Skip to contentSkip to conversation

befs: redundant test on unsigned in befs_get_block()?

PATCH In review Roel Kluin 2 participants 3 messages Oct 16, 2009 – Oct 20, 2009 Befs file system

lore
Share & tools
lore

Roel Kluin

author
#0 lore raw

PATCH 1/1 befs: redundant test on unsigned in befs_get_block()?

block is unsigned, this test appears redundant.

---
Or is a different test required?

fs/befs/linuxvfs.c +0 -7
index 33baf27..3ab1336 100644
@@ -128,13 +128,6 @@ befs_get_block(struct inode *inode, sector_t block,
 	befs_debug(sb, "---> befs_get_block() for inode %lu, block %ld",
 		   inode->i_ino, block);
 
-	if (block < 0) {
-		befs_error(sb, "befs_get_block() was asked for a block "
-			   "number less than zero: block %ld in inode %lu",
-			   block, inode->i_ino);
-		return -EIO;
-	}
-
 	if (create) {
 		befs_error(sb, "befs_get_block() was asked to write to "
 			   "block %ld in inode %lu", block, inode->i_ino);
#1 lore raw
Quoted 14 lines · On Fri, 16 Oct 2009 17:14:41 +0200 Roel Kluin <[email protected]> wrote:

block is unsigned, this test appears redundant.

---
Or is a different test required?

fs/befs/linuxvfs.c +0 -0
index 33baf27..3ab1336 100644
@@ -128,13 +128,6 @@ befs_get_block(struct inode *inode, sector_t block,
 	befs_debug(sb, "---> befs_get_block() for inode %lu, block %ld",
 		   inode->i_ino, block);
 

fs/befs/linuxvfs.c befs_get_block(struct inode *inode, sector_t block, · line 133

Quoted 10 lines
fs/befs/linuxvfs.c +0 -7
-	if (block < 0) {
-		befs_error(sb, "befs_get_block() was asked for a block "
-			   "number less than zero: block %ld in inode %lu",
-			   block, inode->i_ino);
-		return -EIO;
-	}
-
 	if (create) {
 		befs_error(sb, "befs_get_block() was asked to write to "
 			   "block %ld in inode %lu", block, inode->i_ino);

As far as the VFS is concerned, `block' is indeed unsigned and may well be in the range 2G-4G with a 32-bit sector_t. Perhaps not possible on befs but still legal to the VFS.

So the test is wrong from that POV.

However it is possible that befs is defending itself here. Perhaps code internal to befs will explode if passed a "negative" block number. Due to coding errors within the fs implementation.

So really, we'd need to check all code paths called by befs_get_block() and check that they are signednessly clean.

author
#2 lore raw

block is unsigned, check whether it is not too large.

---

Quoted 4 lines

As far as the VFS is concerned, `block' is indeed unsigned and may well be in the range 2G-4G with a 32-bit sector_t. Perhaps not possible on befs but still legal to the VFS.

Quoted 8 lines

So the test is wrong from that POV.

However it is possible that befs is defending itself here. Perhaps code internal to befs will explode if passed a "negative" block number. Due to coding errors within the fs implementation.

So really, we'd need to check all code paths called by befs_get_block() and check that they are signednessly clean.

This appears to be already noted by Jesper Juhl in 2004, however it was never fixed: http://search.luky.org/linux-kernel.2004/msg01392.html

It's getting late here, but what do you think about this:
befs_get_block() calls befs_fblock2brun() and there occurs a

pos = fblock << BEFS_SB(sb)->block_shift;

and in effect:
if (pos >= data->max_double_indirect_range)
error out.

So if I'm not mistaken, this should provide protection:

fs/befs/linuxvfs.c +3 -3
index 33baf27..eeb4625 100644
@@ -128,9 +128,9 @@ befs_get_block(struct inode *inode, sector_t block,
 	befs_debug(sb, "---> befs_get_block() for inode %lu, block %ld",
 		   inode->i_ino, block);
 
-	if (block < 0) {
-		befs_error(sb, "befs_get_block() was asked for a block "
-			   "number less than zero: block %ld in inode %lu",
+	if (block >= ds->max_double_indirect_range >>
+			BEFS_SB(sb)->block_shift) {
+		befs_error(sb, "befs_get_block() was asked for a too large block: block %ld in inode %lu",
 			   block, inode->i_ino);
 		return -EIO;
 	}