RE: What are the chances of GCC for PRIMOS?

["Dave Korn" <dave dot korn at artimi dot com>]

On 18 April 2007 11:00, Andrew Marlow wrote:

[ list Cc'd back in as I see you sent this reply there as well as off-list ]

>> But the
>> odds of some random gcc hacker with no special interest in
>> PRIMOS suddenly deciding to volunteer that amount of time and
>> effort for no reason are ... small.
> 
> Er, yes. I don't have the skills myself. I was hoping that maybe some
> ex-primates hung around in the GCC ml. Perhaps this was wildly optimistic. 

  Not /wildly/ optimistic, but optimistic yes.  There are quite a lot of
pretty old-school big iron guys among the older crew, as it happens.  Even so,
they'd have to be pretty keen to want to do this work...
 
> There may be some chance of this. In comp.sys.prime it was just announced
> that the source for a very early unofficial version of the C compiler has
> been found. I wonder if this might be a useful starting point. Does this ml
> have a pointer (URL maybe) for any GCC implementing a new backend
> documentation that I can point my fellow primates to? Maybe one of them
> could use this info in conjunction with the source that was just
> discovered.      

  The main documentation is the gcc internals manual:
http://gcc.gnu.org/onlinedocs/gccint/, other formats available at
http://gcc.gnu.org/onlinedocs/.  After that, the next step is to look at the
other backend ports, pick one that's architecturally close to your target, and
study how it makes things work.  See also http://gcc.gnu.org/readings.html,
particularly HP's "Porting gcc for dunces".

  There's no chance the code of the old compiler will be /directly/ useful in
a gcc context, but much of the information it tells you will surely be vital.

>>> There is an emerald here the size of a plover's egg!
>>   plugh!
> 
> Nothing happens.

  A hollow voice says "Well, I wasn't really expecting it to work here
anyway."


    cheers,
      DaveK
-- 
Can't think of a witty .sigline today....