Fix how NCCL existence is checked #135
Merged
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
The bug is discovered by @mrava87 when the system is installed with CuPy but without NCCL.
Scenario:
CuPy always comes with
cupy.cuda.ncclso checking the import error ofcupy.cuda.ncclis not sufficient to verify the existence of NCCL. Under the original code, with CuPy installed without NCCL, thedeps.nccl_enabledwill beTrue(as no import error occurs). This will lead to crashimport pylops_mpibecause it can't find nccl.NCCL_FLOAT32Fix:
If CuPy is present, this import statement
import cupy.cuda.nccl as ncclwill not throw an error and we can check the NCCL's existence through thisnccl.availableflag