bin2uf2.js and uf2conv.py script appstart same for samd21 and samd51? about uf2-samdx1

[tannewt]

Ya, it should be 0x4000 for samd51. In CircuitPython we use the py and provide the offset on the command line IIRC.

[FHeilmann]

@ladyada @dhalbert what is your opinion on this? These false offsets bear the risk of people nuking their bootloaders on M4 boards.

[dhalbert]

As @tannewt notes, the Makefile explicitly specifies the offset argument for uf2conv.py with -b ~$(BOOTLOADER_SIZE). So unless you run this by hand, or use binuf2.js, it will not produce a UF2 with a bad offset.

We can, however, take away the default values for these arguments. bin2uf2.js doesn't take arguments but we could add those (or just get rid of this script in favor of the python one).

Also I'd note that the Makefile tries to use python2 to run uf2conv.py. It should be made to be version independent or py3 only. @TG-Techie noticed that this morning.

[FHeilmann]

I see, thanks for clearing that up for me. Removing the default values from these scripts would make them more foolproof in my opinion since you would not be able to execute them without specifically specifying an offset. In addition to your notes I also want to mention sketch.cpp again where bootloader_size is passed as argument, but BOOTLOADER_K is hardcoded to be 8k

I'm not sure how to go about this (maybe you can point me in the right direction) but the same issues are present in the upstream microsoft repository. Should I open a separate issue over there?

Thanks again for your time!

[dhalbert]

We send PR's to upstream regularly, and we're the only non-8K bootloader right now, so just here is fine.