Re: Oops in ReiserFS (2.4.17)

[Andrew Morton]

Steve Hill wrote:
> 
> I keep getting ResierFS oopses on my test box here.  It's a Cobalt Raq4i
> running the 2.4.17 kernel (with some of my own experimental code in the
> kernel but I don'tthink this is the cause of the problem).  The box has
> been powercycled and paniced more times than I've had hot dinners so I
> suspect that these oopses are being caused by hard drive corruption.  The
> ksymoops output is below:
> 
> ...
> 
> Unable to handle kernel NULL pointer dereference at virtual address
> 00000006
> ...
> 
> >>EIP; c0134c9d <get_hash_table+5d/90>   <=====
> Trace; c013536f <getblk+1f/50>

One of the buffer_head addresses (which should have been either a
valid address or zero) in fact had a value of 0x00000002.  Single
bit error; probably bad memory/CPU/etc.

-