Re: Xen 4.4 development update, qemu pci hole start address

[George Dunlap <george.dunlap@eu.citrix.com>]

On 08/10/13 19:13, Pasi Kärkkäinen wrote:
> On Tue, Oct 08, 2013 at 09:05:55PM +0300, Pasi Kärkkäinen wrote:
>> On Thu, Sep 26, 2013 at 05:47:40PM +0100, George Dunlap wrote:
>> http://lists.xen.org/archives/html/xen-devel/2013-05/msg02813.html
>>
>> "quick" workaround was proposed for Xen 4.3 release, and the "proper" solution
>> to be done later for Xen 4.4. Is that still the case? Does this issue still need to be resolved?
>>
>> Quote from the above email/url, Stefano writing:
>>
>> "this is what I would do:
>>
>> 1) modify upstream QEMU to start the PCI hole at 0xe0000000, to match
>> qemu-xen-unstable in terms of configuration and not to introduce any
>> regressions. Do this for the Xen 4.3 release.
>>
>> 2) for Xen 4.4 rework the two patches above and improve
>> i440fx_update_pci_mem_hole: resizing the pci_hole subregion is not
>> enough, it also needs to be able to resize the system memory region
>> (xen.ram) to make room for the bigger pci_hole"
>>
>>
> Hmm, or maybe this series fixed the issue already properly for Xen 4.3.0?

No, this series is still a work-around; see 8/8 of that series: "It's 
too late in the release to do a proper fix, so we try to do damage 
control."  It is fairly likely that there are OS / device combinations 
that currently will not work for PCI pass-through, because either the 
OS, the driver, or the device won't work when relocated into 64-bit PCI 
space.

(I think the problem Gordan has been working on might be related, but 
I'm not sure.)

So yes, good catch -- there should still be an outstanding issue related 
to qemu-xen.

  -George