[vfio-users] Guest does not boot OS with GPU passed through, stuck at TianoCore boot splash

[Stefan Seil]

As far as I can tell, the other PCI device is the HDMI audio controller 
of the graphics card. It is listed as a seperate PCI device and I 
assigned it to the VM, too.
I haven't explicitly added any USB devices, although it seems like there 
are some per default. I have linked the respective domain XMLs below.
I have played around with the configuration a bit and removed the rom 
file for the HDMI audio controller of the graphics card (previously I 
had given it the same rom file as the one I passed to the GPU) and now 
the VM actually does not stop at the TianoCore logo anymore. The monitor 
now displays the GRUB menu and turns black after choosing the first boot 
entry. This is the new log now: http://pastebin.com/2dxBqP6h

I have also configured the VM with emulated graphics using QXL (just the 
standard configuration you get with the setup wizard). If I try to boot 
with the GPU passed through in addition to having the emulated graphics, 
I can see it progressing further on the emulated display. It looks like 
it starts booting the OS and then stops at this line:

intel_rapl: no valid rapl domains found in package 0

I don't see any of that on the monitor connected to the graphics card, 
though.


These are the two domain XMLs now:
Without additional emulated graphics: http://pastebin.com/hBTdP6Sq
With additional emulated graphics: http://pastebin.com/Wm3E23nW

On Mon, Apr 11, 2016 at 10:15 PM, Laszlo Ersek wrote:
> On 04/11/16 22:11, Laszlo Ersek wrote:
>> On 04/11/16 21:48, Stefan Seil wrote:
>>> Thank you for the detailed explanation!
>>> I did this and have uploaded the log here: http://pastebin.com/BXU9UqiG
>>> I don't really see any apparent errors in there...
>> Right, I expected a failed ASSERT() or something similar, but I'm not
>> seeing a direct error anywhere.
>>
>> What I can see from the log is:
>>
>> - Two instances of "AmdGop.efi" starting. (Note: that doesn't mean they
>>    are actually binding the card(s) yet.) The number two is strange. Do
>>    you have two assigned GPUs? Can you share your domain XML?
>>
>> - I can see three instances of EfiGraphicsOutputProtocol being
>>    produced. One of them belongs to the QXL driver built into OVMF.
>>    Again, do you have two assigned GPUs?
> Hm, one of those GOPs is probably produced by the console splitter
> driver, so please ignore this one paragraph.
>
>> - The boot process advances pretty far beyond the above. The last
>>    message is related to the USB bus driver
>>    (MdeModulePkg/Bus/Usb/UsbBusDxe). Are you assigning USB devices too?
>>
>>    In your email, you wrote, "If I remove the PCI devices from the VM,
>>    the OS boots without problems." -- device*s*, plural. Did you perhaps
>>    mean "PCI devices" generally as "assigned devices"? If so, what
>>    happens if you keep the GPU(s), but remove the assigned USB devices
>>    from the domain XML?
>>
>> Thanks
>> Laszlo
>>
>> _______________________________________________
>> vfio-users mailing list
>> vfio-users at redhat.com
>> https://www.redhat.com/mailman/listinfo/vfio-users
>>