Add lowering of aten.floor op

e2e_testing/torchscript/elementwise.py

@@ -378,3 +378,19 @@ class ElementwiseSqrtModule(torch.nn.Module):
 @register_test_case(module_factory=lambda: ElementwiseSqrtModule())
 def ElementwiseSqrtModule_basic(module, tu: TestUtils):
     module.forward(tu.rand(3, 4))
+
+class ElementwiseFloorModule(torch.nn.Module):
+    def __init__(self):
+        super().__init__()
+    @export
+    @annotate_args([
+        None,
+        ([-1, -1], torch.float32, True),
+    ])
+
+    def forward(self, a):
+        return torch.floor(a)
+
+@register_test_case(module_factory=lambda: ElementwiseFloorModule())
+def ElementwiseFloorModule_basic(module, tu: TestUtils):
+    module.forward(tu.rand(3, 4))

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

@@ -240,6 +240,34 @@ def Torch_AtenNeg_Op : Torch_Op<"aten.neg_", [
   let assemblyFormat = "$self attr-dict `:` type($self) `->` type($result)";
 }
 
+def Torch_AtenFloorOp : Torch_Op<"aten.floor", [
+    AllowsTypeRefinement,
+    HasValueSemantics
+  ]> {
+  let summary = "Generated op for `aten::floor : (Tensor) -> (Tensor)`";
+  let arguments = (ins
+    AnyTorchTensorType:$self
+  );
+  let results = (outs
+    AnyTorchTensorType:$result
+  );
+  let assemblyFormat = "$self attr-dict `:` type($self) `->` type($result)";
+}
+
+def Torch_AtenFloor_Op : Torch_Op<"aten.floor_", [
+    IsTrailingUnderscoreInplaceVariant,
+    AllowsTypeRefinement
+  ]> {
+  let summary = "Generated op for `aten::floor_ : (Tensor) -> (Tensor)`";
+  let arguments = (ins
+    AnyTorchTensorType:$self
+  );
+  let results = (outs
+    AnyTorchTensorType:$result
+  );
+  let assemblyFormat = "$self attr-dict `:` type($self) `->` type($result)";
+}
+
 def Torch_AtenBitwiseNotOp : Torch_Op<"aten.bitwise_not", [
     AllowsTypeRefinement,
     HasValueSemantics

lib/Conversion/TorchToLinalg/TorchToLinalg.cpp

@@ -1278,6 +1278,8 @@ static Value createLinalgPayloadCalculationForElementwiseOp(
     return b.create<math::TanhOp>(loc, payloadArgs[0]);
   if (isa<AtenExpOp>(op))
     return b.create<math::ExpOp>(loc, payloadArgs[0]);
+  if (isa<AtenFloorOp>(op))
+    return b.create<math::FloorOp>(loc, payloadArgs[0]);
   if (isa<AtenLogOp>(op))
     return b.create<math::LogOp>(loc, payloadArgs[0]);
   if (isa<AtenSqrtOp>(op))
@@ -1666,7 +1668,7 @@ struct ConvertElementwiseOp : ConversionPattern {
              AtenMulTensorOp, AtenDivTensorOp, AtenSubTensorOp,
              AtenLerpTensorOp, AtenSigmoidOp, AtenExpOp, AtenMinimumOp,
              AtenMaximumOp, AtenClampOp, AtenRsubScalarOp, AtenLogOp,
-             AtenSqrtOp>(op))
+             AtenSqrtOp, AtenFloorOp>(op))
       return rewriter.notifyMatchFailure(op, "not a supported elementwise op");
 
     if (failed(verifyLinalgCompatibleTypes(op, rewriter)))
@@ -2803,7 +2805,7 @@ public:
                         AtenMulTensorOp, AtenDivTensorOp, AtenSubTensorOp,
                         AtenLerpTensorOp, AtenSigmoidOp, AtenMinimumOp,
                         AtenMaximumOp, AtenClampOp, AtenRsubScalarOp, AtenLogOp,
-                        AtenSqrtOp>();
+                        AtenSqrtOp, AtenFloorOp>();
     patterns.add<ConvertElementwiseOp>(typeConverter, context);
     target.addIllegalOp<AtenUnsqueezeOp>();
     patterns.add<ConvertAtenUnsqueezeOp>(typeConverter, context);

lib/Dialect/Torch/Transforms/RefineTypes.cpp

@@ -231,7 +231,7 @@ public:
             AtenFill_ScalarOp, AtenDetachOp, AtenMaskedFill_ScalarOp,
             AtenCopy_Op, AtenIndexPut_Op, AtenCopy_Op, AtenCumsumOp,
             AtenLayerNormOp, AtenClampOp, AtenRsubScalarOp, AtenLogOp,
-            AtenSqrtOp>(op)) {
+            AtenSqrtOp, AtenFloorOp>(op)) {
       return getLatticeElement(op->getResult(0)).join(*operands[0]);
     }
 

python/torch_mlir/dialects/torch/importer/jit_ir/build_tools/torch_ods_gen.py

@@ -445,6 +445,7 @@ def emit_aten_ops(torch_ir_dir: str, registry: Registry):
                 "aten::exp : (Tensor) -> (Tensor)",
                 "aten::cos : (Tensor) -> (Tensor)",
                 "aten::neg : (Tensor) -> (Tensor)",
+                "aten::floor : (Tensor) -> (Tensor)",
                 "aten::bitwise_not : (Tensor) -> (Tensor)",
                 "aten::add.Tensor : (Tensor, Tensor, Scalar) -> (Tensor)",
                 "aten::sub.Tensor : (Tensor, Tensor, Scalar) -> (Tensor)",