1 files changed, 44 insertions, 77 deletions

diff --git a/fs/btrfs/ctree.c b/fs/btrfs/ctree.c
index 02fae7f7e42c..ed504607d8ec 100644
--- a/fs/btrfs/ctree.c
+++ b/fs/btrfs/ctree.c
@@ -1089,7 +1089,8 @@ static noinline int __btrfs_cow_block(struct btrfs_trans_handle *trans,
                 btrfs_set_node_ptr_generation(parent, parent_slot,
                                               trans->transid);
                 btrfs_mark_buffer_dirty(parent);
-                tree_mod_log_free_eb(root->fs_info, buf);
+                if (last_ref)
+                        tree_mod_log_free_eb(root->fs_info, buf);
                 btrfs_free_tree_block(trans, root, buf, parent_start,
                                       last_ref);
         }
@@ -1161,8 +1162,8 @@ __tree_mod_log_oldest_root(struct btrfs_fs_info *fs_info,
  * time_seq).
  */
 static void
-__tree_mod_log_rewind(struct extent_buffer *eb, u64 time_seq,
-                      struct tree_mod_elem *first_tm)
+__tree_mod_log_rewind(struct btrfs_fs_info *fs_info, struct extent_buffer *eb,
+                      u64 time_seq, struct tree_mod_elem *first_tm)
 {
         u32 n;
         struct rb_node *next;
@@ -1172,6 +1173,7 @@ __tree_mod_log_rewind(struct extent_buffer *eb, u64 time_seq,
         unsigned long p_size = sizeof(struct btrfs_key_ptr);
 
         n = btrfs_header_nritems(eb);
+        tree_mod_log_read_lock(fs_info);
         while (tm && tm->seq >= time_seq) {
                 /*
                  * all the operations are recorded with the operator used for
@@ -1226,6 +1228,7 @@ __tree_mod_log_rewind(struct extent_buffer *eb, u64 time_seq,
                 if (tm->index != first_tm->index)
                         break;
         }
+        tree_mod_log_read_unlock(fs_info);
         btrfs_set_header_nritems(eb, n);
 }
 
@@ -1268,13 +1271,12 @@ tree_mod_log_rewind(struct btrfs_fs_info *fs_info, struct extent_buffer *eb,
                 BUG_ON(!eb_rewin);
         }
 
-        extent_buffer_get(eb_rewin);
         btrfs_tree_read_unlock(eb);
         free_extent_buffer(eb);
 
         extent_buffer_get(eb_rewin);
         btrfs_tree_read_lock(eb_rewin);
-        __tree_mod_log_rewind(eb_rewin, time_seq, tm);
+        __tree_mod_log_rewind(fs_info, eb_rewin, time_seq, tm);
         WARN_ON(btrfs_header_nritems(eb_rewin) >
                 BTRFS_NODEPTRS_PER_BLOCK(fs_info->tree_root));
 
@@ -1350,7 +1352,7 @@ get_old_root(struct btrfs_root *root, u64 time_seq)
                 btrfs_set_header_generation(eb, old_generation);
         }
         if (tm)
-                __tree_mod_log_rewind(eb, time_seq, tm);
+                __tree_mod_log_rewind(root->fs_info, eb, time_seq, tm);
         else
                 WARN_ON(btrfs_header_level(eb) != 0);
         WARN_ON(btrfs_header_nritems(eb) > BTRFS_NODEPTRS_PER_BLOCK(root));
@@ -2178,12 +2180,8 @@ static void reada_for_search(struct btrfs_root *root,
         }
 }
 
-/*
- * returns -EAGAIN if it had to drop the path, or zero if everything was in
- * cache
- */
-static noinline int reada_for_balance(struct btrfs_root *root,
-                                      struct btrfs_path *path, int level)
+static noinline void reada_for_balance(struct btrfs_root *root,
+                                       struct btrfs_path *path, int level)
 {
         int slot;
         int nritems;
@@ -2192,12 +2190,11 @@ static noinline int reada_for_balance(struct btrfs_root *root,
         u64 gen;
         u64 block1 = 0;
         u64 block2 = 0;
-        int ret = 0;
         int blocksize;
 
         parent = path->nodes[level + 1];
         if (!parent)
-                return 0;
+                return;
 
         nritems = btrfs_header_nritems(parent);
         slot = path->slots[level + 1];
@@ -2224,28 +2221,11 @@ static noinline int reada_for_balance(struct btrfs_root *root,
                         block2 = 0;
                 free_extent_buffer(eb);
         }
-        if (block1 || block2) {
-                ret = -EAGAIN;
-
-                /* release the whole path */
-                btrfs_release_path(path);
 
-                /* read the blocks */
-                if (block1)
-                        readahead_tree_block(root, block1, blocksize, 0);
-                if (block2)
-                        readahead_tree_block(root, block2, blocksize, 0);
-
-                if (block1) {
-                        eb = read_tree_block(root, block1, blocksize, 0);
-                        free_extent_buffer(eb);
-                }
-                if (block2) {
-                        eb = read_tree_block(root, block2, blocksize, 0);
-                        free_extent_buffer(eb);
-                }
-        }
-        return ret;
+        if (block1)
+                readahead_tree_block(root, block1, blocksize, 0);
+        if (block2)
+                readahead_tree_block(root, block2, blocksize, 0);
 }
 
 
@@ -2359,35 +2339,28 @@ read_block_for_search(struct btrfs_trans_handle *trans,
         tmp = btrfs_find_tree_block(root, blocknr, blocksize);
         if (tmp) {
                 /* first we do an atomic uptodate check */
-                if (btrfs_buffer_uptodate(tmp, 0, 1) > 0) {
-                        if (btrfs_buffer_uptodate(tmp, gen, 1) > 0) {
-                                /*
-                                 * we found an up to date block without
-                                 * sleeping, return
-                                 * right away
-                                 */
-                                *eb_ret = tmp;
-                                return 0;
-                        }
-                        /* the pages were up to date, but we failed
-                         * the generation number check.  Do a full
-                         * read for the generation number that is correct.
-                         * We must do this without dropping locks so
-                         * we can trust our generation number
-                         */
-                        free_extent_buffer(tmp);
-                        btrfs_set_path_blocking(p);
+                if (btrfs_buffer_uptodate(tmp, gen, 1) > 0) {
+                        *eb_ret = tmp;
+                        return 0;
+                }
 
-                        /* now we're allowed to do a blocking uptodate check */
-                        tmp = read_tree_block(root, blocknr, blocksize, gen);
-                        if (tmp && btrfs_buffer_uptodate(tmp, gen, 0) > 0) {
-                                *eb_ret = tmp;
-                                return 0;
-                        }
-                        free_extent_buffer(tmp);
-                        btrfs_release_path(p);
-                        return -EIO;
+                /* the pages were up to date, but we failed
+                 * the generation number check.  Do a full
+                 * read for the generation number that is correct.
+                 * We must do this without dropping locks so
+                 * we can trust our generation number
+                 */
+                btrfs_set_path_blocking(p);
+
+                /* now we're allowed to do a blocking uptodate check */
+                ret = btrfs_read_buffer(tmp, gen);
+                if (!ret) {
+                        *eb_ret = tmp;
+                        return 0;
                 }
+                free_extent_buffer(tmp);
+                btrfs_release_path(p);
+                return -EIO;
         }
 
         /*
@@ -2448,11 +2421,8 @@ setup_nodes_for_search(struct btrfs_trans_handle *trans,
                         goto again;
                 }
 
-                sret = reada_for_balance(root, p, level);
-                if (sret)
-                        goto again;
-
                 btrfs_set_path_blocking(p);
+                reada_for_balance(root, p, level);
                 sret = split_node(trans, root, p, level);
                 btrfs_clear_path_blocking(p, NULL, 0);
 
@@ -2472,11 +2442,8 @@ setup_nodes_for_search(struct btrfs_trans_handle *trans,
                         goto again;
                 }
 
-                sret = reada_for_balance(root, p, level);
-                if (sret)
-                        goto again;
-
                 btrfs_set_path_blocking(p);
+                reada_for_balance(root, p, level);
                 sret = balance_level(trans, root, p, level);
                 btrfs_clear_path_blocking(p, NULL, 0);
 
@@ -3143,7 +3110,7 @@ static int balance_node_right(struct btrfs_trans_handle *trans,
  */
 static noinline int insert_new_root(struct btrfs_trans_handle *trans,
                            struct btrfs_root *root,
-                           struct btrfs_path *path, int level, int log_removal)
+                           struct btrfs_path *path, int level)
 {
         u64 lower_gen;
         struct extent_buffer *lower;
@@ -3194,7 +3161,7 @@ static noinline int insert_new_root(struct btrfs_trans_handle *trans,
         btrfs_mark_buffer_dirty(c);
 
         old = root->node;
-        tree_mod_log_set_root_pointer(root, c, log_removal);
+        tree_mod_log_set_root_pointer(root, c, 0);
         rcu_assign_pointer(root->node, c);
 
         /* the super has an extra ref to root->node */
@@ -3278,14 +3245,14 @@ static noinline int split_node(struct btrfs_trans_handle *trans,
                 /*
                  * trying to split the root, lets make a new one
                  *
-                 * tree mod log: We pass 0 as log_removal parameter to
+                 * tree mod log: We don't log_removal old root in
                  * insert_new_root, because that root buffer will be kept as a
                  * normal node. We are going to log removal of half of the
                  * elements below with tree_mod_log_eb_copy. We're holding a
                  * tree lock on the buffer, which is why we cannot race with
                  * other tree_mod_log users.
                  */
-                ret = insert_new_root(trans, root, path, level + 1, 0);
+                ret = insert_new_root(trans, root, path, level + 1);
                 if (ret)
                         return ret;
         } else {
@@ -3986,7 +3953,7 @@ static noinline int split_leaf(struct btrfs_trans_handle *trans,
                 return -EOVERFLOW;
 
         /* first try to make some room by pushing left and right */
-        if (data_size) {
+        if (data_size && path->nodes[1]) {
                 wret = push_leaf_right(trans, root, path, data_size,
                                        data_size, 0, 0);
                 if (wret < 0)
@@ -4005,7 +3972,7 @@ static noinline int split_leaf(struct btrfs_trans_handle *trans,
         }
 
         if (!path->nodes[1]) {
-                ret = insert_new_root(trans, root, path, 1, 1);
+                ret = insert_new_root(trans, root, path, 1);
                 if (ret)
                         return ret;
         }
@@ -4430,7 +4397,7 @@ void btrfs_truncate_item(struct btrfs_root *root, struct btrfs_path *path,
 }
 
 /*
- * make the item pointed to by the path bigger, data_size is the new size.
+ * make the item pointed to by the path bigger, data_size is the added size.
  */
 void btrfs_extend_item(struct btrfs_root *root, struct btrfs_path *path,
                        u32 data_size)