Open-IOV: initial public release

Open-IOV virtualizes NVIDIA GPUs so that unmodified guest drivers run with
real GPU functionality - CUDA first, 3D and video to follow - on unmodified
hosts, without a proprietary vGPU license and without patching any driver,
firmware or hypervisor.

Core principle: emulate the interface, pass through the work.

  kernel/     open-iov-vfio.ko - host VFIO mdev driver: virtual GPU device
              (PCI config space, BAR shadows, MSI-X), GSP boot emulation,
              GSP-RM message-queue/RPC emulation, memory model (GPU-VA to
              guest-PA discovery) and the semaphore-completion service that
              advances guest UVM progress.
  userspace/  libiov.so management API and examples.
  tools/      guest diagnostics, firmware analysis wrappers, RM probes.
  scripts/    stack manager, VM harness, build/load/test automation.
  config/     hardware/version profiles (the portability surface).
  docs/       documentation tree - start at docs/README.md.
  wiki/       long-form firmware and protocol deep dives.

Status: initialization works end to end for a stock 570.144 Linux guest on
Turing (TU104): driver probe, nvidia-smi, the full GSP-RM control battery,
object creation and the early CUDA/UVM start-up path. GPU work forwarding -
the step that makes CUDA kernels execute on real hardware - is the active
work item; see docs/04-implementation/01-known-knowns-and-unknowns.md.

License: BSD-3-Clause for the project; GPL-2.0 for kernel modules that link
GPL-only kernel interfaces. See LICENSE and docs/09-reference/legal-and-licensing.md.

1 files changed, 75 insertions, 0 deletions

diff --git a/kernel/include/migration.h b/kernel/include/migration.h
new file mode 100644
index 0000000..17cf8db
--- /dev/null
+++ b/kernel/include/migration.h
@@ -0,0 +1,75 @@
+#ifndef _NV_VGPU_MIGRATION_H_
+#define _NV_VGPU_MIGRATION_H_
+
+#include "nvidia-vgpu-vfio.h"
+
+#define NV_VGPU_MIGRATION_MAGIC 0x4E564750
+#define NV_VGPU_MIGRATION_VERSION 1
+
+struct nv_vgpu_migration_ctx {
+	struct nv_vgpu_device *vgpu;
+	struct mutex lock;
+	struct nv_vgpu_migration_state state;
+	struct task_struct *save_thread;
+	struct task_struct *restore_thread;
+	bool in_progress;
+};
+
+struct nv_vgpu_migration_header {
+	u32 magic;
+	u32 version;
+	u32 flags;
+	u32 num_sections;
+	u64 total_size;
+	u64 timestamp;
+};
+
+enum nv_vgpu_migration_section_type {
+	NV_VGPU_MIG_SECTION_DEVICE_STATE = 0x01,
+	NV_VGPU_MIG_SECTION_BAR_STATE = 0x02,
+	NV_VGPU_MIG_SECTION_MEMORY_STATE = 0x03,
+	NV_VGPU_MIG_SECTION_IRQ_STATE = 0x04,
+	NV_VGPU_MIG_SECTION_GSP_STATE = 0x05,
+	NV_VGPU_MIG_SECTION_ENGINE_STATE = 0x06,
+	NV_VGPU_MIG_SECTION_CHANNEL_STATE = 0x07,
+	NV_VGPU_MIG_SECTION_CONTEXT_STATE = 0x08,
+};
+
+struct nv_vgpu_migration_section {
+	u32 type;
+	u32 flags;
+	u64 offset;
+	u64 size;
+	u32 checksum;
+};
+
+struct nv_vgpu_migration_device_state {
+	u32 vgpu_type;
+	u32 vgpu_id;
+	u32 gpu_instance_id;
+	u32 compute_instance_id;
+	u32 state;
+	u64 fb_base;
+	u64 fb_size;
+	u64 bar0_config_base;
+};
+
+struct nv_vgpu_migration_bar_state {
+	u32 bar_index;
+	u64 gpa;
+	u64 size;
+	u64 iova;
+	u8 data[];
+};
+
+int nv_vgpu_migration_init(struct nv_vgpu_device *vgpu);
+void nv_vgpu_migration_cleanup(struct nv_vgpu_device *vgpu);
+
+int nv_vgpu_migration_save_state(struct nv_vgpu_device *vgpu, u8 **data, size_t *size);
+int nv_vgpu_migration_restore_state(struct nv_vgpu_device *vgpu, u8 *data, size_t size);
+
+int nv_vgpu_migration_prepare(struct nv_vgpu_device *vgpu);
+int nv_vgpu_migration_complete(struct nv_vgpu_device *vgpu);
+int nv_vgpu_migration_cancel(struct nv_vgpu_device *vgpu);
+
+#endif
\ No newline at end of file