1 files changed, 75 insertions, 61 deletions

diff --git a/fs/btrfs/tree-log.c b/fs/btrfs/tree-log.c
index a3c43f0b1c95..200cea6e49e5 100644
--- a/fs/btrfs/tree-log.c
+++ b/fs/btrfs/tree-log.c
@@ -279,12 +279,6 @@ void btrfs_end_log_trans(struct btrfs_root *root)
 	}
 }
 
-static void btrfs_wait_tree_block_writeback(struct extent_buffer *buf)
-{
-	filemap_fdatawait_range(buf->pages[0]->mapping,
-			        buf->start, buf->start + buf->len - 1);
-}
-
 /*
  * the walk control struct is used to pass state down the chain when
  * processing the log tree.  The stage field tells us which part
@@ -2623,11 +2617,12 @@ static noinline int walk_down_log_tree(struct btrfs_trans_handle *trans,
 					return ret;
 				}
 
+				btrfs_tree_lock(next);
+				btrfs_clear_buffer_dirty(trans, next);
+				wait_on_extent_buffer_writeback(next);
+				btrfs_tree_unlock(next);
+
 				if (trans) {
-					btrfs_tree_lock(next);
-					btrfs_clean_tree_block(next);
-					btrfs_wait_tree_block_writeback(next);
-					btrfs_tree_unlock(next);
 					ret = btrfs_pin_reserved_extent(trans,
 							bytenr, blocksize);
 					if (ret) {
@@ -2637,8 +2632,6 @@ static noinline int walk_down_log_tree(struct btrfs_trans_handle *trans,
 					btrfs_redirty_list_add(
 						trans->transaction, next);
 				} else {
-					if (test_and_clear_bit(EXTENT_BUFFER_DIRTY, &next->bflags))
-						clear_extent_buffer_dirty(next);
 					unaccount_log_buffer(fs_info, bytenr);
 				}
 			}
@@ -2693,11 +2686,12 @@ static noinline int walk_up_log_tree(struct btrfs_trans_handle *trans,
 
 				next = path->nodes[*level];
 
+				btrfs_tree_lock(next);
+				btrfs_clear_buffer_dirty(trans, next);
+				wait_on_extent_buffer_writeback(next);
+				btrfs_tree_unlock(next);
+
 				if (trans) {
-					btrfs_tree_lock(next);
-					btrfs_clean_tree_block(next);
-					btrfs_wait_tree_block_writeback(next);
-					btrfs_tree_unlock(next);
 					ret = btrfs_pin_reserved_extent(trans,
 						     path->nodes[*level]->start,
 						     path->nodes[*level]->len);
@@ -2706,9 +2700,6 @@ static noinline int walk_up_log_tree(struct btrfs_trans_handle *trans,
 					btrfs_redirty_list_add(trans->transaction,
 							       next);
 				} else {
-					if (test_and_clear_bit(EXTENT_BUFFER_DIRTY, &next->bflags))
-						clear_extent_buffer_dirty(next);
-
 					unaccount_log_buffer(fs_info,
 						path->nodes[*level]->start);
 				}
@@ -2776,19 +2767,18 @@ static int walk_log_tree(struct btrfs_trans_handle *trans,
 
 			next = path->nodes[orig_level];
 
+			btrfs_tree_lock(next);
+			btrfs_clear_buffer_dirty(trans, next);
+			wait_on_extent_buffer_writeback(next);
+			btrfs_tree_unlock(next);
+
 			if (trans) {
-				btrfs_tree_lock(next);
-				btrfs_clean_tree_block(next);
-				btrfs_wait_tree_block_writeback(next);
-				btrfs_tree_unlock(next);
 				ret = btrfs_pin_reserved_extent(trans,
 						next->start, next->len);
 				if (ret)
 					goto out;
 				btrfs_redirty_list_add(trans->transaction, next);
 			} else {
-				if (test_and_clear_bit(EXTENT_BUFFER_DIRTY, &next->bflags))
-					clear_extent_buffer_dirty(next);
 				unaccount_log_buffer(fs_info, next->start);
 			}
 		}
@@ -2980,7 +2970,6 @@ int btrfs_sync_log(struct btrfs_trans_handle *trans,
 		ret = 0;
 	if (ret) {
 		blk_finish_plug(&plug);
-		btrfs_abort_transaction(trans, ret);
 		btrfs_set_log_full_commit(trans);
 		mutex_unlock(&root->log_mutex);
 		goto out;
@@ -3045,15 +3034,12 @@ int btrfs_sync_log(struct btrfs_trans_handle *trans,
 
 		blk_finish_plug(&plug);
 		btrfs_set_log_full_commit(trans);
-
-		if (ret != -ENOSPC) {
-			btrfs_abort_transaction(trans, ret);
-			mutex_unlock(&log_root_tree->log_mutex);
-			goto out;
-		}
+		if (ret != -ENOSPC)
+			btrfs_err(fs_info,
+				  "failed to update log for root %llu ret %d",
+				  root->root_key.objectid, ret);
 		btrfs_wait_tree_log_extents(log, mark);
 		mutex_unlock(&log_root_tree->log_mutex);
-		ret = BTRFS_LOG_FORCE_COMMIT;
 		goto out;
 	}
 
@@ -3112,7 +3098,6 @@ int btrfs_sync_log(struct btrfs_trans_handle *trans,
 		goto out_wake_log_root;
 	} else if (ret) {
 		btrfs_set_log_full_commit(trans);
-		btrfs_abort_transaction(trans, ret);
 		mutex_unlock(&log_root_tree->log_mutex);
 		goto out_wake_log_root;
 	}
@@ -3581,17 +3566,19 @@ static noinline int insert_dir_log_key(struct btrfs_trans_handle *trans,
 }
 
 static int flush_dir_items_batch(struct btrfs_trans_handle *trans,
-				 struct btrfs_root *log,
+				 struct btrfs_inode *inode,
 				 struct extent_buffer *src,
 				 struct btrfs_path *dst_path,
 				 int start_slot,
 				 int count)
 {
+	struct btrfs_root *log = inode->root->log_root;
 	char *ins_data = NULL;
 	struct btrfs_item_batch batch;
 	struct extent_buffer *dst;
 	unsigned long src_offset;
 	unsigned long dst_offset;
+	u64 last_index;
 	struct btrfs_key key;
 	u32 item_size;
 	int ret;
@@ -3649,6 +3636,18 @@ static int flush_dir_items_batch(struct btrfs_trans_handle *trans,
 	src_offset = btrfs_item_ptr_offset(src, start_slot + count - 1);
 	copy_extent_buffer(dst, src, dst_offset, src_offset, batch.total_data_size);
 	btrfs_release_path(dst_path);
+
+	last_index = batch.keys[count - 1].offset;
+	ASSERT(last_index > inode->last_dir_index_offset);
+
+	/*
+	 * If for some unexpected reason the last item's index is not greater
+	 * than the last index we logged, warn and force a transaction commit.
+	 */
+	if (WARN_ON(last_index <= inode->last_dir_index_offset))
+		ret = BTRFS_LOG_FORCE_COMMIT;
+	else
+		inode->last_dir_index_offset = last_index;
 out:
 	kfree(ins_data);
 
@@ -3698,7 +3697,6 @@ static int process_dir_items_leaf(struct btrfs_trans_handle *trans,
 		}
 
 		di = btrfs_item_ptr(src, i, struct btrfs_dir_item);
-		ctx->last_dir_item_offset = key.offset;
 
 		/*
 		 * Skip ranges of items that consist only of dir item keys created
@@ -3761,7 +3759,7 @@ static int process_dir_items_leaf(struct btrfs_trans_handle *trans,
 	if (batch_size > 0) {
 		int ret;
 
-		ret = flush_dir_items_batch(trans, log, src, dst_path,
+		ret = flush_dir_items_batch(trans, inode, src, dst_path,
 					    batch_start, batch_size);
 		if (ret < 0)
 			return ret;
@@ -3785,7 +3783,6 @@ static noinline int log_dir_items(struct btrfs_trans_handle *trans,
 	struct btrfs_key min_key;
 	struct btrfs_root *root = inode->root;
 	struct btrfs_root *log = root->log_root;
-	int err = 0;
 	int ret;
 	u64 last_old_dentry_offset = min_offset - 1;
 	u64 last_offset = (u64)-1;
@@ -3826,7 +3823,10 @@ static noinline int log_dir_items(struct btrfs_trans_handle *trans,
 					      path->slots[0]);
 			if (tmp.type == BTRFS_DIR_INDEX_KEY)
 				last_old_dentry_offset = tmp.offset;
+		} else if (ret > 0) {
+			ret = 0;
 		}
+
 		goto done;
 	}
 
@@ -3846,20 +3846,37 @@ static noinline int log_dir_items(struct btrfs_trans_handle *trans,
 		 */
 		if (tmp.type == BTRFS_DIR_INDEX_KEY)
 			last_old_dentry_offset = tmp.offset;
+	} else if (ret < 0) {
+		goto done;
 	}
+
 	btrfs_release_path(path);
 
 	/*
-	 * Find the first key from this transaction again.  See the note for
-	 * log_new_dir_dentries, if we're logging a directory recursively we
-	 * won't be holding its i_mutex, which means we can modify the directory
-	 * while we're logging it.  If we remove an entry between our first
-	 * search and this search we'll not find the key again and can just
-	 * bail.
+	 * Find the first key from this transaction again or the one we were at
+	 * in the loop below in case we had to reschedule. We may be logging the
+	 * directory without holding its VFS lock, which happen when logging new
+	 * dentries (through log_new_dir_dentries()) or in some cases when we
+	 * need to log the parent directory of an inode. This means a dir index
+	 * key might be deleted from the inode's root, and therefore we may not
+	 * find it anymore. If we can't find it, just move to the next key. We
+	 * can not bail out and ignore, because if we do that we will simply
+	 * not log dir index keys that come after the one that was just deleted
+	 * and we can end up logging a dir index range that ends at (u64)-1
+	 * (@last_offset is initialized to that), resulting in removing dir
+	 * entries we should not remove at log replay time.
 	 */
 search:
 	ret = btrfs_search_slot(NULL, root, &min_key, path, 0, 0);
-	if (ret != 0)
+	if (ret > 0) {
+		ret = btrfs_next_item(root, path);
+		if (ret > 0) {
+			/* There are no more keys in the inode's root. */
+			ret = 0;
+			goto done;
+		}
+	}
+	if (ret < 0)
 		goto done;
 
 	/*
@@ -3870,8 +3887,8 @@ search:
 		ret = process_dir_items_leaf(trans, inode, path, dst_path, ctx,
 					     &last_old_dentry_offset);
 		if (ret != 0) {
-			if (ret < 0)
-				err = ret;
+			if (ret > 0)
+				ret = 0;
 			goto done;
 		}
 		path->slots[0] = btrfs_header_nritems(path->nodes[0]);
@@ -3882,10 +3899,10 @@ search:
 		 */
 		ret = btrfs_next_leaf(root, path);
 		if (ret) {
-			if (ret == 1)
+			if (ret == 1) {
 				last_offset = (u64)-1;
-			else
-				err = ret;
+				ret = 0;
+			}
 			goto done;
 		}
 		btrfs_item_key_to_cpu(path->nodes[0], &min_key, path->slots[0]);
@@ -3916,7 +3933,7 @@ done:
 	btrfs_release_path(path);
 	btrfs_release_path(dst_path);
 
-	if (err == 0) {
+	if (ret == 0) {
 		*last_offset_ret = last_offset;
 		/*
 		 * In case the leaf was changed in the current transaction but
@@ -3927,15 +3944,13 @@ done:
 		 * a range, last_old_dentry_offset is == to last_offset.
 		 */
 		ASSERT(last_old_dentry_offset <= last_offset);
-		if (last_old_dentry_offset < last_offset) {
+		if (last_old_dentry_offset < last_offset)
 			ret = insert_dir_log_key(trans, log, path, ino,
 						 last_old_dentry_offset + 1,
 						 last_offset);
-			if (ret)
-				err = ret;
-		}
 	}
-	return err;
+
+	return ret;
 }
 
 /*
@@ -4031,7 +4046,6 @@ static noinline int log_directory_changes(struct btrfs_trans_handle *trans,
 
 	min_key = BTRFS_DIR_START_INDEX;
 	max_key = 0;
-	ctx->last_dir_item_offset = inode->last_dir_index_offset;
 
 	while (1) {
 		ret = log_dir_items(trans, inode, path, dst_path,
@@ -4043,8 +4057,6 @@ static noinline int log_directory_changes(struct btrfs_trans_handle *trans,
 		min_key = max_key + 1;
 	}
 
-	inode->last_dir_index_offset = ctx->last_dir_item_offset;
-
 	return 0;
 }
 
@@ -6440,7 +6452,6 @@ static int btrfs_log_inode(struct btrfs_trans_handle *trans,
 	 * result in losing the file after a log replay.
 	 */
 	if (full_dir_logging && inode->last_unlink_trans >= trans->transid) {
-		btrfs_set_log_full_commit(trans);
 		ret = BTRFS_LOG_FORCE_COMMIT;
 		goto out_unlock;
 	}
@@ -7459,8 +7470,11 @@ void btrfs_log_new_name(struct btrfs_trans_handle *trans,
 		 * not fail, but if it does, it's not serious, just bail out and
 		 * mark the log for a full commit.
 		 */
-		if (WARN_ON_ONCE(ret < 0))
+		if (WARN_ON_ONCE(ret < 0)) {
+			fscrypt_free_filename(&fname);
 			goto out;
+		}
+
 		log_pinned = true;
 
 		path = btrfs_alloc_path();