Linux kernel mirror (for testing)
git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git
kernel
os
linux
Merge tag 'sched-urgent-2025-11-08' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip
Pull scheduler fix from Ingo Molnar:
"Fix a group-throttling bug in the fair scheduler"
* tag 'sched-urgent-2025-11-08' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip:
sched/fair: Prevent cfs_rq from being unthrottled with zero runtime_remaining
+7
-10
+1
-1
kernel/sched/core.c
+1
-1
kernel/sched/core.c
+6
-9
kernel/sched/fair.c
+6
-9
kernel/sched/fair.c
···
6024
6024
struct sched_entity *se = cfs_rq->tg->se[cpu_of(rq)];
6025
6025
6026
6026
/*
6027
-
* It's possible we are called with !runtime_remaining due to things
6028
-
* like user changed quota setting(see tg_set_cfs_bandwidth()) or async
6029
-
* unthrottled us with a positive runtime_remaining but other still
6030
-
* running entities consumed those runtime before we reached here.
6027
+
* It's possible we are called with runtime_remaining < 0 due to things
6028
+
* like async unthrottled us with a positive runtime_remaining but other
6029
+
* still running entities consumed those runtime before we reached here.
6031
6030
*
6032
-
* Anyway, we can't unthrottle this cfs_rq without any runtime remaining
6033
-
* because any enqueue in tg_unthrottle_up() will immediately trigger a
6034
-
* throttle, which is not supposed to happen on unthrottle path.
6031
+
* We can't unthrottle this cfs_rq without any runtime remaining because
6032
+
* any enqueue in tg_unthrottle_up() will immediately trigger a throttle,
6033
+
* which is not supposed to happen on unthrottle path.
6035
6034
*/
6036
6035
if (cfs_rq->runtime_enabled && cfs_rq->runtime_remaining <= 0)
6037
6036
return;
6038
-
6039
-
se = cfs_rq->tg->se[cpu_of(rq)];
6040
6037
6041
6038
cfs_rq->throttled = 0;
6042
6039