Skip to content

[MLIR][Conversion] Vector to LLVM: Remove unneeded vector shuffle #162946

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Merged
merged 4 commits into from
Oct 15, 2025
+42 −3
Merged

[MLIR][Conversion] Vector to LLVM: Remove unneeded vector shuffle #162946

Show file tree
Hide file tree
Changes from 3 commits
Commits
Show all changes
4 commits
Select commit Hold shift + click to select a range
c44404b
[MLIR][Conversion] Vector to LLVM: Remove unneeded vectorshuffle
silee2 Oct 11, 2025
f75fe90
Update comment.
silee2 Oct 13, 2025
b0d476e
Update test.
silee2 Oct 13, 2025
a33e198
Add op fold pattern for llvm.shufflevector and use it for optimizing …
silee2 Oct 14, 2025
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
7 changes: 6 additions & 1 deletion mlir/lib/Conversion/VectorToLLVM/ConvertVectorToLLVM.cpp
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -1723,12 +1723,17 @@ struct VectorBroadcastScalarToLowRankLowering
return success();
}

// For 1-d vector, we additionally do a `vectorshuffle`.
auto v =
LLVM::InsertElementOp::create(rewriter, broadcast.getLoc(), vectorType,
poison, adaptor.getSource(), zero);

// For 1-d vector, if vector width > 1, we additionally do a
// `vector shuffle`
int64_t width = cast<VectorType>(broadcast.getType()).getDimSize(0);
if (width == 1) {
rewriter.replaceOp(broadcast, v);
return success();
}
Copy link
Member

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Shouldn't this really just be a canonicalization for the shufflevector operation? Does this actually matter in practice where llvm isn't able to canonicalize it? I'm sure we can pre-canonicalize every piece of IR, but why are we doing it?

Copy link
Member

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Imagine if we start doing this for every operation that generates shufflevector, where we add special casing to always generate the canonicalized form. Why not just add it on the llvm operation or check if llvm does it for you?

Copy link
Member

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Why not implement it as a folding pattern on shufflevector and use createOrFold when creating the shufflevecotr operation?

Copy link
Contributor Author

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Added folding pattern for llvm.shufflevector and using createOrFold.

SmallVector<int32_t> zeroValues(width, 0);

// Shuffle the value across the desired number of elements.
Expand Down
12 changes: 12 additions & 0 deletions mlir/test/Conversion/VectorToLLVM/vector-to-llvm.mlir
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -76,6 +76,18 @@ func.func @broadcast_vec1d_from_f32(%arg0: f32) -> vector<2xf32> {

// -----

func.func @broadcast_single_elem_vec1d_from_f32(%arg0: f32) -> vector<1xf32> {
%0 = vector.broadcast %arg0 : f32 to vector<1xf32>
return %0 : vector<1xf32>
}
// CHECK-LABEL: @broadcast_single_elem_vec1d_from_f32
// CHECK-SAME: %[[A:.*]]: f32)
// CHECK: %[[T0:.*]] = llvm.insertelement %[[A]]
// CHECK-NOT: llvm.shufflevector
// CHECK: return %[[T0]] : vector<1xf32>

// -----

func.func @broadcast_vec1d_from_f32_scalable(%arg0: f32) -> vector<[2]xf32> {
%0 = vector.broadcast %arg0 : f32 to vector<[2]xf32>
return %0 : vector<[2]xf32>
Expand Down