try fixes and understanding the problem ...

ngam commented 1 year ago

Checklist

[ ] Used a personal fork of the feedstock to propose changes
[ ] Bumped the build number (if the version is unchanged)
[ ] Reset the build number to 0 (if the version changed)
[ ] Re-rendered with the latest conda-smithy (Use the phrase code>@<space/conda-forge-admin, please rerender in a comment in this PR for automated rerendering)
[ ] Ensured the license file is being packaged.

conda-forge-webservices[bot] commented 1 year ago

Hi! This is the friendly automated conda-forge-linting service.

I just wanted to let you know that I linted all conda-recipes in your PR (recipe) and found it was in an excellent condition.

ngam commented 1 year ago

@conda-forge-admin, please rerender.

mkitti commented 1 year ago

This passes?

ngam commented 1 year ago

Yes, but the logs don’t look good…

@mkitti and @MilesCranmer should review the logs more carefully

ngam commented 1 year ago

Cc @h-vetinari

mkitti commented 1 year ago

I see a lot of clobber warnings. What else should I focus on?

ngam commented 1 year ago

Things like:

Error:import PyCallfailed: https://dev.azure.com/conda-forge/feedstock-builds/_build/results?buildId=719527&view=logs&j=d0d954b5-f111-5dc4-4d76-03b6c9d0cf7e&t=841356e0-85bb-57d8-dbbc-852e683d1642&l=379

I believe you explained this to me as "no problem" previously, but I am not sure I remember exactly. I also think this problem is now resolved, no?

h-vetinari commented 1 year ago

I see a lot of clobber warnings.

The reason for this is most likely as described here, i.e. reinstallation of numpy.

MilesCranmer commented 1 year ago

Once this merges: https://github.com/JuliaPy/PyCall.jl/pull/1040 we can disable the re-installation with a build flag. Then hopefully the build warnings will disappear.

conda-forge-webservices[bot] commented 1 year ago

Hi! This is the friendly automated conda-forge-linting service.

I was trying to look for recipes to lint for you, but it appears we have a merge conflict. Please try to merge or rebase with the base branch to resolve this conflict.

Please ping the 'conda-forge/core' team (using the @ notation in a comment) if you believe this is a bug.

conda-forge-webservices[bot] commented 1 year ago

Hi! This is the friendly automated conda-forge-linting service.

I just wanted to let you know that I linted all conda-recipes in your PR (recipe) and found it was in an excellent condition.

MilesCranmer commented 1 year ago

@conda-forge-admin, please rerender

github-actions[bot] commented 1 year ago

Hi! This is the friendly automated conda-forge-webservice.

I tried to rerender for you, but it looks like there was nothing to do.

This message was generated by GitHub actions workflow run https://github.com/conda-forge/pysr-feedstock/actions/runs/5240960713.

MilesCranmer commented 1 year ago

Not sure if this is a clue, but with the libstdcxx-ng<13 setting in run_constraints:, the mamba build is now working, but the conda build fails:

https://github.com/MilesCranmer/PySR/actions/runs/5242217963

MilesCranmer commented 1 year ago

It seems like the conda builds are finally working: https://github.com/MilesCranmer/PySR/actions/runs/5245593703

The band-aid fix is a combination of the following:

    - numpy
    - numpy >=1.14.0
    - ...
    - libstdcxx-ng <13  # [linux]
  run:
    - ...
    - {{ pin_compatible('numpy') }}
  run_constrained:
    - libstdcxx-ng <13

all of these changes were required. Even lowering the requested numpy constraint to >=1.14.0 was weirdly required to fix things (though I thought it didn't affect things).

It also required libstdcxx-ng <13 in the host: section as well, otherwise hitting some linking issues.

MilesCranmer commented 1 year ago

Once this is merged: https://github.com/JuliaRegistries/General/pull/85755

Then we can re-run these tests, and see if it solves the clobber problem

MilesCranmer commented 1 year ago

Cool, it looks like the clobber issue is gone? Are there any other warnings we need to worry about?

conda-forge-webservices[bot] commented 1 year ago

Hi! This is the friendly automated conda-forge-linting service.

I was trying to look for recipes to lint for you, but it appears we have a merge conflict. Please try to merge or rebase with the base branch to resolve this conflict.

Please ping the 'conda-forge/core' team (using the @ notation in a comment) if you believe this is a bug.

conda-forge-webservices[bot] commented 1 year ago

Hi! This is the friendly automated conda-forge-linting service.

I just wanted to let you know that I linted all conda-recipes in your PR (recipe) and found it was in an excellent condition.

github-actions[bot] commented 1 year ago

Hi! This is the friendly automated conda-forge-webservice.

I tried to rerender for you, but it looks like there was nothing to do.

This message was generated by GitHub actions workflow run https://github.com/conda-forge/pysr-feedstock/actions/runs/5326517222.

MilesCranmer commented 1 year ago

Fingers crossed.

h-vetinari commented 1 year ago

This looks better now, though there's what looks like a segfault in one of the linux jobs on main.

Also, this doesn't look great:

WARNING (pysr,share/pysr/depot/artifacts/9f6f87cc2114df918ff9c21dff9572f690e7dab2/lib/libLLVMExtra-14.so): runpaths ['$ORIGIN'] found in /home/conda/feedstock_root/build_artifacts/pysr_1687299957695/_h_env_placehold_placehold_placehold_placehold_placehold_placehold_placehold_placehold_placehold_placehold_placehold_placehold_placehold_placehold_placehold_placehold_placehold_placehold_placeh/share/pysr/depot/artifacts/9f6f87cc2114df918ff9c21dff9572f690e7dab2/lib/libLLVMExtra-14.so
WARNING (pysr,share/pysr/depot/artifacts/9f6f87cc2114df918ff9c21dff9572f690e7dab2/lib/libLLVMExtra-14.so): $RPATH/libLLVM-14jl.so not found in packages, sysroot(s) nor the missing_dso_whitelist.
.. is this binary repackaging?
WARNING (pysr,share/pysr/depot/artifacts/9f6f87cc2114df918ff9c21dff9572f690e7dab2/lib/libLLVMExtra-14.so): /lib64/libstdc++.so.6 not found in packages, sysroot(s) nor the missing_dso_whitelist.
.. is this binary repackaging?
WARNING (pysr,share/pysr/depot/artifacts/9f6f87cc2114df918ff9c21dff9572f690e7dab2/lib/libLLVMExtra-14.so): /lib64/libm.so.6 not found in packages, sysroot(s) nor the missing_dso_whitelist.
.. is this binary repackaging?
WARNING (pysr,share/pysr/depot/artifacts/9f6f87cc2114df918ff9c21dff9572f690e7dab2/lib/libLLVMExtra-14.so): /lib64/libgcc_s.so.1 not found in packages, sysroot(s) nor the missing_dso_whitelist.
.. is this binary repackaging?
WARNING (pysr,share/pysr/depot/artifacts/9f6f87cc2114df918ff9c21dff9572f690e7dab2/lib/libLLVMExtra-14.so): /lib64/libc.so.6 not found in packages, sysroot(s) nor the missing_dso_whitelist.
.. is this binary repackaging?

I don't know where libLLVMExtra-14.so comes from - I help package LLVM for conda-forge, but I've never seen that library, and I cannot find it upstream either.

ngam commented 1 year ago

Julia uses a special LLVM… like their heavily modified fork. They do try to upstream changes though, but it takes a while.

we once discussed trying to package the Julia LLVM fork…

MilesCranmer commented 1 year ago

Regarding the segfault, the full set of tests is fairly compute/memory intensive compared to the compute used in the build jobs. So maybe we could use a lighter test, since we just want a smoke test for the integration all working.

I’m also unsure why the OSX jobs freeze during the precompilation: https://dev.azure.com/conda-forge/feedstock-builds/_build/results?buildId=731490&view=logs&j=a8dee1cf-d17f-5cb0-b53f-4a93dfdbfb0c&t=d9dbf137-9a7f-5666-c60c-9e1b2fc31445 — this seems repeatable, and only occurs on Python 3.9+OSX…

ngam commented 1 year ago

If CI fails on the main branch, the package doesn’t get uploaded to anaconda dot org

conda-forge / pysr-feedstock

try fixes and understanding the problem ... #90