Re: HVM crash system on AMD APU A8-6600K

[Vitaliy Tomin <highwaystar.ru@gmail.com>]

I've tested system with memtest and it show no issue.

On Wed, Apr 9, 2014 at 10:43 AM, Vitaliy Tomin <highwaystar.ru@gmail.com> wrote:
>>Are you by any chance running your machine at 'overclocked' mode?
>
> Machine is not overckocked and not overheated. It is looks stable all
> the time expect when I run hvm guest with no debug xen. Debug xen runs
> hvm guests with no issues.
>
> There is other reporter with AMD APU A10-6800 with same issue
> http://lists.xen.org/archives/html/xen-devel/2013-08/msg01395.html
>
>>- Doing a memtest with memtest86+ for example to see if everything is ok ?
>>- Do some CPU burnin/stresstest on baremetal linux / dom0 ?
>
> I will try memtest and stresstests later today to be sure.
>
> On Wed, Apr 9, 2014 at 1:17 AM, Konrad Rzeszutek Wilk
> <konrad.wilk@oracle.com> wrote:
>> On Tue, Apr 08, 2014 at 06:10:46PM +0200, Sander Eikelenboom wrote:
>>>
>>> Tuesday, April 8, 2014, 5:29:06 PM, you wrote:
>>>
>>> >> Hmm from the logs you seem to have a serial console attached,
>>> >> perhaps wise to also add a "noreboot" to the xen line in grub so it doesn't automatically
>>> >> reboot on crash. There is a small chance it will print something usefull then.
>>> >>
>>> > I've tried noreboot option with last log, my system still reboots with
>>> > no notice.
>>>
>>> Please keep people who try to help you CC'ed on replies, otherwise you probably won't
>>> get much response (added Jan again).
>>>
>>> So that seems to imply it's not a Xen induced reboot.
>>
>> Are you by any chance running your machine at 'overclocked' mode?
>>
>> That bit me with unexplained reboots and crashes.
>>>
>>> >> Another thing to check could be if you are running the latest bios, since the logs
>>> >> do report problems with setting up CPU's (APIC and MTRR warnings).
>>>
>>> > I'm running latest bios.
>>>
>>> OK, have you also done some basic other checks to try to rule out hardware?
>>>
>>> - Doing a memtest with memtest86+ for example to see if everything is ok ?
>>> - Do some CPU burnin/stresstest on baremetal linux / dom0 ?
>>>
>>>
>>> >> medium == a disk specified in the config file with "disk=" ?
>>> >>
>>> >> The automatic restarts are by a script by you ?
>>> >> From http://wiki.xen.org/wiki/Xen_Configuration_File_Options#HVM:
>>> >> on_crash
>>> >>     Behavior when a domain exits with reason 'crash'
>>>
>>> > Yes config without "disk=". All hvm config is next 3 lines:
>>> > name = 'blank'
>>> > builder = 'hvm'
>>> > memory = 1024
>>>
>>>
>>>
>>>
>>> _______________________________________________
>>> Xen-devel mailing list
>>> Xen-devel@lists.xen.org
>>> http://lists.xen.org/xen-devel