From: Dave Kleikamp on
On Wed, 2010-04-14 at 15:56 -0400, Bill Pemberton wrote:
> Resizing the filesystem would result in an diAllocExt error in some
> instances because changes in bmp->db_agsize would not get noticed if
> goto extendBmap was called.

Awesome! This bug has been frustrating me for a while, but I hadn't
chased it down to the root cause. I'll give a bit of a test and push it
upstream!

Thanks,
Shaggy

>
> Signed-off-by: Bill Pemberton <wfp5p(a)virginia.edu>
> ---
> fs/jfs/resize.c | 6 +++++-
> 1 files changed, 5 insertions(+), 1 deletions(-)
>
> diff --git a/fs/jfs/resize.c b/fs/jfs/resize.c
> index 7f24a0b..1aba003 100644
> --- a/fs/jfs/resize.c
> +++ b/fs/jfs/resize.c
> @@ -81,6 +81,7 @@ int jfs_extendfs(struct super_block *sb, s64 newLVSize, int newLogSize)
> struct inode *iplist[1];
> struct jfs_superblock *j_sb, *j_sb2;
> uint old_agsize;
> + int agsizechanged = 0;
> struct buffer_head *bh, *bh2;
>
> /* If the volume hasn't grown, get out now */
> @@ -333,6 +334,9 @@ int jfs_extendfs(struct super_block *sb, s64 newLVSize, int newLogSize)
> */
> if ((rc = dbExtendFS(ipbmap, XAddress, nblocks)))
> goto error_out;
> +
> + agsizechanged |= (bmp->db_agsize != old_agsize);
> +
> /*
> * the map now has extended to cover additional nblocks:
> * dn_mapsize = oldMapsize + nblocks;
> @@ -432,7 +436,7 @@ int jfs_extendfs(struct super_block *sb, s64 newLVSize, int newLogSize)
> * will correctly identify the new ag);
> */
> /* if new AG size the same as old AG size, done! */
> - if (bmp->db_agsize != old_agsize) {
> + if (agsizechanged) {
> if ((rc = diExtendFS(ipimap, ipbmap)))
> goto error_out;
>
--
Dave Kleikamp
IBM Linux Technology Center

--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo(a)vger.kernel.org
More majordomo info at http://vger.kernel.org/majordomo-info.html
Please read the FAQ at http://www.tux.org/lkml/