This is the mail archive of the
java@gcc.gnu.org
mailing list for the Java project.
Re: gcj-4.2 (20060826) and gnu classpath
- From: Hanno Meyer-Thurow <h dot mth at web dot de>
- To: Andrew Haley <aph at redhat dot com>
- Cc: Tom Tromey <tromey at redhat dot com>, java at gcc dot gnu dot org
- Date: Sat, 9 Sep 2006 11:22:39 +0200
- Subject: Re: gcj-4.2 (20060826) and gnu classpath
- References: <20060828161138.d5e73fdb.h.mth@web.de> <m3lkp8trye.fsf@localhost.localdomain> <20060830161546.fa478b1a.h.mth@web.de> <m3irjy2aw1.fsf@localhost.localdomain> <17666.34194.487720.447877@zebedee.pink>
On Sat, 9 Sep 2006 10:12:50 +0100
Andrew Haley <aph@redhat.com> wrote:
> Tom Tromey writes:
> > >>>>> "Hanno" == Hanno Meyer-Thurow <h.mth@web.de> writes:
> >
> > Sorry for leaving this ... I've been distracted.
> >
> > >> Could you try to make a reduced test case? That might help.
> >
> > Hanno> Maybe this one attached helps?
> >
> > Yes, thank you very much.
> >
> > When I compile fdlibm with -O2, I see the bug.
> > When I compile it without, it works.
> >
> > I suspect a strict aliasing bug in the new mprec code.
>
> I don't think so. Are we compiling with -ffloat-store? I can't see
> it in the makefile. If we don't use -ffloat-store the code won't
> work.
I see it in here:
http://gcc.gnu.org/viewcvs/trunk/libjava/configure.host?view=markup
Hanno