Commit a2a892a2 authored by Andreas Steinmetz's avatar Andreas Steinmetz Committed by David S. Miller

[CRYPTO] Add x86_64 asm AES

Implementation:
===============
The encrypt/decrypt code is based on an x86 implementation I did a while
ago which I never published. This unpublished implementation does
include an assembler based key schedule and precomputed tables. For
simplicity and best acceptance, however, I took Gladman's in-kernel code
for table generation and key schedule for the kernel port of my
assembler code and modified this code to produce the key schedule as
required by my assembler implementation. File locations and Kconfig are
kept similar to the i586 AES assembler implementation.
It may seem a little bit strange to use 32 bit I/O and registers in the
assembler implementation but this gives the best code size. My
implementation takes one instruction more per round compared to
Gladman's x86 assembler but it doesn't require any stack for local
variables or saved registers and it is less serialized than Gladman's
code.
Note that all comparisons to Gladman's code were done after my code was
implemented. I did only use FIPS PUB 197 for the implementation so my
implementation is independent work.
If anybody has a better assembler solution for x86_64 I'll be pleased to
have my code replaced with the better solution.

Testing:
========
The implementation passes the in-kernel crypto testing module and I'm
running it without any problems on my laptop where it is mainly used for
dm-crypt.

Microbenchmark:
===============
The microbenchmark was done in userspace with similar compile flags as
used during kernel compile.
Encrypt/decrypt is about 35% faster than the generic C implementation.
As the generic C as well as my assembler implementation are both table
I don't really expect that there is much room for further
improvements though I'll be glad to be corrected here.
The key schedule is about 5% slower than the generic C implementation.
This is due to the fact that some more work has to be done in the key
schedule routine to fit the schedule to the assembler implementation.

Code Size:
==========
Encrypt and decrypt are together about 2.1 Kbytes smaller than the
generic C implementation which is important with regard to L1 cache
usage. The key schedule routine is about 100 bytes larger than the
generic C implementation.

Data Size:
==========
There's no difference in data size requirements between the assembler
implementation and the generic C implementation.

License:
========
Gladmans's code is dual BSD/GPL whereas my assembler code is GPLv2 only
(I'm  not going to change the license for my code). So I had to change
the module license for the x86_64 aes module from 'Dual BSD/GPL' to
'GPL' to reflect the most restrictive license within the module.
Signed-off-by: default avatarAndreas Steinmetz <ast@domdv.de>
Signed-off-by: default avatarHerbert Xu <herbert@gondor.apana.org.au>
Signed-off-by: default avatarDavid S. Miller <davem@davemloft.net>
parent a61cc448
......@@ -65,7 +65,9 @@ CFLAGS += $(call cc-option,-mno-sse -mno-mmx -mno-sse2 -mno-3dnow,)
head-y := arch/x86_64/kernel/head.o arch/x86_64/kernel/head64.o arch/x86_64/kernel/init_task.o
libs-y += arch/x86_64/lib/
core-y += arch/x86_64/kernel/ arch/x86_64/mm/
core-y += arch/x86_64/kernel/ \
arch/x86_64/mm/ \
arch/x86_64/crypto/
core-$(CONFIG_IA32_EMULATION) += arch/x86_64/ia32/
drivers-$(CONFIG_PCI) += arch/x86_64/pci/
drivers-$(CONFIG_OPROFILE) += arch/x86_64/oprofile/
......
#
# x86_64/crypto/Makefile
#
# Arch-specific CryptoAPI modules.
#
obj-$(CONFIG_CRYPTO_AES_X86_64) += aes-x86_64.o
aes-x86_64-y := aes-x86_64-asm.o aes.o
/* AES (Rijndael) implementation (FIPS PUB 197) for x86_64
*
* Copyright (C) 2005 Andreas Steinmetz, <ast@domdv.de>
*
* License:
* This code can be distributed under the terms of the GNU General Public
* License (GPL) Version 2 provided that the above header down to and
* including this sentence is retained in full.
*/
.extern aes_ft_tab
.extern aes_it_tab
.extern aes_fl_tab
.extern aes_il_tab
.text
#define R1 %rax
#define R1E %eax
#define R1X %ax
#define R1H %ah
#define R1L %al
#define R2 %rbx
#define R2E %ebx
#define R2X %bx
#define R2H %bh
#define R2L %bl
#define R3 %rcx
#define R3E %ecx
#define R3X %cx
#define R3H %ch
#define R3L %cl
#define R4 %rdx
#define R4E %edx
#define R4X %dx
#define R4H %dh
#define R4L %dl
#define R5 %rsi
#define R5E %esi
#define R6 %rdi
#define R6E %edi
#define R7 %rbp
#define R7E %ebp
#define R8 %r8
#define R9 %r9
#define R10 %r10
#define R11 %r11
#define prologue(FUNC,BASE,B128,B192,r1,r2,r3,r4,r5,r6,r7,r8,r9,r10,r11) \
.global FUNC; \
.type FUNC,@function; \
.align 8; \
FUNC: movq r1,r2; \
movq r3,r4; \
leaq BASE+52(r8),r9; \
movq r10,r11; \
movl (r7),r5 ## E; \
movl 4(r7),r1 ## E; \
movl 8(r7),r6 ## E; \
movl 12(r7),r7 ## E; \
movl (r8),r10 ## E; \
xorl -48(r9),r5 ## E; \
xorl -44(r9),r1 ## E; \
xorl -40(r9),r6 ## E; \
xorl -36(r9),r7 ## E; \
cmpl $24,r10 ## E; \
jb B128; \
leaq 32(r9),r9; \
je B192; \
leaq 32(r9),r9;
#define epilogue(r1,r2,r3,r4,r5,r6,r7,r8,r9) \
movq r1,r2; \
movq r3,r4; \
movl r5 ## E,(r9); \
movl r6 ## E,4(r9); \
movl r7 ## E,8(r9); \
movl r8 ## E,12(r9); \
ret;
#define round(TAB,OFFSET,r1,r2,r3,r4,r5,r6,r7,r8,ra,rb,rc,rd) \
movzbl r2 ## H,r5 ## E; \
movzbl r2 ## L,r6 ## E; \
movl TAB+1024(,r5,4),r5 ## E;\
movw r4 ## X,r2 ## X; \
movl TAB(,r6,4),r6 ## E; \
roll $16,r2 ## E; \
shrl $16,r4 ## E; \
movzbl r4 ## H,r7 ## E; \
movzbl r4 ## L,r4 ## E; \
xorl OFFSET(r8),ra ## E; \
xorl OFFSET+4(r8),rb ## E; \
xorl TAB+3072(,r7,4),r5 ## E;\
xorl TAB+2048(,r4,4),r6 ## E;\
movzbl r1 ## L,r7 ## E; \
movzbl r1 ## H,r4 ## E; \
movl TAB+1024(,r4,4),r4 ## E;\
movw r3 ## X,r1 ## X; \
roll $16,r1 ## E; \
shrl $16,r3 ## E; \
xorl TAB(,r7,4),r5 ## E; \
movzbl r3 ## H,r7 ## E; \
movzbl r3 ## L,r3 ## E; \
xorl TAB+3072(,r7,4),r4 ## E;\
xorl TAB+2048(,r3,4),r5 ## E;\
movzbl r1 ## H,r7 ## E; \
movzbl r1 ## L,r3 ## E; \
shrl $16,r1 ## E; \
xorl TAB+3072(,r7,4),r6 ## E;\
movl TAB+2048(,r3,4),r3 ## E;\
movzbl r1 ## H,r7 ## E; \
movzbl r1 ## L,r1 ## E; \
xorl TAB+1024(,r7,4),r6 ## E;\
xorl TAB(,r1,4),r3 ## E; \
movzbl r2 ## H,r1 ## E; \
movzbl r2 ## L,r7 ## E; \
shrl $16,r2 ## E; \
xorl TAB+3072(,r1,4),r3 ## E;\
xorl TAB+2048(,r7,4),r4 ## E;\
movzbl r2 ## H,r1 ## E; \
movzbl r2 ## L,r2 ## E; \
xorl OFFSET+8(r8),rc ## E; \
xorl OFFSET+12(r8),rd ## E; \
xorl TAB+1024(,r1,4),r3 ## E;\
xorl TAB(,r2,4),r4 ## E;
#define move_regs(r1,r2,r3,r4) \
movl r3 ## E,r1 ## E; \
movl r4 ## E,r2 ## E;
#define entry(FUNC,BASE,B128,B192) \
prologue(FUNC,BASE,B128,B192,R2,R8,R7,R9,R1,R3,R4,R6,R10,R5,R11)
#define return epilogue(R8,R2,R9,R7,R5,R6,R3,R4,R11)
#define encrypt_round(TAB,OFFSET) \
round(TAB,OFFSET,R1,R2,R3,R4,R5,R6,R7,R10,R5,R6,R3,R4) \
move_regs(R1,R2,R5,R6)
#define encrypt_final(TAB,OFFSET) \
round(TAB,OFFSET,R1,R2,R3,R4,R5,R6,R7,R10,R5,R6,R3,R4)
#define decrypt_round(TAB,OFFSET) \
round(TAB,OFFSET,R2,R1,R4,R3,R6,R5,R7,R10,R5,R6,R3,R4) \
move_regs(R1,R2,R5,R6)
#define decrypt_final(TAB,OFFSET) \
round(TAB,OFFSET,R2,R1,R4,R3,R6,R5,R7,R10,R5,R6,R3,R4)
/* void aes_encrypt(void *ctx, u8 *out, const u8 *in) */
entry(aes_encrypt,0,enc128,enc192)
encrypt_round(aes_ft_tab,-96)
encrypt_round(aes_ft_tab,-80)
enc192: encrypt_round(aes_ft_tab,-64)
encrypt_round(aes_ft_tab,-48)
enc128: encrypt_round(aes_ft_tab,-32)
encrypt_round(aes_ft_tab,-16)
encrypt_round(aes_ft_tab, 0)
encrypt_round(aes_ft_tab, 16)
encrypt_round(aes_ft_tab, 32)
encrypt_round(aes_ft_tab, 48)
encrypt_round(aes_ft_tab, 64)
encrypt_round(aes_ft_tab, 80)
encrypt_round(aes_ft_tab, 96)
encrypt_final(aes_fl_tab,112)
return
/* void aes_decrypt(void *ctx, u8 *out, const u8 *in) */
entry(aes_decrypt,240,dec128,dec192)
decrypt_round(aes_it_tab,-96)
decrypt_round(aes_it_tab,-80)
dec192: decrypt_round(aes_it_tab,-64)
decrypt_round(aes_it_tab,-48)
dec128: decrypt_round(aes_it_tab,-32)
decrypt_round(aes_it_tab,-16)
decrypt_round(aes_it_tab, 0)
decrypt_round(aes_it_tab, 16)
decrypt_round(aes_it_tab, 32)
decrypt_round(aes_it_tab, 48)
decrypt_round(aes_it_tab, 64)
decrypt_round(aes_it_tab, 80)
decrypt_round(aes_it_tab, 96)
decrypt_final(aes_il_tab,112)
return
/*
* Cryptographic API.
*
* AES Cipher Algorithm.
*
* Based on Brian Gladman's code.
*
* Linux developers:
* Alexander Kjeldaas <astor@fast.no>
* Herbert Valerio Riedel <hvr@hvrlab.org>
* Kyle McMartin <kyle@debian.org>
* Adam J. Richter <adam@yggdrasil.com> (conversion to 2.5 API).
* Andreas Steinmetz <ast@domdv.de> (adapted to x86_64 assembler)
*
* This program is free software; you can redistribute it and/or modify
* it under the terms of the GNU General Public License as published by
* the Free Software Foundation; either version 2 of the License, or
* (at your option) any later version.
*
* ---------------------------------------------------------------------------
* Copyright (c) 2002, Dr Brian Gladman <brg@gladman.me.uk>, Worcester, UK.
* All rights reserved.
*
* LICENSE TERMS
*
* The free distribution and use of this software in both source and binary
* form is allowed (with or without changes) provided that:
*
* 1. distributions of this source code include the above copyright
* notice, this list of conditions and the following disclaimer;
*
* 2. distributions in binary form include the above copyright
* notice, this list of conditions and the following disclaimer
* in the documentation and/or other associated materials;
*
* 3. the copyright holder's name is not used to endorse products
* built using this software without specific written permission.
*
* ALTERNATIVELY, provided that this notice is retained in full, this product
* may be distributed under the terms of the GNU General Public License (GPL),
* in which case the provisions of the GPL apply INSTEAD OF those given above.
*
* DISCLAIMER
*
* This software is provided 'as is' with no explicit or implied warranties
* in respect of its properties, including, but not limited to, correctness
* and/or fitness for purpose.
* ---------------------------------------------------------------------------
*/
/* Some changes from the Gladman version:
s/RIJNDAEL(e_key)/E_KEY/g
s/RIJNDAEL(d_key)/D_KEY/g
*/
#include <asm/byteorder.h>
#include <linux/bitops.h>
#include <linux/crypto.h>
#include <linux/errno.h>
#include <linux/init.h>
#include <linux/module.h>
#include <linux/types.h>
#define AES_MIN_KEY_SIZE 16
#define AES_MAX_KEY_SIZE 32
#define AES_BLOCK_SIZE 16
/*
* #define byte(x, nr) ((unsigned char)((x) >> (nr*8)))
*/
static inline u8 byte(const u32 x, const unsigned n)
{
return x >> (n << 3);
}
#define u32_in(x) le32_to_cpu(*(const __le32 *)(x))
struct aes_ctx
{
u32 key_length;
u32 E[60];
u32 D[60];
};
#define E_KEY ctx->E
#define D_KEY ctx->D
static u8 pow_tab[256] __initdata;
static u8 log_tab[256] __initdata;
static u8 sbx_tab[256] __initdata;
static u8 isb_tab[256] __initdata;
static u32 rco_tab[10];
u32 aes_ft_tab[4][256];
u32 aes_it_tab[4][256];
u32 aes_fl_tab[4][256];
u32 aes_il_tab[4][256];
static inline u8 f_mult(u8 a, u8 b)
{
u8 aa = log_tab[a], cc = aa + log_tab[b];
return pow_tab[cc + (cc < aa ? 1 : 0)];
}
#define ff_mult(a, b) (a && b ? f_mult(a, b) : 0)
#define ls_box(x) \
(aes_fl_tab[0][byte(x, 0)] ^ \
aes_fl_tab[1][byte(x, 1)] ^ \
aes_fl_tab[2][byte(x, 2)] ^ \
aes_fl_tab[3][byte(x, 3)])
static void __init gen_tabs(void)
{
u32 i, t;
u8 p, q;
/* log and power tables for GF(2**8) finite field with
0x011b as modular polynomial - the simplest primitive
root is 0x03, used here to generate the tables */
for (i = 0, p = 1; i < 256; ++i) {
pow_tab[i] = (u8)p;
log_tab[p] = (u8)i;
p ^= (p << 1) ^ (p & 0x80 ? 0x01b : 0);
}
log_tab[1] = 0;
for (i = 0, p = 1; i < 10; ++i) {
rco_tab[i] = p;
p = (p << 1) ^ (p & 0x80 ? 0x01b : 0);
}
for (i = 0; i < 256; ++i) {
p = (i ? pow_tab[255 - log_tab[i]] : 0);
q = ((p >> 7) | (p << 1)) ^ ((p >> 6) | (p << 2));
p ^= 0x63 ^ q ^ ((q >> 6) | (q << 2));
sbx_tab[i] = p;
isb_tab[p] = (u8)i;
}
for (i = 0; i < 256; ++i) {
p = sbx_tab[i];
t = p;
aes_fl_tab[0][i] = t;
aes_fl_tab[1][i] = rol32(t, 8);
aes_fl_tab[2][i] = rol32(t, 16);
aes_fl_tab[3][i] = rol32(t, 24);
t = ((u32)ff_mult(2, p)) |
((u32)p << 8) |
((u32)p << 16) | ((u32)ff_mult(3, p) << 24);
aes_ft_tab[0][i] = t;
aes_ft_tab[1][i] = rol32(t, 8);
aes_ft_tab[2][i] = rol32(t, 16);
aes_ft_tab[3][i] = rol32(t, 24);
p = isb_tab[i];
t = p;
aes_il_tab[0][i] = t;
aes_il_tab[1][i] = rol32(t, 8);
aes_il_tab[2][i] = rol32(t, 16);
aes_il_tab[3][i] = rol32(t, 24);
t = ((u32)ff_mult(14, p)) |
((u32)ff_mult(9, p) << 8) |
((u32)ff_mult(13, p) << 16) |
((u32)ff_mult(11, p) << 24);
aes_it_tab[0][i] = t;
aes_it_tab[1][i] = rol32(t, 8);
aes_it_tab[2][i] = rol32(t, 16);
aes_it_tab[3][i] = rol32(t, 24);
}
}
#define star_x(x) (((x) & 0x7f7f7f7f) << 1) ^ ((((x) & 0x80808080) >> 7) * 0x1b)
#define imix_col(y, x) \
u = star_x(x); \
v = star_x(u); \
w = star_x(v); \
t = w ^ (x); \
(y) = u ^ v ^ w; \
(y) ^= ror32(u ^ t, 8) ^ \
ror32(v ^ t, 16) ^ \
ror32(t, 24)
/* initialise the key schedule from the user supplied key */
#define loop4(i) \
{ \
t = ror32(t, 8); t = ls_box(t) ^ rco_tab[i]; \
t ^= E_KEY[4 * i]; E_KEY[4 * i + 4] = t; \
t ^= E_KEY[4 * i + 1]; E_KEY[4 * i + 5] = t; \
t ^= E_KEY[4 * i + 2]; E_KEY[4 * i + 6] = t; \
t ^= E_KEY[4 * i + 3]; E_KEY[4 * i + 7] = t; \
}
#define loop6(i) \
{ \
t = ror32(t, 8); t = ls_box(t) ^ rco_tab[i]; \
t ^= E_KEY[6 * i]; E_KEY[6 * i + 6] = t; \
t ^= E_KEY[6 * i + 1]; E_KEY[6 * i + 7] = t; \
t ^= E_KEY[6 * i + 2]; E_KEY[6 * i + 8] = t; \
t ^= E_KEY[6 * i + 3]; E_KEY[6 * i + 9] = t; \
t ^= E_KEY[6 * i + 4]; E_KEY[6 * i + 10] = t; \
t ^= E_KEY[6 * i + 5]; E_KEY[6 * i + 11] = t; \
}
#define loop8(i) \
{ \
t = ror32(t, 8); ; t = ls_box(t) ^ rco_tab[i]; \
t ^= E_KEY[8 * i]; E_KEY[8 * i + 8] = t; \
t ^= E_KEY[8 * i + 1]; E_KEY[8 * i + 9] = t; \
t ^= E_KEY[8 * i + 2]; E_KEY[8 * i + 10] = t; \
t ^= E_KEY[8 * i + 3]; E_KEY[8 * i + 11] = t; \
t = E_KEY[8 * i + 4] ^ ls_box(t); \
E_KEY[8 * i + 12] = t; \
t ^= E_KEY[8 * i + 5]; E_KEY[8 * i + 13] = t; \
t ^= E_KEY[8 * i + 6]; E_KEY[8 * i + 14] = t; \
t ^= E_KEY[8 * i + 7]; E_KEY[8 * i + 15] = t; \
}
static int aes_set_key(void *ctx_arg, const u8 *in_key, unsigned int key_len,
u32 *flags)
{
struct aes_ctx *ctx = ctx_arg;
u32 i, j, t, u, v, w;
if (key_len != 16 && key_len != 24 && key_len != 32) {
*flags |= CRYPTO_TFM_RES_BAD_KEY_LEN;
return -EINVAL;
}
ctx->key_length = key_len;
D_KEY[key_len + 24] = E_KEY[0] = u32_in(in_key);
D_KEY[key_len + 25] = E_KEY[1] = u32_in(in_key + 4);
D_KEY[key_len + 26] = E_KEY[2] = u32_in(in_key + 8);
D_KEY[key_len + 27] = E_KEY[3] = u32_in(in_key + 12);
switch (key_len) {
case 16:
t = E_KEY[3];
for (i = 0; i < 10; ++i)
loop4(i);
break;
case 24:
E_KEY[4] = u32_in(in_key + 16);
t = E_KEY[5] = u32_in(in_key + 20);
for (i = 0; i < 8; ++i)
loop6 (i);
break;
case 32:
E_KEY[4] = u32_in(in_key + 16);
E_KEY[5] = u32_in(in_key + 20);
E_KEY[6] = u32_in(in_key + 24);
t = E_KEY[7] = u32_in(in_key + 28);
for (i = 0; i < 7; ++i)
loop8(i);
break;
}
D_KEY[0] = E_KEY[key_len + 24];
D_KEY[1] = E_KEY[key_len + 25];
D_KEY[2] = E_KEY[key_len + 26];
D_KEY[3] = E_KEY[key_len + 27];
for (i = 4; i < key_len + 24; ++i) {
j = key_len + 24 - (i & ~3) + (i & 3);
imix_col(D_KEY[j], E_KEY[i]);
}
return 0;
}
extern void aes_encrypt(void *ctx_arg, u8 *out, const u8 *in);
extern void aes_decrypt(void *ctx_arg, u8 *out, const u8 *in);
static struct crypto_alg aes_alg = {
.cra_name = "aes",
.cra_flags = CRYPTO_ALG_TYPE_CIPHER,
.cra_blocksize = AES_BLOCK_SIZE,
.cra_ctxsize = sizeof(struct aes_ctx),
.cra_module = THIS_MODULE,
.cra_list = LIST_HEAD_INIT(aes_alg.cra_list),
.cra_u = {
.cipher = {
.cia_min_keysize = AES_MIN_KEY_SIZE,
.cia_max_keysize = AES_MAX_KEY_SIZE,
.cia_setkey = aes_set_key,
.cia_encrypt = aes_encrypt,
.cia_decrypt = aes_decrypt
}
}
};
static int __init aes_init(void)
{
gen_tabs();
return crypto_register_alg(&aes_alg);
}
static void __exit aes_fini(void)
{
crypto_unregister_alg(&aes_alg);
}
module_init(aes_init);
module_exit(aes_fini);
MODULE_DESCRIPTION("Rijndael (AES) Cipher Algorithm");
MODULE_LICENSE("GPL");
......@@ -146,7 +146,7 @@ config CRYPTO_SERPENT
config CRYPTO_AES
tristate "AES cipher algorithms"
depends on CRYPTO && !((X86 || UML_X86) && !64BIT)
depends on CRYPTO && !(X86 || UML_X86)
help
AES cipher algorithms (FIPS-197). AES uses the Rijndael
algorithm.
......@@ -184,6 +184,26 @@ config CRYPTO_AES_586
See <http://csrc.nist.gov/encryption/aes/> for more information.
config CRYPTO_AES_X86_64
tristate "AES cipher algorithms (x86_64)"
depends on CRYPTO && ((X86 || UML_X86) && 64BIT)
help
AES cipher algorithms (FIPS-197). AES uses the Rijndael
algorithm.
Rijndael appears to be consistently a very good performer in
both hardware and software across a wide range of computing
environments regardless of its use in feedback or non-feedback
modes. Its key setup time is excellent, and its key agility is
good. Rijndael's very low memory requirements make it very well
suited for restricted-space environments, in which it also
demonstrates excellent performance. Rijndael's operations are
among the easiest to defend against power and timing attacks.
The AES specifies three key sizes: 128, 192 and 256 bits
See <http://csrc.nist.gov/encryption/aes/> for more information.
config CRYPTO_CAST5
tristate "CAST5 (CAST-128) cipher algorithm"
depends on CRYPTO
......
Markdown is supported
0%
or
You are about to add 0 people to the discussion. Proceed with caution.
Finish editing this message first!
Please register or to comment