshold
+-----------------------------
+
+.. kernel-doc:: drivers/gpu/drm/amd/pm/amdgpu_pm.c
+   :doc: acoustic_target_rpm_threshold
+
+fan_target_temperature
+----------------------
+
+.. kernel-doc:: drivers/gpu/drm/amd/pm/amdgpu_pm.c
+   :doc: fan_target_temperature
+
+fan_minimum_pwm
+---------------
+
+.. kernel-doc:: drivers/gpu/drm/amd/pm/amdgpu_pm.c
+   :doc: fan_minimum_pwm
+
+fan_zero_rpm_enable
+----------------------
+
+.. kernel-doc:: drivers/gpu/drm/amd/pm/amdgpu_pm.c
+   :doc: fan_zero_rpm_enable
+
+fan_zero_rpm_stop_temperature
+-----------------------------
+
+.. kernel-doc:: drivers/gpu/drm/amd/pm/amdgpu_pm.c
+   :doc: fan_zero_rpm_stop_temperature
+
+GFXOFF
+======
+
+GFXOFF is a feature found in most recent GPUs that saves power at runtime. The
+card's RLC (RunList Controller) firmware powers off the gfx engine
+dynamically when there is no workload on gfx or compute pipes. GFXOFF is on by
+default on supported GPUs.
+
+Userspace can interact with GFXOFF through a debugfs interface (all values in
+`uint32_t`, unless otherwise noted):
+
+``amdgpu_gfxoff``
+-----------------
+
+Use it to enable/disable GFXOFF, and to check if it's current enabled/disabled::
+
+  $ xxd -l1 -p /sys/kernel/debug/dri/0/amdgpu_gfxoff
+  01
+
+- Write 0 to disable it, and 1 to enable it.
+- Read 0 means it's disabled, 1 it's enabled.
+
+If it's enabled, that means that the GPU is free to enter into GFXOFF mode as
+needed. Disabled means that it will never enter GFXOFF mode.
+
+``amdgpu_gfxoff_status``
+------------------------
+
+Read it to check current GFXOFF's status of a GPU::
+
+  $ xxd -l1 -p /sys/kernel/debug/dri/0/amdgpu_gfxoff_status
+  02
+
+- 0: GPU is in GFXOFF state, the gfx engine is powered down.
+- 1: Transition out of GFXOFF state
+- 2: Not in GFXOFF state
+- 3: Transition into GFXOFF state
+
+If GFXOFF is enabled, the value will be transitioning around [0, 3], always
+getting into 0 when possible. When it's disabled, it's always at 2. Returns
+``-EINVAL`` if it's not supported.
+
+``amdgpu_gfxoff_count``
+-----------------------
+
+Read it to get the total GFXOFF entry count at the time of query since system
+power-up. The value is an `uint64_t` type, however, due to firmware limitations,
+it can currently overflow as an `uint32_t`. *Only supported in vangogh*
+
+``amdgpu_gfxoff_residency``
+---------------------------
+
+Write 1 to amdgpu_gfxoff_residency to start logging, and 0 to stop. Read it to
+get average GFXOFF residency % multiplied by 100 during the last logging
+interval. E.g. a value of 7854 means 78.54% of the time in the last logging
+interval the GPU was in GFXOFF mode. *Only supported in vangogh*