populate API references; clean up & polish docs

cuda_core/cuda/core/experimental/_launcher.py

@@ -103,16 +103,14 @@ def _cast_to_3_tuple(self, cfg):
 
 
 def launch(kernel, config, *kernel_args):
-    """Launches a :obj:`Kernel` object with launch-time configuration.
-
-    Invokes a :obj:`Kernel` object with specified launch-time
-    configurations.
+    """Launches a :obj:`~cuda.core.experimental._module.Kernel`
+    object with launch-time configuration.
 
     Parameters
     ----------
-    kernel : :obj:`Kernel`
+    kernel : :obj:`~cuda.core.experimental._module.Kernel`
         Kernel to launch.
-    config : Any
+    config : :obj:`LaunchConfig`
         Launch configurations inline with options provided by
         :obj:`LaunchConfig` dataclass.
     *kernel_args : Any

cuda_core/cuda/core/experimental/_module.py

@@ -48,8 +48,8 @@ def _lazy_init():
 class Kernel:
     """Represent a compiled kernel that had been loaded onto the device.
 
-    Kernel instances can execution when passed directly into a
-    launch function.
+    Kernel instances can execution when passed directly into the
+    :func:`~cuda.core.experimental.launch` function.
 
     Directly creating a :obj:`Kernel` is not supported, and they
     should instead be created through a :obj:`ObjectCode` object.

cuda_core/cuda/core/experimental/_program.py

@@ -8,7 +8,8 @@
 
 
 class Program:
-    """Represent a compilation machinery to process programs into :obj:`ObjectCode`.
+    """Represent a compilation machinery to process programs into
+    :obj:`~cuda.core.experimental._module.ObjectCode`.
 
     This object provides a unified interface to multiple underlying
     compiler libraries. Compilation support is enabled for a wide
@@ -19,7 +20,7 @@ class Program:
     code : Any
         String of the CUDA Runtime Compilation program.
     code_type : Any
-        String of the code type. Only "c++" is currently supported.
+        String of the code type. Currently only ``"c++"`` is supported.
 
     """
 
@@ -74,7 +75,7 @@ def compile(self, target_type, options=(), name_expressions=(), logs=None):
 
         Returns
         -------
-        :obj:`ObjectCode`
+        :obj:`~cuda.core.experimental._module.ObjectCode`
             Newly created code object.
 
         """

cuda_core/docs/source/_templates/autosummary/class.rst

@@ -0,0 +1,26 @@
+{{ fullname | escape | underline}}
+
+.. currentmodule:: {{ module }}
+
+.. autoclass:: {{ objname }}
+
+   {% block methods %}
+   {% if methods %}
+   .. rubric:: {{ _('Methods') }}
+
+   {% for item in methods %}
+   .. automethod:: {{ item }}
+   {%- endfor %}
+
+   {% endif %}
+   {% endblock %}
+
+   {% block attributes %}
+   {% if attributes %}
+   .. rubric:: {{ _('Attributes') }}
+
+   {% for item in attributes %}
+   .. autoattribute:: {{ item }}
+   {%- endfor %}
+   {% endif %}
+   {% endblock %}

cuda_core/docs/source/_templates/autosummary/dataclass.rst

@@ -0,0 +1,10 @@
+{{ fullname | escape | underline}}
+
+.. currentmodule:: {{ module }}
+
+.. autoclass:: {{ objname }}
+
+   {% block methods %}
+   .. automethod:: __init__
+   {% endblock %}
+

cuda_core/docs/source/_templates/autosummary/namedtuple.rst

@@ -0,0 +1,8 @@
+{{ fullname | escape | underline}}
+
+.. currentmodule:: {{ module }}
+
+.. autoclass:: {{ objname }}
+   :members: __new__
+   :special-members: __new__
+   :exclude-members: count, index, __reduce__, __reduce_ex__, __repr__, __hash__, __str__, __getnewargs__

cuda_core/docs/source/api.rst

@@ -3,13 +3,26 @@
 ``cuda.core.experimental`` API Reference
 ========================================
 
+All of the APIs listed (or cross-referenced from) below are considered *experimental*
+and subject to future changes without deprecation notice. Once stablized they will be
+moved out of the ``experimental`` namespace.
+
+
 CUDA runtime
 ------------
 
 .. autosummary::
    :toctree: generated/
 
    Device
+   launch
+
+   :template: dataclass.rst
+
+   EventOptions
+   StreamOptions
+   LaunchConfig
+
 
 CUDA compilation toolchain
 --------------------------

cuda_core/docs/source/api_private.rst

@@ -0,0 +1,28 @@
+:orphan:
+
+.. This page is to generate documentation for private classes exposed to users,
+   i.e., users cannot instantiate it by themselves but may use it's properties
+   or methods via returned values from public APIs. These classes must be referred
+   in public APIs returning their instances.
+
+.. currentmodule:: cuda.core.experimental
+
+CUDA runtime
+------------
+
+.. autosummary::
+   :toctree: generated/
+
+   _memory.Buffer
+   _stream.Stream
+   _event.Event
+
+
+CUDA compilation toolchain
+--------------------------
+
+.. autosummary::
+   :toctree: generated/
+
+   _module.Kernel
+   _module.ObjectCode

cuda_core/docs/source/conf.py

@@ -34,7 +34,8 @@
     'sphinx.ext.autosummary',
 	'sphinx.ext.napoleon',
 	'myst_nb',
-	'enum_tools.autoenum'
+	'enum_tools.autoenum',
+    'sphinx_copybutton',
 ]
 
 # Add any paths that contain templates here, relative to this directory.
@@ -77,3 +78,6 @@
 # relative to this directory. They are copied after the builtin static files,
 # so a file named "default.css" will overwrite the builtin "default.css".
 html_static_path = ['_static']
+
+# skip cmdline prompts
+copybutton_exclude = '.linenos, .gp'

cuda_core/docs/source/install.md

@@ -5,25 +5,19 @@
 `cuda.core` is supported on all platforms that CUDA is supported. Specific
 dependencies are as follows:
 
-* Driver: Linux (450.80.02 or later) Windows (456.38 or later)
-* CUDA Toolkit 12.0 to 12.6
+|                   | CUDA 11      | CUDA 12     |
+|------------------ | ------------ | ----------- |
+| CUDA Toolkit [^1] | 11.2 - 11.8  | 12.0 - 12.6 |
+| Driver            | 450.80.02+ (Linux), 452.39+ (Windows) | 525.60.13+ (Linux), 527.41+ (Windows) |
 
-
-## Installing from PyPI
-
-Coming soon!
-
-
-## Installing from Conda
-
-Coming soon!
+[^1]: Including `cuda-python`.
 
 
 ## Installing from Source
 
-```shell
+```console
 $ git clone https://github.com/NVIDIA/cuda-python
 $ cd cuda-python/cuda_core
 $ pip install .
 ```
-For now `cuda-python` (`cuda-bindings` later) is a required dependency.
+For now `cuda-python` (`cuda-bindings` later) 11.x or 12.x is a required dependency.

cuda_core/docs/source/release/0.1.0-notes.md

@@ -1,15 +1,17 @@
-# ``cuda.core`` Release notes
+# `cuda.core` Release notes
 
-Released on Oct XX, 2024
+Released on Nov XX, 2024
 
 ## Hightlights
-- Initial beta 1 release
+- Initial EA1 (early access) release
 - Supports all platforms that CUDA is supported
-- Supports all CUDA 12.x drivers
-- Supports all CUDA 12.x Toolkits
+- Supports all CUDA 11.x/12.x drivers
+- Supports all CUDA 11.x/12.x Toolkits
 - Pythonic CUDA runtime and other core functionalities
 
 ## Limitations
 
-- Source code release only; Python packages coming in a future release
-- Support for CUDA 11.x coming in the next release
+- All APIs are currently *experimental* and subject to change without deprecation notice.
+  Please kindly share your feedbacks with us so that we can make `cuda.core` better!
+- Source code release only; `pip`/`conda` support is coming in a future release
+- Windows TCC mode is [not yet supported](https://github.com/NVIDIA/cuda-python/issues/206)

cuda_python/docs/environment-docs.yml

@@ -11,6 +11,7 @@ dependencies:
   - pytest
   - scipy
   - sphinx
+  - sphinx-copybutton
   - pip:
     - furo
     - myst-nb