Enable extended lambdas support by default with CUDA #5580
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Outdated suggestions cannot be applied.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.
Rational:
--expt-extended-lambda
is not an experimental flag any more for a long time with NVCC, it is an alias for the--extended-lambda
flag and is officially supported. We had agreed for a while that we would use it by default once it would not be considered experimental but somehow never got around to actually implement it. So here it is.nvcc_wrapper
so it can handle-[-]extended-lambda
(w/o theexpt-
prefix)Kokkos_ENABLE_CUDA_LAMBDA
option defaultOFF -> ON
Preferedit this controversial bit was dropped from this PR-extended-lambda
flag (that change might be problematic if someone was using a pinned version ofnvcc_wrapper
)NOTE: The default value with generated makefiles has not been updated. I am not sure how we would go about that one.