cuda.core: reject register() on a non-IPC memory resource

[fedonman]

Summary

Closes #2568.

DeviceMemoryResource.register() and PinnedMemoryResource.register() terminated the process with a segmentation fault when the memory resource did not have IPC enabled, including the default resource from Device().memory_resource. Both route through MP_register, which read self._ipc_data unconditionally although it is None unless IPC is enabled.

Changes

• _ipc.pyx: check is_ipc_enabled in MP_register and raise RuntimeError("Memory resource is not IPC-enabled"), the wording the allocation_handle property already uses. The check runs before the registry insertion, so a rejected registration no longer leaves an entry behind.
• tests/memory_ipc/test_errors.py: regression test covering the raise and the registry.
• docs/source/release/1.2.0-notes.rst: release note. The defect is present in released 1.0.1, 1.1.0 and 1.1.1, so this is user-visible rather than confined to the current cycle.

Verification

Without the production change the new test does not merely fail, it takes the interpreter down (exit=139, SIGSEGV). With the change:

$ python -m pytest tests/memory_ipc/test_errors.py -k test_register_rejects_non_ipc -v
tests/memory_ipc/test_errors.py::test_register_rejects_non_ipc_memory_resource PASSED [100%]
======================= 1 passed, 14 deselected in 0.28s =======================

No regressions in the surrounding suites:

$ python -m pytest tests/test_memory.py tests/memory_ipc/ -q
192 passed, 133 skipped in 4.48s

[copy-pr-bot]

This pull request requires additional validation before any workflows can run on NVIDIA's runners.

Pull request vetters can view their responsibilities here.

Contributors can view more details about this message here.