refactor(etm): systimer etm test case

previously we use the systimer event to toggling a GPIO and assert the GPIO level,
which is not stable in CI.
Now we swicth to clear the GPIO at the systimer event.
This commit is contained in:
morris 2023-09-20 11:31:21 +08:00
parent 25f031c6e1
commit 1c15c2aded

View File

@ -79,7 +79,7 @@ TEST_CASE("esp_timer_etm_event", "[etm]")
printf("allocate GPIO etm task\r\n");
esp_etm_task_handle_t gpio_task = NULL;
gpio_etm_task_config_t gpio_task_config = {
.action = GPIO_ETM_TASK_ACTION_TOG,
.action = GPIO_ETM_TASK_ACTION_CLR,
};
TEST_ESP_OK(gpio_new_etm_task(&gpio_task_config, &gpio_task));
@ -114,12 +114,14 @@ TEST_CASE("esp_timer_etm_event", "[etm]")
TEST_ESP_OK(esp_timer_create(&periodic_timer_args, &periodic_timer));
TEST_ESP_OK(esp_timer_start_periodic(periodic_timer, 500000));
// should see a 1Hz square wave on the GPIO
esp_rom_delay_us(1200000);
// check the final GPIO level
vTaskDelay(pdMS_TO_TICKS(100));
// etm event not active yet, so the GPIO level should still be the initial state
TEST_ASSERT_EQUAL(1, gpio_get_level(output_gpio));
vTaskDelay(pdMS_TO_TICKS(1000));
// etm event should have cleared the GPIO level
TEST_ASSERT_EQUAL(0, gpio_get_level(output_gpio));
TEST_ESP_OK(esp_timer_stop(periodic_timer));
TEST_ESP_OK(esp_timer_delete(periodic_timer));