From: Paolo Carlini <paolo.carlini@oracle.com>
To: "François Dumont" <frs.dumont@gmail.com>
Cc: "libstdc++@gcc.gnu.org" <libstdc++@gcc.gnu.org>,
gcc-patches <gcc-patches@gcc.gnu.org>
Subject: Re: PR 58148 patch
Date: Fri, 30 Aug 2013 09:47:00 -0000 [thread overview]
Message-ID: <52206964.5040406@oracle.com> (raw)
In-Reply-To: <521FA314.1040702@gmail.com>
[-- Attachment #1: Type: text/plain, Size: 1096 bytes --]
Hi,
On 08/29/2013 09:37 PM, François Dumont wrote:
> Indeed, I check the Standard about const_pointer, so here is another
> attempt.
>
> Tested under Linux x86_64.
>
> 2013-08-29 François Dumont <fdumont@gcc.gnu.org>
>
> PR libstdc++/58148
> * include/debug/functions.h (__foreign_iterator_aux4): Use
> sequence const_pointer as common type to compare pointers. Add a
> fallback overload in case pointers cannot be cast to sequence
> const_pointer.
> * testsuite/23_containers/vector/modifiers/insert/58148.cc: New.
>
> Ok to commit ?
Seems pretty good to me. I have been thinking: when the non-trivial
__foreign_iterator_aux4 is selected it actually has available as the
last two arguments
std::addressof(*(__it._M_get_sequence()->_M_base().begin()))
std::addressof(*__other)
we could as well give the parameters names and avoid passing __other.
Also, I think we can do everything with std::less. I'm attaching below
something I quickly hacked, untested, see if you like it in case commit
something similar.
Thanks!
Paolo.
//////////////////////
[-- Attachment #2: p --]
[-- Type: text/plain, Size: 2917 bytes --]
Index: functions.h
===================================================================
--- functions.h (revision 202068)
+++ functions.h (working copy)
@@ -36,7 +36,7 @@
#include <bits/move.h> // for __addressof and addressof
#if __cplusplus >= 201103L
# include <bits/stl_function.h> // for less and greater_equal
-# include <type_traits> // for common_type
+# include <type_traits> // for is_lvalue_reference and __and_
#endif
#include <debug/formatter.h>
@@ -172,27 +172,28 @@
}
#if __cplusplus >= 201103L
- template<typename _Iterator, typename _Sequence,
- typename _InputIterator,
- typename _PointerType1,
- typename _PointerType2>
+ // Default implementation.
+ template<typename _Iterator, typename _Sequence>
inline bool
__foreign_iterator_aux4(const _Safe_iterator<_Iterator, _Sequence>& __it,
- _InputIterator __other,
- _PointerType1, _PointerType2)
+ typename _Sequence::const_pointer __begin,
+ typename _Sequence::const_pointer __other)
{
- typedef typename std::common_type<_PointerType1,
- _PointerType2>::type _PointerType;
+ typedef typename _Sequence::const_pointer _PointerType;
constexpr std::less<_PointerType> __l{};
- constexpr std::greater_equal<_PointerType> __ge{};
- return (__l(std::addressof(*__other),
- std::addressof(*(__it._M_get_sequence()->_M_base().begin())))
- || __ge(std::addressof(*__other),
- std::addressof(*(__it._M_get_sequence()->_M_base().end()
- - 1)) + 1));
+ return (__l(__other, __begin)
+ || __l(std::addressof(*(__it._M_get_sequence()->_M_base().end()
+ - 1)), __other));
}
-
+
+ // Fallback when address type cannot be implicitely casted to sequence
+ // const_pointer.
+ template<typename _Iterator, typename _Sequence>
+ inline bool
+ __foreign_iterator_aux4(const _Safe_iterator<_Iterator, _Sequence>&, ...)
+ { return true; }
+
template<typename _Iterator, typename _Sequence, typename _InputIterator>
inline bool
__foreign_iterator_aux3(const _Safe_iterator<_Iterator, _Sequence>& __it,
@@ -209,12 +210,12 @@
- std::addressof(*(__it._M_get_sequence()->_M_base().begin()))
== __it._M_get_sequence()->size() - 1)
return (__foreign_iterator_aux4
- (__it, __other,
+ (__it,
std::addressof(*(__it._M_get_sequence()->_M_base().begin())),
std::addressof(*__other)));
return true;
}
-
+
/* Fallback overload for which we can't say, assume it is valid. */
template<typename _Iterator, typename _Sequence, typename _InputIterator>
inline bool
@@ -223,7 +224,7 @@
std::false_type)
{ return true; }
#endif
-
+
/** Checks that iterators do not belong to the same sequence. */
template<typename _Iterator, typename _Sequence, typename _OtherIterator>
inline bool
next prev parent reply other threads:[~2013-08-30 9:44 UTC|newest]
Thread overview: 3+ messages / expand[flat|nested] mbox.gz Atom feed top
[not found] <521BB025.3040700@gmail.com>
[not found] ` <521BBD4F.1040300@oracle.com>
[not found] ` <521D0D5A.3060201@gmail.com>
[not found] ` <521D1001.3080504@oracle.com>
[not found] ` <CAGNvRgBaXi1LYWC8Gq2p20MO9vrTigZhw7HNBZu-X5pit93vBg@mail.gmail.com>
[not found] ` <521E4FD8.9030201@gmail.com>
[not found] ` <521E627A.4060502@oracle.com>
2013-08-29 19:40 ` François Dumont
2013-08-30 9:47 ` Paolo Carlini [this message]
2013-08-30 21:01 ` François Dumont
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=52206964.5040406@oracle.com \
--to=paolo.carlini@oracle.com \
--cc=frs.dumont@gmail.com \
--cc=gcc-patches@gcc.gnu.org \
--cc=libstdc++@gcc.gnu.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).