result_type to raise unless at least one argument is usm_ndarray

dpctl/tensor/_search_functions.py

@@ -35,6 +35,7 @@
     _all_data_types,
     _can_cast,
     _is_weak_dtype,
+    _result_type_fn_impl,
     _strong_dtype_num_kind,
     _to_device_supported_dtype,
     _weak_type_num_kind,
@@ -95,7 +96,13 @@ def _resolve_two_weak_types(o1_dtype, o2_dtype, dev):
 
 
 def _where_result_type(dt1, dt2, dev):
-    res_dtype = dpt.result_type(dt1, dt2)
+    res_dtype = _result_type_fn_impl(
+        (
+            dt1,
+            dt2,
+        ),
+        sycl_device=dev,
+    )
     fp16 = dev.has_aspect_fp16
     fp64 = dev.has_aspect_fp64
 

dpctl/tensor/_type_utils.py

@@ -21,6 +21,13 @@
 import dpctl.tensor._tensor_impl as ti
 
 
+def _supported_dtype(dtypes):
+    for dtype in dtypes:
+        if dtype.char not in "?bBhHiIlLqQefdFD":
+            raise ValueError(f"Dpctl doesn't support dtype {dtype}.")
+    return True
+
+
 def _all_data_types(_fp16, _fp64):
     _non_fp_types = [
         dpt.bool,
@@ -708,26 +715,11 @@ def can_cast(from_, to, /, *, casting="safe") -> bool:
         return _can_cast(dtype_from, dtype_to, True, True, casting=casting)
 
 
-def result_type(*arrays_and_dtypes):
-    """
-    result_type(*arrays_and_dtypes)
-
-    Returns the dtype that results from applying the Type Promotion Rules to \
-        the arguments.
-
-    Args:
-        arrays_and_dtypes (Union[usm_ndarray, dtype]):
-            An arbitrary length sequence of usm_ndarray objects or dtypes.
-
-    Returns:
-        dtype:
-            The dtype resulting from an operation involving the
-            input arrays and dtypes.
-    """
+def _result_type_fn_impl(arrays_and_dtypes_tuple, sycl_device=None):
     dtypes = []
-    devices = []
+    devices = [] if sycl_device is None else [sycl_device]
     weak_dtypes = []
-    for arg_i in arrays_and_dtypes:
+    for arg_i in arrays_and_dtypes_tuple:
         if isinstance(arg_i, dpt.usm_ndarray):
             devices.append(arg_i.sycl_device)
             dtypes.append(arg_i.dtype)
@@ -766,6 +758,10 @@ def result_type(*arrays_and_dtypes):
                 has_fp64 = d.has_aspect_fp64
                 target_dev = d
                 inspected = True
+    else:
+        raise ValueError(
+            "At least only argument must have type `dpctl.tensor.usm_ndarray`"
+        )
 
     if not (has_fp16 and has_fp64):
         for dt in dtypes:
@@ -788,6 +784,25 @@ def result_type(*arrays_and_dtypes):
     return res_dt
 
 
+def result_type(*arrays_and_dtypes):
+    """
+    result_type(*arrays_and_dtypes)
+
+    Returns the dtype that results from applying the Type Promotion Rules to \
+        the arguments.
+
+    Args:
+        arrays_and_dtypes (Union[usm_ndarray, dtype]):
+            An arbitrary length sequence of usm_ndarray objects or dtypes.
+
+    Returns:
+        dtype:
+            The dtype resulting from an operation involving the
+            input arrays and dtypes.
+    """
+    return _result_type_fn_impl(arrays_and_dtypes)
+
+
 def iinfo(dtype, /):
     """iinfo(dtype)
 
@@ -855,13 +870,6 @@ def finfo(dtype, /):
     return finfo_object(dtype)
 
 
-def _supported_dtype(dtypes):
-    for dtype in dtypes:
-        if dtype.char not in "?bBhHiIlLqQefdFD":
-            raise ValueError(f"Dpctl doesn't support dtype {dtype}.")
-    return True
-
-
 def isdtype(dtype, kind):
     """isdtype(dtype, kind)
 

dpctl/tests/helper/_helper.py

@@ -55,11 +55,14 @@ def skip_if_dtype_not_supported(dt, q_or_dev):
     import dpctl.tensor as dpt
 
     dt = dpt.dtype(dt)
-    if type(q_or_dev) is dpctl.SyclQueue:
+    if isinstance(q_or_dev, dpctl.SyclQueue):
         dev = q_or_dev.sycl_device
-    elif type(q_or_dev) is dpctl.SyclDevice:
-        dev = q_or_dev
     else:
+        dev = q_or_dev
+
+    if not hasattr(dev, "has_aspect_fp16") or not hasattr(
+        dev, "has_aspect_fp64"
+    ):
         raise TypeError(
             "Expected dpctl.SyclQueue or dpctl.SyclDevice, "
             f"got {type(q_or_dev)}"

dpctl/tests/test_usm_ndarray_manipulation.py

@@ -998,11 +998,6 @@ def test_result_type():
 
     assert dpt.result_type(*X) == np.result_type(*X_np)
 
-    X = [dpt.int32, "int64", 2]
-    X_np = [np.int32, "int64", 2]
-
-    assert dpt.result_type(*X) == np.result_type(*X_np)
-
     X = [usm_ar, dpt.int32, "int64", 2.0]
     X_np = [np_ar, np.int32, "int64", 2.0]
 
@@ -1013,6 +1008,10 @@ def test_result_type():
 
     assert dpt.result_type(*X).kind == np.result_type(*X_np).kind
 
+    X = [dpt.int32, "int64", 2]
+    with pytest.raises(ValueError):
+        dpt.result_type(*X)
+
 
 def test_swapaxes_1d():
     get_queue_or_skip()

dpctl/tests/test_usm_ndarray_search_functions.py

@@ -48,6 +48,7 @@
 
 class mock_device:
     def __init__(self, fp16, fp64):
+        self.name = "Mock device"
         self.has_aspect_fp16 = fp16
         self.has_aspect_fp64 = fp64
 
@@ -101,14 +102,17 @@ def test_where_result_types(dt1, dt2, fp16, fp64):
     dev = mock_device(fp16, fp64)
 
     dt1 = dpt.dtype(dt1)
+    skip_if_dtype_not_supported(dt1, dev)
     dt2 = dpt.dtype(dt2)
+    skip_if_dtype_not_supported(dt2, dev)
+
     res_t = _where_result_type(dt1, dt2, dev)
 
     if fp16 and fp64:
-        assert res_t == dpt.result_type(dt1, dt2)
+        assert res_t == np.result_type(dt1, dt2)
     else:
         if res_t:
-            assert res_t.kind == dpt.result_type(dt1, dt2).kind
+            assert res_t.kind == np.result_type(dt1, dt2).kind
         else:
             # some illegal cases are covered above, but
             # this guarantees that _where_result_type