[Bug libstdc++/61795] New: [C++11] return type of std::pow(std::complex<float>, int) should be std::complex<double>

[Bug libstdc++/61795] New: [C++11] return type of std::pow(std::complex<float>, int) should be std::complex<double>

[kariya_mitsuru at hotmail dot com]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=61795

            Bug ID: 61795
           Summary: [C++11] return type of std::pow(std::complex<float>,
                    int) should be std::complex<double>
           Product: gcc
           Version: 4.10.0
            Status: UNCONFIRMED
          Severity: normal
          Priority: P3
         Component: libstdc++
          Assignee: unassigned at gcc dot gnu.org
          Reporter: kariya_mitsuru at hotmail dot com

With libstdc++, the return type of std::pow(std::complex<float>, int) is
std::complex<float>.
However, In C++11 mode, the return type of std::pow(std::complex<float>, int)
should be std::complex<double>.


According to C++11 standard 26.4.9[cmplx.over] paragraph 3, "if either argument
has type complex<double>, double, or an integer type, then both arguments are
effectively cast to complex<double>."

The return type of std::pow(std::complex<double>, std::complex<double>) is
std::complex<double>, So I think that std::pow(std::complex<float>, int) should
be std::complex<double> in C++11 mode.


The sample code below can show whether the return type is std::complex<double>.
================================================================================
#include <iostream>
#include <typeinfo>
#include <complex>

int main()
{
  std::cout << std::boolalpha
            << (typeid(std::pow(std::complex<float>(1), 1)) ==
typeid(std::complex<double>))
            << std::endl;
}
================================================================================
cf. http://melpon.org/wandbox/permlink/zW8TWZe9kKzKWqFq


While in C++03 mode, the return type of std::pow(std::complex<float>, int)
should be std::complex<float>, I think.

Note that this problem does not occur in std::complex<double> and
std::complex<long double> because there is no difference between C++03 and
C++11.


See also: PR56106, PR57974

[Bug libstdc++/61795] [C++11] return type of std::pow(std::complex<float>, int) should be std::complex<double>

[daniel.kruegler at googlemail dot com]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=61795

Daniel Krügler <daniel.kruegler at googlemail dot com> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
                 CC|                            |daniel.kruegler@googlemail.
                   |                            |com

--- Comment #1 from Daniel Krügler <daniel.kruegler at googlemail dot com> ---
I agree with your analysis. I believe that the observed behaviour is due to a
lack of applying yet the corresponding library defect

http://www.open-std.org/jtc1/sc22/wg21/docs/lwg-defects.html#844

to libstdc++, because in an earlier working draft there was the required
overload:

template<class T> complex<T> pow(const complex<T>& x, int y);

which had lead to the observed result.
>From gcc-bugs-return-456232-listarch-gcc-bugs=gcc.gnu.org@gcc.gnu.org Sun Jul 13 18:37:34 2014
Return-Path: <gcc-bugs-return-456232-listarch-gcc-bugs=gcc.gnu.org@gcc.gnu.org>
Delivered-To: listarch-gcc-bugs@gcc.gnu.org
Received: (qmail 2195 invoked by alias); 13 Jul 2014 18:37:34 -0000
Mailing-List: contact gcc-bugs-help@gcc.gnu.org; run by ezmlm
Precedence: bulk
List-Id: <gcc-bugs.gcc.gnu.org>
List-Archive: <http://gcc.gnu.org/ml/gcc-bugs/>
List-Post: <mailto:gcc-bugs@gcc.gnu.org>
List-Help: <mailto:gcc-bugs-help@gcc.gnu.org>
Sender: gcc-bugs-owner@gcc.gnu.org
Delivered-To: mailing list gcc-bugs@gcc.gnu.org
Received: (qmail 1909 invoked by uid 48); 13 Jul 2014 18:37:30 -0000
From: "anlauf at gmx dot de" <gcc-bugzilla@gcc.gnu.org>
To: gcc-bugs@gcc.gnu.org
Subject: [Bug fortran/61615] Failure to resolve correct generic with TYPE(C_PTR) arguments
Date: Sun, 13 Jul 2014 18:37:00 -0000
X-Bugzilla-Reason: CC
X-Bugzilla-Type: changed
X-Bugzilla-Watch-Reason: None
X-Bugzilla-Product: gcc
X-Bugzilla-Component: fortran
X-Bugzilla-Version: 4.10.0
X-Bugzilla-Keywords: wrong-code
X-Bugzilla-Severity: normal
X-Bugzilla-Who: anlauf at gmx dot de
X-Bugzilla-Status: WAITING
X-Bugzilla-Priority: P3
X-Bugzilla-Assigned-To: unassigned at gcc dot gnu.org
X-Bugzilla-Target-Milestone: ---
X-Bugzilla-Flags:
X-Bugzilla-Changed-Fields:
Message-ID: <bug-61615-4-4iRxVJ52Vm@http.gcc.gnu.org/bugzilla/>
In-Reply-To: <bug-61615-4@http.gcc.gnu.org/bugzilla/>
References: <bug-61615-4@http.gcc.gnu.org/bugzilla/>
Content-Type: text/plain; charset="UTF-8"
Content-Transfer-Encoding: 7bit
X-Bugzilla-URL: http://gcc.gnu.org/bugzilla/
Auto-Submitted: auto-generated
MIME-Version: 1.0
X-SW-Source: 2014-07/txt/msg00823.txt.bz2
Content-length: 723

https://gcc.gnu.org/bugzilla/show_bug.cgi?ida615

--- Comment #7 from Harald Anlauf <anlauf at gmx dot de> ---
(In reply to Dominique d'Humieres from comment #6)

Well, the problem is in the resolution of the generic,
not in the invocation in the main.

If you comment out

!   MODULE PROCEDURE bar_a1d

you get the result desired by the author:

 in bar_s

Now if you comment out the module procedure bar_s,
gfortran does *not* complain, but resolve to the *wrong* procedure.
So for some strange reason, gfortran fails to recognise the proper rank.

OTOH, NAG complains:

Error: pr61615.f90, line 28: No specific match for reference to generic BAR

I get corresponding errors from PGI or Intel.

IMO a bug in gfortran.

[Bug libstdc++/61795] [C++11] return type of std::pow(std::complex<float>, int) should be std::complex<double>

[kariya_mitsuru at hotmail dot com]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=61795

--- Comment #2 from Mitsuru Kariya <kariya_mitsuru at hotmail dot com> ---
I think that this behaviour is caused by r201253 (for PR57974, Comment 11).
DR844 was fixed by r136694 but reverted by r201253.

diff r135878 r136694
https://gcc.gnu.org/viewcvs/gcc/trunk/libstdc%2B%2B-v3/include/std/complex?r1=135878&r2=136694

diff r199924 r201253
https://gcc.gnu.org/viewcvs/gcc/trunk/libstdc%2B%2B-v3/include/std/complex?r1=199924&r2=201253


Moreover, I think that I mistook.

> Note that this problem does not occur in std::complex<double> and
> std::complex<long double> because there is no difference between
> C++03 and C++11.

This is not true.

In C++03, the 2nd argument of std::pow can cause implicit conversions.
(Because it is the trivial int type.)
However, I believe that it should cause no implicit conversion in C++11.
(I think so from C++11 standard text quoted above.)

Therefore, I think that the sample code below should be compiled successfully
in C++03 mode but should cause compilation error in C++11 mode.

===================================================================
#include <complex>

struct S {
  operator int() { return 1; }
};

int main()
{
  std::complex<double> d = std::pow(std::complex<double>(0), S());
}
===================================================================

[Bug libstdc++/61795] [C++11] return type of std::pow(std::complex<float>, int) should be std::complex<double>

[redi at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=61795

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

           What    |Removed                     |Added
----------------------------------------------------------------------------
             Status|UNCONFIRMED                 |NEW
   Last reconfirmed|                            |2015-04-09
     Ever confirmed|0                           |1