Add OMP static scheduling throughout, don't depend on heuristics

src/daxpy.c

@@ -41,7 +41,7 @@ double daxpy(
 #pragma omp parallel
     {
         LIKWID_MARKER_START("DAXPY");
-#pragma omp for
+#pragma omp for schedule(static)
         for (int i=0; i<N; i++) {
             a[i] = a[i] + scalar * b[i];
         }