Merge tag 'ext4_for_linus-6.15-rc2' of git://git.kernel.org/pub/scm/l…

…inux/kernel/git/tytso/ext4

Pull ext4 fixes from Ted Ts'o:
 "A few more miscellaneous ext4 bug fixes and cleanups including some
  syzbot failures and fixing a stale file handing refeencing an inode
  previously used as a regular file, but which has been deleted and
  reused as an ea_inode would result in ext4 erroneously considering
  this a case of fs corruption"

* tag 'ext4_for_linus-6.15-rc2' of git://git.kernel.org/pub/scm/linux/kernel/git/tytso/ext4:
  ext4: fix off-by-one error in do_split
  ext4: make block validity check resistent to sb bh corruption
  ext4: avoid -Wflex-array-member-not-at-end warning
  Documentation: ext4: Add fields to ext4_super_block documentation
  ext4: don't treat fhandle lookup of ea_inode as FS corruption

Documentation/filesystems/ext4/super.rst

@@ -328,9 +328,13 @@ The ext4 superblock is laid out as follows in
      - s_checksum_type
      - Metadata checksum algorithm type. The only valid value is 1 (crc32c).
    * - 0x176
-     - __le16
-     - s_reserved_pad
-     -
+     - \_\_u8
+     - s\_encryption\_level
+     - Versioning level for encryption.
+   * - 0x177
+     - \_\_u8
+     - s\_reserved\_pad
+     - Padding to next 32bits.
    * - 0x178
      - __le64
      - s_kbytes_written
@@ -466,9 +470,13 @@ The ext4 superblock is laid out as follows in
      - s_last_error_time_hi
      - Upper 8 bits of the s_last_error_time field.
    * - 0x27A
-     - __u8
-     - s_pad[2]
-     - Zero padding.
+     - \_\_u8
+     - s\_first\_error\_errcode
+     -
+   * - 0x27B
+     - \_\_u8
+     - s\_last\_error\_errcode
+     -
    * - 0x27C
      - __le16
      - s_encoding

fs/ext4/block_validity.c

@@ -351,10 +351,9 @@ int ext4_check_blockref(const char *function, unsigned int line,
 {
 	__le32 *bref = p;
 	unsigned int blk;
+	journal_t *journal = EXT4_SB(inode->i_sb)->s_journal;
 
-	if (ext4_has_feature_journal(inode->i_sb) &&
-	    (inode->i_ino ==
-	     le32_to_cpu(EXT4_SB(inode->i_sb)->s_es->s_journal_inum)))
+	if (journal && inode == journal->j_inode)
 		return 0;
 
 	while (bref < p+max) {

fs/ext4/inode.c

@@ -386,10 +386,11 @@ static int __check_block_validity(struct inode *inode, const char *func,
 				unsigned int line,
 				struct ext4_map_blocks *map)
 {
-	if (ext4_has_feature_journal(inode->i_sb) &&
-	    (inode->i_ino ==
-	     le32_to_cpu(EXT4_SB(inode->i_sb)->s_es->s_journal_inum)))
+	journal_t *journal = EXT4_SB(inode->i_sb)->s_journal;
+
+	if (journal && inode == journal->j_inode)
 		return 0;
+
 	if (!ext4_inode_block_valid(inode, map->m_pblk, map->m_len)) {
 		ext4_error_inode(inode, func, line, map->m_pblk,
 				 "lblock %lu mapped to illegal pblock %llu "
@@ -4724,22 +4725,43 @@ static inline void ext4_inode_set_iversion_queried(struct inode *inode, u64 val)
 		inode_set_iversion_queried(inode, val);
 }
 
-static const char *check_igot_inode(struct inode *inode, ext4_iget_flags flags)
-
+static int check_igot_inode(struct inode *inode, ext4_iget_flags flags,
+			    const char *function, unsigned int line)
 {
+	const char *err_str;
+
 	if (flags & EXT4_IGET_EA_INODE) {
-		if (!(EXT4_I(inode)->i_flags & EXT4_EA_INODE_FL))
-			return "missing EA_INODE flag";
+		if (!(EXT4_I(inode)->i_flags & EXT4_EA_INODE_FL)) {
+			err_str = "missing EA_INODE flag";
+			goto error;
+		}
 		if (ext4_test_inode_state(inode, EXT4_STATE_XATTR) ||
-		    EXT4_I(inode)->i_file_acl)
-			return "ea_inode with extended attributes";
+		    EXT4_I(inode)->i_file_acl) {
+			err_str = "ea_inode with extended attributes";
+			goto error;
+		}
 	} else {
-		if ((EXT4_I(inode)->i_flags & EXT4_EA_INODE_FL))
-			return "unexpected EA_INODE flag";
+		if ((EXT4_I(inode)->i_flags & EXT4_EA_INODE_FL)) {
+			/*
+			 * open_by_handle_at() could provide an old inode number
+			 * that has since been reused for an ea_inode; this does
+			 * not indicate filesystem corruption
+			 */
+			if (flags & EXT4_IGET_HANDLE)
+				return -ESTALE;
+			err_str = "unexpected EA_INODE flag";
+			goto error;
+		}
+	}
+	if (is_bad_inode(inode) && !(flags & EXT4_IGET_BAD)) {
+		err_str = "unexpected bad inode w/o EXT4_IGET_BAD";
+		goto error;
 	}
-	if (is_bad_inode(inode) && !(flags & EXT4_IGET_BAD))
-		return "unexpected bad inode w/o EXT4_IGET_BAD";
-	return NULL;
+	return 0;
+
+error:
+	ext4_error_inode(inode, function, line, 0, err_str);
+	return -EFSCORRUPTED;
 }
 
 struct inode *__ext4_iget(struct super_block *sb, unsigned long ino,
@@ -4751,7 +4773,6 @@ struct inode *__ext4_iget(struct super_block *sb, unsigned long ino,
 	struct ext4_inode_info *ei;
 	struct ext4_super_block *es = EXT4_SB(sb)->s_es;
 	struct inode *inode;
-	const char *err_str;
 	journal_t *journal = EXT4_SB(sb)->s_journal;
 	long ret;
 	loff_t size;
@@ -4780,10 +4801,10 @@ struct inode *__ext4_iget(struct super_block *sb, unsigned long ino,
 	if (!inode)
 		return ERR_PTR(-ENOMEM);
 	if (!(inode->i_state & I_NEW)) {
-		if ((err_str = check_igot_inode(inode, flags)) != NULL) {
-			ext4_error_inode(inode, function, line, 0, err_str);
+		ret = check_igot_inode(inode, flags, function, line);
+		if (ret) {
 			iput(inode);
-			return ERR_PTR(-EFSCORRUPTED);
+			return ERR_PTR(ret);
 		}
 		return inode;
 	}
@@ -5065,13 +5086,21 @@ struct inode *__ext4_iget(struct super_block *sb, unsigned long ino,
 		ret = -EFSCORRUPTED;
 		goto bad_inode;
 	}
-	if ((err_str = check_igot_inode(inode, flags)) != NULL) {
-		ext4_error_inode(inode, function, line, 0, err_str);
-		ret = -EFSCORRUPTED;
-		goto bad_inode;
+	ret = check_igot_inode(inode, flags, function, line);
+	/*
+	 * -ESTALE here means there is nothing inherently wrong with the inode,
+	 * it's just not an inode we can return for an fhandle lookup.
+	 */
+	if (ret == -ESTALE) {
+		brelse(iloc.bh);
+		unlock_new_inode(inode);
+		iput(inode);
+		return ERR_PTR(-ESTALE);
 	}
-
+	if (ret)
+		goto bad_inode;
 	brelse(iloc.bh);
+
 	unlock_new_inode(inode);
 	return inode;
 

fs/ext4/mballoc.c

@@ -3037,18 +3037,16 @@ static int ext4_mb_seq_groups_show(struct seq_file *seq, void *v)
 	unsigned char blocksize_bits = min_t(unsigned char,
 					     sb->s_blocksize_bits,
 					     EXT4_MAX_BLOCK_LOG_SIZE);
-	struct sg {
-		struct ext4_group_info info;
-		ext4_grpblk_t counters[EXT4_MAX_BLOCK_LOG_SIZE + 2];
-	} sg;
+	DEFINE_RAW_FLEX(struct ext4_group_info, sg, bb_counters,
+			EXT4_MAX_BLOCK_LOG_SIZE + 2);
 
 	group--;
 	if (group == 0)
 		seq_puts(seq, "#group: free  frags first ["
 			      " 2^0   2^1   2^2   2^3   2^4   2^5   2^6  "
 			      " 2^7   2^8   2^9   2^10  2^11  2^12  2^13  ]\n");
 
-	i = (blocksize_bits + 2) * sizeof(sg.info.bb_counters[0]) +
+	i = (blocksize_bits + 2) * sizeof(sg->bb_counters[0]) +
 		sizeof(struct ext4_group_info);
 
 	grinfo = ext4_get_group_info(sb, group);
@@ -3068,14 +3066,14 @@ static int ext4_mb_seq_groups_show(struct seq_file *seq, void *v)
 	 * We care only about free space counters in the group info and
 	 * these are safe to access even after the buddy has been unloaded
 	 */
-	memcpy(&sg, grinfo, i);
-	seq_printf(seq, "#%-5u: %-5u %-5u %-5u [", group, sg.info.bb_free,
-			sg.info.bb_fragments, sg.info.bb_first_free);
+	memcpy(sg, grinfo, i);
+	seq_printf(seq, "#%-5u: %-5u %-5u %-5u [", group, sg->bb_free,
+			sg->bb_fragments, sg->bb_first_free);
 	for (i = 0; i <= 13; i++)
 		seq_printf(seq, " %-5u", i <= blocksize_bits + 1 ?
-				sg.info.bb_counters[i] : 0);
+				sg->bb_counters[i] : 0);
 	seq_puts(seq, " ]");
-	if (EXT4_MB_GRP_BBITMAP_CORRUPT(&sg.info))
+	if (EXT4_MB_GRP_BBITMAP_CORRUPT(sg))
 		seq_puts(seq, " Block bitmap corrupted!");
 	seq_putc(seq, '\n');
 	return 0;

fs/ext4/namei.c

@@ -1971,7 +1971,7 @@ static struct ext4_dir_entry_2 *do_split(handle_t *handle, struct inode *dir,
 	 * split it in half by count; each resulting block will have at least
 	 * half the space free.
 	 */
-	if (i > 0)
+	if (i >= 0)
 		split = count - move;
 	else
 		split = count/2;