Gregor Jasny [Wed, 31 Jul 2013 19:44:11 +0000 (21:44 +0200)]
libdvbv5: Actually check for lock
Signed-off-by: Gregor Jasny <gjasny@googlemail.com>
CC: Mauro Carvalho Chehab <mchehab@infradead.org>
Gregor Jasny [Wed, 31 Jul 2013 19:37:39 +0000 (21:37 +0200)]
xc3028: Remove duplicate const
Signed-off-by: Gregor Jasny <gjasny@googlemail.com>
Gregor Jasny [Wed, 31 Jul 2013 19:32:49 +0000 (21:32 +0200)]
libv4lconvert: Fix resource leak in sq905 error path
Detected by Coverity.
Signed-off-by: Gregor Jasny <gjasny@googlemail.com>
CC: Hans de Goede <hdegoede@redhat.com>
Gregor Jasny [Thu, 25 Jul 2013 21:27:17 +0000 (23:27 +0200)]
v4l2-ctl: Do not read past end of num_planes array
Detected by Coverity.
Signed-off-by: Gregor Jasny <gjasny@googlemail.com>
CC: Hans Verkuil <hverkuil@xs4all.nl>
Gregor Jasny [Thu, 25 Jul 2013 21:00:20 +0000 (23:00 +0200)]
pixfmt-test: Explicitely mention fall-through
Spotted by Coverity.
Signed-off-by: Gregor Jasny <gjasny@googlemail.com>
Gregor Jasny [Thu, 25 Jul 2013 20:43:14 +0000 (22:43 +0200)]
libv4lconvert: Avoid division by 0 if gamma value is corrupted
Detected by Coverity.
Signed-off-by: Gregor Jasny <gjasny@googlemail.com>
CC: Hans de Goede <hdegoede@redhat.com>
Gregor Jasny [Thu, 25 Jul 2013 12:46:58 +0000 (14:46 +0200)]
libdvbv5: Fix copy and paste error in parse_service()
Detected by Coverity.
Signed-off-by: Gregor Jasny <gjasny@googlemail.com>
CC: Mauro Carvalho Chehab <mchehab@infradead.org>
CC: André Roth <neolynx@gmail.com>
Gregor Jasny [Mon, 22 Jul 2013 20:27:02 +0000 (22:27 +0200)]
rds-ctl: Always terminate strings properly
Detected by Coverity.
Signed-off-by: Gregor Jasny <gjasny@googlemail.com>
Reviewed-by: Hans Verkuil <hans.verkuil@cisco.com>
Gregor Jasny [Mon, 22 Jul 2013 20:08:09 +0000 (22:08 +0200)]
libdvbv5: Fix reallocation in parse_lcn
Detected by Coverity.
Signed-off-by: Gregor Jasny <gjasny@googlemail.com>
CC: Mauro Carvalho Chehab <mchehab@infradead.org>
Gregor Jasny [Mon, 22 Jul 2013 19:52:52 +0000 (21:52 +0200)]
xc3082: Fix use after free in free_firmware()
Detected by Coverity Scanner.
CC: Mauro Carvalho Chehab <mchehab@infradead.org>
Signed-off-by: Gregor Jasny <gjasny@googlemail.com>
Ricardo Ribalda Delgado [Tue, 30 Jul 2013 12:59:23 +0000 (14:59 +0200)]
v4l2_compliance: -EINVAL is expected when ret is not 0
Otherwise the driver can never return a register
Signed-off-by: Ricardo Ribalda Delgado <ricardo.ribalda@gmail.com>
Signed-off-by: Hans Verkuil <hans.verkuil@cisco.com>
Hans Verkuil [Tue, 30 Jul 2013 12:01:15 +0000 (14:01 +0200)]
v4l2-ctl: fix coverity issues.
Fixed a number of obvious coding issues found by Coverity.
Signed-off-by: Hans Verkuil <hans.verkuil@cisco.com>
Hans Verkuil [Tue, 30 Jul 2013 11:48:21 +0000 (13:48 +0200)]
v4l2-dbg: fix coverity issues.
Fixed a number of obvious coding issues found by Coverity.
Signed-off-by: Hans Verkuil <hans.verkuil@cisco.com>
Hans Verkuil [Tue, 30 Jul 2013 11:35:01 +0000 (13:35 +0200)]
rds-ctl: fix coverity issues.
Fixed a number of obvious coding issues found by Coverity.
Signed-off-by: Hans Verkuil <hans.verkuil@cisco.com>
Hans Verkuil [Tue, 30 Jul 2013 11:16:56 +0000 (13:16 +0200)]
libv4l2rds: fix coverity issues.
Fixed a number of obvious coding issues found by Coverity.
Signed-off-by: Hans Verkuil <hans.verkuil@cisco.com>
Hans Verkuil [Mon, 29 Jul 2013 14:20:56 +0000 (16:20 +0200)]
qv4l2: always update controls by default
The 'Update on change' checkbox was by default only checked for the user controls
tab, not for the others. This is however very confusing and in general you always
want to update a control immediately once it is changed, unless you explicitly
uncheck the checkbox.
The original idea was that for e.g. the MPEG controls you want to set them all up
just the way you want it, and then press Update to update all the changes using
extended controls. This is particularly useful for controls that depend on one
another.
In practice however, and certanly with lots of the new controls that appeared,
you just want to have an immediate effect when changing a control.
Having 'Update on change' disabled just leads to confused users, which frankly
includes myself :-)
Signed-off-by: Hans Verkuil <hans.verkuil@cisco.com>
Hans Verkuil [Mon, 29 Jul 2013 14:16:59 +0000 (16:16 +0200)]
Sync with 3.11 kernel.
Signed-off-by: Hans Verkuil <hans.verkuil@cisco.com>
Mauro Carvalho Chehab [Fri, 26 Jul 2013 10:32:51 +0000 (07:32 -0300)]
parse_string: fix encodings for Taiwan
As reported on:
http://www.mail-archive.com/linux-media@vger.kernel.org/msg64804.html
by Huey-Horng Yo <hiroshiyui@gmail.com>, the proper handling for 0x14
region code is UTF-16BE. Fix it, in order to get the proper channel
names.
Reported-by: Huey-Horng Yo <hiroshiyui@gmail.com>
Signed-off-by: Mauro Carvalho Chehab <m.chehab@samsung.com>
Guy Martin [Tue, 18 Jun 2013 13:19:09 +0000 (10:19 -0300)]
dvbv5-zap: Parse the LNB from the channel file
Parsing the LNB needs to be done for proper tuning.
Signed-off-by: Guy Martin <gmsoft@tuxicoman.be>
Acked-by: André Roth <neolynx@gmail.com>
Signed-off-by: Mauro Carvalho Chehab <m.chehab@samsung.com>
Guy Martin [Tue, 18 Jun 2013 13:19:08 +0000 (10:19 -0300)]
libdvbv5: Use a temporary copy of the dvb parameters when tuning
This patch copies the parms provided into a temporary buffer. This buffer will then
be used for any modification that needs to be performed. It makes the function
dvb_fe_set_parms() thread-safe. Also, since the DTV_FREQUENCY is not modified, it fixes
a bug where dvbv5-scan retrieves the frequency from the parms and write it to the
channel file.
Signed-off-by: Guy Martin <gmsoft@tuxicoman.be>
Acked-by: André Roth <neolynx@gmail.com>
Signed-off-by: Mauro Carvalho Chehab <m.chehab@samsung.com>
Guy Martin [Tue, 18 Jun 2013 13:19:07 +0000 (10:19 -0300)]
libdvbv5: Fix satellite handling and apply polarization parameter to the frontend
Apply polarization parameters even if a satellite number is not provided. This allow
proper setup of the tone and voltage.
Signed-off-by: Guy Martin <gmsoft@tuxicoman.be>
Acked-by: André Roth <neolynx@gmail.com>
Signed-off-by: Mauro Carvalho Chehab <m.chehab@samsung.com>
Guy Martin [Tue, 18 Jun 2013 13:19:06 +0000 (10:19 -0300)]
libdvbv5: Export dvb_fe_is_satellite()
This patch makes the function dvb_fe_is_satellite() availble from libdvbv5. This function is simple
but yet very handful to have around.
Signed-off-by: Guy Martin <gmsoft@tuxicoman.be>
Acked-by: André Roth <neolynx@gmail.com>
Signed-off-by: Mauro Carvalho Chehab <m.chehab@samsung.com>
Guy Martin [Tue, 18 Jun 2013 13:19:05 +0000 (10:19 -0300)]
libdvbv5: Add parsing of POLARIZATION
This patch add parsing support for the POLARIZATION parameter for the DVBv5 file format.
Signed-off-by: Guy Martin <gmsoft@tuxicoman.be>
Acked-by: André Roth <neolynx@gmail.com>
Signed-off-by: Mauro Carvalho Chehab <m.chehab@samsung.com>
Guy Martin [Tue, 18 Jun 2013 13:19:04 +0000 (10:19 -0300)]
libdvbv5: Remove buggy parsing of extra DTV_foo parameters
The parsing of those extra parameters is buggy and completely useless since they are parsed
individually later on in the code.
Signed-off-by: Guy Martin <gmsoft@tuxicoman.be>
Acked-by: André Roth <neolynx@gmail.com>
Signed-off-by: Mauro Carvalho Chehab <m.chehab@samsung.com>
Gregor Jasny [Thu, 27 Jun 2013 20:09:00 +0000 (22:09 +0200)]
libv4lconvert: Prevent integer overflow by checking width and height
The Mayhem Team found a crash caused by an integer overflow.
Details are here:
http://www.forallsecure.com/bug-reports/
8aae67d864bce76993f3f9812b4a2aeea0eb38da/
Signed-off-by: Gregor Jasny <gjasny@googlemail.com>
Acked-by: Hans de Goede <hdegoede@redhat.com>
Acked-by: Mauro Carvalho Chehab <mchehab@redhat.com>
Gregor Jasny [Thu, 27 Jun 2013 20:57:53 +0000 (22:57 +0200)]
keytable: Always check if strtok return value is null
The Mayhem Team found a crash caused by a nullptr.
Details are here:
http://www.forallsecure.com/bug-reports/
567323cd26f180910beb03ae26afb40c432a0c6a/
Signed-off-by: Gregor Jasny <gjasny@googlemail.com>
Acked-by: Hans de Goede <hdegoede@redhat.com>
Acked-by: Mauro Carvalho Chehab <mchehab@redhat.com>
Hans Verkuil [Fri, 28 Jun 2013 09:04:13 +0000 (11:04 +0200)]
Remove committed v4l2-chip-ident.h headers.
This header is no longer needed and will be removed altogether in kernel 3.11.
Signed-off-by: Hans Verkuil <hans.verkuil@cisco.com>
Hans Verkuil [Fri, 28 Jun 2013 09:03:48 +0000 (11:03 +0200)]
libv4l2/log.c: add entry for new VIDIOC_DBG_G_CHIP_INFO ioctl.
Signed-off-by: Hans Verkuil <hans.verkuil@cisco.com>
Hans Verkuil [Fri, 28 Jun 2013 08:50:39 +0000 (10:50 +0200)]
Sync with latest 3.10.0-rc7 kernel.
This also added two new keymaps.
Signed-off-by: Hans Verkuil <hans.verkuil@cisco.com>
Hans Verkuil [Fri, 28 Jun 2013 08:49:25 +0000 (10:49 +0200)]
Remove references to v4l2-chip-ident.h
This header will be removed soon and v4l2-dbg no longer uses it.
So remove any references to it.
Signed-off-by: Hans Verkuil <hans.verkuil@cisco.com>
Gregor Jasny [Sat, 22 Jun 2013 15:21:20 +0000 (17:21 +0200)]
libdvb: Disable library installation by default
Signed-off-by: Gregor Jasny <gjasny@googlemail.com>
Hans Verkuil [Fri, 21 Jun 2013 05:41:02 +0000 (07:41 +0200)]
v4l2-ctl: V4L2_BUF_TYPE_PRIVATE is no longer used.
Remove references to this obsolete buffer type.
Signed-off-by: Hans Verkuil <hans.verkuil@cisco.com>
Hans de Goede [Wed, 19 Jun 2013 22:11:05 +0000 (00:11 +0200)]
libv4l2: do not log a ton of errors on device unplug
Signed-off-by: Hans de Goede <hdegoede@redhat.com>
Hans de Goede [Wed, 19 Jun 2013 21:59:13 +0000 (23:59 +0200)]
libv4l2: Add V4L2_PERROR convenience macro
Signed-off-by: Hans de Goede <hdegoede@redhat.com>
Hans Verkuil [Wed, 19 Jun 2013 15:43:19 +0000 (17:43 +0200)]
v4l2-ctl: fix wrong array accesses in selection code.
The code to print, get and set selections was broken: the while condition
increased the index as well, so when accessing elements inside the while the
array index was one too high.
Signed-off-by: Hans Verkuil <hans.verkuil@cisco.com>
Gregor Jasny [Tue, 18 Jun 2013 20:47:26 +0000 (22:47 +0200)]
buildsystem: Do not bootstrap complete gettext for testing iconv
Automake 1.14 will break with gettext older than 0.18.2 because the
deprecated AM_PROG_MKDIR_P has been removed.
Signed-off-by: Gregor Jasny <gjasny@googlemail.com>
Gregor Jasny [Tue, 18 Jun 2013 19:31:01 +0000 (21:31 +0200)]
qv4l2: Silence Qt build tools invocation if silent mode is enabled
Signed-off-by: Gregor Jasny <gjasny@googlemail.com>
Hans Verkuil [Sat, 15 Jun 2013 09:09:35 +0000 (11:09 +0200)]
v4l2-compliance: fix typo
fmt1 -> fmt2
Signed-off-by: Hans Verkuil <hans.verkuil@cisco.com>
Devin Heitmueller [Fri, 14 Jun 2013 13:59:26 +0000 (09:59 -0400)]
Don't call G_TUNER unless actually performing a tuning related call
Making G_TUNER calls can take a long time on some tuners, in
particular those that load firmware or do power management. As a
result, we don't want to call G_TUNER unless the user is actually
doing a tuning related call. The current code makes a G_TUNER
call regardless of what command the user is attempting to perform.
Problem originally identified on the HVR-950q, where even doing
operations like toggling from the composite to the s-video input
would take over 1000ms.
Signed-off-by: Devin Heitmueller <dheitmueller@kernellabs.com>
Signed-off-by: Hans Verkuil <hans.verkuil@cisco.com>
Hans de Goede [Sun, 9 Jun 2013 14:31:55 +0000 (16:31 +0200)]
libv4l2: Add logging of dqbuf timestamps to debug logging
Signed-off-by: Hans de Goede <hdegoede@redhat.com>
Hans Verkuil [Fri, 7 Jun 2013 08:22:51 +0000 (10:22 +0200)]
rds-ctl: support -d10 to refer to radio10.
In order to be consistent with other v4l utilities support the '-d<num>' option
for num >= 10.
Signed-off-by: Hans Verkuil <hans.verkuil@cisco.com>
Konke Radlow [Tue, 4 Jun 2013 18:15:03 +0000 (18:15 +0000)]
libv4l2rds.c: moving functions to get rid of declarations
Signed-off-by: Konke Radlow <koradlow@gmail.com>
Signed-off-by: Hans Verkuil <hans.verkuil@cisco.com>
Konke Radlow [Tue, 4 Jun 2013 18:15:02 +0000 (18:15 +0000)]
rds-ctl: support RDS-EON and TMC-tuning info
- added functionality to print RDS-EON information
- added functionality to print RDS-TMC tuning information
- clarify option description, change
trigger condition for printing TMC Tuning information
Signed-off-by: Konke Radlow <koradlow@gmail.com>
Signed-off-by: Hans Verkuil <hans.verkuil@cisco.com>
Konke Radlow [Tue, 4 Jun 2013 18:15:01 +0000 (18:15 +0000)]
libv4l2rds: support RDS-EON and TMC-tuning info
- added support to decode RDS-EON information
- added support to decode RDS-TMC tuning information
- fixing compiler warnings due to missing pointer dereferencing
Signed-off-by: Konke Radlow <koradlow@gmail.com>
Signed-off-by: Hans Verkuil <hans.verkuil@cisco.com>
Hans Verkuil [Sun, 2 Jun 2013 11:32:01 +0000 (13:32 +0200)]
rds-ctl: fix percentage handling.
The block_cnt can be 0, and in that case the percentage becomes -nan.
Fix this.
Signed-off-by: Hans Verkuil <hans.verkuil@cisco.com>
Hans Verkuil [Fri, 31 May 2013 14:54:01 +0000 (16:54 +0200)]
v4l2-compliance: check that g_register fills the size field.
Signed-off-by: Hans Verkuil <hans.verkuil@cisco.com>
Hans Verkuil [Fri, 31 May 2013 14:46:23 +0000 (16:46 +0200)]
v4l2-compliance: improve overlay clipping tests.
Signed-off-by: Hans Verkuil <hans.verkuil@cisco.com>
Hans Verkuil [Fri, 31 May 2013 14:20:51 +0000 (16:20 +0200)]
v4l2-compliance: refactor code to create invalid formats.
In order to test the format ioctls we fill v4l2_format with invalid
values. But for the overlay buffer types the clips and bitmap fields
must be NULL. This was done in one place, but not in the others.
Make a function that does this and call it in all places where it
is needed.
Signed-off-by: Hans Verkuil <hans.verkuil@cisco.com>
Diego Viola [Sun, 26 May 2013 08:15:51 +0000 (04:15 -0400)]
Fix spelling of Qt in .desktop file (typo)
Proper spelling of Qt is Qt, not QT. "QT" is often confused with
QuickTime, here is a minor patch to fix this issue in the .desktop file.
Signed-off-by: Diego Viola <diego.viola@gmail.com>
Signed-off-by: Hans Verkuil <hans.verkuil@cisco.com>
Mauro Carvalho Chehab [Sat, 27 Apr 2013 15:37:00 +0000 (12:37 -0300)]
dvbv5-zap: add an option to not filter pids
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
Mauro Carvalho Chehab [Sat, 27 Apr 2013 15:02:46 +0000 (12:02 -0300)]
dvbv5-zap: sort the params by the key letter
That helps to see what letters are already used.
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
Hans Verkuil [Sat, 20 Apr 2013 21:33:18 +0000 (23:33 +0200)]
v4l2-dbg: remove unused source.
v4l2-chipids.cpp is no longer used, delete the source.
Signed-off-by: Hans Verkuil <hans.verkuil@cisco.com>
Gregor Jasny [Sat, 20 Apr 2013 20:54:31 +0000 (22:54 +0200)]
Prepare for 0.9.5 release
Signed-off-by: Gregor Jasny <gjasny@googlemail.com>
Gregor Jasny [Sat, 20 Apr 2013 20:48:22 +0000 (22:48 +0200)]
contrib: Add missing dist files for ioctl-test
Signed-off-by: Gregor Jasny <gjasny@googlemail.com>
Gregor Jasny [Sat, 20 Apr 2013 19:43:20 +0000 (21:43 +0200)]
Prepare for 0.9.4 release
Signed-off-by: Gregor Jasny <gjasny@googlemail.com>
Gregor Jasny [Sat, 20 Apr 2013 19:57:52 +0000 (21:57 +0200)]
contrib: Add missing files to distribution list
Signed-off-by: Gregor Jasny <gjasny@googlemail.com>
Hans Verkuil [Wed, 17 Apr 2013 07:04:49 +0000 (09:04 +0200)]
v4l2-ctl: introduce buffers class to simplify the streaming code
There is a lot of shared code between the capture/output/m2m paths
and creating a buffers class to encapsulate that simplifies the
code.
Signed-off-by: Hans Verkuil <hans.verkuil@cisco.com>
Hans Verkuil [Mon, 15 Apr 2013 06:40:06 +0000 (08:40 +0200)]
v4l2-dbg: replace CHIP_IDENT with CHIP_INFO.
The new VIDIOC_DBG_G_CHIP_INFO ioctl is more powerful than the old CHIP_IDENT
ioctl. In particular, the v4l2-chip-ident.h header parsing is no longer needed
to use it. Since that header will be removed in the near future that simplifies
matters.
Signed-off-by: Hans Verkuil <hans.verkuil@cisco.com>
Hans Verkuil [Mon, 15 Apr 2013 06:19:26 +0000 (08:19 +0200)]
v4l-utils: sync with latest kernel.
Signed-off-by: Hans Verkuil <hans.verkuil@cisco.com>
Mauro Carvalho Chehab [Sun, 14 Apr 2013 15:31:57 +0000 (12:31 -0300)]
dvbv5-zap: print elapsed time when buffer overflow happens
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
Mauro Carvalho Chehab [Sun, 14 Apr 2013 15:19:38 +0000 (12:19 -0300)]
dvbv5-zap: better represent low_traffic data in monitor mode
Some PIDs have a very low traffic and may disappear as time
goes by:
1255 0.00 p/s 0.0 Kbps 0 KB
1261 0.00 p/s 0.0 Kbps 0 KB
1282 0.00 p/s 0.0 Kbps 0 KB
14dd 0.00 p/s 0.0 Kbps 0 KB
That makes the output weird. So, instead, show them as:
11d2 0.01 p/s 0.0 Kbps 188 B
1279 0.01 p/s 0.0 Kbps 188 B
1343 0.01 p/s 0.0 Kbps 188 B
1368 0.01 p/s 0.0 Kbps 188 B
1485 0.01 p/s 0.0 Kbps 188 B
171c 0.01 p/s 0.0 Kbps 188 B
19b9 0.01 p/s 0.0 Kbps 188 B
1b38 0.01 p/s 0.0 Kbps 188 B
1bc2 0.01 p/s 0.0 Kbps 188 B
1ca0 0.01 p/s 0.0 Kbps 188 B
1cf3 0.01 p/s 0.0 Kbps 188 B
However, as most of those are spurious traffic, by default,
only show if packets per second is 1 or bigger.
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
Mauro Carvalho Chehab [Sun, 14 Apr 2013 15:18:58 +0000 (12:18 -0300)]
dvbv5-zap: don't stop monitoring for buffer underrun
If buffer underrun is detected, a transitory EOVERFLOW is
returned:
read: Value too large for defined data type
Don't stop monitoring due to that.
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
Mauro Carvalho Chehab [Sun, 14 Apr 2013 14:10:32 +0000 (11:10 -0300)]
dvbv5-zap: Better output at monitor mode
Improve output while on monitor mode, to show data as:
PID FREQ SPEED TOTAL
0000 2.00 p/s 2.9 Kbps 8 KB
...
TOT 25530.54 p/s 37498.0 Kbps 98465 KB
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
Mauro Carvalho Chehab [Sun, 14 Apr 2013 13:49:41 +0000 (10:49 -0300)]
dvbv5-zap: output stats to stdout on monitor mode
On monitor mode, sometimes it is interesting to redirect its
output to a file. By having both monitor prints and status prints
at stdout, the output looks better.
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
Hans Verkuil [Sun, 14 Apr 2013 12:53:05 +0000 (14:53 +0200)]
v4l2-ctl: find_pixel_format couldn't handle output streams.
find_pixel_format only worked for capture streams, not output.
This broke the --set-fmt-video-out option.
Signed-off-by: Hans Verkuil <hans.verkuil@cisco.com>
Hans Verkuil [Thu, 11 Apr 2013 09:08:35 +0000 (11:08 +0200)]
v4l2-ctl: poll/non-poll behavior fixes
The 'v4l2-ctl: add eos and decoder-stop support.' commit broke non-poll
streaming. This is now fixed.
In addition, the eos handling didn't work either in the non-poll case and
we still want to get hold of the eos event. So poll the exception state
even in the non-poll case.
Finally, instead of using + and - to denote capture and output frames use
'<' and '>' respectively. This matches the semantics of < and > in sh.
And I also find it easier to read.
Signed-off-by: Hans Verkuil <hans.verkuil@cisco.com>
Hans Verkuil [Thu, 11 Apr 2013 08:35:28 +0000 (10:35 +0200)]
v4l2-ctl: small improvements.
do_handle_cap doesn't use buffer_lengths, so drop that argument.
Add static to internal functions where applicable.
Signed-off-by: Hans Verkuil <hans.verkuil@cisco.com>
Hans Verkuil [Thu, 11 Apr 2013 08:23:45 +0000 (10:23 +0200)]
v4l2-ctl: split off test pattern generation in a separate source.
A lot more test patterns are planned, so having this in a separate source
simplified development.
Signed-off-by: Hans Verkuil <hans.verkuil@cisco.com>
Hans Verkuil [Thu, 11 Apr 2013 08:13:14 +0000 (10:13 +0200)]
v4l2-ctl: fix a small bug in the stream-count/skip handling.
The capture part of an m2m streaming case didn't ignore stream_skip as
it should. Fixed this.
Signed-off-by: Hans Verkuil <hans.verkuil@cisco.com>
Hans Verkuil [Thu, 11 Apr 2013 08:07:53 +0000 (10:07 +0200)]
v4l2-ctl: use references instead of pointers.
This makes it a bit more readable.
Signed-off-by: Hans Verkuil <hans.verkuil@cisco.com>
Tzu-Jung Lee [Wed, 10 Apr 2013 10:57:06 +0000 (18:57 +0800)]
v4l2-ctl: initial attempt to support M2M device streaming
Signed-off-by: Tzu-Jung Lee <tjlee@ambarella.com>
Signed-off-by: Hans Verkuil <hans.verkuil@cisco.com>
Tzu-Jung Lee [Wed, 10 Apr 2013 10:57:05 +0000 (18:57 +0800)]
v4l2-ctl: break down the streaming_set()
This patch breaks down the streaming_set() into smaller
ones, which can be resued for supporting m2m devices.
Further cleanup or consolidation can be applied with
separate patches, since this one tries not to modify
logics.
Signed-off-by: Tzu-Jung Lee <tjlee@ambarella.com>
Signed-off-by: Hans Verkuil <hans.verkuil@cisco.com>
Hans Verkuil [Tue, 9 Apr 2013 06:46:49 +0000 (08:46 +0200)]
v4l2-ctl: add eos and decoder-stop support.
When dealing with compressed streams it is useful if the capture streaming
code would check for V4L2_EVENT_EOS and if the output streaming side would
be able to use a STOP DECODER command before calling STREAMOFF to give the
decoder the time to flush any pending data.
This patch adds support for that.
Thanks to Tzu-Jung Lee for testing this.
Signed-off-by: Hans Verkuil <hans.verkuil@cisco.com>
Hans Verkuil [Mon, 8 Apr 2013 19:04:23 +0000 (21:04 +0200)]
v4l2-compliance: really fix the RDS and V4L2_CAP_READWRITE tests
Signed-off-by: Hans Verkuil <hans.verkuil@cisco.com>
Hans Verkuil [Mon, 8 Apr 2013 10:40:51 +0000 (12:40 +0200)]
v4l2-compliance: fix RDS + CAP_READWRITE tests.
The READWRITE capability should only be set if RDS_BLOCK_IO is set.
Signed-off-by: Hans Verkuil <hans.verkuil@cisco.com>
Hans Verkuil [Mon, 8 Apr 2013 10:31:55 +0000 (12:31 +0200)]
v4l2-ctl: set bytesused for streaming to the output.
For output stream I/O bytesused should be filled in.
Note that v4l2-ctl only supports uncompressed formats for output streaming
or compressed streams where each buffer is fully filled up (i.e. bytesused
== buffer length).
Signed-off-by: Hans Verkuil <hans.verkuil@cisco.com>
André Roth [Sat, 6 Apr 2013 16:36:22 +0000 (18:36 +0200)]
libdvbv5: fix anonymous unions on older gcc
Hans Verkuil [Wed, 3 Apr 2013 06:31:50 +0000 (08:31 +0200)]
v4l2-compliance: remove ChipIdent tests
These tests are no longer necessary: VIDIOC_DBG_G_CHIP_IDENT is deprecated
and is replaced by VIDIOC_DBG_G_CHIP_NAME, which is a core ioctl that is
always available.
Signed-off-by: Hans Verkuil <hans.verkuil@cisco.com>
Hans Verkuil [Tue, 2 Apr 2013 09:19:47 +0000 (11:19 +0200)]
v4l-utils: sync-with-kernel.
Update to the latest kernel headers.
Signed-off-by: Hans Verkuil <hans.verkuil@cisco.com>
Hans Verkuil [Tue, 26 Mar 2013 08:25:17 +0000 (09:25 +0100)]
v4l-utils: sync to the latest kernel.
Signed-off-by: Hans Verkuil <hans.verkuil@cisco.com>
Hans Verkuil [Mon, 25 Mar 2013 17:09:57 +0000 (18:09 +0100)]
v4l-utils: sync with kernel, remove all references to the DV_PRESET API.
The DV_PRESET API has been removed from the kernel, so remove all references
to that API.
It has been superceded by the DV TIMINGS API.
Signed-off-by: Hans Verkuil <hans.verkuil@cisco.com>
Hans Verkuil [Wed, 20 Mar 2013 18:24:30 +0000 (19:24 +0100)]
v4l2-ctl: add the --concise option.
With the concise option you can get a more concise overview of some of the
lists. Implemented this for list-dv-timings and list-standards.
Signed-off-by: Hans Verkuil <hans.verkuil@cisco.com>
Hans Verkuil [Wed, 20 Mar 2013 17:59:31 +0000 (18:59 +0100)]
qv4l2/v4l2-ctl: fix rounding error in field-per-second calculation
The total height was divided by 2, since height is odd we should make sure the
height is a double, not unsigned or you get a rounding error.
Signed-off-by: Hans Verkuil <hans.verkuil@cisco.com>
Hans Verkuil [Wed, 20 Mar 2013 17:11:22 +0000 (18:11 +0100)]
qv4l2/v4l2-ctl: fix the fields-per-second calculation
The calculation did not take the vertical blanking of the second field
into account, which is typically one line longer. So that will slightly
affect the fields-per-second value.
Signed-off-by: Hans Verkuil <hans.verkuil@cisco.com>
Jean Delvare [Wed, 20 Mar 2013 14:08:36 +0000 (11:08 -0300)]
dvbv5-zap: Fix build warning
dvbv5-zap.c: In function "do_traffic_monitor":
dvbv5-zap.c:582:12: warning: format "%d" expects argument of type "int",
but argument 5 has type "long long unsigned int" [-Wformat]
Signed-off-by: Jean Delvare <khali@linux-fr.org>
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
Hans Verkuil [Wed, 20 Mar 2013 08:15:57 +0000 (09:15 +0100)]
v4l2-ctl: fix the field-per-second calculation
The height in the bt_timings struct is the height of the full frame,
so for the fields-per-second calculation the height should be divided
by two.
Signed-off-by: Hans Verkuil <hans.verkuil@cisco.com>
Hans Verkuil [Wed, 20 Mar 2013 08:15:07 +0000 (09:15 +0100)]
qv4l2: fix the way interlaced formats are displayed
The timings string shows frames instead of fields, which is non-standard.
Switch to fields for interlaced formats.
Signed-off-by: Hans Verkuil <hans.verkuil@cisco.com>
Mauro Carvalho Chehab [Tue, 19 Mar 2013 21:28:43 +0000 (18:28 -0300)]
dvbv5-zap: add traffic monitor support
Add some code to do traffic monitor. The code there came from
dvb-apps dvbtraffic.c file, released under public domain.
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
Hans Verkuil [Tue, 19 Mar 2013 11:50:06 +0000 (12:50 +0100)]
v4l2-ctl: don't produce verbose output for QBUF while streaming.
You are not interested in the verbose logging of QBUF while streaming.
It's not done for DQBUF, so silence QBUF as well.
Signed-off-by: Hans Verkuil <hans.verkuil@cisco.com>
Hans Verkuil [Tue, 19 Mar 2013 11:45:56 +0000 (12:45 +0100)]
v4l2-compliance: fix s_std test: ENODATA is a valid error code.
The S_STD test failed because the driver returned ENODATA. This is a
valid error code, but the test didn't check for it.
Signed-off-by: Hans Verkuil <hans.verkuil@cisco.com>
Hans Verkuil [Tue, 12 Mar 2013 07:42:13 +0000 (08:42 +0100)]
libv4l2/log.c: update list of ioctls.
This list was out of date, update it.
Also update a very small log printf (missing colon).
Signed-off-by: Hans Verkuil <hans.verkuil@cisco.com>
Hans Verkuil [Tue, 12 Mar 2013 06:40:24 +0000 (07:40 +0100)]
qv4l2: handle control range updates.
Signed-off-by: Hans Verkuil <hans.verkuil@cisco.com>
Hans Verkuil [Mon, 11 Mar 2013 21:55:11 +0000 (22:55 +0100)]
qv4l2: do nothing if querystd does not detect a standard.
Signed-off-by: Hans Verkuil <hans.verkuil@cisco.com>
Hans Verkuil [Mon, 11 Mar 2013 20:45:14 +0000 (21:45 +0100)]
v4l2-compliance: improve tests when TRY/S_FMT returns EINVAL.
If a wrong pixelformat is specified, then TRY/S_FMT can return EINVAL.
However, part of the test sequence was skipped in that case, which led
to missed errors.
Now get a valid pixelformat and put that in and try again. That way all
other tests are executed as well.
Signed-off-by: Hans Verkuil <hans.verkuil@cisco.com>
Hans Verkuil [Sun, 10 Mar 2013 10:40:00 +0000 (11:40 +0100)]
v4l2-compliance: missing check on the frameinterval
A wrong discrete frameinterval as returned by ENUM_FRAMEINTERVALS
wasn't detected.
Signed-off-by: Hans Verkuil <hans.verkuil@cisco.com>
Hans Verkuil [Sun, 10 Mar 2013 09:31:50 +0000 (10:31 +0100)]
v4l2-compliance: add support for the new 'COPY' timestamp type.
Signed-off-by: Hans Verkuil <hans.verkuil@cisco.com>
Hans Verkuil [Sun, 10 Mar 2013 09:29:38 +0000 (10:29 +0100)]
v4l2-ctl: add support for 'COPY' timestamp and ctrl range events.
Signed-off-by: Hans Verkuil <hans.verkuil@cisco.com>
Hans Verkuil [Sun, 10 Mar 2013 09:25:24 +0000 (10:25 +0100)]
v4l-utils: run sync-with-kernel.
Signed-off-by: Hans Verkuil <hans.verkuil@cisco.com>
Hans Verkuil [Sat, 9 Mar 2013 17:59:11 +0000 (18:59 +0100)]
v4l2-* utils: allow -d11 to refer to /dev/video11
Lately I have been working with devices with lots of video nodes. Various
v4l2 utilities allow -dX to refer to /dev/videoX, but 'X' was limited to
a single digit. Extend this to support up to three digits. So -d11 would
refer to /dev/video11 and -d128 to /dev/video128 (that really should be
enough!)
Signed-off-by: Hans Verkuil <hans.verkuil@cisco.com>
Hans Verkuil [Mon, 4 Mar 2013 22:45:09 +0000 (23:45 +0100)]
v4l2-ctl: show buffers or key/P/B-frames when capturing.
Add support in v4l2-ctl to show any captured buffer data if --verbose is
given, or to show whether a key, P or B frame was captured.
Signed-off-by: Hans Verkuil <hans.verkuil@cisco.com>