[mlpack-git] [mlpack] std::size_t warning on 32-bit machines (#553)

Ryan Curtin notifications at github.com
Mon Mar 7 12:11:52 EST 2016


When you compile mlpack on 32-bit machines, you get lots and lots of warnings of this ilk:

```
[  4%] Building CXX object src/mlpack/CMakeFiles/mlpack.dir/core/math/columns_to_blocks.cpp.o
cd /home/ryan/rpmbuild/BUILD/mlpack-2.0.1/src/mlpack && /usr/lib/ccache/c++   -DARMA_NO_DEBUG -DBACKTRACE_FOUND -DBACKTRACE_HEADER=\"execinfo.h\" -DBOOST_TEST_DYN_LINK -DNDEBUG -Dmlpack_EXPORTS -O2 -g -pipe -Wall -Werror=format-security -Wp,-D_FORTIFY_SOURCE=2 -fexceptions -fstack-protector-strong --param=ssp-buffer-size=4 -grecord-gcc-switches -specs=/usr/lib/rpm/redhat/redhat-hardened-cc1 -m32 -march=i686 -mtune=atom -fasynchronous-unwind-tables  -std=c++0x -Wall -Wextra -ftemplate-depth=1000 -O3 -fopenmp -fPIC -I/home/ryan/rpmbuild/BUILD/mlpack-2.0.1 -I/home/ryan/rpmbuild/BUILD/mlpack-2.0.1/src/mlpack/..    -o CMakeFiles/mlpack.dir/core/math/columns_to_blocks.cpp.o -c /home/ryan/rpmbuild/BUILD/mlpack-2.0.1/src/mlpack/core/math/columns_to_blocks.cpp
In file included from /usr/include/armadillo:48:0,
                 from /home/ryan/rpmbuild/BUILD/mlpack-2.0.1/src/mlpack/../mlpack/core/arma_extend/arma_extend.hpp:33,
                 from /home/ryan/rpmbuild/BUILD/mlpack-2.0.1/src/mlpack/../mlpack/prereqs.hpp:76,
                 from /home/ryan/rpmbuild/BUILD/mlpack-2.0.1/src/mlpack/../mlpack/core.hpp:196,
                 from /home/ryan/rpmbuild/BUILD/mlpack-2.0.1/src/mlpack/core/dists/gaussian_distribution.hpp:18,
                 from /home/ryan/rpmbuild/BUILD/mlpack-2.0.1/src/mlpack/core/dists/gaussian_distribution.cpp:15:
/usr/include/armadillo_bits/compiler_setup.hpp:87:100: note: #pragma message: WARNING: disabled use of 64 bit integers, as std::size_t is smaller than 64 bits
       #pragma message ("WARNING: disabled use of 64 bit integers, as std::size_t is smaller than 64 bits")
```

We should try and figure out what is triggering these warnings and whether or not we can disable it, so that we can have a less noisy 32-bit build.  These warnings aren't an issue---mlpack doesn't use any 64-bit types if it's on a 32-bit build.

---
Reply to this email directly or view it on GitHub:
https://github.com/mlpack/mlpack/issues/553
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://mailman.cc.gatech.edu/pipermail/mlpack-git/attachments/20160307/0ae363c0/attachment.html>


More information about the mlpack-git mailing list