hdf5 v1.14.1

[regro-cf-autotick-bot]

It is very likely that the current package version for this feedstock is out of date.

Checklist before merging this PR:

• Dependencies have been updated if changed: see upstream
• Tests have passed
• Updated license if changed and license_file is packaged

Information about this PR:

1. Feel free to push to the bot's branch to update this PR if needed.
2. The bot will almost always only open one PR per version.
3. The bot will stop issuing PRs if more than 3 version bump PRs generated by the bot are open. If you don't want to package a particular version please close the PR.
4. If you want these PRs to be merged automatically, make an issue with @conda-forge-admin,please add bot automerge in the title and merge the resulting PR. This command will add our bot automerge feature to your feedstock.
5. 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.

Dependency Analysis

We couldn't run dependency analysis due to an internal error in the bot. :/ Help is very welcome!

_{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/4928588102, 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.

[hmaarrfk]

@mkitti are you still needing the openssl 1.1.1 builds? If so, there seems to be some missing dependencies.

[mkitti]

@mkitti are you still needing the openssl 1.1.1 builds? If so, there seems to be some missing dependencies.

No. We do not need openssl 1.1.1 builds. Thanks.

[mkitti]

We do not need 2407-include-shlwapi-explicitely-on-windows.patch. It was applied upstream to the 1.14 branch in HDFGroup/hdf5#2790 from the original pull request in HDFGroup/hdf5#2407

[hmaarrfk]

Thakn you. lets see what happens.

Any clue for the unix mpich failures?

[mkitti]

On macOS a time.h header is missing. I'm unsure how. On Windows, try dropping the other patch as well. Did they some work making libaec the default SZIP implementation. We might need to rework the patch or tell it to use an external szip through some option.

[Tobias-Fischer]

For clock_gettime we need macos sdk 10.12 (see e.g. zeromq/libzmq#2175).

I'm not sure what the linux test failures are about ..

[hmaarrfk]

Of course!

[minrk]

I managed to run the mac arm cross-compiles locally for both mpich and nompi, and tested that the binaries work (the ph5diff tests notwithstanding).

Assuming builds pass, I think the only remaining choice ought to be between:

1. release with known broken ph5diff on mpich (no change)
2. build-only pin mpich 4.0, which will give most folks broken ph5diff by default with mpich 4.1, but gives folks the option of pinning mpich 4.0 at runtime if they need ph5diff (many other packages will conflict mpich 4.0 anyway if they were built with 4.1).
3. build-and-run pin mpich 4.0 - ensures ph5diff works, but also ensures incompatibility with other packages recently built against mpich, which require mpich >=4.1.

[minrk]

ha, mac arm builds succeeded. linux mpi builds started failing mysteriously, which appears to have been caused by gfortran 13, and fixed upstream by HDFGroup/hdf5#2957 so I pulled in that patch. We'll see if that makes everything green.

[hmaarrfk]

Do we think the ABI has changed from 1.14.0? Do we need a full migration?

[mkitti]

I asked about the ABI here:
HDFGroup/hdf5#3132 (comment)

[hmaarrfk]

Great. Seems compatible!

[minrk]

huzzah, all green!

Any preference for the handling of mpich pinning or not, and if so build-only or runtime? As it is right now, I think it will be the most compatible with other conda-forge packages, but ph5diff will always exit with a finalization error when run with mpich (the diff is still correct). This would probably be my inclination, and would mean this is ready to merge.

[minrk]

@hmaarrfk anything more you want to do here?

[hmaarrfk]

i'm somewhat worried about the transition we will need to unpin "future" builds that converge to using HDF5 1.14.0 accidentally instead of 1.14.1 to continue to be compatible with both.

[minrk]

Yeah, I'm not sure if we can update the hdf5 migrator to 1.14.1? Or if we need to start a new migration?

Should the migrator have pinned 1.14 instead of 1.14.1?

[hmaarrfk]

We should repodata patch, and force the pin up.

[minrk]

it's only a build-time pin, so I don't think there should be a need to patch the packages that worked, right? Unless some packages have an x.y.z pin for hdf5?

[hmaarrfk]

Hdf5 has a run export on x.x.x

[minrk]

Ah, gotcha. Then we should just stop the 1.14.0 migration and start a new one for 1.14.1?

[hmaarrfk]

Or just edit it to do 1.14.1

[minrk]

But due to the x.x.x pin, it needs to start over and do all the packages that have already been migrated, right?

[jakirkham]

Migrators can be restarted

[minrk]

Since it needs to be re-run with a different version, isn't that just the same as a new migrator?