[Bug libstdc++/82417] Macros from defined in C++11

[redi at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=82417

Jonathan Wakely  changed:

   What|Removed |Added

 Status|ASSIGNED|RESOLVED
   See Also||https://gcc.gnu.org/bugzill
   ||a/show_bug.cgi?id=82503
 Resolution|--- |FIXED
   Target Milestone|--- |8.0

--- Comment #6 from Jonathan Wakely  ---
Fixed on trunk. Bug 82503 opened for the request to add a warning.

[Bug libstdc++/82417] Macros from defined in C++11

[redi at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=82417

--- Comment #5 from Jonathan Wakely  ---
Author: redi
Date: Tue Oct 10 11:25:45 2017
New Revision: 253581

URL: https://gcc.gnu.org/viewcvs?rev=253581=gcc=rev
Log:
PR libstdc++/82417 do not include C99  in strict modes

As an extension the C++ version of  includes the C version,
but that defines macros with non-reserved names that should not be
defined in ISO C++. Only include the C header for non-strict modes, or
for pre-C++11 (because C++98 doesn't mention  at all).

PR libstdc++/59087
PR libstdc++/82417
* include/c_compatibility/complex.h [!C++98 && __STRICT_ANSI__]: Do
not include C library's .
* testsuite/26_numerics/complex/c99.cc: Depend on __STRICT_ANSI__.
* testsuite/26_numerics/headers/complex.h/std_c++11.h: New test.
* testsuite/26_numerics/headers/complex.h/std_gnu++11.h: New test.
* testsuite/26_numerics/headers/complex.h/std_c++98.h: New test.

Added:
trunk/libstdc++-v3/testsuite/26_numerics/headers/complex.h/
trunk/libstdc++-v3/testsuite/26_numerics/headers/complex.h/std_c++11.h
  - copied, changed from r253550,
trunk/libstdc++-v3/testsuite/26_numerics/complex/c99.cc
trunk/libstdc++-v3/testsuite/26_numerics/headers/complex.h/std_c++98.h
  - copied, changed from r253550,
trunk/libstdc++-v3/testsuite/26_numerics/complex/c99.cc
trunk/libstdc++-v3/testsuite/26_numerics/headers/complex.h/std_gnu++11.h
  - copied, changed from r253550,
trunk/libstdc++-v3/testsuite/26_numerics/complex/c99.cc
Modified:
trunk/libstdc++-v3/ChangeLog
trunk/libstdc++-v3/include/c_compatibility/complex.h
trunk/libstdc++-v3/testsuite/26_numerics/complex/c99.cc

[Bug libstdc++/82417] Macros from defined in C++11

[redi at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=82417

--- Comment #4 from Jonathan Wakely  ---
I'm not opposed to a warning in principle, but I'd prefer it to be one
controllable by a -Wxxx option. I don't know how to do that using the
preprocessor, so we'd have to do something like:

#ifndef _GLIBCXX_NO_CCOMPLEX_WARNING
# warning "Including  makes no sense. Define
_GLIBCXX_NO_CCOMPLEX_WARNING to disable this warning."
#endif

And that is a bit ugly.

[Bug libstdc++/82417] Macros from defined in C++11

[glisse at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=82417

--- Comment #3 from Marc Glisse  ---
(In reply to Jonathan Wakely from comment #2)
> Thinking about this further, I think we must not include  at all
> for strict -std=c++1* modes,

Yes.

Can we get a #warning in that case which explains that including  in
strict C++11+ mode makes no sense? Actually,  could also do with a
#warning explaining that it never makes sense to include it.

[Bug libstdc++/82417] Macros from defined in C++11

[redi at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=82417

Jonathan Wakely  changed:

   What|Removed |Added

 Status|NEW |ASSIGNED
   Assignee|unassigned at gcc dot gnu.org  |redi at gcc dot gnu.org

--- Comment #2 from Jonathan Wakely  ---
Thinking about this further, I think we must not include  at all for
strict -std=c++1* modes, because users could have defined their own macros with
those names, but the C header #undefs them and then redefines them with
different meaning:

#define I 1
#define imaginary imag
#include 
std::complex n{0, I};
double zero = n.imaginary();

This explodes horribly.

We can't just #undef the names after including  because the user
macros will already have been trashed by then.

[Bug libstdc++/82417] Macros from defined in C++11

[glisse at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=82417

--- Comment #1 from Marc Glisse  ---
(In reply to Jonathan Wakely from comment #0)
> The C++11 standard says that  should just include the C++
>  header and completely ignore the C library's header.

I am very surprised that nobody has cared enough to get the standard fixed. But
I can't complain, I didn't write a proposal either.

> For C++11 mode we should #undef the macros that  defines with
> non-reserved names, and maybe consider not including  at all for
> -std=c++1* modes.

I guess so.

[Bug libstdc++/82417] Macros from defined in C++11

[redi at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=82417

Jonathan Wakely  changed:

   What|Removed |Added

 Status|UNCONFIRMED |NEW
   Last reconfirmed||2017-10-03
   See Also||https://gcc.gnu.org/bugzill
   ||a/show_bug.cgi?id=56111
 Ever confirmed|0   |1