This is the mail archive of the gcc-patches@gcc.gnu.org mailing list for the GCC project.


Index Nav: [Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav: [Date Prev] [Date Next] [Thread Prev] [Thread Next]
Other format: [Raw text]

Re: stormy16: limit SI reload regs


> X-Original-To: geoffk@foam.wonderslug.com
> Date: Wed, 30 Apr 2003 21:12:29 -0400
> From: DJ Delorie <dj@redhat.com>
> Cc: gcc-patches@gcc.gnu.org
> X-OriginalArrivalTime: 01 May 2003 01:12:05.0562 (UTC) FILETIME=[ADF9A1A0:01C30F7E]
> 
> 
> > This was in response to a different change to the xstormy16 backend,
> > not to a change in reload.
> 
> Ok, I'm confused.
....

Maybe it'd be better if I explained the basic objection:

* You have found a bug in reload.

* You are proposing patches that try to work around the reload bug in
  one port.

* I would prefer that you fix reload.

> So if I do fix reload to honor HARD_REGNO_NREGS when considering
> candidates, what assurances do I have that nowhere else in gcc will
> choose that register?

You fix those places too, in the event that you find any.

>  And if it's safe to just fix reload, then
> what's wrong with just telling reload to use a different class?

Because that leaves reload broken for the next port to stumble over.

-- 
- Geoffrey Keating <geoffk@geoffk.org>


Index Nav: [Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav: [Date Prev] [Date Next] [Thread Prev] [Thread Next]