Daniel Scally [Thu, 7 Jan 2021 13:28:29 +0000 (14:28 +0100)]
media: software_node: unregister software_nodes in reverse order
To maintain consistency with software_node_unregister_nodes(), reverse
the order in which the software_node_unregister_node_group() function
unregisters nodes.
Reported-by: kernel test robot <lkp@intel.com> Reported-by: Dan Carpenter <dan.carpenter@oracle.com> Reviewed-by: Laurent Pinchart <laurent.pinchart@ideasonboard.com> Suggested-by: Andy Shevchenko <andriy.shevchenko@linux.intel.com> Signed-off-by: Daniel Scally <djrscally@gmail.com> Reviewed-by: Heikki Krogerus <heikki.krogerus@linux.intel.com> Acked-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org> Acked-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com> Signed-off-by: Sakari Ailus <sakari.ailus@linux.intel.com> Signed-off-by: Mauro Carvalho Chehab <mchehab+huawei@kernel.org>
Daniel Scally [Thu, 7 Jan 2021 13:28:28 +0000 (14:28 +0100)]
media: software_node: Enforce parent before child ordering of nodes arrays
Registering software_nodes with the .parent member set to point to a
currently unregistered software_node has the potential for problems,
so enforce parent -> child ordering in arrays passed in to
software_node_register_nodes().
Software nodes that are children of another software node should be
unregistered before their parent. To allow easy unregistering of an array
of software_nodes ordered parent to child, reverse the order in which
software_node_unregister_nodes() unregisters software_nodes.
Suggested-by: Andy Shevchenko <andriy.shevchenko@linux.intel.com> Reviewed-by: Laurent Pinchart <laurent.pinchart@ideasonboard.com> Signed-off-by: Daniel Scally <djrscally@gmail.com> Reviewed-by: Heikki Krogerus <heikki.krogerus@linux.intel.com> Acked-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org> Acked-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com> Signed-off-by: Sakari Ailus <sakari.ailus@linux.intel.com> Signed-off-by: Mauro Carvalho Chehab <mchehab+huawei@kernel.org>
Daniel Scally [Thu, 7 Jan 2021 13:28:27 +0000 (14:28 +0100)]
media: device property: Call fwnode_graph_get_endpoint_by_id() for fwnode->secondary
This function is used to find fwnode endpoints against a device. In
some instances those endpoints are software nodes which are children of
fwnode->secondary. Add support to fwnode_graph_get_endpoint_by_id() to
find those endpoints by recursively calling itself passing the ptr to
fwnode->secondary in the event no endpoint is found for the primary.
Reviewed-by: Andy Shevchenko <andriy.shevchenko@linux.intel.com> Reviewed-by: Laurent Pinchart <laurent.pinchart@ideasonboard.com> Signed-off-by: Daniel Scally <djrscally@gmail.com> Acked-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org> Acked-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com> Signed-off-by: Sakari Ailus <sakari.ailus@linux.intel.com> Signed-off-by: Mauro Carvalho Chehab <mchehab+huawei@kernel.org>
Daniel Scally [Thu, 7 Jan 2021 13:28:26 +0000 (14:28 +0100)]
media: device property: Return true in fwnode_device_is_available for NULL ops
Some types of fwnode_handle do not implement the device_is_available()
check, such as those created by software_nodes. There isn't really a
meaningful way to check for the availability of a device that doesn't
actually exist, so if the check isn't implemented just assume that the
"device" is present.
Suggested-by: Laurent Pinchart <laurent.pinchart@ideasonboard.com> Reviewed-by: Laurent Pinchart <laurent.pinchart@ideasonboard.com> Reviewed-by: Andy Shevchenko <andriy.shevchenko@linux.intel.com> Signed-off-by: Daniel Scally <djrscally@gmail.com> Reviewed-by: Heikki Krogerus <heikki.krogerus@linux.intel.com> Acked-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org> Acked-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com> Signed-off-by: Sakari Ailus <sakari.ailus@linux.intel.com> Signed-off-by: Mauro Carvalho Chehab <mchehab+huawei@kernel.org>
Daniel Scally [Thu, 7 Jan 2021 13:28:24 +0000 (14:28 +0100)]
media: software_node: Fix refcounts in software_node_get_next_child()
The software_node_get_next_child() function currently does not hold
references to the child software_node that it finds or put the ref that
is held against the old child - fix that.
Fixes: 59abd83672f7 ("drivers: base: Introducing software nodes to the firmware node framework") Reviewed-by: Andy Shevchenko <andriy.shevchenko@linux.intel.com> Reviewed-by: Laurent Pinchart <laurent.pinchart@ideasonboard.com> Signed-off-by: Daniel Scally <djrscally@gmail.com> Reviewed-by: Heikki Krogerus <heikki.krogerus@linux.intel.com> Acked-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org> Acked-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com> Signed-off-by: Sakari Ailus <sakari.ailus@linux.intel.com> Signed-off-by: Mauro Carvalho Chehab <mchehab+huawei@kernel.org>
A reference to the sub-device pad ops was not follwed by a whitespace,
resulting in a warning during documentation build. Fix it.
Reported-by: Stephen Rothwell <sfr@canb.auug.org.au> Fixes: 25c8d9a7689e ("media: Documentation: v4l: Document that link_validate op is valid for sink only") Signed-off-by: Sakari Ailus <sakari.ailus@linux.intel.com> Signed-off-by: Mauro Carvalho Chehab <mchehab+huawei@kernel.org>
Rob Herring [Mon, 4 Jan 2021 16:58:07 +0000 (17:58 +0100)]
media: dt-bindings: Convert video-interfaces.txt properties to schemas
Convert video-interfaces.txt to DT schema. As it contains a mixture of
device level and endpoint properties, split it up into 2 schemas.
Binding schemas will need to reference both the graph.yaml and
video-interfaces.yaml schemas. The exact schema depends on how many
ports and endpoints for the binding. A single port with a single
endpoint looks similar to this:
Tomi Valkeinen [Wed, 13 Jan 2021 09:00:27 +0000 (10:00 +0100)]
media: ti-vpe: cal: fix write to unallocated memory
The asd allocated with v4l2_async_notifier_add_fwnode_subdev() must be
of size cal_v4l2_async_subdev, otherwise access to
cal_v4l2_async_subdev->phy will go to unallocated memory.
Fixes: 8fcb7576ad19 ("media: ti-vpe: cal: Allow multiple contexts per subdev notifier") Signed-off-by: Tomi Valkeinen <tomi.valkeinen@ideasonboard.com> Signed-off-by: Hans Verkuil <hverkuil-cisco@xs4all.nl> Signed-off-by: Mauro Carvalho Chehab <mchehab+huawei@kernel.org>
Rui Miguel Silva [Thu, 7 Jan 2021 10:47:26 +0000 (11:47 +0100)]
media: imx7: csi: Fix pad link validation
We can not make the assumption that the bound subdev is always a CSI
mux, in i.MX6UL/i.MX6ULL that is not the case. So, just get the entity
selected by source directly upstream from the CSI.
Fixes: 86e02d07871c ("media: imx5/6/7: csi: Mark a bound video mux as a CSI mux") Reported-by: Fabio Estevam <festevam@gmail.com> Signed-off-by: Rui Miguel Silva <rmfrfs@gmail.com> Tested-by: Fabio Estevam <festevam@gmail.com> Tested-by: Sébastien Szymanski <sebastien.szymanski@armadeus.com> Signed-off-by: Hans Verkuil <hverkuil-cisco@xs4all.nl> Signed-off-by: Mauro Carvalho Chehab <mchehab+huawei@kernel.org>
Fabio Estevam [Thu, 7 Jan 2021 10:47:25 +0000 (11:47 +0100)]
media: imx7: csi: Fix regression for parallel cameras on i.MX6UL
Commit 86e02d07871c ("media: imx5/6/7: csi: Mark a bound video mux as
a CSI mux") made an incorrect assumption that for imx7-media-csi,
the bound subdev must always be a CSI mux. On i.MX6UL/i.MX6ULL there
is no CSI mux at all, so do not return an error when the entity is not a
video mux and assign the IMX_MEDIA_GRP_ID_CSI_MUX group id only when
appropriate.
This is the same approach as done in imx-media-csi.c and it fixes the
csi probe regression on i.MX6UL.
Tested on a imx6ull-evk board.
Fixes: 86e02d07871c ("media: imx5/6/7: csi: Mark a bound video mux as a CSI mux") Signed-off-by: Fabio Estevam <festevam@gmail.com> Signed-off-by: Rui Miguel Silva <rmfrfs@gmail.com> Signed-off-by: Hans Verkuil <hverkuil-cisco@xs4all.nl> Signed-off-by: Mauro Carvalho Chehab <mchehab+huawei@kernel.org>
Dikshita Agarwal [Thu, 24 Dec 2020 11:25:33 +0000 (12:25 +0100)]
media: v4l2-ctrl: Add frame-specific min/max qp controls for hevc
- Adds min/max qp controls for B frame for h264.
- Adds min/max qp controls for I/P/B frames for hevc similar to h264.
- Update valid range of min/max qp for hevc to accommodate 10 bit.
Sakari Ailus [Tue, 5 Jan 2021 17:12:28 +0000 (18:12 +0100)]
media: ccs: Make (non-)use of uninitialised variables more robust
GCC with W=2 level of kernel compiler warnings warns about the use of
uninitialised variables in a few locations. While these uninitialised
variables were not used in reality, this still produced compiler warnings.
Address this by assigning the variables to NULL and checking for NULL in
places it is not expected, returning -EIO in that case. This provides
at least some sanity checking at runtime as the compiler appears unable to
do that at compile time.
Reported-by: Hans Verkuil <hverkuil-cisco@xs4all.nl> Signed-off-by: Sakari Ailus <sakari.ailus@linux.intel.com> Signed-off-by: Hans Verkuil <hverkuil-cisco@xs4all.nl> Signed-off-by: Mauro Carvalho Chehab <mchehab+huawei@kernel.org>
Stanimir Varbanov [Wed, 5 Aug 2020 07:36:06 +0000 (09:36 +0200)]
media: venus: Create hfi platform and move vpp/vsp there
Introduce a new hfi platform to cover differences between hfi
versions. As a start move vpp/vsp freq data in that hfi
platform, more platform data will come later.
For resolutions below 720p the size of the compressed buffer must
be bigger. Correct this by checking the resolution when calculating
buffer size and multiply by eight.
The hfi_session_init can be called many times and it returns
EINVAL when the session was already initialized. This error code
(EINVAL) is confusing for the callers. Change hfi_session_init to
return EALREADY error code when the session has been already
initialized.
Vikash Garodia [Fri, 4 Dec 2020 10:01:38 +0000 (11:01 +0100)]
media: venus: request for interrupt from venus
For synchronous commands, update the message queue variable.
This would inform video firmware to raise interrupt on host
CPU whenever there is a response for such commands.
Stanimir Varbanov [Fri, 4 Dec 2020 10:01:37 +0000 (11:01 +0100)]
media: venus: Limit HFI sessions to the maximum supported
Currently we rely on firmware to return error when we reach the maximum
supported number of sessions. But this errors are happened at reqbuf
time which is a bit later. The more reasonable way looks like is to
return the error on driver open.
To achieve that modify hfi_session_create to return error when we reach
maximum count of sessions and thus refuse open.
Fritz Koenig [Tue, 1 Dec 2020 04:23:23 +0000 (05:23 +0100)]
media: venus: vdec: Handle DRC after drain
If the DRC is near the end of the stream the client
may send a V4L2_DEC_CMD_STOP before the DRC occurs.
V4L2_DEC_CMD_STOP puts the driver into the
VENUS_DEC_STATE_DRAIN state. DRC must be aware so
that after the DRC event the state can be restored
correctly.
Alexandre Courbot [Wed, 2 Dec 2020 05:34:24 +0000 (06:34 +0100)]
media: venus: preserve DRC state across seeks
DRC events can happen virtually at anytime, including when we are
starting a seek. Should this happen, we must make sure to return to the
DRC state, otherwise the firmware will expect buffers of the new
resolution whereas userspace will still work with the old one.
Returning to the DRC state upon resume for seeking makes sure that the
client will get the DRC event and will reallocate the buffers to fit the
firmware's expectations.
media: venus: helpers: Lock outside of buffer queue helper
After adding more logic in vdec buf_queue vb2 op it is not
practical to have two lock/unlock for one decoder buf_queue.
So move the instance lock in encoder and decoder vb2 buf_queue
operations.
media: venus: vdec: Make decoder return LAST flag for sufficient event
This makes the decoder to behaives equally for sufficient and
insufficient events. After this change the LAST buffer flag will be set
when the new resolution (in dynamic-resolution-change state) is smaller
then the old bitstream resolution.
media: venus: vdec: Fix non reliable setting of LAST flag
In real use of dynamic-resolution-change it is observed that the
LAST buffer flag (which marks the last decoded buffer with the
resolution before the resolution-change event) is not reliably set.
Fix this by set the LAST buffer flag on next queued capture buffer
after the resolution-change event.
Alexandre Courbot [Mon, 14 Dec 2020 12:57:03 +0000 (13:57 +0100)]
media: venus: use contig vb2 ops
This driver uses the SG vb2 ops, but effectively only ever accesses the
first entry of the SG table, indicating that it expects a flat layout.
Switch it to use the contiguous ops to make sure this expected invariant
is always enforced. Since the device is supposed to be behind an IOMMU
this should have little to none practical consequences beyond making the
driver not rely on a particular behavior of the SG implementation.
Dinghao Liu [Mon, 28 Dec 2020 13:02:05 +0000 (14:02 +0100)]
media: em28xx: Fix use-after-free in em28xx_alloc_urbs
When kzalloc() fails, em28xx_uninit_usb_xfer() will free
usb_bufs->buf and set it to NULL. Thus the later access
to usb_bufs->buf[i] will lead to null pointer dereference.
Also the kfree(usb_bufs->buf) after that is redundant.
Fixes: d571b592c6206 ("media: em28xx: don't use coherent buffer for DMA transfers") Signed-off-by: Dinghao Liu <dinghao.liu@zju.edu.cn> Signed-off-by: Hans Verkuil <hverkuil-cisco@xs4all.nl> Signed-off-by: Mauro Carvalho Chehab <mchehab+huawei@kernel.org>
Sebastian Fricke [Sat, 12 Dec 2020 18:53:06 +0000 (19:53 +0100)]
media: rkisp1: isp: Add the enum_frame_size ioctl
Implement the VIDIOC_SUBDEV_ENUM_FRAME_SIZE ioctl for the isp entity,
check if the mbus code is valid for the given pad.
This call is not available for the parameter or metadata pads of the
RkISP1.
Signed-off-by: Sebastian Fricke <sebastian.fricke.linux@gmail.com> Acked-by: Dafna Hirschfeld <dafna.hirschfeld@collabora.com> Acked-by: Helen Koike <helen.koike@collabora.com> Signed-off-by: Hans Verkuil <hverkuil-cisco@xs4all.nl> Signed-off-by: Mauro Carvalho Chehab <mchehab+huawei@kernel.org>
Geert Uytterhoeven [Fri, 11 Dec 2020 13:38:06 +0000 (14:38 +0100)]
media: sh_vou: Drop bogus __refdata annotation
Since commit 4c62e9764ab403d4 ("Drivers: media: remove __dev*
attributes.") in v3.8, the SuperH Video Output Unit driver no longer has
any code or data located in initmem, hence there is no need to annotate
the sh_vou structure with __refdata. Drop the annotation, to avoid
suppressing future section warnings.
Lukas Bulwahn [Wed, 9 Dec 2020 07:46:58 +0000 (08:46 +0100)]
media: MAINTAINERS: correct entry in Amlogic GE2D driver section
Commit aa821b2b9269 ("media: MAINTAINERS: Add myself as maintainer of the
Amlogic GE2D driver") introduced a new MAINTAINERS section, but the file
entry points to the wrong location.
Dan Carpenter [Wed, 9 Dec 2020 06:50:34 +0000 (07:50 +0100)]
media: camss: Fix signedness bug in video_enum_fmt()
This test has a problem because we want to know if "k" is -1 or a
positive value less than "f->index". But the "f->index" variable is a
u32 so if "k == -1" then -1 gets type promoted to UINT_MAX which is
larger than "f->index". I've added an explicit test to check for -1.
Fixes: a3d412d4b9f3 ("media: Revert "media: camss: Make use of V4L2_CAP_IO_MC"") Signed-off-by: Dan Carpenter <dan.carpenter@oracle.com> Reviewed-by: Robert Foss <robert.foss@linaro.org> Reviewed-by: Andrey Konovalov <andrey.konovalov@linaro.org> Signed-off-by: Hans Verkuil <hverkuil-cisco@xs4all.nl> Signed-off-by: Mauro Carvalho Chehab <mchehab+huawei@kernel.org>
Arnd Bergmann [Thu, 3 Dec 2020 23:07:30 +0000 (00:07 +0100)]
media: ti-vpe: cal: avoid FIELD_GET assertion
FIELD_GET() must only be used with a mask that is a compile-time
constant:
drivers/media/platform/ti-vpe/cal.h: In function 'cal_read_field':
include/linux/compiler_types.h:320:38: error: call to '__compiletime_assert_247' declared with attribute error: FIELD_GET: mask is not constant
include/linux/bitfield.h:46:3: note: in expansion of macro 'BUILD_BUG_ON_MSG'
46 | BUILD_BUG_ON_MSG(!__builtin_constant_p(_mask), \
| ^~~~~~~~~~~~~~~~
drivers/media/platform/ti-vpe/cal.h:220:9: note: in expansion of macro 'FIELD_GET'
220 | return FIELD_GET(mask, cal_read(cal, offset));
| ^~~~~~~~~
The problem here is that the function is not always inlined. Mark it
__always_inline to avoid the problem.
Remove MODULE_VERSION(), as it doesn't seem to serve any practical purpose.
For in-tree drivers, the kernel version matters. The code received lots of
changes, but module version remained constant, since the driver landed in
mainline. So, this version doesn't seem have any practical meaning anymore.
Ezequiel Garcia [Mon, 4 Jan 2021 20:34:41 +0000 (21:34 +0100)]
media: imx: Clean capture unregister
No locking is needed to call video_unregister_device(). Drop it.
Also, drop the superfluous video_is_registered() call, which is
done by video_unregister_device(), and re-order media_entity_cleanup()
and video_unregister_device() calls.
Signed-off-by: Ezequiel Garcia <ezequiel@collabora.com> Reviewed-by: Philipp Zabel <p.zabel@pengutronix.de> Signed-off-by: Hans Verkuil <hverkuil-cisco@xs4all.nl> Signed-off-by: Mauro Carvalho Chehab <mchehab+huawei@kernel.org>
Ezequiel Garcia [Mon, 4 Jan 2021 20:34:40 +0000 (21:34 +0100)]
media: imx: Fix csc/scaler unregister
The csc/scaler device private struct is released by
ipu_csc_scaler_video_device_release(), which can be called
by video_unregister_device() if there are no users
of the underlying struct video device.
Therefore, the mutex can't be held when calling
video_unregister_device() as its memory may be freed
by it, leading to a kernel oops.
Fortunately, the fix is quite simple as no locking
is needed when calling video_unregister_device(): v4l2-core
already has its own internal locking, and the structures
are also properly refcounted.
Ezequiel Garcia [Mon, 4 Jan 2021 20:34:39 +0000 (21:34 +0100)]
media: imx: Unregister csc/scaler only if registered
The csc/scaler device pointer (imxmd->m2m_vdev) is assigned
after the imx media device v4l2-async probe completes,
therefore we need to check if the device is non-NULL
before trying to unregister it.
This can be the case if the non-completed imx media device
is unbinded (or the driver is removed), leading to a kernel oops.
Martin Kepplinger [Wed, 2 Dec 2020 13:12:51 +0000 (14:12 +0100)]
media: staging: media: imx: Kconfig: support VIDEO_IMX7_CSI for imx8m
As described in NXPs' linux tree, the imx8m SoC includes the same
CSI bridge hardware that is part of imx7d. We should be able to
use the "fsl,imx7-csi" driver for imx8m directly.
Since ipuv3 is not relevant for imx8m, move that dependency to
VIDEO_IMX_CSI. That makes VIDEO_IMX7_CSI available to support imx8m
too.
Signed-off-by: Martin Kepplinger <martin.kepplinger@puri.sm> Signed-off-by: Hans Verkuil <hverkuil-cisco@xs4all.nl> Signed-off-by: Mauro Carvalho Chehab <mchehab+huawei@kernel.org>
Sakari Ailus [Fri, 11 Dec 2020 11:30:09 +0000 (12:30 +0100)]
media: ccs: Small definition cleanup
struct device is used in ccs-data.h but no definition is included. Add a
forward definition.
Also ccs-data.c includes linux/types.h but this is already included in
ccs-data.h which ccs-data.c includes. Do don't include linux/types.h in
ccs-data.c.
Signed-off-by: Sakari Ailus <sakari.ailus@linux.intel.com> Signed-off-by: Mauro Carvalho Chehab <mchehab+huawei@kernel.org>
Sakari Ailus [Fri, 9 Oct 2020 10:57:06 +0000 (12:57 +0200)]
media: ccs: Support and default to auto PHY control
CCS supports three variants of PHY timing control, auto, UI based and
manual. The driver previously assumed UI based control that requires
updating the link rate to the sensor. Now default to automatic control
instead, and only write the link rate to the sensor in UI mode.
If neither auto or UI control is supported, return an error in probe.
Signed-off-by: Sakari Ailus <sakari.ailus@linux.intel.com> Signed-off-by: Mauro Carvalho Chehab <mchehab+huawei@kernel.org>
Sakari Ailus [Thu, 8 Oct 2020 16:24:39 +0000 (18:24 +0200)]
media: ccs: Hardware requires a delay after starting the clock of lifting reset
A CCS compliant device requires a delay before the first I²C transaction
after pulling xshutdown up or starting the external clock. This is what
the driver does. However, if neither is actually there, there's no need
for the delay.
This also has the effect of removing an unnecessary delay on ACPI systems
where ACPI is responsible for the power-up sequence.
Signed-off-by: Sakari Ailus <sakari.ailus@linux.intel.com> Signed-off-by: Mauro Carvalho Chehab <mchehab+huawei@kernel.org>
Sakari Ailus [Thu, 8 Oct 2020 08:59:26 +0000 (10:59 +0200)]
media: ccs: Only do software reset if we have no hardware reset
The driver always used software reset after the sensor's power sequence
that includes a hardware reset if we have a reset GPIO. Do not use
software reset if we just brought the sensor up from hardware reset state.
Signed-off-by: Sakari Ailus <sakari.ailus@linux.intel.com> Signed-off-by: Mauro Carvalho Chehab <mchehab+huawei@kernel.org>
Sakari Ailus [Thu, 24 Sep 2020 21:34:27 +0000 (23:34 +0200)]
media: ccs: Add debug prints for MSR registers
Also print out MSR registers written to the sensor. This isn't entirely
optimal as the debug strings are produced even if they're not used but
that isn't really a grave issue --- the I²C bus is very slow anyway.
Signed-off-by: Sakari Ailus <sakari.ailus@linux.intel.com> Signed-off-by: Mauro Carvalho Chehab <mchehab+huawei@kernel.org>
Sakari Ailus [Wed, 23 Sep 2020 22:21:32 +0000 (00:21 +0200)]
media: v4l: uapi: ccs: Add controls for analogue gain constants
Add V4L2 controls for analogue gain constants required to control
analogue gain. The values are device specific and thus need to be obtained
from the driver.
Signed-off-by: Sakari Ailus <sakari.ailus@linux.intel.com> Signed-off-by: Mauro Carvalho Chehab <mchehab+huawei@kernel.org>
Paul Kocialkowski [Thu, 31 Dec 2020 14:27:01 +0000 (15:27 +0100)]
media: i2c: Add support for the OV8865 image sensor
The OV8865 is a 8 Mpx CMOS image sensor producing 3264x2448 at 30 fps.
Other modes (including some with sub-sampling) are available too.
It outputs 10-bit bayer CFA data through a MIPI CSI-2 interface with
up to 4 lanes supported.
Some register initialisation sequences are still needed for this driver,
as they cover registers for which no documentation is available.
This work is based on the first version of the driver submitted by
Kévin L'hôpital, which was adapted to mainline from the Allwinner BSP.
This version is a rewrite of the first version that matches the structure
of the OV5648 driver, with explicit PLL configuration, all the necessary
mode-specific fields, associatied registers and reduced static sequences.
It was tested with the Banana Pi Camera Board v3 and the Banana Pi M3.