Re: [C--] C-- development

Ning Wang <[email protected]> Wed, 13 Mar 2013 22:51:21 -0700
Newsgroups gmane.comp.lang.c--
Message-ID <[email protected]>
On 03/13/2013 07:53 PM, Hendrik Boom wrote:
> On Wed, 13 Mar 2013 17:42:22 -0400, Norman Ramsey wrote:
>
>>>  From my outsider perspective, it appears that C-- is moribund until
>>   > someone picks it up again. However, I also think that, rather than
>>   > starting from the current C-- code base, it would be better to
>>   > migrate the distinctive features of C-- into the LLVM project.
>>
>> My opinion is that this would be nearly impossible.
> Well, with enough casts you can achieve almost anything.  :-)
>
> But the I might as well be using C itself, and have the advantage of
> multiple compilers written to an exacting standard.
>
>> LLVM and C-- make
>> fundamentally different assumptions about what knowledge is preserved
>> during code generation.  LLVM is a fantastic infrastructure but it
>> essentially aimed to do what gcc does, only better.  C-- had very
>> different goals.   I think it would be easier to do a new project from
>> scratch.
> It might be easier to compile LLVM to C--.
I have this same idea for a while, but compiling C-- to LLVM is easier 
than compiling LLVM to C--.   I already have the front-end to parse LLVM 
code,  and my next planned optimization pass is constant folding. Once 
this is done, compiling LLVM to C-- should be possible except C-- still 
does not support variadic functions.

Ning

_______________________________________________
Cminusminus mailing list
[email protected]
https://lists.eecs.harvard.edu/mailman/listinfo/cminusminus