XScreenSaver for dom0 pops up

506 views
Skip to first unread message

John Maher

unread,
Oct 26, 2016, 8:43:12 AM10/26/16
to qubes-users
I'm getting the strangest thing on my screen. I'll be working and the XScreenSaver dialog pops up (indicating the screen is locked) and the screen goes black. However, the screen is not locked. I have to move a window around to redraw my screen. This has always happened since I started using Qubes about 3 weeks ago. There appears to be no pattern that prompts this response. Bizarre! Any thoughts?

Thanks.

John

Gaijin

unread,
Oct 26, 2016, 9:05:48 AM10/26/16
to Qubes Users
Yeah. I had the same thing happen when I upgraded to R3.2 as well. I
asked the same thing but nobody replied.
https://groups.google.com/forum/#!topic/qubes-users/xjHi2TcYBAQ

It still happens from time to time. Seems to happen less if I have a lot
of free memory. Closing down some memory hungry AppVMs seems to lessen
it, but I haven't been able to specifically replicate it.

John Maher

unread,
Oct 26, 2016, 9:38:05 AM10/26/16
to qubes-users, gai...@riseup.net

Interesting, because I think I have more than enough memory (32 GB). Thanks.

Gaijin

unread,
Oct 26, 2016, 6:53:59 PM10/26/16
to Qubes Users
I've got 16GB myself. It's not that the memory is full, but it seems to
happen more when I have a lot of VMs open in one desktop. That may just
be a coincidence though.

I'm glad to see it wasn't just me though. I never had this happen with
earlier versions of Qubes on the same hardware. Not sure if it might be
the XFCE or a video driver issue. (I've never installed the nVidia
drivers.)

Andrew David Wong

unread,
Oct 27, 2016, 12:42:18 AM10/27/16
to Gaijin, Qubes Users
-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA512
Thank you both for reporting this. Gaijin, I'm sorry I missed your thread.

Just to confirm: Are you both using Xfce4?

Tracking the issue here:

https://github.com/QubesOS/qubes-issues/issues/2399

- --
Andrew David Wong (Axon)
Community Manager, Qubes OS
https://www.qubes-os.org
-----BEGIN PGP SIGNATURE-----
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=MANy
-----END PGP SIGNATURE-----

Gaijin

unread,
Oct 28, 2016, 5:25:19 PM10/28/16
to Qubes Users
I used the default install parameters of R3.2. xfce4-about in dom0 shows
I'm using version 4.12. The same hardware since R2 has not experienced
this.

Andrew David Wong

unread,
Oct 29, 2016, 12:10:41 AM10/29/16
to Gaijin, Qubes Users
-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA512

> I used the default install parameters of R3.2. xfce4-about in dom0 shows I'm using version 4.12. The same hardware since R2 has not experienced this.
>

Understood. Thanks.

- --
Andrew David Wong (Axon)
Community Manager, Qubes OS
https://www.qubes-os.org
-----BEGIN PGP SIGNATURE-----
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=sE6O
-----END PGP SIGNATURE-----

John Maher

unread,
Nov 2, 2016, 11:56:06 AM11/2/16
to qubes-users, gai...@riseup.net

Sorry for the delay in responding. I used the default install for 3.2.

I also just noticed that it appears that memory is not being purged when apps are closed. Today when I "woke up" the screen from one of these odd screen saver like issues, it very briefly displayed a browser window with a site I was on yesterday. I looked carefully and confirmed that I did not actually have that window open on my system. It's as if something from yesterday's memory was sent to the screen and then closed. Rather disconcerting.

John

Andrew David Wong

unread,
Nov 3, 2016, 12:50:13 AM11/3/16
to John Maher, qubes-users, gai...@riseup.net
-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA512

> Sorry for the delay in responding. I used the default install for 3.2.
>
> I also just noticed that it appears that memory is not being purged when apps are closed. Today when I "woke up" the screen from one of these odd screen saver like issues, it very briefly displayed a browser window with a site I was on yesterday. I looked carefully and confirmed that I did not actually have that window open on my system. It's as if something from yesterday's memory was sent to the screen and then closed. Rather disconcerting.
>
> John
>

I don't think that should necessarily be disconcerting. The Qubes security model does not include or rely upon purging memory at any particular time interval. It assumes that you, the authentic single user, should be the only one with access to dom0's screen content. So, the fact that you're allowed to see your screen content from yesterday doesn't constitute any violation of the security model. You're still the same trusted user as you were yesterday. (If I've misunderstood your concern, please let me know.)

- --
Andrew David Wong (Axon)
Community Manager, Qubes OS
https://www.qubes-os.org
-----BEGIN PGP SIGNATURE-----

iQIcBAEBCgAGBQJYGsH3AAoJENtN07w5UDAwKtIP/2vblyc4Rf9uwU3M+WGv2K7V
T7VoTDfGjB8XWyX6LZW/TrtameFmAj/Rb0bINOrIWpkrP9RdNUm+0/BR10NkcLul
FqaxMOcl2u2Tk775VjYtC+Z7y1ycJDQjaMvtqrdDQkeGhMumzcDOHD9RufTFSIRm
8ke7GxfMQBH7R0DJ5E26B7HZJg73k1RKFT5BbpmKrfHxBBEaJTUALeNFZnp2ekl+
HrRV4u8+T/Tbwzha5e/iTvJEVTMcMxzD4uziaN+TfiK2Iwp40w0W8IdmRjPzdkLI
+PDhAfjpQCEcZIPT/V+u6GsMhUDJo5ABPs/as17YY4b8VMwm9F7/J2Oo6nfwl/Rh
gOnwVLFQKUtq3iaNbORDzAjBSEny6wYKlfvpL3IWGAHhH0mbP5j1ivSJoF+navuD
qPMvMyAuhFh7hSsTPJ0CTMrDZYTGVlSryLrvXUwl5Lf6zplXRxO6uGu3ruUnepxR
9izVXApPBv/Cc41G3wrCUGN3deE1OpzNSxhQS+NK4B3kJSGAm4+Zi/F/B8yiR6Lt
L/DiLI04X1LZ/XW+ZrKKiQbqVKAzeDqkDPW1D4POEckQnEfPqMcc32LzQVf6ORvX
Xzjn+UhUcdTzINAuISwYeBdvBTk6Wk80T80nChZJdnSdWi6klvbcof96UNdD3sja
swBLAb3aU4hobRxFAcL7
=Fv0E
-----END PGP SIGNATURE-----

Jean-Philippe Ouellet

unread,
Nov 3, 2016, 1:44:20 AM11/3/16
to qubes-users
On Thu, Nov 3, 2016 at 12:50 AM, Andrew David Wong <a...@qubes-os.org> wrote:
> So, the fact that you're allowed to see your screen content from yesterday doesn't constitute any violation of the security model. You're still the same trusted user as you were yesterday. (If I've misunderstood your concern, please let me know.)

This is concerning from the perspective of one who expects a lock
screen to protect the confidentiality of your activities from
untrusted people who may temporarily have limited access to your
machine while it is locked.

But perhaps more seriously, if I understand the report correctly, this
also suggests to me the potential existence of some code path which
renders stale content from untrusted appvms in a full-screen
undecorated context? If that is true, and if it is reliably
triggerable from an appvm, then this would be a useful tool for one
attempting to trick a user with fake UI. This is purely theoretical,
and standard mitigations apply as normal (e.g. trusted window-manager
actions to differentiate true windows) but this still does cause some
concern.

John Maher

unread,
Nov 4, 2016, 1:34:57 PM11/4/16
to qubes-users, john.e...@gmail.com, gai...@riseup.net

If the memory that temporarily renders to the screen is as isolated from other appvms as expected rendered content (like an open windows), then, yes, I'm not concerned about leakage from one environment to another.

I will say that it is simply disconcerting to see it display, security implications aside. More important to me, though, is the hope that this "screen saver" triggering will go away on future updates. Let me know if there is anything I can do to help that process.

John

Opal Raava

unread,
Feb 26, 2017, 9:12:22 AM2/26/17
to qubes-users, john.e...@gmail.com, gai...@riseup.net

Hi all, I'm having the exact same issue both on my previous qubes installation and the current one. It seems to get worse when I have browser windows open, and it continues even if I disable hardware acceleration.

There is a bug tracker entry for this issue:
https://github.com/QubesOS/qubes-issues/issues/2399

The screenshot shows exactly what I'm seeing too. It's very annoying and probably not an issue with graphics drivers. If there is anything I can do to help solve it, I'd be happy to do so.

--Opal

ianr...@gmail.com

unread,
Feb 27, 2017, 3:41:20 PM2/27/17
to qubes-users

I found that this happened when I had the Qubes VM Manager set to appear on all desktops ("Always on Visible Worksace"). When I stopped that, it stopped the problem. Tried this 5 or 6 times on 2 different machines. Running v.3.2 with and without Feb. updates; no difference.

HTH,

--Ian

Opal Raava

unread,
Feb 27, 2017, 7:10:57 PM2/27/17
to qubes-users, ianr...@gmail.com

Oh that's interesting, I found that disabling the compositor (Menu>System Tools>Window Manager Tweaks>Compositor) solves the problem for me and also for somebody else on the github tracker entry.

I can't find the 'always on visible workspaces' settings, but I also lost the tray icon for the Qubes VM Manager somehow.

Pete Howell

unread,
Jan 24, 2019, 12:15:40 PM1/24/19
to qubes-users
Thanks Opal! Disabling the Compositor definitely fixes this issue.
Reply all
Reply to author
Forward
0 new messages