This is the mail archive of the
gcc-patches@gcc.gnu.org
mailing list for the GCC project.
Re: Darwin FP constant peephole
- From: Aldy Hernandez <aldyh at redhat dot com>
- To: Dale Johannesen <dalej at apple dot com>
- Cc: Richard Henderson <rth at redhat dot com>, gcc-patches at gcc dot gnu dot org
- Date: 15 Dec 2001 23:55:39 -0800
- Subject: Re: Darwin FP constant peephole
- References: <6D10E4B9-F0C3-11D5-86F6-003065C86F94@apple.com>
>>>>> "Dale" == Dale Johannesen <dalej@apple.com> writes:
> On Tuesday, December 4, 2001, at 04:13 PM, Richard Henderson wrote:
>> On Tue, Dec 04, 2001 at 11:30:48AM -0800, Dale Johannesen wrote:
>>> + (define_peephole2 <details snipped>
>> You should be able to do this with LEGITIMIZE_RELOAD_ADDRESS,
>> if I am not mistaking where this construct comes from.
> Thanks, that was helpful. This replaces the original peephole2 patch
> (catches some more cases, too). The problem is Darwin specific; I'd
> prefer to put this in rs6000.h, but there is a reason not to, as shown
> in the comment. Bootstrapped and tested on Darwin.
> 2001-12-14 Dale Johannesen <dalej@apple.com>
> * config/rs6000/darwin.h: LEGITIMIZE_RELOAD_ADDRESS: Add
> Darwin-specific version containing code for load of FP
> constant.
I know richard already approved this, but i would really like
LEGITIMIZE_RELOAD_ADDRESS in rs6000.h to be rewritten as a function
and then have it include both darwin and non darwin code in just one
function.
Aldy