ext4: fix deadlock in ext4_symlink() in ENOSPC conditions

ext4_symlink() cannot call __page_symlink() with transaction open.
__page_symlink() calls ext4_write_begin() which can wait for
transaction commit if we are running out of space thus causing a
deadlock. Also error recovery in ext4_truncate_failed_write() does not
count with the transaction being already started (although I'm not
aware of any particular deadlock here).

Fix the problem by stopping a transaction before calling
__page_symlink() (we have to be careful and put inode to orphan list
so that it gets deleted in case of crash) and starting another one
after __page_symlink() returns for addition of symlink into a
directory.

Signed-off-by: Jan Kara <jack@suse.cz>
Signed-off-by: "Theodore Ts'o" <tytso@mit.edu>

fs/ext4/namei.c

@@ -2250,6 +2250,7 @@ static int ext4_symlink(struct inode *dir,
 	handle_t *handle;
 	struct inode *inode;
 	int l, err, retries = 0;
+	int credits;
 
 	l = strlen(symname)+1;
 	if (l > dir->i_sb->s_blocksize)
@@ -2257,10 +2258,26 @@ static int ext4_symlink(struct inode *dir,
 
 	dquot_initialize(dir);
 
+	if (l > EXT4_N_BLOCKS * 4) {
+		/*
+		 * For non-fast symlinks, we just allocate inode and put it on
+		 * orphan list in the first transaction => we need bitmap,
+		 * group descriptor, sb, inode block, quota blocks.
+		 */
+		credits = 4 + EXT4_MAXQUOTAS_INIT_BLOCKS(dir->i_sb);
+	} else {
+		/*
+		 * Fast symlink. We have to add entry to directory
+		 * (EXT4_DATA_TRANS_BLOCKS + EXT4_INDEX_EXTRA_TRANS_BLOCKS),
+		 * allocate new inode (bitmap, group descriptor, inode block,
+		 * quota blocks, sb is already counted in previous macros).
+		 */
+		credits = EXT4_DATA_TRANS_BLOCKS(dir->i_sb) +
+			  EXT4_INDEX_EXTRA_TRANS_BLOCKS + 3 +
+			  EXT4_MAXQUOTAS_INIT_BLOCKS(dir->i_sb);
+	}
 retry:
-	handle = ext4_journal_start(dir, EXT4_DATA_TRANS_BLOCKS(dir->i_sb) +
-					EXT4_INDEX_EXTRA_TRANS_BLOCKS + 5 +
-					EXT4_MAXQUOTAS_INIT_BLOCKS(dir->i_sb));
+	handle = ext4_journal_start(dir, credits);
 	if (IS_ERR(handle))
 		return PTR_ERR(handle);
 
@@ -2273,21 +2290,44 @@ retry:
 	if (IS_ERR(inode))
 		goto out_stop;
 
-	if (l > sizeof(EXT4_I(inode)->i_data)) {
+	if (l > EXT4_N_BLOCKS * 4) {
 		inode->i_op = &ext4_symlink_inode_operations;
 		ext4_set_aops(inode);
 		/*
-		 * page_symlink() calls into ext4_prepare/commit_write.
-		 * We have a transaction open.  All is sweetness.  It also sets
-		 * i_size in generic_commit_write().
+		 * We cannot call page_symlink() with transaction started
+		 * because it calls into ext4_write_begin() which can wait
+		 * for transaction commit if we are running out of space
+		 * and thus we deadlock. So we have to stop transaction now
+		 * and restart it when symlink contents is written.
+		 * 
+		 * To keep fs consistent in case of crash, we have to put inode
+		 * to orphan list in the mean time.
 		 */
+		drop_nlink(inode);
+		err = ext4_orphan_add(handle, inode);
+		ext4_journal_stop(handle);
+		if (err)
+			goto err_drop_inode;
 		err = __page_symlink(inode, symname, l, 1);
+		if (err)
+			goto err_drop_inode;
+		/*
+		 * Now inode is being linked into dir (EXT4_DATA_TRANS_BLOCKS
+		 * + EXT4_INDEX_EXTRA_TRANS_BLOCKS), inode is also modified
+		 */
+		handle = ext4_journal_start(dir,
+				EXT4_DATA_TRANS_BLOCKS(dir->i_sb) +
+				EXT4_INDEX_EXTRA_TRANS_BLOCKS + 1);
+		if (IS_ERR(handle)) {
+			err = PTR_ERR(handle);
+			goto err_drop_inode;
+		}
+		inc_nlink(inode);
+		err = ext4_orphan_del(handle, inode);
 		if (err) {
+			ext4_journal_stop(handle);
 			clear_nlink(inode);
-			unlock_new_inode(inode);
-			ext4_mark_inode_dirty(handle, inode);
-			iput(inode);
-			goto out_stop;
+			goto err_drop_inode;
 		}
 	} else {
 		/* clear the extent format for fast symlink */
@@ -2303,6 +2343,10 @@ out_stop:
 	if (err == -ENOSPC && ext4_should_retry_alloc(dir->i_sb, &retries))
 		goto retry;
 	return err;
+err_drop_inode:
+	unlock_new_inode(inode);
+	iput(inode);
+	return err;
 }
 
 static int ext4_link(struct dentry *old_dentry,