Commit 81d091a2 authored by Ard Biesheuvel's avatar Ard Biesheuvel Committed by Herbert Xu
Browse files

crypto: serpent - use unaligned accessors instead of alignmask 



Instead of using an alignmask of 0x3 to ensure 32-bit alignment of the
Serpent input and output blocks, which propagates to mode drivers, and
results in pointless copying on architectures that don't care about
alignment, use the unaligned accessors, which will do the right thing on
each respective architecture, avoiding the need for double buffering.

Signed-off-by: default avatarArd Biesheuvel <ardb@kernel.org>
Signed-off-by: default avatarHerbert Xu <herbert@gondor.apana.org.au>
parent 784506a1

crypto/serpent_generic.c

+17 −27
Original line number Diff line number Diff line
@@ -10,7 +10,7 @@ 
#include <linux/init.h>

#include <linux/module.h>

#include <linux/errno.h>

#include <asm/byteorder.h>

#include <asm/unaligned.h>

#include <linux/crypto.h>

#include <linux/types.h>

#include <crypto/serpent.h>
@@ -448,19 +448,12 @@ void __serpent_encrypt(const void *c, u8 *dst, const u8 *src) 
{

	const struct serpent_ctx *ctx = c;

	const u32 *k = ctx->expkey;

	const __le32 *s = (const __le32 *)src;

	__le32	*d = (__le32 *)dst;

	u32	r0, r1, r2, r3, r4;



/*

 * Note: The conversions between u8* and u32* might cause trouble

 * on architectures with stricter alignment rules than x86

 */



	r0 = le32_to_cpu(s[0]);

	r1 = le32_to_cpu(s[1]);

	r2 = le32_to_cpu(s[2]);

	r3 = le32_to_cpu(s[3]);

	r0 = get_unaligned_le32(src);

	r1 = get_unaligned_le32(src + 4);

	r2 = get_unaligned_le32(src + 8);

	r3 = get_unaligned_le32(src + 12);



					K(r0, r1, r2, r3, 0);

	S0(r0, r1, r2, r3, r4);		LK(r2, r1, r3, r0, r4, 1);
@@ -496,10 +489,10 @@ void __serpent_encrypt(const void *c, u8 *dst, const u8 *src) 
	S6(r0, r1, r3, r2, r4);		LK(r3, r4, r1, r2, r0, 31);

	S7(r3, r4, r1, r2, r0);		K(r0, r1, r2, r3, 32);



	d[0] = cpu_to_le32(r0);

	d[1] = cpu_to_le32(r1);

	d[2] = cpu_to_le32(r2);

	d[3] = cpu_to_le32(r3);

	put_unaligned_le32(r0, dst);

	put_unaligned_le32(r1, dst + 4);

	put_unaligned_le32(r2, dst + 8);

	put_unaligned_le32(r3, dst + 12);

}

EXPORT_SYMBOL_GPL(__serpent_encrypt);
@@ -514,14 +507,12 @@ void __serpent_decrypt(const void *c, u8 *dst, const u8 *src) 
{

	const struct serpent_ctx *ctx = c;

	const u32 *k = ctx->expkey;

	const __le32 *s = (const __le32 *)src;

	__le32	*d = (__le32 *)dst;

	u32	r0, r1, r2, r3, r4;



	r0 = le32_to_cpu(s[0]);

	r1 = le32_to_cpu(s[1]);

	r2 = le32_to_cpu(s[2]);

	r3 = le32_to_cpu(s[3]);

	r0 = get_unaligned_le32(src);

	r1 = get_unaligned_le32(src + 4);

	r2 = get_unaligned_le32(src + 8);

	r3 = get_unaligned_le32(src + 12);



					K(r0, r1, r2, r3, 32);

	SI7(r0, r1, r2, r3, r4);	KL(r1, r3, r0, r4, r2, 31);
@@ -557,10 +548,10 @@ void __serpent_decrypt(const void *c, u8 *dst, const u8 *src) 
	SI1(r3, r1, r2, r0, r4);	KL(r4, r1, r2, r0, r3, 1);

	SI0(r4, r1, r2, r0, r3);	K(r2, r3, r1, r4, 0);



	d[0] = cpu_to_le32(r2);

	d[1] = cpu_to_le32(r3);

	d[2] = cpu_to_le32(r1);

	d[3] = cpu_to_le32(r4);

	put_unaligned_le32(r2, dst);

	put_unaligned_le32(r3, dst + 4);

	put_unaligned_le32(r1, dst + 8);

	put_unaligned_le32(r4, dst + 12);

}

EXPORT_SYMBOL_GPL(__serpent_decrypt);
@@ -578,7 +569,6 @@ static struct crypto_alg srp_alg = { 
	.cra_flags		=	CRYPTO_ALG_TYPE_CIPHER,

	.cra_blocksize		=	SERPENT_BLOCK_SIZE,

	.cra_ctxsize		=	sizeof(struct serpent_ctx),

	.cra_alignmask		=	3,

	.cra_module		=	THIS_MODULE,

	.cra_u			=	{ .cipher = {

	.cia_min_keysize	=	SERPENT_MIN_KEY_SIZE,