Merge tag 'drm-next-2025-05-28' of https://gitlab.freedesktop.org/drm/kernel

Pull drm updates from Dave Airlie:
 "As part of building up nova-core/nova-drm pieces we've brought in some
  rust abstractions through this tree, aux bus being the main one, with
  devres changes also in the driver-core tree. Along with the drm core
  abstractions and enough nova-core/nova-drm to use them. This is still
  all stub work under construction, to build the nova driver upstream.

  The other big NVIDIA related one is nouveau adds support for
  Hopper/Blackwell GPUs, this required a new GSP firmware update to
  570.144, and a bunch of rework in order to support multiple fw
  interfaces.

  There is also the introduction of an asahi uapi header file as a
  precursor to getting the real driver in later, but to unblock
  userspace mesa packages while the driver is trapped behind rust
  enablement.

  Otherwise it's the usual mixture of stuff all over, amdgpu, i915/xe,
  and msm being the main ones, and some changes to vsprintf.

  new drivers:
   - bring in the asahi uapi header standalone
   - nova-drm: stub driver

  rust dependencies (for nova-core):
   - auxiliary
       - bus abstractions
       - driver registration
       - sample driver
   - devres changes from driver-core
   - revocable changes

  core:
   - add Apple fourcc modifiers
   - add virtio capset definitions
   - extend EXPORT_SYNC_FILE for timeline syncobjs
   - convert to devm_platform_ioremap_resource
   - refactor shmem helper page pinning
   - DP powerup/down link helpers
   - extended %p4cc in vsprintf.c to support fourcc prints
   - change vsprintf %p4cn to %p4chR, remove %p4cn
   - Add drm_file_err function
   - IN_FORMATS_ASYNC property
   - move sitronix from tiny to their own subdir

  rust:
   - add drm core infrastructure rust abstractions
     (device/driver, ioctl, file, gem)

  dma-buf:
   - adjust sg handling to not cache map on attach
   - allow setting dma-device for import
   - Add a helper to sort and deduplicate dma_fence arrays

  docs:
   - updated drm scheduler docs
   - fbdev todo update
   - fb rendering
   - actual brightness

  ttm:
   - fix delayed destroy resv object

  bridge:
   - add kunit tests
   - convert tc358775 to atomic
   - convert drivers to devm_drm_bridge_alloc
   - convert rk3066_hdmi to bridge driver

  scheduler:
   - add kunit tests

  panel:
   - refcount panels to improve lifetime handling
   - Powertip PH128800T004-ZZA01
   - NLT NL13676BC25-03F, Tianma TM070JDHG34-00
   - Himax HX8279/HX8279-D DDIC
   - Visionox G2647FB105
   - Sitronix ST7571
   - ZOTAC rotation quirk

  vkms:
   - allow attaching more displays

  i915:
   - xe3lpd display updates
   - vrr refactor
   - intel_display struct conversions
   - xe2hpd memory type identification
   - add link rate/count to i915_display_info
   - cleanup VGA plane handling
   - refactor HDCP GSC
   - fix SLPC wait boosting reference counting
   - add 20ms delay to engine reset
   - fix fence release on early probe errors

  xe:
   - SRIOV updates
   - BMG PCI ID update
   - support separate firmware for each GT
   - SVM fix, prelim SVM multi-device work
   - export fan speed
   - temp disable d3cold on BMG
   - backup VRAM in PM notifier instead of suspend/freeze
   - update xe_ttm_access_memory to use GPU for non-visible access
   - fix guc_info debugfs for VFs
   - use copy_from_user instead of __copy_from_user
   - append PCIe gen5 limitations to xe_firmware document

  amdgpu:
   - DSC cleanup
   - DC Scaling updates
   - Fused I2C-over-AUX updates
   - DMUB updates
   - Use drm_file_err in amdgpu
   - Enforce isolation updates
   - Use new dma_fence helpers
   - USERQ fixes
   - Documentation updates
   - SR-IOV updates
   - RAS updates
   - PSP 12 cleanups
   - GC 9.5 updates
   - SMU 13.x updates
   - VCN / JPEG SR-IOV updates

  amdkfd:
   - Update error messages for SDMA
   - Userptr updates
   - XNACK fixes

  radeon:
   - CIK doorbell cleanup

  nouveau:
   - add support for NVIDIA r570 GSP firmware
   - enable Hopper/Blackwell support

  nova-core:
   - fix task list
   - register definition infrastructure
   - move firmware into own rust module
   - register auxiliary device for nova-drm

  nova-drm:
   - initial driver skeleton

  msm:
   - GPU:
       - ACD (adaptive clock distribution) for X1-85
       - drop fictional address_space_size
       - improve GMU HFI response time out robustness
       - fix crash when throttling during boot
   - DPU:
       - use single CTL path for flushing on DPU 5.x+
       - improve SSPP allocation code for better sharing
       - Enabled SmartDMA on SM8150, SC8180X, SC8280XP, SM8550
       - Added SAR2130P support
       - Disabled DSC support on MSM8937, MSM8917, MSM8953, SDM660
   - DP:
       - switch to new audio helpers
       - better LTTPR handling
   - DSI:
       - Added support for SA8775P
       - Added SAR2130P support
   - HDMI:
       - Switched to use new helpers for ACR data
       - Fixed old standing issue of HPD not working in some cases

  amdxdna:
   - add dma-buf support
   - allow empty command submits

  renesas:
   - add dma-buf support
   - add zpos, alpha, blend support

  panthor:
   - fail properly for NO_MMAP bos
   - add SET_LABEL ioctl
   - debugfs BO dumping support

  imagination:
   - update DT bindings
   - support TI AM68 GPU

  hibmc:
   - improve interrupt handling and HPD support

  virtio:
   - add panic handler support

  rockchip:
   - add RK3588 support
   - add DP AUX bus panel support

  ivpu:
   - add heartbeat based hangcheck

  mediatek:
   - prepares support for MT8195/99 HDMIv2/DDCv2

  anx7625:
   - improve HPD

  tegra:
   - speed up firmware loading

* tag 'drm-next-2025-05-28' of https://gitlab.freedesktop.org/drm/kernel: (1627 commits)
  drm/nouveau/tegra: Fix error pointer vs NULL return in nvkm_device_tegra_resource_addr()
  drm/xe: Default auto_link_downgrade status to false
  drm/xe/guc: Make creation of SLPC debugfs files conditional
  drm/i915/display: Add check for alloc_ordered_workqueue() and alloc_workqueue()
  drm/i915/dp_mst: Work around Thunderbolt sink disconnect after SINK_COUNT_ESI read
  drm/i915/ptl: Use everywhere the correct DDI port clock select mask
  drm/nouveau/kms: add support for GB20x
  drm/dp: add option to disable zero sized address only transactions.
  drm/nouveau: add support for GB20x
  drm/nouveau/gsp: add hal for fifo.chan.doorbell_handle
  drm/nouveau: add support for GB10x
  drm/nouveau/gf100-: track chan progress with non-WFI semaphore release
  drm/nouveau/nv50-: separate CHANNEL_GPFIFO handling out from CHANNEL_DMA
  drm/nouveau: add helper functions for allocating pinned/cpu-mapped bos
  drm/nouveau: add support for GH100
  drm/nouveau: improve handling of 64-bit BARs
  drm/nouveau/gv100-: switch to volta semaphore methods
  drm/nouveau/gsp: support deeper page tables in COPY_SERVER_RESERVED_PDES
  drm/nouveau/gsp: init client VMMs with NV0080_CTRL_DMA_SET_PAGE_DIRECTORY
  drm/nouveau/gsp: fetch level shift and PDE from BAR2 VMM
  ...

10 files changed, 1702 insertions, 2 deletions

diff --git a/include/uapi/drm/amdgpu_drm.h b/include/uapi/drm/amdgpu_drm.h
index 25d5c6e90a99..45c4fa13499c 100644
--- a/include/uapi/drm/amdgpu_drm.h
+++ b/include/uapi/drm/amdgpu_drm.h
@@ -54,6 +54,9 @@ extern "C" {
 #define DRM_AMDGPU_VM			0x13
 #define DRM_AMDGPU_FENCE_TO_HANDLE	0x14
 #define DRM_AMDGPU_SCHED		0x15
+#define DRM_AMDGPU_USERQ		0x16
+#define DRM_AMDGPU_USERQ_SIGNAL		0x17
+#define DRM_AMDGPU_USERQ_WAIT		0x18
 
 #define DRM_IOCTL_AMDGPU_GEM_CREATE	DRM_IOWR(DRM_COMMAND_BASE + DRM_AMDGPU_GEM_CREATE, union drm_amdgpu_gem_create)
 #define DRM_IOCTL_AMDGPU_GEM_MMAP	DRM_IOWR(DRM_COMMAND_BASE + DRM_AMDGPU_GEM_MMAP, union drm_amdgpu_gem_mmap)
@@ -71,6 +74,9 @@ extern "C" {
 #define DRM_IOCTL_AMDGPU_VM		DRM_IOWR(DRM_COMMAND_BASE + DRM_AMDGPU_VM, union drm_amdgpu_vm)
 #define DRM_IOCTL_AMDGPU_FENCE_TO_HANDLE DRM_IOWR(DRM_COMMAND_BASE + DRM_AMDGPU_FENCE_TO_HANDLE, union drm_amdgpu_fence_to_handle)
 #define DRM_IOCTL_AMDGPU_SCHED		DRM_IOW(DRM_COMMAND_BASE + DRM_AMDGPU_SCHED, union drm_amdgpu_sched)
+#define DRM_IOCTL_AMDGPU_USERQ		DRM_IOWR(DRM_COMMAND_BASE + DRM_AMDGPU_USERQ, union drm_amdgpu_userq)
+#define DRM_IOCTL_AMDGPU_USERQ_SIGNAL	DRM_IOWR(DRM_COMMAND_BASE + DRM_AMDGPU_USERQ_SIGNAL, struct drm_amdgpu_userq_signal)
+#define DRM_IOCTL_AMDGPU_USERQ_WAIT	DRM_IOWR(DRM_COMMAND_BASE + DRM_AMDGPU_USERQ_WAIT, struct drm_amdgpu_userq_wait)
 
 /**
  * DOC: memory domains
@@ -319,6 +325,260 @@ union drm_amdgpu_ctx {
 	union drm_amdgpu_ctx_out out;
 };
 
+/* user queue IOCTL operations */
+#define AMDGPU_USERQ_OP_CREATE	1
+#define AMDGPU_USERQ_OP_FREE	2
+
+/* queue priority levels */
+/* low < normal low < normal high < high */
+#define AMDGPU_USERQ_CREATE_FLAGS_QUEUE_PRIORITY_MASK  0x3
+#define AMDGPU_USERQ_CREATE_FLAGS_QUEUE_PRIORITY_SHIFT 0
+#define AMDGPU_USERQ_CREATE_FLAGS_QUEUE_PRIORITY_NORMAL_LOW 0
+#define AMDGPU_USERQ_CREATE_FLAGS_QUEUE_PRIORITY_LOW 1
+#define AMDGPU_USERQ_CREATE_FLAGS_QUEUE_PRIORITY_NORMAL_HIGH 2
+#define AMDGPU_USERQ_CREATE_FLAGS_QUEUE_PRIORITY_HIGH 3 /* admin only */
+/* for queues that need access to protected content */
+#define AMDGPU_USERQ_CREATE_FLAGS_QUEUE_SECURE  (1 << 2)
+
+/*
+ * This structure is a container to pass input configuration
+ * info for all supported userqueue related operations.
+ * For operation AMDGPU_USERQ_OP_CREATE: user is expected
+ *  to set all fields, excep the parameter 'queue_id'.
+ * For operation AMDGPU_USERQ_OP_FREE: the only input parameter expected
+ *  to be set is 'queue_id', eveything else is ignored.
+ */
+struct drm_amdgpu_userq_in {
+	/** AMDGPU_USERQ_OP_* */
+	__u32	op;
+	/** Queue id passed for operation USERQ_OP_FREE */
+	__u32	queue_id;
+	/** the target GPU engine to execute workload (AMDGPU_HW_IP_*) */
+	__u32   ip_type;
+	/**
+	 * @doorbell_handle: the handle of doorbell GEM object
+	 * associated with this userqueue client.
+	 */
+	__u32   doorbell_handle;
+	/**
+	 * @doorbell_offset: 32-bit offset of the doorbell in the doorbell bo.
+	 * Kernel will generate absolute doorbell offset using doorbell_handle
+	 * and doorbell_offset in the doorbell bo.
+	 */
+	__u32   doorbell_offset;
+	/**
+	 * @flags: flags used for queue parameters
+	 */
+	__u32 flags;
+	/**
+	 * @queue_va: Virtual address of the GPU memory which holds the queue
+	 * object. The queue holds the workload packets.
+	 */
+	__u64   queue_va;
+	/**
+	 * @queue_size: Size of the queue in bytes, this needs to be 256-byte
+	 * aligned.
+	 */
+	__u64   queue_size;
+	/**
+	 * @rptr_va : Virtual address of the GPU memory which holds the ring RPTR.
+	 * This object must be at least 8 byte in size and aligned to 8-byte offset.
+	 */
+	__u64   rptr_va;
+	/**
+	 * @wptr_va : Virtual address of the GPU memory which holds the ring WPTR.
+	 * This object must be at least 8 byte in size and aligned to 8-byte offset.
+	 *
+	 * Queue, RPTR and WPTR can come from the same object, as long as the size
+	 * and alignment related requirements are met.
+	 */
+	__u64   wptr_va;
+	/**
+	 * @mqd: MQD (memory queue descriptor) is a set of parameters which allow
+	 * the GPU to uniquely define and identify a usermode queue.
+	 *
+	 * MQD data can be of different size for different GPU IP/engine and
+	 * their respective versions/revisions, so this points to a __u64 *
+	 * which holds IP specific MQD of this usermode queue.
+	 */
+	__u64 mqd;
+	/**
+	 * @size: size of MQD data in bytes, it must match the MQD structure
+	 * size of the respective engine/revision defined in UAPI for ex, for
+	 * gfx11 workloads, size = sizeof(drm_amdgpu_userq_mqd_gfx11).
+	 */
+	__u64 mqd_size;
+};
+
+/* The structure to carry output of userqueue ops */
+struct drm_amdgpu_userq_out {
+	/**
+	 * For operation AMDGPU_USERQ_OP_CREATE: This field contains a unique
+	 * queue ID to represent the newly created userqueue in the system, otherwise
+	 * it should be ignored.
+	 */
+	__u32	queue_id;
+	__u32 _pad;
+};
+
+union drm_amdgpu_userq {
+	struct drm_amdgpu_userq_in in;
+	struct drm_amdgpu_userq_out out;
+};
+
+/* GFX V11 IP specific MQD parameters */
+struct drm_amdgpu_userq_mqd_gfx11 {
+	/**
+	 * @shadow_va: Virtual address of the GPU memory to hold the shadow buffer.
+	 * Use AMDGPU_INFO_IOCTL to find the exact size of the object.
+	 */
+	__u64   shadow_va;
+	/**
+	 * @csa_va: Virtual address of the GPU memory to hold the CSA buffer.
+	 * Use AMDGPU_INFO_IOCTL to find the exact size of the object.
+	 */
+	__u64   csa_va;
+};
+
+/* GFX V11 SDMA IP specific MQD parameters */
+struct drm_amdgpu_userq_mqd_sdma_gfx11 {
+	/**
+	 * @csa_va: Virtual address of the GPU memory to hold the CSA buffer.
+	 * This must be a from a separate GPU object, and use AMDGPU_INFO IOCTL
+	 * to get the size.
+	 */
+	__u64   csa_va;
+};
+
+/* GFX V11 Compute IP specific MQD parameters */
+struct drm_amdgpu_userq_mqd_compute_gfx11 {
+	/**
+	 * @eop_va: Virtual address of the GPU memory to hold the EOP buffer.
+	 * This must be a from a separate GPU object, and use AMDGPU_INFO IOCTL
+	 * to get the size.
+	 */
+	__u64   eop_va;
+};
+
+/* userq signal/wait ioctl */
+struct drm_amdgpu_userq_signal {
+	/**
+	 * @queue_id: Queue handle used by the userq fence creation function
+	 * to retrieve the WPTR.
+	 */
+	__u32	queue_id;
+	__u32	pad;
+	/**
+	 * @syncobj_handles: The list of syncobj handles submitted by the user queue
+	 * job to be signaled.
+	 */
+	__u64	syncobj_handles;
+	/**
+	 * @num_syncobj_handles: A count that represents the number of syncobj handles in
+	 * @syncobj_handles.
+	 */
+	__u64	num_syncobj_handles;
+	/**
+	 * @bo_read_handles: The list of BO handles that the submitted user queue job
+	 * is using for read only. This will update BO fences in the kernel.
+	 */
+	__u64	bo_read_handles;
+	/**
+	 * @bo_write_handles: The list of BO handles that the submitted user queue job
+	 * is using for write only. This will update BO fences in the kernel.
+	 */
+	__u64	bo_write_handles;
+	/**
+	 * @num_bo_read_handles: A count that represents the number of read BO handles in
+	 * @bo_read_handles.
+	 */
+	__u32	num_bo_read_handles;
+	/**
+	 * @num_bo_write_handles: A count that represents the number of write BO handles in
+	 * @bo_write_handles.
+	 */
+	__u32	num_bo_write_handles;
+};
+
+struct drm_amdgpu_userq_fence_info {
+	/**
+	 * @va: A gpu address allocated for each queue which stores the
+	 * read pointer (RPTR) value.
+	 */
+	__u64	va;
+	/**
+	 * @value: A 64 bit value represents the write pointer (WPTR) of the
+	 * queue commands which compared with the RPTR value to signal the
+	 * fences.
+	 */
+	__u64	value;
+};
+
+struct drm_amdgpu_userq_wait {
+	/**
+	 * @waitq_id: Queue handle used by the userq wait IOCTL to retrieve the
+	 * wait queue and maintain the fence driver references in it.
+	 */
+	__u32	waitq_id;
+	__u32	pad;
+	/**
+	 * @syncobj_handles: The list of syncobj handles submitted by the user queue
+	 * job to get the va/value pairs.
+	 */
+	__u64	syncobj_handles;
+	/**
+	 * @syncobj_timeline_handles: The list of timeline syncobj handles submitted by
+	 * the user queue job to get the va/value pairs at given @syncobj_timeline_points.
+	 */
+	__u64	syncobj_timeline_handles;
+	/**
+	 * @syncobj_timeline_points: The list of timeline syncobj points submitted by the
+	 * user queue job for the corresponding @syncobj_timeline_handles.
+	 */
+	__u64	syncobj_timeline_points;
+	/**
+	 * @bo_read_handles: The list of read BO handles submitted by the user queue
+	 * job to get the va/value pairs.
+	 */
+	__u64	bo_read_handles;
+	/**
+	 * @bo_write_handles: The list of write BO handles submitted by the user queue
+	 * job to get the va/value pairs.
+	 */
+	__u64	bo_write_handles;
+	/**
+	 * @num_syncobj_timeline_handles: A count that represents the number of timeline
+	 * syncobj handles in @syncobj_timeline_handles.
+	 */
+	__u16	num_syncobj_timeline_handles;
+	/**
+	 * @num_fences: This field can be used both as input and output. As input it defines
+	 * the maximum number of fences that can be returned and as output it will specify
+	 * how many fences were actually returned from the ioctl.
+	 */
+	__u16	num_fences;
+	/**
+	 * @num_syncobj_handles: A count that represents the number of syncobj handles in
+	 * @syncobj_handles.
+	 */
+	__u32	num_syncobj_handles;
+	/**
+	 * @num_bo_read_handles: A count that represents the number of read BO handles in
+	 * @bo_read_handles.
+	 */
+	__u32	num_bo_read_handles;
+	/**
+	 * @num_bo_write_handles: A count that represents the number of write BO handles in
+	 * @bo_write_handles.
+	 */
+	__u32	num_bo_write_handles;
+	/**
+	 * @out_fences: The field is a return value from the ioctl containing the list of
+	 * address/value pairs to wait for.
+	 */
+	__u64	out_fences;
+};
+
 /* vm ioctl */
 #define AMDGPU_VM_OP_RESERVE_VMID	1
 #define AMDGPU_VM_OP_UNRESERVE_VMID	2
@@ -599,6 +859,19 @@ struct drm_amdgpu_gem_va {
 	__u64 offset_in_bo;
 	/** Specify mapping size. Must be correctly aligned. */
 	__u64 map_size;
+	/**
+	 * vm_timeline_point is a sequence number used to add new timeline point.
+	 */
+	__u64 vm_timeline_point;
+	/**
+	 * The vm page table update fence is installed in given vm_timeline_syncobj_out
+	 * at vm_timeline_point.
+	 */
+	__u32 vm_timeline_syncobj_out;
+	/** the number of syncobj handles in @input_fence_syncobj_handles */
+	__u32 num_syncobj_handles;
+	/** Array of sync object handle to wait for given input fences */
+	__u64 input_fence_syncobj_handles;
 };
 
 #define AMDGPU_HW_IP_GFX          0
@@ -940,6 +1213,8 @@ struct drm_amdgpu_cs_chunk_cp_gfx_shadow {
 #define AMDGPU_INFO_MAX_IBS			0x22
 /* query last page fault info */
 #define AMDGPU_INFO_GPUVM_FAULT			0x23
+/* query FW object size and alignment */
+#define AMDGPU_INFO_UQ_FW_AREAS			0x24
 
 #define AMDGPU_INFO_MMR_SE_INDEX_SHIFT	0
 #define AMDGPU_INFO_MMR_SE_INDEX_MASK	0xff
@@ -1093,6 +1368,7 @@ struct drm_amdgpu_info_vbios {
 #define AMDGPU_VRAM_TYPE_DDR5  10
 #define AMDGPU_VRAM_TYPE_LPDDR4 11
 #define AMDGPU_VRAM_TYPE_LPDDR5 12
+#define AMDGPU_VRAM_TYPE_HBM3E 13
 
 struct drm_amdgpu_info_device {
 	/** PCI Device ID */
@@ -1198,6 +1474,9 @@ struct drm_amdgpu_info_device {
 	__u32 csa_size;
 	/* context save area base virtual alignment for gfx11 */
 	__u32 csa_alignment;
+	/* Userq IP mask (1 << AMDGPU_HW_IP_*) */
+	__u32 userq_ip_mask;
+	__u32 pad;
 };
 
 struct drm_amdgpu_info_hw_ip {
@@ -1216,6 +1495,27 @@ struct drm_amdgpu_info_hw_ip {
 	__u32  ip_discovery_version;
 };
 
+/* GFX metadata BO sizes and alignment info (in bytes) */
+struct drm_amdgpu_info_uq_fw_areas_gfx {
+	/* shadow area size */
+	__u32 shadow_size;
+	/* shadow area base virtual mem alignment */
+	__u32 shadow_alignment;
+	/* context save area size */
+	__u32 csa_size;
+	/* context save area base virtual mem alignment */
+	__u32 csa_alignment;
+};
+
+/* IP specific fw related information used in the
+ * subquery AMDGPU_INFO_UQ_FW_AREAS
+ */
+struct drm_amdgpu_info_uq_fw_areas {
+	union {
+		struct drm_amdgpu_info_uq_fw_areas_gfx gfx;
+	};
+};
+
 struct drm_amdgpu_info_num_handles {
 	/** Max handles as supported by firmware for UVD */
 	__u32  uvd_max_handles;
@@ -1279,6 +1579,23 @@ struct drm_amdgpu_info_gpuvm_fault {
 	__u32 vmhub;
 };
 
+struct drm_amdgpu_info_uq_metadata_gfx {
+	/* shadow area size for gfx11 */
+	__u32 shadow_size;
+	/* shadow area base virtual alignment for gfx11 */
+	__u32 shadow_alignment;
+	/* context save area size for gfx11 */
+	__u32 csa_size;
+	/* context save area base virtual alignment for gfx11 */
+	__u32 csa_alignment;
+};
+
+struct drm_amdgpu_info_uq_metadata {
+	union {
+		struct drm_amdgpu_info_uq_metadata_gfx gfx;
+	};
+};
+
 /*
  * Supported GPU families
  */
diff --git a/include/uapi/drm/asahi_drm.h b/include/uapi/drm/asahi_drm.h
new file mode 100644
index 000000000000..de67f1c603af
--- /dev/null
+++ b/include/uapi/drm/asahi_drm.h
@@ -0,0 +1,1194 @@
+/* SPDX-License-Identifier: MIT */
+/*
+ * Copyright (C) The Asahi Linux Contributors
+ * Copyright (C) 2018-2023 Collabora Ltd.
+ * Copyright (C) 2014-2018 Broadcom
+ */
+#ifndef _ASAHI_DRM_H_
+#define _ASAHI_DRM_H_
+
+#include "drm.h"
+
+#if defined(__cplusplus)
+extern "C" {
+#endif
+
+/**
+ * DOC: Introduction to the Asahi UAPI
+ *
+ * This documentation describes the Asahi IOCTLs.
+ *
+ * Just a few generic rules about the data passed to the Asahi IOCTLs (cribbed
+ * from Panthor):
+ *
+ * - Structures must be aligned on 64-bit/8-byte. If the object is not
+ *   naturally aligned, a padding field must be added.
+ * - Fields must be explicitly aligned to their natural type alignment with
+ *   pad[0..N] fields.
+ * - All padding fields will be checked by the driver to make sure they are
+ *   zeroed.
+ * - Flags can be added, but not removed/replaced.
+ * - New fields can be added to the main structures (the structures
+ *   directly passed to the ioctl). Those fields can be added at the end of
+ *   the structure, or replace existing padding fields. Any new field being
+ *   added must preserve the behavior that existed before those fields were
+ *   added when a value of zero is passed.
+ * - New fields can be added to indirect objects (objects pointed by the
+ *   main structure), iff those objects are passed a size to reflect the
+ *   size known by the userspace driver (see
+ *   drm_asahi_cmd_header::size).
+ * - If the kernel driver is too old to know some fields, those will be
+ *   ignored if zero, and otherwise rejected (and so will be zero on output).
+ * - If userspace is too old to know some fields, those will be zeroed
+ *   (input) before the structure is parsed by the kernel driver.
+ * - Each new flag/field addition must come with a driver version update so
+ *   the userspace driver doesn't have to guess which flags are supported.
+ * - Structures should not contain unions, as this would defeat the
+ *   extensibility of such structures.
+ * - IOCTLs can't be removed or replaced. New IOCTL IDs should be placed
+ *   at the end of the drm_asahi_ioctl_id enum.
+ */
+
+/**
+ * enum drm_asahi_ioctl_id - IOCTL IDs
+ *
+ * Place new ioctls at the end, don't re-order, don't replace or remove entries.
+ *
+ * These IDs are not meant to be used directly. Use the DRM_IOCTL_ASAHI_xxx
+ * definitions instead.
+ */
+enum drm_asahi_ioctl_id {
+	/** @DRM_ASAHI_GET_PARAMS: Query device properties. */
+	DRM_ASAHI_GET_PARAMS = 0,
+
+	/** @DRM_ASAHI_GET_TIME: Query device time. */
+	DRM_ASAHI_GET_TIME,
+
+	/** @DRM_ASAHI_VM_CREATE: Create a GPU VM address space. */
+	DRM_ASAHI_VM_CREATE,
+
+	/** @DRM_ASAHI_VM_DESTROY: Destroy a VM. */
+	DRM_ASAHI_VM_DESTROY,
+
+	/** @DRM_ASAHI_VM_BIND: Bind/unbind memory to a VM. */
+	DRM_ASAHI_VM_BIND,
+
+	/** @DRM_ASAHI_GEM_CREATE: Create a buffer object. */
+	DRM_ASAHI_GEM_CREATE,
+
+	/**
+	 * @DRM_ASAHI_GEM_MMAP_OFFSET: Get offset to pass to mmap() to map a
+	 * given GEM handle.
+	 */
+	DRM_ASAHI_GEM_MMAP_OFFSET,
+
+	/** @DRM_ASAHI_GEM_BIND_OBJECT: Bind memory as a special object */
+	DRM_ASAHI_GEM_BIND_OBJECT,
+
+	/** @DRM_ASAHI_QUEUE_CREATE: Create a scheduling queue. */
+	DRM_ASAHI_QUEUE_CREATE,
+
+	/** @DRM_ASAHI_QUEUE_DESTROY: Destroy a scheduling queue. */
+	DRM_ASAHI_QUEUE_DESTROY,
+
+	/** @DRM_ASAHI_SUBMIT: Submit commands to a queue. */
+	DRM_ASAHI_SUBMIT,
+};
+
+#define DRM_ASAHI_MAX_CLUSTERS	64
+
+/**
+ * struct drm_asahi_params_global - Global parameters.
+ *
+ * This struct may be queried by drm_asahi_get_params.
+ */
+struct drm_asahi_params_global {
+	/** @features: Feature bits from drm_asahi_feature */
+	__u64 features;
+
+	/** @gpu_generation: GPU generation, e.g. 13 for G13G */
+	__u32 gpu_generation;
+
+	/** @gpu_variant: GPU variant as a character, e.g. 'C' for G13C */
+	__u32 gpu_variant;
+
+	/**
+	 * @gpu_revision: GPU revision in BCD, e.g. 0x00 for 'A0' or
+	 * 0x21 for 'C1'
+	 */
+	__u32 gpu_revision;
+
+	/** @chip_id: Chip ID in BCD, e.g. 0x8103 for T8103 */
+	__u32 chip_id;
+
+	/** @num_dies: Number of dies in the SoC */
+	__u32 num_dies;
+
+	/** @num_clusters_total: Number of GPU clusters (across all dies) */
+	__u32 num_clusters_total;
+
+	/**
+	 * @num_cores_per_cluster: Number of logical cores per cluster
+	 * (including inactive/nonexistent)
+	 */
+	__u32 num_cores_per_cluster;
+
+	/** @max_frequency_khz: Maximum GPU core clock frequency */
+	__u32 max_frequency_khz;
+
+	/** @core_masks: Bitmask of present/enabled cores per cluster */
+	__u64 core_masks[DRM_ASAHI_MAX_CLUSTERS];
+
+	/**
+	 * @vm_start: VM range start VMA. Together with @vm_end, this defines
+	 * the window of valid GPU VAs. Userspace is expected to subdivide VAs
+	 * out of this window.
+	 *
+	 * This window contains all virtual addresses that userspace needs to
+	 * know about. There may be kernel-internal GPU VAs outside this range,
+	 * but that detail is not relevant here.
+	 */
+	__u64 vm_start;
+
+	/** @vm_end: VM range end VMA */
+	__u64 vm_end;
+
+	/**
+	 * @vm_kernel_min_size: Minimum kernel VMA window size.
+	 *
+	 * When creating a VM, userspace is required to carve out a section of
+	 * virtual addresses (within the range given by @vm_start and
+	 * @vm_end). The kernel will allocate various internal structures
+	 * within the specified VA range.
+	 *
+	 * Allowing userspace to choose the VA range for the kernel, rather than
+	 * the kernel reserving VAs and requiring userspace to cope, can assist
+	 * in implementing SVM.
+	 */
+	__u64 vm_kernel_min_size;
+
+	/**
+	 * @max_commands_per_submission: Maximum number of supported commands
+	 * per submission. This mirrors firmware limits. Userspace must split up
+	 * larger command buffers, which may require inserting additional
+	 * synchronization.
+	 */
+	__u32 max_commands_per_submission;
+
+	/**
+	 * @max_attachments: Maximum number of drm_asahi_attachment's per
+	 * command
+	 */
+	__u32 max_attachments;
+
+	/**
+	 * @command_timestamp_frequency_hz: Timebase frequency for timestamps
+	 * written during command execution, specified via drm_asahi_timestamp
+	 * structures. As this rate is controlled by the firmware, it is a
+	 * queryable parameter.
+	 *
+	 * Userspace must divide by this frequency to convert timestamps to
+	 * seconds, rather than hardcoding a particular firmware's rate.
+	 */
+	__u64 command_timestamp_frequency_hz;
+};
+
+/**
+ * enum drm_asahi_feature - Feature bits
+ *
+ * This covers only features that userspace cannot infer from the architecture
+ * version. Most features don't need to be here.
+ */
+enum drm_asahi_feature {
+	/**
+	 * @DRM_ASAHI_FEATURE_SOFT_FAULTS: GPU has "soft fault" enabled. Shader
+	 * loads of unmapped memory will return zero. Shader stores to unmapped
+	 * memory will be silently discarded. Note that only shader load/store
+	 * is affected. Other hardware units are not affected, notably including
+	 * texture sampling.
+	 *
+	 * Soft fault is set when initializing the GPU and cannot be runtime
+	 * toggled. Therefore, it is exposed as a feature bit and not a
+	 * userspace-settable flag on the VM. When soft fault is enabled,
+	 * userspace can speculate memory accesses more aggressively.
+	 */
+	DRM_ASAHI_FEATURE_SOFT_FAULTS = (1UL) << 0,
+};
+
+/**
+ * struct drm_asahi_get_params - Arguments passed to DRM_IOCTL_ASAHI_GET_PARAMS
+ */
+struct drm_asahi_get_params {
+	/** @param_group: Parameter group to fetch (MBZ) */
+	__u32 param_group;
+
+	/** @pad: MBZ */
+	__u32 pad;
+
+	/** @pointer: User pointer to write parameter struct */
+	__u64 pointer;
+
+	/**
+	 * @size: Size of the user buffer. In case of older userspace, this may
+	 * be less than sizeof(struct drm_asahi_params_global). The kernel will
+	 * not write past the length specified here, allowing extensibility.
+	 */
+	__u64 size;
+};
+
+/**
+ * struct drm_asahi_vm_create - Arguments passed to DRM_IOCTL_ASAHI_VM_CREATE
+ */
+struct drm_asahi_vm_create {
+	/**
+	 * @kernel_start: Start of the kernel-reserved address range. See
+	 * drm_asahi_params_global::vm_kernel_min_size.
+	 *
+	 * Both @kernel_start and @kernel_end must be within the range of
+	 * valid VAs given by drm_asahi_params_global::vm_start and
+	 * drm_asahi_params_global::vm_end. The size of the kernel range
+	 * (@kernel_end - @kernel_start) must be at least
+	 * drm_asahi_params_global::vm_kernel_min_size.
+	 *
+	 * Userspace must not bind any memory on this VM into this reserved
+	 * range, it is for kernel use only.
+	 */
+	__u64 kernel_start;
+
+	/**
+	 * @kernel_end: End of the kernel-reserved address range. See
+	 * @kernel_start.
+	 */
+	__u64 kernel_end;
+
+	/** @vm_id: Returned VM ID */
+	__u32 vm_id;
+
+	/** @pad: MBZ */
+	__u32 pad;
+};
+
+/**
+ * struct drm_asahi_vm_destroy - Arguments passed to DRM_IOCTL_ASAHI_VM_DESTROY
+ */
+struct drm_asahi_vm_destroy {
+	/** @vm_id: VM ID to be destroyed */
+	__u32 vm_id;
+
+	/** @pad: MBZ */
+	__u32 pad;
+};
+
+/**
+ * enum drm_asahi_gem_flags - Flags for GEM creation
+ */
+enum drm_asahi_gem_flags {
+	/**
+	 * @DRM_ASAHI_GEM_WRITEBACK: BO should be CPU-mapped as writeback.
+	 *
+	 * Map as writeback instead of write-combine. This optimizes for CPU
+	 * reads.
+	 */
+	DRM_ASAHI_GEM_WRITEBACK = (1L << 0),
+
+	/**
+	 * @DRM_ASAHI_GEM_VM_PRIVATE: BO is private to this GPU VM (no exports).
+	 */
+	DRM_ASAHI_GEM_VM_PRIVATE = (1L << 1),
+};
+
+/**
+ * struct drm_asahi_gem_create - Arguments passed to DRM_IOCTL_ASAHI_GEM_CREATE
+ */
+struct drm_asahi_gem_create {
+	/** @size: Size of the BO */
+	__u64 size;
+
+	/** @flags: Combination of drm_asahi_gem_flags flags. */
+	__u32 flags;
+
+	/**
+	 * @vm_id: VM ID to assign to the BO, if DRM_ASAHI_GEM_VM_PRIVATE is set
+	 */
+	__u32 vm_id;
+
+	/** @handle: Returned GEM handle for the BO */
+	__u32 handle;
+
+	/** @pad: MBZ */
+	__u32 pad;
+};
+
+/**
+ * struct drm_asahi_gem_mmap_offset - Arguments passed to
+ * DRM_IOCTL_ASAHI_GEM_MMAP_OFFSET
+ */
+struct drm_asahi_gem_mmap_offset {
+	/** @handle: Handle for the object being mapped. */
+	__u32 handle;
+
+	/** @flags: Must be zero */
+	__u32 flags;
+
+	/** @offset: The fake offset to use for subsequent mmap call */
+	__u64 offset;
+};
+
+/**
+ * enum drm_asahi_bind_flags - Flags for GEM binding
+ */
+enum drm_asahi_bind_flags {
+	/**
+	 * @DRM_ASAHI_BIND_UNBIND: Instead of binding a GEM object to the range,
+	 * simply unbind the GPU VMA range.
+	 */
+	DRM_ASAHI_BIND_UNBIND = (1L << 0),
+
+	/** @DRM_ASAHI_BIND_READ: Map BO with GPU read permission */
+	DRM_ASAHI_BIND_READ = (1L << 1),
+
+	/** @DRM_ASAHI_BIND_WRITE: Map BO with GPU write permission */
+	DRM_ASAHI_BIND_WRITE = (1L << 2),
+
+	/**
+	 * @DRM_ASAHI_BIND_SINGLE_PAGE: Map a single page of the BO repeatedly
+	 * across the VA range.
+	 *
+	 * This is useful to fill a VA range with scratch pages or zero pages.
+	 * It is intended as a mechanism to accelerate sparse.
+	 */
+	DRM_ASAHI_BIND_SINGLE_PAGE = (1L << 3),
+};
+
+/**
+ * struct drm_asahi_gem_bind_op - Description of a single GEM bind operation.
+ */
+struct drm_asahi_gem_bind_op {
+	/** @flags: Combination of drm_asahi_bind_flags flags. */
+	__u32 flags;
+
+	/** @handle: GEM object to bind (except for UNBIND) */
+	__u32 handle;
+
+	/**
+	 * @offset: Offset into the object (except for UNBIND).
+	 *
+	 * For a regular bind, this is the beginning of the region of the GEM
+	 * object to bind.
+	 *
+	 * For a single-page bind, this is the offset to the single page that
+	 * will be repeatedly bound.
+	 *
+	 * Must be page-size aligned.
+	 */
+	__u64 offset;
+
+	/**
+	 * @range: Number of bytes to bind/unbind to @addr.
+	 *
+	 * Must be page-size aligned.
+	 */
+	__u64 range;
+
+	/**
+	 * @addr: Address to bind to.
+	 *
+	 * Must be page-size aligned.
+	 */
+	__u64 addr;
+};
+
+/**
+ * struct drm_asahi_vm_bind - Arguments passed to
+ * DRM_IOCTL_ASAHI_VM_BIND
+ */
+struct drm_asahi_vm_bind {
+	/** @vm_id: The ID of the VM to bind to */
+	__u32 vm_id;
+
+	/** @num_binds: number of binds in this IOCTL. */
+	__u32 num_binds;
+
+	/**
+	 * @stride: Stride in bytes between consecutive binds. This allows
+	 * extensibility of drm_asahi_gem_bind_op.
+	 */
+	__u32 stride;
+
+	/** @pad: MBZ */
+	__u32 pad;
+
+	/**
+	 * @userptr: User pointer to an array of @num_binds structures of type
+	 * @drm_asahi_gem_bind_op and size @stride bytes.
+	 */
+	__u64 userptr;
+};
+
+/**
+ * enum drm_asahi_bind_object_op - Special object bind operation
+ */
+enum drm_asahi_bind_object_op {
+	/** @DRM_ASAHI_BIND_OBJECT_OP_BIND: Bind a BO as a special GPU object */
+	DRM_ASAHI_BIND_OBJECT_OP_BIND = 0,
+
+	/** @DRM_ASAHI_BIND_OBJECT_OP_UNBIND: Unbind a special GPU object */
+	DRM_ASAHI_BIND_OBJECT_OP_UNBIND = 1,
+};
+
+/**
+ * enum drm_asahi_bind_object_flags - Special object bind flags
+ */
+enum drm_asahi_bind_object_flags {
+	/**
+	 * @DRM_ASAHI_BIND_OBJECT_USAGE_TIMESTAMPS: Map a BO as a timestamp
+	 * buffer.
+	 */
+	DRM_ASAHI_BIND_OBJECT_USAGE_TIMESTAMPS = (1L << 0),
+};
+
+/**
+ * struct drm_asahi_gem_bind_object - Arguments passed to
+ * DRM_IOCTL_ASAHI_GEM_BIND_OBJECT
+ */
+struct drm_asahi_gem_bind_object {
+	/** @op: Bind operation (enum drm_asahi_bind_object_op) */
+	__u32 op;
+
+	/** @flags: Combination of drm_asahi_bind_object_flags flags. */
+	__u32 flags;
+
+	/** @handle: GEM object to bind/unbind (BIND) */
+	__u32 handle;
+
+	/** @vm_id: The ID of the VM to operate on (MBZ currently) */
+	__u32 vm_id;
+
+	/** @offset: Offset into the object (BIND only) */
+	__u64 offset;
+
+	/** @range: Number of bytes to bind/unbind (BIND only) */
+	__u64 range;
+
+	/** @object_handle: Object handle (out for BIND, in for UNBIND) */
+	__u32 object_handle;
+
+	/** @pad: MBZ */
+	__u32 pad;
+};
+
+/**
+ * enum drm_asahi_cmd_type - Command type
+ */
+enum drm_asahi_cmd_type {
+	/**
+	 * @DRM_ASAHI_CMD_RENDER: Render command, executing on the render
+	 * subqueue. Combined vertex and fragment operation.
+	 *
+	 * Followed by a @drm_asahi_cmd_render payload.
+	 */
+	DRM_ASAHI_CMD_RENDER = 0,
+
+	/**
+	 * @DRM_ASAHI_CMD_COMPUTE: Compute command on the compute subqueue.
+	 *
+	 * Followed by a @drm_asahi_cmd_compute payload.
+	 */
+	DRM_ASAHI_CMD_COMPUTE = 1,
+
+	/**
+	 * @DRM_ASAHI_SET_VERTEX_ATTACHMENTS: Software command to set
+	 * attachments for subsequent vertex shaders in the same submit.
+	 *
+	 * Followed by (possibly multiple) @drm_asahi_attachment payloads.
+	 */
+	DRM_ASAHI_SET_VERTEX_ATTACHMENTS = 2,
+
+	/**
+	 * @DRM_ASAHI_SET_FRAGMENT_ATTACHMENTS: Software command to set
+	 * attachments for subsequent fragment shaders in the same submit.
+	 *
+	 * Followed by (possibly multiple) @drm_asahi_attachment payloads.
+	 */
+	DRM_ASAHI_SET_FRAGMENT_ATTACHMENTS = 3,
+
+	/**
+	 * @DRM_ASAHI_SET_COMPUTE_ATTACHMENTS: Software command to set
+	 * attachments for subsequent compute shaders in the same submit.
+	 *
+	 * Followed by (possibly multiple) @drm_asahi_attachment payloads.
+	 */
+	DRM_ASAHI_SET_COMPUTE_ATTACHMENTS = 4,
+};
+
+/**
+ * enum drm_asahi_priority - Scheduling queue priority.
+ *
+ * These priorities are forwarded to the firmware to influence firmware
+ * scheduling. The exact policy is ultimately decided by firmware, but
+ * these enums allow userspace to communicate the intentions.
+ */
+enum drm_asahi_priority {
+	/** @DRM_ASAHI_PRIORITY_LOW: Low priority queue. */
+	DRM_ASAHI_PRIORITY_LOW = 0,
+
+	/** @DRM_ASAHI_PRIORITY_MEDIUM: Medium priority queue. */
+	DRM_ASAHI_PRIORITY_MEDIUM = 1,
+
+	/**
+	 * @DRM_ASAHI_PRIORITY_HIGH: High priority queue.
+	 *
+	 * Reserved for future extension.
+	 */
+	DRM_ASAHI_PRIORITY_HIGH = 2,
+
+	/**
+	 * @DRM_ASAHI_PRIORITY_REALTIME: Real-time priority queue.
+	 *
+	 * Reserved for future extension.
+	 */
+	DRM_ASAHI_PRIORITY_REALTIME = 3,
+};
+
+/**
+ * struct drm_asahi_queue_create - Arguments passed to
+ * DRM_IOCTL_ASAHI_QUEUE_CREATE
+ */
+struct drm_asahi_queue_create {
+	/** @flags: MBZ */
+	__u32 flags;
+
+	/** @vm_id: The ID of the VM this queue is bound to */
+	__u32 vm_id;
+
+	/** @priority: One of drm_asahi_priority */
+	__u32 priority;
+
+	/** @queue_id: The returned queue ID */
+	__u32 queue_id;
+
+	/**
+	 * @usc_exec_base: GPU base address for all USC binaries (shaders) on
+	 * this queue. USC addresses are 32-bit relative to this 64-bit base.
+	 *
+	 * This sets the following registers on all queue commands:
+	 *
+	 *	USC_EXEC_BASE_TA  (vertex)
+	 *	USC_EXEC_BASE_ISP (fragment)
+	 *	USC_EXEC_BASE_CP  (compute)
+	 *
+	 * While the hardware lets us configure these independently per command,
+	 * we do not have a use case for this. Instead, we expect userspace to
+	 * fix a 4GiB VA carveout for USC memory and pass its base address here.
+	 */
+	__u64 usc_exec_base;
+};
+
+/**
+ * struct drm_asahi_queue_destroy - Arguments passed to
+ * DRM_IOCTL_ASAHI_QUEUE_DESTROY
+ */
+struct drm_asahi_queue_destroy {
+	/** @queue_id: The queue ID to be destroyed */
+	__u32 queue_id;
+
+	/** @pad: MBZ */
+	__u32 pad;
+};
+
+/**
+ * enum drm_asahi_sync_type - Sync item type
+ */
+enum drm_asahi_sync_type {
+	/** @DRM_ASAHI_SYNC_SYNCOBJ: Binary sync object */
+	DRM_ASAHI_SYNC_SYNCOBJ = 0,
+
+	/** @DRM_ASAHI_SYNC_TIMELINE_SYNCOBJ: Timeline sync object */
+	DRM_ASAHI_SYNC_TIMELINE_SYNCOBJ = 1,
+};
+
+/**
+ * struct drm_asahi_sync - Sync item
+ */
+struct drm_asahi_sync {
+	/** @sync_type: One of drm_asahi_sync_type */
+	__u32 sync_type;
+
+	/** @handle: The sync object handle */
+	__u32 handle;
+
+	/** @timeline_value: Timeline value for timeline sync objects */
+	__u64 timeline_value;
+};
+
+/**
+ * define DRM_ASAHI_BARRIER_NONE - Command index for no barrier
+ *
+ * This special value may be passed in to drm_asahi_command::vdm_barrier or
+ * drm_asahi_command::cdm_barrier to indicate that the respective subqueue
+ * should not wait on any previous work.
+ */
+#define DRM_ASAHI_BARRIER_NONE (0xFFFFu)
+
+/**
+ * struct drm_asahi_cmd_header - Top level command structure
+ *
+ * This struct is core to the command buffer definition and therefore is not
+ * extensible.
+ */
+struct drm_asahi_cmd_header {
+	/** @cmd_type: One of drm_asahi_cmd_type */
+	__u16 cmd_type;
+
+	/**
+	 * @size: Size of this command, not including this header.
+	 *
+	 * For hardware commands, this enables extensibility of commands without
+	 * requiring extra command types. Passing a command that is shorter
+	 * than expected is explicitly allowed for backwards-compatibility.
+	 * Truncated fields will be zeroed.
+	 *
+	 * For the synthetic attachment setting commands, this implicitly
+	 * encodes the number of attachments. These commands take multiple
+	 * fixed-size @drm_asahi_attachment structures as their payload, so size
+	 * equals number of attachments * sizeof(struct drm_asahi_attachment).
+	 */
+	__u16 size;
+
+	/**
+	 * @vdm_barrier: VDM (render) command index to wait on.
+	 *
+	 * Barriers are indices relative to the beginning of a given submit. A
+	 * barrier of 0 waits on commands submitted to the respective subqueue
+	 * in previous submit ioctls. A barrier of N waits on N previous
+	 * commands on the subqueue within the current submit ioctl. As a
+	 * special case, passing @DRM_ASAHI_BARRIER_NONE avoids waiting on any
+	 * commands in the subqueue.
+	 *
+	 * Examples:
+	 *
+	 *   0: This waits on all previous work.
+	 *
+	 *   NONE: This does not wait for anything on this subqueue.
+	 *
+	 *   1: This waits on the first render command in the submit.
+	 *   This is valid only if there are multiple render commands in the
+	 *   same submit.
+	 *
+	 * Barriers are valid only for hardware commands. Synthetic software
+	 * commands to set attachments must pass NONE here.
+	 */
+	__u16 vdm_barrier;
+
+	/**
+	 * @cdm_barrier: CDM (compute) command index to wait on.
+	 *
+	 * See @vdm_barrier, and replace VDM/render with CDM/compute.
+	 */
+	__u16 cdm_barrier;
+};
+
+/**
+ * struct drm_asahi_submit - Arguments passed to DRM_IOCTL_ASAHI_SUBMIT
+ */
+struct drm_asahi_submit {
+	/**
+	 * @syncs: An optional pointer to an array of drm_asahi_sync. The first
+	 * @in_sync_count elements are in-syncs, then the remaining
+	 * @out_sync_count elements are out-syncs. Using a single array with
+	 * explicit partitioning simplifies handling.
+	 */
+	__u64 syncs;
+
+	/**
+	 * @cmdbuf: Pointer to the command buffer to submit.
+	 *
+	 * This is a flat command buffer. By design, it contains no CPU
+	 * pointers, which makes it suitable for a virtgpu wire protocol without
+	 * requiring any serializing/deserializing step.
+	 *
+	 * It consists of a series of commands. Each command begins with a
+	 * fixed-size @drm_asahi_cmd_header header and is followed by a
+	 * variable-length payload according to the type and size in the header.
+	 *
+	 * The combined count of "real" hardware commands must be nonzero and at
+	 * most drm_asahi_params_global::max_commands_per_submission.
+	 */
+	__u64 cmdbuf;
+
+	/** @flags: Flags for command submission (MBZ) */
+	__u32 flags;
+
+	/** @queue_id: The queue ID to be submitted to */
+	__u32 queue_id;
+
+	/**
+	 * @in_sync_count: Number of sync objects to wait on before starting
+	 * this job.
+	 */
+	__u32 in_sync_count;
+
+	/**
+	 * @out_sync_count: Number of sync objects to signal upon completion of
+	 * this job.
+	 */
+	__u32 out_sync_count;
+
+	/** @cmdbuf_size: Command buffer size in bytes */
+	__u32 cmdbuf_size;
+
+	/** @pad: MBZ */
+	__u32 pad;
+};
+
+/**
+ * struct drm_asahi_attachment - Describe an "attachment".
+ *
+ * Attachments are any memory written by shaders, notably including render
+ * target attachments written by the end-of-tile program. This is purely a hint
+ * about the accessed memory regions. It is optional to specify, which is
+ * fortunate as it cannot be specified precisely with bindless access anyway.
+ * But where possible, it's probably a good idea for userspace to include these
+ * hints, forwarded to the firmware.
+ *
+ * This struct is implicitly sized and therefore is not extensible.
+ */
+struct drm_asahi_attachment {
+	/** @pointer: Base address of the attachment */
+	__u64 pointer;
+
+	/** @size: Size of the attachment in bytes */
+	__u64 size;
+
+	/** @pad: MBZ */
+	__u32 pad;
+
+	/** @flags: MBZ */
+	__u32 flags;
+};
+
+enum drm_asahi_render_flags {
+	/**
+	 * @DRM_ASAHI_RENDER_VERTEX_SCRATCH: A vertex stage shader uses scratch
+	 * memory.
+	 */
+	DRM_ASAHI_RENDER_VERTEX_SCRATCH = (1U << 0),
+
+	/**
+	 * @DRM_ASAHI_RENDER_PROCESS_EMPTY_TILES: Process even empty tiles.
+	 * This must be set when clearing render targets.
+	 */
+	DRM_ASAHI_RENDER_PROCESS_EMPTY_TILES = (1U << 1),
+
+	/**
+	 * @DRM_ASAHI_RENDER_NO_VERTEX_CLUSTERING: Run vertex stage on a single
+	 * cluster (on multi-cluster GPUs)
+	 *
+	 * This harms performance but can workaround certain sync/coherency
+	 * bugs, and therefore is useful for debugging.
+	 */
+	DRM_ASAHI_RENDER_NO_VERTEX_CLUSTERING = (1U << 2),
+
+	/**
+	 * @DRM_ASAHI_RENDER_DBIAS_IS_INT: Use integer depth bias formula.
+	 *
+	 * Graphics specifications contain two alternate formulas for depth
+	 * bias, a float formula used with floating-point depth buffers and an
+	 * integer formula using with unorm depth buffers. This flag specifies
+	 * that the integer formula should be used. If omitted, the float
+	 * formula is used instead.
+	 *
+	 * This corresponds to bit 18 of the relevant hardware control register,
+	 * so we match that here for efficiency.
+	 */
+	DRM_ASAHI_RENDER_DBIAS_IS_INT = (1U << 18),
+};
+
+/**
+ * struct drm_asahi_zls_buffer - Describe a depth or stencil buffer.
+ *
+ * These fields correspond to hardware registers in the ZLS (Z Load/Store) unit.
+ * There are three hardware registers for each field respectively for loads,
+ * stores, and partial renders. In practice, it makes sense to set all to the
+ * same values, except in exceptional cases not yet implemented in userspace, so
+ * we do not duplicate here for simplicity/efficiency.
+ *
+ * This struct is embedded in other structs and therefore is not extensible.
+ */
+struct drm_asahi_zls_buffer {
+	/** @base: Base address of the buffer */
+	__u64 base;
+
+	/**
+	 * @comp_base: If the load buffer is compressed, address of the
+	 * compression metadata section.
+	 */
+	__u64 comp_base;
+
+	/**
+	 * @stride: If layered rendering is enabled, the number of bytes
+	 * between each layer of the buffer.
+	 */
+	__u32 stride;
+
+	/**
+	 * @comp_stride: If layered rendering is enabled, the number of bytes
+	 * between each layer of the compression metadata.
+	 */
+	__u32 comp_stride;
+};
+
+/**
+ * struct drm_asahi_timestamp - Describe a timestamp write.
+ *
+ * The firmware can optionally write the GPU timestamp at render pass
+ * granularities, but it needs to be mapped specially via
+ * DRM_IOCTL_ASAHI_GEM_BIND_OBJECT. This structure therefore describes where to
+ * write as a handle-offset pair, rather than a GPU address like normal.
+ *
+ * This struct is embedded in other structs and therefore is not extensible.
+ */
+struct drm_asahi_timestamp {
+	/**
+	 * @handle: Handle of the timestamp buffer, or 0 to skip this
+	 * timestamp. If nonzero, this must equal the value returned in
+	 * drm_asahi_gem_bind_object::object_handle.
+	 */
+	__u32 handle;
+
+	/** @offset: Offset to write into the timestamp buffer */
+	__u32 offset;
+};
+
+/**
+ * struct drm_asahi_timestamps - Describe timestamp writes.
+ *
+ * Each operation that can be timestamped, can be timestamped at the start and
+ * end. Therefore, drm_asahi_timestamp structs always come in pairs, bundled
+ * together into drm_asahi_timestamps.
+ *
+ * This struct is embedded in other structs and therefore is not extensible.
+ */
+struct drm_asahi_timestamps {
+	/** @start: Timestamp recorded at the start of the operation */
+	struct drm_asahi_timestamp start;
+
+	/** @end: Timestamp recorded at the end of the operation */
+	struct drm_asahi_timestamp end;
+};
+
+/**
+ * struct drm_asahi_helper_program - Describe helper program configuration.
+ *
+ * The helper program is a compute-like kernel required for various hardware
+ * functionality. Its most important role is dynamically allocating
+ * scratch/stack memory for individual subgroups, by partitioning a static
+ * allocation shared for the whole device. It is supplied by userspace via
+ * drm_asahi_helper_program and internally dispatched by the hardware as needed.
+ *
+ * This struct is embedded in other structs and therefore is not extensible.
+ */
+struct drm_asahi_helper_program {
+	/**
+	 * @binary: USC address to the helper program binary. This is a tagged
+	 * pointer with configuration in the bottom bits.
+	 */
+	__u32 binary;
+
+	/** @cfg: Additional configuration bits for the helper program. */
+	__u32 cfg;
+
+	/**
+	 * @data: Data passed to the helper program. This value is not
+	 * interpreted by the kernel, firmware, or hardware in any way. It is
+	 * simply a sideband for userspace, set with the submit ioctl and read
+	 * via special registers inside the helper program.
+	 *
+	 * In practice, userspace will pass a 64-bit GPU VA here pointing to the
+	 * actual arguments, which presumably don't fit in 64-bits.
+	 */
+	__u64 data;
+};
+
+/**
+ * struct drm_asahi_bg_eot - Describe a background or end-of-tile program.
+ *
+ * The background and end-of-tile programs are dispatched by the hardware at the
+ * beginning and end of rendering. As the hardware "tilebuffer" is simply local
+ * memory, these programs are necessary to implement API-level render targets.
+ * The fragment-like background program is responsible for loading either the
+ * clear colour or the existing render target contents, while the compute-like
+ * end-of-tile program stores the tilebuffer contents to memory.
+ *
+ * This struct is embedded in other structs and therefore is not extensible.
+ */
+struct drm_asahi_bg_eot {
+	/**
+	 * @usc: USC address of the hardware USC words binding resources
+	 * (including images and uniforms) and the program itself. Note this is
+	 * an additional layer of indirection compared to the helper program,
+	 * avoiding the need for a sideband for data. This is a tagged pointer
+	 * with additional configuration in the bottom bits.
+	 */
+	__u32 usc;
+
+	/**
+	 * @rsrc_spec: Resource specifier for the program. This is a packed
+	 * hardware data structure describing the required number of registers,
+	 * uniforms, bound textures, and bound samplers.
+	 */
+	__u32 rsrc_spec;
+};
+
+/**
+ * struct drm_asahi_cmd_render - Command to submit 3D
+ *
+ * This command submits a single render pass. The hardware control stream may
+ * include many draws and subpasses, but within the command, the framebuffer
+ * dimensions and attachments are fixed.
+ *
+ * The hardware requires the firmware to set a large number of Control Registers
+ * setting up state at render pass granularity before each command rendering 3D.
+ * The firmware bundles this state into data structures. Unfortunately, we
+ * cannot expose either any of that directly to userspace, because the
+ * kernel-firmware ABI is not stable. Although we can guarantee the firmware
+ * updates in tandem with the kernel, we cannot break old userspace when
+ * upgrading the firmware and kernel. Therefore, we need to abstract well the
+ * data structures to avoid tying our hands with future firmwares.
+ *
+ * The bulk of drm_asahi_cmd_render therefore consists of values of hardware
+ * control registers, marshalled via the firmware interface.
+ *
+ * The framebuffer/tilebuffer dimensions are also specified here. In addition to
+ * being passed to the firmware/hardware, the kernel requires these dimensions
+ * to calculate various essential tiling-related data structures. It is
+ * unfortunate that our submits are heavier than on vendors with saner
+ * hardware-software interfaces. The upshot is all of this information is
+ * readily available to userspace with all current APIs.
+ *
+ * It looks odd - but it's not overly burdensome and it ensures we can remain
+ * compatible with old userspace.
+ */
+struct drm_asahi_cmd_render {
+	/** @flags: Combination of drm_asahi_render_flags flags. */
+	__u32 flags;
+
+	/**
+	 * @isp_zls_pixels: ISP_ZLS_PIXELS register value. This contains the
+	 * depth/stencil width/height, which may differ from the framebuffer
+	 * width/height.
+	 */
+	__u32 isp_zls_pixels;
+
+	/**
+	 * @vdm_ctrl_stream_base: VDM_CTRL_STREAM_BASE register value. GPU
+	 * address to the beginning of the VDM control stream.
+	 */
+	__u64 vdm_ctrl_stream_base;
+
+	/** @vertex_helper: Helper program used for the vertex shader */
+	struct drm_asahi_helper_program vertex_helper;
+
+	/** @fragment_helper: Helper program used for the fragment shader */
+	struct drm_asahi_helper_program fragment_helper;
+
+	/**
+	 * @isp_scissor_base: ISP_SCISSOR_BASE register value. GPU address of an
+	 * array of scissor descriptors indexed in the render pass.
+	 */
+	__u64 isp_scissor_base;
+
+	/**
+	 * @isp_dbias_base: ISP_DBIAS_BASE register value. GPU address of an
+	 * array of depth bias values indexed in the render pass.
+	 */
+	__u64 isp_dbias_base;
+
+	/**
+	 * @isp_oclqry_base: ISP_OCLQRY_BASE register value. GPU address of an
+	 * array of occlusion query results written by the render pass.
+	 */
+	__u64 isp_oclqry_base;
+
+	/** @depth: Depth buffer */
+	struct drm_asahi_zls_buffer depth;
+
+	/** @stencil: Stencil buffer */
+	struct drm_asahi_zls_buffer stencil;
+
+	/** @zls_ctrl: ZLS_CTRL register value */
+	__u64 zls_ctrl;
+
+	/** @ppp_multisamplectl: PPP_MULTISAMPLECTL register value */
+	__u64 ppp_multisamplectl;
+
+	/**
+	 * @sampler_heap: Base address of the sampler heap. This heap is used
+	 * for both vertex shaders and fragment shaders. The registers are
+	 * per-stage, but there is no known use case for separate heaps.
+	 */
+	__u64 sampler_heap;
+
+	/** @ppp_ctrl: PPP_CTRL register value */
+	__u32 ppp_ctrl;
+
+	/** @width_px: Framebuffer width in pixels */
+	__u16 width_px;
+
+	/** @height_px: Framebuffer height in pixels */
+	__u16 height_px;
+
+	/** @layers: Number of layers in the framebuffer */
+	__u16 layers;
+
+	/** @sampler_count: Number of samplers in the sampler heap. */
+	__u16 sampler_count;
+
+	/** @utile_width_px: Width of a logical tilebuffer tile in pixels */
+	__u8 utile_width_px;
+
+	/** @utile_height_px: Height of a logical tilebuffer tile in pixels */
+	__u8 utile_height_px;
+
+	/** @samples: # of samples in the framebuffer. Must be 1, 2, or 4. */
+	__u8 samples;
+
+	/** @sample_size_B: # of bytes in the tilebuffer required per sample. */
+	__u8 sample_size_B;
+
+	/**
+	 * @isp_merge_upper_x: 32-bit float used in the hardware triangle
+	 * merging. Calculate as: tan(60 deg) * width.
+	 *
+	 * Making these values UAPI avoids requiring floating-point calculations
+	 * in the kernel in the hot path.
+	 */
+	__u32 isp_merge_upper_x;
+
+	/**
+	 * @isp_merge_upper_y: 32-bit float. Calculate as: tan(60 deg) * height.
+	 * See @isp_merge_upper_x.
+	 */
+	__u32 isp_merge_upper_y;
+
+	/** @bg: Background program run for each tile at the start */
+	struct drm_asahi_bg_eot bg;
+
+	/** @eot: End-of-tile program ran for each tile at the end */
+	struct drm_asahi_bg_eot eot;
+
+	/**
+	 * @partial_bg: Background program ran at the start of each tile when
+	 * resuming the render pass during a partial render.
+	 */
+	struct drm_asahi_bg_eot partial_bg;
+
+	/**
+	 * @partial_eot: End-of-tile program ran at the end of each tile when
+	 * pausing the render pass during a partial render.
+	 */
+	struct drm_asahi_bg_eot partial_eot;
+
+	/**
+	 * @isp_bgobjdepth: ISP_BGOBJDEPTH register value. This is the depth
+	 * buffer clear value, encoded in the depth buffer's format: either a
+	 * 32-bit float or a 16-bit unorm (with upper bits zeroed).
+	 */
+	__u32 isp_bgobjdepth;
+
+	/**
+	 * @isp_bgobjvals: ISP_BGOBJVALS register value. The bottom 8-bits
+	 * contain the stencil buffer clear value.
+	 */
+	__u32 isp_bgobjvals;
+
+	/** @ts_vtx: Timestamps for the vertex portion of the render */
+	struct drm_asahi_timestamps ts_vtx;
+
+	/** @ts_frag: Timestamps for the fragment portion of the render */
+	struct drm_asahi_timestamps ts_frag;
+};
+
+/**
+ * struct drm_asahi_cmd_compute - Command to submit compute
+ *
+ * This command submits a control stream consisting of compute dispatches. There
+ * is essentially no limit on how many compute dispatches may be included in a
+ * single compute command, although timestamps are at command granularity.
+ */
+struct drm_asahi_cmd_compute {
+	/** @flags: MBZ */
+	__u32 flags;
+
+	/** @sampler_count: Number of samplers in the sampler heap. */
+	__u32 sampler_count;
+
+	/**
+	 * @cdm_ctrl_stream_base: CDM_CTRL_STREAM_BASE register value. GPU
+	 * address to the beginning of the CDM control stream.
+	 */
+	__u64 cdm_ctrl_stream_base;
+
+	/**
+	 * @cdm_ctrl_stream_end: GPU base address to the end of the hardware
+	 * control stream. Note this only considers the first contiguous segment
+	 * of the control stream, as the stream might jump elsewhere.
+	 */
+	__u64 cdm_ctrl_stream_end;
+
+	/** @sampler_heap: Base address of the sampler heap. */
+	__u64 sampler_heap;
+
+	/** @helper: Helper program used for this compute command */
+	struct drm_asahi_helper_program helper;
+
+	/** @ts: Timestamps for the compute command */
+	struct drm_asahi_timestamps ts;
+};
+
+/**
+ * struct drm_asahi_get_time - Arguments passed to DRM_IOCTL_ASAHI_GET_TIME
+ */
+struct drm_asahi_get_time {
+	/** @flags: MBZ. */
+	__u64 flags;
+
+	/** @gpu_timestamp: On return, the GPU timestamp in nanoseconds. */
+	__u64 gpu_timestamp;
+};
+
+/**
+ * DRM_IOCTL_ASAHI() - Build an Asahi IOCTL number
+ * @__access: Access type. Must be R, W or RW.
+ * @__id: One of the DRM_ASAHI_xxx id.
+ * @__type: Suffix of the type being passed to the IOCTL.
+ *
+ * Don't use this macro directly, use the DRM_IOCTL_ASAHI_xxx
+ * values instead.
+ *
+ * Return: An IOCTL number to be passed to ioctl() from userspace.
+ */
+#define DRM_IOCTL_ASAHI(__access, __id, __type) \
+	DRM_IO ## __access(DRM_COMMAND_BASE + DRM_ASAHI_ ## __id, \
+			   struct drm_asahi_ ## __type)
+
+/* Note: this is an enum so that it can be resolved by Rust bindgen. */
+enum {
+	DRM_IOCTL_ASAHI_GET_PARAMS       = DRM_IOCTL_ASAHI(W, GET_PARAMS, get_params),
+	DRM_IOCTL_ASAHI_GET_TIME         = DRM_IOCTL_ASAHI(WR, GET_TIME, get_time),
+	DRM_IOCTL_ASAHI_VM_CREATE        = DRM_IOCTL_ASAHI(WR, VM_CREATE, vm_create),
+	DRM_IOCTL_ASAHI_VM_DESTROY       = DRM_IOCTL_ASAHI(W, VM_DESTROY, vm_destroy),
+	DRM_IOCTL_ASAHI_VM_BIND          = DRM_IOCTL_ASAHI(W, VM_BIND, vm_bind),
+	DRM_IOCTL_ASAHI_GEM_CREATE       = DRM_IOCTL_ASAHI(WR, GEM_CREATE, gem_create),
+	DRM_IOCTL_ASAHI_GEM_MMAP_OFFSET  = DRM_IOCTL_ASAHI(WR, GEM_MMAP_OFFSET, gem_mmap_offset),
+	DRM_IOCTL_ASAHI_GEM_BIND_OBJECT  = DRM_IOCTL_ASAHI(WR, GEM_BIND_OBJECT, gem_bind_object),
+	DRM_IOCTL_ASAHI_QUEUE_CREATE     = DRM_IOCTL_ASAHI(WR, QUEUE_CREATE, queue_create),
+	DRM_IOCTL_ASAHI_QUEUE_DESTROY    = DRM_IOCTL_ASAHI(W, QUEUE_DESTROY, queue_destroy),
+	DRM_IOCTL_ASAHI_SUBMIT           = DRM_IOCTL_ASAHI(W, SUBMIT, submit),
+};
+
+#if defined(__cplusplus)
+}
+#endif
+
+#endif /* _ASAHI_DRM_H_ */
diff --git a/include/uapi/drm/drm.h b/include/uapi/drm/drm.h
index 7fba37b94401..e63a71d3c607 100644
--- a/include/uapi/drm/drm.h
+++ b/include/uapi/drm/drm.h
@@ -905,13 +905,17 @@ struct drm_syncobj_destroy {
 };
 
 #define DRM_SYNCOBJ_FD_TO_HANDLE_FLAGS_IMPORT_SYNC_FILE (1 << 0)
+#define DRM_SYNCOBJ_FD_TO_HANDLE_FLAGS_TIMELINE         (1 << 1)
 #define DRM_SYNCOBJ_HANDLE_TO_FD_FLAGS_EXPORT_SYNC_FILE (1 << 0)
+#define DRM_SYNCOBJ_HANDLE_TO_FD_FLAGS_TIMELINE         (1 << 1)
 struct drm_syncobj_handle {
 	__u32 handle;
 	__u32 flags;
 
 	__s32 fd;
 	__u32 pad;
+
+	__u64 point;
 };
 
 struct drm_syncobj_transfer {
diff --git a/include/uapi/drm/drm_fourcc.h b/include/uapi/drm/drm_fourcc.h
index e41a3cec6a9e..81202a50dc9e 100644
--- a/include/uapi/drm/drm_fourcc.h
+++ b/include/uapi/drm/drm_fourcc.h
@@ -422,6 +422,7 @@ extern "C" {
 #define DRM_FORMAT_MOD_VENDOR_ALLWINNER 0x09
 #define DRM_FORMAT_MOD_VENDOR_AMLOGIC 0x0a
 #define DRM_FORMAT_MOD_VENDOR_MTK     0x0b
+#define DRM_FORMAT_MOD_VENDOR_APPLE   0x0c
 
 /* add more to the end as needed */
 
@@ -1495,6 +1496,50 @@ drm_fourcc_canonicalize_nvidia_format_mod(__u64 modifier)
 #define DRM_FORMAT_MOD_MTK_16L_32S_TILE  DRM_FORMAT_MOD_MTK(MTK_FMT_MOD_TILE_16L32S)
 
 /*
+ * Apple GPU-tiled layouts.
+ *
+ * Apple GPUs support nonlinear tilings with optional lossless compression.
+ *
+ * GPU-tiled images are divided into 16KiB tiles:
+ *
+ *     Bytes per pixel  Tile size
+ *     ---------------  ---------
+ *                   1  128x128
+ *                   2  128x64
+ *                   4  64x64
+ *                   8  64x32
+ *                  16  32x32
+ *
+ * Tiles are raster-order. Pixels within a tile are interleaved (Morton order).
+ *
+ * Compressed images pad the body to 128-bytes and are immediately followed by a
+ * metadata section. The metadata section rounds the image dimensions to
+ * powers-of-two and contains 8 bytes for each 16x16 compression subtile.
+ * Subtiles are interleaved (Morton order).
+ *
+ * All images are 128-byte aligned.
+ *
+ * These layouts fundamentally do not have meaningful strides. No matter how we
+ * specify strides for these layouts, userspace unaware of Apple image layouts
+ * will be unable to use correctly the specified stride for any purpose.
+ * Userspace aware of the image layouts do not use strides. The most "correct"
+ * convention would be setting the image stride to 0. Unfortunately, some
+ * software assumes the stride is at least (width * bytes per pixel). We
+ * therefore require that stride equals (width * bytes per pixel). Since the
+ * stride is arbitrary here, we pick the simplest convention.
+ *
+ * Although containing two sections, compressed image layouts are treated in
+ * software as a single plane. This is modelled after AFBC, a similar
+ * scheme. Attempting to separate the sections to be "explicit" in DRM would
+ * only generate more confusion, as software does not treat the image this way.
+ *
+ * For detailed information on the hardware image layouts, see
+ * https://docs.mesa3d.org/drivers/asahi.html#image-layouts
+ */
+#define DRM_FORMAT_MOD_APPLE_GPU_TILED fourcc_mod_code(APPLE, 1)
+#define DRM_FORMAT_MOD_APPLE_GPU_TILED_COMPRESSED fourcc_mod_code(APPLE, 2)
+
+/*
  * AMD modifiers
  *
  * Memory layout:
diff --git a/include/uapi/drm/nova_drm.h b/include/uapi/drm/nova_drm.h
new file mode 100644
index 000000000000..3ca90ed9d2bb
--- /dev/null
+++ b/include/uapi/drm/nova_drm.h
@@ -0,0 +1,101 @@
+/* SPDX-License-Identifier: MIT */
+
+#ifndef __NOVA_DRM_H__
+#define __NOVA_DRM_H__
+
+#include "drm.h"
+
+/* DISCLAIMER: Do not use, this is not a stable uAPI.
+ *
+ * This uAPI serves only testing purposes as long as this driver is still in
+ * development. It is required to implement and test infrastructure which is
+ * upstreamed in the context of this driver. See also [1].
+ *
+ * [1] https://lore.kernel.org/dri-devel/Zfsj0_tb-0-tNrJy@cassiopeiae/T/#u
+ */
+
+#if defined(__cplusplus)
+extern "C" {
+#endif
+
+/*
+ * NOVA_GETPARAM_VRAM_BAR_SIZE
+ *
+ * Query the VRAM BAR size in bytes.
+ */
+#define NOVA_GETPARAM_VRAM_BAR_SIZE	0x1
+
+/**
+ * struct drm_nova_getparam - query GPU and driver metadata
+ */
+struct drm_nova_getparam {
+	/**
+	 * @param: The identifier of the parameter to query.
+	 */
+	__u64 param;
+
+	/**
+	 * @value: The value for the specified parameter.
+	 */
+	__u64 value;
+};
+
+/**
+ * struct drm_nova_gem_create - create a new DRM GEM object
+ */
+struct drm_nova_gem_create {
+	/**
+	 * @handle: The handle of the new DRM GEM object.
+	 */
+	__u32 handle;
+
+	/**
+	 * @pad: 32 bit padding, should be 0.
+	 */
+	__u32 pad;
+
+	/**
+	 * @size: The size of the new DRM GEM object.
+	 */
+	__u64 size;
+};
+
+/**
+ * struct drm_nova_gem_info - query DRM GEM object metadata
+ */
+struct drm_nova_gem_info {
+	/**
+	 * @handle: The handle of the DRM GEM object to query.
+	 */
+	__u32 handle;
+
+	/**
+	 * @pad: 32 bit padding, should be 0.
+	 */
+	__u32 pad;
+
+	/**
+	 * @size: The size of the DRM GEM obejct.
+	 */
+	__u64 size;
+};
+
+#define DRM_NOVA_GETPARAM		0x00
+#define DRM_NOVA_GEM_CREATE		0x01
+#define DRM_NOVA_GEM_INFO		0x02
+
+/* Note: this is an enum so that it can be resolved by Rust bindgen. */
+enum {
+	DRM_IOCTL_NOVA_GETPARAM		= DRM_IOWR(DRM_COMMAND_BASE + DRM_NOVA_GETPARAM,
+						   struct drm_nova_getparam),
+	DRM_IOCTL_NOVA_GEM_CREATE	= DRM_IOWR(DRM_COMMAND_BASE + DRM_NOVA_GEM_CREATE,
+						   struct drm_nova_gem_create),
+	DRM_IOCTL_NOVA_GEM_INFO		= DRM_IOWR(DRM_COMMAND_BASE + DRM_NOVA_GEM_INFO,
+						   struct drm_nova_gem_info),
+};
+
+#if defined(__cplusplus)
+}
+#endif
+
+#endif /* __NOVA_DRM_H__ */
diff --git a/include/uapi/drm/panthor_drm.h b/include/uapi/drm/panthor_drm.h
index 97e2c4510e69..ad9a70afea6c 100644
--- a/include/uapi/drm/panthor_drm.h
+++ b/include/uapi/drm/panthor_drm.h
@@ -127,6 +127,9 @@ enum drm_panthor_ioctl_id {
 
 	/** @DRM_PANTHOR_TILER_HEAP_DESTROY: Destroy a tiler heap. */
 	DRM_PANTHOR_TILER_HEAP_DESTROY,
+
+	/** @DRM_PANTHOR_BO_SET_LABEL: Label a BO. */
+	DRM_PANTHOR_BO_SET_LABEL,
 };
 
 /**
@@ -978,6 +981,24 @@ struct drm_panthor_tiler_heap_destroy {
 };
 
 /**
+ * struct drm_panthor_bo_set_label - Arguments passed to DRM_IOCTL_PANTHOR_BO_SET_LABEL
+ */
+struct drm_panthor_bo_set_label {
+	/** @handle: Handle of the buffer object to label. */
+	__u32 handle;
+
+	/**  @pad: MBZ. */
+	__u32 pad;
+
+	/**
+	 * @label: User pointer to a NUL-terminated string
+	 *
+	 * Length cannot be greater than 4096
+	 */
+	__u64 label;
+};
+
+/**
  * DRM_IOCTL_PANTHOR() - Build a Panthor IOCTL number
  * @__access: Access type. Must be R, W or RW.
  * @__id: One of the DRM_PANTHOR_xxx id.
@@ -1019,6 +1040,8 @@ enum {
 		DRM_IOCTL_PANTHOR(WR, TILER_HEAP_CREATE, tiler_heap_create),
 	DRM_IOCTL_PANTHOR_TILER_HEAP_DESTROY =
 		DRM_IOCTL_PANTHOR(WR, TILER_HEAP_DESTROY, tiler_heap_destroy),
+	DRM_IOCTL_PANTHOR_BO_SET_LABEL =
+		DRM_IOCTL_PANTHOR(WR, BO_SET_LABEL, bo_set_label),
 };
 
 #if defined(__cplusplus)
diff --git a/include/uapi/drm/virtgpu_drm.h b/include/uapi/drm/virtgpu_drm.h
index c2ce71987e9b..9debb320c34b 100644
--- a/include/uapi/drm/virtgpu_drm.h
+++ b/include/uapi/drm/virtgpu_drm.h
@@ -163,6 +163,12 @@ struct drm_virtgpu_3d_wait {
 	__u32 flags;
 };
 
+#define VIRTGPU_DRM_CAPSET_VIRGL 1
+#define VIRTGPU_DRM_CAPSET_VIRGL2 2
+#define VIRTGPU_DRM_CAPSET_GFXSTREAM_VULKAN 3
+#define VIRTGPU_DRM_CAPSET_VENUS 4
+#define VIRTGPU_DRM_CAPSET_CROSS_DOMAIN 5
+#define VIRTGPU_DRM_CAPSET_DRM 6
 struct drm_virtgpu_get_caps {
 	__u32 cap_set_id;
 	__u32 cap_set_ver;
diff --git a/include/uapi/drm/xe_drm.h b/include/uapi/drm/xe_drm.h
index 616916985e3f..9c08738c3b91 100644
--- a/include/uapi/drm/xe_drm.h
+++ b/include/uapi/drm/xe_drm.h
@@ -917,7 +917,11 @@ struct drm_xe_gem_mmap_offset {
  * struct drm_xe_vm_create - Input of &DRM_IOCTL_XE_VM_CREATE
  *
  * The @flags can be:
- *  - %DRM_XE_VM_CREATE_FLAG_SCRATCH_PAGE
+ *  - %DRM_XE_VM_CREATE_FLAG_SCRATCH_PAGE - Map the whole virtual address
+ *    space of the VM to scratch page. A vm_bind would overwrite the scratch
+ *    page mapping. This flag is mutually exclusive with the
+ *    %DRM_XE_VM_CREATE_FLAG_FAULT_MODE flag, with an exception of on x2 and
+ *    xe3 platform.
  *  - %DRM_XE_VM_CREATE_FLAG_LR_MODE - An LR, or Long Running VM accepts
  *    exec submissions to its exec_queues that don't have an upper time
  *    limit on the job execution time. But exec submissions to these
diff --git a/include/uapi/linux/kfd_ioctl.h b/include/uapi/linux/kfd_ioctl.h
index 1e59344c5673..04c7d283dc7d 100644
--- a/include/uapi/linux/kfd_ioctl.h
+++ b/include/uapi/linux/kfd_ioctl.h
@@ -536,6 +536,8 @@ enum kfd_smi_event {
 	KFD_SMI_EVENT_QUEUE_EVICTION = 9,
 	KFD_SMI_EVENT_QUEUE_RESTORE = 10,
 	KFD_SMI_EVENT_UNMAP_FROM_GPU = 11,
+	KFD_SMI_EVENT_PROCESS_START = 12,
+	KFD_SMI_EVENT_PROCESS_END = 13,
 
 	/*
 	 * max event number, as a flag bit to get events from all processes,
@@ -651,6 +653,9 @@ struct kfd_ioctl_smi_events_args {
 		"%lld -%d @%lx(%lx) %x %d\n", (ns), (pid), (addr), (size),\
 		(node), (unmap_trigger)
 
+#define KFD_EVENT_FMT_PROCESS(pid, task_name)\
+		"%x %s\n", (pid), (task_name)
+
 /**************************************************************************************************
  * CRIU IOCTLs (Checkpoint Restore In Userspace)
  *
diff --git a/include/uapi/linux/virtio_gpu.h b/include/uapi/linux/virtio_gpu.h
index bf2c9cabd207..be109777d10d 100644
--- a/include/uapi/linux/virtio_gpu.h
+++ b/include/uapi/linux/virtio_gpu.h
@@ -309,8 +309,9 @@ struct virtio_gpu_cmd_submit {
 
 #define VIRTIO_GPU_CAPSET_VIRGL 1
 #define VIRTIO_GPU_CAPSET_VIRGL2 2
-/* 3 is reserved for gfxstream */
+#define VIRTIO_GPU_CAPSET_GFXSTREAM_VULKAN 3
 #define VIRTIO_GPU_CAPSET_VENUS 4
+#define VIRTIO_GPU_CAPSET_CROSS_DOMAIN 5
 #define VIRTIO_GPU_CAPSET_DRM 6
 
 /* VIRTIO_GPU_CMD_GET_CAPSET_INFO */