Merge tag 'media/v6.4-1' of git://git.kernel.org/pub/scm/linux/kernel/git/mchehab/linux-media

Pull media updates from Mauro Carvalho Chehab: - Removal of some old unused sensor drivers: ad9389b, m5mols, mt9m032, mt9t001, noon010pc30, s5k6aa, sr030pc30 and vs6624 - New i.MX8 image sensor interface driver - Some new RC keymaps - lots of cleanups at atomisp driver to make it support standard features present on other webcam drivers - the cx18 and saa7146 now uses VB2 - lots of cleanups and driver improvements * tag 'media/v6.4-1' of git://git.kernel.org/pub/scm/linux/kernel/git/mchehab/linux-media: (460 commits) media: ov5670: Fix probe on ACPI media: nxp: imx8-isi: Remove 300ms sleep after enabling channel media: nxp: imx8-isi: Replace udelay() with fsleep() media: nxp: imx8-isi: Drop partial support for i.MX8QM and i.MX8QXP media: nxp: Add i.MX8 ISI driver media: dt-bindings: media: Add i.MX8 ISI DT bindings media: atomisp: gmin_platform: Add Lenovo Ideapad Miix 310 gmin_vars media: atomisp: gmin_platform: Make DMI quirks take precedence over the _DSM table media: atomisp: Remove struct atomisp_sub_device index field media: atomisp: Drop support for streaming from 2 sensors at once media: atomisp: Remove atomisp_try_fmt() call from atomisp_set_fmt() media: atomisp: Remove unused ATOM_ISP_MAX_WIDTH_TMP and ATOM_ISP_MAX_HEIGHT_TMP media: atomisp: Remove snr_mbus_fmt local var from atomisp_try_fmt() media: atomisp: Remove custom V4L2_CID_FMT_AUTO control media: atomisp: Remove continuous mode related code from atomisp_set_fmt() media: atomisp: Remove duplicate atomisp_[start|stop]_streaming() prototypes media: atomisp: gc0310: Switch over to ACPI powermanagement media: atomisp: gc0310: Use devm_kzalloc() for data struct media: atomisp: gc0310: Add runtime-pm support media: atomisp: gc0310: Delay power-on till streaming is started ...

Linus Torvalds committed 3 years ago

Commit 4ea95692 parents c8cc58e 73b41dc442 files · +15626 −21778Browse files

Showing the changes to Documentation/userspace-api (12 of 442 files). Show all

Unified · Split

12 files

Documentation/userspace-api/media/v4l/dev-overlay.rst +8 −2
Documentation/userspace-api/media/v4l/pixfmt-compressed.rst +25 −0
Documentation/userspace-api/media/v4l/pixfmt-packed-yuv.rst +28 −0
Documentation/userspace-api/media/v4l/pixfmt-rgb.rst +42 −0
Documentation/userspace-api/media/v4l/pixfmt-yuv-luma.rst +15 −0
Documentation/userspace-api/media/v4l/pixfmt-yuv-planar.rst +94 −0
Documentation/userspace-api/media/v4l/user-func.rst +1 −0
Documentation/userspace-api/media/v4l/vidioc-g-ext-ctrls.rst +10 −0
Documentation/userspace-api/media/v4l/vidioc-g-fbuf.rst +16 −36
Documentation/userspace-api/media/v4l/vidioc-subdev-enum-frame-size.rst +29 −15
Documentation/userspace-api/media/v4l/vidioc-subdev-enum-mbus-code.rst +27 −12
Documentation/userspace-api/media/v4l/vidioc-subdev-g-client-cap.rst +83 −0

Documentation/userspace-api/media/v4l/dev-overlay.rst+8−2

@@ -67,6 +67,7 @@ ioctls must be supported by all video overlay devices. · 66 unmodified lines
67	67	Setup
68	68	=====
69	69
	70	+Note: support for this has been removed.
70	71	Before overlay can commence applications must program the driver with
71	72	frame buffer parameters, namely the address and size of the frame buffer
72	73	and the image format, for example RGB 5:6:5. The
@@ -92,11 +93,13 @@ A driver may support any (or none) of five clipping/blending methods: · 19 unmodified lines
92	93	1. Chroma-keying displays the overlaid image only where pixels in the
93	94	primary graphics surface assume a certain color.
94	95
95		−2. A bitmap can be specified where each bit corresponds to a pixel in
	96	+2. Note: support for this has been removed.
	97	+ A bitmap can be specified where each bit corresponds to a pixel in
96	98	the overlaid image. When the bit is set, the corresponding video
97	99	pixel is displayed, otherwise a pixel of the graphics surface.
98	100
99		−3. A list of clipping rectangles can be specified. In these regions no
	101	+3. Note: support for this has been removed.
	102	+ A list of clipping rectangles can be specified. In these regions no
100	103	video is displayed, so the graphics surface can be seen here.
101	104
102	105	4. The framebuffer has an alpha channel that can be used to clip or
@@ -185,6 +188,7 @@ struct v4l2_window · 82 unmodified lines
185	188	be 0xRRGGBB on a little endian, 0xBBGGRR on a big endian host.
186	189
187	190	``struct v4l2_clip * clips``
	191	+ Note: support for this has been removed.
188	192	When chroma-keying has not been negotiated and
189	193	:ref:`VIDIOC_G_FBUF <VIDIOC_G_FBUF>` indicated this capability,
190	194	applications can set this field to point to an array of clipping
@@ -201,6 +205,7 @@ struct v4l2_window · 10 unmodified lines
201	205	are undefined.
202	206
203	207	``__u32 clipcount``
	208	+ Note: support for this has been removed.
204	209	When the application set the ``clips`` field, this field must
205	210	contain the number of clipping rectangles in the list. When clip
206	211	lists are not supported the driver ignores this field, its contents
@@ -208,6 +213,7 @@ struct v4l2_window · 1 unmodified line
208	213	supported but no clipping is desired this field must be set to zero.
209	214
210	215	``void * bitmap``
	216	+ Note: support for this has been removed.
211	217	When chroma-keying has not been negotiated and
212	218	:ref:`VIDIOC_G_FBUF <VIDIOC_G_FBUF>` indicated this capability,
213	219	applications can set this field to point to a clipping bit mask.

Documentation/userspace-api/media/v4l/pixfmt-compressed.rst+25−0

@@ -88,6 +88,11 @@ Compressed Formats · 87 unmodified lines
88	88	- ``V4L2_PIX_FMT_H263``
89	89	- 'H263'
90	90	- H263 video elementary stream.
	91	+ * .. _V4L2-PIX-FMT-SPK:
	92	+
	93	+ - ``V4L2_PIX_FMT_SPK``
	94	+ - 'SPK0'
	95	+ - Sorenson Spark is an implementation of H.263 for use in Flash Video and Adobe Flash files
91	96	* .. _V4L2-PIX-FMT-MPEG1:
92	97
93	98	- ``V4L2_PIX_FMT_MPEG1``
@@ -232,6 +237,26 @@ Compressed Formats · 138 unmodified lines
232	237	Metadata associated with the frame to decode is required to be passed
233	238	through the ``V4L2_CID_STATELESS_FWHT_PARAMS`` control.
234	239	See the :ref:`associated Codec Control ID <codec-stateless-fwht>`.
	240	+ * .. _V4L2-PIX-FMT-RV30:
	241	+
	242	+ - ``V4L2_PIX_FMT_RV30``
	243	+ - 'RV30'
	244	+ - RealVideo, or also spelled as Real Video, is a suite of
	245	+ proprietary video compression formats developed by
	246	+ RealNetworks - the specific format changes with the version.
	247	+ RealVideo codecs are identified by four-character codes.
	248	+ RV30 corresponds to RealVideo 8, suspected to be based
	249	+ largely on an early draft of H.264
	250	+ * .. _V4L2-PIX-FMT-RV40:
	251	+
	252	+ - ``V4L2_PIX_FMT_RV40``
	253	+ - 'RV40'
	254	+ - RV40 represents RealVideo 9 and RealVideo 10.
	255	+ RealVideo 9, suspected to be based on H.264.
	256	+ RealVideo 10, aka RV9 EHQ, This refers to an improved encoder
	257	+ for the RV9 format that is fully backwards compatible with
	258	+ RV9 players - the format and decoder did not change, only
	259	+ the encoder did. As a result, it uses the same FourCC.
235	260
236	261	.. raw:: latex
237	262

Documentation/userspace-api/media/v4l/pixfmt-packed-yuv.rst+28−0

@@ -257,6 +257,34 @@ the second byte and Y'\ :sub:`7-0` in the third byte. · 256 unmodified lines
257	257	- The padding bits contain undefined values that must be ignored by all
258	258	applications and drivers.
259	259
	260	+The next table lists the packed YUV 4:4:4 formats with 12 bits per component.
	261	+Expand the bits per component to 16 bits, data in the high bits, zeros in the low bits,
	262	+arranged in little endian order, storing 1 pixel in 6 bytes.
	263	+
	264	+.. flat-table:: Packed YUV 4:4:4 Image Formats (12bpc)
	265	+ :header-rows: 1
	266	+ :stub-columns: 0
	267	+
	268	+ * - Identifier
	269	+ - Code
	270	+ - Byte 1-0
	271	+ - Byte 3-2
	272	+ - Byte 5-4
	273	+ - Byte 7-6
	274	+ - Byte 9-8
	275	+ - Byte 11-10
	276	+
	277	+ * .. _V4L2-PIX-FMT-YUV48-12:
	278	+
	279	+ - ``V4L2_PIX_FMT_YUV48_12``
	280	+ - 'Y312'
	281	+
	282	+ - Y'\ :sub:`0`
	283	+ - Cb\ :sub:`0`
	284	+ - Cr\ :sub:`0`
	285	+ - Y'\ :sub:`1`
	286	+ - Cb\ :sub:`1`
	287	+ - Cr\ :sub:`1`
260	288
261	289	4:2:2 Subsampling
262	290	=================

Documentation/userspace-api/media/v4l/pixfmt-rgb.rst+42−0

@@ -953,6 +953,48 @@ number of bits for each component. · 952 unmodified lines
953	953
954	954	\endgroup
955	955
	956	+12 Bits Per Component
	957	+==============================
	958	+
	959	+These formats store an RGB triplet in six or eight bytes, with 12 bits per component.
	960	+Expand the bits per component to 16 bits, data in the high bits, zeros in the low bits,
	961	+arranged in little endian order.
	962	+
	963	+.. raw:: latex
	964	+
	965	+ \small
	966	+
	967	+.. flat-table:: RGB Formats With 12 Bits Per Component
	968	+ :header-rows: 1
	969	+
	970	+ * - Identifier
	971	+ - Code
	972	+ - Byte 1-0
	973	+ - Byte 3-2
	974	+ - Byte 5-4
	975	+ - Byte 7-6
	976	+ * .. _V4L2-PIX-FMT-BGR48-12:
	977	+
	978	+ - ``V4L2_PIX_FMT_BGR48_12``
	979	+ - 'B312'
	980	+
	981	+ - B\ :sub:`15-4`
	982	+ - G\ :sub:`15-4`
	983	+ - R\ :sub:`15-4`
	984	+ -
	985	+ * .. _V4L2-PIX-FMT-ABGR64-12:
	986	+
	987	+ - ``V4L2_PIX_FMT_ABGR64_12``
	988	+ - 'B412'
	989	+
	990	+ - B\ :sub:`15-4`
	991	+ - G\ :sub:`15-4`
	992	+ - R\ :sub:`15-4`
	993	+ - A\ :sub:`15-4`
	994	+
	995	+.. raw:: latex
	996	+
	997	+ \normalsize
956	998
957	999	Deprecated RGB Formats
958	1000	======================

Documentation/userspace-api/media/v4l/pixfmt-yuv-luma.rst+15−0

@@ -103,6 +103,17 @@ are often referred to as greyscale formats. · 102 unmodified lines
103	103	- ...
104	104	- ...
105	105
	106	+ * .. _V4L2-PIX-FMT-Y012:
	107	+
	108	+ - ``V4L2_PIX_FMT_Y012``
	109	+ - 'Y012'
	110	+
	111	+ - Y'\ :sub:`0`\ [3:0] `0000`
	112	+ - Y'\ :sub:`0`\ [11:4]
	113	+ - ...
	114	+ - ...
	115	+ - ...
	116	+
106	117	* .. _V4L2-PIX-FMT-Y14:
107	118
108	119	- ``V4L2_PIX_FMT_Y14``
@@ -146,3 +157,7 @@ are often referred to as greyscale formats. · 37 unmodified lines
146	157	than 16 bits. For example, 10 bits per pixel uses values in the range 0 to
147	158	1023. For the IPU3_Y10 format 25 pixels are packed into 32 bytes, which
148	159	leaves the 6 most significant bits of the last byte padded with 0.
	160	+
	161	+ For Y012 and Y12 formats, Y012 places its data in the 12 high bits, with
	162	+ padding zeros in the 4 low bits, in contrast to the Y12 format, which has
	163	+ its padding located in the most significant bits of the 16 bit word.

Documentation/userspace-api/media/v4l/pixfmt-yuv-planar.rst+94−0

@@ -123,6 +123,20 @@ All components are stored with the same number of bits per component. · 122 unmodified lines
123	123	- Cb, Cr
124	124	- Yes
125	125	- 4x4 tiles
	126	+ * - V4L2_PIX_FMT_P012
	127	+ - 'P012'
	128	+ - 12
	129	+ - 4:2:0
	130	+ - Cb, Cr
	131	+ - Yes
	132	+ - Linear
	133	+ * - V4L2_PIX_FMT_P012M
	134	+ - 'PM12'
	135	+ - 12
	136	+ - 4:2:0
	137	+ - Cb, Cr
	138	+ - No
	139	+ - Linear
126	140	* - V4L2_PIX_FMT_NV16
127	141	- 'NV16'
128	142	- 8
@@ -586,6 +600,86 @@ Data in the 10 high bits, zeros in the 6 low bits, arranged in little endian ord · 457 unmodified lines
586	600	- Cb\ :sub:`11`
587	601	- Cr\ :sub:`11`
588	602
	603	+.. _V4L2-PIX-FMT-P012:
	604	+.. _V4L2-PIX-FMT-P012M:
	605	+
	606	+P012 and P012M
	607	+--------------
	608	+
	609	+P012 is like NV12 with 12 bits per component, expanded to 16 bits.
	610	+Data in the 12 high bits, zeros in the 4 low bits, arranged in little endian order.
	611	+
	612	+.. flat-table:: Sample 4x4 P012 Image
	613	+ :header-rows: 0
	614	+ :stub-columns: 0
	615	+
	616	+ * - start + 0:
	617	+ - Y'\ :sub:`00`
	618	+ - Y'\ :sub:`01`
	619	+ - Y'\ :sub:`02`
	620	+ - Y'\ :sub:`03`
	621	+ * - start + 8:
	622	+ - Y'\ :sub:`10`
	623	+ - Y'\ :sub:`11`
	624	+ - Y'\ :sub:`12`
	625	+ - Y'\ :sub:`13`
	626	+ * - start + 16:
	627	+ - Y'\ :sub:`20`
	628	+ - Y'\ :sub:`21`
	629	+ - Y'\ :sub:`22`
	630	+ - Y'\ :sub:`23`
	631	+ * - start + 24:
	632	+ - Y'\ :sub:`30`
	633	+ - Y'\ :sub:`31`
	634	+ - Y'\ :sub:`32`
	635	+ - Y'\ :sub:`33`
	636	+ * - start + 32:
	637	+ - Cb\ :sub:`00`
	638	+ - Cr\ :sub:`00`
	639	+ - Cb\ :sub:`01`
	640	+ - Cr\ :sub:`01`
	641	+ * - start + 40:
	642	+ - Cb\ :sub:`10`
	643	+ - Cr\ :sub:`10`
	644	+ - Cb\ :sub:`11`
	645	+ - Cr\ :sub:`11`
	646	+
	647	+.. flat-table:: Sample 4x4 P012M Image
	648	+ :header-rows: 0
	649	+ :stub-columns: 0
	650	+
	651	+ * - start0 + 0:
	652	+ - Y'\ :sub:`00`
	653	+ - Y'\ :sub:`01`
	654	+ - Y'\ :sub:`02`
	655	+ - Y'\ :sub:`03`
	656	+ * - start0 + 8:
	657	+ - Y'\ :sub:`10`
	658	+ - Y'\ :sub:`11`
	659	+ - Y'\ :sub:`12`
	660	+ - Y'\ :sub:`13`
	661	+ * - start0 + 16:
	662	+ - Y'\ :sub:`20`
	663	+ - Y'\ :sub:`21`
	664	+ - Y'\ :sub:`22`
	665	+ - Y'\ :sub:`23`
	666	+ * - start0 + 24:
	667	+ - Y'\ :sub:`30`
	668	+ - Y'\ :sub:`31`
	669	+ - Y'\ :sub:`32`
	670	+ - Y'\ :sub:`33`
	671	+ * -
	672	+ * - start1 + 0:
	673	+ - Cb\ :sub:`00`
	674	+ - Cr\ :sub:`00`
	675	+ - Cb\ :sub:`01`
	676	+ - Cr\ :sub:`01`
	677	+ * - start1 + 8:
	678	+ - Cb\ :sub:`10`
	679	+ - Cr\ :sub:`10`
	680	+ - Cb\ :sub:`11`
	681	+ - Cr\ :sub:`11`
	682	+
589	683
590	684	Fully Planar YUV Formats
591	685	========================

Documentation/userspace-api/media/v4l/user-func.rst+1−0

@@ -72,6 +72,7 @@ Function Reference · 71 unmodified lines
72	72	vidioc-subdev-g-frame-interval
73	73	vidioc-subdev-g-routing
74	74	vidioc-subdev-g-selection
	75	+ vidioc-subdev-g-client-cap
75	76	vidioc-subdev-querycap
76	77	vidioc-subscribe-event
77	78	func-mmap

Documentation/userspace-api/media/v4l/vidioc-g-ext-ctrls.rst+10−0

@@ -185,6 +185,16 @@ still cause this situation. · 184 unmodified lines
185	185	- ``p_u32``
186	186	- A pointer to a matrix control of unsigned 32-bit values. Valid if
187	187	this control is of type ``V4L2_CTRL_TYPE_U32``.
	188	+ * - __u32 *
	189	+ - ``p_s32``
	190	+ - A pointer to a matrix control of signed 32-bit values. Valid if
	191	+ this control is of type ``V4L2_CTRL_TYPE_INTEGER`` and
	192	+ ``V4L2_CTRL_FLAG_HAS_PAYLOAD`` is set.
	193	+ * - __u32 *
	194	+ - ``p_s64``
	195	+ - A pointer to a matrix control of signed 64-bit values. Valid if
	196	+ this control is of type ``V4L2_CTRL_TYPE_INTEGER64`` and
	197	+ ``V4L2_CTRL_FLAG_HAS_PAYLOAD`` is set.
188	198	* - struct :c:type:`v4l2_area` *
189	199	- ``p_area``
190	200	- A pointer to a struct :c:type:`v4l2_area`. Valid if this control is

Documentation/userspace-api/media/v4l/vidioc-g-fbuf.rst+16−36

@@ -49,6 +49,9 @@ of a graphics card. A non-destructive overlay blends video images into a · 48 unmodified lines
49	49	VGA signal or graphics into a video signal. Video Output Overlays are
50	50	always non-destructive.
51	51
	52	+Destructive overlay support has been removed: with modern GPUs and CPUs
	53	+this is no longer needed, and it was always a very dangerous feature.
	54	+
52	55	To get the current parameters applications call the :ref:`VIDIOC_G_FBUF <VIDIOC_G_FBUF>`
53	56	ioctl with a pointer to a struct :c:type:`v4l2_framebuffer`
54	57	structure. The driver fills all fields of the structure or returns an
@@ -63,18 +66,12 @@ this structure, the driver prepares for the overlay and returns the · 8 unmodified lines
63	66	framebuffer parameters as :ref:`VIDIOC_G_FBUF <VIDIOC_G_FBUF>` does, or it returns an error
64	67	code.
65	68
66		−To set the parameters for a non-destructive Video Overlay,
	69	+To set the parameters for a Video Capture Overlay
67	70	applications must initialize the ``flags`` field, the ``fmt``
68	71	substructure, and call :ref:`VIDIOC_S_FBUF <VIDIOC_G_FBUF>`. Again the driver prepares for
69	72	the overlay and returns the framebuffer parameters as :ref:`VIDIOC_G_FBUF <VIDIOC_G_FBUF>`
70	73	does, or it returns an error code.
71	74
72		−For a destructive Video Overlay applications must additionally provide
73		−a ``base`` address. Setting up a DMA to a random memory location can
74		−jeopardize the system security, its stability or even damage the
75		−hardware, therefore only the superuser can set the parameters for a
76		−destructive video overlay.
77		−
78	75	.. tabularcolumns:: \|p{3.5cm}\|p{3.5cm}\|p{3.5cm}\|p{6.6cm}\|
79	76
80	77	.. c:type:: v4l2_framebuffer
@@ -100,17 +97,14 @@ destructive video overlay. · 19 unmodified lines
100	97	- ``base``
101	98	-
102	99	- Physical base address of the framebuffer, that is the address of
103		− the pixel in the top left corner of the framebuffer. [#f1]_
104		− * -
105		− -
106		− -
107		− - This field is irrelevant to non-destructive Video Overlays. For
108		− destructive Video Overlays applications must provide a base
109		− address. The driver may accept only base addresses which are a
110		− multiple of two, four or eight bytes. For Video Output Overlays
111		− the driver must return a valid base address, so applications can
	100	+ the pixel in the top left corner of the framebuffer.
	101	+ For :ref:`VIDIOC_S_FBUF <VIDIOC_G_FBUF>` this field is no longer supported
	102	+ and the kernel will always set this to NULL.
	103	+ For Video Output Overlays
	104	+ the driver will return a valid base address, so applications can
112	105	find the corresponding Linux framebuffer device (see
113		− :ref:`osd`).
	106	+ :ref:`osd`). For Video Capture Overlays this field will always be
	107	+ NULL.
114	108	* - struct
115	109	- ``fmt``
116	110	-
@@ -136,8 +130,7 @@ destructive video overlay. · 19 unmodified lines
136	130	* -
137	131	-
138	132	-
139		− - For destructive Video Overlays applications must initialize this
140		− field. For Video Output Overlays the driver must return a valid
	133	+ - For Video Output Overlays the driver must return a valid
141	134	format.
142	135	* -
143	136	-
@@ -165,13 +158,6 @@ destructive video overlay. · 21 unmodified lines
165	158
166	159	This field is irrelevant to non-destructive Video Overlays.
167	160
168		− For destructive Video Overlays both applications and drivers can
169		− set this field to request padding bytes at the end of each line.
170		− Drivers however may ignore the requested value, returning
171		− ``width`` times bytes-per-pixel or a larger value required by the
172		− hardware. That implies applications can just set this field to
173		− zero to get a reasonable default.
174		−
175	161	For Video Output Overlays the driver must return a valid value.
176	162
177	163	Video hardware may access padding bytes, therefore they must
@@ -190,9 +176,8 @@ destructive video overlay. · 12 unmodified lines
190	176	* -
191	177	- __u32
192	178	- ``sizeimage``
193		− - This field is irrelevant to non-destructive Video Overlays. For
194		− destructive Video Overlays applications must initialize this
195		− field. For Video Output Overlays the driver must return a valid
	179	+ - This field is irrelevant to non-destructive Video Overlays.
	180	+ For Video Output Overlays the driver must return a valid
196	181	format.
197	182
198	183	Together with ``base`` it defines the framebuffer memory
@@ -232,9 +217,11 @@ destructive video overlay. · 33 unmodified lines
232	217	* - ``V4L2_FBUF_CAP_LIST_CLIPPING``
233	218	- 0x0004
234	219	- The device supports clipping using a list of clip rectangles.
	220	+ Note that this is no longer supported.
235	221	* - ``V4L2_FBUF_CAP_BITMAP_CLIPPING``
236	222	- 0x0008
237	223	- The device supports clipping using a bit mask.
	224	+ Note that this is no longer supported.
238	225	* - ``V4L2_FBUF_CAP_LOCAL_ALPHA``
239	226	- 0x0010
240	227	- The device supports clipping/blending using the alpha channel of
@@ -342,10 +329,3 @@ EPERM · 101 unmodified lines
342	329
343	330	EINVAL
344	331	The :ref:`VIDIOC_S_FBUF <VIDIOC_G_FBUF>` parameters are unsuitable.
345		−
346		−.. [#f1]
347		− A physical base address may not suit all platforms. GK notes in
348		− theory we should pass something like PCI device + memory region +
349		− offset instead. If you encounter problems please discuss on the
350		− linux-media mailing list:
351		− `https://linuxtv.org/lists.php <https://linuxtv.org/lists.php>`__.

Documentation/userspace-api/media/v4l/vidioc-subdev-enum-frame-size.rst+29−15

@@ -31,18 +31,30 @@ Arguments · 30 unmodified lines
31	31	Description
32	32	===========
33	33
34		−This ioctl allows applications to enumerate all frame sizes supported by
35		−a sub-device on the given pad for the given media bus format. Supported
36		−formats can be retrieved with the
	34	+This ioctl allows applications to access the enumeration of frame sizes
	35	+supported by a sub-device on the specified pad
	36	+for the specified media bus format.
	37	+Supported formats can be retrieved with the
37	38	:ref:`VIDIOC_SUBDEV_ENUM_MBUS_CODE`
38	39	ioctl.
39	40
40		−To enumerate frame sizes applications initialize the ``pad``, ``which``
41		−, ``code`` and ``index`` fields of the struct
42		−:c:type:`v4l2_subdev_mbus_code_enum` and
43		−call the :ref:`VIDIOC_SUBDEV_ENUM_FRAME_SIZE` ioctl with a pointer to the
44		−structure. Drivers fill the minimum and maximum frame sizes or return an
45		−EINVAL error code if one of the input parameters is invalid.
	41	+The enumerations are defined by the driver, and indexed using the ``index`` field
	42	+of the struct :c:type:`v4l2_subdev_frame_size_enum`.
	43	+Each pair of ``pad`` and ``code`` correspond to a separate enumeration.
	44	+Each enumeration starts with the ``index`` of 0, and
	45	+the lowest invalid index marks the end of the enumeration.
	46	+
	47	+Therefore, to enumerate frame sizes allowed on the specified pad
	48	+and using the specified mbus format, initialize the
	49	+``pad``, ``which``, and ``code`` fields to desired values,
	50	+and set ``index`` to 0.
	51	+Then call the :ref:`VIDIOC_SUBDEV_ENUM_FRAME_SIZE` ioctl with a pointer to the
	52	+structure.
	53	+
	54	+A successful call will return with minimum and maximum frame sizes filled in.
	55	+Repeat with increasing ``index`` until ``EINVAL`` is received.
	56	+``EINVAL`` means that either no more entries are available in the enumeration,
	57	+or that an input parameter was invalid.
46	58
47	59	Sub-devices that only support discrete frame sizes (such as most
48	60	sensors) will return one or more frame sizes with identical minimum and
@@ -72,26 +84,28 @@ information about try formats. · 23 unmodified lines
72	84
73	85	* - __u32
74	86	- ``index``
75		− - Number of the format in the enumeration, set by the application.
	87	+ - Index of the frame size in the enumeration belonging to the given pad
	88	+ and format. Filled in by the application.
76	89	* - __u32
77	90	- ``pad``
78	91	- Pad number as reported by the media controller API.
	92	+ Filled in by the application.
79	93	* - __u32
80	94	- ``code``
81	95	- The media bus format code, as defined in
82		− :ref:`v4l2-mbus-format`.
	96	+ :ref:`v4l2-mbus-format`. Filled in by the application.
83	97	* - __u32
84	98	- ``min_width``
85		− - Minimum frame width, in pixels.
	99	+ - Minimum frame width, in pixels. Filled in by the driver.
86	100	* - __u32
87	101	- ``max_width``
88		− - Maximum frame width, in pixels.
	102	+ - Maximum frame width, in pixels. Filled in by the driver.
89	103	* - __u32
90	104	- ``min_height``
91		− - Minimum frame height, in pixels.
	105	+ - Minimum frame height, in pixels. Filled in by the driver.
92	106	* - __u32
93	107	- ``max_height``
94		− - Maximum frame height, in pixels.
	108	+ - Maximum frame height, in pixels. Filled in by the driver.
95	109	* - __u32
96	110	- ``which``
97	111	- Frame sizes to be enumerated, from enum

Documentation/userspace-api/media/v4l/vidioc-subdev-enum-mbus-code.rst+27−12

@@ -31,15 +31,28 @@ Arguments · 30 unmodified lines
31	31	Description
32	32	===========
33	33
34		−To enumerate media bus formats available at a given sub-device pad
35		−applications initialize the ``pad``, ``which`` and ``index`` fields of
36		−struct
37		−:c:type:`v4l2_subdev_mbus_code_enum` and
38		−call the :ref:`VIDIOC_SUBDEV_ENUM_MBUS_CODE` ioctl with a pointer to this
39		−structure. Drivers fill the rest of the structure or return an ``EINVAL``
40		−error code if either the ``pad`` or ``index`` are invalid. All media bus
41		−formats are enumerable by beginning at index zero and incrementing by
42		−one until ``EINVAL`` is returned.
	34	+This call is used by the application to access the enumeration
	35	+of media bus formats for the selected pad.
	36	+
	37	+The enumerations are defined by the driver, and indexed using the ``index`` field
	38	+of struct :c:type:`v4l2_subdev_mbus_code_enum`.
	39	+Each enumeration starts with the ``index`` of 0, and
	40	+the lowest invalid index marks the end of enumeration.
	41	+
	42	+Therefore, to enumerate media bus formats available at a given sub-device pad,
	43	+initialize the ``pad``, and ``which`` fields to desired values,
	44	+and set ``index`` to 0.
	45	+Then call the :ref:`VIDIOC_SUBDEV_ENUM_MBUS_CODE` ioctl
	46	+with a pointer to this structure.
	47	+
	48	+A successful call will return with the ``code`` field filled in
	49	+with a mbus code value.
	50	+Repeat with increasing ``index`` until ``EINVAL`` is received.
	51	+``EINVAL`` means that either ``pad`` is invalid,
	52	+or that there are no more codes available at this pad.
	53	+
	54	+The driver must not return the same value of ``code`` for different indices
	55	+at the same pad.
43	56
44	57	Available media bus formats may depend on the current 'try' formats at
45	58	other pads of the sub-device, as well as on the current active links.
@@ -57,14 +70,16 @@ information about the try formats. · 11 unmodified lines
57	70
58	71	* - __u32
59	72	- ``pad``
60		− - Pad number as reported by the media controller API.
	73	+ - Pad number as reported by the media controller API. Filled in by the
	74	+ application.
61	75	* - __u32
62	76	- ``index``
63		− - Number of the format in the enumeration, set by the application.
	77	+ - Index of the mbus code in the enumeration belonging to the given pad.
	78	+ Filled in by the application.
64	79	* - __u32
65	80	- ``code``
66	81	- The media bus format code, as defined in
67		− :ref:`v4l2-mbus-format`.
	82	+ :ref:`v4l2-mbus-format`. Filled in by the driver.
68	83	* - __u32
69	84	- ``which``
70	85	- Media bus format codes to be enumerated, from enum

Documentation/userspace-api/media/v4l/vidioc-subdev-g-client-cap.rst+83−0

@@ -0,0 +1,83 @@
	1	+.. SPDX-License-Identifier: GFDL-1.1-no-invariants-or-later
	2	+.. c:namespace:: V4L
	3	+
	4	+.. _VIDIOC_SUBDEV_G_CLIENT_CAP:
	5	+
	6	+************************************************************
	7	+ioctl VIDIOC_SUBDEV_G_CLIENT_CAP, VIDIOC_SUBDEV_S_CLIENT_CAP
	8	+************************************************************
	9	+
	10	+Name
	11	+====
	12	+
	13	+VIDIOC_SUBDEV_G_CLIENT_CAP - VIDIOC_SUBDEV_S_CLIENT_CAP - Get or set client
	14	+capabilities.
	15	+
	16	+Synopsis
	17	+========
	18	+
	19	+.. c:macro:: VIDIOC_SUBDEV_G_CLIENT_CAP
	20	+
	21	+``int ioctl(int fd, VIDIOC_SUBDEV_G_CLIENT_CAP, struct v4l2_subdev_client_capability *argp)``
	22	+
	23	+.. c:macro:: VIDIOC_SUBDEV_S_CLIENT_CAP
	24	+
	25	+``int ioctl(int fd, VIDIOC_SUBDEV_S_CLIENT_CAP, struct v4l2_subdev_client_capability *argp)``
	26	+
	27	+Arguments
	28	+=========
	29	+
	30	+``fd``
	31	+ File descriptor returned by :ref:`open() <func-open>`.
	32	+
	33	+``argp``
	34	+ Pointer to struct :c:type:`v4l2_subdev_client_capability`.
	35	+
	36	+Description
	37	+===========
	38	+
	39	+These ioctls are used to get and set the client (the application using the
	40	+subdevice ioctls) capabilities. The client capabilities are stored in the file
	41	+handle of the opened subdev device node, and the client must set the
	42	+capabilities for each opened subdev separately.
	43	+
	44	+By default no client capabilities are set when a subdev device node is opened.
	45	+
	46	+The purpose of the client capabilities are to inform the kernel of the behavior
	47	+of the client, mainly related to maintaining compatibility with different
	48	+kernel and userspace versions.
	49	+
	50	+The ``VIDIOC_SUBDEV_G_CLIENT_CAP`` ioctl returns the current client capabilities
	51	+associated with the file handle ``fd``.
	52	+
	53	+The ``VIDIOC_SUBDEV_S_CLIENT_CAP`` ioctl sets client capabilities for the file
	54	+handle ``fd``. The new capabilities fully replace the current capabilities, the
	55	+ioctl can therefore also be used to remove capabilities that have previously
	56	+been set.
	57	+
	58	+``VIDIOC_SUBDEV_S_CLIENT_CAP`` modifies the struct
	59	+:c:type:`v4l2_subdev_client_capability` to reflect the capabilities that have
	60	+been accepted. A common case for the kernel not accepting a capability is that
	61	+the kernel is older than the headers the userspace uses, and thus the capability
	62	+is unknown to the kernel.
	63	+
	64	+.. flat-table:: Client Capabilities
	65	+ :header-rows: 1
	66	+
	67	+ * - Capability
	68	+ - Description
	69	+ * - ``V4L2_SUBDEV_CLIENT_CAP_STREAMS``
	70	+ - The client is aware of streams. Setting this flag enables the use
	71	+ of 'stream' fields (referring to the stream number) with various
	72	+ ioctls. If this is not set (which is the default), the 'stream' fields
	73	+ will be forced to 0 by the kernel.
	74	+
	75	+Return Value
	76	+============
	77	+
	78	+On success 0 is returned, on error -1 and the ``errno`` variable is set
	79	+appropriately. The generic error codes are described at the
	80	+:ref:`Generic Error Codes <gen-errors>` chapter.
	81	+
	82	+ENOIOCTLCMD
	83	+ The kernel does not support this ioctl.