Temporarily disable arm64 intrinsics in UTF-16 validation code paths #42052
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Outdated suggestions cannot be applied.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.
Fixes #41699.
Disables the ARM64-intrinsicified UTF-16 validation code paths since they ended up having poorer performance characteristics for non-ASCII data compared to their 3.x counterparts. With this PR, we skip the intrinsics entirely but stay within vectorized code as much as possible. The codegen and perf should be on-par with what shipped in 3.x.
Once this change is in, we can re-introduce tweaked versions of the arm64 code path, validate the perf numbers, and cherry-pick the "real" fix into the 5.x servicing branches.
/cc @jeffhandley