[Bug c++/45642] New: g++ 4.6 regression, c++0x, weird mismatch for arguments with forwarded declaration when attributes are involved

[Bug c++/45642] New: g++ 4.6 regression, c++0x, weird mismatch for arguments with forwarded declaration when attributes are involved

[tbptbp at gmail dot com]

Since around 2010/09/07 g++ started to barf on the attached source about a
mismatched prototype. Apparently the culprit is the forward declaration + later
definition with additional attribute of one of the argument, and, if the
forward declaration is sprinkled with said attribute, the error goes away.

I've failed to reduce both manually and with delta (mismatched prototypes are
too easily produced). Sorry.

$ /usr/local/gcc-4.6-20100910/bin/g++ -std=c++0x -c synth.ii
src/audio/synth_voice_impl.h:140:6: error: prototype for 'bool
synth::voice::voice_t::produce_chunk(const synth::semitones_lut_t&, const
synth::midi_channels_params_t&, synth::chunk_array_mono_t&)' does not match any
in class 'synth::voice::voice_t'
src/audio/synth_voice.h:153:156: error: candidate is:
template<synth::voice::waves::waves_enum_t wave_shape> bool
synth::voice::voice_t::produce_chunk(const synth::semitones_lut_t&, const
synth::midi_channels_params_t&, synth::chunk_array_mono_t&)
...
$ /usr/local/gcc-4.6-20100910/bin/g++ -v
Using built-in specs.
COLLECT_GCC=/usr/local/gcc-4.6-20100910/bin/g++
COLLECT_LTO_WRAPPER=/usr/local/gcc-4.6-20100910/bin/../libexec/gcc/x86_64-unknown-linux-gnu/4.6.0/lto-wrapper
Target: x86_64-unknown-linux-gnu
Configured with: ../configure --prefix=/usr/local/gcc-4.6.0
--enable-languages=c,c++ --enable-threads=posix --disable-checking
--disable-nls --with-system-zlib --disable-bootstrap --enable-checking=none
--enable-mpfr --enable-gold --enable-lto --with-ppl --with-cloog
--with-arch=native
Thread model: posix
gcc version 4.6.0 20100910 (experimental) (GCC) 

midi_channels_params_t is the trigger.
Known to work on g++ 4.4.x, 4.5.x and up to and including 4.6-20100901.

PS: http://gcc.gnu.org/ml/gcc/2010-09/msg00152.html


-- 
           Summary: g++ 4.6 regression, c++0x, weird mismatch for arguments
                    with forwarded declaration when attributes are involved
           Product: gcc
           Version: 4.6.0
            Status: UNCONFIRMED
          Severity: normal
          Priority: P3
         Component: c++
        AssignedTo: unassigned at gcc dot gnu dot org
        ReportedBy: tbptbp at gmail dot com


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=45642

[Bug c++/45642] g++ 4.6 regression, c++0x, weird mismatch for arguments with forwarded declaration when attributes are involved

[tbptbp at gmail dot com]

------- Comment #1 from tbptbp at gmail dot com  2010-09-10 17:34 -------
Created an attachment (id=21767)
 --> (http://gcc.gnu.org/bugzilla/attachment.cgi?id=21767&action=view)
large & fugly testcase


-- 


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=45642

[Bug c++/45642] g++ 4.6 regression, c++0x, weird mismatch for arguments with forwarded declaration when attributes are involved

[pinskia at gcc dot gnu dot org]

------- Comment #3 from pinskia at gcc dot gnu dot org  2010-09-10 17:35 -------
This seems related to PR 45112.


-- 


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=45642

[Bug c++/45642] g++ 4.6 regression, c++0x, weird mismatch for arguments with forwarded declaration when attributes are involved

[pinskia at gcc dot gnu dot org]

------- Comment #2 from pinskia at gcc dot gnu dot org  2010-09-10 17:34 -------
I think you need __attribute((aligned(16))) on the original forward declared
class too.


-- 


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=45642

[Bug c++/45642] g++ 4.6 regression, c++0x, weird mismatch for arguments with forwarded declaration when attributes are involved

[tbptbp at gmail dot com]

------- Comment #4 from tbptbp at gmail dot com  2010-09-10 18:00 -------
(In reply to comment #2)
> I think you need __attribute((aligned(16))) on the original forward declared
> class too.
As stated that does, indeed, fix it.
So, ok, let's say previous versions were too permissive, then, the problem is
with the extremely confusing and useless diagnostic: there's no warning with
-Wall -Wextra about the declaration/definition mismatch, and a puzzling error
about an unrelated foo(x, y) not matching foo(x, y).


-- 


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=45642