From: Patrick Palka <ppalka@redhat.com>
To: Tim Song <t.canens.cpp@gmail.com>
Cc: Patrick Palka <ppalka@redhat.com>,
gcc-patches <gcc-patches@gcc.gnu.org>,
libstdc++ <libstdc++@gcc.gnu.org>
Subject: Re: [PATCH] libstdc++: Avoid hard error in ranges::unique_copy [PR100770]
Date: Thu, 27 May 2021 09:50:09 -0400 (EDT) [thread overview]
Message-ID: <bfedf2da-e0cb-491-559f-1d99e8c7651@idea> (raw)
In-Reply-To: <CAPQZVxs5dKajpF=1JH_unJkpsr_+RXr0a8U46LOAX9Wkk1TGwg@mail.gmail.com>
On Wed, 26 May 2021, Tim Song wrote:
> On Wed, May 26, 2021 at 1:43 PM Patrick Palka <ppalka@redhat.com> wrote:
> >
> > On Wed, 26 May 2021, Tim Song wrote:
> > >
> > > On Wed, May 26, 2021 at 12:00 PM Patrick Palka via Libstdc++
> > > <libstdc++@gcc.gnu.org> wrote:
> > > >
> > > > - else if constexpr (input_iterator<_Out>
> > > > - && same_as<iter_value_t<_Iter>, iter_value_t<_Out>>)
> > > > + else if constexpr (requires { requires (input_iterator<_Out>
> > > > + && same_as<iter_value_t<_Iter>,
> > > > + iter_value_t<_Out>>); })
> > >
> > > It's arguably cleaner to extract this into a concept which can then
> > > also be used in the constraint.
> >
> > Sounds good, though I'm not sure what name to give to this relatively
> > ad-hoc set of requirements. Any suggestions? :)
> >
>
> Something along the lines of "__can_reread_output", perhaps?
Works for me. Here's v2, which factors out the condition into a concept
and defines the value_type, pointer and reference of output_iterator_wrapper
to void but leaves alone its difference_type. Tested on x86_64-pc-linux-gnu.
-- >8 --
Subject: [PATCH] libstdc++: Avoid hard error in ranges::unique_copy [PR100770]
Here, in the constexpr if condition within unique_copy, when
input_iterator<_Out> isn't satisfied we must avoid substituting into
iter_value_t<_Out> because the latter isn't necessarily well-formed in
this case. To that end, this patch factors out the condition into a
concept and uses it throughout.
This patch also makes the definition of our testsuite
output_iterator_wrapper more minimal by defining its value_type, pointer
and reference member types to void. This means our existing tests for
unique_copy already exercise the fix for this bug, so we don't need
to add another test. The only other fallout of this testsuite iterator
change appears in std/ranges/range.cc, where the use of range_value_t
on a test_output_range is now ill-formed.
libstdc++-v3/ChangeLog:
* include/bits/ranges_algo.h (__detail::__can_reread_output):
Factor out this concept from ...
(__unique_copy_fn::operator()): ... here. Use the concept
throughout.
* testsuite/std/ranges/range.cc: Remove now ill-formed use
of range_value_t on an output_range.
* testsuite/util/testsuite_iterators.h (output_iterator_wrapper):
Define value_type, pointer and reference member types to void.
---
libstdc++-v3/include/bits/ranges_algo.h | 16 ++++++++++------
libstdc++-v3/testsuite/std/ranges/range.cc | 3 ---
.../testsuite/util/testsuite_iterators.h | 2 +-
3 files changed, 11 insertions(+), 10 deletions(-)
diff --git a/libstdc++-v3/include/bits/ranges_algo.h b/libstdc++-v3/include/bits/ranges_algo.h
index cda3042c11f..ecf1378742d 100644
--- a/libstdc++-v3/include/bits/ranges_algo.h
+++ b/libstdc++-v3/include/bits/ranges_algo.h
@@ -1396,6 +1396,13 @@ namespace ranges
inline constexpr __unique_fn unique{};
+ namespace __detail
+ {
+ template<typename _Out, typename _Tp>
+ concept __can_reread_output = input_iterator<_Out>
+ && same_as<_Tp, iter_value_t<_Out>>;
+ }
+
template<typename _Iter, typename _Out>
using unique_copy_result = in_out_result<_Iter, _Out>;
@@ -1407,8 +1414,7 @@ namespace ranges
projected<_Iter, _Proj>> _Comp = ranges::equal_to>
requires indirectly_copyable<_Iter, _Out>
&& (forward_iterator<_Iter>
- || (input_iterator<_Out>
- && same_as<iter_value_t<_Iter>, iter_value_t<_Out>>)
+ || __detail::__can_reread_output<_Out, iter_value_t<_Iter>>
|| indirectly_copyable_storable<_Iter, _Out>)
constexpr unique_copy_result<_Iter, _Out>
operator()(_Iter __first, _Sent __last, _Out __result,
@@ -1432,8 +1438,7 @@ namespace ranges
}
return {__next, std::move(++__result)};
}
- else if constexpr (input_iterator<_Out>
- && same_as<iter_value_t<_Iter>, iter_value_t<_Out>>)
+ else if constexpr (__detail::__can_reread_output<_Out, iter_value_t<_Iter>>)
{
*__result = *__first;
while (++__first != __last)
@@ -1467,8 +1472,7 @@ namespace ranges
projected<iterator_t<_Range>, _Proj>> _Comp = ranges::equal_to>
requires indirectly_copyable<iterator_t<_Range>, _Out>
&& (forward_iterator<iterator_t<_Range>>
- || (input_iterator<_Out>
- && same_as<range_value_t<_Range>, iter_value_t<_Out>>)
+ || __detail::__can_reread_output<_Out, range_value_t<_Range>>
|| indirectly_copyable_storable<iterator_t<_Range>, _Out>)
constexpr unique_copy_result<borrowed_iterator_t<_Range>, _Out>
operator()(_Range&& __r, _Out __result,
diff --git a/libstdc++-v3/testsuite/std/ranges/range.cc b/libstdc++-v3/testsuite/std/ranges/range.cc
index 795aca472e5..aa29af471a4 100644
--- a/libstdc++-v3/testsuite/std/ranges/range.cc
+++ b/libstdc++-v3/testsuite/std/ranges/range.cc
@@ -75,9 +75,6 @@ static_assert( same_as<std::ranges::range_difference_t<C>,
static_assert( same_as<std::ranges::range_difference_t<O>,
std::ptrdiff_t> );
-static_assert( same_as<std::ranges::range_value_t<O>,
- char> );
-
static_assert( same_as<std::ranges::range_reference_t<I>,
char&> );
static_assert( same_as<std::ranges::range_reference_t<O>,
diff --git a/libstdc++-v3/testsuite/util/testsuite_iterators.h b/libstdc++-v3/testsuite/util/testsuite_iterators.h
index 67da06b8482..4e668d648f2 100644
--- a/libstdc++-v3/testsuite/util/testsuite_iterators.h
+++ b/libstdc++-v3/testsuite/util/testsuite_iterators.h
@@ -122,7 +122,7 @@ namespace __gnu_test
*/
template<class T>
struct output_iterator_wrapper
- : public std::iterator<std::output_iterator_tag, T, std::ptrdiff_t, T*, T&>
+ : public std::iterator<std::output_iterator_tag, void, std::ptrdiff_t, void, void>
{
protected:
output_iterator_wrapper() : ptr(0), SharedInfo(0)
--
2.32.0.rc0
next prev parent reply other threads:[~2021-05-27 13:50 UTC|newest]
Thread overview: 9+ messages / expand[flat|nested] mbox.gz Atom feed top
2021-05-26 16:45 Patrick Palka
2021-05-26 17:27 ` Tim Song
2021-05-26 18:43 ` Patrick Palka
2021-05-26 19:07 ` Tim Song
2021-05-27 13:50 ` Patrick Palka [this message]
2021-06-03 11:54 ` Jonathan Wakely
2021-05-26 19:55 ` Jonathan Wakely
2021-05-27 0:38 ` Tim Song
2021-05-26 17:28 ` 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=bfedf2da-e0cb-491-559f-1d99e8c7651@idea \
--to=ppalka@redhat.com \
--cc=gcc-patches@gcc.gnu.org \
--cc=libstdc++@gcc.gnu.org \
--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).