ext4: add ext4_sb_bread() to disambiguate ENOMEM cases

Today, when sb_bread() returns NULL, this can either be because of an
I/O error or because the system failed to allocate the buffer.  Since
it's an old interface, changing would require changing many call
sites.

So instead we create our own ext4_sb_bread(), which also allows us to
set the REQ_META flag.

Also fixed a problem in the xattr code where a NULL return in a
function could also mean that the xattr was not found, which could
lead to the wrong error getting returned to userspace.

Fixes: ac27a0ec11 ("ext4: initial copy of files from ext3")
Cc: stable@kernel.org # 2.6.19
Signed-off-by: Theodore Ts'o <tytso@mit.edu>

fs/ext4/ext4.h

@@ -2538,6 +2538,8 @@ extern int ext4_group_extend(struct super_block *sb,
 extern int ext4_resize_fs(struct super_block *sb, ext4_fsblk_t n_blocks_count);
 
 /* super.c */
+extern struct buffer_head *ext4_sb_bread(struct super_block *sb,
+					 sector_t block, int op_flags);
 extern int ext4_seq_options_show(struct seq_file *seq, void *offset);
 extern int ext4_calculate_overhead(struct super_block *sb);
 extern void ext4_superblock_csum_set(struct super_block *sb);

fs/ext4/migrate.c

@@ -116,9 +116,9 @@ static int update_ind_extent_range(handle_t *handle, struct inode *inode,
 	int i, retval = 0;
 	unsigned long max_entries = inode->i_sb->s_blocksize >> 2;
 
-	bh = sb_bread(inode->i_sb, pblock);
+	bh = ext4_sb_bread(inode->i_sb, pblock, 0);
-	if (!bh)
+	if (IS_ERR(bh))
-		return -EIO;
+		return PTR_ERR(bh);
 
 	i_data = (__le32 *)bh->b_data;
 	for (i = 0; i < max_entries; i++) {
@@ -145,9 +145,9 @@ static int update_dind_extent_range(handle_t *handle, struct inode *inode,
 	int i, retval = 0;
 	unsigned long max_entries = inode->i_sb->s_blocksize >> 2;
 
-	bh = sb_bread(inode->i_sb, pblock);
+	bh = ext4_sb_bread(inode->i_sb, pblock, 0);
-	if (!bh)
+	if (IS_ERR(bh))
-		return -EIO;
+		return PTR_ERR(bh);
 
 	i_data = (__le32 *)bh->b_data;
 	for (i = 0; i < max_entries; i++) {
@@ -175,9 +175,9 @@ static int update_tind_extent_range(handle_t *handle, struct inode *inode,
 	int i, retval = 0;
 	unsigned long max_entries = inode->i_sb->s_blocksize >> 2;
 
-	bh = sb_bread(inode->i_sb, pblock);
+	bh = ext4_sb_bread(inode->i_sb, pblock, 0);
-	if (!bh)
+	if (IS_ERR(bh))
-		return -EIO;
+		return PTR_ERR(bh);
 
 	i_data = (__le32 *)bh->b_data;
 	for (i = 0; i < max_entries; i++) {
@@ -224,9 +224,9 @@ static int free_dind_blocks(handle_t *handle,
 	struct buffer_head *bh;
 	unsigned long max_entries = inode->i_sb->s_blocksize >> 2;
 
-	bh = sb_bread(inode->i_sb, le32_to_cpu(i_data));
+	bh = ext4_sb_bread(inode->i_sb, le32_to_cpu(i_data), 0);
-	if (!bh)
+	if (IS_ERR(bh))
-		return -EIO;
+		return PTR_ERR(bh);
 
 	tmp_idata = (__le32 *)bh->b_data;
 	for (i = 0; i < max_entries; i++) {
@@ -254,9 +254,9 @@ static int free_tind_blocks(handle_t *handle,
 	struct buffer_head *bh;
 	unsigned long max_entries = inode->i_sb->s_blocksize >> 2;
 
-	bh = sb_bread(inode->i_sb, le32_to_cpu(i_data));
+	bh = ext4_sb_bread(inode->i_sb, le32_to_cpu(i_data), 0);
-	if (!bh)
+	if (IS_ERR(bh))
-		return -EIO;
+		return PTR_ERR(bh);
 
 	tmp_idata = (__le32 *)bh->b_data;
 	for (i = 0; i < max_entries; i++) {
@@ -382,9 +382,9 @@ static int free_ext_idx(handle_t *handle, struct inode *inode,
 	struct ext4_extent_header *eh;
 
 	block = ext4_idx_pblock(ix);
-	bh = sb_bread(inode->i_sb, block);
+	bh = ext4_sb_bread(inode->i_sb, block, 0);
-	if (!bh)
+	if (IS_ERR(bh))
-		return -EIO;
+		return PTR_ERR(bh);
 
 	eh = (struct ext4_extent_header *)bh->b_data;
 	if (eh->eh_depth != 0) {

fs/ext4/resize.c

@@ -127,10 +127,12 @@ static int verify_group_input(struct super_block *sb,
 	else if (free_blocks_count < 0)
 		ext4_warning(sb, "Bad blocks count %u",
 			     input->blocks_count);
-	else if (!(bh = sb_bread(sb, end - 1)))
+	else if (IS_ERR(bh = ext4_sb_bread(sb, end - 1, 0))) {
+		err = PTR_ERR(bh);
+		bh = NULL;
 		ext4_warning(sb, "Cannot read last block (%llu)",
 			     end - 1);
-	else if (outside(input->block_bitmap, start, end))
+	} else if (outside(input->block_bitmap, start, end))
 		ext4_warning(sb, "Block bitmap not in group (block %llu)",
 			     (unsigned long long)input->block_bitmap);
 	else if (outside(input->inode_bitmap, start, end))
@@ -781,11 +783,11 @@ static int add_new_gdb(handle_t *handle, struct inode *inode,
 	struct ext4_super_block *es = EXT4_SB(sb)->s_es;
 	unsigned long gdb_num = group / EXT4_DESC_PER_BLOCK(sb);
 	ext4_fsblk_t gdblock = EXT4_SB(sb)->s_sbh->b_blocknr + 1 + gdb_num;
-	struct buffer_head **o_group_desc, **n_group_desc;
+	struct buffer_head **o_group_desc, **n_group_desc = NULL;
-	struct buffer_head *dind;
+	struct buffer_head *dind = NULL;
-	struct buffer_head *gdb_bh;
+	struct buffer_head *gdb_bh = NULL;
 	int gdbackups;
-	struct ext4_iloc iloc;
+	struct ext4_iloc iloc = { .bh = NULL };
 	__le32 *data;
 	int err;
 
@@ -794,21 +796,22 @@ static int add_new_gdb(handle_t *handle, struct inode *inode,
 		       "EXT4-fs: ext4_add_new_gdb: adding group block %lu\n",
 		       gdb_num);
 
-	gdb_bh = sb_bread(sb, gdblock);
+	gdb_bh = ext4_sb_bread(sb, gdblock, 0);
-	if (!gdb_bh)
+	if (IS_ERR(gdb_bh))
-		return -EIO;
+		return PTR_ERR(gdb_bh);
 
 	gdbackups = verify_reserved_gdb(sb, group, gdb_bh);
 	if (gdbackups < 0) {
 		err = gdbackups;
-		goto exit_bh;
+		goto errout;
 	}
 
 	data = EXT4_I(inode)->i_data + EXT4_DIND_BLOCK;
-	dind = sb_bread(sb, le32_to_cpu(*data));
+	dind = ext4_sb_bread(sb, le32_to_cpu(*data), 0);
-	if (!dind) {
+	if (IS_ERR(dind)) {
-		err = -EIO;
+		err = PTR_ERR(dind);
-		goto exit_bh;
+		dind = NULL;
+		goto errout;
 	}
 
 	data = (__le32 *)dind->b_data;
@@ -816,18 +819,18 @@ static int add_new_gdb(handle_t *handle, struct inode *inode,
 		ext4_warning(sb, "new group %u GDT block %llu not reserved",
 			     group, gdblock);
 		err = -EINVAL;
-		goto exit_dind;
+		goto errout;
 	}
 
 	BUFFER_TRACE(EXT4_SB(sb)->s_sbh, "get_write_access");
 	err = ext4_journal_get_write_access(handle, EXT4_SB(sb)->s_sbh);
 	if (unlikely(err))
-		goto exit_dind;
+		goto errout;
 
 	BUFFER_TRACE(gdb_bh, "get_write_access");
 	err = ext4_journal_get_write_access(handle, gdb_bh);
 	if (unlikely(err))
-		goto exit_dind;
+		goto errout;
 
 	BUFFER_TRACE(dind, "get_write_access");
 	err = ext4_journal_get_write_access(handle, dind);
@@ -837,7 +840,7 @@ static int add_new_gdb(handle_t *handle, struct inode *inode,
 	/* ext4_reserve_inode_write() gets a reference on the iloc */
 	err = ext4_reserve_inode_write(handle, inode, &iloc);
 	if (unlikely(err))
-		goto exit_dind;
+		goto errout;
 
 	n_group_desc = ext4_kvmalloc((gdb_num + 1) *
 				     sizeof(struct buffer_head *),
@@ -846,7 +849,7 @@ static int add_new_gdb(handle_t *handle, struct inode *inode,
 		err = -ENOMEM;
 		ext4_warning(sb, "not enough memory for %lu groups",
 			     gdb_num + 1);
-		goto exit_inode;
+		goto errout;
 	}
 
 	/*
@@ -862,7 +865,7 @@ static int add_new_gdb(handle_t *handle, struct inode *inode,
 	err = ext4_handle_dirty_metadata(handle, NULL, dind);
 	if (unlikely(err)) {
 		ext4_std_error(sb, err);
-		goto exit_inode;
+		goto errout;
 	}
 	inode->i_blocks -= (gdbackups + 1) * sb->s_blocksize >>
 			   (9 - EXT4_SB(sb)->s_cluster_bits);
@@ -871,8 +874,7 @@ static int add_new_gdb(handle_t *handle, struct inode *inode,
 	err = ext4_handle_dirty_metadata(handle, NULL, gdb_bh);
 	if (unlikely(err)) {
 		ext4_std_error(sb, err);
-		iloc.bh = NULL;
+		goto errout;
-		goto exit_inode;
 	}
 	brelse(dind);
 
@@ -888,15 +890,11 @@ static int add_new_gdb(handle_t *handle, struct inode *inode,
 	err = ext4_handle_dirty_super(handle, sb);
 	if (err)
 		ext4_std_error(sb, err);
-
 	return err;
-
+errout:
-exit_inode:
 	kvfree(n_group_desc);
 	brelse(iloc.bh);
-exit_dind:
 	brelse(dind);
-exit_bh:
 	brelse(gdb_bh);
 
 	ext4_debug("leaving with error %d\n", err);
@@ -916,9 +914,9 @@ static int add_new_gdb_meta_bg(struct super_block *sb,
 
 	gdblock = ext4_meta_bg_first_block_no(sb, group) +
 		   ext4_bg_has_super(sb, group);
-	gdb_bh = sb_bread(sb, gdblock);
+	gdb_bh = ext4_sb_bread(sb, gdblock, 0);
-	if (!gdb_bh)
+	if (IS_ERR(gdb_bh))
-		return -EIO;
+		return PTR_ERR(gdb_bh);
 	n_group_desc = ext4_kvmalloc((gdb_num + 1) *
 				     sizeof(struct buffer_head *),
 				     GFP_NOFS);
@@ -975,9 +973,10 @@ static int reserve_backup_gdb(handle_t *handle, struct inode *inode,
 		return -ENOMEM;
 
 	data = EXT4_I(inode)->i_data + EXT4_DIND_BLOCK;
-	dind = sb_bread(sb, le32_to_cpu(*data));
+	dind = ext4_sb_bread(sb, le32_to_cpu(*data), 0);
-	if (!dind) {
+	if (IS_ERR(dind)) {
-		err = -EIO;
+		err = PTR_ERR(dind);
+		dind = NULL;
 		goto exit_free;
 	}
 
@@ -996,9 +995,10 @@ static int reserve_backup_gdb(handle_t *handle, struct inode *inode,
 			err = -EINVAL;
 			goto exit_bh;
 		}
-		primary[res] = sb_bread(sb, blk);
+		primary[res] = ext4_sb_bread(sb, blk, 0);
-		if (!primary[res]) {
+		if (IS_ERR(primary[res])) {
-			err = -EIO;
+			err = PTR_ERR(primary[res]);
+			primary[res] = NULL;
 			goto exit_bh;
 		}
 		gdbackups = verify_reserved_gdb(sb, group, primary[res]);

fs/ext4/super.c

@@ -140,6 +140,29 @@ MODULE_ALIAS_FS("ext3");
 MODULE_ALIAS("ext3");
 #define IS_EXT3_SB(sb) ((sb)->s_bdev->bd_holder == &ext3_fs_type)
 
+/*
+ * This works like sb_bread() except it uses ERR_PTR for error
+ * returns.  Currently with sb_bread it's impossible to distinguish
+ * between ENOMEM and EIO situations (since both result in a NULL
+ * return.
+ */
+struct buffer_head *
+ext4_sb_bread(struct super_block *sb, sector_t block, int op_flags)
+{
+	struct buffer_head *bh = sb_getblk(sb, block);
+
+	if (bh == NULL)
+		return ERR_PTR(-ENOMEM);
+	if (buffer_uptodate(bh))
+		return bh;
+	ll_rw_block(REQ_OP_READ, REQ_META | op_flags, 1, &bh);
+	wait_on_buffer(bh);
+	if (buffer_uptodate(bh))
+		return bh;
+	put_bh(bh);
+	return ERR_PTR(-EIO);
+}
+
 static int ext4_verify_csum_type(struct super_block *sb,
 				 struct ext4_super_block *es)
 {

fs/ext4/xattr.c

@@ -522,14 +522,13 @@ ext4_xattr_block_get(struct inode *inode, int name_index, const char *name,
 	ea_idebug(inode, "name=%d.%s, buffer=%p, buffer_size=%ld",
 		  name_index, name, buffer, (long)buffer_size);
 
-	error = -ENODATA;
 	if (!EXT4_I(inode)->i_file_acl)
-		goto cleanup;
+		return -ENODATA;
 	ea_idebug(inode, "reading block %llu",
 		  (unsigned long long)EXT4_I(inode)->i_file_acl);
-	bh = sb_bread(inode->i_sb, EXT4_I(inode)->i_file_acl);
+	bh = ext4_sb_bread(inode->i_sb, EXT4_I(inode)->i_file_acl, REQ_PRIO);
-	if (!bh)
+	if (IS_ERR(bh))
-		goto cleanup;
+		return PTR_ERR(bh);
 	ea_bdebug(bh, "b_count=%d, refcount=%d",
 		atomic_read(&(bh->b_count)), le32_to_cpu(BHDR(bh)->h_refcount));
 	error = ext4_xattr_check_block(inode, bh);
@@ -696,26 +695,23 @@ ext4_xattr_block_list(struct dentry *dentry, char *buffer, size_t buffer_size)
 	ea_idebug(inode, "buffer=%p, buffer_size=%ld",
 		  buffer, (long)buffer_size);
 
-	error = 0;
 	if (!EXT4_I(inode)->i_file_acl)
-		goto cleanup;
+		return 0;
 	ea_idebug(inode, "reading block %llu",
 		  (unsigned long long)EXT4_I(inode)->i_file_acl);
-	bh = sb_bread(inode->i_sb, EXT4_I(inode)->i_file_acl);
+	bh = ext4_sb_bread(inode->i_sb, EXT4_I(inode)->i_file_acl, REQ_PRIO);
-	error = -EIO;
+	if (IS_ERR(bh))
-	if (!bh)
+		return PTR_ERR(bh);
-		goto cleanup;
 	ea_bdebug(bh, "b_count=%d, refcount=%d",
 		atomic_read(&(bh->b_count)), le32_to_cpu(BHDR(bh)->h_refcount));
 	error = ext4_xattr_check_block(inode, bh);
 	if (error)
 		goto cleanup;
 	ext4_xattr_block_cache_insert(EA_BLOCK_CACHE(inode), bh);
-	error = ext4_xattr_list_entries(dentry, BFIRST(bh), buffer, buffer_size);
+	error = ext4_xattr_list_entries(dentry, BFIRST(bh), buffer,
-
+					buffer_size);
 cleanup:
 	brelse(bh);
-
 	return error;
 }
 
@@ -830,9 +826,9 @@ int ext4_get_inode_usage(struct inode *inode, qsize_t *usage)
 	}
 
 	if (EXT4_I(inode)->i_file_acl) {
-		bh = sb_bread(inode->i_sb, EXT4_I(inode)->i_file_acl);
+		bh = ext4_sb_bread(inode->i_sb, EXT4_I(inode)->i_file_acl, REQ_PRIO);
-		if (!bh) {
+		if (IS_ERR(bh)) {
-			ret = -EIO;
+			ret = PTR_ERR(bh);
 			goto out;
 		}
 
@@ -1821,16 +1817,15 @@ ext4_xattr_block_find(struct inode *inode, struct ext4_xattr_info *i,
 
 	if (EXT4_I(inode)->i_file_acl) {
 		/* The inode already has an extended attribute block. */
-		bs->bh = sb_bread(sb, EXT4_I(inode)->i_file_acl);
+		bs->bh = ext4_sb_bread(sb, EXT4_I(inode)->i_file_acl, REQ_PRIO);
-		error = -EIO;
+		if (IS_ERR(bs->bh))
-		if (!bs->bh)
+			return PTR_ERR(bs->bh);
-			goto cleanup;
 		ea_bdebug(bs->bh, "b_count=%d, refcount=%d",
 			atomic_read(&(bs->bh->b_count)),
 			le32_to_cpu(BHDR(bs->bh)->h_refcount));
 		error = ext4_xattr_check_block(inode, bs->bh);
 		if (error)
-			goto cleanup;
+			return error;
 		/* Find the named attribute. */
 		bs->s.base = BHDR(bs->bh);
 		bs->s.first = BFIRST(bs->bh);
@@ -1839,13 +1834,10 @@ ext4_xattr_block_find(struct inode *inode, struct ext4_xattr_info *i,
 		error = xattr_find_entry(inode, &bs->s.here, bs->s.end,
 					 i->name_index, i->name, 1);
 		if (error && error != -ENODATA)
-			goto cleanup;
+			return error;
 		bs->s.not_found = error;
 	}
-	error = 0;
+	return 0;
-
-cleanup:
-	return error;
 }
 
 static int
@@ -2274,9 +2266,9 @@ static struct buffer_head *ext4_xattr_get_block(struct inode *inode)
 
 	if (!EXT4_I(inode)->i_file_acl)
 		return NULL;
-	bh = sb_bread(inode->i_sb, EXT4_I(inode)->i_file_acl);
+	bh = ext4_sb_bread(inode->i_sb, EXT4_I(inode)->i_file_acl, REQ_PRIO);
-	if (!bh)
+	if (IS_ERR(bh))
-		return ERR_PTR(-EIO);
+		return bh;
 	error = ext4_xattr_check_block(inode, bh);
 	if (error) {
 		brelse(bh);
@@ -2746,10 +2738,11 @@ retry:
 	if (EXT4_I(inode)->i_file_acl) {
 		struct buffer_head *bh;
 
-		bh = sb_bread(inode->i_sb, EXT4_I(inode)->i_file_acl);
+		bh = ext4_sb_bread(inode->i_sb, EXT4_I(inode)->i_file_acl, REQ_PRIO);
-		error = -EIO;
+		if (IS_ERR(bh)) {
-		if (!bh)
+			error = PTR_ERR(bh);
 			goto cleanup;
+		}
 		error = ext4_xattr_check_block(inode, bh);
 		if (error) {
 			brelse(bh);
@@ -2903,11 +2896,12 @@ int ext4_xattr_delete_inode(handle_t *handle, struct inode *inode,
 	}
 
 	if (EXT4_I(inode)->i_file_acl) {
-		bh = sb_bread(inode->i_sb, EXT4_I(inode)->i_file_acl);
+		bh = ext4_sb_bread(inode->i_sb, EXT4_I(inode)->i_file_acl, REQ_PRIO);
-		if (!bh) {
+		if (IS_ERR(bh)) {
+			error = PTR_ERR(bh);
+			if (error == -EIO)
 				EXT4_ERROR_INODE(inode, "block %llu read error",
 						 EXT4_I(inode)->i_file_acl);
-			error = -EIO;
 			goto cleanup;
 		}
 		error = ext4_xattr_check_block(inode, bh);
@@ -3060,8 +3054,10 @@ ext4_xattr_block_cache_find(struct inode *inode,
 	while (ce) {
 		struct buffer_head *bh;
 
-		bh = sb_bread(inode->i_sb, ce->e_value);
+		bh = ext4_sb_bread(inode->i_sb, ce->e_value, REQ_PRIO);
-		if (!bh) {
+		if (IS_ERR(bh)) {
+			if (PTR_ERR(bh) == -ENOMEM)
+				return NULL;
 			EXT4_ERROR_INODE(inode, "block %lu read error",
 					 (unsigned long)ce->e_value);
 		} else if (ext4_xattr_cmp(header, BHDR(bh)) == 0) {