public inbox for gcc-bugs@sourceware.org
help / color / mirror / Atom feed
* [Bug c++/102345] New: [modules] Cannot define a module interface unit for anything in <new>
@ 2021-09-15 11:21 redi at gcc dot gnu.org
  2021-09-15 11:22 ` [Bug c++/102345] " redi at gcc dot gnu.org
                   ` (5 more replies)
  0 siblings, 6 replies; 7+ messages in thread
From: redi at gcc dot gnu.org @ 2021-09-15 11:21 UTC (permalink / raw)
  To: gcc-bugs

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=102345

            Bug ID: 102345
           Summary: [modules] Cannot define a module interface unit for
                    anything in <new>
           Product: gcc
           Version: 12.0
            Status: UNCONFIRMED
          Keywords: rejects-valid
          Severity: normal
          Priority: P3
         Component: c++
          Assignee: unassigned at gcc dot gnu.org
          Reporter: redi at gcc dot gnu.org
  Target Milestone: ---

Similar to PR 48396, the implicit declaration of std::align_val_t means you
can't include <new> in a module interface unit.

There are also apparently conflicts between the (correct according to the
standard) new/delete operators, and GCC's built-ins.


export module newdel;
#include <new>



new.cc:5:28: error: cannot define ‘enum class std::align_val_t’ in different
module
    5 |   enum class align_val_t : size_t;
      |                            ^~~~~~
<built-in>: note: declared here
new.cc:5: confused by earlier errors, bailing out
Preprocessed source stored into /tmp/ccZNf6X0.out file, please attach this to
your bugreport.
tmp$ g++20 mod2.cc -fmodules-ts
In file included from mod2.cc:2:
/home/jwakely/gcc/12/include/c++/12.0.0/new:127:41: error: declaration 'void*
operator new(std::size_t)' conflicts with builtin
  127 |   __attribute__((__externally_visible__));
      |                                         ^
/home/jwakely/gcc/12/include/c++/12.0.0/new:129:41: error: declaration 'void*
operator new [](std::size_t)' conflicts with builtin
  129 |   __attribute__((__externally_visible__));
      |                                         ^
/home/jwakely/gcc/12/include/c++/12.0.0/new:131:41: error: declaration 'void
operator delete(void*)' conflicts with builtin
  131 |   __attribute__((__externally_visible__));
      |                                         ^
/home/jwakely/gcc/12/include/c++/12.0.0/new:133:41: error: declaration 'void
operator delete [](void*)' conflicts with builtin
  133 |   __attribute__((__externally_visible__));
      |                                         ^
/home/jwakely/gcc/12/include/c++/12.0.0/new:136:41: error: declaration 'void
operator delete(void*, std::size_t)' conflicts with builtin
  136 |   __attribute__((__externally_visible__));
      |                                         ^
/home/jwakely/gcc/12/include/c++/12.0.0/new:138:41: error: declaration 'void
operator delete [](void*, std::size_t)' conflicts with builtin
  138 |   __attribute__((__externally_visible__));
      |                                         ^
/home/jwakely/gcc/12/include/c++/12.0.0/new:150:73: error: declaration 'void*
operator new(std::size_t, std::align_val_t)' conflicts with builtin
  150 |   __attribute__((__externally_visible__, __alloc_size__ (1),
__malloc__));
      |                                                                        
^
/home/jwakely/gcc/12/include/c++/12.0.0/new:154:63: error: declaration 'void
operator delete(void*, std::align_val_t)' conflicts with builtin
  154 |   _GLIBCXX_USE_NOEXCEPT __attribute__((__externally_visible__));
      |                                                               ^
/home/jwakely/gcc/12/include/c++/12.0.0/new:158:73: error: declaration 'void*
operator new [](std::size_t, std::align_val_t)' conflicts with builtin
  158 |   __attribute__((__externally_visible__, __alloc_size__ (1),
__malloc__));
      |                                                                        
^
/home/jwakely/gcc/12/include/c++/12.0.0/new:162:63: error: declaration 'void
operator delete [](void*, std::align_val_t)' conflicts with builtin
  162 |   _GLIBCXX_USE_NOEXCEPT __attribute__((__externally_visible__));
      |                                                               ^
/home/jwakely/gcc/12/include/c++/12.0.0/new:167:63: error: declaration 'void
operator delete(void*, std::size_t, std::align_val_t)' conflicts with builtin
  167 |   _GLIBCXX_USE_NOEXCEPT __attribute__((__externally_visible__));
      |                                                               ^
/home/jwakely/gcc/12/include/c++/12.0.0/new:169:63: error: declaration 'void
operator delete [](void*, std::size_t, std::align_val_t)' conflicts with
builtin
  169 |   _GLIBCXX_USE_NOEXCEPT __attribute__((__externally_visible__));
      |                                                               ^
mod2.cc:1:8: warning: not writing module 'newdel' due to errors
    1 | export module newdel;
      |        ^~~~~~

^ permalink raw reply	[flat|nested] 7+ messages in thread

* [Bug c++/102345] [modules] Cannot define a module interface unit for anything in <new>
  2021-09-15 11:21 [Bug c++/102345] New: [modules] Cannot define a module interface unit for anything in <new> redi at gcc dot gnu.org
@ 2021-09-15 11:22 ` redi at gcc dot gnu.org
  2021-09-15 11:24 ` redi at gcc dot gnu.org
                   ` (4 subsequent siblings)
  5 siblings, 0 replies; 7+ messages in thread
From: redi at gcc dot gnu.org @ 2021-09-15 11:22 UTC (permalink / raw)
  To: gcc-bugs

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=102345

Jonathan Wakely <redi at gcc dot gnu.org> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
   Last reconfirmed|                            |2021-09-15
             Status|UNCONFIRMED                 |NEW
     Ever confirmed|0                           |1

--- Comment #1 from Jonathan Wakely <redi at gcc dot gnu.org> ---
The first problem is demonstrated by:

export module newdel;
namespace std
{
  using size_t = decltype(sizeof(0));
  enum class align_val_t : size_t;
}


new.cc:5:28: error: cannot define ‘enum class std::align_val_t’ in different
module
    5 |   enum class align_val_t : size_t;
      |                            ^~~~~~
<built-in>: note: declared here
new.cc:5: confused by earlier errors, bailing out
Preprocessed source stored into /tmp/cc4qIyxE.out file, please attach this to
your bugreport.


This is fixed by Jason's patch in PR 48396 comment 4.

^ permalink raw reply	[flat|nested] 7+ messages in thread

* [Bug c++/102345] [modules] Cannot define a module interface unit for anything in <new>
  2021-09-15 11:21 [Bug c++/102345] New: [modules] Cannot define a module interface unit for anything in <new> redi at gcc dot gnu.org
  2021-09-15 11:22 ` [Bug c++/102345] " redi at gcc dot gnu.org
@ 2021-09-15 11:24 ` redi at gcc dot gnu.org
  2021-09-15 11:24 ` redi at gcc dot gnu.org
                   ` (3 subsequent siblings)
  5 siblings, 0 replies; 7+ messages in thread
From: redi at gcc dot gnu.org @ 2021-09-15 11:24 UTC (permalink / raw)
  To: gcc-bugs

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=102345

--- Comment #2 from Jonathan Wakely <redi at gcc dot gnu.org> ---
The second problem is demonstrated by:

export module newdel;
export void* operator new(__SIZE_TYPE__);


new.cc:2:40: error: declaration ‘void* operator new(long unsigned int)’
conflicts with builtin
    2 | export void* operator new(__SIZE_TYPE__);
      |                                        ^
new.cc:1:8: warning: not writing module ‘newdel’ due to errors
    1 | export module newdel;
      |        ^~~~~~

^ permalink raw reply	[flat|nested] 7+ messages in thread

* [Bug c++/102345] [modules] Cannot define a module interface unit for anything in <new>
  2021-09-15 11:21 [Bug c++/102345] New: [modules] Cannot define a module interface unit for anything in <new> redi at gcc dot gnu.org
  2021-09-15 11:22 ` [Bug c++/102345] " redi at gcc dot gnu.org
  2021-09-15 11:24 ` redi at gcc dot gnu.org
@ 2021-09-15 11:24 ` redi at gcc dot gnu.org
  2024-03-13 18:26 ` ppalka at gcc dot gnu.org
                   ` (2 subsequent siblings)
  5 siblings, 0 replies; 7+ messages in thread
From: redi at gcc dot gnu.org @ 2021-09-15 11:24 UTC (permalink / raw)
  To: gcc-bugs

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=102345

--- Comment #3 from Jonathan Wakely <redi at gcc dot gnu.org> ---
All testcases above need to be compiled with -fmodules-ts

^ permalink raw reply	[flat|nested] 7+ messages in thread

* [Bug c++/102345] [modules] Cannot define a module interface unit for anything in <new>
  2021-09-15 11:21 [Bug c++/102345] New: [modules] Cannot define a module interface unit for anything in <new> redi at gcc dot gnu.org
                   ` (2 preceding siblings ...)
  2021-09-15 11:24 ` redi at gcc dot gnu.org
@ 2024-03-13 18:26 ` ppalka at gcc dot gnu.org
  2024-05-25  0:32 ` cvs-commit at gcc dot gnu.org
  2024-05-25  0:34 ` nshead at gcc dot gnu.org
  5 siblings, 0 replies; 7+ messages in thread
From: ppalka at gcc dot gnu.org @ 2024-03-13 18:26 UTC (permalink / raw)
  To: gcc-bugs

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=102345

Patrick Palka <ppalka at gcc dot gnu.org> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
                 CC|                            |ppalka at gcc dot gnu.org
           Keywords|rejects-valid               |diagnostic

--- Comment #4 from Patrick Palka <ppalka at gcc dot gnu.org> ---
IIUC we're correct to reject since built-ins are implicitly attached to the
global module and here we're trying to redeclare one in another module?

Perhaps the diagnostic could be improved here though.  Clang gives

  error: declaration of 'operator new' in module newdel follows declaration in
the global module

^ permalink raw reply	[flat|nested] 7+ messages in thread

* [Bug c++/102345] [modules] Cannot define a module interface unit for anything in <new>
  2021-09-15 11:21 [Bug c++/102345] New: [modules] Cannot define a module interface unit for anything in <new> redi at gcc dot gnu.org
                   ` (3 preceding siblings ...)
  2024-03-13 18:26 ` ppalka at gcc dot gnu.org
@ 2024-05-25  0:32 ` cvs-commit at gcc dot gnu.org
  2024-05-25  0:34 ` nshead at gcc dot gnu.org
  5 siblings, 0 replies; 7+ messages in thread
From: cvs-commit at gcc dot gnu.org @ 2024-05-25  0:32 UTC (permalink / raw)
  To: gcc-bugs

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=102345

--- Comment #5 from GCC Commits <cvs-commit at gcc dot gnu.org> ---
The master branch has been updated by Nathaniel Shead <nshead@gcc.gnu.org>:

https://gcc.gnu.org/g:28b508233a12c13295f960a2cb8a4864879acfb4

commit r15-830-g28b508233a12c13295f960a2cb8a4864879acfb4
Author: Nathaniel Shead <nathanieloshead@gmail.com>
Date:   Sat May 25 01:03:54 2024 +1000

    c++/modules: Improve diagnostic when redeclaring builtin in module
[PR102345]

    If a user mistakenly includes a standard library header within the
    module purview, they currently get a confusing "declaration conflicts
    with builtin" error.  This patch updates the message to include "in
    module", to help guide the user towards the likely cause.

            PR c++/102345

    gcc/cp/ChangeLog:

            * module.cc (module_may_redeclare): Update error message.

    gcc/testsuite/ChangeLog:

            * g++.dg/modules/enum-12.C: Test for updated error.

    Signed-off-by: Nathaniel Shead <nathanieloshead@gmail.com>

^ permalink raw reply	[flat|nested] 7+ messages in thread

* [Bug c++/102345] [modules] Cannot define a module interface unit for anything in <new>
  2021-09-15 11:21 [Bug c++/102345] New: [modules] Cannot define a module interface unit for anything in <new> redi at gcc dot gnu.org
                   ` (4 preceding siblings ...)
  2024-05-25  0:32 ` cvs-commit at gcc dot gnu.org
@ 2024-05-25  0:34 ` nshead at gcc dot gnu.org
  5 siblings, 0 replies; 7+ messages in thread
From: nshead at gcc dot gnu.org @ 2024-05-25  0:34 UTC (permalink / raw)
  To: gcc-bugs

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=102345

Nathaniel Shead <nshead at gcc dot gnu.org> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
             Status|NEW                         |RESOLVED
                 CC|                            |nshead at gcc dot gnu.org
         Resolution|---                         |FIXED

--- Comment #6 from Nathaniel Shead <nshead at gcc dot gnu.org> ---
The diagnostic now reads:

e.cpp:2:14: error: declaring ‘void* operator new(long unsigned int)’ in module
‘newdel’ conflicts with builtin in global module
    2 | export void* operator new(__SIZE_TYPE__);
      |              ^~~~~~~~
e.cpp:1:8: warning: not writing module ‘newdel’ due to errors
    1 | export module newdel;
      |        ^~~~~~

which should be clearer as to what the issue is.

^ permalink raw reply	[flat|nested] 7+ messages in thread

end of thread, other threads:[~2024-05-25  0:34 UTC | newest]

Thread overview: 7+ messages (download: mbox.gz / follow: Atom feed)
-- links below jump to the message on this page --
2021-09-15 11:21 [Bug c++/102345] New: [modules] Cannot define a module interface unit for anything in <new> redi at gcc dot gnu.org
2021-09-15 11:22 ` [Bug c++/102345] " redi at gcc dot gnu.org
2021-09-15 11:24 ` redi at gcc dot gnu.org
2021-09-15 11:24 ` redi at gcc dot gnu.org
2024-03-13 18:26 ` ppalka at gcc dot gnu.org
2024-05-25  0:32 ` cvs-commit at gcc dot gnu.org
2024-05-25  0:34 ` nshead at gcc dot gnu.org

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).