Arch Migrator

[regro-cf-autotick-bot]

This feedstock is being rebuilt as part of the aarch64/ppc64le migration.

Feel free to merge the PR if CI is all green, but please don't close it
without reaching out the the ARM migrators first at @conda-forge/arm-arch.

If this PR was opened in error or needs to be updated please add the bot-rerun label to this PR. The bot will close this PR and schedule another one. If you do not have permissions to add this label, you can use the phrase @conda-forge-admin, please rerun bot in a PR comment to have the conda-forge-admin add it for you.

_{This PR was created by the regro-cf-autotick-bot. The regro-cf-autotick-bot is a service to automatically track the dependency graph, migrate packages, and propose package version updates for conda-forge. Feel free to drop us a line if there are any issues! This PR was generated by https://github.com/regro/cf-scripts/actions/runs/4546936792, please use this URL for debugging.}

[conda-forge-webservices]

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.

[stuarteberg]

This PR will need some changes to conda-forge.yml, similar to the ones we did in sparsehash and cairomm:

build_platform:
  osx_arm64: osx_64
  linux_ppc64le: linux_64
  linux_aarch64: linux_64

[stuarteberg]

@conda-forge-admin, please rerender

[stuarteberg]

The failures are due to the fact that cairomm-1.0 has no packages for aarch64. Only the newer cairomm-1.16 is available. This will be fixed via conda-forge/cairomm-feedstock#32.

Alternatively, the graph-tool configure script could be patched to accept cairomm-1.16, rather than cairomm-1.0 exclusively. Our recipe links to some discussion of the issue.

[count0]

@stuarteberg It already does! The current graph-tool version will check first for cairomm-1.16 and fall back to cairomm-1.0 if that is not available.

[count0]

The current failure is more annoying... The configure check fails for a valid test involving boost.

[stuarteberg]

@conda-forge-admin please rerender

[github-actions]

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/graph-tool-feedstock/actions/runs/4554785848.

[stuarteberg]

Looks like the problem with boost::graph is related to a deprecation-cum-removal of std::unary_function, which is used by boost::container:

• std::unary_function deprecation warning with GCC 12 boostorg/container_hash#22
• unary_function has been removed from libc++ boostorg/container_hash#24
• The std lib unary/binary_function base classes are deprecated/removed… boostorg/config#440

In that thread, it's claimed that std::unary_function was supposed to be removed entirely in std=c++17, but I guess neither libstdc++ nor libc++ obeyed that directive... until now.

It's a bit strange that this feedstock built successfully last week, but now it's failing. My guess is that it's related to the fact that last week's build used libcxx-15.0.7, whereas yesterday's build uses libcxx-16.0.0.

---

Edit: Someone else reported this problem yesterday:

• removal of deprecated std::unary_function in libcxx 16 libcxx-feedstock#114
• Incompatibility between boost 1.78.0 and libcxx 16 boost-cpp-feedstock#134

In the latter thread, one proposed solution is to constrain the libcxx version. That would prevent any graph-tool users from using the brand new libcxx-16.0.0 in their conda environments until conda-forge eventually migrates to a newer boost. But that doesn't sound like a huge problem.

Ideally, the libcxx <16 constraint should occur in the boost-cpp feedstock, I think. (And the older, underconstrained boost packages should be deleted from anaconda.org.) But we can add the constraint here for immediate relief.

Another suggestion is to set _LIBCPP_ENABLE_CXX17_REMOVED_UNARY_BINARY, but one user who tried that claims it did not fix the problem.

[stuarteberg]

I don't think that change will require rerendering, but just in case...

@conda-forge-admin, please rerender

[github-actions]

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/graph-tool-feedstock/actions/runs/4566019972.

[stuarteberg]

Assuming that fixes the issue, then the only remaining failures will be specifically for the PPC architecture. There's some sort of linking error. It might be worth inspecting to see if there's an obvious solution, but if not, I'd propose simply skipping that architecture for now, and merging this with only aarch64 support.

[count0]

I believe we need to add -fPIC to CXXFLAGS in that architecture. Let's see.

[count0]

I don't really understand this rerender stuff, but maybe we need this, or maybe we don't, we'll see.

@conda-forge-admin, please rerender

[github-actions]

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/graph-tool-feedstock/actions/runs/4566757203.

[stuarteberg]

PPC is still failing with the same error...

[count0]

Tried a few things but nothing worked. Let's just remove ppc for the moment.

[count0]

@conda-forge-admin, please rerender