On Sat, 23 Sep 2023 01:14:08 +0000
"Joel Fernandes (Google)" <joel@xxxxxxxxxxxxxxxxx> wrote:
> During RCU-boost testing with the TREE03 rcutorture config, I found that
> after a few hours, the machine locks up.
>
> On tracing, I found that there is a live lock happening between 2 CPUs.
> One CPU has an RT task running, while another CPU is being offlined
> which also has an RT task running. During this offlining, all threads
> are migrated. The migration thread is repeatedly scheduled to migrate
> actively running tasks on the CPU being offlined. This results in a live
> lock because select_fallback_rq() keeps picking the CPU that an RT task
> is already running on only to get pushed back to the CPU being offlined.
>
> It is anyway pointless to pick CPUs for pushing tasks to if they are
> being offlined only to get migrated away to somewhere else. This could
> also add unwanted latency to this task.
>
> Fix these issues by not selecting CPUs in RT if they are not 'active'
> for scheduling, using the cpu_active_mask. Other parts in core.c already
> use cpu_active_mask to prevent tasks from being put on CPUs going
> offline.
>
> Tested-by: Paul E. McKenney <paulmck@xxxxxxxxxx>
> Cc: stable@xxxxxxxxxxxxxxx
> Signed-off-by: Joel Fernandes (Google) <joel@xxxxxxxxxxxxxxxxx>
> ---
> kernel/sched/cpupri.c | 1 +
> 1 file changed, 1 insertion(+)
>
> diff --git a/kernel/sched/cpupri.c b/kernel/sched/cpupri.c
> index a286e726eb4b..42c40cfdf836 100644
> --- a/kernel/sched/cpupri.c
> +++ b/kernel/sched/cpupri.c
> @@ -101,6 +101,7 @@ static inline int __cpupri_find(struct cpupri *cp, struct task_struct *p,
>
> if (lowest_mask) {
> cpumask_and(lowest_mask, &p->cpus_mask, vec->mask);
> + cpumask_and(lowest_mask, lowest_mask, cpu_active_mask);
What happens if the cpu_active_mask changes right here?
Is this just making the race window smaller?
Something tells me the fix is going to be something a bit more involved.
But as I'm getting ready for Paris, I can't look at it at the moment.
-- Steve
>
> /*
> * We have to ensure that we have at least one bit
