Integrate vendor VFIO vGPUs into the instance lifecycle

[yummybomb]

Summary

Top layer of the vendor VFIO vGPU stack (generalize-vgpu-device #322 ← hypervisor-liveness #363 ← vendor-vfio-backend #364 ← this). The backend itself is in #364; this layer wires it into the instance lifecycle and adds the release guards that make reusable-VF assignments safe:

• Live-instance claim guard on release — an assignment goes stale when its release succeeds but the metadata save does not (or start fails between the release and its first save). Since vendor VFIO reuses the same VF path across assignments, honoring a stale claim could clear a VF out from under the live instance that owns it now. Every release consults live instance metadata (liveness verified via Harden hypervisor process liveness checks #363's socket-ownership check) and drops stale metadata without touching the device.
• Assignment durability — assignments are tagged with the owning instance ID, persisted before booting a started instance without carrying over the previous VMM identity, and retained in metadata when rollback release fails in create/start so later release paths can still find the device.
• Fail-closed startup reconciliation — the protected set of live VFs is built from ListInstancesForReconcile, which fails on any unreadable metadata instead of silently skipping it. Recent assignments receive bounded startup protection when their PID is absent or stale; when the inventory is unavailable, vendor VFIO reconciliation is skipped entirely while mdev reconciliation still runs.
• Hypervisor scope — hypervisor selection remains a caller policy, preserving the framework layer's existing mdev behavior. The production rollout continues to place vGPU instances on QEMU; the Cloud Hypervisor vendor-VFIO limitation (upstream cloud-hypervisor#7572) is documented in lib/devices/GPU.md.
• Failure reporting — a failed create or start whose vGPU release also failed surfaces as vgpu_cleanup_pending with retained/unretained guidance; the retention record keeps identity fields (name, image, hypervisor) so it lists as a recognizable, deletable instance. The instance API exposes gpu.device_path, the assignment identity on vendor VFIO hosts where mdev_uuid is empty.

Testing

• go build ./..., go vet ./... clean
• go test -race passes for lib/instances targeted suites, lib/devices, lib/resources, lib/builds (TestCreateInstanceWithNetwork needs image pulls + iptables and fails in this environment on the unmodified base as well)

---

Note

High Risk
Touches GPU assignment persistence, release, and startup reconciliation on shared VF paths; incorrect claim or reconcile logic could detach live workloads or leak assignments.

Overview
Wires vendor VFIO vGPUs into create/start/delete and enables devices.CreateVGPU for that framework (integration tests no longer skip vendor VFIO).

Lifecycle and safety: Assignments persist GPUDevicePath, GPUAssignedAt, and optional delete-only retention stubs (GPURetainedForCleanup) when rollback cannot release a VF. Failed create/start rollbacks surface vgpu_cleanup_pending on the API (checked before wrapped errors.Is mappings). Vendor VFIO release scans other instances’ metadata and fails closed when a live claimant or ambiguous inventory exists; mdev skips that scan. Start clears stale hypervisor identity before assigning a new vGPU; fork/snapshot/start reject retention stubs.

Startup: Mdev-only reconcile is replaced by reconcileVGPUs, which builds a protected VF set via ListInstancesForReconcile (strict metadata reads), live hypervisor checks, and a bounded grace for recent assignments, then calls devices.ReconcileVGPUs with optional delayed retry.

API: Instance GPU responses add optional device_path alongside mdev_uuid.

^{Reviewed by Cursor Bugbot for commit 6afdcbb. Bugbot is set up for automated code reviews on this repo. Configure here.}

[yummybomb]

Trimmed over-engineering after an architectural review against the parent task (host kernel 6.8 / Ubuntu 24.04 support). Most of the removed weight came from earlier review-round additions rather than the original design: