/* * Copyright 2013 Advanced Micro Devices, Inc. * * 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 (including the next * paragraph) 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. * * Authors: Tom Stellard * */ char *source = "// This file is taken and modified from the public-domain poclbm project, and\n" "// I have therefore decided to keep it public-domain.\n" "// Modified version copyright 2011-2012 Con Kolivas\n" "#ifdef VECTORS4\n" " typedef uint4 u;\n" "#elif defined VECTORS2\n" " typedef uint2 u;\n" "#else\n" " typedef uint u;\n" "#endif\n" "__constant uint K[64] = { \n" " 0x428a2f98, 0x71374491, 0xb5c0fbcf, 0xe9b5dba5, 0x3956c25b, 0x59f111f1, 0x923f82a4, 0xab1c5ed5,\n" " 0xd807aa98, 0x12835b01, 0x243185be, 0x550c7dc3, 0x72be5d74, 0x80deb1fe, 0x9bdc06a7, 0xc19bf174,\n" " 0xe49b69c1, 0xefbe4786, 0x0fc19dc6, 0x240ca1cc, 0x2de92c6f, 0x4a7484aa, 0x5cb0a9dc, 0x76f988da,\n" " 0x983e5152, 0xa831c66d, 0xb00327c8, 0xbf597fc7, 0xc6e00bf3, 0xd5a79147, 0x06ca6351, 0x14292967,\n" " 0x27b70a85, 0x2e1b2138, 0x4d2c6dfc, 0x53380d13, 0x650a7354, 0x766a0abb, 0x81c2c92e, 0x92722c85,\n" " 0xa2bfe8a1, 0xa81a664b, 0xc24b8b70, 0xc76c51a3, 0xd192e819, 0xd6990624, 0xf40e3585, 0x106aa070,\n" " 0x19a4c116, 0x1e376c08, 0x2748774c, 0x34b0bcb5, 0x391c0cb3, 0x4ed8aa4a, 0x5b9cca4f, 0x682e6ff3,\n" " 0x748f82ee, 0x78a5636f, 0x84c87814, 0x8cc70208, 0x90befffa, 0xa4506ceb, 0xbef9a3f7, 0xc67178f2\n" "};\n" "__constant uint ConstW[128] = {\n" "0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x80000000U, 0x00000000, 0x00000000, 0x00000000,\n" "0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000280U,\n" "0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000,\n" "0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000,\n" "0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000,\n" "0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000,\n" "0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000,\n" "0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000,\n" "0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000,\n" "0x80000000U, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000100U,\n" "0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000,\n" "0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000,\n" "0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000,\n" "0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000,\n" "0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000,\n" "0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000\n" "};\n" "__constant uint H[8] = { \n" " 0x6a09e667, 0xbb67ae85, 0x3c6ef372, 0xa54ff53a, 0x510e527f, 0x9b05688c, 0x1f83d9ab, 0x5be0cd19\n" "};\n" "#ifdef BITALIGN\n" " #pragma OPENCL EXTENSION cl_amd_media_ops : enable\n" " #define rot(x, y) amd_bitalign(x, x, (uint)(32 - y))\n" "// This part is not from the stock poclbm kernel. It's part of an optimization\n" "// added in the Phoenix Miner.\n" "// Some AMD devices have Vals[0] BFI_INT opcode, which behaves exactly like the\n" "// SHA-256 Ch function, but provides it in exactly one instruction. If\n" "// detected, use it for Ch. Otherwise, construct Ch out of simpler logical\n" "// primitives.\n" " #ifdef BFI_INT\n" " // Well, slight problem... It turns out BFI_INT isn't actually exposed to\n" " // OpenCL (or CAL IL for that matter) in any way. However, there is \n" " // a similar instruction, BYTE_ALIGN_INT, which is exposed to OpenCL via\n" " // amd_bytealign, takes the same inputs, and provides the same output. \n" " // We can use that as a placeholder for BFI_INT and have the application \n" " // patch it after compilation.\n" " \n" " // This is the BFI_INT function\n" " #define Ch(x, y, z) amd_bytealign(x,y,z)\n" " // Ma can also be implemented in terms of BFI_INT...\n" " #define Ma(z, x, y) amd_bytealign(z^x,y,x)\n" " #else // BFI_INT\n" " // Later SDKs optimise this to BFI INT without patching and GCN\n" " // actually fails if manually patched with BFI_INT\n" " #define Ch(x, y, z) bitselect((u)z, (u)y, (u)x)\n" " #define Ma(x, y, z) bitselect((u)x, (u)y, (u)z ^ (u)x)\n" " #define rotr(x, y) amd_bitalign((u)x, (u)x, (u)y)\n" " #endif\n" "#else // BITALIGN\n" " #define Ch(x, y, z) (z ^ (x & (y ^ z)))\n" " #define Ma(x, y, z) ((x & z) | (y & (x | z)))\n" " #define rot(x, y) rotate(x, y)\n" " #define rotr(x, y) rotate(x, (32-y))\n" "#endif\n" "//Various intermediate calculations for each SHA round\n" "#define s0(n) (S0(Vals[(0 + 128 - (n)) % 8]))\n" "#define S0(n) (rot(n, 30u)^rot(n, 19u)^rot(n,10u))\n" "#define s1(n) (S1(Vals[(4 + 128 - (n)) % 8]))\n" "#define S1(n) (rot(n, 26u)^rot(n, 21u)^rot(n, 7u))\n" "#define ch(n) Ch(Vals[(4 + 128 - (n)) % 8],Vals[(5 + 128 - (n)) % 8],Vals[(6 + 128 - (n)) % 8])\n" "#define maj(n) Ma(Vals[(1 + 128 - (n)) % 8],Vals[(2 + 128 - (n)) % 8],Vals[(0 + 128 - (n)) % 8])\n" "//t1 calc when W is already calculated\n" "#define t1(n) K[(n) % 64] + Vals[(7 + 128 - (n)) % 8] + W[(n)] + s1(n) + ch(n) \n" "//t1 calc which calculates W\n" "#define t1W(n) K[(n) % 64] + Vals[(7 + 128 - (n)) % 8] + W(n) + s1(n) + ch(n)\n" "//Used for constant W Values (the compiler optimizes out zeros)\n" "#define t1C(n) (K[(n) % 64]+ ConstW[(n)]) + Vals[(7 + 128 - (n)) % 8] + s1(n) + ch(n)\n" "//t2 Calc\n" "#define t2(n) maj(n) + s0(n)\n" "#define rotC(x,n) (x<> (32-n))\n" "//W calculation used for SHA round\n" "#define W(n) (W[n] = P4(n) + P3(n) + P2(n) + P1(n))\n" "//Partial W calculations (used for the begining where only some values are nonzero)\n" "#define P1(n) ((rot(W[(n)-2],15u)^rot(W[(n)-2],13u)^((W[(n)-2])>>10U)))\n" "#define P2(n) ((rot(W[(n)-15],25u)^rot(W[(n)-15],14u)^((W[(n)-15])>>3U)))\n" "#define p1(x) ((rot(x,15u)^rot(x,13u)^((x)>>10U)))\n" "#define p2(x) ((rot(x,25u)^rot(x,14u)^((x)>>3U)))\n" "#define P3(n) W[n-7]\n" "#define P4(n) W[n-16]\n" "//Partial Calcs for constant W values\n" "#define P1C(n) ((rotC(ConstW[(n)-2],15)^rotC(ConstW[(n)-2],13)^((ConstW[(n)-2])>>10U)))\n" "#define P2C(n) ((rotC(ConstW[(n)-15],25)^rotC(ConstW[(n)-15],14)^((ConstW[(n)-15])>>3U)))\n" "#define P3C(x) ConstW[x-7]\n" "#define P4C(x) ConstW[x-16]\n" "//SHA round with built in W calc\n" "#define sharoundW(n) Barrier1(n); Vals[(3 + 128 - (n)) % 8] += t1W(n); Vals[(7 + 128 - (n)) % 8] = t1W(n) + t2(n); \n" "//SHA round without W calc\n" "#define sharound(n) Barrier2(n); Vals[(3 + 128 - (n)) % 8] += t1(n); Vals[(7 + 128 - (n)) % 8] = t1(n) + t2(n);\n" "//SHA round for constant W values\n" "#define sharoundC(n) Barrier3(n); Vals[(3 + 128 - (n)) % 8] += t1C(n); Vals[(7 + 128 - (n)) % 8] = t1C(n) + t2(n);\n" "//The compiler is stupid... I put this in there only to stop the compiler from (de)optimizing the order\n" "#define Barrier1(n) t1 = t1C((n+1))\n" "#define Barrier2(n) t1 = t1C((n))\n" "#define Barrier3(n) t1 = t1C((n))\n" "//#define WORKSIZE 256\n" "#define MAXBUFFERS (4095)\n" "__kernel \n" " __attribute__((reqd_work_group_size(WORKSIZE, 1, 1)))\n" "void search( const uint state0, const uint state1, const uint state2, const uint state3,\n" " const uint state4, const uint state5, const uint state6, const uint state7,\n" " const uint B1, const uint C1, const uint D1,\n" " const uint F1, const uint G1, const uint H1,\n" " const u base,\n" " const uint W16, const uint W17,\n" " const uint PreVal4, const uint PreVal0,\n" " const uint PreW18, const uint PreW19,\n" " const uint PreW31, const uint PreW32,\n" " \n" " volatile __global uint * output)\n" "{\n" " u W[124];\n" " u Vals[8];\n" "//Dummy Variable to prevent compiler from reordering between rounds\n" " u t1;\n" " //Vals[0]=state0;\n" " Vals[1]=B1;\n" " Vals[2]=C1;\n" " Vals[3]=D1;\n" " //Vals[4]=PreVal4;\n" " Vals[5]=F1;\n" " Vals[6]=G1;\n" " Vals[7]=H1;\n" " W[16] = W16;\n" " W[17] = W17;\n" "#ifdef VECTORS4\n" " //Less dependencies to get both the local id and group id and then add them\n" " W[3] = base + (uint)(get_local_id(0)) * 4u + (uint)(get_group_id(0)) * (WORKSIZE * 4u);\n" " uint r = rot(W[3].x,25u)^rot(W[3].x,14u)^((W[3].x)>>3U);\n" " //Since only the 2 LSB is opposite between the nonces, we can save an instruction by flipping the 4 bits in W18 rather than the 1 bit in W3\n" " W[18] = PreW18 + (u){r, r ^ 0x2004000U, r ^ 0x4008000U, r ^ 0x600C000U};\n" "#elif defined VECTORS2\n" " W[3] = base + (uint)(get_local_id(0)) * 2u + (uint)(get_group_id(0)) * (WORKSIZE * 2u);\n" " uint r = rot(W[3].x,25u)^rot(W[3].x,14u)^((W[3].x)>>3U);\n" " W[18] = PreW18 + (u){r, r ^ 0x2004000U};\n" "#else\n" " W[3] = base + get_local_id(0) + get_group_id(0) * (WORKSIZE);\n" " u r = rot(W[3],25u)^rot(W[3],14u)^((W[3])>>3U);\n" " W[18] = PreW18 + r;\n" "#endif\n" " //the order of the W calcs and Rounds is like this because the compiler needs help finding how to order the instructions\n" " Vals[4] = PreVal4 + W[3];\n" " Vals[0] = PreVal0 + W[3];\n" " sharoundC(4);\n" " W[19] = PreW19 + W[3];\n" " sharoundC(5);\n" " W[20] = P4C(20) + P1(20);\n" " sharoundC(6);\n" " W[21] = P1(21);\n" " sharoundC(7);\n" " W[22] = P3C(22) + P1(22);\n" " sharoundC(8);\n" " W[23] = W[16] + P1(23);\n" " sharoundC(9);\n" " W[24] = W[17] + P1(24);\n" " sharoundC(10);\n" " W[25] = P1(25) + P3(25);\n" " W[26] = P1(26) + P3(26);\n" " sharoundC(11);\n" " W[27] = P1(27) + P3(27);\n" " W[28] = P1(28) + P3(28);\n" " sharoundC(12);\n" " W[29] = P1(29) + P3(29);\n" " sharoundC(13);\n" " W[30] = P1(30) + P2C(30) + P3(30);\n" " W[31] = PreW31 + (P1(31) + P3(31));\n" " sharoundC(14);\n" " W[32] = PreW32 + (P1(32) + P3(32));\n" " sharoundC(15);\n" " sharound(16);\n" " sharound(17);\n" " sharound(18);\n" " sharound(19);\n" " sharound(20);\n" " sharound(21);\n" " sharound(22);\n" " sharound(23);\n" " sharound(24);\n" " sharound(25);\n" " sharound(26);\n" " sharound(27);\n" " sharound(28);\n" " sharound(29);\n" " sharound(30);\n" " sharound(31);\n" " sharound(32);\n" " sharoundW(33);\n" " sharoundW(34);\n" " sharoundW(35);\n" " sharoundW(36);\n" " sharoundW(37); \n" " sharoundW(38);\n" " sharoundW(39);\n" " sharoundW(40);\n" " sharoundW(41);\n" " sharoundW(42);\n" " sharoundW(43);\n" " sharoundW(44);\n" " sharoundW(45);\n" " sharoundW(46);\n" " sharoundW(47);\n" " sharoundW(48);\n" " sharoundW(49);\n" " sharoundW(50);\n" " sharoundW(51);\n" " sharoundW(52);\n" " sharoundW(53);\n" " sharoundW(54);\n" " sharoundW(55);\n" " sharoundW(56);\n" " sharoundW(57);\n" " sharoundW(58);\n" " sharoundW(59);\n" " sharoundW(60);\n" " sharoundW(61);\n" " sharoundW(62);\n" " sharoundW(63);\n" " W[64]=state0+Vals[0];\n" " W[65]=state1+Vals[1];\n" " W[66]=state2+Vals[2];\n" " W[67]=state3+Vals[3];\n" " W[68]=state4+Vals[4];\n" " W[69]=state5+Vals[5];\n" " W[70]=state6+Vals[6];\n" " W[71]=state7+Vals[7];\n" " Vals[0]=H[0];\n" " Vals[1]=H[1];\n" " Vals[2]=H[2];\n" " Vals[3]=H[3];\n" " Vals[4]=H[4];\n" " Vals[5]=H[5];\n" " Vals[6]=H[6];\n" " Vals[7]=H[7];\n" " //sharound(64 + 0);\n" " const u Temp = (0xb0edbdd0U + K[0]) + W[64];\n" " Vals[7] = Temp + 0x08909ae5U;\n" " Vals[3] = 0xa54ff53aU + Temp;\n" " \n" "#define P124(n) P2(n) + P1(n) + P4(n)\n" " W[64 + 16] = + P2(64 + 16) + P4(64 + 16);\n" " sharound(64 + 1);\n" " W[64 + 17] = P1C(64 + 17) + P2(64 + 17) + P4(64 + 17);\n" " sharound(64 + 2);\n" " W[64 + 18] = P124(64 + 18);\n" " sharound(64 + 3);\n" " W[64 + 19] = P124(64 + 19);\n" " sharound(64 + 4);\n" " W[64 + 20] = P124(64 + 20);\n" " sharound(64 + 5);\n" " W[64 + 21] = P124(64 + 21);\n" " sharound(64 + 6);\n" " W[64 + 22] = P4(64 + 22) + P3C(64 + 22) + P2(64 + 22) + P1(64 + 22);\n" " sharound(64 + 7);\n" " W[64 + 23] = P4(64 + 23) + P3(64 + 23) + P2C(64 + 23) + P1(64 + 23);\n" " sharoundC(64 + 8);\n" " W[64 + 24] = P1(64 + 24) + P4C(64 + 24) + P3(64 + 24);\n" " sharoundC(64 + 9);\n" " W[64 + 25] = P3(64 + 25) + P1(64 + 25);\n" " sharoundC(64 + 10);\n" " W[64 + 26] = P3(64 + 26) + P1(64 + 26);\n" " sharoundC(64 + 11);\n" " W[64 + 27] = P3(64 + 27) + P1(64 + 27);\n" " sharoundC(64 + 12);\n" " W[64 + 28] = P3(64 + 28) + P1(64 + 28);\n" " sharoundC(64 + 13);\n" " W[64 + 29] = P1(64 + 29) + P3(64 + 29);\n" " W[64 + 30] = P3(64 + 30) + P2C(64 + 30) + P1(64 + 30);\n" " sharoundC(64 + 14);\n" " W[64 + 31] = P4C(64 + 31) + P3(64 + 31) + P2(64 + 31) + P1(64 + 31);\n" " sharoundC(64 + 15);\n" " sharound(64 + 16);\n" " sharound(64 + 17);\n" " sharound(64 + 18);\n" " sharound(64 + 19);\n" " sharound(64 + 20);\n" " sharound(64 + 21);\n" " sharound(64 + 22);\n" " sharound(64 + 23);\n" " sharound(64 + 24);\n" " sharound(64 + 25);\n" " sharound(64 + 26);\n" " sharound(64 + 27);\n" " sharound(64 + 28);\n" " sharound(64 + 29);\n" " sharound(64 + 30);\n" " sharound(64 + 31);\n" " sharoundW(64 + 32);\n" " sharoundW(64 + 33);\n" " sharoundW(64 + 34);\n" " sharoundW(64 + 35);\n" " sharoundW(64 + 36);\n" " sharoundW(64 + 37);\n" " sharoundW(64 + 38);\n" " sharoundW(64 + 39);\n" " sharoundW(64 + 40);\n" " sharoundW(64 + 41);\n" " sharoundW(64 + 42);\n" " sharoundW(64 + 43);\n" " sharoundW(64 + 44);\n" " sharoundW(64 + 45);\n" " sharoundW(64 + 46);\n" " sharoundW(64 + 47);\n" " sharoundW(64 + 48);\n" " sharoundW(64 + 49);\n" " sharoundW(64 + 50);\n" " sharoundW(64 + 51);\n" " sharoundW(64 + 52);\n" " sharoundW(64 + 53);\n" " sharoundW(64 + 54);\n" " sharoundW(64 + 55);\n" " sharoundW(64 + 56);\n" " sharoundW(64 + 57);\n" " sharoundW(64 + 58);\n" " W[117] += W[108] + Vals[3] + Vals[7] + P2(124) + P1(124) + Ch((Vals[0] + Vals[4]) + (K[59] + W(59+64)) + s1(64+59)+ ch(59+64),Vals[1],Vals[2]) -\n" " (-(K[60] + H[7]) - S1((Vals[0] + Vals[4]) + (K[59] + W(59+64)) + s1(64+59)+ ch(59+64)));\n" "#define FOUND (0x0F)\n" "#define SETFOUND(Xnonce) output[output[FOUND]++] = Xnonce\n" "#ifdef VECTORS4\n" " bool result = W[117].x & W[117].y & W[117].z & W[117].w;\n" " if (!result) {\n" " if (!W[117].x)\n" " SETFOUND(W[3].x);\n" " if (!W[117].y)\n" " SETFOUND(W[3].y);\n" " if (!W[117].z)\n" " SETFOUND(W[3].z);\n" " if (!W[117].w)\n" " SETFOUND(W[3].w);\n" " }\n" "#elif defined VECTORS2\n" " bool result = W[117].x & W[117].y;\n" " if (!result) {\n" " if (!W[117].x)\n" " SETFOUND(W[3].x);\n" " if (!W[117].y)\n" " SETFOUND(W[3].y);\n" " }\n" "#else\n" " if (!W[117])\n" " SETFOUND(W[3]);\n" "#endif\n" "}\n"; #include "piglit-util.h" #include "piglit-framework-cl-program.h" #define EXPECTED_NONCE 1322941352 PIGLIT_CL_PROGRAM_TEST_CONFIG_BEGIN config.name = "Bitcoin phatk kernel"; config.program_source = source; config.kernel_name = "search"; config.build_options = "-D WORKSIZE=256"; config.run_per_device = true; PIGLIT_CL_PROGRAM_TEST_CONFIG_END enum piglit_result piglit_cl_test(const int argc, const char** argv, const struct piglit_cl_program_test_config* config, const struct piglit_cl_program_test_env* env) { unsigned i; enum piglit_result result = PIGLIT_PASS; size_t global_size = 32768; size_t local_size = 256; unsigned data[16]; cl_mem buffer = NULL; static const unsigned args[] = { 451767447, 953606276, 3289282709, 1043157133, 3980456748, 2761019621, 3674259583, 4041779413, 1887291860, 3618940936, 4220496293, 2530934726, 204837088, 3991048658, 1322909696, 2820830137, 1604534223, 3318230743, 591787567, 1627450019, 151637226, 4206519064, 47933991, }; /* Create CL objects */ memset(data, 0, sizeof(data)); buffer = piglit_cl_create_buffer(env->context, CL_MEM_WRITE_ONLY, sizeof(data)); /* Initialize the buffer */ piglit_cl_write_whole_buffer(env->context->command_queues[0], buffer, data); /* Set the kernel args */ for (i = 0; i < 23; i++) { piglit_cl_set_kernel_arg(env->kernel, i, 4, &args[i]); } piglit_cl_set_kernel_buffer_arg(env->kernel, 23, &buffer); piglit_cl_execute_ND_range_kernel(env->context->command_queues[0], env->kernel, 1, NULL, &global_size, &local_size); piglit_cl_read_whole_buffer(env->context->command_queues[0], buffer, data); if (data[0] != EXPECTED_NONCE) { fprintf(stderr, "output[0] expected %u, but got %u\n", EXPECTED_NONCE, data[0]); result = PIGLIT_FAIL; } for (i = 1; i < 15; i++) { if (data[i] != 0) { fprintf(stderr, "output[%u] expected %u, but got %u\n", i, 0, data[i]); result = PIGLIT_FAIL; } } if (data[15] != 1) { fprintf(stderr, "output[15] expected 1, but got %u\n", data[15]); result = PIGLIT_FAIL; } return result; }