On Thu, Sep 15, 2005 at 10:29:38PM +0200, Vincent wrote:
> Eric,
> 
> okay, what you suggest is just that this is a normal behaviour after a 
> *bad* crash, and not related to whatever hidden upgrade of the FFS code. 
> I'll make a backup of whatever is important and start over.

	Well, no, it's not normal behaviour, unless by "bad crash" you mean
"corrupt data was written to the disk".  I don't think just pulling
the plug should cause this kind of error.

	If by "hidden upgrade" you mean the recent change to 64-bit ino_t,
then it's possible that there was something that affected you, but
the on-disk size should still be 32-bit, so it might just be a bug in
fsck_ffs reading the value from disk.

eric