[Bug c++/55670] New: This isn't a pure virtual method.

[Bug c++/55670] New: This isn't a pure virtual method.

[qufanat at gmail dot com]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=55670

             Bug #: 55670
           Summary: This isn't a pure virtual method.
    Classification: Unclassified
           Product: gcc
           Version: 4.7.2
            Status: UNCONFIRMED
          Severity: normal
          Priority: P3
         Component: c++
        AssignedTo: unassigned@gcc.gnu.org
        ReportedBy: qufanat@gmail.com
            Target: x86_64-linux-gnu
             Build: gcc version 4.7.2 (Ubuntu/Linaro 4.7.2-11precise2)


Created attachment 28944
  --> http://gcc.gnu.org/bugzilla/attachment.cgi?id=28944
Preprocessed contents of fptest2.cpp

I've been enjoying the new C++11 features in 4.7.  But the combination of
template aliases and non-static data member initialization seems to have caught
a corner-case of pure virtual method parsing.

==== CONTENTS OF fptest.cpp ====
template <class T> using F = T;
struct X {
    F<void ()>* fp = nullptr;
};
int main () { return 0; }
================================

I expected this to compile without error, but instead I got:

==== OUTPUT OF g++-4.7 -std=c++11 fptest.cpp ====
fptest.cpp:3:29: error: invalid pure specifier (only ‘= 0’ is allowed) before
‘;’ token
=================================================

It compiles if I set fp to nullptr in a constructor, and it compiles if I
declare fp with traditional function pointer syntax, " void (* fp ) ()".  It
gives the same error if I initialize fp to the address of a function instead of
nullptr.

Just for kicks, I decided to see what would happen if I tried to initialize fp
to 0.

==== OUTPUT OF g++-4.7 -std=c++11 fptest2.cpp ====
fptest2.cpp:4:1: internal compiler error: Segmentation fault
Please submit a full bug report,
with preprocessed source if appropriate.
See <file:///usr/share/doc/gcc-4.7/README.Bugs> for instructions.
Preprocessed source stored into /tmp/ccjdsX7p.out file, please attach this to
your bugreport.
==================================================

Oh.  Well then.


==== OUTPUT OF g++-4.7 -v ====
Using built-in specs.
COLLECT_GCC=g++-4.7
COLLECT_LTO_WRAPPER=/usr/lib/gcc/x86_64-linux-gnu/4.7/lto-wrapper
Target: x86_64-linux-gnu
Configured with: ../src/configure -v --with-pkgversion='Ubuntu/Linaro
4.7.2-11precise2' --with-bugurl=file:///usr/share/doc/gcc-4.7/README.Bugs
--enable-languages=c,c++,go,fortran,objc,obj-c++ --prefix=/usr
--program-suffix=-4.7 --enable-shared --enable-linker-build-id
--with-system-zlib --libexecdir=/usr/lib --without-included-gettext
--enable-threads=posix --with-gxx-include-dir=/usr/include/c++/4.7
--libdir=/usr/lib --enable-nls --with-sysroot=/ --enable-clocale=gnu
--enable-libstdcxx-debug --enable-libstdcxx-time=yes --enable-gnu-unique-object
--enable-plugin --enable-objc-gc --disable-werror --with-arch-32=i686
--with-abi=m64 --with-multilib-list=m32,m64 --with-tune=generic
--enable-checking=release --build=x86_64-linux-gnu --host=x86_64-linux-gnu
--target=x86_64-linux-gnu
Thread model: posix
gcc version 4.7.2 (Ubuntu/Linaro 4.7.2-11precise2) 
==============================

[Bug c++/55670] [C++11] This isn't a pure virtual method.

[redi at gcc dot gnu.org]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=55670

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

           What    |Removed                     |Added
----------------------------------------------------------------------------
             Target|x86_64-linux-gnu            |
             Status|UNCONFIRMED                 |NEW
           Keywords|                            |ice-on-valid-code,
                   |                            |rejects-valid
   Last reconfirmed|                            |2012-12-13
     Ever Confirmed|0                           |1
            Summary|This isn't a pure virtual   |[C++11] This isn't a pure
                   |method.                     |virtual method.
              Build|gcc version 4.7.2           |
                   |(Ubuntu/Linaro              |
                   |4.7.2-11precise2)           |

--- Comment #1 from Jonathan Wakely <redi at gcc dot gnu.org> 2012-12-13 10:33:01 UTC ---
Confirmed. Not a regression as NSDMI and alias templates are new.

As a workaround you can use a braced-init-list for the initializer:

    F<void ()>* fp { nullptr };

[Bug c++/55670] [C++11] This isn't a pure virtual method.

[jason at gcc dot gnu.org]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=55670

--- Comment #2 from Jason Merrill <jason at gcc dot gnu.org> 2013-02-14 04:31:14 UTC ---
Author: jason
Date: Thu Feb 14 04:31:03 2013
New Revision: 196043

URL: http://gcc.gnu.org/viewcvs?root=gcc&view=rev&rev=196043
Log:
    PR c++/55670
    * parser.c (cp_parser_member_declaration): Check the declarator
    form when detecting a function declaration via typedef.

Added:
    trunk/gcc/testsuite/g++.dg/cpp0x/nsdmi8.C
Modified:
    trunk/gcc/cp/ChangeLog
    trunk/gcc/cp/parser.c

[Bug c++/55670] [C++11] This isn't a pure virtual method.

[jason at gcc dot gnu.org]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=55670

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

           What    |Removed                     |Added
----------------------------------------------------------------------------
             Status|NEW                         |RESOLVED
                 CC|                            |jason at gcc dot gnu.org
         Resolution|                            |FIXED
         AssignedTo|unassigned at gcc dot       |jason at gcc dot gnu.org
                   |gnu.org                     |
   Target Milestone|---                         |4.8.0

--- Comment #3 from Jason Merrill <jason at gcc dot gnu.org> 2013-02-15 19:30:42 UTC ---
Fixed for 4.8.