fix(cuda.core): move VirtualMemoryResource onto the _rt handle layer

[Andy-Jost]

Summary

Moves VirtualMemoryResource onto the _rt handle layer, as planned in #2906 (design). Each physical allocation, address reservation and mapping is a std::shared_ptr handle with a deleter that knows the exact driver call to undo it. A buffer owns a range of mappings through its device pointer handle, so everything a buffer maps is released when the last buffer that maps it closes, and a failed multi-step operation unwinds by letting its local handles die. The module moves from Python to Cython. The design lands as cuda_core/cuda/core/_cpp/rt/VMM_DESIGN.md.

Behavior changes

• modify_allocation returns a new VirtualMemoryBuffer and leaves the buffer passed in open. The two alias the same physical memory, which is freed when the last of them closes. The pointer is preserved when the driver grants the adjacent address range. A request the buffer already covers returns the buffer itself.
• Buffer.size after a grow is the aligned total.
• config= applies to the chunk the call adds and is no longer stored on the resource. It must keep the resource's location and passes the constructor's option checks, including the RDMA support check.
• Buffers from allocate() free themselves when they close and no longer call deallocate(), which now serves pointers wrapped with Buffer.from_handle.
• A buffer records the stream passed to allocate(). The last close of an aliased range synchronizes every recorded stream before it unmaps. An explicit close() on a capturing stream other than a default stream raises; a release ordered on a default stream that would disturb a capture in its context is reported as a CUDAWarning and unmaps without the synchronization. The synchronization runs in relaxed capture mode, so it does not invalidate a capture on an unrelated stream.
• modify_allocation accepts only buffers this resource returned.
• location_type="host" requires handle_type=None, which the driver requires.
• allocate(0) returns an empty buffer without a driver call.

Testing

• The mocked grow test is replaced by real-driver tests: the leak in three modes (allocate, grow, forced move on 2 MiB, which fails rather than skips when the driver declines the placement after the warm-up), grow in place with the pointer preserved, grow that moves with contents preserved, aliasing and cross-alias visibility, unaligned and repeated grows, a failed grow that leaves the input intact, the capture check on close with and without a stream argument, a release from garbage collection during a capture, a close during an unrelated global-mode or thread-local capture, stream synchronization at close in both close orders (deterministic, with a sleep kernel and events), cuMemGetAccess on both chunks after a config= grow, the per-call config checks, the raw-pointer deallocate() path, host location without a current context, size zero with an inherited stream, and buffers alive at interpreter shutdown.
• The graph in-flight retention test is parametrized over the device pool and the virtual memory resource: a memcpy node keeps a virtual memory buffer mapped after close() until the launch completes and the graph is gone. A second test keeps the range mapped across a grow and the close of every alias.
• Before the review round, the full cuda_core suite passed on an H100 with CUDA 13.4: 4288 passed, 106 skipped, 2 xfailed, 0 failed, and the VMM selection passed twice in random order. After the review round, the VMM selection (37 passed, 1 skipped) and the graph retention tests (3 passed) passed on an H200 with CUDA 13.4; CI covers the full suite.

Issues

Fixes #2887
Fixes #2907
Fixes #2908
Fixes #2909
Fixes #2886
Fixes #2345
Fixes #2877

Addresses #2388 items 1, 2 and 3 (the rollback that lost access grants, the dead fast path, and the finalizer warnings; item 4 landed in #2418) and the size-0, misaligned-probe and host handle-type parts of #2910. Part of #2906. Supersedes #2880, #2889 and #2237: they edit the module this PR replaces, and the fixes they carried (#2877, #2886, #2345) are part of the redesign.

🤖 Generated with Claude Code

[github-actions]

Doc Preview CI
🚀 View preview at https://nvidia.github.io/cuda-python/pr-preview/pr-2917/
https://nvidia.github.io/cuda-python/pr-preview/pr-2917/cuda-core/
https://nvidia.github.io/cuda-python/pr-preview/pr-2917/cuda-bindings/
https://nvidia.github.io/cuda-python/pr-preview/pr-2917/cuda-pathfinder/
Preview will be ready when the GitHub Pages deployment is complete.

[sylvesterkaczmarek]

The no-access case can call this with count == 0 and descs.data() from an empty vector. Constructing access(descs, descs + count) then does pointer arithmetic/range construction on a possibly null pointer. Since self_access=None with no peers is supported, could this special-case zero descriptors before forming the range?