Skip to content

fix: add a separate check for pad_dot_general #1446

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

Open
wants to merge 1 commit into
base: main
Choose a base branch
from
Open

fix: add a separate check for pad_dot_general #1446

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
1 commit
Select commit
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
11 changes: 7 additions & 4 deletions src/enzyme_ad/jax/Passes/EnzymeHLOOpt.cpp
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -9667,12 +9667,15 @@ struct PadDotGeneral
nextOtherArg = slice.getResult();
}

auto lhs = otherIsLHS ? nextOtherArg : pad.getOperand();
auto rhs = otherIsLHS ? pad.getOperand() : nextOtherArg;
if (!areValidDotGeneralInputs(lhs, rhs, op.getDotDimensionNumbersAttr()))
Copy link
Member

@wsmoses wsmoses Oct 4, 2025

Choose a reason for hiding this comment

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

this feels like the wrong solution. shouldn't we just slice the inputs properly here?

return failure();

Value res = rewriter.create<stablehlo::DotGeneralOp>(
op.getLoc(),
RankedTensorType::get(resultShape, op.getType().getElementType()),
otherIsLHS ? nextOtherArg : pad.getOperand(),
otherIsLHS ? pad.getOperand() : nextOtherArg,
op.getDotDimensionNumbersAttr(), op.getPrecisionConfigAttr(),
RankedTensorType::get(resultShape, op.getType().getElementType()), lhs,
rhs, op.getDotDimensionNumbersAttr(), op.getPrecisionConfigAttr(),
op.getAlgorithmAttr());

if (!resultDimsToPad.empty()) {
Expand Down
22 changes: 22 additions & 0 deletions src/enzyme_ad/jax/Utils.cpp
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -1248,6 +1248,28 @@ bool reshapeIsTranspose(stablehlo::ReshapeOp reshapeOp) {
return true;
}

bool areValidDotGeneralInputs(
Value lhs, Value rhs, stablehlo::DotDimensionNumbersAttr dotDimNumbers) {
auto lhsShape = cast<ShapedType>(lhs.getType()).getShape();
auto rhsShape = cast<ShapedType>(rhs.getType()).getShape();

for (auto [lhsBatchDim, rhsBatchDim] :
llvm::zip(dotDimNumbers.getLhsBatchingDimensions(),
dotDimNumbers.getRhsBatchingDimensions())) {
if (lhsShape[lhsBatchDim] != rhsShape[rhsBatchDim])
return false;
}

for (auto [lhsContractingDim, rhsContractingDim] :
llvm::zip(dotDimNumbers.getLhsContractingDimensions(),
dotDimNumbers.getRhsContractingDimensions())) {
if (lhsShape[lhsContractingDim] != rhsShape[rhsContractingDim])
return false;
}

return true;
}

} // namespace stablehlo

} // namespace mlir
3 changes: 3 additions & 0 deletions src/enzyme_ad/jax/Utils.h
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -716,6 +716,9 @@ negatedComparisonDirection(stablehlo::ComparisonDirection direction);

bool reshapeIsTranspose(stablehlo::ReshapeOp reshapeOp);

bool areValidDotGeneralInputs(Value lhs, Value rhs,
stablehlo::DotDimensionNumbersAttr dotDimNumbers);

} // namespace stablehlo

} // namespace mlir
17 changes: 17 additions & 0 deletions test/lit_tests/paddotgeneral2.mlir
View file
Open in desktop
Original file line number Diff line number Diff line change
@@ -0,0 +1,17 @@
// RUN: enzymexlamlir-opt --pass-pipeline="any(enzyme-hlo-generate-td{patterns=pad_dot_general<1>(1)},transform-interpreter,enzyme-hlo-remove-transform)" %s | FileCheck %s

func.func @fn(%arg0: tensor<f32>) -> (tensor<32x32xf32>, tensor<f32>) {
%c = stablehlo.constant dense<3> : tensor<32x32xi64>
%c_0 = stablehlo.constant dense<2> : tensor<32x32xi64>
%cst = stablehlo.constant dense<0.000000e+00> : tensor<f32>
%0 = stablehlo.reshape %arg0 : (tensor<f32>) -> tensor<1x1xf32>
%1 = stablehlo.pad %0, %cst, low = [2, 3], high = [29, 28], interior = [0, 0] : (tensor<1x1xf32>, tensor<f32>) -> tensor<32x32xf32>
%2 = stablehlo.convert %c_0 : (tensor<32x32xi64>) -> tensor<32x32xf32>
%3 = stablehlo.multiply %2, %1 : tensor<32x32xf32>
%4 = stablehlo.convert %c : (tensor<32x32xi64>) -> tensor<32x32xf32>
%5 = stablehlo.subtract %3, %4 : tensor<32x32xf32>
%6 = stablehlo.pad %0, %cst, low = [3, 2], high = [28, 29], interior = [0, 0] : (tensor<1x1xf32>, tensor<f32>) -> tensor<32x32xf32>
%7 = stablehlo.dot_general %6, %5, contracting_dims = [0] x [1], precision = [DEFAULT, DEFAULT] : (tensor<32x32xf32>, tensor<32x32xf32>) -> tensor<32x32xf32>
// CHECK: %7 = stablehlo.dot_general %6, %5, contracting_dims = [0] x [1], precision = [DEFAULT, DEFAULT] : (tensor<32x32xf32>, tensor<32x32xf32>) -> tensor<32x32xf32>
return %7, %arg0 : tensor<32x32xf32>, tensor<f32>
}
Loading