From mboxrd@z Thu Jan 1 00:00:00 1970 From: Boris Ostrovsky Subject: Re: HVM crash system on AMD APU A8-6600K Date: Mon, 03 Feb 2014 10:47:06 -0500 Message-ID: <52EFB9FA.7010905@oracle.com> References: <52EF9EEF.8050301@citrix.com> <52EFA304.4000206@citrix.com> <52EFA87A.2000008@citrix.com> <20140203145503.GA3864@phenom.dumpdata.com> Mime-Version: 1.0 Content-Type: text/plain; charset="us-ascii"; Format="flowed" Content-Transfer-Encoding: 7bit Return-path: In-Reply-To: List-Unsubscribe: , List-Post: List-Help: List-Subscribe: , Sender: xen-devel-bounces@lists.xen.org Errors-To: xen-devel-bounces@lists.xen.org To: Vitaliy Tomin Cc: xen-devel@lists.xensource.com List-Id: xen-devel@lists.xenproject.org On 02/03/2014 10:18 AM, Vitaliy Tomin wrote: >> You might want to add 'sync_console' on your Xen line. That should > give you a bit more of data (I hope?) > > Here is log captured with sync_console and empty hvm config. > (XEN) SR-IOV device 0000:00:11.0 has its virtual functions already enabled (01ab) 11.0 is the SATA controller on FCH and I don't believe it's an SR-IOV device. I dont' think it even has extended config space. -boris > > On Mon, Feb 3, 2014 at 11:55 PM, Konrad Rzeszutek Wilk > wrote: >> On Mon, Feb 03, 2014 at 11:36:23PM +0900, Vitaliy Tomin wrote: >>>> What about without PCI passthrough? In all cases you appear to be >>> passing the embedded graphics through to an HVM domain >>> >>> Next log captured with following hvm domain config: >>> >>> name = 'blank' >>> builder = 'hvm' >>> memory = 1024 >>> >>> Only 3 lines. It takes longer to crash, about 5-10 minutes, with OS in >>> hvm it takes less than a minute. >> You might want to add 'sync_console' on your Xen line. That should >> give you a bit more of data (I hope?) >>> Now trying to make log with watchdog added >>> >>> On Mon, Feb 3, 2014 at 11:32 PM, Andrew Cooper >>> wrote: >>>> Can you please reply-to-all to keep this on the list. >>>> >>>> On 03/02/14 14:21, Vitaliy Tomin wrote: >>>>>> Does it start with debug=n, but without trying to passthrough the pci device (the graphics core of the apu?) to the hvm ? >>>>> No it crashes even with empty HVM domain (no os, no disk images no network) >>>> What about without PCI passthrough? In all cases you appear to be >>>> passing the embedded graphics through to an HVM domain >>>> >>>>>> Can you explain "=== whole system crashed ===" a little more. >>>>> II means system instantly rebooted. Black screen, no messages, no >>>>> image on screen next what I see is POST of my real hardware. >>>>> >>>>> Log of xen runned with debug=y attached,hvm dom runned and no crash. >>>> Ok - so something forced a system reset. Even more curious that debug >>>> mode is fine while non-debug is fatal. >>>> >>>> ~Andrew >>>> >>>> >>