Section 2.1 shows the scope and description of each ICV.
| | ||
| ICV | Scope |
Description |
| active-levels-var | data environment |
Number of nested active parallel regions such that all parallel regions are enclosed by the outermost initial task region on the device |
| affinity-format-var | device |
Controls the thread affinity format when displaying thread affinity |
| bind-var | data environment |
Controls the binding of OpenMP threads to places; when binding is requested, indicates that the execution environment is advised not to move threads between places; can also provide default thread affinity policies |
| cancel-var | global |
Controls the desired behavior of the cancel construct and cancellation points |
| debug-var | global |
Controls whether an OpenMP implementation will collect information that an OMPD library can access to satisfy requests from a tool |
| def-allocator-var | implicit task |
Controls the memory allocator used by memory allocation routines, directives and clauses that do not specify one explicitly |
| default-device-var | data environment |
Controls the default target device |
| display-affinity-var | global |
Controls the display of thread affinity |
| dyn-var | data environment |
Enables dynamic adjustment of the number of threads used for encountered parallel regions |
| explicit-task-var | data environment |
Whether a given task is an explicit task |
| final-task-var | data environment |
Whether a given task is a final task |
| levels-var | data environment |
Number of nested parallel regions such that all parallel regions are enclosed by the outermost initial task region on the device |
| max-active-levels-var | data environment |
Controls the maximum number of nested active parallel regions when the innermost parallel region is generated by a given task |
| max-task-priority-var | global |
Controls the maximum value that can be specified in the priority clause |
| nteams-var | device |
Controls the number of teams requested for encountered teams regions |
| nthreads-var | data environment |
Controls the number of threads requested for encountered parallel regions |
| num-procs-var | device |
The number of processors available on the device |
| place-partition-var | implicit task |
Controls the place partition available for encountered parallel regions |
| run-sched-var | data environment |
Controls the schedule used for worksharing-loop regions that specify the runtime schedule kind |
| stacksize-var | device |
Controls the stack size for threads that the OpenMP implementation creates |
| target-offload-var | global |
Controls the offloading behavior |
| team-size-var | data environment |
Size of the current team |
| teams-thread-limit-var | device |
Controls the maximum number of threads in each contention group that a teams construct creates |
| thread-limit-var | data environment |
Controls the maximum number of threads that participate in the contention group |
| thread-num-var | data environment |
Thread number of an implicit task within its binding team |
| tool-libraries-var | global |
List of absolute paths to tool libraries |
| tool-var | global |
Indicates that a tool will be registered |
| tool-verbose-init-var | global |
Controls whether an OpenMP implementation will verbosely log the registration of a tool |
| wait-policy-var | device |
Controls the desired behavior of waiting threads |