When matching and registering a sub-device from a sub-notifier use the
v4l2_device from the root parent notifier. Using the v4l2_dev stored in
the sub-notifier itself is incorrect as it might not be set.

This can be demonstrated by unbinding and rebinding the adv748x driver
and observing that it fails to probe due to the check !v4l2_dev in
v4l2_device_register_subdev().

    # echo 4-0070 > /sys/bus/i2c/drivers/adv748x/unbind
    # echo 4-0070 > /sys/bus/i2c/drivers/adv748x/bind
    adv748x 4-0070: chip found @ 0xe0 revision 2143
    adv748x 4-0070: Failed to probe TXA
    adv748x: probe of 4-0070 failed with error -22

Looking at the commit which adds sub-notifiers to V4L2 it looks like
this is the intended behavior of the original commit. With this fix the
adv748x can be re-bound and still function properly.

Fixes: 2cab00bb ("media: v4l: async: Allow binding notifiers to sub-devices")
Signed-off-by: NNiklas Söderlund <niklas.soderlund+renesas@ragnatech.se>
Signed-off-by: NSakari Ailus <sakari.ailus@linux.intel.com>
Signed-off-by: NMauro Carvalho Chehab <mchehab@s-opensource.com>

The commit aef69d54 ("media: v4l: fwnode: Add a convenience
function for registering sensors") adds the function
v4l2_async_notifier_parse_fwnode_sensor_common() to parse and register a
subdevice and a subdev-notifier by parsing firmware information. This
new subdev-notifier is stored in the new field 'subdev_notifier' in
struct v4l2_subdev.

In v4l2_async_unregister_subdev() this field is used to unregister and
cleanup the subdev-notifier. A check for if the subdev-notifier is
initialized or not was forgotten leading to a NULL pointer dereference
in v4l2_async_notifier_cleanup() if a subdevice do not use the optional
convince function to initialize the field.

Fix this by checking in v4l2_async_notifier_cleanup() that it is
provided whit a notifier making it safe to call with a NULL parameter.

Fixes: aef69d54 ("media: v4l: fwnode: Add a convenience function for registering sensors")
Signed-off-by: NNiklas Söderlund <niklas.soderlund+renesas@ragnatech.se>
Signed-off-by: NSakari Ailus <sakari.ailus@linux.intel.com>
Signed-off-by: NMauro Carvalho Chehab <mchehab@s-opensource.com>

A shadow declaration of variable ret is being assigned a return error
status and this value is being lost when the error exit goto's jump
out of the local scope. This leads to an uninitalized error return value
in the outer scope being returned. Fix this by removing the inner scoped
declaration of variable ret.

Detected by CoverityScan, CID#1460380 ("Uninitialized scalar variable")

Fixes: fb45f436 ("media: v4l: async: Fix notifier complete callback error handling")
Signed-off-by: NColin Ian King <colin.king@canonical.com>
Reviewed-by: NNiklas Söderlund <niklas.soderlund+renesas@ragnatech.se>
Signed-off-by: NSakari Ailus <sakari.ailus@linux.intel.com>
Signed-off-by: NMauro Carvalho Chehab <mchehab@s-opensource.com>

Add a convenience function for parsing firmware for information on related
devices using v4l2_async_notifier_parse_fwnode_sensor_common() registering
the notifier and finally the async sub-device itself.

This should be useful for sensor drivers that do not have device specific
requirements related to firmware information parsing or the async
framework.
Signed-off-by: NSakari Ailus <sakari.ailus@linux.intel.com>
Acked-by: NHans Verkuil <hans.verkuil@cisco.com>
Reviewed-by: NNiklas Söderlund <niklas.soderlund+renesas@ragnatech.se>
Reviewed-by: NSebastian Reichel <sebastian.reichel@collabora.co.uk>
Signed-off-by: NMauro Carvalho Chehab <mchehab@s-opensource.com>

While registering a notifier, check that each newly added fwnode is
unique, and return an error if it is not. Also check that a newly added
notifier does not have the same fwnodes twice.
Signed-off-by: NSakari Ailus <sakari.ailus@linux.intel.com>
Acked-by: NHans Verkuil <hans.verkuil@cisco.com>
Acked-by: NNiklas Söderlund <niklas.soderlund+renesas@ragnatech.se>
Reviewed-by: NSebastian Reichel <sebastian.reichel@collabora.co.uk>
Signed-off-by: NMauro Carvalho Chehab <mchehab@s-opensource.com>

Registering a notifier has required the knowledge of struct v4l2_device
for the reason that sub-devices generally are registered to the
v4l2_device (as well as the media device, also available through
v4l2_device).

This information is not available for sub-device drivers at probe time.

What this patch does is that it allows registering notifiers without
having v4l2_device around. Instead the sub-device pointer is stored in the
notifier. Once the sub-device of the driver that registered the notifier
is registered, the notifier will gain the knowledge of the v4l2_device,
and the binding of async sub-devices from the sub-device driver's notifier
may proceed.

The complete callback of the root notifier will be called only when the
v4l2_device is available and no notifier has pending sub-devices to bind.
No complete callbacks are supported for sub-device notifiers.
Signed-off-by: NSakari Ailus <sakari.ailus@linux.intel.com>
Acked-by: NHans Verkuil <hans.verkuil@cisco.com>
Acked-by: NNiklas Söderlund <niklas.soderlund+renesas@ragnatech.se>
Reviewed-by: NSebastian Reichel <sebastian.reichel@collabora.co.uk>
Signed-off-by: NMauro Carvalho Chehab <mchehab@s-opensource.com>

Refactor the V4L2 async framework a little in preparation for async
sub-device notifiers. This avoids making some structural changes in the
patch actually implementing sub-device notifiers, making that patch easier
to review.
Signed-off-by: NSakari Ailus <sakari.ailus@linux.intel.com>
Acked-by: NHans Verkuil <hans.verkuil@cisco.com>
Acked-by: NNiklas Söderlund <niklas.soderlund+renesas@ragnatech.se>
Reviewed-by: NSebastian Reichel <sebastian.reichel@collabora.co.uk>
Signed-off-by: NMauro Carvalho Chehab <mchehab@s-opensource.com>

The information on how many async sub-devices would be bindable to a
notifier is typically dependent on information from platform firmware and
it's not driver's business to be aware of that.

Many V4L2 main drivers are perfectly usable (and useful) without async
sub-devices and so if there aren't any around, just proceed call the
notifier's complete callback immediately without registering the notifier
itself.

If a driver needs to check whether there are async sub-devices available,
it can be done by inspecting the notifier's num_subdevs field which tells
the number of async sub-devices.
Signed-off-by: NSakari Ailus <sakari.ailus@linux.intel.com>
Acked-by: NHans Verkuil <hans.verkuil@cisco.com>
Acked-by: NNiklas Söderlund <niklas.soderlund+renesas@ragnatech.se>
Reviewed-by: NSebastian Reichel <sebastian.reichel@collabora.co.uk>
Signed-off-by: NMauro Carvalho Chehab <mchehab@s-opensource.com>

Register the sub-device before calling the notifier's bound callback.
Doing this the other way around is problematic as the struct v4l2_device
has not assigned for the sub-device yet and may be required by the bound
callback.
Signed-off-by: NSakari Ailus <sakari.ailus@linux.intel.com>
Acked-by: NHans Verkuil <hans.verkuil@cisco.com>
Acked-by: NPavel Machek <pavel@ucw.cz>
Acked-by: NNiklas Söderlund <niklas.soderlund+renesas@ragnatech.se>
Reviewed-by: NSebastian Reichel <sebastian.reichel@collabora.co.uk>
Signed-off-by: NMauro Carvalho Chehab <mchehab@s-opensource.com>

Add three helper functions to call async operations callbacks. Besides
simplifying callbacks, this allows async notifiers to have no ops set,
i.e. it can be left NULL.
Signed-off-by: NSakari Ailus <sakari.ailus@linux.intel.com>
Acked-by: NHans Verkuil <hans.verkuil@cisco.com>
Acked-by: NPavel Machek <pavel@ucw.cz>
Acked-by: NNiklas Söderlund <niklas.soderlund+renesas@ragnatech.se>
Reviewed-by: NSebastian Reichel <sebastian.reichel@collabora.co.uk>
Signed-off-by: NMauro Carvalho Chehab <mchehab@s-opensource.com>

The async subdev notifier .bound(), .unbind() and .complete() operations
are function pointers stored directly in the v4l2_async_subdev
structure. As the structure isn't immutable, this creates a potential
security risk as the function pointers are mutable.

To fix this, move the function pointers to a new
v4l2_async_subdev_operations structure that can be made const in
drivers.
Signed-off-by: NLaurent Pinchart <laurent.pinchart@ideasonboard.com>
Acked-by: NHans Verkuil <hans.verkuil@cisco.com>
Acked-by: NNiklas Söderlund <niklas.soderlund+renesas@ragnatech.se>
Reviewed-by: NSebastian Reichel <sebastian.reichel@collabora.co.uk>
Signed-off-by: NMauro Carvalho Chehab <mchehab@s-opensource.com>

Add two functions for parsing devices graph endpoints:
v4l2_async_notifier_parse_fwnode_endpoints and
v4l2_async_notifier_parse_fwnode_endpoints_by_port. The former iterates
over all endpoints whereas the latter only iterates over the endpoints in
a given port.

The former is mostly useful for existing drivers that currently implement
the iteration over all the endpoints themselves whereas the latter is
especially intended for devices with both sinks and sources: async
sub-devices for external devices connected to the device's sources will
have already been set up, or the external sub-devices are part of the
master device.

Depends-on: ("device property: preserve usecount for node passed to of_fwnode_graph_get_port_parent()")
Signed-off-by: NSakari Ailus <sakari.ailus@linux.intel.com>
Acked-by: NHans Verkuil <hans.verkuil@cisco.com>
Acked-by: NNiklas Söderlund <niklas.soderlund+renesas@ragnatech.se>
Signed-off-by: NMauro Carvalho Chehab <mchehab@s-opensource.com>

Rename internal functions to make the names of the functions better
describe what they do.

	Old name			New name
	v4l2_async_test_notify	v4l2_async_match_notify
	v4l2_async_belongs	v4l2_async_find_match
Signed-off-by: NSakari Ailus <sakari.ailus@linux.intel.com>
Acked-by: NHans Verkuil <hans.verkuil@cisco.com>
Acked-by: NPavel Machek <pavel@ucw.cz>
Reviewed-by: NLaurent Pinchart <laurent.pinchart@ideasonboard.com>
Reviewed-by: NSebastian Reichel <sebastian.reichel@collabora.co.uk>
Acked-by: NNiklas Söderlund <niklas.soderlund+renesas@ragnatech.se>
Signed-off-by: NMauro Carvalho Chehab <mchehab@s-opensource.com>

The check whether an async sub-device is bound to a notifier was performed
without list_lock held, making it possible for another process to
unbind the async sub-device before the sub-device unregistration function
proceeds to take the lock.

Fix this by first acquiring the lock and then proceeding with the check.
Signed-off-by: NSakari Ailus <sakari.ailus@linux.intel.com>
Reviewed-by: NSebastian Reichel <sebastian.reichel@collabora.co.uk>
Acked-by: NHans Verkuil <hans.verkuil@cisco.com>
Acked-by: NNiklas Söderlund <niklas.soderlund+renesas@ragnatech.se>
Signed-off-by: NMauro Carvalho Chehab <mchehab@s-opensource.com>

The notifier complete callback may return an error. This error code was
simply returned to the caller but never handled properly.

Move calling the complete callback function to the caller from
v4l2_async_test_notify and undo the work that was done either in async
sub-device or async notifier registration.
Reported-by: NRussell King <rmk+kernel@armlinux.org.uk>
Signed-off-by: NSakari Ailus <sakari.ailus@linux.intel.com>
Acked-by: NNiklas Söderlund <niklas.soderlund+renesas@ragnatech.se>
Reviewed-by: NSebastian Reichel <sebastian.reichel@collabora.co.uk>
Signed-off-by: NMauro Carvalho Chehab <mchehab@s-opensource.com>

The call to v4l2_async_cleanup() will set sd->asd to NULL so passing it to
notifier->unbind() has no effect and leaves the notifier confused. Call
the unbind() callback prior to cleaning up the subdevice to avoid this.
Signed-off-by: NNiklas Söderlund <niklas.soderlund+renesas@ragnatech.se>
Signed-off-by: NSakari Ailus <sakari.ailus@linux.intel.com>
Reviewed-by: NSebastian Reichel <sebastian.reichel@collabora.co.uk>
Acked-by: NHans Verkuil <hans.verkuil@cisco.com>
Signed-off-by: NMauro Carvalho Chehab <mchehab@s-opensource.com>

v4l2_async_cleanup() is called when the async sub-device is unbound from
the media device. As the pointer is set by the driver registering the
async sub-device, leave the pointer as set by the driver.
Signed-off-by: NSakari Ailus <sakari.ailus@linux.intel.com>
Reviewed-by: NSebastian Reichel <sebastian.reichel@collabora.co.uk>
Acked-by: NHans Verkuil <hans.verkuil@cisco.com>
Acked-by: NNiklas Söderlund <niklas.soderlund+renesas@ragnatech.se>
Signed-off-by: NMauro Carvalho Chehab <mchehab@s-opensource.com>

Remove V4L2 async re-probing support. The re-probing support has been
there to support cases where the sub-devices require resources provided by
the main driver's hardware to function, such as clocks.

Reprobing has allowed unbinding and again binding the main driver without
explicitly unbinding the sub-device drivers. This is certainly not a
common need, and the responsibility will be the user's going forward.

An alternative could have been to introduce notifier specific locks.
Considering the complexity of the re-probing and that it isn't really a
solution to a problem but a workaround, remove re-probing instead.

If there is a need to support the clock provider unregister/register cycle
while keeping the clock references in the consumers in the future, this
should be implemented in the clock framework instead, not in V4L2.
Signed-off-by: NSakari Ailus <sakari.ailus@linux.intel.com>
Acked-by: NHans Verkuil <hans.verkuil@cisco.com>
Acked-by: NLaurent Pinchart <laurent.pinchart@ideasonboard.com>
Reviewed-by: NSebastian Reichel <sebastian.reichel@collabora.co.uk>
Acked-by: NNiklas Söderlund <niklas.soderlund+renesas@ragnatech.se>
Signed-off-by: NMauro Carvalho Chehab <mchehab@s-opensource.com>

This reverts commit d2180e0c.

The commit was flawed in that if the device_node pointers are different,
then in fact a different device is present and the device node could be
different in ways other than full_name.

As Frank Rowand explained:

"When an overlay (1) is removed, all uses and references to the nodes and
properties in that overlay are no longer valid.  Any driver that uses any
information from the overlay _must_ stop using any data from the overlay.
Any driver that is bound to a new node in the overlay _must_ unbind.  Any
driver that became bound to a pre-existing node that was modified by the
overlay (became bound after the overlay was applied) _must_ adjust itself
to account for any changes to that node when the overlay is removed.  One
way to do this is to unbind when notified that the overlay is about to
be removed, then to re-bind after the overlay is completely removed.

If an overlay (2) is subsequently applied, a node with the same
full_name as from overlay (1) may exist.  There is no guarantee
that overlay (1) and overlay (2) are the same overlay, even if
that node has the same full_name in both cases."

Also, there's not sufficient overlay support in mainline to actually
remove and re-apply an overlay to hit this condition as overlays can
only be applied from in kernel APIs.

Fixes: d2180e0c ("[media] v4l: async: make v4l2 coexist with devicetree nodes in a dt overlay")

Cc: Javier Martinez Canillas <javier@osg.samsung.com>
Cc: Sylwester Nawrocki <s.nawrocki@samsung.com>
Cc: Frank Rowand <frowand.list@gmail.com>
Signed-off-by: NRob Herring <robh@kernel.org>
Acked-by: NJavi Merino <javi.merino@kernel.org>
Signed-off-by: NSakari Ailus <sakari.ailus@linux.intel.com>
Signed-off-by: NMauro Carvalho Chehab <mchehab@s-opensource.com>

There are multiple places where arrays or otherwise variable sized
buffer are allocated through V4L2 core code, including things like
controls, memory pages, staging buffers for ioctls and so on. Such
allocations can potentially require an order > 0 allocation from the
page allocator, which is not guaranteed to be fulfilled and is likely to
fail on a system with severe memory fragmentation (e.g. a system with
very long uptime).

Since the memory being allocated is intended to be used by the CPU
exclusively, we can consider using vmalloc() as a fallback and this is
exactly what the recently merged kvmalloc() helpers do. A kmalloc() call
is still attempted, even for order > 0 allocations, but it is done
with __GFP_NORETRY and __GFP_NOWARN, with expectation of failing if
requested memory is not available instantly. Only then the vmalloc()
fallback is used. This should give us fast and more reliable allocations
even on systems with higher memory pressure and/or more fragmentation,
while still retaining the same performance level on systems not
suffering from such conditions.

While at it, replace explicit array size calculations on changed
allocations with kvmalloc_array().

Purposedly not touching videobuf1, as it is deprecated, has only few
users remaining and would rather be seen removed instead.
Signed-off-by: NTomasz Figa <tfiga@chromium.org>
Acked-by: NMarek Szyprowski <m.szyprowski@samsung.com>
Acked-by: NSakari Ailus <sakari.ailus@linux.intel.com>
Signed-off-by: NHans Verkuil <hans.verkuil@cisco.com>
Signed-off-by: NMauro Carvalho Chehab <mchehab@s-opensource.com>

Add a check for v4l2_dev to v4l2_async_notifier_register() as to fail as
early as possible since this will fail later in v4l2_async_test_notify().
Signed-off-by: NNiklas Söderlund <niklas.soderlund+renesas@ragnatech.se>
Acked-by: NSakari Ailus <sakari.ailus@linux.intel.com>
Signed-off-by: NHans Verkuil <hans.verkuil@cisco.com>
Signed-off-by: NMauro Carvalho Chehab <mchehab@s-opensource.com>

Switch users of the v4l2_of_ APIs to the more generic v4l2_fwnode_ APIs.
Async OF matching is replaced by fwnode matching and OF matching support
is removed.
Signed-off-by: NSakari Ailus <sakari.ailus@linux.intel.com>
Acked-by: Benoit Parrot <bparrot@ti.com> # i2c/ov2569.c, am437x/am437x-vpfe.c and ti-vpe/cal.c
Tested-by: Hans Verkuil <hans.verkuil@cisco.com> # Atmel sama5d3 board + ov2640 sensor
Tested-by: NPhilipp Zabel <p.zabel@pengutronix.de>
Signed-off-by: NMauro Carvalho Chehab <mchehab@s-opensource.com>

Add fwnode matching to complement OF node matching. And fwnode may also be
an OF node.

Do not enable fwnode matching yet. It will replace OF matching soon.
Signed-off-by: NSakari Ailus <sakari.ailus@linux.intel.com>
Tested-by: NHans Verkuil <hans.verkuil@cisco.com>
Tested-by: NPhilipp Zabel <p.zabel@pengutronix.de>
Signed-off-by: NMauro Carvalho Chehab <mchehab@s-opensource.com>

v4l2-async had several functions doing some operations and then
not undoing the operations in a failure situation. For example,
v4l2_async_test_notify() moved a subdev into notifier's done list
even if registering the subdev (v4l2_device_register_subdev) failed.
If the subdev was allocated and v4l2_async_register_subdev() called
from the driver's probe() function, as usually, the probe()
function freed the allocated subdev and returned a failure.
Nevertheless, the subdev was still left into the notifier's done
list, causing an access to already freed memory when the notifier
was later unregistered.

A hand-edited call trace leaving freed subdevs into the notifier:

v4l2_async_register_notifier(notifier, asd)
cameradrv_probe
 sd = devm_kzalloc()
 v4l2_async_register_subdev(sd)
  v4l2_async_test_notify(notifier, sd, asd)
   list_move(sd, &notifier->done)
   v4l2_device_register_subdev(notifier->v4l2_dev, sd)
    cameradrv_registered(sd) -> fails
->v4l2_async_register_subdev returns failure
->cameradrv_probe returns failure
->devres frees the allocated sd
->sd was freed but it still remains in the notifier's list.

This patch fixes this and several other cases where a failing
function could leave nodes into a linked list while the caller
might free the node due to a failure.
Signed-off-by: NTuukka Toivonen <tuukka.toivonen@intel.com>
Acked-by: NSakari Ailus <sakari.ailus@linux.intel.com>
Signed-off-by: NMauro Carvalho Chehab <mchehab@s-opensource.com>

In asds configured with V4L2_ASYNC_MATCH_OF, the v4l2 subdev can be
part of a devicetree overlay, for example:

&media_bridge {
	...
	my_port: port@0 {
		#address-cells = <1>;
		#size-cells = <0>;
		reg = <0>;
		ep: endpoint@0 {
			remote-endpoint = <&camera0>;
		};
	};
};

/ {
	fragment@0 {
		target = <&i2c0>;
		__overlay__ {
			my_cam {
				compatible = "foo,bar";
				port {
					camera0: endpoint {
						remote-endpoint = <&my_port>;
						...
					};
				};
			};
		};
	};
};

Each time the overlay is applied, its of_node pointer will be
different.  We are not interested in matching the pointer, what we
want to match is that the path is the one we are expecting.  Change to
use of_node_cmp() so that we continue matching after the overlay has
been removed and reapplied.

Cc: Mauro Carvalho Chehab <mchehab@kernel.org>
Cc: Sakari Ailus <sakari.ailus@linux.intel.com>
Signed-off-by: NJavi Merino <javi.merino@kernel.org>
Reviewed-by: NJavier Martinez Canillas <javier@osg.samsung.com>
Reviewed-by: NSylwester Nawrocki <s.nawrocki@samsung.com>
Signed-off-by: NMauro Carvalho Chehab <mchehab@s-opensource.com>

A multiplication for the size determination of a memory allocation
indicated that an array data structure should be processed.
Thus use the corresponding function "kmalloc_array".

This issue was detected by using the Coccinelle software.
Signed-off-by: NMarkus Elfring <elfring@users.sourceforge.net>
Acked-by: NSakari Ailus <sakari.ailus@linux.intel.com>
Signed-off-by: NHans Verkuil <hans.verkuil@cisco.com>
Signed-off-by: NMauro Carvalho Chehab <mchehab@s-opensource.com>

The v4l2_subdev_core_ops .registered_async callback was added to notify
a subdev when its entity has been registered with the media device, to
allow for example to modify the media graph (i.e: adding entities/links).

But that's not needed since there is already a .registered callback in
struct v4l2_subdev_internal_ops that's called after the entity has been
registered with the media device in v4l2_device_register_subdev().
Suggested-by: NSakari Ailus <sakari.ailus@linux.intel.com>
Signed-off-by: NJavier Martinez Canillas <javier@osg.samsung.com>
Signed-off-by: NMauro Carvalho Chehab <mchehab@s-opensource.com>

After sub-dev registration in v4l2_async_test_notify(), the v4l2-async
core calls the registered_async callback but if a sub-dev driver does
not implement it, v4l2_subdev_call() will return a -ENOIOCTLCMD which
should not be considered an error.
Reported-by: NBenoit Parrot <bparrot@ti.com>
Signed-off-by: NJavier Martinez Canillas <javier@osg.samsung.com>
Tested-by: NBenoit Parrot <bparrot@ti.com>
Acked-by: NSakari Ailus <sakari.ailus@linux.intel.com>
Signed-off-by: NMauro Carvalho Chehab <mchehab@osg.samsung.com>

V4L2 sub-devices might need to do initialization that depends on being
registered with a V4L2 device. As an example, sub-devices with Media
Controller support may need to register entities and create pad links.

Execute the registered_async callback after the sub-device has been
registered with the V4L2 device so the driver can do any needed init.
Signed-off-by: NJavier Martinez Canillas <javier@osg.samsung.com>
Signed-off-by: NMauro Carvalho Chehab <mchehab@osg.samsung.com>

V4L2 async sub-devices are currently matched (OF case) based on the struct
device_node pointer in struct device. LED devices may have more than one
LED, and in that case the OF node to match is not directly the device's
node, but a LED's node.
Signed-off-by: NLaurent Pinchart <laurent.pinchart@ideasonboard.com>
Signed-off-by: NSakari Ailus <sakari.ailus@iki.fi>
Signed-off-by: NBryan Wu <cooloney@gmail.com>

Dynamic static allocation is evil, as Kernel stack is too low, and
compilation complains about it on some archs:
	drivers/media/v4l2-core/v4l2-async.c:238:1: warning: 'v4l2_async_notifier_unregister' uses dynamic stack allocation [enabled by default]
Instead, let's enforce a limit for the buffer.
In this specific case, there's a hard limit imposed by V4L2_MAX_SUBDEVS,
with is currently 128. That means that the buffer size can be up to
128x8 = 1024 bytes (on a 64bits kernel), with is too big for stack.
Worse than that, someone could increase it and cause real troubles.
So, let's use dynamically allocated data, instead.
Signed-off-by: NMauro Carvalho Chehab <m.chehab@samsung.com>
Reviewed-by: NHans Verkuil <hans.verkuil@cisco.com>
Signed-off-by: NMauro Carvalho Chehab <m.chehab@samsung.com>

This fixes regression introduced in commit b426b3a6
[media] V4L: Merge struct v4l2_async_subdev_list with struct v4l2_subdev
Signed-off-by: NSylwester Nawrocki <s.nawrocki@samsung.com>
Signed-off-by: NKyungmin Park <kyungmin.park@samsung.com>
Signed-off-by: NMauro Carvalho Chehab <m.chehab@samsung.com>

Calling v4l2_async_notifier_unregister() on a notifier that hasn't been
registered leads to a crash. To simplify drivers, make it safe to
unregister a notifier that has not been registered.
Signed-off-by: NLaurent Pinchart <laurent.pinchart@ideasonboard.com>
Tested-by: NSylwester Nawrocki <s.nawrocki@samsung.com>
Acked-by: NSylwester Nawrocki <s.nawrocki@samsung.com>
Signed-off-by: NMauro Carvalho Chehab <m.chehab@samsung.com>

By integrating the v4l2-async API internals a bit more with
the core overall the v4l2-async code becomes a bit simpler
and easier to follow.
Acked-and-tested-by: NLad, Prabhakar <prabhakar.csengg@gmail.com>
Signed-off-by: NSylwester Nawrocki <s.nawrocki@samsung.com>
Signed-off-by: NKyungmin Park <kyungmin.park@samsung.com>
Acked-by: NLaurent Pinchart <laurent.pinchart@ideasonboard.com>
Acked-by: NHans Verkuil <hans.verkuil@cisco.com>
Acked-by: NGuennadi Liakhovetski <g.liakhovetski@gmx.de>
Signed-off-by: NMauro Carvalho Chehab <m.chehab@samsung.com>

This is a purely cosmetic change. Since the 'subdev' member
points to an array of subdevs make it more explicit by
renaming to the plural form.
Acked-and-tested-by: NLad, Prabhakar <prabhakar.csengg@gmail.com>
Signed-off-by: NSylwester Nawrocki <s.nawrocki@samsung.com>
Signed-off-by: NKyungmin Park <kyungmin.park@samsung.com>
Acked-by: NLaurent Pinchart <laurent.pinchart@ideasonboard.com>
Acked-by: NHans Verkuil <hans.verkuil@cisco.com>
Acked-by: NGuennadi Liakhovetski <g.liakhovetski@gmx.de>
Signed-off-by: NMauro Carvalho Chehab <m.chehab@samsung.com>

Add support for matching by device_node pointer. This allows
the notifier user to simply pass a list of device_node pointers
corresponding to sub-devices.
Signed-off-by: NSylwester Nawrocki <s.nawrocki@samsung.com>
Signed-off-by: NKyungmin Park <kyungmin.park@samsung.com>
Acked-by: NLad, Prabhakar <prabhakar.csengg@gmail.com>
Acked-by: NLaurent Pinchart <laurent.pinchart@ideasonboard.com>
Acked-by: NHans Verkuil <hans.verkuil@cisco.com>
Acked-by: NGuennadi Liakhovetski <g.liakhovetski@gmx.de>
Signed-off-by: NMauro Carvalho Chehab <m.chehab@samsung.com>

enum v4l2_async_bus_type also selects a method subdevs are matched
in the notification handlers, rename it to v4l2_async_match_type
so V4L2_ASYNC_MATCH_OF entry can be further added for matching by
device tree node pointer.
Acked-and-tested-by: NLad, Prabhakar <prabhakar.csengg@gmail.com>
Signed-off-by: NSylwester Nawrocki <s.nawrocki@samsung.com>
Signed-off-by: NKyungmin Park <kyungmin.park@samsung.com>
Acked-by: NLaurent Pinchart <laurent.pinchart@ideasonboard.com>
Acked-by: NHans Verkuil <hans.verkuil@cisco.com>
Acked-by: NGuennadi Liakhovetski <g.liakhovetski@gmx.de>
Signed-off-by: NMauro Carvalho Chehab <m.chehab@samsung.com>

These match_* functions are internal callbacks and are always
invoked only after checking asd->bus_type. So drop redundant
checks in match_i2c() and match_platform() functions.
Acked-and-tested-by: NLad, Prabhakar <prabhakar.csengg@gmail.com>
Signed-off-by: NSylwester Nawrocki <s.nawrocki@samsung.com>
Signed-off-by: NKyungmin Park <kyungmin.park@samsung.com>
Acked-by: NLaurent Pinchart <laurent.pinchart@ideasonboard.com>
Acked-by: NHans Verkuil <hans.verkuil@cisco.com>
Acked-by: NGuennadi Liakhovetski <g.liakhovetski@gmx.de>
Signed-off-by: NMauro Carvalho Chehab <m.chehab@samsung.com>

i2c_verify_client() is only available, if I2C is enabled. Fix v4l2-async.c
compilation if I2C is disabled.
Signed-off-by: NGuennadi Liakhovetski <g.liakhovetski@gmx.de>
Signed-off-by: NMauro Carvalho Chehab <mchehab@redhat.com>

Currently bridge device drivers register devices for all subdevices
synchronously, typically, during their probing. E.g. if an I2C CMOS sensor
is attached to a video bridge device, the bridge driver will create an I2C
device and wait for the respective I2C driver to probe. This makes linking
of devices straight forward, but this approach cannot be used with
intrinsically asynchronous and unordered device registration systems like
the Flattened Device Tree. To support such systems this patch adds an
asynchronous subdevice registration framework to V4L2. To use it respective
(e.g. I2C) subdevice drivers must register themselves with the framework.
A bridge driver on the other hand must register notification callbacks,
that will be called upon various related events.
Signed-off-by: NGuennadi Liakhovetski <g.liakhovetski@gmx.de>
Acked-by: NHans Verkuil <hans.verkuil@cisco.com>
Acked-by: NLaurent Pinchart <laurent.pinchart@ideasonboard.com>
Tested-by: NLad, Prabhakar <prabhakar.csengg@gmail.com>
Signed-off-by: NMauro Carvalho Chehab <mchehab@redhat.com>