This is the mail archive of the
gcc-patches@gcc.gnu.org
mailing list for the GCC project.
Re: Patch for libgfortran; broken isnan/isfinite/fpclassify
- From: Steve Ellcey <sje at cup dot hp dot com>
- To: fxcoudert at gmail dot com
- Cc: fortran at gcc dot gnu dot org, gcc-patches at gcc dot gnu dot org
- Date: Mon, 26 Sep 2005 14:36:01 -0700 (PDT)
- Subject: Re: Patch for libgfortran; broken isnan/isfinite/fpclassify
> > ! #if (defined(fpclassify) && defined(HAVE_BROKEN_FPCLASSIFY))
> > ! #undef fpclassify
> > ! #endif
>
> No need to check for defined(fpclassify), as Andrew Pinski told me about
> a recent patch.
>
>
> FX
I am not sure what you mean by this check not being needed. If I don't
find isnan or isfinite (or they are broken) and I do find fpclassify,
then I implement isnan and isfinite by using fpclassify. On HP-UX our
fpclassify has the same problem as isnan and isfinite (it doesn't work
for long doubles), so I need to check if it is broken (it is) and not
use fpclassify on HP-UX to implement isnan and isfinite.
I'll fix the other issues and retest a modified patch before doing a
checkin.
If I didn't use fpclassify to implement isnan and/or isfinite then I
would not need this check as I removed the only use of fpclassify from
the library sources as part of this patch.
Steve Ellcey
sje@cup.hp.com