[Bug c++/51827] New: Error: 'FOO' conflicts with a previous declaration, with PCH/LTO/C++11

[Bug c++/51827] New: Error: 'FOO' conflicts with a previous declaration, with PCH/LTO/C++11

[d.g.gorbachev at gmail dot com]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=51827

             Bug #: 51827
           Summary: Error: 'FOO' conflicts with a previous declaration,
                    with PCH/LTO/C++11
    Classification: Unclassified
           Product: gcc
           Version: 4.7.0
            Status: UNCONFIRMED
          Severity: normal
          Priority: P3
         Component: c++
        AssignedTo: unassigned@gcc.gnu.org
        ReportedBy: d.g.gorbachev@gmail.com
              Host: i686-pc-linux-gnu
            Target: i686-pc-linux-gnu
             Build: i686-pc-linux-gnu


GCC 4.7.0 20120107 (experimental)

TestCase.H:3:29: error: 'T<Y>::T(S<A ...>) [with A = {A ...}; Y = int]'
conflicts with a previous declaration
TestCase.H:3:29: note: previous declaration 'T<Y>::T(S<A ...>)'
TestCase.H:3:29: note: -fabi-version=6 (or =0) avoids this error with a change
in mangling
TestCase.H:3:29: error: 'T<Y>::T(S<A ...>) [with A = {A ...}; Y = int]'
conflicts with a previous declaration
TestCase.H:3:29: note: previous declaration 'T<Y>::T(S<A ...>)'
TestCase.H:3:29: note: -fabi-version=6 (or =0) avoids this error with a change
in mangling

Compile with `-flto -std=gnu++11':

=================== 8< ===================
template<typename X> struct S { };
template<typename Y> struct T {
  template <typename ... A> T(S <A ...>);
};
inline void f(T<int>) { }
=================== >8 ===================

[Bug c++/51827] Error: 'FOO' conflicts with a previous declaration, with PCH/LTO/C++11

[rguenth at gcc dot gnu.org]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=51827

--- Comment #1 from Richard Guenther <rguenth at gcc dot gnu.org> 2012-01-12 09:23:03 UTC ---
I can't reproduce this.

[Bug c++/51827] Error: 'FOO' conflicts with a previous declaration, with PCH/LTO/C++11

[d.g.gorbachev at gmail dot com]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=51827

--- Comment #2 from Dmitry Gorbachev <d.g.gorbachev at gmail dot com> 2012-01-12 10:53:44 UTC ---
I already mentioned "PCH" and ".H" extension, but just to be 100% clear, the
error happens only when compiling the testcase as a c++ header.

Reproduced on i686-pc-linux-gnu and i686-pc-mingw32.

[Bug c++/51827] Error: 'FOO' conflicts with a previous declaration, with PCH/LTO/C++11

[paolo.carlini at oracle dot com]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=51827

Paolo Carlini <paolo.carlini at oracle dot com> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
                 CC|                            |jason at gcc dot gnu.org

--- Comment #3 from Paolo Carlini <paolo.carlini at oracle dot com> 2012-01-16 15:31:15 UTC ---
I can reproduce the error, for the snippet as a .H, but then, -fabi-version=6
avoids it, as expected. Thus I don't see where is the issue.

Jason shall we close this?

[Bug c++/51827] Error: 'FOO' conflicts with a previous declaration, with PCH/LTO/C++11

[d.g.gorbachev at gmail dot com]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=51827

--- Comment #4 from Dmitry Gorbachev <d.g.gorbachev at gmail dot com> 2012-01-16 16:54:43 UTC ---
> Thus I don't see where is the issue.

The issue is a rather confusing error message. It looks as if there was a
compiler bug.

> shall we close this?

Somebody can report it again...

[Bug c++/51827] Error: 'FOO' conflicts with a previous declaration, with PCH/LTO/C++11

[paolo.carlini at oracle dot com]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=51827

--- Comment #5 from Paolo Carlini <paolo.carlini at oracle dot com> 2012-01-16 17:05:03 UTC ---
What's confusing about it, I don't understand. Doesn't '-fabi-version=6' fix it
for you?!? And, by the way, *closing* doesn't mean disintegrating it, would
remain in Bugzilla, for sure.

Unless, there is something *else* entirely about it, like we want to accept it
also without '-fabi-version=6', that's why I asked Jason.

[Bug c++/51827] Error: 'FOO' conflicts with a previous declaration, with PCH/LTO/C++11

[jason at gcc dot gnu.org]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=51827

--- Comment #6 from Jason Merrill <jason at gcc dot gnu.org> 2012-01-16 18:12:10 UTC ---
This should work without -fabi-version=6.

[Bug c++/51827] Error: 'FOO' conflicts with a previous declaration, with PCH/LTO/C++11

[paolo.carlini at oracle dot com]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=51827

Paolo Carlini <paolo.carlini at oracle dot com> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
             Status|UNCONFIRMED                 |NEW
   Last reconfirmed|                            |2012-01-16
     Ever Confirmed|0                           |1

[Bug c++/51827] Error: 'FOO' conflicts with a previous declaration, with PCH/LTO/C++11

[d.g.gorbachev at gmail dot com]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=51827

--- Comment #7 from Dmitry Gorbachev <d.g.gorbachev at gmail dot com> 2012-01-16 18:28:20 UTC ---
(In reply to comment #5)

> What's confusing about it, I don't understand.

There are two errors and notes, all pointing to the same source location --
isn't that odd?

> And, by the way, *closing* doesn't mean disintegrating it,
> would remain in Bugzilla, for sure.

Yeah, but people won't be able to reach it with *ordinary* search and will file
duplicates. It's none of my business, of course...

[Bug c++/51827] [4.7 regression] Error: 'FOO' conflicts with a previous declaration, with PCH/LTO/C++11

[jason at gcc dot gnu.org]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=51827

Jason Merrill <jason at gcc dot gnu.org> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
             Status|NEW                         |ASSIGNED
         AssignedTo|unassigned at gcc dot       |jason at gcc dot gnu.org
                   |gnu.org                     |
            Summary|Error: 'FOO' conflicts with |[4.7 regression] Error:
                   |a previous declaration,     |'FOO' conflicts with a
                   |with PCH/LTO/C++11          |previous declaration, with
                   |                            |PCH/LTO/C++11

[Bug c++/51827] [4.7 regression] Error: 'FOO' conflicts with a previous declaration, with PCH/LTO/C++11

[jason at gcc dot gnu.org]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=51827

--- Comment #8 from Jason Merrill <jason at gcc dot gnu.org> 2012-01-16 21:32:08 UTC ---
Author: jason
Date: Mon Jan 16 21:32:05 2012
New Revision: 183221

URL: http://gcc.gnu.org/viewcvs?root=gcc&view=rev&rev=183221
Log:
    PR c++/51827
    * mangle.c (mangle_decl): Don't mangle uninstantiated templates.

Added:
    trunk/gcc/testsuite/g++.dg/pch/mangle1.C
    trunk/gcc/testsuite/g++.dg/pch/mangle1.Hs
Modified:
    trunk/gcc/cp/ChangeLog
    trunk/gcc/cp/mangle.c
    trunk/gcc/testsuite/ChangeLog

[Bug c++/51827] [4.7 regression] Error: 'FOO' conflicts with a previous declaration, with PCH/LTO/C++11

[jason at gcc dot gnu.org]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=51827

Jason Merrill <jason at gcc dot gnu.org> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
             Status|ASSIGNED                    |RESOLVED
         Resolution|                            |FIXED
   Target Milestone|---                         |4.7.0

--- Comment #9 from Jason Merrill <jason at gcc dot gnu.org> 2012-01-16 21:35:59 UTC ---
Fixed.