Closed
Description
I get the following error when installing in MSYS2 with MinGW-w64.
Reproduction steps:
- Follow all steps at https://www.msys2.org/ to install MSYS2 and mingw-w64.
- In a MSYS2 / MinGW-w64 64-bit console:
pacman -S mingw-w64-x86_64-python-pip
python3 -m pip install wheel
# in numpy git root
python3 -m pip install .
(Note: You may need to apply the 2-line change in #18539 first)
Error message
x86_64-w64-mingw32-gcc -Wno-unused-result -Wsign-compare -DNDEBUG -march=x86-64 -mtune=generic -O2 -pipe -fwrapv -D__USE_MINGW_ANSI_STDIO=1 -D_WIN32_WINNT=0x0601 -DNDEBUG -march=x86-64 -mtune=generic -O2 -pipe -fwrapv -D__USE_MINGW_ANSI_STDIO=1 -D_WIN32_WINNT=0x0601 -DNDEBUG -DNPY_INTERNAL_BUILD=1 -DHAVE_NPY_CONFIG_H=1 -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE=1 -D_LARGEFILE64_SOURCE=1 -Inumpy/core/include -Ibuild/src.mingw-3.8/numpy/core/include/numpy -Ibuild/src.mingw-3.8/numpy/distutils/include -Ibuild/src.mingw-3.8/numpy/core/src/umath -Inumpy/core/src/common -Inumpy/core/src -Inumpy/core -Inumpy/core/src/npymath -Inumpy/core/src/multiarray -Inumpy/core/src/umath -Inumpy/core/src/npysort -Inumpy/core/src/_simd -IC:/msys64/mingw64/include/python3.8 -Ibuild/src.mingw-3.8/numpy/core/src/common -Ibuild/src.mingw-3.8/numpy/core/src/npymath -c build/src.mingw-3.8/numpy/core/src/common/npy_cpu_features.c -o build/temp.mingw-3.8/build/src.mingw-3.8/numpy/core/src/common/npy_cpu_features.o
In file included from numpy/core/src/common/npy_cpu_features.c.src:3:
numpy/core/include/numpy/npy_common.h:185:20: warning: '_fseeki64' redeclared without dllimport attribute: previous dllimport ignored [-Wattributes]
185 | extern int __cdecl _fseeki64(FILE *, long long, int);
| ^~~~~~~~~
numpy/core/include/numpy/npy_common.h:186:26: warning: '_ftelli64' redeclared without dllimport attribute: previous dllimport ignored [-Wattributes]
186 | extern long long __cdecl _ftelli64(FILE *);
| ^~~~~~~~~
cc1.exe: all warnings being treated as errors
Note
This may not even be worth fixing, as it appears that the package maintainers for MinGW are already doing something to work around this; the package mingw64-w64-x86_64-python-numpy provides a fully-up-to-date numpy 1.20.1.
NumPy/Python version information:
- numpy github master (f691b02)
$ python3 --version
Python 3.8.8
$ which python3
/mingw64/bin/python3
$ x86_64-w64-mingw32-gcc --version
x86_64-w64-mingw32-gcc.exe (Rev6, Built by MSYS2 project) 10.2.0
Metadata
Metadata
Assignees
Labels
No labels