回复: [PATCH v2] sched/eevdf: Force propagating min_slice of cfs_rq when a task changing slice
From: 解 咏梅
Date: Wed Nov 13 2024 - 06:50:40 EST
Similar problem as commit d2929762 & 8dafa9d0, but this time heap integrity is corrupted by min_slice attr.
commit eab03c23c fixed it by explicitly calling __dequeue_entity and __enqueue_entity in reweight_entity.
But, it's rare case, it only happens when adjust task's select by setting up scheduler attribute.
Regards,
Yongmei.
________________________________________
发件人: Tianchen Ding <dtcccc@xxxxxxxxxxxxxxxxx>
发送时间: 2024年11月12日 11:25
收件人: linux-kernel@xxxxxxxxxxxxxxx <linux-kernel@xxxxxxxxxxxxxxx>
抄送: Ingo Molnar <mingo@xxxxxxxxxx>; Peter Zijlstra <peterz@xxxxxxxxxxxxx>; Juri Lelli <juri.lelli@xxxxxxxxxx>; Vincent Guittot <vincent.guittot@xxxxxxxxxx>; Dietmar Eggemann <dietmar.eggemann@xxxxxxx>; Steven Rostedt <rostedt@xxxxxxxxxxx>; Ben Segall <bsegall@xxxxxxxxxx>; Mel Gorman <mgorman@xxxxxxx>; Valentin Schneider <vschneid@xxxxxxxxxx>
主题: Re: [PATCH v2] sched/eevdf: Force propagating min_slice of cfs_rq when a task changing slice
On 2024/10/31 17:48, Tianchen Ding wrote:
> When a task changes slice and its cgroup se is already on_rq, the cgroup
> se will not be enqueued again, and hence the root->min_slice leaves
> unchanged.
>
> Force propagating it when se doesn't need to be enqueued (or dequeued).
> Ensure the se hierarchy always get the latest min_slice.
>
> Fixes: aef6987d8954 ("sched/eevdf: Propagate min_slice up the cgroup hierarchy")
> Signed-off-by: Tianchen Ding <dtcccc@xxxxxxxxxxxxxxxxx>
ping for this fix.