jffs2 umount

[Andrew Lunn]

On Wed, Sep 24, 2003 at 12:17:20AM +0200, Thomas Koeller wrote:
> Andrew,
> 
> seems I was too quick to agree with you. A second look at the code
> reveals that the superblock _is_ actually freed, and so my original
> patch was almost correct. The only thing it missed was to zero out the
> data field in the mount table entry, so when the file system was
> mounted again, a stale superblock would have been used.
> 
> The current code is obviously incorrect as it accesses memory that
> has been freed. The attached patch should finally do it the right
> way now.
> 
> tk

Committed

        Andrew