[635] | 1 | /* |
---|
| 2 | * sort.c - Parallel sort |
---|
| 3 | * |
---|
| 4 | * Author Cesar Fuguet Tortolero (2013) |
---|
| 5 | * Alain Greiner (2019) |
---|
| 6 | * |
---|
| 7 | * Copyright (c) UPMC Sorbonne Universites |
---|
| 8 | * |
---|
| 9 | * This is free software; you can redistribute it and/or modify it |
---|
| 10 | * under the terms of the GNU General Public License as published by |
---|
| 11 | * the Free Software Foundation; version 2.0 of the License. |
---|
| 12 | * |
---|
| 13 | * It is distributed in the hope that it will be useful, but |
---|
| 14 | * WITHOUT ANY WARRANTY; without even the implied warranty of |
---|
| 15 | * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU |
---|
| 16 | * General Public License for more details. |
---|
| 17 | * |
---|
| 18 | * You should have received a copy of the GNU General Public License |
---|
| 19 | * along with ALMOS-MKH; if not, write to the Free Software Foundation, |
---|
| 20 | * Inc., 51 Franklin St, Fifth Floor, Boston, MA 02110-1301 USA |
---|
| 21 | */ |
---|
| 22 | |
---|
[417] | 23 | /////////////////////////////////////////////////////////////////////////////// |
---|
[635] | 24 | // This multi-threaded application implement a multi-stage sort. |
---|
| 25 | // It has been writen by Cesar Fuget Tortolero in 2013. |
---|
| 26 | // It has been ported on ALMOS-MKH by Alain Greiner in 2019. |
---|
| 27 | // |
---|
[417] | 28 | // There is one thread per physical cores. |
---|
| 29 | // Computation is organised as a binary tree: |
---|
| 30 | // - All threads execute in parallel a buble sort on a sub-array during the |
---|
| 31 | // the first stage of parallel sort, |
---|
| 32 | // - The number of participating threads is divided by 2 at each next stage, |
---|
| 33 | // to make a merge sort, on two subsets of previous stage. |
---|
| 34 | // |
---|
| 35 | // Number_of_stages = number of barriers = log2(Number_of_threads) |
---|
| 36 | // |
---|
[635] | 37 | // The various stages are separated by synchronisation barriers, and the |
---|
| 38 | // main thread uses the join syscall to check that all threads completed |
---|
| 39 | // before printing the computation time (sequencial & parallel). |
---|
| 40 | // These results can be - optionnaly - registered in an instrumentation file. |
---|
| 41 | // |
---|
[417] | 42 | // Constraints : |
---|
| 43 | // - It supports up to 1024 cores: x_size, y_size, and ncores must be |
---|
| 44 | // power of 2 (max 16*16 clusters / max 4 cores per cluster) |
---|
| 45 | // _ The array of values to be sorted (ARRAY_LENGTH) must be power of 2 |
---|
| 46 | // larger than the number of cores. |
---|
| 47 | /////////////////////////////////////////////////////////////////////////////// |
---|
| 48 | |
---|
| 49 | #include <stdio.h> |
---|
| 50 | #include <stdlib.h> |
---|
[596] | 51 | #include <unistd.h> |
---|
[417] | 52 | #include <pthread.h> |
---|
[445] | 53 | #include <almosmkh.h> |
---|
[459] | 54 | #include <hal_macros.h> |
---|
[417] | 55 | |
---|
[636] | 56 | #define ARRAY_LENGTH 2048 // number of items |
---|
[588] | 57 | #define MAX_THREADS 1024 // 16 * 16 * 4 |
---|
[440] | 58 | |
---|
[623] | 59 | #define USE_DQT_BARRIER 1 // use DQT barrier if non zero |
---|
| 60 | #define DISPLAY_ARRAY 0 // display items values before and after |
---|
[635] | 61 | #define DEBUG_MAIN 0 // trace main function |
---|
| 62 | #define DEBUG_SORT 0 // trace sort function |
---|
[623] | 63 | #define CHECK_RESULT 0 // for debug |
---|
| 64 | #define INSTRUMENTATION 1 // register computation times on file |
---|
| 65 | |
---|
[417] | 66 | ///////////////////////////////////////////////////////////// |
---|
| 67 | // argument for the sort() function (one thread per core) |
---|
| 68 | ///////////////////////////////////////////////////////////// |
---|
| 69 | |
---|
| 70 | typedef struct |
---|
| 71 | { |
---|
[619] | 72 | unsigned int threads; // total number of threads |
---|
[417] | 73 | unsigned int thread_uid; // thread user index (0 to threads -1) |
---|
| 74 | unsigned int main_uid; // main thread user index |
---|
| 75 | } |
---|
| 76 | args_t; |
---|
| 77 | |
---|
| 78 | ////////////////////////////////////////// |
---|
| 79 | // Global variables |
---|
| 80 | ////////////////////////////////////////// |
---|
| 81 | |
---|
| 82 | int array0[ARRAY_LENGTH]; // values to sort |
---|
| 83 | int array1[ARRAY_LENGTH]; |
---|
| 84 | |
---|
| 85 | pthread_barrier_t barrier; // synchronisation variables |
---|
| 86 | |
---|
[635] | 87 | pthread_t trdid[MAX_THREADS]; // kernel identifiers |
---|
| 88 | pthread_attr_t attr[MAX_THREADS]; // thread attributes |
---|
| 89 | args_t arg[MAX_THREADS]; // sort function arguments |
---|
[417] | 90 | |
---|
| 91 | //////////////////////////////////// |
---|
[588] | 92 | static void bubbleSort( int * array, |
---|
| 93 | unsigned int length, |
---|
| 94 | unsigned int init_pos ) |
---|
[417] | 95 | { |
---|
[473] | 96 | unsigned int i; |
---|
| 97 | unsigned int j; |
---|
| 98 | int aux; |
---|
[417] | 99 | |
---|
| 100 | for(i = 0; i < length; i++) |
---|
| 101 | { |
---|
| 102 | for(j = init_pos; j < (init_pos + length - i - 1); j++) |
---|
| 103 | { |
---|
| 104 | if(array[j] > array[j + 1]) |
---|
| 105 | { |
---|
| 106 | aux = array[j + 1]; |
---|
| 107 | array[j + 1] = array[j]; |
---|
| 108 | array[j] = aux; |
---|
| 109 | } |
---|
| 110 | } |
---|
| 111 | } |
---|
| 112 | } // end bubbleSort() |
---|
| 113 | |
---|
| 114 | |
---|
[588] | 115 | /////////////////////////////////// |
---|
[623] | 116 | static void merge( const int * src, // source array |
---|
| 117 | int * dst, // destination array |
---|
| 118 | int length, // number of items in a subset |
---|
| 119 | int init_pos_src_a, // index first item in src subset A |
---|
| 120 | int init_pos_src_b, // index first item in src subset B |
---|
| 121 | int init_pos_dst ) // index first item in destination |
---|
[417] | 122 | { |
---|
| 123 | int i; |
---|
| 124 | int j; |
---|
| 125 | int k; |
---|
| 126 | |
---|
| 127 | i = 0; |
---|
| 128 | j = 0; |
---|
| 129 | k = init_pos_dst; |
---|
| 130 | |
---|
| 131 | while((i < length) || (j < length)) |
---|
| 132 | { |
---|
| 133 | if((i < length) && (j < length)) |
---|
| 134 | { |
---|
| 135 | if(src[init_pos_src_a + i] < src[init_pos_src_b + j]) |
---|
| 136 | { |
---|
| 137 | dst[k++] = src[init_pos_src_a + i]; |
---|
| 138 | i++; |
---|
| 139 | } |
---|
| 140 | else |
---|
| 141 | { |
---|
| 142 | dst[k++] = src[init_pos_src_b + j]; |
---|
| 143 | j++; |
---|
| 144 | } |
---|
| 145 | } |
---|
| 146 | else if(i < length) |
---|
| 147 | { |
---|
| 148 | dst[k++] = src[init_pos_src_a + i]; |
---|
| 149 | i++; |
---|
| 150 | } |
---|
| 151 | else |
---|
| 152 | { |
---|
| 153 | dst[k++] = src[init_pos_src_b + j]; |
---|
| 154 | j++; |
---|
| 155 | } |
---|
| 156 | } |
---|
| 157 | } // end merge() |
---|
| 158 | |
---|
[588] | 159 | ////////////////////////////////////// |
---|
[504] | 160 | static void sort( const args_t * ptr ) |
---|
[417] | 161 | { |
---|
| 162 | unsigned int i; |
---|
| 163 | unsigned long long cycle; |
---|
[440] | 164 | unsigned int cxy; |
---|
| 165 | unsigned int lid; |
---|
[417] | 166 | |
---|
[623] | 167 | int * src_array = NULL; |
---|
| 168 | int * dst_array = NULL; |
---|
[417] | 169 | |
---|
[440] | 170 | // get core coordinates an date |
---|
| 171 | get_core( &cxy , &lid ); |
---|
| 172 | get_cycle( &cycle ); |
---|
| 173 | |
---|
[417] | 174 | unsigned int thread_uid = ptr->thread_uid; |
---|
| 175 | unsigned int threads = ptr->threads; |
---|
| 176 | unsigned int main_uid = ptr->main_uid; |
---|
| 177 | |
---|
[623] | 178 | #if DISPLAY_ARRAY |
---|
| 179 | unsigned int n; |
---|
| 180 | if( thread_uid == main_uid ) |
---|
| 181 | { |
---|
| 182 | printf("\n*** array before sort\n"); |
---|
| 183 | for( n=0; n<ARRAY_LENGTH; n++) printf("array[%d] = %d\n", n , array0[n] ); |
---|
| 184 | } |
---|
| 185 | #endif |
---|
| 186 | |
---|
| 187 | ///////////////////////////////// |
---|
| 188 | pthread_barrier_wait( &barrier ); |
---|
| 189 | |
---|
[627] | 190 | #if DEBUG_SORT |
---|
| 191 | if( thread_uid == 0 ) |
---|
[623] | 192 | printf("\n[sort] thread[%d] exit barrier 0\n", thread_uid ); |
---|
| 193 | #endif |
---|
| 194 | |
---|
[417] | 195 | unsigned int items = ARRAY_LENGTH / threads; |
---|
| 196 | unsigned int stages = __builtin_ctz( threads ) + 1; |
---|
| 197 | |
---|
[627] | 198 | #if DEBUG_SORT |
---|
| 199 | if( thread_uid == 0 ) |
---|
[623] | 200 | printf("\n[sort] thread[%d] : start\n", thread_uid ); |
---|
| 201 | #endif |
---|
[442] | 202 | |
---|
[417] | 203 | bubbleSort( array0, items, items * thread_uid ); |
---|
| 204 | |
---|
[627] | 205 | #if DEBUG_SORT |
---|
| 206 | if( thread_uid == 0 ) |
---|
[623] | 207 | printf("\n[sort] thread[%d] : stage 0 completed\n", thread_uid ); |
---|
| 208 | #endif |
---|
[417] | 209 | |
---|
| 210 | ///////////////////////////////// |
---|
| 211 | pthread_barrier_wait( &barrier ); |
---|
[440] | 212 | |
---|
[627] | 213 | #if DEBUG_SORT |
---|
| 214 | if( thread_uid == 0 ) |
---|
[623] | 215 | printf("\n[sort] thread[%d] exit barrier 0\n", thread_uid ); |
---|
| 216 | #endif |
---|
| 217 | |
---|
| 218 | #if DISPLAY_ARRAY |
---|
| 219 | if( thread_uid == main_uid ) |
---|
| 220 | { |
---|
| 221 | printf("\n*** array after bubble sort\n"); |
---|
| 222 | for( n=0; n<ARRAY_LENGTH; n++) printf("array[%d] = %d\n", n , array0[n] ); |
---|
| 223 | } |
---|
| 224 | #endif |
---|
| 225 | |
---|
| 226 | // the number of threads contributing to sort is divided by 2 |
---|
| 227 | // and the number of items is multiplied by 2 at each next stage |
---|
[417] | 228 | for ( i = 1 ; i < stages ; i++ ) |
---|
| 229 | { |
---|
[623] | 230 | if((i % 2) == 1) // odd stage |
---|
| 231 | { |
---|
| 232 | src_array = array0; |
---|
| 233 | dst_array = array1; |
---|
| 234 | } |
---|
| 235 | else // even stage |
---|
| 236 | { |
---|
| 237 | src_array = array1; |
---|
| 238 | dst_array = array0; |
---|
| 239 | } |
---|
[417] | 240 | |
---|
| 241 | if( (thread_uid & ((1<<i)-1)) == 0 ) |
---|
| 242 | { |
---|
| 243 | |
---|
[627] | 244 | #if DEBUG_SORT |
---|
| 245 | if( thread_uid == 0 ) |
---|
[623] | 246 | printf("\n[sort] thread[%d] : stage %d start\n", thread_uid , i ); |
---|
| 247 | #endif |
---|
[417] | 248 | merge( src_array, |
---|
| 249 | dst_array, |
---|
[623] | 250 | items << (i-1), |
---|
[417] | 251 | items * thread_uid, |
---|
| 252 | items * (thread_uid + (1 << (i-1))), |
---|
| 253 | items * thread_uid ); |
---|
| 254 | |
---|
[627] | 255 | #if DEBUG_SORT |
---|
| 256 | if( thread_uid == 0 ) |
---|
[623] | 257 | printf("\n[sort] thread[%d] : stage %d completed\n", thread_uid , i ); |
---|
| 258 | #endif |
---|
[417] | 259 | } |
---|
| 260 | |
---|
| 261 | ///////////////////////////////// |
---|
| 262 | pthread_barrier_wait( &barrier ); |
---|
| 263 | |
---|
[627] | 264 | #if DEBUG_SORT |
---|
| 265 | if( thread_uid == 0 ) |
---|
[623] | 266 | printf("\n[sort] thread[%d] exit barrier %d\n", thread_uid , i ); |
---|
| 267 | #endif |
---|
[417] | 268 | |
---|
[623] | 269 | #if DISPLAY_ARRAY |
---|
| 270 | if( thread_uid == main_uid ) |
---|
| 271 | { |
---|
| 272 | printf("\n*** array after merge %d\n", i ); |
---|
| 273 | for( n=0; n<ARRAY_LENGTH; n++) printf("array[%d] = %d\n", n , dst_array[n] ); |
---|
| 274 | } |
---|
| 275 | #endif |
---|
| 276 | |
---|
| 277 | } // en for stages |
---|
| 278 | |
---|
[417] | 279 | // all threads but the main thread exit |
---|
| 280 | if( thread_uid != main_uid ) pthread_exit( NULL ); |
---|
| 281 | |
---|
| 282 | } // end sort() |
---|
| 283 | |
---|
| 284 | |
---|
[574] | 285 | ///////////////// |
---|
[475] | 286 | void main( void ) |
---|
[417] | 287 | { |
---|
[619] | 288 | int error; |
---|
[417] | 289 | unsigned int x_size; // number of rows |
---|
| 290 | unsigned int y_size; // number of columns |
---|
| 291 | unsigned int ncores; // number of cores per cluster |
---|
[619] | 292 | unsigned int total_threads; // total number of threads |
---|
[417] | 293 | unsigned int thread_uid; // user defined thread index |
---|
| 294 | unsigned int main_cxy; // cluster identifier for main |
---|
| 295 | unsigned int main_x; // X coordinate for main thread |
---|
| 296 | unsigned int main_y; // Y coordinate for main thread |
---|
| 297 | unsigned int main_lid; // core local index for main thread |
---|
| 298 | unsigned int main_uid; // thread user index for main thread |
---|
| 299 | unsigned int x; // X coordinate for a thread |
---|
| 300 | unsigned int y; // Y coordinate for a thread |
---|
| 301 | unsigned int lid; // core local index for a thread |
---|
| 302 | unsigned int n; // index in array to sort |
---|
| 303 | pthread_barrierattr_t barrier_attr; // barrier attributes |
---|
| 304 | |
---|
[623] | 305 | unsigned long long start_cycle; |
---|
| 306 | unsigned long long seq_end_cycle; |
---|
| 307 | unsigned long long para_end_cycle; |
---|
| 308 | |
---|
| 309 | ///////////////////////// |
---|
| 310 | get_cycle( &start_cycle ); |
---|
| 311 | |
---|
[596] | 312 | // compute number of threads (one thread per core) |
---|
[417] | 313 | get_config( &x_size , &y_size , &ncores ); |
---|
[619] | 314 | total_threads = x_size * y_size * ncores; |
---|
[417] | 315 | |
---|
| 316 | // get core coordinates and user index for the main thread |
---|
| 317 | get_core( &main_cxy , & main_lid ); |
---|
[459] | 318 | main_x = HAL_X_FROM_CXY( main_cxy ); |
---|
| 319 | main_y = HAL_Y_FROM_CXY( main_cxy ); |
---|
[417] | 320 | main_uid = (((main_x * y_size) + main_y) * ncores) + main_lid; |
---|
| 321 | |
---|
| 322 | // checks number of threads |
---|
[619] | 323 | if ( (total_threads != 1) && (total_threads != 2) && (total_threads != 4) && |
---|
| 324 | (total_threads != 8) && (total_threads != 16 ) && (total_threads != 32) && |
---|
| 325 | (total_threads != 64) && (total_threads != 128) && (total_threads != 256) && |
---|
| 326 | (total_threads != 512) && (total_threads != 1024) ) |
---|
[417] | 327 | { |
---|
[623] | 328 | printf("\n[sort error] number of cores must be power of 2\n"); |
---|
[434] | 329 | exit( 0 ); |
---|
[417] | 330 | } |
---|
| 331 | |
---|
| 332 | // check array size |
---|
[619] | 333 | if ( ARRAY_LENGTH % total_threads) |
---|
[417] | 334 | { |
---|
[623] | 335 | printf("\n[sort error] array size must be multiple of number of threads\n"); |
---|
[434] | 336 | exit( 0 ); |
---|
[417] | 337 | } |
---|
| 338 | |
---|
[624] | 339 | printf("\n[sort] main starts / %d threads / %d items / pid %x / cycle %d\n", |
---|
[623] | 340 | total_threads, ARRAY_LENGTH, getpid(), (unsigned int)start_cycle ); |
---|
[417] | 341 | |
---|
[619] | 342 | // initialize barrier |
---|
| 343 | if( USE_DQT_BARRIER ) |
---|
[417] | 344 | { |
---|
[619] | 345 | barrier_attr.x_size = x_size; |
---|
| 346 | barrier_attr.y_size = y_size; |
---|
| 347 | barrier_attr.nthreads = ncores; |
---|
| 348 | error = pthread_barrier_init( &barrier, &barrier_attr , total_threads ); |
---|
| 349 | } |
---|
| 350 | else // use SIMPLE_BARRIER |
---|
| 351 | { |
---|
| 352 | error = pthread_barrier_init( &barrier, NULL , total_threads ); |
---|
| 353 | } |
---|
| 354 | |
---|
| 355 | if( error ) |
---|
| 356 | { |
---|
[623] | 357 | printf("\n[sort error] cannot initialise barrier\n" ); |
---|
[434] | 358 | exit( 0 ); |
---|
[417] | 359 | } |
---|
| 360 | |
---|
[627] | 361 | #if DEBUG_MAIN |
---|
[628] | 362 | if( USE_DQT_BARRIER ) printf("\n[sort] main completes DQT barrier init\n"); |
---|
| 363 | else printf("\n[sort] main completes simple barrier init\n"); |
---|
[623] | 364 | #endif |
---|
[417] | 365 | |
---|
| 366 | // Array to sort initialization |
---|
| 367 | for ( n = 0 ; n < ARRAY_LENGTH ; n++ ) |
---|
| 368 | { |
---|
[596] | 369 | array0[n] = ARRAY_LENGTH - n - 1; |
---|
[417] | 370 | } |
---|
| 371 | |
---|
[627] | 372 | #if DEBUG_MAIN |
---|
[623] | 373 | printf("\n[sort] main completes array init\n"); |
---|
[417] | 374 | #endif |
---|
| 375 | |
---|
| 376 | // launch other threads to execute sort() function |
---|
| 377 | // on cores other than the core running the main thread |
---|
[635] | 378 | for ( x = 0 ; x < x_size ; x++ ) |
---|
[417] | 379 | { |
---|
[635] | 380 | for ( y = 0 ; y < y_size ; y++ ) |
---|
[417] | 381 | { |
---|
[635] | 382 | for ( lid = 0 ; lid < ncores ; lid++ ) |
---|
[417] | 383 | { |
---|
[635] | 384 | // compute thread user index (continuous index) |
---|
[417] | 385 | thread_uid = (((x * y_size) + y) * ncores) + lid; |
---|
| 386 | |
---|
[635] | 387 | // set arguments for all threads |
---|
[619] | 388 | arg[thread_uid].threads = total_threads; |
---|
[417] | 389 | arg[thread_uid].thread_uid = thread_uid; |
---|
| 390 | arg[thread_uid].main_uid = main_uid; |
---|
| 391 | |
---|
| 392 | // set thread attributes for all threads |
---|
[635] | 393 | attr[thread_uid].attributes = PT_ATTR_CLUSTER_DEFINED | PT_ATTR_CORE_DEFINED; |
---|
[459] | 394 | attr[thread_uid].cxy = HAL_CXY_FROM_XY( x , y ); |
---|
[417] | 395 | attr[thread_uid].lid = lid; |
---|
| 396 | |
---|
| 397 | if( thread_uid != main_uid ) |
---|
| 398 | { |
---|
[635] | 399 | if ( pthread_create( &trdid[thread_uid], // buffer for kernel identifier |
---|
[417] | 400 | &attr[thread_uid], // thread attributes |
---|
| 401 | &sort, // entry function |
---|
| 402 | &arg[thread_uid] ) ) // sort arguments |
---|
| 403 | { |
---|
[623] | 404 | printf("\n[sort error] main cannot create thread %x \n", thread_uid ); |
---|
[434] | 405 | exit( 0 ); |
---|
[417] | 406 | } |
---|
[635] | 407 | |
---|
| 408 | #if (DEBUG_MAIN & 1) |
---|
[623] | 409 | printf("\n[sort] main created thread %x \n", thread_uid ); |
---|
| 410 | #endif |
---|
[440] | 411 | } |
---|
[417] | 412 | } |
---|
| 413 | } |
---|
| 414 | } |
---|
[442] | 415 | |
---|
[623] | 416 | /////////////////////////// |
---|
| 417 | get_cycle( &seq_end_cycle ); |
---|
[417] | 418 | |
---|
[627] | 419 | #if DEBUG_MAIN |
---|
[623] | 420 | printf("\n[sort] main completes sequencial init at cycle %d\n", |
---|
| 421 | (unsigned int)seq_end_cycle ); |
---|
| 422 | #endif |
---|
| 423 | |
---|
[445] | 424 | // the main thread run also the sort() function |
---|
| 425 | sort( &arg[main_uid] ); |
---|
| 426 | |
---|
[635] | 427 | // wait other threads completion |
---|
| 428 | for ( x = 0 ; x < x_size ; x++ ) |
---|
| 429 | { |
---|
| 430 | for ( y = 0 ; y < y_size ; y++ ) |
---|
| 431 | { |
---|
| 432 | for ( lid = 0 ; lid < ncores ; lid++ ) |
---|
| 433 | { |
---|
| 434 | // compute thread continuous index |
---|
| 435 | thread_uid = (((x * y_size) + y) * ncores) + lid; |
---|
| 436 | |
---|
| 437 | if( thread_uid != main_uid ) |
---|
| 438 | { |
---|
| 439 | if( pthread_join( trdid[thread_uid] , NULL ) ) |
---|
| 440 | { |
---|
| 441 | printf("\n[fft error] in main thread %d joining thread %d\n", |
---|
| 442 | main_uid , thread_uid ); |
---|
| 443 | exit( 0 ); |
---|
| 444 | } |
---|
| 445 | |
---|
| 446 | #if (DEBUG_MAIN & 1) |
---|
| 447 | printf("\n[fft] main thread %d joined thread %d\n", main_uid, thread_uid ); |
---|
| 448 | #endif |
---|
| 449 | |
---|
| 450 | } |
---|
| 451 | } |
---|
| 452 | } |
---|
| 453 | } |
---|
| 454 | |
---|
[623] | 455 | //////////////////////////// |
---|
| 456 | get_cycle( ¶_end_cycle ); |
---|
[619] | 457 | |
---|
[623] | 458 | printf("\n[sort] main completes parallel sort at cycle %d\n", |
---|
| 459 | (unsigned int)para_end_cycle ); |
---|
| 460 | |
---|
[619] | 461 | // destroy barrier |
---|
| 462 | pthread_barrier_destroy( &barrier ); |
---|
| 463 | |
---|
[625] | 464 | #if CHECK_RESULT |
---|
| 465 | int success = 1; |
---|
| 466 | int * res_array = ( (total_threads == 2) || |
---|
| 467 | (total_threads == 8) || |
---|
| 468 | (total_threads == 32) || |
---|
| 469 | (total_threads == 128) || |
---|
| 470 | (total_threads == 512) ) ? array1 : array0; |
---|
[623] | 471 | |
---|
[625] | 472 | for( n=0 ; n<(ARRAY_LENGTH-2) ; n++ ) |
---|
[417] | 473 | { |
---|
[625] | 474 | if ( res_array[n] > res_array[n+1] ) |
---|
| 475 | { |
---|
| 476 | printf("\n[sort] array[%d] = %d > array[%d] = %d\n", |
---|
| 477 | n , res_array[n] , n+1 , res_array[n+1] ); |
---|
| 478 | success = 0; |
---|
| 479 | break; |
---|
| 480 | } |
---|
[417] | 481 | } |
---|
| 482 | |
---|
[625] | 483 | if ( success ) printf("\n[sort] success\n"); |
---|
| 484 | else printf("\n[sort] failure\n"); |
---|
[417] | 485 | #endif |
---|
| 486 | |
---|
[623] | 487 | #if INSTRUMENTATION |
---|
[626] | 488 | char name[64]; |
---|
| 489 | char path[128]; |
---|
| 490 | unsigned long long instru_cycle; |
---|
[417] | 491 | |
---|
[629] | 492 | // build file name |
---|
| 493 | if( USE_DQT_BARRIER ) |
---|
| 494 | snprintf( name , 64 , "sort_dqt_%d_%d_%d", ARRAY_LENGTH, x_size * y_size, ncores ); |
---|
| 495 | else |
---|
| 496 | snprintf( name , 64 , "sort_smp_%d_%d_%d", ARRAY_LENGTH, x_size * y_size, ncores ); |
---|
[623] | 497 | |
---|
[625] | 498 | // build file pathname |
---|
| 499 | snprintf( path , 128 , "home/%s" , name ); |
---|
[623] | 500 | |
---|
[625] | 501 | // compute results |
---|
| 502 | unsigned int sequencial = (unsigned int)(seq_end_cycle - start_cycle); |
---|
| 503 | unsigned int parallel = (unsigned int)(para_end_cycle - seq_end_cycle); |
---|
[623] | 504 | |
---|
[625] | 505 | // display results on process terminal |
---|
| 506 | printf("\n----- %s -----\n" |
---|
| 507 | " - sequencial : %d cycles\n" |
---|
| 508 | " - parallel : %d cycles\n", |
---|
| 509 | name, sequencial, parallel ); |
---|
[623] | 510 | |
---|
[625] | 511 | // open file |
---|
[626] | 512 | get_cycle( &instru_cycle ); |
---|
[625] | 513 | FILE * stream = fopen( path , NULL ); |
---|
[626] | 514 | |
---|
[625] | 515 | if( stream == NULL ) |
---|
| 516 | { |
---|
[626] | 517 | printf("\n[sort error] cannot open instrumentation file <%s>\n", path ); |
---|
[625] | 518 | exit(0); |
---|
| 519 | } |
---|
[623] | 520 | |
---|
[626] | 521 | printf("\n[sort] file <%s> open at cycle %d\n", path, (unsigned int)instru_cycle ); |
---|
[625] | 522 | |
---|
[626] | 523 | #if IDBG |
---|
| 524 | idbg(); |
---|
| 525 | #endif |
---|
| 526 | |
---|
[625] | 527 | // register results to file |
---|
[626] | 528 | get_cycle( &instru_cycle ); |
---|
[625] | 529 | int ret = fprintf( stream , "\n----- %s -----\n" |
---|
| 530 | " - sequencial : %d cycles\n" |
---|
| 531 | " - parallel : %d cycles\n", name, sequencial, parallel ); |
---|
| 532 | if( ret < 0 ) |
---|
| 533 | { |
---|
[626] | 534 | printf("\n[sort error] cannot write to instrumentation file <%s>\n", path ); |
---|
[625] | 535 | exit(0); |
---|
| 536 | } |
---|
| 537 | |
---|
[626] | 538 | printf("\n[sort] file <%s> written at cycle %d\n", path, (unsigned int)instru_cycle ); |
---|
[625] | 539 | |
---|
[626] | 540 | #if IDBG |
---|
| 541 | idbg(); |
---|
| 542 | #endif |
---|
| 543 | |
---|
[625] | 544 | // close instrumentation file |
---|
[626] | 545 | get_cycle( &instru_cycle ); |
---|
| 546 | ret = fclose( stream ); |
---|
[625] | 547 | |
---|
[626] | 548 | if( ret ) |
---|
[625] | 549 | { |
---|
[626] | 550 | printf("\n[sort error] cannot close instrumentation file <%s>\n", path ); |
---|
[625] | 551 | exit(0); |
---|
| 552 | } |
---|
| 553 | |
---|
[626] | 554 | printf("\n[sort] file <%s> closed at cycle %d\n", path, (unsigned int)instru_cycle ); |
---|
[625] | 555 | |
---|
[619] | 556 | #endif |
---|
| 557 | |
---|
| 558 | exit( 0 ); |
---|
| 559 | |
---|
[417] | 560 | } // end main() |
---|
| 561 | |
---|
| 562 | /* |
---|
| 563 | vim: tabstop=4 : shiftwidth=4 : expandtab |
---|
| 564 | */ |
---|