From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S933901AbbENRiT (ORCPT ); Thu, 14 May 2015 13:38:19 -0400 Received: from mail-wi0-f172.google.com ([209.85.212.172]:37047 "EHLO mail-wi0-f172.google.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S933636AbbENRiN (ORCPT ); Thu, 14 May 2015 13:38:13 -0400 Date: Thu, 14 May 2015 19:38:08 +0200 From: Ingo Molnar To: Jiri Olsa Cc: Milos Vyletel , Peter Zijlstra , Paul Mackerras , Ingo Molnar , Arnaldo Carvalho de Melo , Masami Hiramatsu , Namhyung Kim , Jiri Olsa , He Kuang , Adrian Hunter , "open list:PERFORMANCE EVENT..." Subject: Re: [PATCH 2/2] perf/tools: put new buildid locks to use Message-ID: <20150514173807.GA19338@gmail.com> References: <1431588493-1474-1-git-send-email-milos@redhat.com> <1431588493-1474-3-git-send-email-milos@redhat.com> <20150514104059.GA27771@gmail.com> <20150514113821.GB1313@krava.redhat.com> MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Disposition: inline In-Reply-To: <20150514113821.GB1313@krava.redhat.com> User-Agent: Mutt/1.5.23 (2014-03-12) Sender: linux-kernel-owner@vger.kernel.org List-ID: X-Mailing-List: linux-kernel@vger.kernel.org * Milos Vyletel wrote: > On Thu, May 14, 2015 at 01:38:21PM +0200, Jiri Olsa wrote: > > > On Thu, May 14, 2015 at 12:40:59PM +0200, Ingo Molnar wrote: > > > > > > * Milos Vyletel wrote: > > > > > > > Use new read/write locks when accesing buildid directory on places where > > > > we may race if multiple instances are run simultaneously. > > > > > > Dunno, this will create locking interaction between multiple instances > > > of perf - hanging each other, etc. > > > > > > And it seems unnecessary: the buildid hierarchy is already spread out. > > > What kind of races might there be? > > > > there was just recently one fixed by commit: > > 0635b0f71424 perf tools: Fix race in build_id_cache__add_s() > > > > havent checked the final patch yet, but the idea is to > > protect us from similar bugs > > right. on top of race with EEXIST couple more are possible (EMLINK, > ENOSPC, EDQUOT, ENOMEM... the only way to prevent them all is to > lock this kind of operations and make sure we run one at a time. Yeah, so the race pointed out in 0635b0f71424 can be (and should be) fixed without locking: - first create the file under a process-private name under ~/.debug/tmp/ if the target does not exist yet - then fully fill it in with content - then link(2) it to the public target name, which VFS operation is atomic and may fail safely: at which point it got already created by someone else. - finally unlink() the private instance name and the target will now be the only instance left: either created by us, or by some other perf instance in the rare racy case. Since all of ~/.debug is on the same filesystem this should work fine. Beyond avoiding locking this approach has another advantage: it's transaction safe, so a crashed/interrupted perf instance won't corrupt the debug database, it will only put fully constructed files into the public build-id namespace. It at most leaves a stale private file around in ~/.debug/tmp/. Really, we should be following the example of Git, which is using a similar append-mostly flow to handle data, and generally avoids file locking as much as possible - which is a whole new can of worms. Thanks, Ingo