#!/usr/bin/env perl # Copyright 2024 The BoringSSL Authors # # Licensed under the Apache License, Version 2.0 (the "License"); # you may not use this file except in compliance with the License. # You may obtain a copy of the License at # # https://www.apache.org/licenses/LICENSE-2.0 # # Unless required by applicable law or agreed to in writing, software # distributed under the License is distributed on an "AS IS" BASIS, # WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. # See the License for the specific language governing permissions and # limitations under the License. # #------------------------------------------------------------------------------ # # This is an AES-GCM implementation for x86_64 CPUs that support the following # CPU features: VAES && VPCLMULQDQ && AVX2. # # This is similar to aes-gcm-avx512-x86_64.pl, but it uses AVX2 instead of # AVX512. This means it can only use 16 vector registers instead of 32, the # maximum vector length is 32 bytes, and some instructions such as vpternlogd # and masked loads/stores are unavailable. However, it is able to run on CPUs # that have VAES without AVX512, namely AMD Zen 3 (including "Milan" server # processors) and some Intel client CPUs such as Alder Lake. # # This implementation also uses Karatsuba multiplication instead of schoolbook # multiplication for GHASH in its main loop. This does not help much on Intel, # but it improves performance by ~5% on AMD Zen 3 which is the main target for # this implementation. Other factors weighing slightly in favor of Karatsuba # multiplication in this implementation are the lower maximum vector length # (which means there is space left in the Htable array to cache the halves of # the key powers XOR'd together) and the unavailability of the vpternlogd # instruction (which helped schoolbook a bit more than Karatsuba). use strict; my $flavour = shift; my $output = shift; if ( $flavour =~ /\./ ) { $output = $flavour; undef $flavour; } my $win64; my @argregs; if ( $flavour =~ /[nm]asm|mingw64/ || $output =~ /\.asm$/ ) { $win64 = 1; @argregs = ( "%rcx", "%rdx", "%r8", "%r9" ); } else { $win64 = 0; @argregs = ( "%rdi", "%rsi", "%rdx", "%rcx", "%r8", "%r9" ); } $0 =~ m/(.*[\/\\])[^\/\\]+$/; my $dir = $1; my $xlate; ( $xlate = "${dir}x86_64-xlate.pl" and -f $xlate ) or ( $xlate = "${dir}../../../perlasm/x86_64-xlate.pl" and -f $xlate ) or die "can't locate x86_64-xlate.pl"; open OUT, "| \"$^X\" \"$xlate\" $flavour \"$output\""; *STDOUT = *OUT; my $g_cur_func_name; my $g_cur_func_uses_seh; my @g_cur_func_saved_gpregs; my @g_cur_func_saved_xmmregs; sub _begin_func { my ( $funcname, $uses_seh ) = @_; $g_cur_func_name = $funcname; $g_cur_func_uses_seh = $uses_seh; @g_cur_func_saved_gpregs = (); @g_cur_func_saved_xmmregs = (); return <<___; .globl $funcname .type $funcname,\@abi-omnipotent .align 32 $funcname: .cfi_startproc @{[ $uses_seh ? ".seh_startproc" : "" ]} _CET_ENDBR ___ } # Push a list of general purpose registers onto the stack. sub _save_gpregs { my @gpregs = @_; my $code = ""; die "_save_gpregs requires uses_seh" unless $g_cur_func_uses_seh; die "_save_gpregs can only be called once per function" if @g_cur_func_saved_gpregs; die "Order must be _save_gpregs, then _save_xmmregs" if @g_cur_func_saved_xmmregs; @g_cur_func_saved_gpregs = @gpregs; for my $reg (@gpregs) { $code .= "push $reg\n"; if ($win64) { $code .= ".seh_pushreg $reg\n"; } else { $code .= ".cfi_push $reg\n"; } } return $code; } # Push a list of xmm registers onto the stack if the target is Windows. sub _save_xmmregs { my @xmmregs = @_; my $num_xmmregs = scalar @xmmregs; my $code = ""; die "_save_xmmregs requires uses_seh" unless $g_cur_func_uses_seh; die "_save_xmmregs can only be called once per function" if @g_cur_func_saved_xmmregs; if ( $win64 and $num_xmmregs > 0 ) { @g_cur_func_saved_xmmregs = @xmmregs; my $is_misaligned = ( scalar @g_cur_func_saved_gpregs ) % 2 == 0; my $alloc_size = 16 * $num_xmmregs + ( $is_misaligned ? 8 : 0 ); $code .= "sub \$$alloc_size, %rsp\n"; $code .= ".seh_stackalloc $alloc_size\n"; for my $i ( 0 .. $num_xmmregs - 1 ) { my $reg_num = $xmmregs[$i]; my $pos = 16 * $i; $code .= "vmovdqa %xmm$reg_num, $pos(%rsp)\n"; $code .= ".seh_savexmm %xmm$reg_num, $pos\n"; } } return $code; } sub _end_func { my $code = ""; # Restore any xmm registers that were saved earlier. my $num_xmmregs = scalar @g_cur_func_saved_xmmregs; if ( $win64 and $num_xmmregs > 0 ) { my $need_alignment = ( scalar @g_cur_func_saved_gpregs ) % 2 == 0; my $alloc_size = 16 * $num_xmmregs + ( $need_alignment ? 8 : 0 ); for my $i ( 0 .. $num_xmmregs - 1 ) { my $reg_num = $g_cur_func_saved_xmmregs[$i]; my $pos = 16 * $i; $code .= "vmovdqa $pos(%rsp), %xmm$reg_num\n"; } $code .= "add \$$alloc_size, %rsp\n"; } # Restore any general purpose registers that were saved earlier. for my $reg ( reverse @g_cur_func_saved_gpregs ) { $code .= "pop $reg\n"; if ( !$win64 ) { $code .= ".cfi_pop $reg\n"; } } $code .= <<___; ret @{[ $g_cur_func_uses_seh ? ".seh_endproc" : "" ]} .cfi_endproc .size $g_cur_func_name, . - $g_cur_func_name ___ return $code; } my $code = <<___; .section .rodata .align 16 # A shuffle mask that reflects the bytes of 16-byte blocks .Lbswap_mask: .quad 0x08090a0b0c0d0e0f, 0x0001020304050607 # This is the GHASH reducing polynomial without its constant term, i.e. # x^128 + x^7 + x^2 + x, represented using the backwards mapping # between bits and polynomial coefficients. # # Alternatively, it can be interpreted as the naturally-ordered # representation of the polynomial x^127 + x^126 + x^121 + 1, i.e. the # "reversed" GHASH reducing polynomial without its x^128 term. .Lgfpoly: .quad 1, 0xc200000000000000 # Same as above, but with the (1 << 64) bit set. .Lgfpoly_and_internal_carrybit: .quad 1, 0xc200000000000001 .align 32 # The below constants are used for incrementing the counter blocks. .Lctr_pattern: .quad 0, 0 .quad 1, 0 .Linc_2blocks: .quad 2, 0 .quad 2, 0 .text ___ # We use Htable[0..7] to store H^8 through H^1, and Htable[8..11] to store the # 64-bit halves of the key powers XOR'd together (for Karatsuba multiplication) # in the order 8,6,7,5,4,2,3,1. We do not use Htable[12..15]. my $NUM_H_POWERS = 8; my $OFFSETOFEND_H_POWERS = $NUM_H_POWERS * 16; my $OFFSETOF_H_POWERS_XORED = $OFFSETOFEND_H_POWERS; # Offset to 'rounds' in AES_KEY struct my $OFFSETOF_AES_ROUNDS = 240; # GHASH-multiply the 128-bit lanes of \a by the 128-bit lanes of \b and store # the reduced products in \dst. Uses schoolbook multiplication. sub _ghash_mul { my ( $a, $b, $dst, $gfpoly, $t0, $t1, $t2 ) = @_; return <<___; vpclmulqdq \$0x00, $a, $b, $t0 # LO = a_L * b_L vpclmulqdq \$0x01, $a, $b, $t1 # MI_0 = a_L * b_H vpclmulqdq \$0x10, $a, $b, $t2 # MI_1 = a_H * b_L vpxor $t2, $t1, $t1 # MI = MI_0 + MI_1 vpclmulqdq \$0x01, $t0, $gfpoly, $t2 # LO_L*(x^63 + x^62 + x^57) vpshufd \$0x4e, $t0, $t0 # Swap halves of LO vpxor $t0, $t1, $t1 # Fold LO into MI (part 1) vpxor $t2, $t1, $t1 # Fold LO into MI (part 2) vpclmulqdq \$0x11, $a, $b, $dst # HI = a_H * b_H vpclmulqdq \$0x01, $t1, $gfpoly, $t0 # MI_L*(x^63 + x^62 + x^57) vpshufd \$0x4e, $t1, $t1 # Swap halves of MI vpxor $t1, $dst, $dst # Fold MI into HI (part 1) vpxor $t0, $dst, $dst # Fold MI into HI (part 2) ___ } # This is a specialized version of _ghash_mul that computes \a * \a, i.e. it # squares \a. It skips computing MI = (a_L * a_H) + (a_H * a_L) = 0. sub _ghash_square { my ( $a, $dst, $gfpoly, $t0, $t1 ) = @_; return <<___; vpclmulqdq \$0x00, $a, $a, $t0 # LO = a_L * a_L vpclmulqdq \$0x11, $a, $a, $dst # HI = a_H * a_H vpclmulqdq \$0x01, $t0, $gfpoly, $t1 # LO_L*(x^63 + x^62 + x^57) vpshufd \$0x4e, $t0, $t0 # Swap halves of LO vpxor $t0, $t1, $t1 # Fold LO into MI vpclmulqdq \$0x01, $t1, $gfpoly, $t0 # MI_L*(x^63 + x^62 + x^57) vpshufd \$0x4e, $t1, $t1 # Swap halves of MI vpxor $t1, $dst, $dst # Fold MI into HI (part 1) vpxor $t0, $dst, $dst # Fold MI into HI (part 2) ___ } # void gcm_init_vpclmulqdq_avx2(u128 Htable[16], const uint64_t H[2]); # # Initialize |Htable| with powers of the GHASH subkey |H|. # # We use Htable[0..7] to store H^8 through H^1, and Htable[8..11] to store the # 64-bit halves of the key powers XOR'd together (for Karatsuba multiplication) # in the order 8,6,7,5,4,2,3,1. We do not use Htable[12..15]. $code .= _begin_func "gcm_init_vpclmulqdq_avx2", 1; { my ( $HTABLE, $H_PTR ) = @argregs[ 0 .. 1 ]; my ( $TMP0, $TMP0_XMM ) = ( "%ymm0", "%xmm0" ); my ( $TMP1, $TMP1_XMM ) = ( "%ymm1", "%xmm1" ); my ( $TMP2, $TMP2_XMM ) = ( "%ymm2", "%xmm2" ); my ( $H_CUR, $H_CUR_XMM ) = ( "%ymm3", "%xmm3" ); my ( $H_CUR2, $H_CUR2_XMM ) = ( "%ymm4", "%xmm4" ); my ( $H_INC, $H_INC_XMM ) = ( "%ymm5", "%xmm5" ); my ( $GFPOLY, $GFPOLY_XMM ) = ( "%ymm6", "%xmm6" ); $code .= <<___; @{[ _save_xmmregs (6) ]} .seh_endprologue # Load the byte-reflected hash subkey. BoringSSL provides it in # byte-reflected form except the two halves are in the wrong order. vpshufd \$0x4e, ($H_PTR), $H_CUR_XMM # Finish preprocessing the byte-reflected hash subkey by multiplying it by # x^-1 ("standard" interpretation of polynomial coefficients) or # equivalently x^1 (natural interpretation). This gets the key into a # format that avoids having to bit-reflect the data blocks later. vpshufd \$0xd3, $H_CUR_XMM, $TMP0_XMM vpsrad \$31, $TMP0_XMM, $TMP0_XMM vpaddq $H_CUR_XMM, $H_CUR_XMM, $H_CUR_XMM vpand .Lgfpoly_and_internal_carrybit(%rip), $TMP0_XMM, $TMP0_XMM vpxor $TMP0_XMM, $H_CUR_XMM, $H_CUR_XMM vbroadcasti128 .Lgfpoly(%rip), $GFPOLY # Square H^1 to get H^2. @{[ _ghash_square $H_CUR_XMM, $H_INC_XMM, $GFPOLY_XMM, $TMP0_XMM, $TMP1_XMM ]} # Create H_CUR = [H^2, H^1] and H_INC = [H^2, H^2]. vinserti128 \$1, $H_CUR_XMM, $H_INC, $H_CUR vinserti128 \$1, $H_INC_XMM, $H_INC, $H_INC # Compute H_CUR2 = [H^4, H^3]. @{[ _ghash_mul $H_INC, $H_CUR, $H_CUR2, $GFPOLY, $TMP0, $TMP1, $TMP2 ]} # Store [H^2, H^1] and [H^4, H^3]. vmovdqu $H_CUR, 3*32($HTABLE) vmovdqu $H_CUR2, 2*32($HTABLE) # For Karatsuba multiplication: compute and store the two 64-bit halves of # each key power XOR'd together. Order is 4,2,3,1. vpunpcklqdq $H_CUR, $H_CUR2, $TMP0 vpunpckhqdq $H_CUR, $H_CUR2, $TMP1 vpxor $TMP1, $TMP0, $TMP0 vmovdqu $TMP0, $OFFSETOF_H_POWERS_XORED+32($HTABLE) # Compute and store H_CUR = [H^6, H^5] and H_CUR2 = [H^8, H^7]. @{[ _ghash_mul $H_INC, $H_CUR2, $H_CUR, $GFPOLY, $TMP0, $TMP1, $TMP2 ]} @{[ _ghash_mul $H_INC, $H_CUR, $H_CUR2, $GFPOLY, $TMP0, $TMP1, $TMP2 ]} vmovdqu $H_CUR, 1*32($HTABLE) vmovdqu $H_CUR2, 0*32($HTABLE) # Again, compute and store the two 64-bit halves of each key power XOR'd # together. Order is 8,6,7,5. vpunpcklqdq $H_CUR, $H_CUR2, $TMP0 vpunpckhqdq $H_CUR, $H_CUR2, $TMP1 vpxor $TMP1, $TMP0, $TMP0 vmovdqu $TMP0, $OFFSETOF_H_POWERS_XORED($HTABLE) vzeroupper ___ } $code .= _end_func; # Do one step of the GHASH update of four vectors of data blocks. # $i: the step to do, 0 through 9 # $ghashdata_ptr: pointer to the data blocks (ciphertext or AAD) # $htable: pointer to the Htable for the key # $bswap_mask: mask for reflecting the bytes of blocks # $h_pow[2-1]_xored: XOR'd key powers cached from Htable # $tmp[0-2]: temporary registers. $tmp[1-2] must be preserved across steps. # $lo, $mi: working state for this macro that must be preserved across steps # $ghash_acc: the GHASH accumulator (input/output) sub _ghash_step_4x { my ( $i, $ghashdata_ptr, $htable, $bswap_mask, $h_pow2_xored, $h_pow1_xored, $tmp0, $tmp0_xmm, $tmp1, $tmp2, $lo, $mi, $ghash_acc, $ghash_acc_xmm ) = @_; my ( $hi, $hi_xmm ) = ( $ghash_acc, $ghash_acc_xmm ); # alias if ( $i == 0 ) { return <<___; # First vector vmovdqu 0*32($ghashdata_ptr), $tmp1 vpshufb $bswap_mask, $tmp1, $tmp1 vmovdqu 0*32($htable), $tmp2 vpxor $ghash_acc, $tmp1, $tmp1 vpclmulqdq \$0x00, $tmp2, $tmp1, $lo vpclmulqdq \$0x11, $tmp2, $tmp1, $hi vpunpckhqdq $tmp1, $tmp1, $tmp0 vpxor $tmp1, $tmp0, $tmp0 vpclmulqdq \$0x00, $h_pow2_xored, $tmp0, $mi ___ } elsif ( $i == 1 ) { return <<___; ___ } elsif ( $i == 2 ) { return <<___; # Second vector vmovdqu 1*32($ghashdata_ptr), $tmp1 vpshufb $bswap_mask, $tmp1, $tmp1 vmovdqu 1*32($htable), $tmp2 vpclmulqdq \$0x00, $tmp2, $tmp1, $tmp0 vpxor $tmp0, $lo, $lo vpclmulqdq \$0x11, $tmp2, $tmp1, $tmp0 vpxor $tmp0, $hi, $hi vpunpckhqdq $tmp1, $tmp1, $tmp0 vpxor $tmp1, $tmp0, $tmp0 vpclmulqdq \$0x10, $h_pow2_xored, $tmp0, $tmp0 vpxor $tmp0, $mi, $mi ___ } elsif ( $i == 3 ) { return <<___; # Third vector vmovdqu 2*32($ghashdata_ptr), $tmp1 vpshufb $bswap_mask, $tmp1, $tmp1 vmovdqu 2*32($htable), $tmp2 ___ } elsif ( $i == 4 ) { return <<___; vpclmulqdq \$0x00, $tmp2, $tmp1, $tmp0 vpxor $tmp0, $lo, $lo vpclmulqdq \$0x11, $tmp2, $tmp1, $tmp0 vpxor $tmp0, $hi, $hi ___ } elsif ( $i == 5 ) { return <<___; vpunpckhqdq $tmp1, $tmp1, $tmp0 vpxor $tmp1, $tmp0, $tmp0 vpclmulqdq \$0x00, $h_pow1_xored, $tmp0, $tmp0 vpxor $tmp0, $mi, $mi # Fourth vector vmovdqu 3*32($ghashdata_ptr), $tmp1 vpshufb $bswap_mask, $tmp1, $tmp1 ___ } elsif ( $i == 6 ) { return <<___; vmovdqu 3*32($htable), $tmp2 vpclmulqdq \$0x00, $tmp2, $tmp1, $tmp0 vpxor $tmp0, $lo, $lo vpclmulqdq \$0x11, $tmp2, $tmp1, $tmp0 vpxor $tmp0, $hi, $hi vpunpckhqdq $tmp1, $tmp1, $tmp0 vpxor $tmp1, $tmp0, $tmp0 vpclmulqdq \$0x10, $h_pow1_xored, $tmp0, $tmp0 vpxor $tmp0, $mi, $mi ___ } elsif ( $i == 7 ) { return <<___; # Finalize 'mi' following Karatsuba multiplication. vpxor $lo, $mi, $mi vpxor $hi, $mi, $mi # Fold lo into mi. vbroadcasti128 .Lgfpoly(%rip), $tmp2 vpclmulqdq \$0x01, $lo, $tmp2, $tmp0 vpshufd \$0x4e, $lo, $lo vpxor $lo, $mi, $mi vpxor $tmp0, $mi, $mi ___ } elsif ( $i == 8 ) { return <<___; # Fold mi into hi. vpclmulqdq \$0x01, $mi, $tmp2, $tmp0 vpshufd \$0x4e, $mi, $mi vpxor $mi, $hi, $hi vpxor $tmp0, $hi, $hi ___ } elsif ( $i == 9 ) { return <<___; vextracti128 \$1, $hi, $tmp0_xmm vpxor $tmp0_xmm, $hi_xmm, $ghash_acc_xmm ___ } } sub _ghash_4x { my $code = ""; for my $i ( 0 .. 9 ) { $code .= _ghash_step_4x $i, @_; } return $code; } # void gcm_gmult_vpclmulqdq_avx2(uint8_t Xi[16], const u128 Htable[16]); $code .= _begin_func "gcm_gmult_vpclmulqdq_avx2", 1; { my ( $GHASH_ACC_PTR, $HTABLE ) = @argregs[ 0 .. 1 ]; my ( $GHASH_ACC, $BSWAP_MASK, $H_POW1, $GFPOLY, $T0, $T1, $T2 ) = map( "%xmm$_", ( 0 .. 6 ) ); $code .= <<___; @{[ _save_xmmregs (6) ]} .seh_endprologue vmovdqu ($GHASH_ACC_PTR), $GHASH_ACC vmovdqu .Lbswap_mask(%rip), $BSWAP_MASK vmovdqu $OFFSETOFEND_H_POWERS-16($HTABLE), $H_POW1 vmovdqu .Lgfpoly(%rip), $GFPOLY vpshufb $BSWAP_MASK, $GHASH_ACC, $GHASH_ACC @{[ _ghash_mul $H_POW1, $GHASH_ACC, $GHASH_ACC, $GFPOLY, $T0, $T1, $T2 ]} vpshufb $BSWAP_MASK, $GHASH_ACC, $GHASH_ACC vmovdqu $GHASH_ACC, ($GHASH_ACC_PTR) # No need for vzeroupper, since only xmm registers were used. ___ } $code .= _end_func; # void gcm_ghash_vpclmulqdq_avx2(uint8_t Xi[16], const u128 Htable[16], # const uint8_t *in, size_t len); # # Using the key |Htable|, update the GHASH accumulator |Xi| with the data given # by |in| and |len|. |len| must be a multiple of 16. # # This function handles large amounts of AAD efficiently, while also keeping the # overhead low for small amounts of AAD which is the common case. TLS uses less # than one block of AAD, but (uncommonly) other use cases may use much more. $code .= _begin_func "gcm_ghash_vpclmulqdq_avx2", 1; { # Function arguments my ( $GHASH_ACC_PTR, $HTABLE, $AAD, $AADLEN ) = @argregs[ 0 .. 3 ]; # Additional local variables my ( $TMP0, $TMP0_XMM ) = ( "%ymm0", "%xmm0" ); my ( $TMP1, $TMP1_XMM ) = ( "%ymm1", "%xmm1" ); my ( $TMP2, $TMP2_XMM ) = ( "%ymm2", "%xmm2" ); my ( $LO, $LO_XMM ) = ( "%ymm3", "%xmm3" ); my ( $MI, $MI_XMM ) = ( "%ymm4", "%xmm4" ); my ( $GHASH_ACC, $GHASH_ACC_XMM ) = ( "%ymm5", "%xmm5" ); my ( $BSWAP_MASK, $BSWAP_MASK_XMM ) = ( "%ymm6", "%xmm6" ); my ( $GFPOLY, $GFPOLY_XMM ) = ( "%ymm7", "%xmm7" ); my $H_POW2_XORED = "%ymm8"; my $H_POW1_XORED = "%ymm9"; $code .= <<___; @{[ _save_xmmregs (6 .. 9) ]} .seh_endprologue # Load the bswap_mask and gfpoly constants. Since AADLEN is usually small, # usually only 128-bit vectors will be used. So as an optimization, don't # broadcast these constants to both 128-bit lanes quite yet. vmovdqu .Lbswap_mask(%rip), $BSWAP_MASK_XMM vmovdqu .Lgfpoly(%rip), $GFPOLY_XMM # Load the GHASH accumulator. vmovdqu ($GHASH_ACC_PTR), $GHASH_ACC_XMM vpshufb $BSWAP_MASK_XMM, $GHASH_ACC_XMM, $GHASH_ACC_XMM # Optimize for AADLEN < 32 by checking for AADLEN < 32 before AADLEN < 128. cmp \$32, $AADLEN jb .Lghash_lastblock # AADLEN >= 32, so we'll operate on full vectors. Broadcast bswap_mask and # gfpoly to both 128-bit lanes. vinserti128 \$1, $BSWAP_MASK_XMM, $BSWAP_MASK, $BSWAP_MASK vinserti128 \$1, $GFPOLY_XMM, $GFPOLY, $GFPOLY cmp \$127, $AADLEN jbe .Lghash_loop_1x # Update GHASH with 128 bytes of AAD at a time. vmovdqu $OFFSETOF_H_POWERS_XORED($HTABLE), $H_POW2_XORED vmovdqu $OFFSETOF_H_POWERS_XORED+32($HTABLE), $H_POW1_XORED .Lghash_loop_4x: @{[ _ghash_4x $AAD, $HTABLE, $BSWAP_MASK, $H_POW2_XORED, $H_POW1_XORED, $TMP0, $TMP0_XMM, $TMP1, $TMP2, $LO, $MI, $GHASH_ACC, $GHASH_ACC_XMM ]} sub \$-128, $AAD # 128 is 4 bytes, -128 is 1 byte add \$-128, $AADLEN cmp \$127, $AADLEN ja .Lghash_loop_4x # Update GHASH with 32 bytes of AAD at a time. cmp \$32, $AADLEN jb .Lghash_loop_1x_done .Lghash_loop_1x: vmovdqu ($AAD), $TMP0 vpshufb $BSWAP_MASK, $TMP0, $TMP0 vpxor $TMP0, $GHASH_ACC, $GHASH_ACC vmovdqu $OFFSETOFEND_H_POWERS-32($HTABLE), $TMP0 @{[ _ghash_mul $TMP0, $GHASH_ACC, $GHASH_ACC, $GFPOLY, $TMP1, $TMP2, $LO ]} vextracti128 \$1, $GHASH_ACC, $TMP0_XMM vpxor $TMP0_XMM, $GHASH_ACC_XMM, $GHASH_ACC_XMM add \$32, $AAD sub \$32, $AADLEN cmp \$32, $AADLEN jae .Lghash_loop_1x .Lghash_loop_1x_done: # Update GHASH with the remaining 16-byte block if any. .Lghash_lastblock: test $AADLEN, $AADLEN jz .Lghash_done vmovdqu ($AAD), $TMP0_XMM vpshufb $BSWAP_MASK_XMM, $TMP0_XMM, $TMP0_XMM vpxor $TMP0_XMM, $GHASH_ACC_XMM, $GHASH_ACC_XMM vmovdqu $OFFSETOFEND_H_POWERS-16($HTABLE), $TMP0_XMM @{[ _ghash_mul $TMP0_XMM, $GHASH_ACC_XMM, $GHASH_ACC_XMM, $GFPOLY_XMM, $TMP1_XMM, $TMP2_XMM, $LO_XMM ]} .Lghash_done: # Store the updated GHASH accumulator back to memory. vpshufb $BSWAP_MASK_XMM, $GHASH_ACC_XMM, $GHASH_ACC_XMM vmovdqu $GHASH_ACC_XMM, ($GHASH_ACC_PTR) vzeroupper ___ } $code .= _end_func; sub _vaesenc_4x { my ( $round_key, $aesdata0, $aesdata1, $aesdata2, $aesdata3 ) = @_; return <<___; vaesenc $round_key, $aesdata0, $aesdata0 vaesenc $round_key, $aesdata1, $aesdata1 vaesenc $round_key, $aesdata2, $aesdata2 vaesenc $round_key, $aesdata3, $aesdata3 ___ } sub _ctr_begin_4x { my ( $le_ctr, $bswap_mask, $rndkey0, $aesdata0, $aesdata1, $aesdata2, $aesdata3, $tmp ) = @_; return <<___; # Increment le_ctr four times to generate four vectors of little-endian # counter blocks, swap each to big-endian, and store them in aesdata[0-3]. vmovdqu .Linc_2blocks(%rip), $tmp vpshufb $bswap_mask, $le_ctr, $aesdata0 vpaddd $tmp, $le_ctr, $le_ctr vpshufb $bswap_mask, $le_ctr, $aesdata1 vpaddd $tmp, $le_ctr, $le_ctr vpshufb $bswap_mask, $le_ctr, $aesdata2 vpaddd $tmp, $le_ctr, $le_ctr vpshufb $bswap_mask, $le_ctr, $aesdata3 vpaddd $tmp, $le_ctr, $le_ctr # AES "round zero": XOR in the zero-th round key. vpxor $rndkey0, $aesdata0, $aesdata0 vpxor $rndkey0, $aesdata1, $aesdata1 vpxor $rndkey0, $aesdata2, $aesdata2 vpxor $rndkey0, $aesdata3, $aesdata3 ___ } # Do the last AES round for four vectors of counter blocks, XOR four vectors of # source data with the resulting keystream blocks, and write the result to the # destination buffer. The implementation differs slightly as it takes advantage # of the property vaesenclast(key, a) ^ b == vaesenclast(key ^ b, a) to reduce # latency, but it has the same effect. sub _aesenclast_and_xor_4x { my ( $src, $dst, $rndkeylast, $aesdata0, $aesdata1, $aesdata2, $aesdata3, $t0, $t1, $t2, $t3 ) = @_; return <<___; vpxor 0*32($src), $rndkeylast, $t0 vpxor 1*32($src), $rndkeylast, $t1 vpxor 2*32($src), $rndkeylast, $t2 vpxor 3*32($src), $rndkeylast, $t3 vaesenclast $t0, $aesdata0, $aesdata0 vaesenclast $t1, $aesdata1, $aesdata1 vaesenclast $t2, $aesdata2, $aesdata2 vaesenclast $t3, $aesdata3, $aesdata3 vmovdqu $aesdata0, 0*32($dst) vmovdqu $aesdata1, 1*32($dst) vmovdqu $aesdata2, 2*32($dst) vmovdqu $aesdata3, 3*32($dst) ___ } my $g_update_macro_expansion_count = 0; # void aes_gcm_{enc,dec}_update_vaes_avx2(const uint8_t *in, uint8_t *out, # size_t len, const AES_KEY *key, # const uint8_t ivec[16], # const u128 Htable[16], # uint8_t Xi[16]); # # This macro generates a GCM encryption or decryption update function with the # above prototype (with \enc selecting which one). The function computes the # next portion of the CTR keystream, XOR's it with |len| bytes from |in|, and # writes the resulting encrypted or decrypted data to |out|. It also updates # the GHASH accumulator |Xi| using the next |len| ciphertext bytes. # # |len| must be a multiple of 16. The caller must do any buffering needed to # ensure this. Both in-place and out-of-place en/decryption are supported. # # |ivec| must give the current counter in big-endian format. This function # loads the counter from |ivec| and increments the loaded counter as needed, but # it does *not* store the updated counter back to |ivec|. The caller must # update |ivec| if any more data segments follow. Internally, only the low # 32-bit word of the counter is incremented, following the GCM standard. sub _aes_gcm_update { my $local_label_suffix = "__func" . ++$g_update_macro_expansion_count; my ($enc) = @_; my $code = ""; # Function arguments my ( $SRC, $DST, $DATALEN, $AESKEY, $BE_CTR_PTR, $HTABLE, $GHASH_ACC_PTR ) = $win64 ? ( @argregs[ 0 .. 3 ], "%rsi", "%rdi", "%r12" ) : ( @argregs[ 0 .. 5 ], "%r12" ); # Additional local variables. # %rax is used as a temporary register. BE_CTR_PTR is also available as a # temporary register after the counter is loaded. # AES key length in bytes my ( $AESKEYLEN, $AESKEYLEN64 ) = ( "%r10d", "%r10" ); # Pointer to the last AES round key for the chosen AES variant my $RNDKEYLAST_PTR = "%r11"; # BSWAP_MASK is the shuffle mask for byte-reflecting 128-bit values # using vpshufb, copied to all 128-bit lanes. my ( $BSWAP_MASK, $BSWAP_MASK_XMM ) = ( "%ymm0", "%xmm0" ); # GHASH_ACC is the accumulator variable for GHASH. When fully reduced, # only the lowest 128-bit lane can be nonzero. When not fully reduced, # more than one lane may be used, and they need to be XOR'd together. my ( $GHASH_ACC, $GHASH_ACC_XMM ) = ( "%ymm1", "%xmm1" ); # TMP[0-2] are temporary registers. my ( $TMP0, $TMP0_XMM ) = ( "%ymm2", "%xmm2" ); my ( $TMP1, $TMP1_XMM ) = ( "%ymm3", "%xmm3" ); my ( $TMP2, $TMP2_XMM ) = ( "%ymm4", "%xmm4" ); # LO and MI are used to accumulate unreduced GHASH products. my ( $LO, $LO_XMM ) = ( "%ymm5", "%xmm5" ); my ( $MI, $MI_XMM ) = ( "%ymm6", "%xmm6" ); # Cached key powers from Htable my ( $H_POW2_XORED, $H_POW2_XORED_XMM ) = ( "%ymm7", "%xmm7" ); my ( $H_POW1_XORED, $H_POW1_XORED_XMM ) = ( "%ymm8", "%xmm8" ); # RNDKEY0 caches the zero-th round key, and RNDKEYLAST the last one. my $RNDKEY0 = "%ymm9"; my $RNDKEYLAST = "%ymm10"; # LE_CTR contains the next set of little-endian counter blocks. my $LE_CTR = "%ymm11"; # AESDATA[0-3] hold the counter blocks that are being encrypted by AES. my ( $AESDATA0, $AESDATA0_XMM ) = ( "%ymm12", "%xmm12" ); my ( $AESDATA1, $AESDATA1_XMM ) = ( "%ymm13", "%xmm13" ); my ( $AESDATA2, $AESDATA2_XMM ) = ( "%ymm14", "%xmm14" ); my ( $AESDATA3, $AESDATA3_XMM ) = ( "%ymm15", "%xmm15" ); my @AESDATA = ( $AESDATA0, $AESDATA1, $AESDATA2, $AESDATA3 ); my @ghash_4x_args = ( $enc ? $DST : $SRC, $HTABLE, $BSWAP_MASK, $H_POW2_XORED, $H_POW1_XORED, $TMP0, $TMP0_XMM, $TMP1, $TMP2, $LO, $MI, $GHASH_ACC, $GHASH_ACC_XMM ); if ($win64) { $code .= <<___; @{[ _save_gpregs $BE_CTR_PTR, $HTABLE, $GHASH_ACC_PTR ]} mov 64(%rsp), $BE_CTR_PTR # arg5 mov 72(%rsp), $HTABLE # arg6 mov 80(%rsp), $GHASH_ACC_PTR # arg7 @{[ _save_xmmregs (6 .. 15) ]} .seh_endprologue ___ } else { $code .= <<___; @{[ _save_gpregs $GHASH_ACC_PTR ]} mov 16(%rsp), $GHASH_ACC_PTR # arg7 ___ } if ($enc) { $code .= <<___; #ifdef BORINGSSL_DISPATCH_TEST .extern BORINGSSL_function_hit movb \$1,BORINGSSL_function_hit+6(%rip) #endif ___ } $code .= <<___; vbroadcasti128 .Lbswap_mask(%rip), $BSWAP_MASK # Load the GHASH accumulator and the starting counter. # BoringSSL passes these values in big endian format. vmovdqu ($GHASH_ACC_PTR), $GHASH_ACC_XMM vpshufb $BSWAP_MASK_XMM, $GHASH_ACC_XMM, $GHASH_ACC_XMM vbroadcasti128 ($BE_CTR_PTR), $LE_CTR vpshufb $BSWAP_MASK, $LE_CTR, $LE_CTR # Load the AES key length in bytes. BoringSSL stores number of rounds # minus 1, so convert using: AESKEYLEN = 4 * aeskey->rounds - 20. movl $OFFSETOF_AES_ROUNDS($AESKEY), $AESKEYLEN lea -20(,$AESKEYLEN,4), $AESKEYLEN # Make RNDKEYLAST_PTR point to the last AES round key. This is the # round key with index 10, 12, or 14 for AES-128, AES-192, or AES-256 # respectively. Then load the zero-th and last round keys. lea 6*16($AESKEY,$AESKEYLEN64,4), $RNDKEYLAST_PTR vbroadcasti128 ($AESKEY), $RNDKEY0 vbroadcasti128 ($RNDKEYLAST_PTR), $RNDKEYLAST # Finish initializing LE_CTR by adding 1 to the second block. vpaddd .Lctr_pattern(%rip), $LE_CTR, $LE_CTR # If there are at least 128 bytes of data, then continue into the loop that # processes 128 bytes of data at a time. Otherwise skip it. cmp \$127, $DATALEN jbe .Lcrypt_loop_4x_done$local_label_suffix vmovdqu $OFFSETOF_H_POWERS_XORED($HTABLE), $H_POW2_XORED vmovdqu $OFFSETOF_H_POWERS_XORED+32($HTABLE), $H_POW1_XORED ___ # Main loop: en/decrypt and hash 4 vectors (128 bytes) at a time. if ($enc) { $code .= <<___; # Encrypt the first 4 vectors of plaintext blocks. @{[ _ctr_begin_4x $LE_CTR, $BSWAP_MASK, $RNDKEY0, @AESDATA, $TMP0 ]} lea 16($AESKEY), %rax .Lvaesenc_loop_first_4_vecs$local_label_suffix: vbroadcasti128 (%rax), $TMP0 @{[ _vaesenc_4x $TMP0, @AESDATA ]} add \$16, %rax cmp %rax, $RNDKEYLAST_PTR jne .Lvaesenc_loop_first_4_vecs$local_label_suffix @{[ _aesenclast_and_xor_4x $SRC, $DST, $RNDKEYLAST, @AESDATA, $TMP0, $TMP1, $LO, $MI ]} sub \$-128, $SRC # 128 is 4 bytes, -128 is 1 byte add \$-128, $DATALEN cmp \$127, $DATALEN jbe .Lghash_last_ciphertext_4x$local_label_suffix ___ } $code .= <<___; .align 16 .Lcrypt_loop_4x$local_label_suffix: # Start the AES encryption of the counter blocks. @{[ _ctr_begin_4x $LE_CTR, $BSWAP_MASK, $RNDKEY0, @AESDATA, $TMP0 ]} cmp \$24, $AESKEYLEN jl .Laes128$local_label_suffix je .Laes192$local_label_suffix # AES-256 vbroadcasti128 -13*16($RNDKEYLAST_PTR), $TMP0 @{[ _vaesenc_4x $TMP0, @AESDATA ]} vbroadcasti128 -12*16($RNDKEYLAST_PTR), $TMP0 @{[ _vaesenc_4x $TMP0, @AESDATA ]} .Laes192$local_label_suffix: vbroadcasti128 -11*16($RNDKEYLAST_PTR), $TMP0 @{[ _vaesenc_4x $TMP0, @AESDATA ]} vbroadcasti128 -10*16($RNDKEYLAST_PTR), $TMP0 @{[ _vaesenc_4x $TMP0, @AESDATA ]} .Laes128$local_label_suffix: ___ # Prefetch the source data 512 bytes ahead into the L1 data cache, to # improve performance when the hardware prefetcher is disabled. Assumes the # L1 data cache line size is 64 bytes (de facto standard on x86_64). $code .= "prefetcht0 512($SRC)\n"; $code .= "prefetcht0 512+64($SRC)\n"; # Finish the AES encryption of the counter blocks in AESDATA[0-3], # interleaved with the GHASH update of the ciphertext blocks. for my $i ( reverse 1 .. 9 ) { $code .= <<___; @{[ _ghash_step_4x 9-$i, @ghash_4x_args ]} vbroadcasti128 -$i*16($RNDKEYLAST_PTR), $TMP0 @{[ _vaesenc_4x $TMP0, @AESDATA ]} ___ } $code .= <<___; @{[ _ghash_step_4x 9, @ghash_4x_args ]} @{[ $enc ? "sub \$-128, $DST" : "" ]} # 128 is 4 bytes, -128 is 1 byte @{[ _aesenclast_and_xor_4x $SRC, $DST, $RNDKEYLAST, @AESDATA, $TMP0, $TMP1, $LO, $MI ]} sub \$-128, $SRC @{[ !$enc ? "sub \$-128, $DST" : "" ]} add \$-128, $DATALEN cmp \$127, $DATALEN ja .Lcrypt_loop_4x$local_label_suffix ___ if ($enc) { # Update GHASH with the last set of ciphertext blocks. $code .= <<___; .Lghash_last_ciphertext_4x$local_label_suffix: @{[ _ghash_4x @ghash_4x_args ]} sub \$-128, $DST ___ } my $POWERS_PTR = $BE_CTR_PTR; # BE_CTR_PTR is free to be reused. my ( $HI, $HI_XMM ) = ( $H_POW2_XORED, $H_POW2_XORED_XMM ); # reuse $code .= <<___; .Lcrypt_loop_4x_done$local_label_suffix: # Check whether any data remains. test $DATALEN, $DATALEN jz .Ldone$local_label_suffix # DATALEN is in [16, 32, 48, 64, 80, 96, 112]. # Make POWERS_PTR point to the key powers [H^N, H^(N-1), ...] where N # is the number of blocks that remain. lea $OFFSETOFEND_H_POWERS($HTABLE), $POWERS_PTR sub $DATALEN, $POWERS_PTR # Start collecting the unreduced GHASH intermediate value LO, MI, HI. vpxor $LO_XMM, $LO_XMM, $LO_XMM vpxor $MI_XMM, $MI_XMM, $MI_XMM vpxor $HI_XMM, $HI_XMM, $HI_XMM cmp \$64, $DATALEN jb .Llessthan64bytes$local_label_suffix # DATALEN is in [64, 80, 96, 112]. Encrypt two vectors of counter blocks. vpshufb $BSWAP_MASK, $LE_CTR, $AESDATA0 vpaddd .Linc_2blocks(%rip), $LE_CTR, $LE_CTR vpshufb $BSWAP_MASK, $LE_CTR, $AESDATA1 vpaddd .Linc_2blocks(%rip), $LE_CTR, $LE_CTR vpxor $RNDKEY0, $AESDATA0, $AESDATA0 vpxor $RNDKEY0, $AESDATA1, $AESDATA1 lea 16($AESKEY), %rax .Lvaesenc_loop_tail_1$local_label_suffix: vbroadcasti128 (%rax), $TMP0 vaesenc $TMP0, $AESDATA0, $AESDATA0 vaesenc $TMP0, $AESDATA1, $AESDATA1 add \$16, %rax cmp %rax, $RNDKEYLAST_PTR jne .Lvaesenc_loop_tail_1$local_label_suffix vaesenclast $RNDKEYLAST, $AESDATA0, $AESDATA0 vaesenclast $RNDKEYLAST, $AESDATA1, $AESDATA1 # XOR the data with the two vectors of keystream blocks. vmovdqu 0($SRC), $TMP0 vmovdqu 32($SRC), $TMP1 vpxor $TMP0, $AESDATA0, $AESDATA0 vpxor $TMP1, $AESDATA1, $AESDATA1 vmovdqu $AESDATA0, 0($DST) vmovdqu $AESDATA1, 32($DST) # Update GHASH with two vectors of ciphertext blocks, without reducing. vpshufb $BSWAP_MASK, @{[ $enc ? $AESDATA0 : $TMP0 ]}, $AESDATA0 vpshufb $BSWAP_MASK, @{[ $enc ? $AESDATA1 : $TMP1 ]}, $AESDATA1 vpxor $GHASH_ACC, $AESDATA0, $AESDATA0 vmovdqu ($POWERS_PTR), $TMP0 vmovdqu 32($POWERS_PTR), $TMP1 vpclmulqdq \$0x00, $TMP0, $AESDATA0, $LO vpclmulqdq \$0x01, $TMP0, $AESDATA0, $MI vpclmulqdq \$0x10, $TMP0, $AESDATA0, $TMP2 vpxor $TMP2, $MI, $MI vpclmulqdq \$0x11, $TMP0, $AESDATA0, $HI vpclmulqdq \$0x00, $TMP1, $AESDATA1, $TMP2 vpxor $TMP2, $LO, $LO vpclmulqdq \$0x01, $TMP1, $AESDATA1, $TMP2 vpxor $TMP2, $MI, $MI vpclmulqdq \$0x10, $TMP1, $AESDATA1, $TMP2 vpxor $TMP2, $MI, $MI vpclmulqdq \$0x11, $TMP1, $AESDATA1, $TMP2 vpxor $TMP2, $HI, $HI add \$64, $POWERS_PTR add \$64, $SRC add \$64, $DST sub \$64, $DATALEN jz .Lreduce$local_label_suffix vpxor $GHASH_ACC_XMM, $GHASH_ACC_XMM, $GHASH_ACC_XMM # DATALEN is in [16, 32, 48]. Encrypt two last vectors of counter blocks. .Llessthan64bytes$local_label_suffix: vpshufb $BSWAP_MASK, $LE_CTR, $AESDATA0 vpaddd .Linc_2blocks(%rip), $LE_CTR, $LE_CTR vpshufb $BSWAP_MASK, $LE_CTR, $AESDATA1 vpxor $RNDKEY0, $AESDATA0, $AESDATA0 vpxor $RNDKEY0, $AESDATA1, $AESDATA1 lea 16($AESKEY), %rax .Lvaesenc_loop_tail_2$local_label_suffix: vbroadcasti128 (%rax), $TMP0 vaesenc $TMP0, $AESDATA0, $AESDATA0 vaesenc $TMP0, $AESDATA1, $AESDATA1 add \$16, %rax cmp %rax, $RNDKEYLAST_PTR jne .Lvaesenc_loop_tail_2$local_label_suffix vaesenclast $RNDKEYLAST, $AESDATA0, $AESDATA0 vaesenclast $RNDKEYLAST, $AESDATA1, $AESDATA1 # XOR the remaining data with the keystream blocks, and update GHASH with # the remaining ciphertext blocks without reducing. cmp \$32, $DATALEN jb .Lxor_one_block$local_label_suffix je .Lxor_two_blocks$local_label_suffix .Lxor_three_blocks$local_label_suffix: vmovdqu 0($SRC), $TMP0 vmovdqu 32($SRC), $TMP1_XMM vpxor $TMP0, $AESDATA0, $AESDATA0 vpxor $TMP1_XMM, $AESDATA1_XMM, $AESDATA1_XMM vmovdqu $AESDATA0, 0($DST) vmovdqu $AESDATA1_XMM, 32($DST) vpshufb $BSWAP_MASK, @{[ $enc ? $AESDATA0 : $TMP0 ]}, $AESDATA0 vpshufb $BSWAP_MASK_XMM, @{[ $enc ? $AESDATA1_XMM : $TMP1_XMM ]}, $AESDATA1_XMM vpxor $GHASH_ACC, $AESDATA0, $AESDATA0 vmovdqu ($POWERS_PTR), $TMP0 vmovdqu 32($POWERS_PTR), $TMP1_XMM vpclmulqdq \$0x00, $TMP1_XMM, $AESDATA1_XMM, $TMP2_XMM vpxor $TMP2, $LO, $LO vpclmulqdq \$0x01, $TMP1_XMM, $AESDATA1_XMM, $TMP2_XMM vpxor $TMP2, $MI, $MI vpclmulqdq \$0x10, $TMP1_XMM, $AESDATA1_XMM, $TMP2_XMM vpxor $TMP2, $MI, $MI vpclmulqdq \$0x11, $TMP1_XMM, $AESDATA1_XMM, $TMP2_XMM vpxor $TMP2, $HI, $HI jmp .Lghash_mul_one_vec_unreduced$local_label_suffix .Lxor_two_blocks$local_label_suffix: vmovdqu ($SRC), $TMP0 vpxor $TMP0, $AESDATA0, $AESDATA0 vmovdqu $AESDATA0, ($DST) vpshufb $BSWAP_MASK, @{[ $enc ? $AESDATA0 : $TMP0 ]}, $AESDATA0 vpxor $GHASH_ACC, $AESDATA0, $AESDATA0 vmovdqu ($POWERS_PTR), $TMP0 jmp .Lghash_mul_one_vec_unreduced$local_label_suffix .Lxor_one_block$local_label_suffix: vmovdqu ($SRC), $TMP0_XMM vpxor $TMP0_XMM, $AESDATA0_XMM, $AESDATA0_XMM vmovdqu $AESDATA0_XMM, ($DST) vpshufb $BSWAP_MASK_XMM, @{[ $enc ? $AESDATA0_XMM : $TMP0_XMM ]}, $AESDATA0_XMM vpxor $GHASH_ACC_XMM, $AESDATA0_XMM, $AESDATA0_XMM vmovdqu ($POWERS_PTR), $TMP0_XMM .Lghash_mul_one_vec_unreduced$local_label_suffix: vpclmulqdq \$0x00, $TMP0, $AESDATA0, $TMP2 vpxor $TMP2, $LO, $LO vpclmulqdq \$0x01, $TMP0, $AESDATA0, $TMP2 vpxor $TMP2, $MI, $MI vpclmulqdq \$0x10, $TMP0, $AESDATA0, $TMP2 vpxor $TMP2, $MI, $MI vpclmulqdq \$0x11, $TMP0, $AESDATA0, $TMP2 vpxor $TMP2, $HI, $HI .Lreduce$local_label_suffix: # Finally, do the GHASH reduction. vbroadcasti128 .Lgfpoly(%rip), $TMP0 vpclmulqdq \$0x01, $LO, $TMP0, $TMP1 vpshufd \$0x4e, $LO, $LO vpxor $LO, $MI, $MI vpxor $TMP1, $MI, $MI vpclmulqdq \$0x01, $MI, $TMP0, $TMP1 vpshufd \$0x4e, $MI, $MI vpxor $MI, $HI, $HI vpxor $TMP1, $HI, $HI vextracti128 \$1, $HI, $GHASH_ACC_XMM vpxor $HI_XMM, $GHASH_ACC_XMM, $GHASH_ACC_XMM .Ldone$local_label_suffix: # Store the updated GHASH accumulator back to memory. vpshufb $BSWAP_MASK_XMM, $GHASH_ACC_XMM, $GHASH_ACC_XMM vmovdqu $GHASH_ACC_XMM, ($GHASH_ACC_PTR) vzeroupper ___ return $code; } $code .= _begin_func "aes_gcm_enc_update_vaes_avx2", 1; $code .= _aes_gcm_update 1; $code .= _end_func; $code .= _begin_func "aes_gcm_dec_update_vaes_avx2", 1; $code .= _aes_gcm_update 0; $code .= _end_func; print $code; close STDOUT or die "error closing STDOUT: $!"; exit 0;