From: Jonathan Wakely <jwakely@redhat.com>
To: libstdc++@gcc.gnu.org
Cc: Rainer Orth <ro@CeBiTec.Uni-Bielefeld.DE>,
Jason Merrill <jason@redhat.com>,
Christophe Lyon <christophe.lyon@linaro.org>,
Andreas Schwab <schwab@linux-m68k.org>,
gcc-patches List <gcc-patches@gcc.gnu.org>
Subject: Re: RFA (libstdc++): PATCH to implement C++17 over-aligned new
Date: Fri, 16 Sep 2016 11:12:00 -0000 [thread overview]
Message-ID: <20160916105608.GO17376@redhat.com> (raw)
In-Reply-To: <alpine.DEB.2.20.1609161116130.3301@laptop-mg.saclay.inria.fr>
On 16/09/16 11:37 +0200, Marc Glisse wrote:
>On Fri, 16 Sep 2016, Jonathan Wakely wrote:
>
>>On 16/09/16 09:04 +0200, Rainer Orth wrote:
>>>Hi Jason,
>>>
>>>>OK, one more:
>>>
>>>this works just fine on both sparc-sun-solaris2.12 and
>>>i386-pc-solaris2.12.
>>>
>>>Once Jonathan's patch to heed aligned_alloc's requirement on size being
>>>a multiple of alignment is in, all is fine on Solaris.
>>
>>I've got a slightly different fix now.
>>
>>We only need to make the size a multiple of alignment for
>>aligned_alloc, however for posix_memalign we need to ensure the
>>alignment is a multiple of sizeof(void*).
>>
>>I'm testing this now (but only on x86_64 GNU/Linux where it wasn't
>>failing anyway).
>
>+ // The value of alignment shall be a power of two multiple of sizeof(void *).
>+ if (al < sizeof(void*))
>+ al = sizeof(void*);
>
>The code doesn't exactly match the comment. I can't find the
>precondition in the standard that says operator new can only be called
>on a power of 2... (maybe we can add it if it is really missing?)
[basic.align] says "Every alignment value shall be a non-negative
integral power of two." So asking operator new for any other value
doesn't make sense, but I can't find a restriction on doing so.
I was assuming we only need to ensure it's possible to use valid
alignments such as align_val_t(2) which are not valid arguments to
posix_memalign. For other values such as align_val_t(15) I was
assuming it's OK for posix_memalign to fail, so we throw bad_alloc.
If that's not the case then we need to round up all alignments that
aren't power of two multiples of sizeof(void*). I'd like to avoid
that.
>>Would using __builtin_expect (sz == 0, false) make sense? Surely it's
>>rare to try to allocate zero bytes.
>
>https://gcc.gnu.org/ml/libstdc++/2014-03/msg00001.html
>
>gcc already guesses that a test like sz == 0 is usually false (not
>with as large a probability as if you use __builtin_expect, but enough
>that the generated code is unlikely to differ). But adding
>__builtin_expect cannot hurt...
>
>Is the division (by a non-constant denominator) really necessary?
Probably not, but I've asked the committee for clarification what this
function should do when called with an invalid alignment.
>Since align has to be a power of 2, x % align should be the same as x
>& (align - 1), for instance.
Thanks, if it's UB to call it with alignments that aren't a power of
two then we can do that.
>I guess people interested in performance will do for aligned new the
>same as for the old new: provide an inline version that skips all the
>overhead to forward directly to malloc/aligned_alloc (and avoid
>questionable calls in their code).
>
>--
>Marc Glisse
next prev parent reply other threads:[~2016-09-16 10:56 UTC|newest]
Thread overview: 42+ messages / expand[flat|nested] mbox.gz Atom feed top
2016-09-08 7:10 Jason Merrill
2016-09-08 8:32 ` Marc Glisse
2016-09-08 11:18 ` Jonathan Wakely
2016-09-09 21:40 ` Jason Merrill
2016-09-10 7:03 ` Christophe Lyon
2016-09-10 10:14 ` Marc Glisse
2016-09-10 10:35 ` Jonathan Wakely
2016-09-11 9:14 ` Christophe Lyon
2016-09-11 9:55 ` Jonathan Wakely
2016-09-11 9:56 ` Jonathan Wakely
2016-09-11 10:20 ` Christophe Lyon
2016-09-11 12:09 ` Jonathan Wakely
2016-09-13 13:04 ` Szabolcs Nagy
2016-09-10 10:14 ` Jonathan Wakely
2016-09-11 7:09 ` Andreas Schwab
2016-09-12 21:13 ` Jason Merrill
2016-09-13 8:41 ` Christophe Lyon
2016-09-13 12:37 ` Andreas Schwab
2016-09-13 12:54 ` Jason Merrill
2016-09-13 13:18 ` Andreas Schwab
2016-09-13 13:21 ` Jason Merrill
2016-09-14 12:13 ` Andreas Schwab
2016-09-14 16:11 ` Christophe Lyon
2016-09-14 16:37 ` Jason Merrill
2016-09-15 10:00 ` Rainer Orth
2016-09-15 12:23 ` Christophe Lyon
2016-09-15 20:09 ` Jason Merrill
2016-09-16 7:12 ` Rainer Orth
2016-09-16 8:15 ` Christophe Lyon
2016-09-16 9:14 ` Jonathan Wakely
2016-09-16 9:51 ` Marc Glisse
2016-09-16 11:12 ` Jonathan Wakely [this message]
2016-09-16 13:13 ` Jonathan Wakely
2016-09-16 13:17 ` Rainer Orth
2016-09-16 18:19 ` Jonathan Wakely
2016-09-12 14:15 ` Rainer Orth
2016-09-12 16:19 ` Jonathan Wakely
2016-09-12 18:57 ` Jason Merrill
2016-09-14 12:11 ` Rainer Orth
2016-09-08 11:00 ` Jonathan Wakely
2017-11-24 14:26 ` Marc Glisse
2017-11-29 21:23 ` Jason Merrill
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=20160916105608.GO17376@redhat.com \
--to=jwakely@redhat.com \
--cc=christophe.lyon@linaro.org \
--cc=gcc-patches@gcc.gnu.org \
--cc=jason@redhat.com \
--cc=libstdc++@gcc.gnu.org \
--cc=ro@CeBiTec.Uni-Bielefeld.DE \
--cc=schwab@linux-m68k.org \
/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).