Regression: the update including linux-image-3.2.0-26-generic breaks suspend on Dell m4300

Bug #1018015 reported by selkovjr
6
This bug affects 1 person
Affects Status Importance Assigned to Milestone
pm-utils (Ubuntu)
Expired
Undecided
Unassigned

Bug Description

I am not sure which package is at fault; I chose pm-utils just to prime the reporting process.

Background:

Ubuntu 12.04 LTS
Dell Precision M4300
NVIDIA Corporation G86M [Quadro FX 360M] (rev a1)

Everything worked fine until an auto-update a few days ago wiped out the proprietary nvidia code. The system booted first time following the update in a low-res mode (did not check which). I immediately used the GUI option to re-enable the proprietary driver and rebooted. It started without problems and I suspended it after a day of work. The next day, it did not wake up (or maybe it did, but I had no way of knowing that because the screen was blank). When I touched the keyboard, it started flashing caps lock and num lock simultaneously and after a while rebooted itself. The same happened again today.

I cannot see any relevant error messages in any of the logs I know to check.

ProblemType: Bug
DistroRelease: Ubuntu 12.04
Package: pm-utils 1.4.1-9
ProcVersionSignature: Ubuntu 3.2.0-26.41-generic 3.2.19
Uname: Linux 3.2.0-26-generic x86_64
NonfreeKernelModules: nvidia
ApportVersion: 2.0.1-0ubuntu10
Architecture: amd64
Date: Tue Jun 26 17:13:08 2012
PackageArchitecture: all
ProcEnviron:
 LANGUAGE=en_GB:en
 TERM=xterm
 PATH=(custom, user)
 LANG=en_GB.UTF-8
 SHELL=/bin/bash
SourcePackage: pm-utils
UpgradeStatus: Upgraded to precise on 2012-05-05 (51 days ago)

Revision history for this message
selkovjr (selkovjr-observercentral) wrote :
affects: ubuntu → pm-utils (Ubuntu)
Revision history for this message
selkovjr (selkovjr-observercentral) wrote :
Download full text (4.0 KiB)

Still not sure which package is affected; possibly more than one. Another update just came in; the kernel is now linux-image-3.2.0-27-generic

In the meantime, the following things happened before the last update:

1. Faled to shut down once. Everything seemed to be normal, the usual shutdown messages went by, the screen went blank and the power light started flashing. I closed the lid and went away. When I came back in a few hours, the machine was hot with the fan blowing full speed, bun not responsive to any inputs. With the lid open, the display light came on showing the remnants of the shutdown message. I did not investigate that event.

2. Attempted to suspend. Everything looked normal and the power did go off. On resume, the system started normally except for black screen with an X11 pointer on it and nothing else. The virtual terminals were working, so I could log in and do stuff. After a long while, the X11 screen came back showing all the windows that were there before suspend and a pop-up saying compiz died (or something to that effect). I have saved the logs for that event. The highlights here are:

Xorg.0.log.old:
[ 34.202] (WW) NVIDIA(0): UBB is incompatible with the Composite extension. Disabling
[ 34.202] (WW) NVIDIA(0): UBB.
...
[1006026.436] (WW) Open ACPI failed (/var/run/acpid.socket) (Connection refused)

syslog:
Jul 17 11:26:11 janus gnome-session[1831]: WARNING: Application 'compiz.desktop' killed by signal
Jul 17 11:26:11 janus gnome-session[1831]: WARNING: App 'compiz.desktop' respawning too quickly
Jul 17 11:26:11 janus gnome-session[1831]: CRITICAL: We failed, but the fail whale is dead. Sorry....

There is also nvidia-bug-report.log.gz timestamped a few minutes later.

Then update came in with a lot of things in it, including compiz. Now when I do suspend I am back where I was when I first reported this problem. On resume, I see a black screen, unresponsive I/O, and caps lock + scroll lock flashing (I have incorrectly reported num lock flashing originally; these were caps lock + num lock in both cases). After flashing for about 20 seconds, the machine automatically rebooted itself. I have saved the logs from aronud this event, too, although I can't see anything interesting in them. Except perhaps this:

Jul 17 13:15:01 janus kernel: [ 2.537401] EXT4-fs (sda1): INFO: recovery required on readonly filesystem
Jul 17 13:15:01 janus kernel: [ 2.537405] EXT4-fs (sda1): write access will be enabled during recovery
Jul 17 13:15:01 janus kernel: [ 8.139947] EXT4-fs (sda1): orphan cleanup on readonly fs
Jul 17 13:15:01 janus kernel: [ 8.139957] EXT4-fs (sda1): ext4_orphan_cleanup: deleting unreferenced inode 11280996

... lots of cleanup ...

Jul 17 13:15:01 janus kernel: [ 8.275229] EXT4-fs (sda1): ext4_orphan_cleanup: deleting unreferenced inode 7077899
Jul 17 13:15:01 janus kernel: [ 8.275284] EXT4-fs (sda1): 27 orphan inodes deleted
Jul 17 13:15:01 janus kernel: [ 8.275287] EXT4-fs (sda1): recovery complete

and:

Jul 17 13:15:01 janus kernel: [ 33.033617] ACPI: Video Device [VID] (multi-head: yes rom: no post: no)
Jul 17 13:15:01 janus kernel: [ 33.033678] [Firmware Bug]: Duplicate...

Read more...

Revision history for this message
selkovjr (selkovjr-observercentral) wrote :

A couple kernel updates later, the kernel still panics on resume, but I have made a couple new observations.

It is not nvidia. I have removed the proprietary driver and still could not resume, although the symptom was slightly different: with the nvidia module installed, the panic is indicated by caps lock and scroll lock flashing together, then the machine reboots. Without nvidia, it just hangs forever without any indication of what happened.

It is not wpa_supplicant (although I see this error tightly associated with the "can't resume from suspend" category of discussions:

 pm-suspend.log:Failed to connect to wpa_supplicant - wpa_ctrl_open: No such file or directory

   https://bugs.launchpad.net/ubuntu/+source/linux/+bug/995486
   http://ubuntuforums.org/showthread.php?t=1941346 (have seen it reach the desktop stage and crash there a couple times)
   https://bugs.launchpad.net/ubuntu/+bug/993598

(and plenty more)

I have blacklisted my wireless interface (iwl3945) and tried to suspend without it; the result is the same. Panic without any visible trace.

It may be a simple co-incidence that all these people also have troble with wpa_supplicant , but definitely a regression, and a relatively recent one (5-6 weeks).

Revision history for this message
selkovjr (selkovjr-observercentral) wrote :
Download full text (5.0 KiB)

Last night's update fixed it. Still no idea what it was, but here's the list of packages that have been updated. The usual suspects are:

compiz:amd64 (0.9.7.8-0ubuntu1.2, 0.9.7.8-0ubuntu1.3)
compiz-core:amd64 (0.9.7.8-0ubuntu1.2, 0.9.7.8-0ubuntu1.3)
compiz-gnome:amd64 (0.9.7.8-0ubuntu1.2, 0.9.7.8-0ubuntu1.3)
compiz-plugins:amd64 (0.9.7.8-0ubuntu1.2, 0.9.7.8-0ubuntu1.3)
compiz-plugins-default:amd64 (0.9.7.8-0ubuntu1.2, 0.9.7.8-0ubuntu1.3)
nvidia-current-updates:amd64 (295.49-0ubuntu0.1, 295.49-0ubuntu0.2)
xserver-common:amd64 (1.11.4-0ubuntu10.6, 1.11.4-0ubuntu10.7)
xserver-xorg-core:amd64 (1.11.4-0ubuntu10.6, 1.11.4-0ubuntu10.7)

Don't know which one did it; other things on the list don't seem to be even remotely relevant. Pm-utils is not among them. I tried various versions of nvidia, including the one that was known to work before. The current kernel is 3.2.0-29-generic #46-Ubuntu SMP.

Start-Date: 2012-08-06 22:53:39

accountsservice:amd64 (0.6.15-2ubuntu9.2, 0.6.15-2ubuntu9.3)
base-files:amd64 (6.5ubuntu6, 6.5ubuntu6.2)
compiz:amd64 (0.9.7.8-0ubuntu1.2, 0.9.7.8-0ubuntu1.3)
compiz-core:amd64 (0.9.7.8-0ubuntu1.2, 0.9.7.8-0ubuntu1.3)
compiz-gnome:amd64 (0.9.7.8-0ubuntu1.2, 0.9.7.8-0ubuntu1.3)
compiz-plugins:amd64 (0.9.7.8-0ubuntu1.2, 0.9.7.8-0ubuntu1.3)
compiz-plugins-default:amd64 (0.9.7.8-0ubuntu1.2, 0.9.7.8-0ubuntu1.3)
cups:amd64 (1.5.3-0ubuntu2, 1.5.3-0ubuntu3)
cups-bsd:amd64 (1.5.3-0ubuntu2, 1.5.3-0ubuntu3)
cups-client:amd64 (1.5.3-0ubuntu2, 1.5.3-0ubuntu3)
cups-common:amd64 (1.5.3-0ubuntu2, 1.5.3-0ubuntu3)
cups-ppdc:amd64 (1.5.3-0ubuntu2, 1.5.3-0ubuntu3)
End-Date: 2012-08-06 23:07:45
firefox:amd64 (14.0.1+build1-0ubuntu0.12.04.1, 14.0.1+build1-0ubuntu0.12.04.2)
firefox-globalmenu:amd64 (14.0.1+build1-0ubuntu0.12.04.1, 14.0.1+build1-0ubuntu0.12.04.2)
firefox-gnome-support:amd64 (14.0.1+build1-0ubuntu0.12.04.1, 14.0.1+build1-0ubuntu0.12.04.2)
firefox-locale-en:amd64 (14.0.1+build1-0ubuntu0.12.04.1, 14.0.1+build1-0ubuntu0.12.04.2)
gcalctool:amd64 (6.4.1.1-0ubuntu2, 6.4.1.1-0ubuntu3)
gir1.2-accountsservice-1.0:amd64 (0.6.15-2ubuntu9.2, 0.6.15-2ubuntu9.3)
google-chrome-beta:amd64 (21.0.1180.57-r148591, 21.0.1180.64-r150042)
isc-dhcp-client:amd64 (4.1.ESV-R4-0ubuntu5.2, 4.1.ESV-R4-0ubuntu5.3)
isc-dhcp-common:amd64 (4.1.ESV-R4-0ubuntu5.2, 4.1.ESV-R4-0ubuntu5.3)
language-pack-en:amd64 (12.04+20120618, 12.04+20120801)
language-pack-en-base:amd64 (12.04+20120508, 12.04+20120801)
language-pack-gnome-en:amd64 (12.04+20120618, 12.04+20120801)
language-pack-kde-en:amd64 (12.04+20120618, 12.04+20120801)
language-pack-kde-en-base:amd64 (12.04+20120508, 12.04+20120801)
libaccountsservice0:amd64 (0.6.15-2ubuntu9.2, 0.6.15-2ubuntu9.3)
libbonobo2-0:amd64 (2.32.1-0ubuntu1, 2.32.1-0ubuntu1.1)
libbonobo2-common:amd64 (2.32.1-0ubuntu1, 2.32.1-0ubuntu1.1)
libbonoboui2-0:amd64 (2.24.5-0ubuntu1, 2.24.5-0ubuntu1.1)
libbonoboui2-common:amd64 (2.24.5-0ubuntu1, 2.24.5-0ubuntu1.1)
libcups2:amd64 (1.5.3-0ubuntu2, 1.5.3-0ubuntu3)
libcups2-dev:amd64 (1.5.3-0ubuntu2, 1.5.3-0ubuntu3)
libcups2:i386 (1.5.3-0ubuntu2, 1.5.3-0ubuntu3)
libcupscgi1:amd64 (1.5.3-0ubuntu2, 1.5.3-0ubuntu3)
libcupsdriver1:amd64 (1.5.3-0ubuntu2, 1.5.3-0ubuntu3)
libcupsimage2:amd64 (1.5.3-0ubuntu2...

Read more...

Revision history for this message
Serge Hallyn (serge-hallyn) wrote :

Please reply if this issue still exists.

Changed in pm-utils (Ubuntu):
status: New → Incomplete
Revision history for this message
selkovjr (selkovjr-observercentral) wrote :

The issue came and went multiple times. I was able to avoid it for several months by running with everything current under funtoo with a 3.10 kernel. When an update broke it again (around Dec 2013), I switched back to 12.04 in the form of Elementary OS Luna. Everything has been working fine since then. The machine goes to sleep and wakes up several times a day.

Revision history for this message
Launchpad Janitor (janitor) wrote :

[Expired for pm-utils (Ubuntu) because there has been no activity for 60 days.]

Changed in pm-utils (Ubuntu):
status: Incomplete → Expired
To post a comment you must log in.
This report contains Public information  
Everyone can see this information.

Other bug subscribers

Remote bug watches

Bug watches keep track of this bug in other bug trackers.