Re: CVS commit: src/sys/ufs/ufs

["Darrin B.Jewell" <dbj%netbsd.org@localhost>]

Thanks for catching this.

Can you please request this change to be pulled up to the netbsd-2-0
branch?  The patch looks ok to me.

How did you uncover this problem?  I've been interested in testing
filesystems with large numbers of logical blocks by creating a 2tb
ufs2 filesystem with a 512 byte fragment size, but haven't yet done so
myself.  Do you know if there are other problems lurking for that
case?

Darrin

YAMAMOTO Takashi <yamt%netbsd.org@localhost> writes:

> Module Name:  src
> Committed By: yamt
> Date:         Wed Sep 15 09:52:49 UTC 2004
> 
> Modified Files:
>       src/sys/ufs/ufs: ufs_bmap.c
> 
> Log Message:
> ufs_getlbns:
> - fix an integer overflow when calculating lbns of indirect blocks.
> - remove a redundant calculation of blockcnt.
> 
> 
> To generate a diff of this commit:
> cvs rdiff -r1.32 -r1.33 src/sys/ufs/ufs/ufs_bmap.c
> 
> Please note that diffs are not public domain; they are subject to the
> copyright notices on the relevant files.