Miro Kropáček

> The main purpose of the sysroot is to specify where the target include/lib are stored, instead of the classic $compilation_prefix/m68k-atari-mint? That's my understanding, yes. IMHO the sentence "Print the...

I've wondered about this, too. I've always thought / feared that there's some set of conditions / rules / functions which such a C / C++ compiler must comply to...

That's great stuff, mate! However, I'm experiencing much more unstable behaviour with your version, esp. when I start it in the simulator (iOS 9.3) -- many times I'm unable to...

Any idea what's causing it or better, any hope of fixing it? For me, this is a showstopper, even though I love the other features :-( (I'm still a beginner...

Latest xaaes with the patch above: [xaaes.km.zip](https://github.com/freemint/freemint/files/11189393/xaaes.km.zip)

> to change (imho broken) code that relies on a certain packing That's the problem here (and main argument/wish for keeping the default): you can't change that code easily, if...

> I think we have already clarified that. Defaulting to -malign-int is a no-go. Sure. But for usage in a single cpp file (wrapper) which doesn't interact with mintlib it...

> And i still see no reason to change that. It does not give any benefit, only gives troubles. And it doesn't even help in the case of d2x-rebirth, which...

For the sake of completeness, that bug has been identified and fixed: https://github.com/freemint/m68k-atari-mint-gcc/issues/22#issuecomment-1705697651. Now https://github.com/dxx-rebirth/dxx-rebirth/issues/695 compiles as expected with `-malign-int` and `STRUCTURE_SIZE_BOUNDARY` = 8. This wont be the case with...

Thank you Vincent for the complete overview. It would seem that the m68k-atari-mint is really the only black sheep when handling single `char`s.