Changes between Version 14 and Version 15 of Next-GenOpenMPTransformation

Timestamp:

05/10/19 12:24:46 (7 years ago)

Author:

ziqing

Comment:

--

Next-GenOpenMPTransformation

@@ -112 +112 @@
 == OpenMP Simplifier ==
 * We improve the existing OpenMP simplifier with the informations provided by the [[wiki: StaticAnalysis| Static Analysis]].
+* Example 1: (`DRB067-restrictpointer1-orig-no.c` from [https://github.com/LLNL/dataracebench/releases/tag/v1.2.0 DataRaceBench v1.2.0])
+{{{
+#include <stdlib.h>
+typedef double real8;
+
+void foo(real8 * restrict newSxx, real8 * restrict newSyy, int length)
+{
+  int i;
+
+#pragma omp parallel for private (i) firstprivate (length)
+  for (i = 0; i <= length - 1; i += 1) {
+    newSxx[i] = 0.0;
+    newSyy[i] = 0.0;
+  }
+}
+
+int main()
+{
+  int length=10;
+  real8* newSxx = malloc (length* sizeof (real8));
+  real8* newSyy = malloc (length* sizeof (real8));
+
+  foo(newSxx, newSyy, length);
+
+  free (newSxx);
+  free (newSyy);
+  return 0;
+}
+}}}
+  * The OpenMP simplifier analyzes the parallel region with points-to informations about the two pointer argument `newSxx` and `newSyy` from static analysis.
+  * The OpenMP simplifier can determine that no data race will happen in the parallel region as long as no array out-of-bound error happens in it.
+  * The OpenMP simplifier sequentializes the program which will be checked by CIVL.  Any possible array out-of-bound error will be caught by CIVL.