public inbox for libstdc++@gcc.gnu.org
 help / color / mirror / Atom feed
From: Jonathan Wakely <jwakely@redhat.com>
To: Patrick Palka <ppalka@redhat.com>
Cc: Tim Song <t.canens.cpp@gmail.com>,
	libstdc++ <libstdc++@gcc.gnu.org>,
	gcc-patches <gcc-patches@gcc.gnu.org>
Subject: Re: [PATCH] libstdc++: Implement P2328 changes to join_view
Date: Fri, 30 Apr 2021 23:01:14 +0100	[thread overview]
Message-ID: <20210430220114.GE3008@redhat.com> (raw)
In-Reply-To: <21f9ca23-18bb-fafe-4028-bd5fc1f4d3c@idea>

On 30/04/21 17:34 -0400, Patrick Palka via Libstdc++ wrote:
>On Fri, 30 Apr 2021, Tim Song wrote:
>
>> On Fri, Apr 30, 2021 at 12:11 PM Patrick Palka via Libstdc++
>> <libstdc++@gcc.gnu.org> wrote:
>> >
>> > +       template<typename _Iter>
>> > +         _Tp&
>> > +         _M_emplace_deref(const _Iter& __i)
>> > +         {
>> > +           this->reset();
>> > +           return this->emplace(*__i);
>> > +         }
>>
>> This incurs a move, avoiding of which is the sole reason for
>> emplace-deref's existence.
>
>Ah thanks, I had missed that...  IIUC, if we instead derive from
>optional's internal base class _Optional_base, we can easily get at the
>underlying storage for the object and directly initialize it with '*__i'
>and avoid the move. How does the the following adjustment to the patch
>look?

Looks good to me, thanks.

>-- >8 --
>
>diff --git a/libstdc++-v3/include/std/ranges b/libstdc++-v3/include/std/ranges
>
>index 971898f5492..f1db4a238ef 100644
>--- a/libstdc++-v3/include/std/ranges
>+++ b/libstdc++-v3/include/std/ranges
>@@ -2254,7 +2254,7 @@ namespace views::__adaptor
>
>     template<typename _Tp>
>       requires is_object_v<_Tp>
>-      struct __non_propagating_cache<_Tp> : private optional<_Tp>
>+      struct __non_propagating_cache<_Tp> : private _Optional_base<_Tp>
>       {
> 	__non_propagating_cache() = default;
>
>@@ -2264,32 +2264,36 @@ namespace views::__adaptor
>
> 	constexpr
> 	__non_propagating_cache(__non_propagating_cache&& __other) noexcept
>-	{ __other.reset(); }
>+	{ __other._M_reset(); }
>
> 	constexpr __non_propagating_cache&
> 	operator=(const __non_propagating_cache& __other) noexcept
> 	{
> 	  if (std::__addressof(__other) != this)
>-	    this->reset();
>+	    this->_M_reset();
> 	  return *this;
> 	}
>
> 	constexpr __non_propagating_cache&
> 	operator=(__non_propagating_cache&& __other) noexcept
> 	{
>-	  this->reset();
>-	  __other.reset();
>+	  this->_M_reset();
>+	  __other._M_reset();
> 	  return *this;
> 	}
>
>-	using optional<_Tp>::operator*;
>+	constexpr _Tp&
>+	operator*() noexcept
>+	{ return this->_M_get(); }
>
> 	template<typename _Iter>
> 	  _Tp&
> 	  _M_emplace_deref(const _Iter& __i)
> 	  {
>-	    this->reset();
>-	    return this->emplace(*__i);
>+	    this->_M_reset();
>+	    ::new ((void *) std::__addressof(this->_M_payload._M_payload)) _Tp(*__i);
>+	    this->_M_payload._M_engaged = true;
>+	    return this->_M_get();
> 	  }
>       };
>   }
>


  reply	other threads:[~2021-04-30 22:01 UTC|newest]

Thread overview: 7+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2021-04-30 14:38 Patrick Palka
2021-04-30 16:59 ` Jonathan Wakely
2021-04-30 17:43 ` Tim Song
2021-04-30 21:34   ` Patrick Palka
2021-04-30 22:01     ` Jonathan Wakely [this message]
2021-04-30 22:51       ` Patrick Palka
2021-04-30 22:02   ` Jonathan Wakely

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=20210430220114.GE3008@redhat.com \
    --to=jwakely@redhat.com \
    --cc=gcc-patches@gcc.gnu.org \
    --cc=libstdc++@gcc.gnu.org \
    --cc=ppalka@redhat.com \
    --cc=t.canens.cpp@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).