The sensor supports 8 or 10 bit readout modes, but the
driver only supported 10 bit.

Add 8 bit readout.
Signed-off-by: Dave Stevenson <dave.stevenson@raspberrypi.com>
Reviewed-by: Jacopo Mondi <jacopo@jmondi.org>
Signed-off-by: Sakari Ailus <sakari.ailus@linux.intel.com>
Signed-off-by: Mauro Carvalho Chehab <mchehab@kernel.org>

Adds register settings for additional modes.
Signed-off-by: Dave Stevenson <dave.stevenson@raspberrypi.com>
Reviewed-by: Jacopo Mondi <jacopo@jmondi.org>
Signed-off-by: Sakari Ailus <sakari.ailus@linux.intel.com>
Signed-off-by: Mauro Carvalho Chehab <mchehab@kernel.org>

As required by libcamera, add the relevant cropping targets
to report which portion of the sensor is being read out in
any mode.
Signed-off-by: Dave Stevenson <dave.stevenson@raspberrypi.com>
Reviewed-by: Jacopo Mondi <jacopo@jmondi.org>
Signed-off-by: Sakari Ailus <sakari.ailus@linux.intel.com>
Signed-off-by: Mauro Carvalho Chehab <mchehab@kernel.org>

There's no reason why HBLANK has to be read-only as it
only changes the TIMING_HTS register in the sensor.

Remove the READ_ONLY flag, and add the relevant handling
for it.

The minimum value also varies based on whether continuous clock
mode is being used or not, so allow hblank_min to depend on
that.
Signed-off-by: Dave Stevenson <dave.stevenson@raspberrypi.com>
Reviewed-by: Jacopo Mondi <jacopo@jmondi.org>
Signed-off-by: Sakari Ailus <sakari.ailus@linux.intel.com>
Signed-off-by: Mauro Carvalho Chehab <mchehab@kernel.org>

Adds support for V4L2_CID_HFLIP and V4L2_CID_VFLIP to allow
flipping the image.

The driver previously enabled H & V flips in the register table,
therefore the controls default to the same settings to avoid
changing the behaviour.
Signed-off-by: Dave Stevenson <dave.stevenson@raspberrypi.com>
Reviewed-by: Jacopo Mondi <jacopo@jmondi.org>
Signed-off-by: Sakari Ailus <sakari.ailus@linux.intel.com>
Signed-off-by: Mauro Carvalho Chehab <mchehab@kernel.org>

Programming the sensor with TIMING_VTS (aka LPFR) was done
when triggered by a change in exposure or gain, but not
when V4L2_CID_VBLANK was changed. Dynamic frame rate
changes could therefore not be achieved.

Separate out programming TIMING_VTS so that it is triggered
by set_ctrl(V4L2_CID_VBLANK)
Signed-off-by: Dave Stevenson <dave.stevenson@raspberrypi.com>
Reviewed-by: Jacopo Mondi <jacopo@jmondi.org>
Signed-off-by: Sakari Ailus <sakari.ailus@linux.intel.com>
Signed-off-by: Mauro Carvalho Chehab <mchehab@kernel.org>

Use v4l2_ctrl_new_fwnode_properties to add V4L2_CID_CAMERA_ORIENTATION
and V4L2_CID_CAMERA_SENSOR_ROTATION.
Signed-off-by: Dave Stevenson <dave.stevenson@raspberrypi.com>
Reviewed-by: Jacopo Mondi <jacopo@jmondi.org>
Signed-off-by: Sakari Ailus <sakari.ailus@linux.intel.com>
Signed-off-by: Mauro Carvalho Chehab <mchehab@kernel.org>

The sensor supports either having the CSI2 clock lane free
running, or gated when there is no packet to transmit.
The driver only selected gated (non-continuous) clock mode.

Add code to allow fwnode to configure whether the clock is
gated or free running.
Signed-off-by: Dave Stevenson <dave.stevenson@raspberrypi.com>
Reviewed-by: Jacopo Mondi <jacopo@jmondi.org>
Signed-off-by: Sakari Ailus <sakari.ailus@linux.intel.com>
Signed-off-by: Mauro Carvalho Chehab <mchehab@kernel.org>

The configured vblank_min setting of 151 (meaning VTS of
720 + 151 = 871) is far higher than the setting that works on
the sensor, and there are no obvious restrictions stated in the
datasheet.

Reduce the vblank_min to allow for faster frame rates.
Signed-off-by: Dave Stevenson <dave.stevenson@raspberrypi.com>
Reviewed-by: Jacopo Mondi <jacopo@jmondi.org>
Signed-off-by: Sakari Ailus <sakari.ailus@linux.intel.com>
Signed-off-by: Mauro Carvalho Chehab <mchehab@kernel.org>

The calculations from pixel rate, width+hblank, and height+vblank
do not give the correct framerate - it's half the speed it should
be.

The datasheet lists the default for the TIMING_HTS registers (0x380c/d)
as being 0x2d8 (728) which is less than the width of the image, so
the units clearly can't be pixels.
If TIMING_HTS is considered to be units of 2-pixels, then the
resulting value of 0x5b0 (1456) makes all the calculations correct.

This driver is reporting an HBLANK value of 250, with an image width
of 1280, so TIMING_HTS is 1530 (0x5fa) pixels. However it was also
setting the register to 0x5fa, thereby not taking into account it
being units of 2-pixels.

Correct the register value to 0x2fd so that all the timing calculations
give the correct results.
Signed-off-by: Dave Stevenson <dave.stevenson@raspberrypi.com>
Reviewed-by: Jacopo Mondi <jacopo@jmondi.org>
Signed-off-by: Sakari Ailus <sakari.ailus@linux.intel.com>
Signed-off-by: Mauro Carvalho Chehab <mchehab@kernel.org>

The driver currently has multiple assumptions that there is
only one supported mode.

Convert supported_mode to an array, and fix up all references
to correctly look at that array.
Signed-off-by: Dave Stevenson <dave.stevenson@raspberrypi.com>
Reviewed-by: Jacopo Mondi <jacopo@jmondi.org>
Signed-off-by: Sakari Ailus <sakari.ailus@linux.intel.com>
Signed-off-by: Mauro Carvalho Chehab <mchehab@kernel.org>

The pixel rate is determined by the PLL setup in the common
registers, not by the mode specific registers, therefore
remove it from the mode definition and define it for all modes.
Signed-off-by: Dave Stevenson <dave.stevenson@raspberrypi.com>
Reviewed-by: Jacopo Mondi <jacopo@jmondi.org>
Signed-off-by: Sakari Ailus <sakari.ailus@linux.intel.com>
Signed-off-by: Mauro Carvalho Chehab <mchehab@kernel.org>

The format code is independent of mode, and each mode could
support both Y10 and Y8, so disassociate the code from the
mode.
Signed-off-by: Dave Stevenson <dave.stevenson@raspberrypi.com>
Reviewed-by: Jacopo Mondi <jacopo@jmondi.org>
Signed-off-by: Sakari Ailus <sakari.ailus@linux.intel.com>
Signed-off-by: Mauro Carvalho Chehab <mchehab@kernel.org>

Currently only one mode is supported, so all registers were
dropped in one list.
In preparation for adding more modes, split out the common registers
from those which configure the mode.
Signed-off-by: Dave Stevenson <dave.stevenson@raspberrypi.com>
Reviewed-by: Jacopo Mondi <jacopo@jmondi.org>
Signed-off-by: Sakari Ailus <sakari.ailus@linux.intel.com>
Signed-off-by: Mauro Carvalho Chehab <mchehab@kernel.org>

TIMING_VTS (registers 0x380e/f), EXPOSURE (registers
0x3500/1/2), and GAIN (0x3509) are all set from
ov9282_update_exp_gain as part of the control handler,
therefore they do not need to be in the main list of
registers.

Remove them.
Signed-off-by: Dave Stevenson <dave.stevenson@raspberrypi.com>
Reviewed-by: Jacopo Mondi <jacopo@jmondi.org>
Signed-off-by: Sakari Ailus <sakari.ailus@linux.intel.com>
Signed-off-by: Mauro Carvalho Chehab <mchehab@kernel.org>

Make sure we call ov5645_entity_init_cfg() before registering the subdev
to make sure default formats are set up.
Suggested-by: Sakari Ailus <sakari.ailus@linux.intel.com>
Signed-off-by: Lad Prabhakar <prabhakar.mahadev-lad.rj@bp.renesas.com>
Reviewed-by: Laurent Pinchart <laurent.pinchart@ideasonboard.com>
Signed-off-by: Sakari Ailus <sakari.ailus@linux.intel.com>
Signed-off-by: Mauro Carvalho Chehab <mchehab@kernel.org>

Make sure we call the PM functions while s_stream(0) even in case of
errors in the code flow.

v4l2-core takes care of warning the user so no need to add a warning
message in the driver.
Suggested-by: Sakari Ailus <sakari.ailus@linux.intel.com>
Signed-off-by: Lad Prabhakar <prabhakar.mahadev-lad.rj@bp.renesas.com>
Signed-off-by: Sakari Ailus <sakari.ailus@linux.intel.com>
Signed-off-by: Mauro Carvalho Chehab <mchehab@kernel.org>

Drop empty multiline comment.
Signed-off-by: Lad Prabhakar <prabhakar.mahadev-lad.rj@bp.renesas.com>
Reviewed-by: Laurent Pinchart <laurent.pinchart@ideasonboard.com>
Signed-off-by: Sakari Ailus <sakari.ailus@linux.intel.com>
Signed-off-by: Mauro Carvalho Chehab <mchehab@kernel.org>

Switch to using runtime PM for power management.
Signed-off-by: Lad Prabhakar <prabhakar.mahadev-lad.rj@bp.renesas.com>
Reviewed-by: Laurent Pinchart <laurent.pinchart@ideasonboard.com>
Signed-off-by: Sakari Ailus <sakari.ailus@linux.intel.com>
Signed-off-by: Mauro Carvalho Chehab <mchehab@kernel.org>

Convert the simple OV5645 Device Tree binding to json-schema.

The previous binding marked the below properties as required which was a
driver requirement and not the device requirement so just drop them from
the required list during the conversion.
- clock-frequency
- enable-gpios
- reset-gpios

Also drop the "clock-names" property as we have a single clock source for
the sensor and the driver has been updated to drop the clk referencing by
name.
Signed-off-by: Lad Prabhakar <prabhakar.mahadev-lad.rj@bp.renesas.com>
Reviewed-by: Laurent Pinchart <laurent.pinchart@ideasonboard.com>
Reviewed-by: Rob Herring <robh@kernel.org>
Signed-off-by: Sakari Ailus <sakari.ailus@linux.intel.com>
Signed-off-by: Mauro Carvalho Chehab <mchehab@kernel.org>

Now that the driver has been updated to drop fetching the clk reference by
name we no longer need the clock-names property in the ov5645 sensor node.

This is in preparation for removal for clock-names property from the DT
binding.
Signed-off-by: Lad Prabhakar <prabhakar.mahadev-lad.rj@bp.renesas.com>
Reviewed-by: Geert Uytterhoeven <geert+renesas@glider.be>
Signed-off-by: Sakari Ailus <sakari.ailus@linux.intel.com>
Signed-off-by: Mauro Carvalho Chehab <mchehab@kernel.org>

Now that the driver has been updated to drop fetching the clk reference by
name we no longer need the clock-names property in the ov5645 sensor node.

This is in preparation for removal for clock-names property from the DT
binding.
Signed-off-by: Lad Prabhakar <prabhakar.mahadev-lad.rj@bp.renesas.com>
Acked-by: Shawn Guo <shawnguo@kernel.org>
Signed-off-by: Sakari Ailus <sakari.ailus@linux.intel.com>
Signed-off-by: Mauro Carvalho Chehab <mchehab@kernel.org>

Now that the driver has been updated to drop fetching the clk reference by
name we no longer need the clock-names property in the ov5645 sensor node.

This is in preparation for removal for clock-names property from the
DT binding.
Signed-off-by: Lad Prabhakar <prabhakar.mahadev-lad.rj@bp.renesas.com>
Acked-by: Shawn Guo <shawnguo@kernel.org>
Signed-off-by: Sakari Ailus <sakari.ailus@linux.intel.com>
Signed-off-by: Mauro Carvalho Chehab <mchehab@kernel.org>

The Sony imx577 uses the same silicon enabling reference code in the
available examples provided as the imx412.

Add in compatible strings to enable and differentiate the parts.

Cc: sakari.ailus@iki.fi
Cc: dave.stevenson@raspberrypi.com
Cc: jacopo@jmondi.org
Cc: "Paul J. Murphy" <paul.j.murphy@intel.com>
Cc: Daniele Alessandrelli <daniele.alessandrelli@intel.com>
Cc: Rob Herring <robh+dt@kernel.org>
Cc: Krzysztof Kozlowski <krzysztof.kozlowski+dt@linaro.org>
Cc: devicetree@vger.kernel.org
Cc: linux-kernel@vger.kernel.org
Signed-off-by: Bryan O'Donoghue <bryan.odonoghue@linaro.org>
Signed-off-by: Sakari Ailus <sakari.ailus@linux.intel.com>
Signed-off-by: Mauro Carvalho Chehab <mchehab@kernel.org>

imx412 and imx577 return the same chip-id when interrogated via i2c.
I've confirmed this myself by

Sakari suggested we should add a new compat which should be reflected in
the name of the media entity

https://patchwork.kernel.org/project/linux-media/patch/20220607134057.2427663-3-bryan.odonoghue@linaro.org/#24894500

Set up the .data parameter of of_device_id to pass a string which
we use to set the media entity name. Once done we can add in imx577 as a
compatible chips with the media names reflecting the directed compat string.

Cc: sakari.ailus@iki.fi
Cc: dave.stevenson@raspberrypi.com
Cc: jacopo@jmondi.org
Cc: "Paul J. Murphy" <paul.j.murphy@intel.com>
Cc: Daniele Alessandrelli <daniele.alessandrelli@intel.com>
Cc: Rob Herring <robh+dt@kernel.org>
Cc: Krzysztof Kozlowski <krzysztof.kozlowski+dt@linaro.org>
Cc: devicetree@vger.kernel.org
Cc: linux-kernel@vger.kernel.org
Signed-off-by: Bryan O'Donoghue <bryan.odonoghue@linaro.org>
Signed-off-by: Sakari Ailus <sakari.ailus@linux.intel.com>
Signed-off-by: Mauro Carvalho Chehab <mchehab@kernel.org>

Add compatible bindings for imx577 which uses the same silicon enabling
reference code from Sony in the available examples provided.

Cc: sakari.ailus@iki.fi
Cc: dave.stevenson@raspberrypi.com
Cc: jacopo@jmondi.org
Cc: "Paul J. Murphy" <paul.j.murphy@intel.com>
Cc: Daniele Alessandrelli <daniele.alessandrelli@intel.com>
Cc: Rob Herring <robh+dt@kernel.org>
Cc: Krzysztof Kozlowski <krzysztof.kozlowski+dt@linaro.org>
Cc: devicetree@vger.kernel.org
Cc: linux-kernel@vger.kernel.org
Signed-off-by: Bryan O'Donoghue <bryan.odonoghue@linaro.org>
Acked-by: Rob Herring <robh@kernel.org>
Signed-off-by: Sakari Ailus <sakari.ailus@linux.intel.com>
Signed-off-by: Mauro Carvalho Chehab <mchehab@kernel.org>

Replace custom implementation of the device_match_fwnode().
Signed-off-by: Andy Shevchenko <andriy.shevchenko@linux.intel.com>
Signed-off-by: Sakari Ailus <sakari.ailus@linux.intel.com>
Signed-off-by: Mauro Carvalho Chehab <mchehab@kernel.org>

Replace custom implementation of the device_match_fwnode().
Signed-off-by: Andy Shevchenko <andriy.shevchenko@linux.intel.com>
Signed-off-by: Sakari Ailus <sakari.ailus@linux.intel.com>
Signed-off-by: Mauro Carvalho Chehab <mchehab@kernel.org>

Tag branch

* tag 'br-v6.2d' of git://linuxtv.org/hverkuil/media_tree: (35 commits)
  media: saa7164: remove variable cnt
  atomisp: fix potential NULL pointer dereferences
  radio-terratec: Remove variable p
  media: platform: s5p-mfc: Fix spelling mistake "mmaping" -> "mmapping"
  media: platform: mtk-mdp3: remove unused VIDEO_MEDIATEK_VPU config
  media: vivid: remove redundant assignment to variable checksum
  media: cedrus: h264: Optimize mv col buffer allocation
  media: cedrus: h265: Associate mv col buffers with buffer
  media: mediatek: vcodec: fix h264 cavlc bitstream fail
  media: cedrus: hevc: Fix offset adjustments
  media: imx-jpeg: Fix Coverity issue in probe
  media: v4l2-ioctl.c: Unify YCbCr/YUV terms in format descriptions
  media: atomisp: Fix spelling mistake "mis-match" -> "mismatch"
  media: c8sectpfe: Add missed header(s)
  media: adv748x: afe: Select input port when initializing AFE
  media: vimc: Update device configuration in the documentation
  media: adv748x: Remove dead function declaration
  media: mxl5005s: Make array RegAddr static const
  media: atomisp: Fix spelling mistake "modee" -> "mode"
  media: meson/vdec: always init coef_node_start
  ...

Tag branch

* tag 'br-v6.2e' of git://linuxtv.org/hverkuil/media_tree: (29 commits)
  media: davinci/vpbe: Fix a typo ("defualt_mode")
  media: sun6i-csi: Remove unnecessary print function dev_err()
  media: Documentation: Drop deprecated bytesused == 0
  media: platform: exynos4-is: fix return value check in fimc_md_probe()
  media: dvb-core: remove variable n, turn for-loop to while-loop
  media: vivid: fix compose size exceed boundary
  media: rkisp1: make const arrays ae_wnd_num and hist_wnd_num static
  media: dvb-core: Fix UAF due to refcount races at releasing
  media: rkvdec: Add required padding
  media: aspeed: Extend debug message
  media: aspeed: Support aspeed mode to reduce compressed data
  media: Documentation: aspeed-video: Add user documentation for the aspeed-video driver
  media: v4l2-ctrls: Reserve controls for ASPEED
  media: v4l: Add definition for the Aspeed JPEG format
  staging: media: tegra-video: fix device_node use after free
  staging: media: tegra-video: fix chan->mipi value on error
  media: cedrus: initialize controls a bit later
  media: cedrus: prefer untiled capture format
  media: cedrus: Remove cedrus_codec enum
  media: cedrus: set codec ops immediately
  ...

Venus updates for v6.2

* tag 'tag-venus-for-v6.2' of git://linuxtv.org/svarbanov/media_tree:
  venus: pm_helpers: Fix error check in vcodec_domains_get()
  venus: firmware: Correct assertion of reset bit on remote processor
  venus: firmware: Correct non-pix start and end addresses
  venus: firmware: Correct reset bit
  MAINTAINERS: Change email for Venus driver
  MAINTAINERS: Add Vikash as VENUS video driver co-maintainer

* git://linuxtv.org/sailus/media_tree: (47 commits)
  media: i2c: ov4689: code cleanup
  media: ov9650: Drop platform data code path
  media: ov7670: Drop unused include
  media: ov2640: Drop legacy includes
  media: tc358746: add Toshiba TC358746 Parallel to CSI-2 bridge driver
  media: dt-bindings: add bindings for Toshiba TC358746
  phy: dphy: add support to calculate the timing based on hs_clk_rate
  phy: dphy: refactor get_default_config
  v4l: subdev: Warn if disabling streaming failed, return success
  dw9768: Enable low-power probe on ACPI
  media: i2c: imx290: Replace GAIN control with ANALOGUE_GAIN
  media: i2c: imx290: Add crop selection targets support
  media: i2c: imx290: Factor out format retrieval to separate function
  media: i2c: imx290: Move registers with fixed value to init array
  media: i2c: imx290: Create controls for fwnode properties
  media: i2c: imx290: Implement HBLANK and VBLANK controls
  media: i2c: imx290: Split control initialization to separate function
  media: i2c: imx290: Fix max gain value
  media: i2c: imx290: Add exposure time control
  media: i2c: imx290: Define more register macros
  ...

Tag branch

* tag 'br-v6.2b' of git://linuxtv.org/hverkuil/media_tree: (24 commits)
  media: imx-jpeg: Lock on ioctl encoder/decoder stop cmd
  media: imx-jpeg: Support contiguous and non contiguous format
  media: imx-jpeg: Implement g_selection and s_selection
  mtk-jpegdec: add stop cmd interface for jpgdec
  media: mtk-jpegdec: refactor jpegdec func interface
  media: mtk-jpegdec: add output pic reorder interface
  media: mtk-jpegdec: add jpeg decode worker interface
  media: mtk-jpegdec: add jpegdec timeout func interface
  media: mtk-jpegdec: support jpegdec multi-hardware
  media: mtk-jpegdec: export jpeg decoder functions
  dt-bindings: mediatek: Add mediatek,mt8195-jpgdec compatible
  mtk-jpegenc: add stop cmd interface for jpgenc
  mtk-jpegenc: add output pic reorder interface
  mtk-jpegenc: add jpeg encode worker interface
  mtk-jpegenc: add jpegenc timeout func interface
  mtk-jpegenc: support jpegenc multi-hardware
  mtk-jpegenc: export jpeg encoder functions
  dt-bindings: mediatek: Add mediatek, mt8195-jpgenc compatible
  media: imx-jpeg: Disable useless interrupt to avoid kernel panic
  media: imx-jpeg: Don't clear stop state in handling dynamic resolution change
  ...

Current configuration in the document is outdated and doesn't work. Update
it to match the configuration described in the following commit:

commit 9b4a9b31 ("media: vimc: Enable set resolution at the scaler src pad")

Fixed commit description:
Shuah Khan <skhan@linuxfoundation.org>
Signed-off-by: Dafna Hirschfeld <dafna@fastmail.com>
Signed-off-by: Shuah Khan <skhan@linuxfoundation.org>
Signed-off-by: Mauro Carvalho Chehab <mchehab@kernel.org>

There are three cases of power domain management on supported platforms:
1) CAMSS on MSM8916, where a single VFE power domain is operated outside
   of the camss device driver,
2) CAMSS on MSM8996 and SDM630/SDM660, where two VFE power domains are
   managed separately by the camss device driver, the power domains are
   linked and unlinked on demand by their functions vfe_pm_domain_on()
   and vfe_pm_domain_off() respectively,
3) CAMSS on SDM845 and SM8250 platforms, and there are two VFE power
   domains and their parent power domain TITAN_TOP, the latter one
   shall be turned on prior to turning on any of VFE power domains.

Due to a previously missing link between TITAN_TOP and VFEx power domains
in the latter case, which is now fixed by [1], it was decided always to
turn on all found VFE power domains and TITAN_TOP power domain, even if
just one particular VFE is needed to be enabled or none of VFE power
domains are required, for instance the latter case is when vfe_lite is in
use. This misusage becomes more incovenient and clumsy, if next generations
are to be supported, for instance CAMSS on SM8450 has three VFE power
domains.

The change splits the power management support for platforms with TITAN_TOP
parent power domain, and, since 'power-domain-names' property is not
present in camss device tree nodes, the assumption is that the first
N power domains from the 'power-domains' list correspond to VFE power
domains, and, if the number of power domains is greater than number of
non-lite VFEs, then the last power domain from the list is the TITAN_TOP
power domain.
Signed-off-by: Vladimir Zapolskiy <vladimir.zapolskiy@linaro.org>
Reviewed-by: Robert Foss <robert.foss@linaro.org>
Signed-off-by: Hans Verkuil <hverkuil-cisco@xs4all.nl>
Signed-off-by: Mauro Carvalho Chehab <mchehab@kernel.org>

VFE lite IPs are found on CAMSS with TITAN_TOP power domains, and in
some aspects these types of VFEs are different, in particular there
is no need to enable VFE power domains to operate over VFE lite IPs.
Signed-off-by: Vladimir Zapolskiy <vladimir.zapolskiy@linaro.org>
Reviewed-by: Robert Foss <robert.foss@linaro.org>
Signed-off-by: Hans Verkuil <hverkuil-cisco@xs4all.nl>
Signed-off-by: Mauro Carvalho Chehab <mchehab@kernel.org>

The change to dynamically allocated power domains neglected a case of
CAMSS on MSM8916 platform, where a single VFE power domain is neither
attached, linked or managed in runtime in any way explicitly.

This is a special case and it shall be kept as is, because the power
domain management is done outside of the driver, and it's very different
in comparison to all other platforms supported by CAMSS.

Fixes: 6b1814e2 ("media: camss: Allocate power domain resources dynamically")
Signed-off-by: Vladimir Zapolskiy <vladimir.zapolskiy@linaro.org>
Reviewed-by: Robert Foss <robert.foss@linaro.org>
Signed-off-by: Hans Verkuil <hverkuil-cisco@xs4all.nl>
Signed-off-by: Mauro Carvalho Chehab <mchehab@kernel.org>

It is required to return the received buffers, if streaming can not be
started. For instance media_pipeline_start() may fail with EPIPE, if
a link validation between entities is not passed, and in such a case
a user gets a kernel warning:

  WARNING: CPU: 1 PID: 520 at drivers/media/common/videobuf2/videobuf2-core.c:1592 vb2_start_streaming+0xec/0x160
  <snip>
  Call trace:
   vb2_start_streaming+0xec/0x160
   vb2_core_streamon+0x9c/0x1a0
   vb2_ioctl_streamon+0x68/0xbc
   v4l_streamon+0x30/0x3c
   __video_do_ioctl+0x184/0x3e0
   video_usercopy+0x37c/0x7b0
   video_ioctl2+0x24/0x40
   v4l2_ioctl+0x4c/0x70

The fix is to correct the error path in video_start_streaming() of camss.

Fixes: 0ac2586c ("media: camss: Add files which handle the video device nodes")
Signed-off-by: Vladimir Zapolskiy <vladimir.zapolskiy@linaro.org>
Reviewed-by: Robert Foss <robert.foss@linaro.org>
Signed-off-by: Hans Verkuil <hverkuil-cisco@xs4all.nl>
Signed-off-by: Mauro Carvalho Chehab <mchehab@kernel.org>

Pull cxl fixes from Dan Williams:
 "Several fixes for CXL region creation crashes, leaks and failures.

  This is mainly fallout from the original implementation of dynamic CXL
  region creation (instantiate new physical memory pools) that arrived
  in v6.0-rc1.

  Given the theme of "failures in the presence of pass-through decoders"
  this also includes new regression test infrastructure for that case.

  Summary:

   - Fix region creation crash with pass-through decoders

   - Fix region creation crash when no decoder allocation fails

   - Fix region creation crash when scanning regions to enforce the
     increasing physical address order constraint that CXL mandates

   - Fix a memory leak for cxl_pmem_region objects, track 1:N instead of
     1:1 memory-device-to-region associations.

   - Fix a memory leak for cxl_region objects when regions with active
     targets are deleted

   - Fix assignment of NUMA nodes to CXL regions by CFMWS (CXL Window)
     emulated proximity domains.

   - Fix region creation failure for switch attached devices downstream
     of a single-port host-bridge

   - Fix false positive memory leak of cxl_region objects by recycling
     recently used region ids rather than freeing them

   - Add regression test infrastructure for a pass-through decoder
     configuration

   - Fix some mailbox payload handling corner cases"

* tag 'cxl-fixes-for-6.1-rc4' of git://git.kernel.org/pub/scm/linux/kernel/git/cxl/cxl:
  cxl/region: Recycle region ids
  cxl/region: Fix 'distance' calculation with passthrough ports
  tools/testing/cxl: Add a single-port host-bridge regression config
  tools/testing/cxl: Fix some error exits
  cxl/pmem: Fix cxl_pmem_region and cxl_memdev leak
  cxl/region: Fix cxl_region leak, cleanup targets at region delete
  cxl/region: Fix region HPA ordering validation
  cxl/pmem: Use size_add() against integer overflow
  cxl/region: Fix decoder allocation crash
  ACPI: NUMA: Add CXL CFMWS 'nodes' to the possible nodes set
  cxl/pmem: Fix failure to account for 8 byte header for writes to the device LSA.
  cxl/region: Fix null pointer dereference due to pass through decoder commit
  cxl/mbox: Add a check on input payload size