On Tue, Jan 29, 2019 at 7:16 AM Peter Zijlstra <peterz@xxxxxxxxxxxxx> wrote:
>
> On Tue, Jan 29, 2019 at 01:38:43PM +0100, Peter Zijlstra wrote:
> > On Thu, Jan 24, 2019 at 01:15:18PM -0800, Suren Baghdasaryan wrote:
> > > + atomic_set(&group->polling, polling);
> > > + /*
> > > + * Memory barrier is needed to order group->polling
> > > + * write before times[] read in collect_percpu_times()
> > > + */
> > > + smp_mb__after_atomic();
> >
> > That's broken, smp_mb__{before,after}_atomic() can only be used on
> > atomic RmW operations, something atomic_set() is _not_.
>
> Also; the comment should explain _why_ not only what.
Got it. Will change the comment to something like:
Order group->polling=0 before reading times[] in
collect_percpu_times() to detect possible race with hotpath that
modifies times[] before it sets group->polling=1 (see Race #1 in the
comments at the top).
> --
> You received this message because you are subscribed to the Google Groups "kernel-team" group.
> To unsubscribe from this group and stop receiving emails from it, send an email to kernel-team+unsubscribe@xxxxxxxxxxx.
>
