[vfio-users] failing IGD passthrough on apollo lake - BAR 2 error

Alex Williamson alex.williamson at redhat.com
Wed Jul 26 18:05:29 UTC 2017


On Wed, 26 Jul 2017 19:51:12 +0200
Geert Coulommier <g.coulommier at gmail.com> wrote:

> Hi,
> 
> I've been trying to get IGD passthrough to work on an ubuntu 17.04 into
> linux VM's (Ubuntu 16.04) on an Apollo Lake J3455, using mainly Alex' blog
> and this mailing list as a guide, but I keep ending up with a black screen.
> 
> There is no other graphics card, so I've been trying with Seabios and the
> legacy method. I915 is blacklisted. The host is an Ubuntu 17.04 server with
> the Linux kernel upgraded to 4.12.3. Qemu is on 2.8.
> 
> I also tried to passthrough the audio card with the same method, and this
> worked without a glitch.
> 
> >From an extract of the dmesg logfile (see below in [8]) it seems that I'm  
> also running into the BAR 2 error:
> [  192.797993] vfio-pci 0000:00:02.0: BAR 2: can't reserve [mem
> 0x80000000-0x8fffffff 64bit pref]

I'd say you have two options, 1) look in /proc/iomem and identify the
driver that's still claiming portions of IGD and disable it, or 2)
don't blacklist i915, let the kernel boot with it, then do a 'virsh
nodedev-detach pci_0000_00_02_0' at boot before starting the VM so that
you're not binding it back to i915 after every instance of running the
VM.

> 00:0e.0 Audio device [0403]: Intel Corporation Celeron N3350/Pentium
> N4200/Atom E3900 Series Audio Cluster [8086:5a98] (rev 0b)
>         Subsystem: ASRock Incorporation Atom/Celeron/Pentium Processor
> N4200/N3350/E3900 Series Audio Cluster [1849:c892]
>         Kernel driver in use: vfio-pci
>         Kernel modules: snd_hda_intel, snd_soc_skl
...
> [  192.668847] vfio_cap_init: 0000:00:0e.0 pci config conflict @0x80, was
> cap 0x9 now cap 0x10
> [  192.668851] vfio_cap_init: 0000:00:0e.0 pci config conflict @0x81, was
> cap 0x9 now cap 0x10
> [  192.668853] vfio_cap_init: 0000:00:0e.0 pci config conflict @0x82, was
> cap 0x9 now cap 0x10
> [  192.668855] vfio_cap_init: 0000:00:0e.0 pci config conflict @0x83, was
> cap 0x9 now cap 0x10

Hmm, seems we're not super happy parsing this device's config space,
can you provide 'sudo lspci -xxxxs 0000:00:0e.0'.  I don't think this
is related to the IGD issue though.  Thanks,

Alex




More information about the vfio-users mailing list