This is the mail archive of the
gcc-patches@gcc.gnu.org
mailing list for the GCC project.
Re: [committed] Skip floating-point var-expand tests for soft-float MIPS targets
On Wed, 2007-11-28 at 14:01 -0800, Bob Wilson wrote:
> On Nov 27, 2007, at 4:32 PM, Hans-Peter Nilsson wrote:
> >
> > (Patch in attachment, can't quote; point is that it looks at
> > port-specific preprocessor options so it's not general enough. ;)
> >
> > I think a better method, one that doesn't need tweaking for each
> > target, would be as I suggested in
> > <http://gcc.gnu.org/ml/gcc-patches/2007-05/msg00398.html>:
> > Generate a .expand dump "-dr", of e.g.
> > double a (double b, double c) { return b + c; }
> > and grep the RTL dump for "call".
> >
> > If it's there, the target is not hard_float.
>
> I don't have a strong opinion on how to handle this. Unfortunately,
> I don't have time right now to implement anything other than a simple
> check. If my patch isn't acceptable, I'll just have to ignore those
> failures for now. Maybe sometime in the future I'll be able to fix
> them in a way that is general enough.
The patch is OK. If H-P feels strongly enough about his approach
then he is welcome to submit a patch to replace the guts of
check_e*f*hard_float.
Janis