Rahix

> For instance, the series of the Atmega 2560 have differential ADCs with configurable gain, in addition to the single-ended ones, and two sets even, chosen by the MUX5 bit....

> OK, but I think I won't fully give up on the differential features, trying to not make them harder to implement later; the hope is to avoid another overhaul...

Just a heads up: We need to upgrade `avr-device` to a newer `svd2rust` at some point (https://github.com/Rahix/avr-device/pull/155) and this will produce some slight changes to the API that we need...

You are using the ATmega168 directly, instead of a prebuilt Arduino/SparkFun/etc. dev-board, right? For such situations, you shouldn't use `arduino-hal` and instead depend on `atmega-hal` directly. That way, you are...

The `avr-hal-template` is updated to pin the dependency version. New projects generated from the template will no longer run into this problem. Unclear whether this is enough or whether we...

Examples in this repository now also have the dependency pin (#544).

So `avr-device` will need to deal with this for its own CI, too. So maybe the solution is to simply pin `proc-macro2` in `avr-device-macros` so it will solve the problem...

From #546: The CI of this project (`avr-hal`) does not fail on this... Needs to be investigated.

We have to catch two problems: 1. Wrong rust target 2. Wrong MCU-feature selected in the HAL For the rust target, I guess we can just check the architecture from...

@brainstorm, I think this is not really related to the topic of this issue. If you want to discuss this, please open a separate issue.