Re: x86_pmu_start WARN_ON.

[Peter Zijlstra <peterz@infradead.org>]

On Thu, Feb 20, 2014 at 07:47:23AM -0800, Andi Kleen wrote:
> Peter Zijlstra <peterz@infradead.org> writes:
> 
> > On Wed, Feb 19, 2014 at 05:34:49PM -0500, Vince Weaver wrote:
> >> So where would the NMI counter event get disabled?  Would it never get 
> >> disabled, just because it's always running and always gets the same fixed 
> >> slot?  Why isn't this a problem all the time, not just with corner cases?
> >
> > Well it could get another counter assignment I suppose; there's no
> > guarantee it always ends up in the fixed counter although that's rather
> > likely.
> >
> >> Is somehow n_added getting confused?
> >
> > I'd expect worse than a warning if that happened, but who knows.
> >
> > You could try with the below; make sure you've got CONFIG_FUNCTION_TRACER=y.
> >
> > Then (assuming debugfs is mounted in /debug ; most modern distros have
> > it at /sys/kernel/debug but they're wrong!):
> >
> > # echo 0 > /debug/tracing/tracing_on
> > # echo function > /debug/tracing/current_tracer
> > # echo nop > /debug/tracing/current_tracer
> > # echo 0 > /debug/tracing/trace
> > # echo 1 > /debug/tracing/tracing_on
> >
> > And make it go *BOOM*, then:
> >
> > # cat /debug/tracing/trace | bzip2 -9 > ~/trace.bz2
> 
> This won't work unless you apply this patch
> 
> http://lkml.org/lkml/2013/10/20/94

It will; trace_printk() works without -pg, I think you didn't read the
instructions very well.

And there's a very good reason not to apply your patch; you can route
the function tracer into perf, guess what happens when perf calls the
function tracer again :-)

But yes; you can use it if you don't do that. But I didn't want the
function trace, so who cares.