Add documentation for constants and DLDeviceType enums

Author: ndgrigorian

docs/doc_sources/api_reference/dpctl/tensor.constants.rst

@@ -0,0 +1,35 @@
+.. _dpctl_tensor_constants:
+
+Constants
+========================
+
+The following constants are defined in :py:mod:`dpctl.tensor`:
+
+.. currentmodule:: dpctl.tensor
+
+.. autodata:: DLDeviceType
+
+.. data:: e
+
+   ``float``:
+        IEEE 754 floating-point representation of Euler's constant.
+
+.. data:: inf
+
+   ``float``:
+        IEEE 754 floating-point representation of (positive) infinity.
+
+.. data:: nan
+
+   ``float``:
+        IEEE 754 floating-point representation of Not a Number (NaN).
+
+.. data:: newaxis
+
+   ``NoneType``:
+        Alias for ``None`` which is useful for indexing.
+
+.. data:: pi
+
+   ``float``:
+        IEEE 754 floating-point representation of the mathematical constant π.

docs/doc_sources/api_reference/dpctl/tensor.rst

@@ -29,6 +29,7 @@ This module contains:
 * :ref:`sorting functions <dpctl_tensor_sorting_functions>`
 * :ref:`statistical functions <dpctl_tensor_statistical_functions>`
 * :ref:`utility functions <dpctl_tensor_utility_functions>`
+* :ref:`constants <dpctl_tensor_constants>`
 
 
 .. toctree::
@@ -48,3 +49,4 @@ This module contains:
     tensor.sorting_functions
     tensor.statistical_functions
     tensor.utility_functions
+    tensor.constants

dpctl/tensor/_usmarray.pyx

@@ -1278,6 +1278,8 @@ cdef class usm_ndarray:
         The tuple describes the non-partitioned device where the array has been allocated,
         or the non-partitioned parent device of the allocation device.
 
+        See ``DLDeviceType`` for a list of devices supported by the DLPack protocol.
+
         Raises:
             DLPackCreationError:
                 when the ``device_id`` could not be determined.