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] fix prefetch optimization in loop.c

From: Richard Henderson <rth at redhat dot com>
To: Janis Johnson <janis187 at us dot ibm dot com>
Cc: gcc-patches at gcc dot gnu dot org
Date: Mon, 21 Jan 2002 16:45:57 -0800
Subject: Re: [PATCH] fix prefetch optimization in loop.c
References: <20020121111114.C1890@us.ibm.com> <20020121162047.S19335@redhat.com> <20020121163132.A3505@us.ibm.com>

On Mon, Jan 21, 2002 at 04:31:32PM -0800, Janis Johnson wrote:
> emit_prefetch_instructions() was pulling out the CONST_INT and
> combining it with other index values, but if it decided not to generate
> a prefetch, the original rtx was changed with no good way to put it
> back the way it was.

Out of curiosity, what was emit_prefetch_instructions replacing
the CONST_INT with?  (const (symbol_ref)) isn't happy, neither
is (const (plus (symbol_ref) (const_int 0))).

Looks like remove_constant_addition needs to be smarter about
handling CONST.

r~

Follow-Ups:
- Re: [PATCH] fix prefetch optimization in loop.c
  - From: Janis Johnson
- Re: [PATCH] fix prefetch optimization in loop.c
  - From: Jan Hubicka

References:
- [PATCH] fix prefetch optimization in loop.c
  - From: Janis Johnson
- Re: [PATCH] fix prefetch optimization in loop.c
  - From: Richard Henderson
- Re: [PATCH] fix prefetch optimization in loop.c
  - From: Janis Johnson

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