[Coco] C compiler crashes VCC with Nitros9

[Gene Heskett]

On Thursday 03 October 2013 15:00:24 Bill Pierce did opine:

> Gene, He's running straight from the MW distro
> CC1 is the front end, you are thinking cc.pass1.
> cc1 runs cc.pass1, cc.pass2, c,prep, c.asm, c,opt and c.link in one
> swoop.

I figured someone would come along and help me extract my foot. Thank you.  
:)
> 
> His problem is his nitros9 version (3.2.8). and the Vcc emulator. Vcc
> and nitros9 3.2.8 had problems getting along for some reason. Wasn't
> 3.2.8 the one that sat in the repo for the past 5 years until it
> attracted some cockyroaches?

Something like that. ;-) And one of the reasons I keep copy's of old repo's 
going back to the turn of the century.  And some of that goes back to 1993.

> I think I remember having a problem in
> 3.2.8 with C and that's why I searched and found 3.2.9. Now I got my
> custom C setup and I have no problems... :-)

3.2.9 is several years old, and IMO should have been bumped to 3.3.0 even 
before the lwtools conversion was even thought of.  But its not my boat to 
row.

Now its totally confusing as to which code someone has when they say 3.2.9.

And because of situation like this, we really really need a version number 
control for the C compiler because the only modules that have not ben 
majorly changed or replaced are c.pass1/c.pass2/c.comp. Everything else has 
been tossed and improved, and there seems not to be anyplace on the net 
where a new bee can go get the whole thing in one big tarball that once 
unpacked and placed where it belongs on the coco, is KNOWN to Just 
Work(TM).

With the existing situation, bit and pieces scattered all over the planet, 
and almost as many opinions on whats right as there are members of this 
list, it is totally understandable that the newbie coming here with a 
smidgeon of C knowledge, is going to just throw up his hands & go get a 
cold one, turning on the tv on his way to the fridge.

And we've lost another potentially valuable contributer.

The acid test of whether you have a serviceable compiler setup is to build 
rzsz-3.36 with it.  Most setups will build it, but the final test is to 
actually use it.  The correctly built one will NOT crash, anything else 
will hand you a basket of confetti on the screen at some point.

BUT, while its a nice exercise to prove the compiler is making good code, 
please DO NOT REPLACE the publicly available archive of rzsz-3.36 because 
that was built on my machine with several hand optimizations done by 
stopping the compile and looking at the assembler language intermediate 
file being fed to the assembler that are good for a speedup of 175-200 cps 
in transfer speeds over what you can build.  This compiler can handle bit 
shifts, but does it in very time consuming one bit at a time across 2 
registers loops, whereas the exact same results if the shift is 8 bits or 
more can be had with 3 assembly commands to replace the first 8 looped, 
single bit shifts.  And most of its crc calculations involve fixed 8 bit 
shifts.

> Bill Pierce
> My Music from the Tandy/Radio Shack Color Computer 2 & 3
> https://sites.google.com/site/dabarnstudio/
> Co-Webmaster of The TRS-80 Color Computer Archive
> http://www.colorcomputerarchive.com/
> Co-Contributor, Co-Editor for CocoPedia
> http://www.cocopedia.com/wiki/index.php/Main_Page
> E-Mail: ooogalapasooo at aol.com
> 
> 
> 
> --
> Coco mailing list
> Coco at maltedmedia.com
> http://five.pairlist.net/mailman/listinfo/coco


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)

Revenge is a form of nostalgia.
A pen in the hand of this president is far more
dangerous than 200 million guns in the hands of
         law-abiding citizens.