Support NVIDIA's CUDA Python bindings

docs/source/cuda/bindings.rst

@@ -4,9 +4,12 @@ CUDA Bindings
 Numba supports two bindings to the CUDA Driver APIs: its own internal bindings
 based on ctypes, and the official `NVIDIA CUDA Python bindings
 <https://nvidia.github.io/cuda-python/>`_. Functionality is equivalent between
-the two bindings, with one exception: the NVIDIA bindings presently do not
-support Per-Thread Default Streams (PTDS), and an exception will be raised on
-import if PTDS is enabled along with the NVIDIA bindings. 
+the two bindings, with two exceptions:
+
+* the NVIDIA bindings presently do not support Per-Thread Default Streams
+  (PTDS), and an exception will be raised on import if PTDS is enabled along
+  with the NVIDIA bindings.
+* The profiling APIs are not available with the NVIDIA bindings.
 
 The internal bindings are used by default. If the NVIDIA bindings are installed,
 then they can be used by setting the environment variable

docs/source/cuda/overview.rst

@@ -65,9 +65,9 @@ CUDA Bindings
 Numba supports interacting with the CUDA Driver API via the `NVIDIA CUDA Python
 bindings <https://nvidia.github.io/cuda-python/>`_ and its own ctypes-based
 binding. The ctypes-based binding is presently the default as Per-Thread
-Default Streams and profiler APIs are not supported with the NVIDIA bindings,
-but otherwise functionality is equivalent between the two. You can install the
-NVIDIA bindings with::
+Default Streams and the profiler APIs are not supported with the NVIDIA
+bindings, but otherwise functionality is equivalent between the two. You can
+install the NVIDIA bindings with::
 
    $ conda install nvidia::cuda-python
 
@@ -77,10 +77,8 @@ if you are using Conda, or::
 
 if you are using pip.
 
-The use of NVIDIA bindings is enabled by setting the environment variable
-``NUMBA_CUDA_USE_NVIDIA_BINDING`` to ``"1"``. See
-:ref:`GPU Support Environment Variables <numba-envvars-gpu-support>` for more
-information.
+The use of the NVIDIA bindings is enabled by setting the environment variable
+:envvar:`NUMBA_CUDA_USE_NVIDIA_BINDING` to ``"1"``.
 
 .. _cudatoolkit-lookup:
 

docs/source/reference/envvars.rst

@@ -516,12 +516,12 @@ GPU support
    heuristic needs to check the number of SMs available on the device in the
    current context.
 
-.. envvar:: CUDA_WARN_ON_IMPLICIT_COPY
+.. envvar:: NUMBA_CUDA_WARN_ON_IMPLICIT_COPY
 
    Enable warnings if a kernel is launched with host memory which forces a copy to and
    from the device. This option is on by default (default value is 1).
 
-.. envvar:: CUDA_USE_NVIDIA_BINDING
+.. envvar:: NUMBA_CUDA_USE_NVIDIA_BINDING
 
    When set to 1, Numba will use the `NVIDIA CUDA Python binding
    <https://nvidia.github.io/cuda-python/>`_ to make calls to the driver API