Re: [PATCH] Possible memory leak in arch/cris/arch-v32/drivers/cryptocop.c

[Jesper Nilsson <jesper.nilsson@axis.com>]

Hi,

I'm having trouble with this patch, it seems horribly whitespace damaged.
Also, this leak has already been reported (bugzilla 11363) and
fixed in our local repo.

The following patch is what I have queued for the CRIS-tree:

From: Jesper Nilsson <jesper.nilsson@axis.com>
Subject: [PATCH] CRISv32: Fix potential null reference in cryptocop driver.

The code didn't test the pointer to the newly allocated
memory, but a parameter sent in as value.
Since the input parameter was most often set, the code
would have used a null pointer if the kmalloc failed.
If the input parameter was not set, the code would
leak the allocated buffer.

http://bugzilla.kernel.org/show_bug.cgi?id=11363

Reported-by: Daniel Marjamäki <danielm77@spray.se>
Signed-off-by: Jesper Nilsson <jesper.nilsson@axis.com>
---
 arch/cris/arch-v32/drivers/cryptocop.c |    4 ++--
 1 files changed, 2 insertions(+), 2 deletions(-)

diff --git a/arch/cris/arch-v32/drivers/cryptocop.c b/arch/cris/arch-v32/drivers/cryptocop.c
index 67c61ea..fd529a0 100644
--- a/arch/cris/arch-v32/drivers/cryptocop.c
+++ b/arch/cris/arch-v32/drivers/cryptocop.c
@@ -1395,7 +1395,7 @@ static int create_md5_pad(int alloc_flag, unsigned long long hashed_length, char
 	if (padlen < MD5_MIN_PAD_LENGTH) padlen += MD5_BLOCK_LENGTH;
 
 	p = kmalloc(padlen, alloc_flag);
-	if (!pad) return -ENOMEM;
+	if (!p) return -ENOMEM;
 
 	*p = 0x80;
 	memset(p+1, 0, padlen - 1);
@@ -1427,7 +1427,7 @@ static int create_sha1_pad(int alloc_flag, unsigned long long hashed_length, cha
 	if (padlen < SHA1_MIN_PAD_LENGTH) padlen += SHA1_BLOCK_LENGTH;
 
 	p = kmalloc(padlen, alloc_flag);
-	if (!pad) return -ENOMEM;
+	if (!p) return -ENOMEM;
 
 	*p = 0x80;
 	memset(p+1, 0, padlen - 1);
-- 
1.6.1

On Thu, Jun 11, 2009 at 06:30:17PM +0200, Giangiacomo Mariotti wrote:
> From ee8e71f0ae1356b324906efc8e858b29025c60ef Mon Sep 17 00:00:00 2001
> From: Giangiacomo Mariotti <giangiacomo.mariotti@gmail.com>
> Date: Thu, 11 Jun 2009 18:13:52 +0200
> Subject: [PATCH] Possible mem leak
> 
> ---
>  arch/cris/arch-v32/drivers/cryptocop.c |    4 ++--
>  1 files changed, 2 insertions(+), 2 deletions(-)
> 
> diff --git a/arch/cris/arch-v32/drivers/cryptocop.c
> b/arch/cris/arch-v32/drivers/cryptocop.c
> index 67c61ea..fd529a0 100644
> --- a/arch/cris/arch-v32/drivers/cryptocop.c
> +++ b/arch/cris/arch-v32/drivers/cryptocop.c
> @@ -1395,7 +1395,7 @@ static int create_md5_pad(int alloc_flag,
> unsigned long long hashed_length, char
>  	if (padlen < MD5_MIN_PAD_LENGTH) padlen += MD5_BLOCK_LENGTH;
> 
>  	p = kmalloc(padlen, alloc_flag);
> -	if (!pad) return -ENOMEM;
> +	if (!p) return -ENOMEM;
> 
>  	*p = 0x80;
>  	memset(p+1, 0, padlen - 1);
> @@ -1427,7 +1427,7 @@ static int create_sha1_pad(int alloc_flag,
> unsigned long long hashed_length, cha
>  	if (padlen < SHA1_MIN_PAD_LENGTH) padlen += SHA1_BLOCK_LENGTH;
> 
>  	p = kmalloc(padlen, alloc_flag);
> -	if (!pad) return -ENOMEM;
> +	if (!p) return -ENOMEM;
> 
>  	*p = 0x80;
>  	memset(p+1, 0, padlen - 1);
> --
> 
> -- 
> You highest men whom my eyes have seen, this is my doubt about you and
> my secret laughter: I guess that you would call my overman—devil!
> What is great is so alien to your souls that the overman would be
> terrifying to you in his goodness ...
> --
> To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
> the body of a message to majordomo@vger.kernel.org
> More majordomo info at  http://vger.kernel.org/majordomo-info.html
> Please read the FAQ at  http://www.tux.org/lkml/

/^JN - Jesper Nilsson
-- 
               Jesper Nilsson -- jesper.nilsson@axis.com