* Initial firmware support for Axiometa Genesis Mini
* Report the real AXIOMETA_GENESIS_MINI hardware model
The protobufs now carry AXIOMETA_GENESIS_MINI = 148, so the board no
longer has to masquerade as private hardware.
On ESP32 the -D PRIVATE_HW flag never selected the model on its own -
architecture.h has no arm for it, so the board fell through to the
PRIVATE_HW default at the end of the chain. Give it its own arm and drop
the flag.
* fix(input): sample the encoder button after light-sleep wake
The edge that wakes the device lands while beforeLightSleep() has the
interrupts detached, and a button that is still held produces no further edge
until it is released. The thread stayed parked at INT32_MAX, so the press was
never sampled - no event was emitted, PowerFSM's GPIO-wake branch reads
BUTTON_PIN rather than the encoder pin, and the node dropped straight back into
light sleep with the press swallowed entirely. The second press worked, the
first did not.
Sample once on wake, and only when the button is asserted, so a timer or radio
wake leaves the thread alone. The press then follows the ordinary path and
InputBroker drops the event because the screen was off, so it wakes the screen
and does nothing more - the same behaviour every other input device has.
Rotation stays deliberately non-waking: only the button pin is armed in
doLightSleep(), and the abState re-seed discards a shaft moved during sleep
rather than replaying it as detents.
Reported by CodeRabbit on #11852.
Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
---------
Co-authored-by: rcarteraz <robert.l.carter2@gmail.com>
Co-authored-by: Claude Opus 5 <noreply@anthropic.com>
