[PATCH 5.12 384/384] thermal/core/fair share: Lock the thermal zone while looping over instances

From: Greg Kroah-Hartman
Date: Mon May 10 2021 - 08:50:51 EST


From: Lukasz Luba <lukasz.luba@xxxxxxx>

commit fef05776eb02238dcad8d5514e666a42572c3f32 upstream.

The tz->lock must be hold during the looping over the instances in that
thermal zone. This lock was missing in the governor code since the
beginning, so it's hard to point into a particular commit.

CC: stable@xxxxxxxxxxxxxxx # 4.4+
Signed-off-by: Lukasz Luba <lukasz.luba@xxxxxxx>
Signed-off-by: Daniel Lezcano <daniel.lezcano@xxxxxxxxxx>
Link: https://lore.kernel.org/r/20210422153624.6074-2-lukasz.luba@xxxxxxx
Signed-off-by: Greg Kroah-Hartman <gregkh@xxxxxxxxxxxxxxxxxxx>
---
drivers/thermal/gov_fair_share.c | 4 ++++
1 file changed, 4 insertions(+)

--- a/drivers/thermal/gov_fair_share.c
+++ b/drivers/thermal/gov_fair_share.c
@@ -82,6 +82,8 @@ static int fair_share_throttle(struct th
int total_instance = 0;
int cur_trip_level = get_trip_level(tz);

+ mutex_lock(&tz->lock);
+
list_for_each_entry(instance, &tz->thermal_instances, tz_node) {
if (instance->trip != trip)
continue;
@@ -110,6 +112,8 @@ static int fair_share_throttle(struct th
mutex_unlock(&instance->cdev->lock);
thermal_cdev_update(cdev);
}
+
+ mutex_unlock(&tz->lock);
return 0;
}