cutensor v2.1.0.9 #65

regro-cf-autotick-bot · 2025-01-28T22:54:41Z

It is very likely that the current package version for this feedstock is out of date.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

Feel free to push to the bot's branch to update this PR if needed.
The bot will almost always only open one PR per version.
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.
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.
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
cudatoolkit	12.0.0
git	2.48.1

_{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/13020924503 - please use this URL for debugging.}

…nda-forge-pinning 2025.01.28.22.50.30

conda-forge-admin · 2025-01-28T22:56:13Z

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

I wanted to let you know that I linted all conda-recipes in your PR (recipe/meta.yaml) and found some lint.

Here's what I've got...

For recipe/meta.yaml:

❌ You are setting c_stdlib_version below the current global baseline in conda-forge (10.13). If this is your intention, you also need to override MACOSX_DEPLOYMENT_TARGET (with the same value) locally.

For recipe/meta.yaml:

ℹ️ The recipe is not parsable by parser conda-souschef (grayskull). This parser is not currently used by conda-forge, but may be in the future. We are collecting information to see which recipes are compatible with grayskull.

_{This message was generated by GitHub Actions workflow run https://github.com/conda-forge/conda-forge-webservices/actions/runs/13056533239. Examine the logs at this URL for more detail.}

carterbox · 2025-01-30T00:27:28Z

The problem is the bot didn't update the hashes correctly, so couldn't download any of the tarballs.

carterbox · 2025-01-30T16:15:10Z

@conda-forge-admin, please rerender

…nda-forge-pinning 2025.01.28.22.50.30

jakirkham · 2025-01-30T19:17:55Z

Thanks Daniel! 🙏

Looks like we need GLIBC 2.28

Basing this on this CI failure:

binary glibc 2.27.0 <= recipe glibc 2.17 <= system glibc 2.34 $PREFIX/lib/libcutensorMg.so.2.1.0

carterbox · 2025-01-30T20:57:34Z

The Windows build fails because the runner doesn't have enough space to unpack the zip. Probably this is related to the fact that all builds of cutensor are shipped in the same archive instead of separately for CTK 11 and CTK 12.

carterbox · 2025-01-30T20:59:13Z

@conda-forge-admin, please rerender

conda-forge-admin · 2025-01-30T21:00:18Z

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/meta.yaml) and found it was in an excellent condition.

I do have some suggestions for making it better though...

For recipe/meta.yaml:

ℹ️ The recipe is not parsable by parser conda-souschef (grayskull). This parser is not currently used by conda-forge, but may be in the future. We are collecting information to see which recipes are compatible with grayskull.

_{This message was generated by GitHub Actions workflow run https://github.com/conda-forge/conda-forge-webservices/actions/runs/13080657240. Examine the logs at this URL for more detail.}

…nda-forge-pinning 2025.01.28.22.50.30

jakirkham · 2025-01-31T00:51:53Z

recipe/bld.bat

+:: Hack to free up additional space on Azure
+:: Replace with conda-smithy solution when available
+:: xref: https://github.com/conda-forge/conda-smithy/pull/1966
+if "%CI%" == "azure" (
+    set CLEANUP_DIRS=^
+    C:\hostedtoolcache\windows;^
+    ;
+
+    mkdir C:\empty
+    for %%f in (%CLEANUP_DIRS:;= %) do (
+        if not [%%f] == [] (
+        echo Removing %%f
+        dir %%f
+        robocopy /mir /mt /zb /ns /nc /nfl /ndl /np /njh /njs C:\empty %%f > nul 2>&1
+        rmdir /q %%f
+        )
+    )
+    rmdir /q C:\empty
+)


Have found C:\hostedtoolcache\windows takes up a fair bit of space on Azure. Removing it would free up that space

Had put this solution together in this conda-smithy PR: conda-forge/conda-smithy#1966

Though have not found a good method for deleting this directory quickly on Windows. So the deletion takes a while. Maybe acceptable if there is no other option (like we see here)

That said, if we could find a faster way to do this deletion, it would be easier to integrate this approach in conda-smithy and just enable with a config parameter

Though have not found a good method for deleting this directory quickly on Windows.

Which options have you tried?

Did the bulk of the exploration in these PRs:

[NO MRG] Test Azure Windows image cleanup miktex-feedstock#31

Build Matplotlib 3.9.0rc1 matplotlib-feedstock#373

With matplotlib, it was picking up some VM Python during its build and getting hosed. We were not able to get the build to stop picking up the VM Python and couldn't figure out why it was showing up. So we went with the deletion option to get the build fixed up. Lots of commits in there trying things

Though running out of space is an often enough problem (like with this update), that I found the simplest out-of-the-way Windows only build and used that for testing, which is why MiKTeX was used. It was also simpler than going through a full matplotlib build each iteration

Tried different deletion methods. Found robocopy was better than the others. So then tweaked options to eek out more performance

If you spot something I've missed, would be curious to learn about it

recipe/meta.yaml

Avoids a bug in some recent `patchelf` versions. Ideally would relax this in the future once that issue is fixed.

As `patchelf` is a build tool, it can be in `requirements/build`.

recipe/meta.yaml

jakirkham · 2025-01-31T01:04:38Z

recipe/meta.yaml

@@ -50,8 +41,10 @@ requirements:
    - {{ compiler('cxx') }}
    - {{ compiler('cuda') }}
    - arm-variant * {{ arm_variant_type }}  # [aarch64]
+    - cf-nvidia-tools                       # [linux]


Added cf-nvidia-tools and switched to it for our GLIBC check

Think I got the related bits replaced. Though please let me know if I've missed something

Suggested change

- cf-nvidia-tools # [linux]

- cf-nvidia-tools 1 # [linux]

Also, maybe we want to be conservative and pin to the major version to protect against breaking changes?

Are there particular breaking changes you have in mind?

Not currently, but I've promised to use actual semantic versioning for this package. I suspect there may be breaks once we start adding more features to cf-nvidia-tools and the complexity increases.

Ok let's make sure this goes in the migrator and other used places

Bots already will sort the keys in `conda-forge.yml` when changing them. So go ahead and sort them proactively.

jakirkham · 2025-01-31T01:50:28Z

@conda-forge-admin , please re-render

conda-forge-admin · 2025-01-31T01:52:45Z

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/conda-forge-webservices/actions/runs/13064835195. Examine the logs at this URL for more detail.}

This cleanup needs to happen before the `redist` is downloaded. Otherwise we run out of space when downloading or uncompressing it. So use a custom CI setup script to do this cleanup during CI setup before the recipe build has started.

jakirkham · 2025-01-31T03:28:10Z

@conda-forge-admin , please re-render

…nda-forge-pinning 2025.01.28.22.50.30

Something funky is happening with this variable where nothing is actually iterated over and removed in the `for`-loop. Try setting this outside the `for`-loop in case the indenting is somehow causing syntax issues when parsing the list.

This drive has more space. So hopefully switching to it avoids the space issues we have seen.

jakirkham · 2025-01-31T04:48:33Z

@conda-forge-admin , please re-render

…nda-forge-pinning 2025.01.28.22.50.30

jakirkham · 2025-01-31T06:50:50Z

@conda-forge-admin , please re-render

…nda-forge-pinning 2025.01.28.22.50.30

carterbox

Once patchelf is removed from the deps, LGTM.

Co-authored-by: Daniel Ching <[email protected]>

regro-cf-autotick-bot added 2 commits January 28, 2025 22:53

updated v2.1.0.9

ab441f0

MNT: Re-rendered with conda-build 25.1.1, conda-smithy 3.45.4, and co…

612b2ea

…nda-forge-pinning 2025.01.28.22.50.30

regro-cf-autotick-bot requested review from JeremyWangNVDA, jakirkham, leofang, mtjrider, v0i0, yangcal and a team as code owners January 28, 2025 22:54

Update hashes

af78451

MNT: Re-rendered with conda-build 25.1.1, conda-smithy 3.45.4, and co…

f8827ce

…nda-forge-pinning 2025.01.28.22.50.30

Use glibc 2.28 for all linux archs

df55175

conda-forge-webservices[bot] and others added 3 commits January 30, 2025 21:01

MNT: Re-rendered with conda-build 25.1.1, conda-smithy 3.45.4, and co…

8302649

…nda-forge-pinning 2025.01.28.22.50.30

On Windows, switch to bld.bat

3bcc3df

Add workaround to free up space on Azure image

e932e16

jakirkham reviewed Jan 31, 2025

View reviewed changes

recipe/meta.yaml Outdated Show resolved Hide resolved

jakirkham added 3 commits January 30, 2025 16:56

Constrain patchelf to <0.18.0

6f8a131

Avoids a bug in some recent `patchelf` versions. Ideally would relax this in the future once that issue is fixed.

Move patchelf to requirements/build

fe95b3e

As `patchelf` is a build tool, it can be in `requirements/build`.

Switch to cf-nvidia-tools for GLIBC check

a85852b

jakirkham reviewed Jan 31, 2025

View reviewed changes

recipe/meta.yaml Outdated Show resolved Hide resolved

jakirkham reviewed Jan 31, 2025

View reviewed changes

Set & use environment variable for cuda_major

398e193

jakirkham added 2 commits January 30, 2025 17:46

Sort conda-forge.yml's keys

86c88d4

Bots already will sort the keys in `conda-forge.yml` when changing them. So go ahead and sort them proactively.

Simplify skip

94641d3

Move Azure Windows cleanup into CI setup step

e67026a

This cleanup needs to happen before the `redist` is downloaded. Otherwise we run out of space when downloading or uncompressing it. So use a custom CI setup script to do this cleanup during CI setup before the recipe build has started.

conda-forge-webservices[bot] and others added 5 commits January 31, 2025 03:30

MNT: Re-rendered with conda-build 25.1.1, conda-smithy 3.45.4, and co…

aa8ae12

…nda-forge-pinning 2025.01.28.22.50.30

Set CLEANUP_DIRS outside if

1ec96e7

Something funky is happening with this variable where nothing is actually iterated over and removed in the `for`-loop. Try setting this outside the `for`-loop in case the indenting is somehow causing syntax issues when parsing the list.

Add errorlevel checks in bld.bat

09ed4af

Fix indentation

b5b5ee0

Use C: on Azure Windows

aa21e4a

This drive has more space. So hopefully switching to it avoids the space issues we have seen.

conda-forge-webservices[bot] and others added 2 commits January 31, 2025 04:51

MNT: Re-rendered with conda-build 25.1.1, conda-smithy 3.45.4, and co…

320d20b

…nda-forge-pinning 2025.01.28.22.50.30

Try removing Azure CI image custom cleanup step

2229b3f

MNT: Re-rendered with conda-build 25.1.1, conda-smithy 3.45.4, and co…

9cbcf49

…nda-forge-pinning 2025.01.28.22.50.30

carterbox mentioned this pull request Jan 31, 2025

Should there be an upper bound on patchelf? conda-forge/cf-nvidia-tools-feedstock#2

Open

carterbox requested changes Jan 31, 2025

View reviewed changes

jakirkham and others added 2 commits January 31, 2025 12:56

Drop patchelf from requirements/host

73002fb

Co-authored-by: Daniel Ching <[email protected]>

Pin cf-nvidia-tools to 1

940c41d

Co-authored-by: Daniel Ching <[email protected]>

jakirkham requested a review from carterbox January 31, 2025 20:58

carterbox merged commit 5c5dd0c into conda-forge:main Jan 31, 2025
9 checks passed

regro-cf-autotick-bot deleted the 2.1.0.9_h98a7bf branch January 31, 2025 21:55

Uh oh!

cutensor v2.1.0.9 #65

cutensor v2.1.0.9 #65

Uh oh!

Conversation

regro-cf-autotick-bot commented Jan 28, 2025

Pending Dependency Version Updates

Uh oh!

conda-forge-admin commented Jan 28, 2025 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

carterbox commented Jan 30, 2025

Uh oh!

carterbox commented Jan 30, 2025

Uh oh!

jakirkham commented Jan 30, 2025

Uh oh!

carterbox commented Jan 30, 2025

Uh oh!

carterbox commented Jan 30, 2025

Uh oh!

conda-forge-admin commented Jan 30, 2025 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Uh oh!

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

carterbox Jan 31, 2025 • edited by jakirkham Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

jakirkham commented Jan 31, 2025

Uh oh!

conda-forge-admin commented Jan 31, 2025

Uh oh!

jakirkham commented Jan 31, 2025

Uh oh!

jakirkham commented Jan 31, 2025

Uh oh!

jakirkham commented Jan 31, 2025

Uh oh!

carterbox left a comment

Choose a reason for hiding this comment

Uh oh!

Uh oh!

Uh oh!

conda-forge-admin commented Jan 28, 2025 •

edited

Loading

conda-forge-admin commented Jan 30, 2025 •

edited

Loading

carterbox Jan 31, 2025 •

edited by jakirkham

Loading