ocfs2: Rename ocfs2_cp_xattr_cluster() to ocfs2_mv_xattr_buckets().
authorJoel Becker <joel.becker@oracle.com>
Wed, 26 Nov 2008 21:02:18 +0000 (13:02 -0800)
committerMark Fasheh <mfasheh@suse.com>
Mon, 5 Jan 2009 16:40:29 +0000 (08:40 -0800)
ocfs2_cp_xattr_cluster() takes the last cluster of an xattr extent,
copies its buckets to the front of a new extent, and then shrinks the bucket
count of the original extent.  So it's really moving the data, not
copying it.

While we're here, the function doesn't need a buffer_head for the old
extent, just the block number.

Signed-off-by: Joel Becker <joel.becker@oracle.com>
Signed-off-by: Mark Fasheh <mfasheh@suse.com>
fs/ocfs2/xattr.c

index 5be99666f02ced9127b330f9b63c62c68d6b1e3d..c1f2e0690747e3e94b0237904a2354e4d823cf8c 100644 (file)
@@ -3965,11 +3965,12 @@ static int ocfs2_cp_xattr_bucket(struct inode *inode,
        /*
         * Hey, if we're overwriting t_bucket, what difference does
         * ACCESS_CREATE vs ACCESS_WRITE make?  Well, if we allocated a new
-        * cluster to fill, we came here from ocfs2_cp_xattr_cluster(), and
-        * it is really new - ACCESS_CREATE is required.  But we also
-        * might have moved data out of t_bucket before extending back
-        * into it.  ocfs2_add_new_xattr_bucket() can do this - its call
-        * to ocfs2_add_new_xattr_cluster() may have created a new extent
+        * cluster to fill, we came here from
+        * ocfs2_mv_xattr_buckets(), and it is really new -
+        * ACCESS_CREATE is required.  But we also might have moved data
+        * out of t_bucket before extending back into it.
+        * ocfs2_add_new_xattr_bucket() can do this - its call to
+        * ocfs2_add_new_xattr_cluster() may have created a new extent
         * and copied out the end of the old extent.  Then it re-extends
         * the old extent back to create space for new xattrs.  That's
         * how we get here, and the bucket isn't really new.
@@ -3992,17 +3993,16 @@ out:
 }
 
 /*
- * src_blk points to the last cluster of an existing extent.  to_blk
- * points to a newly allocated extent.  We copy the cluster over to the
- * new extent, initializing its xh_num_buckets.  The old extent's
- * xh_num_buckets shrinks by the same amount.
+ * src_blk points to the start of an existing extent.  last_blk points to
+ * last cluster in that extent.  to_blk points to a newly allocated
+ * extent.  We copy the buckets from cluster at last_blk to the new extent,
+ * initializing its xh_num_buckets.  The old extent's xh_num_buckets
+ * shrinks by the same amount.
  */
-static int ocfs2_cp_xattr_cluster(struct inode *inode,
+static int ocfs2_mv_xattr_buckets(struct inode *inode,
                                  handle_t *handle,
-                                 struct buffer_head *first_bh,
-                                 u64 src_blk,
-                                 u64 to_blk,
-                                 u32 *first_hash)
+                                 u64 src_blk, u64 last_blk,
+                                 u64 to_blk, u32 *first_hash)
 {
        int i, ret, credits;
        struct ocfs2_super *osb = OCFS2_SB(inode->i_sb);
@@ -4011,8 +4011,8 @@ static int ocfs2_cp_xattr_cluster(struct inode *inode,
        int num_buckets = ocfs2_xattr_buckets_per_cluster(osb);
        struct ocfs2_xattr_bucket *old_first, *new_first;
 
-       mlog(0, "cp xattrs from cluster %llu to %llu\n",
-            (unsigned long long)src_blk, (unsigned long long)to_blk);
+       mlog(0, "mv xattrs from cluster %llu to %llu\n",
+            (unsigned long long)last_blk, (unsigned long long)to_blk);
 
        /* The first bucket of the original extent */
        old_first = ocfs2_xattr_bucket_new(inode);
@@ -4024,7 +4024,7 @@ static int ocfs2_cp_xattr_cluster(struct inode *inode,
                goto out;
        }
 
-       ret = ocfs2_read_xattr_bucket(old_first, first_bh->b_blocknr);
+       ret = ocfs2_read_xattr_bucket(old_first, src_blk);
        if (ret) {
                mlog_errno(ret);
                goto out;
@@ -4050,7 +4050,7 @@ static int ocfs2_cp_xattr_cluster(struct inode *inode,
 
        for (i = 0; i < num_buckets; i++) {
                ret = ocfs2_cp_xattr_bucket(inode, handle,
-                                           src_blk + (i * blks_per_bucket),
+                                           last_blk + (i * blks_per_bucket),
                                            to_blk + (i * blks_per_bucket),
                                            1);
                if (ret) {
@@ -4175,8 +4175,10 @@ static int ocfs2_adjust_xattr_cross_cluster(struct inode *inode,
                u64 last_blk = prev_blk + bpc * (prev_clusters - 1);
 
                if (prev_clusters > 1 && (*header_bh)->b_blocknr != last_blk)
-                       ret = ocfs2_cp_xattr_cluster(inode, handle, *first_bh,
-                                                    last_blk, new_blk,
+                       ret = ocfs2_mv_xattr_buckets(inode, handle,
+                                                    (*first_bh)->b_blocknr,
+                                                    last_blk,
+                                                    new_blk,
                                                     v_start);
                else {
                        ret = ocfs2_divide_xattr_cluster(inode, handle,