[Torch] Support Aten_CastLongOp. (#3160)

By canonicalize Aten_CastLongOp into AtenToDtypeOp

include/torch-mlir/Dialect/Torch/IR/GeneratedTorchOps.td

@@ -11047,6 +11047,31 @@ def Torch_Aten_CastFloatOp : Torch_Op<"aten._cast_Float", [
   let hasCanonicalizer = 1;
 }
 
+def Torch_Aten_CastLongOp : Torch_Op<"aten._cast_Long", [
+    AllowsTypeRefinement,
+    HasValueSemantics,
+    ReadOnly
+  ]> {
+  let summary = "Generated op for `aten::_cast_Long : (Tensor, bool) -> (Tensor)`";
+  let arguments = (ins
+    AnyTorchTensorType:$self,
+    Torch_BoolType:$non_blocking
+  );
+  let results = (outs
+    AnyTorchOptionalTensorType:$result
+  );
+  let hasCustomAssemblyFormat = 1;
+  let extraClassDefinition = [{
+    ParseResult Aten_CastLongOp::parse(OpAsmParser &parser, OperationState &result) {
+      return parseDefaultTorchOp(parser, result, 2, 1);
+    }
+    void Aten_CastLongOp::print(OpAsmPrinter &printer) {
+      printDefaultTorchOp(printer, *this, 2, 1);
+    }
+  }];
+  let hasCanonicalizer = 1;
+}
+
 def Torch_AtenTypeAsOp : Torch_Op<"aten.type_as", [
     AllowsTypeRefinement,
     HasValueSemantics,

lib/Dialect/Torch/IR/TorchOps.cpp

@@ -959,6 +959,27 @@ void Aten_CastFloatOp::getCanonicalizationPatterns(RewritePatternSet &patterns,
   });
 }
 
+//===----------------------------------------------------------------------===//
+// Aten_CastLongOp
+//===----------------------------------------------------------------------===//
+
+void Aten_CastLongOp::getCanonicalizationPatterns(RewritePatternSet &patterns,
+                                                  MLIRContext *context) {
+  // `aten.cast_long` -> `aten.to.dtype`
+  patterns.add(+[](Aten_CastLongOp op, PatternRewriter &rewriter) {
+    auto self = op.getSelf();
+    auto loc = op.getLoc();
+    Value constNone = rewriter.create<ConstantNoneOp>(loc);
+    Value longType = rewriter.create<ConstantIntOp>(
+        loc, (int)torch_upstream::ScalarType::Long);
+    Value constFalse = rewriter.create<ConstantBoolOp>(loc, false);
+    rewriter.replaceOpWithNewOp<AtenToDtypeOp>(op, op.getType(), self, longType,
+                                               op.getNonBlocking(), constFalse,
+                                               constNone);
+    return success();
+  });
+}
+
 //===----------------------------------------------------------------------===//
 // AtenViewOp
 //===----------------------------------------------------------------------===//

lib/Dialect/Torch/Transforms/AbstractInterpLibrary.cpp

@@ -6790,6 +6790,10 @@ StringRef mlir::torch::Torch::getAbstractInterpLibrary() {
 "    %0 = call @__torch__.torch.jit._shape_functions.unary(%arg0) : (!torch.list<int>) -> !torch.list<int>\n"
 "    return %0 : !torch.list<int>\n"
 "  }\n"
+"  func.func @\"__torch_mlir_shape_fn.aten._cast_Long\"(%arg0: !torch.list<int>, %arg1: !torch.bool) -> !torch.list<int> {\n"
+"    %0 = call @__torch__.torch.jit._shape_functions.unary(%arg0) : (!torch.list<int>) -> !torch.list<int>\n"
+"    return %0 : !torch.list<int>\n"
+"  }\n"
 "  func.func @\"__torch_mlir_shape_fn.aten.type_as\"(%arg0: !torch.list<int>, %arg1: !torch.list<int>) -> !torch.list<int> {\n"
 "    %0 = call @__torch__.torch.jit._shape_functions.unary(%arg0) : (!torch.list<int>) -> !torch.list<int>\n"
 "    return %0 : !torch.list<int>\n"
@@ -12793,6 +12797,10 @@ StringRef mlir::torch::Torch::getAbstractInterpLibrary() {
 "    %int6 = torch.constant.int 6\n"
 "    return %int6 : !torch.int\n"
 "  }\n"
+"  func.func @\"__torch_mlir_dtype_fn.aten._cast_Long\"(%arg0: !torch.tuple<int, int>, %arg1: !torch.bool) -> !torch.int {\n"
+"    %int4 = torch.constant.int 4\n"
+"    return %int4 : !torch.int\n"
+"  }\n"
 "  func.func @\"__torch_mlir_dtype_fn.aten.type_as\"(%arg0: !torch.tuple<int, int>, %arg1: !torch.tuple<int, int>) -> !torch.int {\n"
 "    %0:2 = torch.prim.TupleUnpack %arg1 : !torch.tuple<int, int> -> !torch.int, !torch.int\n"
 "    return %0#1 : !torch.int\n"

projects/pt1/e2e_testing/xfail_sets.py

@@ -577,6 +577,7 @@ STABLEHLO_PASS_SET = {
     "AtenSubFloatModule_basic",
     "AtenToDeviceModule_basic",
     "Aten_CastFloatModule_basic",
+    "Aten_CastLongModule_basic",
     "AvgPool1dStaticModule_basic",
     "AvgPool2dStaticModule_basic",
     "BaddbmmBroadcast1DInputModule_basic",

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

@@ -394,6 +394,9 @@ def aten〇to〇other〡shape(self: List[int], other: List[int], non_blocking: b
 def aten〇_cast_Float〡shape(self: List[int], non_blocking: bool = False) -> List[int]:
     return upstream_shape_functions.unary(self)
 
+def aten〇_cast_Long〡shape(self: List[int], non_blocking: bool = False) -> List[int]:
+    return upstream_shape_functions.unary(self)
+
 def aten〇type_as〡shape(self: List[int], other: List[int]) -> List[int]:
     return upstream_shape_functions.unary(self)
 
@@ -4366,6 +4369,9 @@ def aten〇to〇other〡dtype(self_rank_dtype: Tuple[int, int], other_rank_dtype
 def aten〇_cast_Float〡dtype(self_rank_dtype: Tuple[int, int], non_blocking: bool = False) -> int:
     return torch.float32
 
+def aten〇_cast_Long〡dtype(self_rank_dtype: Tuple[int, int], non_blocking: bool = False) -> int:
+    return torch.int64
+
 @check_dtype_function(_check_two_tensor_op())
 def aten〇type_as〡dtype(self_rank_dtype: Tuple[int, int], other_rank_dtype: Tuple[int, int]) -> int:
     other_rank, other_dtype = other_rank_dtype

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

@@ -673,6 +673,7 @@ def emit_ops(emitter_td: TextEmitter, registry: Registry):
     emit("aten::to.prim_Device : (Tensor, Device?, int?, bool, bool) -> (Tensor)")
     emit("aten::to.device : (Tensor, Device, int, bool, bool, int?) -> (Tensor)")
     emit("aten::_cast_Float : (Tensor, bool) -> (Tensor)", has_canonicalizer=True)
+    emit("aten::_cast_Long : (Tensor, bool) -> (Tensor)", has_canonicalizer=True)
     emit("aten::type_as : (Tensor, Tensor) -> (Tensor)")
     emit("aten::view : (Tensor, int[]) -> (Tensor)", has_folder=True)
     emit("aten::_unsafe_view : (Tensor, int[]) -> (Tensor)")

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

@@ -4257,7 +4257,25 @@ class Aten_CastFloatModule(torch.nn.Module):
 def Aten_CastFloatModule_basic(module, tu: TestUtils):
     module.forward(tu.randint(2, 4))
 
+
+class Aten_CastLongModule(torch.nn.Module):
+    def __init__(self):
+        super().__init__()
+
+    @export
+    @annotate_args([
+        None,
+        ([2, 4], torch.float32, True),
+    ])
+
+    def forward(self, val):
+        return torch.ops.aten._cast_Long(val)
     
+@register_test_case(module_factory=lambda: Aten_CastLongModule())
+def Aten_CastLongModule_basic(module, tu: TestUtils):
+    module.forward(tu.rand(2, 4))
+
+
 # ==============================================================================
 
 class UpSampleNearest2dBackward(torch.nn.Module):