ESP32-C6: does machine.deepsleep() work with esp32.wake_on_ext1() or esp32.wake_on_ext0() for anyone?

[ned-pcs]

For context, see #17334 .

I've been successfully using machine.deepsleep() with esp32.wake_on_ext1() and/or esp32.wake_on_ext0() on the ESP32-S3 for some time (in an upcoming mass-produced consumer device).

But recently I tried using MicroPython 1.25.0 with a Seeed Studio XIAO ESP32-C6 board on a personal project (see https://github.com/bikeNomad/micropython-tv-b-gone), and found that the GPIO wakeup from deepsleep() was broken in several ways:

• esp32.wake_on_ext1([button], esp32.WAKEUP_ALL_LOW) caused the GPIO to get stuck at 0
• esp32.wake_on_ext0(button, esp32.WAKEUP_ALL_LOW) didn't wake on the GPIO transition
• esp32.wake_on_ext1([button], esp32.WAKEUP_ANY_HIGH) worked (but my button pulls low)

Maybe there's something I'm missing here.
I have been using one of the pins (0-7) that are available to the LP IO MUX.
I have tried using hold=False after waking up, though that shouldn't be necessary.

Has anyone gotten this to work on the ESP32-C6?

[meirarmon]

I know that it does not work for the esp32c3 (I am working on a fix). Check with the espressif documentation or code if the boards supports ext0 and ext1 capabilities (SOC_PM_SUPPORT_EXT1_WAKEUP).
The fix that I am working on right now, adds support for waking with gpio instead.

[ned-pcs]

It looks like the ESP32-C6 does not support wake_on_ext0. It does support the ext1 and GPIO wakeup sources, though. I don't know why the GPIO got stuck at 0 though.

[meirarmon]

I will probably have a pr for wake on gpio within a couple weeks (it's wip), when I do, if you can test it (it will help you as well as help me) that would be much appreciated :-)

[meirarmon]

@ned-pcs do you wanto to take a look at this PR (#17518) and test it on your chip to see if it helps you? I added a wake_on_gpio method similar to wake_on_ext1, and it works on my ESP32C3.