I am having troubles with repeated Kernel Panic so that I am unable now to get any Linux system to load and update without loads of greif
Motherboard in Asus M2N-SLI DELUXE
Possessor was 5200+ tried another with same problem
Have test memory with memtest and another tester memory is OK
Tested the hard drive which came up as alright
Tested the memory on the graphics card which was OK but tried another card anyway for good measure same result system crashes during update
I dont think it’s software as it is doing it to all Linux distros I have tried it with some are crashing more easily than others.
I also had a kernel panic trying the cpu tester and identifier on the ultimate boot disc with both cpu’s
I have tested the power supply which seems OK.
Has anyone got any bright idea.
Would I be right in swapping the mother board or is it something else I have missed
Photos of screen print out after crash (I hope)
By the way the first line if I remember rightly ends with all the zero’s followed by 118 http://http://i1282.photobucket.com/albums/a539/Zdirtyrascals_Zdirtyrascals/img_0021.jpg http://http://i1282.photobucket.com/albums/a539/Zdirtyrascals_Zdirtyrascals/img_0020.jpg
I hope I’ve done this right and apologies if it’s gone wrong.
Thanking all in anticipation
On 11/27/2012 09:26 AM, Ralf1e wrote:
> I am having troubles with repeated Kernel Panic so that I am unable now
> to get any Linux system to load and update without loads of greif
> Motherboard in Asus M2N-SLI DELUXE
> Possessor was 5200+ tried another with same problem
> Have test memory with memtest and another tester memory is OK
> Tested the hard drive which came up as alright
> Tested the memory on the graphics card which was OK but tried another
> card anyway for good measure same result system crashes during update
> I dont think it’s software as it is doing it to all Linux distros I
> have tried it with some are crashing more easily than others.
> I also had a kernel panic trying the cpu tester and identifier on the
> ultimate boot disc with both cpu’s
> I have tested the power supply which seems OK.
> Has anyone got any bright idea.
> Would I be right in swapping the mother board or is it something else I
> have missed
> Photos of screen print out after crash (I hope)
> By the way the first line if I remember rightly ends with all the
> zero’s followed by 118
> I hope I’ve done this right and apologies if it’s gone wrong.
The first line of your first image shows a kernel NULL pointer dereference at an
offset of 0x118 in rt_worker_func. The size of that routine is 0x2f8, and the
problem was at 0x104 relative to the start of that routine.
All of the above is easy to discover. The real problem is that a standard kernel
does not contain a routine named rt_worker_func()! Thus the problem happens for
some non-standard module you are loading. The “taint” flags of “GBO” are
instructive. The “G” can be ignored, the “O” means that an out-of-kernel module
has been loaded, and the “B” means that “A process has been found in a Bad page
state, indicating a corruption of the virtual memory subsystem, possibly caused
by malfunctioning RAM or cache memory.” The modules-loaded list shows that the
out-of-kernel one is unblock.ko.
There are two possibilities: (1) Your module unblock.ko has a bug, or (2) your
hardware is malfunctioning. As you have tested the hardware, I think it is
likely that some pointer has not been initialized correctly in that module
unblock. That can be tested to see if the system still crashes when unblock has
never been loaded. If it does, then yo have a memory problem. If no crash
without unblock, then you can find the offending source line with the command
objdump -d -l <path_to_object_file_of rt_worker_func> | less
Once the display is up, search to the entry point of rt_worker_func and note the
starting address of the routine. Add 0x104 to that number (add in hexadecimal)
and find the line in the source code that corresponds. Then go to the source and
see what pointer is dereferenced there.
There are no kernel panics with live cd’s
It gives the same symptoms with openSUSE, mepis. zorin and mint LMDE & 13 which are especially bad It has run all but zorin very well in the past the trouble started with a system check showing a file system fault on mepis when it was double booted with open SUSE
I have formatted and wiped the hard drive with the manufacturers drive tool twice since then as well as checking it with the same tool.
Hope this all helps
Kind regards ralph
On 11/27/2012 05:06 PM, Ralf1e wrote:
> This is a further photo of the message when I tried to see if it would
> update using apper just this evening after trying the above code in
> terminal there are 2 more closeups of the same screen on my photobucket
> account that may help but I don’t suppose you want them filling up here
> ’ (http://)[image:
> Kind regards ralph
That image still has the out-of-kernel module “unblock” loaded. I know that
the Live CD’s will not have this module available. That driver is looking more
and more guilty.
> Copied and pasted code gives this
> bash: syntax error near unexpected token `|’
You did do the correct substitution for the stuff inside the <> marks. did you not?
Thank you for your help I have tested with a spare cheap memory stick bought from a car boot sale for about £2 lol and have been able to load opensuse then update it and do the packman tricks (one click) all of which it did quite happily:-)
During that time I have had 2 very expensive (when new) innocent looking but very guilty Corsair Dominator XMS2 memory sticks sat on my desk.
The moral being dont waste good money on expensive stuff it can fail as well as the cheap stuff and dont trust memtest lol
How do i mark as solved?
If I cant find it pleas do so for me
Thank you for your patience and help, much appreciated
On 11/28/2012 09:56 AM, Ralf1e wrote:
> The moral being dont waste good money on expensive stuff it can fail
AND, always when you boot from the install disk or usb stick before
you install, -=always=- self test the install medium by selecting the
line “Check Installation Media” on the first green screen (like here http://tinyurl.com/b856ekd), and if the media checks good THEN select