diff options
| author | Yingping Lu <yingping@sgi.com> | 2005-11-03 10:31:47 +1100 |
|---|---|---|
| committer | Nathan Scott <nathans@sgi.com> | 2005-11-03 10:31:47 +1100 |
| commit | bf6f05aa0bb6f1fe9a9ce35287678c3916bddfef (patch) | |
| tree | 0707dd164ab0172a922b52e10212334f4675235e /fs | |
| parent | 19d5bcf370c64b1acb1e928b24baf3b68cec914b (diff) | |
[XFS] Fixed the inconsistency between attribute b-tree intermidiate node
and leaf blocks. The problem cam from xfsqa test 117.
SGI-PV: 940655
SGI-Modid: xfs-linux:xfs-kern:201527a
Signed-off-by: Yingping Lu <yingping@sgi.com>
Signed-off-by: Nathan Scott <nathans@sgi.com>
Diffstat (limited to 'fs')
| -rw-r--r-- | fs/xfs/xfs_da_btree.c | 3 |
1 files changed, 2 insertions, 1 deletions
diff --git a/fs/xfs/xfs_da_btree.c b/fs/xfs/xfs_da_btree.c index 6098aff7bd5..473671fa5c1 100644 --- a/fs/xfs/xfs_da_btree.c +++ b/fs/xfs/xfs_da_btree.c @@ -429,7 +429,8 @@ xfs_da_node_split(xfs_da_state_t *state, xfs_da_state_blk_t *oldblk, /* * With V2 the extra block is data or freespace. */ - useextra = state->extravalid && XFS_DIR_IS_V1(state->mp); + useextra = state->extravalid && (XFS_DIR_IS_V1(state->mp) || + state->args->whichfork == XFS_ATTR_FORK); newcount = 1 + useextra; /* * Do we have to split the node? |