Update XC app to show performance of library

Useful to see when it's worth parallelising matrix ops.
Will help guide optimisation efforts.

app_matrix_c_demo/src/main.c

@@ -40,6 +40,6 @@ int main(void)
 	MATRIX_PRINT("A",A);
 	MATRIX_PRINT("B",B);
 	MATRIX_PRINT("C",C);
-	printf("Ops: %u\n",ops);
+	printf("MATRIX_NTHREADS: %u, Ops: %u\n",MATRIX_NTHREADS,ops);
 	return 0;
 }

app_matrix_xc_demo/src/main.xc

@@ -11,8 +11,11 @@
 #include <stdio.h>
 #include "matrix.h"
 
+void perfcheck(void);
+
 int main(void)
 {
+#ifndef SKIP_BASIC_DEMO
 	MATRIX_CREATE(A,8,8,	1, 4, 3, 1, 5, 1, 2, 1,
 				1, 0, 3, 1, 3, 3, 3, 5,
 				4, 4, 4, 0, 2, 2, 2, 3,
@@ -42,11 +45,62 @@ int main(void)
 	ops += matrix_sca_op(MUL,MATRIX_REF(B),1798,MATRIX_NULL(),0);
 	/* Matrix-multiplication into destination matrix */
 	ops += matrix_mul(MATRIX_REF(A),MATRIX_REF(B),MATRIX_REF(C),0);
-	ops += matrix_arr_op(ADD,MATRIX_REF(A),MATRIX_REF(B),MATRIX_NULL(),0);
 	t :> tvB;
 	MATRIX_PRINT("A",A);
 	MATRIX_PRINT("B",B);
 	MATRIX_PRINT("C",C);
 	printf("Ticks: %u, Ops: %u\n",tvB - tvA, ops);
+#endif	
+	/* That was a basic demo. Now let's compare performance */
+	perfcheck();
 	return 0;
 }
+
+void perfcheck(void)
+{
+	/* The greatest irony of the performance check is that it doesn't
+	 reuse memory or resources from the basic demo. Good job it's only
+	 a demonstration of computational speed, then. */
+	MATRIX_CREATE(A,16,16,	1, 4, 3, 1, 5, 1, 2, 1, 1, 4, 3, 1, 5, 1, 2, 1,
+				1, 0, 3, 1, 3, 3, 3, 5, 1, 0, 3, 1, 3, 3, 3, 5, 
+				4, 4, 4, 0, 2, 2, 2, 3, 4, 4, 4, 0, 2, 2, 2, 3, 
+				3, 4, 3, 2, 2, 1, 0, 2, 3, 4, 3, 2, 2, 1, 0, 2, 
+				2, 3, 3, 2, 2, 0, 1, 4, 2, 3, 3, 2, 2, 0, 1, 4, 
+				4, 2, 2, 2, 4, 4, 3, 0, 4, 2, 2, 2, 4, 4, 3, 0, 
+				3, 4, 2, 5, 0, 4, 2, 5, 3, 4, 2, 5, 0, 4, 2, 5, 
+				4, 2, 1, 0, 1, 2, 4, 0, 4, 2, 1, 0, 1, 2, 4, 0,
+				1, 4, 3, 1, 5, 1, 2, 1, 1, 4, 3, 1, 5, 1, 2, 1,
+				1, 0, 3, 1, 3, 3, 3, 5, 1, 0, 3, 1, 3, 3, 3, 5, 
+				4, 4, 4, 0, 2, 2, 2, 3, 4, 4, 4, 0, 2, 2, 2, 3, 
+				3, 4, 3, 2, 2, 1, 0, 2, 3, 4, 3, 2, 2, 1, 0, 2, 
+				2, 3, 3, 2, 2, 0, 1, 4, 2, 3, 3, 2, 2, 0, 1, 4, 
+				4, 2, 2, 2, 4, 4, 3, 0, 4, 2, 2, 2, 4, 4, 3, 0, 
+				3, 4, 2, 5, 0, 4, 2, 5, 3, 4, 2, 5, 0, 4, 2, 5, 
+				4, 2, 1, 0, 1, 2, 4, 0, 4, 2, 1, 0, 1, 2, 4, 0 );
+	MATRIX_CREATE(B,16,16,0);
+	timer t;
+	unsigned int tvA, tvB;
+	int r,c;
+	printf("MATRIX_NTHREADS: %u\n",MATRIX_NTHREADS);
+	for (int i = 1, ops = 0; i <= 16; i++, ops = 0)
+	{
+		MATRIX_REDIM(A,i,i);
+		printf("\nMatrix size: %dx%d\n",i,i);
+		/* Scalar multiplication by 2 */
+		t :> tvA;
+		for (r = 0; r < MATRIX_ROWS(A); r += 1)
+		{
+			for (c = 0; c < MATRIX_COLS(A); c += 1)
+			{
+				B[r * c] = A[r * c] * 2;
+				ops += 1;
+			}
+		}
+		t :> tvB;
+		printf("By-hand scalar * 2 :: Ticks: %u, Ops: %u\n",tvB - tvA, ops);
+		t :> tvA;
+		ops = matrix_sca_op(MUL,MATRIX_REF(A),2,MATRIX_REF(B),0);
+		t :> tvB;
+		printf("sc_matrix scalar * 2 :: Ticks: %u, Ops: %u\n",tvB - tvA, ops);
+	}
+}

module_matrix/src/matrix.xc

@@ -15,7 +15,7 @@
 
 int matrix_redim(short dims[4],short rows, short columns)
 {
-	if (dims[2] > rows || dims[3] > columns)
+	if (dims[2] < rows || dims[3] < columns)
 	{
 		return -1; //Larger than initial size
 	}

module_matrix/src/matrix_worker.c

@@ -9,8 +9,6 @@
  */
 
 #include "matrix_worker.h"
-
-
 #include <stdio.h>
 
 void matrix_mul_worker(int ptA, int ptDimA, int ptB, int ptDimB, int ptC,