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: [PATCH, MIPS] Fix build after no_new_pseudos

From: Ian Lance Taylor <iant at google dot com>
To: kenner at vlsi1 dot ultra dot nyu dot edu (Richard Kenner)
Cc: anemet at caviumnetworks dot com, gcc-patches at gcc dot gnu dot org, richard at codesourcery dot com, sandra at codesourcery dot com
Date: 16 Jul 2007 10:25:13 -0700
Subject: Re: [PATCH, MIPS] Fix build after no_new_pseudos
References: <18074.28991.32902.226849@localhost.localdomain> <87lkdghf3g.fsf@firetop.home> <m3lkdg9uhe.fsf@localhost.localdomain> <10707161721.AA21942@vlsi1.ultra.nyu.edu>

kenner@vlsi1.ultra.nyu.edu (Richard Kenner) writes:

> > I think we should fix combine to permit splitters to create new
> > pseudos.  I don't think it does us any useful service to prohibit
> > them.
> 
> What if the splitter doesn't result in a recognized insn (most don't).
> We could potentially be creating a lot of unneeded pseudos.

I'm not sure I agree.  I suspect that most splitters, when they
succeed, do result in recognized insns.

Conversely, some splitters may require a new pseudo, and yet be quite
useful if they can be run during combine.

In any case, this is something we can measure.

Ian

References:
- [PATCH, MIPS] Fix build after no_new_pseudos
  - From: Adam Nemet
- Re: [PATCH, MIPS] Fix build after no_new_pseudos
  - From: Richard Sandiford
- Re: [PATCH, MIPS] Fix build after no_new_pseudos
  - From: Ian Lance Taylor
- Re: [PATCH, MIPS] Fix build after no_new_pseudos
  - From: Richard Kenner

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