We have numerous lines over 80 chars, or oddly split. Many
of these are due to using long enum names such as
MMAL_COMPONENT_CAMERA.
Reduce the length of these enum names.
Signed-off-by: Dave Stevenson <dave.stevenson@raspberrypi.org>
Fix checkpatch warnings over spaces around operators.
Many were around operations that can be replaced with the
BIT(x) macro, so replace with that where appropriate.
Signed-off-by: Dave Stevenson <dave.stevenson@raspberrypi.org>
None of the listed author email addresses were valid.
Keep list of authors and the companies they represented.
Update my email address.
Signed-off-by: Dave Stevenson <dave.stevenson@raspberrypi.org>
Fix several instances where it is easier to return
early on error conditions than handle it as an else
clause.
As requested by Mauro.
Signed-off-by: Dave Stevenson <dave.stevenson@raspberrypi.org>
H264 header come from VC with 0 timestamps, which means they get a
strange timestamp when processed with VC/kernel start times,
particularly if used with the inline header option.
Remember the last frame timestamp and use that if set, or otherwise
use the kernel start time.
https://github.com/raspberrypi/linux/issues/1836
Signed-off-by: Dave Stevenson <dave.stevenson@raspberrypi.org>
vchi_bulk_queue_receive will queue up to a default of 4
bulk receives on a connection before blocking.
If called from the VCHI service_callback thread, then
that thread is unable to service the VCHI_CALLBACK_BULK_RECEIVED
events that would enable the queue call to succeed.
Add a workqueue to schedule the call vchi_bulk_queue_receive
in an alternate context to avoid the lock up.
Signed-off-by: Dave Stevenson <dave.stevenson@raspberrypi.org>
The handling of loading platform drivers requires checking IS_ERR
for the pointer on unload.
If the driver fails to load, NULL the pointer during probe as
platform_device_unregister already checks for NULL.
Signed-off-by: Dave Stevenson <dave.stevenson@raspberrypi.org>
It is a waste of resources to load the camera driver in case there isn't
a camera actually connected to the Raspberry Pi. This solution also
avoids a NULL ptr dereference of mmal instance on driver unload.
Fixes: 7b3ad5abf0 ("staging: Import the BCM2835 MMAL-based V4L2 camera driver.")
Signed-off-by: Stefan Wahren <stefan.wahren@i2se.com>
As per default the GPU memory config of the Raspberry Pi isn't sufficient
for the camera usage. Even worse the bcm2835 camera doesn't provide a
helpful error message in this case. So let's add a hint to point the user
to the likely cause.
Signed-off-by: Stefan Wahren <stefan.wahren@i2se.com>
Currently there is only a catch-all info message which print the
relevant error code without any context. So add more specific error
messages in order to narrow down possible issues.
Signed-off-by: Stefan Wahren <stefan.wahren@i2se.com>
Just like the bcm2835-video make this a platform driver which is probed
by vchiq. In order to change the number of channels use a module
parameter instead, but use the maximum as default.
Signed-off-by: Stefan Wahren <stefan.wahren@i2se.com>
Since the camera driver isn't probed via DT, we need to properly setup DMA.
Fixes: 37b7b3087a ("staging/vc04_services: Register a platform device for the camera driver.")
Signed-off-by: Stefan Wahren <stefan.wahren@i2se.com>
In error case platform_device_register_data would return an ERR_PTR
instead of NULL. So we better check this before unregistration.
Fixes: 37b7b3087a ("staging/vc04_services: Register a platform device for the camera driver.")
Signed-off-by: Stefan Wahren <stefan.wahren@i2se.com>
In order to have this more consistent between the vc04 services move
the module information to the end of the file.
Signed-off-by: Stefan Wahren <stefan.wahren@i2se.com>
In order to make the module bcm2835-camera load automatically, we need to
add a module alias.
Fixes: 4bebb0312e ("staging/bcm2835-camera: Set ourselves up as a platform driver.")
Signed-off-by: Stefan Wahren <stefan.wahren@i2se.com>
When calling tc358743_set_fmt, the code was calling tc358743_get_fmt
to choose a valid format. However that sets the colorspace
based on what was read back from the chip. When you set the format,
then the driver would choose and program the colorspace based
on the format code.
The result was that if you called try or set format for UYVY
when the current format was RGB3 then you would get told sRGB,
and try RGB3 when current was UYVY and you would get told
SMPTE170M.
The value programmed into the chip is determined by this driver,
therefore there is no need to read back the value. Return the
colorspace based on the format set/tried instead.
Signed-off-by: Dave Stevenson <dave.stevenson@raspberrypi.org>
The current colorspace was always returned from try_fmt for no
good reason.
Return what the source subdevice returns instead.
Signed-off-by: Dave Stevenson <dave.stevenson@raspberrypi.org>
Create a semi-static mapping for the USB registers early in the boot
process, before additional kernel threads are started, so all threads
will have the mappings from the start. This avoids the need for
data aborts to lazily update them.
See: https://github.com/raspberrypi/linux/issues/2450
Signed-off-by: Floris Bos <bos@je-eigen-domein.nl>
Add "eth_downshift_after" DT parameter to allow the delay before the
downshift to be specified. The default is 2 auto-negotiation cycles,
and legal values are 2, 3, 4, 5 and 0 (disabled).
Add "eth_max_speed" DT parameter as a way of prohibiting 1000Mbps
links. This can be used to avoid the delay until the downshift mechanism
activates. Legal values are 10, 100 and 1000, where the default is
unlimited (effectively 1000Mbps).
Signed-off-by: Phil Elwell <phil@raspberrypi.org>
Enable the auto-downshift feature on Raspberry Pi 3B+ so that a link
can eventually be established using a cable with pairs C and/or D
missing or broken in a 1000Mbps-capable port.
Signed-off-by: Phil Elwell <phil@raspberrypi.org>
Ethernet cables with faulty or missing pairs (specifically pairs C and
D) allow auto-negotiation to 1000Mbs, but do not support the successful
establishment of a link. Add a DT property, "microchip,downshift-after",
to configure the number of auto-negotiation failures after which it
falls back to 100Mbs. Valid values are 2, 3, 4, 5 and 0, where 0 means
never downshift.
Signed-off-by: Phil Elwell <phil@raspberrypi.org>
The DT node describing the LAN7800s PHY has now moved inside an "mdio"
node. Update the DT declarations accordingly.
Signed-off-by: Phil Elwell <phil@raspberrypi.org>
The upstreamed driver for the GPIO expander expects to be a children of
the "firmware" node.
The patch also removes the "firmware" phandle as it's useless.
Signed-off-by: Nicolas Saenz Julienne <nsaenzjulienne@suse.de>
The upstreamed version of the firmware node has been updated to present
it as a "simple-bus". We need to get this in order to accomodate other
device bindings, namely RPi's firmware based gpio expander.
Signed-off-by: Nicolas Saenz Julienne <nsaenzjulienne@suse.de>
The vc4-kms-v3d overlay gained an extra fragment enabling the txp node,
so rebuild the upstream overlay to match.
Signed-off-by: Phil Elwell <phil@raspberrypi.org>
Adds the option of configuring i2c0 to be on GPIOs 0&1 as
this is of use on the Compute Module.
Also fixes the ov5647 overlay where the override enabled the wrong
fragments.
Signed-off-by: Dave Stevenson <dave.stevenson@raspberrypi.org>
This should avoid clicks when stopping and immediately afterwards
starting a stream with the same samplerate as before.
Signed-off-by: Matthias Reichl <hias@horus.com>
Since kernel 4.0 the PWRDN register bits are under DAPM
control from the wm8804 driver.
Drop code that modifies that register to avoid interfering
with DAPM.
Signed-off-by: Matthias Reichl <hias@horus.com>
This reverts commit 3b12dcf797f5a4635aecd7f5c090dc507b124ffd.
Despite the commit message being wrong idle_bias changes
were already reverted in the 4.14 tree.
So drop the commit to bring the wm8804 driver back in line with
the rpi-4.14.y and upstream linux trees.
Signed-off-by: Matthias Reichl <hias@horus.com>
Newer versions of dtc warn about unnecessary usage of #address-cells
and #size-cells, so remove them.
Signed-off-by: Phil Elwell <phil@raspberrypi.org>
A previous change moved away from variable stack
allocation of a data buffer to a fixed maximum size.
However, some mailbox calls use larger data buffers
than the maximum allowed. This change moves from
stack storage to kmalloc to ensure all sizes are
catered for.
Signed-off-by: James Hughes <james.hughes@raspberrypi.org>
The driver that supports the ADV7282-M also supports the ADV7280-M,
ADV7281-M, and ADV7281-MA.
The 7280-M exposes 8 analogue inputs. The 7281-M doesn't have the
I2P deinterlacing block. The 7281-MA has 8 inputs but no I2P.
Otherwise they are the same as ADV7282-M.
Adds a new overlay "adv728x" that includes the existing adv7282
overlay but adds several parameters to modify the behaviour.
Adds a new addr parameter to allow the I2C address to be changed.
(the chip has an address select pin to change between 0x20 and 0x21).
Signed-off-by: Dave Stevenson <dave.stevenson@raspberrypi.org>
DT overlays to setup the above devices via i2c_arm and csi1.
(This currently does not use the i2c-mux-pinctrl driver to
dynamically switch the pinctrl)
tc358743 is tc358743 running at a default link frequency
of 972Mbit/s. This allows up to 1080P50 UYVY on 2 lanes.
There is a parameter to allow changing the link frequency,
but the only values supported by the driver are 297000000
for 594Mbit/s, and 486000000 for 972Mbit/s.
There is also a parameter to enable 4 lane mode (only
relevant to Compute Module (1 or 3) csi1).
tc358743-audio overlay enables I2S audio from the TC358743
to the Pi (SD to GPIO20, SCK to GPIO18, WFS to GPIO19).
ADV7282M is the Analog Devices analogue video to CSI bridge
chip.
OV5647 is the Pi V1.3 camera module. Currently the driver only
supports VGA 8bit Bayer and very few controls.
Signed-off-by: Dave Stevenson <dave.stevenson@raspberrypi.org>
Whilst the adv7180 driver support s_routing, nothing else
does, and there is a missing lump of framework code to
define the mapping from connectors on a board to the inputs
they represent on the ADV7180.
Add a nasty hack to take a module parameter that is passed in
to s_routing on any call to G_STD, or S_STD (or subdev
g_input_status call).
Signed-off-by: Dave Stevenson <dave.stevenson@raspberrypi.org>
Enable:
VIDEO_V4L2_SUBDEV_API=y
VIDEO_BCM2835_UNICAM=m
VIDEO_TC358743=m
VIDEO_ADV7180=m
VIDEO_OV5647=m
so that we can receive CSI data from these devices.
Signed-off-by: Dave Stevenson <dave.stevenson@raspberrypi.org>
