Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1760179Ab3EXJIz (ORCPT ); Fri, 24 May 2013 05:08:55 -0400 Received: from mx1.redhat.com ([209.132.183.28]:36324 "EHLO mx1.redhat.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1760064Ab3EXJIx (ORCPT ); Fri, 24 May 2013 05:08:53 -0400 Date: Fri, 24 May 2013 11:08:17 +0200 From: Jiri Olsa To: David Ahern Cc: acme@ghostprotocols.net, linux-kernel@vger.kernel.org, Mike Galbraith , Ingo Molnar , Frederic Weisbecker , Peter Zijlstra , Namhyung Kim , Stephane Eranian Subject: Re: [PATCH] perf record: handle death by SIGTERM Message-ID: <20130524090817.GA1020@krava.brq.redhat.com> References: <1367864663-1309-1-git-send-email-dsahern@gmail.com> MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Disposition: inline In-Reply-To: <1367864663-1309-1-git-send-email-dsahern@gmail.com> User-Agent: Mutt/1.5.21 (2010-09-15) Sender: linux-kernel-owner@vger.kernel.org List-ID: X-Mailing-List: linux-kernel@vger.kernel.org Content-Length: 2594 Lines: 70 On Mon, May 06, 2013 at 12:24:23PM -0600, David Ahern wrote: > perf data files cannot be processed until the header file is update > which is done via an on_exit handler. If perf is killed due to a SIGTERM > it does not run the on_exit hooks leaving the perf.data file in a > random state which perf-report will happily spin on trying to read. As > noted by Mike an easy reproducer is: > perf record -a -g & sleep 1; killall perf > > Fix by catching SIGTERM like it does SIGINT. Also need to remove the > kill which was added via commit f7b7c26e. > > Signed-off-by: David Ahern > Cc: Mike Galbraith > Cc: Arnaldo Carvalho de Melo > Cc: Ingo Molnar > Cc: Frederic Weisbecker > Cc: Peter Zijlstra > Cc: Jiri Olsa > Cc: Namhyung Kim > Cc: Stephane Eranian > --- > tools/perf/builtin-record.c | 2 +- > 1 file changed, 1 insertion(+), 1 deletion(-) > > diff --git a/tools/perf/builtin-record.c b/tools/perf/builtin-record.c > index cdf58ec..fff985c 100644 > --- a/tools/perf/builtin-record.c > +++ b/tools/perf/builtin-record.c > @@ -198,7 +198,6 @@ static void perf_record__sig_exit(int exit_status __maybe_unused, void *arg) > return; > > signal(signr, SIG_DFL); > - kill(getpid(), signr); > } > > static bool perf_evlist__equal(struct perf_evlist *evlist, > @@ -404,6 +403,7 @@ static int __cmd_record(struct perf_record *rec, int argc, const char **argv) > signal(SIGCHLD, sig_handler); > signal(SIGINT, sig_handler); > signal(SIGUSR1, sig_handler); > + signal(SIGTERM, sig_handler); > > if (!output_name) { > if (!fstat(STDOUT_FILENO, &st) && S_ISFIFO(st.st_mode)) > -- > 1.7.10.1 > hi, got a perf test hanging on me on latest acme's perf/core and bisected it to this one. I separated reproducer so far but probably wont get to it this week: [jolsa@krava2 perf]$ cat /proc/sys/kernel/perf_event_paranoid 1 [jolsa@krava2 perf]$ ./perf record -C 0 kill Error: You may not have permission to collect %sstats. Consider tweaking /proc/sys/kernel/perf_event_paranoid: -1 - Not paranoid at all 0 - Disallow raw tracepoint access for unpriv 1 - Disallow cpu events for unpriv 2 - Disallow kernel profiling for unpriv jirka -- 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/