#include <stdlib.h>
#include <stdint.h>
#include <string.h>
-#include "pbkdf2.h"
+#include <openssl/evp.h>
#include <emmintrin.h>
static inline void xor_salsa8_sse2(__m128i B[4], const __m128i Bx[4])
{
- __m128i X0, X1, X2, X3;
- __m128i T;
- int i;
-
- X0 = B[0] = _mm_xor_si128(B[0], Bx[0]);
- X1 = B[1] = _mm_xor_si128(B[1], Bx[1]);
- X2 = B[2] = _mm_xor_si128(B[2], Bx[2]);
- X3 = B[3] = _mm_xor_si128(B[3], Bx[3]);
-
- for (i = 0; i < 8; i += 2) {
- /* Operate on "columns". */
- T = _mm_add_epi32(X0, X3);
- X1 = _mm_xor_si128(X1, _mm_slli_epi32(T, 7));
- X1 = _mm_xor_si128(X1, _mm_srli_epi32(T, 25));
- T = _mm_add_epi32(X1, X0);
- X2 = _mm_xor_si128(X2, _mm_slli_epi32(T, 9));
- X2 = _mm_xor_si128(X2, _mm_srli_epi32(T, 23));
- T = _mm_add_epi32(X2, X1);
- X3 = _mm_xor_si128(X3, _mm_slli_epi32(T, 13));
- X3 = _mm_xor_si128(X3, _mm_srli_epi32(T, 19));
- T = _mm_add_epi32(X3, X2);
- X0 = _mm_xor_si128(X0, _mm_slli_epi32(T, 18));
- X0 = _mm_xor_si128(X0, _mm_srli_epi32(T, 14));
-
- /* Rearrange data. */
- X1 = _mm_shuffle_epi32(X1, 0x93);
- X2 = _mm_shuffle_epi32(X2, 0x4E);
- X3 = _mm_shuffle_epi32(X3, 0x39);
-
- /* Operate on "rows". */
- T = _mm_add_epi32(X0, X1);
- X3 = _mm_xor_si128(X3, _mm_slli_epi32(T, 7));
- X3 = _mm_xor_si128(X3, _mm_srli_epi32(T, 25));
- T = _mm_add_epi32(X3, X0);
- X2 = _mm_xor_si128(X2, _mm_slli_epi32(T, 9));
- X2 = _mm_xor_si128(X2, _mm_srli_epi32(T, 23));
- T = _mm_add_epi32(X2, X3);
- X1 = _mm_xor_si128(X1, _mm_slli_epi32(T, 13));
- X1 = _mm_xor_si128(X1, _mm_srli_epi32(T, 19));
- T = _mm_add_epi32(X1, X2);
- X0 = _mm_xor_si128(X0, _mm_slli_epi32(T, 18));
- X0 = _mm_xor_si128(X0, _mm_srli_epi32(T, 14));
-
- /* Rearrange data. */
- X1 = _mm_shuffle_epi32(X1, 0x39);
- X2 = _mm_shuffle_epi32(X2, 0x4E);
- X3 = _mm_shuffle_epi32(X3, 0x93);
- }
-
- B[0] = _mm_add_epi32(B[0], X0);
- B[1] = _mm_add_epi32(B[1], X1);
- B[2] = _mm_add_epi32(B[2], X2);
- B[3] = _mm_add_epi32(B[3], X3);
+ __m128i X0, X1, X2, X3;
+ __m128i T;
+ int i;
+
+ X0 = B[0] = _mm_xor_si128(B[0], Bx[0]);
+ X1 = B[1] = _mm_xor_si128(B[1], Bx[1]);
+ X2 = B[2] = _mm_xor_si128(B[2], Bx[2]);
+ X3 = B[3] = _mm_xor_si128(B[3], Bx[3]);
+
+ for (i = 0; i < 8; i += 2) {
+ /* Operate on "columns". */
+ T = _mm_add_epi32(X0, X3);
+ X1 = _mm_xor_si128(X1, _mm_slli_epi32(T, 7));
+ X1 = _mm_xor_si128(X1, _mm_srli_epi32(T, 25));
+ T = _mm_add_epi32(X1, X0);
+ X2 = _mm_xor_si128(X2, _mm_slli_epi32(T, 9));
+ X2 = _mm_xor_si128(X2, _mm_srli_epi32(T, 23));
+ T = _mm_add_epi32(X2, X1);
+ X3 = _mm_xor_si128(X3, _mm_slli_epi32(T, 13));
+ X3 = _mm_xor_si128(X3, _mm_srli_epi32(T, 19));
+ T = _mm_add_epi32(X3, X2);
+ X0 = _mm_xor_si128(X0, _mm_slli_epi32(T, 18));
+ X0 = _mm_xor_si128(X0, _mm_srli_epi32(T, 14));
+
+ /* Rearrange data. */
+ X1 = _mm_shuffle_epi32(X1, 0x93);
+ X2 = _mm_shuffle_epi32(X2, 0x4E);
+ X3 = _mm_shuffle_epi32(X3, 0x39);
+
+ /* Operate on "rows". */
+ T = _mm_add_epi32(X0, X1);
+ X3 = _mm_xor_si128(X3, _mm_slli_epi32(T, 7));
+ X3 = _mm_xor_si128(X3, _mm_srli_epi32(T, 25));
+ T = _mm_add_epi32(X3, X0);
+ X2 = _mm_xor_si128(X2, _mm_slli_epi32(T, 9));
+ X2 = _mm_xor_si128(X2, _mm_srli_epi32(T, 23));
+ T = _mm_add_epi32(X2, X3);
+ X1 = _mm_xor_si128(X1, _mm_slli_epi32(T, 13));
+ X1 = _mm_xor_si128(X1, _mm_srli_epi32(T, 19));
+ T = _mm_add_epi32(X1, X2);
+ X0 = _mm_xor_si128(X0, _mm_slli_epi32(T, 18));
+ X0 = _mm_xor_si128(X0, _mm_srli_epi32(T, 14));
+
+ /* Rearrange data. */
+ X1 = _mm_shuffle_epi32(X1, 0x39);
+ X2 = _mm_shuffle_epi32(X2, 0x4E);
+ X3 = _mm_shuffle_epi32(X3, 0x93);
+ }
+
+ B[0] = _mm_add_epi32(B[0], X0);
+ B[1] = _mm_add_epi32(B[1], X1);
+ B[2] = _mm_add_epi32(B[2], X2);
+ B[3] = _mm_add_epi32(B[3], X3);
}
uint256 scrypt_blockhash__sse2(const uint8_t* input)
{
uint256 result = 0;
uint8_t scratchpad[SCRYPT_BUFFER_SIZE];
- uint8_t B[128];
- union {
- __m128i i128[8];
- uint32_t u32[32];
- } X;
- __m128i *V;
- uint32_t i, j, k;
-
- V = (__m128i *)(((uintptr_t)(scratchpad) + 63) & ~ (uintptr_t)(63));
-
- PBKDF2_SHA256((const uint8_t *)input, 80, (const uint8_t *)input, 80, 1, B, 128);
-
- for (k = 0; k < 2; k++) {
- for (i = 0; i < 16; i++) {
- X.u32[k * 16 + i] = le32dec(&B[(k * 16 + (i * 5 % 16)) * 4]);
- }
- }
-
- for (i = 0; i < 1024; i++) {
- for (k = 0; k < 8; k++)
- V[i * 8 + k] = X.i128[k];
- xor_salsa8_sse2(&X.i128[0], &X.i128[4]);
- xor_salsa8_sse2(&X.i128[4], &X.i128[0]);
- }
- for (i = 0; i < 1024; i++) {
- j = 8 * (X.u32[16] & 1023);
- for (k = 0; k < 8; k++)
- X.i128[k] = _mm_xor_si128(X.i128[k], V[j + k]);
- xor_salsa8_sse2(&X.i128[0], &X.i128[4]);
- xor_salsa8_sse2(&X.i128[4], &X.i128[0]);
- }
-
- for (k = 0; k < 2; k++) {
- for (i = 0; i < 16; i++) {
- le32enc(&B[(k * 16 + (i * 5 % 16)) * 4], X.u32[k * 16 + i]);
- }
- }
-
- PBKDF2_SHA256((const uint8_t *)input, 80, B, 128, 1, (uint8_t *)&result, 32);
+ uint8_t B[128];
+ union {
+ __m128i i128[8];
+ uint32_t u32[32];
+ } X;
+ __m128i *V;
+ uint32_t i, j, k;
+
+ V = (__m128i *)(((uintptr_t)(scratchpad) + 63) & ~ (uintptr_t)(63));
+
+ PKCS5_PBKDF2_HMAC((const int8_t *)input, 80, (const int8_t *)input, 80, 1, EVP_sha256(), 128, B);
+
+ for (k = 0; k < 2; k++) {
+ for (i = 0; i < 16; i++) {
+ X.u32[k * 16 + i] = le32dec(&B[(k * 16 + (i * 5 % 16)) * 4]);
+ }
+ }
+
+ for (i = 0; i < 1024; i++) {
+ for (k = 0; k < 8; k++)
+ V[i * 8 + k] = X.i128[k];
+ xor_salsa8_sse2(&X.i128[0], &X.i128[4]);
+ xor_salsa8_sse2(&X.i128[4], &X.i128[0]);
+ }
+ for (i = 0; i < 1024; i++) {
+ j = 8 * (X.u32[16] & 1023);
+ for (k = 0; k < 8; k++)
+ X.i128[k] = _mm_xor_si128(X.i128[k], V[j + k]);
+ xor_salsa8_sse2(&X.i128[0], &X.i128[4]);
+ xor_salsa8_sse2(&X.i128[4], &X.i128[0]);
+ }
+
+ for (k = 0; k < 2; k++) {
+ for (i = 0; i < 16; i++) {
+ le32enc(&B[(k * 16 + (i * 5 % 16)) * 4], X.u32[k * 16 + i]);
+ }
+ }
+
+ PKCS5_PBKDF2_HMAC((const int8_t *)input, 80, B, 128, 1, EVP_sha256(), 32, (int8_t*)&result);
return result;
}
git://git.novaco.in / novacoin.git / blobdiff