CI: add pytest durations and skip patterngen seeds under sanitizer (N…

…VIDIA#1400)

- Add --durations=0 to CI pytest invocations to report per-test timings.
- Add cuda_python_test_helpers.under_compute_sanitizer() and use it to skip
  the UVM-heavy PatternGen seed test under compute-sanitizer.

Add experimental NVML bindings to 12.9.x branch (NVIDIA#1349)

* Add experimental NVML bindings to 12.9.x branch

* Add tests from main to 12.9.x branch

* Remove newer APIs

* Remove hand-written 13.0 bindings

* Queue up 'skip reasons' (NVIDIA#1352)

* Handle backport correctly

* Fix cimport

* More versioned struct fixes

* Fix test for 12.9

* Add 13.1 bindings

Prepare for cuda-bindings & cuda-python v13.1.1 release (NVIDIA#1348)

* Prepare for cuda-bindings & cuda-python 13.1.1 release

* Also mention Experimental NVML bindings in cuda_bindings/docs/source/release/13.1.1-notes.rst

* Change 13.1.1 release date to Dec 9, 2025

* Add mention of Experimental NVML bindings to cuda-bindings 12.9.5 release notes.

* Fix edit accident (missing closing >)

Merge changes for cuda-bindings 13.1.0 release (NVIDIA#1315)

* automatic code-gen changes: driver, runtime, nvrtc

* automatic code-gen changes: cufile, nvjitlink, nvvm

* accumulated changes from QA testing period

* cuda_pathfinder release notes and version

* cuda_bindings release notes (without changing cuda_bindings/docs/versions.json, with 12.9.5 release notes)

* ci/ changes: adds 19 test jobs (for a new total of 52 test jobs)

* Add nvidia-cufftmp-cu13 in cuda_pathfinder/pyproject.toml

Merge changes for cuda-bindings 13.1.0 release (NVIDIA#1315)

* automatic code-gen changes: driver, runtime, nvrtc

* automatic code-gen changes: cufile, nvjitlink, nvvm

* accumulated changes from QA testing period

* cuda_pathfinder release notes and version

* cuda_bindings release notes (without changing cuda_bindings/docs/versions.json, with 12.9.5 release notes)

* ci/ changes: adds 19 test jobs (for a new total of 52 test jobs)

* Add nvidia-cufftmp-cu13 in cuda_pathfinder/pyproject.toml

[cuda.core v0.4.2] Ensure custom actions can run on the `release/*` b…

…ranches (NVIDIA#1261)

* relax to allow release/* branches in get_pr_number

* relax to allow release/* branches in doc_preview

Add cuTENSOR support & bug fixes discovered while working on conda te…

…sting (NVIDIA#1194)

* Add support for cutensor. Still works for existing libnames, cutensor testing is INCOMPLETE.

* Generalize tests/test_find_nvidia_headers.py to also cover cutensor

* test_find_nvidia_headers.py conda testing and fix

* test_load_nvidia_dynamic_lib.py fix conda testing

* Add conda_create_for_pathfinder_testing.ps1

* Bug fix: SITE_PACKAGES_LIBDIRS_WINDOWS_OTHER cutensor, cutensorMg paths

* Add cudss paths to SITE_PACKAGES_LIBDIRS_WINDOWS_OTHER

* Add SUPPORTED_HEADERS_NON_CTK_ALL to fix Windows site-packages tests

* Bug fix (existing code): conda cccl header directory

* test_find_nvidia_headers.py: refer to toolshed/conda_create_for_pathfinder_testing.*

* nvidia-libmathdx-... only exists for cu12: tolerate abs_path=None in test_load_nvidia_dynamic_lib.py

* find_nvidia_headers.py cccl IS_WINDOWS: fall-through after checking for conda anomaly, to restore proper functioning for standard CTK installations

* Add cublasmp DIRECT_DEPENDENCIES (closes NVIDIA#1116)

* Add SUPPORTED_HEADERS_NON_CTK to cuda_pathfinder/docs/source/api.rst

* Add 1.3.2-notes.rst

* Add nvidia-cufftmp-cu13 data in supported_nvidia_libs.py

Manually tested with:

pip install nvidia-cufftmp-cu13==12.1.3.1

That wheel was yanked, therefore not adding to pyproject.toml

* Add missing comma in toolshed/conda_create_for_pathfinder_testing.ps1

* Systematically add _abs_norm() in find_nvidia_header_directory()

* Move "conda has this anomaly" comment to the end of the line

* Add empty SUPPORTED_HEADERS_NON_CTK_WINDOWS_ONLY in supported_nvidia_headers.py (mirrors supported_nvidia_libs.py)

* Remove SUPPORTED_HEADERS_NON_CTK from public API, as requested by Leo offline

Prepare for v0.4.1 patch release (NVIDIA#1195)

* Initial plan

* Bump cuda.core version to 0.4.1 and add release notes

Co-authored-by: leofang <[email protected]>

* Update release notes for version 0.4.1

* fix typo

Co-authored-by: greptile-apps[bot] <165735046+greptile-apps[bot]@users.noreply.github.com>

* nit: fix crossref

---------

Co-authored-by: copilot-swe-agent[bot] <[email protected]>
Co-authored-by: leofang <[email protected]>
Co-authored-by: greptile-apps[bot] <165735046+greptile-apps[bot]@users.noreply.github.com>

Update the version number in cuda_bindings (NVIDIA#1158)

bump to 12.9.4 (NVIDIA#1159)