Just so this is documented somewhere:

Currently the same54-xpro board has 2 mdt based interfaces, the spi_nor and the at24mac... This should be described by the board somehow but maybe abstracting the a HAVE_ is not the best way since , only for this board we need to select the spi_on_qspi for the spi_nor. With an abstracted HAVE_MTD_SPI_NOR this should normally bring in just the periph_spi but not all boards will need this extra condition.

Thanks for the review @benpicco I adapted your changes and hopefully cleared up the MODULE_PERIPH_INIT_SPI_ON_QSPI issue.

I also fixed the structure of the mtd and eui by resolving them in the modules based on the HAVE_ convention. The only module selection that occurs in the board is based on a peripheral specific configuration, similar to the LPUART stuff needed for selection STDIO.

rebased to fix the atmac conflict thing. Also running the whole test suite to see if I broke other things.

doh, I didn't test the right board... Working on it

Things we getting a bit hairy so a rebased and squashed.

This PR will only add the missing modules needed for the same54-xpro board, add the same54-xpro board, and apply the missing modules to any affected boards. This does not mean the the additional boards will be working just that modules added are correctly applied.

That being said, the adafruit-itsybitsy-m4adafruit-itsybitsy-m4 still has some missing modules (something with the USB) and is not being added to the list yet but will be done later.

Looks good from my side, feel free to squash

Done! Should we try the automerge?

seems like some memory was full in the CI, rerunning.

Thanks!