public inbox for gcc-bugs@sourceware.org
help / color / mirror / Atom feed
From: "webrown.cpp at gmail dot com" <gcc-bugzilla@gcc.gnu.org>
To: gcc-bugs@gcc.gnu.org
Subject: [Bug c++/100687] New: [modules, concepts] imported concept gives different result
Date: Wed, 19 May 2021 22:56:45 +0000	[thread overview]
Message-ID: <bug-100687-4@http.gcc.gnu.org/bugzilla/> (raw)

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=100687

            Bug ID: 100687
           Summary: [modules, concepts] imported concept gives different
                    result
           Product: gcc
           Version: 12.0
            Status: UNCONFIRMED
          Severity: normal
          Priority: P3
         Component: c++
          Assignee: unassigned at gcc dot gnu.org
          Reporter: webrown.cpp at gmail dot com
  Target Milestone: ---

Using recent trunk [g++-HEAD (Homebrew GCC HEAD-5380e3c) 12.0.0 20210513],
compiling with significant flags -std=c++23 -fmodules-ts -c.

I have two source files, as follows:

//////////////////////
export module bug_a;

template< class T >
inline constexpr bool probe = false;
//
template< class R, class C >
inline constexpr bool probe<R C::* > = true;

export
template< class T >
concept mbr_ptr = probe<T>;

struct S { int f( ); };
using mf_t = decltype( &S::f );

static_assert( mbr_ptr< mf_t > );
//////////////////////

This bug_a module compiles successfully.  In particular, the static_assert
passes.

//////////////////////
export module bug_b;

import bug_a;

struct S { int f( ); };
using mf_t = decltype( &S::f );

static_assert( mbr_ptr< mf_t > );
//////////////////////

This bug_b module does not compile.  The diagnostics are reproduced below,
including the somewhat mysterious lone apostrophe:

bug_b.cc:8:16: error: static assertion failed
    8 | static_assert( mbr_ptr< mf_t > );
      |                ^~~~~~~~~~~~~~~
bug_b.cc:8:16: note: constraints not satisfied
'
bug_a.cc:12: confused by earlier errors, bailing out

Please note that the static_assert and supporting declarations in module bug_a
are identical to those in module bug_a, yet one compiles while the other does
not.

Thus we seem to have two problems:  (1) the lone apostrophe amongst the
diagnostics, and (2) the inconsistent evaluation of identical expressions.

Finally, please note that if I use a class template (instead of the variable
template) in module bug_a, both modules' static_asserts now compile without
complaint:

//////////////////////
template< class T >
struct probe { static constexpr bool value = false; };
//
template< class R, class C >
struct probe<R C::*> { static constexpr bool value = true; };

export
template< class T >
concept
  mbr_ptr = probe<T>::value;
//////////////////////

             reply	other threads:[~2021-05-19 22:56 UTC|newest]

Thread overview: 4+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2021-05-19 22:56 webrown.cpp at gmail dot com [this message]
2021-11-25  0:37 ` [Bug c++/100687] " johelegp at gmail dot com
2022-03-23 22:23 ` johelegp at gmail dot com
2022-09-22 12:57 ` ppalka at gcc dot gnu.org

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=bug-100687-4@http.gcc.gnu.org/bugzilla/ \
    --to=gcc-bugzilla@gcc.gnu.org \
    --cc=gcc-bugs@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).