[clr-interp] Fix argument alignment of 16 byte or greater aligned structures in the interpreter #120723
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
…ures in the interpreter - Arguments are up to 16 byte aligned - Handle this in the call stub generator, and call emission. (We already did this in the argument input processing in the interpreter We need to align these at 16byte boundaries as the various ABIs tend to have a hard requirement around 16 byte alignment, when they don't have such a requirement around 32 or 64 byte alignment.
There was a problem hiding this comment.
Pull Request Overview
This PR fixes argument alignment of 16-byte or greater aligned structures in the CLR interpreter to properly handle structures that require 16-byte alignment boundaries as mandated by various ABIs.
Key Changes
- Updated call stub generator to allocate space for alignment injection routines (3x args instead of 2x)
- Added alignment logic in call stub computation to handle 16-byte aligned value types
- Implemented
InjectInterpStackAlignassembly routines across x64 and ARM64 platforms
Reviewed Changes
Copilot reviewed 9 out of 9 changed files in this pull request and generated 2 comments.
Show a summary per file
| File | Description |
|---|---|
src/tests/Interop/PInvoke/Int128/Int128Test.cs |
Removed FIXME comment indicating the interpreter issue is now resolved |
src/coreclr/vm/callstubgenerator.h |
Updated comment and allocation size calculation for call stub routines array |
src/coreclr/vm/callstubgenerator.cpp |
Added alignment logic and InjectInterpStackAlign function declaration |
src/coreclr/vm/arm64/asmhelpers.asm |
Added ARM64 assembly implementation of InjectInterpStackAlign |
src/coreclr/vm/arm64/asmhelpers.S |
Added ARM64 Unix assembly implementation of InjectInterpStackAlign |
src/coreclr/vm/amd64/asmhelpers.S |
Added x64 Unix assembly implementation of InjectInterpStackAlign |
src/coreclr/vm/amd64/AsmHelpers.asm |
Added x64 Windows assembly implementation of InjectInterpStackAlign |
src/coreclr/interpreter/compileropt.cpp |
Added alignment logic for variable allocation in interpreter |
src/coreclr/interpreter/compiler.cpp |
Added alignment constraint and fixed stack type for unbox operation |
davidwrighton
added
area-CodeGen-Interpreter-coreclr
and removed
area-VM-coreclr
labels
|
Tagging subscribers to this area: @BrzVlad, @janvorli, @kg |
kg
reviewed
Oct 14, 2025
kg
reviewed
Oct 14, 2025
kg
reviewed
Oct 14, 2025
kg
reviewed
Oct 14, 2025
kg
reviewed
Oct 14, 2025
kg
reviewed
Oct 14, 2025
|
Thanks for figuring this out and fixing it - that test failure had me stumped! |
kg
approved these changes
Oct 14, 2025
janvorli
approved these changes
Oct 14, 2025
This was referenced Oct 15, 2025
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.
We need to align these at 16byte boundaries as the various ABIs tend to have a hard requirement around 16 byte alignment, when they don't have such a requirement around 32 or 64 byte alignment.