debian+phosh running on PinePhone

[a-wai]

I thought not waking from suspend on call or text was down to hardware:
https://wiki.pine64.org/index.php/PinePh..._connected
https://wiki.pine64.org/index.php/PinePh...nts_wakeup
I can imagine how CRUST might be able to change which pins can receive interrupts in suspend, but a pin that's not connected  is another matter entirely. Have I misunderstood the nature of the problem, or does 'figured it out' mean 'found an undocumented behaviour we can use as a workaround'?

That surely complicates things, but testing s2idle shows that we can do without AP_READY, although more work is needed to make wakeup from RI reliable.
It should also be possible to hack the crust firmware and linux kernel to keep GPIO port B powered on and able to process interrupts during suspend, which would allow wakeup from modem. Admittedly we're still a long way from there, and I may be wrong about that, but I really believe it should be possible.