You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
The Linux builds pass successfully in these setups:
cp37-manylinux_x86_64 wheel
cp38-manylinux_x86_64 wheel
cp39-manylinux_x86_64 wheel
cp310-manylinux_x86_64 wheel
but then it moves to musllinux and fails at:
cp37-musllinux_x86_64 wheel
The reported error is that BLAS is not available when attempting to build Scipy and Numpy. Apparently pip doesn't locate pre-built wheels on PyPi. As far as I understand, this is expected since there are no musllinux wheels for Numpy for example, as discussed here. In the case of manylinux containers, there is no such problem, for example:
If I understand things right, the issue of musllinux having trouble with dependencies has already been raised here.
So I'm trying to understand if my analysis looks right, and if the straightforward solution is to disable musllinux at least until there will be numpy and spicy wheels for this image?
reacted with thumbs up emoji reacted with thumbs down emoji reacted with laugh emoji reacted with hooray emoji reacted with confused emoji reacted with heart emoji reacted with rocket emoji reacted with eyes emoji
-
I recently updated to cibuildwheels 2.9 and to action-based build.yml and encountered some issues related to musllinux.
This is my project: https://github.com/IBM/sib
It depends on SciPy and NumPy among others.
The Linux builds pass successfully in these setups:
cp37-manylinux_x86_64 wheel
cp38-manylinux_x86_64 wheel
cp39-manylinux_x86_64 wheel
cp310-manylinux_x86_64 wheel
but then it moves to musllinux and fails at:
cp37-musllinux_x86_64 wheel
The reported error is that BLAS is not available when attempting to build Scipy and Numpy. Apparently pip doesn't locate pre-built wheels on PyPi. As far as I understand, this is expected since there are no musllinux wheels for Numpy for example, as discussed here. In the case of manylinux containers, there is no such problem, for example:
If I understand things right, the issue of musllinux having trouble with dependencies has already been raised here.
So I'm trying to understand if my analysis looks right, and if the straightforward solution is to disable musllinux at least until there will be numpy and spicy wheels for this image?
Thank you!
Beta Was this translation helpful? Give feedback.
All reactions