Re: [PATCH 0/3] perf, bts: Fallout from the fuzzer for perf/urgent

[Vince Weaver]

On Tue, 23 Aug 2016, Alexander Shishkin wrote:

> Vince Weaver <vince@xxxxxxxxxx> writes:
> 
> > On Tue, 23 Aug 2016, Alexander Shishkin wrote:
> >
> >> Recently Vince has reported warnings and panics coming from the
> >> general direction of AUX tracing. I found two bugs which manifest
> >> similarly, one in intel_bts driver and one in AUX unmapping path.
> >> 
> >> Both are triggered by racing SET_OUTPUT against mmap_close while
> >> running AUX tracing. I have a test case that set fire to the kernel
> >> within a few seconds by doing this, which I can share if anyone
> >> cares.
> >
> > I've applied these and am fuzzing on my Haswell machine and I'm getting 
> > lots and lots of
> >
> > [  945.536374] Dazed and confused, but trying to continue
> > [  945.542391] Uhhuh. NMI received for unknown reason 21 on CPU 2.
> > [  945.549360] Do you have a strange power saving mode enabled?
> > [  945.556701] Dazed and confused, but trying to continue
> > [  945.563643] Uhhuh. NMI received for unknown reason 31 on CPU 2.
> > [  945.571379] Do you have a strange power saving mode enabled?
> > [  945.578754] Dazed and confused, but trying to continue
> > [  945.630889] Uhhuh. NMI received for unknown reason 31 on CPU 2.
> > [  945.638141] Do you have a strange power saving mode enabled?
> > [  945.645523] Dazed and confused, but trying to continue
> >
> > messages that I wasn't getting before.  Of course, I updated from -rc2 to 
> > -rc3 before applying the patch so it could also be related to that too, I 
> > can try backing out the patches and see if it still happens.
> 
> Uhhuh, that's not cool, let me look some more. But it should stop dying
> on bts/aux stuff at least.

Just wanted to say that when I backed out these patches the NMI errors 
went away, so they do seem to be a result of this patch series.

Vince