Re: Ooops with kernel 2.4.22 and reiserfs

[Oleg Drokin <green@linuxhacker.ru>]

Hello!

On Tue, Dec 23, 2003 at 11:04:59AM +0100, Carlo wrote:
> > C> hda: set_drive_speed_status: status=0x58 { DriveReady SeekComplete
> > C> DataRequest }
> > C> ide0: Drive 0 didn't accept speed setting. Oh, well.
> > C> hda: dma_intr: status=0x58 { DriveReady SeekComplete DataRequest }
> > C> hda: CHECK for good STATUS
> > Do you always get these IDE errors prior to oops?
> No, on 6 oops only one has this error!

Ah, I see.

> > C> Unable to handle kernel paging request at virtual address ffffffe0
> ............
> > Also please run your oops through ksymoops.
> >>EIP; c0146553 <select_parent+33/a0>   <=====
> Trace; c01465dd <shrink_dcache_parent+1d/30>
> Trace; c013fa0f <d_unhash+2f/50>
> Trace; c018d840 <reiserfs_rmdir+0/270>
> Trace; c013fb69 <vfs_rmdir+139/1c0>
> Trace; c013fc84 <sys_rmdir+94/e0>
> Trace; c0114d00 <do_page_fault+0/48c>
> Trace; c01088a3 <system_call+33/38>

Sounds like a list corruption to me.
Are you sure the memory in that box is ok?

> If it's necessary i can send the full /var/log/messages  passed through
> ksymoops with a lot of "Unable to handle kernel paging request at
> virtual address" but it's aboute 500Kb.

Run it through ksymoops and see if backtrace is different each time.
Usually only first oops per boot is useful.
If backtrace is different each time, perhaps list all unique oopses?

Thank you.

Bye,
    Oleg