[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: [Xen-devel] SLES10 Big smp HPET domain_crash() c/s 13161

  • To: "Woller, Thomas" <thomas.woller@xxxxxxx>, <xen-devel@xxxxxxxxxxxxxxxxxxx>
  • From: Keir Fraser <Keir.Fraser@xxxxxxxxxxxx>
  • Date: Fri, 29 Dec 2006 15:32:30 +0000
  • Delivery-date: Fri, 29 Dec 2006 07:32:12 -0800
  • List-id: Xen developer discussion <xen-devel.lists.xensource.com>
  • Thread-index: Accq0SL2Dj7aZnwpTKGdoklUY7dbnAAYBqi3AArkIBAAAG/Kgw==
  • Thread-topic: [Xen-devel] SLES10 Big smp HPET domain_crash() c/s 13161

On 29/12/06 3:25 pm, "Woller, Thomas" <thomas.woller@xxxxxxx> wrote:

>> This should be easily reproducible. I'll take a look -- I
>> expect the hpet code just needs some 'loosening' in what it accepts.
> thanks. repo'ed on VT and AMD-V, but not tried on 32bpae hv.
> The BUG() in the SPT common.c code is interesting. FYI, I tried adding
> additional shadow_memory in guest config, but that didn't help. seems
> like a domain_crash() in the hpet code causes an eventual SPT BUG(), at
> least on my 2 test machines w/ 64bit hv.
>  -- Tom

Default action on domain crash is to restart/reboot the domain. And there
seems to be a shadow bug which currently guarantees that HVM domain reboot
will crash the host. I expect it'll be quickly fixed next week when Tim is

 -- Keir

Xen-devel mailing list



Lists.xenproject.org is hosted with RackSpace, monitoring our
servers 24x7x365 and backed by RackSpace's Fanatical Support®.