[PATCH 2/3] ocfs2: reject dinodes whose i_rdev disagrees with the file type

[Michael Bommarito <michael.bommarito@gmail.com>]

id1.dev1.i_rdev is the device-number arm of the ocfs2_dinode id1
union and is only meaningful for character and block device
inodes.  For any other user-visible file type the on-disk value
must be zero.

ocfs2_populate_inode() currently runs

    inode->i_rdev = huge_decode_dev(le64_to_cpu(fe->id1.dev1.i_rdev));

unconditionally, before the S_IFMT switch decides whether the
inode is a special file.  As a result, an i_rdev value present on
a non-device inode is silently published into the in-core inode.
A subsequent forced re-read or in-core mode mutation (cluster
peer with raw write access to the shared LUN, on-disk corruption,
or a separately forged dinode) can then expose the attacker-
controlled device number to init_special_inode() without ever
showing an unusual i_mode at validation time.

System inodes (OCFS2_SYSTEM_FL) legitimately use the bitmap1 and
journal1 arms of the same union: allocator inodes encode i_used
/ i_total in the bitmap1 arm and the journal encodes ij_flags /
ij_recovery_generation in the journal1 arm.  Those byte
sequences are not an i_rdev and a non-zero pattern there is the
on-disk norm, not an integrity violation.  Restrict the cross-
check to non-system inodes; that is the full surface where
i_rdev semantics apply and is also the full surface an
unprivileged consumer of the volume can see.

Following the i_mode canonicalisation in patch 1, S_ISCHR /
S_ISBLK covers the whole device-inode space; this check operates
correctly on its own, but the canonicalised i_mode makes the
predicate exhaustive.

Fixes: b657c95c1108 ("ocfs2: Wrap inode block reads in a dedicated function.")
Cc: stable@vger.kernel.org
Signed-off-by: Michael Bommarito <michael.bommarito@gmail.com>
Assisted-by: Claude:claude-opus-4-7
---
 fs/ocfs2/inode.c | 38 ++++++++++++++++++++++++++++++++++++++
 1 file changed, 38 insertions(+)

diff --git a/fs/ocfs2/inode.c b/fs/ocfs2/inode.c
index fb592bf3e5f31..305e22cc9b1d9 100644
--- a/fs/ocfs2/inode.c
+++ b/fs/ocfs2/inode.c
@@ -1533,6 +1533,44 @@ int ocfs2_validate_inode_block(struct super_block *sb,
 		}
 	}
 
+	/*
+	 * id1.dev1.i_rdev is the device-number arm of the id1 union and
+	 * is only meaningful for character and block device inodes.  For
+	 * any other regular user-visible file type the on-disk value
+	 * must be zero.  ocfs2_populate_inode() currently runs
+	 *
+	 *     inode->i_rdev = huge_decode_dev(le64_to_cpu(fe->id1.dev1.i_rdev));
+	 *
+	 * unconditionally, before the S_IFMT switch decides whether the
+	 * inode is a special file.  As a result, an i_rdev value present
+	 * on a non-device inode is silently published into the in-core
+	 * inode; a subsequent forced re-read or in-core mode mutation
+	 * (cluster peer with raw write access to the shared LUN,
+	 * on-disk corruption, or a separately forged dinode) can then
+	 * expose the attacker-controlled device number to
+	 * init_special_inode() without ever showing an unusual i_mode
+	 * at validation time.
+	 *
+	 * System inodes (OCFS2_SYSTEM_FL) legitimately use the bitmap1
+	 * and journal1 arms of the same union (allocator i_used /
+	 * i_total counters and the journal ij_flags /
+	 * ij_recovery_generation pair); those bytes are not an i_rdev
+	 * and must not be checked here.  Restrict the cross-check to
+	 * non-system inodes, which is the full attacker-controllable
+	 * surface.
+	 */
+	if (!(le32_to_cpu(di->i_flags) & OCFS2_SYSTEM_FL) &&
+	    !S_ISCHR(le16_to_cpu(di->i_mode)) &&
+	    !S_ISBLK(le16_to_cpu(di->i_mode)) &&
+	    di->id1.dev1.i_rdev != 0) {
+		rc = ocfs2_error(sb,
+				 "Invalid dinode #%llu: non-device mode 0%o with i_rdev %llu\n",
+				 (unsigned long long)bh->b_blocknr,
+				 le16_to_cpu(di->i_mode),
+				 (unsigned long long)le64_to_cpu(di->id1.dev1.i_rdev));
+		goto bail;
+	}
+
 	if (le16_to_cpu(di->i_dyn_features) & OCFS2_INLINE_DATA_FL) {
 		struct ocfs2_inline_data *data = &di->id2.i_data;
 
-- 
2.53.0