[Coco] Re: [Color Computer] GCC Update

[Gene Heskett]

On Thursday 28 July 2005 03:23, James Diffendaffer wrote:
>It's been a few days so I thought I'd let people know what was up
> with my work on GCC.
>
>I've been too busy with other things to work on GCC for any real
>length of time so I'm not making code changes at the moment.  I
> don't want to get in the middle of a bunch of changes and have to
> abandon work for a few days and then spend a lot of time trying to
> figure out where I was at.
>
>In addition to GCC I need to scan some magazines and send those off
> to join the DVD collection.  So many projects, so little time.
>
>For those looking at the GCC source or that want to... most of the
>changes from the 68hc11 are in the ".md" file.  It contains most of
>the code generation logic and peephole optimizer logic.
>
>The changes required to the .md file are due to differences in
>instruction names/set.
>The code makes use of slightly different stack instructions, there
>load and compare instruction differences... stuff like that.
>Nothing looks complicated.
>
>Once I have time to sit down and make all those changes at one time
> I will and I should be pretty close to done.  I'll need to look
> through all the instruction generation code again to be sure and
> then make the proper config files to compile the code with 6809
> names instead of hc11.  After that I'm generating code to see if
> all is well.
>
>Once the code generated looks ok I'll start working on making the
> code more OS-9 friendly as I have time.

Are you familiar with the diff command from the *nix camp?  If so, 
then you should work against an unpatched tree for gcc, doing your 
changes to a copy so that when the copy works, you can then run the 
diff command and generate a patch that will apply to anybodies gcc 
tree to effect the updates.  This is how progress is made, and makes 
complying with the GPL a piece of cake.  This also means you should 
start with the current version 4.0.1 srcs available from gnu.org, 
which is currently in a state of flux as they look for reasons why it 
won't compile the x.org X suite among other things.

As there will likely be a 4.0.2 shortly because of this, use the diff 
command to make your local patch so that if and when 4.0.2 or later 
is out, your patch can be brought forward with little or no hassle 
just by running a 'patch -p1 <your-patch-file' to bring the work 
you've done into the next release.  No use re-inventing the 
wheel. :-)

>I'd like to revisit SDCC at some point to get it working as well. 
> GCC relies heavily on ram based registers on small CPUs and SDCC
> doesn't. I don't think it will be a huge slowdown because a CPU
> with so few registers needs to go to RAM a lot anyway and by
> storing them on page 0 or the dp page it may actually speed some
> code up.
>
>Besides, the advantage of having all the compilers that go with GCC
>(C/C++/ObjectiveC/Fortran/Ada...) would make a minor performance hit
>worth it.  Since the number of ram based registers is supposedly
>configurable, a lot of performance testing will probably be required
>to generate optimal code.
>
>
>
>
>
>
>Brought to you by the 6809, the 6803 and their cousins!
>Yahoo! Groups Links
>
>
>

-- 
Cheers, Gene
"There are four boxes to be used in defense of liberty:
 soap, ballot, jury, and ammo. Please use in that order."
-Ed Howdershelt (Author)
99.35% setiathome rank, not too shabby for a WV hillbilly
Yahoo.com and AOL/TW attorneys please note, additions to the above
message by Gene Heskett are:
Copyright 2005 by Maurice Eugene Heskett, all rights reserved.


Brought to you by the 6809, the 6803 and their cousins! 
Yahoo! Groups Links

<*> To visit your group on the web, go to:
    http://groups.yahoo.com/group/ColorComputer/

<*> To unsubscribe from this group, send an email to:
    ColorComputer-unsubscribe at yahoogroups.com

<*> Your use of Yahoo! Groups is subject to:
    http://docs.yahoo.com/info/terms/