Re: [Xen-users] Can't pass through onboard Intel HDA, works in KVM

["Jason O'Brien" <jason23@xxxxxxxx>]

On Sat, Sep 1, 2012 at 7:54 PM, Jason O'Brien <jason23@xxxxxxxx> wrote:
> With the kernel option 'xen-pciback.hide=(00:14.2)
> xen-pciback.permissive', Xen-4.2.0-rc3 is unable to pass through my
> onboard Intel HDA audio adapter to a Windows7 guest. I tried doing the
> same in KVM and it worked immediately without issue. On the guest, I
> get a "High Definition Audio Controller" with the correct
> vendor/device ID, and an error of "This device cannot start. (Code
> 10)". I've tried a bunch of combinations, including late binding to
> pciback, with no luck.
>
> Has anyone seen this before? It seems to be the only thing that won't
> work.. secondary passthrough of my AMD HD6850 in Xen works great (but
> not in KVM unfortunately), as does passing through the USB
> controllers. There's another google hit for a guy with a similar
> problem, but it turned out he was using pci-stub instead of pciback
> (which, as the below info shows, I am not).
>
> Thanks for your time,
> Jason
>
I went ahead and tested this on kernel 3.4.6 (vs 3.5.2, which is what
I was running) and it works fine, so this is likely a PCI passthrough
regression in kernel 3.5. The patch on the xen-devel thread didn't do
the trick for me, I'll give reverting the questionable commit a try.

_______________________________________________
Xen-users mailing list
Xen-users@xxxxxxxxxxxxx
http://lists.xen.org/xen-users