From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: from dkim1.fusionio.com ([66.114.96.53]:45698 "EHLO dkim1.fusionio.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1752093Ab3GARS3 convert rfc822-to-8bit (ORCPT ); Mon, 1 Jul 2013 13:18:29 -0400 Received: from mx2.fusionio.com (unknown [10.101.1.160]) by dkim1.fusionio.com (Postfix) with ESMTP id 8AD277C035B for ; Mon, 1 Jul 2013 11:18:28 -0600 (MDT) Content-Type: text/plain; charset="utf-8" MIME-Version: 1.0 To: Zach Brown , Josef Bacik From: Chris Mason In-Reply-To: <20130701161002.GO10265@lenny.home.zabbo.net> CC: "linux-btrfs@vger.kernel.org" References: <1370384280-28652-1-git-send-email-zab@redhat.com> <20130701125435.GE4288@localhost.localdomain> <20130701161002.GO10265@lenny.home.zabbo.net> Message-ID: <20130701171826.14981.597@localhost.localdomain> Subject: Re: [PATCH 0/6] fix INT_MAX readdir hang, plus cleanups Date: Mon, 1 Jul 2013 13:18:26 -0400 Sender: linux-btrfs-owner@vger.kernel.org List-ID: Quoting Zach Brown (2013-07-01 12:10:02) > > > code. It's all lightly tested with xfstests but it wouldn't surprise > > > me if I missed something so review is appreciated. > > *mmm, hmmm* > > > One of these patches is making new entries not show up in readdir. This was > > discovered while running stress.sh overnight, it complained about files not > > matching but when they were checked the files matched. Dropping the entire > > series made stress.sh run fine. So I'm dropping these for the next merge window > > but I'll dig into it and try and figure out what was causing the problem. > > Nerts. It's got to be the delayed inode stuff. > > Maybe it's some unlink/recreate pattern? Is this a thing that stress.sh > does? (Where's stress.sh live?) It's an old namesys tool, I've copied it here: http://masoncoding.com/mason/tools/stress.sh My command line: stress.sh -n 50 -s -c /build/linux /mnt Basically its: 1) Make a list of all files in /build/linux and their md5sums 2) Start 50 procs 3) Each proc copies /build/linux into /mnt/stress/proc_num/ 4) Each proc compares the md5sums of its private copy with the original master 5) Each proc deletes the private copy 6) Repeat steps 2-5 forever. The most likely cause of the bug I'm seeing is readdir not finding new files. -chris