Bug 84599 - CentOS 7 - Automatic resizing does not work
Summary: CentOS 7 - Automatic resizing does not work
Alias: None
Product: Spice
Classification: Unclassified
Component: xorg qxl (show other bugs)
Version: unspecified
Hardware: Other Linux (All)
: medium normal
Assignee: Spice Bug List
QA Contact:
Depends on:
Reported: 2014-10-02 14:13 UTC by Lyude Paul
Modified: 2014-10-02 17:31 UTC (History)
0 users

See Also:
i915 platform:
i915 features:


Description Lyude Paul 2014-10-02 14:13:59 UTC
So, out of the box I've noticed that on CentOS 7, the guest fails to automatically resize it's screen when I resize the window of my spice client. Instead of changing the resolution, the entire desktop (I'm using MATE on my guest) just flashes for a second like it's trying to change resolution but it can't. I use virt-manager, but I've tried with other spice clients too, and the behavior is exactly the same. I've also tried other guest kernels including the more up to date kernel-ml from the elrepo, but to no avail. The behavior always remains the same.

The current kernel version I'm using is 3.10.0-123.8.1.el7.x86_64, qxl driver version on the guest is the latest upstream qxl driver in the CentOS repo, 0.1.1 release 9.el7. If you need any other information please let me know.
Comment 1 Jonathon Jongsma 2014-10-02 14:36:53 UTC
I seem to recall that MATE had trouble with resizing... What is the behavior when running gnome-shell?
Comment 2 Lyude Paul 2014-10-02 17:20:36 UTC
(In reply to Jonathon Jongsma from comment #1)
> I seem to recall that MATE had trouble with resizing... What is the behavior
> when running gnome-shell?

It looks like that it actually works perfectly with GNOME, I'm guessing I should report this to the MATE developers?
Comment 3 Jonathon Jongsma 2014-10-02 17:31:16 UTC
Yes, it appears that this is a MATE bug then. 

A bit of background: In the past, the spice-vdagent reconfigured the displays directly. However, some time ago this changed, and now a 'hotplug' event is generated in the guest OS (indicating that the device has changed in some way) and the desktop environment is reponsible for querying the new configuration and reconfiguring the displays appropriately. I believe this is the bug where support for this method was added to GNOME (actually, mutter): https://bugzilla.gnome.org/show_bug.cgi?id=711216

Use of freedesktop.org services, including Bugzilla, is subject to our Code of Conduct. How we collect and use information is described in our Privacy Policy.