This is the mail archive of the
fortran@gcc.gnu.org
mailing list for the GNU Fortran project.
Re: Segfault on O2 optimisation for test case for PR18689
- From: Tobias Schlüter <tobias dot schlueter at physik dot uni-muenchen dot de>
- To: Paul Thomas <paulthomas2 at wanadoo dot fr>
- Cc: fortran at gcc dot gnu dot org
- Date: Wed, 11 May 2005 23:24:08 +0200
- Subject: Re: Segfault on O2 optimisation for test case for PR18689
- References: <008601c5565b$30089a50$0400000a@Paul> <428258B5.5050701@physik.uni-muenchen.de> <002301c5566d$ffd1a300$0400000a@Paul>
Paul Thomas wrote:
> This discussion and some of the remarks made today remind me of something
> that has struck me more than once; we need a testsuite component of correct
> fortran; not just things that we have found are broken.
>
> Is there somebody out there, ready to write an f95 NIST - or even the
> beginnings of one?
I was thinking the same thing today, given the regression with character
functions that the f2c patch introduced.
Unfortunately, that's a large body of things to cover, and I don't think
there's anything freely available, and the time is probably better spent at
fixing the known shortcomings of gfortran.
But if someone wants to help gfortran's development without learning the
compiler's internals, please go ahead :-) Actually, I vaguely remember someone
offering to do this in the not-so-distant past.
- Tobi