On 05/16/2016 09:57 AM, Chun Yan Liu wrote:
On 5/14/2016 at 12:58 AM, in message <5736079B.5050308@suse.com>, Jim Fehlig <jfehlig@suse.com> wrote: On 05/13/2016 07:08 AM, Joao Martins wrote: I am not sure what's the etiquette in these cases but I sent you some patches that fixes the tests and makes some adjustments to help out (all changelog-ed). There were a couple of failures lately regarding vram defaults and what not so some of the tests would fail as vram defaults would be wrongly calculated. Jim sent a series fixing that which is Acked already but still to be pushed.
I've pushed those patches now. I think we've finally resolved all the default vram issues.
Joao, I applied your updated patch series, it looks good. Maybe you can send out to mailing list?
Sure! But maybe it's not worth such a big change involving the tests and we should probably wait until something more substantial is required from the vif parsing side [0]. So the initial approach is the one to go - requiring mainly a slight interface change. Jim suggested having an enum to identify which flavor ParseCommon is invoked - which is generic enough representation to handle any trivial disparities between XL and XM. Joao [0] https://www.redhat.com/archives/libvir-list/2016-May/msg01002.html