linux-stable.git/kernel/power/energy_model.c, branch linux-6.12.y

PM: EM: Fix late boot with holes in CPU topology

2025-10-29T13:08:53+00:00

[ Upstream commit 1ebe8f7e782523e62cd1fa8237f7afba5d1dae83 ]

Commit e3f1164fc9ee ("PM: EM: Support late CPUs booting and capacity
adjustment") added a mechanism to handle CPUs that come up late by
retrying when any of the `cpufreq_cpu_get()` call fails.

However, if there are holes in the CPU topology (offline CPUs, e.g.
nosmt), the first missing CPU causes the loop to break, preventing
subsequent online CPUs from being updated.

Instead of aborting on the first missing CPU policy, loop through all
and retry if any were missing.

Fixes: e3f1164fc9ee ("PM: EM: Support late CPUs booting and capacity adjustment")
Suggested-by: Kenneth Crudup 
Reported-by: Kenneth Crudup 
Link: https://lore.kernel.org/linux-pm/40212796-734c-4140-8a85-854f72b8144d@panix.com/
Cc: 6.9+  # 6.9+
Signed-off-by: Christian Loehle 
Link: https://patch.msgid.link/20250831214357.2020076-1-christian.loehle@arm.com
[ rjw: Drop the new pr_debug() message which is not very useful ]
Signed-off-by: Rafael J. Wysocki 
Signed-off-by: Sasha Levin 
Signed-off-by: Greg Kroah-Hartman

PM: EM: Move CPU capacity check to em_adjust_new_capacity()

2025-10-29T13:08:53+00:00

[ Upstream commit 3e3ba654d3097e0031f2add215b12ff81c23814e ]

Move the check of the CPU capacity currently stored in the energy model
against the arch_scale_cpu_capacity() value to em_adjust_new_capacity()
so it will be done regardless of where the latter is called from.

This will be useful when a new em_adjust_new_capacity() caller is added
subsequently.

While at it, move the pd local variable declaration in
em_check_capacity_update() into the loop in which it is used.

No intentional functional impact.

Signed-off-by: Rafael J. Wysocki 
Reviewed-by: Lukasz Luba 
Tested-by: Christian Loehle 
Reviewed-by: Dietmar Eggemann 
Link: https://patch.msgid.link/7810787.EvYhyI6sBW@rjwysocki.net
Stable-dep-of: 1ebe8f7e7825 ("PM: EM: Fix late boot with holes in CPU topology")
Signed-off-by: Sasha Levin 
Signed-off-by: Greg Kroah-Hartman

PM: EM: Slightly reduce em_check_capacity_update() overhead

2025-10-29T13:08:53+00:00

[ Upstream commit a8e62726ac0dd7b610c87ba1a938a5a9091c34df ]

Every iteration of the loop over all possible CPUs in
em_check_capacity_update() causes get_cpu_device() to be called twice
for the same CPU, once indirectly via em_cpu_get() and once directly.

Get rid of the indirect get_cpu_device() call by moving the direct
invocation of it earlier and using em_pd_get() instead of em_cpu_get()
to get a pd pointer for the dev one returned by it.

This also exposes the fact that dev is needed to get a pd, so the code
becomes somewhat easier to follow after it.

No functional impact.

Signed-off-by: Rafael J. Wysocki 
Reviewed-by: Lukasz Luba 
Link: https://patch.msgid.link/1925950.tdWV9SEqCh@rjwysocki.net
Stable-dep-of: 1ebe8f7e7825 ("PM: EM: Fix late boot with holes in CPU topology")
Signed-off-by: Sasha Levin 
Signed-off-by: Greg Kroah-Hartman

PM: EM: Drop unused parameter from em_adjust_new_capacity()

2025-10-29T13:08:53+00:00

[ Upstream commit 5fad775d432c6c9158ea12e7e00d8922ef8d3dfc ]

The max_cap parameter is never used in em_adjust_new_capacity(), so
drop it.

No functional impact.

Signed-off-by: Rafael J. Wysocki 
Reviewed-by: Lukasz Luba 
Link: https://patch.msgid.link/2369979.ElGaqSPkdT@rjwysocki.net
Stable-dep-of: 1ebe8f7e7825 ("PM: EM: Fix late boot with holes in CPU topology")
Signed-off-by: Sasha Levin 
Signed-off-by: Greg Kroah-Hartman

PM: EM: Fix potential division-by-zero error in em_compute_costs()

2025-06-19T13:31:32+00:00

[ Upstream commit 179c0c7044a378198adb36f2a12410ab68cc730a ]

When the device is of a non-CPU type, table[i].performance won't be
initialized in the previous em_init_performance(), resulting in division
by zero when calculating costs in em_compute_costs().

Since the 'cost' algorithm is only used for EAS energy efficiency
calculations and is currently not utilized by other device drivers, we
should add the _is_cpu_device(dev) check to prevent this division-by-zero
issue.

Fixes: 1b600da51073 ("PM: EM: Optimize em_cpu_energy() and remove division")
Signed-off-by: Yaxiong Tian 
Reviewed-by: Lukasz Luba 
Link: https://patch.msgid.link/tencent_7F99ED4767C1AF7889D0D8AD50F34859CE06@qq.com
Signed-off-by: Rafael J. Wysocki 
Signed-off-by: Sasha Levin

PM: EM: Address RCU-related sparse warnings

2025-05-02T05:58:52+00:00

[ Upstream commit 3ee7be9e10dd5f79448788b899591d4bd2bf0c19 ]

The usage of __rcu in the Energy Model code is quite inconsistent
which causes the following sparse warnings to trigger:

kernel/power/energy_model.c:169:15: warning: incorrect type in assignment (different address spaces)
kernel/power/energy_model.c:169:15:    expected struct em_perf_table [noderef] __rcu *table
kernel/power/energy_model.c:169:15:    got struct em_perf_table *
kernel/power/energy_model.c:171:9: warning: incorrect type in argument 1 (different address spaces)
kernel/power/energy_model.c:171:9:    expected struct callback_head *head
kernel/power/energy_model.c:171:9:    got struct callback_head [noderef] __rcu *
kernel/power/energy_model.c:171:9: warning: cast removes address space '__rcu' of expression
kernel/power/energy_model.c:182:19: warning: incorrect type in argument 1 (different address spaces)
kernel/power/energy_model.c:182:19:    expected struct kref *kref
kernel/power/energy_model.c:182:19:    got struct kref [noderef] __rcu *
kernel/power/energy_model.c:200:15: warning: incorrect type in assignment (different address spaces)
kernel/power/energy_model.c:200:15:    expected struct em_perf_table [noderef] __rcu *table
kernel/power/energy_model.c:200:15:    got void *[assigned] _res
kernel/power/energy_model.c:204:20: warning: incorrect type in argument 1 (different address spaces)
kernel/power/energy_model.c:204:20:    expected struct kref *kref
kernel/power/energy_model.c:204:20:    got struct kref [noderef] __rcu *
kernel/power/energy_model.c:320:19: warning: incorrect type in argument 1 (different address spaces)
kernel/power/energy_model.c:320:19:    expected struct kref *kref
kernel/power/energy_model.c:320:19:    got struct kref [noderef] __rcu *
kernel/power/energy_model.c:325:45: warning: incorrect type in argument 2 (different address spaces)
kernel/power/energy_model.c:325:45:    expected struct em_perf_state *table
kernel/power/energy_model.c:325:45:    got struct em_perf_state [noderef] __rcu *
kernel/power/energy_model.c:425:45: warning: incorrect type in argument 3 (different address spaces)
kernel/power/energy_model.c:425:45:    expected struct em_perf_state *table
kernel/power/energy_model.c:425:45:    got struct em_perf_state [noderef] __rcu *
kernel/power/energy_model.c:442:15: warning: incorrect type in argument 1 (different address spaces)
kernel/power/energy_model.c:442:15:    expected void const *objp
kernel/power/energy_model.c:442:15:    got struct em_perf_table [noderef] __rcu *[assigned] em_table
kernel/power/energy_model.c:626:55: warning: incorrect type in argument 2 (different address spaces)
kernel/power/energy_model.c:626:55:    expected struct em_perf_state *table
kernel/power/energy_model.c:626:55:    got struct em_perf_state [noderef] __rcu *
kernel/power/energy_model.c:681:16: warning: incorrect type in assignment (different address spaces)
kernel/power/energy_model.c:681:16:    expected struct em_perf_state *new_ps
kernel/power/energy_model.c:681:16:    got struct em_perf_state [noderef] __rcu *
kernel/power/energy_model.c:699:37: warning: incorrect type in argument 2 (different address spaces)
kernel/power/energy_model.c:699:37:    expected struct em_perf_state *table
kernel/power/energy_model.c:699:37:    got struct em_perf_state [noderef] __rcu *
kernel/power/energy_model.c:733:38: warning: incorrect type in argument 3 (different address spaces)
kernel/power/energy_model.c:733:38:    expected struct em_perf_state *table
kernel/power/energy_model.c:733:38:    got struct em_perf_state [noderef] __rcu *
kernel/power/energy_model.c:855:53: warning: dereference of noderef expression
kernel/power/energy_model.c:864:32: warning: dereference of noderef expression

This is because the __rcu annotation for sparse is only applicable to
pointers that need rcu_dereference() or equivalent for protection, which
basically means pointers assigned with rcu_assign_pointer().

Make all of the above sparse warnings go away by cleaning up the usage
of __rcu and using rcu_dereference_protected() where applicable.

Cc: All applicable 
Signed-off-by: Rafael J. Wysocki 
Reviewed-by: Lukasz Luba 
Link: https://patch.msgid.link/5885405.DvuYhMxLoT@rjwysocki.net
Signed-off-by: Sasha Levin

PM: EM: use kfree_rcu() to simplify the code

2025-05-02T05:58:52+00:00

[ Upstream commit 1618f635bdf56f3ac158171114e9bf18db234cbf ]

The callback function of call_rcu() just calls kfree(), so use
kfree_rcu() instead of call_rcu() + callback function.

Signed-off-by: Li RongQing 
Reviewed-by: Lukasz Luba 
Link: https://patch.msgid.link/20250218082021.2766-1-lirongqing@baidu.com
Signed-off-by: Rafael J. Wysocki 
Stable-dep-of: 3ee7be9e10dd ("PM: EM: Address RCU-related sparse warnings")
Signed-off-by: Sasha Levin

PM: EM: Add em_dev_update_chip_binning()

2024-04-08T14:05:14+00:00

Add a function which allows to modify easily the EM after the new voltage
information is available. The device drivers for the chip can adjust
the voltage values after setup. The voltage for the same frequency in OPP
can be different due to chip binning. The voltage impacts the power usage
and the EM power values can be updated to reflect that.

Reviewed-by: Dietmar Eggemann 
Signed-off-by: Lukasz Luba 
Signed-off-by: Rafael J. Wysocki

PM: EM: Refactor em_adjust_new_capacity()

2024-04-08T14:05:14+00:00

Extract em_table_dup() and em_recalc_and_update() from
em_adjust_new_capacity(). Both functions will be later reused by the
'update EM due to chip binning' functionality.

Reviewed-by: Dietmar Eggemann 
Signed-off-by: Lukasz Luba 
Signed-off-by: Rafael J. Wysocki

PM: EM: Force device drivers to provide power in uW

2024-03-13T19:48:38+00:00

The EM only supports power in uW. Make sure that it is not possible to
register some downstream driver which doesn't provide power in uW.
The only exception is artificial EM, but that EM is ignored by the rest of
kernel frameworks (thermal, powercap, etc).

Reported-by: PoShao Chen 
Signed-off-by: Lukasz Luba 
Signed-off-by: Rafael J. Wysocki