Initial commit: Open-IOV vGPU-IO complete stack

- Firmware decompilation: GSP 570.144W (28.5MB ELF, RV64IMAC), 132 decompiled C files
- Kernel modules: nvidia.ko (34MB, OpenRM), open-iov-vfio.ko (2.8MB, VFIO mdev), gsp-vgpu.ko (1.8MB, guest driver)
- Userspace: libiov.so/libvgpu.so (76KB), example_vgpu (28KB), rpc_emulator.c
- PCI config space emulation (256B with PM, MSI, MSI-X, PCIe caps)
- 12 RPC types fully mapped from firmware decompilation
- BAR trap, WPR, scheduling, notification, migration emulation
- Helper scripts: load_vgpu_stack, create_mdev, launch_qemu_vgpu, test_vgpu_stack, setup_guest_driver, unload_vgpu_stack
- Architecture documentation
- Wiki: 10 technical documents
- Git LFS for large binaries (.ko, .bin, .iso, .img, .o, ghidra repos)
- nvidia-open-gpu-kernel-modules as submodule

1 files changed, 208 insertions, 0 deletions

diff --git a/userspace/sdks/nv570/inc/ctrl/ctrl2080/ctrl2080fla.h b/userspace/sdks/nv570/inc/ctrl/ctrl2080/ctrl2080fla.h
new file mode 100644
index 0000000..1f0c86a
--- /dev/null
+++ b/userspace/sdks/nv570/inc/ctrl/ctrl2080/ctrl2080fla.h
@@ -0,0 +1,208 @@
+/*
+ * SPDX-FileCopyrightText: Copyright (c) 2006-2018 NVIDIA CORPORATION & AFFILIATES. All rights reserved.
+ * SPDX-License-Identifier: MIT
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ * and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in
+ * all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL
+ * THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
+ * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING
+ * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER
+ * DEALINGS IN THE SOFTWARE.
+ */
+
+#pragma once
+
+#include <nvtypes.h>
+
+//
+// This file was generated with FINN, an NVIDIA coding tool.
+// Source file:      ctrl/ctrl2080/ctrl2080fla.finn
+//
+
+#include "ctrl/ctrl2080/ctrl2080base.h"
+
+/* NV20_SUBDEVICE_XX FLA control commands and parameters */
+
+/*
+ * NV2080_CTRL_CMD_FLA_RANGE
+ *
+ * This command is used to initialize/destroy FLA VAS for a GPU.This is intended
+ * to be used by RM clients that manages the FLA VASpace range. The mode of the
+ * command is decided based on the parameter passed by the client.
+ *
+ *    base
+ *         This parameter specifies the base of the FLA VAS that needs to be allocated
+ *         for this GPU
+ *
+ *    size
+ *         This parameter specifies the size of the FLA VAS that needs to be allocated
+ *         for this GPU
+ *
+ *    mode
+ *         This parameter specifies the functionality of the command.
+ *           MODE_INITIALIZE
+ *             Setting this mode, will initialize the FLA VASpace for the gpu with
+ *             base and size passed as arguments. FLA VASpace will be owned by RM.
+ *             if the client calls the command more than once before destroying
+ *             the FLA VAS, then this command will verify the range exported before and
+ *             return success if it matches. If FLA is not supported for the platform,
+ *             will return NV_ERR_NOT_SUPPORTED.
+ *           MODE_DESTROY (deprecated)
+ *             This command is NOP.
+ *           MODE_HOST_MANAGED_VAS_INITIALIZE
+ *             This mode will initialize the FLA VASpace for the gpu with hVASpace
+ *             handle in addition to base and size arguments. FLA VASpace will be initiated 
+ *             and owned by guest RM. Used only in virtualization platforms by internal clients.
+ *           MODE_HOST_MANAGED_VAS_DESTROY
+ *              This mode will destroy the FLA VAS associated with the device. It will destruct
+ *              only the resources associated with host RM side. Used only in virtualization platforms
+ *              by internal clients.
+ *
+ *    hVASpace 
+ *         This paramete specifies the FLA VAspace that needs to be associated with 
+ *         device. This parameter takes effect only for internal client in virtualization
+ *         platforms. For any other platform and external clients, this parameter has no effect.
+ *
+ * Possible status values returned are:
+ *   NV_OK
+ *   NV_ERR_IN_USE
+ *   NV_ERR_INVALID_OWNER
+ *   NV_ERR_NOT_SUPPORTED
+ */
+
+#define NV2080_CTRL_CMD_FLA_RANGE (0x20803501) /* finn: Evaluated from "(FINN_NV20_SUBDEVICE_0_FLA_INTERFACE_ID << 8) | NV2080_CTRL_FLA_RANGE_PARAMS_MESSAGE_ID" */
+
+#define NV2080_CTRL_FLA_RANGE_PARAMS_MESSAGE_ID (0x1U)
+
+typedef struct NV2080_CTRL_FLA_RANGE_PARAMS {
+    NV_DECLARE_ALIGNED(NvU64 base, 8);
+    NV_DECLARE_ALIGNED(NvU64 size, 8);
+    NvU32    mode;
+    NvHandle hVASpace;
+} NV2080_CTRL_FLA_RANGE_PARAMS;
+
+#define NV2080_CTRL_FLA_RANGE_PARAMS_MODE_NONE 0x00000000
+#define NV2080_CTRL_FLA_RANGE_PARAMS_MODE_INITIALIZE                         NVBIT(0)
+#define NV2080_CTRL_FLA_RANGE_PARAMS_MODE_DESTROY                            NVBIT(1)
+#define NV2080_CTRL_FLA_RANGE_PARAMS_MODE_HOST_MANAGED_VAS_INITIALIZE        NVBIT(2)
+#define NV2080_CTRL_FLA_RANGE_PARAMS_MODE_HOST_MANAGED_VAS_DESTROY           NVBIT(3)
+
+
+/*
+ * NV2080_CTRL_CMD_FLA_SETUP_INSTANCE_MEM_BLOCK
+ *
+ * This command is used to (un)bind FLA Instance Memory Block(IMB) with MMU.
+ * This control call is created for vGPU platform, when a FLA VAS is created/destroyed
+ * by Guest RM. Guest RM doesn't have privilege to (un)bind the IMB with MMU, hence
+ * need to be RPC-ed to Host RM to (un)bind
+ * The mode of the command is decided based on the actionParam passed by the client.
+ *
+ *    imbPhysAddr
+ *         This parameter specifies the FLA Instance Memory Block PA to be programmed
+ *         to MMU. IMB address should be 4k aligned. This parameter is needed only
+ *         for ACTION_BIND.
+ *
+ *    addrSpace
+ *         This parameter specifies the address space of FLA Instance Memory Block. This
+ *         parmater is needed only for ACTION_BIND.
+ *         Available options are:
+ *           NV2080_CTRL_FLA_ADDRSPACE_SYSMEM
+ *               Clients need to use this address space if the IMB is located in sysmem
+ *           NV2080_CTRL_FLA_ADDRSPACE_FBMEM
+ *               Clients need to use this address space if the IMB is located in FB
+ *
+ *    actionParam
+ *         This parameter specifies the functionality of the command.
+ *           NV2080_CTRL_FLA_ACTION_BIND
+ *             Setting this type, will call busBindFla helper HAL
+ *           NV2080_CTRL_FLA_ACTION_UNBIND
+ *             Setting this type, will call busUnbindFla helper HAL
+ *
+ * Possible status values returned are:
+ *   NV_OK
+ *   NV_ERR_INVALID_ARGUMENT
+ *   NV_ERR_NOT_SUPPORTED
+ *   NV_ERR_INSUFFICIENT_PERMISSIONS
+ */
+typedef enum NV2080_CTRL_FLA_ADDRSPACE {
+    NV2080_CTRL_FLA_ADDRSPACE_SYSMEM = 0,
+    NV2080_CTRL_FLA_ADDRSPACE_FBMEM = 1,
+} NV2080_CTRL_FLA_ADDRSPACE;
+
+typedef enum NV2080_CTRL_FLA_ACTION {
+    NV2080_CTRL_FLA_ACTION_BIND = 0,
+    NV2080_CTRL_FLA_ACTION_UNBIND = 1,
+} NV2080_CTRL_FLA_ACTION;
+
+#define NV2080_CTRL_CMD_FLA_SETUP_INSTANCE_MEM_BLOCK (0x20803502) /* finn: Evaluated from "(FINN_NV20_SUBDEVICE_0_FLA_INTERFACE_ID << 8) | NV2080_CTRL_FLA_SETUP_INSTANCE_MEM_BLOCK_PARAMS_MESSAGE_ID" */
+
+#define NV2080_CTRL_FLA_SETUP_INSTANCE_MEM_BLOCK_PARAMS_MESSAGE_ID (0x2U)
+
+typedef struct NV2080_CTRL_FLA_SETUP_INSTANCE_MEM_BLOCK_PARAMS {
+    NV_DECLARE_ALIGNED(NvU64 imbPhysAddr, 8);
+    NV2080_CTRL_FLA_ADDRSPACE addrSpace;
+    NV2080_CTRL_FLA_ACTION    flaAction;
+} NV2080_CTRL_FLA_SETUP_INSTANCE_MEM_BLOCK_PARAMS;
+
+
+/*
+ * NV2080_CTRL_CMD_FLA_GET_RANGE
+ *
+ * This command is used to query the FLA base and size from plugin to return as static info to Guest RM. 
+ *
+ *    base
+ *         This parameter returns the base address of FLA range registered to the subdevice. 
+ *    size
+ *         This parameter returns the size of FLA range registered to the subdevice. 
+ *
+ * Possible status values returned are:
+ *   NV_OK
+ *   NV_ERR_NOT_SUPPORTED
+ */
+#define NV2080_CTRL_CMD_FLA_GET_RANGE (0x20803503) /* finn: Evaluated from "(FINN_NV20_SUBDEVICE_0_FLA_INTERFACE_ID << 8) | NV2080_CTRL_FLA_GET_RANGE_PARAMS_MESSAGE_ID" */
+
+#define NV2080_CTRL_FLA_GET_RANGE_PARAMS_MESSAGE_ID (0x3U)
+
+typedef struct NV2080_CTRL_FLA_GET_RANGE_PARAMS {
+    NV_DECLARE_ALIGNED(NvU64 base, 8);
+    NV_DECLARE_ALIGNED(NvU64 size, 8);
+} NV2080_CTRL_FLA_GET_RANGE_PARAMS;
+
+/*
+ * NV2080_CTRL_CMD_FLA_GET_FABRIC_MEM_STATS
+ *
+ * This command returns the total size and the free size of the fabric vaspace.
+ * Note: This returns the information for the FABRIC_VASPACE_A class.
+ *
+ *   totalSize[OUT]
+ *      - Total fabric vaspace.
+ *
+ *   freeSize [OUT]
+ *      - Available fabric vaspace.
+ *
+ * Possible status values returned are:
+ *
+ *   NV_OK
+ *   NV_ERR_NOT_SUPPORTED
+ */
+#define NV2080_CTRL_CMD_FLA_GET_FABRIC_MEM_STATS (0x20803504) /* finn: Evaluated from "(FINN_NV20_SUBDEVICE_0_FLA_INTERFACE_ID << 8) | NV2080_CTRL_FLA_GET_FABRIC_MEM_STATS_PARAMS_MESSAGE_ID" */
+
+#define NV2080_CTRL_FLA_GET_FABRIC_MEM_STATS_PARAMS_MESSAGE_ID (0x4U)
+
+typedef struct NV2080_CTRL_FLA_GET_FABRIC_MEM_STATS_PARAMS {
+    NV_DECLARE_ALIGNED(NvU64 totalSize, 8);
+    NV_DECLARE_ALIGNED(NvU64 freeSize, 8);
+} NV2080_CTRL_FLA_GET_FABRIC_MEM_STATS_PARAMS;
+
+// _ctrl2080fla_h_