openssl/include/crypto/sm4_platform.h
Daniel Hu 4908787f21 SM4 optimization for ARM by ASIMD
This patch optimizes SM4 for ARM processor using ASIMD instruction

It will improve performance if both of following conditions are met:
1) Input data equal to or more than 4 blocks
2) Cipher mode allows parallelism, including ECB,CTR,GCM or CBC decryption

This patch implements SM4 SBOX lookup in vector registers, with the
benefit of constant processing time over existing C implementation.

It is only enabled for micro-architecture N1/V1. In the ideal scenario,
performance can reach up to 2.7X

When either of above two conditions is not met, e.g. single block input
or CFB/OFB mode, CBC encryption, performance could drop about 50%.

The assembly code has been reviewed internally by ARM engineer
Fangming.Fang@arm.com

Signed-off-by: Daniel Hu <Daniel.Hu@arm.com>

Reviewed-by: Paul Dale <pauli@openssl.org>
Reviewed-by: Tomas Mraz <tomas@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/17951)
2022-04-12 10:37:42 +02:00

78 lines
3.3 KiB
C

/*
* Copyright 2022 The OpenSSL Project Authors. All Rights Reserved.
*
* Licensed under the Apache License 2.0 (the "License"). You may not use
* this file except in compliance with the License. You can obtain a copy
* in the file LICENSE in the source distribution or at
* https://www.openssl.org/source/license.html
*/
#ifndef OSSL_SM4_PLATFORM_H
# define OSSL_SM4_PLATFORM_H
# pragma once
# if defined(OPENSSL_CPUID_OBJ)
# if (defined(__arm__) || defined(__arm) || defined(__aarch64__))
# include "arm_arch.h"
# if __ARM_MAX_ARCH__>=8
extern unsigned int OPENSSL_arm_midr;
static inline int vpsm4_capable(void)
{
return (OPENSSL_armcap_P & ARMV8_CPUID) &&
(MIDR_IS_CPU_MODEL(OPENSSL_arm_midr, ARM_CPU_IMP_ARM, ARM_CPU_PART_V1) ||
MIDR_IS_CPU_MODEL(OPENSSL_arm_midr, ARM_CPU_IMP_ARM, ARM_CPU_PART_N1));
}
# if defined(VPSM4_ASM)
# define VPSM4_CAPABLE vpsm4_capable()
# endif
# define HWSM4_CAPABLE (OPENSSL_armcap_P & ARMV8_SM4)
# define HWSM4_set_encrypt_key sm4_v8_set_encrypt_key
# define HWSM4_set_decrypt_key sm4_v8_set_decrypt_key
# define HWSM4_encrypt sm4_v8_encrypt
# define HWSM4_decrypt sm4_v8_decrypt
# define HWSM4_cbc_encrypt sm4_v8_cbc_encrypt
# define HWSM4_ecb_encrypt sm4_v8_ecb_encrypt
# define HWSM4_ctr32_encrypt_blocks sm4_v8_ctr32_encrypt_blocks
# endif
# endif
# endif /* OPENSSL_CPUID_OBJ */
# if defined(HWSM4_CAPABLE)
int HWSM4_set_encrypt_key(const unsigned char *userKey, SM4_KEY *key);
int HWSM4_set_decrypt_key(const unsigned char *userKey, SM4_KEY *key);
void HWSM4_encrypt(const unsigned char *in, unsigned char *out,
const SM4_KEY *key);
void HWSM4_decrypt(const unsigned char *in, unsigned char *out,
const SM4_KEY *key);
void HWSM4_cbc_encrypt(const unsigned char *in, unsigned char *out,
size_t length, const SM4_KEY *key,
unsigned char *ivec, const int enc);
void HWSM4_ecb_encrypt(const unsigned char *in, unsigned char *out,
size_t length, const SM4_KEY *key,
const int enc);
void HWSM4_ctr32_encrypt_blocks(const unsigned char *in, unsigned char *out,
size_t len, const void *key,
const unsigned char ivec[16]);
# endif /* HWSM4_CAPABLE */
#ifdef VPSM4_CAPABLE
int vpsm4_set_encrypt_key(const unsigned char *userKey, SM4_KEY *key);
int vpsm4_set_decrypt_key(const unsigned char *userKey, SM4_KEY *key);
void vpsm4_encrypt(const unsigned char *in, unsigned char *out,
const SM4_KEY *key);
void vpsm4_decrypt(const unsigned char *in, unsigned char *out,
const SM4_KEY *key);
void vpsm4_cbc_encrypt(const unsigned char *in, unsigned char *out,
size_t length, const SM4_KEY *key,
unsigned char *ivec, const int enc);
void vpsm4_ecb_encrypt(const unsigned char *in, unsigned char *out,
size_t length, const SM4_KEY *key,
const int enc);
void vpsm4_ctr32_encrypt_blocks(const unsigned char *in, unsigned char *out,
size_t len, const void *key,
const unsigned char ivec[16]);
# endif /* VPSM4_CAPABLE */
#endif /* OSSL_SM4_PLATFORM_H */