Re: [PATCH HACK RFC] cpu: Prevent late-arriving interrupts from disrupting offline

From: Paul E. McKenney
Date: Tue Jun 11 2019 - 15:30:37 EST


On Tue, Jun 11, 2019 at 04:39:34PM +0200, Dietmar Eggemann wrote:
> On 6/11/19 3:54 PM, Paul E. McKenney wrote:
> >On Tue, Jun 11, 2019 at 03:14:54PM +0200, Dietmar Eggemann wrote:
> >>On 6/8/19 6:41 PM, Paul E. McKenney wrote:
> >>>On Tue, Jun 04, 2019 at 03:29:32PM +0200, Dietmar Eggemann wrote:
> >>>>On 6/4/19 9:45 AM, Paul E. McKenney wrote:
> >>>>>On Mon, Jun 03, 2019 at 03:39:18PM +0200, Dietmar Eggemann wrote:
> >>>>>>On 6/3/19 1:44 PM, Mark Rutland wrote:
> >>>>>>>On Mon, Jun 03, 2019 at 10:38:48AM +0200, Peter Zijlstra wrote:
> >>>>>>>>On Sat, Jun 01, 2019 at 06:12:53PM -0700, Paul E. McKenney wrote:
> >>>>>>>>>Scheduling-clock interrupts can arrive late in the CPU-offline process,
>
> [...]
>
> >>Tested your patch on top of v5.2-rc4* on Arm TC2 (32bit) and CPU
> >>hotplug stress test. W/o your patch, the test fails within seconds
> >>since CPUs are not coming up again. W/ your patch, the test runs for
> >>hours just fine.
> >>
> >>You can add my:
> >>
> >>Tested-by: Dietmar Eggemann <dietmar.eggemann@xxxxxxx>
> >
> >Thank you!!!
> >
> >>* just for the record: one additional unrelated patch (to disable
> >>the NOR flash) is necessary on Arm TC2:
> >>https://patchwork.kernel.org/patch/10968391 .
> >
> >Is this progressing, or does it also need help getting to mainline?
>
> This is an unrelated specific issue w/ the TC2 platform which will
> progress independently. Other Arm32 platforms should profit from
> your patch independently of that. I just wanted to mention it here
> in case people try to recreate the test on this specific platform.
> >Left to myself, I will push my patch and assume that the NOR flash patch
> >will make it in its own good time -- or, alternatively, that there is
> >someone better positioned than me to push it.
>
> IMHO, the best thing is you push your patch.

I just now sent it. Here is hoping. ;-)

Thanx, Paul