mirror of https://lore.kernel.org/lkml/
 help / color / mirror / Atom feed
From: Peter Zijlstra <a.p.zijlstra@chello.nl>
To: Ingo Molnar <mingo@elte.hu>
Cc: Brice Goglin <Brice.Goglin@inria.fr>,
	paulus@samba.org, LKML <linux-kernel@vger.kernel.org>
Subject: Re: [perf] howto switch from pfmon
Date: Tue, 23 Jun 2009 15:22:31 +0200	[thread overview]
Message-ID: <1245763351.19816.1779.camel@twins> (raw)
In-Reply-To: <20090623131450.GA31519@elte.hu>

On Tue, 2009-06-23 at 15:14 +0200, Ingo Molnar wrote:
> * Brice Goglin <Brice.Goglin@inria.fr> wrote:
> 
> > Hello,
> >
> > I am trying to play with perfcounters in current git (actually in
> > latest mmotm). I'd like to reproduce what I previously did with
> > pfmon, but I couldn't so far.
> >
> > Something like
> >     pfmon --follow-exec 'foobar' -e
> > CPU_TO_DRAM_REQUESTS_TO_TARGET_NODE:LOCAL_TO_0,CPU_TO_DRAM_REQUESTS_TO_TARGET_NODE:LOCAL_TO_1
> > -- <shell script>
> > gives the number of memory accesses to dram node #0 and #1 for all
> > processes whose name matches 'foobar'.
> >
> > So there are several questions here:
> > 1) is it possible to specify counter names like the above or do we have
> > to use raw counter numbers? I tried raw numbers from [1] without
> > success. How am I supposed to find and specify these raw numbers?
> > 2) how do we specify "subevents"?
> > 3) is there anything similar to --follow-exec, or --follow-pthreads for
> > getting separated outputs for each thread?
> >
> > I guess there are still a lot of things on the TODOlist but I'd 
> > like to understand a bit more where things are going. Sorry I 
> > didn't read all the archives about this, there are way too many of 
> > them recently :)
> 
> Yeah, there's indeed still a lot on the TODO list :-)
> 
> CPU_TO_DRAM_REQUESTS_TO_TARGET_NODE is a Barcelona hardware event, 
> so if you know that it maps to raw ID 0x100000e0 then you can always 
> extend the events that 'perf' knows about via raw events:
> 
>  $ perf stat -e cycles -e instructions -e r1000ffe0 ./hackbench 10
>  Time: 0.186
> 
>  Performance counter stats for './hackbench 10':
> 
>      4381248335  cycles
>      1964394846  instructions         #      0.448 IPC
>             838  raw 0x1000ffe0
> 
>     0.215382037  seconds time elapsed.

Just to clarify, The event code is 1E0h, and Ingo used a FFh unit mask.
These are combined using the arch masks below:

#define K7_EVNTSEL_EVENT_MASK   0x7000000FFULL
#define K7_EVNTSEL_UNIT_MASK    0x00000FF00ULL

to form the raw event code used: 0x1000ffe0


  reply	other threads:[~2009-06-23 13:22 UTC|newest]

Thread overview: 41+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2009-06-22 20:54 Brice Goglin
2009-06-23 12:12 ` Andi Kleen
2009-06-23 12:23   ` Peter Zijlstra
2009-06-23 13:57   ` Ingo Molnar
2009-06-23 13:14 ` Ingo Molnar
2009-06-23 13:22   ` Peter Zijlstra [this message]
2009-06-23 13:38     ` Ingo Molnar
2009-06-23 13:25   ` Ingo Molnar
2009-06-23 13:47   ` Ingo Molnar
2009-06-23 14:00     ` Brice Goglin
2009-06-23 14:36       ` Ingo Molnar
2009-06-23 15:22         ` Brice Goglin
2009-06-29 19:29           ` Ingo Molnar
2009-08-06 16:59             ` Brice Goglin
2009-08-06 17:40               ` Peter Zijlstra
2009-08-06 17:48                 ` Brice Goglin
2009-08-06 17:59                   ` Peter Zijlstra
2009-08-06 18:57                   ` [PATCH] perf tools: Fix reading of perf.data file header Peter Zijlstra
2009-08-06 19:03                     ` Brice Goglin
2009-08-06 19:59                       ` Ingo Molnar
2009-08-06 20:03                         ` Brice Goglin
2009-08-06 23:35                         ` Brice Goglin
2009-08-07  6:13                           ` Brice Goglin
2009-08-07  6:32                           ` Ingo Molnar
2009-08-07  7:38                             ` Brice Goglin
2009-08-07  7:45                               ` Ingo Molnar
2009-08-07  8:18                                 ` Brice Goglin
2009-08-07  8:23                                   ` Ingo Molnar
2009-08-07  8:27                                   ` Ingo Molnar
2009-08-07  8:30                                   ` [tip:perfcounters/core] perf stat: Rename -S/--scale to -c/--scale tip-bot for Brice Goglin
2009-08-07 11:55                             ` [PATCH] perf report: Display per-thread event counters Brice Goglin
2009-08-08 11:54                               ` [tip:perfcounters/core] perf report: Fix and improve the displaying of " tip-bot for Brice Goglin
2009-08-08 12:14                               ` [PATCH] perf report: Display " Ingo Molnar
2009-08-08 16:10                                 ` Brice Goglin
2009-08-08 16:13                                   ` Ingo Molnar
2009-08-07  6:37                     ` [tip:perfcounters/urgent] perf tools: Fix multi-counter stat bug caused by incorrect reading of perf.data file header tip-bot for Peter Zijlstra
2009-08-07  7:39                     ` tip-bot for Peter Zijlstra
2009-08-06 19:01                 ` [perf] howto switch from pfmon Brice Goglin
2009-06-23 14:21   ` Brice Goglin
2009-06-23 14:51     ` Ingo Molnar
2009-06-23 15:29       ` Jaswinder Singh Rajput

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=1245763351.19816.1779.camel@twins \
    --to=a.p.zijlstra@chello.nl \
    --cc=Brice.Goglin@inria.fr \
    --cc=linux-kernel@vger.kernel.org \
    --cc=mingo@elte.hu \
    --cc=paulus@samba.org \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox

all inboxes | Powered by JetHome®