openblas v0.3.22

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It is very likely that the current package version for this feedstock is out of date.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

Pending Dependency Version Updates

Here is a list of all the pending dependency version updates for this repo. Please double check all dependencies before merging.

Name	Upstream Version	Current Version
openblas	0.3.22

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/4527083694, 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.

[h-vetinari]

@martin-frbg, all unix builds fail with:

make[1]: *** No rule to make target 'cspmv.o', needed by 'libs'.  Stop.

Where should this object file be coming from? 🤔

[martin-frbg]

interface/zspmv.c - maybe your download of the source tarball was truncated ?

[isuruf]

@martin-frbg, it seems like there are two different code patterns

ifneq ($(NO_LAPACK), 1)

and

ifndef NO_LAPACK

We have NO_LAPACK=0 in the build script here which works fine with the first pattern, but is not for the second pattern.

[martin-frbg]

Eh, thought the unqualified ifdefs had died out already, and no recent change in this function (though the interval for this release was definitely too long)

[martin-frbg]

Ok, I see I added them myself two months ago :(

[h-vetinari]

Thank you both! Seems the patch I came up with here is working, so I opened OpenMathLib/OpenBLAS#3972.

[h-vetinari]

Hm, seems osx-arm64 cross-compilation is unhappy as well...

Copying the static library to $PREFIX/lib
install: libopenblas_sandybridgep-r0.3.22.a: No such file or directory
make[1]: *** [Makefile.install:66: install] Error 71

[martin-frbg]

You'll need my 3975 on top of your patch, but it is now looking as if there is something else broken in 0.3.22 - maybe "just" corner cases of ZGETF2 plus a few changed accuracies from LAPACK 3.11.0, maybe more. I'll probably release 0.3.23 as soon as the dust settles.

[martin-frbg]

now you'd need PR 3980 for the regression in GETF2 as well, may make sense to wait for 0.2.23 unless you want to support checking 0.3.22 for further defects

[h-vetinari]

now you'd need PR 3980 for the regression in GETF2 as well, may make sense to wait for 0.2.23 unless you want to support checking 0.3.22 for further defects

I don't mind backporting / testing further until we're sure that all things are sorted out for 0.3.23 (PR 3980 needed 3978 as well).

[h-vetinari]

osx-arm still unhappy...

Copying the static library to $PREFIX/lib
install: libopenblas64__sandybridgep-r0.3.22.a: No such file or directory
make[1]: *** [Makefile.install:71: install] Error 71

[martin-frbg]

how come osx-arm is named sandybridge ? Is the install step not repeating the build arguments (and thus going with an autodetect of the build host) ?

[h-vetinari]

how come osx-arm is named sandybridge ? Is the install step not repeating the build arguments (and thus going with an autodetect of the build host) ?

We didn't change the build instructions since 0.3.21 (where this didn't happen), so either we were implicitly relying on something undocumented, or something broke.

[martin-frbg]

Well, the build is undoubtedly for an ARM64 system (using kernels from the arm64 subtree and culminating in the success message mentioning libopenblas64_vortexp as the name of the library. But the subsequent make install you are attempting appears to lack all parameters, so no concept of cross-compile or -target exists and getarch is run again to identify the build host. This step has never been automated, and in fact a warning is printed at the end of compilation that you need to repeat the exact same set of compile arguments in the install step.

[martin-frbg]

released 0.3.23 now with the fixes (plus a few Raptor Lake cpuids)

[h-vetinari]

Cross-compilation still unhappy in #151; I don't have time to look at it this week.

[h-vetinari]

Superseded by #151