From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1755247Ab0CCXqX (ORCPT ); Wed, 3 Mar 2010 18:46:23 -0500 Received: from mail-vw0-f46.google.com ([209.85.212.46]:38206 "EHLO mail-vw0-f46.google.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1753153Ab0CCXqQ (ORCPT ); Wed, 3 Mar 2010 18:46:16 -0500 DomainKey-Signature: a=rsa-sha1; c=nofws; d=gmail.com; s=gamma; h=sender:date:from:to:cc:subject:message-id:references:mime-version :content-type:content-disposition:in-reply-to:x-url:user-agent; b=KQvBWfXZ5cTM08O9vyiNrwx0Ee2zI4ddSjdBkcSMQag34HL2QZIcAvMYpTLOHu0AGs W9SV/sF9xjBwTMbf81zcFaAHiPjmcwjKONp+S3Hk9K9MNWCu1QWqaQPNQB0wO3pQSphG GSqYYPbldNEeRrwk8nWIMOwubM2SXGFpJJfpY= Date: Wed, 3 Mar 2010 20:46:10 -0300 From: Arnaldo Carvalho de Melo To: Eric B Munson Cc: a.p.zijlstra@chello.nl, paulus@samba.org, mingo@elte.hu, linux-kernel@vger.kernel.org, David Miller Subject: Re: [PATCH] Enable the enable_on_exec flag if record forks the target Message-ID: <20100303234610.GA17271@ghostprotocols.net> References: <1267653987-11022-1-git-send-email-ebmunson@us.ibm.com> MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Disposition: inline In-Reply-To: <1267653987-11022-1-git-send-email-ebmunson@us.ibm.com> X-Url: http://acmel.wordpress.com User-Agent: Mutt/1.5.19 (2009-01-05) Sender: linux-kernel-owner@vger.kernel.org List-ID: X-Mailing-List: linux-kernel@vger.kernel.org Em Wed, Mar 03, 2010 at 10:06:27PM +0000, Eric B Munson escreveu: > When forking its target, perf record can capture data from before the target > application is started. Perf stat uses the enable_on_exec flag in the event > attributes to keep from displaying events from before the target program > starts, this patch adds the same functionality to perf record when it is will > fork the target process. Good catch, Dave, this will help that case too, or at least avoid exercising that bug you fixed by flushing maps when a comm is received :-) Eric, I'll test your patches for handling multiple counters in 'perf report'and integrate then, most likely tomorrow, thanks! - Arnaldo Acked-by: Arnaldo Carvalho de Melo > Signed-off-by: Eric B Munson > --- > tools/perf/builtin-record.c | 13 ++++++++----- > 1 files changed, 8 insertions(+), 5 deletions(-) > > diff --git a/tools/perf/builtin-record.c b/tools/perf/builtin-record.c > index 771533c..625076e 100644 > --- a/tools/perf/builtin-record.c > +++ b/tools/perf/builtin-record.c > @@ -224,7 +224,7 @@ static struct perf_header_attr *get_header_attr(struct perf_event_attr *a, int n > return h_attr; > } > > -static void create_counter(int counter, int cpu, pid_t pid) > +static void create_counter(int counter, int cpu, pid_t pid, bool forks) > { > char *filter = filters[counter]; > struct perf_event_attr *attr = attrs + counter; > @@ -273,6 +273,9 @@ static void create_counter(int counter, int cpu, pid_t pid) > attr->inherit = inherit; > attr->disabled = 1; > > + if (forks) > + attr->enable_on_exec = 1; > + > try_again: > fd[nr_cpu][counter] = sys_perf_event_open(attr, pid, cpu, group_fd, 0); > > @@ -376,13 +379,13 @@ try_again: > ioctl(fd[nr_cpu][counter], PERF_EVENT_IOC_ENABLE); > } > > -static void open_counters(int cpu, pid_t pid) > +static void open_counters(int cpu, pid_t pid, bool forks) > { > int counter; > > group_fd = -1; > for (counter = 0; counter < nr_counters; counter++) > - create_counter(counter, cpu, pid); > + create_counter(counter, cpu, pid, forks); > > nr_cpu++; > } > @@ -542,10 +545,10 @@ static int __cmd_record(int argc, const char **argv) > > > if ((!system_wide && !inherit) || profile_cpu != -1) { > - open_counters(profile_cpu, target_pid); > + open_counters(profile_cpu, target_pid, forks); > } else { > for (i = 0; i < nr_cpus; i++) > - open_counters(i, target_pid); > + open_counters(i, target_pid, forks); > } > > if (file_new) { > -- > 1.6.3.3