lkml.org 
[lkml]   [2010]   [Jan]   [25]   [last100]   RSS Feed
Views: [wrap][no wrap]   [headers]  [forward] 
 
Messages in this thread
/
Date
From
SubjectRe: [Stable-review] [28/29] perf events: Dont report side-band events on each cpu for per-task-per-cpu events
On Sat, Jan 23, 2010 at 12:38:41PM +0100, Stefan Bader wrote:
> Greg KH wrote:
> > 2.6.32-stable review patch. If anyone has any objections, please let us know.
> >
> > ------------------
> >
> > From: Peter Zijlstra <a.p.zijlstra@chello.nl>
> >
> > commit 5d27c23df09b702868d9a3bff86ec6abd22963ac upstream.
> >
> > Acme noticed that his FORK/MMAP numbers were inflated by about
> > the same factor as his cpu-count.
> >
> > This led to the discovery of a few more sites that need to
> > respect the event->cpu filter.
> >
> > Reported-by: Arnaldo Carvalho de Melo <acme@ghostprotocols.net>
> > Signed-off-by: Peter Zijlstra <a.p.zijlstra@chello.nl>
> > Cc: Paul Mackerras <paulus@samba.org>
> > LKML-Reference: <20091217121830.215333434@chello.nl>
> > Signed-off-by: Ingo Molnar <mingo@elte.hu>
> > Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
> >
> > ---
> > kernel/perf_event.c | 20 ++++++++++++++++++--
> > 1 file changed, 18 insertions(+), 2 deletions(-)
> >
> > --- a/kernel/perf_event.c
> > +++ b/kernel/perf_event.c
> > @@ -1359,6 +1359,9 @@ static void perf_ctx_adjust_freq(struct
> > if (event->state != PERF_EVENT_STATE_ACTIVE)
> > continue;
> >
> > + if (event->cpu != -1 && event->cpu != smp_processor_id())
> > + continue;
> > +
> > hwc = &event->hw;
> >
> > interrupts = hwc->interrupts;
> > @@ -3226,6 +3229,9 @@ static void perf_event_task_output(struc
> >
> > static int perf_event_task_match(struct perf_event *event)
> > {
> > + if (event->cpu != -1 && event->cpu != smp_processor_id())
> > + return 0;
> > +
> > if (event->attr.comm || event->attr.mmap || event->attr.task)
> > return 1;
> >
>
> > @@ -3262,6 +3268,7 @@ static void perf_event_task_event(struct
> > ctx = rcu_dereference(task_event->task->perf_event_ctxp);
> > if (ctx)
> > perf_event_task_ctx(ctx, task_event);
> > + put_cpu_var(perf_cpu_context);
> > rcu_read_unlock();
> > }
>
> I believe this hunk drops the move of put_cpu_var. The upstream hunk looks like
> this:
>
> @ -3290,12 +3296,11 @@ static void perf_event_task_event(struct perf_task_event
> rcu_read_lock();
> cpuctx = &get_cpu_var(perf_cpu_context);
> perf_event_task_ctx(&cpuctx->ctx, task_event);
> - put_cpu_var(perf_cpu_context);
> -
> if (!ctx)
> ctx = rcu_dereference(task_event->task->perf_event_ctxp);
> if (ctx)
> perf_event_task_ctx(ctx, task_event);
> + put_cpu_var(perf_cpu_context);
> rcu_read_unlock();
> }

All fixed up now, thanks again for the review.

greg k-h


\
 
 \ /
  Last update: 2010-01-25 19:09    [W:0.057 / U:1.340 seconds]
©2003-2020 Jasper Spaans|hosted at Digital Ocean and TransIP|Read the blog|Advertise on this site