sched/deadline: Add more reschedule cases to prio_changed_dl()
Valentin Schneider
reviewerRFC PATCH v2 1/1 sched/deadline: Add more reschedule cases to prio_changed_dl()
I've been tracking down an issue on a ~5.17ish kernel where:
CPUx CPUy
<DL task p0 owns an rtmutex M>
<p0 depletes its runtime, gets throttled>
<rq switches to the idle task>
<DL task p1 blocks on M, boost/replenish p0> <No call to resched_curr() happens here>
[idle task keeps running here until *something*
accidentally sets TIF_NEED_RESCHED]
On that kernel, it is quite easy to trigger using rt-tests's deadline_test
[1] with the test running on isolated CPUs (this reduces the chance of something unrelated setting TIF_NEED_RESCHED on the idle tasks, making the issue even more obvious as the hung task detector chimes in).
I haven't been able to reproduce this using a mainline kernel, even if I revert
2972e3050e35 ("tracing: Make trace_marker{,_raw} stream-like")
which gets rid of the lock involved in the above test, *but* I cannot convince myself the issue isn't there from looking at the code.
Make prio_changed_dl() issue a reschedule if the current task isn't a deadline one. While at it, ensure a reschedule is emitted when a queued-but-not-current task gets boosted with an earlier deadline that current's.
[1]: https://git.kernel.org/pub/scm/utils/rt-tests/rt-tests.git
---
RFCv1 -> RFCv2
++++++++++++++
o Fixed UP build issue (Juri)
--- kernel/sched/deadline.c | 42 ++++++++++++++++++++++++++--------------- 1 file changed, 27 insertions(+), 15 deletions(-)
kernel/sched/deadline.c +27 -15
@@ -2663,17 +2663,20 @@ static void switched_to_dl(struct rq *rq, struct task_struct *p)static void prio_changed_dl (struct rq *rq,struct task_struct *p,int oldprio) { -if (task_on_rq_queued (p) ||task_current (rq, p)) { +if (!task_on_rq_queued (p)) +return ; +#ifdef CONFIG_SMP -/* -* This might be too much, but unfortunately -* we don't have the old deadline value, and -* we can't argue if the task is increasing -* or lowering its prio, so... -*/ -if (!rq->dl.overloaded) -deadline_queue_pull_task (rq); +/* +* This might be too much, but unfortunately +* we don't have the old deadline value, and +* we can't argue if the task is increasing +* or lowering its prio, so... +*/ +if (!rq->dl.overloaded) +deadline_queue_pull_task (rq); +if (task_current (rq, p)) {/* * If we now have a earlier deadline task than p, * then reschedule, provided p is still on this
@@ -2681,15 +2684,24 @@ static void prio_changed_dl(struct rq *rq, struct task_struct *p,*/ if (dl_time_before (rq->dl.earliest_dl.curr, p->dl.deadline))resched_curr (rq); -#else + } else {/* -* Again, we don't know if p has a earlier -* or later deadline, so let's blindly set a -* (maybe not needed) rescheduling point. +* Current may not be deadline in case p was throttled but we +* have just replenished it (e.g. rt_mutex_setprio()). +* +* Otherwise, if p was given an earlier deadline, reschedule. */ -resched_curr (rq); -#endif /* CONFIG_SMP */ +if (!dl_task (rq->curr) || +dl_time_before (p->dl.deadline, rq->curr->dl.deadline)) +resched_curr (rq); } +#else +/* +* We don't know if p has a earlier or later deadline, so let's blindly +* set a (maybe not needed) rescheduling point. +*/ +resched_curr (rq); +#endif } DEFINE_SCHED_CLASS(dl) = {
Signature · 1 line
2.31.1
Hi,
Quoted 26 lines · Valentin Schneider
I've been tracking down an issue on a ~5.17ish kernel where:
CPUx CPUy
<DL task p0 owns an rtmutex M>
<p0 depletes its runtime, gets throttled>
<rq switches to the idle task>
<DL task p1 blocks on M, boost/replenish p0> <No call toresched_curr()happens here>[idle task keeps running here until *something*
accidentally setsTIF_NEED_RESCHED]On that kernel, it is quite easy to trigger using rt-tests's
deadline_test
[1] with the test running on isolated CPUs (this reduces the chance of something unrelated settingTIF_NEED_RESCHEDon the idle tasks, making the issue even more obvious as the hung task detector chimes in).I haven't been able to reproduce this using a mainline kernel, even if I revert
2972e3050e35("tracing: Make trace_marker{,_raw} stream-like")which gets rid of the lock involved in the above test, *but* I cannot convince myself the issue isn't there from looking at the code.
Quoted 8 lines
Make
prio_changed_dl()issue a reschedule if the current task isn't a deadline one. While at it, ensure a reschedule is emitted when a queued-but-not-current task gets boosted with an earlier deadline that current's.[1]: https://git.kernel.org/pub/scm/utils/rt-tests/rt-tests.git
Signed-off-by Valentin Schneider <[email protected]>---
This looks now good to me, thanks!
Best,
Juri
Daniel Bristot de Oliveira
Reviewed-byQuoted 34 lines · Juri Lelli
Hi,
Quoted 31 lines · Valentin Schneider
I've been tracking down an issue on a ~5.17ish kernel where:
CPUx CPUy
<DL task p0 owns an rtmutex M>
<p0 depletes its runtime, gets throttled>
<rq switches to the idle task>
<DL task p1 blocks on M, boost/replenish p0> <No call toresched_curr()happens here>[idle task keeps running here until *something*
accidentally setsTIF_NEED_RESCHED]On that kernel, it is quite easy to trigger using rt-tests's
deadline_test
[1] with the test running on isolated CPUs (this reduces the chance of something unrelated settingTIF_NEED_RESCHEDon the idle tasks, making the issue even more obvious as the hung task detector chimes in).I haven't been able to reproduce this using a mainline kernel, even if I revert
2972e3050e35("tracing: Make trace_marker{,_raw} stream-like")which gets rid of the lock involved in the above test, *but* I cannot convince myself the issue isn't there from looking at the code.
Make
prio_changed_dl()issue a reschedule if the current task isn't a deadline one. While at it, ensure a reschedule is emitted when a queued-but-not-current task gets boosted with an earlier deadline that current's.
Quoted 6 lines
Quoted 3 lines
[1]: https://git.kernel.org/pub/scm/utils/rt-tests/rt-tests.git
Signed-off-by Valentin Schneider <[email protected]>---
This looks now good to me, thanks!
Acked-by Juri Lelli <[email protected]>
-- Daniel
The following commit has been merged into the sched/core branch of tip:
The following commit has been merged into the sched/core branch of tip:
Commit-ID: 7ea98dfa44917a201e76d4fe96bf61d76e60f524 Gitweb: https://git.kernel.org/tip/7ea98dfa44917a201e76d4fe96bf61d76e60f524 Author: Valentin Schneider <[email protected]> AuthorDate: Mon, 06 Feb 2023 14:06:12 Committer: Peter Zijlstra <[email protected]> CommitterDate: Sat, 11 Feb 2023 11:18:10 +01:00
sched/deadline: Add more reschedule cases to prio_changed_dl()
I've been tracking down an issue on a ~5.17ish kernel where:
CPUx CPUy
<DL task p0 owns an rtmutex M>
<p0 depletes its runtime, gets throttled>
<rq switches to the idle task>
<DL task p1 blocks on M, boost/replenish p0> <No call to resched_curr() happens here>
[idle task keeps running here until *something*
accidentally sets TIF_NEED_RESCHED]
On that kernel, it is quite easy to trigger using rt-tests's deadline_test
[1] with the test running on isolated CPUs (this reduces the chance of something unrelated setting TIF_NEED_RESCHED on the idle tasks, making the issue even more obvious as the hung task detector chimes in).
I haven't been able to reproduce this using a mainline kernel, even if I revert
2972e3050e35 ("tracing: Make trace_marker{,_raw} stream-like")
which gets rid of the lock involved in the above test, *but* I cannot convince myself the issue isn't there from looking at the code.
Make prio_changed_dl() issue a reschedule if the current task isn't a deadline one. While at it, ensure a reschedule is emitted when a queued-but-not-current task gets boosted with an earlier deadline that current's.
[1]: https://git.kernel.org/pub/scm/utils/rt-tests/rt-tests.git
--- kernel/sched/deadline.c | 42 +++++++++++++++++++++++++--------------- 1 file changed, 27 insertions(+), 15 deletions(-)
kernel/sched/deadline.c +27 -15
@@ -2663,17 +2663,20 @@ static void switched_to_dl(struct rq *rq, struct task_struct *p)static void prio_changed_dl (struct rq *rq,struct task_struct *p,int oldprio) { -if (task_on_rq_queued (p) ||task_current (rq, p)) { +if (!task_on_rq_queued (p)) +return ; +#ifdef CONFIG_SMP -/* -* This might be too much, but unfortunately -* we don't have the old deadline value, and -* we can't argue if the task is increasing -* or lowering its prio, so... -*/ -if (!rq->dl.overloaded) -deadline_queue_pull_task (rq); +/* +* This might be too much, but unfortunately +* we don't have the old deadline value, and +* we can't argue if the task is increasing +* or lowering its prio, so... +*/ +if (!rq->dl.overloaded) +deadline_queue_pull_task (rq); +if (task_current (rq, p)) {/* * If we now have a earlier deadline task than p, * then reschedule, provided p is still on this
@@ -2681,15 +2684,24 @@ static void prio_changed_dl(struct rq *rq, struct task_struct *p,*/ if (dl_time_before (rq->dl.earliest_dl.curr, p->dl.deadline))resched_curr (rq); -#else + } else {/* -* Again, we don't know if p has a earlier -* or later deadline, so let's blindly set a -* (maybe not needed) rescheduling point. +* Current may not be deadline in case p was throttled but we +* have just replenished it (e.g. rt_mutex_setprio()). +* +* Otherwise, if p was given an earlier deadline, reschedule. */ -resched_curr (rq); -#endif /* CONFIG_SMP */ +if (!dl_task (rq->curr) || +dl_time_before (p->dl.deadline, rq->curr->dl.deadline)) +resched_curr (rq); } +#else +/* +* We don't know if p has a earlier or later deadline, so let's blindly +* set a (maybe not needed) rescheduling point. +*/ +resched_curr (rq); +#endif } DEFINE_SCHED_CLASS(dl) = {