This is the mail archive of the
gcc@gcc.gnu.org
mailing list for the GCC project.
Re: vectorizer data dependency graph
- From: Dorit Nuzman <DORIT at il dot ibm dot com>
- To: "Sebastian Pop" <sebastian dot pop at inria dot fr>
- Cc: "Daniel Berlin" <dberlin at dberlin dot org>, gcc at gcc dot gnu dot org, "Sashan Govender" <sashang at gmail dot com>, "Georges-André SILBER" <silber at cri dot ensmp dot fr>, "Steven Bosscher" <stevenb dot gcc at gmail dot com>
- Date: Thu, 16 Nov 2006 18:20:53 +0200
- Subject: Re: vectorizer data dependency graph
> On 11/15/06, Steven Bosscher <stevenb.gcc@gmail.com> wrote:
> > On 11/15/06, Sebastian Pop <sebastian.pop@inria.fr> wrote:
> >
> > > There is a ddg in this patch if somebody wants the classic
Allen&Kennedy
> > > way to look at the dependences:
> > > http://gcc.gnu.org/wiki/OptimizationCourse?
> action=AttachFile&do=get&target=loop-distribution-patch-against-
> gcc-4.1.0-release.patch
> > >
> >
> > Any plans to merge this into the FSF trunk?
> >
>
> Georges-Andre Silber wants to submit this patch in a better shape after
> some more cleanups.
>
> Another plan is to just commit this version of the patch in a branch,
> and let everybody improve the loop distribution pass, then submit itto
trunk.
>
I'd suggest autovect-branch, except it's in a bad shape right now (we let
it go way out of synch with mainline; I plan to open a new branch as soon
as the strided-accesses-vectorization stuff is merged to mainline (next
week I guess)).
dorit
> Sebastian