Hans Verkuil [Sat, 2 Jul 2011 10:04:36 +0000 (12:04 +0200)]
v4l2-ctl: fix -w support.
The -w option (use the libv4l2 wrapper library) never worked since the v4l2_open
and v4l2_close functions were never called. Fixed this.
Signed-off-by: Hans Verkuil <hans.verkuil@cisco.com>
Hans Verkuil [Sat, 2 Jul 2011 10:01:59 +0000 (12:01 +0200)]
v4l2-ctl: print the v4l2_fmtdesc flags field.
Signed-off-by: Hans Verkuil <hans.verkuil@cisco.com>
Hans Verkuil [Sat, 2 Jul 2011 10:00:01 +0000 (12:00 +0200)]
v4l2-ctl: move the doioctl_name function up to a more logical place.
doioctl_name was defined fairly late in the source, limiting where it
could be used. Move it up.
Signed-off-by: Hans Verkuil <hans.verkuil@cisco.com>
Hans Verkuil [Tue, 28 Jun 2011 10:07:26 +0000 (12:07 +0200)]
v4l2-compliance: add VIDIOC_SLICED_VBI_CAP compliance test.
Signed-off-by: Hans Verkuil <hans.verkuil@cisco.com>
Hans Verkuil [Mon, 27 Jun 2011 12:47:47 +0000 (14:47 +0200)]
v4l2-compliance: add file/line for all fails. Do some cleanups.
Signed-off-by: Hans Verkuil <hans.verkuil@cisco.com>
Hans Verkuil [Mon, 27 Jun 2011 12:25:10 +0000 (14:25 +0200)]
v4l2-compliance: add VIDIOC_G_FBUF compliance test.
Signed-off-by: Hans Verkuil <hans.verkuil@cisco.com>
Hans Verkuil [Mon, 27 Jun 2011 09:36:09 +0000 (11:36 +0200)]
v4l2-compliance: add VIDIOC_G_FMT compliance tests.
Signed-off-by: Hans Verkuil <hans.verkuil@cisco.com>
Hans Verkuil [Fri, 24 Jun 2011 10:34:28 +0000 (12:34 +0200)]
v4l2-compliance: add VIDIOC_ENUM_FRAMESIZES and VIDIOC_ENUM_FRAMEINTERVALS tests.
Signed-off-by: Hans Verkuil <hans.verkuil@cisco.com>
Hans Verkuil [Thu, 23 Jun 2011 14:13:37 +0000 (16:13 +0200)]
libv4lconvert.c: fix memset bug discovered by v4l2-compliance.
The size of the reserved array is 4 * sizeof(u32) = 16. So only the first
element is now zeroed. Use sizeof(reserved) instead.
Signed-off-by: Hans Verkuil <hans.verkuil@cisco.com>
Hans Verkuil [Thu, 23 Jun 2011 14:13:18 +0000 (16:13 +0200)]
v4l2-compliance: add VIDIOC_ENUM_FMT tests.
Signed-off-by: Hans Verkuil <hans.verkuil@cisco.com>
Hans Verkuil [Wed, 22 Jun 2011 14:27:52 +0000 (16:27 +0200)]
v4l2-compliance: fix QUERYCTRL test when there are no controls.
Signed-off-by: Hans Verkuil <hverkuil@xs4all.nl>
Hans de Goede [Sat, 18 Jun 2011 14:03:10 +0000 (16:03 +0200)]
get_media_devices: white space fixup
My bad.
Signed-off-by: Hans de Goede <hdegoede@redhat.com>
Hans de Goede [Sat, 18 Jun 2011 13:58:43 +0000 (15:58 +0200)]
get_media_devices: Get rid if asprintf
It is a GNU extension and thus not declared in the std headers when
compiling without _GNU_SOURCE
Signed-off-by: Hans de Goede <hdegoede@redhat.com>
Hans de Goede [Sat, 18 Jun 2011 13:42:22 +0000 (15:42 +0200)]
get_media_devices: Consider multiple functions on the same pci dev as one dev
This fixes get_media_devices not seeing the video and the alsa input on
my bt878 card as associated devices.
Signed-off-by: Hans de Goede <hdegoede@redhat.com>
Hans de Goede [Sat, 18 Jun 2011 13:38:08 +0000 (15:38 +0200)]
get_media_devices: Only strip usb interface from usb device path
The old code would strip to much, leading to different usb devices on the
same hub getting seen as the same device.
Signed-off-by: Hans de Goede <hdegoede@redhat.com>
Hans de Goede [Sat, 18 Jun 2011 13:29:45 +0000 (15:29 +0200)]
get_media_devices: Don't complain if there are no devices of a certain class
Signed-off-by: Hans de Goede <hdegoede@redhat.com>
Hans de Goede [Sat, 18 Jun 2011 13:27:57 +0000 (15:27 +0200)]
get_meda_devices: more robust detection of device bus
Signed-off-by: Hans de Goede <hdegoede@redhat.com>
Hans de Goede [Sat, 18 Jun 2011 13:13:40 +0000 (15:13 +0200)]
get_media_device: Continue with the next dir entry on a sysfs parse error
Rather then completely aborting the sysfs-scanning.
Signed-off-by: Hans de Goede <hdegoede@redhat.com>
Hans de Goede [Sat, 18 Jun 2011 11:52:37 +0000 (13:52 +0200)]
get_media_devices: Properly canonicalize names
Signed-off-by: Hans de Goede <hdegoede@redhat.com>
Hans de Goede [Sat, 18 Jun 2011 11:39:40 +0000 (13:39 +0200)]
get_medida_devices: use strdup
Signed-off-by: Hans de Goede <hdegoede@redhat.com>
Hans de Goede [Sat, 18 Jun 2011 11:36:21 +0000 (13:36 +0200)]
get_meda_devices: Use PATH_MAX rather then some arbritary limit
Signed-off-by: Hans de Goede <hdegoede@redhat.com>
Hans de Goede [Sat, 18 Jun 2011 11:29:15 +0000 (13:29 +0200)]
get_media_devices: Remove unnecessary class_node parsing
Signed-off-by: Hans de Goede <hdegoede@redhat.com>
Hans de Goede [Tue, 7 Jun 2011 14:38:31 +0000 (16:38 +0200)]
Add support for SE401 pixelformat
Signed-off-by: Hans de Goede <hdegoede@redhat.com>
Hans Verkuil [Tue, 7 Jun 2011 14:47:05 +0000 (16:47 +0200)]
v4l2-ctl: if V4L2_CTRL_DRIVER_PRIV(id) then use EXT_CTRLS API.
Signed-off-by: Hans Verkuil <hans.verkuil@cisco.com>
Hans Verkuil [Wed, 8 Jun 2011 12:44:41 +0000 (14:44 +0200)]
v4l2-ctl: Add support for multi-planar API
Based on a patch from Sylwester Nawrocki <snjw23@gmail.com>.
Signed-off-by: Hans Verkuil <hans.verkuil@cisco.com>
Hans Verkuil [Wed, 8 Jun 2011 11:31:13 +0000 (13:31 +0200)]
videodev2.h: update to latest version.
Needed for the multiplanar support.
Signed-off-by: Hans Verkuil <hans.verkuil@cisco.com>
Hans de Goede [Sat, 4 Jun 2011 09:09:23 +0000 (11:09 +0200)]
libv4l: autogain tweaks
-Take a bit larger steps when modifying very fine grain controls, even when
steps < 3
-Don't do step multiplication when changing exposure in the exposure low
range
Signed-off-by: Hans de Goede <hdegoede@redhat.com>
Darren Tucker [Sat, 4 Jun 2011 01:50:07 +0000 (22:50 -0300)]
Fix the highest bit of IR keytable util
I have a Nuvoton IR interface (part of an ION board) which has scancodes
with the high bit set, and ir-keytable won't accept these (since strtol
doesn't). Since "codes" is an unsigned, using strtoul fixes this.
I also added a signedness fix in debug output and fixed a typo.
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
Cédric Schieli [Sat, 28 May 2011 09:35:16 +0000 (09:35 +0000)]
keytable: fix segfault when RC driver's module_name is null
Some dvb-usb drivers do not set a proper module_name in their rc.core struct
(e.g. the ttusb2 module and various dib0700 submodules as of 2.6.39).
Auto-load mode was segfaulting if trying to match those drivers with a * value.
Signed-off-by: Cédric Schieli <cschieli@gmail.com>
Hans de Goede [Wed, 1 Jun 2011 10:47:02 +0000 (12:47 +0200)]
Release 0.8.4
Hans de Goede [Wed, 1 Jun 2011 11:08:40 +0000 (13:08 +0200)]
All: update FSF address to their latest address
Signed-off-by: Hans de Goede <hdegoede@redhat.com>
Hans de Goede [Wed, 1 Jun 2011 10:55:05 +0000 (12:55 +0200)]
qv4l2: Add rules to Makefile to actually install the desktop and icon file
Signed-off-by: Hans de Goede <hdegoede@redhat.com>
Hans de Goede [Wed, 1 Jun 2011 10:51:19 +0000 (12:51 +0200)]
qv4l2.desktop: fix desktop-file-validate warnings
[hans@shalem v4l-utils]$ desktop-file-validate utils/qv4l2/qv4l2.desktop
utils/qv4l2/qv4l2.desktop: warning: key "Encoding" in group "Desktop Entry" is deprecated
utils/qv4l2/qv4l2.desktop: warning: value "Application;AudioVideo;" for key "Categories" in group "Desktop Entry" contains a deprecated value "Application"
Signed-off-by: Hans de Goede <hdegoede@redhat.com>
Hans de Goede [Wed, 1 Jun 2011 10:14:12 +0000 (12:14 +0200)]
libv4l: Remove lenovo x201 from upside down list
It is a laptop / tablet convertible, and the webcam is not upside
down when used normally in laptop mode, when holding it upside down
in tablet mode, yes the webcam will be upside down, but that does not count.
Signed-off-by: Hans de Goede <hdegoede@redhat.com>
Mauro Carvalho Chehab [Sun, 29 May 2011 14:18:44 +0000 (11:18 -0300)]
get_media_devices: Add types for sound sequencer/timer
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
Mauro Carvalho Chehab [Sun, 29 May 2011 14:16:18 +0000 (11:16 -0300)]
get_media_devices: Prints also device major/minor
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
Mauro Carvalho Chehab [Sun, 29 May 2011 14:13:34 +0000 (11:13 -0300)]
v4l2-sysfs-path: Allow choosing the display mode
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
Mauro Carvalho Chehab [Sun, 29 May 2011 13:58:51 +0000 (10:58 -0300)]
get_media_devices: don't discard virtual devices
Instead of discarding virtual devices, just consider each of
them as a new hardware.
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
Mauro Carvalho Chehab [Sun, 29 May 2011 13:54:09 +0000 (10:54 -0300)]
get_media_devices: Constify arguments
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
Mauro Carvalho Chehab [Sun, 29 May 2011 13:44:47 +0000 (10:44 -0300)]
get_media_devices: drop a note that some alsa interfaces are missing
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
Mauro Carvalho Chehab [Sun, 29 May 2011 13:33:59 +0000 (10:33 -0300)]
get_media_devices: Add the remaining DVB nodes
From drivers/media/dvb/dvb-core/dvbdev.c
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
Mauro Carvalho Chehab [Sun, 29 May 2011 13:26:30 +0000 (10:26 -0300)]
get_media_devices: Implement the remaining V4L nodes
in order to match drivers/media/video/v4l2-dev.c
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
Mauro Carvalho Chehab [Sun, 29 May 2011 13:17:47 +0000 (10:17 -0300)]
get_media_devices: Reserve space for new devices on each type
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
Mauro Carvalho Chehab [Sun, 29 May 2011 00:55:09 +0000 (21:55 -0300)]
utils/libmedia_dev/README: add an example to list all video nodes
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
Mauro Carvalho Chehab [Sun, 29 May 2011 00:51:59 +0000 (21:51 -0300)]
get_media_devices: Add a proper documentation about the library
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
Mauro Carvalho Chehab [Sun, 29 May 2011 00:12:10 +0000 (21:12 -0300)]
get_media_devices: increment API version
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
Mauro Carvalho Chehab [Sun, 29 May 2011 00:04:51 +0000 (21:04 -0300)]
get_media_devices: get associated devices from a filestream
Sometimes, it is useful to get the alsa device associated with
a filestream. A new function adds this feature, using fstat()
call: fget_associated_device().
Tested with the code bellow:
int fd = open("/dev/video0", O_RDWR);
if (fd < 0)
perror("Couldn't open");
char *alsa = fget_associated_device(md, NULL, MEDIA_SND_CAP, fd, MEDIA_V4L_VIDEO);
printf("\n\nAlsa device = %s\n", alsa);
close(fd);
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
Mauro Carvalho Chehab [Sat, 28 May 2011 21:02:37 +0000 (18:02 -0300)]
get_media_devices: code cleanup
whitespacing & coding style fixes, and a few trivial cleanups.
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
Mauro Carvalho Chehab [Sat, 28 May 2011 20:53:22 +0000 (17:53 -0300)]
get_media_devices: While not too late, change enum definitions
Enum definitions like "NONE" and "UNKNOWN" are dangerous, as
they may conflict with something else. So add MEDIA_ prefix
to all enums to avoid it.
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
Mauro Carvalho Chehab [Sat, 28 May 2011 20:45:25 +0000 (17:45 -0300)]
get_media_devices: Extract major/minor info from the uevent nodes
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
Mauro Carvalho Chehab [Sat, 28 May 2011 20:00:33 +0000 (17:00 -0300)]
get_media_devices: fix get_associated_device() logic
If both last_seek and seek_device is specified and there are multiple
matches, an endless loop were happening.
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
Mauro Carvalho Chehab [Sat, 28 May 2011 19:49:57 +0000 (16:49 -0300)]
v4l2-sysfs-path: Show devices ordered by the video node number
Instead of calling the default print function, it implements its
own logic that will print all devices video ordered, plus the
audio playback devices. The output will be something like:
Video device: video0
vbi: vbi0
dvb frontend: dvb0.frontend0
dvb demux: dvb0.demux0
dvb dvr: dvb0.dvr0
dvb net: dvb0.net0
sound card: hw:2
pcm capture: hw:2,0
mixer: hw:2
Video device: video1
sound card: hw:1
pcm capture: hw:1,0
mixer: hw:1
Alsa playback device(s): hw:0,0 hw:0,1
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
Mauro Carvalho Chehab [Sat, 28 May 2011 19:40:57 +0000 (16:40 -0300)]
get_media_devices: fix device enumberation routines
When called inside a loop, the get_* functions weren't working.
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
Mauro Carvalho Chehab [Sat, 28 May 2011 16:13:01 +0000 (13:13 -0300)]
get_media_devices: Use a more generic way to get associated devices
Instead of having a method that works just for alsa, extend them
to be generic enough to show all types of devices.
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
Mauro Carvalho Chehab [Sat, 28 May 2011 14:09:00 +0000 (11:09 -0300)]
get_media_devices: Improve usecase example and put it at the beginning
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
Mauro Carvalho Chehab [Sat, 28 May 2011 14:08:31 +0000 (11:08 -0300)]
get_media_devices: Discard path when handling a video node name
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
Mauro Carvalho Chehab [Sat, 28 May 2011 13:54:00 +0000 (10:54 -0300)]
get_media_devices: fix a typo and remove an unused var
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
Mauro Carvalho Chehab [Sat, 28 May 2011 13:48:21 +0000 (10:48 -0300)]
get_media_devices: Make internal structs opaque
The internal structs used to retrieve sysfs information were
designed to fulfill internal design requirements in a way that
the data is easily stored inside it. They're not meant to
be used outside the library. Other methods will export the
retrieved information in a way that they'll be useful for
the applications.
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
Hans de Goede [Sat, 28 May 2011 08:10:00 +0000 (10:10 +0200)]
libv4l: update upside down device list
Signed-off-by: Hans de Goede <hdegoede@redhat.com>
Hans Verkuil [Fri, 27 May 2011 09:59:51 +0000 (11:59 +0200)]
qv4l2: filter unknown control types.
qv4l2 would crash if it encounters a control of an unknown (new?) type.
Skip such controls instead.
Signed-off-by: Hans Verkuil <hans.verkuil@cisco.com>
Hans Verkuil [Fri, 27 May 2011 09:53:10 +0000 (11:53 +0200)]
qv4l2: fix handling of private controls
Extended controls should be used for the user class if there are string
controls, integer64 controls or if V4L2_CTRL_DRIVER_PRIV(id) is true.
That last test was missing.
Signed-off-by: Hans Verkuil <hans.verkuil@cisco.com>
Hans Verkuil [Fri, 27 May 2011 09:49:40 +0000 (11:49 +0200)]
v4l2-compliance: fix S_CTRL test for read-only controls.
Fix a faulty failure report for setting read-only controls with
S_CTRL that correctly returned -EACCES.
Signed-off-by: Hans Verkuil <hans.verkuil@cisco.com>
Hans de Goede [Fri, 27 May 2011 09:39:24 +0000 (11:39 +0200)]
libv4l: update upside down device list
Signed-off-by: Hans de Goede <hdegoede@redhat.com>
Hans de Goede [Thu, 26 May 2011 10:46:09 +0000 (12:46 +0200)]
libv4l: update upside down device list
Signed-off-by: Hans de Goede <hdegoede@redhat.com>
Mauro Carvalho Chehab [Tue, 24 May 2011 10:57:14 +0000 (07:57 -0300)]
libmedia_dev: Comment the install methods for now
As a new version of v4l-utils is about to be released, for now, comment
the install procedure for the new library, as it is just too new.
I'd like to keep this library being tested for a while before exporting
it on a version release.
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
Mauro Carvalho Chehab [Tue, 24 May 2011 10:41:15 +0000 (07:41 -0300)]
libmedia_dev: Create a new library to handle media devices
The libmedia_dev library contains API's meant to be used by media
applications that needs to share some common code.
For now, it contains the routines that enumerates the media devices
and provides the associated audio devices that are needed in order
to allow playing audio for a given V4L source.
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
Mauro Carvalho Chehab [Tue, 24 May 2011 10:35:40 +0000 (07:35 -0300)]
Add proper documentation to get_media_devices API
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
Mauro Carvalho Chehab [Mon, 23 May 2011 13:07:12 +0000 (10:07 -0300)]
get_media_devices: Fix a small typo at alsa hw descriptors
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
Mauro Carvalho Chehab [Mon, 23 May 2011 10:00:00 +0000 (07:00 -0300)]
Add an install target to libv4l2util
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
Mauro Carvalho Chehab [Mon, 23 May 2011 09:43:01 +0000 (06:43 -0300)]
get_media_devices: add a function to provide an audio output device
Provide the first audio output device not related to a v4l device, assuming
that the sound card is a different device. If none found, just get any
output device, even it inside the same board that provides video4linux
interface.
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
Mauro Carvalho Chehab [Mon, 23 May 2011 09:34:04 +0000 (06:34 -0300)]
get_media_devices: Provide alsa names as expected by userspace
Instead of using card0, controlC2, etc, use the proper hw:0, hw:2, etc
names, as expected by userspace.
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
Mauro Carvalho Chehab [Mon, 23 May 2011 09:24:21 +0000 (06:24 -0300)]
get_media_devices: Find an alsa capture device associated with a video node
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
Mauro Carvalho Chehab [Mon, 23 May 2011 09:11:31 +0000 (06:11 -0300)]
get_media_devices: don't bail out if one of the clases is not available
For now, it requires that all dvb, v4l and sound classes to be available,
otherwise, an error happens.
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
Mauro Carvalho Chehab [Mon, 23 May 2011 09:10:14 +0000 (06:10 -0300)]
get_media_devices: return alsa hw as expected by alsa utilities
Instead of using the internal HWC?D? representation, convert alsa hw type
into hw:?.?. This makes easier for other utils to use it.
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
Hans de Goede [Mon, 23 May 2011 08:06:47 +0000 (10:06 +0200)]
libv4l: update upside down device list
Signed-off-by: Hans de Goede <hdegoede@redhat.com>
Mauro Carvalho Chehab [Sun, 22 May 2011 16:16:32 +0000 (13:16 -0300)]
v4l2-sysfs-path: Rewrite it to get better results
From the old v4l2-sysfs-path, we've kept just the name. The
entire old code were sent to /dev/null. On its place, a new
code were added, that reads direcly the /sys/class/<subsystem>
device links, and associates devices from video4linux, alsa
and dvb subsystems.
Also, to make easier to be used inside applications, the
discover methods were added under a LGPL library. The end
result is that v4l2-sysfs-path is now just 3 function calls.
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
Hans de Goede [Fri, 20 May 2011 11:40:55 +0000 (13:40 +0200)]
libv4l: update upside down device list
Signed-off-by: Hans de Goede <hdegoede@redhat.com>
Antonio Ospite [Fri, 20 May 2011 07:43:50 +0000 (09:43 +0200)]
libv4lconvert: Add Y10B, a 10 bpp bit-packed greyscale format.
Add a 10 bits per pixel greyscale format in a bit-packed array representation,
naming it Y10B. Such pixel format is supplied for instance by the Kinect
sensor device.
Cc: Steven Toth<stoth@kernellabs.com>
Signed-off-by: Antonio Ospite<ospite@studenti.unina.it>
Signed-off-by: Hans de Goede <hdegoede@redhat.com>
Hans de Goede [Fri, 20 May 2011 07:23:00 +0000 (09:23 +0200)]
libv4l: update upside down device table
Signed-off-by: Hans de Goede <hdegoede@redhat.com>
Hans de Goede [Tue, 17 May 2011 19:27:16 +0000 (21:27 +0200)]
libv4l: update TODO
Hans de Goede [Tue, 17 May 2011 18:59:38 +0000 (20:59 +0200)]
libv4l: update upside down device table
Signed-off-by: Hans de Goede <hdegoede@redhat.com>
Hans de Goede [Sat, 14 May 2011 15:15:25 +0000 (17:15 +0200)]
libv4l: Add error handling to libjpeg calling code
Signed-off-by: Hans de Goede <hdegoede@redhat.com>
Hans de Goede [Sat, 14 May 2011 05:29:03 +0000 (07:29 +0200)]
libv4l: use libjpeg[-turbo] for decompressing jpeg-s
libjpeg-turbo is much faster then tinyjpeg, and with modern cameras
doing 1920x1080@30fps with jpeg compression, we really need that speed.
Todo: libjpeg error handling (fixed by next patch in series)
Signed-off-by: Hans de Goede <hdegoede@redhat.com>
Hans de Goede [Fri, 13 May 2011 13:51:33 +0000 (15:51 +0200)]
libv4l: Move tinyjpeg invocation to helper function
Preperation patch for adding support for decoding JPEG-s using libjpeg
(which is faster).
Signed-off-by: Hans de Goede <hdegoede@redhat.com>
Hans de Goede [Fri, 13 May 2011 13:53:43 +0000 (15:53 +0200)]
libv4lconvert: s/jdev/tinyjpeg/
Preperation patch for adding support for decoding JPEG-s using libjpeg
(which is faster).
Signed-off-by: Hans de Goede <hdegoede@redhat.com>
Hans de Goede [Wed, 11 May 2011 19:23:23 +0000 (21:23 +0200)]
libv4l: detect (usb) connection speed
And use it for determining the optimal source format.
Signed-off-by: Hans de Goede <hdegoede@redhat.com>
Hans de Goede [Tue, 10 May 2011 08:55:56 +0000 (10:55 +0200)]
qv4l2: remove some redundancy from inputChanged
updateAudioInput already does a g_audio and m_audioInput->setCurrentIndex,
no need to do it twice.
Signed-off-by: Hans de Goede <hdegoede@redhat.com>
Hans de Goede [Tue, 10 May 2011 08:52:47 +0000 (10:52 +0200)]
qv4l2: m_freq listens to valueChanged block it to avoid recursion
Signed-off-by: Hans de Goede <hdegoede@redhat.com>
Hans de Goede [Mon, 9 May 2011 20:36:30 +0000 (22:36 +0200)]
qv4l2: Add and hookup set / get interval code
Signed-off-by: Hans de Goede <hdegoede@redhat.com>
Hans de Goede [Mon, 9 May 2011 18:37:48 +0000 (20:37 +0200)]
qv4l2: always update capture format after a set cap fmt
Changing just the width / height may also change other fmt settings,
atleast it will influence the available intervals.
Also:
* only do a s_fmt if the preceeding try_fmt succeeded
* even update capture format if the try or s_fmt failed, so that the
ui element gets restored to the current setting
Signed-off-by: Hans de Goede <hdegoede@redhat.com>
Hans de Goede [Mon, 9 May 2011 18:23:14 +0000 (20:23 +0200)]
qv4l2: general-tab: make sure m_width and height get a proper init value
m_width and m_height were not getting assigned any meaningful value on
startup, this causes the enum_frameintervals call inframeIntervalChanged
to fail. This patch makes sure that m_width and m_height get assigned
the actual width and height on startup, by setting them from updateVidCapFormat
this also removes the need for passing the actual width and height into
updateFrameSize and updateFrameInterval, they can now simply use
m_width and m_height since these are now always accurate.
Signed-off-by: Hans de Goede <hdegoede@redhat.com>
Hans de Goede [Mon, 9 May 2011 18:01:26 +0000 (20:01 +0200)]
libv4l: Adjust src fmt for fps when app does a S_PARM with new fps
Signed-off-by: Hans de Goede <hdegoede@redhat.com>
Hans de Goede [Mon, 9 May 2011 11:56:14 +0000 (13:56 +0200)]
libv4l: s/v4l2_pix_fmt_identical/v4l2_pix_fmt_compat/
The current identical function checks for compat pix formats, not for identical
ones (the padding may be different), so rename it.
Signed-off-by: Hans de Goede <hdegoede@redhat.com>
Hans de Goede [Tue, 10 May 2011 07:44:06 +0000 (09:44 +0200)]
libv4l: Honor V4L2_DISABLE_CONVERSION flag for TRY_FMT ioctls
Signed-off-by: Hans de Goede <hdegoede@redhat.com>
Hans de Goede [Mon, 9 May 2011 08:14:32 +0000 (10:14 +0200)]
libv4l: rewrite src format selection algorithm
When multiple src formats are available for the best resolution match
we need to decide which one to use. The old algorithm was rather crude:
- it depended on the formats being sorted in a certain way in the
source formats array
- it did not really take bus bandwidth / fps into account other then
always preferring compressed formats except for resolutions of quarter
CIF and lower
- it did not differentiate between converting to YUV420 or to RGB32, while
some src formats are clearly a better match for one then for the other
This rewritten src format ranking algorithm takes all of the above into
account.
Signed-off-by: Hans de Goede <hdegoede@redhat.com>
Hans de Goede [Sun, 8 May 2011 17:28:12 +0000 (19:28 +0200)]
libv4l: Make the supported_src_formats bitfield 64 bits
We've been having > 32 src formats for a while now, so we really should
have done this earlier already ...
Signed-off-by: Hans de Goede <hdegoede@redhat.com>
Hans de Goede [Sun, 8 May 2011 17:25:08 +0000 (19:25 +0200)]
libv4l: add support for M420 pixel-format
Signed-off-by: Hans de Goede <hdegoede@redhat.com>
Hans de Goede [Sun, 8 May 2011 10:12:47 +0000 (12:12 +0200)]
libv4l: Webcam image upside down on Asus Eee PC T101MT (13d3:5122)
The caption pretty much says it all. Owing to different IDs, the flags
for similar Asus products don't cut in.
Originally reported for Ubuntu at
https://bugs.launchpad.net/ubuntu/+source/libv4l/+bug/774123.
Signed-off-by: Stefan Löffler <st.loeffler@gmail.com>
Signed-off-by: Hans de Goede <hdegoede@redhat.com>
Robby Workman [Tue, 12 Apr 2011 14:26:57 +0000 (11:26 -0300)]
keytable: Allow override of manpage installation directory
This creates MANDIR in Make.rules and keeps the preexisting
default of $(PREFIX)/share/man, but allows packagers to easily
override via e.g. "make MANDIR=/usr/man"
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
Robby Workman [Tue, 3 May 2011 04:28:37 +0000 (01:28 -0300)]
keytable: Install udev rules to /lib/udev/ instead of /etc/udev
In moderately recent versions of udev (since 125 in 7/2008),
packages should install rules files to /lib/udev/rules.d/
instead of /etc/udev/rules.d/, as /etc/udev/rules.d/ is now
for generated rules and overrides of the packaged rules.
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
projects / platform / upstream / v4l-utils.git / log