On Thu, 12 Sep 2024, Reinette Chatre wrote:
> The resctrl selftests drop the results from every first test run
> to avoid (per comment) "inaccurate due to monitoring setup transition
> phase" data. Previously inaccurate data resulted from workloads needing
> some time to "settle" and also the measurements themselves to
> account for earlier measurements to measure across needed timeframe.
>
> commit da50de0a92f3 ("selftests/resctrl: Calculate resctrl FS derived mem
> bw over sleep(1) only")
>
> ensured that measurements accurately measure just the time frame of
> interest. The default "fill_buf" benchmark since separated the buffer
> prepare phase from the benchmark run phase reducing the need for the
> tests themselves to accommodate the benchmark's "settle" time.
>
> With these enhancements there are no remaining portions needing
> to "settle" and the first test run can contribute to measurements.
>
> Signed-off-by: Reinette Chatre <reinette.cha...@intel.com>
> ---
> Changes since V1:
> - Remove comment about needing results from first run removed.
> - Fix existing incorrect spacing while changing line.
> ---
> tools/testing/selftests/resctrl/cmt_test.c | 5 ++---
> tools/testing/selftests/resctrl/mba_test.c | 10 +++-------
> tools/testing/selftests/resctrl/mbm_test.c | 10 +++-------
> 3 files changed, 8 insertions(+), 17 deletions(-)
>
> diff --git a/tools/testing/selftests/resctrl/cmt_test.c
> b/tools/testing/selftests/resctrl/cmt_test.c
> index a7effe76b419..d4b85d144985 100644
> --- a/tools/testing/selftests/resctrl/cmt_test.c
> +++ b/tools/testing/selftests/resctrl/cmt_test.c
> @@ -99,14 +99,13 @@ static int check_results(struct resctrl_val_param *param,
> size_t span, int no_of
> }
>
> /* Field 3 is llc occ resc value */
> - if (runs > 0)
> - sum_llc_occu_resc += strtoul(token_array[3], NULL, 0);
> + sum_llc_occu_resc += strtoul(token_array[3], NULL, 0);
> runs++;
> }
> fclose(fp);
>
> return show_results_info(sum_llc_occu_resc, no_of_bits, span,
> - MAX_DIFF, MAX_DIFF_PERCENT, runs - 1, true);
> + MAX_DIFF, MAX_DIFF_PERCENT, runs, true);
> }
>
> static void cmt_test_cleanup(void)
> diff --git a/tools/testing/selftests/resctrl/mba_test.c
> b/tools/testing/selftests/resctrl/mba_test.c
> index 5c6063d0a77c..89c2446b9f80 100644
> --- a/tools/testing/selftests/resctrl/mba_test.c
> +++ b/tools/testing/selftests/resctrl/mba_test.c
> @@ -86,18 +86,14 @@ static bool show_mba_info(unsigned long *bw_imc, unsigned
> long *bw_resc)
> int avg_diff_per;
> float avg_diff;
>
> - /*
> - * The first run is discarded due to inaccurate value from
> - * phase transition.
> - */
> - for (runs = NUM_OF_RUNS * allocation + 1;
> + for (runs = NUM_OF_RUNS * allocation;
> runs < NUM_OF_RUNS * allocation + NUM_OF_RUNS ; runs++) {
> sum_bw_imc += bw_imc[runs];
> sum_bw_resc += bw_resc[runs];
> }
>
> - avg_bw_imc = sum_bw_imc / (NUM_OF_RUNS - 1);
> - avg_bw_resc = sum_bw_resc / (NUM_OF_RUNS - 1);
> + avg_bw_imc = sum_bw_imc / NUM_OF_RUNS;
> + avg_bw_resc = sum_bw_resc / NUM_OF_RUNS;
> if (avg_bw_imc < THROTTLE_THRESHOLD || avg_bw_resc <
> THROTTLE_THRESHOLD) {
> ksft_print_msg("Bandwidth below threshold (%d MiB).
> Dropping results from MBA schemata %u.\n",
> THROTTLE_THRESHOLD,
> diff --git a/tools/testing/selftests/resctrl/mbm_test.c
> b/tools/testing/selftests/resctrl/mbm_test.c
> index 7635ee6b9339..8c818e292dce 100644
> --- a/tools/testing/selftests/resctrl/mbm_test.c
> +++ b/tools/testing/selftests/resctrl/mbm_test.c
> @@ -22,17 +22,13 @@ show_bw_info(unsigned long *bw_imc, unsigned long
> *bw_resc, size_t span)
> int runs, ret, avg_diff_per;
> float avg_diff = 0;
>
> - /*
> - * Discard the first value which is inaccurate due to monitoring setup
> - * transition phase.
> - */
> - for (runs = 1; runs < NUM_OF_RUNS ; runs++) {
> + for (runs = 0; runs < NUM_OF_RUNS; runs++) {
> sum_bw_imc += bw_imc[runs];
> sum_bw_resc += bw_resc[runs];
> }
>
> - avg_bw_imc = sum_bw_imc / 4;
> - avg_bw_resc = sum_bw_resc / 4;
> + avg_bw_imc = sum_bw_imc / NUM_OF_RUNS;
> + avg_bw_resc = sum_bw_resc / NUM_OF_RUNS;
> avg_diff = (float)labs(avg_bw_resc - avg_bw_imc) / avg_bw_imc;
> avg_diff_per = (int)(avg_diff * 100);
While the patch itself is fine, I notice the code has this magic number
gem too:
unsigned long bw_imc[1024], bw_resc[1024];
Reviewed-by: Ilpo Järvinen <ilpo.jarvi...@linux.intel.com>
--
i.
