From mboxrd@z Thu Jan 1 00:00:00 1970 From: Andrew Cooper Subject: Re: HVM domains crash after upgrade from XEN 4.5.1 to 4.5.2 Date: Thu, 19 Nov 2015 01:06:16 +0000 Message-ID: <564D2088.3020604@citrix.com> References: <5644A248.1060505@web2web.at> <5644C1CD.3020202@citrix.com> <56451A2B.9090706@web2web.at> <56459E5F02000078000B4944@prv-mh.provo.novell.com> <5645B6BC.6030603@citrix.com> <56467D44.5040205@web2web.at> <56479A6B.6080102@citrix.com> <5647CE57.50209@web2web.at> <5648E727.6080204@cardoe.com> <56492BDF.5030208@web2web.at> <20151116153107.GD13720@char.us.oracle.com> <564A2B91.2090501@web2web.at> <564A6064.4080800@citrix.com> <564A626E.6010305@web2web.at> <564D00E0.8080004@web2web.at> <564D0720.9020506@citrix.com> <564D1875.2060601@web2web.at> Mime-Version: 1.0 Content-Type: text/plain; charset="us-ascii" Content-Transfer-Encoding: 7bit Return-path: In-Reply-To: <564D1875.2060601@web2web.at> List-Unsubscribe: , List-Post: List-Help: List-Subscribe: , Sender: xen-devel-bounces@lists.xen.org Errors-To: xen-devel-bounces@lists.xen.org To: Atom2 , Konrad Rzeszutek Wilk Cc: Doug Goldstein , xen-devel@lists.xen.org List-Id: xen-devel@lists.xenproject.org On 19/11/2015 00:31, Atom2 wrote: > Am 19.11.15 um 00:17 schrieb Andrew Cooper: >> On 18/11/2015 22:51, Atom2 wrote: >>> Am 17.11.15 um 00:10 schrieb Atom2: > [big snip] >>> Hi Andrew, >>> as promised I have again tried with a debug build and the results are >>> very mixed. I initially tried to better understand what the debug USE >>> flag actually does in gentoo and my understanding (after reading the >>> so called ebuilds) is now that the XEN hypervisor will be built by >>> adding a gcc option of "debug=y" - and that's what should compile a >>> debug build - right? >> Yes indeed. >> >>> So I went on and again enabled the debug USE flag plus gdb symbols and >>> rebuilt the hypervisor in the hope that this created a valid and >>> working debug build. >>> >>> It, however, seems there's another problem lurking somewhere which >>> only manifests itself when I boot from the debug build of the >>> hypervisor. >> You did manage to get at least one decent log from a properly >> debugbuild. >> >> However, all we need is the hvm_debug output. This patch: >> >> ---8<--- >> diff --git a/xen/include/asm-x86/hvm/support.h >> b/xen/include/asm-x86/hvm/support.h >> index 05ef5c5..7a8fbb5 100644 >> --- a/xen/include/asm-x86/hvm/support.h >> +++ b/xen/include/asm-x86/hvm/support.h >> @@ -28,7 +28,7 @@ >> >> #define HVM_DELIVER_NO_ERROR_CODE -1 >> >> -#ifndef NDEBUG >> +#if 1 >> #define DBG_LEVEL_0 (1 << 0) >> #define DBG_LEVEL_1 (1 << 1) >> #define DBG_LEVEL_2 (1 << 2) >> ---8<--- >> >> Will enable hvm_debug in a non-debug build of hypervisor. Can you try >> that please? > Done. Please find the xl dmesg output attached to this mail. I guess > this time it is really what you were expecting. Whether it does make > sense though, might be a different issue. But I am confident in your > abilities to figure out what's going on. Thanks! That is what I was looking for. Sadly, it is less useful than I was hoping. The guest is not appearing to do anything interesting which causes the bad state; it is almost a full second between the previous action of note, and the crash. Can you email me the bad HVMLoader binary please? ~Andrew