--- /dev/null
+#! /usr/bin/env perl
+# Copyright 2025 The OpenSSL Project Authors. All Rights Reserved.
+# Copyright (c) 2025, Intel Corporation. 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
+#
+#
+# This module implements support for Intel(R) SM4 instructions
+# from Intel(R) Multi-Buffer Crypto for IPsec Library
+# (https://github.com/intel/intel-ipsec-mb).
+# Original author is Tomasz Kantecki <tomasz.kantecki@intel.com>
+
+# $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 && $ARGV[0] !~ m|\.| ? shift : undef;
+
+$win64=0;
+$win64=1 if ($flavour =~ /[nm]asm|mingw64/ || $output =~ /\.asm$/);
+
+$0 =~ m/(.*[\/\\])[^\/\\]+$/;
+$dir=$1;
+( $xlate="${dir}x86_64-xlate.pl" and -f $xlate ) or
+( $xlate="${dir}../../perlasm/x86_64-xlate.pl" and -f $xlate) or
+die "can't locate x86_64-xlate.pl";
+
+# Check Intel(R) SM4 instructions support
+if (`$ENV{CC} -Wa,-v -c -o /dev/null -x assembler /dev/null 2>&1`
+ =~ /GNU assembler version ([2-9]\.[0-9]+)/) {
+ $avx2_sm4_ni = ($1>=2.22); # minimal avx2 supported version, binary translation for SM4 instructions (sub sm4op) is used
+ $avx2_sm4_ni_native = ($1>=2.42); # support added at GNU asm 2.42
+}
+
+if (!$avx2_sm4_ni && $win64 && ($flavour =~ /nasm/ || $ENV{ASM} =~ /nasm/) &&
+ `nasm -v 2>&1` =~ /NASM version ([2-9])\.([0-9]+)\.([0-9]+)/) {
+ my ($major, $minor, $patch) = ($1, $2, $3);
+ $avx2_sm4_ni = ($major > 2) || ($major == 2 && $minor > 10); # minimal avx2 supported version, binary translation for SM4 instructions (sub sm4op) is used
+ $avx2_sm4_ni_native = ($major > 2) || ($major == 2 && $minor > 16) || ($major == 2 && $minor == 16 && $patch >= 2); # support added at NASM 2.16.02
+}
+
+if (!$avx2_sm4_ni && `$ENV{CC} -v 2>&1` =~ /((?:clang|LLVM) version|.*based on LLVM) ([0-9]+\.[0-9]+)/) {
+ $avx2_sm4_ni = ($2>=7.0); # minimal tested version, binary translation for SM4 instructions (sub sm4op) is used
+ $avx2_sm4_ni_native = ($2>=17.0); # support added at LLVM 17.0.1
+}
+
+open OUT,"| \"$^X\" \"$xlate\" $flavour \"$output\""
+ or die "can't call $xlate: $!";
+*STDOUT=*OUT;
+
+$prefix="hw_x86_64_sm4";
+
+if ($avx2_sm4_ni>0) {
+
+$code.= ".text\n";
+{
+# input arguments aliases for set_key
+my ($userKey,$key) = ("%rdi","%rsi");
+
+# input arguments aliases for encrypt/decrypt
+my ($in,$out,$ks) = ("%rdi","%rsi","%rdx");
+
+$code.=<<___;
+.section .rodata align=64
+.align 16
+SM4_FK:
+.long 0xa3b1bac6, 0x56aa3350, 0x677d9197, 0xb27022dc
+
+.align 16
+SM4_CK:
+.long 0x00070E15, 0x1C232A31, 0x383F464D, 0x545B6269
+.long 0x70777E85, 0x8C939AA1, 0xA8AFB6BD, 0xC4CBD2D9
+.long 0xE0E7EEF5, 0xFC030A11, 0x181F262D, 0x343B4249
+.long 0x50575E65, 0x6C737A81, 0x888F969D, 0xA4ABB2B9
+.long 0xC0C7CED5, 0xDCE3EAF1, 0xF8FF060D, 0x141B2229
+.long 0x30373E45, 0x4C535A61, 0x686F767D, 0x848B9299
+.long 0xA0A7AEB5, 0xBCC3CAD1, 0xD8DFE6ED, 0xF4FB0209
+.long 0x10171E25, 0x2C333A41, 0x484F565D, 0x646B7279
+
+IN_SHUFB:
+.byte 0x03, 0x02, 0x01, 0x00, 0x07, 0x06, 0x05, 0x04
+.byte 0x0b, 0x0a, 0x09, 0x08, 0x0f, 0x0e, 0x0d, 0x0c
+.byte 0x03, 0x02, 0x01, 0x00, 0x07, 0x06, 0x05, 0x04
+.byte 0x0b, 0x0a, 0x09, 0x08, 0x0f, 0x0e, 0x0d, 0x0c
+
+OUT_SHUFB:
+.byte 0x0f, 0x0e, 0x0d, 0x0c, 0x0b, 0x0a, 0x09, 0x08
+.byte 0x07, 0x06, 0x05, 0x04, 0x03, 0x02, 0x01, 0x00
+.byte 0x0f, 0x0e, 0x0d, 0x0c, 0x0b, 0x0a, 0x09, 0x08
+.byte 0x07, 0x06, 0x05, 0x04, 0x03, 0x02, 0x01, 0x00
+
+.text
+
+# int ${prefix}_set_key(const unsigned char *userKey, SM4_KEY *key)
+#
+# input: $userKey secret key
+# $key round keys
+#
+
+.globl ${prefix}_set_key
+.type ${prefix}_set_key,\@function,2
+.align 32
+${prefix}_set_key:
+.cfi_startproc
+ endbranch
+# Prolog
+ push %rbp
+.cfi_push %rbp
+# Prolog ends here.
+.Lossl_${prefix}_set_key_seh_prolog_end:
+
+ vmovdqu ($userKey), %xmm0
+ vpshufb IN_SHUFB(%rip), %xmm0, %xmm0
+ vpxor SM4_FK(%rip), %xmm0, %xmm0
+
+ vmovdqu SM4_CK(%rip), %xmm1
+ vsm4key4 %xmm1, %xmm0, %xmm0
+ vmovdqu %xmm0, ($key)
+ vmovdqu SM4_CK + 16(%rip), %xmm1
+ vsm4key4 %xmm1, %xmm0, %xmm0
+ vmovdqu %xmm0, 16($key)
+ vmovdqu SM4_CK + 32(%rip), %xmm1
+ vsm4key4 %xmm1, %xmm0, %xmm0
+ vmovdqu %xmm0, 32($key)
+ vmovdqu SM4_CK + 48(%rip), %xmm1
+ vsm4key4 %xmm1, %xmm0, %xmm0
+ vmovdqu %xmm0, 48($key)
+ vmovdqu SM4_CK + 64(%rip), %xmm1
+ vsm4key4 %xmm1, %xmm0, %xmm0
+ vmovdqu %xmm0, 64($key)
+ vmovdqu SM4_CK + 80(%rip), %xmm1
+ vsm4key4 %xmm1, %xmm0, %xmm0
+ vmovdqu %xmm0, 80($key)
+ vmovdqu SM4_CK + 96(%rip), %xmm1
+ vsm4key4 %xmm1, %xmm0, %xmm0
+ vmovdqu %xmm0, 96($key)
+ vmovdqu SM4_CK + 112(%rip), %xmm1
+ vsm4key4 %xmm1, %xmm0, %xmm0
+ vmovdqu %xmm0, 112($key)
+
+ vpxor %xmm0, %xmm0, %xmm0 # clear register
+ mov \$1, %eax
+ pop %rbp
+.cfi_pop %rbp
+ ret
+.cfi_endproc
+
+# void ${prefix}_encrypt(const uint8_t *in, uint8_t *out, const SM4_KEY *ks)
+
+.globl ${prefix}_encrypt
+.type ${prefix}_encrypt,\@function,3
+.align 32
+${prefix}_encrypt:
+.cfi_startproc
+ endbranch
+# Prolog
+ push %rbp
+.cfi_push %rbp
+# Prolog ends here.
+.Lossl_${prefix}_encrypt_seh_prolog_end:
+
+ vmovdqu ($in), %xmm0
+ vpshufb IN_SHUFB(%rip), %xmm0, %xmm0
+
+ # note: to simplify binary instructions translation
+ mov $ks, %r10
+
+ vsm4rnds4 (%r10), %xmm0, %xmm0
+ vsm4rnds4 16(%r10), %xmm0, %xmm0
+ vsm4rnds4 32(%r10), %xmm0, %xmm0
+ vsm4rnds4 48(%r10), %xmm0, %xmm0
+ vsm4rnds4 64(%r10), %xmm0, %xmm0
+ vsm4rnds4 80(%r10), %xmm0, %xmm0
+ vsm4rnds4 96(%r10), %xmm0, %xmm0
+ vsm4rnds4 112(%r10), %xmm0, %xmm0
+
+ vpshufb OUT_SHUFB(%rip), %xmm0, %xmm0
+ vmovdqu %xmm0, ($out)
+ vpxor %xmm0, %xmm0, %xmm0 # clear register
+ pop %rbp
+.cfi_pop %rbp
+ ret
+.cfi_endproc
+
+# void ${prefix}_decrypt(const uint8_t *in, uint8_t *out, const SM4_KEY *ks)
+
+.globl ${prefix}_decrypt
+.type ${prefix}_decrypt,\@function,3
+.align 32
+${prefix}_decrypt:
+.cfi_startproc
+ endbranch
+# Prolog
+ push %rbp
+.cfi_push %rbp
+# Prolog ends here.
+.Lossl_${prefix}_decrypt_seh_prolog_end:
+
+ vmovdqu ($in), %xmm0
+ vpshufb IN_SHUFB(%rip), %xmm0, %xmm0
+
+ vmovdqu 112($ks), %xmm1
+ vpshufd \$27, %xmm1, %xmm1
+ vsm4rnds4 %xmm1, %xmm0, %xmm0
+ vmovdqu 96($ks), %xmm1
+ vpshufd \$27, %xmm1, %xmm1
+ vsm4rnds4 %xmm1, %xmm0, %xmm0
+ vmovdqu 80($ks), %xmm1
+ vpshufd \$27, %xmm1, %xmm1
+ vsm4rnds4 %xmm1, %xmm0, %xmm0
+ vmovdqu 64($ks), %xmm1
+ vpshufd \$27, %xmm1, %xmm1
+ vsm4rnds4 %xmm1, %xmm0, %xmm0
+ vmovdqu 48($ks), %xmm1
+ vpshufd \$27, %xmm1, %xmm1
+ vsm4rnds4 %xmm1, %xmm0, %xmm0
+ vmovdqu 32($ks), %xmm1
+ vpshufd \$27, %xmm1, %xmm1
+ vsm4rnds4 %xmm1, %xmm0, %xmm0
+ vmovdqu 16($ks), %xmm1
+ vpshufd \$27, %xmm1, %xmm1
+ vsm4rnds4 %xmm1, %xmm0, %xmm0
+ vmovdqu ($ks), %xmm1
+ vpshufd \$27, %xmm1, %xmm1
+ vsm4rnds4 %xmm1, %xmm0, %xmm0
+
+ vpshufb OUT_SHUFB(%rip), %xmm0, %xmm0
+ vmovdqu %xmm0, ($out)
+ vpxor %xmm0, %xmm0, %xmm0 # clear registers
+ vpxor %xmm1, %xmm1, %xmm1
+ pop %rbp
+.cfi_pop %rbp
+ ret
+.cfi_endproc
+___
+}
+
+} else { # fallback for the unsupported configurations with undefined instruction
+
+$code .= <<___;
+.text
+
+.globl ${prefix}_set_key
+.type ${prefix}_set_key,\@abi-omnipotent
+${prefix}_set_key:
+ .byte 0x0f,0x0b # ud2
+ ret
+.size ${prefix}_set_key, .-${prefix}_set_key
+
+.globl ${prefix}_encrypt
+.type ${prefix}_encrypt,\@abi-omnipotent
+${prefix}_encrypt:
+ .byte 0x0f,0x0b # ud2
+ ret
+.size ${prefix}_encrypt, .-${prefix}_encrypt
+
+.globl ${prefix}_decrypt
+.type ${prefix}_decrypt,\@abi-omnipotent
+${prefix}_decrypt:
+ .byte 0x0f,0x0b # ud2
+ ret
+.size ${prefix}_decrypt, .-${prefix}_decrypt
+___
+} # avx2_sm4_ni
+
+if ($avx2_sm4_ni_native > 0) { # SM4 instructions are supported in asm
+ $code =~ s/\`([^\`]*)\`/eval $1/gem;
+ print $code;
+} else { # binary translation for SM4 instructions
+ sub sm4op {
+ my $instr = shift;
+ my $args = shift;
+ if ($args =~ /^(.+)\s*#/) {
+ $args = $1; # drop comment and its leading whitespace
+ }
+ if (($instr eq "vsm4key4") && ($args =~ /%xmm(\d{1,2})\s*,\s*%xmm(\d{1,2})\s*,\s*%xmm(\d{1,2})/)) {
+ my $b1 = sprintf("0x%02x", 0x62 | ((1-int($1/8))<<5) | ((1-int($3/8))<<7) );
+ my $b2 = sprintf("0x%02x", 0x02 | (15 - $2 & 15)<<3 );
+ my $b3 = sprintf("0x%02x", 0xc0 | ($1 & 7) | (($3 & 7)<<3) );
+ return ".byte 0xc4,".$b1.",".$b2.",0xda,".$b3;
+ }
+ elsif (($instr eq "vsm4rnds4") && ($args =~ /(\d*)\(([^)]+)\)\s*,\s*%xmm(\d{1,2})\s*,\s*%xmm(\d{1,2})/)) {
+ my $shift = $1;
+ my $b3_offset = 0x00;
+ if ($shift) {
+ $shift = ",0x".sprintf("%02x", $shift);
+ $b3_offset = 0x40;
+ }
+ my $b1 = sprintf("0x%02x", 0x42 | ((1-int($4/8))<<7) );
+ my $b2 = sprintf("0x%02x", 0x03 | (15 - $3 & 15)<<3 );
+ my $b3 = sprintf("0x%02x", 0x02 | ($4 & 7)<<3 | $b3_offset );
+ return ".byte 0xc4,".$b1.",".$b2.",0xda,".$b3.$shift;
+ }
+ elsif (($instr eq "vsm4rnds4") && ($args =~ /%xmm(\d{1,2})\s*,\s*%xmm(\d{1,2})\s*,\s*%xmm(\d{1,2})/)) {
+ my $b1 = sprintf("0x%02x", 0x62 | ((1-int($1/8))<<5) | ((1-int($3/8))<<7) );
+ my $b2 = sprintf("0x%02x", 0x03 | (15 - $2 & 15)<<3 );
+ my $b3 = sprintf("0x%02x", 0xc0 | ($1 & 7) | (($3 & 7)<<3) );
+ return ".byte 0xc4,".$b1.",".$b2.",0xda,".$b3;
+ }
+ return $instr."\t".$args;
+ }
+
+ foreach (split("\n",$code)) {
+ s/\`([^\`]*)\`/eval $1/geo;
+ s/\b(vsm4[^\s]*)\s+(.*)/sm4op($1,$2)/geo;
+ print $_,"\n";
+ }
+} # avx2_sm4_ni_native > 0
+
+close STDOUT or die "error closing STDOUT: $!";
projects / thirdparty / openssl.git / commitdiff
