nix-community / nix-index

Quickly locate nix packages with specific files [maintainers=@bennofs @figsoda @raitobezarius]
Other
821 stars 50 forks source link

Fails with blas=mkl #133

Open tbenst opened 3 years ago

tbenst commented 3 years ago

On release-20.09, If I use the overlay:

self: super:
{
  pythonOverrides = python-self: python-super: {
    numpy = python-super.numpy.override { blas = super.mkl; };
  };
  python37 = super.python37.override {packageOverrides = self.pythonOverrides;};
}

Then nix-index fails with:

nix-index --show-trace
+ querying available packages
error: querying available packages failed
caused by: nix-env failed with error: nix-env failed with exit code 1:
error: while evaluating 'toPythonModule' at /Computer/nixpkgs/pkgs/top-level/python-packages.nix:76:20, called from /Computer/nixpkgs/pkgs/top-level/python-packages.nix:1011:11:
while evaluating anonymous function at /Computer/nixpkgs/lib/customisation.nix:77:32, called from /Computer/nixpkgs/pkgs/top-level/python-packages.nix:1011:27:
while evaluating 'makeOverridable' at /Computer/nixpkgs/lib/customisation.nix:67:24, called from /Computer/nixpkgs/lib/customisation.nix:77:41:
while evaluating anonymous function at /Computer/nixpkgs/pkgs/applications/science/math/caffe/default.nix:1:1, called from /Computer/nixpkgs/lib/customisation.nix:69:16:
while evaluating the attribute 'numpy' at /home/tyler/.config/nixpkgs/overlays/python.nix:33:5:
while evaluating anonymous function at /Computer/nixpkgs/lib/customisation.nix:77:32, called from /home/tyler/.config/nixpkgs/overlays/python.nix:33:13:
while evaluating 'makeOverridable' at /Computer/nixpkgs/lib/customisation.nix:67:24, called from /Computer/nixpkgs/lib/customisation.nix:77:41:
while evaluating anonymous function at /Computer/nixpkgs/pkgs/development/python-modules/numpy/default.nix:1:1, called from /Computer/nixpkgs/lib/customisation.nix:69:16:
attribute 'isILP64' missing, at /Computer/nixpkgs/pkgs/development/python-modules/numpy/default.nix:16:10

This is despite nix-shell -p python3Packages.numpy succeeding.

cc @matthewbauer