lp:libv4l

Branch merges

This import branch has no branches proposed for merge into it.

No branches dependent on this one.

Related source package recipes

2 recipes using this branch. (?)

Related snap packages

Related bugs

Related blueprints

5326. By Hans Verkuil <email address hidden> on 2025-10-29

v4l2-compliance: skip V4L2_CID_PRIVATE_BASE check for UVC

UVC doesn't support V4L2_CID_PRIVATE_BASE, so skip the check.

Signed-off-by: Hans Verkuil <email address hidden>
Reviewed-by: Ricardo Ribalda <email address hidden>
Tested-by: Ricardo Ribalda <email address hidden>

5325. By Hans Verkuil <email address hidden> on 2025-10-28

edid-decode: reading EDID from non-0 segment failed

When reading the EDID from a non-zero segment the EDID offset
wasn't written ('write_message'), so the read data was shifted.

Make sure to include the offset in the i2c transaction.

Signed-off-by: Hans Verkuil <email address hidden>

5324. By Hans Verkuil <email address hidden> on 2025-10-22

edid-decode: fix speaker mappings

The SADB, HDMI ADB and RCDB all have slightly different speaker
mappings, unfortunately. So create three separate mappings for the
three DBs.

The SADB deprecation check was also incorrect, one bit was marked
as invalid, when it was really valid.

Signed-off-by: Hans Verkuil <email address hidden>

5323. By Hans Verkuil <email address hidden> on 2025-10-22

edid-decode: sync with CTA-861-J standard

A new CTA-861-J standard was released. Update edid-decode accordingly:

- Added one check for empty Audio Data Block
- Updated speaker allocation checks

Signed-off-by: Hans Verkuil <email address hidden>

5322. By Quentin Schulz <email address hidden> on 2025-10-17

README.md: couple of typo fixes

- Should be "mailing list" instead.
- Derivated isn't appropriate here and we should be saying derivative
  instead, c.f. https://www.debian.org/derivatives/.
- requiried has one too many i.

Signed-off-by: Quentin Schulz <email address hidden>
Signed-off-by: Hans Verkuil <email address hidden>

5321. By Quentin Schulz <email address hidden> on 2025-10-17

add basic b4 config file

b4[1] is a very nice tool for mail-based contribution. A config[2] file
exists to set up a few defaults. We can use it to set the Cc recipients
to always add, in our case the mailing list.

This shouldn't be necessary if we had a script that b4 prep --auto-to-cc
could call to find the mail address(es) to send to (cc), à-la
scripts/get_maintainer.pl from the kernel tree.

This allows us to not have to look for the mailing list to send the
patches to, just use b4 and you're all set.

[1] https://pypi.org/project/b4/
[2] https://b4.docs.kernel.org/en/latest/config.html

Signed-off-by: Quentin Schulz <email address hidden>
Signed-off-by: Hans Verkuil <email address hidden>

5320. By Quentin Schulz <email address hidden> on 2025-10-17

README.md: migrate instruction to "meson setup"

meson [options] is deprecated since 0.64.0 (commit 3c7ab542c0c4
("deprecate running "meson builddir" without setup subcommand")) and
throws a warning telling you to use meson setup [options] instead, so
let's do that.

While 0.64.0 is higher than the current minimum requirement of 0.60,
meson setup is, according to the aforementioned commit, supported since
0.42, so this change should be fine for meson versions between 0.60 and
0.64.

Signed-off-by: Quentin Schulz <email address hidden>
Signed-off-by: Hans Verkuil <email address hidden>

5319. By Quentin Schulz <email address hidden> on 2025-10-16

v4l2-ctl: do not fail on kernel not implementing VIDIOC_SUBDEV_QUERYCAP

The comment above the code snippet specifies that we do not want to fail
on kernels not implementing VIDIOC_SUBDEV_QUERYCAP (5.9 and earlier) but
calling doioctl will still set app_result to a negative value if the
kernel doesn't implement it (this ioctl returning -1 with errno set to
ENOTTY). Considering app_result is used as the exit code of the program,
ignoring the return value of doioctl function is not enough to
make this ioctl truly optional.

Let's handle this ioctl as optional to avoid kernels lacking it making
v4l2-ctl command fail (but still doing what it's been asked to do) by
calling test_ioctl directly.

Fixes: fc89f67d9fe0 ("v4l2-ctl: add support for VIDIOC_SUBDEV_QUERYCAP")
Signed-off-by: Quentin Schulz <email address hidden>
Signed-off-by: Hans Verkuil <email address hidden>

5318. By Quentin Schulz <email address hidden> on 2025-10-16

v4l2-ctl: do not fail on kernel not implementing VIDIOC_SUBDEV_S_CLIENT_CAP

VIDIOC_SUBDEV_S_CLIENT_CAP is only available in kernels 6.4 and later.

The current logic has a fallback in case the ioctl fails by simply
stating there's no capabilities available for the subdev.

However, doioctl sets app_result on any ioctl fail (regardless of errno,
which is set to ENOIOCTLCMD when this ioctl doesn't exist). Considering
app_result is used as the exit code of the program, ignoring the return
value of doioctl function is not enough to make this ioctl truly
optional.

Let's handle this ioctl as optional to avoid kernels lacking it making
v4l2-ctl command fail (but still doing what it's been asked to do) by
calling test_ioctl directly.

Fixes: 2b2ba7e0e86c ("v4l2-ctl/compliance: read and show the subdev client capabilities")
Signed-off-by: Quentin Schulz <email address hidden>
Signed-off-by: Hans Verkuil <email address hidden>

5317. By Hans Verkuil <email address hidden> on 2025-10-16

edid-decode: check for UHD VIC bit

Check if the UHD VIC bit is set correctly.

I thought this was checked already, but it wasn't. Add the new test.

Signed-off-by: Hans Verkuil <email address hidden>