Re: [PATCH v6 4/5] sched: Handle set_cpus_allowed_ptr() & sched_setaffinity() race

[Waiman Long]

On 8/31/22 05:26, Peter Zijlstra wrote:
> On Thu, Aug 25, 2022 at 09:01:18PM -0400, Waiman Long wrote:
> > Racing is possible between set_cpus_allowed_ptr() and sched_setaffinity()
> > or between multiple sched_setaffinity() calls from different CPUs. To
> > resolve these race conditions, we need to update both user_cpus_ptr
> > and cpus_mask in a single lock critical section instead of separated
> > ones. This requires moving the user_cpus_ptr update to
> > affine_move_task() before doing task_rq_unlock().
> >
> > A new argument puser_mask is added to affine_move_task(),
> > __set_cpus_allowed_ptr_locked() and __set_cpus_allowed_ptr() to do that.
> >
> > Ideally, user_cpus_ptr should only be updated if the sched_setaffinity()
> > is successful. However, this patch will update user_cpus_ptr when the
> > first call to __set_cpus_allowed_ptr() is successful. However, if there
> > is racing between sched_setaffinity() and cpuset update, the subsequent
> > calls to __set_cpus_allowed_ptr() may fail but the user_cpus_ptr will
> > still be updated in this corner case.
> Urgh, this is a bit weird, to have a fix for a patch in the same series.

This is just to make each patch simpler and easier to read.

Cheers,
Longman