AtenCumprodOp (#3737)

Author: renxida

include/torch-mlir/Conversion/Utils/Utils.h

@@ -40,6 +40,8 @@ Value createInitTensor(OpBuilder &b, Location loc, ValueRange sizes,
 
 Value createZeroInitTensor(OpBuilder &b, Location loc, ValueRange sizes,
                            Type elemTy);
+Value createOneInitTensor(OpBuilder &b, Location loc, ValueRange sizes,
+                          Type elemTy);
 
 Value castIntToIndex(OpBuilder &b, Location loc, Value v);
 

lib/Conversion/TorchToTMTensor/TorchToTMTensor.cpp

@@ -1497,6 +1497,79 @@ class ConvertAtenSortOp : public OpConversionPattern<AtenSortOp> {
 };
 } // namespace
 
+namespace {
+class ConvertAtenCumprodOp : public OpConversionPattern<AtenCumprodOp> {
+public:
+  using OpConversionPattern::OpConversionPattern;
+  LogicalResult
+  matchAndRewrite(AtenCumprodOp op, OpAdaptor adaptor,
+                  ConversionPatternRewriter &rewriter) const override {
+
+    Location loc = op.getLoc();
+    Value input = adaptor.getSelf();
+    auto resultType = cast<RankedTensorType>(
+        getTypeConverter()->convertType(op->getResult(0).getType()));
+    Type elementType = resultType.getElementType();
+    Type inputElementType =
+        cast<RankedTensorType>(input.getType()).getElementType();
+
+    // Converting the input element type to the result's element type.
+    // The only possible mismatch would be when the input element type is an
+    // integer but not `si64`. Therefore, we directly convert the input to
+    // `si64`. Rest all cases are handled in the dtype definition for this op.
+    if (elementType != inputElementType) {
+      Value torchInput = convertTensorToDtype(
+          rewriter, loc, op.getSelf(),
+          rewriter.getIntegerType(64, IntegerType::Signed));
+      input = typeConverter->materializeTargetConversion(
+          rewriter, loc, typeConverter->convertType(torchInput.getType()),
+          torchInput);
+    }
+
+    int64_t inputRank = resultType.getRank();
+    Value dtype = op.getDtype();
+    if (!isa<Torch::NoneType>(dtype.getType()))
+      return rewriter.notifyMatchFailure(
+          op, "unsupported: dtype argument not supported");
+
+    int64_t dim;
+    if (!matchPattern(op.getDim(), m_TorchConstantInt(&dim)))
+      return rewriter.notifyMatchFailure(
+          op, "unimplemented: only constant dim value is supported");
+    dim = toPositiveDim(dim, inputRank);
+    if (!isValidDim(dim, inputRank))
+      return rewriter.notifyMatchFailure(op, "invalid dim");
+
+    SmallVector<Value> sizes = getTensorSizes(rewriter, loc, input);
+    Value output = createOneInitTensor(rewriter, loc, sizes, elementType);
+    output = rewriter.create<tensor::CastOp>(loc, resultType, output);
+
+    SmallVector<Value> accSizes(sizes);
+    accSizes.erase(accSizes.begin() + dim);
+    SmallVector<int64_t> accStatic(
+        makeShapeTorchCompatible(resultType.getShape()));
+    accStatic.erase(accStatic.begin() + dim);
+    Value acc = createOneInitTensor(rewriter, loc, accSizes, elementType);
+    Type accType =
+        RankedTensorType::get(makeShapeLLVMCompatible(accStatic), elementType);
+    acc = rewriter.create<tensor::CastOp>(loc, accType, acc);
+
+    Value result = createTMTensorScanOp(
+        rewriter, loc, input, output, acc, dim, /*inclusive=*/true,
+        [](OpBuilder &b, Location loc, Value input, Value acc) {
+          Value prod =
+              (isa<mlir::FloatType>(input.getType())
+                   ? b.create<arith::MulFOp>(loc, input, acc)->getResult(0)
+                   : b.create<arith::MulIOp>(loc, input, acc)->getResult(0));
+          b.create<TMTensor::YieldOp>(loc, prod);
+        });
+
+    rewriter.replaceOpWithNewOp<tensor::CastOp>(op, resultType, result);
+    return success();
+  }
+};
+} // namespace
+
 namespace {
 class ConvertAtenCumsumOp : public OpConversionPattern<AtenCumsumOp> {
 public:
@@ -2240,6 +2313,8 @@ class ConvertTorchToTMTensor
     patterns.add<ConvertAtenSortOp>(typeConverter, context);
     target.addIllegalOp<AtenCumsumOp>();
     patterns.add<ConvertAtenCumsumOp>(typeConverter, context);
+    target.addIllegalOp<AtenCumprodOp>();
+    patterns.add<ConvertAtenCumprodOp>(typeConverter, context);
     target.addIllegalOp<AtenScaledDotProductAttentionOp>();
     patterns.add<ConvertAtenScaledDotProductAttentionOp>(typeConverter,
                                                          context);

lib/Conversion/Utils/Utils.cpp

@@ -138,6 +138,16 @@ Value createZeroInitTensor(OpBuilder &b, Location loc, ValueRange sizes,
   return b.create<linalg::FillOp>(loc, c0, initTensor).getResult(0);
 }
 
+Value createOneInitTensor(OpBuilder &b, Location loc, ValueRange sizes,
+                          Type elemTy) {
+  Value initTensor =
+      b.create<tensor::EmptyOp>(loc, getAsOpFoldResult(sizes), elemTy);
+  RankedTensorType type = cast<RankedTensorType>(initTensor.getType());
+  Value c1 =
+      b.create<arith::ConstantOp>(loc, b.getOneAttr(type.getElementType()));
+  return b.create<linalg::FillOp>(loc, c1, initTensor).getResult(0);
+}
+
 Value castIntToIndex(OpBuilder &b, Location loc, Value v) {
   assert(isa<IntegerType>(v.getType()) && "must be called with integer type");
   return b.createOrFold<arith::IndexCastOp>(loc, b.getIndexType(), v);

lib/Dialect/Torch/Transforms/AbstractInterpLibrary.cpp

@@ -9134,6 +9134,9 @@ StringRef mlir::torch::Torch::getAbstractInterpLibrary() {
 "  func.func @\"__torch_mlir_shape_fn.aten.cumsum\"(%arg0: !torch.list<int>, %arg1: !torch.int, %arg2: !torch.optional<int>) -> !torch.list<int> {\n"
 "    return %arg0 : !torch.list<int>\n"
 "  }\n"
+"  func.func @\"__torch_mlir_shape_fn.aten.cumprod\"(%arg0: !torch.list<int>, %arg1: !torch.int, %arg2: !torch.optional<int>) -> !torch.list<int> {\n"
+"    return %arg0 : !torch.list<int>\n"
+"  }\n"
 "  func.func @\"__torch_mlir_shape_fn.aten.rand_like\"(%arg0: !torch.list<int>, %arg1: !torch.optional<int>, %arg2: !torch.optional<int>, %arg3: !torch.optional<Device>, %arg4: !torch.optional<bool>, %arg5: !torch.optional<int>) -> !torch.list<int> {\n"
 "    return %arg0 : !torch.list<int>\n"
 "  }\n"
@@ -11844,6 +11847,25 @@ StringRef mlir::torch::Torch::getAbstractInterpLibrary() {
 "    }\n"
 "    return %1 : !torch.int\n"
 "  }\n"
+"  func.func @\"__torch_mlir_dtype_fn.aten.cumprod\"(%arg0: !torch.tuple<int, int>, %arg1: !torch.int, %arg2: !torch.optional<int>) -> !torch.int {\n"
+"    %int4 = torch.constant.int 4\n"
+"    %none = torch.constant.none\n"
+"    %0 = torch.aten.__isnot__ %arg2, %none : !torch.optional<int>, !torch.none -> !torch.bool\n"
+"    %1 = torch.prim.If %0 -> (!torch.int) {\n"
+"      %2 = torch.prim.unchecked_cast %arg2 : !torch.optional<int> -> !torch.int\n"
+"      torch.prim.If.yield %2 : !torch.int\n"
+"    } else {\n"
+"      %2:2 = torch.prim.TupleUnpack %arg0 : !torch.tuple<int, int> -> !torch.int, !torch.int\n"
+"      %3 = func.call @__torch__.torch_mlir.jit_ir_importer.build_tools.library_generator.is_integer_dtype(%2#1) : (!torch.int) -> !torch.bool\n"
+"      %4 = torch.prim.If %3 -> (!torch.int) {\n"
+"        torch.prim.If.yield %int4 : !torch.int\n"
+"      } else {\n"
+"        torch.prim.If.yield %2#1 : !torch.int\n"
+"      }\n"
+"      torch.prim.If.yield %4 : !torch.int\n"
+"    }\n"
+"    return %1 : !torch.int\n"
+"  }\n"
 "  func.func @\"__torch_mlir_dtype_fn.aten.detach\"(%arg0: !torch.tuple<int, int>) -> !torch.int {\n"
 "    %0:2 = torch.prim.TupleUnpack %arg0 : !torch.tuple<int, int> -> !torch.int, !torch.int\n"
 "    return %0#1 : !torch.int\n"

projects/pt1/e2e_testing/xfail_sets.py

@@ -79,6 +79,7 @@
     #### General TorchDynamo/PyTorch errors
     # torch._dynamo.exc.Unsupported: Tensor.item
     "CumsumModule_basic",
+    "CumprodModule_basic",
     # TypeError: new_empty(): argument 'size' (position 1) must be tuple of ints, but found element of type NoneType at pos 0
     # RuntimeError: Failed running call_function aten.convolution_backward(...
     # https://github.com/pytorch/pytorch/issues/89629
@@ -432,6 +433,7 @@
     "ConvolutionBackwardModule2DStrided_basic",
     "ConvolutionBackwardModule2D_basic",
     "CumsumModule_basic",
+    "CumprodModule_basic",
     "DeformConv2D_basic",
     "DivFloatModule_basic",
     "DivIntModule_basic",
@@ -667,6 +669,10 @@
     "ConvolutionBackwardModule2DStrided_basic",
     "ConvolutionBackwardModule2D_basic",
     "CumsumModule_basic",
+    "CumprodModule_basic",
+    "CumprodInputDtypeInt32Module_basic",
+    "CumprodStaticModule_basic",
+    "CumprodStaticNegativeDimModule_basic",
     "DeformConv2D_basic",
     "DeterminantBatchedModule_F32",
     "DeterminantDynamicModule_F32",
@@ -1077,6 +1083,9 @@
     "CumsumInputDtypeInt32Module_basic",
     "CumsumStaticModule_basic",
     "CumsumStaticNegativeDimModule_basic",
+    "CumprodInputDtypeInt32Module_basic",
+    "CumprodStaticModule_basic",
+    "CumprodStaticNegativeDimModule_basic",
     "DetachModule_basic",
     "DivFloatModule_basic",
     "DivIntModule_basic",
@@ -3105,6 +3114,10 @@
     "CopyWithDifferentDTypesModule_basic",
     "CosineSimilarityStaticBroadcastModule_basic",
     "CumsumInputDtypeInt32Module_basic",
+    "CumprodModule_basic",
+    "CumprodInputDtypeInt32Module_basic",
+    "CumprodStaticModule_basic",
+    "CumprodStaticNegativeDimModule_basic",
     "ElementwiseAcosIntModule_basic",
     "ElementwiseAsinIntModule_basic",
     "ElementwiseAtanTensorIntModule_basic",
@@ -3378,6 +3391,10 @@
     "CumsumModule_basic",
     "CumsumStaticModule_basic",
     "CumsumStaticNegativeDimModule_basic",
+    "CumprodModule_basic",
+    "CumprodInputDtypeInt32Module_basic",
+    "CumprodStaticModule_basic",
+    "CumprodStaticNegativeDimModule_basic",
     "DeformConv2D_basic",
     "DeterminantBatchedModule_F32",
     "DeterminantDynamicModule_F32",
@@ -4110,6 +4127,10 @@
     "CumsumModule_basic",
     "CumsumStaticModule_basic",
     "CumsumStaticNegativeDimModule_basic",
+    "CumprodModule_basic",
+    "CumprodInputDtypeInt32Module_basic",
+    "CumprodStaticModule_basic",
+    "CumprodStaticNegativeDimModule_basic",
     "DeformConv2D_basic",
     "DeterminantModule_F32",
     "DeterminantBatchedModule_F32",

projects/pt1/python/torch_mlir/jit_ir_importer/build_tools/abstract_interp_lib_gen.py

@@ -1434,6 +1434,9 @@ def aten〇multinomial〡shape(self: List[int], num_samples: int, replacement: b
 def aten〇cumsum〡shape(self: List[int], dim: int, dtype: Optional[int] = None) -> List[int]:
     return self
 
+def aten〇cumprod〡shape(self: List[int], dim: int, dtype: Optional[int] = None) -> List[int]:
+    return self
+
 def aten〇rand_like〡shape(self: List[int], dtype: Optional[int] = None, layout: Optional[int] = None, device: Optional[device] = None, pin_memory: Optional[bool] = None, memory_format: Optional[int] = None) -> List[int]:
     return self
 
@@ -2926,6 +2929,18 @@ def aten〇cumsum〡dtype(self_rank_dtype: Tuple[int, int], dim: int, dtype: Opt
         return torch.int64
     return self_dtype
 
+
+@check_dtype_function(
+    _check_tensors_with_the_same_dtype(num_of_tensors=1, dim=0) +
+    _check_tensors_with_the_same_dtype(num_of_tensors=1, dim=0, dtype=torch.float32))
+def aten〇cumprod〡dtype(self_rank_dtype: Tuple[int, int], dim: int, dtype: Optional[int] = None) -> int:
+    if dtype is not None:
+        return dtype
+    self_rank, self_dtype = self_rank_dtype
+    if is_integer_dtype(self_dtype):
+        return torch.int64
+    return self_dtype
+
 @check_dtype_function(_check_tensors_with_the_same_dtype(num_of_tensors=1))
 def aten〇detach〡dtype(self_rank_dtype: Tuple[int, int]) -> int:
     self_rank, self_dtype = self_rank_dtype

projects/pt1/python/torch_mlir_e2e_test/test_suite/basic.py

@@ -4830,6 +4830,90 @@ def CumsumInputDtypeInt32Module_basic(module, tu: TestUtils):
 # ==============================================================================
 
 
+class CumprodModule(torch.nn.Module):
+    def __init__(self):
+        super().__init__()
+
+    @export
+    @annotate_args(
+        [
+            None,
+            ([-1, -1, -1], torch.float32, True),
+        ]
+    )
+    def forward(self, val):
+        ones = torch.ones([1], dtype=torch.int32)
+        return torch.ops.aten.cumprod(val, ones.item())
+
+
+@register_test_case(module_factory=lambda: CumprodModule())
+def CumprodModule_basic(module, tu: TestUtils):
+    module.forward(tu.rand(2, 7, 4))
+
+
+class CumprodStaticModule(torch.nn.Module):
+    def __init__(self):
+        super().__init__()
+
+    @export
+    @annotate_args(
+        [
+            None,
+            ([2, 7, 4], torch.float32, True),
+        ]
+    )
+    def forward(self, val):
+        return torch.ops.aten.cumprod(val, 1)
+
+
+@register_test_case(module_factory=lambda: CumprodStaticModule())
+def CumprodStaticModule_basic(module, tu: TestUtils):
+    module.forward(tu.rand(2, 7, 4))
+
+
+class CumprodStaticNegativeDimModule(torch.nn.Module):
+    def __init__(self):
+        super().__init__()
+
+    @export
+    @annotate_args(
+        [
+            None,
+            ([2, 7, 4], torch.float32, True),
+        ]
+    )
+    def forward(self, val):
+        return torch.ops.aten.cumprod(val, dim=-1)
+
+
+@register_test_case(module_factory=lambda: CumprodStaticNegativeDimModule())
+def CumprodStaticNegativeDimModule_basic(module, tu: TestUtils):
+    module.forward(tu.rand(2, 7, 4))
+
+
+class CumprodInputDtypeInt32Module(torch.nn.Module):
+    def __init__(self):
+        super().__init__()
+
+    @export
+    @annotate_args(
+        [
+            None,
+            ([2, 7, 4], torch.int32, True),
+        ]
+    )
+    def forward(self, val):
+        return torch.ops.aten.cumprod(val, 1)
+
+
+@register_test_case(module_factory=lambda: CumprodInputDtypeInt32Module())
+def CumprodInputDtypeInt32Module_basic(module, tu: TestUtils):
+    module.forward(tu.randint(2, 7, 4).to(torch.int32))
+
+
+# ==============================================================================
+
+
 class AtenToDeviceModule(torch.nn.Module):
     def __init__(self):
         super().__init__()