Re: [PATCH 0/9] perf: Adding better precise_ip field handling

[Peter Zijlstra <peterz@infradead.org>]

On Mon, May 13, 2013 at 09:43:13PM +0200, Ingo Molnar wrote:
> 
> * Peter Zijlstra <peterz@infradead.org> wrote:
> 
> > On Sat, May 11, 2013 at 09:50:08AM +0200, Ingo Molnar wrote:
> > > That's really a red herring: there's absolutely no reason why the 
> > > kernel could not pass back the level of precision it provided.
> > 
> > All I've been saying is that doing random precision without feedback is 
> > confusing.
> 
> I agree with that.
> 
> > We also don't really have a good feedback channel for this kind of 
> > thing. The best I can come up with is tagging each and every sample with 
> > the quality it represents. I think we can do with only one extra 
> > PERF_RECORD_MISC bit, but it looks like we're quickly running out of 
> > those things.
> 
> Hm, how about passing precision back to user-space at creation time, in 
> the perf_attr data structure? There's no need to pass it back in every 
> sample, precision will not really change during the life-time of an event.

Ah indeed, we talked about modifying the attr structure before (error details
or so). Did something like that ever make it in, or would this be the first
use now?

> > But I think the biggest problem is PEBS's inability do deal with REP 
> > prefixes; see this email from Stephane: 
> > https://lkml.org/lkml/2011/2/1/177
> >
> > It is really unfortunate for PEBS to have such a side-effect; but it 
> > makes all memset/memcpy/memmove things appear like they have no cost. 
> > I'm very sure that will surprise a number of people.
> 
> I'd expect PEBS to get gradually better.
> 
> Note that at least for user-space, REP MOVS is getting rarer. libc uses 
> SSE based memcpy/memset variants - which is not miscounted by PEBS. The 
> kernel still uses REP MOVS - but it's a special case because it cannot 
> cheaply use vector registers.

What's the rep_good cpu feature flag for? I thought Intel was putting more
effort into making REP MOVS doing the right thing again. No need to worry your
pretty head about the best way to move bytes around any longer.

> The vast majority of code gets measured by cycles:pp more accurately than 
> cycles.
> 
> We could try and see how many people complain. It's not like it's hard to 
> undo such a change of the default event?

I suppose so.. Alternatively we can have the PEBS event read a 'real' cycles
counter and weight the sample based on that. Bit cumbersome, esp if you want to
implement it kernel side, but it could possibly work around this issue.