[SPIRV] Rework on rootOp selection in kernel config.#23685
Merged
Conversation
The revision adds a getRootOperation utility mirroring LLVMGPU's multi-pass root selection strategy. The root is found by reverse-iterating compute ops with priority tiers: (1) named ops or generics with reductions, (2) elementwise generics, (3) fills. It also switches the softmax config from SPIRVBaseVectorize pipeline to SPIRVSubgroupReduce pipeline; it is a preparation for scf.forall migration. The scf.forall distribution prefers starting from root op for distribution. The SPIRVSubgroupReduce pipeline sets lowering config on root op only which matches the expectation. Signed-off-by: hanhanW <hanhan0912@gmail.com>
kuhar
approved these changes
Mar 6, 2026
Signed-off-by: hanhanW <hanhan0912@gmail.com>
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
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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.
The revision adds a getRootOperation utility mirroring LLVMGPU's multi-pass root selection strategy. The root is found by reverse-iterating compute ops with priority tiers: (1) named ops or generics with reductions, (2) elementwise generics, (3) fills.
It also switches the softmax config from SPIRVBaseVectorize pipeline to SPIRVSubgroupReduce pipeline; it is a preparation for scf.forall migration. The scf.forall distribution prefers starting from root op for distribution. The SPIRVSubgroupReduce pipeline sets lowering config on root op only which matches the expectation.