Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Reduce cuda complexity by dropping 11.8 #6917

Closed
wants to merge 1 commit into from
Closed

Reduce cuda complexity by dropping 11.8 #6917

Changes from all commits
Commits
Show all changes
1 commit
Select commit
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
28 changes: 3 additions & 25 deletions recipe/conda_build_config.yaml
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -6,7 +6,6 @@ c_compiler:
c_compiler_version: # [unix]
- 13 # [linux]
- 18 # [osx]
- 11 # [linux and os.environ.get("CF_CUDA_ENABLED", "False") == "True"]
# CUDA 12.4 on PPC does not support GCC 13
- 12 # [linux and ppc64le and os.environ.get("CF_CUDA_ENABLED", "False") == "True"]
- 13 # [linux and not ppc64le and os.environ.get("CF_CUDA_ENABLED", "False") == "True"]
Expand All @@ -28,7 +27,6 @@ cxx_compiler:
cxx_compiler_version: # [unix]
- 13 # [linux]
- 18 # [osx]
- 11 # [linux and os.environ.get("CF_CUDA_ENABLED", "False") == "True"]
- 12 # [linux and ppc64le and os.environ.get("CF_CUDA_ENABLED", "False") == "True"]
- 13 # [linux and not ppc64le and os.environ.get("CF_CUDA_ENABLED", "False") == "True"]
llvm_openmp: # [osx]
Expand All @@ -41,7 +39,6 @@ fortran_compiler_version: # [unix or win64]
- 13 # [linux]
- 13 # [osx]
- 5 # [win64]
- 11 # [linux and os.environ.get("CF_CUDA_ENABLED", "False") == "True"]
- 12 # [linux and ppc64le and os.environ.get("CF_CUDA_ENABLED", "False") == "True"]
- 13 # [linux and not ppc64le and os.environ.get("CF_CUDA_ENABLED", "False") == "True"]
m2w64_c_compiler: # [win]
Expand All @@ -51,18 +48,13 @@ m2w64_cxx_compiler: # [win]
m2w64_fortran_compiler: # [win]
- m2w64-toolchain # [win]

cuda_compiler:
- None
- nvcc # [(linux or win64) and os.environ.get("CF_CUDA_ENABLED", "False") == "True"]
- cuda-nvcc # [(linux or win64) and os.environ.get("CF_CUDA_ENABLED", "False") == "True"]
Copy link
Contributor Author

Choose a reason for hiding this comment

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

do we really need to use this key? could we have users simply depend on cuda-nvcc? I likely don't understand the downstream implications.

Copy link
Member

Choose a reason for hiding this comment

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

Users shouldn't depend on cuda-nvcc but continue to depend on {{ compiler("cuda") }}. For that, we need to set cuda_compiler in the pinning, also in a 12+ world

cuda_compiler_version:
- None
- 11.8 # [(linux or win64) and os.environ.get("CF_CUDA_ENABLED", "False") == "True"]
- 12.4 # [linux and ppc64le and os.environ.get("CF_CUDA_ENABLED", "False") == "True"]
- 12.6 # [((linux and not ppc64le) or win64) and os.environ.get("CF_CUDA_ENABLED", "False") == "True"]
- 12 # [(linux or win64) and os.environ.get("CF_CUDA_ENABLED", "False") == "True"]
Comment on lines 51 to +53
Copy link
Member

Choose a reason for hiding this comment

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

I don't want to switch to free-floating minor for 12.x in an unrelated PR. If you think that's a good idea, please open an issue and we can discuss it there.

Copy link
Member

@jakirkham jakirkham Jan 14, 2025 •
edited
Loading

Choose a reason for hiding this comment

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

Right we also discussed this previously. The reason we pin this here is to avoid picking up incomplete CUDA Toolkit's during upgrades

We do plan to bump this version whenever the next CUDA Toolkit is fully available on conda-forge


cuda_compiler_version_min:
- None # [osx]
- 11.8 # [linux or win64]
- 12 # [linux or win64]

arm_variant_type: # [aarch64]
- sbsa # [aarch64]
Expand Down Expand Up @@ -147,16 +139,6 @@ docker_image: # [os.environ.get("BUILD_PLA
- quay.io/condaforge/linux-anvil-aarch64:alma9 # [os.environ.get("BUILD_PLATFORM") == "linux-aarch64" and os.environ.get("DEFAULT_LINUX_VERSION", "alma9") == "alma9"]
- quay.io/condaforge/linux-anvil-ppc64le:alma9 # [os.environ.get("BUILD_PLATFORM") == "linux-ppc64le" and os.environ.get("DEFAULT_LINUX_VERSION", "alma9") == "alma9"]

# CUDA 11.8 builds (only x64 has a DEFAULT_LINUX_VERSION choice; alma9 not available)
- quay.io/condaforge/linux-anvil-x86_64-cuda11.8:cos7 # [linux64 and os.environ.get("CF_CUDA_ENABLED", "False") == "True" and os.environ.get("BUILD_PLATFORM") == "linux-64" and os.environ.get("DEFAULT_LINUX_VERSION", "ubi8") == "cos7"]
- quay.io/condaforge/linux-anvil-x86_64-cuda11.8:ubi8 # [linux64 and os.environ.get("CF_CUDA_ENABLED", "False") == "True" and os.environ.get("BUILD_PLATFORM") == "linux-64" and os.environ.get("DEFAULT_LINUX_VERSION", "ubi8") in ("ubi8", "alma8", "alma9")]
# CUDA 11.8 arch: native compilation (build == target)
- quay.io/condaforge/linux-anvil-aarch64-cuda11.8:ubi8 # [aarch64 and os.environ.get("CF_CUDA_ENABLED", "False") == "True" and os.environ.get("BUILD_PLATFORM") == "linux-aarch64"]
- quay.io/condaforge/linux-anvil-ppc64le-cuda11.8:ubi8 # [ppc64le and os.environ.get("CF_CUDA_ENABLED", "False") == "True" and os.environ.get("BUILD_PLATFORM") == "linux-ppc64le"]
# CUDA 11.8 arch: cross-compilation (build != target)
- quay.io/condaforge/linux-anvil-x86_64-cuda11.8:ubi8 # [aarch64 and os.environ.get("CF_CUDA_ENABLED", "False") == "True" and os.environ.get("BUILD_PLATFORM") == "linux-64"]
- quay.io/condaforge/linux-anvil-x86_64-cuda11.8:ubi8 # [ppc64le and os.environ.get("CF_CUDA_ENABLED", "False") == "True" and os.environ.get("BUILD_PLATFORM") == "linux-64"]

# CUDA 12 builds on CentOS 7
- quay.io/condaforge/linux-anvil-x86_64:cos7 # [os.environ.get("CF_CUDA_ENABLED", "False") == "True" and os.environ.get("BUILD_PLATFORM") == "linux-64" and os.environ.get("DEFAULT_LINUX_VERSION", "alma9") == "cos7"]
- quay.io/condaforge/linux-anvil-aarch64:cos7 # [os.environ.get("CF_CUDA_ENABLED", "False") == "True" and os.environ.get("BUILD_PLATFORM") == "linux-aarch64" and os.environ.get("DEFAULT_LINUX_VERSION", "alma9") == "cos7"]
Expand All @@ -177,12 +159,8 @@ zip_keys:
- c_compiler_version # [unix]
- cxx_compiler_version # [unix]
- fortran_compiler_version # [unix]
- cuda_compiler # [linux and os.environ.get("CF_CUDA_ENABLED", "False") == "True"]
- cuda_compiler_version # [linux and os.environ.get("CF_CUDA_ENABLED", "False") == "True"]
- docker_image # [linux and os.environ.get("CF_CUDA_ENABLED", "False") == "True" and os.environ.get("BUILD_PLATFORM", "").startswith("linux-")]
- # [win64 and os.environ.get("CF_CUDA_ENABLED", "False") == "True"]
- cuda_compiler # [win64 and os.environ.get("CF_CUDA_ENABLED", "False") == "True"]
- cuda_compiler_version # [win64 and os.environ.get("CF_CUDA_ENABLED", "False") == "True"]
Comment on lines -180 to -185
Copy link
Member

Choose a reason for hiding this comment

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

This will thus continue to be necessary

-
- python
- numpy
Expand Down
Loading