Re: [RFC GIT PULL] perf/trace/lock optimization/scalability improvements

[Frederic Weisbecker <fweisbec@gmail.com>]

On Thu, Feb 04, 2010 at 08:40:47PM +0100, Jens Axboe wrote:
> On Wed, Feb 03 2010, Frederic Weisbecker wrote:
> > Ok, thanks a lot, the fact you can test on a 64 threads box is critically
> > helpful.
> > 
> > I also wonder what happens after this patch applied:
> > 
> > diff --git a/kernel/perf_event.c b/kernel/perf_event.c
> > index 98fd360..254b3d4 100644
> > --- a/kernel/perf_event.c
> > +++ b/kernel/perf_event.c
> > @@ -3094,7 +3094,8 @@ static u32 perf_event_tid(struct perf_event *event, struct task_struct *p)
> >         if (event->parent)
> >                 event = event->parent;
> >  
> > -       return task_pid_nr_ns(p, event->ns);
> > +       return p->pid;
> >  }
> > 
> > In my box it has increased the speed from 2x this patchset.
> 
> Doesn't seem to change anything, same runtime for a ls.


Ah.


 
> > I wonder if the tool becomes usable for you with that.
> > Otherwise, it means we have other things to fix, and
> > the result of:
> > 
> > 	perf record -g -f perf lock record sleep 6
> > 	perf report
> > 
> > would be very nice to have.
> 
> root@nehalem:/dev/shm # perf record -g -f perf lock record sleep 6
> [ perf record: Woken up 0 times to write data ]
> [ perf record: Captured and wrote 446.208 MB perf.data (~19495127 samples) ]
> [ perf record: Woken up 9 times to write data ]
> [ perf record: Captured and wrote 1.135 MB perf.data (~49609 samples) ]
> 
> It's huuuge. Thankfully the output isn't so big, I've attached it.



Yeah. I made a mistake, both perf record and perf lock are writing to the
same perf.data with this cmdline.

Actually we'd have needed:

perf record -o perflock.data -g -f perf lock record sleep 6
perf report -i perflock.data

That's why you don't have callchains in your profile, probably because the
file has been messed up.

Anyway, we should retry after my next iteration, I have various
reviews to address :)

Thanks.