Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1751393Ab2FETEI (ORCPT ); Tue, 5 Jun 2012 15:04:08 -0400 Received: from mail-lpp01m010-f46.google.com ([209.85.215.46]:46095 "EHLO mail-lpp01m010-f46.google.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1751087Ab2FETEG convert rfc822-to-8bit (ORCPT ); Tue, 5 Jun 2012 15:04:06 -0400 MIME-Version: 1.0 In-Reply-To: <4FCE3DE3.9060805@gmail.com> References: <4FCE25C6.2080601@gmail.com> <1338912052.2749.2.camel@twins> <1338912312.2749.6.camel@twins> <4FCE3DE3.9060805@gmail.com> Date: Tue, 5 Jun 2012 21:04:04 +0200 Message-ID: Subject: Re: git bisect and perf From: Stephane Eranian To: David Ahern Cc: Peter Zijlstra , Ingo Molnar , LKML , Arnaldo Carvalho de Melo Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8BIT X-System-Of-Record: true Sender: linux-kernel-owner@vger.kernel.org List-ID: X-Mailing-List: linux-kernel@vger.kernel.org Content-Length: 2049 Lines: 72 On Tue, Jun 5, 2012 at 7:12 PM, David Ahern wrote: > On 6/5/12 10:05 AM, Peter Zijlstra wrote: >> >> On Tue, 2012-06-05 at 18:00 +0200, Peter Zijlstra wrote: >>> >>> On Tue, 2012-06-05 at 09:29 -0600, David Ahern wrote: >>>> >>>> I am trying find out when (ie. which commit) the pebs feature >>>> actually >>>> started working on a westmere system >>> >>> >>> What do you mean with working? The whole cycles:pp thing is magic and >>> unrelated to 'regular' PEBS stuff. > > > :p requires PEBS? > > >> .35 would indeed be the version we >>> >>> introduced PEBS support and it should actually work. > > > I'm referring to cycles:p > > model name      : Intel(R) Xeon(R) CPU           E5620  @ 2.40GHz That's not as helpful as /proc/cpuinfo. I suspect this is a Westmere system, isn't it? > Fedora 14, 2.6.35.14-106.fc14.x86_64 (few perf_event* differences from > 2.6.35.14). > > # perf record -e cycles:p -v -- sleep 1 > >  Warning:  ... trying to fall back to cpu-clock-ticks > > ... > > Cycles is supported; it's the :p that is not. perf userspace only recently > gained the knowledge for discriminating the two. Using perf from > acme/perf/urgent: > > # /tmp/pbuild/perf record -e cycles:p -v -- sleep 1 > >  Error: sys_perf_event_open() syscall returned with 28 (No space left on > device).  /bin/dmesg may provide additional information. > > I chased this to intel_pebs_constraints() returning &emptyconstraint which > has a weight of 0 (validate_event -> intel_get_event_constraints -> > intel_pebs_constraints). > > > >>> >>> But if you're looking for the cycles:pp stuff (commit 7639dae0ca1) >>> that's .38-rc1-ish. >> >> >> So on kernels pre that commit you could still do: >> >>   perf record -e r108000c0:pp > > > Indeed that works on the F14 kernel. > > David -- To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to majordomo@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html Please read the FAQ at http://www.tux.org/lkml/