Deploy doc preview for PR 1969 (d9c8839)

Author: cuda-python-bot

docs/pr-preview/pr-1969/cuda-bindings/latest/_sources/api.rst.txt

@@ -15,6 +15,7 @@ CUDA Python API Reference
    module/nvjitlink
    module/nvvm
    module/nvfatbin
+   module/cudla
    module/cufile
    module/nvml
    module/utils

docs/pr-preview/pr-1969/cuda-bindings/latest/_sources/module/cudla.rst.txt

@@ -0,0 +1,64 @@
+.. SPDX-FileCopyrightText: Copyright (c) 2026 NVIDIA CORPORATION & AFFILIATES. All rights reserved.
+.. SPDX-License-Identifier: LicenseRef-NVIDIA-SOFTWARE-LICENSE
+
+.. default-role:: cpp:any
+
+cudla
+=====
+
+Note
+----
+
+The cuDLA bindings require a Jetson platform with DLA hardware (Xavier or Orin).
+cuDLA is not available on desktop GPUs.
+
+Functions
+---------
+
+cuDLA defines the following functions for DLA device management and inference.
+
+.. autofunction:: cuda.bindings.cudla.get_version
+.. autofunction:: cuda.bindings.cudla.device_get_count
+.. autofunction:: cuda.bindings.cudla.create_device
+.. autofunction:: cuda.bindings.cudla.destroy_device
+.. autofunction:: cuda.bindings.cudla.mem_register
+.. autofunction:: cuda.bindings.cudla.mem_unregister
+.. autofunction:: cuda.bindings.cudla.module_load_from_memory
+.. autofunction:: cuda.bindings.cudla.module_get_attributes
+.. autofunction:: cuda.bindings.cudla.module_unload
+.. autofunction:: cuda.bindings.cudla.submit_task
+.. autofunction:: cuda.bindings.cudla.device_get_attribute
+.. autofunction:: cuda.bindings.cudla.get_last_error
+.. autofunction:: cuda.bindings.cudla.set_task_timeout_in_ms
+
+Types
+-----
+
+.. autoclass:: cuda.bindings.cudla.ExternalMemoryHandleDesc
+.. autoclass:: cuda.bindings.cudla.ExternalSemaphoreHandleDesc
+.. autoclass:: cuda.bindings.cudla.ModuleTensorDescriptor
+.. autoclass:: cuda.bindings.cudla.Fence
+.. autoclass:: cuda.bindings.cudla.DevAttribute
+.. autoclass:: cuda.bindings.cudla.ModuleAttribute
+.. autoclass:: cuda.bindings.cudla.WaitEvents
+.. autoclass:: cuda.bindings.cudla.SignalEvents
+.. autoclass:: cuda.bindings.cudla.Task
+
+Enums
+-----
+
+.. autoclass:: cuda.bindings.cudla.Status
+
+   .. autoattribute:: cuda.bindings.cudla.Status.Success
+
+.. autoclass:: cuda.bindings.cudla.Mode
+
+   .. autoattribute:: cuda.bindings.cudla.Mode.CUDA_DLA
+   .. autoattribute:: cuda.bindings.cudla.Mode.STANDALONE
+
+.. autoclass:: cuda.bindings.cudla.ModuleAttributeType
+.. autoclass:: cuda.bindings.cudla.FenceType
+.. autoclass:: cuda.bindings.cudla.ModuleLoadFlags
+.. autoclass:: cuda.bindings.cudla.SubmissionFlags
+.. autoclass:: cuda.bindings.cudla.AccessPermissionFlags
+.. autoclass:: cuda.bindings.cudla.DevAttributeType

docs/pr-preview/pr-1969/cuda-bindings/latest/_sources/release/13.2.1-notes.rst.txt

@@ -0,0 +1,14 @@
+.. SPDX-FileCopyrightText: Copyright (c) 2026 NVIDIA CORPORATION & AFFILIATES. All rights reserved.
+.. SPDX-License-Identifier: LicenseRef-NVIDIA-SOFTWARE-LICENSE
+
+.. module:: cuda.bindings
+
+``cuda-bindings`` 13.2.1 Release notes
+======================================
+
+Bugfixes
+--------
+
+* Per-thread default stream mode would be used whenever the
+  ``CUDA_PYTHON_CUDA_PER_THREAD_DEFAULT_STREAM`` environment variable was set,
+  even if it was set to ``0``.

docs/pr-preview/pr-1969/cuda-bindings/latest/_sources/release/13.3.0-notes.rst.txt

@@ -0,0 +1,41 @@
+.. SPDX-FileCopyrightText: Copyright (c) 2025-2026 NVIDIA CORPORATION & AFFILIATES. All rights reserved.
+.. SPDX-License-Identifier: LicenseRef-NVIDIA-SOFTWARE-LICENSE
+
+.. module:: cuda.bindings
+
+``cuda-bindings`` 13.3.0 Release notes
+======================================
+
+Highlights
+----------
+
+
+Bugfixes
+--------
+
+* Fixed a use-after-free in ``cudaGraphGetEdges``, ``cudaGraphNodeGetDependencies``,
+  ``cudaGraphNodeGetDependentNodes``, ``cudaStreamGetCaptureInfo``, and their
+  driver-API counterparts (``cuGraphGetEdges``, ``cuGraphNodeGetDependencies``,
+  ``cuGraphNodeGetDependentNodes``, ``cuStreamGetCaptureInfo``). The returned
+  ``cudaGraphEdgeData``/``CUgraphEdgeData`` wrappers were backed by a scratch
+  buffer that was freed before the call returned, leaving every wrapper holding
+  a dangling pointer. The returned wrappers now own deep copies of the edge
+  data.
+  (`Issue #1804 <https://github.com/NVIDIA/cuda-python/issues/1804>`_)
+
+* Fixed a double-free in the generated setters for list-valued struct members
+  (e.g. ``CUlaunchConfig.attrs``, ``CUDA_MEM_ALLOC_NODE_PARAMS.accessDescs``,
+  external-semaphore and batch-mem-op node parameter arrays, and their runtime
+  counterparts). Assigning an empty list freed the internal buffer but left
+  the cached pointer non-NULL, so a subsequent assignment or ``__dealloc__``
+  would call ``free()`` again on the dangling pointer.
+
+
+Miscellaneous
+-------------
+
+
+Known issues
+------------
+
+* Updating from older versions (v12.6.2.post1 and below) via ``pip install -U cuda-python`` might not work. Please do a clean re-installation by uninstalling ``pip uninstall -y cuda-python`` followed by installing ``pip install cuda-python``.

docs/pr-preview/pr-1969/cuda-bindings/latest/_static/scripts/bootstrap.js

@@ -1,5 +1,5 @@
 /*!
- * Font Awesome Free 6.5.2 by @fontawesome - https://fontawesome.com
+ * Font Awesome Free 7.2.0 by @fontawesome - https://fontawesome.com
  * License - https://fontawesome.com/license/free (Icons: CC BY 4.0, Fonts: SIL OFL 1.1, Code: MIT License)
- * Copyright 2024 Fonticons, Inc.
+ * Copyright 2026 Fonticons, Inc.
  */