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

Re: [Xen-devel] probe_intel_cpuid_faulting() behavior

  • To: Jan Beulich <JBeulich@xxxxxxxx>
  • From: "Liu, Jinsong" <jinsong.liu@xxxxxxxxx>
  • Date: Thu, 12 Sep 2013 17:12:43 +0000
  • Accept-language: en-US
  • Cc: xen-devel <xen-devel@xxxxxxxxxxxxxxxxxxxx>, Keir Fraser <keir@xxxxxxx>
  • Delivery-date: Thu, 12 Sep 2013 17:13:13 +0000
  • List-id: Xen developer discussion <xen-devel.lists.xen.org>
  • Thread-index: AQHOrtItj3n//s/qmkiGRDCjvqQrHZnCVRpA
  • Thread-topic: probe_intel_cpuid_faulting() behavior

Jan Beulich wrote:
> Jinsong,
> somehow it escaped my attention so far that this probing, on a CPU
> not supporting the respective MSR, necessarily produces a GPF log
> message. Since such instances are generally a reason to worry,
> would it be possible to at least avoid this for all CPU
> families/models below a certain level (the system I'm seeing this on
> is family 0xf model 6)? If so, could you provide precise cut-off
> values (I can of course easily enough create a patch myself once I
> have the 
> numbers)?
> Jan

Sorry, no. It's non-architectural but no official/precise family/model/stepping 
provided, so enumeration has to tolerate GPF log on legacy processors. What is 
more, no guarantee for future processors also support it.

Xen-devel mailing list



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