1 .. Copyright 2020 DisplayLink (UK) Ltd. 1 .. Copyright 2020 DisplayLink (UK) Ltd.
2 2
3 =================== 3 ===================
4 Userland interfaces 4 Userland interfaces
5 =================== 5 ===================
6 6
7 The DRM core exports several interfaces to app 7 The DRM core exports several interfaces to applications, generally
8 intended to be used through corresponding libd 8 intended to be used through corresponding libdrm wrapper functions. In
9 addition, drivers export device-specific inter 9 addition, drivers export device-specific interfaces for use by userspace
10 drivers & device-aware applications through io 10 drivers & device-aware applications through ioctls and sysfs files.
11 11
12 External interfaces include: memory mapping, c 12 External interfaces include: memory mapping, context management, DMA
13 operations, AGP management, vblank control, fe 13 operations, AGP management, vblank control, fence management, memory
14 management, and output management. 14 management, and output management.
15 15
16 Cover generic ioctls and sysfs layout here. We 16 Cover generic ioctls and sysfs layout here. We only need high-level
17 info, since man pages should cover the rest. 17 info, since man pages should cover the rest.
18 18
19 libdrm Device Lookup 19 libdrm Device Lookup
20 ==================== 20 ====================
21 21
22 .. kernel-doc:: drivers/gpu/drm/drm_ioctl.c 22 .. kernel-doc:: drivers/gpu/drm/drm_ioctl.c
23 :doc: getunique and setversion story 23 :doc: getunique and setversion story
24 24
25 25
26 .. _drm_primary_node: 26 .. _drm_primary_node:
27 27
28 Primary Nodes, DRM Master and Authentication 28 Primary Nodes, DRM Master and Authentication
29 ============================================ 29 ============================================
30 30
31 .. kernel-doc:: drivers/gpu/drm/drm_auth.c 31 .. kernel-doc:: drivers/gpu/drm/drm_auth.c
32 :doc: master and authentication 32 :doc: master and authentication
33 33
34 .. kernel-doc:: drivers/gpu/drm/drm_auth.c 34 .. kernel-doc:: drivers/gpu/drm/drm_auth.c
35 :export: 35 :export:
36 36
37 .. kernel-doc:: include/drm/drm_auth.h 37 .. kernel-doc:: include/drm/drm_auth.h
38 :internal: 38 :internal:
39 39
40 <<
41 .. _drm_leasing: <<
42 <<
43 DRM Display Resource Leasing <<
44 ============================ <<
45 <<
46 .. kernel-doc:: drivers/gpu/drm/drm_lease.c <<
47 :doc: drm leasing <<
48 <<
49 Open-Source Userspace Requirements 40 Open-Source Userspace Requirements
50 ================================== 41 ==================================
51 42
52 The DRM subsystem has stricter requirements th 43 The DRM subsystem has stricter requirements than most other kernel subsystems on
53 what the userspace side for new uAPI needs to 44 what the userspace side for new uAPI needs to look like. This section here
54 explains what exactly those requirements are, 45 explains what exactly those requirements are, and why they exist.
55 46
56 The short summary is that any addition of DRM 47 The short summary is that any addition of DRM uAPI requires corresponding
57 open-sourced userspace patches, and those patc 48 open-sourced userspace patches, and those patches must be reviewed and ready for
58 merging into a suitable and canonical upstream 49 merging into a suitable and canonical upstream project.
59 50
60 GFX devices (both display and render/GPU side) 51 GFX devices (both display and render/GPU side) are really complex bits of
61 hardware, with userspace and kernel by necessi 52 hardware, with userspace and kernel by necessity having to work together really
62 closely. The interfaces, for rendering and mo 53 closely. The interfaces, for rendering and modesetting, must be extremely wide
63 and flexible, and therefore it is almost alway 54 and flexible, and therefore it is almost always impossible to precisely define
64 them for every possible corner case. This in t 55 them for every possible corner case. This in turn makes it really practically
65 infeasible to differentiate between behaviour 56 infeasible to differentiate between behaviour that's required by userspace, and
66 which must not be changed to avoid regressions 57 which must not be changed to avoid regressions, and behaviour which is only an
67 accidental artifact of the current implementat 58 accidental artifact of the current implementation.
68 59
69 Without access to the full source code of all 60 Without access to the full source code of all userspace users that means it
70 becomes impossible to change the implementatio 61 becomes impossible to change the implementation details, since userspace could
71 depend upon the accidental behaviour of the cu 62 depend upon the accidental behaviour of the current implementation in minute
72 details. And debugging such regressions withou 63 details. And debugging such regressions without access to source code is pretty
73 much impossible. As a consequence this means: 64 much impossible. As a consequence this means:
74 65
75 - The Linux kernel's "no regression" policy ho 66 - The Linux kernel's "no regression" policy holds in practice only for
76 open-source userspace of the DRM subsystem. 67 open-source userspace of the DRM subsystem. DRM developers are perfectly fine
77 if closed-source blob drivers in userspace u 68 if closed-source blob drivers in userspace use the same uAPI as the open
78 drivers, but they must do so in the exact sa 69 drivers, but they must do so in the exact same way as the open drivers.
79 Creative (ab)use of the interfaces will, and 70 Creative (ab)use of the interfaces will, and in the past routinely has, lead
80 to breakage. 71 to breakage.
81 72
82 - Any new userspace interface must have an ope 73 - Any new userspace interface must have an open-source implementation as
83 demonstration vehicle. 74 demonstration vehicle.
84 75
85 The other reason for requiring open-source use 76 The other reason for requiring open-source userspace is uAPI review. Since the
86 kernel and userspace parts of a GFX stack must 77 kernel and userspace parts of a GFX stack must work together so closely, code
87 review can only assess whether a new interface 78 review can only assess whether a new interface achieves its goals by looking at
88 both sides. Making sure that the interface ind 79 both sides. Making sure that the interface indeed covers the use-case fully
89 leads to a few additional requirements: 80 leads to a few additional requirements:
90 81
91 - The open-source userspace must not be a toy/ 82 - The open-source userspace must not be a toy/test application, but the real
92 thing. Specifically it needs to handle all t 83 thing. Specifically it needs to handle all the usual error and corner cases.
93 These are often the places where new uAPI fa 84 These are often the places where new uAPI falls apart and hence essential to
94 assess the fitness of a proposed interface. 85 assess the fitness of a proposed interface.
95 86
96 - The userspace side must be fully reviewed an 87 - The userspace side must be fully reviewed and tested to the standards of that
97 userspace project. For e.g. mesa this means 88 userspace project. For e.g. mesa this means piglit testcases and review on the
98 mailing list. This is again to ensure that t 89 mailing list. This is again to ensure that the new interface actually gets the
99 job done. The userspace-side reviewer shoul 90 job done. The userspace-side reviewer should also provide an Acked-by on the
100 kernel uAPI patch indicating that they belie 91 kernel uAPI patch indicating that they believe the proposed uAPI is sound and
101 sufficiently documented and validated for us 92 sufficiently documented and validated for userspace's consumption.
102 93
103 - The userspace patches must be against the ca 94 - The userspace patches must be against the canonical upstream, not some vendor
104 fork. This is to make sure that no one cheat 95 fork. This is to make sure that no one cheats on the review and testing
105 requirements by doing a quick fork. 96 requirements by doing a quick fork.
106 97
107 - The kernel patch can only be merged after al 98 - The kernel patch can only be merged after all the above requirements are met,
108 but it **must** be merged to either drm-next 99 but it **must** be merged to either drm-next or drm-misc-next **before** the
109 userspace patches land. uAPI always flows fr 100 userspace patches land. uAPI always flows from the kernel, doing things the
110 other way round risks divergence of the uAPI 101 other way round risks divergence of the uAPI definitions and header files.
111 102
112 These are fairly steep requirements, but have 103 These are fairly steep requirements, but have grown out from years of shared
113 pain and experience with uAPI added hastily, a 104 pain and experience with uAPI added hastily, and almost always regretted about
114 just as fast. GFX devices change really fast, 105 just as fast. GFX devices change really fast, requiring a paradigm shift and
115 entire new set of uAPI interfaces every few ye 106 entire new set of uAPI interfaces every few years at least. Together with the
116 Linux kernel's guarantee to keep existing user 107 Linux kernel's guarantee to keep existing userspace running for 10+ years this
117 is already rather painful for the DRM subsyste 108 is already rather painful for the DRM subsystem, with multiple different uAPIs
118 for the same thing co-existing. If we add a fe 109 for the same thing co-existing. If we add a few more complete mistakes into the
119 mix every year it would be entirely unmanageab 110 mix every year it would be entirely unmanageable.
120 111
121 .. _drm_render_node: 112 .. _drm_render_node:
122 113
123 Render nodes 114 Render nodes
124 ============ 115 ============
125 116
126 DRM core provides multiple character-devices f 117 DRM core provides multiple character-devices for user-space to use.
127 Depending on which device is opened, user-spac 118 Depending on which device is opened, user-space can perform a different
128 set of operations (mainly ioctls). The primary 119 set of operations (mainly ioctls). The primary node is always created
129 and called card<num>. Additionally, a currentl 120 and called card<num>. Additionally, a currently unused control node,
130 called controlD<num> is also created. The prim 121 called controlD<num> is also created. The primary node provides all
131 legacy operations and historically was the onl 122 legacy operations and historically was the only interface used by
132 userspace. With KMS, the control node was intr 123 userspace. With KMS, the control node was introduced. However, the
133 planned KMS control interface has never been w 124 planned KMS control interface has never been written and so the control
134 node stays unused to date. 125 node stays unused to date.
135 126
136 With the increased use of offscreen renderers 127 With the increased use of offscreen renderers and GPGPU applications,
137 clients no longer require running compositors 128 clients no longer require running compositors or graphics servers to
138 make use of a GPU. But the DRM API required un 129 make use of a GPU. But the DRM API required unprivileged clients to
139 authenticate to a DRM-Master prior to getting 130 authenticate to a DRM-Master prior to getting GPU access. To avoid this
140 step and to grant clients GPU access without a 131 step and to grant clients GPU access without authenticating, render
141 nodes were introduced. Render nodes solely ser 132 nodes were introduced. Render nodes solely serve render clients, that
142 is, no modesetting or privileged ioctls can be 133 is, no modesetting or privileged ioctls can be issued on render nodes.
143 Only non-global rendering commands are allowed 134 Only non-global rendering commands are allowed. If a driver supports
144 render nodes, it must advertise it via the DRI 135 render nodes, it must advertise it via the DRIVER_RENDER DRM driver
145 capability. If not supported, the primary node 136 capability. If not supported, the primary node must be used for render
146 clients together with the legacy drmAuth authe 137 clients together with the legacy drmAuth authentication procedure.
147 138
148 If a driver advertises render node support, DR 139 If a driver advertises render node support, DRM core will create a
149 separate render node called renderD<num>. Ther 140 separate render node called renderD<num>. There will be one render node
150 per device. No ioctls except PRIME-related ioc 141 per device. No ioctls except PRIME-related ioctls will be allowed on
151 this node. Especially GEM_OPEN will be explici !! 142 this node. Especially GEM_OPEN will be explicitly prohibited. Render
152 complete list of driver-independent ioctls tha <<
153 nodes, see the ioctls marked DRM_RENDER_ALLOW <<
154 nodes are designed to avoid the buffer-leaks, 143 nodes are designed to avoid the buffer-leaks, which occur if clients
155 guess the flink names or mmap offsets on the l 144 guess the flink names or mmap offsets on the legacy interface.
156 Additionally to this basic interface, drivers 145 Additionally to this basic interface, drivers must mark their
157 driver-dependent render-only ioctls as DRM_REN 146 driver-dependent render-only ioctls as DRM_RENDER_ALLOW so render
158 clients can use them. Driver authors must be c 147 clients can use them. Driver authors must be careful not to allow any
159 privileged ioctls on render nodes. 148 privileged ioctls on render nodes.
160 149
161 With render nodes, user-space can now control 150 With render nodes, user-space can now control access to the render node
162 via basic file-system access-modes. A running 151 via basic file-system access-modes. A running graphics server which
163 authenticates clients on the privileged primar 152 authenticates clients on the privileged primary/legacy node is no longer
164 required. Instead, a client can open the rende 153 required. Instead, a client can open the render node and is immediately
165 granted GPU access. Communication between clie 154 granted GPU access. Communication between clients (or servers) is done
166 via PRIME. FLINK from render node to legacy no 155 via PRIME. FLINK from render node to legacy node is not supported. New
167 clients must not use the insecure FLINK interf 156 clients must not use the insecure FLINK interface.
168 157
169 Besides dropping all modeset/global ioctls, re 158 Besides dropping all modeset/global ioctls, render nodes also drop the
170 DRM-Master concept. There is no reason to asso 159 DRM-Master concept. There is no reason to associate render clients with
171 a DRM-Master as they are independent of any gr 160 a DRM-Master as they are independent of any graphics server. Besides,
172 they must work without any running master, any 161 they must work without any running master, anyway. Drivers must be able
173 to run without a master object if they support 162 to run without a master object if they support render nodes. If, on the
174 other hand, a driver requires shared state bet 163 other hand, a driver requires shared state between clients which is
175 visible to user-space and accessible beyond op 164 visible to user-space and accessible beyond open-file boundaries, they
176 cannot support render nodes. 165 cannot support render nodes.
177 166
178 Device Hot-Unplug 167 Device Hot-Unplug
179 ================= 168 =================
180 169
181 .. note:: 170 .. note::
182 The following is the plan. Implementation i 171 The following is the plan. Implementation is not there yet
183 (2020 May). 172 (2020 May).
184 173
185 Graphics devices (display and/or render) may b 174 Graphics devices (display and/or render) may be connected via USB (e.g.
186 display adapters or docking stations) or Thund 175 display adapters or docking stations) or Thunderbolt (e.g. eGPU). An end
187 user is able to hot-unplug this kind of device 176 user is able to hot-unplug this kind of devices while they are being
188 used, and expects that the very least the mach 177 used, and expects that the very least the machine does not crash. Any
189 damage from hot-unplugging a DRM device needs 178 damage from hot-unplugging a DRM device needs to be limited as much as
190 possible and userspace must be given the chanc 179 possible and userspace must be given the chance to handle it if it wants
191 to. Ideally, unplugging a DRM device still let 180 to. Ideally, unplugging a DRM device still lets a desktop continue to
192 run, but that is going to need explicit suppor 181 run, but that is going to need explicit support throughout the whole
193 graphics stack: from kernel and userspace driv 182 graphics stack: from kernel and userspace drivers, through display
194 servers, via window system protocols, and in a 183 servers, via window system protocols, and in applications and libraries.
195 184
196 Other scenarios that should lead to the same a 185 Other scenarios that should lead to the same are: unrecoverable GPU
197 crash, PCI device disappearing off the bus, or 186 crash, PCI device disappearing off the bus, or forced unbind of a driver
198 from the physical device. 187 from the physical device.
199 188
200 In other words, from userspace perspective eve 189 In other words, from userspace perspective everything needs to keep on
201 working more or less, until userspace stops us 190 working more or less, until userspace stops using the disappeared DRM
202 device and closes it completely. Userspace wil 191 device and closes it completely. Userspace will learn of the device
203 disappearance from the device removed uevent, 192 disappearance from the device removed uevent, ioctls returning ENODEV
204 (or driver-specific ioctls returning driver-sp 193 (or driver-specific ioctls returning driver-specific things), or open()
205 returning ENXIO. 194 returning ENXIO.
206 195
207 Only after userspace has closed all relevant D 196 Only after userspace has closed all relevant DRM device and dmabuf file
208 descriptors and removed all mmaps, the DRM dri 197 descriptors and removed all mmaps, the DRM driver can tear down its
209 instance for the device that no longer exists. 198 instance for the device that no longer exists. If the same physical
210 device somehow comes back in the mean time, it 199 device somehow comes back in the mean time, it shall be a new DRM
211 device. 200 device.
212 201
213 Similar to PIDs, chardev minor numbers are not 202 Similar to PIDs, chardev minor numbers are not recycled immediately. A
214 new DRM device always picks the next free mino 203 new DRM device always picks the next free minor number compared to the
215 previous one allocated, and wraps around when 204 previous one allocated, and wraps around when minor numbers are
216 exhausted. 205 exhausted.
217 206
218 The goal raises at least the following require 207 The goal raises at least the following requirements for the kernel and
219 drivers. 208 drivers.
220 209
221 Requirements for KMS UAPI 210 Requirements for KMS UAPI
222 ------------------------- 211 -------------------------
223 212
224 - KMS connectors must change their status to d 213 - KMS connectors must change their status to disconnected.
225 214
226 - Legacy modesets and pageflips, and atomic co 215 - Legacy modesets and pageflips, and atomic commits, both real and
227 TEST_ONLY, and any other ioctls either fail 216 TEST_ONLY, and any other ioctls either fail with ENODEV or fake
228 success. 217 success.
229 218
230 - Pending non-blocking KMS operations deliver 219 - Pending non-blocking KMS operations deliver the DRM events userspace
231 is expecting. This applies also to ioctls th 220 is expecting. This applies also to ioctls that faked success.
232 221
233 - open() on a device node whose underlying dev 222 - open() on a device node whose underlying device has disappeared will
234 fail with ENXIO. 223 fail with ENXIO.
235 224
236 - Attempting to create a DRM lease on a disapp 225 - Attempting to create a DRM lease on a disappeared DRM device will
237 fail with ENODEV. Existing DRM leases remain 226 fail with ENODEV. Existing DRM leases remain and work as listed
238 above. 227 above.
239 228
240 Requirements for Render and Cross-Device UAPI 229 Requirements for Render and Cross-Device UAPI
241 --------------------------------------------- 230 ---------------------------------------------
242 231
243 - All GPU jobs that can no longer run must hav 232 - All GPU jobs that can no longer run must have their fences
244 force-signalled to avoid inflicting hangs on 233 force-signalled to avoid inflicting hangs on userspace.
245 The associated error code is ENODEV. 234 The associated error code is ENODEV.
246 235
247 - Some userspace APIs already define what shou 236 - Some userspace APIs already define what should happen when the device
248 disappears (OpenGL, GL ES: `GL_KHR_robustnes 237 disappears (OpenGL, GL ES: `GL_KHR_robustness`_; `Vulkan`_:
249 VK_ERROR_DEVICE_LOST; etc.). DRM drivers are 238 VK_ERROR_DEVICE_LOST; etc.). DRM drivers are free to implement this
250 behaviour the way they see best, e.g. return 239 behaviour the way they see best, e.g. returning failures in
251 driver-specific ioctls and handling those in 240 driver-specific ioctls and handling those in userspace drivers, or
252 rely on uevents, and so on. 241 rely on uevents, and so on.
253 242
254 - dmabuf which point to memory that has disapp 243 - dmabuf which point to memory that has disappeared will either fail to
255 import with ENODEV or continue to be success 244 import with ENODEV or continue to be successfully imported if it would
256 have succeeded before the disappearance. See 245 have succeeded before the disappearance. See also about memory maps
257 below for already imported dmabufs. 246 below for already imported dmabufs.
258 247
259 - Attempting to import a dmabuf to a disappear 248 - Attempting to import a dmabuf to a disappeared device will either fail
260 with ENODEV or succeed if it would have succ 249 with ENODEV or succeed if it would have succeeded without the
261 disappearance. 250 disappearance.
262 251
263 - open() on a device node whose underlying dev 252 - open() on a device node whose underlying device has disappeared will
264 fail with ENXIO. 253 fail with ENXIO.
265 254
266 .. _GL_KHR_robustness: https://www.khronos.org 255 .. _GL_KHR_robustness: https://www.khronos.org/registry/OpenGL/extensions/KHR/KHR_robustness.txt
267 .. _Vulkan: https://www.khronos.org/vulkan/ 256 .. _Vulkan: https://www.khronos.org/vulkan/
268 257
269 Requirements for Memory Maps 258 Requirements for Memory Maps
270 ---------------------------- 259 ----------------------------
271 260
272 Memory maps have further requirements that app 261 Memory maps have further requirements that apply to both existing maps
273 and maps created after the device has disappea 262 and maps created after the device has disappeared. If the underlying
274 memory disappears, the map is created or modif 263 memory disappears, the map is created or modified such that reads and
275 writes will still complete successfully but th 264 writes will still complete successfully but the result is undefined.
276 This applies to both userspace mmap()'d memory 265 This applies to both userspace mmap()'d memory and memory pointed to by
277 dmabuf which might be mapped to other devices 266 dmabuf which might be mapped to other devices (cross-device dmabuf
278 imports). 267 imports).
279 268
280 Raising SIGBUS is not an option, because users 269 Raising SIGBUS is not an option, because userspace cannot realistically
281 handle it. Signal handlers are global, which m 270 handle it. Signal handlers are global, which makes them extremely
282 difficult to use correctly from libraries like 271 difficult to use correctly from libraries like those that Mesa produces.
283 Signal handlers are not composable, you can't 272 Signal handlers are not composable, you can't have different handlers
284 for GPU1 and GPU2 from different vendors, and 273 for GPU1 and GPU2 from different vendors, and a third handler for
285 mmapped regular files. Threads cause additiona 274 mmapped regular files. Threads cause additional pain with signal
286 handling as well. 275 handling as well.
287 276
288 Device reset <<
289 ============ <<
290 <<
291 The GPU stack is really complex and is prone t <<
292 faulty applications and everything in between <<
293 require resetting the device in order to make <<
294 section describes the expectations for DRM and <<
295 device resets and how to propagate the reset s <<
296 <<
297 Device resets can not be disabled without tain <<
298 hanging the entire kernel through shrinkers/mm <<
299 device resets is to propagate the message to t <<
300 special policy for blocking guilty application <<
301 debugging a hung GPU context require hardware <<
302 a GPU context while it's stopped. <<
303 <<
304 Kernel Mode Driver <<
305 ------------------ <<
306 <<
307 The KMD is responsible for checking if the dev <<
308 it as needed. Usually a hang is detected when <<
309 should keep track of resets, because userspace <<
310 reset status for a specific context. This is n <<
311 the stack that a reset has happened. Currently <<
312 driver separately, with no common DRM interfac <<
313 integrated at DRM scheduler to provide a commo <<
314 reset, KMD should reject new command submissio <<
315 <<
316 User Mode Driver <<
317 ---------------- <<
318 <<
319 After command submission, UMD should check if <<
320 rejected. After a reset, KMD should reject sub <<
321 ioctl to the KMD to check the reset status, an <<
322 if the UMD requires it. After detecting a rese <<
323 it to the application using the appropriate AP <<
324 section below about robustness. <<
325 <<
326 Robustness <<
327 ---------- <<
328 <<
329 The only way to try to keep a graphical API co <<
330 it complies with the robustness aspects of the <<
331 <<
332 Graphical APIs provide ways to applications to <<
333 there is no guarantee that the app will use su <<
334 userspace that doesn't support robust interfac <<
335 OpenGL context or API without any robustness s <<
336 robustness handling entirely to the userspace <<
337 community consensus on what the userspace driv <<
338 since all reasonable approaches have some clea <<
339 <<
340 OpenGL <<
341 ~~~~~~ <<
342 <<
343 Apps using OpenGL should use the available rob <<
344 extension ``GL_ARB_robustness`` (or ``GL_EXT_r <<
345 interface tells if a reset has happened, and i <<
346 considered lost and the app proceeds by creati <<
347 on what to do to if robustness is not in use. <<
348 <<
349 Vulkan <<
350 ~~~~~~ <<
351 <<
352 Apps using Vulkan should check for ``VK_ERROR_ <<
353 This error code means, among other things, tha <<
354 it needs to recreate the contexts to keep goin <<
355 <<
356 Reporting causes of resets <<
357 -------------------------- <<
358 <<
359 Apart from propagating the reset through the s <<
360 really useful for driver developers to learn m <<
361 the first place. DRM devices should make use o <<
362 information about the reset, so this informati <<
363 reports. <<
364 <<
365 .. _drm_driver_ioctl: 277 .. _drm_driver_ioctl:
366 278
367 IOCTL Support on Device Nodes 279 IOCTL Support on Device Nodes
368 ============================= 280 =============================
369 281
370 .. kernel-doc:: drivers/gpu/drm/drm_ioctl.c 282 .. kernel-doc:: drivers/gpu/drm/drm_ioctl.c
371 :doc: driver specific ioctls 283 :doc: driver specific ioctls
372 284
373 Recommended IOCTL Return Values 285 Recommended IOCTL Return Values
374 ------------------------------- 286 -------------------------------
375 287
376 In theory a driver's IOCTL callback is only al 288 In theory a driver's IOCTL callback is only allowed to return very few error
377 codes. In practice it's good to abuse a few mo 289 codes. In practice it's good to abuse a few more. This section documents common
378 practice within the DRM subsystem: 290 practice within the DRM subsystem:
379 291
380 ENOENT: 292 ENOENT:
381 Strictly this should only be used when 293 Strictly this should only be used when a file doesn't exist e.g. when
382 calling the open() syscall. We reuse t 294 calling the open() syscall. We reuse that to signal any kind of object
383 lookup failure, e.g. for unknown GEM b 295 lookup failure, e.g. for unknown GEM buffer object handles, unknown KMS
384 object handles and similar cases. 296 object handles and similar cases.
385 297
386 ENOSPC: 298 ENOSPC:
387 Some drivers use this to differentiate 299 Some drivers use this to differentiate "out of kernel memory" from "out
388 of VRAM". Sometimes also applies to ot 300 of VRAM". Sometimes also applies to other limited gpu resources used for
389 rendering (e.g. when you have a specia 301 rendering (e.g. when you have a special limited compression buffer).
390 Sometimes resource allocation/reservat 302 Sometimes resource allocation/reservation issues in command submission
391 IOCTLs are also signalled through EDEA 303 IOCTLs are also signalled through EDEADLK.
392 304
393 Simply running out of kernel/system me 305 Simply running out of kernel/system memory is signalled through ENOMEM.
394 306
395 EPERM/EACCES: 307 EPERM/EACCES:
396 Returned for an operation that is vali 308 Returned for an operation that is valid, but needs more privileges.
397 E.g. root-only or much more common, DR 309 E.g. root-only or much more common, DRM master-only operations return
398 this when called by unpriviledged clie 310 this when called by unpriviledged clients. There's no clear
399 difference between EACCES and EPERM. 311 difference between EACCES and EPERM.
400 312
401 ENODEV: 313 ENODEV:
402 The device is not present anymore or i 314 The device is not present anymore or is not yet fully initialized.
403 315
404 EOPNOTSUPP: 316 EOPNOTSUPP:
405 Feature (like PRIME, modesetting, GEM) 317 Feature (like PRIME, modesetting, GEM) is not supported by the driver.
406 318
407 ENXIO: 319 ENXIO:
408 Remote failure, either a hardware tran 320 Remote failure, either a hardware transaction (like i2c), but also used
409 when the exporting driver of a shared 321 when the exporting driver of a shared dma-buf or fence doesn't support a
410 feature needed. 322 feature needed.
411 323
412 EINTR: 324 EINTR:
413 DRM drivers assume that userspace rest 325 DRM drivers assume that userspace restarts all IOCTLs. Any DRM IOCTL can
414 return EINTR and in such a case should 326 return EINTR and in such a case should be restarted with the IOCTL
415 parameters left unchanged. 327 parameters left unchanged.
416 328
417 EIO: 329 EIO:
418 The GPU died and couldn't be resurrect 330 The GPU died and couldn't be resurrected through a reset. Modesetting
419 hardware failures are signalled throug 331 hardware failures are signalled through the "link status" connector
420 property. 332 property.
421 333
422 EINVAL: 334 EINVAL:
423 Catch-all for anything that is an inva 335 Catch-all for anything that is an invalid argument combination which
424 cannot work. 336 cannot work.
425 337
426 IOCTL also use other error codes like ETIME, E 338 IOCTL also use other error codes like ETIME, EFAULT, EBUSY, ENOTTY but their
427 usage is in line with the common meanings. The 339 usage is in line with the common meanings. The above list tries to just document
428 DRM specific patterns. Note that ENOTTY has th 340 DRM specific patterns. Note that ENOTTY has the slightly unintuitive meaning of
429 "this IOCTL does not exist", and is used exact 341 "this IOCTL does not exist", and is used exactly as such in DRM.
430 342
431 .. kernel-doc:: include/drm/drm_ioctl.h 343 .. kernel-doc:: include/drm/drm_ioctl.h
432 :internal: 344 :internal:
433 345
434 .. kernel-doc:: drivers/gpu/drm/drm_ioctl.c 346 .. kernel-doc:: drivers/gpu/drm/drm_ioctl.c
435 :export: 347 :export:
436 348
437 .. kernel-doc:: drivers/gpu/drm/drm_ioc32.c 349 .. kernel-doc:: drivers/gpu/drm/drm_ioc32.c
438 :export: 350 :export:
439 351
440 Testing and validation 352 Testing and validation
441 ====================== 353 ======================
442 354
443 Testing Requirements for userspace API 355 Testing Requirements for userspace API
444 -------------------------------------- 356 --------------------------------------
445 357
446 New cross-driver userspace interface extension 358 New cross-driver userspace interface extensions, like new IOCTL, new KMS
447 properties, new files in sysfs or anything els 359 properties, new files in sysfs or anything else that constitutes an API change
448 should have driver-agnostic testcases in IGT f 360 should have driver-agnostic testcases in IGT for that feature, if such a test
449 can be reasonably made using IGT for the targe 361 can be reasonably made using IGT for the target hardware.
450 362
451 Validating changes with IGT 363 Validating changes with IGT
452 --------------------------- 364 ---------------------------
453 365
454 There's a collection of tests that aims to cov 366 There's a collection of tests that aims to cover the whole functionality of
455 DRM drivers and that can be used to check that 367 DRM drivers and that can be used to check that changes to DRM drivers or the
456 core don't regress existing functionality. Thi 368 core don't regress existing functionality. This test suite is called IGT and
457 its code and instructions to build and run can 369 its code and instructions to build and run can be found in
458 https://gitlab.freedesktop.org/drm/igt-gpu-too 370 https://gitlab.freedesktop.org/drm/igt-gpu-tools/.
459 371
460 Using VKMS to test DRM API 372 Using VKMS to test DRM API
461 -------------------------- 373 --------------------------
462 374
463 VKMS is a software-only model of a KMS driver 375 VKMS is a software-only model of a KMS driver that is useful for testing
464 and for running compositors. VKMS aims to enab 376 and for running compositors. VKMS aims to enable a virtual display without
465 the need for a hardware display capability. Th 377 the need for a hardware display capability. These characteristics made VKMS
466 a perfect tool for validating the DRM core beh 378 a perfect tool for validating the DRM core behavior and also support the
467 compositor developer. VKMS makes it possible t 379 compositor developer. VKMS makes it possible to test DRM functions in a
468 virtual machine without display, simplifying t 380 virtual machine without display, simplifying the validation of some of the
469 core changes. 381 core changes.
470 382
471 To Validate changes in DRM API with VKMS, star 383 To Validate changes in DRM API with VKMS, start setting the kernel: make
472 sure to enable VKMS module; compile the kernel 384 sure to enable VKMS module; compile the kernel with the VKMS enabled and
473 install it in the target machine. VKMS can be 385 install it in the target machine. VKMS can be run in a Virtual Machine
474 (QEMU, virtme or similar). It's recommended th 386 (QEMU, virtme or similar). It's recommended the use of KVM with the minimum
475 of 1GB of RAM and four cores. 387 of 1GB of RAM and four cores.
476 388
477 It's possible to run the IGT-tests in a VM in 389 It's possible to run the IGT-tests in a VM in two ways:
478 390
479 1. Use IGT inside a VM 391 1. Use IGT inside a VM
480 2. Use IGT from the host machine and w 392 2. Use IGT from the host machine and write the results in a shared directory.
481 393
482 Following is an example of using a VM with a s !! 394 As follow, there is an example of using a VM with a shared directory with
483 the host machine to run igt-tests. This exampl !! 395 the host machine to run igt-tests. As an example it's used virtme::
484 396
485 $ virtme-run --rwdir /path/for/shared_ 397 $ virtme-run --rwdir /path/for/shared_dir --kdir=path/for/kernel/directory --mods=auto
486 398
487 Run the igt-tests in the guest machine. This e !! 399 Run the igt-tests in the guest machine, as example it's ran the 'kms_flip'
488 tests:: 400 tests::
489 401
490 $ /path/for/igt-gpu-tools/scripts/run- 402 $ /path/for/igt-gpu-tools/scripts/run-tests.sh -p -s -t "kms_flip.*" -v
491 403
492 In this example, instead of building the igt_r !! 404 In this example, instead of build the igt_runner, Piglit is used
493 (-p option). It creates an HTML summary of the !! 405 (-p option); it's created html summary of the tests results and it's saved
494 them in the folder "igt-gpu-tools/results". It !! 406 in the folder "igt-gpu-tools/results"; it's executed only the igt-tests
495 matching the -t option. 407 matching the -t option.
496 408
497 Display CRC Support 409 Display CRC Support
498 ------------------- 410 -------------------
499 411
500 .. kernel-doc:: drivers/gpu/drm/drm_debugfs_cr 412 .. kernel-doc:: drivers/gpu/drm/drm_debugfs_crc.c
501 :doc: CRC ABI 413 :doc: CRC ABI
502 414
503 .. kernel-doc:: drivers/gpu/drm/drm_debugfs_cr 415 .. kernel-doc:: drivers/gpu/drm/drm_debugfs_crc.c
504 :export: 416 :export:
505 417
506 Debugfs Support 418 Debugfs Support
507 --------------- 419 ---------------
508 420
509 .. kernel-doc:: include/drm/drm_debugfs.h 421 .. kernel-doc:: include/drm/drm_debugfs.h
510 :internal: 422 :internal:
511 423
512 .. kernel-doc:: drivers/gpu/drm/drm_debugfs.c 424 .. kernel-doc:: drivers/gpu/drm/drm_debugfs.c
513 :export: 425 :export:
514 426
515 Sysfs Support 427 Sysfs Support
516 ============= 428 =============
517 429
518 .. kernel-doc:: drivers/gpu/drm/drm_sysfs.c 430 .. kernel-doc:: drivers/gpu/drm/drm_sysfs.c
519 :doc: overview 431 :doc: overview
520 432
521 .. kernel-doc:: drivers/gpu/drm/drm_sysfs.c 433 .. kernel-doc:: drivers/gpu/drm/drm_sysfs.c
522 :export: 434 :export:
523 435
524 436
525 VBlank event handling 437 VBlank event handling
526 ===================== 438 =====================
527 439
528 The DRM core exposes two vertical blank relate 440 The DRM core exposes two vertical blank related ioctls:
529 441
530 :c:macro:`DRM_IOCTL_WAIT_VBLANK` !! 442 DRM_IOCTL_WAIT_VBLANK
531 This takes a struct drm_wait_vblank struct 443 This takes a struct drm_wait_vblank structure as its argument, and
532 it is used to block or request a signal wh 444 it is used to block or request a signal when a specified vblank
533 event occurs. 445 event occurs.
534 446
535 :c:macro:`DRM_IOCTL_MODESET_CTL` !! 447 DRM_IOCTL_MODESET_CTL
536 This was only used for user-mode-settind d 448 This was only used for user-mode-settind drivers around modesetting
537 changes to allow the kernel to update the 449 changes to allow the kernel to update the vblank interrupt after
538 mode setting, since on many devices the ve 450 mode setting, since on many devices the vertical blank counter is
539 reset to 0 at some point during modeset. M 451 reset to 0 at some point during modeset. Modern drivers should not
540 call this any more since with kernel mode 452 call this any more since with kernel mode setting it is a no-op.
541 453
542 Userspace API Structures 454 Userspace API Structures
543 ======================== 455 ========================
544 456
545 .. kernel-doc:: include/uapi/drm/drm_mode.h 457 .. kernel-doc:: include/uapi/drm/drm_mode.h
546 :doc: overview 458 :doc: overview
547 459
548 .. _crtc_index: <<
549 <<
550 CRTC index <<
551 ---------- <<
552 <<
553 CRTC's have both an object ID and an index, an <<
554 The index is used in cases where a densely pac <<
555 needed, for instance a bitmask of CRTC's. The <<
556 drm_mode_get_plane is an example. <<
557 <<
558 :c:macro:`DRM_IOCTL_MODE_GETRESOURCES` populat <<
559 CRTC ID's, and the CRTC index is its position <<
560 <<
561 .. kernel-doc:: include/uapi/drm/drm.h 460 .. kernel-doc:: include/uapi/drm/drm.h
562 :internal: 461 :internal:
563 462
564 .. kernel-doc:: include/uapi/drm/drm_mode.h 463 .. kernel-doc:: include/uapi/drm/drm_mode.h
565 :internal: 464 :internal:
566 <<
567 <<
568 dma-buf interoperability <<
569 ======================== <<
570 <<
571 Please see Documentation/userspace-api/dma-buf <<
572 information on how dma-buf is integrated and e <<
Linux® is a registered trademark of Linus Torvalds in the United States and other countries.
TOMOYO® is a registered trademark of NTT DATA CORPORATION.