Re: [Xen-devel] Xen-unstable: pci-passthrough of device using MSI-X interrupts not working after commit x86/MSI: track host and guest masking separately

[Sander Eikelenboom <linux@xxxxxxxxxxxxxx>]

Thursday, June 25, 2015, 10:48:40 AM, you wrote:

>>>> On 24.06.15 at 21:38, <linux@xxxxxxxxxxxxxx> wrote:
>> I'm having some trouble with a xhci controller passed through with 
>> pci-passthrough to one of my HVM guests.
>> It uses MSI-X for interrupts, a bisection turned up the following commit:
>> 
>>     x86/MSI: track host and guest masking separately
>> 
>> Although from a first glance it looks as if the controller is correctly 
>> initialize during the boot of the HVM guest (no worrying messages in dmesg 
>> yet).
>> It utterly fails a simple "lsusb" this results in the hang pasted below.
>> 
>> Other devices  i passthrough which use legacy or MSI interrupts seem to be 
>> unaffected.

> Odd enough, since I'm having a hard time testing MSI (no suitable
> devices), but did a lot of testing with MSI-X.

>> Please say so if you need any specific output from Xen debug keys or 
>> anything 
>> else !

> M and i debug key output would be the first thing. I'd suspect host
> masking to be wrongly active for some reason.

> Jan

Hi Jan,

Attached is the xl-dmesg output of:

- debug-keys M and i before guest boot
- guest boot
- debug-keys M and i after lsusb in the guest that hangs.

The not working controller is 0000:08:00.0.

--
Sander

Attachment: xl-dmesg.txt
Description: Text document

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