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: Add missing memory clobbers to atomic memory insns.

From: Hans-Peter Nilsson <hp at bitrange dot com>
To: Richard Sandiford <richard at codesourcery dot com>
Cc: David Daney <ddaney at avtrex dot com>, gcc-patches at gcc dot gnu dot org
Date: Thu, 13 Sep 2007 05:08:53 -0400 (EDT)
Subject: Re: [Patch] MIPS: Add missing memory clobbers to atomic memory insns.
References: <46E378C5.9050904@avtrex.com> <46E379FF.6050602@avtrex.com> <873axo5go3.fsf@firetop.home> <Pine.BSF.4.58.0709121841040.52805@dair.pair.com> <87sl5jt4mx.fsf@firetop.home>

On Thu, 13 Sep 2007, Richard Sandiford wrote:
> See files like alias.c and dse.c for why (mem:BLK (reg:SI CRIS_CP_REGNUM))

CRIS_SP_REGNUM, SP as in stack-pointer.

> is different from (mem:BLK (scratch)).

Yeah, looks like *I* should use that, thereby hopefully enabling
scheduling of non-memory insns across the deallocation barrier.

But I really wish there was mechanism to conflict only with
stack reads and writes, not "everything".  My observations were
b.df. (before dataflow) anyway; I need to refresh them.

brgds, H-P

References:
- [Patch] MIPS: Add missing memory clobbers to atomic memory insns.
  - From: David Daney
- Re: [Patch] MIPS: Add missing memory clobbers to atomic memory insns.
  - From: Richard Sandiford
- Re: [Patch] MIPS: Add missing memory clobbers to atomic memory insns.
  - From: Hans-Peter Nilsson
- Re: [Patch] MIPS: Add missing memory clobbers to atomic memory insns.
  - From: Richard Sandiford

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