MinGW: C++11 perfect forwarding support is broken

David Ostrovsky d.ostrovsky at idaia.de
Sun Feb 24 09:33:15 PST 2013


Am 24.02.2013 02:23, schrieb Gábor Stefanik:
> On Sun, Feb 24, 2013 at 12:43 AM, David Ostrovsky
> <david.ostrovsky at gmx.de>  wrote:
>> >So we have weird breakage on MinGW:
>> >
>> >in configure phase i686-w64-mingw32-g++ reports back success for that check:
>> >
>> >configure:15247: checking whether ccache i686-w64-mingw32-g++ supports C++11
>> >perfect forwarding
>> >configure:15275: ccache i686-w64-mingw32-g++ -c  -std=gnu++0x conftest.cpp
>>> >>&5
> Apparently gnu++0x is no longer recognized as a synonym of gnu++11.
> There are several references in the error log to features "only
> available with -std=c++11 or -std=gnu++11".
>
no it does, it was late yesterday:

diff --git a/solenv/gbuild/platform/WNT_INTEL_GCC.mk 
b/solenv/gbuild/platform/WNT_INTEL_GCC.mk
index b9edab0..70a4608 100644
--- a/solenv/gbuild/platform/WNT_INTEL_GCC.mk
+++ b/solenv/gbuild/platform/WNT_INTEL_GCC.mk
@@ -87,6 +87,7 @@ endif
  ifeq ($(HAVE_CXX0X),TRUE)
  # We can turn on additional useful checks with c++0x
  # FIXME still does not compile fully gb_CXXFLAGS += -std=gnu++0x
+gb_CXXFLAGS += -std=gnu++0x
  endif

did the trick. It was activated during the configure, but deactivated 
during compilation.
Not sure if i should commit that?

David



More information about the LibreOffice mailing list