This is the mail archive of the cygwin mailing list for the Cygwin project.

Index Nav: [Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav: [Date Prev] [Date Next] [Thread Prev] [Thread Next]
Other format: [Raw text]

RE: 'Missing isnand' still not fixed in 2006-03-01 snapshot

On 02 March 2006 15:08, Corinna Vinschen wrote:

> However, as another datapoint, I just unpacked cygwin-inst-20060301.tar.bz2
> into a directory and tried to build a testapplication(TM):
>   $ tar xjf cygwin-inst-20060301.tar.bz2
>   $ cat > x.c <<EOF
>   #include <math.h>
>   int
>   main ()
>   {
>     double f = 3.14;
>     return isnan (f);
>   }
>   EOF
>   $ gcc -I./usr/include -L./usr/lib -o x x.c
>   $ ./x
>   $
> So, that works.  

  And I've built and installed from CVS and get the same results: test app
compiles ok and is correctly linked against ___isnand (looked at in objdump)
and I've verified that the headers and libs in /usr/i686-pc-cygwin are the new

> Jerry, you have -a -L//usr/local/lib when linking
> miniperl.  Could you have older Cygwin libs in /usr/local/lib?

  I seem to have gotten myself a /usr/include which is full of old headers, is
that meant to be there?  It didn't interfere with the testcase but I dunno if
I just messed up a $prefix some earlier time when I was building from cvs...

Can't think of a witty .sigline today....

Unsubscribe info:
Problem reports:

Index Nav: [Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav: [Date Prev] [Date Next] [Thread Prev] [Thread Next]