On Thu, Nov 17, 2022 at 10:39 PM Frederic Weisbecker
<frederic@xxxxxxxxxx> wrote:
>
> On Mon, Nov 07, 2022 at 08:07:26AM -0800, Paul E. McKenney wrote:
> > > I ran 200 hours of TREE04 and got an RCU CPU stall warning. I ran 2000
> > > hours on v6.0, which precedes these commits, and everything passed.
> > >
> > > I will run more, primarily on v6.0, but that is what I have thus far.
> > > At the moment, I have some concerns about this change.
> >
> > OK, so I have run a total of 8000 hours on v6.0 without failure. I have
> > run 4200 hours on rcu#revert_tick_dep with 15 failures. The ones I
> > looked at were RCU CPU stall warnings with timer failures.
> >
> > This data suggests that the kernel is not yet ready for that commit
> > to be reverted.
>
> But that branch has the three commits reverted:
>
> 1) tick: Detect and fix jiffies update stall
> 2) timers/nohz: Last resort update jiffies on nohz_full IRQ entry*
> 3) rcu: Make CPU-hotplug removal operations enable tick
>
> Reverting all of them is expected to fail anyway.
>
> What we would like to know is if reverting just 3) is fine. Because
> 1) and 2) are supposed to fix the underlying issue.
>
> I personally didn't manage to trigger failures with just reverting 3)
> after thousands hours. But it failed with reverting all of them.
>
> Has someone managed to trigger a failure with only 3) reverted?
>
Oh, a long long thread hides the history! Once it was done [1]. But
I have not kept the data (IIRC, it occupied ~1GB).
If the information and attachment in [1] are not enough for you, I am
happy to redo the test.
Thanks,
Pingfan
[1]: https://lore.kernel.org/all/CAFgQCTtLm-JXRyQfKo6-+P00SShVGujZGau+khmtCe1AiRodQA@xxxxxxxxxxxxxx/
> Thanks.
