mirror of
https://github.com/openssl/openssl.git
synced 2024-12-15 06:01:37 +08:00
15b7175f55
This patch implements the SM4 optimization for ARM processor, using SM4 HW instruction, which is an optional feature of crypto extension for aarch64 V8. Tested on some modern ARM micro-architectures with SM4 support, the performance uplift can be observed around 8X~40X over existing C implementation in openssl. Algorithms that can be parallelized (like CTR, ECB, CBC decryption) are on higher end, with algorithm like CBC encryption on lower end (due to inter-block dependency) Perf data on Yitian-710 2.75GHz hardware, before and after optimization: Before: type 16 bytes 64 bytes 256 bytes 1024 bytes 8192 bytes 16384 bytes SM4-CTR 105787.80k 107837.87k 108380.84k 108462.08k 108549.46k 108554.92k SM4-ECB 111924.58k 118173.76k 119776.00k 120093.70k 120264.02k 120274.94k SM4-CBC 106428.09k 109190.98k 109674.33k 109774.51k 109827.41k 109827.41k After (7.4x - 36.6x faster): type 16 bytes 64 bytes 256 bytes 1024 bytes 8192 bytes 16384 bytes SM4-CTR 781979.02k 2432994.28k 3437753.86k 3834177.88k 3963715.58k 3974556.33k SM4-ECB 937590.69k 2941689.02k 3945751.81k 4328655.87k 4459181.40k 4468692.31k SM4-CBC 890639.88k 1027746.58k 1050621.78k 1056696.66k 1058613.93k 1058701.31k 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/17455)
244 lines
5.0 KiB
Perl
Executable File
244 lines
5.0 KiB
Perl
Executable File
#! /usr/bin/env perl
|
|
# Copyright 2015-2020 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
|
|
|
|
|
|
# $output is the last argument if it looks like a file (it has an extension)
|
|
# $flavour is the first argument if it doesn't look like a file
|
|
$output = $#ARGV >= 0 && $ARGV[$#ARGV] =~ m|\.\w+$| ? pop : undef;
|
|
$flavour = $#ARGV >= 0 && $ARGV[0] !~ m|\.| ? shift : undef;
|
|
|
|
$0 =~ m/(.*[\/\\])[^\/\\]+$/; $dir=$1;
|
|
( $xlate="${dir}arm-xlate.pl" and -f $xlate ) or
|
|
( $xlate="${dir}perlasm/arm-xlate.pl" and -f $xlate) or
|
|
die "can't locate arm-xlate.pl";
|
|
|
|
open OUT,"| \"$^X\" $xlate $flavour \"$output\""
|
|
or die "can't call $xlate: $!";
|
|
*STDOUT=*OUT;
|
|
|
|
$code.=<<___;
|
|
#include "arm_arch.h"
|
|
|
|
.text
|
|
.arch armv8-a+crypto
|
|
|
|
.align 5
|
|
.globl _armv7_neon_probe
|
|
.type _armv7_neon_probe,%function
|
|
_armv7_neon_probe:
|
|
AARCH64_VALID_CALL_TARGET
|
|
orr v15.16b, v15.16b, v15.16b
|
|
ret
|
|
.size _armv7_neon_probe,.-_armv7_neon_probe
|
|
|
|
.globl _armv7_tick
|
|
.type _armv7_tick,%function
|
|
_armv7_tick:
|
|
AARCH64_VALID_CALL_TARGET
|
|
#ifdef __APPLE__
|
|
mrs x0, CNTPCT_EL0
|
|
#else
|
|
mrs x0, CNTVCT_EL0
|
|
#endif
|
|
ret
|
|
.size _armv7_tick,.-_armv7_tick
|
|
|
|
.globl _armv8_aes_probe
|
|
.type _armv8_aes_probe,%function
|
|
_armv8_aes_probe:
|
|
AARCH64_VALID_CALL_TARGET
|
|
aese v0.16b, v0.16b
|
|
ret
|
|
.size _armv8_aes_probe,.-_armv8_aes_probe
|
|
|
|
.globl _armv8_sha1_probe
|
|
.type _armv8_sha1_probe,%function
|
|
_armv8_sha1_probe:
|
|
AARCH64_VALID_CALL_TARGET
|
|
sha1h s0, s0
|
|
ret
|
|
.size _armv8_sha1_probe,.-_armv8_sha1_probe
|
|
|
|
.globl _armv8_sha256_probe
|
|
.type _armv8_sha256_probe,%function
|
|
_armv8_sha256_probe:
|
|
AARCH64_VALID_CALL_TARGET
|
|
sha256su0 v0.4s, v0.4s
|
|
ret
|
|
.size _armv8_sha256_probe,.-_armv8_sha256_probe
|
|
|
|
.globl _armv8_pmull_probe
|
|
.type _armv8_pmull_probe,%function
|
|
_armv8_pmull_probe:
|
|
AARCH64_VALID_CALL_TARGET
|
|
pmull v0.1q, v0.1d, v0.1d
|
|
ret
|
|
.size _armv8_pmull_probe,.-_armv8_pmull_probe
|
|
|
|
.globl _armv8_sm4_probe
|
|
.type _armv8_sm4_probe,%function
|
|
_armv8_sm4_probe:
|
|
AARCH64_VALID_CALL_TARGET
|
|
.long 0xcec08400 // sm4e v0.4s, v0.4s
|
|
ret
|
|
.size _armv8_sm4_probe,.-_armv8_sm4_probe
|
|
|
|
.globl _armv8_sha512_probe
|
|
.type _armv8_sha512_probe,%function
|
|
_armv8_sha512_probe:
|
|
AARCH64_VALID_CALL_TARGET
|
|
.long 0xcec08000 // sha512su0 v0.2d,v0.2d
|
|
ret
|
|
.size _armv8_sha512_probe,.-_armv8_sha512_probe
|
|
|
|
.globl _armv8_cpuid_probe
|
|
.type _armv8_cpuid_probe,%function
|
|
_armv8_cpuid_probe:
|
|
AARCH64_VALID_CALL_TARGET
|
|
mrs x0, midr_el1
|
|
ret
|
|
.size _armv8_cpuid_probe,.-_armv8_cpuid_probe
|
|
|
|
.globl _armv8_sm3_probe
|
|
.type _armv8_sm3_probe,%function
|
|
_armv8_sm3_probe:
|
|
AARCH64_VALID_CALL_TARGET
|
|
.long 0xce63c004 // sm3partw1 v4.4s, v0.4s, v3.4s
|
|
ret
|
|
.size _armv8_sm3_probe,.-_armv8_sm3_probe
|
|
|
|
.globl OPENSSL_cleanse
|
|
.type OPENSSL_cleanse,%function
|
|
.align 5
|
|
OPENSSL_cleanse:
|
|
AARCH64_VALID_CALL_TARGET
|
|
cbz x1,.Lret // len==0?
|
|
cmp x1,#15
|
|
b.hi .Lot // len>15
|
|
nop
|
|
.Little:
|
|
strb wzr,[x0],#1 // store byte-by-byte
|
|
subs x1,x1,#1
|
|
b.ne .Little
|
|
.Lret: ret
|
|
|
|
.align 4
|
|
.Lot: tst x0,#7
|
|
b.eq .Laligned // inp is aligned
|
|
strb wzr,[x0],#1 // store byte-by-byte
|
|
sub x1,x1,#1
|
|
b .Lot
|
|
|
|
.align 4
|
|
.Laligned:
|
|
str xzr,[x0],#8 // store word-by-word
|
|
sub x1,x1,#8
|
|
tst x1,#-8
|
|
b.ne .Laligned // len>=8
|
|
cbnz x1,.Little // len!=0?
|
|
ret
|
|
.size OPENSSL_cleanse,.-OPENSSL_cleanse
|
|
|
|
.globl CRYPTO_memcmp
|
|
.type CRYPTO_memcmp,%function
|
|
.align 4
|
|
CRYPTO_memcmp:
|
|
AARCH64_VALID_CALL_TARGET
|
|
eor w3,w3,w3
|
|
cbz x2,.Lno_data // len==0?
|
|
cmp x2,#16
|
|
b.ne .Loop_cmp
|
|
ldp x8,x9,[x0]
|
|
ldp x10,x11,[x1]
|
|
eor x8,x8,x10
|
|
eor x9,x9,x11
|
|
orr x8,x8,x9
|
|
mov x0,#1
|
|
cmp x8,#0
|
|
csel x0,xzr,x0,eq
|
|
ret
|
|
|
|
.align 4
|
|
.Loop_cmp:
|
|
ldrb w4,[x0],#1
|
|
ldrb w5,[x1],#1
|
|
eor w4,w4,w5
|
|
orr w3,w3,w4
|
|
subs x2,x2,#1
|
|
b.ne .Loop_cmp
|
|
|
|
.Lno_data:
|
|
neg w0,w3
|
|
lsr w0,w0,#31
|
|
ret
|
|
.size CRYPTO_memcmp,.-CRYPTO_memcmp
|
|
|
|
.globl _armv8_rng_probe
|
|
.type _armv8_rng_probe,%function
|
|
_armv8_rng_probe:
|
|
mrs x0, s3_3_c2_c4_0 // rndr
|
|
mrs x0, s3_3_c2_c4_1 // rndrrs
|
|
ret
|
|
.size _armv8_rng_probe,.-_armv8_rng_probe
|
|
___
|
|
|
|
sub gen_random {
|
|
my $rdop = shift;
|
|
my $rand_reg = $rdop eq "rndr" ? "s3_3_c2_c4_0" : "s3_3_c2_c4_1";
|
|
|
|
print<<___;
|
|
// Fill buffer with Randomly Generated Bytes
|
|
// inputs: char * in x0 - Pointer to buffer
|
|
// size_t in x1 - Number of bytes to write to buffer
|
|
// outputs: size_t in x0 - Number of bytes successfully written to buffer
|
|
.globl OPENSSL_${rdop}_asm
|
|
.type OPENSSL_${rdop}_asm,%function
|
|
.align 4
|
|
OPENSSL_${rdop}_asm:
|
|
mov x2,xzr
|
|
mov x3,xzr
|
|
|
|
.align 4
|
|
.Loop_${rdop}:
|
|
cmp x1,#0
|
|
b.eq .${rdop}_done
|
|
mov x3,xzr
|
|
mrs x3,$rand_reg
|
|
b.eq .${rdop}_done
|
|
|
|
cmp x1,#8
|
|
b.lt .Loop_single_byte_${rdop}
|
|
|
|
str x3,[x0]
|
|
add x0,x0,#8
|
|
add x2,x2,#8
|
|
subs x1,x1,#8
|
|
b.ge .Loop_${rdop}
|
|
|
|
.align 4
|
|
.Loop_single_byte_${rdop}:
|
|
strb w3,[x0]
|
|
lsr x3,x3,#8
|
|
add x2,x2,#1
|
|
add x0,x0,#1
|
|
subs x1,x1,#1
|
|
b.gt .Loop_single_byte_${rdop}
|
|
|
|
.align 4
|
|
.${rdop}_done:
|
|
mov x0,x2
|
|
ret
|
|
.size OPENSSL_${rdop}_asm,.-OPENSSL_${rdop}_asm
|
|
___
|
|
}
|
|
gen_random("rndr");
|
|
gen_random("rndrrs");
|
|
|
|
print $code;
|
|
close STDOUT or die "error closing STDOUT: $!";
|