Hello,
Found the same here, just get the "kernel BUG at raid5.c:211"
When trying to mkraid command.
raid0 seems to work OK, but raid[15] not
BTW i (try) to use raid* on Logical Volumes, and my machine is not running
low on memory (just 100MB free)
I _did_ apply the (latest) reiserfs patch, so that _could_ be related
(sorry haven't tested that yet)
Greatings,
On Wed, 31 May 2000, Neil Brown wrote:
> On Tuesday May 30, nils@ipe.uni-stuttgart.de wrote:
> > What I did:
> >
> > create /etc/raidtab like this:
> >
> > raiddev /dev/md0
> > raid-level 5
> ...
> >
> > Then ran mkraid /dev/md0 and got:
> >
> > [...]
> > kernel BUG at raid5.c:211
> > invalid operand: 0000
> > CPU: 0
> > EIP: 0010:[<c01bed54>]
> > EFLAGS: 00010086
> > [...]
>
> Looks like a bug all right.
>
> When a raid5 array is enabled, a few megabytes of kernel memory is
> allocated to provide a stripe cache.
> If you get a kmalloc failure, it can call
> __put_free_stripe
> which expects the device to be locked, but at this stage it isn't.
>
> Could your machine be running low on memory?
>
> Probably the simplest fix is to initialise conf->device_lock
> to MD_SPIN_LOCK_LOCKED on line 1768, and the unlock the device when it
> is all set up, but I haven't looked closely.
>
> I will add this to my todo list.
>
> Thanks,
> NeilBrown
>
> Please read the FAQ at http://www.tux.org/lkml/
>
> ------------------------------
>
Arjan Filius
mailto:iafilius@xs4all.nl
-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.rutgers.edu
Please read the FAQ at http://www.tux.org/lkml/
This archive was generated by hypermail 2b29 : Wed Jun 07 2000 - 21:00:12 EST