Only add GVMDependencies for types with used GVMs #118704
Merged
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
Ever since GVM support was added to native AOT, we were generating the GVM resolution metadata for every type considered allocated. This included GVMs that were never even called: see `TypeGVMEntriesNode` that simply goes over everything on the type - we were adding a `TypeGVMEntriesNode` for all allocated types that have something to do with GVMs. This PR changes it so that we only generate `TypeGVMEntries` for types that have at least one used GVM. This is a scoped down version of dotnet#118632 (that tries to track things per-method) with a lot less risk.
There was a problem hiding this comment.
Pull Request Overview
This PR optimizes Generic Virtual Method (GVM) dependency generation in the native AOT compiler by only creating GVM resolution metadata for types that actually have used GVMs, rather than for all allocated types that could potentially have GVMs.
Key changes:
- Remove eager GVM metadata generation from type allocation
- Add conditional GVM metadata generation based on actual method usage
- Track implementation discovery to determine when GVM metadata is needed
Reviewed Changes
Copilot reviewed 3 out of 3 changed files in this pull request and generated no comments.
| File | Description |
|---|---|
| GVMDependenciesNode.cs | Adds conditional TypeGVMEntries dependency generation based on found implementations and tracks inheritance hierarchy |
| EETypeNode.cs | Removes automatic TypeGVMEntries dependency for types with GVM flags |
| ConstructedEETypeNode.cs | Removes automatic TypeGVMEntries dependency for canonical subtypes with GVM flags |
You can also share your feedback on Copilot code review for a chance to win a $100 gift card. Take the survey.
|
Tagging subscribers to this area: @agocke, @MichalStrehovsky, @jkotas |
|
/azp run runtime-nativeaot-outerloop |
|
Azure Pipelines successfully started running 1 pipeline(s). |
jkotas
approved these changes
Aug 13, 2025
This was referenced Aug 14, 2025
Open
|
/ba-g arm32 failures are dotnet/dnceng#6034 but infra doesn't pair it |
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.
Ever since GVM support was added to native AOT, we were generating the GVM resolution metadata for every type considered allocated. This included GVMs that were never even called: see
TypeGVMEntriesNodethat simply goes over everything on the type - we were adding aTypeGVMEntriesNodefor all allocated types that have something to do with GVMs.This PR changes it so that we only generate
TypeGVMEntriesfor types that have at least one used GVM. This is a scoped down version of #118632 (that tries to track things per-method) with a lot less risk.Cc @dotnet/ilc-contrib