[onnx] Fix onnx.gather lowering to use torch.aten.index_select (#2913)

Onnx's gather maps directly to `torch.aten.index_select`. We should just
use that path.

lib/Conversion/TorchOnnxToTorch/DefaultDomainGtoP.cpp

@@ -471,146 +471,66 @@ void mlir::torch::onnx_c::populateDefaultDomainGtoP(
             binder.s64IntegerAttr(axis, "axis", 0))
           return failure();
         Location loc = binder.getLoc();
+        auto ctx = binder.op->getContext();
+        auto indicesTy = cast<Torch::ValueTensorType>(indices.getType());
+        auto dataTy = cast<Torch::ValueTensorType>(data.getType());
+        if (!dataTy || !dataTy.hasSizes())
+          return failure();
+        if (axis < 0)
+          axis += dataTy.getSizes().size();
 
-        // 1. Get data shape and rank.
+        Value index = rewriter.create<Torch::ConstantIntOp>(
-        auto dataTensorType = data.getType().cast<Torch::ValueTensorType>();
+            loc, Torch::IntType::get(ctx), rewriter.getI64IntegerAttr(axis));
-        if (!dataTensorType || !dataTensorType.hasSizes()) {
-          return rewriter.notifyMatchFailure(binder.op,
-                                             "Expect non empty input data");
-        }
-        ArrayRef<int64_t> dataShape = dataTensorType.getSizes();
-        unsigned dataRank = dataShape.size();
 
-        // 2. Get indices shape and rank.
+        // Apply bounds checking on the input:
-        auto indexType = indices.getType().cast<Torch::ValueTensorType>();
+        auto intTy = rewriter.getType<Torch::IntType>();
-        if (!indexType || !indexType.hasSizes()) {
+        auto boolTy = rewriter.getType<Torch::ValueTensorType>(
-          return rewriter.notifyMatchFailure(binder.op,
+            indicesTy.getSizes(), rewriter.getI1Type());
-                                             "Expect non empty index tensor");
+        Value zero = rewriter.create<Torch::ConstantIntOp>(
-        }
+            loc, intTy, rewriter.getI64IntegerAttr(0));
-        ArrayRef<int64_t> indexShape = indexType.getSizes();
+        Value one = rewriter.create<Torch::ConstantIntOp>(
-        unsigned indexRank = indexShape.size();
+            loc, intTy, rewriter.getI64IntegerAttr(1));
+        Value lt =
+            rewriter.create<Torch::AtenLeScalarOp>(loc, boolTy, indices, zero);
+        Value dim =
+            rewriter.create<Torch::AtenSizeIntOp>(loc, intTy, data, index);
+        Value add = rewriter.create<Torch::AtenAddScalarOp>(loc, indicesTy,
+                                                            indices, dim, one);
+        indices = rewriter.create<Torch::AtenWhereSelfOp>(loc, indicesTy, lt,
+                                                          add, indices);
 
-        // 3. Compute total elements in the indices tensor, as we will collapse
+        auto intListTy = rewriter.getType<Torch::ListType>(
-        // the indices tensor to a unary tensor. Also compute index shape and
+            rewriter.getType<Torch::IntType>());
-        // data shape tensors as they will be used for creating output types.
+        auto indicesSize =
-        int64_t indexElemCount = 1;
+            rewriter.create<Torch::AtenSizeOp>(loc, intListTy, indices);
-        for (int64_t dim : indexShape) {
+
-          if (dim == Torch::kUnknownSize) {
+        // Determine the collapsed dim size:
-            indexElemCount = Torch::kUnknownSize;
+        auto indicesCt = 1;
+        for (auto sz : indicesTy.getSizes()) {
+          if (sz == Torch::kUnknownSize) {
+            indicesCt = Torch::kUnknownSize;
             break;
           }
-          indexElemCount *= dim;
+
+          indicesCt *= sz;
         }
 
-        Value constOne = rewriter.create<Torch::ConstantIntOp>(
+        auto flattenTy = rewriter.getType<Torch::ValueTensorType>(
-            loc, rewriter.getI64IntegerAttr(1));
+            SmallVector<int64_t>{indicesCt}, indicesTy.getOptionalDtype());
-        SmallVector<Value> indexShapeTensor;
+        Value rank = rewriter.create<Torch::AtenDimOp>(loc, intTy, indices);
-        Value indexElemCountVal = constOne;
+        Value end = rewriter.create<Torch::AtenSubIntOp>(loc, rank, one);
-        for (unsigned i = 0; i < indexRank; ++i) {
+        indices = rewriter.create<Torch::AtenFlattenUsingIntsOp>(
-          Value indexDimVal = rewriter.create<Torch::AtenSizeIntOp>(
+            loc, flattenTy, indices, zero, end);
-              loc, indices,
-              rewriter.create<Torch::ConstantIntOp>(
-                  loc, rewriter.getI64IntegerAttr(i)));
-          indexShapeTensor.emplace_back(indexDimVal);
-          indexElemCountVal = rewriter.create<Torch::AtenMulIntOp>(
-              loc, indexElemCountVal, indexDimVal);
-        }
 
-        SmallVector<Value> dataShapeTensor;
+        llvm::SmallVector<int64_t> gatherShape(dataTy.getSizes());
-        for (unsigned i = 0; i < dataRank; ++i) {
+        gatherShape[axis] = indicesCt;
-          dataShapeTensor.emplace_back(rewriter.create<Torch::AtenSizeIntOp>(
-              loc, data,
-              rewriter.create<Torch::ConstantIntOp>(
-                  loc, rewriter.getI64IntegerAttr(i))));
-        }
 
-        // Correct for negative axis:
+        auto gatherTy = rewriter.getType<Torch::ValueTensorType>(
-        if (axis < 0)
+            gatherShape, dataTy.getOptionalDtype());
-          axis += dataRank;
+        Value gather = rewriter.create<Torch::AtenIndexSelectOp>(
-
+            loc, gatherTy, data, index, indices);
-        // 4. We can not directly perform torch.gather as the onnx.gather op
+        rewriter.replaceOpWithNewOp<Torch::AtenUnflattenIntOp>(
-        // collects the input data at different location of output compared to
+            binder.op, resultType, gather, index, indicesSize);
-        // torch.gather op. The output of torch.gather and onnx.gather ops are
-        // indexed differently.
-        // check https://onnx.ai/onnx/operators/onnx__Gather.html for more
-        // details. So we will collapse indices tensor to a unary tensor and
-        // materialize to non-axis dimension of data tensor. For example,
-        // assuming indices is of shape (4, 5, 6), data is (8, 10, 11, 12) and
-        // axis=1. we will collapse indices into a (120,) unary tensor,
-        // materialize to non-axis dimension of data i.e. reshaping the unary
-        // indices tensor to (1, 120, 1, 1) and then perform the torch.gather
-        // operation. Now broadcast the output of gather operation to non-axis
-        // dimensions of data tensor. This would make the result of shape (8,
-        // 10, 120, 12). Post the broadcasting, expand the indices dimensions by
-        // reshaping (8, 10, 120, 12) to (8, 10, 4, 5, 6, 12) tensor, which is
-        // our expected final result.
-        SmallVector<int64_t> collapsedIndexShape(dataRank, 1);
-        collapsedIndexShape[axis] = indexElemCount;
-        Type collapsedIndexType = Torch::ValueTensorType::get(
-            indexType.getContext(), llvm::ArrayRef(collapsedIndexShape),
-            indexType.getOptionalDtype());
-
-        SmallVector<Value> collapsedIndexSize(dataRank, constOne);
-        collapsedIndexSize[axis] = indexElemCountVal;
-        auto collapsedIndexSizeList =
-            rewriter.create<Torch::PrimListConstructOp>(
-                loc,
-                rewriter.getType<Torch::ListType>(
-                    rewriter.getType<Torch::IntType>()),
-                collapsedIndexSize);
-
-        auto collapsedIndices = rewriter.create<Torch::AtenViewOp>(
-            loc, collapsedIndexType, indices, collapsedIndexSizeList);
-
-        // 5. Compute gather result type and perform gather operation.
-        Type gatherResultType = Torch::ValueTensorType::get(
-            dataTensorType.getContext(), llvm::ArrayRef(collapsedIndexShape),
-            dataTensorType.getOptionalDtype());
-        Value constAxis = rewriter.create<Torch::ConstantIntOp>(
-            binder.getLoc(), rewriter.getType<Torch::IntType>(),
-            rewriter.getIntegerAttr(rewriter.getIntegerType(64), axis));
-        Value constFalse = rewriter.create<Torch::ConstantBoolOp>(
-            binder.getLoc(), rewriter.getType<Torch::BoolType>(),
-            rewriter.getBoolAttr(false));
-        auto gatherOp = rewriter.create<Torch::AtenGatherOp>(
-            loc, gatherResultType, data, constAxis, collapsedIndices,
-            /*sparseGrad=*/constFalse);
-
-        // 6. Broadcast the gather output to non-axis dimensions of data tensor.
-        SmallVector<int64_t> dataShapeVector(dataShape);
-        dataShapeVector[axis] = indexElemCount;
-        Type expandResultType = Torch::ValueTensorType::get(
-            dataTensorType.getContext(), llvm::ArrayRef(dataShapeVector),
-            dataTensorType.getOptionalDtype());
-
-        dataShapeTensor[axis] = indexElemCountVal;
-        auto expandSizeList = rewriter.create<Torch::PrimListConstructOp>(
-            loc, Torch::ListType::get(Torch::IntType::get(data.getContext())),
-            dataShapeTensor);
-        auto expandedGather = rewriter.create<Torch::AtenExpandOp>(
-            loc, expandResultType, gatherOp, expandSizeList,
-            /*implicit=*/constFalse);
-
-        // 7. Compute the result type of reshape op which expands the collapsed
-        // indices shapes back to the original indices shapes and reshape the
-        // output produced at step 6. This will produce our expected result of
-        // onnx.gather op.
-        SmallVector<Value> resultShapeTensor;
-        for (unsigned i = 0; i < dataRank; ++i) {
-          if (i == axis) {
-            resultShapeTensor.insert(resultShapeTensor.end(),
-                                     indexShapeTensor.begin(),
-                                     indexShapeTensor.end());
-            continue;
-          }
-          resultShapeTensor.emplace_back(dataShapeTensor[i]);
-        }
-        auto resultSizeList = rewriter.create<Torch::PrimListConstructOp>(
-            loc, Torch::ListType::get(Torch::IntType::get(data.getContext())),
-            resultShapeTensor);
-
-        rewriter.replaceOpWithNewOp<Torch::AtenViewOp>(
-            binder.op, resultType, expandedGather, resultSizeList);
         return success();
       });
   patterns.onOp(

projects/pt1/e2e_testing/xfail_sets.py

@@ -2134,17 +2134,14 @@ ONNX_XFAIL_SET = {
     "AdaptiveAvgPool2dUnitOutputSizeDynamicModule_basic",
     "ElementwiseSeluModule_basic",
     "EmbeddingModule1DIndices_basic",
-    "EmbeddingModuleI32Static_basic",
     "FlipNegativeIndexModule_basic",
     "HardsigmoidModule_basic",
     "HardsigmoidRandomModule_basic",
     "IndexSelectDynamicIndexSizeModule_basic",
     "IndexSelectDynamicInputSizeModule_basic",
     "IndexSelectDynamicModulebasic",
-    "IndexSelectNegativeDimModule_basic",
-    "IndexSelectSingleIdxModule_basic",
-    "IndexSelectTwoIdxModule_basic",
     "IndexSelectWholeDimensionModule_basic",
+    "IndexSelectWholeTensorModule_basic",
     "IndexTensorStaticModule_basic",
     "IndexTensorStaticNonContiguousWithNoneModule_basic",
     "PixelShuffleModuleStaticRank4Float32_basic",

test/Conversion/TorchOnnxToTorch/simple_ops_g_to_p.mlir

@@ -39,35 +39,20 @@ func.func @test_less(%arg0: !torch.vtensor<[3,4,5],f32>, %arg1: !torch.vtensor<[
 
 // CHECK-LABEL: func.func @test_gather
 func.func @test_gather(%arg0: !torch.vtensor<[3,4,5],f32>, %arg1: !torch.vtensor<[8,10,20,40], si64>) -> !torch.vtensor<[8,10,20,40,4,5],f32> attributes {torch.onnx_meta.opset_version = 13 : si64} {
-  // CHECK-DAG: %[[INT1:.+]] = torch.constant.int 1
+  // CHECK: %[[AXIS:.+]] = torch.constant.int 0
-  // CHECK-DAG: %[[INT0:.+]] = torch.constant.int 0
+  // CHECK: %[[ZERO:.+]] = torch.constant.int 0
-  // CHECK: %[[ARG1_SIZE0:.+]] = torch.aten.size.int %arg1, %[[INT0]]
+  // CHECK: %[[ONE:.+]] = torch.constant.int 1
-  // CHECK: %[[MUL1:.+]] = torch.aten.mul.int %[[INT1]], %[[ARG1_SIZE0]]
+  // CHECK: %[[LT:.+]] = torch.aten.le.Scalar %arg1, %[[ZERO]]
-  // CHECK: %[[INT1_0:.+]] = torch.constant.int 1
+  // CHECK: %[[SZ:.+]] = torch.aten.size.int %arg0, %[[AXIS]]
-  // CHECK: %[[ARG1_SIZE1:.+]] = torch.aten.size.int %arg1, %[[INT1_0]]
+  // CHECK: %[[ADD:.+]] = torch.aten.add.Scalar %arg1, %[[SZ]], %[[ONE]]
-  // CHECK: %[[MUL2:.+]] = torch.aten.mul.int %[[MUL1]], %[[ARG1_SIZE1]]
+  // CHECK: %[[SEL:.+]] = torch.aten.where.self %[[LT]], %[[ADD]], %arg1
-  // CHECK-DAG: %[[INT2:.+]] = torch.constant.int 2
+  // CHECK: %[[SZ:.+]] = torch.aten.size %[[SEL]]
-  // CHECK: %[[ARG1_SIZE2:.+]] = torch.aten.size.int %arg1, %[[INT2]] : !torch.vtensor<[8,10,20,40],si64>, !torch.int -> !torch.int
+  // CHECK: %[[DIM:.+]] = torch.aten.dim %[[SEL]]
-  // CHECK: %[[MUL3:.+]] = torch.aten.mul.int %[[MUL2]], %[[ARG1_SIZE2]] : !torch.int, !torch.int -> !torch.int
+  // CHECK: %[[SUB:.+]] = torch.aten.sub.int %[[DIM]], %[[ONE]]
-  // CHECK-DAG: %[[INT3:.+]] = torch.constant.int 3
+  // CHECK: %[[FLAT:.+]] = torch.aten.flatten.using_ints %[[SEL]], %[[ZERO]], %[[SUB]]
-  // CHECK: %[[ARG1_SIZE3:.+]] = torch.aten.size.int %arg1, %[[INT3]] : !torch.vtensor<[8,10,20,40],si64>, !torch.int -> !torch.int
+  // CHECK: %[[ISEL:.+]] = torch.aten.index_select %arg0, %[[AXIS]], %[[FLAT]]
-  // CHECK: %[[MUL4:.+]] = torch.aten.mul.int %[[MUL3]], %[[ARG1_SIZE3]] : !torch.int, !torch.int -> !torch.int
+  // CHECK: %[[RES:.+]] = torch.aten.unflatten.int %[[ISEL]], %[[AXIS]], %[[SZ]]
-  // CHECK: %[[INT0_2:.+]] = torch.constant.int 0
+  // CHECK: return %[[RES]]
-  // CHECK: %[[ARG0_SIZE0:.+]] = torch.aten.size.int %arg0, %[[INT0_2]] : !torch.vtensor<[3,4,5],f32>, !torch.int -> !torch.int
-  // CHECK: %[[INT1_3:.+]] = torch.constant.int 1
-  // CHECK: %[[ARG0_SIZE1:.+]] = torch.aten.size.int %arg0, %[[INT1_3]] : !torch.vtensor<[3,4,5],f32>, !torch.int -> !torch.int
-  // CHECK: %[[INT2_4:.+]] = torch.constant.int 2
-  // CHECK: %[[ARG0_SIZE2:.+]] = torch.aten.size.int %arg0, %[[INT2_4]] : !torch.vtensor<[3,4,5],f32>, !torch.int -> !torch.int
-  // CHECK: %[[LIST1:.+]] = torch.prim.ListConstruct %[[MUL4]], %[[INT1]], %[[INT1]] : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
-  // CHECK: %[[VIEW1:.+]] = torch.aten.view %arg1, %[[LIST1]] : !torch.vtensor<[8,10,20,40],si64>, !torch.list<int> -> !torch.vtensor<[64000,1,1],si64>
-  // CHECK: %[[INT0_1:.+]] = torch.constant.int 0
-  // CHECK: %[[FALSE:.+]] = torch.constant.bool false
-  // CHECK: %[[GATHER:.+]] = torch.aten.gather %arg0, %[[INT0_1]], %[[VIEW1]], %[[FALSE]] : !torch.vtensor<[3,4,5],f32>, !torch.int, !torch.vtensor<[64000,1,1],si64>, !torch.bool -> !torch.vtensor<[64000,1,1],f32>
-  // CHECK: %[[LIST2:.+]] = torch.prim.ListConstruct %[[MUL4]], %[[ARG0_SIZE1]], %[[ARG0_SIZE2]] : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
-  // CHECK: %[[EXPAND:.+]] = torch.aten.expand %[[GATHER]], %[[LIST2]], %[[FALSE]] : !torch.vtensor<[64000,1,1],f32>, !torch.list<int>, !torch.bool -> !torch.vtensor<[64000,4,5],f32>
-  // CHECK: %[[LIST3:.+]] = torch.prim.ListConstruct %[[ARG1_SIZE0]], %[[ARG1_SIZE1]], %[[ARG1_SIZE2]], %[[ARG1_SIZE3]], %[[ARG0_SIZE1]], %[[ARG0_SIZE2]] : (!torch.int, !torch.int, !torch.int, !torch.int, !torch.int, !torch.int) -> !torch.list<int>
-  // CHECK: %[[RES:.+]] = torch.aten.view %[[EXPAND]], %[[LIST3]] : !torch.vtensor<[64000,4,5],f32>, !torch.list<int> -> !torch.vtensor<[8,10,20,40,4,5],f32>
-  // CHECK: return %[[RES]] : !torch.vtensor<[8,10,20,40,4,5],f32>
   %0 = torch.operator "onnx.Gather"(%arg0, %arg1) {torch.onnx.axis = 0 : si64} : (!torch.vtensor<[3,4,5],f32>, !torch.vtensor<[8,10,20,40], si64>) -> !torch.vtensor<[8,10,20,40,4,5],f32>
   return %0 : !torch.vtensor<[8,10,20,40,4,5],f32>
 }