On Thu, Dec 13, 2018 at 04:03:58PM +0100, Sebastian Andrzej Siewior wrote:
> On 2018-12-12 22:05:19 [+0800], Boqun Feng wrote:
> > So Jiangshan and TJ, what's your opion on this one? If we call a
> > queue_work_on() at a place where that target cpu may be offlined, I
> > think we have the guarantee that the work will be eventually executed
> > even if the cpu is never online again, right? In other words, if a cpu
> > has been online once, queue_work_on() on it will be free from racing
> > with cpu hotplug.
> >
> > Am I right about this, or did I miss something subtle?
>
> tj answered this one:
>
> https://lkml.kernel.org/r/20180919205521.ge902...@devbig004.ftw2.facebook.com
I must confess that I would have felt better about that email had it
been more definite than "is might just work already". ;-)
Thanx, Paul
