This is the mail archive of the
mailing list for the GCC project.
Re: vectorization test changes for alpha
- From: Richard Henderson <rth at redhat dot com>
- To: Janis Johnson <janis187 at us dot ibm dot com>
- Cc: gcc-patches at gcc dot gnu dot org, dorit at il dot ibm dot com
- Date: Tue, 23 Nov 2004 15:53:12 -0800
- Subject: Re: vectorization test changes for alpha
- References: <20041123222010.GA28344@redhat.com> <20041123233108.GA7528@us.ibm.com>
On Tue, Nov 23, 2004 at 03:31:08PM -0800, Janis Johnson wrote:
> On Tue, Nov 23, 2004 at 02:20:10PM -0800, Richard Henderson wrote:
> > Janis, you might consider adding some amount of infrastructure such that
> > the bulk of the check-for-hardware test can be shared between various
> > targets. Certainly I think that it's a mistake for only powerpc and
> > alpha to have such a test; I think x86_64 is the only target for which
> > all versions of the processor have vector instructions.
> I have a plan to make this simpler but no one has looked at it yet. See
> http://gcc.gnu.org/ml/gcc-patches/2004-11/msg01507.html ...
Um, this is something else entirely. I'm talking about the *specific*
problem of determining if vector operations are available for the target,
and how to avoid mucho code duplication for each.