feat(pathfinder): centralize CUDA env var handling, prioritize CUDA_PATH over CUDA_HOME#1801
feat(pathfinder): centralize CUDA env var handling, prioritize CUDA_PATH over CUDA_HOME#1801rwgk wants to merge 8 commits intoNVIDIA:mainfrom
CUDA_PATH over CUDA_HOME#1801Conversation
…d onto main. Adds cuda.pathfinder._utils.env_var_constants with canonical search order, enhances get_cuda_home_or_path() with robust path comparison and caching, and updates documentation across all packages to reflect the new priority. Co-authored-by: Rob Parolin <rparolin@nvidia.com> Made-with: Cursor
|
Auto-sync is disabled for draft pull requests in this repository. Workflows must be run manually. Contributors can view more details about this message here. |
|
Decision: drop Both
Both Related commit: 2164c33 |
Drop os.pathsep splitting of CUDA_PATH/CUDA_HOME in both build_hooks.py files. Both functions now delegate to get_cuda_home_or_path() from cuda.pathfinder, returning a single path. See NVIDIA#1801 (comment) Made-with: Cursor
|
Decision: treat empty Rob's original implementation preserved empty strings (
This addresses my review comment on Rob's PR: #1519 (comment) |
See NVIDIA#1801 (comment) for the rationale Made-with: Cursor
Made-with: Cursor
…lti-path test Made-with: Cursor
Made-with: Cursor
Made-with: Cursor
Safe: currently an internal-only API (not yet public). Made-with: Cursor
Closes #1433
WIP-WIP-WIP
Current state: Squash-merge of PR #1519 rebased onto main. Untested (not even locally).