This is the mail archive of the
gcc-patches@gcc.gnu.org
mailing list for the GCC project.
Re: [tree-ssa] cfg.texi needs reviewing by a native speaker (Was: Re: "Documentation by paper")
- From: Jan Hubicka <jh at suse dot cz>
- To: law at redhat dot com
- Cc: Steven Bosscher <stevenb at suse dot de>, Robert Dewar <dewar at gnat dot com>,Richard Kenner <kenner at vlsi1 dot ultra dot nyu dot edu>, gcc-patches at gcc dot gnu dot org,jh at suse dot cz, dnovillo at redhat dot com
- Date: Fri, 6 Feb 2004 00:14:58 +0100
- Subject: Re: [tree-ssa] cfg.texi needs reviewing by a native speaker (Was: Re: "Documentation by paper")
- References: <200402042105.28601.stevenb@suse.de> <200402042237.i14MbCDs003070@speedy.slc.redhat.com>
>
> Funny -- I was looking for some of this just the other day (frequency stuff).
>
> Now what we need is better documentation of things like the algorithms we're
> using to do prediction & frequency computation/propagation. :-)
Also since you seem to be poking around the stuff in tree-ssa, you shall
be warned that the predictions are poorer than on mainline.
The tree predictors are very limited at the moment (doing just opcode
based predictions and very basic loop predictions) and gimplification
broke high level predictors in RTL, so they will have worse results too.
I don't have exact numbers, but if you think the predictors are problem,
you can use analyze_brprob script in the contrib directory.
I do have plans to get this back into full funcitonality, but I need the
CFG transparency stuff first so I need to do highlevel predictions only
on tree level.
Honza
>
>
> Jeff
>