ext4: convert symlink external data block mapping to bdev

Symlink's external data block is one kind of metadata block, and now
that almost all ext4 metadata block's page cache (e.g. directory blocks,
quota blocks...) belongs to bdev backing inode except the symlink. It
is essentially worked in data=journal mode like other regular file's
data block because probably in order to make it simple for generic VFS
code handling symlinks or some other historical reasons, but the logic
of creating external data block in ext4_symlink() is complicated. and it
also make things confused if user do not want to let the filesystem
worked in data=journal mode. This patch convert the final exceptional
case and make things clean, move the mapping of the symlink's external
data block to bdev like any other metadata block does.

Signed-off-by: Zhang Yi <yi.zhang@huawei.com>
Link: https://lore.kernel.org/r/20220424140936.1898920-3-yi.zhang@huawei.com

1 files changed, 55 insertions, 68 deletions

diff --git a/fs/ext4/namei.c b/fs/ext4/namei.c
index b202626391ff..7aca8944901d 100644
--- a/fs/ext4/namei.c
+++ b/fs/ext4/namei.c
@@ -3249,6 +3249,32 @@ out_trace:
 	return retval;
 }
 
+static int ext4_init_symlink_block(handle_t *handle, struct inode *inode,
+				   struct fscrypt_str *disk_link)
+{
+	struct buffer_head *bh;
+	char *kaddr;
+	int err = 0;
+
+	bh = ext4_bread(handle, inode, 0, EXT4_GET_BLOCKS_CREATE);
+	if (IS_ERR(bh))
+		return PTR_ERR(bh);
+
+	BUFFER_TRACE(bh, "get_write_access");
+	err = ext4_journal_get_write_access(handle, inode->i_sb, bh, EXT4_JTR_NONE);
+	if (err)
+		goto out;
+
+	kaddr = (char *)bh->b_data;
+	memcpy(kaddr, disk_link->name, disk_link->len);
+	inode->i_size = disk_link->len - 1;
+	EXT4_I(inode)->i_disksize = inode->i_size;
+	err = ext4_handle_dirty_metadata(handle, inode, bh);
+out:
+	brelse(bh);
+	return err;
+}
+
 static int ext4_symlink(struct user_namespace *mnt_userns, struct inode *dir,
 			struct dentry *dentry, const char *symname)
 {
@@ -3257,6 +3283,7 @@ static int ext4_symlink(struct user_namespace *mnt_userns, struct inode *dir,
 	int err, len = strlen(symname);
 	int credits;
 	struct fscrypt_str disk_link;
+	int retries = 0;
 
 	if (unlikely(ext4_forced_shutdown(EXT4_SB(dir->i_sb))))
 		return -EIO;
@@ -3270,26 +3297,15 @@ static int ext4_symlink(struct user_namespace *mnt_userns, struct inode *dir,
 	if (err)
 		return err;
 
-	if ((disk_link.len > 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, and
-		 * possibly selinux xattr blocks.
-		 */
-		credits = 4 + EXT4_MAXQUOTAS_INIT_BLOCKS(dir->i_sb) +
-			  EXT4_XATTR_TRANS_BLOCKS;
-	} 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_INDEX_EXTRA_TRANS_BLOCKS for addition of entry into the
+	 * directory. +3 for inode, inode bitmap, group descriptor allocation.
+	 * EXT4_DATA_TRANS_BLOCKS for the data block allocation and
+	 * modification.
+	 */
+	credits = EXT4_DATA_TRANS_BLOCKS(dir->i_sb) +
+		  EXT4_INDEX_EXTRA_TRANS_BLOCKS + 3;
+retry:
 	inode = ext4_new_inode_start_handle(mnt_userns, dir, S_IFLNK|S_IRWXUGO,
 					    &dentry->d_name, 0, NULL,
 					    EXT4_HT_DIR, credits);
@@ -3297,7 +3313,8 @@ static int ext4_symlink(struct user_namespace *mnt_userns, struct inode *dir,
 	if (IS_ERR(inode)) {
 		if (handle)
 			ext4_journal_stop(handle);
-		return PTR_ERR(inode);
+		err = PTR_ERR(inode);
+		goto out_retry;
 	}
 
 	if (IS_ENCRYPTED(inode)) {
@@ -3305,74 +3322,44 @@ static int ext4_symlink(struct user_namespace *mnt_userns, struct inode *dir,
 		if (err)
 			goto err_drop_inode;
 		inode->i_op = &ext4_encrypted_symlink_inode_operations;
+	} else {
+		if ((disk_link.len > EXT4_N_BLOCKS * 4)) {
+			inode->i_op = &ext4_symlink_inode_operations;
+		} else {
+			inode->i_op = &ext4_fast_symlink_inode_operations;
+			inode->i_link = (char *)&EXT4_I(inode)->i_data;
+		}
 	}
 
 	if ((disk_link.len > EXT4_N_BLOCKS * 4)) {
-		if (!IS_ENCRYPTED(inode))
-			inode->i_op = &ext4_symlink_inode_operations;
-		inode_nohighmem(inode);
-		ext4_set_aops(inode);
-		/*
-		 * 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);
-		if (handle)
-			ext4_journal_stop(handle);
-		handle = NULL;
-		if (err)
-			goto err_drop_inode;
-		err = __page_symlink(inode, disk_link.name, disk_link.len, 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_HT_DIR,
-				EXT4_DATA_TRANS_BLOCKS(dir->i_sb) +
-				EXT4_INDEX_EXTRA_TRANS_BLOCKS + 1);
-		if (IS_ERR(handle)) {
-			err = PTR_ERR(handle);
-			handle = NULL;
-			goto err_drop_inode;
-		}
-		set_nlink(inode, 1);
-		err = ext4_orphan_del(handle, inode);
+		/* alloc symlink block and fill it */
+		err = ext4_init_symlink_block(handle, inode, &disk_link);
 		if (err)
 			goto err_drop_inode;
 	} else {
 		/* clear the extent format for fast symlink */
 		ext4_clear_inode_flag(inode, EXT4_INODE_EXTENTS);
-		if (!IS_ENCRYPTED(inode)) {
-			inode->i_op = &ext4_fast_symlink_inode_operations;
-			inode->i_link = (char *)&EXT4_I(inode)->i_data;
-		}
 		memcpy((char *)&EXT4_I(inode)->i_data, disk_link.name,
 		       disk_link.len);
 		inode->i_size = disk_link.len - 1;
+		EXT4_I(inode)->i_disksize = inode->i_size;
 	}
-	EXT4_I(inode)->i_disksize = inode->i_size;
 	err = ext4_add_nondir(handle, dentry, &inode);
 	if (handle)
 		ext4_journal_stop(handle);
 	iput(inode);
-	goto out_free_encrypted_link;
+	goto out_retry;
 
 err_drop_inode:
-	if (handle)
-		ext4_journal_stop(handle);
 	clear_nlink(inode);
+	ext4_orphan_add(handle, inode);
 	unlock_new_inode(inode);
+	if (handle)
+		ext4_journal_stop(handle);
 	iput(inode);
-out_free_encrypted_link:
+out_retry:
+	if (err == -ENOSPC && ext4_should_retry_alloc(dir->i_sb, &retries))
+		goto retry;
 	if (disk_link.name != (unsigned char *)symname)
 		kfree(disk_link.name);
 	return err;