On Fri, Jan 14, 2022, 08:53 Matti Picus <matti.pi...@gmail.com <mailto:matti.pi...@gmail.com>> wrote:


    ...
    If we want to move off travis for aarch64, I think the only option
    right
    now is to use emulation, conda-forge does this on Azure machines
    [0] and
    the build + tests takes ~40 minutes. I can only imagine how long it
    would take to do move SciPy to a QEMU.


    ...
    Matti


    [0]
    
https://dev.azure.com/conda-forge/feedstock-builds/_build/results?buildId=434709&view=logs&jobId=e9afaa34-1a0f-534e-78fc-fae528ccd915&j=e9afaa34-1a0f-534e-78fc-fae528ccd915&t=48395e60-bc5f-5e45-7455-d404f0c95f3a
    
<https://dev.azure.com/conda-forge/feedstock-builds/_build/results?buildId=434709&view=logs&jobId=e9afaa34-1a0f-534e-78fc-fae528ccd915&j=e9afaa34-1a0f-534e-78fc-fae528ccd915&t=48395e60-bc5f-5e45-7455-d404f0c95f3a>

On 14/1/22 12:08 pm, Kevin Sheppard wrote:
As a data point against relying on QEMU: statsmodels times out (5+ hours) when trying to build with QEMU rising cibuildwheel and GitHub actions.

Kevin


Interesting. It seems to work on the azure builds from conda-forge, I wonder what is different. Are the time out build logs still around?

Matti


[1] https://github.com/conda-forge/statsmodels-feedstock/runs/4196021348

_______________________________________________
NumPy-Discussion mailing list -- numpy-discussion@python.org
To unsubscribe send an email to numpy-discussion-le...@python.org
https://mail.python.org/mailman3/lists/numpy-discussion.python.org/
Member address: arch...@mail-archive.com

Reply via email to