Revert nvJitLink driver-major compatibility fallback

[isVoid]

Summary

• Revert the linker backend-selection portion of Check nvJitLink driver major compatibility #2320.
• Stop falling back to cuLink solely because the CUDA driver major version is newer than the available nvJitLink major version.
• Retain Check nvJitLink driver major compatibility #2320's cuLink option-array and log-buffer lifetime fix, including its regression tests.

Rationale

The backend-selection policy and the cuLink storage-lifetime fix are independent. This change rolls back the compatibility fallback while preserving the memory-safety fix that keeps storage referenced by CUlinkState alive until cuLinkDestroy.

User impact

cuda.core will use nvJitLink whenever it is available and exposes the required version symbol, even when the CUDA driver has a newer major version. When the cuLink backend is used, its option arrays and log buffers remain valid for the full lifetime of the link state.

Testing

• pixi run --as-is -e cu13 pytest tests/test_linker.py tests/test_optional_dependency_imports.py tests/test_program_cache.py -q (279 passed, 3 skipped)
• Simulated cuLink selection for the two affected test groups (4 passed, 3 skipped with -time is not supported in cuLink)
• git diff --check upstream/main...HEAD

Partially reverts #2320.

[copy-pr-bot]

Auto-sync is disabled for draft pull requests in this repository. Workflows must be run manually.

Contributors can view more details about this message here.

[github-actions]

Doc Preview CI
🚀 View preview at https://nvidia.github.io/cuda-python/pr-preview/pr-2355/
https://nvidia.github.io/cuda-python/pr-preview/pr-2355/cuda-core/
https://nvidia.github.io/cuda-python/pr-preview/pr-2355/cuda-bindings/
https://nvidia.github.io/cuda-python/pr-preview/pr-2355/cuda-pathfinder/
Preview will be ready when the GitHub Pages deployment is complete.

[leofang]

LGTM. The inline comments below are nitpicks about tests for the cuLink backend path, which does not appear to have CI coverage today — feel free to defer.

-- Leo's bot

[leofang]

nit: on a cuLink-backend host, _LinkerBackend.validate() rejects non-None time, so this case fails there:

cuda-python/cuda_core/cuda/core/utils/_program_cache/_keys.py

Lines 509 to 522 in f835032

	if self._decide_driver() is True:
	# Mirror ``_prepare_driver_options``'s exact gate: ``is not
	# None`` for these fields, so ``time=False`` or
	# ``ptxas_options=[]`` is still a rejection. Do NOT use the
	# truthiness-based ``_option_is_set`` helper here.
	unsupported = [
	name for name in _DRIVER_LINKER_UNSUPPORTED_FIELDS if getattr(options, name, None) is not None
	]
	if unsupported:
	raise ValueError(
	f"the cuLink driver linker does not support these options: "
	f"{', '.join(unsupported)}; Program.compile() would reject this "
	f"configuration before producing an ObjectCode."
	)

The monkeypatch in this test only pins the version probe, not the backend decision. Pinning the decision too (same idiom as test_make_program_cache_key_ptx_driver_ignored_fields_collapse) keeps the case portable:

    from cuda.core import _linker

    monkeypatch.setattr(_linker, "_decide_nvjitlink_or_driver", lambda: False)  # nvJitLink

[leofang]

nit: same as the comment above — without the skipif, time_true/time_path hit the driver-unsupported-options rejection on cuLink hosts. Pinning the backend avoids both the failure and the skip:

Suggested change

	def test_make_program_cache_key_accepts_side_effect_options_for_ptx(option_kw):
	"""nvJitLink ``-time`` writes only to its info log, not the filesystem."""
	"""The side-effect guard is NVRTC-specific: PTX (linker) and NVVM must
	not be blocked by options whose side effects only apply under NVRTC."""
	_make_key(code=".version 7.0", code_type="ptx", options=_opts(**option_kw)) # no raise
	def test_make_program_cache_key_accepts_side_effect_options_for_ptx(option_kw, monkeypatch):
	"""The side-effect guard is NVRTC-specific: PTX (linker) and NVVM must
	not be blocked by options whose side effects only apply under NVRTC."""
	from cuda.core import _linker
	monkeypatch.setattr(_linker, "_decide_nvjitlink_or_driver", lambda: False) # nvJitLink
	_make_key(code=".version 7.0", code_type="ptx", options=_opts(**option_kw)) # no raise

[isVoid]

I kinda feel like instead of explicitly pinning, we should mark the test with time argument with skipif when the backend is driver. Even though currently we don't have a setup in the CI matrix that touches this test, in the future if we augment the matrix and this path hits, we can at least see the skip reason in the test, which means the un-reverted change is better.