Re: [PATCH RFC] VT-d: honor firmware-first mode in XSA-59 workaround code

[Andrew Cooper <andrew.cooper3@citrix.com>]

On 22/05/14 08:13, Jan Beulich wrote:
>>>> On 21.05.14 at 18:33, <andrew.cooper3@citrix.com> wrote:
>> On 21/05/14 17:09, Jan Beulich wrote:
>>> When firmware-first mode is being indicated by firmware, we shouldn't
>>> be modifying AER registers - these ar considered to be owned by
>> "these are"
>>
>>> firmware in that case. Violating this is being reported to result in
>>> SMI (or was it SCI?) storms.
>> SMM livelock.
> IOW "SMI storm".

It wasn't completely clear from the BIOS engineer's description whether
it was a single infinite loop in SMM mode, or an SMI storm, but the
overall effects are still the same.

>
>> The exact problem was a 2.6.32 kernel ignoring firmware-first and
>> disabling reporting on the root ports but not endpoint devices, causing
>> the SMM handler to fail to find the issue, and fail to clear the SMI.
>>
>> However, as part of the investigation, we identified that Xen was also
>> violating firmware-first as part of XSA-59
> So has a problem solely with the XSA-59 workaround (i.e. without an
> ignorant kernel) then been observed at all?
>
> Jan
>

Correct

~Andrew