[PATCH v2] ocfs2: don't BUG_ON an invalid journal dinode

[ZhengYuan Huang]

[BUG]
A fuzzed OCFS2 image can corrupt the current slot journal dinode while
mount is still in progress. The mount path first reports the invalid
journal block and then crashes in shutdown:

kernel BUG at fs/ocfs2/journal.c:1034!
Oops: invalid opcode: 0000 [#1] SMP KASAN NOPTI
RIP: 0010:ocfs2_journal_toggle_dirty+0x2d6/0x340 fs/ocfs2/journal.c:1034
Call Trace:
 ocfs2_journal_shutdown+0x414/0xc30 fs/ocfs2/journal.c:1116
 ocfs2_mount_volume fs/ocfs2/super.c:1785 [inline]
 ocfs2_fill_super+0x30a9/0x3cd0 fs/ocfs2/super.c:1083
 get_tree_bdev_flags+0x38b/0x640 fs/super.c:1698
 get_tree_bdev+0x24/0x40 fs/super.c:1721
 ocfs2_get_tree+0x21/0x30 fs/ocfs2/super.c:1184
 vfs_get_tree+0x9a/0x370 fs/super.c:1758
 fc_mount fs/namespace.c:1199 [inline]
 do_new_mount_fc fs/namespace.c:3642 [inline]
 do_new_mount fs/namespace.c:3718 [inline]
 path_mount+0x5b8/0x1ea0 fs/namespace.c:4028
 do_mount fs/namespace.c:4041 [inline]
 __do_sys_mount fs/namespace.c:4229 [inline]
 __se_sys_mount fs/namespace.c:4206 [inline]
 __x64_sys_mount+0x282/0x320 fs/namespace.c:4206
 ...

[CAUSE]
ocfs2_journal_toggle_dirty() used to return -EIO when journal->j_bh no
longer contained a valid dinode, because the startup and shutdown paths
already handled that failure. Commit 10995aa2451a
("ocfs2: Morph the haphazard OCFS2_IS_VALID_DINODE() checks.") changed
the check to a BUG_ON() under the assumption that the journal dinode had
already been validated. That turns an unexpected invalid journal dinode
during mount teardown into a kernel crash instead of a normal mount
failure.

[FIX]
Replace the BUG_ON() with WARN_ON() and return -EIO. This keeps the
invariant warning for debugging, but restores the original behavior of
failing startup or shutdown cleanly instead of panicking the kernel.

Fixes: 10995aa2451a ("ocfs2: Morph the haphazard OCFS2_IS_VALID_DINODE() checks.")
Signed-off-by: ZhengYuan Huang <gality369@xxxxxxxxx>
---
v2:
- Drop the full dinode revalidation and use WARN_ON plus -EIO instead
- Rework the rationale around the historical BUG_ON conversion
- Add a Fixes tag for the commit that replaced the old error return
---
 fs/ocfs2/journal.c | 7 ++-----
 1 file changed, 2 insertions(+), 5 deletions(-)

diff --git a/fs/ocfs2/journal.c b/fs/ocfs2/journal.c
index f9bf3bac085d..fc54cc798ce3 100644
--- a/fs/ocfs2/journal.c
+++ b/fs/ocfs2/journal.c
@@ -1022,11 +1022,8 @@ static int ocfs2_journal_toggle_dirty(struct ocfs2_super *osb,
 	struct ocfs2_dinode *fe;
 
 	fe = (struct ocfs2_dinode *)bh->b_data;
-
-	/* The journal bh on the osb always comes from ocfs2_journal_init()
-	 * and was validated there inside ocfs2_inode_lock_full().  It's a
-	 * code bug if we mess it up. */
-	BUG_ON(!OCFS2_IS_VALID_DINODE(fe));
+	if (WARN_ON(!OCFS2_IS_VALID_DINODE(fe)))
+		return -EIO;
 
 	flags = le32_to_cpu(fe->id1.journal1.ij_flags);
 	if (dirty)
-- 
2.49.0