[Async lightbulb performance] Move suppression fixes to 'Low' priority bucket#56959
Merged
mavasani merged 5 commits intodotnet:mainfrom Dec 1, 2021
Merged
[Async lightbulb performance] Move suppression fixes to 'Low' priority bucket#56959mavasani merged 5 commits intodotnet:mainfrom
mavasani merged 5 commits intodotnet:mainfrom
Conversation
a715801 to
cdb9ac6
Compare
Contributor
Author
|
/azp run |
|
Azure Pipelines successfully started running 3 pipeline(s). |
cdb9ac6 to
f2e3b8f
Compare
1236b74 to
928ccb4
Compare
mavasani
commented
Nov 8, 2021
src/VisualStudio/IntegrationTest/IntegrationTests/CSharp/CSharpCodeActions.cs
Show resolved
Hide resolved
This was referenced Nov 11, 2021
Addresses second item in dotnet#56843 Prior to this change, all the analyzers that are not high-pri are added to the normal priority bucket and executed at once. This PR further breaks down the normal priority bucket into two buckets as follow: 1. `Normal`: Analyzers that produce at least one fixable diagnostic, i.e. have a corresponding code fixer. 2. `Low`: Analyzers without any fixable diagnostics, i.e. have no corresponding code fixer. These diagnostics will only have Suppression/Configuration quick actions, which are lower priority and always show up at the bottom of the light bulb. This new bucketing allows us to reduce the number of analyzers that are executed in the Normal priority bucket and hence generate all the non-suppression/configuration code fixes faster in async light bulb.
d4ee79c to
0a2f5d7
Compare
Contributor
Author
|
@CyrusNajmabadi this should be ready for review |
src/EditorFeatures/Core.Wpf/Suggestions/AsyncSuggestedActionsSource.cs
Outdated
Show resolved
Hide resolved
src/EditorFeatures/Test/EditAndContinue/Helpers/MockDiagnosticAnalyzerService.cs
Outdated
Show resolved
Hide resolved
src/Features/Core/Portable/Diagnostics/DiagnosticAnalyzerService.cs
Outdated
Show resolved
Hide resolved
...es/Core/Portable/Diagnostics/EngineV2/DiagnosticIncrementalAnalyzer_GetDiagnosticsForSpan.cs
Outdated
Show resolved
Hide resolved
...es/Core/Portable/Diagnostics/EngineV2/DiagnosticIncrementalAnalyzer_GetDiagnosticsForSpan.cs
Outdated
Show resolved
Hide resolved
...es/Core/Portable/Diagnostics/EngineV2/DiagnosticIncrementalAnalyzer_GetDiagnosticsForSpan.cs
Show resolved
Hide resolved
...es/Core/Portable/Diagnostics/EngineV2/DiagnosticIncrementalAnalyzer_GetDiagnosticsForSpan.cs
Outdated
Show resolved
Hide resolved
src/Features/Core/Portable/Diagnostics/IDiagnosticAnalyzerService.cs
Outdated
Show resolved
Hide resolved
CyrusNajmabadi
suggested changes
Nov 19, 2021
Contributor
CyrusNajmabadi
left a comment
There was a problem hiding this comment.
overall i'm highly supportive of this. however, i think we shoudl takea slightly longer view and try to get this benefit without forcing the 'low' group only for suppressions. we have the 'low' concept today for many fixes and refactorings. Indeed, internally we have 'low' and 'lowest' (the latter is used for config/suppress). So i think we shoudl model the entire stack from top to bottom this way.
It will prevent a lot of confusing mismatches about which sort of 'low' we're talking about.
Contributor
Author
|
@CyrusNajmabadi Thanks for the feedback, I have addressed your comments, please take a look. |
590238c to
c75be31
Compare
c75be31 to
da34e59
Compare
...es/Core/Portable/Diagnostics/EngineV2/DiagnosticIncrementalAnalyzer_GetDiagnosticsForSpan.cs
Show resolved
Hide resolved
...rkspaces/SharedUtilitiesAndExtensions/Compiler/Core/CodeActions/CodeActionRequestPriority.cs
Show resolved
Hide resolved
CyrusNajmabadi
approved these changes
Nov 29, 2021
Contributor
CyrusNajmabadi
left a comment
There was a problem hiding this comment.
nice. this got a lot cleaner. i like it. just some basic suggestions for clarity. tehre are some places that could use some comments to help with understanding.
Contributor
Author
|
Thank you for the review @CyrusNajmabadi! |
mavasani
added a commit
to mavasani/roslyn
that referenced
this pull request
Dec 17, 2021
…edit/1450689) This regression was introduced with async lightbulb performance improvement PR: dotnet#56959. With that PR, we now optimize/reduce the number of analyzers to execute in the "Normal" priority bucket for async light bulb. We earlier executed all analyzers in this bucket, but now only try to execute analyzers which report at least one diagnostic ID which has a corresponding code fix provider that can fix that diagnostic ID. To identify this reduced set of analyzers, we execute this newly added perf optimization to skip analyzers that don't have a code fix: https://github.com/dotnet/roslyn/blob/88f829939583d49d9e08104a71a7645d712bcfbc/src/Features/Core/Portable/Diagnostics/EngineV2/DiagnosticIncrementalAnalyzer_GetDiagnosticsForSpan.cs#L151-L155 `_owner.DiagnosticAnalyzerInfoCache.GetDiagnosticDescriptors` call in this code is problematic for TS analyzers, because these are not regular DiagnosticAnalyzer sub-types, but instead derive from our internal DocumentDiagnosticAnalyzer and ProjectDiagnosticAnalyzer types, and this analyzer lies to us and returns 0 supported diagnostic descriptors: https://github.com/dotnet/roslyn/blob/315c2e149ba7889b0937d872274c33fcbfe9af5f/src/Features/Core/Portable/ExternalAccess/VSTypeScript/VSTypeScriptDocumentDiagnosticAnalyzer.cs#L16 and https://github.com/dotnet/roslyn/blob/315c2e149ba7889b0937d872274c33fcbfe9af5f/src/Features/Core/Portable/ExternalAccess/VSTypeScript/VSTypeScriptProjectDiagnosticAnalyzer.cs#L16. So, we skip TS analyzers for "Normal" priority bucket. This PR implements a short term fix to update the newly added code to special case DocumentDiagnosticAnalyzers and skip the newly added perf optimization for these analyzers. Note that ProjectDiagnosticAnalyzers are never executed in lightbulb/code fix path. I will create a separate issue to track designing a better long term solution for TS analyzers to avoid such breaking changes in future.
mavasani
added a commit
to mavasani/roslyn
that referenced
this pull request
Dec 17, 2021
…edit/1450689) This regression was introduced with async lightbulb performance improvement PR: dotnet#56959. With that PR, we now optimize/reduce the number of analyzers to execute in the "Normal" priority bucket for async light bulb. We earlier executed all analyzers in this bucket, but now only try to execute analyzers which report at least one diagnostic ID which has a corresponding code fix provider that can fix that diagnostic ID. To identify this reduced set of analyzers, we execute this newly added perf optimization to skip analyzers that don't have a code fix: https://github.com/dotnet/roslyn/blob/88f829939583d49d9e08104a71a7645d712bcfbc/src/Features/Core/Portable/Diagnostics/EngineV2/DiagnosticIncrementalAnalyzer_GetDiagnosticsForSpan.cs#L151-L155 `_owner.DiagnosticAnalyzerInfoCache.GetDiagnosticDescriptors` call in this code is problematic for TS analyzers, because these are not regular DiagnosticAnalyzer sub-types, but instead derive from our internal DocumentDiagnosticAnalyzer and ProjectDiagnosticAnalyzer types, and this analyzer lies to us and returns 0 supported diagnostic descriptors: https://github.com/dotnet/roslyn/blob/315c2e149ba7889b0937d872274c33fcbfe9af5f/src/Features/Core/Portable/ExternalAccess/VSTypeScript/VSTypeScriptDocumentDiagnosticAnalyzer.cs#L16 and https://github.com/dotnet/roslyn/blob/315c2e149ba7889b0937d872274c33fcbfe9af5f/src/Features/Core/Portable/ExternalAccess/VSTypeScript/VSTypeScriptProjectDiagnosticAnalyzer.cs#L16. So, we skip TS analyzers for "Normal" priority bucket. This PR implements a short term fix to update the newly added code to special case DocumentDiagnosticAnalyzers and skip the newly added perf optimization for these analyzers. Note that ProjectDiagnosticAnalyzers are never executed in lightbulb/code fix path. I will create a separate issue to track designing a better long term solution for TS analyzers to avoid such breaking changes in future.
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
3 participants
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 second item in #56843
Prior to this change, all the analyzers that are not high-pri are added to the normal priority bucket and executed at once. This PR further breaks down the normal priority bucket into two buckets as follow:
Normal: Analyzers that produce at least one fixable diagnostic, i.e. have a corresponding code fixer.Low: Analyzers without any fixable diagnostics, i.e. have no corresponding code fixer. These diagnostics will only have Suppression/Configuration quick actions, which are lower priority and always show up at the bottom of the light bulb.This new bucketing allows us to reduce the number of analyzers that are executed in the Normal priority bucket and hence generate all the non-suppression/configuration code fixes faster in async light bulb.