[release/7.0] Fix bug in String.Equals unrolling for CJK single-char strings #78234
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Outdated suggestions cannot be applied.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.
Manual backport of #78190 to release/7.0
Customer Impact
Fixes #78169 regression introduced in the String.Equals/StartsWith unrolling optimization in NET 7.0.
Prints different values depending on code being optimized or not (by jit). All comparisons against single-char literals of certain symbols (all two-bytes chars which pass this check:
bool IsAffected(char c) => (((ushort)c) >> 15) == 1;
) of CJK languages are affected so we probably want to land the fix it as soon as possible.Testing
We had a pretty big test coverage for all sorts of inputs for that optimizations, unfortunately, for the non-ASCII one we didn't use any of CJK symbols. Added CJK and did a fuzzy testing locally via a hand-written test generator for this API.
Risk
Low