Date: Wed, 20 Sep 2006 07:41:18 -0400
From: Prarit Bhargava <prarit@redhat.com>
Subject: Re: [RHEL5 PATCH]: BZ 198572 Module load unaligned access errors



David Howells wrote:
>This should really use get_unaligned() not memcpy() - after all, some archs
>can handle it directly without the need to throw an exception.
>  
Good point -- thanks David.

New patch ...

P.

Patch to fix unaligned access during module loads.

Tested by Doug (dchapman@redhat.com) on HP ia64 boxes, and by me on SGI ia64
box.  Additionally tested by me on x86_64.

Addresses and resolves BZ 198572.

Signed-off-by: Prarit Bhargava <prarit@redhat.com>
Signed-off-by: Doug Chapman <dchapman@redhat.com>


--- linux-2.6.17.ia64/lib/sha1.c.orig	2006-09-19 17:07:59.000000000 -0400
+++ linux-2.6.17.ia64/lib/sha1.c	2006-09-19 17:07:48.000000000 -0400
@@ -3,6 +3,7 @@
  * Peter Gutmann, and placed in the public domain.
  */
 
+#include <asm/unaligned.h>
 #include <linux/kernel.h>
 #include <linux/module.h>
 #include <linux/cryptohash.h>
@@ -41,7 +42,7 @@
 	__u32 a, b, c, d, e, t, i;
 
 	for (i = 0; i < 16; i++)
-		W[i] = be32_to_cpu(((const __be32 *)in)[i]);
+		W[i] = be32_to_cpu(get_unaligned((const __be32 *)in+i));
 
 	for (i = 0; i < 64; i++)
 		W[i+16] = rol32(W[i+13] ^ W[i+8] ^ W[i+2] ^ W[i], 1);