gcl-devel
[Top][All Lists]
Advanced

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: [Gcl-devel] Re: gcl on ms windows


From: Camm Maguire
Subject: Re: [Gcl-devel] Re: gcl on ms windows
Date: 08 Mar 2002 11:13:17 -0500

Greetings!

"Mike Thomas" <address@hidden> writes:

> OK, more progress on the Mingw32 build of GCL and Maxima.
> 
> I completely rebuilt GCL using my own build of GCL (including .c, .data and
> .o files).  The differences below occurred both before and after the
> rebuild.
> 
> Looking at the .data files in the Cygwin builds of GCL and Maxima put up by
> Dan Stanger, I noticed significant differences with those produced when I
> compiled Maxima (NOT GCL however).
> 
> The three most notable differences are:
> 
>    1. Mine use '0xa' as an end of line marker (as do the CVS GCL .data
> files). Dan's use 'CR/LF' separators. This seems to be of no import and is
> probably the result of my setting default I/O to be binary in main.c, as
> done by Rainer.
> 

Agreed, this is likely of no import.  Who is Rainer, and where can I
find out what he's been doing?

> The CVS .data files (for example, "cmpnew/*.data") for GCL are identical to
> mine (according to "cvs diff") after being rebuilt from scratch here.
> 

Great!!!  Just to clarify, you've built a Mingw32 gcl from cvs, and
then used that to recompile the gcl lisp files, and got no
differences.  Wonderful!  If true, then it would seem that *any*
subsequent discrepancy in the .c, .h or .data files your gcl produces
on and lisp file as compared to the i386 Linux gcl *would indicate
some failure outside the compiler, most likely in the gc*.


>    2. My scratch Maxima .data files are substantially different to Dan's
> (and the GCL ones) in that they seem to be a completely different format.
> For example, my new Maxima files lack a leading line of spaces followed by
> some kind of vector? format header which is present in Mingw32 GCL and the
> Cygwin builds:
> 
>       #(#! blah blah....
> 

Can you provide a sample?

>    3. They are also upper case, whereas Dan's and the GCL .data files I made
> (and the CVS GCL ones) are all lower-case!!!
> 

Very odd, but isn't lisp case insensitive?  I wonder if this could
also be a system (i.e. Mingw) matter.  Being completely ignorant, but
guessing that there is a DOS legacy, perhaps 'printf' statements are
compiled to produce upper case?  Try a hello world program in C.

> These last two differences seem to be critical to my Maxima build, as it is
> during fasload of "defcal.o" that the NULL pointer dereference occurs.
> 

I suspect the data files.  Would like to see your corrupted one.

> I will look further into why there is such a substantial difference between
> the .data files generated for GCL and Maxima in the near future, clearly
> something radically different is happening.  My guess is that the problem
> may be a really simple Makefile or configure issue, which I think would be
> quite a positive outcome!

Agreed, this would be great!

> 
> Meanwhile, any insights would be appreciated.
> 
> Cheers
> 
> Mike Thomas
> 
> 
> 
> 
> 

-- 
Camm Maguire                                            address@hidden
==========================================================================
"The earth is but one country, and mankind its citizens."  --  Baha'u'llah



reply via email to

[Prev in Thread] Current Thread [Next in Thread]