Re: [PATCHv2 00/14] perf diff: Factor diff command

2012-10-03 Thread Arnaldo Carvalho de Melo
Em Wed, Oct 03, 2012 at 06:53:42PM +0200, Andi Kleen escreveu: > The problem is with the perf-archive script. After some hacking I got it > running > now. I suspect the perf tool needs to pass its own executable name to > the script to run. > > Also would be nice if another file than perf.data co

Re: [PATCHv2 00/14] perf diff: Factor diff command

2012-10-03 Thread Jiri Olsa
On Wed, Oct 03, 2012 at 06:55:32PM +0200, Andi Kleen wrote: > > > hm, I missed this one.. hopefully it should be no problem to add it > > > > Here's more issues I found (again no show stopper): > > And another issue (this one is more serious): > > I was trying to compare the same program with tw

Re: [PATCHv2 00/14] perf diff: Factor diff command

2012-10-03 Thread Andi Kleen
> > hm, I missed this one.. hopefully it should be no problem to add it > > Here's more issues I found (again no show stopper): And another issue (this one is more serious): I was trying to compare the same program with two different kernels. But perf diff does not recognize the same functions i

Re: [PATCHv2 00/14] perf diff: Factor diff command

2012-10-03 Thread Andi Kleen
On Wed, Oct 03, 2012 at 06:18:16PM +0200, Andi Kleen wrote: > On Wed, Oct 03, 2012 at 06:47:57AM -0700, Arnaldo Carvalho de Melo wrote: > > Em Tue, Oct 02, 2012 at 06:30:02PM +0200, Andi Kleen escreveu: > > > The other problem I ran into is that perf archive doesn't seem to work > > > very > > > w

Re: [PATCHv2 00/14] perf diff: Factor diff command

2012-10-03 Thread Andi Kleen
On Wed, Oct 03, 2012 at 06:47:57AM -0700, Arnaldo Carvalho de Melo wrote: > Em Tue, Oct 02, 2012 at 06:30:02PM +0200, Andi Kleen escreveu: > > The other problem I ran into is that perf archive doesn't seem to work very > > well with kernels, so it's hard to move profiles from different kernels > >

Re: [PATCHv2 00/14] perf diff: Factor diff command

2012-10-03 Thread Arnaldo Carvalho de Melo
Em Tue, Oct 02, 2012 at 06:30:02PM +0200, Andi Kleen escreveu: > The other problem I ran into is that perf archive doesn't seem to work very > well with kernels, so it's hard to move profiles from different kernels > around to diff them (e.g. for a performance regression) One way around this > woul

Re: [PATCHv2 00/14] perf diff: Factor diff command

2012-10-02 Thread Andi Kleen
> > Some minor issues I ran into (but no show stoppers): > > - The error messages for bad -c expressions could be better > > - I found the requirement for no space after -c unintuitive. > > I'll see to that > > > - It would be nice to have support for doing the bucketizing per line > > instead of

Re: [PATCHv2 00/14] perf diff: Factor diff command

2012-10-01 Thread Jiri Olsa
On Thu, Sep 27, 2012 at 11:31:02PM +0200, Andi Kleen wrote: > On Thu, Sep 27, 2012 at 01:09:21PM +0200, Jiri Olsa wrote: > > hi, > > this is v2 of diff command changes proposed in here: > > https://lkml.org/lkml/2012/9/6/344 > > > > It's now rebased on new hists hpp* interface plus few > > more ad

Re: [PATCHv2 00/14] perf diff: Factor diff command

2012-09-27 Thread Andi Kleen
On Thu, Sep 27, 2012 at 01:09:21PM +0200, Jiri Olsa wrote: > hi, > this is v2 of diff command changes proposed in here: > https://lkml.org/lkml/2012/9/6/344 > > It's now rebased on new hists hpp* interface plus few > more additional changes. FWIW I've been playing around with it a bit. It seems u

[PATCHv2 00/14] perf diff: Factor diff command

2012-09-27 Thread Jiri Olsa
hi, this is v2 of diff command changes proposed in here: https://lkml.org/lkml/2012/9/6/344 It's now rebased on new hists hpp* interface plus few more additional changes. I'm still testing/using it to come up with some proved usability, but would be nice to have it in anyway. v2 changes: - bas