Re: [PATCH] perf, x86: try to handle unknown nmis with runningperfctrs

From: Don Zickus
Date: Wed Aug 11 2010 - 10:33:01 EST


On Wed, Aug 11, 2010 at 04:03:07PM +0200, Robert Richter wrote:
> On 11.08.10 08:44:43, Don Zickus wrote:
> > On Wed, Aug 11, 2010 at 01:10:46PM +0200, Robert Richter wrote:
> > > > > + *
> > > > > + * NOTE: We have no way of knowing if a second NMI was
> > > > > + * actually triggered, so we may accidentally skip a valid
> > > > > + * unknown nmi later.
> > > > > + */
> > > > > + __get_cpu_var(perfctr_skip) +=1;
> > >
> > > ... but this will not work. You have to mark the *absolute* nmi number
> > > here. If you only raise a flag, the next unknown nmi will be dropped,
> > > every. Because, in between there could have been other nmis that
> > > stopped the chain and thus the 'unknown' path is not executed. The
> > > trick in my patch is that you *know*, which nmi you want to skip.
> >
> > I guess I am confused. The way I read your patch was that you assumed the
> > next NMI would be the one you skip and if there was another NMI in between
> > the handled one and the one to skip, you would not skip it (nmi count !=
> > prev + 1) and it would produce an accidental unknown nmi.
>
> That's how it works.
>
> > I tried to change that with my patch by setting the skip flag which would
> > be drained on the next unknown nmi, independent of where it is in the NMI
> > backlog of NMIs.
>
> "setting the skip flag which would be drained on the next unknown nmi"
>
> That's what is wrong, it drops every unknown nmi, no matter when it is

Well as Frederic pointed out the skip variable will never go past one, so
it will only drop at most one unknown nmi.

> detected. In between there could be 1000's of valid other nmis
> handled. You even could have been returned from nmi mode. But still,
> the next unknown nmi will be dropped. Your patch could accumulate also

That was the intent. Can we guarantee that in the rare cases where the
perfctr is generating two nmis, that they will be back-to-back?

I think Huang tried to cap my approach even further my creating a time
window in which the two nmis had to happen. That gives us the flexibility
to handle nmis that are not back to back, but yet deal with the case where
two perfctrs fired but we are unsure if it generated a second nmi and we
falsely set the skip flag.

Cheers,
Don

--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at http://vger.kernel.org/majordomo-info.html
Please read the FAQ at http://www.tux.org/lkml/