-
Notifications
You must be signed in to change notification settings - Fork 76
Set compiler versions in context #755
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.
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
base: branch-25.08
Are you sure you want to change the base?
Conversation
Simplify the logic necessary to handle compiler versions by folding it into the context.
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Gave a rough overview of the relevant changes below
Generally the idea is push more things into context
where it can be written compactly. Keeping only the relevant differentiator (CUDA version) in the variant file
cat > variants.yaml << EOF | ||
cuda_version: | ||
- ${RAPIDS_CUDA_VERSION} |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
This writes out the common variant file content for all cases
if [[ "$(arch)" == "aarch64" ]]; then | ||
cat >> variants.yaml << EOF | ||
- 12.1 # The last version to not support cufile |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Then the legacy CUDA ARM builds are appended to the file
This also makes it easier to drop this section of code when it no longer applies
should_use_cufile: ${{ x86_64 or (aarch64 and cuda_version >= "12.2") }} | ||
c_compiler_version: ${{ 13 if should_use_cufile else 12 }} | ||
cxx_compiler_version: ${{ c_compiler_version }} |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Instead of setting different compiler versions in the CUDA variant file, we just set them here in the context. Also this ties them to the relevant thing. Namely whether we can build with cuFile or not
We can see this works as intended with the CUDA 12.1 ARM build
│ │ │ cuda-version ┆ 12.1 ┆ h1d6eff3_3 ┆ conda-forge ┆ 20.56 KiB │
│ │ │ gcc_impl_linux-aarch64 ┆ 12.4.0 ┆ h628656a_2 ┆ conda-forge ┆ 56.19 MiB │
│ │ │ gcc_linux-aarch64 ┆ 12.4.0 ┆ heb3b579_10 ┆ conda-forge ┆ 31.88 KiB │
│ │ │ gxx_impl_linux-aarch64 ┆ 12.4.0 ┆ h0bf7a72_2 ┆ conda-forge ┆ 11.36 MiB │
│ │ │ gxx_linux-aarch64 ┆ 12.4.0 ┆ h3f57e68_10 ┆ conda-forge ┆ 30.23 KiB │
Simplify the logic necessary to handle compiler versions by folding it into the context.