This message is generated daily by a cron job that builds media_tree for
the kernels and architectures in the list below.
Results of the daily build of media_tree:
date: Tue Jul 24 05:00:11 CEST 2018
media-tree git hash:39fbb88165b2bbbc77ea7acab5f10632a31526e6
media_build gi
Switch all media platform drivers to call v4l2_async_notifier_add_subdev()
to add asd's to a notifier, in place of referencing the notifier->subdevs[]
array. These drivers also must now call v4l2_async_notifier_init() before
adding asd's to their notifiers.
There may still be cases where a platfor
On 07/23/2018 12:31 AM, Philipp Zabel wrote:
ipu_image_convert_adjust tries to adjust both input and output image at
the same time, here we just have the format of either input or output
image. Do you suggest to split this function into an input and an output
version?
See b4362162c0 ("media:
Make use of the default CSI-2 lane mapping from caller-passed
configuration.
Signed-off-by: Sakari Ailus
---
drivers/media/v4l2-core/v4l2-fwnode.c | 20
1 file changed, 12 insertions(+), 8 deletions(-)
diff --git a/drivers/media/v4l2-core/v4l2-fwnode.c
b/drivers/media/v4l2
Allow specifying the bus type explicitly for MIPI D-PHY, parallel and
Bt.656 busses. This is useful for devices that can make use of different
bus types. There are CSI-2 transmitters and receivers but the PHY
selection needs to be made between C-PHY and D-PHY; many devices also
support parallel and
In case the device supports multiple video bus types on an endpoint, the
V4L2 fwnode framework attempts to detect the type based on the available
information. This wasn't working really well, and sometimes could lead to
the V4L2 fwnode endpoint struct as being mishandled between the bus types.
Sig
Print debug info as standard V4L2 endpoint are parsed.
Signed-off-by: Sakari Ailus
Reviewed-by: Niklas Söderlund
---
drivers/media/v4l2-core/v4l2-fwnode.c | 108 ++
1 file changed, 85 insertions(+), 23 deletions(-)
diff --git a/drivers/media/v4l2-core/v4l2-fwnod
The caller may provide default flags for the endpoint. Change the
configuration based on what is available through the fwnode property API.
Signed-off-by: Sakari Ailus
---
drivers/media/v4l2-core/v4l2-fwnode.c | 19 +++
1 file changed, 19 insertions(+)
diff --git a/drivers/media
Use fwnode_graph_for_each_endpoint iterator for better readability.
Signed-off-by: Sakari Ailus
Reviewed-by: Niklas Söderlund
---
drivers/media/v4l2-core/v4l2-fwnode.c | 3 +--
1 file changed, 1 insertion(+), 2 deletions(-)
diff --git a/drivers/media/v4l2-core/v4l2-fwnode.c
b/drivers/media/v4
Hello everyone,
I've long thought the V4L2 fwnode framework requires some work (it's buggy
and it does not adequately serve common needs). This set should address in
particular these matters:
- Most devices support a particular media bus type but the V4L2 fwnode
framework was not able to use su
Instead of opportunistically trying to gather some information from the
V4L2 endpoint, set the bus type and let the V4L2 fwnode framework figure
out the configuration.
Signed-off-by: Sakari Ailus
---
drivers/media/i2c/smiapp/smiapp-core.c | 6 ++
1 file changed, 6 insertions(+)
diff --git a
Add definitions corresponding to DT bindings to the CSI-2 D-PHY, parallel
and Bt.656 busses.
Signed-off-by: Sakari Ailus
---
drivers/media/v4l2-core/v4l2-fwnode.c | 3 +++
1 file changed, 3 insertions(+)
diff --git a/drivers/media/v4l2-core/v4l2-fwnode.c
b/drivers/media/v4l2-core/v4l2-fwnode.c
The continuous clock flag was only set if there was a clock or data lanes.
This isn't needed as such a configuration is invalid to begin with. Always
set the continuous clock flag if the non-continuous property is not found.
Still don't consider the continuous clock flag as an indication that this
Use the given media bus type set by the caller. If none is given (i.e. the
mbus type is V4L2_MBUS_UNKNOWN, or 0), fall back to the old behaviour.
This is to obtain the information from the DT or try to guess the bus
type.
-ENXIO is returned if the caller sets the bus type but that does not match
w
Read the lane inversion independently of whether the "data-lanes" property
exists. This makes sense since the caller may pass the number of lanes as
the default configuration while the lane inversion configuration may still
be available in firmware.
Signed-off-by: Sakari Ailus
---
drivers/media/
Only assign endpoint configuration if the endpoint is parsed successfully.
Signed-off-by: Sakari Ailus
---
drivers/media/v4l2-core/v4l2-fwnode.c | 30 ++
1 file changed, 18 insertions(+), 12 deletions(-)
diff --git a/drivers/media/v4l2-core/v4l2-fwnode.c
b/drivers/m
Instead of allocating the V4L2 fwnode endpoint in
v4l2_fwnode_endpoint_alloc_parse, let the caller to do this. This allows
setting default parameters for the endpoint which is a very common need
for drivers.
Signed-off-by: Sakari Ailus
---
drivers/media/i2c/ov2659.c | 14
dr
Use the media bus types instead of the fwnode bus types internally. This
is the interface to the drivers as well, making the use of the fwnode bus
types more localised to the V4L2 fwnode framework.
Signed-off-by: Sakari Ailus
---
drivers/media/v4l2-core/v4l2-fwnode.c | 98 +++
Print bus type either as set by the driver or as parsed from the bus-type
property, as well as the guessed V4L2 media bus type.
Signed-off-by: Sakari Ailus
---
drivers/media/v4l2-core/v4l2-fwnode.c | 38 +++
1 file changed, 38 insertions(+)
diff --git a/drivers/m
Make use of bus type specified for an endpoint.
Signed-off-by: Sakari Ailus
---
drivers/media/v4l2-core/v4l2-fwnode.c | 48 ++-
1 file changed, 36 insertions(+), 12 deletions(-)
diff --git a/drivers/media/v4l2-core/v4l2-fwnode.c
b/drivers/media/v4l2-core/v4l2-fw
Most hardware doesn't support re-mapping of the CSI-2 lanes. Especially
sensor drivers have a default number of lanes. Instead of requiring the
caller (the driver) to provide such a unit mapping, provide one if no
mapping is configured.
Signed-off-by: Sakari Ailus
---
drivers/media/v4l2-core/v4l
The semantics of v4l2_fwnode_endpoint_parse() and
v4l2_fwnode_endpoint_alloc_parse() have changed slightly: they now take
the bus type from the user as well as a default configuration for the bus
that shall reflect the DT binding defaults. Document this.
Signed-off-by: Sakari Ailus
---
drivers/m
The V4L2 fwnode framework only parsed CSI-2 D-PHY endpoints while C-PHY
support wasn't there. Also parse endpoints for media bus type
V4L2_MBUS_CSI2_CPHY.
Signed-off-by: Sakari Ailus
---
drivers/media/v4l2-core/v4l2-fwnode.c | 10 +++---
1 file changed, 7 insertions(+), 3 deletions(-)
diff
The CSI-2 bus may use either D-PHY or C-PHY. Make this visible in media
bus enum.
Signed-off-by: Sakari Ailus
---
drivers/gpu/ipu-v3/ipu-csi.c | 2 +-
drivers/media/i2c/adv7180.c | 2 +-
drivers/media/i2c/ov5640.c | 4 ++--
drivers/m
Parsing the graph endpoint is always successful; therefore parse it as
last.
Signed-off-by: Sakari Ailus
---
drivers/media/v4l2-core/v4l2-fwnode.c | 26 +-
1 file changed, 21 insertions(+), 5 deletions(-)
diff --git a/drivers/media/v4l2-core/v4l2-fwnode.c
b/drivers/medi
Hi Steve,
On Sun, 2018-07-22 at 11:30 -0700, Steve Longerbeam wrote:
[...]
> To aid in debugging this I created branch 'imx-mem2mem.stevel' in my
> mediatree fork on github. I moved the mem2mem driver to the beginning
> and added a few patches:
>
> d317a7771c ("gpu: ipu-cpmem: add WARN_ON_ONCE()
Hi Matt,
On Sun, Jul 15, 2018 at 11:05:42PM -0700, Matt Ranostay wrote:
> Hello et all,
>
> So currently working with some thermal sensors that have coefficients
> that needs to be passed back to userspace that aren't related to the
> pixel data but are required to normalize to remove scan patter
The dw9807 contains a voice coil lens driver as well as an EEPROM. This
driver is just for the VCM. Reflect this in the driver's name --- this is
already the case for the compatible string, for instance.
Signed-off-by: Sakari Ailus
---
drivers/media/i2c/Kconfig| 2 +-
drivers
Add the DT binding documentation for dw9714 and dw9807-vcm to the
MAINTAINERS file. The dw9807-vcm binding documentation file is renamed to
match the dw9807's VCM bit's compatible string.
Signed-off-by: Sakari Ailus
---
.../bindings/media/i2c/{dongwoon,dw9807.txt => dongwoon,dw9807-vcm.txt} | 0
Hi,
I though of writing these two small patches to rename the dw9807 bindings
and driver to reflect what's apparent in the compatible string: this is
just the VCM part.
Sakari Ailus (2):
dt-bindings: dw9714, dw9807-vcm: Add files to MAINTAINERS, rename
files
dw9807-vcm: Recognise this is
On Mon, 2018-07-23 at 11:29 +0200, Philipp Zabel wrote:
[...]
> > Also, I'm trying to parse the functions find_best_seam() and
> > find_seams(). Can
> > you provide some more background on the behavior of those functions?
>
> The hardware limits us to restart linear sampling at zero with each
> t
Hi Mauro,
This is the pull request for the new vicodec driver. The cover letter is
here:
https://www.mail-archive.com/linux-media@vger.kernel.org/msg133594.html
One request: the vicodec-codec.h header refers to Tom's report on the Fast Walsh
Hadamard Transform:
https://hverkuil.home.xs4all.nl/f
Hi Steve,
On Sun, 2018-07-22 at 11:30 -0700, Steve Longerbeam wrote:
> Hi Philipp,
>
> On 07/19/2018 08:30 AM, Philipp Zabel wrote:
> > Hi,
> >
> > this is the second version of the i.MX mem2mem scaler series.
> > Patches 8 and 16 have been modified.
> >
> > Changes since v1:
> > - Fix invert
Some displays disable CEC if the HDMI 5V is not detected. In order
to test issues related to this you want to be able to optionally
detect when the 5V line changes in the cec-gpio driver.
This patch series adds support for this feature.
Regards,
Hans
The following changes since commit 3
In order to debug the HDMI 5V line we need to add a new v5-gpios
property.
Signed-off-by: Hans Verkuil
Reviewed-by: Rob Herring
---
Changes since v1:
- Document that hpd-gpios and 5v-gpios are meant for debugging those lines.
---
.../devicetree/bindings/media/cec-gpio.txt| 22 --
On 21/07/18 19:51, Tom aan de Wiel wrote:
> Hi!
>
> As of now, the version of the document I have locally is pretty much
> identical to the one you already linked to.
> I haven't had the time to shape it into a more documentation kind of document
> instead of the report format it has now.
> Howe
On Sun, 2018-07-22 at 11:02 -0700, Steve Longerbeam wrote:
> On 07/16/2018 07:12 AM, Philipp Zabel wrote:
[...]
> > > > + /*
> > > > +* The IC burst reads 8 pixels at a time. Reading
> > > > beyond the
> > > > +* end of the line is usually acceptable.
37 matches
Mail list logo
