msgpack-node
msgpack-node copied to clipboard
Published
20 hours ago •
msgpack
msgpack
electron-rebuild and g++ issue
I using electron and when I run electron-rebuild command it shows such error:
CXX(target) Release/obj.target/msgpackBinding/src/msgpack.o
In file included from /home/pi/.electron-gyp/11.2.3/include/node/v8.h:30,
from ../src/msgpack.cc:1:
/home/pi/.electron-gyp/11.2.3/include/node/v8-internal.h: In function ‘void v8::internal::PerformCastCheck(T*)’:
/home/pi/.electron-gyp/11.2.3/include/node/v8-internal.h:418:38: error: ‘remove_cv_t’ is not a member of ‘std’
!std::is_same<Data, std::remove_cv_t<T>>::value>::Perform(data);
^~~~~~~~~~~
/home/pi/.electron-gyp/11.2.3/include/node/v8-internal.h:418:38: note: suggested alternative: ‘remove_cv’
!std::is_same<Data, std::remove_cv_t<T>>::value>::Perform(data);
^~~~~~~~~~~
remove_cv
/home/pi/.electron-gyp/11.2.3/include/node/v8-internal.h:418:38: error: ‘remove_cv_t’ is not a member of ‘std’
/home/pi/.electron-gyp/11.2.3/include/node/v8-internal.h:418:38: note: suggested alternative: ‘remove_cv’
!std::is_same<Data, std::remove_cv_t<T>>::value>::Perform(data);
^~~~~~~~~~~
remove_cv
/home/pi/.electron-gyp/11.2.3/include/node/v8-internal.h:418:50: error: template argument 2 is invalid
!std::is_same<Data, std::remove_cv_t<T>>::value>::Perform(data);
^
/home/pi/.electron-gyp/11.2.3/include/node/v8-internal.h:418:63: error: ‘::Perform’ has not been declared
!std::is_same<Data, std::remove_cv_t<T>>::value>::Perform(data);
^~~~~~~
/home/pi/.electron-gyp/11.2.3/include/node/v8-internal.h:418:63: note: suggested alternative: ‘perror’
!std::is_same<Data, std::remove_cv_t<T>>::value>::Perform(data);
^~~~~~~
perror
This issue appears because we have -std=c++11 flag in binding.gyp file:
'cflags_cc': [
'-Wall',
'-O3',
'-std=c++11',
],
Changing this flag to -std=c++14 or removing it fixes the issue.
I am working on linux with gcc-8.
I have a similar issue here when building with node-gyp manually from the source.
- Operating system: macOS 11.3.1
- Apple clang version: 12.0.5 (clang-1205.0.22.9)
# build from the root directory
$ node-gyp rebuild
# or module directory
$ npx node-gyp rebuild -C ./node_modules/msgpack
gyp info it worked if it ends with ok
gyp info using [email protected]
gyp info using [email protected] | darwin | x64
gyp info chdir ./node_modules/msgpack
[...]
gyp info spawn args [ 'BUILDTYPE=Release', '-C', 'build' ]
CC(target) Release/obj.target/libmsgpack/deps/msgpack/objectc.o
CC(target) Release/obj.target/libmsgpack/deps/msgpack/unpack.o
CC(target) Release/obj.target/libmsgpack/deps/msgpack/vrefbuffer.o
CC(target) Release/obj.target/libmsgpack/deps/msgpack/zone.o
CC(target) Release/obj.target/libmsgpack/deps/msgpack/version.o
LIBTOOL-STATIC Release/msgpack.a
CXX(target) Release/obj.target/msgpackBinding/src/msgpack.o
In file included from ../src/msgpack.cc:1:
In file included from /Users/user/Library/Caches/node-gyp/16.1.0/include/node/v8.h:30:
/Users/user/Library/Caches/node-gyp/16.1.0/include/node/v8-internal.h:452:38: error: no template named 'remove_cv_t' in namespace 'std'; did you mean 'remove_cv'?
!std::is_same<Data, std::remove_cv_t<T>>::value>::Perform(data);
~~~~~^~~~~~~~~~~
remove_cv
/Applications/Xcode.app/Contents/Developer/Platforms/MacOSX.platform/Developer/SDKs/MacOSX.sdk/usr/include/c++/v1/type_traits:776:50: note: 'remove_cv' declared here
template <class _Tp> struct _LIBCPP_TEMPLATE_VIS remove_cv
^
1 error generated.
make: *** [Release/obj.target/msgpackBinding/src/msgpack.o] Error 1
As @systemmind said, this issue could be fixed by removing the -std=c++11 in cflags_cc attribute in binding.gyp file. In macOS, I had to remove the 'OTHER_CFLAGS': ['-std=c++11'] options as well.
Fix:
{
"targets": [
{
[...]
'cflags_cc': [
'-Wall',
'-O3'
],
[...]
'conditions': [
['OS=="mac"', {
'configurations': {
'Debug': {
'xcode_settings': {
'GCC_ENABLE_CPP_EXCEPTIONS': 'YES',
'WARNING_CFLAGS': ['-Wall', '-Wno-unused-function']
}
},
'Release': {
'xcode_settings': {
'GCC_ENABLE_CPP_EXCEPTIONS': 'YES',
'WARNING_CFLAGS': ['-Wall', '-Wno-unused-function']
},
},
},
}],
[...]
@godsflaw Can I send a PR for this issue?
