Skip to content

Commit

Permalink
[MHLO] Add convolution op pattern (#1152)
Browse files Browse the repository at this point in the history 
Co-authored-by: Bairen Yi <yibairen.byron@bytedance.com>
Co-authored-by: Jiawei Wu <xremold@gmail.com>
Co-authored-by: Tianyou Guo <tianyou.gty@alibaba-inc.com>
Co-authored-by: Xu Yan <yancey.yx@alibaba-inc.com>
Co-authored-by: Ziheng Jiang <ziheng.jiang@bytedance.com>
  • Loading branch information
@Vremold @Yancey1989 @ZihengJiang
6 people authored Aug 4, 2022
1 parent 08fc2d8 commit c94431f
Show file tree
Hide file tree
Showing 2 changed files with 250 additions and 0 deletions.
171 changes: 171 additions & 0 deletions lib/Conversion/TorchToMhlo/Linear.cpp
View file
Original file line number Diff line number Diff line change
Expand Up @@ -379,6 +379,171 @@ class ConvertAtenLinearOp : public ConvertAtenMatmulBaseOp<AtenOpT> {

} // namespace

// AtenConvolutionOp
namespace {
class ConvertAtenConvlutionOp : public OpConversionPattern<AtenConvolutionOp> {
public:
using OpConversionPattern<AtenConvolutionOp>::OpConversionPattern;
using OpAdaptor = typename AtenConvolutionOp::Adaptor;

LogicalResult
matchAndRewrite(AtenConvolutionOp op, OpAdaptor adaptor,
ConversionPatternRewriter &rewriter) const override {
Value input = adaptor.input();
Value weight = adaptor.weight();

// The input shape is [N, C, H, W]
auto inputTy = input.getType().template cast<RankedTensorType>();
// The weight shape is [OC, (IC // groups), KH, KW]
// If tranposed is set to true, the weight shape changes to [IC, (OC //
// groups), KH, KW]
auto weightTy = weight.getType().template cast<RankedTensorType>();
auto outTy = getTypeConverter()
->convertType(op.getType())
.template cast<RankedTensorType>();

if (!inputTy || !weightTy || !outTy) {
return op.emitError("input, weight and output must be ranked tensors");
}

if (inputTy.getRank() < 3)
return op.emitError("only input with at least 3 dims valid");

SmallVector<int64_t> stride;
if (!matchPattern(op.stride(), m_TorchConstantIntList(stride))) {
return rewriter.notifyMatchFailure(op,
"non-const stride list unsupported");
}

SmallVector<int64_t> padding;
if (!matchPattern(op.padding(), m_TorchConstantIntList(padding))) {
return rewriter.notifyMatchFailure(op,
"non-const padding list unsupported");
}

SmallVector<int64_t> dilation;
if (!matchPattern(op.dilation(), m_TorchConstantIntList(dilation))) {
return rewriter.notifyMatchFailure(op,
"non-const dilation list unsupported");
}
SmallVector<int64_t> outputPadding;
if (!matchPattern(op.output_padding(),
m_TorchConstantIntList(outputPadding))) {
return rewriter.notifyMatchFailure(
op, "non-const output_padding list unsupported");
}
// Just ignore the outputPadding attribute
for (int64_t item : outputPadding) {
if (item != 0)
return rewriter.notifyMatchFailure(
op, "only zero output_padding list supported");
}

int64_t groups;
if (!matchPattern(op.groups(), m_TorchConstantInt(&groups))) {
return rewriter.notifyMatchFailure(op, "non-int groups unsupported");
}

bool transposed;
if (!matchPattern(op.transposed(), m_TorchConstantBool(&transposed))) {
return rewriter.notifyMatchFailure(op, "non-bool transposed unsupported");
}
if (transposed) {
return rewriter.notifyMatchFailure(
op, "only param tranposed of value 'false' supported!");
}

assert(padding.size() == dilation.size() &&
padding.size() == stride.size() &&
padding.size() == static_cast<size_t>(inputTy.getRank()) - 2);
int64_t nSpatialDims = padding.size();

// Get mhlo::ConvolutionOp attributes
DenseIntElementsAttr mhloWindowStride = DenseIntElementsAttr::get(
RankedTensorType::get({static_cast<long int>(stride.size())},
rewriter.getI64Type()),
stride);
std::vector<int64_t> mhloPaddingVec;
for (size_t i = 0; i < padding.size(); i++) {
mhloPaddingVec.emplace_back(padding[i]);
mhloPaddingVec.emplace_back(padding[i]);
}

DenseIntElementsAttr mhloPadding = DenseIntElementsAttr::get(
RankedTensorType::get(
{static_cast<long int>(padding.size()), static_cast<long int>(2)},
rewriter.getI64Type()),
mhloPaddingVec);

DenseIntElementsAttr mhloRhsDilation = DenseIntElementsAttr::get(
RankedTensorType::get({static_cast<long int>(dilation.size())},
rewriter.getI64Type()),
dilation);

SmallVector<int64_t> spatialDimensions;
for (int64_t i = 2; i < inputTy.getRank(); i++) {
spatialDimensions.emplace_back(i);
}
mhlo::ConvDimensionNumbersAttr dimensionNumbers =
mhlo::ConvDimensionNumbersAttr::get(
/*context=*/rewriter.getContext(), /*inputBatchDimension=*/0,
/*inputFeatureDimension=*/1,
/*inputSpatialDimensions=*/spatialDimensions,
/*kernelInputFeatureDimension=*/1,
/*kernelOutputFeatureDimension=*/0,
/*kernelSpatialDimensions=*/spatialDimensions,
/*outputBatchDimension=*/0, /*outputFeatureDimension=*/1,
/*outputSpatialDimensions=*/spatialDimensions);

IntegerAttr featureGroupCount =
IntegerAttr::get(rewriter.getI64Type(), groups);
IntegerAttr batchGroupCount = IntegerAttr::get(rewriter.getI64Type(), 1);

// mhlo::ConvolutionOp's optional attributes, leave them as default
DenseIntElementsAttr mhloLhsDilation;
DenseElementsAttr windowReversal;
ArrayAttr precisionConfig;

auto mhloConvOp = rewriter.create<mhlo::ConvolutionOp>(
op->getLoc(), outTy, input, weight, mhloWindowStride, mhloPadding,
mhloLhsDilation, mhloRhsDilation, windowReversal, dimensionNumbers,
featureGroupCount, batchGroupCount, precisionConfig);

auto bias = adaptor.bias();

// No bias provided
if (failed(checkNotNone(rewriter, op, op.bias()))) {
rewriter.replaceOp(op, mhloConvOp.getResult());
return success();
}

// Handle bias
if (!bias.getType().cast<RankedTensorType>()) {
return op.emitError("bias provided but not a ranked tensor");
}

auto biasTy = bias.getType().template cast<RankedTensorType>();
if (!biasTy.getElementType().isIntOrFloat()) {
return op.emitError("only floating-point or integer datatype "
"legalization for bias supported");
}

assert(biasTy.getRank() <= 1);

// Reshape and promote bias
auto inputUnsqzDims =
llvm::to_vector<4>(llvm::seq<int64_t>(-nSpatialDims, 0));
bias = *mhlo::unsqueezeTensor(rewriter, op, bias, inputUnsqzDims);
bias = mhlo::promoteType(rewriter, bias, outTy);

DenseIntElementsAttr bcastDimensions;
rewriter.replaceOpWithNewOp<chlo::BroadcastAddOp>(
op, outTy, mhloConvOp.getResult(), bias, bcastDimensions);
return success();
}
};
} // namespace

void mlir::torch::torch_to_mhlo::populateLinearOpPatternsAndLegality(
TypeConverter &typeConverter, RewritePatternSet &patterns,
ConversionTarget &target) {
Expand All @@ -402,4 +567,10 @@ void mlir::torch::torch_to_mhlo::populateLinearOpPatternsAndLegality(
patterns.add<ConvertAtenLinearOp<AtenOp>>(typeConverter, context);
INSERT_LINEAR_ATENOP_PATTERN(AtenLinearOp);
#undef INSERT_LINEAR_ATEMOP_PATTERN

#define INSERT_CONVOLUTION_ATENOP_PATTERN(AtenOp) \
target.addIllegalOp<AtenOp>(); \
patterns.add<ConvertAtenConvlutionOp>(typeConverter, context);
INSERT_CONVOLUTION_ATENOP_PATTERN(AtenConvolutionOp);
#undef INSERT_CONVOLUTION_ATENOP_PATTERN
}
79 changes: 79 additions & 0 deletions test/Conversion/TorchToMhlo/linear.mlir
View file
Original file line number Diff line number Diff line change
Expand Up @@ -266,3 +266,82 @@ func.func @torch.aten.mm$proj(%arg0: !torch.vtensor<[?,256],f32>) -> !torch.vten
return %1 : !torch.vtensor<[?,256],f32>
}

// -----

// CHECK-LABEL: func.func @torch.aten.convolution(
// CHECK-SAME: %[[ARG_0:.*]]: !torch.vtensor<[?,?,?,?],f32>,
// CHECK-SAME: %[[ARG_1:.*]]: !torch.vtensor<[?,?,?,?],f32>) -> !torch.vtensor<[?,?,?,?],f32> {
// CHECK: %[[T_0:.*]] = torch_c.to_builtin_tensor %[[ARG_0]] : !torch.vtensor<[?,?,?,?],f32> -> tensor<?x?x?x?xf32>
// CHECK: %[[T_1:.*]] = torch_c.to_builtin_tensor %[[ARG_1]] : !torch.vtensor<[?,?,?,?],f32> -> tensor<?x?x?x?xf32>
// CHECK: %[[T_2:.*]] = torch.constant.none
// CHECK: %[[T_4:.*]] = torch.constant.int 2
// CHECK: %[[T_5:.*]] = torch.constant.int 1
// CHECK: %[[T_6:.*]] = torch.constant.int 4
// CHECK: %[[T_7:.*]] = torch.constant.int 3
// CHECK: %[[T_8:.*]] = torch_c.to_i64 %[[T_7]]
// CHECK: %[[T_9:.*]] = torch.prim.ListConstruct %[[T_4]], %[[T_5]] : (!torch.int, !torch.int) -> !torch.list<int>
// CHECK: %[[T_10:.*]] = torch.prim.ListConstruct %[[T_6]], %[[T_4]] : (!torch.int, !torch.int) -> !torch.list<int>
// CHECK: %[[T_11:.*]] = torch.prim.ListConstruct %[[T_7]], %[[T_5]] : (!torch.int, !torch.int) -> !torch.list<int>
// CHECK: %[[T_12:.*]] = torch.prim.ListConstruct : () -> !torch.list<int>
// CHECK: %[[T_13:.*]] = torch.constant.bool false
// CHECK: %[[T_14:.*]] = mhlo.convolution(%[[T_0]], %[[T_1]])
// CHECK-SAME{LITERAL}: dim_numbers = [b, f, 0, 1]x[o, i, 0, 1]->[b, f, 0, 1], window = {stride = [2, 1], pad = [[4, 4], [2, 2]], rhs_dilate = [3, 1]} {batch_group_count = 1 : i64, feature_group_count = 3 : i64} : (tensor<?x?x?x?xf32>, tensor<?x?x?x?xf32>) -> tensor<?x?x?x?xf32>
// CHECK: %[[T_15:.*]] = torch_c.from_builtin_tensor %[[T_14]] : tensor<?x?x?x?xf32> -> !torch.vtensor<[?,?,?,?],f32>
// CHECK: return %[[T_15]] : !torch.vtensor<[?,?,?,?],f32>
func.func @torch.aten.convolution(%arg0: !torch.vtensor<[?,?,?,?],f32>, %arg1: !torch.vtensor<[?,?,?,?],f32>) -> !torch.vtensor<[?,?,?,?],f32> {
%none = torch.constant.none
%int2 = torch.constant.int 2
%int1 = torch.constant.int 1
%int4 = torch.constant.int 4
%int3 = torch.constant.int 3
%1 = torch.prim.ListConstruct %int2, %int1 : (!torch.int, !torch.int) -> !torch.list<int>
%2 = torch.prim.ListConstruct %int4, %int2 : (!torch.int, !torch.int) -> !torch.list<int>
%3 = torch.prim.ListConstruct %int3, %int1 : (!torch.int, !torch.int) -> !torch.list<int>
%4 = torch.prim.ListConstruct : () -> !torch.list<int>
%false = torch.constant.bool false
%5 = torch.aten.convolution %arg0, %arg1, %none, %1, %2, %3, %false, %4, %int3 : !torch.vtensor<[?,?,?,?],f32>, !torch.vtensor<[?,?,?,?],f32>, !torch.none, !torch.list<int>, !torch.list<int>, !torch.list<int>, !torch.bool, !torch.list<int>, !torch.int -> !torch.vtensor<[?,?,?,?],f32>
return %5 : !torch.vtensor<[?,?,?,?],f32>
}

// -----

// CHECK-LABEL: func.func @torch.aten.convolution$bias(
// CHECK-SAME: %[[ARG_0:.*]]: !torch.vtensor<[?,?,?,?],f32>, %[[ARG_1:.*]]: !torch.vtensor<[?,?,?,?],f32>,
// CHECK-SAME: %[[ARG_2:.*]]: !torch.vtensor<[?],f32>) -> !torch.vtensor<[?,?,?,?],f32> {
// CHECK: %[[T_0:.*]] = torch_c.to_builtin_tensor %[[ARG_0]] : !torch.vtensor<[?,?,?,?],f32> -> tensor<?x?x?x?xf32>
// CHECK: %[[T_1:.*]] = torch_c.to_builtin_tensor %[[ARG_1]] : !torch.vtensor<[?,?,?,?],f32> -> tensor<?x?x?x?xf32>
// CHECK: %[[T_2:.*]] = torch_c.to_builtin_tensor %[[ARG_2]] : !torch.vtensor<[?],f32> -> tensor<?xf32>
// CHECK: %int2 = torch.constant.int 2
// CHECK: %int1 = torch.constant.int 1
// CHECK: %int4 = torch.constant.int 4
// CHECK: %int3 = torch.constant.int 3
// CHECK: %[[T_3:.*]] = torch_c.to_i64 %int3
// CHECK: %[[T_4:.*]] = torch.prim.ListConstruct %int2, %int1 : (!torch.int, !torch.int) -> !torch.list<int>
// CHECK: %[[T_5:.*]] = torch.prim.ListConstruct %int4, %int2 : (!torch.int, !torch.int) -> !torch.list<int>
// CHECK: %[[T_6:.*]] = torch.prim.ListConstruct %int3, %int1 : (!torch.int, !torch.int) -> !torch.list<int>
// CHECK: %[[T_7:.*]] = torch.prim.ListConstruct : () -> !torch.list<int>
// CHECK: %false = torch.constant.bool false
// CHECK: %[[T_8:.*]] = mhlo.convolution(%[[T_0]], %[[T_1]])
// CHECK{LITERAL}: dim_numbers = [b, f, 0, 1]x[o, i, 0, 1]->[b, f, 0, 1], window = {stride = [2, 1], pad = [[4, 4], [2, 2]], rhs_dilate = [3, 1]} {batch_group_count = 1 : i64, feature_group_count = 3 : i64} : (tensor<?x?x?x?xf32>, tensor<?x?x?x?xf32>) -> tensor<?x?x?x?xf32>
// CHECK: %[[IDX_0:.*]] = arith.constant 0 : index
// CHECK: %[[T_9:.*]] = tensor.dim %[[T_2]], %[[IDX_0]] : tensor<?xf32>
// CHECK: %[[T_10:.*]] = arith.index_cast %[[T_9]] : index to i64
// CHECK: %[[VAL_0:.*]] = arith.constant 1 : i64
// CHECK: %[[T_11:.*]] = tensor.from_elements %[[T_10]], %[[VAL_0]], %[[VAL_0]] : tensor<3xi64>
// CHECK: %[[T_12:.*]] = "mhlo.dynamic_reshape"(%[[T_2]], %[[T_11]]) : (tensor<?xf32>, tensor<3xi64>) -> tensor<?x1x1xf32>
// CHECK: %[[T_13:.*]] = chlo.broadcast_add %[[T_8]], %[[T_12]] : (tensor<?x?x?x?xf32>, tensor<?x1x1xf32>) -> tensor<?x?x?x?xf32>
// CHECK: %[[T_14:.*]] = torch_c.from_builtin_tensor %[[T_13]] : tensor<?x?x?x?xf32> -> !torch.vtensor<[?,?,?,?],f32>
// CHECK: return %[[T_14]] : !torch.vtensor<[?,?,?,?],f32>
func.func @torch.aten.convolution$bias(%arg0: !torch.vtensor<[?,?,?,?],f32>, %arg1: !torch.vtensor<[?,?,?,?],f32>, %arg2: !torch.vtensor<[?],f32>) -> !torch.vtensor<[?,?,?,?],f32> {
%int2 = torch.constant.int 2
%int1 = torch.constant.int 1
%int4 = torch.constant.int 4
%int3 = torch.constant.int 3
%1 = torch.prim.ListConstruct %int2, %int1 : (!torch.int, !torch.int) -> !torch.list<int>
%2 = torch.prim.ListConstruct %int4, %int2 : (!torch.int, !torch.int) -> !torch.list<int>
%3 = torch.prim.ListConstruct %int3, %int1 : (!torch.int, !torch.int) -> !torch.list<int>
%4 = torch.prim.ListConstruct : () -> !torch.list<int>
%false = torch.constant.bool false
%5 = torch.aten.convolution %arg0, %arg1, %arg2, %1, %2, %3, %false, %4, %int3 : !torch.vtensor<[?,?,?,?],f32>, !torch.vtensor<[?,?,?,?],f32>, !torch.vtensor<[?],f32>, !torch.list<int>, !torch.list<int>, !torch.list<int>, !torch.bool, !torch.list<int>, !torch.int -> !torch.vtensor<[?,?,?,?],f32>
return %5 : !torch.vtensor<[?,?,?,?],f32>
}

0 comments on commit c94431f

Please sign in to comment.