[Bug c++/56111] New: {float,double} complex not accepted by g++-4.8 anymore

[Bug c++/56111] New: {float,double} complex not accepted by g++-4.8 anymore

[jtaylor.debian at gmail dot com]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=56111

             Bug #: 56111
           Summary: {float,double} complex not accepted by g++-4.8 anymore
    Classification: Unclassified
           Product: gcc
           Version: 4.8.0
            Status: UNCONFIRMED
          Severity: normal
          Priority: P3
         Component: c++
        AssignedTo: unassigned@gcc.gnu.org
        ReportedBy: jtaylor.debian@gmail.com


the C99 float and double complex is not accepted by g++ 4.8 r195467 anymore in
c++98 or c++11.
Even though its technically not part of c++98 it worked in all gcc versions I
ever used.
Is this change intentional? It will break the compilation of a bunch of
scientific c++ software using C libraries.

$ cat test.c
#include <complex.h>
int testing(float complex a);

int main()
{
float complex a = 1;
return testing(a);
}
$ gcc-4.7 test.c -c
$ gcc-4.8 test.c -c
$ g++-4.7 test.c -c

$ g++-4.8 test.c -c
test.c:2:27: error: expected ',' or '...' before 'a'
 int testing(float complex a);
                           ^
test.c: In function 'int main()':
test.c:6:15: error: expected initializer before 'a'
 float complex a = 1;
               ^
test.c:7:16: error: 'a' was not declared in this scope
 return testing(a);
                ^
$ g++-4.8 -std=c++98 test.c -c
test.c:2:27: error: expected ',' or '...' before 'a'
 int testing(float complex a);
                           ^
test.c: In function 'int main()':
test.c:6:15: error: expected initializer before 'a'
 float complex a = 1;
               ^
test.c:7:16: error: 'a' was not declared in this scope
 return testing(a);
                ^
$ g++-4.8 --version
g$-4.8 (Debian 4.8-20130123-1) 4.8.0 20130123 (experimental) [trunk revision
195400]
Copyright (C) 2013 Free Software Foundation, Inc.
This is free software; see the source for copying conditions.  There is NO
warranty; not even for MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.


also reproduced with svn trunk r195467 built on a fedora 11 machine.
using ccomplex and -std=c++11 does not work either.

[Bug c++/56111] {float,double} complex not accepted by g++-4.8 anymore

[glisse at gcc dot gnu.org]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=56111

Marc Glisse <glisse at gcc dot gnu.org> changed:

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

--- Comment #1 from Marc Glisse <glisse at gcc dot gnu.org> 2013-01-25 19:14:03 UTC ---
You need to write _Complex or __complex__. complex is a macro that breaks code
from the standard library (think std::complex<double>) so we are forced to
undefine it. See PR54112 for details of the change.

If we really wanted to, I guess we could #undef complex at the beginning of
<complex> and condition the #undef in <complex.h> to the fact that <complex>
has been included before...

[Bug c++/56111] {float,double} complex not accepted by g++-4.8 anymore

[paolo.carlini at oracle dot com]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=56111

--- Comment #2 from Paolo Carlini <paolo.carlini at oracle dot com> 2013-01-26 10:36:19 UTC ---
What happens if we simply #undef at the beginning of <complex> and nothing
else? It seems to me that <complex> is the *real* place where such a macro
doesn't make sense and the primary change should go there (in case with further
adjustments).

[Bug c++/56111] {float,double} complex not accepted by g++-4.8 anymore

[paolo.carlini at oracle dot com]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=56111

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

           What    |Removed                     |Added
----------------------------------------------------------------------------
                 CC|                            |jakub at redhat dot com

--- Comment #3 from Paolo Carlini <paolo.carlini at oracle dot com> 2013-01-26 10:38:55 UTC ---
Let's add Jakub in CC.

[Bug libstdc++/56111] {float,double,long double} complex not accepted anymore

[glisse at gcc dot gnu.org]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=56111

Marc Glisse <glisse at gcc dot gnu.org> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
          Component|c++                         |libstdc++

--- Comment #4 from Marc Glisse <glisse at gcc dot gnu.org> 2013-01-26 10:58:22 UTC ---
(In reply to comment #2)
> What happens if we simply #undef at the beginning of <complex> and nothing
> else?

My idea in comment #1 was that if people include <complex> and then <complex.h>
(the order is important), with your suggestion, they won't be able to use
std::complex, so there is little point including <complex>.

Now in C++11, complex.h actually first includes ccomplex, which seems to be an
alias for complex... so there is no way to include complex.h and not complex.
Which seems to imply that at least in C++11, we should just always undef
complex.

[Bug libstdc++/56111] {float,double,long double} complex not accepted anymore

[paolo.carlini at oracle dot com]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=56111

--- Comment #5 from Paolo Carlini <paolo.carlini at oracle dot com> 2013-01-26 11:11:24 UTC ---
Ok, I understand. Looks like, however, we are back to my original observation
in PR54112 that in C++03 mode we have less constraints and with more leeway we
could manage to keep on supporting this? I think submitter has a point that
it's weird that we don't support anymore in C++03 mode the very simple and old
kind of code attached here. I have no idea how complex it would be and whether
it's worth it, I would like to ear from the Release managers too.

[Bug libstdc++/56111] {float,double,long double} complex not accepted anymore

[glisse at gcc dot gnu.org]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=56111

--- Comment #6 from Marc Glisse <glisse at gcc dot gnu.org> 2013-01-26 11:35:52 UTC ---
(In reply to comment #5)
> Ok, I understand. Looks like, however, we are back to my original observation
> in PR54112 that in C++03 mode we have less constraints and with more leeway we
> could manage to keep on supporting this? I think submitter has a point that
> it's weird that we don't support anymore in C++03 mode the very simple and old
> kind of code attached here. I have no idea how complex it would be and whether
> it's worth it, I would like to ear from the Release managers too.

What I said in comment #1 should take about 3-5 lines of macros and support
this in C++03 mode, I hope. Or we could partially revert my change and only
undef complex in C++11, leaving complex.h and complex as incompatible in C++03.
Or even never undef complex, though it doesn't seem very conforming.

Note while we are talking about this that I think it was a very bad decision
from the committee to make <complex.h> an alias for <complex>. A few minor
changes like in string.h are acceptable (they already cause trouble), but
complex.h is a C header, and C++ has no business replacing it with something
that doesn't define the same interface at all. It would have been better to not
mention complex.h at all, or only as not supported, leaving it up to
implementors to decide how to handle it.

[Bug libstdc++/56111] [4.8 Regression] {float,double,long double} complex not accepted anymore

[glisse at gcc dot gnu.org]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=56111

Marc Glisse <glisse at gcc dot gnu.org> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
             Status|UNCONFIRMED                 |NEW
   Last reconfirmed|                            |2013-01-31
            Summary|{float,double,long double}  |[4.8 Regression]
                   |complex not accepted        |{float,double,long double}
                   |anymore                     |complex not accepted
                   |                            |anymore
     Ever Confirmed|0                           |1

--- Comment #7 from Marc Glisse <glisse at gcc dot gnu.org> 2013-01-31 12:31:18 UTC ---
Confirming as regression (doesn't mean it can't be marked wontfix later).

[Bug libstdc++/56111] [4.8 Regression] {float,double,long double} complex not accepted anymore

[jakub at gcc dot gnu.org]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=56111

Jakub Jelinek <jakub at gcc dot gnu.org> changed:

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

--- Comment #8 from Jakub Jelinek <jakub at gcc dot gnu.org> 2013-01-31 12:37:11 UTC ---
Yeah, I think my preference would be to #undef complex at the beginning of
<complex>, or ignore PR54112 altogether.

[Bug libstdc++/56111] [4.8 Regression] {float,double,long double} complex not accepted anymore

[glisse at gcc dot gnu.org]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=56111

--- Comment #9 from Marc Glisse <glisse at gcc dot gnu.org> 2013-01-31 12:53:45 UTC ---
Created attachment 29315
  --> http://gcc.gnu.org/bugzilla/attachment.cgi?id=29315
#undef complex only if <complex> is included

Completely untested patch just to make comment #1 more clear.

[Bug libstdc++/56111] [4.8 Regression] {float,double,long double} complex not accepted anymore

[paolo.carlini at oracle dot com]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=56111

--- Comment #10 from Paolo Carlini <paolo.carlini at oracle dot com> 2013-01-31 12:57:46 UTC ---
Just to confirm that I like this sort of change. Thanks Marc.

[Bug libstdc++/56111] [4.8 Regression] {float,double,long double} complex not accepted anymore

[jakub at gcc dot gnu.org]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=56111

--- Comment #11 from Jakub Jelinek <jakub at gcc dot gnu.org> 2013-01-31 13:05:52 UTC ---
Yeah, looks nice.

[Bug libstdc++/56111] [4.8 Regression] {float,double,long double} complex not accepted anymore

[glisse at gcc dot gnu.org]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=56111

--- Comment #12 from Marc Glisse <glisse at gcc dot gnu.org> 2013-01-31 16:53:42 UTC ---
Ok, I'll add a testcase and test the patch in a couple weeks if nobody beats me
to it (I am leaving tonight).

[Bug libstdc++/56111] [4.8 Regression] {float,double,long double} complex not accepted anymore

[paolo.carlini at oracle dot com]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=56111

--- Comment #13 from Paolo Carlini <paolo.carlini at oracle dot com> 2013-01-31 16:59:52 UTC ---
Great, thanks.

[Bug libstdc++/56111] [4.8 Regression] {float,double,long double} complex not accepted anymore

[rguenth at gcc dot gnu.org]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=56111

Richard Biener <rguenth at gcc dot gnu.org> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
           Keywords|                            |rejects-valid
           Priority|P3                          |P1
   Target Milestone|---                         |4.8.0

[Bug libstdc++/56111] [4.8 Regression] {float,double,long double} complex not accepted anymore

[jakub at gcc dot gnu.org]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=56111

--- Comment #14 from Jakub Jelinek <jakub at gcc dot gnu.org> 2013-02-09 18:42:48 UTC ---
Given this is now P1, could the #c9 patch be committed soon, and perhaps just
defer a testcase for it when Marc returns?

[Bug libstdc++/56111] [4.8 Regression] {float,double,long double} complex not accepted anymore

[paolo.carlini at oracle dot com]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=56111

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

           What    |Removed                     |Added
----------------------------------------------------------------------------
                 CC|                            |paolo.carlini at oracle dot
                   |                            |com

--- Comment #15 from Paolo Carlini <paolo.carlini at oracle dot com> 2013-02-12 16:53:42 UTC ---
Marc, any news on this?

[Bug libstdc++/56111] [4.8 Regression] {float,double,long double} complex not accepted anymore

[paolo.carlini at oracle dot com]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=56111

--- Comment #16 from Paolo Carlini <paolo.carlini at oracle dot com> 2013-02-12 16:59:39 UTC ---
Jakub, if Marc remains unreachable I'll take care of carefully checking and
committing the patch in #c9 first thing tomorrow.

[Bug libstdc++/56111] [4.8 Regression] {float,double,long double} complex not accepted anymore

[glisse at gcc dot gnu.org]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=56111

--- Comment #17 from Marc Glisse <glisse at gcc dot gnu.org> 2013-02-12 20:55:09 UTC ---
(In reply to comment #15)
> Marc, any news on this?

I landed less than 2 hours ago...
I'll see if I can handle it tomorrow.

[Bug libstdc++/56111] [4.8 Regression] {float,double,long double} complex not accepted anymore

[glisse at gcc dot gnu.org]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=56111

--- Comment #18 from Marc Glisse <glisse at gcc dot gnu.org> 2013-02-13 21:58:57 UTC ---
Author: glisse
Date: Wed Feb 13 21:58:53 2013
New Revision: 196034

URL: http://gcc.gnu.org/viewcvs?root=gcc&view=rev&rev=196034
Log:
2013-02-13  Marc Glisse  <marc.glisse@inria.fr>

    PR libstdc++/56111
    * include/std/complex (complex): Undefine.
    * include/c_compatibility/complex.h (complex): Only undefine if
    <complex> has been included.
    * testsuite/26_numerics/complex/56111.cc: New testcase.


Added:
    trunk/libstdc++-v3/testsuite/26_numerics/complex/56111.cc   (with props)
Modified:
    trunk/libstdc++-v3/ChangeLog
    trunk/libstdc++-v3/include/c_compatibility/complex.h
    trunk/libstdc++-v3/include/std/complex

Propchange: trunk/libstdc++-v3/testsuite/26_numerics/complex/56111.cc
            ('svn:eol-style' added)

Propchange: trunk/libstdc++-v3/testsuite/26_numerics/complex/56111.cc
            ('svn:keywords' added)

[Bug libstdc++/56111] [4.8 Regression] {float,double,long double} complex not accepted anymore

[glisse at gcc dot gnu.org]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=56111

Marc Glisse <glisse at gcc dot gnu.org> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
             Status|NEW                         |RESOLVED
         Resolution|                            |FIXED
         AssignedTo|unassigned at gcc dot       |glisse at gcc dot gnu.org
                   |gnu.org                     |

--- Comment #19 from Marc Glisse <glisse at gcc dot gnu.org> 2013-02-13 22:02:33 UTC ---
Done. Users should still convert their code to use _Complex instead if they
want it to work in C++11.