Re: A92: Re: Suggestion: A simple C compiler
[Prev][Next][Index][Thread]
Re: A92: Re: Suggestion: A simple C compiler
>cc68k is a shitty K&R style compiler: it will not understand the modern
>ansi way to code function protoyping... And the code it generates is not
>very usefull for a68k...
Oh course it isn't ideal. But that's why it's called RESEARCH!! I've
used cc68k and actually made working programs. Unfortunately, it required
a post-processing stage on the ASM. Then it was ready for A68k.
>However: an idea: once, gcc could do cross compiling on 68000 machines...
>Has anyone tried to get ideas about this ? My linux box crashed two days
>ago so i cannot be of much help here but i will definitely try to search
>in this directrion during my Xmas holidays.
ARGH!!! Do not even bring that up! Noah and I have spent so many hours
trying to get it to work. GCC (and DJGPP for DOS) is a GREAT compiler.
However, we cannot use the 68000 cross-compilers. Because in essence, our
target-platform is not a 68000. It is a TI-92 (or 92II, 92+, or 89).
David found that the Amiga platform was the closest match. This is why
our code in assembled to AmigaDOS object code. And even if we could find
the cross-compiler for AmigaDOS for GCC (or DJGPP) we still couldn't use
it because, again, we are not a Amiga computer. To make this GCC and DJGPP
thing work, we will need to write our own binaries to define our own target-
platform. Considering we have a great assembler and linker, we will only
need to rewrite the compiler proper. (The preprocessor may need modification
for FARGO/PlusShell specifics) Currently the compilers all output the GCC/
DJGPP common assembler format. It was their goal to make all output very
generic. (Grutsch, it's already there... don't waste your time re-inventing
the wheel) We simply shape the output to A68K compatible assembler, and we
are done.
Have fun, whoever is stu... uh... willing to do it. ;)
====
Aaron Hill (Redmond, WA)
E-Mail: SeracOhw24@msn.com
IRC-Nick: Serac (EF-Net) (was SeracOhw)