GMP-4.1.3 Built as Win 32 DLL with MPFR-2.0.3 and Tests

[Patrick Pelissier]

Hello,

>If there is general interest, where should I post the files?
 Send then at mpfr at loria.fr.

> Are diffs better?
 Yes.
 
 Sincerely,
  Patrick Pelissier

On Wed, Aug 25, 2004 at 19:31:33 -0400, Mike Loehr wrote:
> Greetings;
> 
> I've been following the posts concerning GMP in a Win32 DLL.
> 
> Using the excellent info from Brian Gladman, Sysiphus, and the Core project
> sites, I was able to create a VC-6 project and work space that compiles
> GMP-4.1.3 with the newest MPFR-2.0.3 (several edits in mpfr.h).  The biggest
> problem seems to be that the functions in mpfr.h and mpfr-impl.h aren't
> setup to be exported into a DLL.  I've added an mpfr_declspec and
> mpfrt_declspec macro (based on the gmp macro) in front of the mpfr functions
> and the impl functions that are required to be exported to build the test
> programs.  I've edited mpfr.h, and mpfr-impl.h to get this set up so that
> the test functions may or may not be included in the DLL.  I've also added
> some logic that should (that means it is untested) allow the Linux/gcc users
> to compile as they currently are by defining two new macros that set the
> declspec to blank.
> 
> I've been able to build the MPFR Test programs on the command line using a
> makefile and the GMP DLL (I chose to call it gmp-4.1.dll in the VC project
> based on the glib naming convention) produced by the VC project.  The tests
> are run from a .bat file and they all (except for the str_out which crashes
> after the first output) seem to run, however there are many that only fail
> on the memory test at program exit, not on the internal mpfr functions.
> 
> If there is general interest, where should I post the files?  Are diffs
> better?  Anyone interest in testing my creations? Pardon any dummy
> questions, I'm new to working on these projects.
> 
> Mike


> _______________________________________________
> gmp-discuss mailing list
> gmp-discuss at swox.com
> https://gmplib.org/mailman/listinfo/gmp-discuss