From mboxrd@z Thu Jan 1 00:00:00 1970 From: Andrew Cooper Subject: Re: HVM crash system on AMD APU A8-6600K Date: Mon, 3 Feb 2014 14:41:56 +0000 Message-ID: <52EFAAB4.4010607@citrix.com> References: <52EF9EEF.8050301@citrix.com> <52EFA304.4000206@citrix.com> <52EFA87A.2000008@citrix.com> Mime-Version: 1.0 Content-Type: text/plain; charset="us-ascii" 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 03/02/14 14:36, 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. > > Now trying to make log with watchdog added Don't bother - my suggestion of watchdog was for a hang, not a reset. Looking further through the logs, (XEN) APIC error on CPU1: 00(40) (XEN) APIC error on CPU0: 00(40) (XEN) APIC error on CPU2: 00(40) (XEN) APIC error on CPU3: 00(40) Show that "Received illegal vector" has been indicated on all cpus, beside a comment in the code which states "This interrupt should never happen with our APIC/SMP architecture" This appears to consistently happen after updating the PCI routing entries on the first IOAPIC. ~Andrew