From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: from dkim2.fusionio.com ([66.114.96.54]:50869 "EHLO dkim2.fusionio.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S933026Ab3FRQgl convert rfc822-to-8bit (ORCPT ); Tue, 18 Jun 2013 12:36:41 -0400 Received: from mx2.fusionio.com (unknown [10.101.1.160]) by dkim2.fusionio.com (Postfix) with ESMTP id 082239A03DD for ; Tue, 18 Jun 2013 10:36:40 -0600 (MDT) Content-Type: text/plain; charset="utf-8" MIME-Version: 1.0 To: Sage Weil From: Chris Mason In-Reply-To: CC: "linux-btrfs@vger.kernel.org" References: <20130611162129.4914.54461@localhost.localdomain> Message-ID: <20130618163637.9494.34639@localhost.localdomain> Subject: Re: hang on 3.9, 3.10-rc5 Date: Tue, 18 Jun 2013 12:36:37 -0400 Sender: linux-btrfs-owner@vger.kernel.org List-ID: Quoting Sage Weil (2013-06-18 11:56:37) > On Wed, 12 Jun 2013, Sage Weil wrote: > > On Tue, 11 Jun 2013, Chris Mason wrote: > > > Quoting Sage Weil (2013-06-11 11:43:30) > > > > I'm also seeing this hang regularly with both 3.9 and 3.10-rc5. Is this > > > > is a known problem? In this case there is no powercycling; just a regular > > > > ceph-osd workload. > > > > > > Everyone here is waiting for the root node, but it isn't immediately > > > clear who has the lock. log_one_extent is the most likely suspect, but > > > I can't see how it would be scheduling with the root lock held. > > > > > > Could you please sysrq-w? > > > > Sorry for the slow reply; had to wait for it to reproduce again. Attached > > both the original dmesg output and the blocked tasks. I'll leave the box > > wedged this time in case there is any other info that can help. > > Still seeing this hang with the latest from linus' tree. Is there another > tree I should try testing against? No, I don't see anything in -next that should fix this. Trying to reproduce here... -chris