This is the mail archive of the
gcc-patches@gcc.gnu.org
mailing list for the GCC project.
Re: [PATCH, rs6000] Add missing logical-op interfaces to emmintrin.h
- From: Segher Boessenkool <segher at kernel dot crashing dot org>
- To: Bill Schmidt <wschmidt at linux dot ibm dot com>
- Cc: GCC Patches <gcc-patches at gcc dot gnu dot org>, David Edelsohn <dje dot gcc at gmail dot com>
- Date: Fri, 13 Jul 2018 12:21:20 -0500
- Subject: Re: [PATCH, rs6000] Add missing logical-op interfaces to emmintrin.h
- References: <a57f8d4c-bb20-5c1b-72ac-694b34f7b750@linux.ibm.com>
Hi!
On Wed, Jul 11, 2018 at 12:26:24PM -0500, Bill Schmidt wrote:
> It was recently brought to our attention that the existing emmintrin.h
> header, which was believed to be feature-complete for SSE2 support, is
> actually missing four logical-op interfaces:
>
> _mm_and_si128
> _mm_andnot_si128
> _mm_or_si128
> _mm_xor_si128
>
> This patch provides those with the obvious implementations, along with
> test cases. I've bootstrapped it on powerpc64le-linux-gnu (P8, P9)
> and powerpc64-linux-gnu (P7, P8) and tested it with no regressions.
> Is this okay for trunk?
>
> Although this isn't a regression, it is an oversight that leaves the
> SSE2 support incomplete. Thus I'd like to ask permission to also
> backport this to gcc-8-branch after a short waiting period. It's
> passed regstrap on P8 and P9 LE, and P7/P8 BE testing is underway.
> Is that backport okay if testing succeeds?
>
> [BTW, I'm shepherding this patch on behalf of Steve Munroe.]
This looks fine. Okay for trunk. Also okay for 8 (as we discussed, you
probably should check if 8 hasn't diverged from trunk here; it shouldn't
have).
Thanks to both of you,
Segher