1============================ 2 Core Driver Infrastructure 3============================ 4 5GPU Hardware Structure 6====================== 7 8Each ASIC is a collection of hardware blocks. We refer to them as 9"IPs" (Intellectual Property blocks). Each IP encapsulates certain 10functionality. IPs are versioned and can also be mixed and matched. 11E.g., you might have two different ASICs that both have System DMA (SDMA) 5.x IPs. 12The driver is arranged by IPs. There are driver components to handle 13the initialization and operation of each IP. There are also a bunch 14of smaller IPs that don't really need much if any driver interaction. 15Those end up getting lumped into the common stuff in the soc files. 16The soc files (e.g., vi.c, soc15.c nv.c) contain code for aspects of 17the SoC itself rather than specific IPs. E.g., things like GPU resets 18and register access functions are SoC dependent. 19 20An APU contains more than just CPU and GPU, it also contains all of 21the platform stuff (audio, usb, gpio, etc.). Also, a lot of 22components are shared between the CPU, platform, and the GPU (e.g., 23SMU, PSP, etc.). Specific components (CPU, GPU, etc.) usually have 24their interface to interact with those common components. For things 25like S0i3 there is a ton of coordination required across all the 26components, but that is probably a bit beyond the scope of this 27section. 28 29With respect to the GPU, we have the following major IPs: 30 31GMC (Graphics Memory Controller) 32 This was a dedicated IP on older pre-vega chips, but has since 33 become somewhat decentralized on vega and newer chips. They now 34 have dedicated memory hubs for specific IPs or groups of IPs. We 35 still treat it as a single component in the driver however since 36 the programming model is still pretty similar. This is how the 37 different IPs on the GPU get the memory (VRAM or system memory). 38 It also provides the support for per process GPU virtual address 39 spaces. 40 41IH (Interrupt Handler) 42 This is the interrupt controller on the GPU. All of the IPs feed 43 their interrupts into this IP and it aggregates them into a set of 44 ring buffers that the driver can parse to handle interrupts from 45 different IPs. 46 47PSP (Platform Security Processor) 48 This handles security policy for the SoC and executes trusted 49 applications, and validates and loads firmwares for other blocks. 50 51SMU (System Management Unit) 52 This is the power management microcontroller. It manages the entire 53 SoC. The driver interacts with it to control power management 54 features like clocks, voltages, power rails, etc. 55 56DCN (Display Controller Next) 57 This is the display controller. It handles the display hardware. 58 It is described in more details in :ref:`Display Core <amdgpu-display-core>`. 59 60SDMA (System DMA) 61 This is a multi-purpose DMA engine. The kernel driver uses it for 62 various things including paging and GPU page table updates. It's also 63 exposed to userspace for use by user mode drivers (OpenGL, Vulkan, 64 etc.) 65 66GC (Graphics and Compute) 67 This is the graphics and compute engine, i.e., the block that 68 encompasses the 3D pipeline and and shader blocks. This is by far the 69 largest block on the GPU. The 3D pipeline has tons of sub-blocks. In 70 addition to that, it also contains the CP microcontrollers (ME, PFP, CE, 71 MEC) and the RLC microcontroller. It's exposed to userspace for user mode 72 drivers (OpenGL, Vulkan, OpenCL, etc.). More details in :ref:`Graphics (GFX) 73 and Compute <amdgpu-gc>`. 74 75VCN (Video Core Next) 76 This is the multi-media engine. It handles video and image encode and 77 decode. It's exposed to userspace for user mode drivers (VA-API, 78 OpenMAX, etc.) 79 80.. _pipes-and-queues-description: 81 82GFX, Compute, and SDMA Overall Behavior 83======================================= 84 85.. note:: For simplicity, whenever the term block is used in this section, it 86 means GFX, Compute, and SDMA. 87 88GFX, Compute and SDMA share a similar form of operation that can be abstracted 89to facilitate understanding of the behavior of these blocks. See the figure 90below illustrating the common components of these blocks: 91 92.. kernel-figure:: pipe_and_queue_abstraction.svg 93 94In the central part of this figure, you can see two hardware elements, one called 95**Pipes** and another called **Queues**; it is important to highlight that Queues 96must be associated with a Pipe and vice-versa. Every specific hardware IP may have 97a different number of Pipes and, in turn, a different number of Queues; for 98example, GFX 11 has two Pipes and two Queues per Pipe for the GFX front end. 99 100Pipe is the hardware that processes the instructions available in the Queues; 101in other words, it is a thread executing the operations inserted in the Queue. 102One crucial characteristic of Pipes is that they can only execute one Queue at 103a time; no matter if the hardware has multiple Queues in the Pipe, it only runs 104one Queue per Pipe. 105 106Pipes have the mechanics of swapping between queues at the hardware level. 107Nonetheless, they only make use of Queues that are considered mapped. Pipes can 108switch between queues based on any of the following inputs: 109 1101. Command Stream; 1112. Packet by Packet; 1123. Other hardware requests the change (e.g., MES). 113 114Queues within Pipes are defined by the Hardware Queue Descriptors (HQD). 115Associated with the HQD concept, we have the Memory Queue Descriptor (MQD), 116which is responsible for storing information about the state of each of the 117available Queues in the memory. The state of a Queue contains information such 118as the GPU virtual address of the queue itself, save areas, doorbell, etc. The 119MQD also stores the HQD registers, which are vital for activating or 120deactivating a given Queue. The scheduling firmware (e.g., MES) is responsible 121for loading HQDs from MQDs and vice versa. 122 123The Queue-switching process can also happen with the firmware requesting the 124preemption or unmapping of a Queue. The firmware waits for the HQD_ACTIVE bit 125to change to low before saving the state into the MQD. To make a different 126Queue become active, the firmware copies the MQD state into the HQD registers 127and loads any additional state. Finally, it sets the HQD_ACTIVE bit to high to 128indicate that the queue is active. The Pipe will then execute work from active 129Queues. 130 131Driver Structure 132================ 133 134In general, the driver has a list of all of the IPs on a particular 135SoC and for things like init/fini/suspend/resume, more or less just 136walks the list and handles each IP. 137 138Some useful constructs: 139 140KIQ (Kernel Interface Queue) 141 This is a control queue used by the kernel driver to manage other gfx 142 and compute queues on the GFX/compute engine. You can use it to 143 map/unmap additional queues, etc. This is replaced by MES on 144 GFX 11 and newer hardware. 145 146IB (Indirect Buffer) 147 A command buffer for a particular engine. Rather than writing 148 commands directly to the queue, you can write the commands into a 149 piece of memory and then put a pointer to the memory into the queue. 150 The hardware will then follow the pointer and execute the commands in 151 the memory, then returning to the rest of the commands in the ring. 152 153.. _amdgpu_memory_domains: 154 155Memory Domains 156============== 157 158.. kernel-doc:: include/uapi/drm/amdgpu_drm.h 159 :doc: memory domains 160 161Buffer Objects 162============== 163 164.. kernel-doc:: drivers/gpu/drm/amd/amdgpu/amdgpu_object.c 165 :doc: amdgpu_object 166 167.. kernel-doc:: drivers/gpu/drm/amd/amdgpu/amdgpu_object.c 168 :internal: 169 170PRIME Buffer Sharing 171==================== 172 173.. kernel-doc:: drivers/gpu/drm/amd/amdgpu/amdgpu_dma_buf.c 174 :doc: PRIME Buffer Sharing 175 176.. kernel-doc:: drivers/gpu/drm/amd/amdgpu/amdgpu_dma_buf.c 177 :internal: 178 179MMU Notifier 180============ 181 182.. kernel-doc:: drivers/gpu/drm/amd/amdgpu/amdgpu_hmm.c 183 :doc: MMU Notifier 184 185.. kernel-doc:: drivers/gpu/drm/amd/amdgpu/amdgpu_hmm.c 186 :internal: 187 188AMDGPU Virtual Memory 189===================== 190 191.. kernel-doc:: drivers/gpu/drm/amd/amdgpu/amdgpu_vm.c 192 :doc: GPUVM 193 194.. kernel-doc:: drivers/gpu/drm/amd/amdgpu/amdgpu_vm.c 195 :internal: 196 197Interrupt Handling 198================== 199 200.. kernel-doc:: drivers/gpu/drm/amd/amdgpu/amdgpu_irq.c 201 :doc: Interrupt Handling 202 203.. kernel-doc:: drivers/gpu/drm/amd/amdgpu/amdgpu_irq.c 204 :internal: 205 206IP Blocks 207========= 208 209.. kernel-doc:: drivers/gpu/drm/amd/include/amd_shared.h 210 :doc: IP Blocks 211 212.. kernel-doc:: drivers/gpu/drm/amd/include/amd_shared.h 213 :identifiers: amd_ip_block_type amd_ip_funcs DC_DEBUG_MASK 214