3
0
Fork 0
forked from mirrors/nixpkgs
nixpkgs/pkgs/development/python-modules/numpy/default.nix
Matthew Bauer 8bdb2a058e numpy: set lapack, blas sections
When using netlib lapack/blas, the section name doesn’t match what
numpy expects. So we need to add extra sections for both so that the
right directory is found. The original “blas.implementation” section
may not actually be still required, but it is still a good idea so
that numpy know whether to apply any blas-implementation specific
quirks.

Fixes #86613
2020-05-08 06:47:30 +02:00

95 lines
2.3 KiB
Nix

{ lib
, fetchPypi
, python
, buildPythonPackage
, gfortran
, pytest
, blas
, lapack
, writeTextFile
, isPyPy
, cython
, setuptoolsBuildHook
}:
assert (!blas.isILP64) && (!lapack.isILP64);
let
cfg = writeTextFile {
name = "site.cfg";
text = (lib.generators.toINI {} {
${blas.implementation} = {
include_dirs = "${lib.getDev blas}/include:${lib.getDev lapack}/include";
library_dirs = "${blas}/lib:${lapack}/lib";
libraries = "lapack,lapacke,blas,cblas";
};
lapack = {
include_dirs = "${lib.getDev lapack}/include";
library_dirs = "${lapack}/lib";
};
blas = {
include_dirs = "${lib.getDev blas}/include";
library_dirs = "${blas}/lib";
};
});
};
in buildPythonPackage rec {
pname = "numpy";
version = "1.18.3";
format = "pyproject.toml";
src = fetchPypi {
inherit pname version;
extension = "zip";
sha256 = "e46e2384209c91996d5ec16744234d1c906ab79a701ce1a26155c9ec890b8dc8";
};
nativeBuildInputs = [ gfortran pytest cython setuptoolsBuildHook ];
buildInputs = [ blas lapack ];
patches = lib.optionals python.hasDistutilsCxxPatch [
# We patch cpython/distutils to fix https://bugs.python.org/issue1222585
# Patching of numpy.distutils is needed to prevent it from undoing the
# patch to distutils.
./numpy-distutils-C++.patch
];
preConfigure = ''
sed -i 's/-faltivec//' numpy/distutils/system_info.py
export NPY_NUM_BUILD_JOBS=$NIX_BUILD_CORES
'';
preBuild = ''
ln -s ${cfg} site.cfg
'';
enableParallelBuilding = true;
doCheck = !isPyPy; # numpy 1.16+ hits a bug in pypy's ctypes, using either numpy or pypy HEAD fixes this (https://github.com/numpy/numpy/issues/13807)
checkPhase = ''
runHook preCheck
pushd dist
${python.interpreter} -c 'import numpy; numpy.test("fast", verbose=10)'
popd
runHook postCheck
'';
passthru = {
# just for backwards compatibility
blas = blas.provider;
blasImplementation = blas.implementation;
inherit cfg;
};
# Disable test
# - test_large_file_support: takes a long time and can cause the machine to run out of disk space
NOSE_EXCLUDE="test_large_file_support";
meta = {
description = "Scientific tools for Python";
homepage = "https://numpy.org/";
maintainers = with lib.maintainers; [ fridh ];
};
}