| View previous topic :: View next topic |
| Author |
Message |
jagdpanther
l33t
Joined: 22 Nov 2003
Posts: 768
|
 Posted: Wed Feb 18, 2015 2:10 pm Post subject: [Solved] Boot Fails 25% of Attempts with kernel BUG |
 |
|
About 25% of the times I attempt to boot Gentoo I receive a "BUG: unable to handle kernel NULL pointer dereference at (null)".
This appears just after "Waiting for uevents to be processed ..." appears on the screen.
https://www.dropbox.com/s/8ivfq4wrc82la45/DSCN0993.jpg is a picture of the screen after the 'Oops'. This all appears to happen before syslogng starts up so no log files.
I notice that if the "Waiting for uevents to be processed ..." takes about a second, then everything is fine and my new system boots up and runs without issue. During a failure there is no time between the "Waiting for uevents to be processed ..." message and the BUG messages.
System: Intel Core i7 based system on an Asus X99-A motherboard.
I see a few references to nvidia on the screen: I am using nvidia-drivers-343.36
I masked out nvidia-drivers-346.35 because consolefonts (other than the lowres default) don't work.
I am using gentoo-sources-3.17.8-r1.
Any suggestions on how to stop this intermittent BUG would be greatly appreciated.
(If I need to supply more information, please ask.)
Last edited by jagdpanther on Mon Feb 23, 2015 1:47 am; edited 1 time in total |
|
| Back to top |
|
 |
Jaglover
Watchman
Joined: 29 May 2005
Posts: 8291
Location: Saint Amant, Acadiana
|
|
| Back to top |
|
|
jagdpanther
l33t
Joined: 22 Nov 2003
Posts: 768
|
| Posted: Thu Feb 19, 2015 1:39 pm Post subject: |
|
|
Thanks for the reply.
| Quote: | | you should tell what init system are you using. |
OpenRC
| Quote: | | can you reproduce it without nvidia blob? |
I'll try tonight and post the results. |
|
| Back to top |
|
|
jagdpanther
l33t
Joined: 22 Nov 2003
Posts: 768
|
| Posted: Thu Feb 19, 2015 8:37 pm Post subject: |
|
|
| Quote: | | can you reproduce it without nvidia blob? |
I can not reproduce the error after removing the nvidia-drivers package.
I unmerged nvidia-drivers and then booted the system five times. All five succeeded.
(of course XDM did not start up either, but that is ok.)
After re-emergeing nvidia-drivers, startx worked fine.
If nvidia-drivers is causing this intermittent issue at boot-time and no-one has a better idea, I can live with starting in console mode and then running 'exec startx' on the console when I want X. In addition to removing 'nvidia' from /etc/conf.d/modules and having rc-update NOT start XDM, what else do I need to do to NOT have the nvidia module load a boot time? (I'll let it auto-load when I run 'startx') |
|
| Back to top |
|
|
Jaglover
Watchman
Joined: 29 May 2005
Posts: 8291
Location: Saint Amant, Acadiana
|
|
| Back to top |
|
|
jagdpanther
l33t
Joined: 22 Nov 2003
Posts: 768
|
| Posted: Mon Feb 23, 2015 1:46 am Post subject: |
|
|
| blacklisting the nvidia module did the trick. Now the nvidia module loads later in boot (when xdm starts) and I haven't had a boot failure since adding "blacklist nvidia" to /etc/modprobe.d/blackllist.conf (I created that file too.) |
|
| Back to top |
|
|
|
Kernel & Hardware
