Linux kernel mirror (for testing)
git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git
kernel
os
linux
sched_ext: Skip tasks with stale task_rq in bypass_lb_cpu()
bypass_lb_cpu() transfers tasks between per-CPU bypass DSQs without
migrating them - task_cpu() only updates when the donee later consumes the
task via move_remote_task_to_local_dsq(). If the LB timer fires again before
consumption and the new DSQ becomes a donor, @p is still on the previous CPU
and task_rq(@p) != donor_rq. @p can't be moved without its own rq locked.
Skip such tasks.
Fixes: 95d1df610cdc ("sched_ext: Implement load balancer for bypass mode")
Cc: stable@vger.kernel.org # v6.19+
Reported-by: Chris Mason <clm@meta.com>
Signed-off-by: Tejun Heo <tj@kernel.org>
Reviewed-by: Andrea Righi <arighi@nvidia.com>
+9
+9
kernel/sched/ext.c
+9
kernel/sched/ext.c
···
5023
5023
if (cpumask_empty(donee_mask))
5024
5024
break;
5025
5025
5026
+
/*
5027
+
* If an earlier pass placed @p on @donor_dsq from a different
5028
+
* CPU and the donee hasn't consumed it yet, @p is still on the
5029
+
* previous CPU and task_rq(@p) != @donor_rq. @p can't be moved
5030
+
* without its rq locked. Skip.
5031
+
*/
5032
+
if (task_rq(p) != donor_rq)
5033
+
continue;
5034
+
5026
5035
donee = cpumask_any_and_distribute(donee_mask, p->cpus_ptr);
5027
5036
if (donee >= nr_cpu_ids)
5028
5037
continue;