[vfio-users] Intel IGD VFIO passthrough success on i3-6100, but failed on i3-6100U

[Acewind]

If I leave vnc graphic in libvirt xml, and set target address of IGD
hostdev to another dbsf (not 0000:0:02.0)，win 10 guest can work.
Then I connect into win10 guest with vnc viewer, Intel graphic device can
be seen ok and I can install driver for it.

Acewind <acewind at gmail.com> 于2018年10月13日周六 上午7:07写道：

> Is OVMF recommended? My host has only one IGD display device.
> I haven't tried OVMF. Will it replace legacy mode vfio solution in future?
>
> Alex Williamson <alex.williamson at redhat.com> 于2018年10月12日周五 下午11:56写道：
>
>> On Fri, 12 Oct 2018 23:17:17 +0800
>> Acewind <acewind at gmail.com> wrote:
>>
>> > I remeber your reply here:
>> > https://www.mail-archive.com/vfio-users@redhat.com/msg03415.html
>> > You said QEMU will fix the rom with correct device id automatically when
>> > it's reading the rom directly from the device.
>> >
>> > In today's env, the host is already booted in legacy mode. Do I still
>> need
>> > to dump the rom and fix the device id?
>> >
>> > If I need, can I dump with these commands?
>> >     echo 1 > /sys/devices/pci0000:00/0000:00:02.0/rom
>> >     cat /sys/devices/pci0000:00/0000:00:02.0/rom > vbios.dump
>> >     echo 0 > /sys/devices/pci0000:00/0000:00:02.0/rom
>> >
>>
>> Those would be the correct commands, yes, but if the host is already
>> booting in legacy BIOS mode, I wouldn't expect that passing the ROM via
>> file rather than reading it from the device will make a difference.  I
>> don't know what's wrong, sorry.  Thanks,
>>
>> Alex
>>
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://listman.redhat.com/archives/vfio-users/attachments/20181013/d325ff0b/attachment.htm>