Port-amd64 archive

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index][Old Index]

Re: Changing x87 precision to full 63bit as default

On Thu, Nov 07, 2013 at 08:23:43AM -0500, Greg Troxel wrote:
> Joerg Sonnenberger <joerg%britannica.bec.de@localhost> writes:
> >> Have you run paranoia (with no math options in CFLAGS) with the patch?
> >> Is it happy with everything?  If not, can you explain what it objects
> >> to, and why it is wrong?
> >
> > I haven't and I don't plan to.
> Then I object; that's a well-known standard test of floating point,
> arguably the standard test.  A change to floating point behavior that
> doesn't at least understand how that is affected has not been
> sufficiently tested and considered.  It's trivial to build and run, so
> it seems unreasonable not to.

Why haven't you then? The changes return NetBSD to the much more well
understood default behavior and in fact kill some very dubious
port-specific aspects in GCC. In fact, it is a wonder if GCC behaved
correctly at all.


Home | Main Index | Thread Index | Old Index