Fix regression in IsPrefix / IsSuffix benchmarks#61935
Merged
elinor-fung merged 1 commit intodotnet:mainfrom Nov 22, 2021
Merged
Fix regression in IsPrefix / IsSuffix benchmarks#61935elinor-fung merged 1 commit intodotnet:mainfrom
elinor-fung merged 1 commit intodotnet:mainfrom
Conversation
jkoritzinsky
approved these changes
Nov 22, 2021
AaronRobinsonMSFT
approved these changes
Nov 22, 2021
Member
|
I'm not usually a fan of putting bug ID's into the product code, but will we ever know to remove these when the JIT is improved? |
Member
|
@danmoseley This is part of an existing JIT issue we are tracking to improve this scenario—#61821 (comment). We could add a comment too I guess. |
Member
Ah a note in an active issue works too. |
This was referenced Nov 29, 2021
This was referenced Nov 30, 2021
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 subscribe to this conversation on GitHub.
Already have an account?
Sign in.
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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.
The regressed scenarios actually did not call into any p/invokes - which is actually why they are the ones that regressed. With the switch to
GeneratedDllImportin #61640, the generated stub and blittable p/invoke forInterop.Globalization.EndsWithandInterop.Globalization.StartsWithended up inlined into their callers, resulting in a p/invoke frame being initialized even in the cases that did not use the p/invoke.This PR marks those two functions as no inlining and gets us back to where it was before #61640. The issue was filed for arm64, but it is the same problem on x64 (just by less). Ran on Windows x64:
Results for System.Globalization.Tests.StringSearch*
Fixes #61821
@AaronRobinsonMSFT @jkoritzinsky