41 lines
1.3 KiB
ReStructuredText
41 lines
1.3 KiB
ReStructuredText
.. _omp170:
|
|
|
|
OpenMP runtime call <call> deduplicated. [OMP170]
|
|
====================================================================
|
|
|
|
This optimization remark indicates that a call to an OpenMP runtime call was
|
|
replaced with the result of an existing one. This occurs when the compiler knows
|
|
that the result of a runtime call is immutable. Removing duplicate calls is done
|
|
by replacing all calls to that function with the result of the first call. This
|
|
cannot be done automatically by the compiler because the implementations of the
|
|
OpenMP runtime calls live in a separate library the compiler cannot see.
|
|
|
|
Example
|
|
-------
|
|
|
|
This optimization will trigger for known OpenMP runtime calls whose return value
|
|
will not change.
|
|
|
|
.. code-block:: c++
|
|
|
|
void foo(int N) {
|
|
double *A = malloc(N * omp_get_thread_limit());
|
|
double *B = malloc(N * omp_get_thread_limit());
|
|
|
|
#pragma omp parallel
|
|
work(&A[omp_get_thread_num() * N]);
|
|
#pragma omp parallel
|
|
work(&B[omp_get_thread_num() * N]);
|
|
}
|
|
|
|
.. code-block:: console
|
|
|
|
$ clang -fopenmp -O2 -Rpass=openmp-opt omp170.c
|
|
ompi170.c:2:26: remark: OpenMP runtime call omp_get_thread_limit deduplicated. [OMP170]
|
|
double *A = malloc(N * omp_get_thread_limit());
|
|
^
|
|
|
|
Diagnostic Scope
|
|
----------------
|
|
|
|
OpenMP optimization remark.
|