jakoddannudau/rig
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity

Plain C "cn/r" implementation.

Browse Source
This commit is contained in:
XMRig 2019-03-03 20:19:17 +07:00
parent 5e6a69e16f
commit 27980f24f8
11 changed files with 559 additions and 26 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

1
CMakeLists.txt
View File

@ -11,6 +11,7 @@ set(HEADERS
algo/cryptonight/cryptonight_monero.h
algo/cryptonight/cryptonight_softaes.h
algo/cryptonight/cryptonight_test.h
algo/cryptonight/variant4_random_math.h
compat.h
cpu.h
donate.h

1
algo/cryptonight/cryptonight.h
View File

@ -68,6 +68,7 @@ struct cryptonight_ctx {
cn_mainloop_double_fun_ms_abi generated_code_double;
struct cryptonight_r_data generated_code_data;
struct cryptonight_r_data generated_code_double_data;
uint64_t height;
};

52
algo/cryptonight/cryptonight_monero.h
View File

@ -6,8 +6,8 @@
* Copyright 2016 Jay D Dee <jayddee246@gmail.com>
* Copyright 2017-2018 XMR-Stak <https://github.com/fireice-uk>, <https://github.com/psychocrypt>
* Copyright 2018 Lee Clagett <https://github.com/vtnerd>
* Copyright 2018 SChernykh <https://github.com/SChernykh>
* Copyright 2016-2018 XMRig <https://github.com/xmrig>, <support@xmrig.com>
* Copyright 2018-2019 SChernykh <https://github.com/SChernykh>
* Copyright 2016-2019 XMRig <https://github.com/xmrig>, <support@xmrig.com>
*
* This program is free software: you can redistribute it and/or modify
* it under the terms of the GNU General Public License as published by
@ -29,6 +29,8 @@
#include <fenv.h>
#include <math.h>
#include <stdint.h>
#include <x86intrin.h>
static inline __m128i int_sqrt_v2(const uint64_t n0)
@ -87,6 +89,17 @@ static inline __m128i int_sqrt_v2(const uint64_t n0)
_mm_store_si128((__m128i *)((base_ptr) + ((offset) ^ 0x30)), _mm_add_epi64(chunk2, _a)); \
}
# define VARIANT4_SHUFFLE(base_ptr, offset, _a, _b, _b1, _c) \
{ \
const __m128i chunk1 = _mm_load_si128((__m128i *)((base_ptr) + ((offset) ^ 0x10))); \
const __m128i chunk2 = _mm_load_si128((__m128i *)((base_ptr) + ((offset) ^ 0x20))); \
const __m128i chunk3 = _mm_load_si128((__m128i *)((base_ptr) + ((offset) ^ 0x30))); \
_mm_store_si128((__m128i *)((base_ptr) + ((offset) ^ 0x10)), _mm_add_epi64(chunk3, _b1)); \
_mm_store_si128((__m128i *)((base_ptr) + ((offset) ^ 0x20)), _mm_add_epi64(chunk1, _b)); \
_mm_store_si128((__m128i *)((base_ptr) + ((offset) ^ 0x30)), _mm_add_epi64(chunk2, _a)); \
_c = _mm_xor_si128(_mm_xor_si128(_c, chunk3), _mm_xor_si128(chunk1, chunk2)); \
}
# define VARIANT2_SHUFFLE2(base_ptr, offset, _a, _b, _b1, hi, lo) \
{ \
const __m128i chunk1 = _mm_xor_si128(_mm_load_si128((__m128i *)((base_ptr) + ((offset) ^ 0x10))), _mm_set_epi64x(lo, hi)); \
@ -99,4 +112,39 @@ static inline __m128i int_sqrt_v2(const uint64_t n0)
_mm_store_si128((__m128i *)((base_ptr) + ((offset) ^ 0x30)), _mm_add_epi64(chunk2, _a)); \
}
#ifndef NOINLINE
#ifdef __GNUC__
#define NOINLINE __attribute__ ((noinline))
#elif _MSC_VER
#define NOINLINE __declspec(noinline)
#else
#define NOINLINE
#endif
#endif
#include "variant4_random_math.h"
#define VARIANT4_RANDOM_MATH_INIT(part) \
uint32_t r##part[9]; \
struct V4_Instruction code##part[256]; \
{ \
r##part[0] = (uint32_t)(h##part[12]); \
r##part[1] = (uint32_t)(h##part[12] >> 32); \
r##part[2] = (uint32_t)(h##part[13]); \
r##part[3] = (uint32_t)(h##part[13] >> 32); \
} \
v4_random_math_init(code##part, ctx[part]->height);
#define VARIANT4_RANDOM_MATH(part, al, ah, cl, bx0, bx1) \
{ \
cl ^= (r##part[0] + r##part[1]) | ((uint64_t)(r##part[2] + r##part[3]) << 32); \
r##part[4] = (uint32_t)(al); \
r##part[5] = (uint32_t)(ah); \
r##part[6] = (uint32_t)(_mm_cvtsi128_si32(bx0)); \
r##part[7] = (uint32_t)(_mm_cvtsi128_si32(bx1)); \
r##part[8] = (uint32_t)(_mm_cvtsi128_si32(_mm_srli_si128(bx1, 8))); \
v4_random_math(code##part, r##part); \
}
#endif /* XMRIG_CRYPTONIGHT_MONERO_H */

10
algo/cryptonight/cryptonight_r_av1.c
View File

@ -44,6 +44,7 @@ void cryptonight_r_av1(const uint8_t *restrict input, size_t size, uint8_t *rest
VARIANT2_INIT(0);
VARIANT2_SET_ROUNDING_MODE();
VARIANT4_RANDOM_MATH_INIT(0);
uint64_t al0 = h0[0] ^ h0[4];
uint64_t ah0 = h0[1] ^ h0[5];
@ -58,7 +59,7 @@ void cryptonight_r_av1(const uint8_t *restrict input, size_t size, uint8_t *rest
cx = _mm_aesenc_si128(cx, ax0);
VARIANT2_SHUFFLE(l0, idx0 & 0x1FFFF0, ax0, bx0, bx1);
VARIANT4_SHUFFLE(l0, idx0 & 0x1FFFF0, ax0, bx0, bx1, cx);
_mm_store_si128((__m128i *) &l0[idx0 & 0x1FFFF0], _mm_xor_si128(bx0, cx));
idx0 = _mm_cvtsi128_si64(cx);
@ -67,9 +68,12 @@ void cryptonight_r_av1(const uint8_t *restrict input, size_t size, uint8_t *rest
cl = ((uint64_t*) &l0[idx0 & 0x1FFFF0])[0];
ch = ((uint64_t*) &l0[idx0 & 0x1FFFF0])[1];
VARIANT2_INTEGER_MATH(0, cl, cx);
VARIANT4_RANDOM_MATH(0, al0, ah0, cl, bx0, bx1);
al0 ^= r0[2] | ((uint64_t)(r0[3]) << 32);
ah0 ^= r0[0] | ((uint64_t)(r0[1]) << 32);
lo = _umul128(idx0, cl, &hi);
VARIANT2_SHUFFLE2(l0, idx0 & 0x1FFFF0, ax0, bx0, bx1, hi, lo);
VARIANT4_SHUFFLE(l0, idx0 & 0x1FFFF0, ax0, bx0, bx1, cx);
al0 += hi;
ah0 += lo;

20
algo/cryptonight/cryptonight_r_av2.c
View File

@ -46,6 +46,8 @@ void cryptonight_r_av2(const uint8_t *restrict input, size_t size, uint8_t *rest
VARIANT2_INIT(0);
VARIANT2_INIT(1);
VARIANT2_SET_ROUNDING_MODE();
VARIANT4_RANDOM_MATH_INIT(0);
VARIANT4_RANDOM_MATH_INIT(1);
cn_explode_scratchpad((__m128i*) h0, (__m128i*) l0);
cn_explode_scratchpad((__m128i*) h1, (__m128i*) l1);
@ -73,10 +75,10 @@ void cryptonight_r_av2(const uint8_t *restrict input, size_t size, uint8_t *rest
cx0 = _mm_aesenc_si128(cx0, ax0);
cx1 = _mm_aesenc_si128(cx1, ax1);
VARIANT2_SHUFFLE(l0, idx0 & 0x1FFFF0, ax0, bx00, bx01);
VARIANT4_SHUFFLE(l0, idx0 & 0x1FFFF0, ax0, bx00, bx01, cx0);
_mm_store_si128((__m128i *) &l0[idx0 & 0x1FFFF0], _mm_xor_si128(bx00, cx0));
VARIANT2_SHUFFLE(l1, idx1 & 0x1FFFF0, ax1, bx10, bx11);
VARIANT4_SHUFFLE(l1, idx1 & 0x1FFFF0, ax1, bx10, bx11, cx1);
_mm_store_si128((__m128i *) &l1[idx1 & 0x1FFFF0], _mm_xor_si128(bx10, cx1));
idx0 = _mm_cvtsi128_si64(cx0);
@ -86,9 +88,12 @@ void cryptonight_r_av2(const uint8_t *restrict input, size_t size, uint8_t *rest
cl = ((uint64_t*) &l0[idx0 & 0x1FFFF0])[0];
ch = ((uint64_t*) &l0[idx0 & 0x1FFFF0])[1];
VARIANT2_INTEGER_MATH(0, cl, cx0);
VARIANT4_RANDOM_MATH(0, al0, ah0, cl, bx00, bx01);
al0 ^= r0[2] | ((uint64_t)(r0[3]) << 32);
ah0 ^= r0[0] | ((uint64_t)(r0[1]) << 32);
lo = _umul128(idx0, cl, &hi);
VARIANT2_SHUFFLE2(l0, idx0 & 0x1FFFF0, ax0, bx00, bx01, hi, lo);
VARIANT4_SHUFFLE(l0, idx0 & 0x1FFFF0, ax0, bx00, bx01, cx0);
al0 += hi;
ah0 += lo;
@ -103,9 +108,12 @@ void cryptonight_r_av2(const uint8_t *restrict input, size_t size, uint8_t *rest
cl = ((uint64_t*) &l1[idx1 & 0x1FFFF0])[0];
ch = ((uint64_t*) &l1[idx1 & 0x1FFFF0])[1];
VARIANT2_INTEGER_MATH(1, cl, cx1);
VARIANT4_RANDOM_MATH(1, al1, ah1, cl, bx10, bx11);
al1 ^= r1[2] | ((uint64_t)(r1[3]) << 32);
ah1 ^= r1[0] | ((uint64_t)(r1[1]) << 32);
lo = _umul128(idx1, cl, &hi);
VARIANT2_SHUFFLE2(l1, idx1 & 0x1FFFF0, ax1, bx10, bx11, hi, lo);
VARIANT4_SHUFFLE(l1, idx1 & 0x1FFFF0, ax1, bx10, bx11, cx1);
al1 += hi;
ah1 += lo;

10
algo/cryptonight/cryptonight_r_av3.c
View File

@ -44,6 +44,7 @@ void cryptonight_r_av3(const uint8_t *restrict input, size_t size, uint8_t *rest
VARIANT2_INIT(0);
VARIANT2_SET_ROUNDING_MODE();
VARIANT4_RANDOM_MATH_INIT(0);
uint64_t al0 = h0[0] ^ h0[4];
uint64_t ah0 = h0[1] ^ h0[5];
@ -58,7 +59,7 @@ void cryptonight_r_av3(const uint8_t *restrict input, size_t size, uint8_t *rest
cx = soft_aesenc(cx, ax0);
VARIANT2_SHUFFLE(l0, idx0 & 0x1FFFF0, ax0, bx0, bx1);
VARIANT4_SHUFFLE(l0, idx0 & 0x1FFFF0, ax0, bx0, bx1, cx);
_mm_store_si128((__m128i *) &l0[idx0 & 0x1FFFF0], _mm_xor_si128(bx0, cx));
idx0 = _mm_cvtsi128_si64(cx);
@ -67,9 +68,12 @@ void cryptonight_r_av3(const uint8_t *restrict input, size_t size, uint8_t *rest
cl = ((uint64_t*) &l0[idx0 & 0x1FFFF0])[0];
ch = ((uint64_t*) &l0[idx0 & 0x1FFFF0])[1];
VARIANT2_INTEGER_MATH(0, cl, cx);
VARIANT4_RANDOM_MATH(0, al0, ah0, cl, bx0, bx1);
al0 ^= r0[2] | ((uint64_t)(r0[3]) << 32);
ah0 ^= r0[0] | ((uint64_t)(r0[1]) << 32);
lo = _umul128(idx0, cl, &hi);
VARIANT2_SHUFFLE2(l0, idx0 & 0x1FFFF0, ax0, bx0, bx1, hi, lo);
VARIANT4_SHUFFLE(l0, idx0 & 0x1FFFF0, ax0, bx0, bx1, cx);
al0 += hi;
ah0 += lo;

20
algo/cryptonight/cryptonight_r_av4.c
View File

@ -46,6 +46,8 @@ void cryptonight_r_av4(const uint8_t *restrict input, size_t size, uint8_t *rest
VARIANT2_INIT(0);
VARIANT2_INIT(1);
VARIANT2_SET_ROUNDING_MODE();
VARIANT4_RANDOM_MATH_INIT(0);
VARIANT4_RANDOM_MATH_INIT(1);
cn_explode_scratchpad((__m128i*) h0, (__m128i*) l0);
cn_explode_scratchpad((__m128i*) h1, (__m128i*) l1);
@ -73,10 +75,10 @@ void cryptonight_r_av4(const uint8_t *restrict input, size_t size, uint8_t *rest
cx0 = soft_aesenc(cx0, ax0);
cx1 = soft_aesenc(cx1, ax1);
VARIANT2_SHUFFLE(l0, idx0 & 0x1FFFF0, ax0, bx00, bx01);
VARIANT4_SHUFFLE(l0, idx0 & 0x1FFFF0, ax0, bx00, bx01, cx0);
_mm_store_si128((__m128i *) &l0[idx0 & 0x1FFFF0], _mm_xor_si128(bx00, cx0));
VARIANT2_SHUFFLE(l1, idx1 & 0x1FFFF0, ax1, bx10, bx11);
VARIANT4_SHUFFLE(l1, idx1 & 0x1FFFF0, ax1, bx10, bx11, cx1);
_mm_store_si128((__m128i *) &l1[idx1 & 0x1FFFF0], _mm_xor_si128(bx10, cx1));
idx0 = _mm_cvtsi128_si64(cx0);
@ -86,9 +88,12 @@ void cryptonight_r_av4(const uint8_t *restrict input, size_t size, uint8_t *rest
cl = ((uint64_t*) &l0[idx0 & 0x1FFFF0])[0];
ch = ((uint64_t*) &l0[idx0 & 0x1FFFF0])[1];
VARIANT2_INTEGER_MATH(0, cl, cx0);
VARIANT4_RANDOM_MATH(0, al0, ah0, cl, bx00, bx01);
al0 ^= r0[2] | ((uint64_t)(r0[3]) << 32);
ah0 ^= r0[0] | ((uint64_t)(r0[1]) << 32);
lo = _umul128(idx0, cl, &hi);
VARIANT2_SHUFFLE2(l0, idx0 & 0x1FFFF0, ax0, bx00, bx01, hi, lo);
VARIANT4_SHUFFLE(l0, idx0 & 0x1FFFF0, ax0, bx00, bx01, cx0);
al0 += hi;
ah0 += lo;
@ -103,9 +108,12 @@ void cryptonight_r_av4(const uint8_t *restrict input, size_t size, uint8_t *rest
cl = ((uint64_t*) &l1[idx1 & 0x1FFFF0])[0];
ch = ((uint64_t*) &l1[idx1 & 0x1FFFF0])[1];
VARIANT2_INTEGER_MATH(1, cl, cx1);
VARIANT4_RANDOM_MATH(1, al1, ah1, cl, bx10, bx11);
al1 ^= r1[2] | ((uint64_t)(r1[3]) << 32);
ah1 ^= r1[0] | ((uint64_t)(r1[1]) << 32);
lo = _umul128(idx1, cl, &hi);
VARIANT2_SHUFFLE2(l1, idx1 & 0x1FFFF0, ax1, bx10, bx11, hi, lo);
VARIANT4_SHUFFLE(l1, idx1 & 0x1FFFF0, ax1, bx10, bx11, cx1);
al1 += hi;
ah1 += lo;

448
algo/cryptonight/variant4_random_math.h Normal file
View File

File diff suppressed because it is too large Load Diff

7
stratum.c
View File

@ -4,8 +4,9 @@
* Copyright 2014 Lucas Jones <https://github.com/lucasjones>
* Copyright 2014-2016 Wolf9466 <https://github.com/OhGodAPet>
* Copyright 2016 Jay D Dee <jayddee246@gmail.com>
* Copyright 2016-2017 XMRig <support@xmrig.com>
*
* Copyright 2017-2018 XMR-Stak <https://github.com/fireice-uk>, <https://github.com/psychocrypt>
* Copyright 2018-2019 SChernykh <https://github.com/SChernykh>
* Copyright 2016-2019 XMRig <https://github.com/xmrig>, <support@xmrig.com>
*
* This program is free software: you can redistribute it and/or modify
* it under the terms of the GNU General Public License as published by
@ -709,6 +710,8 @@ static bool job_decode(const json_t *job) {
memset(work.job_id, 0, sizeof(work.job_id));
memcpy(work.job_id, job_id, strlen(job_id));
work.height = (uint64_t) json_integer_value(json_object_get(job, "height"));
return true;
}

6
stratum.h
View File

@ -4,8 +4,9 @@
* Copyright 2014 Lucas Jones <https://github.com/lucasjones>
* Copyright 2014-2016 Wolf9466 <https://github.com/OhGodAPet>
* Copyright 2016 Jay D Dee <jayddee246@gmail.com>
* Copyright 2016-2017 XMRig <support@xmrig.com>
*
* Copyright 2017-2018 XMR-Stak <https://github.com/fireice-uk>, <https://github.com/psychocrypt>
* Copyright 2018-2019 SChernykh <https://github.com/SChernykh>
* Copyright 2016-2019 XMRig <https://github.com/xmrig>, <support@xmrig.com>
*
* This program is free software: you can redistribute it and/or modify
* it under the terms of the GNU General Public License as published by
@ -42,6 +43,7 @@ struct work {
uint32_t target __attribute__((aligned(16)));
uint32_t hash[8] __attribute__((aligned(16)));
char job_id[64] __attribute__((aligned(16)));
uint64_t height;
};

10
xmrig.c
View File

@ -4,8 +4,9 @@
* Copyright 2014 Lucas Jones <https://github.com/lucasjones>
* Copyright 2014-2016 Wolf9466 <https://github.com/OhGodAPet>
* Copyright 2016 Jay D Dee <jayddee246@gmail.com>
* Copyright 2016-2017 XMRig <support@xmrig.com>
*
* Copyright 2017-2018 XMR-Stak <https://github.com/fireice-uk>, <https://github.com/psychocrypt>
* Copyright 2018-2019 SChernykh <https://github.com/SChernykh>
* Copyright 2016-2019 XMRig <https://github.com/xmrig>, <support@xmrig.com>
*
* This program is free software: you can redistribute it and/or modify
* it under the terms of the GNU General Public License as published by
@ -306,6 +307,8 @@ static void *miner_thread(void *userdata) {
struct timeval tv_start;
gettimeofday(&tv_start, NULL);
persistentctx[0]->height = work.height;
/* scan nonces for a proof-of-work hash */
const int rc = scanhash_cryptonight(thr_id, hash, (uint8_t *) work.blob, work.blob_size, work.target, max_nonce, &hashes_done, persistentctx);
stats_add_hashes(thr_id, &tv_start, hashes_done);
@ -391,6 +394,9 @@ static void *miner_thread_double(void *userdata) {
struct timeval tv_start;
gettimeofday(&tv_start, NULL);
persistentctx[0]->height = work.height;
persistentctx[1]->height = work.height;
/* scan nonces for a proof-of-work hash */
const int rc = scanhash_cryptonight_double(thr_id, (uint32_t *) double_hash, double_blob, work.blob_size, work.target, max_nonce, &hashes_done, persistentctx);
stats_add_hashes(thr_id, &tv_start, hashes_done);