Merge branch 'main' into exir-flatbuffer-serialize-fastpath_v2

Author: zingo

backends/arm/test/ops/test_amax.py

@@ -143,7 +143,7 @@ def test_amax_u55_INT(test_data: Amax.input_t):
     pipeline = EthosU55PipelineINT[Amax.input_t](
         Amax(dim, keep_dims),
         data,
-        Amax.aten_op,
+        amax_aten_op,
     )
     pipeline.run()
 

backends/nxp/backend/edge_program_converter.py

@@ -40,6 +40,7 @@
     exir_ops.edge.aten.mean.dim: MeanDimConverter,  # noqa F405
     exir_ops.edge.aten.mm.default: MMConverter,  # noqa F405
     exir_ops.edge.aten.mul.Tensor: MulTensorConverter,  # noqa F405
+    exir_ops.edge.aten.neg.default: NegConverter,  # noqa F405
     exir_ops.edge.aten.permute_copy.default: PermuteCopyConverter,  # noqa F405
     exir_ops.edge.aten.relu.default: ReLUConverter,  # noqa F405
     exir_ops.edge.aten.sigmoid.default: SigmoidConverter,  # noqa F405

backends/nxp/backend/ir/converter/node_converters/ops_converters/__init__.py

@@ -40,6 +40,9 @@
 from executorch.backends.nxp.backend.ir.converter.node_converters.ops_converters.mul_tensor_converter import (
     MulTensorConverter,
 )
+from executorch.backends.nxp.backend.ir.converter.node_converters.ops_converters.neg_converter import (
+    NegConverter,
+)
 from executorch.backends.nxp.backend.ir.converter.node_converters.ops_converters.permute_copy_converter import (
     PermuteCopyConverter,
 )
@@ -93,6 +96,7 @@
     "MeanDimConverter",
     "MMConverter",
     "MulTensorConverter",
+    "NegConverter",
     "PermuteCopyConverter",
     "QDQPerChannelDequantizeConverter",
     "QDQPerTensorDequantizeConverter",

backends/nxp/backend/ir/converter/node_converters/ops_converters/neg_converter.py

@@ -0,0 +1,70 @@
+# Copyright 2026 NXP
+#
+# This source code is licensed under the BSD-style license found in the
+# LICENSE file in the root directory of this source tree.
+
+import numpy as np
+
+from executorch.backends.nxp.backend import edge_helper
+from executorch.backends.nxp.backend.ir.converter.node_converter import (
+    CustomDelegationOptions,
+    NodeConverter,
+)
+from executorch.backends.nxp.backend.ir.tflite_generator.builtin_options import (
+    sub_options,
+)
+from executorch.backends.nxp.backend.ir.tflite_generator.tflite_model import (
+    Quantization,
+    Scale,
+    ZeroPoint,
+)
+from torch.fx import Node
+from torch.nn import Parameter
+
+
+class NegConverter(NodeConverter):
+
+    @staticmethod
+    def _is_supported_in_IR(
+        node: Node,
+        parameters_mapping: dict[str, Parameter],
+        custom_delegation_options: CustomDelegationOptions,
+    ) -> bool:
+        if len(node.args) != 1:
+            # Should never happen
+            return False
+
+        # The conversion code below expects a per tensor quantized operator.
+        scale, zp = edge_helper.get_quantization_parameters_for(node.args[0])
+        match scale, zp:
+            case [float(), int()]:
+                pass  # Atomic quantization parameters -> per tensor quantization.
+            case _:
+                return False  # Everything else is unexpected.
+
+        return True
+
+    def convert(self, node: Node):
+        """Convert 'aten.neg.default' operator to NeutronIR  0 - 'Sub'.
+
+        The ExecuTorch schema is 'aten::neg(Tensor self) -> Tensor'
+        """
+        self.assert_convertible(node)
+
+        t_op = self._create_tflite_op_with_io_tensors(node)
+
+        x = t_op.tmp_inputs[0]
+
+        # Extract the zero_point, to use as the first input of the `Sub`.
+        scale = x.quantization.scale.vector
+        zp = x.quantization.zero_point.vector
+        zero_tensor = self.builder.create_tensor_for_data(np.array(zp, "int8"), "zero")
+        zero_tensor.quantization = Quantization(
+            scale=Scale(list(scale)), zero_point=ZeroPoint(list(zp))
+        )
+
+        # Assign the NeutronIR operator its builtin options and inputs.
+        t_op.builtin_options = sub_options.Sub()
+        t_op.tmp_inputs = [zero_tensor, x]
+
+        self.builder.append_operators([t_op])

backends/nxp/neutron_partitioner.py

@@ -213,6 +213,7 @@ def tag_qdq_clusters(self, nodes: list[torch.fx.Node]):
     exir_ops.edge.aten.mean.dim: MeanDimConverter,  # noqa F405
     exir_ops.edge.aten.mm.default: MMConverter,  # noqa F405
     exir_ops.edge.aten.mul.Tensor: MulTensorConverter,  # noqa F405
+    exir_ops.edge.aten.neg.default: NegConverter,  # noqa F405
     exir_ops.edge.aten.permute_copy.default: PermuteCopyConverter,  # noqa F405
     exir_ops.edge.aten.relu.default: ReLUConverter,  # noqa F405
     exir_ops.edge.aten.sigmoid.default: SigmoidConverter,  # noqa F405

backends/nxp/quantizer/neutron_quantizer.py

@@ -32,6 +32,7 @@
     MeanDimPattern,
     MmPattern,
     MulTensorPattern,
+    NegPattern,
     NodeArgsIdx,
     PadPattern,
     PermutePattern,
@@ -263,6 +264,7 @@ def __init__(self, neutron_target_spec: NeutronTargetSpec, is_qat: bool = False)
                 OpQuantizer(MeanDimPattern(is_qat=is_qat), static_qconfig),
                 OpQuantizer(MmPattern(self, is_qat=is_qat), static_qconfig),
                 OpQuantizer(MulTensorPattern(is_qat=is_qat), static_qconfig),
+                OpQuantizer(NegPattern(is_qat=is_qat), static_qconfig),
                 OpQuantizer(PadPattern(is_qat=is_qat), static_qconfig),
                 OpQuantizer(PermutePattern(is_qat=is_qat), static_qconfig),
                 OpQuantizer(ReluPattern(is_qat=is_qat), static_qconfig),

backends/nxp/quantizer/patterns.py

@@ -800,6 +800,15 @@ def get_anchors(
         )
 
 
+class NegPattern(SharedSpecPattern):
+    """
+    Quantizer for the `aten.neg.default` operator.
+    """
+
+    def partition_types(self):
+        return [torch.ops.aten.neg.default]
+
+
 class PadPattern(SharedSpecPattern):
     """
     Quantizer for Pad operator.

backends/nxp/tests/ir/converter/node_converter/test_neg_converter.py

@@ -0,0 +1,123 @@
+# Copyright 2026 NXP
+#
+# This source code is licensed under the BSD-style license found in the
+# LICENSE file in the root directory of this source tree.
+
+import numpy as np
+import pytest
+import torch
+
+from executorch.backends.nxp.backend.edge_program_converter import (
+    EdgeProgramToIRConverter,
+)
+from executorch.backends.nxp.tests.executorch_pipeline import to_quantized_edge_program
+from executorch.backends.nxp.tests.executors import (
+    convert_run_compare,
+    graph_contains_any_of_ops,
+    ToChannelFirstPreprocess,
+    ToChannelLastPreprocess,
+)
+from executorch.exir.dialects._ops import ops as exir_ops
+
+
+@pytest.fixture(autouse=True)
+def reseed_model_per_test_run():
+    torch.manual_seed(42)
+    np.random.seed(23)
+
+
+# noinspection PyProtectedMember
+ExecutorchDelegateCall = torch._higher_order_ops.executorch_call_delegate
+Neg = exir_ops.edge.aten.neg.default
+
+
+class NegModule(torch.nn.Module):
+
+    def __init__(self):
+        super().__init__()
+
+    # noinspection PyMethodMayBeStatic
+    def forward(self, x):
+        return -x
+
+
+class ConvNegModule(torch.nn.Module):
+
+    def __init__(self):
+        super().__init__()
+        self.conv = torch.nn.Conv2d(3, 3, 1)
+
+    # noinspection PyMethodMayBeStatic
+    def forward(self, x):
+        x = self.conv(x)
+        return -x
+
+
+@pytest.mark.parametrize(
+    "input_shape",
+    [
+        pytest.param((8,), id="1D"),
+        pytest.param((4, 2), id="2D"),
+        pytest.param((1, 2, 3), id="3D"),
+        pytest.param((1, 2, 3, 4), id="4D"),
+    ],
+)
+def test_convert_neg(mocker, input_shape):
+    model = NegModule()
+
+    converter_spy = mocker.spy(EdgeProgramToIRConverter, "convert_program")
+    delegated_ep = to_quantized_edge_program(model, input_shape).exported_program()
+
+    # Make sure the `neg` was delegated.
+    assert graph_contains_any_of_ops(delegated_ep.graph, [ExecutorchDelegateCall])
+    assert not graph_contains_any_of_ops(delegated_ep.graph, [Neg])
+
+    # Verify correct behavior of the converted NeutronIR model.
+    intermediate_ep = converter_spy.call_args.args[1]
+    neutron_ir_model, _ = converter_spy.spy_return
+
+    input_data = (
+        np.random.random(input_shape).astype(np.float32) * 256.0 - 128.0
+    ).astype(np.int8)
+
+    # Make sure the tested program contains the `neg`.
+    assert graph_contains_any_of_ops(intermediate_ep.graph, [Neg])
+
+    convert_run_compare(
+        intermediate_ep,
+        tfl_model=neutron_ir_model,
+        input_data=input_data,
+    )
+
+
+def test_convert_neg__channels_last(mocker):
+    model = ConvNegModule()
+    input_shape = (1, 3, 4, 5)
+
+    converter_spy = mocker.spy(EdgeProgramToIRConverter, "convert_program")
+    delegated_ep = to_quantized_edge_program(
+        model, input_shape, use_neutron_for_format_conversion=False
+    ).exported_program()
+
+    # Make sure the `neg` was delegated.
+    assert graph_contains_any_of_ops(delegated_ep.graph, [ExecutorchDelegateCall])
+    assert not graph_contains_any_of_ops(delegated_ep.graph, [Neg])
+
+    # Verify correct behavior of the converted NeutronIR model.
+    intermediate_ep = converter_spy.call_args.args[1]
+    neutron_ir_model, _ = converter_spy.spy_return
+
+    input_data = (
+        np.random.random(input_shape).astype(np.float32) * 256.0 - 128.0
+    ).astype(np.int8)
+
+    # Make sure the tested program contains the `neg`.
+    assert graph_contains_any_of_ops(intermediate_ep.graph, [Neg])
+
+    convert_run_compare(
+        intermediate_ep,
+        tfl_model=neutron_ir_model,
+        input_data=input_data,
+        tflite_input_preprocess=ToChannelLastPreprocess(),
+        tflite_output_preprocess=ToChannelFirstPreprocess(),
+    )

docs/source/backends/nxp/op-support.csv

@@ -14,6 +14,7 @@ aten.max_pool2d_with_indices.default,int8,static int8,"dilation=1, ceil_mode=Fal
 aten.mean.dim,int8,static int8,"4D tensor only, dims = [-1,-2] or [-2,-1]"
 aten.mul.Tensor, int8, static int8, "tensor-size % 8 = 0"
 aten.mm.default,int8,static int8,"2D tensor only"
+aten.neg.default,int8,static int8,
 aten.relu.default,int8,static int8,
 aten.sigmoid.default,int8,static int8,
 aten.slice_copy.Tensor, int8, static int8

kernels/aten/functions.yaml

@@ -4,6 +4,8 @@
 
 - op: _cdist_forward.out
 
+- op: _conj_physical.out
+
 - op: _fake_quantize_per_tensor_affine_cachemask_tensor_qparams.out
 
 - op: _fft_c2r.out