This is the mail archive of the
gcc-patches@gcc.gnu.org
mailing list for the GCC project.
Re: [PATCH] Cleanup fold_rtx, 1/n
- From: Steven Bosscher <stevenb at suse dot de>
- To: Paolo Bonzini <paolo dot bonzini at lu dot unisi dot ch>
- Cc: GCC Patches <gcc-patches at gcc dot gnu dot org>, Roger Sayle <roger at eyesopen dot com>
- Date: Wed, 13 Apr 2005 17:43:31 +0200
- Subject: Re: [PATCH] Cleanup fold_rtx, 1/n
- Organization: SUSE Labs
- References: <425D3C77.201@lu.unisi.ch>
On Wednesday 13 April 2005 17:36, Paolo Bonzini wrote:
> This patch removes an unnecessary optimization in fold_rtx_subreg, that
> is (quoting a comment) "folding arithmetic in narrow modes when the
> machine only supports word-sized arithmetic". But the only relevant
> arithmetic optimizations that CSE performs are on multi-word values,
> and tree optimizers take care of the rest: indeed this part of the code
> does not trigger anymore. I tested this claim by building an i686->alpha
> cross with 4.0 as the system compiler, and looking at code coverage traces
> (attached to PR17921).
My gcov attachment to PR19721 shows that the code does not
trigger on amd64 either.
Nice cleanup, and only step 1 of n. Whee!
Gr.
Steven