Skip to content

[WIP] Add CUDA 11.8 manual migrator#7432

Merged
hmaarrfk merged 2 commits into
conda-forge:mainfrom
jakirkham-feedstocks:add_cuda118_manual_migrator
Jun 13, 2025
Merged

[WIP] Add CUDA 11.8 manual migrator#7432
hmaarrfk merged 2 commits into
conda-forge:mainfrom
jakirkham-feedstocks:add_cuda118_manual_migrator

Conversation

@jakirkham

@jakirkham jakirkham commented May 29, 2025

Copy link
Copy Markdown
Member

We are planning to drop CUDA 11.8 from the global pinnings as discussed in issue: #7404

Following the pattern that we have used in the past to readd old CUDA version like with PR: #1657

Provide a manual migrator that recipe maintainers can use to readd CUDA 11.8 in their feedstocks.

@conda-forge-admin

Copy link
Copy Markdown
Contributor

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.

Comment on lines +17 to +28
ordering:
cuda_compiler:
- None
- cuda-nvcc
- nvcc
cuda_compiler_version:
- None
- 12.4
- 12.6
- 12.8
- 12.9
- 11.8

Copy link
Copy Markdown
Member Author

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

It looks like we need to add CUDA 11.8 last in the ordering for this to work correctly

With that in place this does appear to do the right thing

@jakirkham jakirkham force-pushed the add_cuda118_manual_migrator branch 5 times, most recently from dbc4c7a to 06d37f7 Compare May 29, 2025 06:39
jakirkham added 2 commits May 28, 2025 23:40
Provide a manual migrator that recipe maintainers can use to readd CUDA
11.8 after we drop it from the global pinnings.
@jakirkham jakirkham force-pushed the add_cuda118_manual_migrator branch from 06d37f7 to 7372cd4 Compare May 29, 2025 06:44
Comment thread .yamllint
document-start: disable
line-length:
max: 180
max: 260

Copy link
Copy Markdown
Member Author

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

The Docker image lines with selectors get kind of long. Already squeezed them down quite a bit

Needed to bump this to keep the linter happy

@h-vetinari

Copy link
Copy Markdown
Member

Thanks for doing this. I agree it's nice to provide a migrator for people to copy (even though I suspect it won't see much use 🤷). Did you test it with the state of #74311? Due to the zipping change (removing docker_image), there's potential for things to behave differently once it's been merged.

Footnotes

  1. e.g. conda smithy rerender -e <local_pinning>

@jakirkham

Copy link
Copy Markdown
Member Author

Yeah exactly was testing with your PR removing CUDA 11.8 (and the zip_keys)

Think this is why we needed the ordering flipped for adding CUDA 11.8 above

Previously CUDA 11.8 would have been earlier in the list, but that causes problems without zip_keys. Moving it to the end seemed to resolve these issues

@h-vetinari

Copy link
Copy Markdown
Member

Could you test it one last time now that cuda_compiler has also changed in #7431?

@hmaarrfk hmaarrfk merged commit 9e4ce93 into conda-forge:main Jun 13, 2025
3 checks passed
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

4 participants