Complete examples for parallel for with and without nounroll.

Author: ludgerpaehler

P5_omp_parallel_for/omp_parallel_for.c

@@ -0,0 +1,38 @@
+#include <omp.h>
+
+
+// Do parallel for 
+void omp(float *x, int npoints) {
+
+#pragma omp parallel for
+    for (int i = 0; i < npoints; i++) {
+        x[i] *= x[i];
+    }
+}
+
+
+double __enzyme_autodiff(void*, ...);
+
+int main() {
+
+    // Initialize array
+    float array[1000];
+    for(int i=0, i<1000; i++) {
+        array[i] = i + 0.5;
+    }
+
+    // Set up the array to host the gradients
+    float d_array[1000];
+    for(int i=0, i<1000; i++) {
+        d_array[i] = 1.0f;
+    }
+
+    // Alter the entries
+#ifdef FORWARD
+    sub(array, 1000);
+#else
+    __enzyme_autodiff((void*)omp, array, d_array, 1000);
+#endif
+
+    return 0;
+}

P5_omp_parallel_for/omp_parallel_for_firstprivate.c

@@ -0,0 +1,39 @@
+#include <omp.h>
+
+
+// Do parallel for 
+void omp(float *x, int npoints) {
+
+#pragma omp parallel for
+#pragma nounroll
+    for (int i = 0; i < npoints; i++) {
+        x[i] *= x[i];
+    }
+}
+
+
+double __enzyme_autodiff(void*, ...);
+
+int main() {
+
+    // Initialize array
+    float array[1000];
+    for(int i=0, i<1000; i++) {
+        array[i] = i + 0.5;
+    }
+
+    // Set up the array to host the gradients
+    float d_array[1000];
+    for(int i=0, i<1000; i++) {
+        d_array[i] = 1.0f;
+    }
+
+    // Alter the entries
+#ifdef FORWARD
+    sub(array, 1000);
+#else
+    __enzyme_autodiff((void*)omp, array, d_array, 1000);
+#endif
+
+    return 0;
+}