This is the mail archive of the
gcc-patches@gcc.gnu.org
mailing list for the GCC project.
Re: [PATCH][libgcc-math] Vectorized intrinsics for x86_64
- From: Richard Guenther <rguenther at suse dot de>
- To: Richard Henderson <rth at redhat dot com>
- Cc: gcc-patches at gcc dot gnu dot org, Michael Matz <matz at suse dot de>
- Date: Fri, 7 Apr 2006 09:32:12 +0200 (CEST)
- Subject: Re: [PATCH][libgcc-math] Vectorized intrinsics for x86_64
- References: <Pine.LNX.4.64.0603281022000.3982@t148.fhfr.qr> <20060331155757.GB15017@redhat.com> <Pine.LNX.4.64.0604041737170.3938@t148.fhfr.qr> <20060406174426.GA17231@redhat.com>
On Thu, 6 Apr 2006, Richard Henderson wrote:
> On Tue, Apr 04, 2006 at 05:40:07PM +0200, Richard Guenther wrote:
> > Here's an updated patch with your suggestions applied (but with .align 16,
> > to have it cacheline aligned), like so
> >
> > .section .rodata.cst16, "M", @progbits, 16
> > .align 16
>
> Not ok for 8 byte data.
I believe it's all 16byte data, i.e. always full sse registers with
either 2 doubles or 4 floats. Matz did the annotation, so I guess
he knows better if there are any partial loads from these data (it
might be f.i. the .L__two_to_jby32_table - which is seemingly accessed
in 4 byte quantities f.i.)
> And you didn't attach the patch.
Yeah - it was in the followup :/
Richard.