v3.20.1

Makefile.am

@@ -289,7 +289,7 @@ cpuminer_SOURCES = \
   algo/yescrypt/yescrypt-best.c \
   algo/yespower/yespower-gate.c \
   algo/yespower/yespower-blake2b.c \
-  algo/yespower/crypto/blake2b-yp.c \
+  algo/yespower/crypto/hmac-blake2b.c \
   algo/yespower/yescrypt-r8g.c \
   algo/yespower/yespower-opt.c
 

RELEASE_NOTES

@@ -65,6 +65,15 @@ If not what makes it happen or not happen?
 Change Log
 ----------
 
+v3.20.1
+
+sph_blake2b optimized 1-way SSSE3 & AVX2.
+Removed duplicate Blake2b used by Power2b algo, will now use optimized sph_blake2b.
+Removed imprecise hash & target display from rejected share log.
+Share and target difficulty is now displayed only for low diificulty shares.
+Updated configure.ac to check for AVX512 asm support.
+Small optimization to Lyra2 SSE2.
+
 v3.20.0
 
 #375 Fixed segfault in algos using Groestl VAES due to use of uninitialized data.

algo/blake/blake2b-hash-4way.c

@@ -52,6 +52,180 @@ static const uint8_t sigma[12][16] =
 };
 
 
+#define Z00   0
+#define Z01   1
+#define Z02   2
+#define Z03   3
+#define Z04   4
+#define Z05   5
+#define Z06   6
+#define Z07   7
+#define Z08   8
+#define Z09   9
+#define Z0A   A
+#define Z0B   B
+#define Z0C   C
+#define Z0D   D
+#define Z0E   E
+#define Z0F   F
+
+#define Z10   E
+#define Z11   A
+#define Z12   4
+#define Z13   8
+#define Z14   9
+#define Z15   F
+#define Z16   D
+#define Z17   6
+#define Z18   1
+#define Z19   C
+#define Z1A   0
+#define Z1B   2
+#define Z1C   B
+#define Z1D   7
+#define Z1E   5
+#define Z1F   3
+
+#define Z20   B
+#define Z21   8
+#define Z22   C
+#define Z23   0
+#define Z24   5
+#define Z25   2
+#define Z26   F
+#define Z27   D
+#define Z28   A
+#define Z29   E
+#define Z2A   3
+#define Z2B   6
+#define Z2C   7
+#define Z2D   1
+#define Z2E   9
+#define Z2F   4
+
+#define Z30   7
+#define Z31   9
+#define Z32   3
+#define Z33   1
+#define Z34   D
+#define Z35   C
+#define Z36   B
+#define Z37   E
+#define Z38   2
+#define Z39   6
+#define Z3A   5
+#define Z3B   A
+#define Z3C   4
+#define Z3D   0
+#define Z3E   F
+#define Z3F   8
+
+#define Z40   9
+#define Z41   0
+#define Z42   5
+#define Z43   7
+#define Z44   2
+#define Z45   4
+#define Z46   A
+#define Z47   F
+#define Z48   E
+#define Z49   1
+#define Z4A   B
+#define Z4B   C
+#define Z4C   6
+#define Z4D   8
+#define Z4E   3
+#define Z4F   D
+
+#define Z50   2
+#define Z51   C
+#define Z52   6
+#define Z53   A
+#define Z54   0
+#define Z55   B
+#define Z56   8
+#define Z57   3
+#define Z58   4
+#define Z59   D
+#define Z5A   7
+#define Z5B   5
+#define Z5C   F
+#define Z5D   E
+#define Z5E   1
+#define Z5F   9
+
+#define Z60   C
+#define Z61   5
+#define Z62   1
+#define Z63   F
+#define Z64   E
+#define Z65   D
+#define Z66   4
+#define Z67   A
+#define Z68   0
+#define Z69   7
+#define Z6A   6
+#define Z6B   3
+#define Z6C   9
+#define Z6D   2
+#define Z6E   8
+#define Z6F   B
+
+#define Z70   D
+#define Z71   B
+#define Z72   7
+#define Z73   E
+#define Z74   C
+#define Z75   1
+#define Z76   3
+#define Z77   9
+#define Z78   5
+#define Z79   0
+#define Z7A   F
+#define Z7B   4
+#define Z7C   8
+#define Z7D   6
+#define Z7E   2
+#define Z7F   A
+
+#define Z80   6
+#define Z81   F
+#define Z82   E
+#define Z83   9
+#define Z84   B
+#define Z85   3
+#define Z86   0
+#define Z87   8
+#define Z88   C
+#define Z89   2
+#define Z8A   D
+#define Z8B   7
+#define Z8C   1
+#define Z8D   4
+#define Z8E   A
+#define Z8F   5
+
+#define Z90   A
+#define Z91   2
+#define Z92   8
+#define Z93   4
+#define Z94   7
+#define Z95   6
+#define Z96   1
+#define Z97   5
+#define Z98   F
+#define Z99   B
+#define Z9A   9
+#define Z9B   E
+#define Z9C   3
+#define Z9D   C
+#define Z9E   D
+#define Z9F   0
+
+#define Mx(r, i)    Mx_(Z ## r ## i)
+#define Mx_(n)      Mx__(n)
+#define Mx__(n)     M ## n
+
 #if defined(__AVX512F__) && defined(__AVX512VL__) && defined(__AVX512DQ__) && defined(__AVX512BW__)
 
 #define B2B8W_G(a, b, c, d, x, y) \

algo/blake/blake512-hash-4way.c

@@ -757,7 +757,6 @@ void blake512_8way_final_le( blake_8way_big_context *sc, void *hash,
    GB_8WAY(Mx(1, C), Mx(1, D), CBx(1, C), CBx(1, D), V2, V7, V8, VD);
    GB_8WAY(Mx(1, E), Mx(1, F), CBx(1, E), CBx(1, F), V3, V4, V9, VE);
 
-
    // remaining rounds  
    ROUND_B_8WAY(2);
    ROUND_B_8WAY(3);

algo/blake/sph_blake2b.c

@@ -30,16 +30,10 @@
 #include <stdlib.h>
 #include <stdint.h>
 #include <string.h>
-
+#include "simd-utils.h"
 #include "algo/sha/sph_types.h"
 #include "sph_blake2b.h"
 
-// Cyclic right rotation.
-
-#ifndef ROTR64
-#define ROTR64(x, y)  (((x) >> (y)) ^ ((x) << (64 - (y))))
-#endif
-
 // Little-endian byte access.
 
 #define B2B_GET64(p)                            \
@@ -54,45 +48,131 @@
 
 // G Mixing function.
 
-#define B2B_G(a, b, c, d, x, y) {   \
-	v[a] = v[a] + v[b] + x;         \
-	v[d] = ROTR64(v[d] ^ v[a], 32); \
-	v[c] = v[c] + v[d];             \
-	v[b] = ROTR64(v[b] ^ v[c], 24); \
-	v[a] = v[a] + v[b] + y;         \
-	v[d] = ROTR64(v[d] ^ v[a], 16); \
-	v[c] = v[c] + v[d];             \
-	v[b] = ROTR64(v[b] ^ v[c], 63); }
+#if defined(__AVX2__)
+
+#define BLAKE2B_G( R, Sa, Sb, Sc, Sd, Na, Nb ) \
+{ \
+  V[0] = _mm256_add_epi64( V[0], _mm256_add_epi64( V[1], \
+              _mm256_set_epi64x( m[ sigma[R][Sd] ], m[ sigma[R][Sc] ], \
+                                 m[ sigma[R][Sb] ], m[ sigma[R][Sa] ] ) ) ); \
+  V[3] = mm256_ror_64( _mm256_xor_si256( V[3], V[0] ), Na ); \
+  V[2] = _mm256_add_epi64( V[2], V[3] ); \
+  V[1] = mm256_ror_64( _mm256_xor_si256( V[1], V[2] ), Nb ); \
+}
+
+#define BLAKE2B_ROUND( R ) \
+{ \
+  __m256i *V = (__m256i*)v; \
+  BLAKE2B_G( R,  0,  2,  4,  6, 32, 24 ); \
+  BLAKE2B_G( R,  1,  3,  5,  7, 16, 63 ); \
+  V[3] = mm256_shufll_64( V[3] ); \
+  V[2] = mm256_swap_128( V[2] ); \
+  V[1] = mm256_shuflr_64( V[1] ); \
+  BLAKE2B_G( R,  8, 10, 12, 14, 32, 24 ); \
+  BLAKE2B_G( R,  9, 11, 13, 15, 16, 63 ); \
+  V[3] = mm256_shuflr_64( V[3] ); \
+  V[2] = mm256_swap_128( V[2] ); \
+  V[1] = mm256_shufll_64( V[1] ); \
+}
+
+#elif defined(__SSSE3__)
+
+#define BLAKE2B_G( R, Va, Vb, Vc, Vd, Sa, Sb, Na, Nb ) \
+{ \
+   Va = _mm_add_epi64( Va, _mm_add_epi64( Vb, \
+                 _mm_set_epi64x( m[ sigma[R][Sb] ], m[ sigma[R][Sa] ] ) ) ); \
+   Vd = mm128_ror_64( _mm_xor_si128( Vd, Va ), Na ); \
+   Vc = _mm_add_epi64( Vc, Vd ); \
+   Vb = mm128_ror_64( _mm_xor_si128( Vb, Vc ), Nb ); \
+}
+
+#define BLAKE2B_ROUND( R ) \
+{ \
+   __m128i *V = (__m128i*)v; \
+   __m128i V2, V3, V6, V7; \
+   BLAKE2B_G( R, V[0], V[2], V[4], V[6], 0, 2, 32, 24 ); \
+   BLAKE2B_G( R, V[0], V[2], V[4], V[6], 1, 3, 16, 63 ); \
+   BLAKE2B_G( R, V[1], V[3], V[5], V[7], 4, 6, 32, 24 ); \
+   BLAKE2B_G( R, V[1], V[3], V[5], V[7], 5, 7, 16, 63 ); \
+   V2 = mm128_shufl2r_64( V[2], V[3] ); \
+   V3 = mm128_shufl2r_64( V[3], V[2] ); \
+   V6 = mm128_shufl2l_64( V[6], V[7] ); \
+   V7 = mm128_shufl2l_64( V[7], V[6] ); \
+   BLAKE2B_G( R, V[0], V2, V[5], V6,  8, 10, 32, 24 ); \
+   BLAKE2B_G( R, V[0], V2, V[5], V6,  9, 11, 16, 63 ); \
+   BLAKE2B_G( R, V[1], V3, V[4], V7, 12, 14, 32, 24 ); \
+   BLAKE2B_G( R, V[1], V3, V[4], V7, 13, 15, 16, 63 ); \
+   V[2] = mm128_shufl2l_64( V2, V3 ); \
+   V[3] = mm128_shufl2l_64( V3, V2 ); \
+   V[6] = mm128_shufl2r_64( V6, V7 ); \
+   V[7] = mm128_shufl2r_64( V7, V6 ); \
+}
+
+#else
+
+#ifndef ROTR64
+#define ROTR64(x, y)  (((x) >> (y)) ^ ((x) << (64 - (y))))
+#endif
+
+#define BLAKE2B_G( R, Va, Vb, Vc, Vd, Sa, Sb ) \
+{ \
+   Va = Va + Vb + m[ sigma[R][Sa] ]; \
+   Vd = ROTR64( Vd ^ Va, 32 ); \
+   Vc = Vc + Vd; \
+   Vb = ROTR64( Vb ^ Vc, 24 ); \
+   Va = Va + Vb + m[ sigma[R][Sb] ]; \
+   Vd = ROTR64( Vd ^ Va, 16 ); \
+   Vc = Vc + Vd; \
+   Vb = ROTR64( Vb ^ Vc, 63 ); \
+}
+
+#define BLAKE2B_ROUND( R ) \
+{ \
+   BLAKE2B_G( R, v[ 0], v[ 4], v[ 8], v[12],  0,  1 ); \
+   BLAKE2B_G( R, v[ 1], v[ 5], v[ 9], v[13],  2,  3 ); \
+   BLAKE2B_G( R, v[ 2], v[ 6], v[10], v[14],  4,  5 ); \
+   BLAKE2B_G( R, v[ 3], v[ 7], v[11], v[15],  6,  7 ); \
+   BLAKE2B_G( R, v[ 0], v[ 5], v[10], v[15],  8,  9 ); \
+   BLAKE2B_G( R, v[ 1], v[ 6], v[11], v[12], 10, 11 ); \
+   BLAKE2B_G( R, v[ 2], v[ 7], v[ 8], v[13], 12, 13 ); \
+   BLAKE2B_G( R, v[ 3], v[ 4], v[ 9], v[14], 14, 15 ); \
+}
+
+#endif
 
 // Initialization Vector.
 
-static const uint64_t blake2b_iv[8] = {
+static const uint64_t blake2b_iv[8] __attribute__ ((aligned (32))) =
+{
 	0x6A09E667F3BCC908, 0xBB67AE8584CAA73B,
 	0x3C6EF372FE94F82B, 0xA54FF53A5F1D36F1,
 	0x510E527FADE682D1, 0x9B05688C2B3E6C1F,
 	0x1F83D9ABFB41BD6B, 0x5BE0CD19137E2179
 };
 
+static const uint8_t sigma[12][16] __attribute__ ((aligned (32))) =
+{
+      { 0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15 },
+      { 14, 10, 4, 8, 9, 15, 13, 6, 1, 12, 0, 2, 11, 7, 5, 3 },
+      { 11, 8, 12, 0, 5, 2, 15, 13, 10, 14, 3, 6, 7, 1, 9, 4 },
+      { 7, 9, 3, 1, 13, 12, 11, 14, 2, 6, 5, 10, 4, 0, 15, 8 },
+      { 9, 0, 5, 7, 2, 4, 10, 15, 14, 1, 11, 12, 6, 8, 3, 13 },
+      { 2, 12, 6, 10, 0, 11, 8, 3, 4, 13, 7, 5, 15, 14, 1, 9 },
+      { 12, 5, 1, 15, 14, 13, 4, 10, 0, 7, 6, 3, 9, 2, 8, 11 },
+      { 13, 11, 7, 14, 12, 1, 3, 9, 5, 0, 15, 4, 8, 6, 2, 10 },
+      { 6, 15, 14, 9, 11, 3, 0, 8, 12, 2, 13, 7, 1, 4, 10, 5 },
+      { 10, 2, 8, 4, 7, 6, 1, 5, 15, 11, 9, 14, 3, 12, 13, 0 },
+      { 0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15 },
+      { 14, 10, 4, 8, 9, 15, 13, 6, 1, 12, 0, 2, 11, 7, 5, 3 }
+};
+
 // Compression function. "last" flag indicates last block.
 
 static void blake2b_compress( sph_blake2b_ctx *ctx, int last )
 {
-	const uint8_t sigma[12][16] = {
-		{ 0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15 },
-		{ 14, 10, 4, 8, 9, 15, 13, 6, 1, 12, 0, 2, 11, 7, 5, 3 },
-		{ 11, 8, 12, 0, 5, 2, 15, 13, 10, 14, 3, 6, 7, 1, 9, 4 },
-		{ 7, 9, 3, 1, 13, 12, 11, 14, 2, 6, 5, 10, 4, 0, 15, 8 },
-		{ 9, 0, 5, 7, 2, 4, 10, 15, 14, 1, 11, 12, 6, 8, 3, 13 },
-		{ 2, 12, 6, 10, 0, 11, 8, 3, 4, 13, 7, 5, 15, 14, 1, 9 },
-		{ 12, 5, 1, 15, 14, 13, 4, 10, 0, 7, 6, 3, 9, 2, 8, 11 },
-		{ 13, 11, 7, 14, 12, 1, 3, 9, 5, 0, 15, 4, 8, 6, 2, 10 },
-		{ 6, 15, 14, 9, 11, 3, 0, 8, 12, 2, 13, 7, 1, 4, 10, 5 },
-		{ 10, 2, 8, 4, 7, 6, 1, 5, 15, 11, 9, 14, 3, 12, 13, 0 },
-		{ 0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15 },
-		{ 14, 10, 4, 8, 9, 15, 13, 6, 1, 12, 0, 2, 11, 7, 5, 3 }
-	};
-	int i;
-	uint64_t v[16], m[16];
+	uint64_t v[16] __attribute__ ((aligned (32)));
+   uint64_t m[16] __attribute__ ((aligned (32)));
+   int i;
 
 	for (i = 0; i < 8; i++) {           // init work variables
 		v[i] = ctx->h[i];
@@ -106,16 +186,8 @@ static void blake2b_compress( sph_blake2b_ctx *ctx, int last )
 	for (i = 0; i < 16; i++)            // get little-endian words
 		m[i] = B2B_GET64(&ctx->b[8 * i]);
 
-	for (i = 0; i < 12; i++) {          // twelve rounds
-		B2B_G( 0, 4,  8, 12, m[sigma[i][ 0]], m[sigma[i][ 1]]);
-		B2B_G( 1, 5,  9, 13, m[sigma[i][ 2]], m[sigma[i][ 3]]);
-		B2B_G( 2, 6, 10, 14, m[sigma[i][ 4]], m[sigma[i][ 5]]);
-		B2B_G( 3, 7, 11, 15, m[sigma[i][ 6]], m[sigma[i][ 7]]);
-		B2B_G( 0, 5, 10, 15, m[sigma[i][ 8]], m[sigma[i][ 9]]);
-		B2B_G( 1, 6, 11, 12, m[sigma[i][10]], m[sigma[i][11]]);
-		B2B_G( 2, 7,  8, 13, m[sigma[i][12]], m[sigma[i][13]]);
-		B2B_G( 3, 4,  9, 14, m[sigma[i][14]], m[sigma[i][15]]);
-	}
+	for (i = 0; i < 12; i++)
+      BLAKE2B_ROUND( i );   
 
 	for( i = 0; i < 8; ++i )
 		ctx->h[i] ^= v[i] ^ v[i + 8];

algo/lyra2/sponge.h

@@ -150,12 +150,10 @@ static inline uint64_t rotr64( const uint64_t w, const unsigned c ){
    G_2X64( s1, s3, s5, s7 ); \
    mm128_vrol256_64( s6, s7 ); \
    mm128_vror256_64( s2, s3 ); \
-   mm128_swap256_128( s4, s5 ); \
-   G_2X64( s0, s2, s4, s6 ); \
-   G_2X64( s1, s3, s5, s7 ); \
+   G_2X64( s0, s2, s5, s6 ); \
+   G_2X64( s1, s3, s4, s7 ); \
    mm128_vror256_64( s6, s7 ); \
-   mm128_vrol256_64( s2, s3 ); \
-   mm128_swap256_128( s4, s5 );
+   mm128_vrol256_64( s2, s3 );
 
 #define LYRA_12_ROUNDS_AVX(s0,s1,s2,s3,s4,s5,s6,s7) \
    LYRA_ROUND_AVX(s0,s1,s2,s3,s4,s5,s6,s7) \

algo/yespower/crypto/blake2b-yp.c

@@ -1,323 +0,0 @@
-/*
- * Copyright 2009 Colin Percival, 2014 savale
- * All rights reserved.
- *
- * Redistribution and use in source and binary forms, with or without
- * modification, are permitted provided that the following conditions
- * are met:
- * 1. Redistributions of source code must retain the above copyright
- *  notice, this list of conditions and the following disclaimer.
- * 2. Redistributions in binary form must reproduce the above copyright
- *  notice, this list of conditions and the following disclaimer in the
- *  documentation and/or other materials provided with the distribution.
- *
- * THIS SOFTWARE IS PROVIDED BY THE AUTHOR AND CONTRIBUTORS ``AS IS'' AND
- * ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
- * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
- * ARE DISCLAIMED.  IN NO EVENT SHALL THE AUTHOR OR CONTRIBUTORS BE LIABLE
- * FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
- * DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS
- * OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
- * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT
- * LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY
- * OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF
- * SUCH DAMAGE.
- *
- * This file was originally written by Colin Percival as part of the Tarsnap
- * online backup system.
- */
-
-#include <stdlib.h>
-#include <stdint.h>
-#include <string.h>
-#include "simd-utils.h"
-#include <algo/yespower/crypto/sph_types.h>
-#include "blake2b-yp.h"
-
-// Cyclic right rotation.
-//#ifndef ROTR64
-//#define ROTR64(x, y)  (((x) >> (y)) ^ ((x) << (64 - (y))))
-//#endif
-
-#define ROTR64(x, y) ror64( x, y )
-
-// Little-endian byte access.
-#define B2B_GET64(p)                            \
-    (((uint64_t) ((uint8_t *) (p))[0]) ^        \
-    (((uint64_t) ((uint8_t *) (p))[1]) << 8) ^  \
-    (((uint64_t) ((uint8_t *) (p))[2]) << 16) ^ \
-    (((uint64_t) ((uint8_t *) (p))[3]) << 24) ^ \
-    (((uint64_t) ((uint8_t *) (p))[4]) << 32) ^ \
-    (((uint64_t) ((uint8_t *) (p))[5]) << 40) ^ \
-    (((uint64_t) ((uint8_t *) (p))[6]) << 48) ^ \
-    (((uint64_t) ((uint8_t *) (p))[7]) << 56))
-
-// G Mixing function.
-#define B2B_G(a, b, c, d, x, y) {   \
-    v[a] = v[a] + v[b] + x;      \
-    v[d] = ROTR64(v[d] ^ v[a], 32); \
-    v[c] = v[c] + v[d];          \
-    v[b] = ROTR64(v[b] ^ v[c], 24); \
-    v[a] = v[a] + v[b] + y;      \
-    v[d] = ROTR64(v[d] ^ v[a], 16); \
-    v[c] = v[c] + v[d];          \
-    v[b] = ROTR64(v[b] ^ v[c], 63); }
-
-// Initialization Vector.
-static const uint64_t blake2b_iv[8] = {
-    0x6A09E667F3BCC908, 0xBB67AE8584CAA73B,
-    0x3C6EF372FE94F82B, 0xA54FF53A5F1D36F1,
-    0x510E527FADE682D1, 0x9B05688C2B3E6C1F,
-    0x1F83D9ABFB41BD6B, 0x5BE0CD19137E2179
-};
-
-// Compression function. "last" flag indicates last block.
-static void blake2b_compress(blake2b_yp_ctx *ctx, int last)
-{
-    const uint8_t sigma[12][16] = {
-        { 0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15 },
-        { 14, 10, 4, 8, 9, 15, 13, 6, 1, 12, 0, 2, 11, 7, 5, 3 },
-        { 11, 8, 12, 0, 5, 2, 15, 13, 10, 14, 3, 6, 7, 1, 9, 4 },
-        { 7, 9, 3, 1, 13, 12, 11, 14, 2, 6, 5, 10, 4, 0, 15, 8 },
-        { 9, 0, 5, 7, 2, 4, 10, 15, 14, 1, 11, 12, 6, 8, 3, 13 },
-        { 2, 12, 6, 10, 0, 11, 8, 3, 4, 13, 7, 5, 15, 14, 1, 9 },
-        { 12, 5, 1, 15, 14, 13, 4, 10, 0, 7, 6, 3, 9, 2, 8, 11 },
-        { 13, 11, 7, 14, 12, 1, 3, 9, 5, 0, 15, 4, 8, 6, 2, 10 },
-        { 6, 15, 14, 9, 11, 3, 0, 8, 12, 2, 13, 7, 1, 4, 10, 5 },
-        { 10, 2, 8, 4, 7, 6, 1, 5, 15, 11, 9, 14, 3, 12, 13, 0 },
-        { 0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15 },
-        { 14, 10, 4, 8, 9, 15, 13, 6, 1, 12, 0, 2, 11, 7, 5, 3 }
-    };
-    int i;
-    uint64_t v[16], m[16];
-
-    // init work variables
-    for (i = 0; i < 8; i++) {
-        v[i] = ctx->h[i];
-        v[i + 8] = blake2b_iv[i];
-    }
-
-    v[12] ^= ctx->t[0]; // low 64 bits of offset
-    v[13] ^= ctx->t[1]; // high 64 bits
-
-    // last block flag set ?
-    if (last) { 
-        v[14] = ~v[14];
-    }
-
-    // get little-endian words
-    for (i = 0; i < 16; i++) {
-        m[i] = B2B_GET64(&ctx->b[8 * i]);
-    }
-
-    // twelve rounds
-    for (i = 0; i < 12; i++) {
-        B2B_G( 0, 4,  8, 12, m[sigma[i][ 0]], m[sigma[i][ 1]]);
-        B2B_G( 1, 5,  9, 13, m[sigma[i][ 2]], m[sigma[i][ 3]]);
-        B2B_G( 2, 6, 10, 14, m[sigma[i][ 4]], m[sigma[i][ 5]]);
-        B2B_G( 3, 7, 11, 15, m[sigma[i][ 6]], m[sigma[i][ 7]]);
-        B2B_G( 0, 5, 10, 15, m[sigma[i][ 8]], m[sigma[i][ 9]]);
-        B2B_G( 1, 6, 11, 12, m[sigma[i][10]], m[sigma[i][11]]);
-        B2B_G( 2, 7,  8, 13, m[sigma[i][12]], m[sigma[i][13]]);
-        B2B_G( 3, 4,  9, 14, m[sigma[i][14]], m[sigma[i][15]]);
-    }
-
-    for(i = 0; i < 8; ++i) {
-        ctx->h[i] ^= v[i] ^ v[i + 8];
-    }
-}
-
-// Initialize the hashing context "ctx" with optional key "key".
-// 1 <= outlen <= 64 gives the digest size in bytes.
-// Secret key (also <= 64 bytes) is optional (keylen = 0).
-int blake2b_yp_init(blake2b_yp_ctx *ctx, size_t outlen,
-    const void *key, size_t keylen) // (keylen=0: no key)
-{
-    size_t i;
-
-    // illegal parameters
-    if (outlen == 0 || outlen > 64 || keylen > 64) {
-        return -1;
-    }
-
-    // state, "param block"
-    for (i = 0; i < 8; i++) {
-        ctx->h[i] = blake2b_iv[i];
-    }
-
-    ctx->h[0] ^= 0x01010000 ^ (keylen << 8) ^ outlen;
-
-    ctx->t[0] = 0; // input count low word
-    ctx->t[1] = 0; // input count high word
-    ctx->c = 0; // pointer within buffer
-    ctx->outlen = outlen;
-
-    // zero input block
-    for (i = keylen; i < 128; i++) {
-        ctx->b[i] = 0;
-    }
-
-    if (keylen > 0) {
-        blake2b_yp_update(ctx, key, keylen);
-        ctx->c = 128; // at the end
-    }
-
-    return 0;
-}
-
-// Add "inlen" bytes from "in" into the hash.
-void blake2b_yp_update(blake2b_yp_ctx *ctx,
-    const void *in, size_t inlen) // data bytes
-{
-    size_t i;
-    for (i = 0; i < inlen; i++) {
-        if (ctx->c == 128) { // buffer full ?
-            ctx->t[0] += ctx->c; // add counters
-            if (ctx->t[0] < ctx->c) // carry overflow ?
-                ctx->t[1]++; // high word
-            blake2b_compress(ctx, 0); // compress (not last)
-            ctx->c = 0; // counter to zero
-        }
-        ctx->b[ctx->c++] = ((const uint8_t *) in)[i];
-    }
-}
-
-// Generate the message digest (size given in init).
-// Result placed in "out".
-void blake2b_yp_final(blake2b_yp_ctx *ctx, void *out)
-{
-    size_t i;
-
-    ctx->t[0] += ctx->c; // mark last block offset
-    // carry overflow
-    if (ctx->t[0] < ctx->c) {
-        ctx->t[1]++; // high word
-    }
-
-    // fill up with zeros
-    while (ctx->c < 128) {
-        ctx->b[ctx->c++] = 0;
-    }
-
-    blake2b_compress(ctx, 1); // final block flag = 1
-
-    // little endian convert and store
-    for (i = 0; i < ctx->outlen; i++) {
-        ((uint8_t *) out)[i] =
-            (ctx->h[i >> 3] >> (8 * (i & 7))) & 0xFF;
-    }
-}
-
-// inlen = number of bytes
-void blake2b_yp_hash(void *out, const void *in, size_t inlen) {
-    blake2b_yp_ctx ctx;
-    blake2b_yp_init(&ctx, 32, NULL, 0);
-    blake2b_yp_update(&ctx, in, inlen);
-    blake2b_yp_final(&ctx, out);
-}
-
-// // keylen = number of bytes
-void hmac_blake2b_yp_init(hmac_yp_ctx *hctx, const void *_key, size_t keylen) {
-    const uint8_t *key = _key;
-    uint8_t keyhash[32];
-    uint8_t pad[64];
-    uint64_t i;
-
-    if (keylen > 64) {
-        blake2b_yp_hash(keyhash, key, keylen);
-        key = keyhash;
-        keylen = 32;
-    }
-
-    blake2b_yp_init(&hctx->inner, 32, NULL, 0);
-    memset(pad, 0x36, 64);
-    for (i = 0; i < keylen; ++i) {
-        pad[i] ^= key[i];
-    }
-
-    blake2b_yp_update(&hctx->inner, pad, 64);
-    blake2b_yp_init(&hctx->outer, 32, NULL, 0);
-    memset(pad, 0x5c, 64);
-    for (i = 0; i < keylen; ++i) {
-        pad[i] ^= key[i];
-    }
-
-    blake2b_yp_update(&hctx->outer, pad, 64);
-    memset(keyhash, 0, 32);
-}
-
-// datalen = number of bits
-void hmac_blake2b_yp_update(hmac_yp_ctx *hctx, const void *data, size_t datalen) {
-    // update the inner state
-    blake2b_yp_update(&hctx->inner, data, datalen);
-}
-
-void hmac_blake2b_yp_final(hmac_yp_ctx *hctx, uint8_t *digest) {
-    uint8_t ihash[32];
-    blake2b_yp_final(&hctx->inner, ihash);
-    blake2b_yp_update(&hctx->outer, ihash, 32);
-    blake2b_yp_final(&hctx->outer, digest);
-    memset(ihash, 0, 32);
-}
-
-// // keylen = number of bytes; inlen = number of bytes
-void hmac_blake2b_yp_hash(void *out, const void *key, size_t keylen, const void *in, size_t inlen) {
-    hmac_yp_ctx hctx;
-    hmac_blake2b_yp_init(&hctx, key, keylen);
-    hmac_blake2b_yp_update(&hctx, in, inlen);
-    hmac_blake2b_yp_final(&hctx, out);
-}
-
-void pbkdf2_blake2b_yp(const uint8_t * passwd, size_t passwdlen, const uint8_t * salt,
-    size_t saltlen, uint64_t c, uint8_t * buf, size_t dkLen)
-{
-    hmac_yp_ctx PShctx, hctx;
-    size_t i;
-    uint32_t ivec;
-    uint8_t U[32];
-    uint8_t T[32];
-    uint64_t j;
-    int k;
-    size_t clen;
-
-    /* Compute HMAC state after processing P and S. */
-    hmac_blake2b_yp_init(&PShctx, passwd, passwdlen);
-    hmac_blake2b_yp_update(&PShctx, salt, saltlen);
-
-    /* Iterate through the blocks. */
-    for (i = 0; i * 32 < dkLen; i++) {
-        /* Generate INT(i + 1). */
-        ivec = bswap_32( i+1 );
-
-        /* Compute U_1 = PRF(P, S || INT(i)). */
-        memcpy(&hctx, &PShctx, sizeof(hmac_yp_ctx));
-        hmac_blake2b_yp_update(&hctx, &ivec, 4);
-        hmac_blake2b_yp_final(&hctx, U);
-
-        /* T_i = U_1 ... */
-        memcpy(T, U, 32);
-
-        for (j = 2; j <= c; j++) {
-            /* Compute U_j. */
-            hmac_blake2b_yp_init(&hctx, passwd, passwdlen);
-            hmac_blake2b_yp_update(&hctx, U, 32);
-            hmac_blake2b_yp_final(&hctx, U);
-
-            /* ... xor U_j ... */
-            for (k = 0; k < 32; k++) {
-                T[k] ^= U[k];
-            }
-        }
-
-        /* Copy as many bytes as necessary into buf. */
-        clen = dkLen - i * 32;
-        if (clen > 32) {
-            clen = 32;
-        }
-
-        memcpy(&buf[i * 32], T, clen);
-    }
-
-    /* Clean PShctx, since we never called _Final on it. */
-    memset(&PShctx, 0, sizeof(hmac_yp_ctx));
-}

algo/yespower/crypto/blake2b-yp.h

@@ -1,42 +0,0 @@
-#pragma once
-#ifndef __BLAKE2B_H__
-#define __BLAKE2B_H__
-
-#include <stddef.h>
-#include <stdint.h>
-
-#if defined(_MSC_VER) || defined(__x86_64__) || defined(__x86__)
-#define NATIVE_LITTLE_ENDIAN
-#endif
-
-// state context
-typedef struct {
-    uint8_t b[128]; // input buffer
-    uint64_t h[8];  // chained state
-    uint64_t t[2];  // total number of bytes
-    size_t c;       // pointer for b[]
-    size_t outlen;  // digest size
-} blake2b_yp_ctx;
-
-typedef struct {
-    blake2b_yp_ctx inner;
-    blake2b_yp_ctx outer;
-} hmac_yp_ctx;
-
-#if defined(__cplusplus)
-extern "C" {
-#endif
-
-int blake2b_yp_init(blake2b_yp_ctx *ctx, size_t outlen, const void *key, size_t keylen);
-void blake2b_yp_update(blake2b_yp_ctx *ctx, const void *in, size_t inlen);
-void blake2b_yp_final(blake2b_yp_ctx *ctx, void *out);
-void blake2b_yp_hash(void *out, const void *in, size_t inlen);
-void hmac_blake2b_yp_hash(void *out, const void *key, size_t keylen, const void *in, size_t inlen);
-void pbkdf2_blake2b_yp(const uint8_t * passwd, size_t passwdlen, const uint8_t * salt,
-    size_t saltlen, uint64_t c, uint8_t * buf, size_t dkLen);
-
-#if defined(__cplusplus)
-}
-#endif
-
-#endif

algo/yespower/crypto/hmac-blake2b.c

@@ -0,0 +1,150 @@
+/*
+ * Copyright 2009 Colin Percival, 2014 savale
+ * All rights reserved.
+ *
+ * Redistribution and use in source and binary forms, with or without
+ * modification, are permitted provided that the following conditions
+ * are met:
+ * 1. Redistributions of source code must retain the above copyright
+ *  notice, this list of conditions and the following disclaimer.
+ * 2. Redistributions in binary form must reproduce the above copyright
+ *  notice, this list of conditions and the following disclaimer in the
+ *  documentation and/or other materials provided with the distribution.
+ *
+ * THIS SOFTWARE IS PROVIDED BY THE AUTHOR AND CONTRIBUTORS ``AS IS'' AND
+ * ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
+ * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
+ * ARE DISCLAIMED.  IN NO EVENT SHALL THE AUTHOR OR CONTRIBUTORS BE LIABLE
+ * FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
+ * DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS
+ * OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
+ * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT
+ * LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY
+ * OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF
+ * SUCH DAMAGE.
+ *
+ * This file was originally written by Colin Percival as part of the Tarsnap
+ * online backup system.
+ */
+
+#include <stdlib.h>
+#include <stdint.h>
+#include <string.h>
+#include "simd-utils.h"
+#include "hmac-blake2b.h"
+
+// keylen = number of bytes
+void hmac_blake2b_init( hmac_blake2b_ctx *hctx, const void *_key,
+                        size_t keylen )
+{
+    const uint8_t *key = _key;
+    uint8_t keyhash[32];
+    uint8_t pad[64];
+    uint64_t i;
+
+    if (keylen > 64)
+    {
+       sph_blake2b_ctx ctx;
+       sph_blake2b_init( &ctx, 32, NULL, 0 );
+       sph_blake2b_update( &ctx, key, keylen );
+       sph_blake2b_final( &ctx, keyhash );
+       key = keyhash;
+       keylen = 32;
+    }
+
+    sph_blake2b_init( &hctx->inner, 32, NULL, 0 );
+    memset( pad, 0x36, 64 );
+    for ( i = 0; i < keylen; ++i )
+        pad[i] ^= key[i];
+
+    sph_blake2b_update( &hctx->inner, pad, 64 );
+    sph_blake2b_init( &hctx->outer, 32, NULL, 0 );
+    memset( pad, 0x5c, 64 );
+    for ( i = 0; i < keylen; ++i )
+        pad[i] ^= key[i];
+
+    sph_blake2b_update( &hctx->outer, pad, 64 );
+    memset( keyhash, 0, 32 );
+}
+
+// datalen = number of bits
+void hmac_blake2b_update( hmac_blake2b_ctx *hctx, const void *data,
+                          size_t datalen )
+{
+    // update the inner state
+    sph_blake2b_update( &hctx->inner, data, datalen );
+}
+
+void hmac_blake2b_final( hmac_blake2b_ctx *hctx, uint8_t *digest )
+{
+    uint8_t ihash[32];
+    sph_blake2b_final( &hctx->inner, ihash );
+    sph_blake2b_update( &hctx->outer, ihash, 32 );
+    sph_blake2b_final( &hctx->outer, digest );
+    memset( ihash, 0, 32 );
+}
+
+// // keylen = number of bytes; inlen = number of bytes
+void hmac_blake2b_hash( void *out, const void *key, size_t keylen,
+                        const void *in, size_t inlen )
+{
+    hmac_blake2b_ctx hctx;
+    hmac_blake2b_init( &hctx, key, keylen );
+    hmac_blake2b_update( &hctx, in, inlen );
+    hmac_blake2b_final( &hctx, out );
+}
+
+void pbkdf2_blake2b( const uint8_t *passwd, size_t passwdlen,
+                     const uint8_t *salt, size_t saltlen, uint64_t c,
+                     uint8_t *buf, size_t dkLen )
+{
+    hmac_blake2b_ctx PShctx, hctx;
+    size_t i;
+    uint32_t ivec;
+    uint8_t U[32];
+    uint8_t T[32];
+    uint64_t j;
+    int k;
+    size_t clen;
+
+    /* Compute HMAC state after processing P and S. */
+    hmac_blake2b_init( &PShctx, passwd, passwdlen );
+    hmac_blake2b_update( &PShctx, salt, saltlen );
+
+    /* Iterate through the blocks. */
+    for ( i = 0; i * 32 < dkLen; i++ )
+    {
+        /* Generate INT(i + 1). */
+        ivec = bswap_32( i+1 );
+
+        /* Compute U_1 = PRF(P, S || INT(i)). */
+        memcpy( &hctx, &PShctx, sizeof(hmac_blake2b_ctx) );
+        hmac_blake2b_update( &hctx, &ivec, 4 );
+        hmac_blake2b_final( &hctx, U );
+
+        /* T_i = U_1 ... */
+        memcpy( T, U, 32 );
+
+        for ( j = 2; j <= c; j++ )
+        {
+            /* Compute U_j. */
+            hmac_blake2b_init( &hctx, passwd, passwdlen );
+            hmac_blake2b_update( &hctx, U, 32 );
+            hmac_blake2b_final( &hctx, U );
+
+            /* ... xor U_j ... */
+            for ( k = 0; k < 32; k++ )
+                T[k] ^= U[k];
+        }
+
+        /* Copy as many bytes as necessary into buf. */
+        clen = dkLen - i * 32;
+        if (clen > 32)
+            clen = 32;
+
+        memcpy( &buf[i * 32], T, clen );
+    }
+
+    /* Clean PShctx, since we never called _Final on it. */
+    memset( &PShctx, 0, sizeof(hmac_blake2b_ctx) );
+}

algo/yespower/crypto/hmac-blake2b.h

@@ -0,0 +1,34 @@
+#pragma once
+#ifndef __HMAC_BLAKE2B_H__
+#define __HMAC_BLAKE2B_H__
+
+#include <stddef.h>
+#include <stdint.h>
+#include "algo/blake/sph_blake2b.h"
+
+#if defined(_MSC_VER) || defined(__x86_64__) || defined(__x86__)
+#define NATIVE_LITTLE_ENDIAN
+#endif
+
+typedef struct
+{
+    sph_blake2b_ctx inner;
+    sph_blake2b_ctx outer;
+} hmac_blake2b_ctx;
+
+#if defined(__cplusplus)
+extern "C" {
+#endif
+
+void hmac_blake2b_hash( void *out, const void *key, size_t keylen,
+                        const void *in, size_t inlen );
+
+void pbkdf2_blake2b( const uint8_t * passwd, size_t passwdlen,
+                     const uint8_t * salt, size_t saltlen, uint64_t c,
+                     uint8_t * buf, size_t dkLen );
+
+#if defined(__cplusplus)
+}
+#endif
+
+#endif

algo/yespower/yespower-blake2b.c

@@ -95,7 +95,7 @@
 #include <stdint.h>
 #include <stdlib.h>
 #include <string.h>
-#include "crypto/blake2b-yp.h"
+#include "crypto/hmac-blake2b.h"
 #include "yespower.h"
 
 #ifdef __unix__
@@ -1136,6 +1136,7 @@ int yespower_b2b(yespower_local_t *local,
     salsa20_blk_t *V, *XY;
     pwxform_ctx_t ctx;
     uint8_t init_hash[32];
+    sph_blake2b_ctx blake2b_ctx;
 
     /* Sanity-check parameters */
     if ((N < 1024 || N > 512 * 1024 || r < 8 || r > 32 ||
@@ -1167,7 +1168,9 @@ int yespower_b2b(yespower_local_t *local,
     ctx.S0 = S;
     ctx.S1 = S + Swidth_to_Sbytes1(Swidth);
 
-    blake2b_yp_hash(init_hash, src, srclen);
+    sph_blake2b_init( &blake2b_ctx, 32, NULL, 0 );
+    sph_blake2b_update( &blake2b_ctx, src, srclen );
+    sph_blake2b_final( &blake2b_ctx, init_hash );
 
     ctx.S2 = S + 2 * Swidth_to_Sbytes1(Swidth);
     ctx.w = 0;
@@ -1181,7 +1184,7 @@ int yespower_b2b(yespower_local_t *local,
 
     if ( work_restart[thrid].restart ) return false;
     
-    pbkdf2_blake2b_yp(init_hash, sizeof(init_hash), src, srclen, 1, B, 128);
+    pbkdf2_blake2b(init_hash, sizeof(init_hash), src, srclen, 1, B, 128);
 
     if ( work_restart[thrid].restart ) return false;
 
@@ -1190,7 +1193,7 @@ int yespower_b2b(yespower_local_t *local,
 
     if ( work_restart[thrid].restart ) return false;
 
-    hmac_blake2b_yp_hash((uint8_t *)dst, B + B_size - 64, 64, init_hash, sizeof(init_hash));
+    hmac_blake2b_hash((uint8_t *)dst, B + B_size - 64, 64, init_hash, sizeof(init_hash));
 
     /* Success! */
     return 1;

algo/yespower/yespower-gate.c

@@ -249,7 +249,7 @@ bool register_power2b_algo( algo_gate_t* gate )
   applog( LOG_NOTICE,"Key= \"%s\"", yespower_params.pers );
   applog( LOG_NOTICE,"Key length= %d\n", yespower_params.perslen );
 
-  gate->optimizations = SSE2_OPT;
+  gate->optimizations = SSE2_OPT | AVX2_OPT;
   gate->scanhash      = (void*)&scanhash_yespower_b2b;
   gate->hash          = (void*)&yespower_b2b_hash;
   opt_target_factor = 65536.0;

configure

@@ -1,6 +1,6 @@
 #! /bin/sh
 # Guess values for system-dependent variables and create Makefiles.
-# Generated by GNU Autoconf 2.69 for cpuminer-opt 3.20.0.
+# Generated by GNU Autoconf 2.69 for cpuminer-opt 3.20.1.
 #
 #
 # Copyright (C) 1992-1996, 1998-2012 Free Software Foundation, Inc.
@@ -577,8 +577,8 @@ MAKEFLAGS=
 # Identity of this package.
 PACKAGE_NAME='cpuminer-opt'
 PACKAGE_TARNAME='cpuminer-opt'
-PACKAGE_VERSION='3.20.0'
-PACKAGE_STRING='cpuminer-opt 3.20.0'
+PACKAGE_VERSION='3.20.1'
+PACKAGE_STRING='cpuminer-opt 3.20.1'
 PACKAGE_BUGREPORT=''
 PACKAGE_URL=''
 
@@ -1332,7 +1332,7 @@ if test "$ac_init_help" = "long"; then
   # Omit some internal or obsolete options to make the list less imposing.
   # This message is too long to be a string in the A/UX 3.1 sh.
   cat <<_ACEOF
-\`configure' configures cpuminer-opt 3.20.0 to adapt to many kinds of systems.
+\`configure' configures cpuminer-opt 3.20.1 to adapt to many kinds of systems.
 
 Usage: $0 [OPTION]... [VAR=VALUE]...
 
@@ -1404,7 +1404,7 @@ fi
 
 if test -n "$ac_init_help"; then
   case $ac_init_help in
-     short | recursive ) echo "Configuration of cpuminer-opt 3.20.0:";;
+     short | recursive ) echo "Configuration of cpuminer-opt 3.20.1:";;
    esac
   cat <<\_ACEOF
 
@@ -1509,7 +1509,7 @@ fi
 test -n "$ac_init_help" && exit $ac_status
 if $ac_init_version; then
   cat <<\_ACEOF
-cpuminer-opt configure 3.20.0
+cpuminer-opt configure 3.20.1
 generated by GNU Autoconf 2.69
 
 Copyright (C) 2012 Free Software Foundation, Inc.
@@ -2012,7 +2012,7 @@ cat >config.log <<_ACEOF
 This file contains any messages produced by compilers while
 running configure, to aid debugging if configure makes a mistake.
 
-It was created by cpuminer-opt $as_me 3.20.0, which was
+It was created by cpuminer-opt $as_me 3.20.1, which was
 generated by GNU Autoconf 2.69.  Invocation command line was
 
   $ $0 $@
@@ -2993,7 +2993,7 @@ fi
 
 # Define the identity of the package.
  PACKAGE='cpuminer-opt'
- VERSION='3.20.0'
+ VERSION='3.20.1'
 
 
 cat >>confdefs.h <<_ACEOF
@@ -5820,6 +5820,34 @@ $as_echo "#define USE_AVX2 1" >>confdefs.h
 
       { $as_echo "$as_me:${as_lineno-$LINENO}: result: yes" >&5
 $as_echo "yes" >&6; }
+      { $as_echo "$as_me:${as_lineno-$LINENO}: checking whether we can compile AVX512 code" >&5
+$as_echo_n "checking whether we can compile AVX512 code... " >&6; }
+      cat confdefs.h - <<_ACEOF >conftest.$ac_ext
+/* end confdefs.h.  */
+
+int
+main ()
+{
+asm ("vpaddd %zmm0, %zmm1, %zmm2{%k1}");
+  ;
+  return 0;
+}
+_ACEOF
+if ac_fn_c_try_compile "$LINENO"; then :
+
+$as_echo "#define USE_AVX512 1" >>confdefs.h
+
+        { $as_echo "$as_me:${as_lineno-$LINENO}: result: yes" >&5
+$as_echo "yes" >&6; }
+
+else
+  { $as_echo "$as_me:${as_lineno-$LINENO}: result: no" >&5
+$as_echo "no" >&6; }
+        { $as_echo "$as_me:${as_lineno-$LINENO}: WARNING: The assembler does not support the AVX512 instruction set." >&5
+$as_echo "$as_me: WARNING: The assembler does not support the AVX512 instruction set." >&2;}
+
+fi
+rm -f core conftest.err conftest.$ac_objext conftest.$ac_ext
 
 else
   { $as_echo "$as_me:${as_lineno-$LINENO}: result: no" >&5
@@ -6690,7 +6718,7 @@ cat >>$CONFIG_STATUS <<\_ACEOF || ac_write_fail=1
 # report actual input values of CONFIG_FILES etc. instead of their
 # values after options handling.
 ac_log="
-This file was extended by cpuminer-opt $as_me 3.20.0, which was
+This file was extended by cpuminer-opt $as_me 3.20.1, which was
 generated by GNU Autoconf 2.69.  Invocation command line was
 
   CONFIG_FILES    = $CONFIG_FILES
@@ -6756,7 +6784,7 @@ _ACEOF
 cat >>$CONFIG_STATUS <<_ACEOF || ac_write_fail=1
 ac_cs_config="`$as_echo "$ac_configure_args" | sed 's/^ //; s/[\\""\`\$]/\\\\&/g'`"
 ac_cs_version="\\
-cpuminer-opt config.status 3.20.0
+cpuminer-opt config.status 3.20.1
 configured by $0, generated by GNU Autoconf 2.69,
   with options \\"\$ac_cs_config\\"
 

configure.ac

@@ -1,4 +1,4 @@
-AC_INIT([cpuminer-opt], [3.20.0])
+AC_INIT([cpuminer-opt], [3.20.1])
 
 AC_PREREQ([2.59c])
 AC_CANONICAL_SYSTEM
@@ -93,6 +93,14 @@ then
     AC_COMPILE_IFELSE([AC_LANG_PROGRAM(,[asm ("vpaddd %ymm0, %ymm1, %ymm2");])],
       AC_DEFINE(USE_AVX2, 1, [Define to 1 if AVX2 assembly is available.])
       AC_MSG_RESULT(yes)
+      AC_MSG_CHECKING(whether we can compile AVX512 code)
+      AC_COMPILE_IFELSE([AC_LANG_PROGRAM(,[asm ("vpaddd %zmm0, %zmm1, %zmm2{%k1}");])],
+        AC_DEFINE(USE_AVX512, 1, [Define to 1 if AVX512 assembly is available.])
+        AC_MSG_RESULT(yes)
+      ,
+        AC_MSG_RESULT(no)
+        AC_MSG_WARN([The assembler does not support the AVX512 instruction set.])
+      )
     ,
       AC_MSG_RESULT(no)
       AC_MSG_WARN([The assembler does not support the AVX2 instruction set.])

cpu-miner.c

@@ -1300,6 +1300,7 @@ static int share_result( int result, struct work *work,
            my_stats.share_count, acol, ares, scol, sres, rcol, rres, bcol,
            bres, CL_N, share_time, latency );
 
+/*   
    if ( unlikely( opt_debug || !result || solved ) )
    {
       if ( have_stratum )
@@ -1309,14 +1310,27 @@ static int share_result( int result, struct work *work,
          applog2( LOG_INFO, "Diff %.5g, Block %d",
                my_stats.share_diff, work ? work->height : last_block_height );
    }
+*/
 
    if ( unlikely( !( opt_quiet || result || stale ) ) )
    {
-      uint32_t str[8];
-      uint32_t *targ;
+//      uint32_t str[8];
+//      uint32_t *targ;
 
-      if ( reason ) applog( LOG_MINR, "Reject reason: %s", reason );
-         
+      if ( reason ) applog2( LOG_MINR, "Reject reason: %s", reason );
+      {
+         // The exact hash is not avaiable here, it's just an imprecise
+         // approximation calculated from the share difficulty. It's useless
+         // for anything other than low diff rejects. Until and unless a
+         // solution is implemented to make the hash and targets avaiable
+         // don't bother displaying them. In the meantime display the diff for
+         // low diff rejects.
+
+         if ( strstr( reason, "difficulty" ) )
+            applog2( LOG_MINR, "Share diff: %.5g, Target: %.5g",
+                               my_stats.share_diff, my_stats.target_diff );
+
+/*
       diff_to_hash( str, my_stats.share_diff );
       applog2( LOG_INFO, "Hash:   %08x%08x%08x%08x%08x%08x", str[7], str[6],
                str[5], str[4], str[3],str[2], str[1], str[0] );
@@ -1330,6 +1344,8 @@ static int share_result( int result, struct work *work,
       }
       applog2( LOG_INFO, "Target: %08x%08x%08x%08x%08x%08x", targ[7], targ[6],
                targ[5], targ[4], targ[3], targ[2], targ[1], targ[0] );
+*/
+      }
    }
    return 1;
 }

simd-utils/simd-128.h

@@ -546,14 +546,13 @@ static inline void mm128_block_bswap_32( __m128i *d, const __m128i *s )
 
 
 // Two input shuffle-rotate.
-// Concatenate v1 & v2 and rotate as one 256 bit vector.
-// Continue to use vror/vrol for now to avoid confusion with
-// shufl2r/shufl2l function macros available with AVX512.
+// Concatenate v1 & v2 and bit rotate as one 256 bit vector.
 
 #if defined(__SSSE3__)
 
-// Function macro with two inputs and one output, inputs are preserved.
-// Two input functions are not available without SSSE3. Use procedure
+// Function macros with two inputs and one output, inputs are preserved.
+// Returns the high 128 bits, ie updated v1.
+// These two-input functions are not available without SSSE3. Use procedure
 // macros below instead.
 
 #define mm128_shufl2r_64( v1, v2 )     _mm_alignr_epi8( v2, v1, 8 )
@@ -568,12 +567,9 @@ static inline void mm128_block_bswap_32( __m128i *d, const __m128i *s )
 #define mm128_shufl2r_8( v1, v2 )      _mm_alignr_epi8( v2, v1, 8 )
 #define mm128_shufl2l_8( v1, v2 )      _mm_alignr_epi8( v1, v2, 8 )
 
-// Procedure macros with 2 inputs and 2 outputs, inputs args are overwritten.
-
-// These macros retain the vrol/vror name for now to avoid
-// confusion with the shufl2r/shuffle2l function macros above.
-// These may be renamed to something like shufl2r2 for 2 nputs and
-// 2 outputs, ie SHUFfLe 2 inputs Right with 2 outputs.
+// Procedure macros with 2 inputs and 2 outputs, input args are overwritten.
+// Deprecated for SSSE3 and above, they exist for SSSE3 only for compatibility
+// with existing code. The function macros above can be used more effciently.
 
 #define mm128_vror256_64( v1, v2 ) \
 do { \