Stop packaging GPL libquadmath.so

cd/mxnet_lib/static/Jenkins_pipeline.groovy

@@ -32,8 +32,8 @@ libmxnet = 'lib/libmxnet.so'
 licenses = 'licenses/*'
 
 // libmxnet dependencies
-mx_native_deps = 'lib/libgfortran.so.4, lib/libquadmath.so.0, lib/libopenblas.so.0'
-mx_deps = 'lib/libgfortran.so.4, lib/libquadmath.so.0, lib/libopenblas.so.0, include/mkldnn/dnnl_version.h, include/mkldnn/dnnl_config.h'
+mx_native_deps = 'lib/libgfortran.so.4, lib/libopenblas.so.0'
+mx_deps = 'lib/libgfortran.so.4, lib/libopenblas.so.0, include/mkldnn/dnnl_version.h, include/mkldnn/dnnl_config.h'
 
 // library type
 // either static or dynamic - depending on how it links to its dependencies

ci/jenkins/Jenkins_steps.groovy

@@ -37,7 +37,7 @@ mx_tensorrt_lib = 'build/libmxnet.so, build/3rdparty/tvm/libtvm_runtime.so, buil
 mx_lib_cpp_examples = 'build/libmxnet.so, build/3rdparty/tvm/libtvm_runtime.so, build/libtvmop.so, build/tvmop.conf, build/3rdparty/openmp/runtime/src/libomp.so, build/libcustomop_lib.so, build/libcustomop_gpu_lib.so, build/libsubgraph_lib.so, python/mxnet/_cy3/*.so, python/mxnet/_ffi/_cy3/*.so'
 mx_lib_cpp_examples_no_tvm_op = 'build/libmxnet.so, build/libcustomop_lib.so, build/libcustomop_gpu_lib.so, build/libsubgraph_lib.so, build/3rdparty/openmp/runtime/src/libomp.so, python/mxnet/_cy3/*.so, python/mxnet/_ffi/_cy3/*.so'
 mx_lib_cpp_examples_cpu = 'build/libmxnet.so, build/3rdparty/tvm/libtvm_runtime.so, build/libtvmop.so, build/tvmop.conf, build/3rdparty/openmp/runtime/src/libomp.so'
-mx_cd_lib = 'lib/libmxnet.so, licenses/*, lib/libgfortran.so.4, lib/libquadmath.so.0, lib/libopenblas.so.0, include/mkldnn/dnnl_version.h, include/mkldnn/dnnl_config.h'
+mx_cd_lib = 'lib/libmxnet.so, licenses/*, lib/libgfortran.so.4, lib/libopenblas.so.0, include/mkldnn/dnnl_version.h, include/mkldnn/dnnl_config.h'
 
 // Python unittest for CPU
 // Python 3

tools/pip/doc/CPU_ADDITIONAL.md

@@ -25,6 +25,14 @@ This package supports Linux, Mac OSX, and Windows platforms. You may also want t
 - [mxnet](https://pypi.python.org/pypi/mxnet/).
 - [mxnet-native](https://pypi.python.org/pypi/mxnet-native/) CPU variant without MKLDNN.
 
+To use this package on Linux you need the `libquadmath.so.0` shared library. On
+Debian based systems, including Ubuntu, run `sudo apt install libquadmath0` to
+install the shared library. On RHEL based systems, including CentOS, run `sudo
+yum install libquadmath` to install the shared library. As `libquadmath.so.0` is
+a GPL library and MXNet part of the Apache Software Foundation, MXNet must not
+redistribute `libquadmath.so.0` as part of the Pypi package and users must
+manually install it.
+
 To install for other platforms (e.g. Windows, Raspberry Pi/ARM) or other versions, check [Installing MXNet](https://mxnet.apache.org/versions/master/install/index.html) for instructions on building from source.
 
 Installation

tools/pip/doc/CU100_ADDITIONAL.md

@@ -27,6 +27,14 @@ This package supports Linux and Windows platforms. You may also want to check:
 
 To download CUDA, check [CUDA download](https://developer.nvidia.com/cuda-downloads). For more instructions, check [CUDA Toolkit online documentation](http://docs.nvidia.com/cuda/index.html).
 
+To use this package on Linux you need the `libquadmath.so.0` shared library. On
+Debian based systems, including Ubuntu, run `sudo apt install libquadmath0` to
+install the shared library. On RHEL based systems, including CentOS, run `sudo
+yum install libquadmath` to install the shared library. As `libquadmath.so.0` is
+a GPL library and MXNet part of the Apache Software Foundation, MXNet must not
+redistribute `libquadmath.so.0` as part of the Pypi package and users must
+manually install it.
+
 To install for other platforms (e.g. Windows, Raspberry Pi/ARM) or other versions, check [Installing MXNet](https://mxnet.apache.org/versions/master/install/index.html) for instructions on building from source.
 
 Installation

tools/pip/doc/CU101_ADDITIONAL.md

@@ -27,6 +27,14 @@ This package supports Linux and Windows platforms. You may also want to check:
 
 To download CUDA, check [CUDA download](https://developer.nvidia.com/cuda-downloads). For more instructions, check [CUDA Toolkit online documentation](http://docs.nvidia.com/cuda/index.html).
 
+To use this package on Linux you need the `libquadmath.so.0` shared library. On
+Debian based systems, including Ubuntu, run `sudo apt install libquadmath0` to
+install the shared library. On RHEL based systems, including CentOS, run `sudo
+yum install libquadmath` to install the shared library. As `libquadmath.so.0` is
+a GPL library and MXNet part of the Apache Software Foundation, MXNet must not
+redistribute `libquadmath.so.0` as part of the Pypi package and users must
+manually install it.
+
 To install for other platforms (e.g. Windows, Raspberry Pi/ARM) or other versions, check [Installing MXNet](https://mxnet.incubator.apache.org/versions/master/install/index.html) for instructions on building from source.
 
 Installation

tools/pip/doc/CU102_ADDITIONAL.md

@@ -27,6 +27,14 @@ This package supports Linux and Windows platforms. You may also want to check:
 
 To download CUDA, check [CUDA download](https://developer.nvidia.com/cuda-downloads). For more instructions, check [CUDA Toolkit online documentation](http://docs.nvidia.com/cuda/index.html).
 
+To use this package on Linux you need the `libquadmath.so.0` shared library. On
+Debian based systems, including Ubuntu, run `sudo apt install libquadmath0` to
+install the shared library. On RHEL based systems, including CentOS, run `sudo
+yum install libquadmath` to install the shared library. As `libquadmath.so.0` is
+a GPL library and MXNet part of the Apache Software Foundation, MXNet must not
+redistribute `libquadmath.so.0` as part of the Pypi package and users must
+manually install it.
+
 To install for other platforms (e.g. Windows, Raspberry Pi/ARM) or other versions, check [Installing MXNet](https://mxnet.incubator.apache.org/versions/master/install/index.html) for instructions on building from source.
 
 Installation

tools/pip/doc/CU92_ADDITIONAL.md

@@ -26,6 +26,14 @@ This package supports Linux and Windows platforms. You may also want to check:
 
 To download CUDA, check [CUDA download](https://developer.nvidia.com/cuda-downloads). For more instructions, check [CUDA Toolkit online documentation](http://docs.nvidia.com/cuda/index.html).
 
+To use this package on Linux you need the `libquadmath.so.0` shared library. On
+Debian based systems, including Ubuntu, run `sudo apt install libquadmath0` to
+install the shared library. On RHEL based systems, including CentOS, run `sudo
+yum install libquadmath` to install the shared library. As `libquadmath.so.0` is
+a GPL library and MXNet part of the Apache Software Foundation, MXNet must not
+redistribute `libquadmath.so.0` as part of the Pypi package and users must
+manually install it.
+
 To install for other platforms (e.g. Windows, Raspberry Pi/ARM) or other versions, check [Installing MXNet](https://mxnet.apache.org/versions/master/install/index.html) for instructions on building from source.
 
 Installation

tools/pip/doc/NATIVE_ADDITIONAL.md

@@ -27,6 +27,14 @@ This package supports Linux and Windows platforms. You may also want to check:
 
 To download CUDA, check [CUDA download](https://developer.nvidia.com/cuda-downloads). For more instructions, check [CUDA Toolkit online documentation](http://docs.nvidia.com/cuda/index.html).
 
+To use this package on Linux you need the `libquadmath.so.0` shared library. On
+Debian based systems, including Ubuntu, run `sudo apt install libquadmath0` to
+install the shared library. On RHEL based systems, including CentOS, run `sudo
+yum install libquadmath` to install the shared library. As `libquadmath.so.0` is
+a GPL library and MXNet part of the Apache Software Foundation, MXNet must not
+redistribute `libquadmath.so.0` as part of the Pypi package and users must
+manually install it.
+
 To install for other platforms (e.g. Windows, Raspberry Pi/ARM) or other versions, check [Installing MXNet](https://mxnet.apache.org/versions/master/install/index.html) for instructions on building from source.
 
 Installation

tools/pip/setup.py

@@ -155,8 +155,6 @@ def skip_markdown_comments(md):
     else:
         shutil.copy(os.path.join(libdir, 'libgfortran.so.4'), mxdir)
         package_data['mxnet'].append('mxnet/libgfortran.so.4')
-    shutil.copy(os.path.join(libdir, 'libquadmath.so.0'), mxdir)
-    package_data['mxnet'].append('mxnet/libquadmath.so.0')
     if os.path.exists(os.path.join(libdir, 'libopenblas.so.0')):
         shutil.copy(os.path.join(libdir, 'libopenblas.so.0'), mxdir)
         package_data['mxnet'].append('mxnet/libopenblas.so.0')

tools/staticbuild/build_lib.sh

@@ -40,7 +40,6 @@ if [[ $PLATFORM == 'linux' ]]; then
     cp -L build/libmxnet.so lib/libmxnet.so
     cp -L staticdeps/lib/libopenblas.so lib/libopenblas.so.0
     cp -L $(ldd lib/libmxnet.so | grep libgfortran |  awk '{print $3}') lib/
-    cp -L $(ldd lib/libmxnet.so | grep libquadmath |  awk '{print $3}') lib/
 elif [[ $PLATFORM == 'darwin' ]]; then
     cp -L build/libmxnet.dylib lib/libmxnet.dylib
 fi