Laurent Pinchart [Sat, 18 Jun 2022 18:54:19 +0000 (19:54 +0100)]
media: Replace dependency on VIDEO_V4L2_SUBDEV_API with select
The VIDEO_V4L2_SUBDEV_API Kconfig symbol is mostly selected
automatically, with a a handful of drivers still using it as a
dependency. Fix them to use selection, and drop the symbol title text to
not expose it for manual selection.
Sakari Ailus [Wed, 13 Jul 2022 15:18:08 +0000 (16:18 +0100)]
media: sunxi: Depend on GENERIC_PHY_MIPI_DPHY
The sun6i-mipi-csi2 and sun6i-a83t-mipi-csi2 drivers depend on the generic
MIPI D-PHY support. Select it. This fixes a linking problem when either of
these drivers is enabled and GENERIC_PHY_MIPI_DPHY is disabled.
Laurent Pinchart [Thu, 17 Mar 2022 12:37:12 +0000 (12:37 +0000)]
media: v4l2: Make colorspace validity checks more future-proof
The helper functions that test validity of colorspace-related fields
use the last value of the corresponding enums. This isn't very
future-proof, as there's a high chance someone adding a new value may
forget to update the helpers. Add new "LAST" entries to the enumerations
to improve this, and keep them private to the kernel.
Signed-off-by: Laurent Pinchart <laurent.pinchart@ideasonboard.com> Acked-by: Sakari Ailus <sakari.ailus@linux.intel.com> Reviewed-by: Hans Verkuil <hverkuil-cisco@xs4all.nl> Signed-off-by: Mauro Carvalho Chehab <mchehab@kernel.org>
media: mediatek: vcodec: Make encoder capability fields fit requirements
This partially reverts commit fd9f8050e355d7fd1e126cd207b06c96cde7f783,
and changes things so that the capability string fields of the encoder
conform to their requirements.
The driver name field should contain the actual driver name, not some
otherwise unused string macro from the driver. To make this clear,
copy the name from the driver's name field.
The card name for the video encoder previously held a static platform
name that was fixed to match MT8173. This obviously doesn't make sense
for newer chips. Since commit fd9f8050e355 ("media: mediatek: vcodec:
Change encoder v4l2 capability value"), this field was changed to hold
the driver's name, or "mtk-vcodec-dec". This doesn't make much sense
either, since this still doesn't reflect what chip this is.
Instead, fill in the card name with "MTxxxx video encoder" with the
proper chip number.
Since commit f2d8b6917f3b ("media: v4l: ioctl: Set bus_info in
v4l_querycap()"), the V4L2 core provides a default value for the
bus_info field for platform and PCI devices. This value will match
the default value for media devices added by commit cef699749f37
("media: mc: Set bus_info in media_device_init()"). These defaults
are stable and device-specific.
Drop the custom capability bus_info from the mtk-vcodec encoder
driver, and use the defaults.
As this patch removes the last usage of MTK_VCODEC_DRV_NAME, remove
the macro as well.
media: mediatek: vcodec: Make decoder capability fields fit requirements
This partially reverts commit a8a7a278c56ad3b4ddd4db9a960e0537d032b0b3,
and changes things so that the capability string fields of the decoder
conform to their requirements.
This recent change caused ChromeOS's decoder to no longer function. This
is due to ChromeOS using the driver name field to match the video device
with its accompanying media device. After the change, they no longer
matched.
The driver name field should contain the actual driver name, not some
otherwise unused string macro from the driver. To make this clear,
copy the name from the driver's name field.
The card name for the video decoder previously held a static platform
name that was fixed to match MT8173. This obviously doesn't make sense
for newer chips. Since commit a8a7a278c56a ("media: mediatek: vcodec:
Change decoder v4l2 capability value"), this field was changed to hold
the driver's name, or "mtk-vcodec-dec". This doesn't make much sense
either, since this still doesn't reflect what chip this is.
Instead, fill in the card name with "MTxxxx video decoder" with the
proper chip number.
Since commit f2d8b6917f3b ("media: v4l: ioctl: Set bus_info in
v4l_querycap()"), the V4L2 core provides a default value for the
bus_info field for platform and PCI devices. This value will match
the default value for media devices added by commit cef699749f37
("media: mc: Set bus_info in media_device_init()"). These defaults
are stable and device-specific.
Drop the custom capability bus_info from the mtk-vcodec decoder
driver, and use the defaults. This also fixes the long standing
issue where the media device used for the stateless decoder didn't
have its bus_info set, and would never match its accompanying video
device.
Marko Mäkelä [Fri, 8 Jul 2022 08:44:59 +0000 (09:44 +0100)]
media: lirc: ensure lirc device receives repeats
Pressing a button on a remote control unit will typically lead to
messages being sent several times per second until the button is released.
Some remote control units indicate long key presses by sending
special "repeat" messages, for which the protocol driver calls
rc_repeat(). Other units repeat the same message over and over,
which will be handled by calling rc_keydown().
The function rc_keydown() never set the LIRC "repeat" flag to distinguish
repeated messages that were sent due to a long keypress, and messages
sent due to repeated short keypresses. While a user-space program may
implement special logic to distinguish long keypresses, it is much simpler
to be able to rely on the flag.
Signed-off-by: Marko Mäkelä <marko.makela@iki.fi> Co-developed-by: Sean Young <sean@mess.org> Signed-off-by: Sean Young <sean@mess.org> Signed-off-by: Mauro Carvalho Chehab <mchehab@kernel.org>
Daniel Oakley [Thu, 16 Jun 2022 10:07:47 +0000 (11:07 +0100)]
media: vimc: use data link entities enum to index the ent_config array
Future additions to the ent_config[] could break the association between
the index of the struct vimc_ent_config entries in the ent_config[] array,
and the index defined by the enum proposed in the previous patch. Using
designated initializers solves this by linking the 2 together clearly in
code and prevents the array not reflecting the enum. There is no
functional change intended.
Daniel Oakley [Thu, 16 Jun 2022 10:07:46 +0000 (11:07 +0100)]
media: vimc: enumerate data link entities for clarity
The data_links array was hard to read and understand. By implementing
enumerated vimc data link entities, clarity has been improved when
defining data_links. This therefore should help new programmers to
understand the codebase better. There is no functional change intended.
Daniel Oakley [Thu, 16 Jun 2022 10:07:45 +0000 (11:07 +0100)]
media: vimc: expand the names of vimc entity types
When introducing the lens controller, it became apparent that the vimc
entity type names were hard to understand, e.g. vimc_len_type refers to the
lens. The names of the vimc entity types have been expanded to make the
code easier to understand. There is no functional change intended.
Yunke Cao [Tue, 28 Jun 2022 00:53:52 +0000 (01:53 +0100)]
media: vimc: add ancillary lens
Add lens to vimc driver and link them with sensors using ancillary links.
Provides an example of ancillary link usage.The lens supports
FOCUS_ABSOLUTE control.
Test example: With default vimc topology
> media-ctl -p
Media controller API version 5.18.0
...
- entity 28: Lens A (0 pad, 0 link)
type V4L2 subdev subtype Lens flags 0
device node name /dev/v4l-subdev6
- entity 29: Lens B (0 pad, 0 link)
type V4L2 subdev subtype Lens flags 0
device node name /dev/v4l-subdev7
> v4l2-ctl -d /dev/v4l-subdev7 -C focus_absolute
focus_absolute: 0
Sakari Ailus [Tue, 5 Jul 2022 17:40:54 +0000 (18:40 +0100)]
media: v4l: async: Also match secondary fwnode endpoints
For camera sensor devices the firmware information of which comes from
non-DT (or some ACPI variants), the kernel makes the information visible
to the drivers in a form similar to DT. This takes place through device's
secondary fwnodes, in which case also the secondary fwnode needs to be
heterogenously (endpoint vs. device) matched.
Fixes: 1f391df44607 ("media: v4l2-async: Use endpoints in __v4l2_async_nf_add_fwnode_remote()") Signed-off-by: Sakari Ailus <sakari.ailus@linux.intel.com> Reviewed-by: Laurent Pinchart <laurent.pinchart+renesas@ideasonboard.com> Signed-off-by: Mauro Carvalho Chehab <mchehab@kernel.org>
Ming Qian [Fri, 8 Jul 2022 08:56:50 +0000 (09:56 +0100)]
media: amphion: only insert the first sequence startcode for vc1l format
For format V4L2_PIX_FMT_VC1_ANNEX_L,
the amphion vpu requires driver to help insert some custom startcode
before sequence and frame.
but only the first sequence startcode is needed,
the extra startcode will cause decoding error.
So after seek, we don't need to insert the sequence startcode.
In other words, for V4L2_PIX_FMT_VC1_ANNEX_L,
the vpu doesn't support dynamic resolution change.
Ezequiel Garcia [Wed, 29 Jun 2022 19:56:23 +0000 (20:56 +0100)]
media: hantro: Fix RK3399 H.264 format advertising
Commit 1f82f2df523cb ("media: hantro: Enable H.264 on Rockchip VDPU2")
enabled H.264 on some SoCs with VDPU2 cores. This had the side-effect
of exposing H.264 coded format as supported on RK3399.
Fix this and clarify how the codec is explicitly disabled on RK3399 on
this driver.
Fixes: 1f82f2df523cb ("media: hantro: Enable H.264 on Rockchip VDPU2") Signed-off-by: Ezequiel Garcia <ezequiel@vanguardiasur.com.ar> Tested-by: Nicolas Dufresne <nicolas.dufresne@collabora.com> Reviewed-by: Nicolas Dufresne <nicolas.dufresne@collabora.com> Signed-off-by: Hans Verkuil <hverkuil-cisco@xs4all.nl> Signed-off-by: Mauro Carvalho Chehab <mchehab@kernel.org>
Akira Yokosawa [Wed, 29 Jun 2022 13:01:17 +0000 (14:01 +0100)]
media: docs: Remove extraneous \endgroup from P010 table
The \endgroup command at the bottom of a table added in commit 5374d8fb75f3 ("media: Add P010 video format") doesn't have a
corresponding \begingroup command preceding it.
This imbalance causes an build error in "make pdfdocs".
Fix the issue by removing it.
Signed-off-by: Akira Yokosawa <akiyks@gmail.com> Fixes: 5374d8fb75f3 ("media: Add P010 video format") Cc: Benjamin Gaignard <benjamin.gaignard@collabora.com> Cc: Nicolas Dufresne <nicolas.dufresne@collabora.com> Signed-off-by: Hans Verkuil <hverkuil-cisco@xs4all.nl> Signed-off-by: Mauro Carvalho Chehab <mchehab@kernel.org>
Krzysztof Kozlowski [Wed, 29 Jun 2022 12:08:03 +0000 (13:08 +0100)]
media: dt-bindings: media: samsung,s5pv210-jpeg: convert to dtschema
Convert the Samsung SoC JPEG codec bindings to DT schema.
The original bindings were quite old and incomplete, so change during
conversion:
1. Add typical (already used) properties like iommus and power domains.
2. Document samsung,exynos4212-jpeg compatible (already used in DTS and
driver).
3. List clocks per each variant.
Signed-off-by: Krzysztof Kozlowski <krzysztof.kozlowski@linaro.org> Reviewed-by: Rob Herring <robh@kernel.org> Signed-off-by: Hans Verkuil <hverkuil-cisco@xs4all.nl> Signed-off-by: Mauro Carvalho Chehab <mchehab@kernel.org>
Ming Qian [Tue, 28 Jun 2022 05:21:12 +0000 (06:21 +0100)]
media: amphion: defer setting last_buffer_dequeued until resolution changes are processed
Don't set last_buffer_dequeued during dynamic resolution change,
otherwise it may be cleared in handling resolution change,
as streamoff may be called in dynamic resolution change.
Normally, this does not happen.
But we encounter a special testcase,
User issue V4L2_DEC_CMD_STOP after enqueue one buffer
that only contains codec config header, but not any frame data.
So VPU report the parsed resolution, then report the eos event.
So driver should notify user to handle resolution change first,
after it's handled, set the last_buffer_dequeued.
then the user can exit decoding normally.
media: mediatek: vcodec: Initialize decoder parameters for each instance
The decoder parameters are stored in each instance's context data. This
needs to be initialized per-instance, but a previous fix incorrectly
changed it to only be initialized for the first opened instance. This
resulted in subsequent instances not correctly signaling the requirement
for the Requests API.
Fix this by calling the initializing function outside of the
v4l2_fh_is_singular() conditional block.
Right now the decoder maintains two separate lists for supported pixel
formats and frame sizes. Getting the supported frame sizes for the
current set format is a bit convoluted, requiring a search through
the separate frame size list. The frame sizes are used to clamp and
align requested resolutions.
Instead, the frame size structure could be embedded inside the pixel
format structure. Getting one also gets the other. And since the
the driver already keeps pointers to the current set format, getting
the frame sizes becomes straightforward.
Do just that. Move v4l2_frmsize_stepwise inside mtk_video_fmt, and get
rid of mtk_codec_framesizes.
Signed-off-by: Chen-Yu Tsai <wenst@chromium.org> Tested-by: AngeloGioacchino Del Regno <angelogioacchino.delregno@collabora.com> Signed-off-by: Hans Verkuil <hverkuil-cisco@xs4all.nl> Signed-off-by: Mauro Carvalho Chehab <mchehab@kernel.org>
media: mediatek: vcodec: decoder: Fix resolution clamping in TRY_FMT
In commit b018be06f3c7 ("media: mediatek: vcodec: Read max resolution
from dec_capability"), TRY_FMT clamps the resolution to the maximum
that was previously set either by default 1080p or the limit set by a
previous S_FMT call. This does not make sense when doing TRY_FMT for
the output side, which may have different capabilities.
Instead, for the output side, find the maximum resolution based on the
pixel format requested. For the capture side, find the maximum
resolution based on the currently set output format.
The maximum resolution is found from the list of per-format frame
sizes, so the patch "media: mediatek: vcodec: dec: Fix 4K frame size
enumeration" is needed.
Fixes: b018be06f3c7 ("media: mediatek: vcodec: Read max resolution from dec_capability") Signed-off-by: Chen-Yu Tsai <wenst@chromium.org> Tested-by: AngeloGioacchino Del Regno <angelogioacchino.delregno@collabora.com> Signed-off-by: Hans Verkuil <hverkuil-cisco@xs4all.nl> Signed-off-by: Mauro Carvalho Chehab <mchehab@kernel.org>
This partially reverts commit b018be06f3c7 ("media: mediatek: vcodec:
Read max resolution from dec_capability"). In this commit, the maximum
resolution ended up being a function of both the firmware capability and
the current set format.
However, frame size enumeration for output (coded) formats should not
depend on the format set, but should return supported resolutions for
the format requested by userspace.
Fix this so that the driver returns the supported resolutions correctly,
even if the instance only has default settings, or if the output format
is currently set to VP8F, which does not support 4K.
This adds an copy of special casing for !VP8 and 4K support. The other
existing copy will be removed when .max_{width,height} are removed from
|struct mtk_vcodec_ctx| in a subsequent patch.
Fixes: b018be06f3c7 ("media: mediatek: vcodec: Read max resolution from dec_capability") Signed-off-by: Chen-Yu Tsai <wenst@chromium.org> Tested-by: AngeloGioacchino Del Regno <angelogioacchino.delregno@collabora.com> Signed-off-by: Hans Verkuil <hverkuil-cisco@xs4all.nl> Signed-off-by: Mauro Carvalho Chehab <mchehab@kernel.org>
Laurent Pinchart [Tue, 10 May 2022 09:47:34 +0000 (10:47 +0100)]
media: staging: media: imx: imx7-media-csi: Fix list of supported formats
The list of supported formats comes from helpers shared with the i.MX6
IPUv3 and is incorrect in multiple regards:
- 10-, 12- and 14-bit Bayer formats are stored in memory as SBGGR10,
SBGGR12 and SBGGR14 respectively (plus components permutations), not
SBGGR16. Same thing for greyscale formats.
- 16-bit RAW formats are not supported by the hardware.
Fix the supported formats table.
Signed-off-by: Laurent Pinchart <laurent.pinchart@ideasonboard.com> Acked-by: Rui Miguel Silva <rmfrfs@gmail.com> Tested-by: Alexander Stein <alexander.stein@ew.tq-group.com> Signed-off-by: Mauro Carvalho Chehab <mchehab@kernel.org>
Define a macro for the default media bus code and use it through the
driver to replace a hardcoded value and a dynamic query from the
pixel_formats table.
Signed-off-by: Laurent Pinchart <laurent.pinchart@ideasonboard.com> Acked-by: Rui Miguel Silva <rmfrfs@gmail.com> Tested-by: Alexander Stein <alexander.stein@ew.tq-group.com> Signed-off-by: Mauro Carvalho Chehab <mchehab@kernel.org>
Laurent Pinchart [Tue, 10 May 2022 02:06:45 +0000 (03:06 +0100)]
media: staging: media: imx: imx7-media-csi: Drop unsupported YUV and RGB formats
A large number of formats defined in the pixel_formats array are not
supported, as shown by the switch-case check in
imx7_csi_pad_link_validate(). Drop them.
Signed-off-by: Laurent Pinchart <laurent.pinchart@ideasonboard.com> Acked-by: Rui Miguel Silva <rmfrfs@gmail.com> Tested-by: Alexander Stein <alexander.stein@ew.tq-group.com> Signed-off-by: Mauro Carvalho Chehab <mchehab@kernel.org>
Laurent Pinchart [Mon, 9 May 2022 22:32:33 +0000 (23:32 +0100)]
media: staging: media: imx: imx7-media-csi: Import format helpers
To prepare for code refactoring, copy the format helpers used by this
driver from imx-media-utils.c. Rename structures and functions to avoid
name clashes, no functional change intended.
Signed-off-by: Laurent Pinchart <laurent.pinchart@ideasonboard.com> Acked-by: Rui Miguel Silva <rmfrfs@gmail.com> Tested-by: Alexander Stein <alexander.stein@ew.tq-group.com> Signed-off-by: Mauro Carvalho Chehab <mchehab@kernel.org>
Laurent Pinchart [Mon, 9 May 2022 22:32:33 +0000 (23:32 +0100)]
media: staging: media: imx: imx7-media-csi: Decouple from shared macros
Decouple from the IMX7_CSI_VIDEO_DEF_PIX_WIDTH,
IMX7_CSI_VIDEO_DEF_PIX_HEIGHT and IMX_MEDIA_EOF_TIMEOUT macros defined
in shared helpers by duplicating them in the imx7-media-csi driver, with
a rename to avoid name clashes. No functional change intended.
Signed-off-by: Laurent Pinchart <laurent.pinchart@ideasonboard.com> Acked-by: Rui Miguel Silva <rmfrfs@gmail.com> Tested-by: Alexander Stein <alexander.stein@ew.tq-group.com> Signed-off-by: Mauro Carvalho Chehab <mchehab@kernel.org>
Laurent Pinchart [Mon, 9 May 2022 22:32:33 +0000 (23:32 +0100)]
media: staging: media: imx: imx7-media-csi: Decouple from imx_media_dma_buf
Decouple from the imx_media_dma_buf structure defined in shared helpers
by duplicating it in the imx7-media-csi driver, along with the two small
alloc and free functions. No functional change intended.
Signed-off-by: Laurent Pinchart <laurent.pinchart@ideasonboard.com> Acked-by: Rui Miguel Silva <rmfrfs@gmail.com> Tested-by: Alexander Stein <alexander.stein@ew.tq-group.com> Signed-off-by: Mauro Carvalho Chehab <mchehab@kernel.org>
Laurent Pinchart [Mon, 9 May 2022 22:32:33 +0000 (23:32 +0100)]
media: staging: media: imx: imx7-media-csi: Fold imx_media_video_dev into imx7_csi
Now that the driver doesn't use any helper that relies on the
imx_media_video_dev structure anymore, merge its fields directly in the
imx7_csi structure. Update all the users accordingly, and drop the list
field that isn't used by the driver.
Signed-off-by: Laurent Pinchart <laurent.pinchart@ideasonboard.com> Acked-by: Rui Miguel Silva <rmfrfs@gmail.com> Tested-by: Alexander Stein <alexander.stein@ew.tq-group.com> Signed-off-by: Mauro Carvalho Chehab <mchehab@kernel.org>
Laurent Pinchart [Mon, 9 May 2022 22:32:33 +0000 (23:32 +0100)]
media: staging: media: imx: imx7-media-csi: Fold imx_media_dev into imx7_csi
Now that the driver doesn't use any helper that relies on the
imx_media_dev structure anymore, merge the three fields of the
imx_media_dev structure directly in the imx7_csi structure. Update all
the users accordingly.
Signed-off-by: Laurent Pinchart <laurent.pinchart@ideasonboard.com> Acked-by: Rui Miguel Silva <rmfrfs@gmail.com> Tested-by: Alexander Stein <alexander.stein@ew.tq-group.com> Signed-off-by: Mauro Carvalho Chehab <mchehab@kernel.org>
The imx7_csi_media_pipeline_set_stream() function has separate branches
for the start and stop paths, surrounded by mutex lock/unlock. That is
very little shared code, inline the locking and corresponding branch in
each of the two callers.
Signed-off-by: Laurent Pinchart <laurent.pinchart@ideasonboard.com> Acked-by: Rui Miguel Silva <rmfrfs@gmail.com> Tested-by: Alexander Stein <alexander.stein@ew.tq-group.com> Signed-off-by: Mauro Carvalho Chehab <mchehab@kernel.org>
Simplify the imx7_csi_media_pipeline_set_stream() function by passing it
the imx7_csi pointer, which avoids going from subdev to entity and back
to subdev.
Signed-off-by: Laurent Pinchart <laurent.pinchart@ideasonboard.com> Acked-by: Rui Miguel Silva <rmfrfs@gmail.com> Tested-by: Alexander Stein <alexander.stein@ew.tq-group.com> Signed-off-by: Mauro Carvalho Chehab <mchehab@kernel.org>
To prepare for code refactoring, copy the
imx_media_pipeline_set_stream() helper used by this driver from
imx-media-utils.c. Rename the function to avoid name clashes, no
functional change intended.
Signed-off-by: Laurent Pinchart <laurent.pinchart@ideasonboard.com> Acked-by: Rui Miguel Silva <rmfrfs@gmail.com> Tested-by: Alexander Stein <alexander.stein@ew.tq-group.com> Signed-off-by: Mauro Carvalho Chehab <mchehab@kernel.org>
Laurent Pinchart [Tue, 10 May 2022 00:25:26 +0000 (01:25 +0100)]
media: staging: media: imx: imx7-media-csi: Drop imx_media_add_video_device call
There's no need to call imx_media_add_video_device() anymore, as the
video devices list it manages is only used by the control inheritance
mechanism in the helpers, which this driver doesn't use.
Signed-off-by: Laurent Pinchart <laurent.pinchart@ideasonboard.com> Acked-by: Rui Miguel Silva <rmfrfs@gmail.com> Tested-by: Alexander Stein <alexander.stein@ew.tq-group.com> Signed-off-by: Mauro Carvalho Chehab <mchehab@kernel.org>
To prepare for code refactoring, copy the imx_media_dev_init() helper
used by this driver from imx-media-dev-common.c. Rename the function to
avoid name clashes, and leave the v4l2_device notify handler out as it
only serves to implement control inheritance, which this driver doesn't
use. No functional change intended.
Signed-off-by: Laurent Pinchart <laurent.pinchart@ideasonboard.com> Acked-by: Rui Miguel Silva <rmfrfs@gmail.com> Tested-by: Alexander Stein <alexander.stein@ew.tq-group.com> Signed-off-by: Mauro Carvalho Chehab <mchehab@kernel.org>
Laurent Pinchart [Mon, 9 May 2022 22:32:33 +0000 (23:32 +0100)]
media: staging: media: imx: imx7-media-csi: Import video device helpers
To prepare for code refactoring, copy the video device helper code used
by this driver verbatim from imx-media-capture. Rename some functions to
avoid name clashes, and leave the legacy ioctls out as they're not used
by the imx7-media-csi driver. No functional change included.
Signed-off-by: Laurent Pinchart <laurent.pinchart@ideasonboard.com> Acked-by: Rui Miguel Silva <rmfrfs@gmail.com> Tested-by: Alexander Stein <alexander.stein@ew.tq-group.com> Signed-off-by: Mauro Carvalho Chehab <mchehab@kernel.org>
imx_media_of_add_csi() from imx-media-of.c validates that the fwnode for
the CSI bridge is enabled, and adds the fwnode to the async notifier of
the imxmd. The former is redundant, as if the CSI bridge driver probes,
clearly it is enabled in the DT. The latter is not needed as the imxmd
notifier isn't used anymore. The call is thus not needed and can be
dropped. This removes the dependency of imx7-media-csi.c on
imx-media-of.c.
Signed-off-by: Paul Elder <paul.elder@ideasonboard.com> Acked-by: Rui Miguel Silva <rmfrfs@gmail.com> Tested-by: Alexander Stein <alexander.stein@ew.tq-group.com> Signed-off-by: Laurent Pinchart <laurent.pinchart@ideasonboard.com> Signed-off-by: Mauro Carvalho Chehab <mchehab@kernel.org>
The imx_media_dev framework maintains a per-pad list of connected video
devices, created once all subdevs have been bound. This is used for two
purposes, updating V4L2 control inheritance when links change, and
relaying subdev events to video nodes. None of these are used by the
imx7-media-csi driver as it implements the MC-centric approach. Drop
them.
Signed-off-by: Laurent Pinchart <laurent.pinchart@ideasonboard.com> Acked-by: Rui Miguel Silva <rmfrfs@gmail.com> Tested-by: Alexander Stein <alexander.stein@ew.tq-group.com> Signed-off-by: Mauro Carvalho Chehab <mchehab@kernel.org>
Paul Elder [Thu, 5 May 2022 12:48:33 +0000 (13:48 +0100)]
media: staging: media: imx: imx7-media-csi: Move misc init out of probe()
There is a chunk of miscellaneous initializations related to the CSI
subdev and media pads directly in the probe function. Move them into the
imx7_csi_media_init() function to clean up the probe function.
Signed-off-by: Paul Elder <paul.elder@ideasonboard.com> Acked-by: Rui Miguel Silva <rmfrfs@gmail.com> Tested-by: Alexander Stein <alexander.stein@ew.tq-group.com> Signed-off-by: Laurent Pinchart <laurent.pinchart@ideasonboard.com> Signed-off-by: Mauro Carvalho Chehab <mchehab@kernel.org>
Laurent Pinchart [Sun, 6 Feb 2022 05:13:53 +0000 (05:13 +0000)]
media: staging: media: imx: imx7-media-csi: Drop the imx_media notifier
The V4L2 async notifier embedded in the imx_media_dev structure is
triggered when the CSI bridge subdev is registered. We don't need an
async notifier for this, as we can call the .complete() handler directly
from the CSI bridge subdev .registered() handler.
Signed-off-by: Laurent Pinchart <laurent.pinchart@ideasonboard.com> Acked-by: Rui Miguel Silva <rmfrfs@gmail.com> Tested-by: Alexander Stein <alexander.stein@ew.tq-group.com> Signed-off-by: Mauro Carvalho Chehab <mchehab@kernel.org>
Laurent Pinchart [Sun, 6 Feb 2022 05:13:53 +0000 (05:13 +0000)]
media: staging: media: imx: imx7-media-csi: Drop duplicate link creation
The imx_media_create_csi2_links() creates a media controller link
between the CSI-2 receiver and the next entity in the pipeline, which
can be either a video mux (handled by the video-mux driver) or the CSI
bridge itself. This isn't needed, as the link is already created either
by the video-mux driver or by the imx7-media-csi driver itself (in
imx7_csi_notify_bound()).
Drop imx_media_create_csi2_links(), which allows dropping the CSI bridge
subdev grp_id.
Signed-off-by: Laurent Pinchart <laurent.pinchart@ideasonboard.com> Acked-by: Rui Miguel Silva <rmfrfs@gmail.com> Tested-by: Alexander Stein <alexander.stein@ew.tq-group.com> Signed-off-by: Mauro Carvalho Chehab <mchehab@kernel.org>
To prepare for code refactoring, copy the V4L2 async notifier helper
code used by this driver verbatim from imx-media-dev-common.c. Rename
some functions to avoid name clashes. No functional change included.
Signed-off-by: Laurent Pinchart <laurent.pinchart@ideasonboard.com> Acked-by: Rui Miguel Silva <rmfrfs@gmail.com> Tested-by: Alexander Stein <alexander.stein@ew.tq-group.com> Signed-off-by: Mauro Carvalho Chehab <mchehab@kernel.org>
Laurent Pinchart [Sat, 5 Feb 2022 04:16:36 +0000 (04:16 +0000)]
media: staging: media: imx: imx7-media-csi: Split imx_media_dev from probe()
Prepare for the decoupling of the imx7-media-csi driver from the
IPUv3-based drivers by moving the imx_media_dev handling from probe()
function to separate functions.
Signed-off-by: Laurent Pinchart <laurent.pinchart@ideasonboard.com> Acked-by: Rui Miguel Silva <rmfrfs@gmail.com> Tested-by: Alexander Stein <alexander.stein@ew.tq-group.com> Signed-off-by: Mauro Carvalho Chehab <mchehab@kernel.org>
Laurent Pinchart [Wed, 25 May 2022 09:16:04 +0000 (10:16 +0100)]
media: renesas: rcar_drif: Drop of_match_ptr()
The device_driver structure's of_match_table field exists
unconditionally, so there's no need for of_match_ptr(). This fixes a
compiler warning when test-compiling on non-OF platforms:
Colin Ian King [Tue, 28 Jun 2022 15:24:51 +0000 (16:24 +0100)]
media: atomisp: clean up for-loop, remove redundant assignment to variable i
There is a for-loop that initializes variable i but does not use it; the
assignment is redundant and can be removed. The proceeding assignment to
pointer fw can also be moved into the for-loop to clean up the code.
Ard Biesheuvel [Mon, 20 Jun 2022 10:08:19 +0000 (11:08 +0100)]
media: atomisp_gmin_platform: stop abusing efivar API
As the code comment already suggests, using the efivar API in this way
is not how it is intended, and so let's switch to the right one, which
is simply to call efi.get_variable() directly after checking whether or
not the GetVariable() runtime service is supported.
Hans de Goede [Sun, 12 Jun 2022 16:05:56 +0000 (17:05 +0100)]
media: atomisp: fix -Wdangling-pointer warning
ia_css_rmgr_acq_vbuf() uses a local on stack
"struct ia_css_rmgr_vbuf_handle v" variable.
When this path using this is hit, either the rmgr_pop_handle() call
will make *handle point to another vbuf-handle, or because
v.count == 0, ia_css_rmgr_refcount_retain_vbuf() will alloc a new
vbuf-handle and make *handle point to it.
So on leaving the function *handle will never point to the on stack
vbuf-handle, but gcc does not know this and emits the following:
drivers/staging/media/atomisp/pci/runtime/rmgr/src/rmgr_vbuf.c: In function ‘ia_css_rmgr_acq_vbuf’:
drivers/staging/media/atomisp/pci/runtime/rmgr/src/rmgr_vbuf.c:276:33: warning: storing the address of local variable ‘h’ in ‘*handle’ [-Wdangling-pointer=]
276 | *handle = &h;
| ~~~~~~~~^~~~
drivers/staging/media/atomisp/pci/runtime/rmgr/src/rmgr_vbuf.c:257:40: note: ‘h’ declared here
257 | struct ia_css_rmgr_vbuf_handle h;
| ^
drivers/staging/media/atomisp/pci/runtime/rmgr/src/rmgr_vbuf.c:257:40: note: ‘handle’ declared here
Rework the code using a new_handle helper to suppress this
false-postive compiler warning.
Hans de Goede [Sun, 12 Jun 2022 16:05:55 +0000 (17:05 +0100)]
media: atomisp: fix uninitialized stack mem usage in ia_css_rmgr_acq_vbuf()
When ia_css_rmgr_acq_vbuf() enters the code path where it uses the local
"struct ia_css_rmgr_vbuf_handle v" on the stack it relies on v.count==0
so that ia_css_rmgr_refcount_retain_vbuf allocates a new handle.
Explicitly set v.count to 0 rather then it being whatever was on the stack.
Hans de Goede [Sun, 12 Jun 2022 16:05:54 +0000 (17:05 +0100)]
media: atomisp: revert "don't pass a pointer to a local variable"
The gcc is warning about returning a pointer to a local variable
is a false positive.
The type of handle is "struct ia_css_rmgr_vbuf_handle **" and
"h.vptr" is left to NULL, so the "if ((*handle)->vptr == 0x0)"
check always succeeds when the "*handle = &h;" statement which
gcc warns about executes. Leading to this statement being executed:
rmgr_pop_handle(pool, handle);
If that succeeds, then *handle has been set to point to one of
the pre-allocated array of handles, so it no longer points to h.
If that fails the following statement will be executed:
/* Note that handle will change to an internally maintained one */
ia_css_rmgr_refcount_retain_vbuf(handle);
Which allocated a new handle from the array of pre-allocated handles
and then makes *handle point to this. So the address of h is actually
never returned.
The fix for the false-postive compiler warning actually breaks the code,
the new:
**handle = h;
is part of a "if (pool->copy_on_write) { ... }" which means that the
handle where *handle points to should be treated read-only, IOW
**handle must never be set, instead *handle must be set to point to
a new handle (with a copy of the contents of the old handle).
The old code correctly did this and the new fixed code gets this wrong.
Note there is another patch in this series, which fixes the warning
in another way.
media: [PATCH] pci: atomisp_cmd: fix three missing checks on list iterator
The three bugs are here:
__func__, s3a_buf->s3a_data->exp_id);
__func__, md_buf->metadata->exp_id);
__func__, dis_buf->dis_data->exp_id);
The list iterator 's3a_buf/md_buf/dis_buf' will point to a bogus
position containing HEAD if the list is empty or no element is found.
This case must be checked before any use of the iterator, otherwise
it will lead to a invalid memory access.
To fix this bug, add an check. Use a new variable '*_iter' as the
list iterator, while use the old variable '*_buf' as a dedicated
pointer to point to the found element.
Fabio M. De Francesco [Wed, 13 Apr 2022 22:55:31 +0000 (23:55 +0100)]
media: staging: media: atomisp: Use kmap_local_page() in hmm_store()
The use of kmap() is being deprecated in favor of kmap_local_page()
where it is feasible. The same is true for kmap_atomic().
In file pci/hmm/hmm.c, function hmm_store() test if we are in atomic
context and, if so, it calls kmap_atomic(), if not, it calls kmap().
First of all, in_atomic() shouldn't be used in drivers. This macro
cannot always detect atomic context; in particular, it cannot know
about held spinlocks in non-preemptible kernels.
Notwithstanding what it is said above, this code doesn't need to care
whether or not it is executing in atomic context. It can simply use
kmap_local_page() / kunmap_local() that can instead do the mapping /
unmapping regardless of the context.
With kmap_local_page(), the mapping is per thread, CPU local and not
globally visible. Therefore, hmm_store()() is a function where the use
of kmap_local_page() in place of both kmap() and kmap_atomic() is
correctly suited.
Convert the calls of kmap() / kunmap() and kmap_atomic() /
kunmap_atomic() to kmap_local_page() / kunmap_local() and drop the
unnecessary tests which test if the code is in atomic context.
Fabio M. De Francesco [Wed, 13 Apr 2022 21:22:10 +0000 (22:22 +0100)]
media: staging: media: atomisp: Use kmap_local_page() in hmm_set()
The use of kmap() is being deprecated in favor of kmap_local_page()
where it is feasible. In file pci/hmm/hmm.c, function hmm_set() calls
kmap() / kunmap() where kmap_local_page() can instead do the mapping.
With kmap_local_page(), the mapping is per thread, CPU local and not
globally visible. Therefore, hmm_set()() is a function where the use
of kmap_local_page() in place of kmap() is correctly suited.
Convert the calls of kmap() / kunmap() to kmap_local_page() /
kunmap_local().
projects / users / jedix / linux-maple.git / log