feat(cuinterpose): implement CUDA memory IPC over tracked VMM

[galletas1712]

Summary

Refs #295 (approved).

Implement synchronous device malloc and CUDA memory IPC using existing VMM records, peer export FDs, and host-carrier reconstruction. Native memory-IPC functions are not called.

Review boundary

A fixed 64-byte ticket identifies the creator and allocation. Repeated opens share an address and count references; final close removes the import. Free synchronizes before unmapping. Address-range lookup preserves the requested size. Supported scope is fully interposed peers and the documented context behavior, not event/pool IPC or general peer-access emulation. Frontend declarations are in #326; focused tests stay in #338.

This PR contains one signed-off commit, 6a64f3d15857f5b4cd1c6d85ba3ec41707e83cd3. Diff: 4 files changed, 402 insertions(+).

Stack and compatibility

The stack starts directly on main; it does not depend on the PageBroker GPU-transfer branch or #323. The shim only saves shared creator bytes through host carriers. Never-shared allocations remain native CUDA state. The stack removes launch-job/jobfile support, and older jobfile-dependent or draft shim artifacts are rejected rather than migrated.

PageBroker and native CustomStorage changes are separate. No PR in this stack adds that implementation.

Order	PR	Scope
1	#326	C preload frontend and private ABI
2	#327	typed identities and peer transport
3	#328	process identity and peer export service
4	#329	VMM ownership and allocation lifecycle
5	#342	implement CUDA memory IPC over tracked VMM
6	#330	host-carrier storage for shared allocation bytes
7	#331	multicast tracking and reconstruction
8	#332	Rust lifecycle coordinator
9	#333	package frontend, Rust backend, and coordinator
10	#334	deliver CUDA tools and enable opt-in preload
11	#335	coordinate cuinterpose capture and restore without jobfiles
12	#336	deliver CUDA tools without rewriting workload commands
13	#337	explain host carriers, memory IPC, and restore ordering
14	#338	unit, integration, and GPU lifecycle coverage

Validation

Validation of the assembled implementation:

• Full API, agent, and operator Go tests; key agent race tests.
• Rust workspace tests, strict Clippy, GNU/musl builds, ABI/ELF checks, packaged frontend/fake-driver tests, and memory-IPC regressions.
• Helm tests/lint, Python manifest/report tests, repository lint, and make check in a clean disposable worktree.
• A small real-agent CRIU cross-node test passed on two GPU nodes, covering memory IPC, private/shared VMM, multicast/graph replay, bytes, and original addresses. It preceded only the final manifest-format guard; that guard has local regression coverage.

GLM testing without CustomStorage was cancelled at the user's request and is not a pass. GLM qualification uses a separate composition with CustomStorage; previous experimental GLM results are not qualification of this rebuilt stack. The installed test driver is not claimed to be a stock-driver qualification.

Tests above were run on the assembled implementation, not claimed independently for every source-only intermediate PR. The standard-library cleanup was validated with Rust workspace tests and strict Clippy, GNU/musl builds and packaged fake-driver tests, plus targeted pod-contract and CUDA-agent Go tests. Physical-GPU tests were not rerun for this cleanup; the earlier GPU results above remain attributed to the pre-cleanup implementation.

[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.

[coderabbitai]

Important

Draft PR not reviewed

Draft PRs are not automatically reviewed by default.

• Trigger a manual review

To automatically review draft PRs, update your CodeRabbit configuration:

reviews:
  auto_review:
    drafts: true

---

_{Comment @coderabbitai help to get the list of available commands.}