updated for verushashv2

verus/haraka.h

@@ -1,3 +1,110 @@
+
+/*
+The MIT License (MIT)
+
+Copyright (c) 2016 kste
+
+Permission is hereby granted, free of charge, to any person obtaining a copy
+of this software and associated documentation files (the "Software"), to deal
+in the Software without restriction, including without limitation the rights
+to use, copy, modify, merge, publish, distribute, sublicense, and/or sell
+copies of the Software, and to permit persons to whom the Software is
+furnished to do so, subject to the following conditions:
+
+The above copyright notice and this permission notice shall be included in all
+copies or substantial portions of the Software.
+
+THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
+AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
+LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
+OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
+SOFTWARE.
+
+Optimized Implementations for Haraka256 and Haraka512
+*/
+#ifndef HARAKA_H_
+#define HARAKA_H_
+
+#include "immintrin.h"
+
+#define NUMROUNDS 5
+
+#ifdef _WIN32
+typedef unsigned long long u64;
+#else
+typedef unsigned long u64;
+#endif
+typedef __m128i u128;
+
+extern u128 rc[40];
+
+#define LOAD(src) _mm_load_si128((u128 *)(src))
+#define STORE(dest,src) _mm_storeu_si128((u128 *)(dest),src)
+
+#define AES2(s0, s1, rci) \
+  s0 = _mm_aesenc_si128(s0, rc[rci]); \
+  s1 = _mm_aesenc_si128(s1, rc[rci + 1]); \
+  s0 = _mm_aesenc_si128(s0, rc[rci + 2]); \
+  s1 = _mm_aesenc_si128(s1, rc[rci + 3]);
+
+#define AES2_4x(s0, s1, s2, s3, rci) \
+  AES2(s0[0], s0[1], rci); \
+  AES2(s1[0], s1[1], rci); \
+  AES2(s2[0], s2[1], rci); \
+  AES2(s3[0], s3[1], rci);
+
+#define AES2_8x(s0, s1, s2, s3, s4, s5, s6, s7, rci) \
+  AES2_4x(s0, s1, s2, s3, rci); \
+  AES2_4x(s4, s5, s6, s7, rci);
+
+#define AES4(s0, s1, s2, s3, rci) \
+  s0 = _mm_aesenc_si128(s0, rc[rci]); \
+  s1 = _mm_aesenc_si128(s1, rc[rci + 1]); \
+  s2 = _mm_aesenc_si128(s2, rc[rci + 2]); \
+  s3 = _mm_aesenc_si128(s3, rc[rci + 3]); \
+  s0 = _mm_aesenc_si128(s0, rc[rci + 4]); \
+  s1 = _mm_aesenc_si128(s1, rc[rci + 5]); \
+  s2 = _mm_aesenc_si128(s2, rc[rci + 6]); \
+  s3 = _mm_aesenc_si128(s3, rc[rci + 7]); \
+
+#define AES4_zero(s0, s1, s2, s3, rci) \
+  s0 = _mm_aesenc_si128(s0, rc0[rci]); \
+  s1 = _mm_aesenc_si128(s1, rc0[rci + 1]); \
+  s2 = _mm_aesenc_si128(s2, rc0[rci + 2]); \
+  s3 = _mm_aesenc_si128(s3, rc0[rci + 3]); \
+  s0 = _mm_aesenc_si128(s0, rc0[rci + 4]); \
+  s1 = _mm_aesenc_si128(s1, rc0[rci + 5]); \
+  s2 = _mm_aesenc_si128(s2, rc0[rci + 6]); \
+  s3 = _mm_aesenc_si128(s3, rc0[rci + 7]); \
+
+#define AES4_4x(s0, s1, s2, s3, rci) \
+  AES4(s0[0], s0[1], s0[2], s0[3], rci); \
+  AES4(s1[0], s1[1], s1[2], s1[3], rci); \
+  AES4(s2[0], s2[1], s2[2], s2[3], rci); \
+  AES4(s3[0], s3[1], s3[2], s3[3], rci);
+
+#define AES4_8x(s0, s1, s2, s3, s4, s5, s6, s7, rci) \
+  AES4_4x(s0, s1, s2, s3, rci); \
+  AES4_4x(s4, s5, s6, s7, rci);
+
+#define MIX2(s0, s1) \
+  tmp = _mm_unpacklo_epi32(s0, s1); \
+  s1 = _mm_unpackhi_epi32(s0, s1); \
+  s0 = tmp;
+
+#define MIX4(s0, s1, s2, s3) \
+  tmp  = _mm_unpacklo_epi32(s0, s1); \
+  s0 = _mm_unpackhi_epi32(s0, s1); \
+  s1 = _mm_unpacklo_epi32(s2, s3); \
+  s2 = _mm_unpackhi_epi32(s2, s3); \
+  s3 = _mm_unpacklo_epi32(s0, s2); \
+  s0 = _mm_unpackhi_epi32(s0, s2); \
+  s2 = _mm_unpackhi_epi32(s1, tmp); \
+  s1 = _mm_unpacklo_epi32(s1, tmp);
+
+
 /*
 The MIT License (MIT)
 
@@ -23,8 +130,8 @@ SOFTWARE.
 
 Optimized Implementations for Haraka256 and Haraka512
 */
-#ifndef HARAKA_H_
-#define HARAKA_H_
+#ifdef HARAKA_H_
+
 
 #include "immintrin.h"
 
@@ -68,15 +175,15 @@ extern u128 rc[40];
   s2 = _mm_aesenc_si128(s2, rc[rci + 6]); \
   s3 = _mm_aesenc_si128(s3, rc[rci + 7]); \
 
-#define AES40(s0, s1, s2, s3) \
-  s0 = _mm_aesenc_si128(s0, zilch); \
-  s1 = _mm_aesenc_si128(s1, zilch); \
-  s2 = _mm_aesenc_si128(s2, zilch); \
-  s3 = _mm_aesenc_si128(s3, zilch); \
-  s0 = _mm_aesenc_si128(s0, zilch); \
-  s1 = _mm_aesenc_si128(s1, zilch); \
-  s2 = _mm_aesenc_si128(s2, zilch); \
-  s3 = _mm_aesenc_si128(s3, zilch); \
+#define AES4LAST(s0, s1, s2, s3, rci) \
+  s0 = _mm_aesenc_si128(s0, rc[rci]); \
+  s1 = _mm_aesenc_si128(s1, rc[rci + 1]); \
+  s2 = _mm_aesenc_si128(s2, rc[rci + 2]); \
+  s3 = _mm_aesenc_si128(s3, rc[rci + 3]); \
+  s0 = _mm_aesenc_si128(s0, rc[rci + 4]); \
+  s1 = _mm_aesenc_si128(s1, rc[rci + 5]); \
+  s2 = _mm_aesenc_si128(s2, rc[rci + 6]); \
+  s3 = _mm_aesenc_si128(s3, rc[rci + 7]); \
 
 #define AES4_zero(s0, s1, s2, s3, rci) \
   s0 = _mm_aesenc_si128(s0, rc0[rci]); \
@@ -88,11 +195,11 @@ extern u128 rc[40];
   s2 = _mm_aesenc_si128(s2, rc0[rci + 6]); \
   s3 = _mm_aesenc_si128(s3, rc0[rci + 7]); \
 
-#define AES4_4x(s0, s1, s2, s3) \
-  AES40(s0[0], s0[1], s0[2], s0[3]); \
-  AES40(s1[0], s1[1], s1[2], s1[3]); \
-  AES40(s2[0], s2[1], s2[2], s2[3]); \
-  AES40(s3[0], s3[1], s3[2], s3[3]);
+#define AES4_4x(s0, s1, s2, s3, rci) \
+  AES4(s0[0], s0[1], s0[2], s0[3], rci); \
+  AES4(s1[0], s1[1], s1[2], s1[3], rci); \
+  AES4(s2[0], s2[1], s2[2], s2[3], rci); \
+  AES4(s3[0], s3[1], s3[2], s3[3], rci);
 
 #define AES4_8x(s0, s1, s2, s3, s4, s5, s6, s7, rci) \
   AES4_4x(s0, s1, s2, s3, rci); \
@@ -113,27 +220,60 @@ extern u128 rc[40];
   s2 = _mm_unpackhi_epi32(s1, tmp); \
   s1 = _mm_unpacklo_epi32(s1, tmp);
 
+#define MIX4LAST(s0, s1, s2, s3) \
+  s0 = _mm_unpackhi_epi32(s0, s1); \
+  s2 = _mm_unpackhi_epi32(s2, s3); \
+  s3 = _mm_unpacklo_epi32(s0, s2); 
+
+#define MIX4_LASTBUT1(s0, s1, s2, s3) \
+  tmp  = _mm_unpacklo_epi32(s0, s1); \
+  s1 = _mm_unpacklo_epi32(s2, s3); \
+  s2 = _mm_unpackhi_epi32(s1, tmp); 
+
+#define AES4_LAST(s3, rci) \
+  _mm_aesenc_si128(s3, rc[rci + 2]); \
+  _mm_aesenc_si128(s3, rc[rci + 6]); \
+
 #define TRUNCSTORE(out, s0, s1, s2, s3) \
   *(u64*)(out) = *(((u64*)&s0 + 1)); \
   *(u64*)(out + 8) = *(((u64*)&s1 + 1)); \
   *(u64*)(out + 16) = *(((u64*)&s2 + 0)); \
-  *(u64*)(out + 24) = *(((u64*)&s3 + 0));
+  *(u64*)(out + 24) = *(((u64*)&s3 + 0));
+
+void load_constants();
+void test_implementations();
+
+void load_constants();
+
+void haraka256(unsigned char *out, const unsigned char *in);
+void haraka256_keyed(unsigned char *out, const unsigned char *in, const u128 *rc);
+void haraka256_4x(unsigned char *out, const unsigned char *in);
+void haraka256_8x(unsigned char *out, const unsigned char *in);
+
+void haraka512(unsigned char *out, const unsigned char *in);
+void haraka512_zero(unsigned char *out, const unsigned char *in);
+void haraka512_keyed(unsigned char *out, const unsigned char *in, const u128 *rc);
+void haraka512_4x(unsigned char *out, const unsigned char *in);
+void haraka512_8x(unsigned char *out, const unsigned char *in);
+
+#endif
 
-#define TRUNCSTORE2(out, s3) \
-  *(u64*)(out + 24) = *(((u64*)&s3 + 0));
 
 void load_constants();
 void test_implementations();
 
 void load_constants();
 
 void haraka256(unsigned char *out, const unsigned char *in);
+void haraka256_keyed(unsigned char *out, const unsigned char *in, const u128 *rc);
 void haraka256_4x(unsigned char *out, const unsigned char *in);
 void haraka256_8x(unsigned char *out, const unsigned char *in);
 
 void haraka512(unsigned char *out, const unsigned char *in);
 void haraka512_zero(unsigned char *out, const unsigned char *in);
+void haraka512_keyed(unsigned char *out, const unsigned char *in, const u128 *rc);
 void haraka512_4x(unsigned char *out, const unsigned char *in);
 void haraka512_8x(unsigned char *out, const unsigned char *in);
 
 #endif
+