[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: [Xen-devel] PCI Passthrough question (xm->xl migration in Xen 4.2.2)

Monday, November 4, 2013, 4:45:25 PM, you wrote:

> create ^
> title it xl does not support specifying virtual function for passthrough 
> device
> thanks

> On Thu, 2013-10-31 at 14:29 -0700, Saurabh Mishra wrote:
>> In XM toolchain, we use to put pci = [ '0000:07:11.6=0@1a' ] but looks
>> like in XL toolchain there is no way to specify function in the guest.

> The code is a bit confusing, but that does appear to be the case. libxl
> seems to have some degree of support but I can't see any code in xl (or
> the libxlu helper library function which parses the PCI options) that
> would support the =N@M syntax.

> Given the lack of xl support there's a good chance that the libxl
> support is relatively untested.

FYI qemu-upstream xen passhthrough currently also doesn't support specifying a 
request for
a certain dev/function nr.

And it is vaguely related to the lacking ability to passthrough multifunction 
devices as multifunction devices.
(the xm + qemu_trad require you to specify the virtual dev/func in this case, 
and additional some checks in libxl can't handle
the multifunction mask and bail out early)

>> Kindly let me know.
>> XM  - pci = [ '0000:07:11.6=0@1a' ]
>> XL   - pci = [ '0000:07:11.6@1a' ]
>> In the guest VM we see that PCI device is getting 0 function in both
>> the cases however, it may not work the same way in the future.

> Do you happen to know (or can you easily do the experiment to find out)
> what happens with '0000:07:11.6@1a' on xm? Does the guest see function 0
> or function 6? If it is 0 then I think we can safely say xl will not
> change, but if it is 6 we'll have to have a think...

> Ian.

Xen-devel mailing list



Lists.xenproject.org is hosted with RackSpace, monitoring our
servers 24x7x365 and backed by RackSpace's Fanatical Support®.