nvidia_drv.so crash during X restart by kdm

[Leo Milano]

Binary package hint: nvidia-glx

In Kubuntu hardy (current as of today), X starts fine with the NVIDIA drivers. You can login as a regular user, but on logout, you get ablack screen, and then you get dropped to the command line. If I login, and then I force a kdm restart, a new session starts, and works just fine. I am attaching my xorg.cond, and X logs.

I have an NVIDIA GeForce2 card. I am using the regular nvidia-glx package, which is the one that gets installed by the restricted driver manager. I tried installing glx-new manually and it wouldn't work, maybe because this card is rather old.

This is in hardy current. It is a regression, previous (K)Ubuntu versions ran fine (including Gutsy).

Please let me know if I can provide any further info. Thanks for the great work.

Cheers,
Leo

[Leo Milano (lmilano)]

Just an update. Current Hardy (April 6 2008, kernel 2.6.24-15, nvidia-glx 1:96.43.05+2.6.24.12-15.3) still has this issue. I have another partition running gutsy and there are no NVIDIA issues there. I also tried copying the exact same xorg.conf from that partition into the Hardy partition, and it didn't help. On the other hand, if I use the xorg "nv" driver there are no issues on logout. This definitely looks like a regression, at least in my machine.

Please let me know if you have ideas on how to debug this. Thanks!

[Leo Milano (lmilano)]

I just tried using gdm, and the crash _disappears_ with gdm. In summary, both kdm and kdm-kde4 fail to restart X when you log out, but gdm is fine, it shows the login window again.

[Ernst Kloppenburg (ernst-kloppenburg)]

I can confirm the problem on two different machines with up to date hardy beta and no foreign software

using kdm and nvidia cards

when a user logs out, kdm tries to start a new xserver, and that one crashes. restarting kdm produces a new working xserver

the nvidia cards are:
nVidia Corporation NV17 [GeForce4 MX 440-SE]
nVidia Corporation NV34 [GeForce FX 5500] (rev a1)

[Ernst Kloppenburg (ernst-kloppenburg)]

Leo, I suggest that you retitle this bug to, e.g. :
"nvidia_drv.so crash during X restart by kdm"

[Leo Milano (lmilano)]

Thank Ernst,

I changed the status to confirmed. and changed the bug title to what you suggested. I think this is pretty bad, one week away from the release. Particularly so for Kubuntu. Do you know if there is a way to point this to the core Kubuntu developers?

Cheers!
Leo

[Ernst Kloppenburg (ernst-kloppenburg)]

this seems to be a kdm problem (it goes away if you use gdm instead)

[Leo Milano (lmilano)]

Here is something odd: KDM is supposed to have bulletproof X built in. Why are we not falling back to a "safe" VGA driver instead of crashing ?

####
kdebase (4:3.5.8-2ubuntu21) hardy; urgency=low

  * Add kubuntu_9921_kdm_bulletproof_x.diff, adds bulletproof X
    feature to KDM.
####

[yknot (dennisthompso)]

I am experiencing a similar problem. On reboot kdm starts fine. I logout and I get a blank screen, no command line no nothing. If I ssh in from another computer and check the kdm.log and Xorg.0.logs I get the same error messages.

Backtrace:
0: /usr/bin/X(xf86SigHandler+0x6a) [0x48402a]
1: /lib/libc.so.6 [0x7f9b91f3d100]
2: /usr/lib/xorg/modules/drivers//nvidia_drv.so [0x7f9b8fc1c780]

Fatal server error:
Caught signal 11. Server aborting

"ps uax | grep kdm" shows that kdm is still running

So the X server is dead but kdm is still running. Now if I kill the running kdm and then restart kdm "/usr/bin/kdm -config /var/run/kdm/kdmrc"

Both the X server and kdm will restart just fine no problems no errors. So the problem is some how related to the restarting of the X server on logout.

nvidia-glx 1:96.43.05+2.6.24.13-19.42
nvidia-kernel-common 20051028+1ubuntu8
nvidia-new-kernel-source 169.12+2.6.24.13-19.42
linux-image-2.6.24-19-generic 2.6.24-19.34
x11-common 1:7.3+10ubuntu10
libc6 2.7-10ubuntu3

I am running dual displays on a Dell Precision 390
Intel(R) Core(TM)2 CPU 6300 @ 1.86GHz
nVidia Corporation NV44 [Quadro NVS 285]

[JETorkildsen (jeta3)]

Have exactly the same problem on my computer using a Nvidia Geforce4 MX 420 card. I'm using hardy and the nvidia driver: nvidia-glx 1:96.43.05+2.6.24.13-19.44. On logout I get a black screen. In this state, I can use alt+F2 and log on using the terminal and then startx to start the GUI.

Here is a the last 15 lines of my /var/log/Xorg.0.log.old:

jan@fuji:~$ tail /var/log/Xorg.0.log.old -n15
Error opening /dev/input/wacom : Success
(II) Open ACPI successful (/var/run/acpid.socket)
(II) APM registered successfully
(II) NVIDIA(0): Initialized GART.
(II) NVIDIA(0): Setting mode "1280x1024@60"

Backtrace:
0: /usr/bin/X(xf86SigHandler+0x7e) [0x80c780e]
1: [0xb7fe2420]
2: /usr/lib/xorg/modules/drivers//nvidia_drv.so(_nv001216X+0xe5) [0xb71c9711]
3: [0x1]

Fatal server error:
Caught signal 11. Server aborting

jan@fuji:~$

[JETorkildsen (jeta3)]

Output from System Log:

07/11/2008 06:48:44 PM fuji kdm[6258] Unable to fire up local display :0; disabling.
07/11/2008 06:48:44 PM fuji kdm[6258] X server for display :0 terminated unexpectedly
07/11/2008 06:48:44 PM fuji kdm[6258] X server terminated: [0, 0, 0]
07/11/2008 06:48:44 PM fuji kernel [ 214.942477] Xorg[6263]: segfault at 00005456 eip 080c6706 esp bfdcfec0 error 4

[Ralf Hildebrandt (ralf-hildebrandt)]

I get the same:

from dmesg:
[825698.383182] Xorg[14659]: segfault at ffffffffffed7b0e rip 4830b3 rsp 7ffff8875210 error 4
[862666.364335] Xorg[18354]: segfault at 446e00004155 rip 4830c6 rsp 7fff5055ff00 error 4

from the Xorg.0.log.old:

...
(II) Open ACPI successful (/var/run/acpid.socket)
(II) APM registered successfully
(II) NVIDIA(0): Initialized GART.
(II) NVIDIA(0): Setting mode "1440x900"

Backtrace:
0: /usr/bin/X(xf86SigHandler+0x6a) [0x48402a]
1: /lib/libc.so.6 [0x7fc946a86100]
2: /usr/lib/xorg/modules/drivers//nvidia_drv.so [0x7fc944765780]

Fatal server error:
Caught signal 11. Server aborting

00:05.0 VGA compatible controller: nVidia Corporation MCP51 PCI-X GeForce Go 6100 (rev a2) (prog-if 00 [VGA controller])
        Subsystem: Acer Incorporated [ALI] Unknown device 0112
        Flags: bus master, 66MHz, fast devsel, latency 0, IRQ 18
        Memory at c2000000 (32-bit, non-prefetchable) [size=16M]
        Memory at d0000000 (64-bit, prefetchable) [size=256M]
        Memory at c1000000 (64-bit, non-prefetchable) [size=16M]
        [virtual] Expansion ROM at 50000000 [disabled] [size=128K]
        Capabilities: [48] Power Management version 2
        Capabilities: [50] Message Signalled Interrupts: Mask- 64bit+ Queue=0/0 Enable-

[Jonathan Thomas (echidnaman)]

This would be a bug in either the driver or the driver packaging that kdm merely triggers. Unfortunately, if the bug is actually in the driver code itself and not the packaging, it cannot be fixed by Kubuntu since the driver is closed-source.

[ashutosh mahajan (asm4)]

a temporary fix is to use:
TerminateServer=true

in /etc/kde3/kdm/kdmrc. this restarts xserver after every session and avoids
the crash.

[markofealing (mark-ferns16)]

I'm also experiencing this problem:

AMD Athlon 4200 x2
nVidia GeForce 6600 (nVidia drivers 185.18.36) dual display.
Kubuntu 9.10 - 64-bit

X randomly crashes (occasionally) and restarts, taking me back t the login window. I login and it will work until the next crash. Shutting down will occasionally give me a black screen.

Occasionally, when I start up I will get the login window displayed briefly and then get a black screen

Checking the system log give me the following at time of crash this evening when X restarted itself:

[10397.372523] NVRM: Xid (0002:00): 26, Ch 00000000 M 00000a2c D 00000000 intr ffffffff
[10397.376599] NVRM: Xid (0002:00): 1, Ch 00000000 M 00000a2c D 00000000 intr ffffffff
[10397.769809] Xorg[1164]: segfault at 0 ip 00007fa2bff7bbc7 sp 00007fffabd81ad0 error 4 in nvidia_drv.so[7fa2bfc61000+3ac000]
X server for display :0 terminated unexpectedly
client 1164[0:0] has disconnected
client 1164[0:0] has disconnected

[penalvch (penalvch)]

Leo Milano, thank you for reporting this bug to Ubuntu. Hardy Desktop reached EOL on May 12, 2011.
See this document for currently supported Ubuntu releases: https://wiki.ubuntu.com/Releases

Is this an issue in a supported release? If so, could you please execute the following command, as it will automatically gather debugging information, in a terminal:
apport-collect 202855

Please ensure you have xdiagnose installed, and that you click the Yes button for attaching additional debugging information.

[Launchpad Janitor (janitor)]

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