intel/llvm
1
0
Fork 0
mirror of https://github.com/intel/llvm.git synced 2026-02-04 11:38:04 +08:00
Code Issues Packages Projects Releases Wiki Activity

[mlir][sparse][NFC] Switch InitOp to bufferization::AllocTensorOp

Browse Source 
Now that we have an AllocTensorOp (previously InitTensorOp) in the bufferization dialect, the InitOp in the sparse dialect is no longer needed.

Differential Revision: https://reviews.llvm.org/D126180
This commit is contained in:
Matthias Springer
2022-06-01 23:59:58 +02:00
parent 902360d21d
commit 6232a8f3d6
34 changed files with 117 additions and 200 deletions
Download Patch File Download Diff File Expand all files Collapse all files

27
mlir/lib/Dialect/SparseTensor/Transforms/SparseTensorConversion.cpp
View File

@@ -459,22 +459,33 @@ class SparseTensorNewConverter : public OpConversionPattern<NewOp> {
}
};
/// Sparse conversion rule for the init operator.
class SparseTensorInitConverter : public OpConversionPattern<InitOp> {
/// Sparse conversion rule for the alloc operator.
class SparseTensorAllocConverter
: public OpConversionPattern<bufferization::AllocTensorOp> {
using OpConversionPattern::OpConversionPattern;
LogicalResult
matchAndRewrite(InitOp op, OpAdaptor adaptor,
matchAndRewrite(bufferization::AllocTensorOp op, OpAdaptor adaptor,
ConversionPatternRewriter &rewriter) const override {
Type resType = op.getType();
RankedTensorType resType = op.getType();
auto enc = getSparseTensorEncoding(resType);
if (!enc)
return failure();
// Gather all dimension sizes as SSA values.
SmallVector<Value> sizes;
unsigned int operandCtr = 0;
for (int64_t i = 0; i < resType.getRank(); ++i) {
if (resType.isDynamicDim(i)) {
sizes.push_back(adaptor.getOperands()[operandCtr++]);
} else {
sizes.push_back(rewriter.create<arith::ConstantIndexOp>(
op.getLoc(), op.getStaticSize(i)));
}
}
// Generate the call to construct empty tensor. The sizes are
// explicitly defined by the arguments to the init operator.
// explicitly defined by the arguments to the alloc operator.
SmallVector<Value, 8> params;
ShapedType stp = resType.cast<ShapedType>();
newParams(rewriter, params, op, stp, enc, Action::kEmpty,
adaptor.getOperands());
newParams(rewriter, params, op, stp, enc, Action::kEmpty, sizes);
rewriter.replaceOp(op, genNewCall(rewriter, op, params));
return success();
}
@@ -912,7 +923,7 @@ void mlir::populateSparseTensorConversionPatterns(
const SparseTensorConversionOptions &options) {
patterns.add<SparseReturnConverter, SparseTensorToDimSizeConverter,
SparseCastConverter, SparseTensorNewConverter,
SparseTensorInitConverter, SparseTensorReleaseConverter,
SparseTensorAllocConverter, SparseTensorReleaseConverter,
SparseTensorToPointersConverter, SparseTensorToIndicesConverter,
SparseTensorToValuesConverter, SparseTensorLoadConverter,
SparseTensorLexInsertConverter, SparseTensorExpandConverter,