From mboxrd@z Thu Jan 1 00:00:00 1970 From: Konrad Rzeszutek Wilk Subject: Re: e008:[] check_lock+0x1b/0x45 Date: Wed, 19 Jun 2013 16:30:30 -0400 Message-ID: <20130619203030.GA18133@phenom.dumpdata.com> References: <20130619160622.GA11787@phenom.dumpdata.com> <51C1DE2D.2010409@citrix.com> <20130619170139.GA13046@phenom.dumpdata.com> <51C1E8EC.9050503@citrix.com> Mime-Version: 1.0 Content-Type: text/plain; charset="us-ascii" Content-Transfer-Encoding: 7bit Return-path: Content-Disposition: inline In-Reply-To: <51C1E8EC.9050503@citrix.com> List-Unsubscribe: , List-Post: List-Help: List-Subscribe: , Sender: xen-devel-bounces@lists.xen.org Errors-To: xen-devel-bounces@lists.xen.org To: Andrew Cooper Cc: "aravindp@cisco.com" , "xen-devel@lists.xensource.com" , "jbeulich@suse.com" , Roger Pau Monne List-Id: xen-devel@lists.xenproject.org On Wed, Jun 19, 2013 at 06:22:52PM +0100, Andrew Cooper wrote: > On 19/06/13 18:01, Konrad Rzeszutek Wilk wrote: > > On Wed, Jun 19, 2013 at 05:37:01PM +0100, Andrew Cooper wrote: > >> On 19/06/13 17:06, Konrad Rzeszutek Wilk wrote: > >>> Hey, > >>> > >>> It looks like one of the patches that went in Xen 4.1 blows up. > >>> > >>> This used to work over the weekend, so I can only surmise > >>> it is one the three patches. This is with a debug=y build > >>> (which has been working nicely for the last month or more). > >>> > >>> It looks like one of the patches that went in Xen 4.1 blows up. > >>> > >>> This used to work a day ago - that is c/s 23551 worked nicely. > > And just to confirm, yes the same kernel but with cs 23551 boots > > (just built it). Let me narrow down which of the c/s is at fault. > > > > There are only 3 changesets to choose from. It can only really be > > hg: 23552:d7ad3e2f1119 > git: df751b6da15afff1f87a68f63013dd96e9563047 > > "x86: fix ordering of operations in destroy_irq()". > > Is it possible for your to run tests with the same dom0 kernel but 4.2 > or unstable around these changesets? Unless it can be proved to only be > a problem with 4.1, I suspect we may want to back out the change, > especially given the proximity of 4.3 > I am not sure what is happening, but I redid the build and now I don't see the crash. I am going to dig in this a bit more but if I cannot reproduce this then please ignore me. > ~Andrew