This is the mail archive of the
gcc@gcc.gnu.org
mailing list for the GCC project.
Re: PRE (sometimes) confuses ivopts/scev?
- From: Daniel Berlin <dberlin at dberlin dot org>
- To: Richard Guenther <rguenth at tat dot physik dot uni-tuebingen dot de>
- Cc: Ulrich Weigand <uweigand at de dot ibm dot com>, gcc at gcc dot gnu dot org
- Date: Tue, 25 Jan 2005 10:33:47 -0500
- Subject: Re: PRE (sometimes) confuses ivopts/scev?
- References: <Pine.LNX.4.44.0501251546090.16464-100000@alwazn.tat.physik.uni-tuebingen.de>
On Tue, 2005-01-25 at 15:46 +0100, Richard Guenther wrote:
> On Tue, 25 Jan 2005, Daniel Berlin wrote:
>
> > >>> if (firstinsideloop ^ secondinsideloop)
> > >>
> > >> This is pretty much what i'm testing now.
> > >
> > > This helps - though for the particular function there is now
> > >
> > > Insertions:0
> > > New PHIs:0
> > > Eliminated:0
> > > Constified:0
> > >
> > > but I didn't expect any obvious redundant computations in the
> > > function anyway.
> >
> > However, this patch causes us to miss the invariant code in ssa-pre-3.c,
> > so we need *some* test there that avoids wraparounds while still catching
> > invariant code.
>
> I created PR tree-optimization/19624 with my testcase (took a day to
> reduce...).
>
Okay, i think i've got it going now.
Can you two test a patch for me if i send it to you?