public inbox for gcc-cvs@sourceware.org help / color / mirror / Atom feed
From: Jonathan Wakely <redi@gcc.gnu.org> To: gcc-cvs@gcc.gnu.org, libstdc++-cvs@gcc.gnu.org Subject: [gcc r11-9908] libstdc++: Ignore cv-quals when std::allocator<void> constructs Date: Thu, 21 Apr 2022 12:33:08 +0000 (GMT) [thread overview] Message-ID: <20220421123308.BEFD73888C4E@sourceware.org> (raw) https://gcc.gnu.org/g:7bb5e5768b79594266468f71ee61c3dfd2d3c3bd commit r11-9908-g7bb5e5768b79594266468f71ee61c3dfd2d3c3bd Author: Jonathan Wakely <jwakely@redhat.com> Date: Thu Jan 13 21:59:13 2022 +0000 libstdc++: Ignore cv-quals when std::allocator<void> constructs When I added the std::allocator_traits<std::allocator<void>> specialization it broke code like this: std::allocate_shared<const int>(std::allocator<void>()); The problem is that allocator_traits<allocator<void>>::construct(a, p) now uses std::_Construct(p), which only does a static_cast<void*>(p) and so fails if the pointer has cv-quals. This changes std::_Construct (and the related std::_Construct_novalue) to use a C-style cast to (void*) which matches the effects of the "voidify" helper in the C++20 standard. libstdc++-v3/ChangeLog: * include/bits/stl_construct.h (_Construct, _Construct_novalue): Also cast away cv-qualifiers when converting pointer to void. * testsuite/20_util/allocator/void.cc: Test construct function with cv-qualified types. (cherry picked from commit fc6f1128ae603164aea6303ce2b3ed0b57e6a378) Diff: --- libstdc++-v3/include/bits/stl_construct.h | 4 ++-- libstdc++-v3/testsuite/20_util/allocator/void.cc | 15 +++++++++++++++ 2 files changed, 17 insertions(+), 2 deletions(-) diff --git a/libstdc++-v3/include/bits/stl_construct.h b/libstdc++-v3/include/bits/stl_construct.h index e53ed0d9f91..f14fba0cfb1 100644 --- a/libstdc++-v3/include/bits/stl_construct.h +++ b/libstdc++-v3/include/bits/stl_construct.h @@ -116,7 +116,7 @@ _GLIBCXX_BEGIN_NAMESPACE_VERSION return; } #endif - ::new(static_cast<void*>(__p)) _Tp(std::forward<_Args>(__args)...); + ::new((void*)__p) _Tp(std::forward<_Args>(__args)...); } #else template<typename _T1, typename _T2> @@ -132,7 +132,7 @@ _GLIBCXX_BEGIN_NAMESPACE_VERSION template<typename _T1> inline void _Construct_novalue(_T1* __p) - { ::new(static_cast<void*>(__p)) _T1; } + { ::new((void*)__p) _T1; } template<typename _ForwardIterator> _GLIBCXX20_CONSTEXPR void diff --git a/libstdc++-v3/testsuite/20_util/allocator/void.cc b/libstdc++-v3/testsuite/20_util/allocator/void.cc index e3d024d525b..3720774abaf 100644 --- a/libstdc++-v3/testsuite/20_util/allocator/void.cc +++ b/libstdc++-v3/testsuite/20_util/allocator/void.cc @@ -87,8 +87,23 @@ static_assert( std::is_same<std::allocator<void>::const_pointer, const void*>(), "const_pointer is const void*" ); #endif // C++20 +void +test02() +{ + std::allocator<void> av; + int* p = std::allocator<int>().allocate(1); + const int* c = p; + std::allocator_traits<std::allocator<void>>::construct(av, c, 0); + volatile int* v = p; + std::allocator_traits<std::allocator<void>>::construct(av, v, 0); + const volatile int* cv = p; + std::allocator_traits<std::allocator<void>>::construct(av, cv, 0); + std::allocator<int>().deallocate(p, 1); +} + int main() { test01(); + test02(); }
reply other threads:[~2022-04-21 12:33 UTC|newest] Thread overview: [no followups] expand[flat|nested] mbox.gz Atom feed
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=20220421123308.BEFD73888C4E@sourceware.org \ --to=redi@gcc.gnu.org \ --cc=gcc-cvs@gcc.gnu.org \ --cc=libstdc++-cvs@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: linkBe 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).