Improve performance of graph-based type-checking - don't validate filepaths unnecessarily#15431
Merged
T-Gro merged 10 commits intodotnet:mainfrom Jun 19, 2023
Merged
Improve performance of graph-based type-checking - don't validate filepaths unnecessarily#15431T-Gro merged 10 commits intodotnet:mainfrom
T-Gro merged 10 commits intodotnet:mainfrom
Conversation
safesparrow
commented
Jun 18, 2023
nojaf
approved these changes
Jun 19, 2023
Contributor
|
Great work! |
0101
approved these changes
Jun 19, 2023
T-Gro
approved these changes
Jun 19, 2023
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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.
Addresses most of the issues described in #15427 .
Change 1: speedup calculating
FSharpCoredependencyWe avoid calculating the same
FSharpCoredependency for every file - we do it once and we avoid using filepath validation which was shown to be slow and is not necessary.Change 2: speedup
queryTrieDo not memoize
queryTrie- memoization is much more costly than the invocation itself, because it involves generating hash keys and comparing keys of typestring list, and usesConcurrentDictionary.Also allow a
dualquery where two parts of theLongIdentifieras passed separately, avoiding the need to concatenate them into a new list.This change vastly reduces the time for


findDependenciescalls.Here is a DotTrace snapshot showing before:
and after: