diff options
Diffstat (limited to 'src/hash/sha1_amd64/sha1_amd64_imp.S')
| -rw-r--r-- | src/hash/sha1_amd64/sha1_amd64_imp.S | 258 |
1 files changed, 258 insertions, 0 deletions
diff --git a/src/hash/sha1_amd64/sha1_amd64_imp.S b/src/hash/sha1_amd64/sha1_amd64_imp.S new file mode 100644 index 000000000..f20494999 --- /dev/null +++ b/src/hash/sha1_amd64/sha1_amd64_imp.S @@ -0,0 +1,258 @@ +/************************************************* +* SHA-160 Source File * +* (C) 1999-2007 Jack Lloyd * +*************************************************/ + +#include <botan/asm_macr.h> + +START_LISTING(sha1_amd64.S) + +START_FUNCTION(botan_sha160_amd64_compress) + +#define DIGEST_ARR %rdi +#define INPUT %rsi +#define W %rdx +#define LOOP_CTR %eax + +#define A %r8d +#define B %r9d +#define C %r10d +#define D %r11d +#define E %ecx + + ZEROIZE(LOOP_CTR) + +ALIGN; +.LOOP_LOAD_INPUT: + addl $8, %eax + + movq ARRAY8(INPUT, 0), %r8 + movq ARRAY8(INPUT, 1), %r9 + movq ARRAY8(INPUT, 2), %r10 + movq ARRAY8(INPUT, 3), %r11 + + bswap %r8 + bswap %r9 + bswap %r10 + bswap %r11 + + rolq $32, %r8 + rolq $32, %r9 + rolq $32, %r10 + rolq $32, %r11 + + movq %r8, ARRAY8(W, 0) + movq %r9, ARRAY8(W, 1) + movq %r10, ARRAY8(W, 2) + movq %r11, ARRAY8(W, 3) + + addq $32, W + addq $32, INPUT + + cmp IMM(16), LOOP_CTR + jne .LOOP_LOAD_INPUT + +/* +#define A %r8d +#define B %r9d +#define C %r10d +#define D %r11d +#define E %ecx +*/ + +ALIGN; +.LOOP_EXPANSION: + addl $4, LOOP_CTR + + ZEROIZE(A) + ASSIGN(B, ARRAY4(W, -1)) + ASSIGN(C, ARRAY4(W, -2)) + ASSIGN(D, ARRAY4(W, -3)) + + XOR(A, ARRAY4(W, -5)) + XOR(B, ARRAY4(W, -6)) + XOR(C, ARRAY4(W, -7)) + XOR(D, ARRAY4(W, -8)) + + XOR(A, ARRAY4(W, -11)) + XOR(B, ARRAY4(W, -12)) + XOR(C, ARRAY4(W, -13)) + XOR(D, ARRAY4(W, -14)) + + XOR(A, ARRAY4(W, -13)) + XOR(B, ARRAY4(W, -14)) + XOR(C, ARRAY4(W, -15)) + XOR(D, ARRAY4(W, -16)) + + ROTL_IMM(D, 1) + ROTL_IMM(C, 1) + ROTL_IMM(B, 1) + XOR(A, D) + ROTL_IMM(A, 1) + + ASSIGN(ARRAY4(W, 0), D) + ASSIGN(ARRAY4(W, 1), C) + ASSIGN(ARRAY4(W, 2), B) + ASSIGN(ARRAY4(W, 3), A) + + addq $16, W + cmp IMM(80), LOOP_CTR + jne .LOOP_EXPANSION + + subq $320, W + +#define MAGIC1 0x5A827999 +#define MAGIC2 0x6ED9EBA1 +#define MAGIC3 0x8F1BBCDC +#define MAGIC4 0xCA62C1D6 + +#define T %esi +#define T2 %eax + +#define F1(A, B, C, D, E, F, N) \ + ASSIGN(T2, ARRAY4(W, N)) ; \ + ASSIGN(A, F) ; \ + ROTL_IMM(F, 5) ; \ + ADD(F, E) ; \ + ASSIGN(E, C) ; \ + XOR(E, D) ; \ + ADD3_IMM(F, T2, MAGIC1) ; \ + AND(E, B) ; \ + XOR(E, D) ; \ + ROTR_IMM(B, 2) ; \ + ADD(E, F) ; + +#define F2_4(A, B, C, D, E, F, N, MAGIC) \ + ASSIGN(T2, ARRAY4(W, N)) ; \ + ASSIGN(A, F) ; \ + ROTL_IMM(F, 5) ; \ + ADD(F, E) ; \ + ASSIGN(E, B) ; \ + XOR(E, C) ; \ + ADD3_IMM(F, T2, MAGIC) ; \ + XOR(E, D) ; \ + ROTR_IMM(B, 2) ; \ + ADD(E, F) ; + +#define F3(A, B, C, D, E, F, N) \ + ASSIGN(T2, ARRAY4(W, N)) ; \ + ASSIGN(A, F) ; \ + ROTL_IMM(F, 5) ; \ + ADD(F, E) ; \ + ASSIGN(E, B) ; \ + OR(E, C) ; \ + AND(E, D) ; \ + ADD3_IMM(F, T2, MAGIC3) ; \ + ASSIGN(T2, B) ; \ + AND(T2, C) ; \ + OR(E, T2) ; \ + ROTR_IMM(B, 2) ; \ + ADD(E, F) ; + +#define F2(A, B, C, D, E, F, W) \ + F2_4(A, B, C, D, E, F, W, MAGIC2) + +#define F4(A, B, C, D, E, F, W) \ + F2_4(A, B, C, D, E, F, W, MAGIC4) + + ASSIGN(T, ARRAY4(DIGEST_ARR, 0)) + ASSIGN(B, ARRAY4(DIGEST_ARR, 1)) + ASSIGN(C, ARRAY4(DIGEST_ARR, 2)) + ASSIGN(D, ARRAY4(DIGEST_ARR, 3)) + ASSIGN(E, ARRAY4(DIGEST_ARR, 4)) + + /* First Round */ + F1(A, B, C, D, E, T, 0) + F1(T, A, B, C, D, E, 1) + F1(E, T, A, B, C, D, 2) + F1(D, E, T, A, B, C, 3) + F1(C, D, E, T, A, B, 4) + F1(B, C, D, E, T, A, 5) + F1(A, B, C, D, E, T, 6) + F1(T, A, B, C, D, E, 7) + F1(E, T, A, B, C, D, 8) + F1(D, E, T, A, B, C, 9) + F1(C, D, E, T, A, B, 10) + F1(B, C, D, E, T, A, 11) + F1(A, B, C, D, E, T, 12) + F1(T, A, B, C, D, E, 13) + F1(E, T, A, B, C, D, 14) + F1(D, E, T, A, B, C, 15) + F1(C, D, E, T, A, B, 16) + F1(B, C, D, E, T, A, 17) + F1(A, B, C, D, E, T, 18) + F1(T, A, B, C, D, E, 19) + + /* Second Round */ + F2(E, T, A, B, C, D, 20) + F2(D, E, T, A, B, C, 21) + F2(C, D, E, T, A, B, 22) + F2(B, C, D, E, T, A, 23) + F2(A, B, C, D, E, T, 24) + F2(T, A, B, C, D, E, 25) + F2(E, T, A, B, C, D, 26) + F2(D, E, T, A, B, C, 27) + F2(C, D, E, T, A, B, 28) + F2(B, C, D, E, T, A, 29) + F2(A, B, C, D, E, T, 30) + F2(T, A, B, C, D, E, 31) + F2(E, T, A, B, C, D, 32) + F2(D, E, T, A, B, C, 33) + F2(C, D, E, T, A, B, 34) + F2(B, C, D, E, T, A, 35) + F2(A, B, C, D, E, T, 36) + F2(T, A, B, C, D, E, 37) + F2(E, T, A, B, C, D, 38) + F2(D, E, T, A, B, C, 39) + + /* Third Round */ + F3(C, D, E, T, A, B, 40) + F3(B, C, D, E, T, A, 41) + F3(A, B, C, D, E, T, 42) + F3(T, A, B, C, D, E, 43) + F3(E, T, A, B, C, D, 44) + F3(D, E, T, A, B, C, 45) + F3(C, D, E, T, A, B, 46) + F3(B, C, D, E, T, A, 47) + F3(A, B, C, D, E, T, 48) + F3(T, A, B, C, D, E, 49) + F3(E, T, A, B, C, D, 50) + F3(D, E, T, A, B, C, 51) + F3(C, D, E, T, A, B, 52) + F3(B, C, D, E, T, A, 53) + F3(A, B, C, D, E, T, 54) + F3(T, A, B, C, D, E, 55) + F3(E, T, A, B, C, D, 56) + F3(D, E, T, A, B, C, 57) + F3(C, D, E, T, A, B, 58) + F3(B, C, D, E, T, A, 59) + + /* Fourth Round */ + F4(A, B, C, D, E, T, 60) + F4(T, A, B, C, D, E, 61) + F4(E, T, A, B, C, D, 62) + F4(D, E, T, A, B, C, 63) + F4(C, D, E, T, A, B, 64) + F4(B, C, D, E, T, A, 65) + F4(A, B, C, D, E, T, 66) + F4(T, A, B, C, D, E, 67) + F4(E, T, A, B, C, D, 68) + F4(D, E, T, A, B, C, 69) + F4(C, D, E, T, A, B, 70) + F4(B, C, D, E, T, A, 71) + F4(A, B, C, D, E, T, 72) + F4(T, A, B, C, D, E, 73) + F4(E, T, A, B, C, D, 74) + F4(D, E, T, A, B, C, 75) + F4(C, D, E, T, A, B, 76) + F4(B, C, D, E, T, A, 77) + F4(A, B, C, D, E, T, 78) + F4(T, A, B, C, D, E, 79) + + ADD(ARRAY4(DIGEST_ARR, 0), D) + ADD(ARRAY4(DIGEST_ARR, 1), T) + ADD(ARRAY4(DIGEST_ARR, 2), A) + ADD(ARRAY4(DIGEST_ARR, 3), B) + ADD(ARRAY4(DIGEST_ARR, 4), C) + +END_FUNCTION(botan_sha160_amd64_compress) |