Re: [Jfs-discussion] [2.4.21-pre3] kernel BUG at jfs_metapage.c:295!

[Dave K. <sh...@au...>]

On Wednesday 29 January 2003 06:45, Axel Siebenwirth wrote:
> Hi all,
>
> long time it's when I saw an oops in JFS :)
> CONFIG_PREEMPT is enabled. If it's not enough informatik I provide,
> please let me know.

You've provided enough.  I can see what's going on.  A quick fix would 
be to just remove the assert(new) at jfs_metapage.c:295.  I'll have the 
correct fix shortly.  (Due to a recent change, add_index() is doing a 
read_metapage() where it used to do a get_metapage().  This sanity 
check would have allowed the get_metapage, but doesn't like the 
read_metapage.)

> Before the oops I also got this msg: assert(new)

That's the same problem.

> And afterwards this:
> <6>note: cvs[537] exited with preempt_count 2

Hitting an oops doesn't allow the code to release locks, etc.

Thanks,
Shaggy
-- 
David Kleikamp
IBM Linux Technology Center