panic on -Syu: "BUG: unable to handle kernel NULL pointer dereference"

cfr · 2012-08-30 23:42:46

Just upgraded the system. On reboot, I got the grub menu fine but I then got a kernel panic. After switching off at the power and back on again, I booted the same menu option without issue. (Well, without issue so far.)

System is currently able to boot with either systemd or initscripts. Both the failed and successful boot involved the default startup to KDM with systemd.

Details of upgrades:

[2012-08-30 23:06] Running 'pacman -Syu'
[2012-08-30 23:06] synchronizing package lists
[2012-08-30 23:06] starting full system upgrade
[2012-08-30 23:10] removed libsystemd (188-2)
[2012-08-30 23:10] removed systemd-tools (188-2)
[2012-08-30 23:10] Generating locales...
[2012-08-30 23:10]   cy_GB.UTF-8... done
[2012-08-30 23:10]   cy_GB.ISO-8859-14... done
[2012-08-30 23:10]   en_GB.UTF-8... done
[2012-08-30 23:10]   en_GB.ISO-8859-1... done
[2012-08-30 23:10]   en_US.UTF-8... done
[2012-08-30 23:10]   en_US.ISO-8859-1... done
[2012-08-30 23:10] Generation complete.
[2012-08-30 23:10] upgraded glibc (2.16.0-3 -> 2.16.0-4)
[2012-08-30 23:10] upgraded cracklib (2.8.18-2 -> 2.8.19-1)
[2012-08-30 23:10] warning: directory permissions differ on media/
filesystem: 750  package: 755
[2012-08-30 23:10] warning: directory permissions differ on mnt/
filesystem: 750  package: 755
[2012-08-30 23:10] warning: directory permissions differ on usr/local/
filesystem: 750  package: 755
[2012-08-30 23:10] upgraded filesystem (2012.7-1 -> 2012.8-1)
[2012-08-30 23:10] * When used with initscripts, /etc/ifplugd/ifplugd.conf now uses
[2012-08-30 23:10]   INTERFACES= again instead of NET_IFS=.
[2012-08-30 23:10] * This package no longer provides a default ifplugd.action script.
[2012-08-30 23:10] * To use ifplugd with systemd, run
[2012-08-30 23:10]    systemctl enable ifplugd@eth0.service
[2012-08-30 23:10] upgraded ifplugd (0.28-11 -> 0.28-13)
[2012-08-30 23:10] upgraded systemd (188-2 -> 189-3)
[2012-08-30 23:10] upgraded initscripts (2012.08.3-1 -> 2012.08.3-2)
[2012-08-30 23:10] upgraded lib32-glibc (2.16.0-3 -> 2.16.0-4)
[2012-08-30 23:10] >>> Updating module dependencies. Please wait ...
[2012-08-30 23:10] >>> Generating initial ramdisk, using mkinitcpio.  Please wait...
[2012-08-30 23:10] ==> Building image from preset: 'default'
[2012-08-30 23:10]   -> -k /boot/vmlinuz-linux-lts -c /etc/mkinitcpio.conf -g /boot/initramfs-linux-lts.img
[2012-08-30 23:10] ==> Starting build: 3.0.42-1-lts
[2012-08-30 23:10]   -> Running build hook: [base]
[2012-08-30 23:10]   -> Running build hook: [udev]
[2012-08-30 23:10]   -> Running build hook: [autodetect]
[2012-08-30 23:10]   -> Running build hook: [pata]
[2012-08-30 23:10]   -> Running build hook: [scsi]
[2012-08-30 23:10]   -> Running build hook: [sata]
[2012-08-30 23:10]   -> Running build hook: [resume]
[2012-08-30 23:10]   -> Running build hook: [filesystems]
[2012-08-30 23:10]   -> Running build hook: [usbinput]
[2012-08-30 23:10]   -> Running build hook: [fsck]
[2012-08-30 23:10] ==> Generating module dependencies
[2012-08-30 23:10] ==> Creating gzip initcpio image: /boot/initramfs-linux-lts.img
[2012-08-30 23:10] ==> Image generation successful
[2012-08-30 23:10] ==> Building image from preset: 'fallback'
[2012-08-30 23:10]   -> -k /boot/vmlinuz-linux-lts -c /etc/mkinitcpio.conf -g /boot/initramfs-linux-lts-fallback.img -S autodetect
[2012-08-30 23:10] ==> Starting build: 3.0.42-1-lts
[2012-08-30 23:10]   -> Running build hook: [base]
[2012-08-30 23:10]   -> Running build hook: [udev]
[2012-08-30 23:10]   -> Running build hook: [pata]
[2012-08-30 23:10]   -> Running build hook: [scsi]
[2012-08-30 23:10]   -> Running build hook: [sata]
[2012-08-30 23:10]   -> Running build hook: [resume]
[2012-08-30 23:10]   -> Running build hook: [filesystems]
[2012-08-30 23:10]   -> Running build hook: [usbinput]
[2012-08-30 23:10]   -> Running build hook: [fsck]
[2012-08-30 23:10] ==> Generating module dependencies
[2012-08-30 23:10] ==> Creating gzip initcpio image: /boot/initramfs-linux-lts-fallback.img
[2012-08-30 23:10] ==> Image generation successful
[2012-08-30 23:10] upgraded linux-lts (3.0.41-1 -> 3.0.42-1)
[2012-08-30 23:11] upgraded unetbootin (578-1 -> 581-1)
[2012-08-30 23:11] upgraded vym (2.2.0-1 -> 2.2.4-1)
[2012-08-30 23:11] Exited with code 0
[2012-08-30 23:17] Running 'pacman --dbpath /dev/shm/fakepacdb -Sy'
[2012-08-30 23:17] synchronizing package lists
[2012-08-31 00:01] Running 'pacman --dbpath /dev/shm/fakepacdb -Sy'
[2012-08-31 00:01] synchronizing package lists

I have no good way of getting the information off the screen following the panic because I'm never sure what is most useful. The last successful stage completed of the boot sequence was "Reached target swap" which was marked as OK. The next line:

BUG: unable to handle kernel NULL pointer dereference at 0000...018

Dots are all zeroes - I didn't actually count them as I'm assuming the total length of the string is standard.

The trace began with:

native_sched_clock+0x13/0x80
tick_nohz_stop_sched_...

And the messages ended with something like:

Shutting down cpus with NMI.
... panic occurred... returning to text console.

I tried to take a picture with my camera but couldn't get the camera to work. (It eats batteries...) I tried taking one with my phone. This worked but I'm not sure how useful it might be. I will be happy to upload it somewhere if anybody wants to take a look. I won't do this otherwise because it means learning how and although this is on my todo list, it isn't my highest priority unless somebody is actually interested in the result.

I tried to notice on the successful boot what came after swap but couldn't catch it and it seems not possible to scroll the console messages back on tty1. I guess it is possible to recover this information from systemd but need to figure out the best way to do that.

Any advice would be much appreciated - as always .

EDIT: Note that although I have the LTS kernel installed neither the failed nor successful boot involve that kernel. So the panic occurred for a kernel (3.5.3-1-ARCH) which I've actually booted successful several times over the past day or two.

Last edited by cfr (2012-08-30 23:45:42)

cfr · 2012-08-31 15:53:20

This is probably a different problem but I just lost X. The system didn't crash and this was certainly no kernel panic - in fact, KDM respawned automatically - but it is nonetheless unsettling and I did lose some work (not much).

Here's the relevant extract from journalctl. It doesn't look very informative to me but I'd be happy to post more info if somebody can suggest what my be useful:

Aug 31 16:17:11  <hostname> kdm[465]: X server for display :0 terminated unexpectedly
Aug 31 16:17:11  <hostname> polkitd(authority=local)[1284]: Unregistered Authentication Agent for unix-session:/org/freedesktop/Cons
Aug 31 16:17:11  <hostname> kdm[1109]: :0[1109]: pam_unix(kde:session): session closed for user cfrees
Aug 31 16:17:11  <hostname> su[6741]: pam_unix(su:session): session closed for user software
Aug 31 16:17:11  <hostname> acpid[436]: client 493[0:0] has disconnected
Aug 31 16:17:11  <hostname> acpid[436]: client connected from 20163[0:0]
Aug 31 16:17:11  <hostname> acpid[436]: 1 client rule loaded

The bit immediately before this is just crond running and exiting. This includes an error about the unavailability of sendmail. (For some reason, although I can get rid of this error with initscripts, I don't seem to be able to with systemd. Even though crond is started correctly with "-s", it still tries to use sendmail which is most frustrating!) Since this happens every time sendmail runs and since the exit happened only after crond was finished - not after the error with sendmail - I assume that's not the cause.

cfr · 2012-09-02 23:24:15

In trying to solve a different problem, I discovered that KDM had logged a backtrace for X's existence failure. I do wish people put dates and times in log files, though!

acktrace:
0: /usr/bin/X (xorg_backtrace+0x36) [0x560366]
1: /usr/bin/X (0x400000+0x1640c9) [0x5640c9]
2: /usr/lib/libpthread.so.0 (0x7f5344ccc000+0xf170) [0x7f5344cdb170]
3: /usr/lib/xorg/modules/drivers/intel_drv.so (0x7f5341c8f000+0x5ec6e) [0x7f5341cedc6e]
4: /usr/lib/xorg/modules/drivers/intel_drv.so (0x7f5341c8f000+0xb8e2a) [0x7f5341d47e2a]
5: /usr/lib/xorg/modules/drivers/intel_drv.so (0x7f5341c8f000+0x615fe) [0x7f5341cf05fe]
6: /usr/bin/X (0x400000+0xf2119) [0x4f2119]
7: /usr/bin/X (0x400000+0xeb084) [0x4eb084]
8: /usr/bin/X (0x400000+0x34531) [0x434531]
9: /usr/bin/X (0x400000+0x23615) [0x423615]
10: /usr/lib/libc.so.6 (__libc_start_main+0xf5) [0x7f5343b78725]
11: /usr/bin/X (0x400000+0x238ed) [0x4238ed]

Segmentation fault at address 0x7a

Fatal server error:
Caught signal 11 (Segmentation fault). Server aborting


Please consult the The X.Org Foundation support 
         at http://wiki.x.org
 for help. 
Please also check the log file at "/var/log/Xorg.0.log" for additional information.

Server terminated with error (1). Closing log file.

graysky · 2012-09-02 23:33:58

You'll find that the right people will see your 'bug' if you open a flyspray task against the package you believe to be affected.

cfr · 2012-09-03 01:14:47

Thanks. The thing is, I'm not sure it is a bug. At least, I assume the kernel shouldn't panic more-or-less no matter what. But I don't have a clue if the problems I'm seeing are related or what packages they are related to. I've got some suspicions about bluetooth and intel video and systemd and... but nothing I can sort out or really get a grip on.

I'm not even entirely sure I'm not seeing some sort of intermittent hardware issue mixed in as well.

I was really hoping to get some advice on what to do in terms of tracking it down or sorting out different issues. Part of it is that I'm pretty sure some of my configuration is making matters worse even though I doubt that's the only problem. Switching to systemd is *not* straightforward! (At least, not for me.)

DarkVenger · 2012-10-30 00:36:50

Just for the record, if this happens right at system boot, with a kernel 3.6.x (at least 3.6.3) and pata_acpi is the culprit, then you may want to check this kernel bug report.

The short version is to exclude pata_acpi from mkinitcpio.conf and/or blacklisted IF from the output of the lspci -v command you are sure than you don't need this module.
Or else, be patient and help the kernel developers in finding a solution.

Arch Linux

#1 2012-08-30 23:42:46

panic on -Syu: "BUG: unable to handle kernel NULL pointer dereference"

#2 2012-08-31 15:53:20

Re: panic on -Syu: "BUG: unable to handle kernel NULL pointer dereference"

#3 2012-09-02 23:24:15

Re: panic on -Syu: "BUG: unable to handle kernel NULL pointer dereference"

#4 2012-09-02 23:33:58

Re: panic on -Syu: "BUG: unable to handle kernel NULL pointer dereference"

#5 2012-09-03 01:14:47

Re: panic on -Syu: "BUG: unable to handle kernel NULL pointer dereference"

#6 2012-10-30 00:36:50

Re: panic on -Syu: "BUG: unable to handle kernel NULL pointer dereference"

Board footer