From: "Dr. Matthias Kretz" <m.kretz@gsi.de>
To: Andrew Pinski <pinskia@gmail.com>
Cc: libstdc++ <libstdc++@gcc.gnu.org>, GCC Patches <gcc-patches@gcc.gnu.org>
Subject: Re: [PATCH] std::experimental::simd
Date: Tue, 07 Jan 2020 13:19:00 -0000 [thread overview]
Message-ID: <2523999.LYGYClEg0C@proxima3> (raw)
In-Reply-To: <CA+=Sn1mTJUgtuvsvXjOi8onQpFMvGOpQOFoJ7CmfAzFOaEE2+g@mail.gmail.com>
On Dienstag, 7. Januar 2020 12:16:57 CET Andrew Pinski wrote:
> On Tue, Jan 7, 2020 at 3:01 AM Matthias Kretz <m.kretz@gsi.de> wrote:
> > Is there any chance left we can get this done for 10.1? If not, can we
> > please get it ready for 10.2 ASAP?
> >
> > Cheers,
> >
> > Matthias
> >
> > On Montag, 14. Oktober 2019 14:12:12 CET Matthias Kretz wrote:
> > > Let me try again to get this patch ready. It will need a few
> > > iterations...
> > > This patch is without documentation and testsuite. I can add them on
> > > request but would prefer a follow-up patch after getting this one
> > > right.
> > >
> > > I recommend to review starting from simd.h + simd_scalar.h, then
> > > simd_builtin.h, simd_x86.h, and simd_fixed_size.h. I sure when we get
> > > this
> > > far we are a few iterations further.
> > >
> > > Regarding the license. The license header is currently just a copy from
> > > my
> > > repo, but we can change it to the libstdc++ license. The paperwork with
> > > the
> > > FSF is done.
>
> Seems like it would be better if we put the x86 and aarch64/arm
> specific parts in their own headers.
Yes. I'm already working on it. It makes me unhappy in some of the generic
parts of the code, but I think it's still a worthwhile reorganization. Last
state is here: https://github.com/VcDevel/std-simd/tree/master/experimental/
bits
I'll prepare a new patch.
> Also all of the x86 conversion should be removed as
> __builtin_convertvector is supported now.
simd_x86_conversions.h is about PR85048 (and more missing optimizations). I'd
prefer to implement simd_x86_conversions.h in the compiler, but I'd need some
guidance. I'd like the first release of std::experimental::simd to have high
performance - because that's the main reason for using it. I'd rather wait a
release than taint the impression of its usefulness.
> libstdc++v3 is only ever supported by the version that comes with the
> compiler.
Right, that's an artifact of having active users of this code. I'll clean it
up.
Thanks for the feedback,
Matthias
--
──────────────────────────────┬────────────────────────────────────────────
Dr. Matthias Kretz │ SDE — Software Development for Experiments
Senior Software Engineer, │ 📞 +49 6159 713084
SIMD Expert, │ 📧 m.kretz@gsi.de
ISO C++ Committee Member │ 🔗 mattkretz.github.io
──────────────────────────────┴────────────────────────────────────────────
GSI Helmholtzzentrum für Schwerionenforschung GmbH
Planckstraße 1, 64291 Darmstadt, Germany, www.gsi.de
Commercial Register / Handelsregister: Amtsgericht Darmstadt, HRB 1528
Managing Directors / Geschäftsführung:
Professor Dr. Paolo Giubellino, Ursula Weyrich, Jörg Blaurock
Chairman of the Supervisory Board / Vorsitzender des GSI-Aufsichtsrats:
State Secretary / Staatssekretär Dr. Georg Schütte
next prev parent reply other threads:[~2020-01-07 13:19 UTC|newest]
Thread overview: 13+ messages / expand[flat|nested] mbox.gz Atom feed top
2019-10-14 12:12 Matthias Kretz
2019-10-15 3:52 ` Thomas Rodgers
2019-10-24 8:26 ` Dr. Matthias Kretz
2020-02-10 16:49 ` Thomas Rodgers
2020-02-10 20:14 ` Thomas Rodgers
2020-01-07 11:01 ` Matthias Kretz
2020-01-07 11:17 ` Andrew Pinski
2020-01-07 13:19 ` Dr. Matthias Kretz [this message]
[not found] ` <3486545.znU0eCzeS4@excalibur>
[not found] ` <xkqeo8qyl8y8.fsf@trodgers.remote>
2020-05-08 19:03 ` Matthias Kretz
2020-11-11 23:43 ` Jonathan Wakely
2020-11-14 1:11 ` Matthias Kretz
2020-11-15 19:11 ` Matthias Kretz
2020-12-10 21:13 ` Matthias Kretz
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=2523999.LYGYClEg0C@proxima3 \
--to=m.kretz@gsi.de \
--cc=gcc-patches@gcc.gnu.org \
--cc=libstdc++@gcc.gnu.org \
--cc=pinskia@gmail.com \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for read-only IMAP folder(s) and NNTP newsgroup(s).