Messages in this thread | | | Date | Fri, 19 Nov 2010 00:19:49 +0300 | From | Cyrill Gorcunov <> | Subject | Re: [V2 PATCH 0/6] x86, NMI: give NMI handler a face-lift |
| |
On Thu, Nov 18, 2010 at 04:02:31PM -0500, Don Zickus wrote: ... > > > Don, Robert, > > > > > > I still have suspicious on ours 'pending' nmi handler. Look what I mean -- > > > (keep in mind that p4 has a a way more counters than others). > > > > > > > To be precise -- it seems this scenario may force the back-to-back > > nmi handler to eat unknown nmi. > > That was the point of the change to do exactly that.
I missed the word, I meant to eat 'real' unknown nmi, not those generated by counters and stand 'in-fly', so that unknown_nmi_error() will be eaten. what is worse the NMI generated by kgdb may be eaten as well and treated as being 'pending' one, though there is quite a small probability of such situation I believe.
> > The problem is/was when you go to check to see if the period expired in > x86_perf_event_set_period(), you refresh the perf counter. The next step > is to see if the event period has expired, if so disabled the 'active' > bit. > > However, there is a race between when you refresh the counter to when you > actually disable it, such that you may cause the counter to overflow again > and thus generate another NMI. The whole ->running thing was implemented > by Robert to try and check for that condition and eat the NMI as we have > no intention of handling it (because it is bogus). > > The alternative is to use another rdmsrl to actually see if we trigger > another NMI. This was deemed a performance hit for such a small case. > > Cheers, > Don >
yeah, I recall now, thanks for refresh Don!
Cyrill
| |