public inbox for gcc-patches@gcc.gnu.org
 help / color / mirror / Atom feed
From: Jonathan Wakely <jwakely@redhat.com>
To: libstdc++@gcc.gnu.org, gcc-patches@gcc.gnu.org
Subject: [PATCH] libstdc++: Add workaround to std::ranges::subrange [PR111948]
Date: Thu, 30 Nov 2023 15:50:50 +0000	[thread overview]
Message-ID: <20231130155155.74430-1-jwakely@redhat.com> (raw)

I think I'll push this to work around the compiler bug. We can revert it
later if the front end gets fixed.

Tested x86_64-linux. Needed on trunk and gcc-13.

-- >8 --

libstdc++-v3/ChangeLog:

	PR libstdc++/111948
	* include/bits/ranges_util.h (subrange): Add constructor to
	_Size to aoid setting member in constructor.
	* testsuite/std/ranges/subrange/111948.cc: New test.
---
 libstdc++-v3/include/bits/ranges_util.h       | 21 ++++++++++++-------
 .../testsuite/std/ranges/subrange/111948.cc   |  8 +++++++
 2 files changed, 21 insertions(+), 8 deletions(-)
 create mode 100644 libstdc++-v3/testsuite/std/ranges/subrange/111948.cc

diff --git a/libstdc++-v3/include/bits/ranges_util.h b/libstdc++-v3/include/bits/ranges_util.h
index ab6c69c57d0..185e46ec7a9 100644
--- a/libstdc++-v3/include/bits/ranges_util.h
+++ b/libstdc++-v3/include/bits/ranges_util.h
@@ -267,13 +267,21 @@ namespace ranges
       using __size_type
 	= __detail::__make_unsigned_like_t<iter_difference_t<_It>>;
 
-      template<typename, bool = _S_store_size>
+      template<typename _Tp, bool = _S_store_size>
 	struct _Size
-	{ };
+	{
+	  [[__gnu__::__always_inline__]]
+	  constexpr _Size(_Tp = {}) { }
+	};
 
       template<typename _Tp>
 	struct _Size<_Tp, true>
-	{ _Tp _M_size; };
+	{
+	  [[__gnu__::__always_inline__]]
+	  constexpr _Size(_Tp __s = {}) : _M_size(__s) { }
+
+	  _Tp _M_size;
+	};
 
       [[no_unique_address]] _Size<__size_type> _M_size = {};
 
@@ -294,11 +302,8 @@ namespace ranges
       noexcept(is_nothrow_constructible_v<_It, decltype(__i)>
 	       && is_nothrow_constructible_v<_Sent, _Sent&>)
 	requires (_Kind == subrange_kind::sized)
-      : _M_begin(std::move(__i)), _M_end(__s)
-      {
-	if constexpr (_S_store_size)
-	  _M_size._M_size = __n;
-      }
+      : _M_begin(std::move(__i)), _M_end(__s), _M_size(__n)
+      { }
 
       template<__detail::__different_from<subrange> _Rng>
 	requires borrowed_range<_Rng>
diff --git a/libstdc++-v3/testsuite/std/ranges/subrange/111948.cc b/libstdc++-v3/testsuite/std/ranges/subrange/111948.cc
new file mode 100644
index 00000000000..dcc64b56def
--- /dev/null
+++ b/libstdc++-v3/testsuite/std/ranges/subrange/111948.cc
@@ -0,0 +1,8 @@
+// { dg-do compile { target c++20 } }
+
+#include <ranges>
+
+// Bug libstdc++/111948 - subrange modifies a const size object
+
+constexpr auto r = std::ranges::subrange(std::views::iota(0), 5);
+static_assert(std::ranges::distance(r));
-- 
2.43.0


             reply	other threads:[~2023-11-30 15:52 UTC|newest]

Thread overview: 2+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2023-11-30 15:50 Jonathan Wakely [this message]
2023-12-05 23:35 ` 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=20231130155155.74430-1-jwakely@redhat.com \
    --to=jwakely@redhat.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).