Re: [PATCH] perf/record: make perf_event__synthesize_mmap_events() scale
From: Stephane Eranian
Date: Wed Mar 15 2017 - 22:34:40 EST
Arnaldo,
On Wed, Mar 15, 2017 at 10:44 AM, Arnaldo Carvalho de Melo
<acme@xxxxxxxxxx> wrote:
>
> Em Wed, Mar 15, 2017 at 10:08:27AM -0700, Stephane Eranian escreveu:
> > On Wed, Mar 15, 2017 at 6:58 AM, Arnaldo Carvalho de Melo
> > <acme@xxxxxxxxxx> wrote:
> > > Em Wed, Mar 15, 2017 at 10:50:59AM -0300, Arnaldo Carvalho de Melo escreveu:
> > >> So, fixing up the "tasks" -> "tasks" we end up with something safe and
> > >> that avoids this by now
> > >
> > > "tasks" -> "task", grrr
> > >
> > Oops, yeah, sorry about that.
> > Let me submit a V2 with the fix.
>
> Ok, but what about the other observations, what is the kernel you're
> using?
>
Ok, yeah, I was using an older kernel which still had [stack:tid] annotations.
With more recent kernel, it seems this optimization is not needed.
I guess it would only be useful to people running a new perf on a older kernel.
>
> I think this will not make any difference for any kernel >= 4.9, which
> isn't to say the patch shouldn't be applied, I'm just curious.
>
correct, see above.
>
> - Arnaldo