C#: Invoke the extractor after the compiler to ensure that source generators have been executed. #20933
+6
−2
Conversation
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
michaelnebel
force-pushed
the
csharp/runtraceraftercompilation
branch
3 times, most recently
from
b3e96c4 to
2bb6c63
Compare
Copilot finished reviewing on behalf of
michaelnebel
December 3, 2025 13:06
Contributor
There was a problem hiding this comment.
Pull request overview
This PR modifies the C# tracer configuration to invoke the extractor after compiler execution rather than before, ensuring source generators have completed before extraction begins. This change addresses a breaking change in .NET 10's compiler invocation pattern and improves extraction reliability across all .NET versions.
Key Changes:
- Updated two
CreatePatternMatcherconfigurations to useORDER_AFTERinstead ofORDER_BEFOREfor CSC compiler patterns - Added change notes documenting the fix for compiler-generated file extraction
Reviewed changes
Copilot reviewed 2 out of 2 changed files in this pull request and generated no comments.
| File | Description |
|---|---|
| csharp/tools/tracing-config.lua | Changed extraction order from ORDER_BEFORE to ORDER_AFTER for Windows (csc.*\.exe) and POSIX (mcs\.exe, csc\.exe, csc) compiler matchers to ensure source generators complete before extraction |
| csharp/ql/lib/change-notes/2025-12-03-run-tracer-after-compilation.md | Added release notes documenting the fix for compiler-generated file extraction |
💡 Add Copilot custom instructions for smarter, more guided reviews. Learn how to get started.
…erators have been executed.
michaelnebel
force-pushed
the
csharp/runtraceraftercompilation
branch
from
9ff2a1b to
a903420
Compare
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
2 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.
In this PR, we modify the tracer to consistently run the extractor after the compiler call (which it already does in some cases), ensuring that all source generators have been executed.
This change addresses a breaking change in how .NET invokes the compiler in .NET 10 and beyond (details available here). Without this modification to the remaining patterns in the tracer, extraction of generated code will be broken in .NET 10+ (and is already broken in some cases for earlier .NET versions).
Two DCA experiments indicate this change is sound:
Autobuild suite: Shows an increase in extracted files for 13 projects and approximately 2,000 new alerts, with roughly 20 alerts removed. Spot-checking indicates that only projects with changed file extraction count have altered alert counts. Notably, all compilation errors for the project
Corona-Studio/ProjBobcathave been resolved—this project also accounts for the 20 removedcs/useless-assignment-to-local alerts(explained by successful compilation).Nightly suite: Exhibits similar patterns to the autobuild experiment. Significantly, all compilation errors for ASP.NET Core have been eliminated. These errors were likely introduced when the SHA was updated to an ASP.NET version compiled with .NET 10.
Neither experiment showed performance regressions (with one exception in the nightly suite—however, this project's extracted file count tripled due to generated source files, which explains the regression).