Topic: Restrict Keyword in C++0x
Author: =?ISO-8859-1?Q?Falk_Tannh=E4user?= <falk.tannhauser@crf.canon.fr>
Date: Wed, 21 Oct 2009 11:57:57 CST Raw View
AJ Guillon wrote:
> I wasn't sure if there was a proposal and it had been rejected, I
> couldn't find any relevant discussions on it. Unfortunately, as I
> understand things, you really need the restrict keyword to help the
> vectorizer do a good job generating efficient code for things like SSE
> (or any other SIMD processor features).
Wasn't std::valarray supposed to address exactly these problems?
Falk
--
[ comp.std.c++ is moderated. To submit articles, try just posting with ]
[ your news-reader. If that fails, use mailto:std-c++@netlab.cs.rpi.edu]
[ --- Please see the FAQ before posting. --- ]
[ FAQ: http://www.comeaucomputing.com/csc/faq.html ]
Author: AJ Guillon <aj.guillon@gmail.com>
Date: Thu, 15 Oct 2009 22:14:37 CST Raw View
Are there plans to include the restrict keyword from C99 into C++0x?
--
[ comp.std.c++ is moderated. To submit articles, try just posting with ]
[ your news-reader. If that fails, use mailto:std-c++@netlab.cs.rpi.edu]
[ --- Please see the FAQ before posting. --- ]
[ FAQ: http://www.comeaucomputing.com/csc/faq.html ]
Author: CornedBee <wasti.redl@gmx.net>
Date: Sun, 18 Oct 2009 08:51:53 CST Raw View
On Oct 16, 6:14 am, AJ Guillon <aj.guil...@gmail.com> wrote:
> Are there plans to include the restrict keyword from C99 into C++0x?
>
Given that all through the C++0x standardization process it has
existed in C99, but there was no proposal for adding it to C++0x, I'd
say the answer is probably no.
Or to put it differently, there's no way restrict is going into C++0x
(too late), and there don't even seem to be any plans to add it to any
later versions of C++ either.
Sebastian
--
[ comp.std.c++ is moderated. To submit articles, try just posting with ]
[ your news-reader. If that fails, use mailto:std-c++@netlab.cs.rpi.edu]
[ --- Please see the FAQ before posting. --- ]
[ FAQ: http://www.comeaucomputing.com/csc/faq.html ]
Author: AJ Guillon <aj.guillon@gmail.com>
Date: Tue, 20 Oct 2009 12:00:30 CST Raw View
I wasn't sure if there was a proposal and it had been rejected, I
couldn't find any relevant discussions on it. Unfortunately, as I
understand things, you really need the restrict keyword to help the
vectorizer do a good job generating efficient code for things like SSE
(or any other SIMD processor features).
> Given that all through the C++0x standardization process it has
> existed in C99, but there was no proposal for adding it to C++0x, I'd
> say the answer is probably no.
> Or to put it differently, there's no way restrict is going into C++0x
> (too late), and there don't even seem to be any plans to add it to any
> later versions of C++ either.
>
> Sebastian
>
> --
> [ comp.std.c++ is moderated. To submit articles, try just posting with ]
> [ your news-reader. If that fails, use mailto:std-...@netlab.cs.rpi.edu]
> [ --- Please see the FAQ before posting. --- ]
> [ FAQ:http://www.comeaucomputing.com/csc/faq.html ]
--
[ comp.std.c++ is moderated. To submit articles, try just posting with ]
[ your news-reader. If that fails, use mailto:std-c++@netlab.cs.rpi.edu]
[ --- Please see the FAQ before posting. --- ]
[ FAQ: http://www.comeaucomputing.com/csc/faq.html ]
Author: Stephen Clamage <stephen.clamage@sun.com>
Date: Tue, 20 Oct 2009 15:23:07 CST Raw View
On 10/15/2009 9:14 PM, AJ Guillon wrote:
Are there plans to include the restrict keyword from C99 into C++0x?
The restrict keyword was discussed a few years ago, and the consensus
in the C++ Committee was that it was not, and probably could not be,
well specified. Having a feature in the standard that can't be tested
for conformance and that has no semantics associated with it (a
compiler can ignore restrict without changing the behavior of a valid
program) seemed inappropriate. At least, nobody pushed for the
feature.
That said, some C++ compilers implement restrict (sometimes spelled
__restrict) as an extension, to provide the expected optimization
opportunities.
---
Steve Clamage
--
[ comp.std.c++ is moderated. To submit articles, try just posting with ]
[ your news-reader. If that fails, use mailto:std-c++@netlab.cs.rpi.edu]
[ --- Please see the FAQ before posting. --- ]
[ FAQ: http://www.comeaucomputing.com/csc/faq.html ]