nilfs2-fix-general-protection-fault-in-nilfs_btree_insert.patch removed from -mm tree

The quilt patch titled
Subject: nilfs2: fix general protection fault in nilfs_btree_insert()
has been removed from the -mm tree. Its filename was
nilfs2-fix-general-protection-fault-in-nilfs_btree_insert.patch

This patch was dropped because it was merged into the mm-hotfixes-stable branch of git://git.kernel.org/pub/scm/linux/kernel/git/akpm/mm

------------------------------------------------------
From: Ryusuke Konishi <[email protected]>
Subject: nilfs2: fix general protection fault in nilfs_btree_insert()
Date: Thu, 5 Jan 2023 14:53:56 +0900

If nilfs2 reads a corrupted disk image and tries to reads a b-tree node block by calling __nilfs_btree_get_block() against an invalid virtual block address, it returns -ENOENT because conversion of the virtual block address to a disk block address fails. However, this return value is the same as the internal code that b-tree lookup routines return to indicate that the block being searched does not exist, so functions that operate on that b-tree may misbehave.

When nilfs_btree_insert() receives this spurious 'not found' code from nilfs_btree_do_lookup(), it misunderstands that the 'not found' check was successful and continues the insert operation using incomplete lookup path data, causing the following crash:

 general protection fault, probably for non-canonical address
 0xdffffc0000000005: 0000 [#1] PREEMPT SMP KASAN
 KASAN: null-ptr-deref in range [0x0000000000000028-0x000000000000002f]
 ...
 RIP: 0010:nilfs_btree_get_nonroot_node fs/nilfs2/btree.c:418 [inline]
 RIP: 0010:nilfs_btree_prepare_insert fs/nilfs2/btree.c:1077 [inline]
 RIP: 0010:nilfs_btree_insert+0x6d3/0x1c10 fs/nilfs2/btree.c:1238
 Code: bc 24 80 00 00 00 4c 89 f8 48 c1 e8 03 42 80 3c 28 00 74 08 4c 89
 ff e8 4b 02 92 fe 4d 8b 3f 49 83 c7 28 4c 89 f8 48 c1 e8 03 <42> 80 3c
 28 00 74 08 4c 89 ff e8 2e 02 92 fe 4d 8b 3f 49 83 c7 02
 ...
 Call Trace:
 <TASK>
  nilfs_bmap_do_insert fs/nilfs2/bmap.c:121 [inline]
  nilfs_bmap_insert+0x20d/0x360 fs/nilfs2/bmap.c:147
  nilfs_get_block+0x414/0x8d0 fs/nilfs2/inode.c:101
  __block_write_begin_int+0x54c/0x1a80 fs/buffer.c:1991
  __block_write_begin fs/buffer.c:2041 [inline]
  block_write_begin+0x93/0x1e0 fs/buffer.c:2102
  nilfs_write_begin+0x9c/0x110 fs/nilfs2/inode.c:261
  generic_perform_write+0x2e4/0x5e0 mm/filemap.c:3772
  __generic_file_write_iter+0x176/0x400 mm/filemap.c:3900
  generic_file_write_iter+0xab/0x310 mm/filemap.c:3932
  call_write_iter include/linux/fs.h:2186 [inline]
  new_sync_write fs/read_write.c:491 [inline]
  vfs_write+0x7dc/0xc50 fs/read_write.c:584
  ksys_write+0x177/0x2a0 fs/read_write.c:637
  do_syscall_x64 arch/x86/entry/common.c:50 [inline]
  do_syscall_64+0x3d/0xb0 arch/x86/entry/common.c:80
  entry_SYSCALL_64_after_hwframe+0x63/0xcd
 ...
 </TASK>

This patch fixes the root cause of this problem by replacing the error code that __nilfs_btree_get_block() returns on block address conversion failure from -ENOENT to another internal code -EINVAL which means that the b-tree metadata is corrupted.

By returning -EINVAL, it propagates without glitches, and for all relevant b-tree operations, functions in the upper bmap layer output an error message indicating corrupted b-tree metadata via nilfs_bmap_convert_error(), and code -EIO will be eventually returned as it should be.

---

 fs/nilfs2/btree.c |   15 ++++++++++++---
 1 file changed, 12 insertions(+), 3 deletions(-)

@@ -480,9 +480,18 @@ static int __nilfs_btree_get_block(const
 	ret = nilfs_btnode_submit_block(btnc, ptr, 0, REQ_OP_READ, &bh,
 					&submit_ptr);
 	if (ret) {
-		if (ret != -EEXIST)
-			return ret;
-		goto out_check;
+		if (likely(ret == -EEXIST))
+			goto out_check;
+		if (ret == -ENOENT) {
+			/*
+			 * Block address translation failed due to invalid
+			 * value of 'ptr'.  In this case, return internal code
+			 * -EINVAL (broken bmap) to notify bmap layer of fatal
+			 * metadata corruption.
+			 */
+			ret = -EINVAL;
+		}
+		return ret;
 	}
 
 	if (ra) {

_

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