Re: [PATCH v1 4/6] perf: Allow using AUX data in perf samples

[Alexander Shishkin <alexander.shishkin@linux.intel.com>]

Peter Zijlstra <peterz@infradead.org> writes:

> On Tue, Jun 19, 2018 at 01:47:25PM +0300, Alexander Shishkin wrote:
>> On Thu, Jun 14, 2018 at 10:20:22PM +0200, Peter Zijlstra wrote:
>
>> > More yuck...
>> > 
>> > You rreally should not be calling these pmu::methods, they're meant to
>> > be used from _interrupt_ not NMI context. Using them like this is asking
>> > for tons of trouble.
>> 
>> Right, the SW stuff may then race with event_function_call() stuff. Hmm.
>> For the HW stuff, I'm hoping that some kind of a sleight of hand may
>> suffice. Let me think some more.
>
> I currently don't see how the SW driven snapshot can ever work, see my
> comment on the last patch.

Unless we explicitly break clone when PERF_SAMPLE_AUX is set. Basically,
if you're asking for AUX samples, full perf context switch is not your
biggest performance penalty.

>> > Why can't you just snapshot the current location and let the thing
>> > 'run' ?
>> 
>> Because the buffer will overwrite itself and the location will be useless.
>
> Not if it's large enough ;-)
>
>> We don't write the AUX data out in this 'mode' at all, only the samples,
>> which allows for much less data in the resulting perf.data, less work for
>> the consumer, less IO bandwidth etc, and as a bonus, no AUX-related
>> interrupts.
>> 
>> But actually, even to snapshot the location we need to stop the event.
>
> Maybe new methods that should only be called from NMI context? 

True. Ideally, the existing ->start()/->stop()/interrupt should already
do all necessary internal bookkeeping.

Regards,
--
Alex