On 12/4/25 12:19 AM, Viacheslav Dubeyko wrote:
On Tue, 2025-12-02 at 11:16 +0100, Mehdi Ben Hadj Khelifa wrote:
On 12/2/25 12:04 AM, Viacheslav Dubeyko wrote:
On Mon, 2025-12-01 at 23:23 +0100, Mehdi Ben Hadj Khelifa wrote:
When hfs was converted to the new mount api a bug was introduced by changing the allocation pattern of sb->s_fs_info. If setup_bdev_super() fails after a new superblock has been allocated by sget_fc(), but before hfs_fill_super() takes ownership of the filesystem-specific s_fs_info data it was leaked.
Fix this by freeing sb->s_fs_info in hfs_kill_super().
Cc: stable@vger.kernel.org Fixes: ffcd06b6d13b ("hfs: convert hfs to use the new mount api") Reported-by: syzbot+ad45f827c88778ff7df6@syzkaller.appspotmail.com Closes: https://syzkaller.appspot.com/bug?extid=ad45f827c88778ff7df6 Tested-by: Viacheslav Dubeyko Slava.Dubeyko@ibm.com Signed-off-by: Christian Brauner brauner@kernel.org Signed-off-by: Mehdi Ben Hadj Khelifa mehdi.benhadjkhelifa@gmail.com
fs/hfs/mdb.c | 35 ++++++++++++++--------------------- fs/hfs/super.c | 10 +++++++++- 2 files changed, 23 insertions(+), 22 deletions(-)
diff --git a/fs/hfs/mdb.c b/fs/hfs/mdb.c index 53f3fae60217..f28cd24dee84 100644 --- a/fs/hfs/mdb.c +++ b/fs/hfs/mdb.c @@ -92,7 +92,7 @@ int hfs_mdb_get(struct super_block *sb) /* See if this is an HFS filesystem */ bh = sb_bread512(sb, part_start + HFS_MDB_BLK, mdb); if (!bh)
goto out;
return -EIO;Frankly speaking, I don't see the point to rework the hfs_mdb_get() method so intensively. We had pretty good pattern before:
int hfs_mdb_get(struct super_block *sb) { if (something_happens) goto out;
if (something_happens_and_we_need_free_buffer) goto out_bh; return 0;out_bh: brelse(bh); out: return -EIO; }
The point here that we have error management logic in one place. Now you have spread this logic through the whole function. It makes function more difficult to manage and we can introduce new bugs. Could you please localize your change without reworking this pattern of error situation management? Also, it will make the patch more compact. Could you please rework the patch?
This change in particular is made by christian. As he mentionned in one of his emails to my patches[1], his logic was that hfs_mdb_put() should only be called in fill_super() which cleans everything up and that the cleanup labels don't make sense here which is why he spread the logic of cleanup across the function. Maybe he can give us more input on this since it wasn't my code.
I am not against of not calling the hfs_mdb_put() in hfs_mdb_get(). But if I am trying to rework some method significantly, guys are not happy at all about it. :) I am slightly worried about such significant rework of hfs_mdb_get() because we potentially could introduce some new bugs. And I definitely will have the conflict with another patch under review. :)
Totally understandable. If I was to make that change I would probably seperate it from the fix (except the part where we delete freeing the s_fs_info struct). But I guess christian wanted to do the whole refactoring since it was related and it made more sense as he explained it.
I've spent some more time for reviewing the patch again. And I think I can accept it as it is. Currently, I don't see any serious issue in hfs_mdb_get(). It is simply my code style preferences. :) But people can see it in different ways.
Thanks for you time and effort!
if (mdb->drSigWord == cpu_to_be16(HFS_SUPER_MAGIC)) break; @@ -102,13 +102,14 @@ int hfs_mdb_get(struct super_block *sb) * (should do this only for cdrom/loop though) */ if (hfs_part_find(sb, &part_start, &part_size))
goto out;
return -EIO;
goto out_bh;
brelse(bh);return -EIO;@@ -125,14 +126,16 @@ int hfs_mdb_get(struct super_block *sb) brelse(bh); if (!sb_set_blocksize(sb, size)) { pr_err("unable to set blocksize to %u\n", size);
goto out;
return -EIO;
goto out;- if (mdb->drSigWord != cpu_to_be16(HFS_SUPER_MAGIC))
goto out_bh;
return -EIO;- if (mdb->drSigWord != cpu_to_be16(HFS_SUPER_MAGIC)) {
brelse(bh);return -EIO;- } HFS_SB(sb)->mdb_bh = bh; HFS_SB(sb)->mdb = mdb;
@@ -174,7 +177,7 @@ int hfs_mdb_get(struct super_block *sb) HFS_SB(sb)->bitmap = kzalloc(8192, GFP_KERNEL); if (!HFS_SB(sb)->bitmap)
goto out;
return -EIO;@@ -185,7 +188,7 @@ int hfs_mdb_get(struct super_block *sb) bh = sb_bread(sb, off >> sb->s_blocksize_bits); if (!bh) { pr_err("unable to read volume bitmap\n");
goto out;
return -EIO; } off2 = off & (sb->s_blocksize - 1); len = min((int)sb->s_blocksize - off2, size);@@ -199,12 +202,12 @@ int hfs_mdb_get(struct super_block *sb) HFS_SB(sb)->ext_tree = hfs_btree_open(sb, HFS_EXT_CNID, hfs_ext_keycmp); if (!HFS_SB(sb)->ext_tree) { pr_err("unable to open extent tree\n");
goto out;
return -EIO;
goto out;
return -EIO;@@ -229,12 +232,6 @@ int hfs_mdb_get(struct super_block *sb) } return 0;
-out_bh:
- brelse(bh);
-out:
- hfs_mdb_put(sb);
- return -EIO; } /*
@@ -359,8 +356,6 @@ void hfs_mdb_close(struct super_block *sb) * Release the resources associated with the in-core MDB. */ void hfs_mdb_put(struct super_block *sb) {
- if (!HFS_SB(sb))
return;@@ -373,6 +368,4 @@ void hfs_mdb_put(struct super_block *sb) unload_nls(HFS_SB(sb)->nls_disk); kfree(HFS_SB(sb)->bitmap);
- kfree(HFS_SB(sb));
- sb->s_fs_info = NULL; }
diff --git a/fs/hfs/super.c b/fs/hfs/super.c index 47f50fa555a4..df289cbdd4e8 100644 --- a/fs/hfs/super.c +++ b/fs/hfs/super.c @@ -431,10 +431,18 @@ static int hfs_init_fs_context(struct fs_context *fc) return 0; } +static void hfs_kill_super(struct super_block *sb) +{
- struct hfs_sb_info *hsb = HFS_SB(sb);
- kill_block_super(sb);
- kfree(hsb);
+}
- static struct file_system_type hfs_fs_type = { .owner = THIS_MODULE, .name = "hfs",
- .kill_sb = kill_block_super,
- .kill_sb = hfs_kill_super, .fs_flags = FS_REQUIRES_DEV, .init_fs_context = hfs_init_fs_context, };
Looks good. Thanks a lot for the fix.
Reviewed-by: Viacheslav Dubeyko slava@dubeyko.com
Thanks, Slava.
Best Regards, Mehdi Ben Hadj khelifa