libgomp.info revision 1.1.1.4
1 This is libgomp.info, produced by makeinfo version 6.1 from 2 libgomp.texi. 3 4 Copyright (C) 2006-2015 Free Software Foundation, Inc. 5 6 Permission is granted to copy, distribute and/or modify this document 7 under the terms of the GNU Free Documentation License, Version 1.3 or 8 any later version published by the Free Software Foundation; with the 9 Invariant Sections being "Funding Free Software", the Front-Cover texts 10 being (a) (see below), and with the Back-Cover Texts being (b) (see 11 below). A copy of the license is included in the section entitled "GNU 12 Free Documentation License". 13 14 (a) The FSF's Front-Cover Text is: 15 16 A GNU Manual 17 18 (b) The FSF's Back-Cover Text is: 19 20 You have freedom to copy and modify this GNU Manual, like GNU 21 software. Copies published by the Free Software Foundation raise funds 22 for GNU development. 23 INFO-DIR-SECTION GNU Libraries 24 START-INFO-DIR-ENTRY 25 * libgomp: (libgomp). GNU Offloading and Multi Processing Runtime Library. 26 END-INFO-DIR-ENTRY 27 28 This manual documents libgomp, the GNU Offloading and Multi 29 Processing Runtime library. This is the GNU implementation of the 30 OpenMP and OpenACC APIs for parallel and accelerator programming in 31 C/C++ and Fortran. 32 33 Published by the Free Software Foundation 51 Franklin Street, Fifth 34 Floor Boston, MA 02110-1301 USA 35 36 Copyright (C) 2006-2015 Free Software Foundation, Inc. 37 38 Permission is granted to copy, distribute and/or modify this document 39 under the terms of the GNU Free Documentation License, Version 1.3 or 40 any later version published by the Free Software Foundation; with the 41 Invariant Sections being "Funding Free Software", the Front-Cover texts 42 being (a) (see below), and with the Back-Cover Texts being (b) (see 43 below). A copy of the license is included in the section entitled "GNU 44 Free Documentation License". 45 46 (a) The FSF's Front-Cover Text is: 47 48 A GNU Manual 49 50 (b) The FSF's Back-Cover Text is: 51 52 You have freedom to copy and modify this GNU Manual, like GNU 53 software. Copies published by the Free Software Foundation raise funds 54 for GNU development. 55 56 57 File: libgomp.info, Node: Top, Next: Enabling OpenMP, Up: (dir) 58 59 Introduction 60 ************ 61 62 This manual documents the usage of libgomp, the GNU Offloading and Multi 63 Processing Runtime Library. This includes the GNU implementation of the 64 OpenMP (http://www.openmp.org) Application Programming Interface (API) 65 for multi-platform shared-memory parallel programming in C/C++ and 66 Fortran, and the GNU implementation of the OpenACC 67 (http://www.openacc.org/) Application Programming Interface (API) for 68 offloading of code to accelerator devices in C/C++ and Fortran. 69 70 Originally, libgomp implemented the GNU OpenMP Runtime Library. 71 Based on this, support for OpenACC and offloading (both OpenACC and 72 OpenMP 4's target construct) has been added later on, and the library's 73 name changed to GNU Offloading and Multi Processing Runtime Library. 74 75 * Menu: 76 77 * Enabling OpenMP:: How to enable OpenMP for your applications. 78 * Runtime Library Routines:: The OpenMP runtime application programming 79 interface. 80 * Environment Variables:: Influencing runtime behavior with environment 81 variables. 82 * The libgomp ABI:: Notes on the external ABI presented by libgomp. 83 * Reporting Bugs:: How to report bugs in the GNU Offloading and 84 Multi Processing Runtime Library. 85 * Copying:: GNU general public license says 86 how you can copy and share libgomp. 87 * GNU Free Documentation License:: 88 How you can copy and share this manual. 89 * Funding:: How to help assure continued work for free 90 software. 91 * Library Index:: Index of this documentation. 92 93 94 File: libgomp.info, Node: Enabling OpenMP, Next: Runtime Library Routines, Prev: Top, Up: Top 95 96 1 Enabling OpenMP 97 ***************** 98 99 To activate the OpenMP extensions for C/C++ and Fortran, the 100 compile-time flag '-fopenmp' must be specified. This enables the OpenMP 101 directive '#pragma omp' in C/C++ and '!$omp' directives in free form, 102 'c$omp', '*$omp' and '!$omp' directives in fixed form, '!$' conditional 103 compilation sentinels in free form and 'c$', '*$' and '!$' sentinels in 104 fixed form, for Fortran. The flag also arranges for automatic linking 105 of the OpenMP runtime library (*note Runtime Library Routines::). 106 107 A complete description of all OpenMP directives accepted may be found 108 in the OpenMP Application Program Interface (http://www.openmp.org) 109 manual, version 4.0. 110 111 112 File: libgomp.info, Node: Runtime Library Routines, Next: Environment Variables, Prev: Enabling OpenMP, Up: Top 113 114 2 Runtime Library Routines 115 ************************** 116 117 The runtime routines described here are defined by Section 3 of the 118 OpenMP specification in version 4.0. The routines are structured in 119 following three parts: 120 121 * Menu: 122 123 Control threads, processors and the parallel environment. They have C 124 linkage, and do not throw exceptions. 125 126 * omp_get_active_level:: Number of active parallel regions 127 * omp_get_ancestor_thread_num:: Ancestor thread ID 128 * omp_get_cancellation:: Whether cancellation support is enabled 129 * omp_get_default_device:: Get the default device for target regions 130 * omp_get_dynamic:: Dynamic teams setting 131 * omp_get_level:: Number of parallel regions 132 * omp_get_max_active_levels:: Maximum number of active regions 133 * omp_get_max_threads:: Maximum number of threads of parallel region 134 * omp_get_nested:: Nested parallel regions 135 * omp_get_num_devices:: Number of target devices 136 * omp_get_num_procs:: Number of processors online 137 * omp_get_num_teams:: Number of teams 138 * omp_get_num_threads:: Size of the active team 139 * omp_get_proc_bind:: Whether theads may be moved between CPUs 140 * omp_get_schedule:: Obtain the runtime scheduling method 141 * omp_get_team_num:: Get team number 142 * omp_get_team_size:: Number of threads in a team 143 * omp_get_thread_limit:: Maximum number of threads 144 * omp_get_thread_num:: Current thread ID 145 * omp_in_parallel:: Whether a parallel region is active 146 * omp_in_final:: Whether in final or included task region 147 * omp_is_initial_device:: Whether executing on the host device 148 * omp_set_default_device:: Set the default device for target regions 149 * omp_set_dynamic:: Enable/disable dynamic teams 150 * omp_set_max_active_levels:: Limits the number of active parallel regions 151 * omp_set_nested:: Enable/disable nested parallel regions 152 * omp_set_num_threads:: Set upper team size limit 153 * omp_set_schedule:: Set the runtime scheduling method 154 155 Initialize, set, test, unset and destroy simple and nested locks. 156 157 * omp_init_lock:: Initialize simple lock 158 * omp_set_lock:: Wait for and set simple lock 159 * omp_test_lock:: Test and set simple lock if available 160 * omp_unset_lock:: Unset simple lock 161 * omp_destroy_lock:: Destroy simple lock 162 * omp_init_nest_lock:: Initialize nested lock 163 * omp_set_nest_lock:: Wait for and set simple lock 164 * omp_test_nest_lock:: Test and set nested lock if available 165 * omp_unset_nest_lock:: Unset nested lock 166 * omp_destroy_nest_lock:: Destroy nested lock 167 168 Portable, thread-based, wall clock timer. 169 170 * omp_get_wtick:: Get timer precision. 171 * omp_get_wtime:: Elapsed wall clock time. 172 173 174 File: libgomp.info, Node: omp_get_active_level, Next: omp_get_ancestor_thread_num, Up: Runtime Library Routines 175 176 2.1 'omp_get_active_level' - Number of parallel regions 177 ======================================================= 178 179 _Description_: 180 This function returns the nesting level for the active parallel 181 blocks, which enclose the calling call. 182 183 _C/C++_ 184 _Prototype_: 'int omp_get_active_level(void);' 185 186 _Fortran_: 187 _Interface_: 'integer function omp_get_active_level()' 188 189 _See also_: 190 *note omp_get_level::, *note omp_get_max_active_levels::, *note 191 omp_set_max_active_levels:: 192 193 _Reference_: 194 OpenMP specification v4.0 (http://www.openmp.org/), Section 3.2.20. 195 196 197 File: libgomp.info, Node: omp_get_ancestor_thread_num, Next: omp_get_cancellation, Prev: omp_get_active_level, Up: Runtime Library Routines 198 199 2.2 'omp_get_ancestor_thread_num' - Ancestor thread ID 200 ====================================================== 201 202 _Description_: 203 This function returns the thread identification number for the 204 given nesting level of the current thread. For values of LEVEL 205 outside zero to 'omp_get_level' -1 is returned; if LEVEL is 206 'omp_get_level' the result is identical to 'omp_get_thread_num'. 207 208 _C/C++_ 209 _Prototype_: 'int omp_get_ancestor_thread_num(int level);' 210 211 _Fortran_: 212 _Interface_: 'integer function omp_get_ancestor_thread_num(level)' 213 'integer level' 214 215 _See also_: 216 *note omp_get_level::, *note omp_get_thread_num::, *note 217 omp_get_team_size:: 218 219 _Reference_: 220 OpenMP specification v4.0 (http://www.openmp.org/), Section 3.2.18. 221 222 223 File: libgomp.info, Node: omp_get_cancellation, Next: omp_get_default_device, Prev: omp_get_ancestor_thread_num, Up: Runtime Library Routines 224 225 2.3 'omp_get_cancellation' - Whether cancellation support is enabled 226 ==================================================================== 227 228 _Description_: 229 This function returns 'true' if cancellation is activated, 'false' 230 otherwise. Here, 'true' and 'false' represent their 231 language-specific counterparts. Unless 'OMP_CANCELLATION' is set 232 true, cancellations are deactivated. 233 234 _C/C++_: 235 _Prototype_: 'int omp_get_cancellation(void);' 236 237 _Fortran_: 238 _Interface_: 'logical function omp_get_cancellation()' 239 240 _See also_: 241 *note OMP_CANCELLATION:: 242 243 _Reference_: 244 OpenMP specification v4.0 (http://www.openmp.org/), Section 3.2.9. 245 246 247 File: libgomp.info, Node: omp_get_default_device, Next: omp_get_dynamic, Prev: omp_get_cancellation, Up: Runtime Library Routines 248 249 2.4 'omp_get_default_device' - Get the default device for target regions 250 ======================================================================== 251 252 _Description_: 253 Get the default device for target regions without device clause. 254 255 _C/C++_: 256 _Prototype_: 'int omp_get_default_device(void);' 257 258 _Fortran_: 259 _Interface_: 'integer function omp_get_default_device()' 260 261 _See also_: 262 *note OMP_DEFAULT_DEVICE::, *note omp_set_default_device:: 263 264 _Reference_: 265 OpenMP specification v4.0 (http://www.openmp.org/), Section 3.2.24. 266 267 268 File: libgomp.info, Node: omp_get_dynamic, Next: omp_get_level, Prev: omp_get_default_device, Up: Runtime Library Routines 269 270 2.5 'omp_get_dynamic' - Dynamic teams setting 271 ============================================= 272 273 _Description_: 274 This function returns 'true' if enabled, 'false' otherwise. Here, 275 'true' and 'false' represent their language-specific counterparts. 276 277 The dynamic team setting may be initialized at startup by the 278 'OMP_DYNAMIC' environment variable or at runtime using 279 'omp_set_dynamic'. If undefined, dynamic adjustment is disabled by 280 default. 281 282 _C/C++_: 283 _Prototype_: 'int omp_get_dynamic(void);' 284 285 _Fortran_: 286 _Interface_: 'logical function omp_get_dynamic()' 287 288 _See also_: 289 *note omp_set_dynamic::, *note OMP_DYNAMIC:: 290 291 _Reference_: 292 OpenMP specification v4.0 (http://www.openmp.org/), Section 3.2.8. 293 294 295 File: libgomp.info, Node: omp_get_level, Next: omp_get_max_active_levels, Prev: omp_get_dynamic, Up: Runtime Library Routines 296 297 2.6 'omp_get_level' - Obtain the current nesting level 298 ====================================================== 299 300 _Description_: 301 This function returns the nesting level for the parallel blocks, 302 which enclose the calling call. 303 304 _C/C++_ 305 _Prototype_: 'int omp_get_level(void);' 306 307 _Fortran_: 308 _Interface_: 'integer function omp_level()' 309 310 _See also_: 311 *note omp_get_active_level:: 312 313 _Reference_: 314 OpenMP specification v4.0 (http://www.openmp.org/), Section 3.2.17. 315 316 317 File: libgomp.info, Node: omp_get_max_active_levels, Next: omp_get_max_threads, Prev: omp_get_level, Up: Runtime Library Routines 318 319 2.7 'omp_get_max_active_levels' - Maximum number of active regions 320 ================================================================== 321 322 _Description_: 323 This function obtains the maximum allowed number of nested, active 324 parallel regions. 325 326 _C/C++_ 327 _Prototype_: 'int omp_get_max_active_levels(void);' 328 329 _Fortran_: 330 _Interface_: 'integer function omp_get_max_active_levels()' 331 332 _See also_: 333 *note omp_set_max_active_levels::, *note omp_get_active_level:: 334 335 _Reference_: 336 OpenMP specification v4.0 (http://www.openmp.org/), Section 3.2.16. 337 338 339 File: libgomp.info, Node: omp_get_max_threads, Next: omp_get_nested, Prev: omp_get_max_active_levels, Up: Runtime Library Routines 340 341 2.8 'omp_get_max_threads' - Maximum number of threads of parallel region 342 ======================================================================== 343 344 _Description_: 345 Return the maximum number of threads used for the current parallel 346 region that does not use the clause 'num_threads'. 347 348 _C/C++_: 349 _Prototype_: 'int omp_get_max_threads(void);' 350 351 _Fortran_: 352 _Interface_: 'integer function omp_get_max_threads()' 353 354 _See also_: 355 *note omp_set_num_threads::, *note omp_set_dynamic::, *note 356 omp_get_thread_limit:: 357 358 _Reference_: 359 OpenMP specification v4.0 (http://www.openmp.org/), Section 3.2.3. 360 361 362 File: libgomp.info, Node: omp_get_nested, Next: omp_get_num_devices, Prev: omp_get_max_threads, Up: Runtime Library Routines 363 364 2.9 'omp_get_nested' - Nested parallel regions 365 ============================================== 366 367 _Description_: 368 This function returns 'true' if nested parallel regions are 369 enabled, 'false' otherwise. Here, 'true' and 'false' represent 370 their language-specific counterparts. 371 372 Nested parallel regions may be initialized at startup by the 373 'OMP_NESTED' environment variable or at runtime using 374 'omp_set_nested'. If undefined, nested parallel regions are 375 disabled by default. 376 377 _C/C++_: 378 _Prototype_: 'int omp_get_nested(void);' 379 380 _Fortran_: 381 _Interface_: 'logical function omp_get_nested()' 382 383 _See also_: 384 *note omp_set_nested::, *note OMP_NESTED:: 385 386 _Reference_: 387 OpenMP specification v4.0 (http://www.openmp.org/), Section 3.2.11. 388 389 390 File: libgomp.info, Node: omp_get_num_devices, Next: omp_get_num_procs, Prev: omp_get_nested, Up: Runtime Library Routines 391 392 2.10 'omp_get_num_devices' - Number of target devices 393 ===================================================== 394 395 _Description_: 396 Returns the number of target devices. 397 398 _C/C++_: 399 _Prototype_: 'int omp_get_num_devices(void);' 400 401 _Fortran_: 402 _Interface_: 'integer function omp_get_num_devices()' 403 404 _Reference_: 405 OpenMP specification v4.0 (http://www.openmp.org/), Section 3.2.25. 406 407 408 File: libgomp.info, Node: omp_get_num_procs, Next: omp_get_num_teams, Prev: omp_get_num_devices, Up: Runtime Library Routines 409 410 2.11 'omp_get_num_procs' - Number of processors online 411 ====================================================== 412 413 _Description_: 414 Returns the number of processors online on that device. 415 416 _C/C++_: 417 _Prototype_: 'int omp_get_num_procs(void);' 418 419 _Fortran_: 420 _Interface_: 'integer function omp_get_num_procs()' 421 422 _Reference_: 423 OpenMP specification v4.0 (http://www.openmp.org/), Section 3.2.5. 424 425 426 File: libgomp.info, Node: omp_get_num_teams, Next: omp_get_num_threads, Prev: omp_get_num_procs, Up: Runtime Library Routines 427 428 2.12 'omp_get_num_teams' - Number of teams 429 ========================================== 430 431 _Description_: 432 Returns the number of teams in the current team region. 433 434 _C/C++_: 435 _Prototype_: 'int omp_get_num_teams(void);' 436 437 _Fortran_: 438 _Interface_: 'integer function omp_get_num_teams()' 439 440 _Reference_: 441 OpenMP specification v4.0 (http://www.openmp.org/), Section 3.2.26. 442 443 444 File: libgomp.info, Node: omp_get_num_threads, Next: omp_get_proc_bind, Prev: omp_get_num_teams, Up: Runtime Library Routines 445 446 2.13 'omp_get_num_threads' - Size of the active team 447 ==================================================== 448 449 _Description_: 450 Returns the number of threads in the current team. In a sequential 451 section of the program 'omp_get_num_threads' returns 1. 452 453 The default team size may be initialized at startup by the 454 'OMP_NUM_THREADS' environment variable. At runtime, the size of 455 the current team may be set either by the 'NUM_THREADS' clause or 456 by 'omp_set_num_threads'. If none of the above were used to define 457 a specific value and 'OMP_DYNAMIC' is disabled, one thread per CPU 458 online is used. 459 460 _C/C++_: 461 _Prototype_: 'int omp_get_num_threads(void);' 462 463 _Fortran_: 464 _Interface_: 'integer function omp_get_num_threads()' 465 466 _See also_: 467 *note omp_get_max_threads::, *note omp_set_num_threads::, *note 468 OMP_NUM_THREADS:: 469 470 _Reference_: 471 OpenMP specification v4.0 (http://www.openmp.org/), Section 3.2.2. 472 473 474 File: libgomp.info, Node: omp_get_proc_bind, Next: omp_get_schedule, Prev: omp_get_num_threads, Up: Runtime Library Routines 475 476 2.14 'omp_get_proc_bind' - Whether theads may be moved between CPUs 477 =================================================================== 478 479 _Description_: 480 This functions returns the currently active thread affinity policy, 481 which is set via 'OMP_PROC_BIND'. Possible values are 482 'omp_proc_bind_false', 'omp_proc_bind_true', 483 'omp_proc_bind_master', 'omp_proc_bind_close' and 484 'omp_proc_bind_spread'. 485 486 _C/C++_: 487 _Prototype_: 'omp_proc_bind_t omp_get_proc_bind(void);' 488 489 _Fortran_: 490 _Interface_: 'integer(kind=omp_proc_bind_kind) function 491 omp_get_proc_bind()' 492 493 _See also_: 494 *note OMP_PROC_BIND::, *note OMP_PLACES::, *note 495 GOMP_CPU_AFFINITY::, 496 497 _Reference_: 498 OpenMP specification v4.0 (http://www.openmp.org/), Section 3.2.22. 499 500 501 File: libgomp.info, Node: omp_get_schedule, Next: omp_get_team_num, Prev: omp_get_proc_bind, Up: Runtime Library Routines 502 503 2.15 'omp_get_schedule' - Obtain the runtime scheduling method 504 ============================================================== 505 506 _Description_: 507 Obtain the runtime scheduling method. The KIND argument will be 508 set to the value 'omp_sched_static', 'omp_sched_dynamic', 509 'omp_sched_guided' or 'omp_sched_auto'. The second argument, 510 MODIFIER, is set to the chunk size. 511 512 _C/C++_ 513 _Prototype_: 'void omp_get_schedule(omp_sched_t *kind, int 514 *modifier);' 515 516 _Fortran_: 517 _Interface_: 'subroutine omp_get_schedule(kind, modifier)' 518 'integer(kind=omp_sched_kind) kind' 519 'integer modifier' 520 521 _See also_: 522 *note omp_set_schedule::, *note OMP_SCHEDULE:: 523 524 _Reference_: 525 OpenMP specification v4.0 (http://www.openmp.org/), Section 3.2.13. 526 527 528 File: libgomp.info, Node: omp_get_team_num, Next: omp_get_team_size, Prev: omp_get_schedule, Up: Runtime Library Routines 529 530 2.16 'omp_get_team_num' - Get team number 531 ========================================= 532 533 _Description_: 534 Returns the team number of the calling thread. 535 536 _C/C++_: 537 _Prototype_: 'int omp_get_team_num(void);' 538 539 _Fortran_: 540 _Interface_: 'integer function omp_get_team_num()' 541 542 _Reference_: 543 OpenMP specification v4.0 (http://www.openmp.org/), Section 3.2.27. 544 545 546 File: libgomp.info, Node: omp_get_team_size, Next: omp_get_thread_limit, Prev: omp_get_team_num, Up: Runtime Library Routines 547 548 2.17 'omp_get_team_size' - Number of threads in a team 549 ====================================================== 550 551 _Description_: 552 This function returns the number of threads in a thread team to 553 which either the current thread or its ancestor belongs. For 554 values of LEVEL outside zero to 'omp_get_level', -1 is returned; if 555 LEVEL is zero, 1 is returned, and for 'omp_get_level', the result 556 is identical to 'omp_get_num_threads'. 557 558 _C/C++_: 559 _Prototype_: 'int omp_get_team_size(int level);' 560 561 _Fortran_: 562 _Interface_: 'integer function omp_get_team_size(level)' 563 'integer level' 564 565 _See also_: 566 *note omp_get_num_threads::, *note omp_get_level::, *note 567 omp_get_ancestor_thread_num:: 568 569 _Reference_: 570 OpenMP specification v4.0 (http://www.openmp.org/), Section 3.2.19. 571 572 573 File: libgomp.info, Node: omp_get_thread_limit, Next: omp_get_thread_num, Prev: omp_get_team_size, Up: Runtime Library Routines 574 575 2.18 'omp_get_thread_limit' - Maximum number of threads 576 ======================================================= 577 578 _Description_: 579 Return the maximum number of threads of the program. 580 581 _C/C++_: 582 _Prototype_: 'int omp_get_thread_limit(void);' 583 584 _Fortran_: 585 _Interface_: 'integer function omp_get_thread_limit()' 586 587 _See also_: 588 *note omp_get_max_threads::, *note OMP_THREAD_LIMIT:: 589 590 _Reference_: 591 OpenMP specification v4.0 (http://www.openmp.org/), Section 3.2.14. 592 593 594 File: libgomp.info, Node: omp_get_thread_num, Next: omp_in_parallel, Prev: omp_get_thread_limit, Up: Runtime Library Routines 595 596 2.19 'omp_get_thread_num' - Current thread ID 597 ============================================= 598 599 _Description_: 600 Returns a unique thread identification number within the current 601 team. In a sequential parts of the program, 'omp_get_thread_num' 602 always returns 0. In parallel regions the return value varies from 603 0 to 'omp_get_num_threads'-1 inclusive. The return value of the 604 master thread of a team is always 0. 605 606 _C/C++_: 607 _Prototype_: 'int omp_get_thread_num(void);' 608 609 _Fortran_: 610 _Interface_: 'integer function omp_get_thread_num()' 611 612 _See also_: 613 *note omp_get_num_threads::, *note omp_get_ancestor_thread_num:: 614 615 _Reference_: 616 OpenMP specification v4.0 (http://www.openmp.org/), Section 3.2.4. 617 618 619 File: libgomp.info, Node: omp_in_parallel, Next: omp_in_final, Prev: omp_get_thread_num, Up: Runtime Library Routines 620 621 2.20 'omp_in_parallel' - Whether a parallel region is active 622 ============================================================ 623 624 _Description_: 625 This function returns 'true' if currently running in parallel, 626 'false' otherwise. Here, 'true' and 'false' represent their 627 language-specific counterparts. 628 629 _C/C++_: 630 _Prototype_: 'int omp_in_parallel(void);' 631 632 _Fortran_: 633 _Interface_: 'logical function omp_in_parallel()' 634 635 _Reference_: 636 OpenMP specification v4.0 (http://www.openmp.org/), Section 3.2.6. 637 638 639 File: libgomp.info, Node: omp_in_final, Next: omp_is_initial_device, Prev: omp_in_parallel, Up: Runtime Library Routines 640 641 2.21 'omp_in_final' - Whether in final or included task region 642 ============================================================== 643 644 _Description_: 645 This function returns 'true' if currently running in a final or 646 included task region, 'false' otherwise. Here, 'true' and 'false' 647 represent their language-specific counterparts. 648 649 _C/C++_: 650 _Prototype_: 'int omp_in_final(void);' 651 652 _Fortran_: 653 _Interface_: 'logical function omp_in_final()' 654 655 _Reference_: 656 OpenMP specification v4.0 (http://www.openmp.org/), Section 3.2.21. 657 658 659 File: libgomp.info, Node: omp_is_initial_device, Next: omp_set_default_device, Prev: omp_in_final, Up: Runtime Library Routines 660 661 2.22 'omp_is_initial_device' - Whether executing on the host device 662 =================================================================== 663 664 _Description_: 665 This function returns 'true' if currently running on the host 666 device, 'false' otherwise. Here, 'true' and 'false' represent 667 their language-specific counterparts. 668 669 _C/C++_: 670 _Prototype_: 'int omp_is_initial_device(void);' 671 672 _Fortran_: 673 _Interface_: 'logical function omp_is_initial_device()' 674 675 _Reference_: 676 OpenMP specification v4.0 (http://www.openmp.org/), Section 3.2.28. 677 678 679 File: libgomp.info, Node: omp_set_default_device, Next: omp_set_dynamic, Prev: omp_is_initial_device, Up: Runtime Library Routines 680 681 2.23 'omp_set_default_device' - Set the default device for target regions 682 ========================================================================= 683 684 _Description_: 685 Set the default device for target regions without device clause. 686 The argument shall be a nonnegative device number. 687 688 _C/C++_: 689 _Prototype_: 'void omp_set_default_device(int device_num);' 690 691 _Fortran_: 692 _Interface_: 'subroutine omp_set_default_device(device_num)' 693 'integer device_num' 694 695 _See also_: 696 *note OMP_DEFAULT_DEVICE::, *note omp_get_default_device:: 697 698 _Reference_: 699 OpenMP specification v4.0 (http://www.openmp.org/), Section 3.2.23. 700 701 702 File: libgomp.info, Node: omp_set_dynamic, Next: omp_set_max_active_levels, Prev: omp_set_default_device, Up: Runtime Library Routines 703 704 2.24 'omp_set_dynamic' - Enable/disable dynamic teams 705 ===================================================== 706 707 _Description_: 708 Enable or disable the dynamic adjustment of the number of threads 709 within a team. The function takes the language-specific equivalent 710 of 'true' and 'false', where 'true' enables dynamic adjustment of 711 team sizes and 'false' disables it. 712 713 _C/C++_: 714 _Prototype_: 'void omp_set_dynamic(int dynamic_threads);' 715 716 _Fortran_: 717 _Interface_: 'subroutine omp_set_dynamic(dynamic_threads)' 718 'logical, intent(in) :: dynamic_threads' 719 720 _See also_: 721 *note OMP_DYNAMIC::, *note omp_get_dynamic:: 722 723 _Reference_: 724 OpenMP specification v4.0 (http://www.openmp.org/), Section 3.2.7. 725 726 727 File: libgomp.info, Node: omp_set_max_active_levels, Next: omp_set_nested, Prev: omp_set_dynamic, Up: Runtime Library Routines 728 729 2.25 'omp_set_max_active_levels' - Limits the number of active parallel regions 730 =============================================================================== 731 732 _Description_: 733 This function limits the maximum allowed number of nested, active 734 parallel regions. 735 736 _C/C++_ 737 _Prototype_: 'void omp_set_max_active_levels(int max_levels);' 738 739 _Fortran_: 740 _Interface_: 'subroutine omp_set_max_active_levels(max_levels)' 741 'integer max_levels' 742 743 _See also_: 744 *note omp_get_max_active_levels::, *note omp_get_active_level:: 745 746 _Reference_: 747 OpenMP specification v4.0 (http://www.openmp.org/), Section 3.2.15. 748 749 750 File: libgomp.info, Node: omp_set_nested, Next: omp_set_num_threads, Prev: omp_set_max_active_levels, Up: Runtime Library Routines 751 752 2.26 'omp_set_nested' - Enable/disable nested parallel regions 753 ============================================================== 754 755 _Description_: 756 Enable or disable nested parallel regions, i.e., whether team 757 members are allowed to create new teams. The function takes the 758 language-specific equivalent of 'true' and 'false', where 'true' 759 enables dynamic adjustment of team sizes and 'false' disables it. 760 761 _C/C++_: 762 _Prototype_: 'void omp_set_nested(int nested);' 763 764 _Fortran_: 765 _Interface_: 'subroutine omp_set_nested(nested)' 766 'logical, intent(in) :: nested' 767 768 _See also_: 769 *note OMP_NESTED::, *note omp_get_nested:: 770 771 _Reference_: 772 OpenMP specification v4.0 (http://www.openmp.org/), Section 3.2.10. 773 774 775 File: libgomp.info, Node: omp_set_num_threads, Next: omp_set_schedule, Prev: omp_set_nested, Up: Runtime Library Routines 776 777 2.27 'omp_set_num_threads' - Set upper team size limit 778 ====================================================== 779 780 _Description_: 781 Specifies the number of threads used by default in subsequent 782 parallel sections, if those do not specify a 'num_threads' clause. 783 The argument of 'omp_set_num_threads' shall be a positive integer. 784 785 _C/C++_: 786 _Prototype_: 'void omp_set_num_threads(int num_threads);' 787 788 _Fortran_: 789 _Interface_: 'subroutine omp_set_num_threads(num_threads)' 790 'integer, intent(in) :: num_threads' 791 792 _See also_: 793 *note OMP_NUM_THREADS::, *note omp_get_num_threads::, *note 794 omp_get_max_threads:: 795 796 _Reference_: 797 OpenMP specification v4.0 (http://www.openmp.org/), Section 3.2.1. 798 799 800 File: libgomp.info, Node: omp_set_schedule, Next: omp_init_lock, Prev: omp_set_num_threads, Up: Runtime Library Routines 801 802 2.28 'omp_set_schedule' - Set the runtime scheduling method 803 =========================================================== 804 805 _Description_: 806 Sets the runtime scheduling method. The KIND argument can have the 807 value 'omp_sched_static', 'omp_sched_dynamic', 'omp_sched_guided' 808 or 'omp_sched_auto'. Except for 'omp_sched_auto', the chunk size 809 is set to the value of MODIFIER if positive, or to the default 810 value if zero or negative. For 'omp_sched_auto' the MODIFIER 811 argument is ignored. 812 813 _C/C++_ 814 _Prototype_: 'void omp_set_schedule(omp_sched_t kind, int modifier);' 815 816 _Fortran_: 817 _Interface_: 'subroutine omp_set_schedule(kind, modifier)' 818 'integer(kind=omp_sched_kind) kind' 819 'integer modifier' 820 821 _See also_: 822 *note omp_get_schedule:: *note OMP_SCHEDULE:: 823 824 _Reference_: 825 OpenMP specification v4.0 (http://www.openmp.org/), Section 3.2.12. 826 827 828 File: libgomp.info, Node: omp_init_lock, Next: omp_set_lock, Prev: omp_set_schedule, Up: Runtime Library Routines 829 830 2.29 'omp_init_lock' - Initialize simple lock 831 ============================================= 832 833 _Description_: 834 Initialize a simple lock. After initialization, the lock is in an 835 unlocked state. 836 837 _C/C++_: 838 _Prototype_: 'void omp_init_lock(omp_lock_t *lock);' 839 840 _Fortran_: 841 _Interface_: 'subroutine omp_init_lock(svar)' 842 'integer(omp_lock_kind), intent(out) :: svar' 843 844 _See also_: 845 *note omp_destroy_lock:: 846 847 _Reference_: 848 OpenMP specification v4.0 (http://www.openmp.org/), Section 3.3.1. 849 850 851 File: libgomp.info, Node: omp_set_lock, Next: omp_test_lock, Prev: omp_init_lock, Up: Runtime Library Routines 852 853 2.30 'omp_set_lock' - Wait for and set simple lock 854 ================================================== 855 856 _Description_: 857 Before setting a simple lock, the lock variable must be initialized 858 by 'omp_init_lock'. The calling thread is blocked until the lock 859 is available. If the lock is already held by the current thread, a 860 deadlock occurs. 861 862 _C/C++_: 863 _Prototype_: 'void omp_set_lock(omp_lock_t *lock);' 864 865 _Fortran_: 866 _Interface_: 'subroutine omp_set_lock(svar)' 867 'integer(omp_lock_kind), intent(inout) :: svar' 868 869 _See also_: 870 *note omp_init_lock::, *note omp_test_lock::, *note 871 omp_unset_lock:: 872 873 _Reference_: 874 OpenMP specification v4.0 (http://www.openmp.org/), Section 3.3.3. 875 876 877 File: libgomp.info, Node: omp_test_lock, Next: omp_unset_lock, Prev: omp_set_lock, Up: Runtime Library Routines 878 879 2.31 'omp_test_lock' - Test and set simple lock if available 880 ============================================================ 881 882 _Description_: 883 Before setting a simple lock, the lock variable must be initialized 884 by 'omp_init_lock'. Contrary to 'omp_set_lock', 'omp_test_lock' 885 does not block if the lock is not available. This function returns 886 'true' upon success, 'false' otherwise. Here, 'true' and 'false' 887 represent their language-specific counterparts. 888 889 _C/C++_: 890 _Prototype_: 'int omp_test_lock(omp_lock_t *lock);' 891 892 _Fortran_: 893 _Interface_: 'logical function omp_test_lock(svar)' 894 'integer(omp_lock_kind), intent(inout) :: svar' 895 896 _See also_: 897 *note omp_init_lock::, *note omp_set_lock::, *note omp_set_lock:: 898 899 _Reference_: 900 OpenMP specification v4.0 (http://www.openmp.org/), Section 3.3.5. 901 902 903 File: libgomp.info, Node: omp_unset_lock, Next: omp_destroy_lock, Prev: omp_test_lock, Up: Runtime Library Routines 904 905 2.32 'omp_unset_lock' - Unset simple lock 906 ========================================= 907 908 _Description_: 909 A simple lock about to be unset must have been locked by 910 'omp_set_lock' or 'omp_test_lock' before. In addition, the lock 911 must be held by the thread calling 'omp_unset_lock'. Then, the 912 lock becomes unlocked. If one or more threads attempted to set the 913 lock before, one of them is chosen to, again, set the lock to 914 itself. 915 916 _C/C++_: 917 _Prototype_: 'void omp_unset_lock(omp_lock_t *lock);' 918 919 _Fortran_: 920 _Interface_: 'subroutine omp_unset_lock(svar)' 921 'integer(omp_lock_kind), intent(inout) :: svar' 922 923 _See also_: 924 *note omp_set_lock::, *note omp_test_lock:: 925 926 _Reference_: 927 OpenMP specification v4.0 (http://www.openmp.org/), Section 3.3.4. 928 929 930 File: libgomp.info, Node: omp_destroy_lock, Next: omp_init_nest_lock, Prev: omp_unset_lock, Up: Runtime Library Routines 931 932 2.33 'omp_destroy_lock' - Destroy simple lock 933 ============================================= 934 935 _Description_: 936 Destroy a simple lock. In order to be destroyed, a simple lock 937 must be in the unlocked state. 938 939 _C/C++_: 940 _Prototype_: 'void omp_destroy_lock(omp_lock_t *lock);' 941 942 _Fortran_: 943 _Interface_: 'subroutine omp_destroy_lock(svar)' 944 'integer(omp_lock_kind), intent(inout) :: svar' 945 946 _See also_: 947 *note omp_init_lock:: 948 949 _Reference_: 950 OpenMP specification v4.0 (http://www.openmp.org/), Section 3.3.2. 951 952 953 File: libgomp.info, Node: omp_init_nest_lock, Next: omp_set_nest_lock, Prev: omp_destroy_lock, Up: Runtime Library Routines 954 955 2.34 'omp_init_nest_lock' - Initialize nested lock 956 ================================================== 957 958 _Description_: 959 Initialize a nested lock. After initialization, the lock is in an 960 unlocked state and the nesting count is set to zero. 961 962 _C/C++_: 963 _Prototype_: 'void omp_init_nest_lock(omp_nest_lock_t *lock);' 964 965 _Fortran_: 966 _Interface_: 'subroutine omp_init_nest_lock(nvar)' 967 'integer(omp_nest_lock_kind), intent(out) :: nvar' 968 969 _See also_: 970 *note omp_destroy_nest_lock:: 971 972 _Reference_: 973 OpenMP specification v4.0 (http://www.openmp.org/), Section 3.3.1. 974 975 976 File: libgomp.info, Node: omp_set_nest_lock, Next: omp_test_nest_lock, Prev: omp_init_nest_lock, Up: Runtime Library Routines 977 978 2.35 'omp_set_nest_lock' - Wait for and set nested lock 979 ======================================================= 980 981 _Description_: 982 Before setting a nested lock, the lock variable must be initialized 983 by 'omp_init_nest_lock'. The calling thread is blocked until the 984 lock is available. If the lock is already held by the current 985 thread, the nesting count for the lock is incremented. 986 987 _C/C++_: 988 _Prototype_: 'void omp_set_nest_lock(omp_nest_lock_t *lock);' 989 990 _Fortran_: 991 _Interface_: 'subroutine omp_set_nest_lock(nvar)' 992 'integer(omp_nest_lock_kind), intent(inout) :: nvar' 993 994 _See also_: 995 *note omp_init_nest_lock::, *note omp_unset_nest_lock:: 996 997 _Reference_: 998 OpenMP specification v4.0 (http://www.openmp.org/), Section 3.3.3. 999 1000 1001 File: libgomp.info, Node: omp_test_nest_lock, Next: omp_unset_nest_lock, Prev: omp_set_nest_lock, Up: Runtime Library Routines 1002 1003 2.36 'omp_test_nest_lock' - Test and set nested lock if available 1004 ================================================================= 1005 1006 _Description_: 1007 Before setting a nested lock, the lock variable must be initialized 1008 by 'omp_init_nest_lock'. Contrary to 'omp_set_nest_lock', 1009 'omp_test_nest_lock' does not block if the lock is not available. 1010 If the lock is already held by the current thread, the new nesting 1011 count is returned. Otherwise, the return value equals zero. 1012 1013 _C/C++_: 1014 _Prototype_: 'int omp_test_nest_lock(omp_nest_lock_t *lock);' 1015 1016 _Fortran_: 1017 _Interface_: 'logical function omp_test_nest_lock(nvar)' 1018 'integer(omp_nest_lock_kind), intent(inout) :: nvar' 1019 1020 _See also_: 1021 *note omp_init_lock::, *note omp_set_lock::, *note omp_set_lock:: 1022 1023 _Reference_: 1024 OpenMP specification v4.0 (http://www.openmp.org/), Section 3.3.5. 1025 1026 1027 File: libgomp.info, Node: omp_unset_nest_lock, Next: omp_destroy_nest_lock, Prev: omp_test_nest_lock, Up: Runtime Library Routines 1028 1029 2.37 'omp_unset_nest_lock' - Unset nested lock 1030 ============================================== 1031 1032 _Description_: 1033 A nested lock about to be unset must have been locked by 1034 'omp_set_nested_lock' or 'omp_test_nested_lock' before. In 1035 addition, the lock must be held by the thread calling 1036 'omp_unset_nested_lock'. If the nesting count drops to zero, the 1037 lock becomes unlocked. If one ore more threads attempted to set 1038 the lock before, one of them is chosen to, again, set the lock to 1039 itself. 1040 1041 _C/C++_: 1042 _Prototype_: 'void omp_unset_nest_lock(omp_nest_lock_t *lock);' 1043 1044 _Fortran_: 1045 _Interface_: 'subroutine omp_unset_nest_lock(nvar)' 1046 'integer(omp_nest_lock_kind), intent(inout) :: nvar' 1047 1048 _See also_: 1049 *note omp_set_nest_lock:: 1050 1051 _Reference_: 1052 OpenMP specification v4.0 (http://www.openmp.org/), Section 3.3.4. 1053 1054 1055 File: libgomp.info, Node: omp_destroy_nest_lock, Next: omp_get_wtick, Prev: omp_unset_nest_lock, Up: Runtime Library Routines 1056 1057 2.38 'omp_destroy_nest_lock' - Destroy nested lock 1058 ================================================== 1059 1060 _Description_: 1061 Destroy a nested lock. In order to be destroyed, a nested lock 1062 must be in the unlocked state and its nesting count must equal 1063 zero. 1064 1065 _C/C++_: 1066 _Prototype_: 'void omp_destroy_nest_lock(omp_nest_lock_t *);' 1067 1068 _Fortran_: 1069 _Interface_: 'subroutine omp_destroy_nest_lock(nvar)' 1070 'integer(omp_nest_lock_kind), intent(inout) :: nvar' 1071 1072 _See also_: 1073 *note omp_init_lock:: 1074 1075 _Reference_: 1076 OpenMP specification v4.0 (http://www.openmp.org/), Section 3.3.2. 1077 1078 1079 File: libgomp.info, Node: omp_get_wtick, Next: omp_get_wtime, Prev: omp_destroy_nest_lock, Up: Runtime Library Routines 1080 1081 2.39 'omp_get_wtick' - Get timer precision 1082 ========================================== 1083 1084 _Description_: 1085 Gets the timer precision, i.e., the number of seconds between two 1086 successive clock ticks. 1087 1088 _C/C++_: 1089 _Prototype_: 'double omp_get_wtick(void);' 1090 1091 _Fortran_: 1092 _Interface_: 'double precision function omp_get_wtick()' 1093 1094 _See also_: 1095 *note omp_get_wtime:: 1096 1097 _Reference_: 1098 OpenMP specification v4.0 (http://www.openmp.org/), Section 3.4.2. 1099 1100 1101 File: libgomp.info, Node: omp_get_wtime, Prev: omp_get_wtick, Up: Runtime Library Routines 1102 1103 2.40 'omp_get_wtime' - Elapsed wall clock time 1104 ============================================== 1105 1106 _Description_: 1107 Elapsed wall clock time in seconds. The time is measured per 1108 thread, no guarantee can be made that two distinct threads measure 1109 the same time. Time is measured from some "time in the past", 1110 which is an arbitrary time guaranteed not to change during the 1111 execution of the program. 1112 1113 _C/C++_: 1114 _Prototype_: 'double omp_get_wtime(void);' 1115 1116 _Fortran_: 1117 _Interface_: 'double precision function omp_get_wtime()' 1118 1119 _See also_: 1120 *note omp_get_wtick:: 1121 1122 _Reference_: 1123 OpenMP specification v4.0 (http://www.openmp.org/), Section 3.4.1. 1124 1125 1126 File: libgomp.info, Node: Environment Variables, Next: The libgomp ABI, Prev: Runtime Library Routines, Up: Top 1127 1128 3 Environment Variables 1129 *********************** 1130 1131 The environment variables which beginning with 'OMP_' are defined by 1132 section 4 of the OpenMP specification in version 4.0, while those 1133 beginning with 'GOMP_' are GNU extensions. 1134 1135 * Menu: 1136 1137 * OMP_CANCELLATION:: Set whether cancellation is activated 1138 * OMP_DISPLAY_ENV:: Show OpenMP version and environment variables 1139 * OMP_DEFAULT_DEVICE:: Set the device used in target regions 1140 * OMP_DYNAMIC:: Dynamic adjustment of threads 1141 * OMP_MAX_ACTIVE_LEVELS:: Set the maximum number of nested parallel regions 1142 * OMP_NESTED:: Nested parallel regions 1143 * OMP_NUM_THREADS:: Specifies the number of threads to use 1144 * OMP_PROC_BIND:: Whether theads may be moved between CPUs 1145 * OMP_PLACES:: Specifies on which CPUs the theads should be placed 1146 * OMP_STACKSIZE:: Set default thread stack size 1147 * OMP_SCHEDULE:: How threads are scheduled 1148 * OMP_THREAD_LIMIT:: Set the maximum number of threads 1149 * OMP_WAIT_POLICY:: How waiting threads are handled 1150 * GOMP_CPU_AFFINITY:: Bind threads to specific CPUs 1151 * GOMP_DEBUG:: Enable debugging output 1152 * GOMP_STACKSIZE:: Set default thread stack size 1153 * GOMP_SPINCOUNT:: Set the busy-wait spin count 1154 1155 1156 File: libgomp.info, Node: OMP_CANCELLATION, Next: OMP_DISPLAY_ENV, Up: Environment Variables 1157 1158 3.1 'OMP_CANCELLATION' - Set whether cancellation is activated 1159 ============================================================== 1160 1161 _Description_: 1162 If set to 'TRUE', the cancellation is activated. If set to 'FALSE' 1163 or if unset, cancellation is disabled and the 'cancel' construct is 1164 ignored. 1165 1166 _See also_: 1167 *note omp_get_cancellation:: 1168 1169 _Reference_: 1170 OpenMP specification v4.0 (http://www.openmp.org/), Section 4.11 1171 1172 1173 File: libgomp.info, Node: OMP_DISPLAY_ENV, Next: OMP_DEFAULT_DEVICE, Prev: OMP_CANCELLATION, Up: Environment Variables 1174 1175 3.2 'OMP_DISPLAY_ENV' - Show OpenMP version and environment variables 1176 ===================================================================== 1177 1178 _Description_: 1179 If set to 'TRUE', the OpenMP version number and the values 1180 associated with the OpenMP environment variables are printed to 1181 'stderr'. If set to 'VERBOSE', it additionally shows the value of 1182 the environment variables which are GNU extensions. If undefined 1183 or set to 'FALSE', this information will not be shown. 1184 1185 _Reference_: 1186 OpenMP specification v4.0 (http://www.openmp.org/), Section 4.12 1187 1188 1189 File: libgomp.info, Node: OMP_DEFAULT_DEVICE, Next: OMP_DYNAMIC, Prev: OMP_DISPLAY_ENV, Up: Environment Variables 1190 1191 3.3 'OMP_DEFAULT_DEVICE' - Set the device used in target regions 1192 ================================================================ 1193 1194 _Description_: 1195 Set to choose the device which is used in a 'target' region, unless 1196 the value is overridden by 'omp_set_default_device' or by a 1197 'device' clause. The value shall be the nonnegative device number. 1198 If no device with the given device number exists, the code is 1199 executed on the host. If unset, device number 0 will be used. 1200 1201 _See also_: 1202 *note omp_get_default_device::, *note omp_set_default_device::, 1203 1204 _Reference_: 1205 OpenMP specification v4.0 (http://www.openmp.org/), Section 4.11 1206 1207 1208 File: libgomp.info, Node: OMP_DYNAMIC, Next: OMP_MAX_ACTIVE_LEVELS, Prev: OMP_DEFAULT_DEVICE, Up: Environment Variables 1209 1210 3.4 'OMP_DYNAMIC' - Dynamic adjustment of threads 1211 ================================================= 1212 1213 _Description_: 1214 Enable or disable the dynamic adjustment of the number of threads 1215 within a team. The value of this environment variable shall be 1216 'TRUE' or 'FALSE'. If undefined, dynamic adjustment is disabled by 1217 default. 1218 1219 _See also_: 1220 *note omp_set_dynamic:: 1221 1222 _Reference_: 1223 OpenMP specification v4.0 (http://www.openmp.org/), Section 4.3 1224 1225 1226 File: libgomp.info, Node: OMP_MAX_ACTIVE_LEVELS, Next: OMP_NESTED, Prev: OMP_DYNAMIC, Up: Environment Variables 1227 1228 3.5 'OMP_MAX_ACTIVE_LEVELS' - Set the maximum number of nested parallel regions 1229 =============================================================================== 1230 1231 _Description_: 1232 Specifies the initial value for the maximum number of nested 1233 parallel regions. The value of this variable shall be a positive 1234 integer. If undefined, the number of active levels is unlimited. 1235 1236 _See also_: 1237 *note omp_set_max_active_levels:: 1238 1239 _Reference_: 1240 OpenMP specification v4.0 (http://www.openmp.org/), Section 4.9 1241 1242 1243 File: libgomp.info, Node: OMP_NESTED, Next: OMP_NUM_THREADS, Prev: OMP_MAX_ACTIVE_LEVELS, Up: Environment Variables 1244 1245 3.6 'OMP_NESTED' - Nested parallel regions 1246 ========================================== 1247 1248 _Description_: 1249 Enable or disable nested parallel regions, i.e., whether team 1250 members are allowed to create new teams. The value of this 1251 environment variable shall be 'TRUE' or 'FALSE'. If undefined, 1252 nested parallel regions are disabled by default. 1253 1254 _See also_: 1255 *note omp_set_nested:: 1256 1257 _Reference_: 1258 OpenMP specification v4.0 (http://www.openmp.org/), Section 4.6 1259 1260 1261 File: libgomp.info, Node: OMP_NUM_THREADS, Next: OMP_PROC_BIND, Prev: OMP_NESTED, Up: Environment Variables 1262 1263 3.7 'OMP_NUM_THREADS' - Specifies the number of threads to use 1264 ============================================================== 1265 1266 _Description_: 1267 Specifies the default number of threads to use in parallel regions. 1268 The value of this variable shall be a comma-separated list of 1269 positive integers; the value specified the number of threads to use 1270 for the corresponding nested level. If undefined one thread per 1271 CPU is used. 1272 1273 _See also_: 1274 *note omp_set_num_threads:: 1275 1276 _Reference_: 1277 OpenMP specification v4.0 (http://www.openmp.org/), Section 4.2 1278 1279 1280 File: libgomp.info, Node: OMP_PROC_BIND, Next: OMP_PLACES, Prev: OMP_NUM_THREADS, Up: Environment Variables 1281 1282 3.8 'OMP_PROC_BIND' - Whether theads may be moved between CPUs 1283 ============================================================== 1284 1285 _Description_: 1286 Specifies whether threads may be moved between processors. If set 1287 to 'TRUE', OpenMP theads should not be moved; if set to 'FALSE' 1288 they may be moved. Alternatively, a comma separated list with the 1289 values 'MASTER', 'CLOSE' and 'SPREAD' can be used to specify the 1290 thread affinity policy for the corresponding nesting level. With 1291 'MASTER' the worker threads are in the same place partition as the 1292 master thread. With 'CLOSE' those are kept close to the master 1293 thread in contiguous place partitions. And with 'SPREAD' a sparse 1294 distribution across the place partitions is used. 1295 1296 When undefined, 'OMP_PROC_BIND' defaults to 'TRUE' when 1297 'OMP_PLACES' or 'GOMP_CPU_AFFINITY' is set and 'FALSE' otherwise. 1298 1299 _See also_: 1300 *note OMP_PLACES::, *note GOMP_CPU_AFFINITY::, *note 1301 omp_get_proc_bind:: 1302 1303 _Reference_: 1304 OpenMP specification v4.0 (http://www.openmp.org/), Section 4.4 1305 1306 1307 File: libgomp.info, Node: OMP_PLACES, Next: OMP_STACKSIZE, Prev: OMP_PROC_BIND, Up: Environment Variables 1308 1309 3.9 'OMP_PLACES' - Specifies on which CPUs the theads should be placed 1310 ====================================================================== 1311 1312 _Description_: 1313 The thread placement can be either specified using an abstract name 1314 or by an explicit list of the places. The abstract names 1315 'threads', 'cores' and 'sockets' can be optionally followed by a 1316 positive number in parentheses, which denotes the how many places 1317 shall be created. With 'threads' each place corresponds to a 1318 single hardware thread; 'cores' to a single core with the 1319 corresponding number of hardware threads; and with 'sockets' the 1320 place corresponds to a single socket. The resulting placement can 1321 be shown by setting the 'OMP_DISPLAY_ENV' environment variable. 1322 1323 Alternatively, the placement can be specified explicitly as 1324 comma-separated list of places. A place is specified by set of 1325 nonnegative numbers in curly braces, denoting the denoting the 1326 hardware threads. The hardware threads belonging to a place can 1327 either be specified as comma-separated list of nonnegative thread 1328 numbers or using an interval. Multiple places can also be either 1329 specified by a comma-separated list of places or by an interval. 1330 To specify an interval, a colon followed by the count is placed 1331 after after the hardware thread number or the place. Optionally, 1332 the length can be followed by a colon and the stride number - 1333 otherwise a unit stride is assumed. For instance, the following 1334 specifies the same places list: '"{0,1,2}, {3,4,6}, {7,8,9}, 1335 {10,11,12}"'; '"{0:3}, {3:3}, {7:3}, {10:3}"'; and '"{0:2}:4:3"'. 1336 1337 If 'OMP_PLACES' and 'GOMP_CPU_AFFINITY' are unset and 1338 'OMP_PROC_BIND' is either unset or 'false', threads may be moved 1339 between CPUs following no placement policy. 1340 1341 _See also_: 1342 *note OMP_PROC_BIND::, *note GOMP_CPU_AFFINITY::, *note 1343 omp_get_proc_bind::, *note OMP_DISPLAY_ENV:: 1344 1345 _Reference_: 1346 OpenMP specification v4.0 (http://www.openmp.org/), Section 4.5 1347 1348 1349 File: libgomp.info, Node: OMP_STACKSIZE, Next: OMP_SCHEDULE, Prev: OMP_PLACES, Up: Environment Variables 1350 1351 3.10 'OMP_STACKSIZE' - Set default thread stack size 1352 ==================================================== 1353 1354 _Description_: 1355 Set the default thread stack size in kilobytes, unless the number 1356 is suffixed by 'B', 'K', 'M' or 'G', in which case the size is, 1357 respectively, in bytes, kilobytes, megabytes or gigabytes. This is 1358 different from 'pthread_attr_setstacksize' which gets the number of 1359 bytes as an argument. If the stack size cannot be set due to 1360 system constraints, an error is reported and the initial stack size 1361 is left unchanged. If undefined, the stack size is system 1362 dependent. 1363 1364 _Reference_: 1365 OpenMP specification v4.0 (http://www.openmp.org/), Section 4.7 1366 1367 1368 File: libgomp.info, Node: OMP_SCHEDULE, Next: OMP_THREAD_LIMIT, Prev: OMP_STACKSIZE, Up: Environment Variables 1369 1370 3.11 'OMP_SCHEDULE' - How threads are scheduled 1371 =============================================== 1372 1373 _Description_: 1374 Allows to specify 'schedule type' and 'chunk size'. The value of 1375 the variable shall have the form: 'type[,chunk]' where 'type' is 1376 one of 'static', 'dynamic', 'guided' or 'auto' The optional 'chunk' 1377 size shall be a positive integer. If undefined, dynamic scheduling 1378 and a chunk size of 1 is used. 1379 1380 _See also_: 1381 *note omp_set_schedule:: 1382 1383 _Reference_: 1384 OpenMP specification v4.0 (http://www.openmp.org/), Sections 2.7.1 1385 and 4.1 1386 1387 1388 File: libgomp.info, Node: OMP_THREAD_LIMIT, Next: OMP_WAIT_POLICY, Prev: OMP_SCHEDULE, Up: Environment Variables 1389 1390 3.12 'OMP_THREAD_LIMIT' - Set the maximum number of threads 1391 =========================================================== 1392 1393 _Description_: 1394 Specifies the number of threads to use for the whole program. The 1395 value of this variable shall be a positive integer. If undefined, 1396 the number of threads is not limited. 1397 1398 _See also_: 1399 *note OMP_NUM_THREADS::, *note omp_get_thread_limit:: 1400 1401 _Reference_: 1402 OpenMP specification v4.0 (http://www.openmp.org/), Section 4.10 1403 1404 1405 File: libgomp.info, Node: OMP_WAIT_POLICY, Next: GOMP_CPU_AFFINITY, Prev: OMP_THREAD_LIMIT, Up: Environment Variables 1406 1407 3.13 'OMP_WAIT_POLICY' - How waiting threads are handled 1408 ======================================================== 1409 1410 _Description_: 1411 Specifies whether waiting threads should be active or passive. If 1412 the value is 'PASSIVE', waiting threads should not consume CPU 1413 power while waiting; while the value is 'ACTIVE' specifies that 1414 they should. If undefined, threads wait actively for a short time 1415 before waiting passively. 1416 1417 _See also_: 1418 *note GOMP_SPINCOUNT:: 1419 1420 _Reference_: 1421 OpenMP specification v4.0 (http://www.openmp.org/), Section 4.8 1422 1423 1424 File: libgomp.info, Node: GOMP_CPU_AFFINITY, Next: GOMP_DEBUG, Prev: OMP_WAIT_POLICY, Up: Environment Variables 1425 1426 3.14 'GOMP_CPU_AFFINITY' - Bind threads to specific CPUs 1427 ======================================================== 1428 1429 _Description_: 1430 Binds threads to specific CPUs. The variable should contain a 1431 space-separated or comma-separated list of CPUs. This list may 1432 contain different kinds of entries: either single CPU numbers in 1433 any order, a range of CPUs (M-N) or a range with some stride 1434 (M-N:S). CPU numbers are zero based. For example, 1435 'GOMP_CPU_AFFINITY="0 3 1-2 4-15:2"' will bind the initial thread 1436 to CPU 0, the second to CPU 3, the third to CPU 1, the fourth to 1437 CPU 2, the fifth to CPU 4, the sixth through tenth to CPUs 6, 8, 1438 10, 12, and 14 respectively and then start assigning back from the 1439 beginning of the list. 'GOMP_CPU_AFFINITY=0' binds all threads to 1440 CPU 0. 1441 1442 There is no libgomp library routine to determine whether a CPU 1443 affinity specification is in effect. As a workaround, 1444 language-specific library functions, e.g., 'getenv' in C or 1445 'GET_ENVIRONMENT_VARIABLE' in Fortran, may be used to query the 1446 setting of the 'GOMP_CPU_AFFINITY' environment variable. A defined 1447 CPU affinity on startup cannot be changed or disabled during the 1448 runtime of the application. 1449 1450 If both 'GOMP_CPU_AFFINITY' and 'OMP_PROC_BIND' are set, 1451 'OMP_PROC_BIND' has a higher precedence. If neither has been set 1452 and 'OMP_PROC_BIND' is unset, or when 'OMP_PROC_BIND' is set to 1453 'FALSE', the host system will handle the assignment of threads to 1454 CPUs. 1455 1456 _See also_: 1457 *note OMP_PLACES::, *note OMP_PROC_BIND:: 1458 1459 1460 File: libgomp.info, Node: GOMP_DEBUG, Next: GOMP_STACKSIZE, Prev: GOMP_CPU_AFFINITY, Up: Environment Variables 1461 1462 3.15 'GOMP_DEBUG' - Enable debugging output 1463 =========================================== 1464 1465 _Description_: 1466 Enable debugging output. The variable should be set to '0' 1467 (disabled, also the default if not set), or '1' (enabled). 1468 1469 If enabled, some debugging output will be printed during execution. 1470 This is currently not specified in more detail, and subject to 1471 change. 1472 1473 1474 File: libgomp.info, Node: GOMP_STACKSIZE, Next: GOMP_SPINCOUNT, Prev: GOMP_DEBUG, Up: Environment Variables 1475 1476 3.16 'GOMP_STACKSIZE' - Set default thread stack size 1477 ===================================================== 1478 1479 _Description_: 1480 Set the default thread stack size in kilobytes. This is different 1481 from 'pthread_attr_setstacksize' which gets the number of bytes as 1482 an argument. If the stack size cannot be set due to system 1483 constraints, an error is reported and the initial stack size is 1484 left unchanged. If undefined, the stack size is system dependent. 1485 1486 _See also_: 1487 *note OMP_STACKSIZE:: 1488 1489 _Reference_: 1490 GCC Patches Mailinglist 1491 (http://gcc.gnu.org/ml/gcc-patches/2006-06/msg00493.html), GCC 1492 Patches Mailinglist 1493 (http://gcc.gnu.org/ml/gcc-patches/2006-06/msg00496.html) 1494 1495 1496 File: libgomp.info, Node: GOMP_SPINCOUNT, Prev: GOMP_STACKSIZE, Up: Environment Variables 1497 1498 3.17 'GOMP_SPINCOUNT' - Set the busy-wait spin count 1499 ==================================================== 1500 1501 _Description_: 1502 Determines how long a threads waits actively with consuming CPU 1503 power before waiting passively without consuming CPU power. The 1504 value may be either 'INFINITE', 'INFINITY' to always wait actively 1505 or an integer which gives the number of spins of the busy-wait 1506 loop. The integer may optionally be followed by the following 1507 suffixes acting as multiplication factors: 'k' (kilo, thousand), 1508 'M' (mega, million), 'G' (giga, billion), or 'T' (tera, trillion). 1509 If undefined, 0 is used when 'OMP_WAIT_POLICY' is 'PASSIVE', 1510 300,000 is used when 'OMP_WAIT_POLICY' is undefined and 30 billion 1511 is used when 'OMP_WAIT_POLICY' is 'ACTIVE'. If there are more 1512 OpenMP threads than available CPUs, 1000 and 100 spins are used for 1513 'OMP_WAIT_POLICY' being 'ACTIVE' or undefined, respectively; unless 1514 the 'GOMP_SPINCOUNT' is lower or 'OMP_WAIT_POLICY' is 'PASSIVE'. 1515 1516 _See also_: 1517 *note OMP_WAIT_POLICY:: 1518 1519 1520 File: libgomp.info, Node: The libgomp ABI, Next: Reporting Bugs, Prev: Environment Variables, Up: Top 1521 1522 4 The libgomp ABI 1523 ***************** 1524 1525 The following sections present notes on the external ABI as presented by 1526 libgomp. Only maintainers should need them. 1527 1528 * Menu: 1529 1530 * Implementing MASTER construct:: 1531 * Implementing CRITICAL construct:: 1532 * Implementing ATOMIC construct:: 1533 * Implementing FLUSH construct:: 1534 * Implementing BARRIER construct:: 1535 * Implementing THREADPRIVATE construct:: 1536 * Implementing PRIVATE clause:: 1537 * Implementing FIRSTPRIVATE LASTPRIVATE COPYIN and COPYPRIVATE clauses:: 1538 * Implementing REDUCTION clause:: 1539 * Implementing PARALLEL construct:: 1540 * Implementing FOR construct:: 1541 * Implementing ORDERED construct:: 1542 * Implementing SECTIONS construct:: 1543 * Implementing SINGLE construct:: 1544 1545 1546 File: libgomp.info, Node: Implementing MASTER construct, Next: Implementing CRITICAL construct, Up: The libgomp ABI 1547 1548 4.1 Implementing MASTER construct 1549 ================================= 1550 1551 if (omp_get_thread_num () == 0) 1552 block 1553 1554 Alternately, we generate two copies of the parallel subfunction and 1555 only include this in the version run by the master thread. Surely this 1556 is not worthwhile though... 1557 1558 1559 File: libgomp.info, Node: Implementing CRITICAL construct, Next: Implementing ATOMIC construct, Prev: Implementing MASTER construct, Up: The libgomp ABI 1560 1561 4.2 Implementing CRITICAL construct 1562 =================================== 1563 1564 Without a specified name, 1565 1566 void GOMP_critical_start (void); 1567 void GOMP_critical_end (void); 1568 1569 so that we don't get COPY relocations from libgomp to the main 1570 application. 1571 1572 With a specified name, use omp_set_lock and omp_unset_lock with name 1573 being transformed into a variable declared like 1574 1575 omp_lock_t gomp_critical_user_<name> __attribute__((common)) 1576 1577 Ideally the ABI would specify that all zero is a valid unlocked 1578 state, and so we wouldn't need to initialize this at startup. 1579 1580 1581 File: libgomp.info, Node: Implementing ATOMIC construct, Next: Implementing FLUSH construct, Prev: Implementing CRITICAL construct, Up: The libgomp ABI 1582 1583 4.3 Implementing ATOMIC construct 1584 ================================= 1585 1586 The target should implement the '__sync' builtins. 1587 1588 Failing that we could add 1589 1590 void GOMP_atomic_enter (void) 1591 void GOMP_atomic_exit (void) 1592 1593 which reuses the regular lock code, but with yet another lock object 1594 private to the library. 1595 1596 1597 File: libgomp.info, Node: Implementing FLUSH construct, Next: Implementing BARRIER construct, Prev: Implementing ATOMIC construct, Up: The libgomp ABI 1598 1599 4.4 Implementing FLUSH construct 1600 ================================ 1601 1602 Expands to the '__sync_synchronize' builtin. 1603 1604 1605 File: libgomp.info, Node: Implementing BARRIER construct, Next: Implementing THREADPRIVATE construct, Prev: Implementing FLUSH construct, Up: The libgomp ABI 1606 1607 4.5 Implementing BARRIER construct 1608 ================================== 1609 1610 void GOMP_barrier (void) 1611 1612 1613 File: libgomp.info, Node: Implementing THREADPRIVATE construct, Next: Implementing PRIVATE clause, Prev: Implementing BARRIER construct, Up: The libgomp ABI 1614 1615 4.6 Implementing THREADPRIVATE construct 1616 ======================================== 1617 1618 In _most_ cases we can map this directly to '__thread'. Except that OMP 1619 allows constructors for C++ objects. We can either refuse to support 1620 this (how often is it used?) or we can implement something akin to 1621 .ctors. 1622 1623 Even more ideally, this ctor feature is handled by extensions to the 1624 main pthreads library. Failing that, we can have a set of entry points 1625 to register ctor functions to be called. 1626 1627 1628 File: libgomp.info, Node: Implementing PRIVATE clause, Next: Implementing FIRSTPRIVATE LASTPRIVATE COPYIN and COPYPRIVATE clauses, Prev: Implementing THREADPRIVATE construct, Up: The libgomp ABI 1629 1630 4.7 Implementing PRIVATE clause 1631 =============================== 1632 1633 In association with a PARALLEL, or within the lexical extent of a 1634 PARALLEL block, the variable becomes a local variable in the parallel 1635 subfunction. 1636 1637 In association with FOR or SECTIONS blocks, create a new automatic 1638 variable within the current function. This preserves the semantic of 1639 new variable creation. 1640 1641 1642 File: libgomp.info, Node: Implementing FIRSTPRIVATE LASTPRIVATE COPYIN and COPYPRIVATE clauses, Next: Implementing REDUCTION clause, Prev: Implementing PRIVATE clause, Up: The libgomp ABI 1643 1644 4.8 Implementing FIRSTPRIVATE LASTPRIVATE COPYIN and COPYPRIVATE clauses 1645 ======================================================================== 1646 1647 This seems simple enough for PARALLEL blocks. Create a private struct 1648 for communicating between the parent and subfunction. In the parent, 1649 copy in values for scalar and "small" structs; copy in addresses for 1650 others TREE_ADDRESSABLE types. In the subfunction, copy the value into 1651 the local variable. 1652 1653 It is not clear what to do with bare FOR or SECTION blocks. The only 1654 thing I can figure is that we do something like: 1655 1656 #pragma omp for firstprivate(x) lastprivate(y) 1657 for (int i = 0; i < n; ++i) 1658 body; 1659 1660 which becomes 1661 1662 { 1663 int x = x, y; 1664 1665 // for stuff 1666 1667 if (i == n) 1668 y = y; 1669 } 1670 1671 where the "x=x" and "y=y" assignments actually have different uids 1672 for the two variables, i.e. not something you could write directly in 1673 C. Presumably this only makes sense if the "outer" x and y are global 1674 variables. 1675 1676 COPYPRIVATE would work the same way, except the structure broadcast 1677 would have to happen via SINGLE machinery instead. 1678 1679 1680 File: libgomp.info, Node: Implementing REDUCTION clause, Next: Implementing PARALLEL construct, Prev: Implementing FIRSTPRIVATE LASTPRIVATE COPYIN and COPYPRIVATE clauses, Up: The libgomp ABI 1681 1682 4.9 Implementing REDUCTION clause 1683 ================================= 1684 1685 The private struct mentioned in the previous section should have a 1686 pointer to an array of the type of the variable, indexed by the thread's 1687 TEAM_ID. The thread stores its final value into the array, and after 1688 the barrier, the master thread iterates over the array to collect the 1689 values. 1690 1691 1692 File: libgomp.info, Node: Implementing PARALLEL construct, Next: Implementing FOR construct, Prev: Implementing REDUCTION clause, Up: The libgomp ABI 1693 1694 4.10 Implementing PARALLEL construct 1695 ==================================== 1696 1697 #pragma omp parallel 1698 { 1699 body; 1700 } 1701 1702 becomes 1703 1704 void subfunction (void *data) 1705 { 1706 use data; 1707 body; 1708 } 1709 1710 setup data; 1711 GOMP_parallel_start (subfunction, &data, num_threads); 1712 subfunction (&data); 1713 GOMP_parallel_end (); 1714 1715 void GOMP_parallel_start (void (*fn)(void *), void *data, unsigned num_threads) 1716 1717 The FN argument is the subfunction to be run in parallel. 1718 1719 The DATA argument is a pointer to a structure used to communicate 1720 data in and out of the subfunction, as discussed above with respect to 1721 FIRSTPRIVATE et al. 1722 1723 The NUM_THREADS argument is 1 if an IF clause is present and false, 1724 or the value of the NUM_THREADS clause, if present, or 0. 1725 1726 The function needs to create the appropriate number of threads and/or 1727 launch them from the dock. It needs to create the team structure and 1728 assign team ids. 1729 1730 void GOMP_parallel_end (void) 1731 1732 Tears down the team and returns us to the previous 1733 'omp_in_parallel()' state. 1734 1735 1736 File: libgomp.info, Node: Implementing FOR construct, Next: Implementing ORDERED construct, Prev: Implementing PARALLEL construct, Up: The libgomp ABI 1737 1738 4.11 Implementing FOR construct 1739 =============================== 1740 1741 #pragma omp parallel for 1742 for (i = lb; i <= ub; i++) 1743 body; 1744 1745 becomes 1746 1747 void subfunction (void *data) 1748 { 1749 long _s0, _e0; 1750 while (GOMP_loop_static_next (&_s0, &_e0)) 1751 { 1752 long _e1 = _e0, i; 1753 for (i = _s0; i < _e1; i++) 1754 body; 1755 } 1756 GOMP_loop_end_nowait (); 1757 } 1758 1759 GOMP_parallel_loop_static (subfunction, NULL, 0, lb, ub+1, 1, 0); 1760 subfunction (NULL); 1761 GOMP_parallel_end (); 1762 1763 #pragma omp for schedule(runtime) 1764 for (i = 0; i < n; i++) 1765 body; 1766 1767 becomes 1768 1769 { 1770 long i, _s0, _e0; 1771 if (GOMP_loop_runtime_start (0, n, 1, &_s0, &_e0)) 1772 do { 1773 long _e1 = _e0; 1774 for (i = _s0, i < _e0; i++) 1775 body; 1776 } while (GOMP_loop_runtime_next (&_s0, _&e0)); 1777 GOMP_loop_end (); 1778 } 1779 1780 Note that while it looks like there is trickiness to propagating a 1781 non-constant STEP, there isn't really. We're explicitly allowed to 1782 evaluate it as many times as we want, and any variables involved should 1783 automatically be handled as PRIVATE or SHARED like any other variables. 1784 So the expression should remain evaluable in the subfunction. We can 1785 also pull it into a local variable if we like, but since its supposed to 1786 remain unchanged, we can also not if we like. 1787 1788 If we have SCHEDULE(STATIC), and no ORDERED, then we ought to be able 1789 to get away with no work-sharing context at all, since we can simply 1790 perform the arithmetic directly in each thread to divide up the 1791 iterations. Which would mean that we wouldn't need to call any of these 1792 routines. 1793 1794 There are separate routines for handling loops with an ORDERED 1795 clause. Bookkeeping for that is non-trivial... 1796 1797 1798 File: libgomp.info, Node: Implementing ORDERED construct, Next: Implementing SECTIONS construct, Prev: Implementing FOR construct, Up: The libgomp ABI 1799 1800 4.12 Implementing ORDERED construct 1801 =================================== 1802 1803 void GOMP_ordered_start (void) 1804 void GOMP_ordered_end (void) 1805 1806 1807 File: libgomp.info, Node: Implementing SECTIONS construct, Next: Implementing SINGLE construct, Prev: Implementing ORDERED construct, Up: The libgomp ABI 1808 1809 4.13 Implementing SECTIONS construct 1810 ==================================== 1811 1812 A block as 1813 1814 #pragma omp sections 1815 { 1816 #pragma omp section 1817 stmt1; 1818 #pragma omp section 1819 stmt2; 1820 #pragma omp section 1821 stmt3; 1822 } 1823 1824 becomes 1825 1826 for (i = GOMP_sections_start (3); i != 0; i = GOMP_sections_next ()) 1827 switch (i) 1828 { 1829 case 1: 1830 stmt1; 1831 break; 1832 case 2: 1833 stmt2; 1834 break; 1835 case 3: 1836 stmt3; 1837 break; 1838 } 1839 GOMP_barrier (); 1840 1841 1842 File: libgomp.info, Node: Implementing SINGLE construct, Prev: Implementing SECTIONS construct, Up: The libgomp ABI 1843 1844 4.14 Implementing SINGLE construct 1845 ================================== 1846 1847 A block like 1848 1849 #pragma omp single 1850 { 1851 body; 1852 } 1853 1854 becomes 1855 1856 if (GOMP_single_start ()) 1857 body; 1858 GOMP_barrier (); 1859 1860 while 1861 1862 #pragma omp single copyprivate(x) 1863 body; 1864 1865 becomes 1866 1867 datap = GOMP_single_copy_start (); 1868 if (datap == NULL) 1869 { 1870 body; 1871 data.x = x; 1872 GOMP_single_copy_end (&data); 1873 } 1874 else 1875 x = datap->x; 1876 GOMP_barrier (); 1877 1878 1879 File: libgomp.info, Node: Reporting Bugs, Next: Copying, Prev: The libgomp ABI, Up: Top 1880 1881 5 Reporting Bugs 1882 **************** 1883 1884 Bugs in the GNU Offloading and Multi Processing Runtime Library should 1885 be reported via Bugzilla (http://gcc.gnu.org/bugzilla/). Please add 1886 "openacc", or "openmp", or both to the keywords field in the bug report, 1887 as appropriate. 1888 1889 1890 File: libgomp.info, Node: Copying, Next: GNU Free Documentation License, Prev: Reporting Bugs, Up: Top 1891 1892 GNU General Public License 1893 ************************** 1894 1895 Version 3, 29 June 2007 1896 1897 Copyright (C) 2007 Free Software Foundation, Inc. <http://fsf.org/> 1898 1899 Everyone is permitted to copy and distribute verbatim copies of this 1900 license document, but changing it is not allowed. 1901 1902 Preamble 1903 ======== 1904 1905 The GNU General Public License is a free, copyleft license for software 1906 and other kinds of works. 1907 1908 The licenses for most software and other practical works are designed 1909 to take away your freedom to share and change the works. By contrast, 1910 the GNU General Public License is intended to guarantee your freedom to 1911 share and change all versions of a program-to make sure it remains free 1912 software for all its users. We, the Free Software Foundation, use the 1913 GNU General Public License for most of our software; it applies also to 1914 any other work released this way by its authors. You can apply it to 1915 your programs, too. 1916 1917 When we speak of free software, we are referring to freedom, not 1918 price. Our General Public Licenses are designed to make sure that you 1919 have the freedom to distribute copies of free software (and charge for 1920 them if you wish), that you receive source code or can get it if you 1921 want it, that you can change the software or use pieces of it in new 1922 free programs, and that you know you can do these things. 1923 1924 To protect your rights, we need to prevent others from denying you 1925 these rights or asking you to surrender the rights. Therefore, you have 1926 certain responsibilities if you distribute copies of the software, or if 1927 you modify it: responsibilities to respect the freedom of others. 1928 1929 For example, if you distribute copies of such a program, whether 1930 gratis or for a fee, you must pass on to the recipients the same 1931 freedoms that you received. You must make sure that they, too, receive 1932 or can get the source code. And you must show them these terms so they 1933 know their rights. 1934 1935 Developers that use the GNU GPL protect your rights with two steps: 1936 (1) assert copyright on the software, and (2) offer you this License 1937 giving you legal permission to copy, distribute and/or modify it. 1938 1939 For the developers' and authors' protection, the GPL clearly explains 1940 that there is no warranty for this free software. For both users' and 1941 authors' sake, the GPL requires that modified versions be marked as 1942 changed, so that their problems will not be attributed erroneously to 1943 authors of previous versions. 1944 1945 Some devices are designed to deny users access to install or run 1946 modified versions of the software inside them, although the manufacturer 1947 can do so. This is fundamentally incompatible with the aim of 1948 protecting users' freedom to change the software. The systematic 1949 pattern of such abuse occurs in the area of products for individuals to 1950 use, which is precisely where it is most unacceptable. Therefore, we 1951 have designed this version of the GPL to prohibit the practice for those 1952 products. If such problems arise substantially in other domains, we 1953 stand ready to extend this provision to those domains in future versions 1954 of the GPL, as needed to protect the freedom of users. 1955 1956 Finally, every program is threatened constantly by software patents. 1957 States should not allow patents to restrict development and use of 1958 software on general-purpose computers, but in those that do, we wish to 1959 avoid the special danger that patents applied to a free program could 1960 make it effectively proprietary. To prevent this, the GPL assures that 1961 patents cannot be used to render the program non-free. 1962 1963 The precise terms and conditions for copying, distribution and 1964 modification follow. 1965 1966 TERMS AND CONDITIONS 1967 ==================== 1968 1969 0. Definitions. 1970 1971 "This License" refers to version 3 of the GNU General Public 1972 License. 1973 1974 "Copyright" also means copyright-like laws that apply to other 1975 kinds of works, such as semiconductor masks. 1976 1977 "The Program" refers to any copyrightable work licensed under this 1978 License. Each licensee is addressed as "you". "Licensees" and 1979 "recipients" may be individuals or organizations. 1980 1981 To "modify" a work means to copy from or adapt all or part of the 1982 work in a fashion requiring copyright permission, other than the 1983 making of an exact copy. The resulting work is called a "modified 1984 version" of the earlier work or a work "based on" the earlier work. 1985 1986 A "covered work" means either the unmodified Program or a work 1987 based on the Program. 1988 1989 To "propagate" a work means to do anything with it that, without 1990 permission, would make you directly or secondarily liable for 1991 infringement under applicable copyright law, except executing it on 1992 a computer or modifying a private copy. Propagation includes 1993 copying, distribution (with or without modification), making 1994 available to the public, and in some countries other activities as 1995 well. 1996 1997 To "convey" a work means any kind of propagation that enables other 1998 parties to make or receive copies. Mere interaction with a user 1999 through a computer network, with no transfer of a copy, is not 2000 conveying. 2001 2002 An interactive user interface displays "Appropriate Legal Notices" 2003 to the extent that it includes a convenient and prominently visible 2004 feature that (1) displays an appropriate copyright notice, and (2) 2005 tells the user that there is no warranty for the work (except to 2006 the extent that warranties are provided), that licensees may convey 2007 the work under this License, and how to view a copy of this 2008 License. If the interface presents a list of user commands or 2009 options, such as a menu, a prominent item in the list meets this 2010 criterion. 2011 2012 1. Source Code. 2013 2014 The "source code" for a work means the preferred form of the work 2015 for making modifications to it. "Object code" means any non-source 2016 form of a work. 2017 2018 A "Standard Interface" means an interface that either is an 2019 official standard defined by a recognized standards body, or, in 2020 the case of interfaces specified for a particular programming 2021 language, one that is widely used among developers working in that 2022 language. 2023 2024 The "System Libraries" of an executable work include anything, 2025 other than the work as a whole, that (a) is included in the normal 2026 form of packaging a Major Component, but which is not part of that 2027 Major Component, and (b) serves only to enable use of the work with 2028 that Major Component, or to implement a Standard Interface for 2029 which an implementation is available to the public in source code 2030 form. A "Major Component", in this context, means a major 2031 essential component (kernel, window system, and so on) of the 2032 specific operating system (if any) on which the executable work 2033 runs, or a compiler used to produce the work, or an object code 2034 interpreter used to run it. 2035 2036 The "Corresponding Source" for a work in object code form means all 2037 the source code needed to generate, install, and (for an executable 2038 work) run the object code and to modify the work, including scripts 2039 to control those activities. However, it does not include the 2040 work's System Libraries, or general-purpose tools or generally 2041 available free programs which are used unmodified in performing 2042 those activities but which are not part of the work. For example, 2043 Corresponding Source includes interface definition files associated 2044 with source files for the work, and the source code for shared 2045 libraries and dynamically linked subprograms that the work is 2046 specifically designed to require, such as by intimate data 2047 communication or control flow between those subprograms and other 2048 parts of the work. 2049 2050 The Corresponding Source need not include anything that users can 2051 regenerate automatically from other parts of the Corresponding 2052 Source. 2053 2054 The Corresponding Source for a work in source code form is that 2055 same work. 2056 2057 2. Basic Permissions. 2058 2059 All rights granted under this License are granted for the term of 2060 copyright on the Program, and are irrevocable provided the stated 2061 conditions are met. This License explicitly affirms your unlimited 2062 permission to run the unmodified Program. The output from running 2063 a covered work is covered by this License only if the output, given 2064 its content, constitutes a covered work. This License acknowledges 2065 your rights of fair use or other equivalent, as provided by 2066 copyright law. 2067 2068 You may make, run and propagate covered works that you do not 2069 convey, without conditions so long as your license otherwise 2070 remains in force. You may convey covered works to others for the 2071 sole purpose of having them make modifications exclusively for you, 2072 or provide you with facilities for running those works, provided 2073 that you comply with the terms of this License in conveying all 2074 material for which you do not control copyright. Those thus making 2075 or running the covered works for you must do so exclusively on your 2076 behalf, under your direction and control, on terms that prohibit 2077 them from making any copies of your copyrighted material outside 2078 their relationship with you. 2079 2080 Conveying under any other circumstances is permitted solely under 2081 the conditions stated below. Sublicensing is not allowed; section 2082 10 makes it unnecessary. 2083 2084 3. Protecting Users' Legal Rights From Anti-Circumvention Law. 2085 2086 No covered work shall be deemed part of an effective technological 2087 measure under any applicable law fulfilling obligations under 2088 article 11 of the WIPO copyright treaty adopted on 20 December 2089 1996, or similar laws prohibiting or restricting circumvention of 2090 such measures. 2091 2092 When you convey a covered work, you waive any legal power to forbid 2093 circumvention of technological measures to the extent such 2094 circumvention is effected by exercising rights under this License 2095 with respect to the covered work, and you disclaim any intention to 2096 limit operation or modification of the work as a means of 2097 enforcing, against the work's users, your or third parties' legal 2098 rights to forbid circumvention of technological measures. 2099 2100 4. Conveying Verbatim Copies. 2101 2102 You may convey verbatim copies of the Program's source code as you 2103 receive it, in any medium, provided that you conspicuously and 2104 appropriately publish on each copy an appropriate copyright notice; 2105 keep intact all notices stating that this License and any 2106 non-permissive terms added in accord with section 7 apply to the 2107 code; keep intact all notices of the absence of any warranty; and 2108 give all recipients a copy of this License along with the Program. 2109 2110 You may charge any price or no price for each copy that you convey, 2111 and you may offer support or warranty protection for a fee. 2112 2113 5. Conveying Modified Source Versions. 2114 2115 You may convey a work based on the Program, or the modifications to 2116 produce it from the Program, in the form of source code under the 2117 terms of section 4, provided that you also meet all of these 2118 conditions: 2119 2120 a. The work must carry prominent notices stating that you 2121 modified it, and giving a relevant date. 2122 2123 b. The work must carry prominent notices stating that it is 2124 released under this License and any conditions added under 2125 section 7. This requirement modifies the requirement in 2126 section 4 to "keep intact all notices". 2127 2128 c. You must license the entire work, as a whole, under this 2129 License to anyone who comes into possession of a copy. This 2130 License will therefore apply, along with any applicable 2131 section 7 additional terms, to the whole of the work, and all 2132 its parts, regardless of how they are packaged. This License 2133 gives no permission to license the work in any other way, but 2134 it does not invalidate such permission if you have separately 2135 received it. 2136 2137 d. If the work has interactive user interfaces, each must display 2138 Appropriate Legal Notices; however, if the Program has 2139 interactive interfaces that do not display Appropriate Legal 2140 Notices, your work need not make them do so. 2141 2142 A compilation of a covered work with other separate and independent 2143 works, which are not by their nature extensions of the covered 2144 work, and which are not combined with it such as to form a larger 2145 program, in or on a volume of a storage or distribution medium, is 2146 called an "aggregate" if the compilation and its resulting 2147 copyright are not used to limit the access or legal rights of the 2148 compilation's users beyond what the individual works permit. 2149 Inclusion of a covered work in an aggregate does not cause this 2150 License to apply to the other parts of the aggregate. 2151 2152 6. Conveying Non-Source Forms. 2153 2154 You may convey a covered work in object code form under the terms 2155 of sections 4 and 5, provided that you also convey the 2156 machine-readable Corresponding Source under the terms of this 2157 License, in one of these ways: 2158 2159 a. Convey the object code in, or embodied in, a physical product 2160 (including a physical distribution medium), accompanied by the 2161 Corresponding Source fixed on a durable physical medium 2162 customarily used for software interchange. 2163 2164 b. Convey the object code in, or embodied in, a physical product 2165 (including a physical distribution medium), accompanied by a 2166 written offer, valid for at least three years and valid for as 2167 long as you offer spare parts or customer support for that 2168 product model, to give anyone who possesses the object code 2169 either (1) a copy of the Corresponding Source for all the 2170 software in the product that is covered by this License, on a 2171 durable physical medium customarily used for software 2172 interchange, for a price no more than your reasonable cost of 2173 physically performing this conveying of source, or (2) access 2174 to copy the Corresponding Source from a network server at no 2175 charge. 2176 2177 c. Convey individual copies of the object code with a copy of the 2178 written offer to provide the Corresponding Source. This 2179 alternative is allowed only occasionally and noncommercially, 2180 and only if you received the object code with such an offer, 2181 in accord with subsection 6b. 2182 2183 d. Convey the object code by offering access from a designated 2184 place (gratis or for a charge), and offer equivalent access to 2185 the Corresponding Source in the same way through the same 2186 place at no further charge. You need not require recipients 2187 to copy the Corresponding Source along with the object code. 2188 If the place to copy the object code is a network server, the 2189 Corresponding Source may be on a different server (operated by 2190 you or a third party) that supports equivalent copying 2191 facilities, provided you maintain clear directions next to the 2192 object code saying where to find the Corresponding Source. 2193 Regardless of what server hosts the Corresponding Source, you 2194 remain obligated to ensure that it is available for as long as 2195 needed to satisfy these requirements. 2196 2197 e. Convey the object code using peer-to-peer transmission, 2198 provided you inform other peers where the object code and 2199 Corresponding Source of the work are being offered to the 2200 general public at no charge under subsection 6d. 2201 2202 A separable portion of the object code, whose source code is 2203 excluded from the Corresponding Source as a System Library, need 2204 not be included in conveying the object code work. 2205 2206 A "User Product" is either (1) a "consumer product", which means 2207 any tangible personal property which is normally used for personal, 2208 family, or household purposes, or (2) anything designed or sold for 2209 incorporation into a dwelling. In determining whether a product is 2210 a consumer product, doubtful cases shall be resolved in favor of 2211 coverage. For a particular product received by a particular user, 2212 "normally used" refers to a typical or common use of that class of 2213 product, regardless of the status of the particular user or of the 2214 way in which the particular user actually uses, or expects or is 2215 expected to use, the product. A product is a consumer product 2216 regardless of whether the product has substantial commercial, 2217 industrial or non-consumer uses, unless such uses represent the 2218 only significant mode of use of the product. 2219 2220 "Installation Information" for a User Product means any methods, 2221 procedures, authorization keys, or other information required to 2222 install and execute modified versions of a covered work in that 2223 User Product from a modified version of its Corresponding Source. 2224 The information must suffice to ensure that the continued 2225 functioning of the modified object code is in no case prevented or 2226 interfered with solely because modification has been made. 2227 2228 If you convey an object code work under this section in, or with, 2229 or specifically for use in, a User Product, and the conveying 2230 occurs as part of a transaction in which the right of possession 2231 and use of the User Product is transferred to the recipient in 2232 perpetuity or for a fixed term (regardless of how the transaction 2233 is characterized), the Corresponding Source conveyed under this 2234 section must be accompanied by the Installation Information. But 2235 this requirement does not apply if neither you nor any third party 2236 retains the ability to install modified object code on the User 2237 Product (for example, the work has been installed in ROM). 2238 2239 The requirement to provide Installation Information does not 2240 include a requirement to continue to provide support service, 2241 warranty, or updates for a work that has been modified or installed 2242 by the recipient, or for the User Product in which it has been 2243 modified or installed. Access to a network may be denied when the 2244 modification itself materially and adversely affects the operation 2245 of the network or violates the rules and protocols for 2246 communication across the network. 2247 2248 Corresponding Source conveyed, and Installation Information 2249 provided, in accord with this section must be in a format that is 2250 publicly documented (and with an implementation available to the 2251 public in source code form), and must require no special password 2252 or key for unpacking, reading or copying. 2253 2254 7. Additional Terms. 2255 2256 "Additional permissions" are terms that supplement the terms of 2257 this License by making exceptions from one or more of its 2258 conditions. Additional permissions that are applicable to the 2259 entire Program shall be treated as though they were included in 2260 this License, to the extent that they are valid under applicable 2261 law. If additional permissions apply only to part of the Program, 2262 that part may be used separately under those permissions, but the 2263 entire Program remains governed by this License without regard to 2264 the additional permissions. 2265 2266 When you convey a copy of a covered work, you may at your option 2267 remove any additional permissions from that copy, or from any part 2268 of it. (Additional permissions may be written to require their own 2269 removal in certain cases when you modify the work.) You may place 2270 additional permissions on material, added by you to a covered work, 2271 for which you have or can give appropriate copyright permission. 2272 2273 Notwithstanding any other provision of this License, for material 2274 you add to a covered work, you may (if authorized by the copyright 2275 holders of that material) supplement the terms of this License with 2276 terms: 2277 2278 a. Disclaiming warranty or limiting liability differently from 2279 the terms of sections 15 and 16 of this License; or 2280 2281 b. Requiring preservation of specified reasonable legal notices 2282 or author attributions in that material or in the Appropriate 2283 Legal Notices displayed by works containing it; or 2284 2285 c. Prohibiting misrepresentation of the origin of that material, 2286 or requiring that modified versions of such material be marked 2287 in reasonable ways as different from the original version; or 2288 2289 d. Limiting the use for publicity purposes of names of licensors 2290 or authors of the material; or 2291 2292 e. Declining to grant rights under trademark law for use of some 2293 trade names, trademarks, or service marks; or 2294 2295 f. Requiring indemnification of licensors and authors of that 2296 material by anyone who conveys the material (or modified 2297 versions of it) with contractual assumptions of liability to 2298 the recipient, for any liability that these contractual 2299 assumptions directly impose on those licensors and authors. 2300 2301 All other non-permissive additional terms are considered "further 2302 restrictions" within the meaning of section 10. If the Program as 2303 you received it, or any part of it, contains a notice stating that 2304 it is governed by this License along with a term that is a further 2305 restriction, you may remove that term. If a license document 2306 contains a further restriction but permits relicensing or conveying 2307 under this License, you may add to a covered work material governed 2308 by the terms of that license document, provided that the further 2309 restriction does not survive such relicensing or conveying. 2310 2311 If you add terms to a covered work in accord with this section, you 2312 must place, in the relevant source files, a statement of the 2313 additional terms that apply to those files, or a notice indicating 2314 where to find the applicable terms. 2315 2316 Additional terms, permissive or non-permissive, may be stated in 2317 the form of a separately written license, or stated as exceptions; 2318 the above requirements apply either way. 2319 2320 8. Termination. 2321 2322 You may not propagate or modify a covered work except as expressly 2323 provided under this License. Any attempt otherwise to propagate or 2324 modify it is void, and will automatically terminate your rights 2325 under this License (including any patent licenses granted under the 2326 third paragraph of section 11). 2327 2328 However, if you cease all violation of this License, then your 2329 license from a particular copyright holder is reinstated (a) 2330 provisionally, unless and until the copyright holder explicitly and 2331 finally terminates your license, and (b) permanently, if the 2332 copyright holder fails to notify you of the violation by some 2333 reasonable means prior to 60 days after the cessation. 2334 2335 Moreover, your license from a particular copyright holder is 2336 reinstated permanently if the copyright holder notifies you of the 2337 violation by some reasonable means, this is the first time you have 2338 received notice of violation of this License (for any work) from 2339 that copyright holder, and you cure the violation prior to 30 days 2340 after your receipt of the notice. 2341 2342 Termination of your rights under this section does not terminate 2343 the licenses of parties who have received copies or rights from you 2344 under this License. If your rights have been terminated and not 2345 permanently reinstated, you do not qualify to receive new licenses 2346 for the same material under section 10. 2347 2348 9. Acceptance Not Required for Having Copies. 2349 2350 You are not required to accept this License in order to receive or 2351 run a copy of the Program. Ancillary propagation of a covered work 2352 occurring solely as a consequence of using peer-to-peer 2353 transmission to receive a copy likewise does not require 2354 acceptance. However, nothing other than this License grants you 2355 permission to propagate or modify any covered work. These actions 2356 infringe copyright if you do not accept this License. Therefore, 2357 by modifying or propagating a covered work, you indicate your 2358 acceptance of this License to do so. 2359 2360 10. Automatic Licensing of Downstream Recipients. 2361 2362 Each time you convey a covered work, the recipient automatically 2363 receives a license from the original licensors, to run, modify and 2364 propagate that work, subject to this License. You are not 2365 responsible for enforcing compliance by third parties with this 2366 License. 2367 2368 An "entity transaction" is a transaction transferring control of an 2369 organization, or substantially all assets of one, or subdividing an 2370 organization, or merging organizations. If propagation of a 2371 covered work results from an entity transaction, each party to that 2372 transaction who receives a copy of the work also receives whatever 2373 licenses to the work the party's predecessor in interest had or 2374 could give under the previous paragraph, plus a right to possession 2375 of the Corresponding Source of the work from the predecessor in 2376 interest, if the predecessor has it or can get it with reasonable 2377 efforts. 2378 2379 You may not impose any further restrictions on the exercise of the 2380 rights granted or affirmed under this License. For example, you 2381 may not impose a license fee, royalty, or other charge for exercise 2382 of rights granted under this License, and you may not initiate 2383 litigation (including a cross-claim or counterclaim in a lawsuit) 2384 alleging that any patent claim is infringed by making, using, 2385 selling, offering for sale, or importing the Program or any portion 2386 of it. 2387 2388 11. Patents. 2389 2390 A "contributor" is a copyright holder who authorizes use under this 2391 License of the Program or a work on which the Program is based. 2392 The work thus licensed is called the contributor's "contributor 2393 version". 2394 2395 A contributor's "essential patent claims" are all patent claims 2396 owned or controlled by the contributor, whether already acquired or 2397 hereafter acquired, that would be infringed by some manner, 2398 permitted by this License, of making, using, or selling its 2399 contributor version, but do not include claims that would be 2400 infringed only as a consequence of further modification of the 2401 contributor version. For purposes of this definition, "control" 2402 includes the right to grant patent sublicenses in a manner 2403 consistent with the requirements of this License. 2404 2405 Each contributor grants you a non-exclusive, worldwide, 2406 royalty-free patent license under the contributor's essential 2407 patent claims, to make, use, sell, offer for sale, import and 2408 otherwise run, modify and propagate the contents of its contributor 2409 version. 2410 2411 In the following three paragraphs, a "patent license" is any 2412 express agreement or commitment, however denominated, not to 2413 enforce a patent (such as an express permission to practice a 2414 patent or covenant not to sue for patent infringement). To "grant" 2415 such a patent license to a party means to make such an agreement or 2416 commitment not to enforce a patent against the party. 2417 2418 If you convey a covered work, knowingly relying on a patent 2419 license, and the Corresponding Source of the work is not available 2420 for anyone to copy, free of charge and under the terms of this 2421 License, through a publicly available network server or other 2422 readily accessible means, then you must either (1) cause the 2423 Corresponding Source to be so available, or (2) arrange to deprive 2424 yourself of the benefit of the patent license for this particular 2425 work, or (3) arrange, in a manner consistent with the requirements 2426 of this License, to extend the patent license to downstream 2427 recipients. "Knowingly relying" means you have actual knowledge 2428 that, but for the patent license, your conveying the covered work 2429 in a country, or your recipient's use of the covered work in a 2430 country, would infringe one or more identifiable patents in that 2431 country that you have reason to believe are valid. 2432 2433 If, pursuant to or in connection with a single transaction or 2434 arrangement, you convey, or propagate by procuring conveyance of, a 2435 covered work, and grant a patent license to some of the parties 2436 receiving the covered work authorizing them to use, propagate, 2437 modify or convey a specific copy of the covered work, then the 2438 patent license you grant is automatically extended to all 2439 recipients of the covered work and works based on it. 2440 2441 A patent license is "discriminatory" if it does not include within 2442 the scope of its coverage, prohibits the exercise of, or is 2443 conditioned on the non-exercise of one or more of the rights that 2444 are specifically granted under this License. You may not convey a 2445 covered work if you are a party to an arrangement with a third 2446 party that is in the business of distributing software, under which 2447 you make payment to the third party based on the extent of your 2448 activity of conveying the work, and under which the third party 2449 grants, to any of the parties who would receive the covered work 2450 from you, a discriminatory patent license (a) in connection with 2451 copies of the covered work conveyed by you (or copies made from 2452 those copies), or (b) primarily for and in connection with specific 2453 products or compilations that contain the covered work, unless you 2454 entered into that arrangement, or that patent license was granted, 2455 prior to 28 March 2007. 2456 2457 Nothing in this License shall be construed as excluding or limiting 2458 any implied license or other defenses to infringement that may 2459 otherwise be available to you under applicable patent law. 2460 2461 12. No Surrender of Others' Freedom. 2462 2463 If conditions are imposed on you (whether by court order, agreement 2464 or otherwise) that contradict the conditions of this License, they 2465 do not excuse you from the conditions of this License. If you 2466 cannot convey a covered work so as to satisfy simultaneously your 2467 obligations under this License and any other pertinent obligations, 2468 then as a consequence you may not convey it at all. For example, 2469 if you agree to terms that obligate you to collect a royalty for 2470 further conveying from those to whom you convey the Program, the 2471 only way you could satisfy both those terms and this License would 2472 be to refrain entirely from conveying the Program. 2473 2474 13. Use with the GNU Affero General Public License. 2475 2476 Notwithstanding any other provision of this License, you have 2477 permission to link or combine any covered work with a work licensed 2478 under version 3 of the GNU Affero General Public License into a 2479 single combined work, and to convey the resulting work. The terms 2480 of this License will continue to apply to the part which is the 2481 covered work, but the special requirements of the GNU Affero 2482 General Public License, section 13, concerning interaction through 2483 a network will apply to the combination as such. 2484 2485 14. Revised Versions of this License. 2486 2487 The Free Software Foundation may publish revised and/or new 2488 versions of the GNU General Public License from time to time. Such 2489 new versions will be similar in spirit to the present version, but 2490 may differ in detail to address new problems or concerns. 2491 2492 Each version is given a distinguishing version number. If the 2493 Program specifies that a certain numbered version of the GNU 2494 General Public License "or any later version" applies to it, you 2495 have the option of following the terms and conditions either of 2496 that numbered version or of any later version published by the Free 2497 Software Foundation. If the Program does not specify a version 2498 number of the GNU General Public License, you may choose any 2499 version ever published by the Free Software Foundation. 2500 2501 If the Program specifies that a proxy can decide which future 2502 versions of the GNU General Public License can be used, that 2503 proxy's public statement of acceptance of a version permanently 2504 authorizes you to choose that version for the Program. 2505 2506 Later license versions may give you additional or different 2507 permissions. However, no additional obligations are imposed on any 2508 author or copyright holder as a result of your choosing to follow a 2509 later version. 2510 2511 15. Disclaimer of Warranty. 2512 2513 THERE IS NO WARRANTY FOR THE PROGRAM, TO THE EXTENT PERMITTED BY 2514 APPLICABLE LAW. EXCEPT WHEN OTHERWISE STATED IN WRITING THE 2515 COPYRIGHT HOLDERS AND/OR OTHER PARTIES PROVIDE THE PROGRAM "AS IS" 2516 WITHOUT WARRANTY OF ANY KIND, EITHER EXPRESSED OR IMPLIED, 2517 INCLUDING, BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF 2518 MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE. THE ENTIRE 2519 RISK AS TO THE QUALITY AND PERFORMANCE OF THE PROGRAM IS WITH YOU. 2520 SHOULD THE PROGRAM PROVE DEFECTIVE, YOU ASSUME THE COST OF ALL 2521 NECESSARY SERVICING, REPAIR OR CORRECTION. 2522 2523 16. Limitation of Liability. 2524 2525 IN NO EVENT UNLESS REQUIRED BY APPLICABLE LAW OR AGREED TO IN 2526 WRITING WILL ANY COPYRIGHT HOLDER, OR ANY OTHER PARTY WHO MODIFIES 2527 AND/OR CONVEYS THE PROGRAM AS PERMITTED ABOVE, BE LIABLE TO YOU FOR 2528 DAMAGES, INCLUDING ANY GENERAL, SPECIAL, INCIDENTAL OR 2529 CONSEQUENTIAL DAMAGES ARISING OUT OF THE USE OR INABILITY TO USE 2530 THE PROGRAM (INCLUDING BUT NOT LIMITED TO LOSS OF DATA OR DATA 2531 BEING RENDERED INACCURATE OR LOSSES SUSTAINED BY YOU OR THIRD 2532 PARTIES OR A FAILURE OF THE PROGRAM TO OPERATE WITH ANY OTHER 2533 PROGRAMS), EVEN IF SUCH HOLDER OR OTHER PARTY HAS BEEN ADVISED OF 2534 THE POSSIBILITY OF SUCH DAMAGES. 2535 2536 17. Interpretation of Sections 15 and 16. 2537 2538 If the disclaimer of warranty and limitation of liability provided 2539 above cannot be given local legal effect according to their terms, 2540 reviewing courts shall apply local law that most closely 2541 approximates an absolute waiver of all civil liability in 2542 connection with the Program, unless a warranty or assumption of 2543 liability accompanies a copy of the Program in return for a fee. 2544 2545 END OF TERMS AND CONDITIONS 2546 =========================== 2547 2548 How to Apply These Terms to Your New Programs 2549 ============================================= 2550 2551 If you develop a new program, and you want it to be of the greatest 2552 possible use to the public, the best way to achieve this is to make it 2553 free software which everyone can redistribute and change under these 2554 terms. 2555 2556 To do so, attach the following notices to the program. It is safest 2557 to attach them to the start of each source file to most effectively 2558 state the exclusion of warranty; and each file should have at least the 2559 "copyright" line and a pointer to where the full notice is found. 2560 2561 ONE LINE TO GIVE THE PROGRAM'S NAME AND A BRIEF IDEA OF WHAT IT DOES. 2562 Copyright (C) YEAR NAME OF AUTHOR 2563 2564 This program is free software: you can redistribute it and/or modify 2565 it under the terms of the GNU General Public License as published by 2566 the Free Software Foundation, either version 3 of the License, or (at 2567 your option) any later version. 2568 2569 This program is distributed in the hope that it will be useful, but 2570 WITHOUT ANY WARRANTY; without even the implied warranty of 2571 MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU 2572 General Public License for more details. 2573 2574 You should have received a copy of the GNU General Public License 2575 along with this program. If not, see <http://www.gnu.org/licenses/>. 2576 2577 Also add information on how to contact you by electronic and paper 2578 mail. 2579 2580 If the program does terminal interaction, make it output a short 2581 notice like this when it starts in an interactive mode: 2582 2583 PROGRAM Copyright (C) YEAR NAME OF AUTHOR 2584 This program comes with ABSOLUTELY NO WARRANTY; for details type 'show w'. 2585 This is free software, and you are welcome to redistribute it 2586 under certain conditions; type 'show c' for details. 2587 2588 The hypothetical commands 'show w' and 'show c' should show the 2589 appropriate parts of the General Public License. Of course, your 2590 program's commands might be different; for a GUI interface, you would 2591 use an "about box". 2592 2593 You should also get your employer (if you work as a programmer) or 2594 school, if any, to sign a "copyright disclaimer" for the program, if 2595 necessary. For more information on this, and how to apply and follow 2596 the GNU GPL, see <http://www.gnu.org/licenses/>. 2597 2598 The GNU General Public License does not permit incorporating your 2599 program into proprietary programs. If your program is a subroutine 2600 library, you may consider it more useful to permit linking proprietary 2601 applications with the library. If this is what you want to do, use the 2602 GNU Lesser General Public License instead of this License. But first, 2603 please read <http://www.gnu.org/philosophy/why-not-lgpl.html>. 2604 2605 2606 File: libgomp.info, Node: GNU Free Documentation License, Next: Funding, Prev: Copying, Up: Top 2607 2608 GNU Free Documentation License 2609 ****************************** 2610 2611 Version 1.3, 3 November 2008 2612 2613 Copyright (C) 2000, 2001, 2002, 2007, 2008 Free Software Foundation, Inc. 2614 <http://fsf.org/> 2615 2616 Everyone is permitted to copy and distribute verbatim copies 2617 of this license document, but changing it is not allowed. 2618 2619 0. PREAMBLE 2620 2621 The purpose of this License is to make a manual, textbook, or other 2622 functional and useful document "free" in the sense of freedom: to 2623 assure everyone the effective freedom to copy and redistribute it, 2624 with or without modifying it, either commercially or 2625 noncommercially. Secondarily, this License preserves for the 2626 author and publisher a way to get credit for their work, while not 2627 being considered responsible for modifications made by others. 2628 2629 This License is a kind of "copyleft", which means that derivative 2630 works of the document must themselves be free in the same sense. 2631 It complements the GNU General Public License, which is a copyleft 2632 license designed for free software. 2633 2634 We have designed this License in order to use it for manuals for 2635 free software, because free software needs free documentation: a 2636 free program should come with manuals providing the same freedoms 2637 that the software does. But this License is not limited to 2638 software manuals; it can be used for any textual work, regardless 2639 of subject matter or whether it is published as a printed book. We 2640 recommend this License principally for works whose purpose is 2641 instruction or reference. 2642 2643 1. APPLICABILITY AND DEFINITIONS 2644 2645 This License applies to any manual or other work, in any medium, 2646 that contains a notice placed by the copyright holder saying it can 2647 be distributed under the terms of this License. Such a notice 2648 grants a world-wide, royalty-free license, unlimited in duration, 2649 to use that work under the conditions stated herein. The 2650 "Document", below, refers to any such manual or work. Any member 2651 of the public is a licensee, and is addressed as "you". You accept 2652 the license if you copy, modify or distribute the work in a way 2653 requiring permission under copyright law. 2654 2655 A "Modified Version" of the Document means any work containing the 2656 Document or a portion of it, either copied verbatim, or with 2657 modifications and/or translated into another language. 2658 2659 A "Secondary Section" is a named appendix or a front-matter section 2660 of the Document that deals exclusively with the relationship of the 2661 publishers or authors of the Document to the Document's overall 2662 subject (or to related matters) and contains nothing that could 2663 fall directly within that overall subject. (Thus, if the Document 2664 is in part a textbook of mathematics, a Secondary Section may not 2665 explain any mathematics.) The relationship could be a matter of 2666 historical connection with the subject or with related matters, or 2667 of legal, commercial, philosophical, ethical or political position 2668 regarding them. 2669 2670 The "Invariant Sections" are certain Secondary Sections whose 2671 titles are designated, as being those of Invariant Sections, in the 2672 notice that says that the Document is released under this License. 2673 If a section does not fit the above definition of Secondary then it 2674 is not allowed to be designated as Invariant. The Document may 2675 contain zero Invariant Sections. If the Document does not identify 2676 any Invariant Sections then there are none. 2677 2678 The "Cover Texts" are certain short passages of text that are 2679 listed, as Front-Cover Texts or Back-Cover Texts, in the notice 2680 that says that the Document is released under this License. A 2681 Front-Cover Text may be at most 5 words, and a Back-Cover Text may 2682 be at most 25 words. 2683 2684 A "Transparent" copy of the Document means a machine-readable copy, 2685 represented in a format whose specification is available to the 2686 general public, that is suitable for revising the document 2687 straightforwardly with generic text editors or (for images composed 2688 of pixels) generic paint programs or (for drawings) some widely 2689 available drawing editor, and that is suitable for input to text 2690 formatters or for automatic translation to a variety of formats 2691 suitable for input to text formatters. A copy made in an otherwise 2692 Transparent file format whose markup, or absence of markup, has 2693 been arranged to thwart or discourage subsequent modification by 2694 readers is not Transparent. An image format is not Transparent if 2695 used for any substantial amount of text. A copy that is not 2696 "Transparent" is called "Opaque". 2697 2698 Examples of suitable formats for Transparent copies include plain 2699 ASCII without markup, Texinfo input format, LaTeX input format, 2700 SGML or XML using a publicly available DTD, and standard-conforming 2701 simple HTML, PostScript or PDF designed for human modification. 2702 Examples of transparent image formats include PNG, XCF and JPG. 2703 Opaque formats include proprietary formats that can be read and 2704 edited only by proprietary word processors, SGML or XML for which 2705 the DTD and/or processing tools are not generally available, and 2706 the machine-generated HTML, PostScript or PDF produced by some word 2707 processors for output purposes only. 2708 2709 The "Title Page" means, for a printed book, the title page itself, 2710 plus such following pages as are needed to hold, legibly, the 2711 material this License requires to appear in the title page. For 2712 works in formats which do not have any title page as such, "Title 2713 Page" means the text near the most prominent appearance of the 2714 work's title, preceding the beginning of the body of the text. 2715 2716 The "publisher" means any person or entity that distributes copies 2717 of the Document to the public. 2718 2719 A section "Entitled XYZ" means a named subunit of the Document 2720 whose title either is precisely XYZ or contains XYZ in parentheses 2721 following text that translates XYZ in another language. (Here XYZ 2722 stands for a specific section name mentioned below, such as 2723 "Acknowledgements", "Dedications", "Endorsements", or "History".) 2724 To "Preserve the Title" of such a section when you modify the 2725 Document means that it remains a section "Entitled XYZ" according 2726 to this definition. 2727 2728 The Document may include Warranty Disclaimers next to the notice 2729 which states that this License applies to the Document. These 2730 Warranty Disclaimers are considered to be included by reference in 2731 this License, but only as regards disclaiming warranties: any other 2732 implication that these Warranty Disclaimers may have is void and 2733 has no effect on the meaning of this License. 2734 2735 2. VERBATIM COPYING 2736 2737 You may copy and distribute the Document in any medium, either 2738 commercially or noncommercially, provided that this License, the 2739 copyright notices, and the license notice saying this License 2740 applies to the Document are reproduced in all copies, and that you 2741 add no other conditions whatsoever to those of this License. You 2742 may not use technical measures to obstruct or control the reading 2743 or further copying of the copies you make or distribute. However, 2744 you may accept compensation in exchange for copies. If you 2745 distribute a large enough number of copies you must also follow the 2746 conditions in section 3. 2747 2748 You may also lend copies, under the same conditions stated above, 2749 and you may publicly display copies. 2750 2751 3. COPYING IN QUANTITY 2752 2753 If you publish printed copies (or copies in media that commonly 2754 have printed covers) of the Document, numbering more than 100, and 2755 the Document's license notice requires Cover Texts, you must 2756 enclose the copies in covers that carry, clearly and legibly, all 2757 these Cover Texts: Front-Cover Texts on the front cover, and 2758 Back-Cover Texts on the back cover. Both covers must also clearly 2759 and legibly identify you as the publisher of these copies. The 2760 front cover must present the full title with all words of the title 2761 equally prominent and visible. You may add other material on the 2762 covers in addition. Copying with changes limited to the covers, as 2763 long as they preserve the title of the Document and satisfy these 2764 conditions, can be treated as verbatim copying in other respects. 2765 2766 If the required texts for either cover are too voluminous to fit 2767 legibly, you should put the first ones listed (as many as fit 2768 reasonably) on the actual cover, and continue the rest onto 2769 adjacent pages. 2770 2771 If you publish or distribute Opaque copies of the Document 2772 numbering more than 100, you must either include a machine-readable 2773 Transparent copy along with each Opaque copy, or state in or with 2774 each Opaque copy a computer-network location from which the general 2775 network-using public has access to download using public-standard 2776 network protocols a complete Transparent copy of the Document, free 2777 of added material. If you use the latter option, you must take 2778 reasonably prudent steps, when you begin distribution of Opaque 2779 copies in quantity, to ensure that this Transparent copy will 2780 remain thus accessible at the stated location until at least one 2781 year after the last time you distribute an Opaque copy (directly or 2782 through your agents or retailers) of that edition to the public. 2783 2784 It is requested, but not required, that you contact the authors of 2785 the Document well before redistributing any large number of copies, 2786 to give them a chance to provide you with an updated version of the 2787 Document. 2788 2789 4. MODIFICATIONS 2790 2791 You may copy and distribute a Modified Version of the Document 2792 under the conditions of sections 2 and 3 above, provided that you 2793 release the Modified Version under precisely this License, with the 2794 Modified Version filling the role of the Document, thus licensing 2795 distribution and modification of the Modified Version to whoever 2796 possesses a copy of it. In addition, you must do these things in 2797 the Modified Version: 2798 2799 A. Use in the Title Page (and on the covers, if any) a title 2800 distinct from that of the Document, and from those of previous 2801 versions (which should, if there were any, be listed in the 2802 History section of the Document). You may use the same title 2803 as a previous version if the original publisher of that 2804 version gives permission. 2805 2806 B. List on the Title Page, as authors, one or more persons or 2807 entities responsible for authorship of the modifications in 2808 the Modified Version, together with at least five of the 2809 principal authors of the Document (all of its principal 2810 authors, if it has fewer than five), unless they release you 2811 from this requirement. 2812 2813 C. State on the Title page the name of the publisher of the 2814 Modified Version, as the publisher. 2815 2816 D. Preserve all the copyright notices of the Document. 2817 2818 E. Add an appropriate copyright notice for your modifications 2819 adjacent to the other copyright notices. 2820 2821 F. Include, immediately after the copyright notices, a license 2822 notice giving the public permission to use the Modified 2823 Version under the terms of this License, in the form shown in 2824 the Addendum below. 2825 2826 G. Preserve in that license notice the full lists of Invariant 2827 Sections and required Cover Texts given in the Document's 2828 license notice. 2829 2830 H. Include an unaltered copy of this License. 2831 2832 I. Preserve the section Entitled "History", Preserve its Title, 2833 and add to it an item stating at least the title, year, new 2834 authors, and publisher of the Modified Version as given on the 2835 Title Page. If there is no section Entitled "History" in the 2836 Document, create one stating the title, year, authors, and 2837 publisher of the Document as given on its Title Page, then add 2838 an item describing the Modified Version as stated in the 2839 previous sentence. 2840 2841 J. Preserve the network location, if any, given in the Document 2842 for public access to a Transparent copy of the Document, and 2843 likewise the network locations given in the Document for 2844 previous versions it was based on. These may be placed in the 2845 "History" section. You may omit a network location for a work 2846 that was published at least four years before the Document 2847 itself, or if the original publisher of the version it refers 2848 to gives permission. 2849 2850 K. For any section Entitled "Acknowledgements" or "Dedications", 2851 Preserve the Title of the section, and preserve in the section 2852 all the substance and tone of each of the contributor 2853 acknowledgements and/or dedications given therein. 2854 2855 L. Preserve all the Invariant Sections of the Document, unaltered 2856 in their text and in their titles. Section numbers or the 2857 equivalent are not considered part of the section titles. 2858 2859 M. Delete any section Entitled "Endorsements". Such a section 2860 may not be included in the Modified Version. 2861 2862 N. Do not retitle any existing section to be Entitled 2863 "Endorsements" or to conflict in title with any Invariant 2864 Section. 2865 2866 O. Preserve any Warranty Disclaimers. 2867 2868 If the Modified Version includes new front-matter sections or 2869 appendices that qualify as Secondary Sections and contain no 2870 material copied from the Document, you may at your option designate 2871 some or all of these sections as invariant. To do this, add their 2872 titles to the list of Invariant Sections in the Modified Version's 2873 license notice. These titles must be distinct from any other 2874 section titles. 2875 2876 You may add a section Entitled "Endorsements", provided it contains 2877 nothing but endorsements of your Modified Version by various 2878 parties--for example, statements of peer review or that the text 2879 has been approved by an organization as the authoritative 2880 definition of a standard. 2881 2882 You may add a passage of up to five words as a Front-Cover Text, 2883 and a passage of up to 25 words as a Back-Cover Text, to the end of 2884 the list of Cover Texts in the Modified Version. Only one passage 2885 of Front-Cover Text and one of Back-Cover Text may be added by (or 2886 through arrangements made by) any one entity. If the Document 2887 already includes a cover text for the same cover, previously added 2888 by you or by arrangement made by the same entity you are acting on 2889 behalf of, you may not add another; but you may replace the old 2890 one, on explicit permission from the previous publisher that added 2891 the old one. 2892 2893 The author(s) and publisher(s) of the Document do not by this 2894 License give permission to use their names for publicity for or to 2895 assert or imply endorsement of any Modified Version. 2896 2897 5. COMBINING DOCUMENTS 2898 2899 You may combine the Document with other documents released under 2900 this License, under the terms defined in section 4 above for 2901 modified versions, provided that you include in the combination all 2902 of the Invariant Sections of all of the original documents, 2903 unmodified, and list them all as Invariant Sections of your 2904 combined work in its license notice, and that you preserve all 2905 their Warranty Disclaimers. 2906 2907 The combined work need only contain one copy of this License, and 2908 multiple identical Invariant Sections may be replaced with a single 2909 copy. If there are multiple Invariant Sections with the same name 2910 but different contents, make the title of each such section unique 2911 by adding at the end of it, in parentheses, the name of the 2912 original author or publisher of that section if known, or else a 2913 unique number. Make the same adjustment to the section titles in 2914 the list of Invariant Sections in the license notice of the 2915 combined work. 2916 2917 In the combination, you must combine any sections Entitled 2918 "History" in the various original documents, forming one section 2919 Entitled "History"; likewise combine any sections Entitled 2920 "Acknowledgements", and any sections Entitled "Dedications". You 2921 must delete all sections Entitled "Endorsements." 2922 2923 6. COLLECTIONS OF DOCUMENTS 2924 2925 You may make a collection consisting of the Document and other 2926 documents released under this License, and replace the individual 2927 copies of this License in the various documents with a single copy 2928 that is included in the collection, provided that you follow the 2929 rules of this License for verbatim copying of each of the documents 2930 in all other respects. 2931 2932 You may extract a single document from such a collection, and 2933 distribute it individually under this License, provided you insert 2934 a copy of this License into the extracted document, and follow this 2935 License in all other respects regarding verbatim copying of that 2936 document. 2937 2938 7. AGGREGATION WITH INDEPENDENT WORKS 2939 2940 A compilation of the Document or its derivatives with other 2941 separate and independent documents or works, in or on a volume of a 2942 storage or distribution medium, is called an "aggregate" if the 2943 copyright resulting from the compilation is not used to limit the 2944 legal rights of the compilation's users beyond what the individual 2945 works permit. When the Document is included in an aggregate, this 2946 License does not apply to the other works in the aggregate which 2947 are not themselves derivative works of the Document. 2948 2949 If the Cover Text requirement of section 3 is applicable to these 2950 copies of the Document, then if the Document is less than one half 2951 of the entire aggregate, the Document's Cover Texts may be placed 2952 on covers that bracket the Document within the aggregate, or the 2953 electronic equivalent of covers if the Document is in electronic 2954 form. Otherwise they must appear on printed covers that bracket 2955 the whole aggregate. 2956 2957 8. TRANSLATION 2958 2959 Translation is considered a kind of modification, so you may 2960 distribute translations of the Document under the terms of section 2961 4. Replacing Invariant Sections with translations requires special 2962 permission from their copyright holders, but you may include 2963 translations of some or all Invariant Sections in addition to the 2964 original versions of these Invariant Sections. You may include a 2965 translation of this License, and all the license notices in the 2966 Document, and any Warranty Disclaimers, provided that you also 2967 include the original English version of this License and the 2968 original versions of those notices and disclaimers. In case of a 2969 disagreement between the translation and the original version of 2970 this License or a notice or disclaimer, the original version will 2971 prevail. 2972 2973 If a section in the Document is Entitled "Acknowledgements", 2974 "Dedications", or "History", the requirement (section 4) to 2975 Preserve its Title (section 1) will typically require changing the 2976 actual title. 2977 2978 9. TERMINATION 2979 2980 You may not copy, modify, sublicense, or distribute the Document 2981 except as expressly provided under this License. Any attempt 2982 otherwise to copy, modify, sublicense, or distribute it is void, 2983 and will automatically terminate your rights under this License. 2984 2985 However, if you cease all violation of this License, then your 2986 license from a particular copyright holder is reinstated (a) 2987 provisionally, unless and until the copyright holder explicitly and 2988 finally terminates your license, and (b) permanently, if the 2989 copyright holder fails to notify you of the violation by some 2990 reasonable means prior to 60 days after the cessation. 2991 2992 Moreover, your license from a particular copyright holder is 2993 reinstated permanently if the copyright holder notifies you of the 2994 violation by some reasonable means, this is the first time you have 2995 received notice of violation of this License (for any work) from 2996 that copyright holder, and you cure the violation prior to 30 days 2997 after your receipt of the notice. 2998 2999 Termination of your rights under this section does not terminate 3000 the licenses of parties who have received copies or rights from you 3001 under this License. If your rights have been terminated and not 3002 permanently reinstated, receipt of a copy of some or all of the 3003 same material does not give you any rights to use it. 3004 3005 10. FUTURE REVISIONS OF THIS LICENSE 3006 3007 The Free Software Foundation may publish new, revised versions of 3008 the GNU Free Documentation License from time to time. Such new 3009 versions will be similar in spirit to the present version, but may 3010 differ in detail to address new problems or concerns. See 3011 <http://www.gnu.org/copyleft/>. 3012 3013 Each version of the License is given a distinguishing version 3014 number. If the Document specifies that a particular numbered 3015 version of this License "or any later version" applies to it, you 3016 have the option of following the terms and conditions either of 3017 that specified version or of any later version that has been 3018 published (not as a draft) by the Free Software Foundation. If the 3019 Document does not specify a version number of this License, you may 3020 choose any version ever published (not as a draft) by the Free 3021 Software Foundation. If the Document specifies that a proxy can 3022 decide which future versions of this License can be used, that 3023 proxy's public statement of acceptance of a version permanently 3024 authorizes you to choose that version for the Document. 3025 3026 11. RELICENSING 3027 3028 "Massive Multiauthor Collaboration Site" (or "MMC Site") means any 3029 World Wide Web server that publishes copyrightable works and also 3030 provides prominent facilities for anybody to edit those works. A 3031 public wiki that anybody can edit is an example of such a server. 3032 A "Massive Multiauthor Collaboration" (or "MMC") contained in the 3033 site means any set of copyrightable works thus published on the MMC 3034 site. 3035 3036 "CC-BY-SA" means the Creative Commons Attribution-Share Alike 3.0 3037 license published by Creative Commons Corporation, a not-for-profit 3038 corporation with a principal place of business in San Francisco, 3039 California, as well as future copyleft versions of that license 3040 published by that same organization. 3041 3042 "Incorporate" means to publish or republish a Document, in whole or 3043 in part, as part of another Document. 3044 3045 An MMC is "eligible for relicensing" if it is licensed under this 3046 License, and if all works that were first published under this 3047 License somewhere other than this MMC, and subsequently 3048 incorporated in whole or in part into the MMC, (1) had no cover 3049 texts or invariant sections, and (2) were thus incorporated prior 3050 to November 1, 2008. 3051 3052 The operator of an MMC Site may republish an MMC contained in the 3053 site under CC-BY-SA on the same site at any time before August 1, 3054 2009, provided the MMC is eligible for relicensing. 3055 3056 ADDENDUM: How to use this License for your documents 3057 ==================================================== 3058 3059 To use this License in a document you have written, include a copy of 3060 the License in the document and put the following copyright and license 3061 notices just after the title page: 3062 3063 Copyright (C) YEAR YOUR NAME. 3064 Permission is granted to copy, distribute and/or modify this document 3065 under the terms of the GNU Free Documentation License, Version 1.3 3066 or any later version published by the Free Software Foundation; 3067 with no Invariant Sections, no Front-Cover Texts, and no Back-Cover 3068 Texts. A copy of the license is included in the section entitled ``GNU 3069 Free Documentation License''. 3070 3071 If you have Invariant Sections, Front-Cover Texts and Back-Cover 3072 Texts, replace the "with...Texts." line with this: 3073 3074 with the Invariant Sections being LIST THEIR TITLES, with 3075 the Front-Cover Texts being LIST, and with the Back-Cover Texts 3076 being LIST. 3077 3078 If you have Invariant Sections without Cover Texts, or some other 3079 combination of the three, merge those two alternatives to suit the 3080 situation. 3081 3082 If your document contains nontrivial examples of program code, we 3083 recommend releasing these examples in parallel under your choice of free 3084 software license, such as the GNU General Public License, to permit 3085 their use in free software. 3086 3087 3088 File: libgomp.info, Node: Funding, Next: Library Index, Prev: GNU Free Documentation License, Up: Top 3089 3090 Funding Free Software 3091 ********************* 3092 3093 If you want to have more free software a few years from now, it makes 3094 sense for you to help encourage people to contribute funds for its 3095 development. The most effective approach known is to encourage 3096 commercial redistributors to donate. 3097 3098 Users of free software systems can boost the pace of development by 3099 encouraging for-a-fee distributors to donate part of their selling price 3100 to free software developers--the Free Software Foundation, and others. 3101 3102 The way to convince distributors to do this is to demand it and 3103 expect it from them. So when you compare distributors, judge them 3104 partly by how much they give to free software development. Show 3105 distributors they must compete to be the one who gives the most. 3106 3107 To make this approach work, you must insist on numbers that you can 3108 compare, such as, "We will donate ten dollars to the Frobnitz project 3109 for each disk sold." Don't be satisfied with a vague promise, such as 3110 "A portion of the profits are donated," since it doesn't give a basis 3111 for comparison. 3112 3113 Even a precise fraction "of the profits from this disk" is not very 3114 meaningful, since creative accounting and unrelated business decisions 3115 can greatly alter what fraction of the sales price counts as profit. If 3116 the price you pay is $50, ten percent of the profit is probably less 3117 than a dollar; it might be a few cents, or nothing at all. 3118 3119 Some redistributors do development work themselves. This is useful 3120 too; but to keep everyone honest, you need to inquire how much they do, 3121 and what kind. Some kinds of development make much more long-term 3122 difference than others. For example, maintaining a separate version of 3123 a program contributes very little; maintaining the standard version of a 3124 program for the whole community contributes much. Easy new ports 3125 contribute little, since someone else would surely do them; difficult 3126 ports such as adding a new CPU to the GNU Compiler Collection contribute 3127 more; major new features or packages contribute the most. 3128 3129 By establishing the idea that supporting further development is "the 3130 proper thing to do" when distributing free software for a fee, we can 3131 assure a steady flow of resources into making more free software. 3132 3133 Copyright (C) 1994 Free Software Foundation, Inc. 3134 Verbatim copying and redistribution of this section is permitted 3135 without royalty; alteration is not permitted. 3136 3137 3138 File: libgomp.info, Node: Library Index, Prev: Funding, Up: Top 3139 3140 Library Index 3141 ************* 3142 3143 [index] 3144 * Menu: 3145 3146 * Environment Variable: OMP_CANCELLATION. (line 6) 3147 * Environment Variable <1>: OMP_DISPLAY_ENV. (line 6) 3148 * Environment Variable <2>: OMP_DEFAULT_DEVICE. (line 6) 3149 * Environment Variable <3>: OMP_DYNAMIC. (line 6) 3150 * Environment Variable <4>: OMP_MAX_ACTIVE_LEVELS. (line 6) 3151 * Environment Variable <5>: OMP_NESTED. (line 6) 3152 * Environment Variable <6>: OMP_NUM_THREADS. (line 6) 3153 * Environment Variable <7>: OMP_PROC_BIND. (line 6) 3154 * Environment Variable <8>: OMP_PLACES. (line 6) 3155 * Environment Variable <9>: OMP_STACKSIZE. (line 6) 3156 * Environment Variable <10>: OMP_SCHEDULE. (line 6) 3157 * Environment Variable <11>: OMP_THREAD_LIMIT. (line 6) 3158 * Environment Variable <12>: OMP_WAIT_POLICY. (line 6) 3159 * Environment Variable <13>: GOMP_CPU_AFFINITY. (line 6) 3160 * Environment Variable <14>: GOMP_DEBUG. (line 6) 3161 * Environment Variable <15>: GOMP_STACKSIZE. (line 6) 3162 * Environment Variable <16>: GOMP_SPINCOUNT. (line 6) 3163 * FDL, GNU Free Documentation License: GNU Free Documentation License. 3164 (line 6) 3165 * Implementation specific setting: OMP_NESTED. (line 6) 3166 * Implementation specific setting <1>: OMP_NUM_THREADS. (line 6) 3167 * Implementation specific setting <2>: OMP_SCHEDULE. (line 6) 3168 * Implementation specific setting <3>: GOMP_STACKSIZE. (line 6) 3169 * Implementation specific setting <4>: GOMP_SPINCOUNT. (line 6) 3170 * Introduction: Top. (line 6) 3171 3172 3173 3174 Tag Table: 3175 Node: Top2083 3176 Node: Enabling OpenMP3878 3177 Node: Runtime Library Routines4665 3178 Node: omp_get_active_level7652 3179 Node: omp_get_ancestor_thread_num8352 3180 Node: omp_get_cancellation9282 3181 Node: omp_get_default_device10096 3182 Node: omp_get_dynamic10772 3183 Node: omp_get_level11647 3184 Node: omp_get_max_active_levels12267 3185 Node: omp_get_max_threads12966 3186 Node: omp_get_nested13723 3187 Node: omp_get_num_devices14635 3188 Node: omp_get_num_procs15156 3189 Node: omp_get_num_teams15695 3190 Node: omp_get_num_threads16211 3191 Node: omp_get_proc_bind17300 3192 Node: omp_get_schedule18221 3193 Node: omp_get_team_num19167 3194 Node: omp_get_team_size19666 3195 Node: omp_get_thread_limit20626 3196 Node: omp_get_thread_num21245 3197 Node: omp_in_parallel22116 3198 Node: omp_in_final22765 3199 Node: omp_is_initial_device23439 3200 Node: omp_set_default_device24132 3201 Node: omp_set_dynamic24923 3202 Node: omp_set_max_active_levels25809 3203 Node: omp_set_nested26586 3204 Node: omp_set_num_threads27478 3205 Node: omp_set_schedule28346 3206 Node: omp_init_lock29397 3207 Node: omp_set_lock30050 3208 Node: omp_test_lock30905 3209 Node: omp_unset_lock31881 3210 Node: omp_destroy_lock32812 3211 Node: omp_init_nest_lock33489 3212 Node: omp_set_nest_lock34224 3213 Node: omp_test_nest_lock35139 3214 Node: omp_unset_nest_lock36166 3215 Node: omp_destroy_nest_lock37181 3216 Node: omp_get_wtick37932 3217 Node: omp_get_wtime38524 3218 Node: Environment Variables39300 3219 Node: OMP_CANCELLATION40692 3220 Node: OMP_DISPLAY_ENV41225 3221 Node: OMP_DEFAULT_DEVICE41928 3222 Node: OMP_DYNAMIC42708 3223 Node: OMP_MAX_ACTIVE_LEVELS43304 3224 Node: OMP_NESTED43943 3225 Node: OMP_NUM_THREADS44548 3226 Node: OMP_PROC_BIND45236 3227 Node: OMP_PLACES46428 3228 Node: OMP_STACKSIZE48603 3229 Node: OMP_SCHEDULE49427 3230 Node: OMP_THREAD_LIMIT50123 3231 Node: OMP_WAIT_POLICY50723 3232 Node: GOMP_CPU_AFFINITY51415 3233 Node: GOMP_DEBUG53145 3234 Node: GOMP_STACKSIZE53652 3235 Node: GOMP_SPINCOUNT54481 3236 Node: The libgomp ABI55653 3237 Node: Implementing MASTER construct56452 3238 Node: Implementing CRITICAL construct56866 3239 Node: Implementing ATOMIC construct57605 3240 Node: Implementing FLUSH construct58086 3241 Node: Implementing BARRIER construct58357 3242 Node: Implementing THREADPRIVATE construct58626 3243 Node: Implementing PRIVATE clause59279 3244 Node: Implementing FIRSTPRIVATE LASTPRIVATE COPYIN and COPYPRIVATE clauses59860 3245 Node: Implementing REDUCTION clause61184 3246 Node: Implementing PARALLEL construct61741 3247 Node: Implementing FOR construct62998 3248 Node: Implementing ORDERED construct64996 3249 Node: Implementing SECTIONS construct65302 3250 Node: Implementing SINGLE construct66068 3251 Node: Reporting Bugs66730 3252 Node: Copying67090 3253 Node: GNU Free Documentation License104637 3254 Node: Funding129760 3255 Node: Library Index132286 3256 3257 End Tag Table 3258
Indexes created Fri Mar 20 00:23:26 UTC 2026