[libvirt] [PATCH] conf: Fix backwards migration of pSeries guests

Peter Krempa pkrempa at redhat.com
Mon Jul 24 08:32:35 UTC 2017


On Fri, Jul 21, 2017 at 10:39:05 +0200, Andrea Bolognani wrote:
> On Fri, 2017-07-21 at 10:07 +0200, Andrea Bolognani wrote:

[...]

> > That makes sense.
>> > Doesn't it also mean that we don't really need to record
> > whether the user set the model name explicitly or not? We
> > can just skip formatting it if it's spapr-pci-host-bridge
> > and all versions of libvirt, past or future, will handle
> > that correctly.
> 
> Wait, that's exactly how this patch behaves, except it's
> not immediately apparent because the check for the model
> name happens inside virDomainControllerIsPCIHostBridge()
> rather than being explicit.
> 
> So maybe I'm still missing something?

You are right. The not-formatting only the first model ever added should
be fine in this case also in the future.

Please send a fixed version of this patch that will make sure that it
states that the PHB thing is a pSeries thing, so that it does not
confuse others as it did me.
-------------- next part --------------
A non-text attachment was scrubbed...
Name: signature.asc
Type: application/pgp-signature
Size: 833 bytes
Desc: Digital signature
URL: <http://listman.redhat.com/archives/libvir-list/attachments/20170724/218e1492/attachment-0001.sig>


More information about the libvir-list mailing list