> For what it's worth, the decision to use unsigned chars on NetBSD/arm32
> _was_ our decision :-) (at least in part).

"Yes, you can always do something gratiutously different."

It's nice that it's helpful for debugging, but efficiency and
compatibility with other compilers for the architecture should be (or
should have been) the primary deciding factor for something like this.

I mean, hell, some architectures cause a _lot_ of pain with their
default choices.  If I'd used ILP32 on the Alpha instead of LP64, i'd
have saved myself months of troubles...  The decision to use LP64
wasn't done because it'd help me debug NetBSD.  8-)


> > The PowerPC also uses unsigned chars, but it doesn't use our in-tree
> > gcc sources, so hasn't tripped over the problem.
> 
> The PowerPC port tripped over the problem for quite some time.

It tripped over problems like this one, but not this specific problem.

Or at least, if it did, I don't see how it did so, since there's no
powerpc support in our in-tree toolchain.

I realize that the PowerPC port solved many of the char-is-unsigned
problems, but that's ... completely irrelevant in this particular
instance.



chris