When logging in to Unity, the Desktop session crashes

[Rainer Rohde]

When launching either Thunderbird or Firefox, my desktop session crashes. I see the underlying command line stuff for a second, and then lightdm launches and I have to re-log in again.

This seems to coincide with my recent issue of not being able to launch Unity or Gnome Shell with nvidia-current drivers installed. The only DE that works for me is Classic Gnome w/o effects.

When launching Unity from lightdm, the behavior is the same as mentioned above -- the desktop crashes, and immediately spawns lightdm again. I can repeat this cycle ad-nauseum.
---
ApportVersion: 2.4-0ubuntu8
Architecture: amd64
DistroRelease: Ubuntu 12.10
EcryptfsInUse: Yes
NonfreeKernelModules: nvidia
Package: unity 6.2.0-0ubuntu2
PackageArchitecture: amd64
ProcEnviron:
 TERM=xterm
 PATH=(custom, no user)
 LANG=en_US.UTF-8
 SHELL=/bin/bash
ProcVersionSignature: Ubuntu 3.5.0-11.11-generic 3.5.2
Tags: third-party-packages quantal third-party-packages quantal
Uname: Linux 3.5.0-11-generic x86_64
UpgradeStatus: No upgrade log present (probably fresh install)
UserGroups: adm cdrom dip lpadmin plugdev sambashare sudo
---
ApportVersion: 2.4-0ubuntu8
Architecture: amd64
DistroRelease: Ubuntu 12.10
EcryptfsInUse: Yes
NonfreeKernelModules: nvidia
Package: unity 6.2.0-0ubuntu2
PackageArchitecture: amd64
ProcEnviron:
 TERM=xterm
 PATH=(custom, no user)
 LANG=en_US.UTF-8
 SHELL=/bin/bash
ProcVersionSignature: Ubuntu 3.5.0-11.11-generic 3.5.2
Tags: third-party-packages quantal third-party-packages quantal
Uname: Linux 3.5.0-11-generic x86_64
UpgradeStatus: No upgrade log present (probably fresh install)
UserGroups: adm cdrom dip lpadmin plugdev sambashare sudo

[Brad Figg (brad-figg)]

This bug is missing log files that will aid in diagnosing the problem. From a terminal window please run:

apport-collect 1039148

and then change the status of the bug to 'Confirmed'.

If, due to the nature of the issue you have encountered, you are unable to run this command, please add a comment stating that fact and change the bug status to 'Confirmed'.

This change has been made by an automated script, maintained by the Ubuntu Kernel Team.

[Rainer Rohde (rainer-rohde)]

apport information

[Rainer Rohde (rainer-rohde)]

apport information

[Rainer Rohde (rainer-rohde)]

I reinstalled Quantal from scratch, and the behavior persists: When trying to launch into Unity, I am immediately dumped to the shell, and then lightdm restarts. Again, the only session that works for me is Classic Gnome, however, even there launching either Firefox or Thunderbird result in the same action -- me being dumped to the shell and then returned to lightdm.

This is really frustrating. Everything worked perfectly on my Dell XPS 15z, and I realize that the only main difference between my current Asus G75 and that Dell is the fact that the Dell had nVidia Optimus on board, which means that it would run on an Intel graphics card most of the time. And even then, when testing Bumblebee's "optirun", I never had an issue with the discrete nVidia card.

However, now, with this Asus laptop having a dedicated nVidia 660M GTX, it seems as if the functionality of Ubuntu (I know -- it's only alpha) has gone to hell.

How can I help test this so that someone can fix this? Heck, I even had to resort to a PPA in order to install nvidia-current in the first place....

Sorry for the rant, but nut being able to use Thunderbird for emailing and calendaring poses a nightmare here at work.

[Rainer Rohde (rainer-rohde)]

Forgot to add this...

[Omer Akram (om26er)]

This may as well be similar to what many people are facing in bug 973096. Please try the driver from here http://ppa.launchpad.net/xorg-edgers/ppa/ubuntu/pool/main/n/nvidia-graphics-drivers/ and see if that fixes the issue for you.

[Rainer Rohde (rainer-rohde)]

Omer, thanks for the pointer, but sadly this didn't fix it. I installed "nvidia-current_304.37-0ubuntu1~xedgers~quantal1_amd64" from that link you provided and my session still crashes when using either Thunderbird, Firefox or attempting to use Unity...

Anything else I can do?

[Rainer Rohde (rainer-rohde)]

Here's some output from my dmesg regarding Xorg as well; not sure if that helps at all...

[ 3.576578] Pid: 1127, comm: Xorg Tainted: P O 3.5.0-11-generic #11-Ubuntu
[ 3.576579] Call Trace:
[ 3.576583] [<ffffffff81051c4f>] warn_slowpath_common+0x7f/0xc0
[ 3.576585] [<ffffffff81051caa>] warn_slowpath_null+0x1a/0x20
[ 3.576587] [<ffffffff81043435>] __ioremap_caller+0x335/0x380
[ 3.576643] [<ffffffffa0de3cc5>] ? os_map_kernel_space+0x85/0xf0 [nvidia]
[ 3.576645] [<ffffffff810434b4>] ioremap_cache+0x14/0x20
[ 3.576686] [<ffffffffa0de3cc5>] os_map_kernel_space+0x85/0xf0 [nvidia]
[ 3.576729] [<ffffffffa0daf918>] _nv014877rm+0x76/0x92 [nvidia]
[ 3.576765] [<ffffffffa07dea1e>] ? _nv009673rm+0x89/0x142 [nvidia]
[ 3.576800] [<ffffffffa07deb8f>] ? _nv014448rm+0xb8/0x102 [nvidia]
[ 3.576835] [<ffffffffa07df042>] ? _nv014486rm+0x58/0x9e [nvidia]
[ 3.576871] [<ffffffffa07daaa8>] ? _nv014458rm+0xbe/0x2f0 [nvidia]
[ 3.576906] [<ffffffffa07dad85>] ? _nv014491rm+0xab/0x174 [nvidia]
[ 3.576940] [<ffffffffa07dae9e>] ? _nv014457rm+0x50/0x5d [nvidia]
[ 3.576976] [<ffffffffa07e1072>] ? _nv014432rm+0x82d/0x96b [nvidia]
[ 3.577018] [<ffffffffa0dbf7ef>] ? _nv012658rm+0x174/0x573 [nvidia]
[ 3.577059] [<ffffffffa0dbf76d>] ? _nv012658rm+0xf2/0x573 [nvidia]
[ 3.577105] [<ffffffffa082b261>] ? _nv004021rm+0x1bb/0x19e9 [nvidia]
[ 3.577196] [<ffffffffa0b82ef9>] ? _nv004046rm+0x8690/0xae8b [nvidia]
[ 3.577284] [<ffffffffa0b818bb>] ? _nv004046rm+0x7052/0xae8b [nvidia]
[ 3.577316] [<ffffffffa07b3d0c>] ? _nv009980rm+0x25/0x40 [nvidia]
[ 3.577358] [<ffffffffa0dbf501>] ? _nv014932rm+0x7c9/0x943 [nvidia]
[ 3.577399] [<ffffffffa0dc04af>] ? _nv001095rm+0x42a/0x6b4 [nvidia]
[ 3.577441] [<ffffffffa0db8ef4>] ? rm_init_adapter+0xac/0x146 [nvidia]
[ 3.577481] [<ffffffffa0dd8f74>] ? nv_kern_open+0x4b4/0x800 [nvidia]

[Rainer Rohde (rainer-rohde)]

I tried something different to generate more meaningful output.

-- Logged in from the command-line and ran "startx". The session started just fine, however there was no Gnome Panel loaded whatsoever.

-- Pressed F12 to launch Guake terminal and typed "unity --replace &" and hit enter. This dropped me immediately to the shell and I observed an error message regarding a segmentation fault pertaining to the nvidia-current driver (please see screen photos attached).

[Rainer Rohde (rainer-rohde)]

Here's some more content to look at for those working towards a fix: I ran "unity --debug" and got this:

[Thread debugging using libthread_db enabled]
Using host libthread_db library "/lib/x86_64-linux-gnu/libthread_db.so.1".
[New Thread 0x7ffff3b5c700 (LWP 10817)]

Program received signal SIGHUP, Hangup.
0x00007ffff7331c07 in ioctl () from /lib/x86_64-linux-gnu/libc.so.6
#0 0x00007ffff7331c07 in ioctl () from /lib/x86_64-linux-gnu/libc.so.6
No symbol table info available.
#1 0x00007fffeabab74c in ?? () from /usr/lib/nvidia-current/libnvidia-glcore.so.304.37
No symbol table info available.
#2 0x00007fffeabb014f in ?? () from /usr/lib/nvidia-current/libnvidia-glcore.so.304.37
No symbol table info available.
#3 0x00007fffea82e651 in ?? () from /usr/lib/nvidia-current/libnvidia-glcore.so.304.37
No symbol table info available.
#4 0x00007fffea79acd2 in ?? () from /usr/lib/nvidia-current/libnvidia-glcore.so.304.37
No symbol table info available.
#5 0x00007fffea79ade4 in ?? () from /usr/lib/nvidia-current/libnvidia-glcore.so.304.37
No symbol table info available.
#6 0x00007fffea79be0b in ?? () from /usr/lib/nvidia-current/libnvidia-glcore.so.304.37
No symbol table info available.
#7 0x00007fffea806fe5 in ?? () from /usr/lib/nvidia-current/libnvidia-glcore.so.304.37
No symbol table info available.
#8 0x00007fffea79561a in ?? () from /usr/lib/nvidia-current/libnvidia-glcore.so.304.37
No symbol table info available.
#9 0x00007fffebb3a44c in ?? () from /usr/lib/nvidia-current/libGL.so.1
No symbol table info available.
#10 0x00007fffebb246eb in ?? () from /usr/lib/nvidia-current/libGL.so.1
No symbol table info available.
#11 0x00007ffff7de992d in ?? () from /lib64/ld-linux-x86-64.so.2
No symbol table info available.
#12 0x00007ffff72828e1 in ?? () from /lib/x86_64-linux-gnu/libc.so.6
No symbol table info available.
#13 0x00007ffff7282965 in exit () from /lib/x86_64-linux-gnu/libc.so.6
No symbol table info available.
#14 0x00007ffff6d35363 in _XDefaultIOError () from /usr/lib/x86_64-linux-gnu/libX11.so.6
No symbol table info available.
#15 0x00007ffff6d3559e in _XIOError () from /usr/lib/x86_64-linux-gnu/libX11.so.6
No symbol table info available.
#16 0x00007ffff6d3357c in _XReply () from /usr/lib/x86_64-linux-gnu/libX11.so.6
No symbol table info available.
#17 0x00007fffebb4bcaa in ?? () from /usr/lib/nvidia-current/libGL.so.1
No symbol table info available.
#18 0x00007fffebb19342 in ?? () from /usr/lib/nvidia-current/libGL.so.1
No symbol table info available.
#19 0x00007fffebb0fac4 in glXBindTexImageEXT () from /usr/lib/nvidia-current/libGL.so.1
No symbol table info available.
#20 0x00007fffebdec212 in TfpTexture::bindPixmapToTexture(unsigned long, int, int, int) () from /usr/lib/compiz/libopengl.so
No symbol table info available.
#21 0x00007fffebde71a4 in boost::detail::function::function_invoker4<GLTexture::List (*)(unsigned long, int, int, int), GLTexture::List, unsigned long, int, int, int>::invoke(boost::detail::function::function_buffer&, unsigned long, int, int, int) () from /usr/lib/compiz/libopengl.so
No symbol table info available.
#22 0x00007fffebdebe24 in GLTexture::bindPixmapToTexture(unsigned long, int, int, int) () from /usr/li...

[Karl Hegbloom (karl.hegbloom)]

I get pretty much the same errors in the dmesg log, but the X server starts up and runs mostly alright, but after it's been running a while, starts getting wonky, like it won't respond for a long time, then suddenly comes back to life, but hangs intermittently from then on. A reboot is necessary to make it usable again.

I suspect this is nVidia's bug, though I do see a few mentions via Google searches of patches made in the past to the Linux kernel surrounding the ioremap and nVidia's proprietary drivers.

I've tried their newer driver, 304.60, and it's the same.

NVRM version: NVIDIA UNIX x86_64 Kernel Module 304.51 Tue Sep 18 17:16:56 PDT 2012
GCC version: gcc version 4.7.2 (Ubuntu/Linaro 4.7.2-2ubuntu1)

Model: GeForce 7600 GT
IRQ: 24
GPU UUID: N/A
Video BIOS: 05.73.22.18.00
Bus Type: PCI-E
DMA Size: 39 bits
DMA Mask: 0x7fffffffff
Bus Location: 0000:01.00.0

[dino99 (9d9)]

That version is no more supported; please open a new bug report if the actual archive found version also has the same issue.