media: ipu-bridge: Serialise calls to IPU bridge init

The IPU bridge initialisation will be called from multiple locations in
the future. Serialise the access to devices' fwnodes in this context.

Signed-off-by: Sakari Ailus <sakari.ailus@linux.intel.com>
Signed-off-by: Hans Verkuil <hverkuil-cisco@xs4all.nl>

media: ipu-bridge: Move graph checking to IPU bridge

Move checking the graph to the IPU bridge. This way the caller won't need
to do it.

Signed-off-by: Sakari Ailus <sakari.ailus@linux.intel.com>
Signed-off-by: Hans Verkuil <hverkuil-cisco@xs4all.nl>

media: ipu-bridge: Add ov01a10 in Dell XPS 9315

Add Omnivision ov01a10 sensor used in Dell XPS 9315, and use the driver
default frequency.

Signed-off-by: Sakari Ailus <sakari.ailus@linux.intel.com>
Reviewed-by: Daniel Scally <dan.scally@ideasonboard.com>
Signed-off-by: Hans Verkuil <hverkuil-cisco@xs4all.nl>

media: ipu-bridge: Change ov2740 link-frequency to 180 MHz

The only known devices that use an ov2740 sensor in combination with
the ipu-bridge code are various Lenovo ThinkPad models, which all
need the link-frequency to be 180 MHz for things to work properly.

The ov2740 driver used to only support 360 MHz link-frequency,
which is why the ipu-bridge entry used 360 MHz, but now the
ov2740 driver has been extended to also support 180 MHz.

The ov2740 is actually used with 360 MHz link-frequency on Chromebooks.
On Chromebooks the camera/sensor fwnode graph is part of the ACPI tables.
The ipu-bridge code is used to dynamically generate the graph when it is
missing, so it is not used on Chromebooks and the ov2740 will keep using
360 MHz link-frequency there as before.

Signed-off-by: Hans de Goede <hdegoede@redhat.com>
Signed-off-by: Sakari Ailus <sakari.ailus@linux.intel.com>
Signed-off-by: Hans Verkuil <hverkuil-cisco@xs4all.nl>

media: ipu-bridge: Add missing acpi_dev_put() in ipu_bridge_get_ivsc_acpi_dev()

In ipu_bridge_get_ivsc_acpi_dev(), the "ivsc_adev" acpi_device pointer
from the outer loop is handed over to the caller, which takes proper
care of its reference count.
However, the "consumer" acpi_device pointer from the inner loop is lost,
without decrementing its reference count.

Fix this by adding the missing call to acpi_dev_put().

Fixes: c66821f381ae ("media: pci: intel: Add IVSC support for IPU bridge driver")
Signed-off-by: Geert Uytterhoeven <geert+renesas@glider.be>
Reviewed-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>
Signed-off-by: Hans Verkuil <hverkuil-cisco@xs4all.nl>

media: ipu-bridge: Add link-frequency to OV2680 ipu_supported_sensors[] entry

Now that the ov2680 driver verifies the bus-cfg from the endpoint
fwnode the link-frequency must be set for things to work.

Signed-off-by: Hans de Goede <hdegoede@redhat.com>
Signed-off-by: Sakari Ailus <sakari.ailus@linux.intel.com>
Signed-off-by: Mauro Carvalho Chehab <mchehab@kernel.org>

media: pci: intel: Add IVSC support for IPU bridge driver

Previously on ACPI platforms, sensors that are intended to be connected
to an IPU device for use with the ipu3-cio2 driver lacking the necessary
connection information in firmware. IPU bridge driver is to connect
sensors to IPU device via software nodes.

Currently IVSC located between IPU device and sensors is available in
existing commercial platforms from multiple OEMs. But the connection
information between them in firmware is also not enough to build V4L2
connection graph. This patch parses the connection properties from the
SSDB buffer in DSDT and build the connection using software nodes.

IVSC driver is based on MEI framework (previously known as HECI), it
has two MEI clients, MEI CSI and MEI ACE. Both clients are used to
communicate messages with IVSC firmware. Linux abstracts MEI client
as a device, whose bus type is MEI. And the device is addressed by a
GUID/UUID which is part of the device name of MEI client. After figured
out MEI CSI via the UUID composed device name, this patch setup the
connection between MEI CSI and IPU, and the connection between MEI CSI
and sensor via software nodes.

Signed-off-by: Wentong Wu <wentong.wu@intel.com>
Reviewed-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>

media: ipu-bridge: Add a runtime-pm device-link between VCM and sensor

In most cases when a VCM is used there is a single integrated module
with the sensor + VCM + lens. This means that the sensor and VCM often
share regulators and possibly also something like a powerdown pin.

In the ACPI tables this is modelled as a single ACPI device with
multiple I2cSerialBus resources.

On atomisp devices the regulators and clks are modelled as ACPI
power-resources, which are controlled by the (ACPI) power state
of the sensor. So the sensor must be in D0 power state for the VCM
to work.

To make this work add a device-link with DL_FLAG_PM_RUNTIME flag
so that the sensor will automatically be runtime-resumed whenever
the VCM is runtime-resumed.

This requires the probing of the VCM and thus the creation of the VCM
I2C-client to be delayed till after the sensor driver has bound.

Move the instantiation of the VCM I2C-client to the v4l2_async_notifier
bound op, so that it is done after the sensor driver has bound; and
add code to add the device-link.

This fixes the problem with the shared ACPI power-resources on atomisp2
and this avoids the need for VCM related workarounds on IPU3 / IPU6.

E.g. until now the dw9719 driver needed to get and control a Vsio
(V sensor IO) regulator since that needs to be enabled to enable I2C
pass-through on the PMIC on the sensor module. So the driver was
controlling this regulator even though the actual dw9719 chip has no
Vsio pin / power-plane.

This also removes the need for ipu_bridge_init() to return
-EPROBE_DEFER since the VCM is now instantiated later.

Reviewed-by: Andy Shevchenko <andy@kernel.org>
Reviewed-by: Daniel Scally <dan.scally@ideasonboard.com>
Tested-by: Daniel Scally <dan.scally@ideasonboard.com>
Signed-off-by: Hans de Goede <hdegoede@redhat.com>
Signed-off-by: Sakari Ailus <sakari.ailus@linux.intel.com>
Signed-off-by: Mauro Carvalho Chehab <mchehab@kernel.org>

media: ipu-bridge: Add GalaxyCore GC0310 to ipu_supported_sensors[]

The GalaxyCore GC0310 is used together with the atomisp on various
devices, add it to ipu_supported_sensors[].

Reviewed-by: Andy Shevchenko <andy.shevchenko@gmail.com>
Reviewed-by: Daniel Scally <dan.scally@ideasonboard.com>
Signed-off-by: Hans de Goede <hdegoede@redhat.com>
Signed-off-by: Sakari Ailus <sakari.ailus@linux.intel.com>
Signed-off-by: Mauro Carvalho Chehab <mchehab@kernel.org>

media: ipu-bridge: Move ipu-bridge.h to include/media/

Move ipu-bridge.h to include/media/, so that it can also be used by
the atomisp code.

Reviewed-by: Andy Shevchenko <andy.shevchenko@gmail.com>
Signed-off-by: Hans de Goede <hdegoede@redhat.com>
Signed-off-by: Sakari Ailus <sakari.ailus@linux.intel.com>
Signed-off-by: Mauro Carvalho Chehab <mchehab@kernel.org>

media: ipu-bridge: Add a parse_sensor_fwnode callback to ipu_bridge_init()

Add a parse_sensor_fwnode() callback to ipu_bridge_init(), so that
ipu_bridge_init() can be used with other sensor fwnode parse functions
then just ipu_bridge_parse_ssdb().

This will allow the ipu3-bridge code to also be used by the atomisp
driver.

Signed-off-by: Hans de Goede <hdegoede@redhat.com>
Reviewed-by: Andy Shevchenko <andy@kernel.org>
Signed-off-by: Sakari Ailus <sakari.ailus@linux.intel.com>
Signed-off-by: Mauro Carvalho Chehab <mchehab@kernel.org>

media: ipu-bridge: Drop early setting of sensor->adev

sensor->adev is no longer dereferenced before it is permanently set by:

sensor->adev = acpi_dev_get(adev);

So the early assignment with a borrowed reference can be dropped.

Signed-off-by: Hans de Goede <hdegoede@redhat.com>
Signed-off-by: Sakari Ailus <sakari.ailus@linux.intel.com>
Signed-off-by: Mauro Carvalho Chehab <mchehab@kernel.org>

media: ipu-bridge: Add a ipu_bridge_parse_ssdb() helper function

The code to go from ACPI sensor info to a fwnode-tree with connector
nodes and endpoint properties is 99% the same for the atomisp2 and
the IPU3.

The main difference is that atomisp2 devices do not have a SSDB table
with various info.

Abstract out the parsing of the sensor's ACPI fwnode into a helper
function and store the parsed results, rather then the raw SSDB
in struct ipu_sensor.

This is a preparation patch for making the ipu-bridge code more generic
so that it can be shared with the atomisp driver.

Reviewed-by: Andy Shevchenko <andy@kernel.org>
Reviewed-by: Daniel Scally <dan.scally@ideasonboard.com>
Signed-off-by: Hans de Goede <hdegoede@redhat.com>
Signed-off-by: Sakari Ailus <sakari.ailus@linux.intel.com>
Signed-off-by: Mauro Carvalho Chehab <mchehab@kernel.org>

media: ipu-bridge: Only keep PLD around while parsing

There is no need to keep a reference to the PLD struct around,
it is only used once the get the sensor orientation.

Make ipu_bridge_parse_orientation() also get + put the PLD.

This is a preparation patch for making the ipu-bridge code more generic
so that it can be shared with the atomisp driver.

Reviewed-by: Andy Shevchenko <andy@kernel.org>
Reviewed-by: Daniel Scally <dan.scally@ideasonboard.com>
Signed-off-by: Hans de Goede <hdegoede@redhat.com>
Signed-off-by: Sakari Ailus <sakari.ailus@linux.intel.com>
Signed-off-by: Mauro Carvalho Chehab <mchehab@kernel.org>

media: ipu-bridge: Store dev pointer in struct ipu_bridge

Store the dev pointer in struct ipu_bridge instead of passing it
around 3 levels deep.

This takes up slightly more memory but further patches in this series
add more data which needs to be passed around making passing
everything as arguments cumbersome and those further patches also
add data to struct ipu_bridge.

To be consistent with these upcoming patches also add
the dev pointer to struct ipu_bridge.

Reviewed-by: Andy Shevchenko <andy@kernel.org>
Reviewed-by: Daniel Scally <dan.scally@ideasonboard.com>
Signed-off-by: Hans de Goede <hdegoede@redhat.com>
Signed-off-by: Sakari Ailus <sakari.ailus@linux.intel.com>
Signed-off-by: Mauro Carvalho Chehab <mchehab@kernel.org>

media: ipu-bridge: Move initialization of node_names.vcm to ipu_bridge_init_swnode_names()

Move initialization of node_names.vcm to ipu_bridge_init_swnode_names()
where it belongs.

And make the initialization of nodes[SWNODE_VCM] unconditional,
ipu_bridge_init_swnode_group() takes care of not registering it
when there is no VCM.

Reviewed-by: Andy Shevchenko <andy@kernel.org>
Reviewed-by: Daniel Scally <dan.scally@ideasonboard.com>
Signed-off-by: Hans de Goede <hdegoede@redhat.com>
Signed-off-by: Sakari Ailus <sakari.ailus@linux.intel.com>
Signed-off-by: Mauro Carvalho Chehab <mchehab@kernel.org>

media: ipu-bridge: Move initialization of node_names.vcm to ipu_bridge_init_swnode_names()

Move initialization of node_names.vcm to ipu_bridge_init_swnode_names()
where it belongs.

And make the initialization of nodes[SWNODE_VCM] unconditional,
ipu_bridge_init_swnode_group() takes care of not registering it
when there is no VCM.

Reviewed-by: Andy Shevchenko <andy@kernel.org>
Reviewed-by: Daniel Scally <dan.scally@ideasonboard.com>
Signed-off-by: Hans de Goede <hdegoede@redhat.com>
Signed-off-by: Sakari Ailus <sakari.ailus@linux.intel.com>
Signed-off-by: Mauro Carvalho Chehab <mchehab@kernel.org>

media: ipu-bridge: Do not use on stack memory for software_node.name field

Commit 567f97bd381f ("media: ipu3-cio2: support multiple sensors and VCMs
with same HID") introduced an on stack vcm_name and then uses this for
the name field of the software_node struct used for the vcm.

But the software_node struct is much longer lived then the current
stack-frame, so this is no good.

Instead extend the ipu_node_names struct with an extra field to store
the vcm software_node name and use that.

Note this also changes the length of the allocated buffer from
ACPI_ID_LEN + 4 to 16. the name is filled with "<ipu_vcm_types[x]>-%u"
where ipu_vcm_types[x] is not an ACPI_ID. The maximum length of
the strings in the ipu_vcm_types[] array is 11 + 5 bytes for "-255\0"
means 16 bytes are needed in the worst case scenario.

Fixes: 567f97bd381f ("media: ipu3-cio2: support multiple sensors and VCMs with same HID")
Cc: Bingbu Cao <bingbu.cao@intel.com>
Reviewed-by: Andy Shevchenko <andy@kernel.org>
Reviewed-by: Daniel Scally <dan.scally@ideasonboard.com>
Signed-off-by: Hans de Goede <hdegoede@redhat.com>
Signed-off-by: Sakari Ailus <sakari.ailus@linux.intel.com>
Signed-off-by: Mauro Carvalho Chehab <mchehab@kernel.org>

media: ipu-bridge: Fix null pointer deref on SSDB/PLD parsing warnings

When ipu_bridge_parse_rotation() and ipu_bridge_parse_orientation() run
sensor->adev is not set yet.

So if either of the dev_warn() calls about unknown values are hit this
will lead to a NULL pointer deref.

Set sensor->adev earlier, with a borrowed ref to avoid making unrolling
on errors harder, to fix this.

Fixes: 485aa3df0dff ("media: ipu3-cio2: Parse sensor orientation and rotation")
Cc: Fabian Wüthrich <me@fabwu.ch>
Signed-off-by: Hans de Goede <hdegoede@redhat.com>
Reviewed-by: Daniel Scally <dan.scally@ideasonboard.com>
Signed-off-by: Sakari Ailus <sakari.ailus@linux.intel.com>
Signed-off-by: Mauro Carvalho Chehab <mchehab@kernel.org>

media: ipu3-cio2: allow ipu_bridge to be a module again

This code was previously part of the VIDEO_IPU3_CIO2 driver, which could
be built-in or a loadable module, but after the move it turned into a
builtin-only driver. This fails to link when the I2C subsystem is a
module:

x86_64-linux-ld: drivers/media/pci/intel/ipu-bridge.o: in function `ipu_bridge_unregister_sensors':
ipu-bridge.c:(.text+0x50): undefined reference to `i2c_unregister_device'
x86_64-linux-ld: drivers/media/pci/intel/ipu-bridge.o: in function `ipu_bridge_init':
ipu-bridge.c:(.text+0x9c9): undefined reference to `i2c_acpi_new_device_by_fwnode'

In general, drivers should not have to be built-in, so change the option
to a tristate with the corresponding dependency. This in turn opens a
new problem with the dependency, as the IPU bridge can be a loadable module
while the ipu3 driver itself is built-in, producing a new link failure:

86_64-linux-ld: drivers/media/pci/intel/ipu3/ipu3-cio2.o: in function `cio2_pci_probe':
ipu3-cio2.c:(.text+0x197e): undefined reference to `ipu_bridge_init'

In order to fix this, restore the old Kconfig option that controlled
the ipu bridge driver before it was split out, but make it select a
hidden symbol that now corresponds to the bridge driver.

When other drivers get added that share ipu-bridge, this should cover
all corner cases, and allow any combination of them to be built-in
or modular.

Link: https://lore.kernel.org/linux-media/20230727122331.2421453-1-arnd@kernel.org

Fixes: 881ca25978c6 ("media: ipu3-cio2: rename cio2 bridge to ipu bridge and move out of ipu3")'
Signed-off-by: Arnd Bergmann <arnd@arndb.de>
Reviewed-by: Andy Shevchenko <andriy.shevchenko@linux.intel.com>
Signed-off-by: Mauro Carvalho Chehab <mchehab@kernel.org>

media: ipu-bridge: use IPU_MAX_PORTS for bridge instead of CIO2_NUM_PORTS

Before bridge driver use CIO2_NUM_PORTS as the maximum supported CSI2 port
number. Current bridge driver is moved out of ipu3, so define a new macro
in ipu bridge for all IPUs instead of including CIO2 definition.
This patch also removes the ipu3-cio2.h inclusion in ipu-bridge.h.

Signed-off-by: Bingbu Cao <bingbu.cao@intel.com>
Reviewed-by: Daniel Scally <dan.scally@ideasonboard.com>
Signed-off-by: Sakari Ailus <sakari.ailus@linux.intel.com>
Signed-off-by: Mauro Carvalho Chehab <mchehab@kernel.org>

media: ipu3-cio2: rename cio2 bridge to ipu bridge and move out of ipu3

cio2 bridge was involved along with IPU3. However, in fact all Intel IPUs
besides IPU3 CIO2 need this bridge driver. This patch move bridge driver
out of ipu3 directory and rename as ipu-bridge. Then it can be worked with
IPU3 and other Intel IPUs.

Signed-off-by: Bingbu Cao <bingbu.cao@intel.com>
Reviewed-by: Daniel Scally <dan.scally@ideasonboard.com>
Signed-off-by: Sakari Ailus <sakari.ailus@linux.intel.com>
Signed-off-by: Mauro Carvalho Chehab <mchehab@kernel.org>