WiFi still works now after light-sleep (w/o powering down modem)

[bjoernQ]

This at least partially addresses #3751 (not sure if we should close that and open new issues instead) - might also address #3873 (at least it seems like it does)

ESP32 seems to have a few more problems - e.g. the UART-baudrate is none of the common baud-rates (and definitely not the default 115200) after returning from light-sleep

Testing:

Add this e.g. after init

    // Start adding sleep code STOPSHIP
    let sleep_config = esp_hal::rtc_cntl::sleep::RtcSleepConfig::default();

    let delay = esp_hal::delay::Delay::new();
    delay.delay_millis(100);

    let timer =
        esp_hal::rtc_cntl::sleep::TimerWakeupSource::new(core::time::Duration::from_secs(5));

    let mut rtc = esp_hal::rtc_cntl::Rtc::new(peripherals.LPWR);
    esp_println::println!("sleep");
    delay.delay_millis(100);

    delay.delay_millis(100);

    rtc.sleep(&sleep_config, &[&timer]);
    delay.delay_millis(100);

With this change WiFi etc. works after a light-sleep w/o powering down the modem (or PLL) - for that I think we probably need to do more (e.g. it seems there are rom functions used as callbacks for enter/exit light-sleep - and more)

[bjoernQ]

Looking at all the other chips this is a better default

[bugadani]

Do we really need to keep the main PLL running in sleep? I'd like to measure the change in power consumption after this.

[bjoernQ]

I'd like to measure the change in power consumption after this

Sure - that's very welcome - IIRC it wasn't much difference but I might be wrong. In any case a user can always opt-out (and end up with non-working radio after the light sleep)

[bjoernQ]

revert the base-settings

[bugadani]

Can we have a QA test for this? I don't agree with these changes but if they work, we need a way to easily see we don't break wifi.