Bug 8148 - init does not load but there is no kernel panic
Summary: init does not load but there is no kernel panic
Status: REJECTED INSUFFICIENT_DATA
Alias: None
Product: Platform Specific/Hardware
Classification: Unclassified
Component: Other (show other bugs)
Hardware: i386 Linux
: P2 normal
Assignee: other_other
URL:
Keywords:
Depends on:
Blocks:
 
Reported: 2007-03-07 22:11 UTC by js
Modified: 2007-11-20 03:13 UTC (History)
2 users (show)

See Also:
Kernel Version: 2.6.18.3
Subsystem:
Regression: ---
Bisected commit-id:


Attachments
This is the config file that is used to compile the kernel (30.40 KB, text/plain)
2007-03-07 22:13 UTC, js
Details

Description js 2007-03-07 22:11:53 UTC
Most recent kernel where this bug did *NOT* occur: not sure
Distribution: none - generic kernel with bash script as init
Hardware Environment: AMD K6 600 mHz
Software Environment: n/a non distribution with bash init
Problem Description:
This problem is quite rare but it does happen on some systems, it seems more
likely to occur on older systems.

This problem seems to only occur on older hardware as far as I can confirm.
The kernel seems to load fine without any errors but it just decides not to load
init at all.  The keyboard is active.

But the last thing I see is "Freeing Unused Kernel Memory: 13292k freed"

Normally the kernel should load but nothing happens, the kernel does not panic
and I can hit enter on the keyboard but that's it.

Steps to reproduce:
Create this kernel with the config I have shown.
On many older systems that run AMD K6 600 Mhz it seems to do this, my suspicion
is that it is more motherboard related.
Comment 1 js 2007-03-07 22:13:26 UTC
Created attachment 10649 [details]
This is the config file that is used to compile the kernel

This is the config file that is used to compile the kernel
Comment 2 Diego Calleja 2007-03-08 07:08:39 UTC
You have a initrd. Maybe the bug is theere....
Comment 3 js 2007-03-08 07:52:50 UTC
No because it never hits the initrd at all.  It does not even load init and it
only happens on this one system so far :(

I would believe you if it would at least execute 1 command from the init but it
does not.  There is similar to another bug with the same symptoms but caused by
a different kernel option that I am not using.

Comment 4 js 2007-03-08 11:34:12 UTC
This post seems like it might be the problem.  Perhaps this is not a kernel bug
but what this poster mentions below seems like it might be my problem.

What do you guys think?

http://www.linuxquestions.org/questions/showthread.php?p=1099989#post1099989

"Thought I would add my plea for help. Seems like this happens often, and
doesn't have much hope. I actually did move my hard drive from an older AMD K-6
to an AMD Athon 900 Mhz board and get this error now. I reinstalled over the
last one, but did not reformat as I have some info I don't want to lose. Could
this be the prob? I am wondering that the kernel is configured wrong, but I
thought by reinstalling it would be configured right? How can I insure that it
is configured right for the new board? Is it a question of swamp memory? Any ideas?
"
Comment 5 js 2007-03-08 11:35:32 UTC
Sorry I meant to post this from the forum:


I got message "Illegal instruction". Digging deeper I found that this message
appeared when system was trying to execute /bin/bash. That means that bash (and
some other parts of system, init for example) was compiled with instructions not
supported by my K6, only on Duron.

I hope this will help you, guys. Good luck 
Comment 6 Diego Calleja 2007-03-09 13:27:10 UTC
If you had such problem it'd be a user problem (not using the proper distro),
not the kernel
Comment 7 Natalie Protasevich 2007-08-26 12:20:07 UTC
js,
Have you resolved your problem? You can try using "debug" and "apic=debug" on the boot line and then attach your boot trace taken with serial console. You can try booting with acpi=off or nosmp, noapic say to see if your problem is mis-configured interrupts.
Thanks.
Comment 8 Natalie Protasevich 2007-11-20 03:13:19 UTC
Since no response from the reporter, I am closing the bug.
js, feel free to reopen it if you still have this problem.
Thanks.

Note You need to log in before you can comment on or make changes to this bug.