screensaver hangs in rare circumstances

Bug #655212 reported by Al Stone on 2010-10-05
6
This bug affects 1 person
Affects Status Importance Assigned to Milestone
fglrx-installer (Ubuntu)
Undecided
Unassigned

Bug Description

Binary package hint: xorg

I hesitate to report this bug because it is (1) complicated to reproduce, and (2) therefore not likely to occur. Regardless, it does happen and the only reliable workaround is to reboot my laptop.

To reproduce this bug, I need to:

   -- run my lucid laptop with two screens, one on :0 and one on :1, that are different sizes, with the second screen connected via HDMI

   -- run virt-manager on the larger display (:1) and attempt to install Windows 7 in a 32-bit virtual machine; all interaction with the virtual machine must occur through the virt-manager GUI, not via VNC or other method.

   -- walk away from the laptop and display long enough for the screen saver to kick in (mine just blanks the screen), AND long enough for power savings mode to kick in

   -- when you attempt to get the machine's attention by wiggling the mouse or hitting shift on the keyboard, the only response is the cursor flashing in the lower left part of the :1 display -- no password screen, no actual control of the cursor (the mouse does not change the cursor location),

At this point, the windowing system is completely unresponsive; logging in from another machine via ssh and restarting gdm can sometimes work (X does not always die), but more often than not it's necessary to reboot the system.

If I run the same scenario but remove the second screen (i.e., do everything on the laptop :0 alone), the screen does not lock up. It's also not clear that virt-manager is a key part of the issue -- I can recreate it that way, but it has occurred under other circumstances where I do not recall if virt-manager had been running or not.

ProblemType: Bug
DistroRelease: Ubuntu 10.04
Package: xorg 1:7.5+5ubuntu1
ProcVersionSignature: Ubuntu 2.6.32-25.44-generic 2.6.32.21+drm33.7
Uname: Linux 2.6.32-25-generic x86_64
NonfreeKernelModules: fglrx
Architecture: amd64
Date: Tue Oct 5 09:43:38 2010
DkmsStatus:
 fglrx, 8.723.1, 2.6.32-24-generic, x86_64: installed
 fglrx, 8.723.1, 2.6.32-25-generic, x86_64: installed
GdmLog1: X: /tmp/.X11-unix has suspicious mode (not 1777) or is not a directory, aborting.
InstallationMedia: Ubuntu 10.04 LTS "Lucid Lynx" - Release amd64 (20100429)
MachineType: System76, Inc. Pangolin Performance
ProcCmdLine: BOOT_IMAGE=/boot/vmlinuz-2.6.32-25-generic root=UUID=625ed789-1cb3-4081-9c7e-f812810b5126 ro crashkernel=384M-2G:64M,2G-:128M quiet splash acpi_os_name=Linux acpi_osi=Linux
ProcEnviron:
 PATH=(custom, user)
 LANG=en_US.utf8
 SHELL=/bin/bash
SourcePackage: xorg
Symptom: display
dmi.bios.date: 03/18/2010
dmi.bios.vendor: Phoenix Technologies LTD
dmi.bios.version: CALPELLACRB.86C.0000.X.0000000000
dmi.board.asset.tag: Tag 12345
dmi.board.name: W76x/M77xCUH
dmi.board.vendor: CLEVO
dmi.board.version: Not Applicable
dmi.chassis.asset.tag: No Asset Tag
dmi.chassis.type: 9
dmi.chassis.vendor: CLEVO
dmi.chassis.version: N/A
dmi.modalias: dmi:bvnPhoenixTechnologiesLTD:bvrCALPELLACRB.86C.0000.X.0000000000:bd03/18/2010:svnSystem76,Inc.:pnPangolinPerformance:pvrpanp7:rvnCLEVO:rnW76x/M77xCUH:rvrNotApplicable:cvnCLEVO:ct9:cvrN/A:
dmi.product.name: Pangolin Performance
dmi.product.version: panp7
dmi.sys.vendor: System76, Inc.
glxinfo: Error: [Errno 2] No such file or directory
system:
 distro: Ubuntu
 codename: lucid
 architecture: x86_64
 kernel: 2.6.32-25-generic

Al Stone (ahs3) wrote :
Al Stone (ahs3) wrote :

Some follow-up: I can eliminate virt-manager as one of the variables. This behavior can occur even if virt-manager has never been run. And note that power saving *must* kick in; if it's just the screensaver, it seems to work fine. If, however, both screens have blanked themselves and powered down the backlight, this can occur (it happens more than 50% of the time but less than 100%).

Bryce Harrington (bryce) on 2010-11-04
affects: xorg (Ubuntu) → fglrx-installer (Ubuntu)

Is it possible that this and #637109 may both be rooted in the same issue?

To post a comment you must log in.
This report contains Public information  Edit
Everyone can see this information.

Other bug subscribers