[Codegen] Support codegen for vectorized tir.ShuffleNode #17748
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
This PR introduces the support for vectorized tir.ShuffleNode, which is useful for extracting bits and converting to float4, since float4 is sub-byte. Prior to this PR, ShuffleNode is not supported in vectorization. This PR allows vectorizing ShuffleNode subject to special patterns, and still throws error for ShuffleNodes that don't meet the pattern requirements.
yongwww
approved these changes
Mar 15, 2025
MasterJH5574
added a commit
that referenced
this pull request
May 4, 2025
) This PR overrides `PrintVecElemLoad()` and `PrintVecElemStore()` for the WebGPU backend. Otherwise, we would generate things like `(QK_local[0i].s0)` for WebGPU, which is not a valid syntax in WGSL. Instead, we generate `(QK_local[0i][0])` after this PR. `QK_local` here is a `array<vec4<f32>, 1>`. This issue prevented WebLLM from generating the correct kernel after #17748 Co-authored-by: Ruihang Lai <ruihangl@cs.cmu.edu>
ShiboXing
pushed a commit
to ShiboXing/tvm
that referenced
this pull request
Aug 10, 2025
This PR introduces the support for vectorized tir.ShuffleNode, which is useful for extracting bits and converting to float4, since float4 is sub-byte. Prior to this PR, ShuffleNode is not supported in vectorization. This PR allows vectorizing ShuffleNode subject to special patterns, and still throws error for ShuffleNodes that don't meet the pattern requirements.
ShiboXing
pushed a commit
to ShiboXing/tvm
that referenced
this pull request
Aug 10, 2025
…che#17917) This PR overrides `PrintVecElemLoad()` and `PrintVecElemStore()` for the WebGPU backend. Otherwise, we would generate things like `(QK_local[0i].s0)` for WebGPU, which is not a valid syntax in WGSL. Instead, we generate `(QK_local[0i][0])` after this PR. `QK_local` here is a `array<vec4<f32>, 1>`. This issue prevented WebLLM from generating the correct kernel after apache#17748 Co-authored-by: Ruihang Lai <ruihangl@cs.cmu.edu>
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.
This PR introduces the support for vectorized tir.ShuffleNode, which is useful for extracting bits and converting to float4, since float4 is sub-byte.
Prior to this PR, ShuffleNode is not supported in vectorization. This PR allows vectorizing ShuffleNode subject to special patterns, and still throws error for ShuffleNodes that don't meet the pattern requirements.