To monitor OpenMP runtime library routine calls, the transformation process replaces these calls by calls to the performance tool interface library. For example, a OpenMP API call to omp_set_lock() is transformed into a call to pomp_set_lock(). In the implementation of the performance interface function, the original corresponding OpenMP runtime library routine must be called. Performance data can be obtained before and after. Currently, we think it is sufficient to use this procedure for the omp_###_lock() and omp_###_nest_lock() routines because they are most relevant for the observation of OpenMP performance behavior.