Skip to content

Commit

Permalink
powerpc: perf: Fix frequency calculation for overflowing counters
Browse files Browse the repository at this point in the history
When profiling a benchmark that is almost 100% userspace, I noticed some wildly
inaccurate profiles that showed almost all time spent in the kernel.

Closer examination shows we were programming a tiny number of cycles into the
PMU after each overflow (about ~200 away from the next overflow). This gets us
stuck in a loop which we eventually break out of by throttling the PMU (there
are regular throttle/unthrottle events in the log).

It looks like we aren't setting event->hw.last_period to something same and the
frequency to period calculations in perf are going haywire.

With the following patch we find the correct period after a few interrupts and
stay there. I also see no more throttle events.

Signed-off-by: Anton Blanchard <anton@samba.org>
Acked-by: Benjamin Herrenschmidt <benh@kernel.crashing.org>
Cc: linuxppc-dev@lists.ozlabs.org
Cc: Paul Mackerras <paulus@samba.org>
Cc: Peter Zijlstra <a.p.zijlstra@chello.nl>
Cc: Arnaldo Carvalho de Melo <acme@ghostprotocols.net>
LKML-Reference: <20110117161742.5feb3761@kryten>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
  • Loading branch information
Anton Blanchard authored and Ingo Molnar committed Jan 17, 2011
1 parent 7c46d8d commit 4bca770
Showing 1 changed file with 1 addition and 0 deletions.
1 change: 1 addition & 0 deletions arch/powerpc/kernel/perf_event.c
Original file line number Diff line number Diff line change
Expand Up @@ -1212,6 +1212,7 @@ static void record_and_restart(struct perf_event *event, unsigned long val,
if (left <= 0)
left = period;
record = 1;
event->hw.last_period = event->hw.sample_period;
}
if (left < 0x80000000LL)
val = 0x80000000LL - left;
Expand Down

0 comments on commit 4bca770

Please sign in to comment.