This is the mail archive of the
gcc-patches@gcc.gnu.org
mailing list for the GCC project.
Re: [autovect] [patch] and aliasing question
- From: Ira Rosen <IRAR at il dot ibm dot com>
- To: Ira Rosen <IRAR at il dot ibm dot com>
- Cc: Daniel Berlin <dberlin at dberlin dot org>, Diego Novillo <dnovillo at redhat dot com>, gcc-patches at gcc dot gnu dot org
- Date: Thu, 26 May 2005 15:31:45 +0300
- Subject: Re: [autovect] [patch] and aliasing question
Ira Rosen/Haifa/IBM wrote on 26/05/2005 15:18:01:
> Diego Novillo <dnovillo@redhat.com> wrote on 26/05/2005 14:02:10:
>
> > On Thu, May 26, 2005 at 12:21:24PM +0300, Ira Rosen wrote:
> >
> > > vect-77v.c (attached)
> > > ib_16 doesn't have a TMT
> > >
> > Because it is never dereferenced. We only create memory tags for
> > dereferenced pointers. The aliaser decides that it is
> > pointless to compute aliasing information for a non-dereferenced
> > pointer, because there are no loads/stores to the pointed-to
> > memory.
> >
> > Is this a problem for your analysis? Do you need points-to
> > information for non-dereferenced pointers as well?
>
> We get to ib_16 with our data-ref analysis for *D.1607_25:
>
> D.1607_25 = ib_16 + D.1606_24;
> D.1608_26 = *D.1607_25;
>
> I guess in such cases, we can get the tag from the original reference.
>
> Thanks,
> Ira
BTW, the original pointer also has no flow-sensitive information.
Ira