8280daad43
Patch adds 3-way parallel x86_64 assembly implementation of twofish as new module. New assembler functions crypt data in three blocks chunks, improving cipher performance on out-of-order CPUs. Patch has been tested with tcrypt and automated filesystem tests. Summary of the tcrypt benchmarks: Twofish 3-way-asm vs twofish asm (128bit 8kb block ECB) encrypt: 1.3x speed decrypt: 1.3x speed Twofish 3-way-asm vs twofish asm (128bit 8kb block CBC) encrypt: 1.07x speed decrypt: 1.4x speed Twofish 3-way-asm vs twofish asm (128bit 8kb block CTR) encrypt: 1.4x speed Twofish 3-way-asm vs AES asm (128bit 8kb block ECB) encrypt: 1.0x speed decrypt: 1.0x speed Twofish 3-way-asm vs AES asm (128bit 8kb block CBC) encrypt: 0.84x speed decrypt: 1.09x speed Twofish 3-way-asm vs AES asm (128bit 8kb block CTR) encrypt: 1.15x speed Full output: http://koti.mbnet.fi/axh/kernel/crypto/tcrypt-speed-twofish-3way-asm-x86_64.txt http://koti.mbnet.fi/axh/kernel/crypto/tcrypt-speed-twofish-asm-x86_64.txt http://koti.mbnet.fi/axh/kernel/crypto/tcrypt-speed-aes-asm-x86_64.txt Tests were run on: vendor_id : AuthenticAMD cpu family : 16 model : 10 model name : AMD Phenom(tm) II X6 1055T Processor Also userspace test were run on: vendor_id : GenuineIntel cpu family : 6 model : 15 model name : Intel(R) Xeon(R) CPU E7330 @ 2.40GHz stepping : 11 Userspace test results: Encryption/decryption of twofish 3-way vs x86_64-asm on AMD Phenom II: encrypt: 1.27x decrypt: 1.25x Encryption/decryption of twofish 3-way vs x86_64-asm on Intel Xeon E7330: encrypt: 1.36x decrypt: 1.36x Signed-off-by: Jussi Kivilinna <jussi.kivilinna@mbnet.fi> Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
316 lines
6.9 KiB
ArmAsm
316 lines
6.9 KiB
ArmAsm
/*
|
|
* Twofish Cipher 3-way parallel algorithm (x86_64)
|
|
*
|
|
* Copyright (C) 2011 Jussi Kivilinna <jussi.kivilinna@mbnet.fi>
|
|
*
|
|
* 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.
|
|
*
|
|
* This program is distributed in the hope that it will be useful,
|
|
* but WITHOUT ANY WARRANTY; without even the implied warranty of
|
|
* MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
|
|
* GNU General Public License for more details.
|
|
*
|
|
* You should have received a copy of the GNU General Public License
|
|
* along with this program; if not, write to the Free Software
|
|
* Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307
|
|
* USA
|
|
*
|
|
*/
|
|
|
|
.file "twofish-x86_64-asm-3way.S"
|
|
.text
|
|
|
|
/* structure of crypto context */
|
|
#define s0 0
|
|
#define s1 1024
|
|
#define s2 2048
|
|
#define s3 3072
|
|
#define w 4096
|
|
#define k 4128
|
|
|
|
/**********************************************************************
|
|
3-way twofish
|
|
**********************************************************************/
|
|
#define CTX %rdi
|
|
#define RIO %rdx
|
|
|
|
#define RAB0 %rax
|
|
#define RAB1 %rbx
|
|
#define RAB2 %rcx
|
|
|
|
#define RAB0d %eax
|
|
#define RAB1d %ebx
|
|
#define RAB2d %ecx
|
|
|
|
#define RAB0bh %ah
|
|
#define RAB1bh %bh
|
|
#define RAB2bh %ch
|
|
|
|
#define RAB0bl %al
|
|
#define RAB1bl %bl
|
|
#define RAB2bl %cl
|
|
|
|
#define RCD0 %r8
|
|
#define RCD1 %r9
|
|
#define RCD2 %r10
|
|
|
|
#define RCD0d %r8d
|
|
#define RCD1d %r9d
|
|
#define RCD2d %r10d
|
|
|
|
#define RX0 %rbp
|
|
#define RX1 %r11
|
|
#define RX2 %r12
|
|
|
|
#define RX0d %ebp
|
|
#define RX1d %r11d
|
|
#define RX2d %r12d
|
|
|
|
#define RY0 %r13
|
|
#define RY1 %r14
|
|
#define RY2 %r15
|
|
|
|
#define RY0d %r13d
|
|
#define RY1d %r14d
|
|
#define RY2d %r15d
|
|
|
|
#define RT0 %rdx
|
|
#define RT1 %rsi
|
|
|
|
#define RT0d %edx
|
|
#define RT1d %esi
|
|
|
|
#define do16bit_ror(rot, op1, op2, T0, T1, tmp1, tmp2, ab, dst) \
|
|
movzbl ab ## bl, tmp2 ## d; \
|
|
movzbl ab ## bh, tmp1 ## d; \
|
|
rorq $(rot), ab; \
|
|
op1##l T0(CTX, tmp2, 4), dst ## d; \
|
|
op2##l T1(CTX, tmp1, 4), dst ## d;
|
|
|
|
/*
|
|
* Combined G1 & G2 function. Reordered with help of rotates to have moves
|
|
* at begining.
|
|
*/
|
|
#define g1g2_3(ab, cd, Tx0, Tx1, Tx2, Tx3, Ty0, Ty1, Ty2, Ty3, x, y) \
|
|
/* G1,1 && G2,1 */ \
|
|
do16bit_ror(32, mov, xor, Tx0, Tx1, RT0, x ## 0, ab ## 0, x ## 0); \
|
|
do16bit_ror(48, mov, xor, Ty1, Ty2, RT0, y ## 0, ab ## 0, y ## 0); \
|
|
\
|
|
do16bit_ror(32, mov, xor, Tx0, Tx1, RT0, x ## 1, ab ## 1, x ## 1); \
|
|
do16bit_ror(48, mov, xor, Ty1, Ty2, RT0, y ## 1, ab ## 1, y ## 1); \
|
|
\
|
|
do16bit_ror(32, mov, xor, Tx0, Tx1, RT0, x ## 2, ab ## 2, x ## 2); \
|
|
do16bit_ror(48, mov, xor, Ty1, Ty2, RT0, y ## 2, ab ## 2, y ## 2); \
|
|
\
|
|
/* G1,2 && G2,2 */ \
|
|
do16bit_ror(32, xor, xor, Tx2, Tx3, RT0, RT1, ab ## 0, x ## 0); \
|
|
do16bit_ror(16, xor, xor, Ty3, Ty0, RT0, RT1, ab ## 0, y ## 0); \
|
|
xchgq cd ## 0, ab ## 0; \
|
|
\
|
|
do16bit_ror(32, xor, xor, Tx2, Tx3, RT0, RT1, ab ## 1, x ## 1); \
|
|
do16bit_ror(16, xor, xor, Ty3, Ty0, RT0, RT1, ab ## 1, y ## 1); \
|
|
xchgq cd ## 1, ab ## 1; \
|
|
\
|
|
do16bit_ror(32, xor, xor, Tx2, Tx3, RT0, RT1, ab ## 2, x ## 2); \
|
|
do16bit_ror(16, xor, xor, Ty3, Ty0, RT0, RT1, ab ## 2, y ## 2); \
|
|
xchgq cd ## 2, ab ## 2;
|
|
|
|
#define enc_round_end(ab, x, y, n) \
|
|
addl y ## d, x ## d; \
|
|
addl x ## d, y ## d; \
|
|
addl k+4*(2*(n))(CTX), x ## d; \
|
|
xorl ab ## d, x ## d; \
|
|
addl k+4*(2*(n)+1)(CTX), y ## d; \
|
|
shrq $32, ab; \
|
|
roll $1, ab ## d; \
|
|
xorl y ## d, ab ## d; \
|
|
shlq $32, ab; \
|
|
rorl $1, x ## d; \
|
|
orq x, ab;
|
|
|
|
#define dec_round_end(ba, x, y, n) \
|
|
addl y ## d, x ## d; \
|
|
addl x ## d, y ## d; \
|
|
addl k+4*(2*(n))(CTX), x ## d; \
|
|
addl k+4*(2*(n)+1)(CTX), y ## d; \
|
|
xorl ba ## d, y ## d; \
|
|
shrq $32, ba; \
|
|
roll $1, ba ## d; \
|
|
xorl x ## d, ba ## d; \
|
|
shlq $32, ba; \
|
|
rorl $1, y ## d; \
|
|
orq y, ba;
|
|
|
|
#define encrypt_round3(ab, cd, n) \
|
|
g1g2_3(ab, cd, s0, s1, s2, s3, s0, s1, s2, s3, RX, RY); \
|
|
\
|
|
enc_round_end(ab ## 0, RX0, RY0, n); \
|
|
enc_round_end(ab ## 1, RX1, RY1, n); \
|
|
enc_round_end(ab ## 2, RX2, RY2, n);
|
|
|
|
#define decrypt_round3(ba, dc, n) \
|
|
g1g2_3(ba, dc, s1, s2, s3, s0, s3, s0, s1, s2, RY, RX); \
|
|
\
|
|
dec_round_end(ba ## 0, RX0, RY0, n); \
|
|
dec_round_end(ba ## 1, RX1, RY1, n); \
|
|
dec_round_end(ba ## 2, RX2, RY2, n);
|
|
|
|
#define encrypt_cycle3(ab, cd, n) \
|
|
encrypt_round3(ab, cd, n*2); \
|
|
encrypt_round3(ab, cd, (n*2)+1);
|
|
|
|
#define decrypt_cycle3(ba, dc, n) \
|
|
decrypt_round3(ba, dc, (n*2)+1); \
|
|
decrypt_round3(ba, dc, (n*2));
|
|
|
|
#define inpack3(in, n, xy, m) \
|
|
movq 4*(n)(in), xy ## 0; \
|
|
xorq w+4*m(CTX), xy ## 0; \
|
|
\
|
|
movq 4*(4+(n))(in), xy ## 1; \
|
|
xorq w+4*m(CTX), xy ## 1; \
|
|
\
|
|
movq 4*(8+(n))(in), xy ## 2; \
|
|
xorq w+4*m(CTX), xy ## 2;
|
|
|
|
#define outunpack3(op, out, n, xy, m) \
|
|
xorq w+4*m(CTX), xy ## 0; \
|
|
op ## q xy ## 0, 4*(n)(out); \
|
|
\
|
|
xorq w+4*m(CTX), xy ## 1; \
|
|
op ## q xy ## 1, 4*(4+(n))(out); \
|
|
\
|
|
xorq w+4*m(CTX), xy ## 2; \
|
|
op ## q xy ## 2, 4*(8+(n))(out);
|
|
|
|
#define inpack_enc3() \
|
|
inpack3(RIO, 0, RAB, 0); \
|
|
inpack3(RIO, 2, RCD, 2);
|
|
|
|
#define outunpack_enc3(op) \
|
|
outunpack3(op, RIO, 2, RAB, 6); \
|
|
outunpack3(op, RIO, 0, RCD, 4);
|
|
|
|
#define inpack_dec3() \
|
|
inpack3(RIO, 0, RAB, 4); \
|
|
rorq $32, RAB0; \
|
|
rorq $32, RAB1; \
|
|
rorq $32, RAB2; \
|
|
inpack3(RIO, 2, RCD, 6); \
|
|
rorq $32, RCD0; \
|
|
rorq $32, RCD1; \
|
|
rorq $32, RCD2;
|
|
|
|
#define outunpack_dec3() \
|
|
rorq $32, RCD0; \
|
|
rorq $32, RCD1; \
|
|
rorq $32, RCD2; \
|
|
outunpack3(mov, RIO, 0, RCD, 0); \
|
|
rorq $32, RAB0; \
|
|
rorq $32, RAB1; \
|
|
rorq $32, RAB2; \
|
|
outunpack3(mov, RIO, 2, RAB, 2);
|
|
|
|
.align 8
|
|
.global __twofish_enc_blk_3way
|
|
.type __twofish_enc_blk_3way,@function;
|
|
|
|
__twofish_enc_blk_3way:
|
|
/* input:
|
|
* %rdi: ctx, CTX
|
|
* %rsi: dst
|
|
* %rdx: src, RIO
|
|
* %rcx: bool, if true: xor output
|
|
*/
|
|
pushq %r15;
|
|
pushq %r14;
|
|
pushq %r13;
|
|
pushq %r12;
|
|
pushq %rbp;
|
|
pushq %rbx;
|
|
|
|
pushq %rcx; /* bool xor */
|
|
pushq %rsi; /* dst */
|
|
|
|
inpack_enc3();
|
|
|
|
encrypt_cycle3(RAB, RCD, 0);
|
|
encrypt_cycle3(RAB, RCD, 1);
|
|
encrypt_cycle3(RAB, RCD, 2);
|
|
encrypt_cycle3(RAB, RCD, 3);
|
|
encrypt_cycle3(RAB, RCD, 4);
|
|
encrypt_cycle3(RAB, RCD, 5);
|
|
encrypt_cycle3(RAB, RCD, 6);
|
|
encrypt_cycle3(RAB, RCD, 7);
|
|
|
|
popq RIO; /* dst */
|
|
popq %rbp; /* bool xor */
|
|
|
|
testb %bpl, %bpl;
|
|
jnz __enc_xor3;
|
|
|
|
outunpack_enc3(mov);
|
|
|
|
popq %rbx;
|
|
popq %rbp;
|
|
popq %r12;
|
|
popq %r13;
|
|
popq %r14;
|
|
popq %r15;
|
|
ret;
|
|
|
|
__enc_xor3:
|
|
outunpack_enc3(xor);
|
|
|
|
popq %rbx;
|
|
popq %rbp;
|
|
popq %r12;
|
|
popq %r13;
|
|
popq %r14;
|
|
popq %r15;
|
|
ret;
|
|
|
|
.global twofish_dec_blk_3way
|
|
.type twofish_dec_blk_3way,@function;
|
|
|
|
twofish_dec_blk_3way:
|
|
/* input:
|
|
* %rdi: ctx, CTX
|
|
* %rsi: dst
|
|
* %rdx: src, RIO
|
|
*/
|
|
pushq %r15;
|
|
pushq %r14;
|
|
pushq %r13;
|
|
pushq %r12;
|
|
pushq %rbp;
|
|
pushq %rbx;
|
|
|
|
pushq %rsi; /* dst */
|
|
|
|
inpack_dec3();
|
|
|
|
decrypt_cycle3(RAB, RCD, 7);
|
|
decrypt_cycle3(RAB, RCD, 6);
|
|
decrypt_cycle3(RAB, RCD, 5);
|
|
decrypt_cycle3(RAB, RCD, 4);
|
|
decrypt_cycle3(RAB, RCD, 3);
|
|
decrypt_cycle3(RAB, RCD, 2);
|
|
decrypt_cycle3(RAB, RCD, 1);
|
|
decrypt_cycle3(RAB, RCD, 0);
|
|
|
|
popq RIO; /* dst */
|
|
|
|
outunpack_dec3();
|
|
|
|
popq %rbx;
|
|
popq %rbp;
|
|
popq %r12;
|
|
popq %r13;
|
|
popq %r14;
|
|
popq %r15;
|
|
ret;
|
|
|