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, 61 insertions, 0 deletions

diff --git a/scripts/build_nvidia_openrm.sh b/scripts/build_nvidia_openrm.sh
new file mode 100755
index 0000000..89e058b
--- /dev/null
+++ b/scripts/build_nvidia_openrm.sh
@@ -0,0 +1,61 @@
+#!/bin/bash
+# Build NVIDIA OpenRM kernel modules with Open-IOV vGPU VFIO symbols
+# This script ensures KBUILD_EXTRA_SYMBOLS is correctly passed to the build
+
+set -e
+
+OPEN_IOV_ROOT="${OPEN_IOV_ROOT:-$(cd "$(dirname "${BASH_SOURCE[0]}")/.." && pwd)}"
+NVIDIA_OPENRM_DIR="$OPEN_IOV_ROOT/nvidia-open-gpu-kernel-modules"
+KERNEL_DIR="$OPEN_IOV_ROOT/kernel"
+SYMVERS_FILE="$KERNEL_DIR/Module.symvers"
+
+echo "=== Building NVIDIA OpenRM with Open-IOV vGPU VFIO symbols ==="
+echo "Open-IOV root: $OPEN_IOV_ROOT"
+echo "NVIDIA OpenRM dir: $NVIDIA_OPENRM_DIR"
+echo "Kernel Module.symvers: $SYMVERS_FILE"
+
+# Check if open-iov-vfio.ko has been built (provides Module.symvers)
+if [ ! -f "$SYMVERS_FILE" ]; then
+    echo "ERROR: $SYMVERS_FILE not found!"
+    echo "Please build open-iov-vfio.ko first: cd $OPEN_IOV_ROOT && make kernel"
+    exit 1
+fi
+
+echo "Found Module.symvers with $(wc -l < "$SYMVERS_FILE") exported symbols"
+echo "Key vGPU symbols:"
+grep -i vgpu "$SYMVERS_FILE" | head -5
+
+# Build NVIDIA OpenRM kernel modules
+cd "$NVIDIA_OPENRM_DIR/kernel-open"
+
+# Clean previous build to ensure fresh symbol resolution
+echo "Cleaning previous build..."
+make clean
+
+echo "Building NVIDIA OpenRM modules with KBUILD_EXTRA_SYMBOLS..."
+KBUILD_EXTRA_SYMBOLS="$SYMVERS_FILE" make -j$(nproc) modules
+
+# Verify the symbols are resolved
+echo ""
+echo "=== Verification ==="
+echo "Checking nvidia.ko for vGPU VFIO symbols..."
+if modinfo nvidia.ko 2>/dev/null | grep -q "nvidia_vgpu_vfio"; then
+    echo "✓ nvidia.ko has vGPU VFIO symbols"
+else
+    echo "⚠ nvidia.ko may not have vGPU VFIO symbols (check modpost output above)"
+fi
+
+echo ""
+echo "Checking Module.symvers for vGPU symbols..."
+grep -i vgpu Module.symvers | head -5
+
+echo ""
+echo "=== Build Complete ==="
+echo "nvidia.ko location: $NVIDIA_OPENRM_DIR/kernel-open/nvidia.ko"
+echo ""
+echo "To load the stack:"
+echo "  sudo insmod $NVIDIA_OPENRM_DIR/kernel-open/nvidia.ko"
+echo "  sudo insmod $KERNEL_DIR/open-iov-vfio.ko"
+echo "  sudo cp $OPEN_IOV_ROOT/firmware/57X-GSP/570.144/gsp-570.144.fwimage.bin /lib/firmware/"
+echo "  ./scripts/create_mdev.sh"
+echo "  ./scripts/launch_qemu_vgpu.sh"
\ No newline at end of file