1 .. SPDX-License-Identifier: GPL-2.0
6 Author: Guennadi Liakhovetski <g.liakhovetski@gmx.de>
11 The following terms are used in this document:
12 - camera / camera device / camera sensor - a video-camera sensor chip, capable
13 of connecting to a variety of systems and interfaces, typically uses i2c for
14 control and configuration, and a parallel or a serial bus for data.
15 - camera host - an interface, to which a camera is connected. Typically a
16 specialised interface, present on many SoCs, e.g. PXA27x and PXA3xx, SuperH,
18 - camera host bus - a connection between a camera host and a camera. Can be
19 parallel or serial, consists of data and control lines, e.g. clock, vertical
20 and horizontal synchronization signals.
22 Purpose of the soc-camera subsystem
23 -----------------------------------
25 The soc-camera subsystem initially provided a unified API between camera host
26 drivers and camera sensor drivers. Later the soc-camera sensor API has been
27 replaced with the V4L2 standard subdev API. This also made camera driver re-use
28 with non-soc-camera hosts possible. The camera host API to the soc-camera core
31 Soc-camera implements a V4L2 interface to the user, currently only the "mmap"
32 method is supported by host drivers. However, the soc-camera core also provides
33 support for the "read" method.
35 The subsystem has been designed to support multiple camera host interfaces and
36 multiple cameras per interface, although most applications have only one camera
42 As of 3.7 there are seven host drivers in the mainline: atmel-isi.c,
43 mx1_camera.c (broken, scheduled for removal), mx2_camera.c, mx3_camera.c,
44 omap1_camera.c, pxa_camera.c, sh_mobile_ceu_camera.c, and multiple sensor
45 drivers under drivers/media/i2c/soc_camera/.
50 A host camera driver is registered using the
54 soc_camera_host_register(struct soc_camera_host *);
56 function. The host object can be initialized as follows:
60 struct soc_camera_host *ici;
61 ici->drv_name = DRV_NAME;
62 ici->ops = &camera_host_ops;
64 ici->v4l2_dev.dev = &pdev->dev;
67 All camera host methods are passed in a struct soc_camera_host_ops:
71 static struct soc_camera_host_ops camera_host_ops = {
73 .add = camera_add_device,
74 .remove = camera_remove_device,
75 .set_fmt = camera_set_fmt_cap,
76 .try_fmt = camera_try_fmt_cap,
77 .init_videobuf2 = camera_init_videobuf2,
79 .querycap = camera_querycap,
80 .set_bus_param = camera_set_bus_param,
81 /* The rest of host operations are optional */
84 .add and .remove methods are called when a sensor is attached to or detached
85 from the host. .set_bus_param is used to configure physical connection
86 parameters between the host and the sensor. .init_videobuf2 is called by
87 soc-camera core when a video-device is opened, the host driver would typically
88 call vb2_queue_init() in this method. Further video-buffer management is
89 implemented completely by the specific camera host driver. If the host driver
90 supports non-standard pixel format conversion, it should implement a
91 .get_formats and, possibly, a .put_formats operations. See below for more
92 details about format conversion. The rest of the methods are called from
93 respective V4L2 operations.
98 Sensor drivers can use struct soc_camera_link, typically provided by the
99 platform, and used to specify to which camera host bus the sensor is connected,
100 and optionally provide platform .power and .reset methods for the camera. This
101 struct is provided to the camera driver via the I2C client device platform data
102 and can be obtained, using the soc_camera_i2c_to_link() macro. Care should be
103 taken, when using soc_camera_vdev_to_subdev() and when accessing struct
104 soc_camera_device, using v4l2_get_subdev_hostdata(): both only work, when
105 running on an soc-camera host. The actual camera driver operation is implemented
106 using the V4L2 subdev API. Additionally soc-camera camera drivers can use
107 auxiliary soc-camera helper functions like soc_camera_power_on() and
108 soc_camera_power_off(), which switch regulators, provided by the platform and call
109 board-specific power switching methods. soc_camera_apply_board_flags() takes
110 camera bus configuration capability flags and applies any board transformations,
111 e.g. signal polarity inversion. soc_mbus_get_fmtdesc() can be used to obtain a
112 pixel format descriptor, corresponding to a certain media-bus pixel format code.
113 soc_camera_limit_side() can be used to restrict beginning and length of a frame
114 side, based on camera capabilities.
116 VIDIOC_S_CROP and VIDIOC_S_FMT behaviour
117 ----------------------------------------
119 Above user ioctls modify image geometry as follows:
121 VIDIOC_S_CROP: sets location and sizes of the sensor window. Unit is one sensor
122 pixel. Changing sensor window sizes preserves any scaling factors, therefore
123 user window sizes change as well.
125 VIDIOC_S_FMT: sets user window. Should preserve previously set sensor window as
126 much as possible by modifying scaling factors. If the sensor window cannot be
127 preserved precisely, it may be changed too.
129 In soc-camera there are two locations, where scaling and cropping can take
130 place: in the camera driver and in the host driver. User ioctls are first passed
131 to the host driver, which then generally passes them down to the camera driver.
132 It is more efficient to perform scaling and cropping in the camera driver to
133 save camera bus bandwidth and maximise the framerate. However, if the camera
134 driver failed to set the required parameters with sufficient precision, the host
135 driver may decide to also use its own scaling and cropping to fulfill the user's
138 Camera drivers are interfaced to the soc-camera core and to host drivers over
139 the v4l2-subdev API, which is completely functional, it doesn't pass any data.
140 Therefore all camera drivers shall reply to .g_fmt() requests with their current
141 output geometry. This is necessary to correctly configure the camera bus.
142 .s_fmt() and .try_fmt() have to be implemented too. Sensor window and scaling
143 factors have to be maintained by camera drivers internally. According to the
144 V4L2 API all capture drivers must support the VIDIOC_CROPCAP ioctl, hence we
145 rely on camera drivers implementing .cropcap(). If the camera driver does not
146 support cropping, it may choose to not implement .s_crop(), but to enable
147 cropping support by the camera host driver at least the .g_crop method must be
150 User window geometry is kept in .user_width and .user_height fields in struct
151 soc_camera_device and used by the soc-camera core and host drivers. The core
152 updates these fields upon successful completion of a .s_fmt() call, but if these
153 fields change elsewhere, e.g. during .s_crop() processing, the host driver is
154 responsible for updating them.
159 V4L2 distinguishes between pixel formats, as they are stored in memory, and as
160 they are transferred over a media bus. Soc-camera provides support to
161 conveniently manage these formats. A table of standard transformations is
162 maintained by soc-camera core, which describes, what FOURCC pixel format will
163 be obtained, if a media-bus pixel format is stored in memory according to
164 certain rules. E.g. if MEDIA_BUS_FMT_YUYV8_2X8 data is sampled with 8 bits per
165 sample and stored in memory in the little-endian order with no gaps between
166 bytes, data in memory will represent the V4L2_PIX_FMT_YUYV FOURCC format. These
167 standard transformations will be used by soc-camera or by camera host drivers to
168 configure camera drivers to produce the FOURCC format, requested by the user,
169 using the VIDIOC_S_FMT ioctl(). Apart from those standard format conversions,
170 host drivers can also provide their own conversion rules by implementing a
171 .get_formats and, if required, a .put_formats methods.