LinuxQuestions.org

LinuxQuestions.org (http://www.linuxquestions.org/questions/index.php)
-   Slackware (http://www.linuxquestions.org/questions/forumdisplay.php?f=14)
-   -   Custom Kernel won't boot (http://www.linuxquestions.org/questions/showthread.php?t=4175435672)

J1NKL3 11-05-2012 01:42 AM

Custom Kernel won't boot
 
So I installed 3.2.30 around the time Slackware 14 was released. Past couple of days I've been having problems launching random programs and upon rebooting it freezes at the "Loading Linux-3.2.30........" screen after about 8 dots or so. Any idea what would cause this or what logs I could check to help me figure it out?

guanx 11-05-2012 04:30 AM

Quote:

Originally Posted by J1NKL3 (Post 4822336)
... I've been having problems launching random programs ...

What problems?

J1NKL3 11-05-2012 04:32 AM

Quote:

Originally Posted by guanx (Post 4822410)
What problems?

Well they just fail to launch... xpdf, gwenview, ktorrent... the 3.2.28 kernel launches fine but the newer kernel freezes during loading

dwblas 11-05-2012 11:55 AM

Possibly, you should have an initrd in order to boot, depending on whether you compiled a huge kernel or not. See /boot/README.initrd and execute mkinitrd with the proper kernel version, etc., and change LILO per the instructions.

J1NKL3 11-06-2012 12:31 AM

Quote:

Originally Posted by dwblas (Post 4822753)
Possibly, you should have an initrd in order to boot, depending on whether you compiled a huge kernel or not. See /boot/README.initrd and execute mkinitrd with the proper kernel version, etc., and change LILO per the instructions.

It was booting fine without an initrd before this weekend. I booted in and out several times without problems and then had it up for 30+ days until this weekend when I started having problems.

BlackRider 11-06-2012 01:29 PM

It does not really seem to be a kernel problem, or at least a fully kernel related trouble.

If you keep a copy of the Slackware stock kernel, you should return to it and do some testing to ensure the problem is kernel related. Usually, when it is a problem with the kernel you get Panics, warnings and all kind of messages and hints. You can try to surf the logs and use "dmsg" to get a sight about what is going on, given you are able to load the system first.

Common applications failing are more likely problems in the userspace, or may indicate your filesystem has been damaged. You may want to check the system binaries and filesystems are not doomed (you can compare the checksums of the system files with the checksums of the files of the slackware packages).

jheengut 11-08-2012 10:22 AM

Quote:

Originally Posted by BlackRider (Post 4823602)
You can try to surf the logs and use "dmsg" to get a sight about what is going on, given you are able to load the system first.


its
"dmesg" not "dmsg"


If you can drop to a rescue shell before the rootfs '/' is mount rw you can perform a fsck."your rootfs type" on the partition and reboot .

I do not remember how but if you can manually add a ` rescue ` option on the boot command line of lilo or grub then boot .

pokute 11-08-2012 05:54 PM

If the system was fine for 30 days, and then you started having random exe failures, and then the kernel hung at boot, you have filesystem corruption. Possible causes are hardware failure, bad kernel (improper options chosen at build time, bad options in /etc/fstab, or possibly bad code in fs driver), or being logged in as root and making a mess.

If you get back to the point where you run fine for N days, and then some exe randomly fails, run the exe under strace to see why it's failing. Usually when people report what you are reporting, there are libraries that are missing or corrupted, or deleted symlinks.

If you have an SSD and you are certain that you didn't F up anything, your SSD is probably ready to go back to China for a vacation.

J1NKL3 11-09-2012 01:07 AM

Quote:

Originally Posted by pokute (Post 4825371)
If the system was fine for 30 days, and then you started having random exe failures, and then the kernel hung at boot, you have filesystem corruption. Possible causes are hardware failure, bad kernel (improper options chosen at build time, bad options in /etc/fstab, or possibly bad code in fs driver), or being logged in as root and making a mess.

If you get back to the point where you run fine for N days, and then some exe randomly fails, run the exe under strace to see why it's failing. Usually when people report what you are reporting, there are libraries that are missing or corrupted, or deleted symlinks.

If you have an SSD and you are certain that you didn't F up anything, your SSD is probably ready to go back to China for a vacation.

The stock kernel works fine though on the same drive. Shouldn't both fail if I had filesystem problems?

pokute 11-09-2012 10:01 AM

Then (assuming everything happened exactly as described) the kernel you BUILT corrupted the FS in a way that the stock kernel was able to recover. If it were my system, I would do destructive badblocks testing on the drive and then reinstall.


All times are GMT -5. The time now is 12:46 PM.