issues builing on windows (and a partial resolution and a possible bug)

["Benjamin Hutz" <bhutz@amherst.edu>]

Title: issues builing on windows (and a partial resolution and a possible bug)

The environment is a windows xp pentium, running the latest cygwin trying to build pari-2.4.4 and gp2c-0.0.5.

When building pari it sometimes works and sometime doesn’t.  When it fails the problem is that the make file is lacking a name for one of the .o files.  Going in and adding the name before the $(_o) fixes it. (I can compile the profiling version and the non-profiling without graphics or readline this way).

Gp2c builds just fine, but when trying to use it the first problem is:

>> /usr/lib/gcc/i686-pc-cygwin/3.4.4/../../../../i686-pc-cygwin/bin/ld:

>> cannot find

>>  -lpari.dll

(with some help from a list member) Changing the gp2c-run script to have “-lpari” instead causes this to work.  (You may or may not need to add "/usr/local/lib” to the path).

I’ve attempted to compile and use the squfof.gp example and it doesn’t seem to behave like the tutorial describes and gives a segmentation fault.

So, I made a very simple test function

test(n) = {

 local(k);

 k=1;

 return(k);

}

It compiles just fine and gp starts up, but again running the function cause a segmentation fault.  Changing the “work” of the function is irrelevant.  The only thing I could get to work was

test(n) = {

 return(n);

}

Any ideas what is causing this segmentation fault and how to fix it?

Thanks,

  Ben