Re: [PATCH v3 2/2] cgroup/cpuset: Skip security check for hotplug induced v1 task migration

[Waiman Long]

On 4/20/26 1:08 PM, Michal Koutný wrote:
> On Tue, Mar 31, 2026 at 11:11:08AM -0400, Waiman Long <longman@xxxxxxxxxx> wrote:
> > If a strict security policy is in place, however, the task migration
> > may fail when security_task_setscheduler() call in cpuset_can_attach()
> > returns a -EACCESS error.
> I think this should be generally safe to skip (since v2 doesn't consider
> the object of cpuset migration at all).
>
> > That will mean that those tasks will have no CPU to run on. The system
> > administrators will have to explicitly intervene to either add CPUs to
> > that cpuset or move the tasks elsewhere if they are aware of it.
> That "no CPU to run on" means the affected tasks would remain in
> schedule() indefinitely?

I believe the scheduler has a fallback mechanism in that particular 
case, but it can be any CPU. So I don't think we should rely on that.

Cheers,
Longman