tests: pwm_api: Add timing check

[raffarost]

Use GPIO interrupts to check PWM output timing.

Depends on #80872

[henrikbrixandersen]

This is what https://github.com/zephyrproject-rtos/zephyr/tree/main/tests/drivers/pwm/pwm_loopback is for?

[raffarost]

This is what https://github.com/zephyrproject-rtos/zephyr/tree/main/tests/drivers/pwm/pwm_loopback is for?

Hi @henrikbrixandersen,
I understand pwm_loopback requires the peripheral to support CC, which is not the case for all drivers (e.g ESP's LEDC pwm).

[henrikbrixandersen]

This is what main/tests/drivers/pwm/pwm_loopback is for?

Hi @henrikbrixandersen, I understand pwm_loopback requires the peripheral to support CC, which is not the case for all drivers (e.g ESP's LEDC pwm).

Does the chip have another PWM-capture capable peripheral? Producer and consumer does not need to be of the same peripheral type in pwm_loopback.

I'm okay with introducing a GPIO-based PWM loopback test, but I don't think it should be in the main PWM API test suite like this (which is also why I placed the initial loopback test in its own test suite).

[raffarost]

This is what main/tests/drivers/pwm/pwm_loopback is for?

Hi @henrikbrixandersen, I understand pwm_loopback requires the peripheral to support CC, which is not the case for all drivers (e.g ESP's LEDC pwm).

Does the chip have another PWM-capture capable peripheral? Producer and consumer does not need to be of the same peripheral type in pwm_loopback.

I'm okay with introducing a GPIO-based PWM loopback test, but I don't think it should be in the main PWM API test suite like this (which is also why I placed the initial loopback test in its own test suite).

Some Espressif chips only have LEDC and no MCPWM, thus no CC available. The test was meant to be a simple timing check, because we've had some regressions after HAL updates. I can create a separate test suite for it. I'd just need to think how to avoid repeating code, as the 'configs', 'sets' and test sequence would be the same.

[raffarost]

@henrikbrixandersen I created a PR for the new implementation (#82258)

[bjarki-andreasen]

should this one be closed in favor of #82258 ?