Re: oops with kjournald in SMP 2.4.16

[Andrew Morton]

Alois Treindl wrote:
> 
> I had an OOPS and kernel crash last night,
> on a dual CPU 2.4.16 (Dell Poweredge 2450, dual 933 Mhz P3, 1.75 gb RAM).
> 
> The system had been up for about 100 days without reboot.
> 
> I use the ext3 file system with 3 SCSI disks and various NFS clients
> attached to this file server.
> 

[ oops in slab allocation for bh_cachep ]

This looks like random memory corruption - someone wrote somewhere
where they shouldn't have.

For a while we were seeing a lot of these.  Around 2.4.14 to 2.4.17.
I have twenty or thirty different reports saved away.

But they seem to have stopped.  It's beginning to look like whatever
it was got fixed somehow.

> 
> Does anyone recognize this problem, and has it been fixed in later kernels?
> 

Possibly so, yes.  2.4.16 is suspect in this regard.

-