LLVM OpenMP* Runtime Library

Generated by Doxygen 1.8.3.1

Wed Sep 23 2015 13:02:39

2

Trademarks The OpenMP name and the OpenMP logo are registered trademarks of the OpenMP Architecture
Review Board.
Intel, Xeon, and Intel Xeon Phi are trademarks of Intel Corporation in the U.S. and/or other countries.
This document is Copyright © 2015the LLVM Project. It is subject to the same license terms as the LLVM OpenMP
runtime.
*
Other names and brands may be claimed as the property of others.

Generated on Wed Sep 23 2015 13:02:36 for LLVM OpenMP∗ Runtime Library by Doxygen
Contents

1 LLVM∗ OpenMP∗ Runtime Library Interface 1

1.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1
1.2 Quickly Building the Runtime . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1
1.3 Supported RTL Build Configurations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1
1.4 Front-end Compilers that work with this RTL . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2
1.5 Outlining . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2
1.5.1 Addressing shared variables . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3
1.5.1.1 Current Technique . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3
1.5.1.2 Future Technique . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3
1.6 Library Interfaces . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3
1.7 Examples . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4
1.7.1 Work Sharing Example . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4

2 Module Index 7
2.1 Modules . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7

3 Hierarchical Index 9
3.1 Class Hierarchy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9

4 Class Index 11
4.1 Class List . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11

5 Module Documentation 13
5.1 Atomic Operations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13
5.1.1 Detailed Description . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13
5.2 Wait/Release operations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20
5.2.1 Detailed Description . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20
5.2.2 Enumeration Type Documentation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20
5.2.2.1 flag_type . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20
5.3 Basic Types . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21
5.3.1 Detailed Description . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21
5.3.2 Macro Definition Documentation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21
5.3.2.1 KMP_IDENT_ATOMIC_REDUCE . . . . . . . . . . . . . . . . . . . . . . . . . 21
ii CONTENTS

5.3.2.2 KMP_IDENT_AUTOPAR . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21
5.3.2.3 KMP_IDENT_BARRIER_EXPL . . . . . . . . . . . . . . . . . . . . . . . . . . 21
5.3.2.4 KMP_IDENT_BARRIER_IMPL . . . . . . . . . . . . . . . . . . . . . . . . . . 21
5.3.2.5 KMP_IDENT_IMB . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21
5.3.2.6 KMP_IDENT_KMPC . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21
5.3.3 Typedef Documentation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 22
5.3.3.1 ident_t . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 22
5.4 Deprecated Functions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 23
5.4.1 Detailed Description . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 23
5.4.2 Function Documentation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 23
5.4.2.1 __kmpc_ok_to_fork . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 23
5.5 Startup and Shutdown . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 24
5.5.1 Detailed Description . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 24
5.5.2 Function Documentation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 24
5.5.2.1 __kmpc_begin . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 24
5.5.2.2 __kmpc_end . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 24
5.6 Parallel (fork/join) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 25
5.6.1 Detailed Description . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 25
5.6.2 Typedef Documentation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 25
5.6.2.1 kmpc_micro . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 25
5.6.3 Function Documentation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 25
5.6.3.1 __kmpc_end_serialized_parallel . . . . . . . . . . . . . . . . . . . . . . . . . . 25
5.6.3.2 __kmpc_fork_call . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 25
5.6.3.3 __kmpc_fork_teams . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 26
5.6.3.4 __kmpc_push_num_teams . . . . . . . . . . . . . . . . . . . . . . . . . . . . 26
5.6.3.5 __kmpc_push_num_threads . . . . . . . . . . . . . . . . . . . . . . . . . . . . 26
5.6.3.6 __kmpc_serialized_parallel . . . . . . . . . . . . . . . . . . . . . . . . . . . . 26
5.7 Thread Information . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 27
5.7.1 Detailed Description . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 27
5.7.2 Function Documentation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 27
5.7.2.1 __kmpc_bound_num_threads . . . . . . . . . . . . . . . . . . . . . . . . . . . 27
5.7.2.2 __kmpc_bound_thread_num . . . . . . . . . . . . . . . . . . . . . . . . . . . 27
5.7.2.3 __kmpc_global_num_threads . . . . . . . . . . . . . . . . . . . . . . . . . . . 27
5.7.2.4 __kmpc_global_thread_num . . . . . . . . . . . . . . . . . . . . . . . . . . . . 28
5.7.2.5 __kmpc_in_parallel . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 28
5.8 Work Sharing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 29
5.8.1 Detailed Description . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 30
5.8.2 Enumeration Type Documentation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 30
5.8.2.1 sched_type . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 30
5.8.3 Function Documentation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 31

Generated on Wed Sep 23 2015 13:02:36 for LLVM OpenMP∗ Runtime Library by Doxygen
CONTENTS iii

5.8.3.1 __kmpc_critical . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 31
5.8.3.2 __kmpc_dispatch_fini_4 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 31
5.8.3.3 __kmpc_dispatch_fini_4u . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 31
5.8.3.4 __kmpc_dispatch_fini_8 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 31
5.8.3.5 __kmpc_dispatch_fini_8u . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 32
5.8.3.6 __kmpc_dispatch_init_4 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 32
5.8.3.7 __kmpc_dispatch_init_4u . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 32
5.8.3.8 __kmpc_dispatch_init_8 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 32
5.8.3.9 __kmpc_dispatch_init_8u . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 32
5.8.3.10 __kmpc_dispatch_next_4 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 32
5.8.3.11 __kmpc_dispatch_next_4u . . . . . . . . . . . . . . . . . . . . . . . . . . . . 33
5.8.3.12 __kmpc_dispatch_next_8 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 33
5.8.3.13 __kmpc_dispatch_next_8u . . . . . . . . . . . . . . . . . . . . . . . . . . . . 33
5.8.3.14 __kmpc_dist_dispatch_init_4 . . . . . . . . . . . . . . . . . . . . . . . . . . . 33
5.8.3.15 __kmpc_dist_for_static_init_4 . . . . . . . . . . . . . . . . . . . . . . . . . . . 33
5.8.3.16 __kmpc_dist_for_static_init_4u . . . . . . . . . . . . . . . . . . . . . . . . . . 34
5.8.3.17 __kmpc_dist_for_static_init_8 . . . . . . . . . . . . . . . . . . . . . . . . . . . 34
5.8.3.18 __kmpc_dist_for_static_init_8u . . . . . . . . . . . . . . . . . . . . . . . . . . 34
5.8.3.19 __kmpc_end_critical . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 34
5.8.3.20 __kmpc_end_master . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 34
5.8.3.21 __kmpc_end_ordered . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 34
5.8.3.22 __kmpc_end_single . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 35
5.8.3.23 __kmpc_for_static_fini . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 35
5.8.3.24 __kmpc_for_static_init_4 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 35
5.8.3.25 __kmpc_for_static_init_4u . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 35
5.8.3.26 __kmpc_for_static_init_8 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 36
5.8.3.27 __kmpc_for_static_init_8u . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 36
5.8.3.28 __kmpc_master . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 36
5.8.3.29 __kmpc_ordered . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 36
5.8.3.30 __kmpc_single . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 36
5.8.3.31 __kmpc_team_static_init_4 . . . . . . . . . . . . . . . . . . . . . . . . . . . . 37
5.8.3.32 __kmpc_team_static_init_4u . . . . . . . . . . . . . . . . . . . . . . . . . . . . 37
5.8.3.33 __kmpc_team_static_init_8 . . . . . . . . . . . . . . . . . . . . . . . . . . . . 37
5.8.3.34 __kmpc_team_static_init_8u . . . . . . . . . . . . . . . . . . . . . . . . . . . . 37
5.9 Synchronization . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 38
5.9.1 Detailed Description . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 38
5.9.2 Function Documentation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 38
5.9.2.1 __kmpc_barrier . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 38
5.9.2.2 __kmpc_barrier_master . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 38
5.9.2.3 __kmpc_barrier_master_nowait . . . . . . . . . . . . . . . . . . . . . . . . . . 38

Generated on Wed Sep 23 2015 13:02:36 for LLVM OpenMP∗ Runtime Library by Doxygen
iv CONTENTS

5.9.2.4 __kmpc_end_barrier_master . . . . . . . . . . . . . . . . . . . . . . . . . . . 39

5.9.2.5 __kmpc_end_reduce . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 39

5.9.2.6 __kmpc_end_reduce_nowait . . . . . . . . . . . . . . . . . . . . . . . . . . . 39

5.9.2.7 __kmpc_flush . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 39

5.9.2.8 __kmpc_reduce . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 39

5.9.2.9 __kmpc_reduce_nowait . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 40

5.10 Thread private data support . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 41

5.10.1 Detailed Description . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 41

5.10.2 Typedef Documentation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 41

5.10.2.1 kmpc_cctor . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 41

5.10.2.2 kmpc_cctor_vec . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 41

5.10.2.3 kmpc_ctor . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 41

5.10.2.4 kmpc_ctor_vec . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 41

5.10.2.5 kmpc_dtor . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 41

5.10.2.6 kmpc_dtor_vec . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 42

5.10.3 Function Documentation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 42

5.10.3.1 __kmpc_copyprivate . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 42

5.10.3.2 __kmpc_threadprivate_cached . . . . . . . . . . . . . . . . . . . . . . . . . . 42

5.10.3.3 __kmpc_threadprivate_register . . . . . . . . . . . . . . . . . . . . . . . . . . 43

5.10.3.4 __kmpc_threadprivate_register_vec . . . . . . . . . . . . . . . . . . . . . . . . 43

5.11 Statistics Gathering from OMPTB . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 44

5.11.1 Detailed Description . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 44

5.11.2 Environment Variables . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 44

5.11.3 Macro Definition Documentation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 45

5.11.3.1 KMP_COUNT_BLOCK . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 45

5.11.3.2 KMP_COUNT_VALUE . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 45

5.11.3.3 KMP_FOREACH_COUNTER . . . . . . . . . . . . . . . . . . . . . . . . . . . 45

5.11.3.4 KMP_FOREACH_EXPLICIT_TIMER . . . . . . . . . . . . . . . . . . . . . . . 46

5.11.3.5 KMP_OUTPUT_STATS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 46

5.11.3.6 KMP_RESET_STATS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 46

5.11.3.7 KMP_START_EXPLICIT_TIMER . . . . . . . . . . . . . . . . . . . . . . . . . 47

5.11.3.8 KMP_STOP_EXPLICIT_TIMER . . . . . . . . . . . . . . . . . . . . . . . . . . 47

5.11.3.9 KMP_TIME_BLOCK . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 47

5.12 Tasking support . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 48

5.12.1 Detailed Description . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 48

5.12.2 Function Documentation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 48

5.12.2.1 __kmpc_omp_task_with_deps . . . . . . . . . . . . . . . . . . . . . . . . . . . 48

5.12.2.2 __kmpc_omp_wait_deps . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 48

5.13 User visible functions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 49

Generated on Wed Sep 23 2015 13:02:36 for LLVM OpenMP∗ Runtime Library by Doxygen
CONTENTS v

6 Class Documentation 51
6.1 hierarchy_info Class Reference . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 51
6.1.1 Detailed Description . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 51
6.1.2 Member Data Documentation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 51
6.1.2.1 depth . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 51
6.1.2.2 maxLeaves . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 51
6.1.2.3 maxLevels . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 52
6.1.2.4 numPerLevel . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 52
6.2 ident Struct Reference . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 52
6.2.1 Detailed Description . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 52
6.2.2 Member Data Documentation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 52
6.2.2.1 flags . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 52
6.2.2.2 psource . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 52
6.2.2.3 reserved_1 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 53
6.2.2.4 reserved_2 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 53
6.2.2.5 reserved_3 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 53
6.3 kmp_flag< P > Class Template Reference . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 53
6.3.1 Detailed Description . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 53
6.3.2 Member Function Documentation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 53
6.3.2.1 get . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 53
6.3.2.2 get_type . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 54
6.3.3 Member Data Documentation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 54
6.3.3.1 loc . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 54
6.3.3.2 t . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 54
6.4 stats_flags_e Class Reference . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 54
6.4.1 Detailed Description . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 55

Index 55

Generated on Wed Sep 23 2015 13:02:36 for LLVM OpenMP∗ Runtime Library by Doxygen
Chapter 1

LLVM∗ OpenMP∗ Runtime Library Interface

1.1 Introduction

This document describes the interface provided by the LLVM OpenMP∗ runtime library to the compiler. Routines
that are directly called as simple functions by user code are not currently described here, since their definition is in
the OpenMP specification available from http://openmp.org
The aim here is to explain the interface from the compiler to the runtime.
The overall design is described, and each function in the interface has its own description. (At least, that’s the
ambition, we may not be there yet).

1.2 Quickly Building the Runtime

For the impatient, we cover building the runtime as the first topic here.
CMake is used to build the OpenMP runtime. For details and a full list of options for the CMake build system, see
Build_With_CMake.txt inside the runtime/ subdirectory. These instructions will provide the most typical
build.
In-LLVM-tree build:.

$ cd where-you-want-to-live
Check out openmp into llvm/projects
$ cd where-you-want-to-build
$ mkdir build && cd build
$ cmake path/to/llvm -DCMAKE_C_COMPILER=<C compiler> -DCMAKE_CXX_COMPILER=<C++ compiler>
$ make omp

Out-of-LLVM-tree build:

$ cd where-you-want-to-live
Check out openmp
$ cd where-you-want-to-live/openmp/runtime
$ mkdir build && cd build
$ cmake path/to/openmp -DCMAKE_C_COMPILER=<C compiler> -DCMAKE_CXX_COMPILER=<C++ compiler>
$ make

1.3 Supported RTL Build Configurations

The architectures supported are IA-32 architecture, Intel® 64, and Intel® Many Integrated Core Architecture. The
build configurations supported are shown in the table below.
2 LLVM∗ OpenMP∗ Runtime Library Interface

icc/icl gcc clang

Linux∗ OS Yes(1,5) Yes(2,4) Yes(4,6,7)
FreeBSD∗ Yes(1,5) Yes(2,4) Yes(4,6,7,8)
OS X∗ Yes(1,3,4) No Yes(4,6,7)
Windows∗ OS Yes(1,4) No No
(1) On IA-32 architecture and Intel® 64, icc/icl versions 12.x are supported (12.1 is recommended).
(2) gcc version 4.7 is supported.
(3) For icc on OS X∗ , OS X∗ version 10.5.8 is supported.
(4) Intel® Many Integrated Core Architecture not supported.
(5) On Intel® Many Integrated Core Architecture, icc/icl versions 13.0 or later are required.
(6) Clang∗ version 3.3 is supported.
(7) Clang∗ currently does not offer a software-implemented 128 bit extended precision type. Thus, all entry points
reliant on this type are removed from the library and cannot be called in the user program. The following functions
are not available:

__kmpc_atomic_cmplx16_*
__kmpc_atomic_float16_*
__kmpc_atomic_*_fp

(8) Community contribution provided AS IS, not tested by Intel.

Supported Architectures: IBM(R) Power 7 and Power 8
gcc clang
Linux∗ OS Yes(1,2) Yes(3,4)
(1) On Power 7, gcc version 4.8.2 is supported.
(2) On Power 8, gcc version 4.8.2 is supported.
(3) On Power 7, clang version 3.7 is supported.
(4) On Power 8, clang version 3.7 is supported.

1.4 Front-end Compilers that work with this RTL

The following compilers are known to do compatible code generation for this RTL: icc/icl, gcc. Code generation is
discussed in more detail later in this document.

1.5 Outlining

The runtime interface is based on the idea that the compiler "outlines" sections of code that are to run in parallel
into separate functions that can then be invoked in multiple threads. For instance, simple code like this

void foo()
{
#pragma omp parallel
{
... do something ...
}
}

is converted into something that looks conceptually like this (where the names used are merely illustrative; the real
library function names will be used later after we’ve discussed some more issues...)

static void outlinedFooBody()

{
... do something ...

Generated on Wed Sep 23 2015 13:02:36 for LLVM OpenMP∗ Runtime Library by Doxygen
1.6 Library Interfaces 3

void foo()
{
__OMP_runtime_fork(outlinedFooBody, (void*)0); // Not the real function name!
}

1.5.1 Addressing shared variables

In real uses of the OpenMP∗ API there are normally references from the outlined code to shared variables that are
in scope in the containing function. Therefore the containing function must be able to address these variables. The
runtime supports two alternate ways of doing this.

1.5.1.1 Current Technique

The technique currently supported by the runtime library is to receive a separate pointer to each shared variable
that can be accessed from the outlined function. This is what is shown in the example below.
We hope soon to provide an alternative interface to support the alternate implementation described in the next
section. The alternative implementation has performance advantages for small parallel regions that have many
shared variables.

1.5.1.2 Future Technique

The idea is to treat the outlined function as though it were a lexically nested function, and pass it a single argument
which is the pointer to the parent’s stack frame. Provided that the compiler knows the layout of the parent frame
when it is generating the outlined function it can then access the up-level variables at appropriate offsets from
the parent frame. This is a classical compiler technique from the 1960s to support languages like Algol (and its
descendants) that support lexically nested functions.
The main benefit of this technique is that there is no code required at the fork point to marshal the arguments to the
outlined function. Since the runtime knows statically how many arguments must be passed to the outlined function,
it can easily copy them to the thread’s stack frame. Therefore the performance of the fork code is independent of
the number of shared variables that are accessed by the outlined function.
If it is hard to determine the stack layout of the parent while generating the outlined code, it is still possible to use
this approach by collecting all of the variables in the parent that are accessed from outlined functions into a single
struct which is placed on the stack, and whose address is passed to the outlined functions. In this way the offsets
of the shared variables are known (since they are inside the struct) without needing to know the complete layout of
the parent stack-frame. From the point of view of the runtime either of these techniques is equivalent, since in either
case it only has to pass a single argument to the outlined function to allow it to access shared variables.
A scheme like this is how gcc∗ generates outlined functions.

1.6 Library Interfaces

The library functions used for specific parts of the OpenMP∗ language implementation are documented in different
modules.

• Basic Types fundamental types used by the runtime in many places

• Deprecated Functions functions that are in the library but are no longer required

• Startup and Shutdown functions for initializing and finalizing the runtime

• Parallel (fork/join) functions for implementing omp parallel

• Thread Information functions for supporting thread state inquiries

• Work Sharing functions for work sharing constructs such as omp for, omp sections

Generated on Wed Sep 23 2015 13:02:36 for LLVM OpenMP∗ Runtime Library by Doxygen
4 LLVM∗ OpenMP∗ Runtime Library Interface

• Thread private data support functions to support thread private data, copyin etc

• Synchronization functions to support omp critical, omp barrier, omp master, reductions etc

• Atomic Operations functions to support atomic operations

• Statistics Gathering from OMPTB macros to support developer profiling of libomp

• Documentation on tasking has still to be written...

1.7 Examples

1.7.1 Work Sharing Example

This example shows the code generated for a parallel for with reduction and dynamic scheduling.

extern float foo( void );

int main () {
int i;
float r = 0.0;
#pragma omp parallel for schedule(dynamic) reduction(+:r)
for ( i = 0; i < 10; i ++ ) {
r += foo();
}
}

The transformed code looks like this.

extern float foo( void );

int main () {
static int zero = 0;
auto int gtid;
auto float r = 0.0;
__kmpc_begin( & loc3, 0 );
// The gtid is not actually required in this example so could be omitted;
// We show its initialization here because it is often required for calls into
// the runtime and should be locally cached like this.
gtid = __kmpc_global thread num( & loc3 );
__kmpc_fork call( & loc7, 1, main_7_parallel_3, & r );
__kmpc_end( & loc0 );
return 0;
}

struct main_10_reduction_t_5 { float r_10_rpr; };

static kmp_critical_name lck = { 0 };

static ident_t loc10; // loc10.flags should contain KMP_IDENT_ATOMIC_REDUCE bit set
// if compiler has generated an atomic reduction.

void main_7_parallel_3( int *gtid, int *btid, float *r_7_shp ) {

auto int i_7_pr;
auto int lower, upper, liter, incr;
auto struct main_10_reduction_t_5 reduce;
reduce.r_10_rpr = 0.F;
liter = 0;
__kmpc_dispatch_init_4( & loc7,*gtid, 35, 0, 9, 1, 1 );
while ( __kmpc_dispatch_next_4( & loc7, *gtid, & liter, & lower, & upper, & incr
) ) {
for( i_7_pr = lower; upper >= i_7_pr; i_7_pr ++ )
reduce.r_10_rpr += foo();
}
switch( __kmpc_reduce_nowait( & loc10, *gtid, 1, 4, & reduce, main_10_reduce_5, &
lck ) ) {
case 1:
*r_7_shp += reduce.r_10_rpr;
__kmpc_end_reduce_nowait( & loc10, *gtid, & lck );
break;
case 2:
__kmpc_atomic_float4_add( & loc10, *gtid, r_7_shp, reduce.r_10_rpr );
break;
default:;
}
}

Generated on Wed Sep 23 2015 13:02:36 for LLVM OpenMP∗ Runtime Library by Doxygen
1.7 Examples 5

void main_10_reduce_5( struct main_10_reduction_t_5 *reduce_lhs,

struct main_10_reduction_t_5 *reduce_rhs )
{
reduce_lhs->r_10_rpr += reduce_rhs->r_10_rpr;
}

Generated on Wed Sep 23 2015 13:02:36 for LLVM OpenMP∗ Runtime Library by Doxygen
6 LLVM∗ OpenMP∗ Runtime Library Interface

Generated on Wed Sep 23 2015 13:02:36 for LLVM OpenMP∗ Runtime Library by Doxygen
Chapter 2

Module Index

2.1 Modules

Here is a list of all modules:

Atomic Operations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13
Wait/Release operations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20
Basic Types . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21
Deprecated Functions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 23
Startup and Shutdown . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 24
Parallel (fork/join) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 25
Thread Information . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 27
Work Sharing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 29
Synchronization . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 38
Thread private data support . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 41
Statistics Gathering from OMPTB . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 44
Tasking support . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 48
User visible functions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 49
8 Module Index

Generated on Wed Sep 23 2015 13:02:36 for LLVM OpenMP∗ Runtime Library by Doxygen
Chapter 3

Hierarchical Index

3.1 Class Hierarchy

This inheritance list is sorted roughly, but not completely, alphabetically:

hierarchy_info . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 51
ident . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 52
kmp_flag< P > . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 53
kmp_flag< FlagType > . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 53
kmp_flag< kmp_uint32 > . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 53
kmp_flag< kmp_uint64 > . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 53
stats_flags_e . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 54
10 Hierarchical Index

Generated on Wed Sep 23 2015 13:02:36 for LLVM OpenMP∗ Runtime Library by Doxygen
Chapter 4

Class Index

4.1 Class List

Here are the classes, structs, unions and interfaces with brief descriptions:
hierarchy_info . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 51
ident . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 52
kmp_flag< P > . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 53
stats_flags_e
Flags to describe the statistic ( timers or counter ) . . . . . . . . . . . . . . . . . . . . . . . . 54
12 Class Index

Generated on Wed Sep 23 2015 13:02:36 for LLVM OpenMP∗ Runtime Library by Doxygen
Chapter 5

Module Documentation

5.1 Atomic Operations

5.1.1 Detailed Description

These functions are used for implementing the many different varieties of atomic operations.
The compiler is at liberty to inline atomic operations that are naturally supported by the target architecture. For
instance on IA-32 architecture an atomic like this can be inlined

static int s = 0;
#pragma omp atomic
s++;

using the single instruction: lock; incl s

However the runtime does provide entrypoints for these operations to support compilers that choose not to inline
them. (For instance, __kmpc_atomic_fixed4_add could be used to perform the increment above.)
The names of the functions are encoded by using the data type name and the operation name, as in these tables.
Data Type Data type encoding
int8_t fixed1
uint8_t fixed1u
int16_t fixed2
uint16_t fixed2u
int32_t fixed4
uint32_t fixed4u
int32_t fixed8
uint32_t fixed8u
float float4
double float8
float 10 (8087 eighty bit float) float10
complex<float> cmplx4
complex<double> cmplx8
complex<float10> cmplx10
Operation Operation encoding
+ add
- sub
∗ mul
/ div
14 Module Documentation

& andb
<< shl
>> shr
| orb
∧ xor
&& andl
|| orl
maximum max
minimum min
.eqv. eqv
.neqv. neqv
For non-commutative operations, _rev can also be added for the reversed operation. For the functions that capture
the result, the suffix _cpt is added.

Update Functions

The general form of an atomic function that just performs an update (without a capture)

void __kmpc_atomic_<datatype>_<operation>( ident_t *id_ref, int gtid, TYPE * lhs, TYPE rhs );

Parameters
ident_t a pointer to source location
gtid the global thread id
lhs a pointer to the left operand
rhs the right operand

capture functions

The capture functions perform an atomic update and return a result, which is either the value before the capture, or
that after. They take an additional argument to determine which result is returned. Their general form is therefore

TYPE __kmpc_atomic_<datatype>_<operation>_cpt( ident_t *id_ref, int gtid, TYPE * lhs, TYPE rhs, int
flag );

Parameters
ident_t a pointer to source location
gtid the global thread id
lhs a pointer to the left operand
rhs the right operand
flag one if the result is to be captured after the operation, zero if captured before.

The one set of exceptions to this is the complex<float> type where the value is not returned, rather an extra
argument pointer is passed.
They look like

void __kmpc_atomic_cmplx4_<op>_cpt( ident_t *id_ref, int gtid, kmp_cmplx32 * lhs, kmp_cmplx32 rhs,
kmp_cmplx32 * out, int flag );

Read and Write Operations

The OpenMP∗ standard now supports atomic operations that simply ensure that the value is read or written atom-
ically, with no modification performed. In many cases on IA-32 architecture these operations can be inlined since

Generated on Wed Sep 23 2015 13:02:36 for LLVM OpenMP∗ Runtime Library by Doxygen
5.1 Atomic Operations 15

the architecture guarantees that no tearing occurs on aligned objects accessed with a single memory operation of
up to 64 bits in size.
The general form of the read operations is

TYPE __kmpc_atomic_<type>_rd ( ident_t *id_ref, int gtid, TYPE * loc );

For the write operations the form is

void __kmpc_atomic_<type>_wr ( ident_t *id_ref, int gtid, TYPE * lhs, TYPE rhs );

Full list of functions

This leads to the generation of 376 atomic functions, as follows.

Functons for integers

There are versions here for integers of size 1,2,4 and 8 bytes both signed and unsigned (where that matters).

__kmpc_atomic_fixed1_add
__kmpc_atomic_fixed1_add_cpt
__kmpc_atomic_fixed1_add_fp
__kmpc_atomic_fixed1_andb
__kmpc_atomic_fixed1_andb_cpt
__kmpc_atomic_fixed1_andl
__kmpc_atomic_fixed1_andl_cpt
__kmpc_atomic_fixed1_div
__kmpc_atomic_fixed1_div_cpt
__kmpc_atomic_fixed1_div_cpt_rev
__kmpc_atomic_fixed1_div_float8
__kmpc_atomic_fixed1_div_fp
__kmpc_atomic_fixed1_div_rev
__kmpc_atomic_fixed1_eqv
__kmpc_atomic_fixed1_eqv_cpt
__kmpc_atomic_fixed1_max
__kmpc_atomic_fixed1_max_cpt
__kmpc_atomic_fixed1_min
__kmpc_atomic_fixed1_min_cpt
__kmpc_atomic_fixed1_mul
__kmpc_atomic_fixed1_mul_cpt
__kmpc_atomic_fixed1_mul_float8
__kmpc_atomic_fixed1_mul_fp
__kmpc_atomic_fixed1_neqv
__kmpc_atomic_fixed1_neqv_cpt
__kmpc_atomic_fixed1_orb
__kmpc_atomic_fixed1_orb_cpt
__kmpc_atomic_fixed1_orl
__kmpc_atomic_fixed1_orl_cpt
__kmpc_atomic_fixed1_rd
__kmpc_atomic_fixed1_shl
__kmpc_atomic_fixed1_shl_cpt
__kmpc_atomic_fixed1_shl_cpt_rev
__kmpc_atomic_fixed1_shl_rev
__kmpc_atomic_fixed1_shr
__kmpc_atomic_fixed1_shr_cpt
__kmpc_atomic_fixed1_shr_cpt_rev
__kmpc_atomic_fixed1_shr_rev
__kmpc_atomic_fixed1_sub
__kmpc_atomic_fixed1_sub_cpt
__kmpc_atomic_fixed1_sub_cpt_rev
__kmpc_atomic_fixed1_sub_fp
__kmpc_atomic_fixed1_sub_rev
__kmpc_atomic_fixed1_swp
__kmpc_atomic_fixed1_wr
__kmpc_atomic_fixed1_xor
__kmpc_atomic_fixed1_xor_cpt
__kmpc_atomic_fixed1u_div
__kmpc_atomic_fixed1u_div_cpt
__kmpc_atomic_fixed1u_div_cpt_rev
__kmpc_atomic_fixed1u_div_fp
__kmpc_atomic_fixed1u_div_rev
__kmpc_atomic_fixed1u_shr
__kmpc_atomic_fixed1u_shr_cpt
__kmpc_atomic_fixed1u_shr_cpt_rev
__kmpc_atomic_fixed1u_shr_rev
__kmpc_atomic_fixed2_add

Generated on Wed Sep 23 2015 13:02:36 for LLVM OpenMP∗ Runtime Library by Doxygen
16 Module Documentation

__kmpc_atomic_fixed2_add_cpt
__kmpc_atomic_fixed2_add_fp
__kmpc_atomic_fixed2_andb
__kmpc_atomic_fixed2_andb_cpt
__kmpc_atomic_fixed2_andl
__kmpc_atomic_fixed2_andl_cpt
__kmpc_atomic_fixed2_div
__kmpc_atomic_fixed2_div_cpt
__kmpc_atomic_fixed2_div_cpt_rev
__kmpc_atomic_fixed2_div_float8
__kmpc_atomic_fixed2_div_fp
__kmpc_atomic_fixed2_div_rev
__kmpc_atomic_fixed2_eqv
__kmpc_atomic_fixed2_eqv_cpt
__kmpc_atomic_fixed2_max
__kmpc_atomic_fixed2_max_cpt
__kmpc_atomic_fixed2_min
__kmpc_atomic_fixed2_min_cpt
__kmpc_atomic_fixed2_mul
__kmpc_atomic_fixed2_mul_cpt
__kmpc_atomic_fixed2_mul_float8
__kmpc_atomic_fixed2_mul_fp
__kmpc_atomic_fixed2_neqv
__kmpc_atomic_fixed2_neqv_cpt
__kmpc_atomic_fixed2_orb
__kmpc_atomic_fixed2_orb_cpt
__kmpc_atomic_fixed2_orl
__kmpc_atomic_fixed2_orl_cpt
__kmpc_atomic_fixed2_rd
__kmpc_atomic_fixed2_shl
__kmpc_atomic_fixed2_shl_cpt
__kmpc_atomic_fixed2_shl_cpt_rev
__kmpc_atomic_fixed2_shl_rev
__kmpc_atomic_fixed2_shr
__kmpc_atomic_fixed2_shr_cpt
__kmpc_atomic_fixed2_shr_cpt_rev
__kmpc_atomic_fixed2_shr_rev
__kmpc_atomic_fixed2_sub
__kmpc_atomic_fixed2_sub_cpt
__kmpc_atomic_fixed2_sub_cpt_rev
__kmpc_atomic_fixed2_sub_fp
__kmpc_atomic_fixed2_sub_rev
__kmpc_atomic_fixed2_swp
__kmpc_atomic_fixed2_wr
__kmpc_atomic_fixed2_xor
__kmpc_atomic_fixed2_xor_cpt
__kmpc_atomic_fixed2u_div
__kmpc_atomic_fixed2u_div_cpt
__kmpc_atomic_fixed2u_div_cpt_rev
__kmpc_atomic_fixed2u_div_fp
__kmpc_atomic_fixed2u_div_rev
__kmpc_atomic_fixed2u_shr
__kmpc_atomic_fixed2u_shr_cpt
__kmpc_atomic_fixed2u_shr_cpt_rev
__kmpc_atomic_fixed2u_shr_rev
__kmpc_atomic_fixed4_add
__kmpc_atomic_fixed4_add_cpt
__kmpc_atomic_fixed4_add_fp
__kmpc_atomic_fixed4_andb
__kmpc_atomic_fixed4_andb_cpt
__kmpc_atomic_fixed4_andl
__kmpc_atomic_fixed4_andl_cpt
__kmpc_atomic_fixed4_div
__kmpc_atomic_fixed4_div_cpt
__kmpc_atomic_fixed4_div_cpt_rev
__kmpc_atomic_fixed4_div_float8
__kmpc_atomic_fixed4_div_fp
__kmpc_atomic_fixed4_div_rev
__kmpc_atomic_fixed4_eqv
__kmpc_atomic_fixed4_eqv_cpt
__kmpc_atomic_fixed4_max
__kmpc_atomic_fixed4_max_cpt
__kmpc_atomic_fixed4_min
__kmpc_atomic_fixed4_min_cpt
__kmpc_atomic_fixed4_mul
__kmpc_atomic_fixed4_mul_cpt
__kmpc_atomic_fixed4_mul_float8
__kmpc_atomic_fixed4_mul_fp
__kmpc_atomic_fixed4_neqv
__kmpc_atomic_fixed4_neqv_cpt
__kmpc_atomic_fixed4_orb
__kmpc_atomic_fixed4_orb_cpt
__kmpc_atomic_fixed4_orl
__kmpc_atomic_fixed4_orl_cpt
__kmpc_atomic_fixed4_rd
__kmpc_atomic_fixed4_shl
__kmpc_atomic_fixed4_shl_cpt

Generated on Wed Sep 23 2015 13:02:36 for LLVM OpenMP∗ Runtime Library by Doxygen
5.1 Atomic Operations 17

__kmpc_atomic_fixed4_shl_cpt_rev
__kmpc_atomic_fixed4_shl_rev
__kmpc_atomic_fixed4_shr
__kmpc_atomic_fixed4_shr_cpt
__kmpc_atomic_fixed4_shr_cpt_rev
__kmpc_atomic_fixed4_shr_rev
__kmpc_atomic_fixed4_sub
__kmpc_atomic_fixed4_sub_cpt
__kmpc_atomic_fixed4_sub_cpt_rev
__kmpc_atomic_fixed4_sub_fp
__kmpc_atomic_fixed4_sub_rev
__kmpc_atomic_fixed4_swp
__kmpc_atomic_fixed4_wr
__kmpc_atomic_fixed4_xor
__kmpc_atomic_fixed4_xor_cpt
__kmpc_atomic_fixed4u_div
__kmpc_atomic_fixed4u_div_cpt
__kmpc_atomic_fixed4u_div_cpt_rev
__kmpc_atomic_fixed4u_div_fp
__kmpc_atomic_fixed4u_div_rev
__kmpc_atomic_fixed4u_shr
__kmpc_atomic_fixed4u_shr_cpt
__kmpc_atomic_fixed4u_shr_cpt_rev
__kmpc_atomic_fixed4u_shr_rev
__kmpc_atomic_fixed8_add
__kmpc_atomic_fixed8_add_cpt
__kmpc_atomic_fixed8_add_fp
__kmpc_atomic_fixed8_andb
__kmpc_atomic_fixed8_andb_cpt
__kmpc_atomic_fixed8_andl
__kmpc_atomic_fixed8_andl_cpt
__kmpc_atomic_fixed8_div
__kmpc_atomic_fixed8_div_cpt
__kmpc_atomic_fixed8_div_cpt_rev
__kmpc_atomic_fixed8_div_float8
__kmpc_atomic_fixed8_div_fp
__kmpc_atomic_fixed8_div_rev
__kmpc_atomic_fixed8_eqv
__kmpc_atomic_fixed8_eqv_cpt
__kmpc_atomic_fixed8_max
__kmpc_atomic_fixed8_max_cpt
__kmpc_atomic_fixed8_min
__kmpc_atomic_fixed8_min_cpt
__kmpc_atomic_fixed8_mul
__kmpc_atomic_fixed8_mul_cpt
__kmpc_atomic_fixed8_mul_float8
__kmpc_atomic_fixed8_mul_fp
__kmpc_atomic_fixed8_neqv
__kmpc_atomic_fixed8_neqv_cpt
__kmpc_atomic_fixed8_orb
__kmpc_atomic_fixed8_orb_cpt
__kmpc_atomic_fixed8_orl
__kmpc_atomic_fixed8_orl_cpt
__kmpc_atomic_fixed8_rd
__kmpc_atomic_fixed8_shl
__kmpc_atomic_fixed8_shl_cpt
__kmpc_atomic_fixed8_shl_cpt_rev
__kmpc_atomic_fixed8_shl_rev
__kmpc_atomic_fixed8_shr
__kmpc_atomic_fixed8_shr_cpt
__kmpc_atomic_fixed8_shr_cpt_rev
__kmpc_atomic_fixed8_shr_rev
__kmpc_atomic_fixed8_sub
__kmpc_atomic_fixed8_sub_cpt
__kmpc_atomic_fixed8_sub_cpt_rev
__kmpc_atomic_fixed8_sub_fp
__kmpc_atomic_fixed8_sub_rev
__kmpc_atomic_fixed8_swp
__kmpc_atomic_fixed8_wr
__kmpc_atomic_fixed8_xor
__kmpc_atomic_fixed8_xor_cpt
__kmpc_atomic_fixed8u_div
__kmpc_atomic_fixed8u_div_cpt
__kmpc_atomic_fixed8u_div_cpt_rev
__kmpc_atomic_fixed8u_div_fp
__kmpc_atomic_fixed8u_div_rev
__kmpc_atomic_fixed8u_shr
__kmpc_atomic_fixed8u_shr_cpt
__kmpc_atomic_fixed8u_shr_cpt_rev
__kmpc_atomic_fixed8u_shr_rev

Generated on Wed Sep 23 2015 13:02:36 for LLVM OpenMP∗ Runtime Library by Doxygen
18 Module Documentation

Functions for floating point

There are versions here for floating point numbers of size 4, 8, 10 and 16 bytes. (Ten byte floats are used by X87,
but are now rare).

__kmpc_atomic_float4_add
__kmpc_atomic_float4_add_cpt
__kmpc_atomic_float4_add_float8
__kmpc_atomic_float4_add_fp
__kmpc_atomic_float4_div
__kmpc_atomic_float4_div_cpt
__kmpc_atomic_float4_div_cpt_rev
__kmpc_atomic_float4_div_float8
__kmpc_atomic_float4_div_fp
__kmpc_atomic_float4_div_rev
__kmpc_atomic_float4_max
__kmpc_atomic_float4_max_cpt
__kmpc_atomic_float4_min
__kmpc_atomic_float4_min_cpt
__kmpc_atomic_float4_mul
__kmpc_atomic_float4_mul_cpt
__kmpc_atomic_float4_mul_float8
__kmpc_atomic_float4_mul_fp
__kmpc_atomic_float4_rd
__kmpc_atomic_float4_sub
__kmpc_atomic_float4_sub_cpt
__kmpc_atomic_float4_sub_cpt_rev
__kmpc_atomic_float4_sub_float8
__kmpc_atomic_float4_sub_fp
__kmpc_atomic_float4_sub_rev
__kmpc_atomic_float4_swp
__kmpc_atomic_float4_wr
__kmpc_atomic_float8_add
__kmpc_atomic_float8_add_cpt
__kmpc_atomic_float8_add_fp
__kmpc_atomic_float8_div
__kmpc_atomic_float8_div_cpt
__kmpc_atomic_float8_div_cpt_rev
__kmpc_atomic_float8_div_fp
__kmpc_atomic_float8_div_rev
__kmpc_atomic_float8_max
__kmpc_atomic_float8_max_cpt
__kmpc_atomic_float8_min
__kmpc_atomic_float8_min_cpt
__kmpc_atomic_float8_mul
__kmpc_atomic_float8_mul_cpt
__kmpc_atomic_float8_mul_fp
__kmpc_atomic_float8_rd
__kmpc_atomic_float8_sub
__kmpc_atomic_float8_sub_cpt
__kmpc_atomic_float8_sub_cpt_rev
__kmpc_atomic_float8_sub_fp
__kmpc_atomic_float8_sub_rev
__kmpc_atomic_float8_swp
__kmpc_atomic_float8_wr
__kmpc_atomic_float10_add
__kmpc_atomic_float10_add_cpt
__kmpc_atomic_float10_add_fp
__kmpc_atomic_float10_div
__kmpc_atomic_float10_div_cpt
__kmpc_atomic_float10_div_cpt_rev
__kmpc_atomic_float10_div_fp
__kmpc_atomic_float10_div_rev
__kmpc_atomic_float10_mul
__kmpc_atomic_float10_mul_cpt
__kmpc_atomic_float10_mul_fp
__kmpc_atomic_float10_rd
__kmpc_atomic_float10_sub
__kmpc_atomic_float10_sub_cpt
__kmpc_atomic_float10_sub_cpt_rev
__kmpc_atomic_float10_sub_fp
__kmpc_atomic_float10_sub_rev
__kmpc_atomic_float10_swp
__kmpc_atomic_float10_wr
__kmpc_atomic_float16_add
__kmpc_atomic_float16_add_cpt
__kmpc_atomic_float16_div
__kmpc_atomic_float16_div_cpt
__kmpc_atomic_float16_div_cpt_rev
__kmpc_atomic_float16_div_rev
__kmpc_atomic_float16_max
__kmpc_atomic_float16_max_cpt
__kmpc_atomic_float16_min
__kmpc_atomic_float16_min_cpt
__kmpc_atomic_float16_mul

Generated on Wed Sep 23 2015 13:02:36 for LLVM OpenMP∗ Runtime Library by Doxygen
5.1 Atomic Operations 19

__kmpc_atomic_float16_mul_cpt
__kmpc_atomic_float16_rd
__kmpc_atomic_float16_sub
__kmpc_atomic_float16_sub_cpt
__kmpc_atomic_float16_sub_cpt_rev
__kmpc_atomic_float16_sub_rev
__kmpc_atomic_float16_swp
__kmpc_atomic_float16_wr

Functions for Complex types

Functions for complex types whose component floating point variables are of size 4,8,10 or 16 bytes. The names
here are based on the size of the component float, not the size of the complex type. So __kmpc_atomc_-
cmplx8_add is an operation on a complex<double> or complex(kind=8), not complex<float>.

__kmpc_atomic_cmplx4_add
__kmpc_atomic_cmplx4_add_cmplx8
__kmpc_atomic_cmplx4_add_cpt
__kmpc_atomic_cmplx4_div
__kmpc_atomic_cmplx4_div_cmplx8
__kmpc_atomic_cmplx4_div_cpt
__kmpc_atomic_cmplx4_div_cpt_rev
__kmpc_atomic_cmplx4_div_rev
__kmpc_atomic_cmplx4_mul
__kmpc_atomic_cmplx4_mul_cmplx8
__kmpc_atomic_cmplx4_mul_cpt
__kmpc_atomic_cmplx4_rd
__kmpc_atomic_cmplx4_sub
__kmpc_atomic_cmplx4_sub_cmplx8
__kmpc_atomic_cmplx4_sub_cpt
__kmpc_atomic_cmplx4_sub_cpt_rev
__kmpc_atomic_cmplx4_sub_rev
__kmpc_atomic_cmplx4_swp
__kmpc_atomic_cmplx4_wr
__kmpc_atomic_cmplx8_add
__kmpc_atomic_cmplx8_add_cpt
__kmpc_atomic_cmplx8_div
__kmpc_atomic_cmplx8_div_cpt
__kmpc_atomic_cmplx8_div_cpt_rev
__kmpc_atomic_cmplx8_div_rev
__kmpc_atomic_cmplx8_mul
__kmpc_atomic_cmplx8_mul_cpt
__kmpc_atomic_cmplx8_rd
__kmpc_atomic_cmplx8_sub
__kmpc_atomic_cmplx8_sub_cpt
__kmpc_atomic_cmplx8_sub_cpt_rev
__kmpc_atomic_cmplx8_sub_rev
__kmpc_atomic_cmplx8_swp
__kmpc_atomic_cmplx8_wr
__kmpc_atomic_cmplx10_add
__kmpc_atomic_cmplx10_add_cpt
__kmpc_atomic_cmplx10_div
__kmpc_atomic_cmplx10_div_cpt
__kmpc_atomic_cmplx10_div_cpt_rev
__kmpc_atomic_cmplx10_div_rev
__kmpc_atomic_cmplx10_mul
__kmpc_atomic_cmplx10_mul_cpt
__kmpc_atomic_cmplx10_rd
__kmpc_atomic_cmplx10_sub
__kmpc_atomic_cmplx10_sub_cpt
__kmpc_atomic_cmplx10_sub_cpt_rev
__kmpc_atomic_cmplx10_sub_rev
__kmpc_atomic_cmplx10_swp
__kmpc_atomic_cmplx10_wr
__kmpc_atomic_cmplx16_add
__kmpc_atomic_cmplx16_add_cpt
__kmpc_atomic_cmplx16_div
__kmpc_atomic_cmplx16_div_cpt
__kmpc_atomic_cmplx16_div_cpt_rev
__kmpc_atomic_cmplx16_div_rev
__kmpc_atomic_cmplx16_mul
__kmpc_atomic_cmplx16_mul_cpt
__kmpc_atomic_cmplx16_rd
__kmpc_atomic_cmplx16_sub
__kmpc_atomic_cmplx16_sub_cpt
__kmpc_atomic_cmplx16_sub_cpt_rev
__kmpc_atomic_cmplx16_swp
__kmpc_atomic_cmplx16_wr

Generated on Wed Sep 23 2015 13:02:36 for LLVM OpenMP∗ Runtime Library by Doxygen
20 Module Documentation

5.2 Wait/Release operations

• enum flag_type { flag32, flag64, flag_oncore }

5.2.1 Detailed Description

The definitions and functions here implement the lowest level thread synchronizations of suspending a thread and
awaking it. They are used to build higher level operations such as barriers and fork/join.

5.2.2 Enumeration Type Documentation

5.2.2.1 enum flag_type

The flag_type describes the storage used for the flag.

Enumerator

flag32 32 bit flags

flag64 64 bit flags
flag_oncore special 64-bit flag for on-core barrier (hierarchical)

Definition at line 38 of file kmp_wait_release.h.

Generated on Wed Sep 23 2015 13:02:36 for LLVM OpenMP∗ Runtime Library by Doxygen
5.3 Basic Types 21

5.3 Basic Types

• typedef struct ident ident_t

• #define KMP_IDENT_IMB 0x01
• #define KMP_IDENT_KMPC 0x02
• #define KMP_IDENT_AUTOPAR 0x08
• #define KMP_IDENT_ATOMIC_REDUCE 0x10
• #define KMP_IDENT_BARRIER_EXPL 0x20
• #define KMP_IDENT_BARRIER_IMPL 0x0040

5.3.1 Detailed Description

Types that are used throughout the runtime.

5.3.2 Macro Definition Documentation

5.3.2.1 #define KMP IDENT ATOMIC REDUCE 0x10

Compiler generates atomic reduction option for kmpc_reduce∗

Definition at line 177 of file kmp.h.

5.3.2.2 #define KMP IDENT AUTOPAR 0x08

Entry point generated by auto-parallelization

Definition at line 175 of file kmp.h.
Referenced by __kmpc_end_serialized_parallel().

5.3.2.3 #define KMP IDENT BARRIER EXPL 0x20

To mark a ’barrier’ directive in user code

Definition at line 179 of file kmp.h.

5.3.2.4 #define KMP IDENT BARRIER IMPL 0x0040

To Mark implicit barriers.

Definition at line 181 of file kmp.h.

5.3.2.5 #define KMP IDENT IMB 0x01

Values for bit flags used in the ident_t to describe the fields.
Use trampoline for internal microtasks
Definition at line 170 of file kmp.h.

5.3.2.6 #define KMP IDENT KMPC 0x02

Use c-style ident structure

Definition at line 172 of file kmp.h.

Generated on Wed Sep 23 2015 13:02:36 for LLVM OpenMP∗ Runtime Library by Doxygen
22 Module Documentation

5.3.3 Typedef Documentation

5.3.3.1 typedef struct ident ident_t

The ident structure that describes a source location.

Generated on Wed Sep 23 2015 13:02:36 for LLVM OpenMP∗ Runtime Library by Doxygen
5.4 Deprecated Functions 23

5.4 Deprecated Functions

Functions

• kmp_int32 __kmpc_ok_to_fork (ident_t ∗loc)

5.4.1 Detailed Description

Functions in this group are for backwards compatibility only, and should not be used in new code.

5.4.2 Function Documentation

5.4.2.1 kmp int32 kmpc ok to fork ( ident_t ∗ loc )

Parameters
loc location description

This function need not be called. It always returns TRUE.

Definition at line 161 of file kmp_csupport.c.

Generated on Wed Sep 23 2015 13:02:36 for LLVM OpenMP∗ Runtime Library by Doxygen
24 Module Documentation

5.5 Startup and Shutdown

Functions

• void __kmpc_begin (ident_t ∗loc, kmp_int32 flags)

• void __kmpc_end (ident_t ∗loc)

5.5.1 Detailed Description

These functions are for library initialization and shutdown.

5.5.2 Function Documentation

5.5.2.1 void kmpc begin ( ident_t ∗ loc, kmp int32 flags )

Parameters
loc in source location information
flags in for future use (currently ignored)

Initialize the runtime library. This call is optional; if it is not made then it will be implicitly called by attempts to use
other library functions.
Definition at line 46 of file kmp_csupport.c.

5.5.2.2 void kmpc end ( ident_t ∗ loc )

Parameters
loc source location information

Shutdown the runtime library. This is also optional, and even if called will not do anything unless the KMP_IGNO-
RE_MPPEND environment variable is set to zero.
Definition at line 64 of file kmp_csupport.c.

Generated on Wed Sep 23 2015 13:02:36 for LLVM OpenMP∗ Runtime Library by Doxygen
5.6 Parallel (fork/join) 25

5.6 Parallel (fork/join)

Typedefs

• typedef void(∗ kmpc_micro )(kmp_int32 ∗global_tid, kmp_int32 ∗bound_tid,...)

Functions

• void __kmpc_push_num_threads (ident_t ∗loc, kmp_int32 global_tid, kmp_int32 num_threads)

• void __kmpc_fork_call (ident_t ∗loc, kmp_int32 argc, kmpc_micro microtask,...)
• void __kmpc_push_num_teams (ident_t ∗loc, kmp_int32 global_tid, kmp_int32 num_teams, kmp_int32 num-
_threads)
• void __kmpc_fork_teams (ident_t ∗loc, kmp_int32 argc, kmpc_micro microtask,...)
• void __kmpc_serialized_parallel (ident_t ∗loc, kmp_int32 global_tid)
• void __kmpc_end_serialized_parallel (ident_t ∗loc, kmp_int32 global_tid)

5.6.1 Detailed Description

These functions are used for implementing #pragma omp parallel.

5.6.2 Typedef Documentation

5.6.2.1 typedef void(∗ kmpc micro)(kmp int32 ∗global tid, kmp int32 ∗bound tid,...)

The type for a microtask which gets passed to __kmpc_fork_call(). The arguments to the outlined function are

Parameters
global_tid the global thread identity of the thread executing the function.
bound_tid the local identitiy of the thread executing the function
... pointers to shared variables accessed by the function.

Definition at line 1283 of file kmp.h.

5.6.3 Function Documentation

5.6.3.1 void kmpc end serialized parallel ( ident_t ∗ loc, kmp int32 global tid )

Parameters
loc source location information
global_tid global thread number

Leave a serialized parallel construct.

Definition at line 489 of file kmp_csupport.c.

5.6.3.2 void kmpc fork call ( ident_t ∗ loc, kmp int32 argc, kmpc_micro microtask, ... )

Parameters
loc source location information
argc total number of arguments in the ellipsis
microtask pointer to callback routine consisting of outlined parallel construct
... pointers to shared variables that aren’t global

Generated on Wed Sep 23 2015 13:02:36 for LLVM OpenMP∗ Runtime Library by Doxygen
26 Module Documentation

Do the actual fork and call the microtask in the relevant number of threads.
Definition at line 281 of file kmp_csupport.c.

5.6.3.3 void kmpc fork teams ( ident_t ∗ loc, kmp int32 argc, kmpc_micro microtask, ... )

Parameters
loc source location information
argc total number of arguments in the ellipsis
microtask pointer to callback routine consisting of outlined teams construct
... pointers to shared variables that aren’t global

Do the actual fork and call the microtask in the relevant number of threads.
Definition at line 385 of file kmp_csupport.c.

5.6.3.4 void kmpc push num teams ( ident_t ∗ loc, kmp int32 global tid, kmp int32 num teams, kmp int32 num threads )

Parameters
loc source location information
global_tid global thread number
num_teams number of teams requested for the teams construct
num_threads number of threads per team requested for the teams construct

Set the number of teams to be used by the teams construct. This call is only required if the teams construct has a
num_teams clause or a thread_limit clause (or both).
Definition at line 367 of file kmp_csupport.c.

5.6.3.5 void kmpc push num threads ( ident_t ∗ loc, kmp int32 global tid, kmp int32 num threads )

Parameters
loc source location information
global_tid global thread number
num_threads number of threads requested for this parallel construct

Set the number of threads to be used by the next fork spawned by this thread. This call is only required if the parallel
construct has a num_threads clause.
Definition at line 240 of file kmp_csupport.c.

5.6.3.6 void kmpc serialized parallel ( ident_t ∗ loc, kmp int32 global tid )

Parameters
loc source location information
global_tid global thread number

Enter a serialized parallel construct. This interface is used to handle a conditional parallel region, like this,

#pragma omp parallel if (condition)

when the condition is false.

Definition at line 474 of file kmp_csupport.c.

Generated on Wed Sep 23 2015 13:02:36 for LLVM OpenMP∗ Runtime Library by Doxygen
5.7 Thread Information 27

5.7 Thread Information

Functions

• kmp_int32 __kmpc_global_thread_num (ident_t ∗loc)

• kmp_int32 __kmpc_global_num_threads (ident_t ∗loc)
• kmp_int32 __kmpc_bound_thread_num (ident_t ∗loc)
• kmp_int32 __kmpc_bound_num_threads (ident_t ∗loc)
• kmp_int32 __kmpc_in_parallel (ident_t ∗loc)

5.7.1 Detailed Description

These functions return information about the currently executing thread.

5.7.2 Function Documentation

5.7.2.1 kmp int32 kmpc bound num threads ( ident_t ∗ loc )

Parameters
loc Source location information.

Returns

The number of threads in the innermost active parallel construct.

Definition at line 147 of file kmp_csupport.c.

5.7.2.2 kmp int32 kmpc bound thread num ( ident_t ∗ loc )

Parameters
loc Source location information.

Returns

The thread number of the calling thread in the innermost active parallel construct.

Definition at line 135 of file kmp_csupport.c.

5.7.2.3 kmp int32 kmpc global num threads ( ident_t ∗ loc )

Parameters
loc Source location information.

Returns

The number of threads under control of the OpenMP∗ runtime

This function can be called in any context. It returns the total number of threads under the control of the OpenMP
runtime. That is not a number that can be determined by any OpenMP standard calls, since the library may be
called from more than one non-OpenMP thread, and this reflects the total over all such calls. Similarly the runtime
maintains underlying threads even when they are not active (since the cost of creating and destroying OS threads
is high), this call counts all such threads even if they are not waiting for work.

Generated on Wed Sep 23 2015 13:02:36 for LLVM OpenMP∗ Runtime Library by Doxygen
28 Module Documentation

Definition at line 121 of file kmp_csupport.c.

5.7.2.4 kmp int32 kmpc global thread num ( ident_t ∗ loc )

Parameters
loc Source location information.

Returns

The global thread index of the active thread.

This function can be called in any context.

If the runtime has ony been entered at the outermost level from a single (necessarily non-OpenMP∗ ) thread,
then the thread number is that which would be returned by omp_get_thread_num() in the outermost active parallel
construct. (Or zero if there is no active parallel construct, since the master thread is necessarily thread zero).
If multiple non-OpenMP threads all enter an OpenMP construct then this will be a unique thread identifier among
all the threads created by the OpenMP runtime (but the value cannote be defined in terms of OpenMP thread ids
returned by omp_get_thread_num()).
Definition at line 98 of file kmp_csupport.c.

5.7.2.5 kmp int32 kmpc in parallel ( ident_t ∗ loc )

Parameters
loc Source location information.

Returns

1 if this thread is executing inside an active parallel region, zero if not.

Definition at line 225 of file kmp_csupport.c.

Referenced by __kmpc_fork_call().

Generated on Wed Sep 23 2015 13:02:36 for LLVM OpenMP∗ Runtime Library by Doxygen
5.8 Work Sharing 29

5.8 Work Sharing

Enumerations

• enum sched_type {
kmp_sch_lower = 32 , kmp_sch_static = 34 , kmp_sch_guided_chunked = 36 , kmp_sch_auto = 38 ,
kmp_sch_static_steal = 44, kmp_sch_upper = 45, kmp_ord_lower = 64 , kmp_ord_static = 66 ,
kmp_ord_auto = 70 , kmp_ord_upper = 72, kmp_distribute_static_chunked = 91, kmp_distribute_static = 92,
kmp_nm_lower = 160 , kmp_nm_static = 162 , kmp_nm_guided_chunked = 164 , kmp_nm_auto = 166 ,
kmp_nm_ord_static = 194 , kmp_nm_ord_auto = 198 , kmp_nm_upper = 200, kmp_sch_default = kmp_sch-
_static }

Functions

• kmp_int32 __kmpc_master (ident_t ∗loc, kmp_int32 global_tid)

• void __kmpc_end_master (ident_t ∗loc, kmp_int32 global_tid)
• void __kmpc_ordered (ident_t ∗loc, kmp_int32 gtid)
• void __kmpc_end_ordered (ident_t ∗loc, kmp_int32 gtid)
• void __kmpc_critical (ident_t ∗loc, kmp_int32 global_tid, kmp_critical_name ∗crit)
• void __kmpc_end_critical (ident_t ∗loc, kmp_int32 global_tid, kmp_critical_name ∗crit)
• kmp_int32 __kmpc_single (ident_t ∗loc, kmp_int32 global_tid)
• void __kmpc_end_single (ident_t ∗loc, kmp_int32 global_tid)
• void __kmpc_for_static_fini (ident_t ∗loc, kmp_int32 global_tid)

• void __kmpc_dispatch_init_4 (ident_t ∗loc, kmp_int32 gtid, enum sched_type schedule, kmp_int32 lb, kmp_-
int32 ub, kmp_int32 st, kmp_int32 chunk)
• void __kmpc_dispatch_init_4u (ident_t ∗loc, kmp_int32 gtid, enum sched_type schedule, kmp_uint32 lb, kmp-
_uint32 ub, kmp_int32 st, kmp_int32 chunk)
• void __kmpc_dispatch_init_8 (ident_t ∗loc, kmp_int32 gtid, enum sched_type schedule, kmp_int64 lb, kmp_-
int64 ub, kmp_int64 st, kmp_int64 chunk)
• void __kmpc_dispatch_init_8u (ident_t ∗loc, kmp_int32 gtid, enum sched_type schedule, kmp_uint64 lb, kmp-
_uint64 ub, kmp_int64 st, kmp_int64 chunk)
• void __kmpc_dist_dispatch_init_4 (ident_t ∗loc, kmp_int32 gtid, enum sched_type schedule, kmp_int32 ∗p_-
last, kmp_int32 lb, kmp_int32 ub, kmp_int32 st, kmp_int32 chunk)
• int __kmpc_dispatch_next_4 (ident_t ∗loc, kmp_int32 gtid, kmp_int32 ∗p_last, kmp_int32 ∗p_lb, kmp_int32
∗p_ub, kmp_int32 ∗p_st)
• int __kmpc_dispatch_next_4u (ident_t ∗loc, kmp_int32 gtid, kmp_int32 ∗p_last, kmp_uint32 ∗p_lb, kmp_-
uint32 ∗p_ub, kmp_int32 ∗p_st)
• int __kmpc_dispatch_next_8 (ident_t ∗loc, kmp_int32 gtid, kmp_int32 ∗p_last, kmp_int64 ∗p_lb, kmp_int64
∗p_ub, kmp_int64 ∗p_st)
• int __kmpc_dispatch_next_8u (ident_t ∗loc, kmp_int32 gtid, kmp_int32 ∗p_last, kmp_uint64 ∗p_lb, kmp_-
uint64 ∗p_ub, kmp_int64 ∗p_st)
• void __kmpc_dispatch_fini_4 (ident_t ∗loc, kmp_int32 gtid)
• void __kmpc_dispatch_fini_8 (ident_t ∗loc, kmp_int32 gtid)
• void __kmpc_dispatch_fini_4u (ident_t ∗loc, kmp_int32 gtid)
• void __kmpc_dispatch_fini_8u (ident_t ∗loc, kmp_int32 gtid)

• void __kmpc_for_static_init_4 (ident_t ∗loc, kmp_int32 gtid, kmp_int32 schedtype, kmp_int32 ∗plastiter, kmp-
_int32 ∗plower, kmp_int32 ∗pupper, kmp_int32 ∗pstride, kmp_int32 incr, kmp_int32 chunk)
• void __kmpc_for_static_init_4u (ident_t ∗loc, kmp_int32 gtid, kmp_int32 schedtype, kmp_int32 ∗plastiter,
kmp_uint32 ∗plower, kmp_uint32 ∗pupper, kmp_int32 ∗pstride, kmp_int32 incr, kmp_int32 chunk)
• void __kmpc_for_static_init_8 (ident_t ∗loc, kmp_int32 gtid, kmp_int32 schedtype, kmp_int32 ∗plastiter, kmp-
_int64 ∗plower, kmp_int64 ∗pupper, kmp_int64 ∗pstride, kmp_int64 incr, kmp_int64 chunk)
• void __kmpc_for_static_init_8u (ident_t ∗loc, kmp_int32 gtid, kmp_int32 schedtype, kmp_int32 ∗plastiter,
kmp_uint64 ∗plower, kmp_uint64 ∗pupper, kmp_int64 ∗pstride, kmp_int64 incr, kmp_int64 chunk)

Generated on Wed Sep 23 2015 13:02:36 for LLVM OpenMP∗ Runtime Library by Doxygen
30 Module Documentation

• void __kmpc_dist_for_static_init_4 (ident_t ∗loc, kmp_int32 gtid, kmp_int32 schedule, kmp_int32 ∗plastiter,
kmp_int32 ∗plower, kmp_int32 ∗pupper, kmp_int32 ∗pupperD, kmp_int32 ∗pstride, kmp_int32 incr, kmp_int32
chunk)
• void __kmpc_dist_for_static_init_4u (ident_t ∗loc, kmp_int32 gtid, kmp_int32 schedule, kmp_int32 ∗plastiter,
kmp_uint32 ∗plower, kmp_uint32 ∗pupper, kmp_uint32 ∗pupperD, kmp_int32 ∗pstride, kmp_int32 incr, kmp-
_int32 chunk)
• void __kmpc_dist_for_static_init_8 (ident_t ∗loc, kmp_int32 gtid, kmp_int32 schedule, kmp_int32 ∗plastiter,
kmp_int64 ∗plower, kmp_int64 ∗pupper, kmp_int64 ∗pupperD, kmp_int64 ∗pstride, kmp_int64 incr, kmp_int64
chunk)
• void __kmpc_dist_for_static_init_8u (ident_t ∗loc, kmp_int32 gtid, kmp_int32 schedule, kmp_int32 ∗plastiter,
kmp_uint64 ∗plower, kmp_uint64 ∗pupper, kmp_uint64 ∗pupperD, kmp_int64 ∗pstride, kmp_int64 incr, kmp-
_int64 chunk)

• void __kmpc_team_static_init_4 (ident_t ∗loc, kmp_int32 gtid, kmp_int32 ∗p_last, kmp_int32 ∗p_lb, kmp_-
int32 ∗p_ub, kmp_int32 ∗p_st, kmp_int32 incr, kmp_int32 chunk)
• void __kmpc_team_static_init_4u (ident_t ∗loc, kmp_int32 gtid, kmp_int32 ∗p_last, kmp_uint32 ∗p_lb, kmp_-
uint32 ∗p_ub, kmp_int32 ∗p_st, kmp_int32 incr, kmp_int32 chunk)
• void __kmpc_team_static_init_8 (ident_t ∗loc, kmp_int32 gtid, kmp_int32 ∗p_last, kmp_int64 ∗p_lb, kmp_-
int64 ∗p_ub, kmp_int64 ∗p_st, kmp_int64 incr, kmp_int64 chunk)
• void __kmpc_team_static_init_8u (ident_t ∗loc, kmp_int32 gtid, kmp_int32 ∗p_last, kmp_uint64 ∗p_lb, kmp_-
uint64 ∗p_ub, kmp_int64 ∗p_st, kmp_int64 incr, kmp_int64 chunk)

5.8.1 Detailed Description

These functions are used for implementing #pragma omp for, #pragma omp sections, #pragma omp
single and #pragma omp master constructs.
When handling loops, there are different functions for each of the signed and unsigned 32 and 64 bit integer types
which have the name suffixes _4, _4u, _8 and _8u. The semantics of each of the functions is the same, so they
are only described once.
Static loop scheduling is handled by __kmpc_for_static_init_4 and friends. Only a single call is needed, since the
iterations to be executed by any give thread can be determined as soon as the loop parameters are known.
Dynamic scheduling is handled by the __kmpc_dispatch_init_4 and __kmpc_dispatch_next_4 functions. The init
function is called once in each thread outside the loop, while the next function is called each time that the previous
chunk of work has been exhausted.

5.8.2 Enumeration Type Documentation

5.8.2.1 enum sched_type

Describes the loop schedule to be used for a parallel for loop.

Enumerator

kmp_sch_lower lower bound for unordered values

kmp_sch_static static unspecialized
kmp_sch_guided_chunked guided unspecialized
kmp_sch_auto auto
kmp_sch_static_steal accessible only through KMP_SCHEDULE environment variable
kmp_sch_upper upper bound for unordered values
kmp_ord_lower lower bound for ordered values, must be power of 2
kmp_ord_static ordered static unspecialized
kmp_ord_auto ordered auto

Generated on Wed Sep 23 2015 13:02:36 for LLVM OpenMP∗ Runtime Library by Doxygen
5.8 Work Sharing 31

kmp_ord_upper upper bound for ordered values

kmp_distribute_static_chunked distribute static chunked
kmp_distribute_static distribute static unspecialized
kmp_nm_lower lower bound for nomerge values
kmp_nm_static static unspecialized
kmp_nm_guided_chunked guided unspecialized
kmp_nm_auto auto
kmp_nm_ord_static ordered static unspecialized
kmp_nm_ord_auto auto
kmp_nm_upper upper bound for nomerge values
kmp_sch_default default scheduling algorithm

Definition at line 294 of file kmp.h.

5.8.3 Function Documentation

5.8.3.1 void kmpc critical ( ident_t ∗ loc, kmp int32 global tid, kmp critical name ∗ crit )

Parameters
loc source location information.
global_tid global thread number .
crit identity of the critical section. This could be a pointer to a lock associated with the critical
section, or some other suitably unique value.

Enter code protected by a critical construct. This function blocks until the executing thread can enter the critical
section.
Definition at line 1119 of file kmp_csupport.c.

5.8.3.2 void kmpc dispatch fini 4 ( ident_t ∗ loc, kmp int32 gtid )

Parameters
loc Source code location
gtid Global thread id

Mark the end of a dynamic loop.

Definition at line 2458 of file kmp_dispatch.cpp.

5.8.3.3 void kmpc dispatch fini 4u ( ident_t ∗ loc, kmp int32 gtid )

See __kmpc_dispatch_fini_4
Definition at line 2476 of file kmp_dispatch.cpp.

5.8.3.4 void kmpc dispatch fini 8 ( ident_t ∗ loc, kmp int32 gtid )

See __kmpc_dispatch_fini_4
Definition at line 2467 of file kmp_dispatch.cpp.

Generated on Wed Sep 23 2015 13:02:36 for LLVM OpenMP∗ Runtime Library by Doxygen
32 Module Documentation

5.8.3.5 void kmpc dispatch fini 8u ( ident_t ∗ loc, kmp int32 gtid )

See __kmpc_dispatch_fini_4
Definition at line 2485 of file kmp_dispatch.cpp.

5.8.3.6 void kmpc dispatch init 4 ( ident_t ∗ loc, kmp int32 gtid, enum sched_type schedule, kmp int32 lb, kmp int32
ub, kmp int32 st, kmp int32 chunk )

Parameters
loc Source location
gtid Global thread id
schedule Schedule type
lb Lower bound
ub Upper bound
st Step (or increment if you prefer)
chunk The chunk size to block with

This function prepares the runtime to start a dynamically scheduled for loop, saving the loop arguments. These
functions are all identical apart from the types of the arguments.
Definition at line 2316 of file kmp_dispatch.cpp.

5.8.3.7 void kmpc dispatch init 4u ( ident_t ∗ loc, kmp int32 gtid, enum sched_type schedule, kmp uint32 lb,
kmp uint32 ub, kmp int32 st, kmp int32 chunk )

See __kmpc_dispatch_init_4
Definition at line 2326 of file kmp_dispatch.cpp.

5.8.3.8 void kmpc dispatch init 8 ( ident_t ∗ loc, kmp int32 gtid, enum sched_type schedule, kmp int64 lb, kmp int64
ub, kmp int64 st, kmp int64 chunk )

See __kmpc_dispatch_init_4
Definition at line 2337 of file kmp_dispatch.cpp.

5.8.3.9 void kmpc dispatch init 8u ( ident_t ∗ loc, kmp int32 gtid, enum sched_type schedule, kmp uint64 lb,
kmp uint64 ub, kmp int64 st, kmp int64 chunk )

See __kmpc_dispatch_init_4
Definition at line 2349 of file kmp_dispatch.cpp.

5.8.3.10 int kmpc dispatch next 4 ( ident_t ∗ loc, kmp int32 gtid, kmp int32 ∗ p last, kmp int32 ∗ p lb, kmp int32 ∗ p ub,
kmp int32 ∗ p st )

Parameters
loc Source code location
gtid Global thread id
p_last Pointer to a flag set to one if this is the last chunk or zero otherwise
p_lb Pointer to the lower bound for the next chunk of work
p_ub Pointer to the upper bound for the next chunk of work
p_st Pointer to the stride for the next chunk of work

Generated on Wed Sep 23 2015 13:02:36 for LLVM OpenMP∗ Runtime Library by Doxygen
5.8 Work Sharing 33

Returns

one if there is work to be done, zero otherwise

Get the next dynamically allocated chunk of work for this thread. If there is no more work, then the lb,ub and stride
need not be modified.
Definition at line 2415 of file kmp_dispatch.cpp.

5.8.3.11 int kmpc dispatch next 4u ( ident_t ∗ loc, kmp int32 gtid, kmp int32 ∗ p last, kmp uint32 ∗ p lb, kmp uint32 ∗
p ub, kmp int32 ∗ p st )

See __kmpc_dispatch_next_4
Definition at line 2425 of file kmp_dispatch.cpp.

5.8.3.12 int kmpc dispatch next 8 ( ident_t ∗ loc, kmp int32 gtid, kmp int32 ∗ p last, kmp int64 ∗ p lb, kmp int64 ∗ p ub,
kmp int64 ∗ p st )

See __kmpc_dispatch_next_4
Definition at line 2435 of file kmp_dispatch.cpp.

5.8.3.13 int kmpc dispatch next 8u ( ident_t ∗ loc, kmp int32 gtid, kmp int32 ∗ p last, kmp uint64 ∗ p lb, kmp uint64 ∗
p ub, kmp int64 ∗ p st )

See __kmpc_dispatch_next_4
Definition at line 2445 of file kmp_dispatch.cpp.

5.8.3.14 void kmpc dist dispatch init 4 ( ident_t ∗ loc, kmp int32 gtid, enum sched_type schedule, kmp int32 ∗ p last,
kmp int32 lb, kmp int32 ub, kmp int32 st, kmp int32 chunk )

See __kmpc_dispatch_init_4
Difference from __kmpc_dispatch_init set of functions is these functions are called for composite distribute parallel
for construct. Thus before regular iterations dispatching we need to calc per-team iteration space.
These functions are all identical apart from the types of the arguments.
Definition at line 2367 of file kmp_dispatch.cpp.

5.8.3.15 void kmpc dist for static init 4 ( ident_t ∗ loc, kmp int32 gtid, kmp int32 schedule, kmp int32 ∗ plastiter,
kmp int32 ∗ plower, kmp int32 ∗ pupper, kmp int32 ∗ pupperD, kmp int32 ∗ pstride, kmp int32 incr, kmp int32
chunk )

Parameters
loc Source code location
gtid Global thread id of this thread
schedule Scheduling type for the parallel loop
plastiter Pointer to the "last iteration" flag
plower Pointer to the lower bound
pupper Pointer to the upper bound of loop chunk
pupperD Pointer to the upper bound of dist_chunk
pstride Pointer to the stride for parallel loop
incr Loop increment
chunk The chunk size for the parallel loop

Generated on Wed Sep 23 2015 13:02:36 for LLVM OpenMP∗ Runtime Library by Doxygen
34 Module Documentation

Each of the four functions here are identical apart from the argument types.
The functions compute the upper and lower bounds and strides to be used for the set of iterations to be executed
by the current thread from the statically scheduled loop that is described by the initial values of the bounds, strides,
increment and chunks for parallel loop and distribute constructs.
Definition at line 809 of file kmp_sched.cpp.

5.8.3.16 void kmpc dist for static init 4u ( ident_t ∗ loc, kmp int32 gtid, kmp int32 schedule, kmp int32 ∗ plastiter,
kmp uint32 ∗ plower, kmp uint32 ∗ pupper, kmp uint32 ∗ pupperD, kmp int32 ∗ pstride, kmp int32 incr, kmp int32
chunk )

See __kmpc_dist_for_static_init_4
Definition at line 822 of file kmp_sched.cpp.

5.8.3.17 void kmpc dist for static init 8 ( ident_t ∗ loc, kmp int32 gtid, kmp int32 schedule, kmp int32 ∗ plastiter,
kmp int64 ∗ plower, kmp int64 ∗ pupper, kmp int64 ∗ pupperD, kmp int64 ∗ pstride, kmp int64 incr, kmp int64
chunk )

See __kmpc_dist_for_static_init_4
Definition at line 835 of file kmp_sched.cpp.

5.8.3.18 void kmpc dist for static init 8u ( ident_t ∗ loc, kmp int32 gtid, kmp int32 schedule, kmp int32 ∗ plastiter,
kmp uint64 ∗ plower, kmp uint64 ∗ pupper, kmp uint64 ∗ pupperD, kmp int64 ∗ pstride, kmp int64 incr, kmp int64
chunk )

See __kmpc_dist_for_static_init_4
Definition at line 848 of file kmp_sched.cpp.

5.8.3.19 void kmpc end critical ( ident_t ∗ loc, kmp int32 global tid, kmp critical name ∗ crit )

Parameters
loc source location information.
global_tid global thread number .
crit identity of the critical section. This could be a pointer to a lock associated with the critical
section, or some other suitably unique value.

Leave a critical section, releasing any lock that was held during its execution.
Definition at line 1219 of file kmp_csupport.c.

5.8.3.20 void kmpc end master ( ident_t ∗ loc, kmp int32 global tid )

Parameters
loc source location information.
global_tid global thread number .

Mark the end of a master region. This should only be called by the thread that executes the master region.
Definition at line 789 of file kmp_csupport.c.

5.8.3.21 void kmpc end ordered ( ident_t ∗ loc, kmp int32 gtid )

Generated on Wed Sep 23 2015 13:02:36 for LLVM OpenMP∗ Runtime Library by Doxygen
5.8 Work Sharing 35

Parameters
loc source location information.
gtid global thread number.

End execution of an ordered construct.

Definition at line 889 of file kmp_csupport.c.

5.8.3.22 void kmpc end single ( ident_t ∗ loc, kmp int32 global tid )

Parameters
loc source location information
global_tid global thread number

Mark the end of a single construct. This function should only be called by the thread that executed the block of
code protected by the single construct.
Definition at line 1456 of file kmp_csupport.c.

5.8.3.23 void kmpc for static fini ( ident_t ∗ loc, kmp int32 global tid )

Parameters
loc Source location
global_tid Global thread id

Mark the end of a statically scheduled loop.

Definition at line 1483 of file kmp_csupport.c.

5.8.3.24 void kmpc for static init 4 ( ident_t ∗ loc, kmp int32 gtid, kmp int32 schedtype, kmp int32 ∗ plastiter, kmp int32
∗ plower, kmp int32 ∗ pupper, kmp int32 ∗ pstride, kmp int32 incr, kmp int32 chunk )

Parameters
loc Source code location
gtid Global thread id of this thread
schedtype Scheduling type
plastiter Pointer to the "last iteration" flag
plower Pointer to the lower bound
pupper Pointer to the upper bound
pstride Pointer to the stride
incr Loop increment
chunk The chunk size

Each of the four functions here are identical apart from the argument types.
The functions compute the upper and lower bounds and stride to be used for the set of iterations to be executed
by the current thread from the statically scheduled loop that is described by the initial values of the bounds, stride,
increment and chunk size.
Definition at line 739 of file kmp_sched.cpp.

5.8.3.25 void kmpc for static init 4u ( ident_t ∗ loc, kmp int32 gtid, kmp int32 schedtype, kmp int32 ∗ plastiter,
kmp uint32 ∗ plower, kmp uint32 ∗ pupper, kmp int32 ∗ pstride, kmp int32 incr, kmp int32 chunk )

See __kmpc_for_static_init_4

Generated on Wed Sep 23 2015 13:02:36 for LLVM OpenMP∗ Runtime Library by Doxygen
36 Module Documentation

Definition at line 751 of file kmp_sched.cpp.

5.8.3.26 void kmpc for static init 8 ( ident_t ∗ loc, kmp int32 gtid, kmp int32 schedtype, kmp int32 ∗ plastiter, kmp int64
∗ plower, kmp int64 ∗ pupper, kmp int64 ∗ pstride, kmp int64 incr, kmp int64 chunk )

See __kmpc_for_static_init_4
Definition at line 763 of file kmp_sched.cpp.

5.8.3.27 void kmpc for static init 8u ( ident_t ∗ loc, kmp int32 gtid, kmp int32 schedtype, kmp int32 ∗ plastiter,
kmp uint64 ∗ plower, kmp uint64 ∗ pupper, kmp int64 ∗ pstride, kmp int64 incr, kmp int64 chunk )

See __kmpc_for_static_init_4
Definition at line 775 of file kmp_sched.cpp.

5.8.3.28 kmp int32 kmpc master ( ident_t ∗ loc, kmp int32 global tid )

Parameters
loc source location information.
global_tid global thread number .

Returns

1 if this thread should execute the master block, 0 otherwise.

Definition at line 733 of file kmp_csupport.c.

Referenced by __kmpc_barrier_master_nowait().

5.8.3.29 void kmpc ordered ( ident_t ∗ loc, kmp int32 gtid )

Parameters
loc source location information.
gtid global thread number.

Start execution of an ordered construct.

Definition at line 825 of file kmp_csupport.c.

5.8.3.30 kmp int32 kmpc single ( ident_t ∗ loc, kmp int32 global tid )

Parameters
loc source location information
global_tid global thread number

Returns

One if this thread should execute the single construct, zero otherwise.

Test whether to execute a single construct. There are no implicit barriers in the two "single" calls, rather the
compiler should introduce an explicit barrier if it is required.
Definition at line 1411 of file kmp_csupport.c.

Generated on Wed Sep 23 2015 13:02:36 for LLVM OpenMP∗ Runtime Library by Doxygen
5.8 Work Sharing 37

5.8.3.31 void kmpc team static init 4 ( ident_t ∗ loc, kmp int32 gtid, kmp int32 ∗ p last, kmp int32 ∗ p lb, kmp int32 ∗
p ub, kmp int32 ∗ p st, kmp int32 incr, kmp int32 chunk )

Parameters
loc Source location
gtid Global thread id
p_last pointer to last iteration flag
p_lb pointer to Lower bound
p_ub pointer to Upper bound
p_st Step (or increment if you prefer)
incr Loop increment
chunk The chunk size to block with

The functions compute the upper and lower bounds and stride to be used for the set of iterations to be executed
by the current team from the statically scheduled loop that is described by the initial values of the bounds, stride,
increment and chunk for the distribute construct as part of composite distribute parallel loop construct. These
functions are all identical apart from the types of the arguments.
Definition at line 886 of file kmp_sched.cpp.

5.8.3.32 void kmpc team static init 4u ( ident_t ∗ loc, kmp int32 gtid, kmp int32 ∗ p last, kmp uint32 ∗ p lb, kmp uint32
∗ p ub, kmp int32 ∗ p st, kmp int32 incr, kmp int32 chunk )

See __kmpc_team_static_init_4
Definition at line 898 of file kmp_sched.cpp.

5.8.3.33 void kmpc team static init 8 ( ident_t ∗ loc, kmp int32 gtid, kmp int32 ∗ p last, kmp int64 ∗ p lb, kmp int64 ∗
p ub, kmp int64 ∗ p st, kmp int64 incr, kmp int64 chunk )

See __kmpc_team_static_init_4
Definition at line 910 of file kmp_sched.cpp.

5.8.3.34 void kmpc team static init 8u ( ident_t ∗ loc, kmp int32 gtid, kmp int32 ∗ p last, kmp uint64 ∗ p lb, kmp uint64
∗ p ub, kmp int64 ∗ p st, kmp int64 incr, kmp int64 chunk )

See __kmpc_team_static_init_4
Definition at line 922 of file kmp_sched.cpp.

Generated on Wed Sep 23 2015 13:02:36 for LLVM OpenMP∗ Runtime Library by Doxygen
38 Module Documentation

5.9 Synchronization

Functions

• void __kmpc_flush (ident_t ∗loc)

• void __kmpc_barrier (ident_t ∗loc, kmp_int32 global_tid)
• kmp_int32 __kmpc_barrier_master (ident_t ∗loc, kmp_int32 global_tid)
• void __kmpc_end_barrier_master (ident_t ∗loc, kmp_int32 global_tid)
• kmp_int32 __kmpc_barrier_master_nowait (ident_t ∗loc, kmp_int32 global_tid)
• kmp_int32 __kmpc_reduce_nowait (ident_t ∗loc, kmp_int32 global_tid, kmp_int32 num_vars, size_t reduce-
_size, void ∗reduce_data, void(∗reduce_func)(void ∗lhs_data, void ∗rhs_data), kmp_critical_name ∗lck)
• void __kmpc_end_reduce_nowait (ident_t ∗loc, kmp_int32 global_tid, kmp_critical_name ∗lck)
• kmp_int32 __kmpc_reduce (ident_t ∗loc, kmp_int32 global_tid, kmp_int32 num_vars, size_t reduce_size,
void ∗reduce_data, void(∗reduce_func)(void ∗lhs_data, void ∗rhs_data), kmp_critical_name ∗lck)
• void __kmpc_end_reduce (ident_t ∗loc, kmp_int32 global_tid, kmp_critical_name ∗lck)

5.9.1 Detailed Description

These functions are used for implementing barriers.

5.9.2 Function Documentation

5.9.2.1 void kmpc barrier ( ident_t ∗ loc, kmp int32 global tid )

Parameters
loc source location information
global_tid thread id.

Execute a barrier.
Definition at line 697 of file kmp_csupport.c.

5.9.2.2 kmp int32 kmpc barrier master ( ident_t ∗ loc, kmp int32 global tid )

Parameters
loc source location information
global_tid thread id.

Returns

one if the thread should execute the master block, zero otherwise

Start execution of a combined barrier and master. The barrier is executed inside this function.
Definition at line 1309 of file kmp_csupport.c.

5.9.2.3 kmp int32 kmpc barrier master nowait ( ident_t ∗ loc, kmp int32 global tid )

Parameters
loc source location information
global_tid thread id.

Generated on Wed Sep 23 2015 13:02:36 for LLVM OpenMP∗ Runtime Library by Doxygen
5.9 Synchronization 39

Returns

one if the thread should execute the master block, zero otherwise

Start execution of a combined barrier and master(nowait) construct. The barrier is executed inside this function.
There is no equivalent "end" function, since the
Definition at line 1357 of file kmp_csupport.c.

5.9.2.4 void kmpc end barrier master ( ident_t ∗ loc, kmp int32 global tid )

Parameters
loc source location information
global_tid thread id.

Complete the execution of a combined barrier and master. This function should only be called at the completion of
the master code. Other threads will still be waiting at the barrier and this call releases them.
Definition at line 1339 of file kmp_csupport.c.

5.9.2.5 void kmpc end reduce ( ident_t ∗ loc, kmp int32 global tid, kmp critical name ∗ lck )

Parameters
loc source location information
global_tid global thread id.
lck pointer to the unique lock data structure

Finish the execution of a blocking reduce. The lck pointer must be the same as that used in the corresponding
start function.
Definition at line 2761 of file kmp_csupport.c.

5.9.2.6 void kmpc end reduce nowait ( ident_t ∗ loc, kmp int32 global tid, kmp critical name ∗ lck )

Parameters
loc source location information
global_tid global thread id.
lck pointer to the unique lock data structure

Finish the execution of a reduce nowait.

Definition at line 2616 of file kmp_csupport.c.

5.9.2.7 void kmpc flush ( ident_t ∗ loc )

Parameters
loc source location information.

Execute flush. This is implemented as a full memory fence. (Though depending on the memory ordering con-
vention obeyed by the compiler even that may not be necessary).
Definition at line 631 of file kmp_csupport.c.

5.9.2.8 kmp int32 kmpc reduce ( ident_t ∗ loc, kmp int32 global tid, kmp int32 num vars, size t reduce size, void ∗
reduce data, void(∗)(void ∗lhs data, void ∗rhs data) reduce func, kmp critical name ∗ lck )

Generated on Wed Sep 23 2015 13:02:36 for LLVM OpenMP∗ Runtime Library by Doxygen
40 Module Documentation

Parameters
loc source location information
global_tid global thread number
num_vars number of items (variables) to be reduced
reduce_size size of data in bytes to be reduced
reduce_data pointer to data to be reduced
reduce_func callback function providing reduction operation on two operands and returning result of reduc-
tion in lhs_data
lck pointer to the unique lock data structure

Returns

1 for the master thread, 0 for all other team threads, 2 for all team threads if atomic reduction needed

A blocking reduce that includes an implicit barrier.

Definition at line 2674 of file kmp_csupport.c.

5.9.2.9 kmp int32 kmpc reduce nowait ( ident_t ∗ loc, kmp int32 global tid, kmp int32 num vars, size t reduce size, void
∗ reduce data, void(∗)(void ∗lhs data, void ∗rhs data) reduce func, kmp critical name ∗ lck )

Parameters
loc source location information
global_tid global thread number
num_vars number of items (variables) to be reduced
reduce_size size of data in bytes to be reduced
reduce_data pointer to data to be reduced
reduce_func callback function providing reduction operation on two operands and returning result of reduc-
tion in lhs_data
lck pointer to the unique lock data structure

Returns

1 for the master thread, 0 for all other team threads, 2 for all team threads if atomic reduction needed

The nowait version is used for a reduce clause with the nowait argument.
Definition at line 2477 of file kmp_csupport.c.

Generated on Wed Sep 23 2015 13:02:36 for LLVM OpenMP∗ Runtime Library by Doxygen
5.10 Thread private data support 41

5.10 Thread private data support

Functions

• void __kmpc_copyprivate (ident_t ∗loc, kmp_int32 gtid, size_t cpy_size, void ∗cpy_data, void(∗cpy_func)(void
∗, void ∗), kmp_int32 didit)
• void __kmpc_threadprivate_register (ident_t ∗loc, void ∗data, kmpc_ctor ctor, kmpc_cctor cctor, kmpc_dtor
dtor)
• void ∗ __kmpc_threadprivate_cached (ident_t ∗loc, kmp_int32 global_tid, void ∗data, size_t size, void
∗∗∗cache)
• void __kmpc_threadprivate_register_vec (ident_t ∗loc, void ∗data, kmpc_ctor_vec ctor, kmpc_cctor_vec cc-
tor, kmpc_dtor_vec dtor, size_t vector_length)

• typedef void ∗(∗ kmpc_ctor )(void ∗)

• typedef void(∗ kmpc_dtor )(void ∗)
• typedef void ∗(∗ kmpc_cctor )(void ∗, void ∗)
• typedef void ∗(∗ kmpc_ctor_vec )(void ∗, size_t)
• typedef void(∗ kmpc_dtor_vec )(void ∗, size_t)
• typedef void ∗(∗ kmpc_cctor_vec )(void ∗, void ∗, size_t)

5.10.1 Detailed Description

These functions support copyin/out and thread private data.

5.10.2 Typedef Documentation

5.10.2.1 typedef void∗(∗ kmpc cctor)(void ∗, void ∗)

Pointer to an alternate constructor. The first argument is the this pointer.

Definition at line 1310 of file kmp.h.

5.10.2.2 typedef void∗(∗ kmpc cctor vec)(void ∗, void ∗, size t)

Array constructor. First argument is the this pointer Third argument the number of array elements.
Definition at line 1332 of file kmp.h.

5.10.2.3 typedef void∗(∗ kmpc ctor)(void ∗)

Pointer to the constructor function. The first argument is the this pointer
Definition at line 1299 of file kmp.h.

5.10.2.4 typedef void∗(∗ kmpc ctor vec)(void ∗, size t)

Array constructor. First argument is the this pointer Second argument the number of array elements.
Definition at line 1320 of file kmp.h.

5.10.2.5 typedef void(∗ kmpc dtor)(void ∗)

Pointer to the destructor function. The first argument is the this pointer
Definition at line 1305 of file kmp.h.

Generated on Wed Sep 23 2015 13:02:36 for LLVM OpenMP∗ Runtime Library by Doxygen
42 Module Documentation

5.10.2.6 typedef void(∗ kmpc dtor vec)(void ∗, size t)

Pointer to the array destructor function. The first argument is the this pointer Second argument the number of
array elements.
Definition at line 1326 of file kmp.h.

5.10.3 Function Documentation

5.10.3.1 void kmpc copyprivate ( ident_t ∗ loc, kmp int32 gtid, size t cpy size, void ∗ cpy data, void(∗)(void ∗, void ∗)
cpy func, kmp int32 didit )

Parameters
loc source location information
gtid global thread number
cpy_size size of the cpy_data buffer
cpy_data pointer to data to be copied
cpy_func helper function to call for copying data
didit flag variable: 1=single thread; 0=not single thread

__kmpc_copyprivate implements the interface for the private data broadcast needed for the copyprivate clause
associated with a single region in an OpenMP∗ program (both C and Fortran). All threads participating in the
parallel region call this routine. One of the threads (called the single thread) should have the didit variable set to
1 and all other threads should have that variable set to 0. All threads pass a pointer to a data buffer (cpy_data) that
they have built.
The OpenMP specification forbids the use of nowait on the single region when a copyprivate clause is present.
However, __kmpc_copyprivate implements a barrier internally to avoid race conditions, so the code generation for
the single region should avoid generating a barrier after the call to __kmpc_copyprivate.
The gtid parameter is the global thread id for the current thread. The loc parameter is a pointer to source
location information.
Internal implementation: The single thread will first copy its descriptor address (cpy_data) to a team-private location,
then the other threads will each call the function pointed to by the parameter cpy_func, which carries out the copy
by copying the data using the cpy_data buffer.
The cpy_func routine used for the copy and the contents of the data area defined by cpy_data and cpy_size may
be built in any fashion that will allow the copy to be done. For instance, the cpy_data buffer can hold the actual data
to be copied or it may hold a list of pointers to the data. The cpy_func routine must interpret the cpy_data buffer
appropriately.
The interface to cpy_func is as follows:

void cpy_func( void *destination, void *source )

where void ∗destination is the cpy_data pointer for the thread being copied to and void ∗source is the cpy_data
pointer for the thread being copied from.
Definition at line 1694 of file kmp_csupport.c.

5.10.3.2 void∗ kmpc threadprivate cached ( ident_t ∗ loc, kmp int32 global tid, void ∗ data, size t size, void ∗∗∗ cache )

Parameters
loc source location information
global_tid global thread number
data pointer to data to privatize
size size of data to privatize
cache pointer to cache

Generated on Wed Sep 23 2015 13:02:36 for LLVM OpenMP∗ Runtime Library by Doxygen
5.10 Thread private data support 43

Returns

pointer to private storage

Allocate private storage for threadprivate data.

Definition at line 630 of file kmp_threadprivate.c.

5.10.3.3 void kmpc threadprivate register ( ident_t ∗ loc, void ∗ data, kmpc_ctor ctor, kmpc_cctor cctor, kmpc_dtor
dtor )

Parameters
loc source location information
data pointer to data being privatized
ctor pointer to constructor function for data
cctor pointer to copy constructor function for data
dtor pointer to destructor function for data

Register constructors and destructors for thread private data. This function is called when executing in parallel,
when we know the thread id.
Definition at line 532 of file kmp_threadprivate.c.

5.10.3.4 void kmpc threadprivate register vec ( ident_t ∗ loc, void ∗ data, kmpc_ctor_vec ctor, kmpc_cctor_vec
cctor, kmpc_dtor_vec dtor, size t vector length )

Parameters
loc source location information
data pointer to data being privatized
ctor pointer to constructor function for data
cctor pointer to copy constructor function for data
dtor pointer to destructor function for data
vector_length length of the vector (bytes or elements?) Register vector constructors and destructors for
thread private data.

Definition at line 699 of file kmp_threadprivate.c.

Generated on Wed Sep 23 2015 13:02:36 for LLVM OpenMP∗ Runtime Library by Doxygen
44 Module Documentation

5.11 Statistics Gathering from OMPTB

Classes

• class stats_flags_e
flags to describe the statistic ( timers or counter )

Macros

• #define KMP_FOREACH_COUNTER(macro, arg)

Add new counters under KMP_FOREACH_COUNTER() macro in kmp_stats.h.
• #define KMP_FOREACH_EXPLICIT_TIMER(macro, arg)
Add new explicit timers under KMP_FOREACH_EXPLICIT_TIMER() macro.
• #define KMP_TIME_BLOCK(name) blockTimer __BLOCKTIME__(__kmp_stats_thread_ptr->getTimer(TIM-
ER_##name), TIMER_##name)
Uses specified timer (name) to time code block.
• #define KMP_COUNT_VALUE(name, value) __kmp_stats_thread_ptr->getTimer(TIMER_##name)->add-
Sample(value)
Adds value to specified timer (name).
• #define KMP_COUNT_BLOCK(name) __kmp_stats_thread_ptr->getCounter(COUNTER_##name)-
>increment()
Increments specified counter (name).
• #define KMP_START_EXPLICIT_TIMER(name) __kmp_stats_thread_ptr->getExplicitTimer(EXPLICIT_TI-
MER_##name)->start(TIMER_##name)
"Starts" an explicit timer which will need a corresponding KMP_STOP_EXPLICIT_TIMER() macro.
• #define KMP_STOP_EXPLICIT_TIMER(name) __kmp_stats_thread_ptr->getExplicitTimer(EXPLICIT_TIM-
ER_##name)->stop(TIMER_##name)
"Stops" an explicit timer.
• #define KMP_OUTPUT_STATS(heading_string) __kmp_output_stats(heading_string)
Outputs the current thread statistics and reset them.
• #define KMP_RESET_STATS() __kmp_reset_stats()
resets all stats (counters to 0, timers to 0 elapsed ticks)

5.11.1 Detailed Description

These macros support profiling the libomp library. Use –stats=on when building with build.pl to enable and then use
the KMP_∗ macros to profile (through counts or clock ticks) libomp during execution of an OpenMP program.

5.11.2 Environment Variables

This section describes the environment variables relevant to stats-gathering in libomp

KMP_STATS_FILE

This environment variable is set to an output filename that will be appended NOT OVERWRITTEN if it exists. If this
environment variable is undefined, the statistics will be output to stderr

KMP_STATS_THREADS

This environment variable indicates to print thread-specific statistics as well as aggregate statistics. Each thread’s
statistics will be shown as well as the collective sum of all threads. The values "true", "on", "1", "yes" will all indicate
to print per thread statistics.

Generated on Wed Sep 23 2015 13:02:36 for LLVM OpenMP∗ Runtime Library by Doxygen
5.11 Statistics Gathering from OMPTB 45

5.11.3 Macro Definition Documentation

5.11.3.1 #define KMP COUNT BLOCK( name ) kmp stats thread ptr->getCounter(COUNTER ##name)->increment()

Increments specified counter (name).

Parameters
name counter name as specified under the KMP_FOREACH_COUNTER() macro

Use KMP_COUNT_BLOCK(name, value) macro to increment a statistics counter for the executing thread.
Definition at line 656 of file kmp_stats.h.
Referenced by __kmpc_barrier(), __kmpc_critical(), __kmpc_fork_call(), __kmpc_fork_teams(), __kmpc_master(),
__kmpc_reduce(), __kmpc_reduce_nowait(), and __kmpc_single().

5.11.3.2 #define KMP COUNT VALUE( name, value ) kmp stats thread ptr->getTimer(TIMER ##name)->addSample(value)

Adds value to specified timer (name).

Parameters
name timer name as specified under the KMP_FOREACH_TIMER() macro
value double precision sample value to add to statistics for the timer

Use KMP_COUNT_VALUE(name, value) macro to add a particular value to a timer statistics.

Definition at line 644 of file kmp_stats.h.

5.11.3.3 #define KMP FOREACH COUNTER( macro, arg )

Value:

macro (OMP_PARALLEL, stats_flags_e::onlyInMaster, arg) \

macro (OMP_NESTED_PARALLEL, 0, arg) \
macro (OMP_FOR_static, 0, arg) \
macro (OMP_FOR_dynamic, 0, arg) \
macro (OMP_DISTRIBUTE, 0, arg) \
macro (OMP_BARRIER, 0, arg) \
macro (OMP_CRITICAL,0, arg) \
macro (OMP_SINGLE, 0, arg) \
macro (OMP_MASTER, 0, arg) \
macro (OMP_TEAMS, 0, arg) \
macro (OMP_set_lock, 0, arg) \
macro (OMP_test_lock, 0, arg) \
macro (REDUCE_wait, 0, arg) \
macro (REDUCE_nowait, 0, arg) \
macro (OMP_TASKYIELD, 0, arg) \
macro (TASK_executed, 0, arg) \
macro (TASK_cancelled, 0, arg) \
macro (TASK_stolen, 0, arg) \
macro (LAST,0,arg)

Add new counters under KMP_FOREACH_COUNTER() macro in kmp_stats.h.

Parameters
macro a user defined macro that takes three arguments - macro(COUNTER_NAME, flags, arg)
arg a user defined argument to send to the user defined macro

A counter counts the occurrence of some event. Each thread accumulates its own count, at the end of execution
the counts are aggregated treating each thread as a separate measurement. (Unless onlyInMaster is set, in which
case there’s only a single measurement). The min,mean,max are therefore the values for the threads. Adding the

Generated on Wed Sep 23 2015 13:02:36 for LLVM OpenMP∗ Runtime Library by Doxygen
46 Module Documentation

counter here and then putting a KMP_BLOCK_COUNTER(name) at the point you want to count is all you need to
do. All of the tables and printing is generated from this macro. Format is "macro(name, flags, arg)"
Definition at line 70 of file kmp_stats.h.

5.11.3.4 #define KMP FOREACH EXPLICIT TIMER( macro, arg )

Value:

macro(OMP_serial, 0, arg) \
macro(OMP_start_end, 0, arg) \
macro(OMP_single, 0, arg) \
macro(OMP_master, 0, arg) \
KMP_FOREACH_EXPLICIT_DEVELOPER_TIMER(macro,arg) \
macro(LAST, 0, arg)

Add new explicit timers under KMP_FOREACH_EXPLICIT_TIMER() macro.

Parameters
macro a user defined macro that takes three arguments - macro(TIMER_NAME, flags, arg)
arg a user defined argument to send to the user defined macro

Warning

YOU MUST HAVE THE SAME NAMED TIMER UNDER KMP_FOREACH_TIMER() OR ELSE BAD THINGS
WILL HAPPEN!

Explicit timers are ones where we need to allocate a timer itself (as well as the accumulated timing statistics). We
allocate these on a per-thread basis, and explicitly start and stop them. Block timers just allocate the timer itself on
the stack, and use the destructor to notice block exit; they don’t need to be defined here. The name here should be
the same as that of a timer above.
Definition at line 198 of file kmp_stats.h.

5.11.3.5 #define KMP OUTPUT STATS( heading string ) kmp output stats(heading string)

Outputs the current thread statistics and reset them.

Parameters
heading_string heading put above the final stats output

Explicitly stops all timers and outputs all stats. Environment variable, OMPTB_STATSFILE=filename, can
be used to output the stats to a filename instead of stderr Environment variable, OMPTB_STATSTHREAD-
S=true|undefined, can be used to output thread specific stats For now the OMPTB_STATSTHREADS en-
vironment variable can either be defined with any value, which will print out thread specific stats, or it can be
undefined (not specified in the environment) and thread specific stats won’t be printed It should be noted that all
statistics are reset when this macro is called.
Definition at line 701 of file kmp_stats.h.

5.11.3.6 #define KMP RESET STATS( ) kmp reset stats()

resets all stats (counters to 0, timers to 0 elapsed ticks)

Reset all stats for all threads.
Definition at line 711 of file kmp_stats.h.

Generated on Wed Sep 23 2015 13:02:36 for LLVM OpenMP∗ Runtime Library by Doxygen
5.11 Statistics Gathering from OMPTB 47

5.11.3.7 #define KMP START EXPLICIT TIMER( name ) kmp stats thread ptr->getExplicitTimer(EXPLICIT TIMER ##name)-
>start(TIMER ##name)

"Starts" an explicit timer which will need a corresponding KMP_STOP_EXPLICIT_TIMER() macro.

Parameters
name explicit timer name as specified under the KMP_FOREACH_EXPLICIT_TIMER() macro

Use to start a timer. This will need a corresponding KMP_STOP_EXPLICIT_TIMER() macro to stop the timer unlike
the KMP_TIME_BLOCK(name) macro which has an implicit stopping macro at the end of the code block. All explicit
timers are stopped at library exit time before the final statistics are outputted.
Definition at line 670 of file kmp_stats.h.
Referenced by __kmpc_fork_call(), __kmpc_master(), and __kmpc_single().

5.11.3.8 #define KMP STOP EXPLICIT TIMER( name ) kmp stats thread ptr->getExplicitTimer(EXPLICIT TIMER ##name)-
>stop(TIMER ##name)

"Stops" an explicit timer.

Parameters
name explicit timer name as specified under the KMP_FOREACH_EXPLICIT_TIMER() macro

Use KMP_STOP_EXPLICIT_TIMER(name) to stop a timer. When this is done, the time between the last KMP_-
START_EXPLICIT_TIMER(name) and this KMP_STOP_EXPLICIT_TIMER(name) will be added to the timer’s stat
value. The timer will then be reset. After the KMP_STOP_EXPLICIT_TIMER(name) macro is called, another call to
KMP_START_EXPLICIT_TIMER(name) will start the timer once again.
Definition at line 684 of file kmp_stats.h.
Referenced by __kmpc_end_master(), __kmpc_end_single(), and __kmpc_fork_call().

5.11.3.9 #define KMP TIME BLOCK( name ) blockTimer BLOCKTIME ( kmp stats thread ptr->getTimer(TIMER ##name),
TIMER ##name)

Uses specified timer (name) to time code block.

Parameters
name timer name as specified under the KMP_FOREACH_TIMER() macro

Use KMP_TIME_BLOCK(name) macro to time a code block. This will record the time taken in the block and use
the destructor to stop the timer. Convenient! With this definition you can’t have more than one KMP_TIME_BLOCK
in the same code block. I don’t think that’s a problem.
Definition at line 631 of file kmp_stats.h.
Referenced by __kmpc_barrier().

Generated on Wed Sep 23 2015 13:02:36 for LLVM OpenMP∗ Runtime Library by Doxygen
48 Module Documentation

5.12 Tasking support

Functions

• kmp_int32 __kmpc_omp_task_with_deps (ident_t ∗loc_ref, kmp_int32 gtid, kmp_task_t ∗new_task, kmp_-

int32 ndeps, kmp_depend_info_t ∗dep_list, kmp_int32 ndeps_noalias, kmp_depend_info_t ∗noalias_dep_-
list)
• void __kmpc_omp_wait_deps (ident_t ∗loc_ref, kmp_int32 gtid, kmp_int32 ndeps, kmp_depend_info_t ∗dep-
_list, kmp_int32 ndeps_noalias, kmp_depend_info_t ∗noalias_dep_list)

5.12.1 Detailed Description

These functions support tasking constructs.

5.12.2 Function Documentation

5.12.2.1 kmp int32 kmpc omp task with deps ( ident_t ∗ loc ref, kmp int32 gtid, kmp task t ∗ new task, kmp int32 ndeps,
kmp depend info t ∗ dep list, kmp int32 ndeps noalias, kmp depend info t ∗ noalias dep list )

Parameters
loc_ref location of the original task directive
gtid Global Thread ID of encountering thread
new_task task thunk allocated by __kmp_omp_task_alloc() for the ”new task”
ndeps Number of depend items with possible aliasing
dep_list List of depend items with possible aliasing
ndeps_noalias Number of depend items with no aliasing
noalias_dep_list List of depend items with no aliasing

Returns

Returns either TASK_CURRENT_NOT_QUEUED if the current task was not suspendend and queued, or TA-
SK_CURRENT_QUEUED if it was suspended and queued

Schedule a non-thread-switchable task with dependences for execution

Definition at line 395 of file kmp_taskdeps.cpp.

5.12.2.2 void kmpc omp wait deps ( ident_t ∗ loc ref, kmp int32 gtid, kmp int32 ndeps, kmp depend info t ∗ dep list,
kmp int32 ndeps noalias, kmp depend info t ∗ noalias dep list )

Parameters
loc_ref location of the original task directive
gtid Global Thread ID of encountering thread
ndeps Number of depend items with possible aliasing
dep_list List of depend items with possible aliasing
ndeps_noalias Number of depend items with no aliasing
noalias_dep_list List of depend items with no aliasing

Blocks the current task until all specifies dependencies have been fulfilled.
Definition at line 463 of file kmp_taskdeps.cpp.
Referenced by __kmpc_omp_task_with_deps().

Generated on Wed Sep 23 2015 13:02:36 for LLVM OpenMP∗ Runtime Library by Doxygen
5.13 User visible functions 49

5.13 User visible functions

These functions can be called directly by the user, but are runtime library specific, rather than being OpenMP
interfaces.

Generated on Wed Sep 23 2015 13:02:36 for LLVM OpenMP∗ Runtime Library by Doxygen
50 Module Documentation

Generated on Wed Sep 23 2015 13:02:36 for LLVM OpenMP∗ Runtime Library by Doxygen
Chapter 6

Class Documentation

6.1 hierarchy info Class Reference

#include <kmp_affinity.h>

Public Attributes

• kmp_uint32 maxLevels
• kmp_uint32 depth
• kmp_uint32 ∗ numPerLevel

Static Public Attributes

• static const kmp_uint32 maxLeaves =4

6.1.1 Detailed Description

A structure for holding machine-specific hierarchy info to be computed once at init. This structure represents a
mapping of threads to the actual machine hierarchy, or to our best guess at what the hierarchy might be, for the
purpose of performing an efficient barrier. In the worst case, when there is no machine hierarchy information, it
produces a tree suitable for a barrier, similar to the tree used in the hyper barrier.
Definition at line 127 of file kmp_affinity.h.

6.1.2 Member Data Documentation

6.1.2.1 kmp uint32 hierarchy info::depth

This is specifically the depth of the machine configuration hierarchy, in terms of the number of levels along the
longest path from root to any leaf. It corresponds to the number of entries in numPerLevel if we exclude all but one
trailing 1.
Definition at line 142 of file kmp_affinity.h.

6.1.2.2 const kmp uint32 hierarchy info::maxLeaves =4 [static]

Good default values for number of leaves and branching factor, given no affinity information. Behaves a bit like hyper
barrier.
Definition at line 131 of file kmp_affinity.h.
52 Class Documentation

6.1.2.3 kmp uint32 hierarchy info::maxLevels

Number of levels in the hierarchy. Typical levels are threads/core, cores/package or socket, packages/node, nodes/-
machine, etc. We don’t want to get specific with nomenclature. When the machine is oversubscribed we add levels
to duplicate the hierarchy, doubling the thread capacity of the hierarchy each time we add a level.
Definition at line 137 of file kmp_affinity.h.

6.1.2.4 kmp uint32∗ hierarchy info::numPerLevel

Level 0 corresponds to leaves. numPerLevel[i] is the number of children the parent of a node at level i has. For
example, if we have a machine with 4 packages, 4 cores/package and 2 HT per core, then numPerLevel = {2, 4, 4,
1, 1}. All empty levels are set to 1.
Definition at line 151 of file kmp_affinity.h.
The documentation for this class was generated from the following file:

• kmp_affinity.h

6.2 ident Struct Reference

#include <kmp.h>

Public Attributes

• kmp_int32 reserved_1
• kmp_int32 flags
• kmp_int32 reserved_2
• kmp_int32 reserved_3
• char const ∗ psource

6.2.1 Detailed Description

The ident structure that describes a source location.

Definition at line 192 of file kmp.h.

6.2.2 Member Data Documentation

6.2.2.1 kmp int32 ident::flags

also f.flags; KMP_IDENT_xxx flags; KMP_IDENT_KMPC identifies this union member

Definition at line 194 of file kmp.h.
Referenced by __kmpc_end_serialized_parallel().

6.2.2.2 char const∗ ident::psource

String describing the source location.

The string is composed of semi-colon separated fields which describe the source file, the function and a pair of line
numbers that delimit the construct.

Generated on Wed Sep 23 2015 13:02:36 for LLVM OpenMP∗ Runtime Library by Doxygen
6.3 kmp_flag< P > Class Template Reference 53

Definition at line 201 of file kmp.h.

Referenced by __kmpc_ok_to_fork().

6.2.2.3 kmp int32 ident::reserved 1

might be used in Fortran; see above

Definition at line 193 of file kmp.h.

6.2.2.4 kmp int32 ident::reserved 2

not really used in Fortran any more; see above

Definition at line 195 of file kmp.h.

6.2.2.5 kmp int32 ident::reserved 3

source[4] in Fortran, do not use for C++

Definition at line 200 of file kmp.h.
The documentation for this struct was generated from the following file:

• kmp.h

6.3 kmp flag< P > Class Template Reference

#include <kmp_wait_release.h>

Public Member Functions

• volatile P ∗ get ()
• flag_type get_type ()

Private Attributes

• volatile P ∗ loc
• flag_type t

6.3.1 Detailed Description

template<typename P>class kmp flag< P >

Base class for wait/release volatile flag

Definition at line 48 of file kmp_wait_release.h.

6.3.2 Member Function Documentation

6.3.2.1 template<typename P> volatile P∗ kmp_flag< P >::get ( ) [inline]

Generated on Wed Sep 23 2015 13:02:36 for LLVM OpenMP∗ Runtime Library by Doxygen
54 Class Documentation

Returns

the pointer to the actual flag

Definition at line 57 of file kmp_wait_release.h.

6.3.2.2 template<typename P> flag_type kmp_flag< P >::get type ( ) [inline]

Returns

the flag_type

Definition at line 61 of file kmp_wait_release.h.

6.3.3 Member Data Documentation

6.3.3.1 template<typename P> volatile P∗ kmp_flag< P >::loc [private]

Pointer to the flag storage that is modified by another thread

Definition at line 49 of file kmp_wait_release.h.
Referenced by kmp_flag< kmp_uint32 >::get().

6.3.3.2 template<typename P> flag_type kmp_flag< P >::t [private]

"Type" of the flag in loc

Definition at line 50 of file kmp_wait_release.h.
Referenced by kmp_flag< kmp_uint32 >::get_type().
The documentation for this class was generated from the following file:

• kmp_wait_release.h

6.4 stats flags e Class Reference

flags to describe the statistic ( timers or counter )

#include <kmp_stats.h>

Static Public Attributes

• static const int onlyInMaster = 1<<0

statistic is valid only for master
• static const int noUnits = 1<<1
statistic doesn’t need units printed next to it in output
• static const int synthesized = 1<<2
statistic’s value is created atexit time in the __kmp_output_stats function
• static const int notInMaster = 1<<3
statistic is valid for non-master threads
• static const int logEvent = 1<<4
statistic can be logged when KMP_STATS_EVENTS is on (valid only for timers)

Generated on Wed Sep 23 2015 13:02:36 for LLVM OpenMP∗ Runtime Library by Doxygen
6.4 stats_flags_e Class Reference 55

6.4.1 Detailed Description

flags to describe the statistic ( timers or counter )

Definition at line 45 of file kmp_stats.h.
The documentation for this class was generated from the following file:

• kmp_stats.h

Generated on Wed Sep 23 2015 13:02:36 for LLVM OpenMP∗ Runtime Library by Doxygen
Index

__kmpc_barrier Work Sharing, 34

Synchronization, 38 __kmpc_end
__kmpc_barrier_master Startup and Shutdown, 24
Synchronization, 38 __kmpc_end_barrier_master
__kmpc_barrier_master_nowait Synchronization, 39
Synchronization, 38 __kmpc_end_critical
__kmpc_begin Work Sharing, 34
Startup and Shutdown, 24 __kmpc_end_master
__kmpc_bound_num_threads Work Sharing, 34
Thread Information, 27 __kmpc_end_ordered
__kmpc_bound_thread_num Work Sharing, 34
Thread Information, 27 __kmpc_end_reduce
__kmpc_copyprivate Synchronization, 39
Thread private data support, 42 __kmpc_end_reduce_nowait
__kmpc_critical Synchronization, 39
Work Sharing, 31 __kmpc_end_serialized_parallel
__kmpc_dispatch_fini_4 Parallel (fork/join), 25
Work Sharing, 31 __kmpc_end_single
__kmpc_dispatch_fini_4u Work Sharing, 35
Work Sharing, 31 __kmpc_flush
__kmpc_dispatch_fini_8 Synchronization, 39
Work Sharing, 31 __kmpc_for_static_fini
__kmpc_dispatch_fini_8u Work Sharing, 35
Work Sharing, 31 __kmpc_for_static_init_4
__kmpc_dispatch_init_4 Work Sharing, 35
Work Sharing, 32 __kmpc_for_static_init_4u
__kmpc_dispatch_init_4u Work Sharing, 35
Work Sharing, 32 __kmpc_for_static_init_8
__kmpc_dispatch_init_8 Work Sharing, 36
Work Sharing, 32 __kmpc_for_static_init_8u
__kmpc_dispatch_init_8u Work Sharing, 36
Work Sharing, 32 __kmpc_fork_call
__kmpc_dispatch_next_4 Parallel (fork/join), 25
Work Sharing, 32 __kmpc_fork_teams
__kmpc_dispatch_next_4u Parallel (fork/join), 26
Work Sharing, 33 __kmpc_global_num_threads
__kmpc_dispatch_next_8 Thread Information, 27
Work Sharing, 33 __kmpc_global_thread_num
__kmpc_dispatch_next_8u Thread Information, 28
Work Sharing, 33 __kmpc_in_parallel
__kmpc_dist_dispatch_init_4 Thread Information, 28
Work Sharing, 33 __kmpc_master
__kmpc_dist_for_static_init_4 Work Sharing, 36
Work Sharing, 33 __kmpc_ok_to_fork
__kmpc_dist_for_static_init_4u Deprecated Functions, 23
Work Sharing, 34 __kmpc_omp_task_with_deps
__kmpc_dist_for_static_init_8 Tasking support, 48
Work Sharing, 34 __kmpc_omp_wait_deps
__kmpc_dist_for_static_init_8u Tasking support, 48
INDEX 57

__kmpc_ordered maxLeaves, 51
Work Sharing, 36 maxLevels, 51
__kmpc_push_num_teams numPerLevel, 52
Parallel (fork/join), 26
__kmpc_push_num_threads ident, 52
Parallel (fork/join), 26 flags, 52
__kmpc_reduce psource, 52
Synchronization, 39 reserved_1, 53
__kmpc_reduce_nowait reserved_2, 53
Synchronization, 40 reserved_3, 53
__kmpc_serialized_parallel ident_t
Parallel (fork/join), 26 Basic Types, 22
__kmpc_single
Work Sharing, 36 KMP_COUNT_BLOCK
__kmpc_team_static_init_4 Statistics Gathering from OMPTB, 45
Work Sharing, 36 KMP_COUNT_VALUE
__kmpc_team_static_init_4u Statistics Gathering from OMPTB, 45
Work Sharing, 37 KMP_IDENT_AUTOPAR
__kmpc_team_static_init_8 Basic Types, 21
Work Sharing, 37 KMP_IDENT_IMB
__kmpc_team_static_init_8u Basic Types, 21
Work Sharing, 37 KMP_IDENT_KMPC
__kmpc_threadprivate_cached Basic Types, 21
Thread private data support, 42 KMP_OUTPUT_STATS
__kmpc_threadprivate_register Statistics Gathering from OMPTB, 46
Thread private data support, 43 KMP_RESET_STATS
__kmpc_threadprivate_register_vec Statistics Gathering from OMPTB, 46
Thread private data support, 43 KMP_TIME_BLOCK
Statistics Gathering from OMPTB, 47
Atomic Operations, 13 kmp_distribute_static
Work Sharing, 31
Basic Types, 21 kmp_distribute_static_chunked
ident_t, 22 Work Sharing, 31
KMP_IDENT_AUTOPAR, 21 kmp_nm_auto
KMP_IDENT_IMB, 21 Work Sharing, 31
KMP_IDENT_KMPC, 21 kmp_nm_guided_chunked
Work Sharing, 31
Deprecated Functions, 23 kmp_nm_lower
__kmpc_ok_to_fork, 23 Work Sharing, 31
depth kmp_nm_ord_auto
hierarchy_info, 51 Work Sharing, 31
kmp_nm_ord_static
flag32
Work Sharing, 31
Wait/Release operations, 20
kmp_nm_static
flag64
Work Sharing, 31
Wait/Release operations, 20
kmp_nm_upper
flag_oncore
Work Sharing, 31
Wait/Release operations, 20
kmp_ord_auto
flag_type
Work Sharing, 30
Wait/Release operations, 20
kmp_ord_lower
flags
Work Sharing, 30
ident, 52
kmp_ord_static
get Work Sharing, 30
kmp_flag, 53 kmp_ord_upper
get_type Work Sharing, 30
kmp_flag, 54 kmp_sch_auto
Work Sharing, 30
hierarchy_info, 51 kmp_sch_default
depth, 51 Work Sharing, 31

Generated on Wed Sep 23 2015 13:02:36 for LLVM OpenMP∗ Runtime Library by Doxygen
58 INDEX

kmp_sch_guided_chunked sched_type
Work Sharing, 30 Work Sharing, 30
kmp_sch_lower Startup and Shutdown, 24
Work Sharing, 30 __kmpc_begin, 24
kmp_sch_static __kmpc_end, 24
Work Sharing, 30 Statistics Gathering from OMPTB, 44
kmp_sch_static_steal KMP_COUNT_BLOCK, 45
Work Sharing, 30 KMP_COUNT_VALUE, 45
kmp_sch_upper KMP_RESET_STATS, 46
Work Sharing, 30 KMP_TIME_BLOCK, 47
kmp_flag stats_flags_e, 54
get, 53 Synchronization, 38
get_type, 54 __kmpc_barrier, 38
loc, 54 __kmpc_barrier_master, 38
t, 54 __kmpc_barrier_master_nowait, 38
kmp_flag< P >, 53 __kmpc_end_barrier_master, 39
kmpc_cctor __kmpc_end_reduce, 39
Thread private data support, 41 __kmpc_end_reduce_nowait, 39
kmpc_cctor_vec __kmpc_flush, 39
Thread private data support, 41 __kmpc_reduce, 39
kmpc_ctor __kmpc_reduce_nowait, 40
Thread private data support, 41
kmpc_ctor_vec t
Thread private data support, 41 kmp_flag, 54
kmpc_dtor Tasking support, 48
Thread private data support, 41 __kmpc_omp_task_with_deps, 48
kmpc_dtor_vec __kmpc_omp_wait_deps, 48
Thread private data support, 41 Thread Information, 27
kmpc_micro __kmpc_bound_num_threads, 27
Parallel (fork/join), 25 __kmpc_bound_thread_num, 27
__kmpc_global_num_threads, 27
loc __kmpc_global_thread_num, 28
kmp_flag, 54 __kmpc_in_parallel, 28
Thread private data support, 41
maxLeaves __kmpc_copyprivate, 42
hierarchy_info, 51 __kmpc_threadprivate_cached, 42
maxLevels __kmpc_threadprivate_register, 43
hierarchy_info, 51 __kmpc_threadprivate_register_vec, 43
kmpc_cctor, 41
numPerLevel kmpc_cctor_vec, 41
hierarchy_info, 52 kmpc_ctor, 41
kmpc_ctor_vec, 41
Parallel (fork/join), 25 kmpc_dtor, 41
__kmpc_end_serialized_parallel, 25 kmpc_dtor_vec, 41
__kmpc_fork_call, 25
__kmpc_fork_teams, 26 User visible functions, 49
__kmpc_push_num_teams, 26
__kmpc_push_num_threads, 26 Wait/Release operations
__kmpc_serialized_parallel, 26 flag32, 20
kmpc_micro, 25 flag64, 20
psource flag_oncore, 20
ident, 52 Wait/Release operations, 20
flag_type, 20
reserved_1 Work Sharing, 29
ident, 53 __kmpc_critical, 31
reserved_2 __kmpc_dispatch_fini_4, 31
ident, 53 __kmpc_dispatch_fini_4u, 31
reserved_3 __kmpc_dispatch_fini_8, 31
ident, 53 __kmpc_dispatch_fini_8u, 31

Generated on Wed Sep 23 2015 13:02:36 for LLVM OpenMP∗ Runtime Library by Doxygen
INDEX 59

__kmpc_dispatch_init_4, 32
__kmpc_dispatch_init_4u, 32
__kmpc_dispatch_init_8, 32
__kmpc_dispatch_init_8u, 32
__kmpc_dispatch_next_4, 32
__kmpc_dispatch_next_4u, 33
__kmpc_dispatch_next_8, 33
__kmpc_dispatch_next_8u, 33
__kmpc_dist_dispatch_init_4, 33
__kmpc_dist_for_static_init_4, 33
__kmpc_dist_for_static_init_4u, 34
__kmpc_dist_for_static_init_8, 34
__kmpc_dist_for_static_init_8u, 34
__kmpc_end_critical, 34
__kmpc_end_master, 34
__kmpc_end_ordered, 34
__kmpc_end_single, 35
__kmpc_for_static_fini, 35
__kmpc_for_static_init_4, 35
__kmpc_for_static_init_4u, 35
__kmpc_for_static_init_8, 36
__kmpc_for_static_init_8u, 36
__kmpc_master, 36
__kmpc_ordered, 36
__kmpc_single, 36
__kmpc_team_static_init_4, 36
__kmpc_team_static_init_4u, 37
__kmpc_team_static_init_8, 37
__kmpc_team_static_init_8u, 37
kmp_distribute_static, 31
kmp_distribute_static_chunked, 31
kmp_nm_auto, 31
kmp_nm_guided_chunked, 31
kmp_nm_lower, 31
kmp_nm_ord_auto, 31
kmp_nm_ord_static, 31
kmp_nm_static, 31
kmp_nm_upper, 31
kmp_ord_auto, 30
kmp_ord_lower, 30
kmp_ord_static, 30
kmp_ord_upper, 30
kmp_sch_auto, 30
kmp_sch_default, 31
kmp_sch_guided_chunked, 30
kmp_sch_lower, 30
kmp_sch_static, 30
kmp_sch_static_steal, 30
kmp_sch_upper, 30
sched_type, 30

Generated on Wed Sep 23 2015 13:02:36 for LLVM OpenMP∗ Runtime Library by Doxygen