xfs: Check for extent overflow when renaming dir entries
A rename operation is essentially a directory entry remove operation from the perspective of parent directory (i.e. src_dp) of rename's source. Hence the only place where we check for extent count overflow for src_dp is in xfs_bmap_del_extent_real(). xfs_bmap_del_extent_real() returns -ENOSPC when it detects a possible extent count overflow and in response, the higher layers of directory handling code do the following: 1. Data/Free blocks: XFS lets these blocks linger until a future remove operation removes them. 2. Dabtree blocks: XFS swaps the blocks with the last block in the Leaf space and unmaps the last block. For target_dp, there are two cases depending on whether the destination directory entry exists or not. When destination directory entry does not exist (i.e. target_ip == NULL), extent count overflow check is performed only when transaction has a non-zero sized space reservation associated with it. With a zero-sized space reservation, XFS allows a rename operation to continue only when the directory has sufficient free space in its data/leaf/free space blocks to hold the new entry. When destination directory entry exists (i.e. target_ip != NULL), all we need to do is change the inode number associated with the already existing entry. Hence there is no need to perform an extent count overflow check. Signed-off-by: Chandan Babu R <chandanrlinux@gmail.com> Reviewed-by: Darrick J. Wong <darrick.wong@oracle.com> Signed-off-by: Darrick J. Wong <darrick.wong@oracle.com>
This commit is contained in:
parent
0dbc5cb1a9
commit
02092a2f03
@ -5160,6 +5160,9 @@ xfs_bmap_del_extent_real(
|
|||||||
* until a future remove operation. Dabtree blocks would be
|
* until a future remove operation. Dabtree blocks would be
|
||||||
* swapped with the last block in the leaf space and then the
|
* swapped with the last block in the leaf space and then the
|
||||||
* new last block will be unmapped.
|
* new last block will be unmapped.
|
||||||
|
*
|
||||||
|
* The above logic also applies to the source directory entry of
|
||||||
|
* a rename operation.
|
||||||
*/
|
*/
|
||||||
error = xfs_iext_count_may_overflow(ip, whichfork, 1);
|
error = xfs_iext_count_may_overflow(ip, whichfork, 1);
|
||||||
if (error) {
|
if (error) {
|
||||||
|
@ -3116,6 +3116,35 @@ xfs_rename(
|
|||||||
/*
|
/*
|
||||||
* Check for expected errors before we dirty the transaction
|
* Check for expected errors before we dirty the transaction
|
||||||
* so we can return an error without a transaction abort.
|
* so we can return an error without a transaction abort.
|
||||||
|
*
|
||||||
|
* Extent count overflow check:
|
||||||
|
*
|
||||||
|
* From the perspective of src_dp, a rename operation is essentially a
|
||||||
|
* directory entry remove operation. Hence the only place where we check
|
||||||
|
* for extent count overflow for src_dp is in
|
||||||
|
* xfs_bmap_del_extent_real(). xfs_bmap_del_extent_real() returns
|
||||||
|
* -ENOSPC when it detects a possible extent count overflow and in
|
||||||
|
* response, the higher layers of directory handling code do the
|
||||||
|
* following:
|
||||||
|
* 1. Data/Free blocks: XFS lets these blocks linger until a
|
||||||
|
* future remove operation removes them.
|
||||||
|
* 2. Dabtree blocks: XFS swaps the blocks with the last block in the
|
||||||
|
* Leaf space and unmaps the last block.
|
||||||
|
*
|
||||||
|
* For target_dp, there are two cases depending on whether the
|
||||||
|
* destination directory entry exists or not.
|
||||||
|
*
|
||||||
|
* When destination directory entry does not exist (i.e. target_ip ==
|
||||||
|
* NULL), extent count overflow check is performed only when transaction
|
||||||
|
* has a non-zero sized space reservation associated with it. With a
|
||||||
|
* zero-sized space reservation, XFS allows a rename operation to
|
||||||
|
* continue only when the directory has sufficient free space in its
|
||||||
|
* data/leaf/free space blocks to hold the new entry.
|
||||||
|
*
|
||||||
|
* When destination directory entry exists (i.e. target_ip != NULL), all
|
||||||
|
* we need to do is change the inode number associated with the already
|
||||||
|
* existing entry. Hence there is no need to perform an extent count
|
||||||
|
* overflow check.
|
||||||
*/
|
*/
|
||||||
if (target_ip == NULL) {
|
if (target_ip == NULL) {
|
||||||
/*
|
/*
|
||||||
@ -3126,6 +3155,12 @@ xfs_rename(
|
|||||||
error = xfs_dir_canenter(tp, target_dp, target_name);
|
error = xfs_dir_canenter(tp, target_dp, target_name);
|
||||||
if (error)
|
if (error)
|
||||||
goto out_trans_cancel;
|
goto out_trans_cancel;
|
||||||
|
} else {
|
||||||
|
error = xfs_iext_count_may_overflow(target_dp,
|
||||||
|
XFS_DATA_FORK,
|
||||||
|
XFS_IEXT_DIR_MANIP_CNT(mp));
|
||||||
|
if (error)
|
||||||
|
goto out_trans_cancel;
|
||||||
}
|
}
|
||||||
} else {
|
} else {
|
||||||
/*
|
/*
|
||||||
@ -3291,9 +3326,16 @@ xfs_rename(
|
|||||||
if (wip) {
|
if (wip) {
|
||||||
error = xfs_dir_replace(tp, src_dp, src_name, wip->i_ino,
|
error = xfs_dir_replace(tp, src_dp, src_name, wip->i_ino,
|
||||||
spaceres);
|
spaceres);
|
||||||
} else
|
} else {
|
||||||
|
/*
|
||||||
|
* NOTE: We don't need to check for extent count overflow here
|
||||||
|
* because the dir remove name code will leave the dir block in
|
||||||
|
* place if the extent count would overflow.
|
||||||
|
*/
|
||||||
error = xfs_dir_removename(tp, src_dp, src_name, src_ip->i_ino,
|
error = xfs_dir_removename(tp, src_dp, src_name, src_ip->i_ino,
|
||||||
spaceres);
|
spaceres);
|
||||||
|
}
|
||||||
|
|
||||||
if (error)
|
if (error)
|
||||||
goto out_trans_cancel;
|
goto out_trans_cancel;
|
||||||
|
|
||||||
|
Loading…
Reference in New Issue
Block a user