Large strings support in cudf::merge #15374
Merged
Conversation
This file contains 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
davidwendt
added
2 - In Progress
3 tasks
davidwendt
added
3 - Ready for Review
robertmaynard
approved these changes
Apr 10, 2024
bdice
reviewed
Apr 10, 2024
| total_bytes < threshold, "Size of output exceeds the column size limit", std::overflow_error); | ||
| } | ||
| if (total_bytes >= get_offset64_threshold()) { | ||
| // recompute as int64 offsets when above the threshold |
There was a problem hiding this comment.
Once large strings are the default behavior, I wonder if we should do a first pass in int64 and then cast the offsets down if it fits in int32, rather than recomputing as int64 on overflow. I don't know how expensive the downcast would be compared to recomputing. Of course this also depends on whether we expect large strings to be common. I don't expect any action here, but would be open to your thoughts.
There was a problem hiding this comment.
I don't really expect large strings to be very common. At least not in the near to mid future.
I'd rather force any performance hit into the int64 path right now if possible.
bdice
approved these changes
Apr 18, 2024
nvdbaranec
reviewed
Apr 18, 2024
ttnghia
reviewed
Apr 18, 2024
vuule
reviewed
Apr 18, 2024
nvdbaranec
approved these changes
Apr 18, 2024
vuule
reviewed
Apr 18, 2024
vuule
approved these changes
Apr 19, 2024
|
/merge |
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Outdated suggestions cannot be applied.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.
Description
Enable large strings support in
cudf::merge.Simplifies the strings specialization to use the gather-based strings factory function which is already optimized for long strings and is now appropriately enabled for large strings.
Also moved source from
include/cudf/strings/detail/merge.cuhtosrc/strings/merge/merge.cufile since the template implemenation was not actually required.Checklist