This is the mail archive of the
gcc-patches@gcc.gnu.org
mailing list for the GCC project.
Re: [tree-ssa] Simple value range propagation for tree-ssa-dom
- From: Toon Moene <toon at moene dot indiv dot nluug dot nl>
- To: law at redhat dot com
- Cc: Andrew Haley <aph at redhat dot com>, gcc-patches at gcc dot gnu dot org
- Date: Wed, 01 Oct 2003 22:47:15 +0200
- Subject: Re: [tree-ssa] Simple value range propagation for tree-ssa-dom
- Organization: Moene Computational Physics, Maartensdijk, The Netherlands
- References: <200310011744.h91Hi6q7004911@speedy.slc.redhat.com>
law@redhat.com wrote:
In message <16250.46575.720716.405975@cuddles.cambridge.redhat.com>,
Andrew Haley wrote:
> for (j=0; j<10;j++)
> {
> if (j < 0)
> abort();
> if (j > 10)
> abort();
> m+=a[j];
>
We don't currently merge ranges at PHI nodes, nor do we track ranges for
outputs of statements within a basic block. To optimize a loop like that
we have to merge ranges as PHI nodes and possibly look at statemetns within
blocks to build ranges for their outputs.
It's on the todo list, but it's after some other (non-VRP) work.
That would be great, as it would speed up g77/g95 -fbounds-check code
enormously.
Cheers,
--
Toon Moene - mailto:toon@moene.indiv.nluug.nl - phoneto: +31 346 214290
Saturnushof 14, 3738 XG Maartensdijk, The Netherlands
Maintainer, GNU Fortran 77: http://gcc.gnu.org/onlinedocs/g77_news.html
GNU Fortran 95: http://gcc-g95.sourceforge.net/ (under construction)