Re: [PATCH v2 0/2] Support of RO MMCFG access for PVH/HVMlite dom0s

[Boris Ostrovsky <boris.ostrovsky@oracle.com>]

On 12/18/2015 09:16 AM, Jan Beulich wrote:
>>>> On 18.12.15 at 14:53, <boris.ostrovsky@oracle.com> wrote:
>> On 12/18/2015 08:02 AM, Andrew Cooper wrote:
>>> On 18/12/15 10:11, Jan Beulich wrote:
>>>>>>> On 17.12.15 at 21:22, <boris.ostrovsky@oracle.com> wrote:
>>>>> * Left non-MMCFG RO accesses unhandled (we havent't encountered those
>> accesses
>>>>>     yet with PVH dom0 and it's probably better to fault on them and
>> investigate
>>>>>     whether they are guest's issues).
>>>> I seriously question this being a good approach,
>>> I concur.  All accesses should be terminated somehow, even if this is
>>> the hypervisor dropping writes and completing reads with ~0's.
>>>
>>> This is the expected behaviour from the x86 architecture.
>> OK, I'll add that in.
>>
>> To answer Jan's question, I did try 'vga=keep' and that worked fine.
> Interesting. Does your system not load/use a DRM driver for the
> card? Or is the driver so minimalistic that no extended config
> space write ever occurs?

It does load the driver:

0b:00.0 VGA compatible controller: Matrox Graphics, Inc. MGA G200e 
[Pilot] ServerEngines (SEP1) (rev 04)
     Subsystem: Intel Corporation Device 0101
     Kernel driver in use: mgag200
..
root@ovs104> lsmod | grep mgag
mgag200                45056  1
ttm                    90112  1 mgag200
drm_kms_helper        122880  1 mgag200
drm                   335872  4 mgag200,ttm,drm_kms_helper
i2c_algo_bit           16384  2 igb,mgag200

but I don't know what it does to the config space. I configured my 
kernel for a fairly minimal configuration so perhaps I disabled 
something that limits what the driver does.

-boris