Re: [Xen-devel] Possible bug with pass-through hot-plug?

[Tom Rotenberg <tom.rotenberg@xxxxxxxxx>]

OK...
So, first try to reproduce my problem with a Windows VM, and if it
dosn't reproduce - instruct me on which tests to do to check your
theory.

On Fri, Oct 9, 2009 at 2:09 PM, Simon Horman <horms@xxxxxxxxxxxx> wrote:
> On Fri, Oct 09, 2009 at 01:49:49PM +0200, Tom Rotenberg wrote:
>> OK, great.
>>
>> I will be very surprised if it will work for u on a Windows VM...
>> BTW - do u use the (small) patch i wrote for "forcing" the device to
>> be on the 00:1b.0 slot?
>
> I haven't needed your patch when testing with Linux guests.
>
> The thing is, during my testing I have observed that devfn
> is set correctly without your patch. I was able to observe this
> using the debugging patch that I posted. I am suspecting that during
> your testing devfn is being set to an incorrect value - which your
> patch forcibly corrects. I strongly suspect that is due to some
> behaviour in xm or xend that isn't related to which OS the
> guest is running.
>
>

_______________________________________________
Xen-devel mailing list
Xen-devel@xxxxxxxxxxxxxxxxxxx
http://lists.xensource.com/xen-devel