Allow more folding of SIMD compares #117099
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
github-actions
bot
added
the
area-CodeGen-coreclr
Contributor
|
Tagging subscribers to this area: @JulieLeeMSFT, @jakobbotsch |
tannergooding
force-pushed
the
better-simd-cmp
branch
from
4c1b4ca to
a44eed5
Compare
Contributor
There was a problem hiding this comment.
Pull Request Overview
This PR extends the constant folding optimizations for SIMD compare operations on XARCH targets. Key changes include:
- Adding new handling for AVX/AVX512 compare intrinsics in rationalize.cpp
- Enhancing the transformation logic in morph.cpp and lowering in lowerxarch.cpp for more SIMD compare cases
- Removing the now-unused float comparison swap helper in hwintrinsicxarch.cpp and its declaration in hwintrinsic.h, and adding additional folding logic in gentree.cpp
Reviewed Changes
Copilot reviewed 6 out of 6 changed files in this pull request and generated no comments.
Show a summary per file
| File | Description |
|---|---|
| src/coreclr/jit/rationalize.cpp | Adds a new switch case for AVX/AVX512 compare constants folding on XARCH |
| src/coreclr/jit/morph.cpp | Introduces transformations for folding compare intrinsics with constant mode adjustments |
| src/coreclr/jit/lowerxarch.cpp | Updates mask intrinsic handling and constant folding logic for compare operations |
| src/coreclr/jit/hwintrinsicxarch.cpp | Removes the obsolete lookupFloatComparisonModeForSwappedArgs function |
| src/coreclr/jit/hwintrinsic.h | Removes the declaration of the obsolete lookupFloatComparisonModeForSwappedArgs |
| src/coreclr/jit/gentree.cpp | Extends folding logic to support additional SIMD compare intrinsics |
Comments suppressed due to low confidence (2)
src/coreclr/jit/morph.cpp:9927
- In this branch, the constant value should be updated to 'newMode' instead of 'mode' to reflect the intended transformation when the mode is modified.
FloatComparisonMode mode =
src/coreclr/jit/lowerxarch.cpp:3077
- This update should assign 'newMode' to cmpOp3 rather than the original 'mode' to ensure the transformation correctly folds the intrinsic.
cmpOp3->AsIntConCommon()->SetIntegralValue(static_cast<uint8_t>(mode));
Member
Author
|
CC. @dotnet/jit-contrib, small improvement to the xarch codegen and constant folding around compares. Normalizes more of the compare constants to require less nodes and does so early, allowing more constant folding and small output in some scenarios. |
EgorBo
reviewed
Jul 1, 2025
EgorBo
approved these changes
Jul 1, 2025
Sign up for free
to subscribe to this conversation on GitHub.
Already have an account?
Sign in.
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.
This ensures that more SIMD compares can undergo constant folding and light up as optimizations alongside the broader handling.