Add DeviceAdjacentDifference::Subtract[Left|Right] overloads with two iterators#8588
Draft
bernhardmgruber wants to merge 7 commits intoNVIDIA:mainfrom
Draft
Add DeviceAdjacentDifference::Subtract[Left|Right] overloads with two iterators#8588bernhardmgruber wants to merge 7 commits intoNVIDIA:mainfrom
DeviceAdjacentDifference::Subtract[Left|Right] overloads with two iterators#8588bernhardmgruber wants to merge 7 commits intoNVIDIA:mainfrom
Conversation
Contributor
|
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. |
DeviceAdjacentDifference::Subtract[Left|Right] overloads with two iterators
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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.
While working on #8492, I realized that Thrust uses a version of
cub::DeviceAdjacentDifferencethat is not exposed in the public API: an overlapping version that uses dedicated input and output iterators. We only have an overload taking a single iterator (in-place) and another one taking two iterators but requiring the input and output range to not overlap.I am still not entirely sure whether we want those overloads and if Thrust's logic is fully correct. Thrust's API mentions:
but it does not mention that the input/output ranges are not allowed to overlap in any other way. The parallel overloads in the PSTL do not allow that. So Thrust seems to support more use cases here.
This PR adds the corresponding CUB overloads for Thrust use case, so we can migrate it away from using the dispatch layer.