Re: [RFC PATCH 0/2] workqueue: Fix hotplug races

[Paul E. McKenney]

On Tue, Nov 30, 2021 at 01:06:10AM +0100, Frederic Weisbecker wrote:
> While triggering rcutorture splats, I couldn't trace down to the exact
> cause due to the issue being tracing-unfriendly. But looking at the code,
> those two potential races look like plausible causes and also after
> these two patches, I couldn't reproduce the rcutorture issues again (yet...).
> 
> Frederic Weisbecker (2):
>   workqueue: Fix unbind_workers() VS wq_worker_running() race
>   workqueue: Fix unbind_workers() VS wq_worker_sleeping() race
> 
>  kernel/workqueue.c | 19 +++++++++++++++++++
>  1 file changed, 19 insertions(+)

This series passed 1500 hours of TREE01 with rcupdate.rcu_expedited=1
and also with Frederic's fix for expedited RCU grace periods, so:

Tested-by: Paul E. McKenney <paulmck@xxxxxxxxxx>

							Thanx, Paul