diff --git a/Makefile.am b/Makefile.am index cbd8de4..b98757b 100644 --- a/Makefile.am +++ b/Makefile.am @@ -119,6 +119,7 @@ cpuminer_SOURCES = \ algo/scrypt.c \ algo/scryptjane/scrypt-jane.c \ algo/sha2/sha2.c \ + algo/sha2/sha256t.c \ algo/simd/sse2/nist.c \ algo/simd/sse2/vector.c \ algo/skein/skein.c \ diff --git a/RELEASE_NOTES b/RELEASE_NOTES index 2581bfc..b6c851e 100644 --- a/RELEASE_NOTES +++ b/RELEASE_NOTES @@ -3,11 +3,17 @@ Compile instruction for Linux and Windows are at the bottom of this file. Change Log ---------- +v3.5.12 + +New algo sha256t for Onecoin (OC), 29% faster than ocminer version. +lyra2zoin algo renamed to lyra2z330, lyra2zoin and zoin still work + as aliases. + v3.5.11 -- fixed hmq1725 crash on Ubuntu 16.04 -- fixed compile error in hodl.cpp with gcc 6.3 -- fixed x11 crash on Windows with AVX2 +Fixed hmq1725 crash on Ubuntu 16.04 +Fixed compile error in hodl.cpp with gcc 6.3 +Fixed x11 crash on Windows with AVX2 v3.5.10 diff --git a/algo-gate-api.c b/algo-gate-api.c index 0901e3e..3cd7549 100644 --- a/algo-gate-api.c +++ b/algo-gate-api.c @@ -179,7 +179,7 @@ bool register_algo_gate( int algo, algo_gate_t *gate ) case ALGO_LYRA2RE: register_lyra2re_algo ( gate ); break; case ALGO_LYRA2REV2: register_lyra2rev2_algo ( gate ); break; case ALGO_LYRA2Z: register_zcoin_algo ( gate ); break; - case ALGO_LYRA2ZOIN: register_zoin_algo ( gate ); break; + case ALGO_LYRA2Z330: register_lyra2z330_algo ( gate ); break; case ALGO_M7M: register_m7m_algo ( gate ); break; case ALGO_MYR_GR: register_myriad_algo ( gate ); break; case ALGO_NEOSCRYPT: register_neoscrypt_algo ( gate ); break; @@ -191,6 +191,7 @@ bool register_algo_gate( int algo, algo_gate_t *gate ) case ALGO_SCRYPT: register_scrypt_algo ( gate ); break; case ALGO_SCRYPTJANE: register_scryptjane_algo ( gate ); break; case ALGO_SHA256D: register_sha256d_algo ( gate ); break; + case ALGO_SHA256T: register_sha256t_algo ( gate ); break; case ALGO_SHAVITE3: register_shavite_algo ( gate ); break; case ALGO_SKEIN: register_skein_algo ( gate ); break; case ALGO_SKEIN2: register_skein2_algo ( gate ); break; @@ -281,13 +282,14 @@ const char* const algo_alias_map[][2] = { "jane", "scryptjane" }, { "lyra2", "lyra2re" }, { "lyra2v2", "lyra2rev2" }, + { "lyra2zoin", "lyra2z330" }, { "myriad", "myr-gr" }, { "neo", "neoscrypt" }, { "sib", "x11gost" }, { "yes", "yescrypt" }, { "ziftr", "zr5" }, { "zcoin", "lyra2z" }, - { "zoin", "lyra2zoin" }, + { "zoin", "lyra2z330" }, { NULL, NULL } }; diff --git a/algo-gate-api.h b/algo-gate-api.h index 316cd1e..26bdd75 100644 --- a/algo-gate-api.h +++ b/algo-gate-api.h @@ -130,7 +130,6 @@ void ( *build_extraheader ) ( struct work*, struct stratum_ctx* ); void ( *build_stratum_request ) ( char*, struct work*, struct stratum_ctx* ); void ( *set_work_data_endian ) ( struct work* ); double ( *calc_network_diff ) ( struct work* ); -//bool ( *prevent_dupes ) ( struct work*, struct stratum_ctx*, int ); bool ( *ready_to_mine ) ( struct work*, struct stratum_ctx*, int ); void ( *resync_threads ) ( struct work* ); bool ( *do_this_thread ) ( int ); diff --git a/algo/groestl/aes_ni/hash-groestl.c b/algo/groestl/aes_ni/hash-groestl.c index c364609..db44820 100644 --- a/algo/groestl/aes_ni/hash-groestl.c +++ b/algo/groestl/aes_ni/hash-groestl.c @@ -189,7 +189,7 @@ HashReturn_gr update_and_final_groestl( hashState_groestl* ctx, void* output, int rem = ctx->rem_ptr; int blocks = len / SIZE512; __m128i* in = (__m128i*)input; - int i, i0; + int i; // --- update --- diff --git a/algo/luffa/sse2/luffa_for_sse2.c b/algo/luffa/sse2/luffa_for_sse2.c index c69f78f..627b746 100644 --- a/algo/luffa/sse2/luffa_for_sse2.c +++ b/algo/luffa/sse2/luffa_for_sse2.c @@ -38,6 +38,15 @@ _mm256_set_epi32( 0, 0, 0, 0, 0, 0, 0, 0x00800800 ) ) ) #endif // __AVX2__ + +#define MULT2(a0,a1) do \ +{ \ + __m128i b = _mm_xor_si128( a0, _mm_shuffle_epi32( _mm_and_si128(a1,MASK), 16 ) ); \ + a0 = _mm_or_si128( _mm_srli_si128(b,4), _mm_slli_si128(a1,12) ); \ + a1 = _mm_or_si128( _mm_srli_si128(a1,4), _mm_slli_si128(b,12) ); \ +} while(0) + +/* #define MULT2(a0,a1) do \ { \ __m128i b; \ @@ -46,6 +55,7 @@ _mm256_set_epi32( 0, 0, 0, 0, 0, 0, 0, 0x00800800 ) ) ) a0 = _mm_or_si128( _mm_srli_si128(a0,4), _mm_slli_si128(a1,12) ); \ a1 = _mm_or_si128( _mm_srli_si128(a1,4), _mm_slli_si128(b,12) ); \ } while(0) +*/ #define STEP_PART(x,c,t)\ SUBCRUMB(*x,*(x+1),*(x+2),*(x+3),*t);\ diff --git a/algo/lyra2/zoin.c b/algo/lyra2/zoin.c index 33f990a..05be115 100644 --- a/algo/lyra2/zoin.c +++ b/algo/lyra2/zoin.c @@ -84,7 +84,7 @@ bool zoin_thread_init() return true; } -bool register_zoin_algo( algo_gate_t* gate ) +bool register_lyra2z330_algo( algo_gate_t* gate ) { gate->optimizations = SSE2_OPT | AES_OPT | AVX_OPT | AVX2_OPT; gate->miner_thread_init = (void*)&zoin_thread_init; diff --git a/algo/sha2/sha256t.c b/algo/sha2/sha256t.c new file mode 100644 index 0000000..6a8818f --- /dev/null +++ b/algo/sha2/sha256t.c @@ -0,0 +1,143 @@ +#include "miner.h" +#include "algo-gate-api.h" + +#include +#include +#include +#include + +#include "sph-sha2.h" + +//#define DEBUG_ALGO + + +static sph_sha256_context sha256t_ctx __attribute__ ((aligned (64))); +static __thread sph_sha256_context sha256t_mid __attribute__ ((aligned (64))); + +void sha256t_midstate( const void* input ) +{ + memcpy( &sha256t_mid, &sha256t_ctx, sizeof sha256t_mid ); + sph_sha256( &sha256t_mid, input, 64 ); +} + +void sha256t_hash(void* output, const void* input, uint32_t len) +{ + sph_sha256_context ctx_sha256 __attribute__ ((aligned (64))); + uint32_t _ALIGN(64) hashA[16]; + + const int midlen = 64; // bytes + const int tail = 80 - midlen; // 16 + + memcpy( &ctx_sha256, &sha256t_mid, sizeof sha256t_mid ); + sph_sha256( &ctx_sha256, input + midlen, tail ); + +// sph_sha256_init(&ctx_sha256); +// sph_sha256 (&ctx_sha256, input, 80); + sph_sha256_close( &ctx_sha256, hashA ); + +// sph_sha256_init(&ctx_sha256); + memcpy( &ctx_sha256, &sha256t_ctx, sizeof sha256t_ctx ); + sph_sha256( &ctx_sha256, hashA, 32 ); + sph_sha256_close( &ctx_sha256, hashA ); + +// sph_sha256_init(&ctx_sha256); + memcpy( &ctx_sha256, &sha256t_ctx, sizeof sha256t_ctx ); + sph_sha256( &ctx_sha256, hashA, 32 ); + sph_sha256_close( &ctx_sha256, hashA ); + + memcpy( output, hashA, 32 ); +} + +int scanhash_sha256t(int thr_id, struct work *work, + uint32_t max_nonce, uint64_t *hashes_done) +{ + uint32_t *pdata = work->data; + uint32_t *ptarget = work->target; + uint32_t len = 80; + + uint32_t n = pdata[19] - 1; + const uint32_t first_nonce = pdata[19]; + const uint32_t Htarg = ptarget[7]; +#ifdef _MSC_VER + uint32_t __declspec(align(32)) hash64[8]; +#else + uint32_t hash64[8] __attribute__((aligned(32))); +#endif + uint32_t endiandata[32]; + + uint64_t htmax[] = { + 0, + 0xF, + 0xFF, + 0xFFF, + 0xFFFF, + 0x10000000 + }; + uint32_t masks[] = { + 0xFFFFFFFF, + 0xFFFFFFF0, + 0xFFFFFF00, + 0xFFFFF000, + 0xFFFF0000, + 0 + }; + + // we need bigendian data... + for (int k = 0; k < 19; k++) + be32enc(&endiandata[k], pdata[k]); + + sha256t_midstate( endiandata ); + +#ifdef DEBUG_ALGO + if (Htarg != 0) + printf("[%d] Htarg=%X\n", thr_id, Htarg); +#endif + for (int m=0; m < 6; m++) { + if (Htarg <= htmax[m]) { + uint32_t mask = masks[m]; + do { + pdata[19] = ++n; + be32enc(&endiandata[19], n); + sha256t_hash(hash64, endiandata, len); +#ifndef DEBUG_ALGO + if ((!(hash64[7] & mask)) && fulltest(hash64, ptarget)) { + *hashes_done = n - first_nonce + 1; + return true; + } +#else + if (!(n % 0x1000) && !thr_id) printf("."); + if (!(hash64[7] & mask)) { + printf("[%d]",thr_id); + if (fulltest(hash64, ptarget)) { + *hashes_done = n - first_nonce + 1; + return true; + } + } +#endif + } while (n < max_nonce && !work_restart[thr_id].restart); + // see blake.c if else to understand the loop on htmax => mask + break; + } + } + + *hashes_done = n - first_nonce + 1; + pdata[19] = n; + return 0; +} + +void sha256t_set_target( struct work* work, double job_diff ) +{ + work_set_target( work, job_diff / (256.0 * opt_diff_factor) ); +} + + +bool register_sha256t_algo( algo_gate_t* gate ) +{ + sph_sha256_init( &sha256t_ctx ); + gate->scanhash = (void*)&scanhash_sha256t; + gate->hash = (void*)&sha256t_hash; + gate->hash_alt = (void*)&sha256t_hash; + gate->set_target = (void*)&sha256t_set_target; + gate->get_max64 = (void*)&get_max64_0x3ffff; + return true; +} diff --git a/configure.ac b/configure.ac index f082631..2e35f36 100644 --- a/configure.ac +++ b/configure.ac @@ -1,4 +1,4 @@ -AC_INIT([cpuminer-opt], [3.5.11]) +AC_INIT([cpuminer-opt], [3.5.12]) AC_PREREQ([2.59c]) AC_CANONICAL_SYSTEM diff --git a/miner.h b/miner.h index 5a1bea3..089bff6 100644 --- a/miner.h +++ b/miner.h @@ -499,7 +499,7 @@ enum algos { ALGO_LYRA2RE, ALGO_LYRA2REV2, ALGO_LYRA2Z, - ALGO_LYRA2ZOIN, + ALGO_LYRA2Z330, ALGO_M7M, ALGO_MYR_GR, ALGO_NEOSCRYPT, @@ -511,6 +511,7 @@ enum algos { ALGO_SCRYPT, ALGO_SCRYPTJANE, ALGO_SHA256D, + ALGO_SHA256T, ALGO_SHAVITE3, ALGO_SKEIN, ALGO_SKEIN2, @@ -559,7 +560,7 @@ static const char* const algo_names[] = { "lyra2re", "lyra2rev2", "lyra2z", - "lyra2zoin", + "lyra2z330", "m7m", "myr-gr", "neoscrypt", @@ -571,6 +572,7 @@ static const char* const algo_names[] = { "scrypt", "scryptjane", "sha256d", + "sha256t", "shavite3", "skein", "skein2", @@ -674,7 +676,7 @@ Options:\n\ lyra2re lyra2\n\ lyra2rev2 lyrav2, Vertcoin\n\ lyra2z Zcoin (XZC)\n\ - lyra2zoin Zoin (ZOI)\n\ + lyra2z330 Zoin (ZOI)\n\ m7m Magi (XMG)\n\ myr-gr Myriad-Groestl\n\ neoscrypt NeoScrypt(128, 2, 1)\n\ @@ -686,7 +688,8 @@ Options:\n\ scrypt scrypt(1024, 1, 1) (default)\n\ scrypt:N scrypt(N, 1, 1)\n\ scryptjane:nf\n\ - sha256d SHA-256d\n\ + sha256d Double SHA-256\n\ + sha256t Triple SHA-256, Onecoin (OC)\n\ shavite3 Shavite3\n\ skein Skein+Sha (Skeincoin)\n\ skein2 Double Skein (Woodcoin)\n\ diff --git a/winbuild-allarch.sh b/winbuild-allarch.sh new file mode 100755 index 0000000..bbb3d81 --- /dev/null +++ b/winbuild-allarch.sh @@ -0,0 +1,45 @@ +#!/bin/bash + +make distclean || echo clean +rm -f config.status +./autogen.sh || echo done +CFLAGS="-O3 -march=core-avx2 -Wall" CXXFLAGS="$CFLAGS -std=gnu++11 -fpermissive" ./configure --with-curl +make -j 4 +strip -s cpuminer.exe +mv cpuminer.exe cpuminer-aes-avx2.exe + +make clean || echo clean +rm -f config.status +./autogen.sh || echo done +CFLAGS="-O3 -march=corei7-avx -Wall" CXXFLAGS="$CFLAGS -std=gnu++11 -fpermissive" ./configure --with-curl +make -j 4 +strip -s cpuminer.exe +mv cpuminer.exe cpuminer-aes-avx.exe + +make clean || echo clean +rm -f config.status +./autogen.sh || echo done +CFLAGS="-O3 -maes -msse4.2 -Wall" CXXFLAGS="$CFLAGS -std=gnu++11 -fpermissive" ./configure --with-curl +make -j 4 +strip -s cpuminer.exe +mv cpuminer.exe cpuminer-aes-sse42.exe + +make clean || echo clean +rm -f config.status +./autogen.sh || echo done +CFLAGS="-O3 -march=corei7 -Wall" CXXFLAGS="$CFLAGS -std=gnu++11 -fpermissive" ./configure --with-curl +make -j 4 +strip -s cpuminer.exe +mv cpuminer.exe cpuminer-sse42.exe + +make clean || echo clean +rm -f config.status +./autogen.sh || echo done +CFLAGS="-O3 -march=core2 -Wall" CXXFLAGS="$CFLAGS -std=gnu++11 -fpermissive" ./configure --with-curl +make -j 4 +strip -s cpuminer.exe +mv cpuminer.exe cpuminer-sse2.exe + +make clean || echo done + +