Preparing test case for GPU

llvm/test/VISC/gemm_opencl/matrixMul/Makefile

@@ -27,4 +27,4 @@ $(HOST:%=%.bin):%.bin:%.c
 	$(LLVM_CC) -O3 -lOpenCL -I /usr/local/cuda/include $< -o $@
 
 clean :
-	rm -f *.ll *.bc *.s *.bin
+	rm -f $(HOST).ll $(KERNELS).ll *.bc *.s *.bin

llvm/test/VISC/gemm_opencl/matrixMul/gemm_opencl.c

@@ -109,52 +109,9 @@ int checkResults(float* A, float* B, float* C) {
   return 1; // Success
 }
 
-// Main
-int main(int argc, char** argv) {
-
-  // seed for rand()
-  srand(2006);
-
-  // Allocate host memory for matrices A and B
-  unsigned int size_A = WA * HA;
-  unsigned int bytes_A = sizeof(float) * size_A;
-  float* h_A = (float*) malloc(bytes_A);
-
-  unsigned int size_B = WB * HB;
-  unsigned int bytes_B = sizeof(float) * size_B;
-  float* h_B = (float*) malloc(bytes_B);
-
-   // Initialize host memory
-   randomInit(h_A, size_A);
-   randomInit(h_B, size_B);
-
-/*
-   // Print A and B
-   printf("\n\nMatrix A\n");
-   for(int i = 0; i < size_A; i++)
-   {
-      printf("%f ", h_A[i]);
-      if(((i + 1) % WA) == 0)
-      printf("\n");
-   }
-
-   printf("\n\nMatrix B\n");
-   for(int i = 0; i < size_B; i++)
-   {
-      printf("%f ", h_B[i]);
-      if(((i + 1) % WB) == 0)
-      printf("\n");
-   }
-*/
-
-  // Allocate host memory for the result matrix C
-  unsigned int size_C = WC * HC;
-  unsigned int bytes_C = sizeof(float) * size_C;
-  float* h_C = (float*) malloc(bytes_C);
-
-   // Initialize OpenCL
-
-   // OpenCL specific variables
+// GPU Computation of MatrixMul
+void computeMatrixMul(float* h_A, unsigned bytes_A, float* h_B, unsigned bytes_B, float* h_C, unsigned bytes_C) {
+ // OpenCL specific variables
   cl_context clGPUContext;
   cl_command_queue clCommandQue;
   cl_program clProgram;
@@ -291,16 +248,7 @@ int main(int argc, char** argv) {
   }
   printf("\n");
  */
-  if(checkResults(h_A, h_B, h_C))
-    printf("\nPass!\n");
-  else
-    printf("\nFailed!\n");
-  printf("\nDone!\n");
-
   // Deallocate memory
-  free(h_A);
-  free(h_B);
-  free(h_C);
 
   clReleaseMemObject(d_A);
   clReleaseMemObject(d_C);
@@ -315,3 +263,61 @@ int main(int argc, char** argv) {
 
 }
 
+// Main
+int main(int argc, char** argv) {
+
+  // seed for rand()
+  srand(2006);
+
+  // Allocate host memory for matrices A and B
+  unsigned int size_A = WA * HA;
+  unsigned int bytes_A = sizeof(float) * size_A;
+  float* h_A = (float*) malloc(bytes_A);
+
+  unsigned int size_B = WB * HB;
+  unsigned int bytes_B = sizeof(float) * size_B;
+  float* h_B = (float*) malloc(bytes_B);
+
+   // Initialize host memory
+   randomInit(h_A, size_A);
+   randomInit(h_B, size_B);
+
+/*
+   // Print A and B
+   printf("\n\nMatrix A\n");
+   for(int i = 0; i < size_A; i++)
+   {
+      printf("%f ", h_A[i]);
+      if(((i + 1) % WA) == 0)
+      printf("\n");
+   }
+
+   printf("\n\nMatrix B\n");
+   for(int i = 0; i < size_B; i++)
+   {
+      printf("%f ", h_B[i]);
+      if(((i + 1) % WB) == 0)
+      printf("\n");
+   }
+*/
+
+  // Allocate host memory for the result matrix C
+  unsigned int size_C = WC * HC;
+  unsigned int bytes_C = sizeof(float) * size_C;
+  float* h_C = (float*) malloc(bytes_C);
+
+   // Compute using OpenCL
+  computeMatrixMul(h_A, bytes_A, h_B, bytes_B, h_C, bytes_C);
+
+  if(checkResults(h_A, h_B, h_C))
+    printf("\nPass!\n");
+  else
+    printf("\nFailed!\n");
+  printf("\nDone!\n");
+
+  // Deallocate memory
+  free(h_A);
+  free(h_B);
+  free(h_C);
+}
+