From: Dmitri Monakhov Subject: Re: Problem with fsstress on powerpc Date: Wed, 9 Apr 2008 14:01:11 +0400 Message-ID: <20080409100111.GD944@dmon-lap.sw.ru> References: <20080408152455.GA10827@skywalker> <20080409074457.GB944@dmon-lap.sw.ru> <20080409082115.GA8238@skywalker> <20080409083341.GC944@dmon-lap.sw.ru> <20080409093934.GB8238@skywalker> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Cc: Eric Sandeen , Mingming Cao , ext4 development To: "Aneesh Kumar K.V" Return-path: Received: from mailhub.sw.ru ([195.214.232.25]:18635 "EHLO relay.sw.ru" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1754862AbYDIKJv (ORCPT ); Wed, 9 Apr 2008 06:09:51 -0400 Content-Disposition: inline In-Reply-To: <20080409093934.GB8238@skywalker> Sender: linux-ext4-owner@vger.kernel.org List-ID: On 15:09 Wed 09 Apr , Aneesh Kumar K.V wrote: > On Wed, Apr 09, 2008 at 12:33:41PM +0400, Dmitri Monakhov wrote: > > On 13:51 Wed 09 Apr , Aneesh Kumar K.V wrote: > > > On Wed, Apr 09, 2008 at 11:44:57AM +0400, Dmitri Monakhov wrote: > > > > On 20:54 Tue 08 Apr , Aneesh Kumar K.V wrote: > > > > > Hi All, > > > > > > > > > > Test case: fsstress with fallocate test case added. > > > > This may happens because of direct write. > > > > Please add exact fsstess command and options. > > > > > > > > > > > fsstress -c -n1000 -p1000 -d /mnt/tmp/ > > > > > > What i found was also related. The file which had errors had the below > > > operations on them. > > > > > > 82/901: dread d16/d20/d22/d2d/d63/d68/d9a/f93 [622592,53248] 0 > > > 82/939: dwrite d16/d20/d22/d2d/d63/d68/d9a/f93 [1093632,77824] 28 > > > > > > > > > 9/863: fallocate d1/d2d/d32/d8f/d97/d108/d9b/ffb [2041,75851] 28 > > > > > > > > > 18/733: fallocate d9/ddd/f86 [481025,18486] 0 > > > 18/735: dwrite d9/ddd/f86 [679936,36864] 0 > > > 18/764: dread d9/ddd/f86 [53248,8192] 0 > > > 18/910: dwrite d9/ddd/f86 [1454080,4096] 28 > > > 18/999: dwrite d9/ddd/f86 [1798144,61440] 28 > > > > > > > > > But what would happen during direct write that cause this ? > > Ohh it is long live bug in direct-io error handling logic: > > http://lkml.org/lkml/2006/12/18/103 > > So IMHO it is beter to disable direct-io writes in your case. > > Why is the patch not yet available in the Linus tree ? Because not everybody was happy about this solution. I'll send new version today. >Does this mean that i will not be able to test direct io write at all. Strictly saying YES, but practically you can use direct-io writes if you have enough disk space. >Right now i did > > #ifdef ENOSPC_BUG_DIRECT_IO_YES > { OP_DREAD, "dread", dread_f, 4, 0 }, > { OP_DWRITE, "dwrite", dwrite_f, 4, 1 }, > #endif > > i am not interested in dread, so commented that also. > > fsstress ran successfully with this change. > > -aneesh > > -- > To unsubscribe from this list: send the line "unsubscribe linux-ext4" in > the body of a message to majordomo@vger.kernel.org > More majordomo info at http://vger.kernel.org/majordomo-info.html