Skip to contentSkip to conversation

gfs2: Validate i_depth for exhash directories

PATCH Applied Ruohan Lan 2 participants 2 messages Apr 23

Subsystems (2)
lore
Share & tools
lore

Ruohan Lan

author
#0 lore raw

PATCH 1/1 gfs2: Validate i_depth for exhash directories

From: Andrew Price <[email protected]>

[ Upstream commit 557c024ca7250bb65ae60f16c02074106c2f197b ]

A fuzzer test introduced corruption that ends up with a depth of 0 in dir_e_read(), causing an undefined shift by 32 at:

  index = hash >> (32 - dip->i_depth);

As calculated in an open-coded way in dir_make_exhash(), the minimum depth for an exhash directory is ilog2(sdp->sd_hash_ptrs) and 0 is invalid as sdp->sd_hash_ptrs is fixed as sdp->bsize / 16 at mount time.

So we can avoid the undefined behaviour by checking for depth values lower than the minimum in gfs2_dinode_in(). Values greater than the maximum are already being checked for there.

Also switch the calculation in dir_make_exhash() to use ilog2() to clarify how the depth is calculated.

Tested with the syzkaller repro.c and xfstests '-g quick'.

[ To maintain consistency in error handling in gfs2_dinode_in(), use "goto corrupt" in v5.10. ]

---
 fs/gfs2/dir.c   | 6 ++----
 fs/gfs2/glops.c | 4 ++++
 2 files changed, 6 insertions(+), 4 deletions(-)
fs/gfs2/dir.c +2 -4
index 4517ffb7c13d..7b11f7b7151a 100644
@@ -60,6 +60,7 @@
 #include <linux/crc32.h>
 #include <linux/vmalloc.h>
 #include <linux/bio.h>
+#include <linux/log2.h>
 
 #include "gfs2.h"
 #include "incore.h"
@@ -910,7 +911,6 @@ static int dir_make_exhash(struct inode *inode)
 	struct qstr args;
 	struct buffer_head *bh, *dibh;
 	struct gfs2_leaf *leaf;
-	int y;
 	u32 x;
 	__be64 *lp;
 	u64 bn;
@@ -977,9 +977,7 @@ static int dir_make_exhash(struct inode *inode)
 	i_size_write(inode, sdp->sd_sb.sb_bsize / 2);
 	gfs2_add_inode_blocks(&dip->i_inode, 1);
 	dip->i_diskflags |= GFS2_DIF_EXHASH;
-
-	for (x = sdp->sd_hash_ptrs, y = -1; x; x >>= 1, y++) ;
-	dip->i_depth = y;
+	dip->i_depth = ilog2(sdp->sd_hash_ptrs);
 
 	gfs2_dinode_out(dip, dibh->b_data);
 
fs/gfs2/glops.c +4 -0
index 87f811088466..a4050468fecc 100644
@@ -11,6 +11,7 @@
 #include <linux/bio.h>
 #include <linux/posix_acl.h>
 #include <linux/security.h>
+#include <linux/log2.h>
 
 #include "gfs2.h"
 #include "incore.h"
@@ -452,6 +453,9 @@ static int gfs2_dinode_in(struct gfs2_inode *ip, const void *buf)
 	depth = be16_to_cpu(str->di_depth);
 	if (unlikely(depth > GFS2_DIR_MAX_DEPTH))
 		goto corrupt;
+	if ((ip->i_diskflags & GFS2_DIF_EXHASH) &&
+	    depth < ilog2(sdp->sd_hash_ptrs))
+		goto corrupt;
 	ip->i_depth = (u8)depth;
 	ip->i_entries = be32_to_cpu(str->di_entries);
 
Signature · 1 line
2.43.0
maintainer Applied
#1 lore raw

Patch "gfs2: Validate i_depth for exhash directories" has been added to the 5.10-stable tree

This is a note to let you know that I've just added the patch titled

gfs2: Validate i_depth for exhash directories

to the 5.10-stable tree which can be found at:
http://www.kernel.org/git/?p=linux/kernel/git/stable/stable-queue.git;a=summary

The filename of the patch is:
gfs2-validate-i_depth-for-exhash-directories.patch
and it can be found in the queue-5.10 subdirectory.

If you, or anyone else, feels it should not be added to the stable tree, please let <[email protected]> know about it.

From [email protected] Thu Apr 23 05:20:44 2026
From: Ruohan Lan <[email protected]>
Date: Thu, 23 Apr 2026 11:20:02 +0800
Subject: gfs2: Validate i_depth for exhash directories
To: [email protected], [email protected], [email protected]
Cc: [email protected], Andrew Price <[email protected]>, [email protected], Andreas Gruenbacher <[email protected]>, Ruohan Lan <[email protected]>
Message-ID: <[email protected]>

From: Andrew Price <[email protected]>

[ Upstream commit 557c024ca7250bb65ae60f16c02074106c2f197b ]

A fuzzer test introduced corruption that ends up with a depth of 0 in dir_e_read(), causing an undefined shift by 32 at:

  index = hash >> (32 - dip->i_depth);

As calculated in an open-coded way in dir_make_exhash(), the minimum depth for an exhash directory is ilog2(sdp->sd_hash_ptrs) and 0 is invalid as sdp->sd_hash_ptrs is fixed as sdp->bsize / 16 at mount time.

So we can avoid the undefined behaviour by checking for depth values lower than the minimum in gfs2_dinode_in(). Values greater than the maximum are already being checked for there.

Also switch the calculation in dir_make_exhash() to use ilog2() to clarify how the depth is calculated.

Tested with the syzkaller repro.c and xfstests '-g quick'.

[ To maintain consistency in error handling in gfs2_dinode_in(), use "goto corrupt" in v5.10. ]

---
 fs/gfs2/dir.c   |    6 ++----
 fs/gfs2/glops.c |    4 ++++
 2 files changed, 6 insertions(+), 4 deletions(-)
fs/gfs2/dir.c +2 -4
@@ -60,6 +60,7 @@
 #include <linux/crc32.h>
 #include <linux/vmalloc.h>
 #include <linux/bio.h>
+#include <linux/log2.h>
 
 #include "gfs2.h"
 #include "incore.h"
@@ -910,7 +911,6 @@ static int dir_make_exhash(struct inode
 	struct qstr args;
 	struct buffer_head *bh, *dibh;
 	struct gfs2_leaf *leaf;
-	int y;
 	u32 x;
 	__be64 *lp;
 	u64 bn;
@@ -977,9 +977,7 @@ static int dir_make_exhash(struct inode
 	i_size_write(inode, sdp->sd_sb.sb_bsize / 2);
 	gfs2_add_inode_blocks(&dip->i_inode, 1);
 	dip->i_diskflags |= GFS2_DIF_EXHASH;
-
-	for (x = sdp->sd_hash_ptrs, y = -1; x; x >>= 1, y++) ;
-	dip->i_depth = y;
+	dip->i_depth = ilog2(sdp->sd_hash_ptrs);
 
 	gfs2_dinode_out(dip, dibh->b_data);
 
fs/gfs2/glops.c +4 -0
@@ -11,6 +11,7 @@
 #include <linux/bio.h>
 #include <linux/posix_acl.h>
 #include <linux/security.h>
+#include <linux/log2.h>
 
 #include "gfs2.h"
 #include "incore.h"
@@ -452,6 +453,9 @@ static int gfs2_dinode_in(struct gfs2_in
 	depth = be16_to_cpu(str->di_depth);
 	if (unlikely(depth > GFS2_DIR_MAX_DEPTH))
 		goto corrupt;
+	if ((ip->i_diskflags & GFS2_DIF_EXHASH) &&
+	    depth < ilog2(sdp->sd_hash_ptrs))
+		goto corrupt;
 	ip->i_depth = (u8)depth;
 	ip->i_entries = be32_to_cpu(str->di_entries);
 

Patches currently in stable-queue which might be from [email protected] are

queue-5.10/gfs2-validate-i_depth-for-exhash-directories.patch