🚀 go-pugleaf

> It contains the type definitions for the Amiga functions.  Using it properly
> should eliminate many ptr->int type errors.  I don't know what Lattice
> does for this.

In Lattice the Amiga protos are in <proto/XXXX.h> where xxx is the
name of the library.  <proto/all.h> will include all libraries.  Note
that these files also have the #pragma's to generate in-line function
called.

>>Smart-ass remarks from Lattice users will be ignored.
>Agreed :^)

>Q#2:Did you guys even TRY compiling any 2.0-specific source? (Ok,maybe I won't
>ask that question.)

SAS 5.1 seems to compile what little 2.0 sources I've stuffed through
it, although I still suffer some minor problems with the 5.1 LMK
occasionally wanting a return to execute.  But overall, its pretty
stable :-).

--
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
Mark Gooderum                 /\     \ | /     H a p p y  
Academic Computing Services  / v\   -- * --         H o l i d a y s ! :-)
University of Kansas        /v  v\   / | \    ///
                           /__v___\   Only  ///  /|         __    _  
Bitnet:   MARKV@UKANVAX       ||     \\\  ///  /__| |\/| | | _   /_\  makes it
Internet: markv@kuhub.cc.ukans.edu     \/\/  /    | |  | | |__| /   \ possible
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~

In article <1951.AA1951@caleb>, Jim Pritchett writes:

> 
> In article <9871@fy.sei.cmu.edu> jol@sei.cmu.edu (Jun Lee) wrote:
> 
> > Can anyone tell me what is contained in <function.h?  I take it this is
> > a Aztec C thing and I was wondering what would be a comparable include
> > file on a Lattice 5.1.  
> 
> It contains the type definitions for the Amiga functions.  Using it properly
> should eliminate many ptr->int type errors.  I don't know what Lattice
> does for this.

   The Prototype files that come with Lattice are the equivalent. This is
great for catching incorrect parameter passings and such. Using prototypes
for your own functions (compiler generated) is very neat, I set the
compiler up to warn me about functions without proto's, and catch quite a
few bugs at compile time rather than at link or execute time.

   Use the pragma include files as well, that's quite a reduction in code
size. Precompiled headers is almost a must for that many header files.

                                         -Henrik

______________________________________________________________________________
| Henrik Clausen, Graffiti Data | If the Doors of Perception where cleansed, |
| ...{pyramid|rutgers}!cbmvax!  | Man would see Reality as it is - Infinite. |
\______cbmehq!adspdk!hclausen___|_________________________________W. Blake___/

In article <27460.2768a8e9@kuhub.cc.ukans.edu>, markv@kuhub.cc.ukans.edu writes:

> In Lattice the Amiga protos are in <proto/XXXX.h> where xxx is the
> name of the library.  <proto/all.h> will include all libraries.  Note
> that these files also have the #pragma's to generate in-line function
> called.
> 
> >Q#2:Did you guys even TRY compiling any 2.0-specific source? (Ok,maybe I won't
> >ask that question.)
> 
> SAS 5.1 seems to compile what little 2.0 sources I've stuffed through
> it, although I still suffer some minor problems with the 5.1 LMK
> occasionally wanting a return to execute.  But overall, its pretty
> stable :-).

   I've encountered a problem with the mathieeedoub*.h pragma files. The
2.0 version of this runs:

#pragma libcall MathIeeeDoubTransBase IEEEDPSin 24 1002

   #pragma experts will notice that this expects two arguments - to do a
Sin()!!!

   I encountered the same when I converted .fd files to pragma statements
with their FD2Pragma program, and it seems FD2Pragma doesn't recognize that
the parameter in D0 & D1 is in fact a double, which means just a single
argument. The syntax of .fd files isn't that helpful in this situation,
though.

   This creates a BLTN109 warning on every math call I did, and that means
quite a few.

   The 5.05 include files looked like:

#pragma libcall MathIeeeDoubTransBase IEEEDPSin 24 1

   Now, I didn't like the sole '1' at the end, it seemed more accurate to
specify the registers, so I changed the line to:

#pragma libcall MathIeeeDoubTransBase IEEEDPSin 24 001


   One of John Toebes' words on this is 'Thou shalt not modify header files'
but there are limits to my patience.

   This has been reported to SAS via the normal channels. The Enforcer hits
by LC2 have not, but I'll avait the 5.1a upgrade to look further at that.


                                                -Henrik
______________________________________________________________________________
| Henrik Clausen, Graffiti Data | If the Doors of Perception where cleansed, |
| ...{pyramid|rutgers}!cbmvax!  | Man would see Reality as it is - Infinite. |
\______cbmehq!adspdk!hclausen___|_________________________________W. Blake___/