QUERY · ISSUE

ESP32 timer wake up repeating

openby AndreoBotelhoopened 2023-10-03updated 2023-10-24

bugport-esp32

I'm using timer wakeup from light sleep with:

machine.lightsleep(30000) #sleep 30s

after 30 the esp32 wakes, disable peripherals and sleep indefinitely with:

machine.lightsleep()

but the 30s wake timer keeps repeating and didn't find any way to disable wake source

CANDIDATE · PULL REQUEST

esp32 : fix for #7158 - machine.deepsleep() not behaving as expected

mergedby PBrunotopened 2023-10-23updated 2023-10-31

port-esp32

This PR fixes machine.deepsleep() waking up the board after a delay d, for example when a machine.lightsleep(d) call was issued previously. This is because previous wake-sources remain active, whereas machine.deepsleep() shall put the board to sleep forever as per the
machine.deepsleep documentation

See #7158 for more details on : issue description, reproduction steps and fix implementation.

Assessment

S Sonnet · medium

Both issues share the same root cause: previously configured wake-up sources persist and continue firing on subsequent sleep calls that should be indefinite. The candidate PR fixes this for machine.deepsleep() by disabling all wake sources in machine_sleep_helper when no timeout is provided. The query reports the identical problem for machine.lightsleep(). Since both sleep modes go through machine_sleep_helper in modmachine.c, the PR's fix likely resolves the lightsleep case too — but the PR explicitly targets machine.deepsleep() and references issue #7158, not this issue. Without seeing the full diff it's unclear whether the fix is conditioned on sleep type or applies broadly.

Suggested action

Review PR #12782's diff in `machine_sleep_helper` to confirm whether the `esp_sleep_disable_wakeup_source` call applies when `lightsleep()` is called without a timeout. If yes, close issue #12579 as resolved by PR #12782. If the fix is deepsleep-only, keep #12579 open and cross-reference #12782 as a related fix targeting the same root cause.

ESP32 timer wake up repeating

esp32 : fix for #7158 - machine.deepsleep() not behaving as expected

Keyboard