Messages in this thread | | | Date | Mon, 10 Dec 2012 01:27:58 +0530 | From | "Srivatsa S. Bhat" <> | Subject | Re: [RFC PATCH v3 7/9] yield_to(), cpu-hotplug: Prevent offlining of other CPUs properly |
| |
On 12/10/2012 01:18 AM, Oleg Nesterov wrote: > On 12/07, Srivatsa S. Bhat wrote: >> >> Once stop_machine() is gone from the CPU offline path, we won't be able to >> depend on local_irq_save() to prevent CPUs from going offline from under us. > > OK, I guess we need to avoid resched_task()->smp_send_reschedule() > after __cpu_disable() and before migrate_tasks(). >
Yes.
> But, whatever problem we have, > >> Use the get/put_online_cpus_atomic() APIs to prevent CPUs from going offline, >> while invoking from atomic context. > > it should be solved, so... > >> - if (preempt && rq != p_rq) >> + if (preempt && rq != p_rq && cpu_online(task_cpu(p))) > > Why do we need this change? > > Afaics, you could add BUG_ON(!cpu_online(...)) instead? > > I am just curious. >
Oh, I think that's a remnant of v1 (which needed readers to use cpu_online_stable()). You're right, we don't need it. Or we could put a BUG_ON instead, like you suggested.
Regards, Srivatsa S. Bhat
| |