cifs: proper fix for integer overflow in parse_dacl()

PATCH In review Dan Carpenter 1 participant 2 messages Sep 27, 2012 – Oct 3, 2012 1 message not in the archive

Subsystems (3)

Kernel janitors SMB client (CIFS) Kernel SMB3 server (ksmbd)

lore

Share & tools

lore

Dan Carpenter

author

2012-09-27 20:49 UTC #0 lore raw

PATCH 1/1 cifs: proper fix for integer overflow in parse_dacl()

I tried to fix this before by adding the ULONG_MAX check, but num_aces is an unsigned int so it should have been UINT_MAX. Sorry for that. These days we can just call kmalloc_array() which has the overflow check built in.

Reported-by pipacs

Signed-off-by Dan Carpenter <[email protected]>

fs/cifs/cifsacl.c +1 -3

index 2ee5c54..bc9fcfb 100644
@@ -910,9 +910,7 @@ static void parse_dacl(struct cifs_acl *pdacl, char *end_of_acl,
 		umode_t group_mask = S_IRWXG;
 		umode_t other_mask = S_IRWXU | S_IRWXG | S_IRWXO;
 
-		if (num_aces > ULONG_MAX / sizeof(struct cifs_ace *))
-			return;
-		ppace = kmalloc(num_aces * sizeof(struct cifs_ace *),
+		ppace = kmalloc_array(num_aces, sizeof(struct cifs_ace *),
 				GFP_KERNEL);
 		if (!ppace) {
 			cERROR(1, "DACL memory allocation error");

A message that is not in the archive (sent off-list or not yet received): [email protected]

Dan Carpenter

author

2012-10-03 12:43 UTC #2 lore raw

Quoted 5 lines · Dan Carpenter

I tried to fix this before by adding the ULONG_MAX check, but num_aces is an unsigned int so it should have been UINT_MAX. Sorry for that. These days we can just call kmalloc_array() which has the overflow check built in.

Uh... It turns out that I still suck at understanding C. My first patch was fine. Sorry for the noise.

regards,
dan carpenter