Merge tag 'ext4_for_linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tytso/ext4

Pull ext4 updates from Ted Ts'o:
 "All cleanups and bug fixes; most notably, fix some problems discovered
  in ext4's NFS support, and fix an ioctl (EXT4_IOC_GROUP_ADD) used by
  old versions of e2fsprogs which we accidentally broke a while back.

  Also fixed some error paths in ext4's quota and inline data support.

  Finally, improve tail latency in jbd2's commit code"

* tag 'ext4_for_linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tytso/ext4:
  ext4: check for shutdown and r/o file system in ext4_write_inode()
  ext4: force inode writes when nfsd calls commit_metadata()
  ext4: avoid declaring fs inconsistent due to invalid file handles
  ext4: include terminating u32 in size of xattr entries when expanding inodes
  ext4: compare old and new mode before setting update_mode flag
  ext4: fix EXT4_IOC_GROUP_ADD ioctl
  ext4: hard fail dax mount on unsupported devices
  jbd2: update locking documentation for transaction_t
  ext4: remove redundant condition check
  jbd2: clean up indentation issue, replace spaces with tab
  ext4: clean up indentation issues, remove extraneous tabs
  ext4: missing unlock/put_page() in ext4_try_to_write_inline_data()
  ext4: fix possible use after free in ext4_quota_enable
  jbd2: avoid long hold times of j_state_lock while committing a transaction
  ext4: add ext4_sb_bread() to disambiguate ENOMEM cases

fs/ext4/acl.c

@@ -248,6 +248,7 @@ ext4_set_acl(struct inode *inode, struct posix_acl *acl, int type)
 		error = posix_acl_update_mode(inode, &mode, &acl);
 		if (error)
 			goto out_stop;
+		if (mode != inode->i_mode)
 			update_mode = 1;
 	}
 

fs/ext4/ext4.h

@@ -2454,8 +2454,19 @@ int do_journal_get_write_access(handle_t *handle,
 #define FALL_BACK_TO_NONDELALLOC 1
 #define CONVERT_INLINE_DATA	 2
 
-extern struct inode *ext4_iget(struct super_block *, unsigned long);
-extern struct inode *ext4_iget_normal(struct super_block *, unsigned long);
+typedef enum {
+	EXT4_IGET_NORMAL =	0,
+	EXT4_IGET_SPECIAL =	0x0001, /* OK to iget a system inode */
+	EXT4_IGET_HANDLE = 	0x0002	/* Inode # is from a handle */
+} ext4_iget_flags;
+
+extern struct inode *__ext4_iget(struct super_block *sb, unsigned long ino,
+				 ext4_iget_flags flags, const char *function,
+				 unsigned int line);
+
+#define ext4_iget(sb, ino, flags) \
+	__ext4_iget((sb), (ino), (flags), __func__, __LINE__)
+
 extern int  ext4_write_inode(struct inode *, struct writeback_control *);
 extern int  ext4_setattr(struct dentry *, struct iattr *);
 extern int  ext4_getattr(const struct path *, struct kstat *, u32, unsigned int);
@@ -2538,6 +2549,8 @@ 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/ialloc.c

@@ -1225,7 +1225,7 @@ struct inode *ext4_orphan_get(struct super_block *sb, unsigned long ino)
 	if (!ext4_test_bit(bit, bitmap_bh->b_data))
 		goto bad_orphan;
 
-	inode = ext4_iget(sb, ino);
+	inode = ext4_iget(sb, ino, EXT4_IGET_NORMAL);
 	if (IS_ERR(inode)) {
 		err = PTR_ERR(inode);
 		ext4_error(sb, "couldn't read orphan inode %lu (err %d)",

fs/ext4/inline.c

@@ -705,9 +705,12 @@ int ext4_try_to_write_inline_data(struct address_space *mapping,
 
 	if (!PageUptodate(page)) {
 		ret = ext4_read_inline_page(inode, page);
-		if (ret < 0)
+		if (ret < 0) {
+			unlock_page(page);
+			put_page(page);
 			goto out_up_read;
 		}
+	}
 
 	ret = 1;
 	handle = NULL;

fs/ext4/inode.c

@@ -4817,7 +4817,9 @@ static inline u64 ext4_inode_peek_iversion(const struct inode *inode)
 		return inode_peek_iversion(inode);
 }
 
-struct inode *ext4_iget(struct super_block *sb, unsigned long ino)
+struct inode *__ext4_iget(struct super_block *sb, unsigned long ino,
+			  ext4_iget_flags flags, const char *function,
+			  unsigned int line)
 {
 	struct ext4_iloc iloc;
 	struct ext4_inode *raw_inode;
@@ -4831,6 +4833,18 @@ struct inode *ext4_iget(struct super_block *sb, unsigned long ino)
 	gid_t i_gid;
 	projid_t i_projid;
 
+	if (((flags & EXT4_IGET_NORMAL) &&
+	     (ino < EXT4_FIRST_INO(sb) && ino != EXT4_ROOT_INO)) ||
+	    (ino < EXT4_ROOT_INO) ||
+	    (ino > le32_to_cpu(EXT4_SB(sb)->s_es->s_inodes_count))) {
+		if (flags & EXT4_IGET_HANDLE)
+			return ERR_PTR(-ESTALE);
+		__ext4_error(sb, function, line,
+			     "inode #%lu: comm %s: iget: illegal inode #",
+			     ino, current->comm);
+		return ERR_PTR(-EFSCORRUPTED);
+	}
+
 	inode = iget_locked(sb, ino);
 	if (!inode)
 		return ERR_PTR(-ENOMEM);
@@ -4846,18 +4860,26 @@ struct inode *ext4_iget(struct super_block *sb, unsigned long ino)
 	raw_inode = ext4_raw_inode(&iloc);
 
 	if ((ino == EXT4_ROOT_INO) && (raw_inode->i_links_count == 0)) {
-		EXT4_ERROR_INODE(inode, "root inode unallocated");
+		ext4_error_inode(inode, function, line, 0,
+				 "iget: root inode unallocated");
 		ret = -EFSCORRUPTED;
 		goto bad_inode;
 	}
 
+	if ((flags & EXT4_IGET_HANDLE) &&
+	    (raw_inode->i_links_count == 0) && (raw_inode->i_mode == 0)) {
+		ret = -ESTALE;
+		goto bad_inode;
+	}
+
 	if (EXT4_INODE_SIZE(inode->i_sb) > EXT4_GOOD_OLD_INODE_SIZE) {
 		ei->i_extra_isize = le16_to_cpu(raw_inode->i_extra_isize);
 		if (EXT4_GOOD_OLD_INODE_SIZE + ei->i_extra_isize >
 			EXT4_INODE_SIZE(inode->i_sb) ||
 		    (ei->i_extra_isize & 3)) {
-			EXT4_ERROR_INODE(inode,
-					 "bad extra_isize %u (inode size %u)",
+			ext4_error_inode(inode, function, line, 0,
+					 "iget: bad extra_isize %u "
+					 "(inode size %u)",
 					 ei->i_extra_isize,
 					 EXT4_INODE_SIZE(inode->i_sb));
 			ret = -EFSCORRUPTED;
@@ -4879,7 +4901,8 @@ struct inode *ext4_iget(struct super_block *sb, unsigned long ino)
 	}
 
 	if (!ext4_inode_csum_verify(inode, raw_inode, ei)) {
-		EXT4_ERROR_INODE(inode, "checksum invalid");
+		ext4_error_inode(inode, function, line, 0,
+				 "iget: checksum invalid");
 		ret = -EFSBADCRC;
 		goto bad_inode;
 	}
@@ -4936,7 +4959,8 @@ struct inode *ext4_iget(struct super_block *sb, unsigned long ino)
 			((__u64)le16_to_cpu(raw_inode->i_file_acl_high)) << 32;
 	inode->i_size = ext4_isize(sb, raw_inode);
 	if ((size = i_size_read(inode)) < 0) {
-		EXT4_ERROR_INODE(inode, "bad i_size value: %lld", size);
+		ext4_error_inode(inode, function, line, 0,
+				 "iget: bad i_size value: %lld", size);
 		ret = -EFSCORRUPTED;
 		goto bad_inode;
 	}
@@ -5012,7 +5036,8 @@ struct inode *ext4_iget(struct super_block *sb, unsigned long ino)
 	ret = 0;
 	if (ei->i_file_acl &&
 	    !ext4_data_block_valid(EXT4_SB(sb), ei->i_file_acl, 1)) {
-		EXT4_ERROR_INODE(inode, "bad extended attribute block %llu",
+		ext4_error_inode(inode, function, line, 0,
+				 "iget: bad extended attribute block %llu",
 				 ei->i_file_acl);
 		ret = -EFSCORRUPTED;
 		goto bad_inode;
@@ -5040,8 +5065,9 @@ struct inode *ext4_iget(struct super_block *sb, unsigned long ino)
 	} else if (S_ISLNK(inode->i_mode)) {
 		/* VFS does not allow setting these so must be corruption */
 		if (IS_APPEND(inode) || IS_IMMUTABLE(inode)) {
-			EXT4_ERROR_INODE(inode,
-			  "immutable or append flags not allowed on symlinks");
+			ext4_error_inode(inode, function, line, 0,
+					 "iget: immutable or append flags "
+					 "not allowed on symlinks");
 			ret = -EFSCORRUPTED;
 			goto bad_inode;
 		}
@@ -5071,7 +5097,8 @@ struct inode *ext4_iget(struct super_block *sb, unsigned long ino)
 		make_bad_inode(inode);
 	} else {
 		ret = -EFSCORRUPTED;
-		EXT4_ERROR_INODE(inode, "bogus i_mode (%o)", inode->i_mode);
+		ext4_error_inode(inode, function, line, 0,
+				 "iget: bogus i_mode (%o)", inode->i_mode);
 		goto bad_inode;
 	}
 	brelse(iloc.bh);
@@ -5085,13 +5112,6 @@ struct inode *ext4_iget(struct super_block *sb, unsigned long ino)
 	return ERR_PTR(ret);
 }
 
-struct inode *ext4_iget_normal(struct super_block *sb, unsigned long ino)
-{
-	if (ino < EXT4_FIRST_INO(sb) && ino != EXT4_ROOT_INO)
-		return ERR_PTR(-EFSCORRUPTED);
-	return ext4_iget(sb, ino);
-}
-
 static int ext4_inode_blocks_set(handle_t *handle,
 				struct ext4_inode *raw_inode,
 				struct ext4_inode_info *ei)
@@ -5380,9 +5400,13 @@ int ext4_write_inode(struct inode *inode, struct writeback_control *wbc)
 {
 	int err;
 
-	if (WARN_ON_ONCE(current->flags & PF_MEMALLOC))
+	if (WARN_ON_ONCE(current->flags & PF_MEMALLOC) ||
+	    sb_rdonly(inode->i_sb))
 		return 0;
 
+	if (unlikely(ext4_forced_shutdown(EXT4_SB(inode->i_sb))))
+		return -EIO;
+
 	if (EXT4_SB(inode->i_sb)->s_journal) {
 		if (ext4_journal_current_handle()) {
 			jbd_debug(1, "called recursively, non-PF_MEMALLOC!\n");
@@ -5398,7 +5422,8 @@ int ext4_write_inode(struct inode *inode, struct writeback_control *wbc)
 		if (wbc->sync_mode != WB_SYNC_ALL || wbc->for_sync)
 			return 0;
 
-		err = ext4_force_commit(inode->i_sb);
+		err = jbd2_complete_transaction(EXT4_SB(inode->i_sb)->s_journal,
+						EXT4_I(inode)->i_sync_tid);
 	} else {
 		struct ext4_iloc iloc;
 

fs/ext4/ioctl.c

@@ -125,7 +125,7 @@ static long swap_inode_boot_loader(struct super_block *sb,
 	    !inode_owner_or_capable(inode) || !capable(CAP_SYS_ADMIN))
 		return -EPERM;
 
-	inode_bl = ext4_iget(sb, EXT4_BOOT_LOADER_INO);
+	inode_bl = ext4_iget(sb, EXT4_BOOT_LOADER_INO, EXT4_IGET_SPECIAL);
 	if (IS_ERR(inode_bl))
 		return PTR_ERR(inode_bl);
 	ei_bl = EXT4_I(inode_bl);

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);
-	if (!bh)
-		return -EIO;
+	bh = ext4_sb_bread(inode->i_sb, pblock, 0);
+	if (IS_ERR(bh))
+		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);
-	if (!bh)
-		return -EIO;
+	bh = ext4_sb_bread(inode->i_sb, pblock, 0);
+	if (IS_ERR(bh))
+		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);
-	if (!bh)
-		return -EIO;
+	bh = ext4_sb_bread(inode->i_sb, pblock, 0);
+	if (IS_ERR(bh))
+		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));
-	if (!bh)
-		return -EIO;
+	bh = ext4_sb_bread(inode->i_sb, le32_to_cpu(i_data), 0);
+	if (IS_ERR(bh))
+		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));
-	if (!bh)
-		return -EIO;
+	bh = ext4_sb_bread(inode->i_sb, le32_to_cpu(i_data), 0);
+	if (IS_ERR(bh))
+		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);
-	if (!bh)
-		return -EIO;
+	bh = ext4_sb_bread(inode->i_sb, block, 0);
+	if (IS_ERR(bh))
+		return PTR_ERR(bh);
 
 	eh = (struct ext4_extent_header *)bh->b_data;
 	if (eh->eh_depth != 0) {

fs/ext4/namei.c

@@ -1571,7 +1571,7 @@ static struct dentry *ext4_lookup(struct inode *dir, struct dentry *dentry, unsi
 					 dentry);
 			return ERR_PTR(-EFSCORRUPTED);
 		}
-		inode = ext4_iget_normal(dir->i_sb, ino);
+		inode = ext4_iget(dir->i_sb, ino, EXT4_IGET_NORMAL);
 		if (inode == ERR_PTR(-ESTALE)) {
 			EXT4_ERROR_INODE(dir,
 					 "deleted inode referenced: %u",
@@ -1613,7 +1613,7 @@ struct dentry *ext4_get_parent(struct dentry *child)
 		return ERR_PTR(-EFSCORRUPTED);
 	}
 
-	return d_obtain_alias(ext4_iget_normal(child->d_sb, ino));
+	return d_obtain_alias(ext4_iget(child->d_sb, ino, EXT4_IGET_NORMAL));
 }
 
 /*

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 *dind;
-	struct buffer_head *gdb_bh;
+	struct buffer_head **o_group_desc, **n_group_desc = NULL;
+	struct buffer_head *dind = NULL;
+	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);
-	if (!gdb_bh)
-		return -EIO;
+	gdb_bh = ext4_sb_bread(sb, gdblock, 0);
+	if (IS_ERR(gdb_bh))
+		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));
-	if (!dind) {
-		err = -EIO;
-		goto exit_bh;
+	dind = ext4_sb_bread(sb, le32_to_cpu(*data), 0);
+	if (IS_ERR(dind)) {
+		err = PTR_ERR(dind);
+		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 exit_inode;
+		goto errout;
 	}
 	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;
-
-exit_inode:
+errout:
 	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);
-	if (!gdb_bh)
-		return -EIO;
+	gdb_bh = ext4_sb_bread(sb, gdblock, 0);
+	if (IS_ERR(gdb_bh))
+		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));
-	if (!dind) {
-		err = -EIO;
+	dind = ext4_sb_bread(sb, le32_to_cpu(*data), 0);
+	if (IS_ERR(dind)) {
+		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);
-		if (!primary[res]) {
-			err = -EIO;
+		primary[res] = ext4_sb_bread(sb, blk, 0);
+		if (IS_ERR(primary[res])) {
+			err = PTR_ERR(primary[res]);
+			primary[res] = NULL;
 			goto exit_bh;
 		}
 		gdbackups = verify_reserved_gdb(sb, group, primary[res]);
@@ -1631,13 +1631,13 @@ int ext4_group_add(struct super_block *sb, struct ext4_new_group_data *input)
 	}
 
 	if (reserved_gdb || gdb_off == 0) {
-		if (ext4_has_feature_resize_inode(sb) ||
+		if (!ext4_has_feature_resize_inode(sb) ||
 		    !le16_to_cpu(es->s_reserved_gdt_blocks)) {
 			ext4_warning(sb,
 				     "No reserved GDT blocks, can't resize");
 			return -EPERM;
 		}
-		inode = ext4_iget(sb, EXT4_RESIZE_INO);
+		inode = ext4_iget(sb, EXT4_RESIZE_INO, EXT4_IGET_SPECIAL);
 		if (IS_ERR(inode)) {
 			ext4_warning(sb, "Error opening resize inode");
 			return PTR_ERR(inode);
@@ -1965,7 +1965,8 @@ int ext4_resize_fs(struct super_block *sb, ext4_fsblk_t n_blocks_count)
 		}
 
 		if (!resize_inode)
-			resize_inode = ext4_iget(sb, EXT4_RESIZE_INO);
+			resize_inode = ext4_iget(sb, EXT4_RESIZE_INO,
+						 EXT4_IGET_SPECIAL);
 		if (IS_ERR(resize_inode)) {
 			ext4_warning(sb, "Error opening resize inode");
 			return PTR_ERR(resize_inode);

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)
 {
@@ -1000,14 +1023,13 @@ static void ext4_put_super(struct super_block *sb)
 		invalidate_bdev(sbi->journal_bdev);
 		ext4_blkdev_remove(sbi);
 	}
-	if (sbi->s_ea_inode_cache) {
+
 	ext4_xattr_destroy_cache(sbi->s_ea_inode_cache);
 	sbi->s_ea_inode_cache = NULL;
-	}
-	if (sbi->s_ea_block_cache) {
+
 	ext4_xattr_destroy_cache(sbi->s_ea_block_cache);
 	sbi->s_ea_block_cache = NULL;
-	}
+
 	if (sbi->s_mmp_tsk)
 		kthread_stop(sbi->s_mmp_tsk);
 	brelse(sbi->s_sbh);
@@ -1151,20 +1173,11 @@ static struct inode *ext4_nfs_get_inode(struct super_block *sb,
 {
 	struct inode *inode;
 
-	if (ino < EXT4_FIRST_INO(sb) && ino != EXT4_ROOT_INO)
-		return ERR_PTR(-ESTALE);
-	if (ino > le32_to_cpu(EXT4_SB(sb)->s_es->s_inodes_count))
-		return ERR_PTR(-ESTALE);
-
-	/* iget isn't really right if the inode is currently unallocated!!
-	 *
-	 * ext4_read_inode will return a bad_inode if the inode had been
-	 * deleted, so we should be safe.
-	 *
+	/*
 	 * Currently we don't know the generation for parent directory, so
 	 * a generation of 0 means "accept any"
 	 */
-	inode = ext4_iget_normal(sb, ino);
+	inode = ext4_iget(sb, ino, EXT4_IGET_HANDLE);
 	if (IS_ERR(inode))
 		return ERR_CAST(inode);
 	if (generation && inode->i_generation != generation) {
@@ -1189,6 +1202,16 @@ static struct dentry *ext4_fh_to_parent(struct super_block *sb, struct fid *fid,
 				    ext4_nfs_get_inode);
 }
 
+static int ext4_nfs_commit_metadata(struct inode *inode)
+{
+	struct writeback_control wbc = {
+		.sync_mode = WB_SYNC_ALL
+	};
+
+	trace_ext4_nfs_commit_metadata(inode);
+	return ext4_write_inode(inode, &wbc);
+}
+
 /*
  * Try to release metadata pages (indirect blocks, directories) which are
  * mapped via the block device.  Since these pages could have journal heads
@@ -1393,6 +1416,7 @@ static const struct export_operations ext4_export_ops = {
 	.fh_to_dentry = ext4_fh_to_dentry,
 	.fh_to_parent = ext4_fh_to_parent,
 	.get_parent = ext4_get_parent,
+	.commit_metadata = ext4_nfs_commit_metadata,
 };
 
 enum {
@@ -3842,12 +3866,12 @@ static int ext4_fill_super(struct super_block *sb, void *data, int silent)
 		if (ext4_has_feature_inline_data(sb)) {
 			ext4_msg(sb, KERN_ERR, "Cannot use DAX on a filesystem"
 					" that may contain inline data");
-			sbi->s_mount_opt &= ~EXT4_MOUNT_DAX;
+			goto failed_mount;
 		}
 		if (!bdev_dax_supported(sb->s_bdev, blocksize)) {
 			ext4_msg(sb, KERN_ERR,
-				"DAX unsupported by block device. Turning off DAX.");
-			sbi->s_mount_opt &= ~EXT4_MOUNT_DAX;
+				"DAX unsupported by block device.");
+			goto failed_mount;
 		}
 	}
 
@@ -4328,7 +4352,7 @@ static int ext4_fill_super(struct super_block *sb, void *data, int silent)
 	 * so we can safely mount the rest of the filesystem now.
 	 */
 
-	root = ext4_iget(sb, EXT4_ROOT_INO);
+	root = ext4_iget(sb, EXT4_ROOT_INO, EXT4_IGET_SPECIAL);
 	if (IS_ERR(root)) {
 		ext4_msg(sb, KERN_ERR, "get root inode failed");
 		ret = PTR_ERR(root);
@@ -4522,14 +4546,12 @@ static int ext4_fill_super(struct super_block *sb, void *data, int silent)
 	if (EXT4_SB(sb)->rsv_conversion_wq)
 		destroy_workqueue(EXT4_SB(sb)->rsv_conversion_wq);
 failed_mount_wq:
-	if (sbi->s_ea_inode_cache) {
 	ext4_xattr_destroy_cache(sbi->s_ea_inode_cache);
 	sbi->s_ea_inode_cache = NULL;
-	}
-	if (sbi->s_ea_block_cache) {
+
 	ext4_xattr_destroy_cache(sbi->s_ea_block_cache);
 	sbi->s_ea_block_cache = NULL;
-	}
+
 	if (sbi->s_journal) {
 		jbd2_journal_destroy(sbi->s_journal);
 		sbi->s_journal = NULL;
@@ -4598,7 +4620,7 @@ static struct inode *ext4_get_journal_inode(struct super_block *sb,
 	 * happen if we iget() an unused inode, as the subsequent iput()
 	 * will try to delete it.
 	 */
-	journal_inode = ext4_iget(sb, journal_inum);
+	journal_inode = ext4_iget(sb, journal_inum, EXT4_IGET_SPECIAL);
 	if (IS_ERR(journal_inode)) {
 		ext4_msg(sb, KERN_ERR, "no journal found");
 		return NULL;
@@ -5680,7 +5702,7 @@ static int ext4_quota_enable(struct super_block *sb, int type, int format_id,
 	if (!qf_inums[type])
 		return -EPERM;
 
-	qf_inode = ext4_iget(sb, qf_inums[type]);
+	qf_inode = ext4_iget(sb, qf_inums[type], EXT4_IGET_SPECIAL);
 	if (IS_ERR(qf_inode)) {
 		ext4_error(sb, "Bad quota inode # %lu", qf_inums[type]);
 		return PTR_ERR(qf_inode);
@@ -5690,9 +5712,9 @@ static int ext4_quota_enable(struct super_block *sb, int type, int format_id,
 	qf_inode->i_flags |= S_NOQUOTA;
 	lockdep_set_quota_inode(qf_inode, I_DATA_SEM_QUOTA);
 	err = dquot_enable(qf_inode, type, format_id, flags);
-	iput(qf_inode);
 	if (err)
 		lockdep_set_quota_inode(qf_inode, I_DATA_SEM_NORMAL);
+	iput(qf_inode);
 
 	return err;
 }

fs/ext4/xattr.c

@@ -384,7 +384,7 @@ static int ext4_xattr_inode_iget(struct inode *parent, unsigned long ea_ino,
 	struct inode *inode;
 	int err;
 
-	inode = ext4_iget(parent->i_sb, ea_ino);
+	inode = ext4_iget(parent->i_sb, ea_ino, EXT4_IGET_NORMAL);
 	if (IS_ERR(inode)) {
 		err = PTR_ERR(inode);
 		ext4_error(parent->i_sb,
@@ -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);
-	if (!bh)
-		goto cleanup;
+	bh = ext4_sb_bread(inode->i_sb, EXT4_I(inode)->i_file_acl, REQ_PRIO);
+	if (IS_ERR(bh))
+		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);
-	error = -EIO;
-	if (!bh)
-		goto cleanup;
+	bh = ext4_sb_bread(inode->i_sb, EXT4_I(inode)->i_file_acl, REQ_PRIO);
+	if (IS_ERR(bh))
+		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);
 	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);
-		if (!bh) {
-			ret = -EIO;
+		bh = ext4_sb_bread(inode->i_sb, EXT4_I(inode)->i_file_acl, REQ_PRIO);
+		if (IS_ERR(bh)) {
+			ret = PTR_ERR(bh);
 			goto out;
 		}
 
@@ -1486,7 +1482,8 @@ ext4_xattr_inode_cache_find(struct inode *inode, const void *value,
 	}
 
 	while (ce) {
-		ea_inode = ext4_iget(inode->i_sb, ce->e_value);
+		ea_inode = ext4_iget(inode->i_sb, ce->e_value,
+				     EXT4_IGET_NORMAL);
 		if (!IS_ERR(ea_inode) &&
 		    !is_bad_inode(ea_inode) &&
 		    (EXT4_I(ea_inode)->i_flags & EXT4_EA_INODE_FL) &&
@@ -1821,16 +1818,15 @@ 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);
-		error = -EIO;
-		if (!bs->bh)
-			goto cleanup;
+		bs->bh = ext4_sb_bread(sb, EXT4_I(inode)->i_file_acl, REQ_PRIO);
+		if (IS_ERR(bs->bh))
+			return PTR_ERR(bs->bh);
 		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 +1835,10 @@ 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;
-
-cleanup:
-	return error;
+	return 0;
 }
 
 static int
@@ -2274,9 +2267,9 @@ 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);
-	if (!bh)
-		return ERR_PTR(-EIO);
+	bh = ext4_sb_bread(inode->i_sb, EXT4_I(inode)->i_file_acl, REQ_PRIO);
+	if (IS_ERR(bh))
+		return bh;
 	error = ext4_xattr_check_block(inode, bh);
 	if (error) {
 		brelse(bh);
@@ -2729,7 +2722,7 @@ int ext4_expand_extra_isize_ea(struct inode *inode, int new_extra_isize,
 	base = IFIRST(header);
 	end = (void *)raw_inode + EXT4_SB(inode->i_sb)->s_inode_size;
 	min_offs = end - base;
-	total_ino = sizeof(struct ext4_xattr_ibody_header);
+	total_ino = sizeof(struct ext4_xattr_ibody_header) + sizeof(u32);
 
 	error = xattr_check_inode(inode, header, end);
 	if (error)
@@ -2746,10 +2739,11 @@ int ext4_expand_extra_isize_ea(struct inode *inode, int new_extra_isize,
 	if (EXT4_I(inode)->i_file_acl) {
 		struct buffer_head *bh;
 
-		bh = sb_bread(inode->i_sb, EXT4_I(inode)->i_file_acl);
-		error = -EIO;
-		if (!bh)
+		bh = ext4_sb_bread(inode->i_sb, EXT4_I(inode)->i_file_acl, REQ_PRIO);
+		if (IS_ERR(bh)) {
+			error = PTR_ERR(bh);
 			goto cleanup;
+		}
 		error = ext4_xattr_check_block(inode, bh);
 		if (error) {
 			brelse(bh);
@@ -2903,11 +2897,12 @@ 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);
-		if (!bh) {
+		bh = ext4_sb_bread(inode->i_sb, EXT4_I(inode)->i_file_acl, REQ_PRIO);
+		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 +3055,10 @@ ext4_xattr_block_cache_find(struct inode *inode,
 	while (ce) {
 		struct buffer_head *bh;
 
-		bh = sb_bread(inode->i_sb, ce->e_value);
-		if (!bh) {
+		bh = ext4_sb_bread(inode->i_sb, ce->e_value, REQ_PRIO);
+		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) {

fs/jbd2/commit.c

@@ -439,6 +439,8 @@ void jbd2_journal_commit_transaction(journal_t *journal)
 		finish_wait(&journal->j_wait_updates, &wait);
 	}
 	spin_unlock(&commit_transaction->t_handle_lock);
+	commit_transaction->t_state = T_SWITCH;
+	write_unlock(&journal->j_state_lock);
 
 	J_ASSERT (atomic_read(&commit_transaction->t_outstanding_credits) <=
 			journal->j_max_transaction_buffers);
@@ -505,6 +507,7 @@ void jbd2_journal_commit_transaction(journal_t *journal)
 	atomic_sub(atomic_read(&journal->j_reserved_credits),
 		   &commit_transaction->t_outstanding_credits);
 
+	write_lock(&journal->j_state_lock);
 	trace_jbd2_commit_flushing(journal, commit_transaction);
 	stats.run.rs_flushing = jiffies;
 	stats.run.rs_locked = jbd2_time_diff(stats.run.rs_locked,

fs/jbd2/transaction.c

@@ -138,9 +138,9 @@ static inline void update_t_max_wait(transaction_t *transaction,
 }
 
 /*
- * Wait until running transaction passes T_LOCKED state. Also starts the commit
- * if needed. The function expects running transaction to exist and releases
- * j_state_lock.
+ * Wait until running transaction passes to T_FLUSH state and new transaction
+ * can thus be started. Also starts the commit if needed. The function expects
+ * running transaction to exist and releases j_state_lock.
  */
 static void wait_transaction_locked(journal_t *journal)
 	__releases(journal->j_state_lock)
@@ -160,6 +160,32 @@ static void wait_transaction_locked(journal_t *journal)
 	finish_wait(&journal->j_wait_transaction_locked, &wait);
 }
 
+/*
+ * Wait until running transaction transitions from T_SWITCH to T_FLUSH
+ * state and new transaction can thus be started. The function releases
+ * j_state_lock.
+ */
+static void wait_transaction_switching(journal_t *journal)
+	__releases(journal->j_state_lock)
+{
+	DEFINE_WAIT(wait);
+
+	if (WARN_ON(!journal->j_running_transaction ||
+		    journal->j_running_transaction->t_state != T_SWITCH))
+		return;
+	prepare_to_wait(&journal->j_wait_transaction_locked, &wait,
+			TASK_UNINTERRUPTIBLE);
+	read_unlock(&journal->j_state_lock);
+	/*
+	 * We don't call jbd2_might_wait_for_commit() here as there's no
+	 * waiting for outstanding handles happening anymore in T_SWITCH state
+	 * and handling of reserved handles actually relies on that for
+	 * correctness.
+	 */
+	schedule();
+	finish_wait(&journal->j_wait_transaction_locked, &wait);
+}
+
 static void sub_reserved_credits(journal_t *journal, int blocks)
 {
 	atomic_sub(blocks, &journal->j_reserved_credits);
@@ -183,7 +209,8 @@ static int add_transaction_credits(journal_t *journal, int blocks,
 	 * If the current transaction is locked down for commit, wait
 	 * for the lock to be released.
 	 */
-	if (t->t_state == T_LOCKED) {
+	if (t->t_state != T_RUNNING) {
+		WARN_ON_ONCE(t->t_state >= T_FLUSH);
 		wait_transaction_locked(journal);
 		return 1;
 	}
@@ -360,8 +387,14 @@ static int start_this_handle(journal_t *journal, handle_t *handle,
 		/*
 		 * We have handle reserved so we are allowed to join T_LOCKED
 		 * transaction and we don't have to check for transaction size
-		 * and journal space.
+		 * and journal space. But we still have to wait while running
+		 * transaction is being switched to a committing one as it
+		 * won't wait for any handles anymore.
 		 */
+		if (transaction->t_state == T_SWITCH) {
+			wait_transaction_switching(journal);
+			goto repeat;
+		}
 		sub_reserved_credits(journal, blocks);
 		handle->h_reserved = 0;
 	}

include/linux/jbd2.h

@@ -575,6 +575,7 @@ struct transaction_s
 	enum {
 		T_RUNNING,
 		T_LOCKED,
+		T_SWITCH,
 		T_FLUSH,
 		T_COMMIT,
 		T_COMMIT_DFLUSH,
@@ -662,13 +663,13 @@ struct transaction_s
 
 	/*
 	 * Number of outstanding updates running on this transaction
-	 * [t_handle_lock]
+	 * [none]
 	 */
 	atomic_t		t_updates;
 
 	/*
 	 * Number of buffers reserved for use by all handles in this transaction
-	 * handle but not yet modified. [t_handle_lock]
+	 * handle but not yet modified. [none]
 	 */
 	atomic_t		t_outstanding_credits;
 
@@ -690,7 +691,7 @@ struct transaction_s
 	ktime_t			t_start_time;
 
 	/*
-	 * How many handles used this transaction? [t_handle_lock]
+	 * How many handles used this transaction? [none]
 	 */
 	atomic_t		t_handle_count;
 

include/trace/events/ext4.h

@@ -226,6 +226,26 @@ TRACE_EVENT(ext4_drop_inode,
 		  (unsigned long) __entry->ino, __entry->drop)
 );
 
+TRACE_EVENT(ext4_nfs_commit_metadata,
+	TP_PROTO(struct inode *inode),
+
+	TP_ARGS(inode),
+
+	TP_STRUCT__entry(
+		__field(	dev_t,	dev			)
+		__field(	ino_t,	ino			)
+	),
+
+	TP_fast_assign(
+		__entry->dev	= inode->i_sb->s_dev;
+		__entry->ino	= inode->i_ino;
+	),
+
+	TP_printk("dev %d,%d ino %lu",
+		  MAJOR(__entry->dev), MINOR(__entry->dev),
+		  (unsigned long) __entry->ino)
+);
+
 TRACE_EVENT(ext4_mark_inode_dirty,
 	TP_PROTO(struct inode *inode, unsigned long IP),