From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1752741Ab1IYNeu (ORCPT ); Sun, 25 Sep 2011 09:34:50 -0400 Received: from mx1.redhat.com ([209.132.183.28]:31742 "EHLO mx1.redhat.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1752649Ab1IYNet (ORCPT ); Sun, 25 Sep 2011 09:34:49 -0400 Date: Sun, 25 Sep 2011 15:34:06 +0200 From: Jiri Olsa To: Frederic Weisbecker Cc: Arnaldo Carvalho de Melo , Eric Dumazet , a.p.zijlstra@chello.nl, mingo@elte.hu, paulus@samba.org, linux-kernel@vger.kernel.org, rostedt@goodmis.org, Neil Horman Subject: Re: [PATCH] perf tools: Fix tracing info recording Message-ID: <20110925133406.GB2702@jolsa.brq.redhat.com> References: <1314022997-9217-1-git-send-email-jolsa@redhat.com> <1314023913.2307.63.camel@edumazet-HP-Compaq-6005-Pro-SFF-PC> <20110822145210.GA8694@jolsa.brq.redhat.com> <1314028266.2307.93.camel@edumazet-HP-Compaq-6005-Pro-SFF-PC> <20110822160713.GA10134@jolsa.brq.redhat.com> <20110829132037.GB27261@ghostprotocols.net> <20110829134147.GD1918@jolsa.redhat.com> <20110829142547.GC27261@ghostprotocols.net> <20110914135840.GC2719@jolsa.brq.redhat.com> <20110921153016.GB1811@somewhere> MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Disposition: inline In-Reply-To: <20110921153016.GB1811@somewhere> 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 On Wed, Sep 21, 2011 at 05:30:24PM +0200, Frederic Weisbecker wrote: > On Wed, Sep 14, 2011 at 03:58:40PM +0200, Jiri Olsa wrote: > > The tracing information is part of the perf data file. It contains > > several files from within the tracing debugfs and procs directories. > > > > Beside some static header files, for each tracing event the format > > file is added. The /proc/kallsyms file is also added. > > > > The tracing data are stored with preceeding size. This is causing some > > dificulties for pipe output, since there's no way to tell debugfs/proc > > file size before reading it. So, for pipe output, all the debugfs files > > were read twice. Once to get the overall size and once to store the > > content itself. This can cause problem in case any of these file > > changed, within the storage time. > > > > Fixing this behaviour by using temp file in case of pipe output. The > > debugfs/proc files are being read only once, ensuring the integrity of > > the tracing data. > > > > Also changing the way the event files are searched by quering specified > > event files directly, instead of walking the events directory. > > > > Signed-off-by: Jiri Olsa > > Looks good overall, but I have some comments about details: hi, thanks for comments, I'll make the changes and send out new version soon jirka