[ExecuTorch] Arm Ethos: Do not depend on torch.testing._internal

This can cuase issues with `disable_global_flags` and internal state of the library, this is something which is set when importing this

Differential Revision: [D70402061](https://our.internmc.facebook.com/intern/diff/D70402061/)

ghstack-source-id: 269356031
Pull Request resolved: pytorch/executorch#8893

Author: digantdesai

backends/arm/test/passes/test_rescale_pass.py

@@ -13,7 +13,6 @@
 from executorch.backends.arm.test import common, conftest
 from executorch.backends.arm.test.tester.arm_tester import ArmTester
 from parameterized import parameterized
-from torch.testing._internal import optests
 
 
 def test_rescale_op():
@@ -64,7 +63,7 @@ def test_nonzero_zp_for_int32():
         ),
     ]
     for sample_input in sample_inputs:
-        with pytest.raises(optests.generate_tests.OpCheckError):
+        with pytest.raises(Exception, match="opcheck"):
             torch.library.opcheck(torch.ops.tosa._rescale, sample_input)
 
 
@@ -87,7 +86,7 @@ def test_zp_outside_range():
         ),
     ]
     for sample_input in sample_inputs:
-        with pytest.raises(optests.generate_tests.OpCheckError):
+        with pytest.raises(Exception, match="opcheck"):
             torch.library.opcheck(torch.ops.tosa._rescale, sample_input)
 
 

backends/arm/test/runner_utils.py

@@ -34,12 +34,33 @@
 from torch.fx.node import Node
 
 from torch.overrides import TorchFunctionMode
-from torch.testing._internal.common_utils import torch_to_numpy_dtype_dict
 from tosa import TosaGraph
 
 logger = logging.getLogger(__name__)
 logger.setLevel(logging.CRITICAL)
 
+# Copied from PyTorch.
+# From torch/testing/_internal/common_utils.py:torch_to_numpy_dtype_dict
+# To avoid a dependency on _internal stuff.
+_torch_to_numpy_dtype_dict = {
+    torch.bool: np.bool_,
+    torch.uint8: np.uint8,
+    torch.uint16: np.uint16,
+    torch.uint32: np.uint32,
+    torch.uint64: np.uint64,
+    torch.int8: np.int8,
+    torch.int16: np.int16,
+    torch.int32: np.int32,
+    torch.int64: np.int64,
+    torch.float16: np.float16,
+    torch.float32: np.float32,
+    torch.float64: np.float64,
+    torch.bfloat16: np.float32,
+    torch.complex32: np.complex64,
+    torch.complex64: np.complex64,
+    torch.complex128: np.complex128,
+}
+
 
 class QuantizationParams:
     __slots__ = ["node_name", "zp", "scale", "qmin", "qmax", "dtype"]
@@ -335,7 +356,7 @@ def run_corstone(
         output_dtype = node.meta["val"].dtype
         tosa_ref_output = np.fromfile(
             os.path.join(intermediate_path, f"out-{i}.bin"),
-            torch_to_numpy_dtype_dict[output_dtype],
+            _torch_to_numpy_dtype_dict[output_dtype],
         )
 
         output_np.append(torch.from_numpy(tosa_ref_output).reshape(output_shape))
@@ -349,7 +370,7 @@ def prep_data_for_save(
 ):
     if isinstance(data, torch.Tensor):
         data_np = np.array(data.detach(), order="C").astype(
-            torch_to_numpy_dtype_dict[data.dtype]
+            _torch_to_numpy_dtype_dict[data.dtype]
         )
     else:
         data_np = np.array(data)