/**************************************************************************** * examples/calib_udelay/calib_udelay_main.c * * Copyright (C) 2017 Haltian Ltd All rights reserved. * Author: Jussi Kivilinna * * Redistribution and use in source and binary forms, with or without * modification, are permitted provided that the following conditions * are met: * * 1. Redistributions of source code must retain the above copyright * notice, this list of conditions and the following disclaimer. * 2. Redistributions in binary form must reproduce the above copyright * notice, this list of conditions and the following disclaimer in * the documentation and/or other materials provided with the * distribution. * 3. Neither the name NuttX nor the names of its contributors may be * used to endorse or promote products derived from this software * without specific prior written permission. * * THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS * "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT * LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS * FOR A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE * COPYRIGHT OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, * INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, * BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS * OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED * AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT * LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN * ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE * POSSIBILITY OF SUCH DAMAGE. * ****************************************************************************/ /**************************************************************************** * Included Files ****************************************************************************/ #include #include #include #include #include #include #include /**************************************************************************** * Pre-processor Definitions ****************************************************************************/ #ifndef CONFIG_EXAMPLES_CALIB_UDELAY_NUM_MEASUREMENTS # define CONFIG_EXAMPLES_CALIB_UDELAY_NUM_MEASUREMENTS 3 #endif #ifndef CONFIG_EXAMPLES_CALIB_UDELAY_NUM_RESULTS # define CONFIG_EXAMPLES_CALIB_UDELAY_NUM_RESULTS 20 #endif #define DELAY_TEST_ITERS 100000 /**************************************************************************** * Private Types ****************************************************************************/ struct measurement_s { int count; uint64_t nsecs; }; /**************************************************************************** * Private Functions ****************************************************************************/ static uint64_t gettime_nsecs(void) { struct timespec ts; uint64_t nsecs; clock_gettime(CLOCK_MONOTONIC, &ts); nsecs = ts.tv_sec; nsecs *= 1000 * 1000 * 1000; nsecs += ts.tv_nsec; return nsecs; } static int compare_measurements(const void *va, const void *vb) { const struct measurement_s *a = va; const struct measurement_s *b = vb; if (a->nsecs == b->nsecs) { return 0; } else if (a->nsecs < b->nsecs) { return -1; } else { return 1; } } static void __attribute__((noinline)) calib_udelay_test(uint32_t count) { volatile int i; while (count > 0) { for (i = 0; i < DELAY_TEST_ITERS; i++) { } count--; } } static void perform_measurements(int loop_count, FAR struct measurement_s *measurements, int num_measurements) { int n; memset(measurements, 0, sizeof(*measurements) * num_measurements); for (n = 0; n < num_measurements; n++) { measurements[n].nsecs = gettime_nsecs(); calib_udelay_test(loop_count); measurements[n].nsecs = gettime_nsecs() - measurements[n].nsecs; measurements[n].count = loop_count; } qsort(measurements, num_measurements, sizeof(measurements[0]), compare_measurements); } static double getx(FAR struct measurement_s *point) { return point->count * (double)DELAY_TEST_ITERS; } static double gety(struct measurement_s *point) { return point->nsecs; } static int linreg(FAR struct measurement_s *point, int num_points, FAR double *m, FAR double *b, FAR double *r2) { double sumx = 0.0; double sumx2 = 0.0; double sumxy = 0.0; double sumy = 0.0; double sumy2 = 0.0; double x; double y; double denom; double inv_denom; int i; for (i = 0; i < num_points; i++) { x = getx(point + i); y = gety(point + i); sumx += x; sumx2 += x * x; sumxy += x * y; sumy += y; sumy2 += y * y; } denom = num_points * sumx2 - sumx * sumx; if (denom == 0) { *m = 0; *b = 0; if (r2) { *r2 = 0; } return ERROR; } inv_denom = 1.0 / denom; *m = ((num_points * sumxy) - (sumx * sumy)) * inv_denom; *b = ((sumy * sumx2) - (sumx * sumxy)) * inv_denom; if (r2) { double term1 = ((num_points * sumxy) - (sumx * sumy)); double term2 = ((num_points * sumx2) - (sumx * sumx)); double term3 = ((num_points * sumy2) - (sumy * sumy)); double term23 = (term2 * term3); *r2 = 1.0; if (fabs(term23) > 1e-10) { *r2 = (term1 * term1) / term23; } } return OK; } /**************************************************************************** * Public Functions ****************************************************************************/ /**************************************************************************** * calib_udelay_main ****************************************************************************/ int main(int argc, FAR char *argv[]) { const int num_measurements = CONFIG_EXAMPLES_CALIB_UDELAY_NUM_MEASUREMENTS; const int num_results = CONFIG_EXAMPLES_CALIB_UDELAY_NUM_RESULTS; const int min_timer_resolution_steps = 3; const int calibration_step_multiplier = 10; struct measurement_s measurements[num_measurements]; struct measurement_s result; struct measurement_s results[num_results]; double iters_per_nsec; double iters_per_msec; uint64_t timer_resolution; double slope_m, slope_b, slope_r2; int min_step; int loop_count; double duration; int i; printf("\n"); sched_lock(); printf("Calibrating timer for main calibration...\n"); usleep(200 * 1000); /* Find out timer resolution. */ loop_count = 0; do { loop_count++; perform_measurements(loop_count, measurements, num_measurements); result = measurements[0]; } while (result.nsecs == 0); timer_resolution = result.nsecs; /* Find first loop count where timer steps five times. */ do { loop_count++; perform_measurements(loop_count, measurements, num_measurements); result = measurements[0]; } while (result.nsecs < timer_resolution * min_timer_resolution_steps); min_step = result.count; /* Get calibration slope for loop function. */ for (duration = 0, loop_count = min_step, i = 0; i < num_results; i++, loop_count += min_step * calibration_step_multiplier) { duration += (double)num_measurements * loop_count * timer_resolution * min_timer_resolution_steps / min_step; } printf("Performing main calibration for udelay. This will take approx. %.3f seconds.\n", duration * 1e-9); usleep(200 * 1000); for (loop_count = min_step, i = 0; i < num_results; i++, loop_count += min_step * calibration_step_multiplier) { perform_measurements(loop_count, measurements, num_measurements); results[i] = measurements[0]; } if (linreg(results, num_results, &slope_m, &slope_b, &slope_r2) == OK) { printf("Calibration slope for udelay:\n" " Y = m*X + b, where\n" " X is loop iterations,\n" " Y is time in nanoseconds,\n" " b is base overhead,\n" " m is nanoseconds per loop iteration.\n\n" " m = %.08f nsec/iter\n" " b = %.08f nsec\n\n" " Correlation coefficient, R² = %.4f\n\n", slope_m, slope_b, slope_r2); iters_per_nsec = (1.0 / slope_m); iters_per_msec = iters_per_nsec * 1000 * 1000; printf("Without overhead, %.8f iterations per nanosecond and %.2f " "iterations per millisecond.\n\n", iters_per_nsec, iters_per_msec); printf("Recommended setting for CONFIG_BOARD_LOOPSPERMSEC:\n" " CONFIG_BOARD_LOOPSPERMSEC=%.0f\n", ceil(iters_per_msec)); } else { printf("cannot solve\n"); } sched_unlock(); return 0; }