Debian Codegraph @

unstable/linux-dmabuf/feedback.rst

raw
1 .. Copyright 2021 Simon Ser
2
3 .. contents::
4
5
6 linux-dmabuf feedback introduction
7 ==================================
8
9 linux-dmabuf feedback allows compositors and clients to negotiate optimal buffer
10 allocation parameters. This document will assume that the compositor is using a
11 rendering API such as OpenGL or Vulkan and KMS as the presentation API: even if
12 linux-dmabuf feedback isn't restricted to this use-case, it's the most common.
13
14 linux-dmabuf feedback introduces the following concepts:
15
16 1. A main device. This is the render device that the compositor is using to
17 perform composition. Compositors should always be able to display a buffer
18 submitted by a client, so this device can be used as a fallback in case none
19 of the more optimized code-paths work. Clients should allocate buffers such
20 that they can be imported and textured from the main device.
21
22 2. One or more tranches. Each tranche consists of a target device, allocation
23 flags and a set of format/modifier pairs. A tranche can be seen as a set of
24 formats/modifier pairs that are compatible with the target device.
25
26 A tranche can have the ``scanout`` flag. It means that the target device is
27 a KMS device, and that buffers allocated with one of the format/modifier
28 pairs in the tranche are eligible for direct scanout.
29
30 Clients should use the tranches in order to allocate buffers with the most
31 appropriate format/modifier and also to avoid allocating in private device
32 memory when cross-device operations are going to happen.
33
34 linux-dmabuf feedback implementation notes
35 ==========================================
36
37 This section contains recommendations for client and compositor implementations.
38
39 For clients
40 -----------
41
42 Clients are expected to either pick a fixed DRM format beforehand, or
43 perform the following steps repeatedly until they find a suitable format.
44
45 Basic clients may only support static buffer allocation on startup. These
46 clients should do the following:
47
48 1. Send a ``get_default_feedback`` request to get global feedback.
49 2. Select the device indicated by ``main_device`` for allocation.
50 3. For each tranche:
51
52 1. If ``tranche_target_device`` doesn't match the allocation device, ignore
53 the tranche.
54 2. Accumulate allocation flags from ``tranche_flags``.
55 3. Accumulate format/modifier pairs received via ``tranche_formats`` in a
56 list.
57 4. When the ``tranche_done`` event is received, try to allocate the buffer
58 with the accumulated list of modifiers and allocation flags. If that
59 fails, proceed with the next tranche. If that succeeds, stop the loop.
60
61 4. Destroy the feedback object.
62
63 Tranches are ordered by preference: the more optimized tranches come first. As
64 such, clients should use the first tranche that happens to work.
65
66 Some clients may have already selected the device they want to use beforehand.
67 These clients can ignore the ``main_device`` event, and ignore tranches whose
68 ``tranche_target_device`` doesn't match the selected device. Such clients need
69 to be prepared for the ``wp_linux_buffer_params.create`` request to potentially
70 fail.
71
72 If the client allocates a buffer without specifying explicit modifiers on a
73 device different from the one indicated by ``main_device``, then the client
74 must force a linear layout.
75
76 Some clients might support re-negotiating the buffer format/modifier on the
77 fly. These clients should send a ``get_surface_feedback`` request and keep the
78 feedback object alive after the initial allocation. Each time a new set of
79 feedback parameters is received (ended by the ``done`` event), they should
80 perform the same steps as basic clients described above. They should detect
81 when the optimal allocation parameters didn't change (same
82 format/modifier/flags) to avoid needlessly re-allocating their buffers.
83
84 Some clients might additionally support switching the device used for
85 allocations on the fly. Such clients should send a ``get_surface_feedback``
86 request. For each tranche, select the device indicated by
87 ``tranche_target_device`` for allocation. Accumulate allocation flags (received
88 via ``tranche_flags``) and format/modifier pairs (received via
89 ``tranche_formats``) as usual. When the ``tranche_done`` event is received, try
90 to allocate the buffer with the accumulated list of modifiers and the
91 allocation flags. Try to import the resulting buffer by sending a
92 ``wp_linux_buffer_params.create`` request (this might fail). Repeat with each
93 tranche until an allocation and import succeeds. Each time a new set of
94 feedback parameters is received, they should perform these steps again. They
95 should detect when the optimal allocation parameters didn't change (same
96 device/format/modifier/flags) to avoid needlessly re-allocating their buffers.
97
98 For compositors
99 ---------------
100
101 Basic compositors may only support texturing the DMA-BUFs via a rendering API
102 such as OpenGL or Vulkan. Such compositors can send a single tranche as a reply
103 to both ``get_default_feedback`` and ``get_surface_feedback``. Set the
104 ``main_device`` to the rendering device. Send the tranche with
105 ``tranche_target_device`` set to the rendering device and all of the DRM
106 format/modifier pairs supported by the rendering API. Do not set the
107 ``scanout`` flag in the ``tranche_flags`` event.
108
109 Some compositors may support direct scan-out for full-screen surfaces. These
110 compositors can re-send the feedback parameters when a surface becomes
111 full-screen or leaves full-screen mode if the client has used the
112 ``get_surface_feedback`` request. The non-full-screen feedback parameters are
113 the same as basic compositors described above. The full-screen feedback
114 parameters have two tranches: one with the format/modifier pairs supported by
115 the KMS plane, with the ``scanout`` flag set in the ``tranche_flags`` event and
116 with ``tranche_target_device`` set to the KMS scan-out device; the other with
117 the rest of the format/modifier pairs (supported for texturing, but not for
118 scan-out), without the ``scanout`` flag set in the ``tranche_flags`` event, and
119 with the ``tranche_target_device`` set to the rendering device.
120
121 Some compositors may support direct scan-out for all surfaces. These
122 compositors can send two tranches for surfaces that become candidates for
123 direct scan-out, similarly to compositors supporting direct scan-out for
124 fullscreen surfaces. When a surface stops being a candidate for direct
125 scan-out, compositors should re-send the feedback parameters optimized for
126 texturing only. The way candidates for direct scan-out are selected is
127 compositor policy, a possible implementation is to select as many surfaces as
128 there are available hardware planes, starting from surfaces closer to the eye.
129
130 Some compositors may support multiple devices at the same time. If the
131 compositor supports rendering with a fixed device and direct scan-out on a
132 secondary device, it may send a separate tranche for surfaces displayed on
133 the secondary device that are candidates for direct scan-out. The
134 ``tranche_target_device`` for this tranche will be the secondary device and
135 will not match the ``main_device``.
136
137 Some compositors may support switching their rendering device at runtime or
138 changing their rendering device depending on the surface. When the rendering
139 device changes for a surface, such compositors may re-send the feedback
140 parameters with a different ``main_device``. However there is a risk that
141 clients don't support switching their device at runtime and continue using the
142 previous device. For this reason, compositors should always have a fallback
143 rendering device that they initially send as ``main_device``, such that these
144 clients use said fallback device.
145
146 Compositors should not change the ``main_device`` on-the-fly when explicit
147 modifiers are not supported, because there's a risk of importing buffers
148 with an implicit non-linear modifier as a linear buffer, resulting in
149 misinterpreted buffer contents.
150
151 Compositors should not send feedback parameters if they don't have a fallback
152 path. For instance, compositors shouldn't send a format/modifier supported for
153 direct scan-out but not supported by the rendering API for texturing.
154
155 Compositors can decide to use multiple tranches to describe the allocation
156 parameters optimized for texturing. For example, if there are formats which
157 have a fast texturing path and formats which have a slower texturing path, the
158 compositor can decide to expose two separate tranches.
159
160 Compositors can decide to use intermediate tranches to describe code-paths
161 slower than direct scan-out but faster than texturing. For instance, a
162 compositor could insert an intermediate tranche if it's possible to use a
163 mem2mem device to convert buffers to be able to use scan-out.
164
165 ``dev_t`` encoding
166 ==================
167
168 The protocol carries ``dev_t`` values on the wire using arrays. A compositor
169 written in C can encode the values as follows:
170
171 .. code-block:: c
172
173 struct stat drm_node_stat;
174 struct wl_array dev_array = {
175 .size = sizeof(drm_node_stat.st_rdev),
176 .data = &drm_node_stat.st_rdev,
177 };
178
179 A client can decode the values as follows:
180
181 .. code-block:: c
182
183 dev_t dev;
184 assert(dev_array->size == sizeof(dev));
185 memcpy(&dev, dev_array->data, sizeof(dev));
186
187 Because two DRM nodes can refer to the same DRM device while having different
188 ``dev_t`` values, clients should use ``drmDevicesEqual`` to compare two
189 devices.
190
191 ``format_table`` encoding
192 =========================
193
194 The ``format_table`` event carries a file descriptor containing a list of
195 format + modifier pairs. The list is an array of pairs which can be accessed
196 with this C structure definition:
197
198 .. code-block:: c
199
200 struct dmabuf_format_modifier {
201 uint32_t format;
202 uint32_t pad; /* unused */
203 uint64_t modifier;
204 };
205
206 Integration with other APIs
207 ===========================
208
209 - libdrm: ``drmGetDeviceFromDevId`` returns a ``drmDevice`` from a device ID.
210 - EGL: the `EGL_EXT_device_drm_render_node`_ extension may be used to query the
211 DRM device render node used by a given EGL display. When unavailable, the
212 older `EGL_EXT_device_drm`_ extension may be used as a fallback.
213 - Vulkan: the `VK_EXT_physical_device_drm`_ extension may be used to query the
214 DRM device used by a given ``VkPhysicalDevice``.
215
216 .. _EGL_EXT_device_drm: https://www.khronos.org/registry/EGL/extensions/EXT/EGL_EXT_device_drm.txt
217 .. _EGL_EXT_device_drm_render_node: https://www.khronos.org/registry/EGL/extensions/EXT/EGL_EXT_device_drm_render_node.txt
218 .. _VK_EXT_physical_device_drm: https://www.khronos.org/registry/vulkan/specs/1.2-extensions/man/html/VK_EXT_physical_device_drm.html