Messages in this thread | | | Date | Wed, 29 Jun 2011 16:10:56 +0200 | From | Robert Richter <> | Subject | Re: [tip:perf/core] perf: Ignore non-sampling overflows |
| |
On 29.06.11 06:50:55, Peter Zijlstra wrote: > On Wed, 2011-06-29 at 12:37 +0200, Robert Richter wrote: > > > I looked at the interrupt handlers. The events are always determined > > from a per-cpu array: > > > > cpuc = &__get_cpu_var(cpu_hw_events); > > ... > > event = cpuc->events[idx]; > > > > In case of interrupts the event should then always be a hw event (or > > uninitialized). Even if the interrupt was triggered by a different > > source, it would always be mapped to the same event and the check > > is_sampling_event() would be meaningless. > > I'm probably not quite getting what you mean, but how is > is_sampling_event() meaningless? the INT bit is enabled for _all_ > events, whether they were configured as a sampling event or not.
Aren't all events that are mapped to counters via cpu_hw_events always sampling events? Then, when calling perf_event_overflow() from an interrupt handler there are no other events than sampling events.
> > Its just that for !sampling events we shouldn't attempt to generate any > output.
If attr.sample_type is null, there is no output to generate. Better use this instead of attr.sample_type in is_sampling_event()? perf_event_overflow() could be used then to generate output also for samples where no period is specified.
> > > There are other occurrences of perf_event_overflow() in > > kernel/events/core.c for events of type PERF_TYPE_SOFTWARE. These > > events are initialized with sample_period set and a check would always > > be true too. > > I'm failing to see what you mean, where do we always set > event->attr.sample_period for software events?
Hmm, I read the code wrong and the check in perf_event_overflow() might be needed for swevents.
> > > For both cases I stil don't see a reason for the check. > > You're going to have to spell things out for me, I'm really not getting > your argument.
I was thinking about to change this check and haven't seen cases for that the check is for. What would happen if the check isn't there and perf_event_overflow() is called from the interrupt handler?
> > > Anyway, would the following extentension of the check above ok? > > > > if (unlikely(!is_sampling_event(event) && !event->attr.sample_type)) > > ... > > > > With no bits set in attr.sample_type the sample would be empty and > > nothing to report. Now, with this change, samples that have data to > > report wouldn't be dropped anymore. > > Also, could you explain in what way data is dropped? Where do > non-sampling events need to write sample data?
I stumbled over this while rebasing my perf ibs patches:
http://git.kernel.org/?p=linux/kernel/git/rric/oprofile.git;a=shortlog;h=refs/heads/perf-ibs
Hope I could explain this to you better now.
-Robert
-- Advanced Micro Devices, Inc. Operating System Research Center
| |