Re: [LTP] [PATCH 10/16] [Realtime] matrix_mult cleanups

[Darren H. <dv...@li...>]

On Thursday 10 January 2008 03:12:30 Sebastien Dugue wrote:
>   Perform some cleanups on matrix_mult:
>
>   - remove unused ops local from matrix_mult().
>   - remove unused timing calculation from matrix_mult().
>   - streamline parenthesis use in matrix_mult().
>   - cleanup stray '\' in the stats_container_save() calls.
>   - remove one commented out join_threads().
>
> Signed-off-by: Sebastien Dugue <seb...@bu...>

Acked-by: Darren Hart <dv...@us...>

> Cc: Darren Hart <dv...@us...>
> Cc: Tim Chavez <ti...@us...>
> ---
>  testcases/realtime/func/matrix_mult/matrix_mult.c |   32
> +++++++------------- 1 files changed, 11 insertions(+), 21
> deletions(-)
>
> diff --git a/testcases/realtime/func/matrix_mult/matrix_mult.c
> b/testcases/realtime/func/matrix_mult/matrix_mult.c index
> 7c3c6b6..5577a51 100644
> --- a/testcases/realtime/func/matrix_mult/matrix_mult.c
> +++ b/testcases/realtime/func/matrix_mult/matrix_mult.c
> @@ -101,24 +101,16 @@ void matrix_mult(void)
>  	float B[MATRIX_SIZE][MATRIX_SIZE];
>  	float C[MATRIX_SIZE][MATRIX_SIZE];
>  	int i, j, k;
> -	nsec_t start, end, delta;
> -	int ops;
>
>  	matrix_init(A, B);
> -	start = rt_gettime();
> -	ops = 0;
> -	for (i = 0; i < MATRIX_SIZE; i++)
> -	{
> -		for (j = 0; j < MATRIX_SIZE; j++)
> -		{
> +
> +	for (i = 0; i < MATRIX_SIZE; i++) {
> +		for (j = 0; j < MATRIX_SIZE; j++) {
>  			for (k = 0; k < MATRIX_SIZE; k++) {
>  				C[i][j] += A[i][k]*B[k][j];
> -				ops++;
>  			}
>  		}
>  	}
> -	end = rt_gettime();
> -	delta = end - start;
>  }
>
>  /* arg: the number of concurrent threads being run */
> @@ -201,10 +193,10 @@ int main(int argc, char *argv[])
>  	printf("StdDev: %.4f us\n", stats_stddev(&sdat));
>  	stats_hist(&shist, &sdat);
>
> -	stats_container_save("sequential", "Matrix Multiplication
> Sequential Execution Runtime Scatter Plot",\ -				"Iteration",
> "Runtime (us)", &sdat, "points");
> -	stats_container_save("sequential_hist", "Matrix Multiplicatoin
> Sequential Execution Runtime Histogram",\ -				"Iteration", "Runtime
> (us)", &shist, "steps");
> +	stats_container_save("sequential", "Matrix Multiplication
> Sequential Execution Runtime Scatter Plot", +			     "Iteration",
> "Runtime (us)", &sdat, "points");
> +	stats_container_save("sequential_hist", "Matrix Multiplicatoin
> Sequential Execution Runtime Histogram", +			     "Iteration",
> "Runtime (us)", &shist, "steps");
>
>  	// run matrix mult operation concurrently
>  	printf("\nConcurrent (%dx):\n", numcpus);
> @@ -244,18 +236,16 @@ int main(int argc, char *argv[])
>  	printf("StdDev: %.4f us\n", stats_stddev(&cdat));
>  	stats_hist(&chist, &cdat);
>
> -	stats_container_save("concurrent", "Matrix Multiplication
> Concurrent Execution Runtime Scatter Plot",\ -				"Iteration",
> "Runtime (us)", &cdat, "points");
> -	stats_container_save("concurrent_hist", "Matrix Multiplication
> Concurrent Execution Runtime Histogram",\ -				"Iteration", "Runtime
> (us)", &chist, "steps");
> +	stats_container_save("concurrent", "Matrix Multiplication
> Concurrent Execution Runtime Scatter Plot", +			     "Iteration",
> "Runtime (us)", &cdat, "points");
> +	stats_container_save("concurrent_hist", "Matrix Multiplication
> Concurrent Execution Runtime Histogram", +			     "Iteration",
> "Runtime (us)", &chist, "steps");
>
>  	printf("\nSeq/Conc Ratios:\n");
>  	printf("Min: %.4f\n", (float)smin/cmin);
>  	printf("Max: %.4f\n", (float)smax/cmax);
>  	printf("Avg: %.4f\n", (float)savg/cavg);
>
> -	//join_threads();
> -
>  	ret = 1;
>  	if (savg > (cavg * criteria))
>  		ret = 0;



-- 
Darren Hart
IBM Linux Technology Center
Real-Time Linux Team