Age | Commit message (Collapse) | Author |
|
From: Jean-Francois Moine <moinejf@free.fr>
Add support for conversion to RGB24 (before we only support BGR24) based
on a patch by Jean-Francois Moine
Signed-off-by: Jean-Francois Moine <moinejf@free.fr>
Signed-off-by: Hans de Goede <j.w.r.degoede@hhs.nl>
|
|
From: Hans de Goede <j.w.r.degoede@hhs.nl>
Be more relaxed in our checks for mixing read and mmap access, we were
being more strict in this then certain kernel drivers (bttv) making xawtv
unhappy
Signed-off-by: Hans de Goede <j.w.r.degoede@hhs.nl>
|
|
From: Brandon Philips <bphilips@suse.de>
Part of the copyright headers refered GPL instead of LGPL due to a copy
and paste error (Brandon Philips)
Signed-off-by: Brandon Philips <bphilips@suse.de>
Signed-off-by: Hans de Goede <j.w.r.degoede@hhs.nl>
|
|
Signed-off-by: Brandon Philips <bphilips@suse.de>
|
|
Signed-off-by: Brandon Philips <bphilips@suse.de>
|
|
flags
From: Hans de Goede <j.w.r.degoede@hhs.nl>
libv4l: query buffer: indicate the mapping state of our (fake) buffer in the flags
Signed-off-by: Hans de Goede <j.w.r.degoede@hhs.nl>
|
|
From: Hans de Goede <j.w.r.degoede@hhs.nl>
libv4l: proper accounting of mmap count of fake mmap buffers
Signed-off-by: Hans de Goede <j.w.r.degoede@hhs.nl>
|
|
From: Hans de Goede <j.w.r.degoede@hhs.nl>
libv4l: intercept get capabilites and report read capability
Signed-off-by: Hans de Goede <j.w.r.degoede@hhs.nl>
|
|
From: Hans de Goede <j.w.r.degoede@hhs.nl>
libv4l: check capabilities for streaming
Signed-off-by: Hans de Goede <j.w.r.degoede@hhs.nl>
|
|
From: Hans de Goede <j.w.r.degoede@hhs.nl>
libv4l: add pkg-config (.pc) files for easier detection if libv4l is available
Signed-off-by: Hans de Goede <j.w.r.degoede@hhs.nl>
|
|
From: Gregor Jasny <jasny@vidsoft.de>
libv4l2: work around wrong REQUEST_BUFFERS ioctl return code
Signed-off-by: Gregor Jasny <jasny@vidsoft.de>
Signed-off-by: Hans de Goede <j.w.r.degoede@hhs.nl>
Acked-by: Mauro Carvalho Chehab <mchehab@infradead.org>
|
|
From: Hans de Goede <j.w.r.degoede@hhs.nl>
* Add spca505/6 and spca508 cam specific formats (YUYV per line variations)
Signed-off-by: Hans de Goede <j.w.r.degoede@hhs.nl>
|
|
|
|
From: Hans de Goede <j.w.r.degoede@hhs.nl>
* Make JPEG decoding more robust
Signed-off-by: Hans de Goede <j.w.r.degoede@hhs.nl>
|
|
From: Hans de Goede <j.w.r.degoede@hhs.nl>
* The mmap64 support in 0.3.3, has caused a bug in libv4l1 when running on
32 bit systems (who uses those now a days?), this bug caused v4l1
compatibility to not work at all, this release fixes this
Signed-off-by: Hans de Goede <j.w.r.degoede@hhs.nl>
Signed-off-by: Thierry MERLE <thierry.merle@free.fr>
|
|
From: Hans de Goede <j.w.r.degoede@hhs.nl>
* The mmap64 support in 0.3.3, has caused a bug in libv4l1 when running on
32 bit systems (who uses those now a days?), this bug caused v4l1
compatibility to not work at all, this release fixes this
* Some apps (xawtv, kopete) use an ioctl wrapper internally for various
reasons. This wrappers request argument is an int, but the real ioctl's
request argument is an unsigned long. Passing the VIDIOC_xxx defines through
to the wrapper, and then to the real ioctl, causes the request to get sign
extended on 64 bit args. The kernel seems to ignore the upper 32 bits,
causing the sign extension to not make a difference. libv4l now also
ignores the upper 32 bits of the libv4lx_ioctl request argument on 64 bit
archs
* Add a bugfix patch for kopete in the appl-patches dir, currently it assumes
that it got the width and height it asked for when doing a S_FMT, which is a
wrong assumption
Signed-off-by: Hans de Goede <j.w.r.degoede@hhs.nl>
Signed-off-by: Thierry MERLE <thierry.merle@free.fr>
|
|
From: Hans de Goede <j.w.r.degoede@hhs.nl>
* Add open64 and mmap64 wrappers to the LD_PRELOAD wrapper libs, so that
they also work for applications compiled with FILE_OFFSET_BITS=64, this
fixes using them with v4l-info
* While looking at xawtv in general, found a few bugs in xawtv itself, added
a patch to fix those to the appl-patches dir
* Talking about the appl-patches dir, restore that as it accidentally got
dropped from 0.3.2
* Be more verbose in various places when it comes to logging (esp errors)
* Change v4lconvert_enum_fmt code a bit, so that it is easier to add more
supported destination formats to libv4lconvert
* Don't return -EINVAL from try_fmt when we cannot convert because the cam
doesn't have any formats we know. Instead just return as format whatever the
cam returns from try_fmt, this new behavior is compliant with the v4l2
api as documented
Signed-off-by: Hans de Goede <j.w.r.degoede@hhs.nl>
Signed-off-by: Thierry MERLE <thierry.merle@free.fr>
|
|
From: Hans de Goede <j.w.r.degoede@hhs.nl>
The current mercurial has been released on my homepage as tarbal release 0.3.2, the only changes from the current mercurial + all patches already send (including the asm patch from Gregor Jasny) is updating of the ChangeLog file and updating the version in the Makefile, these changes are included in the attached patch.
Signed-off-by: Hans de Goede <j.w.r.degoede@hhs.nl>
Signed-off-by: Thierry MERLE <thierry.merle@free.fr>
|
|
From: Gregor Jasny <jasny@vidsoft.de>
This patch fixes the input constraint for the sar instruction. It allows only an
immediate or cl as shift width.
Signed-off-by: Gregor Jasny <jasny@vidsoft.de>
Signed-off-by: Thierry MERLE <thierry.merle@free.fr>
|
|
From: Hans de Goede <j.w.r.degoede@hhs.nl>
Fix all compiler warnings in libv4l
Signed-off-by: Hans de Goede <j.w.r.degoede@hhs.nl>
Signed-off-by: Thierry MERLE <thierry.merle@free.fr>
|
|
From: Hans de Goede <j.w.r.degoede@hhs.nl>
Makefile improvements:
* Split DESTDIR into DESTDIR and PREFIX as used in most makefiles out there
* Add LIBDIR variable to allow installation in <prefix>/lib64 for example
* Install the wrappers in <libdir>/libv4l instead of directly under libdir,
as they are not libraries meant for linking
* preserve timestamps of header files when installing them
Signed-off-by: Hans de Goede <j.w.r.degoede@hhs.nl>
Signed-off-by: Thierry MERLE <thierry.merle@free.fr>
|
|
From: Hans de Goede <j.w.r.degoede@hhs.nl>
Some documentation updates to bring the documentation up2date with the latest
changes.
Signed-off-by: Hans de Goede <j.w.r.degoede@hhs.nl>
Signed-off-by: Thierry MERLE <thierry.merle@free.fr>
|
|
From: Hans de Goede <j.w.r.degoede@hhs.nl>
Add comments to pac207.c about what todays experiments have teached us about
the pac207 compression.
Signed-off-by: Hans de Goede <j.w.r.degoede@hhs.nl>
Signed-off-by: Thierry MERLE <thierry.merle@free.fr>
|
|
From: Hans de Goede <j.w.r.degoede@hhs.nl>
Add license change (GPL -> LGPL) permission notice, and don't claim copyright
over code I didn't write, instead add copyright header of the original author
(oops)
Signed-off-by: Hans de Goede <j.w.r.degoede@hhs.nl>
Signed-off-by: Thierry MERLE <thierry.merle@free.fr>
|
|
by pac207 based cams
From: Hans de Goede <j.w.r.degoede@hhs.nl>
This patch adds support to libv4l for the compressed bayer format emitted by
pac207 based cams.
Signed-off-by: Hans de Goede <j.w.r.degoede@hhs.nl>
Signed-off-by: Thierry MERLE <thierry.merle@free.fr>
|
|
From: Hans de Goede <j.w.r.degoede@hhs.nl>
Add support for sonix (sn9c10x) bayer compression to libv4lconvert
Signed-off-by: Hans de Goede <j.w.r.degoede@hhs.nl>
Signed-off-by: Thierry MERLE <thierry.merle@free.fr>
|
|
From: Hans de Goede <j.w.r.degoede@hhs.nl>
This patch syncs (updates) the libv4l in mercurial with (to) the latest 0.3.1
release.
- Only serialize V4L2_BUF_TYPE_VIDEO_CAPTURE type ioctls
- Do not return an uninitialized variable as result code for GPICT
(fixes vlc, but see below)
- Add an apps-patches directory which includes:
* vlc-0.8.6-libv4l1.patch, modify vlc's v4l1 plugin to directly call into
libv4l1, in the end we want all apps todo this as its better then
LD_PRELOAD tricks, but for vlc this is needed as vlc's plugin system
causes LD_PRELOAD to not work on symbols in the plugins
* camorama-0.19-fixes.patch, small bugfixes to camorama's v4l1 support,
this patch only fixes _real_ bugs in camorama and does not change it to
work with v4l1compat. Although it does work better with these bugs fixed
:) With this patch and LD_PRELOAD=<path>/v4l1compat.so it works
flawless.
Signed-off-by: Hans de Goede <j.w.r.degoede@hhs.nl>
Signed-off-by: Thierry MERLE <thierry.merle@free.fr>
|
|
From: Hans de Goede <j.w.r.degoede at hhs.nl>
Global files to make the libv4l compile, install and exported into a tar.gz file
Signed-off-by: Hans de Goede <j.w.r.degoede at hhs.nl>
Signed-off-by: Thierry MERLE <thierry.merle@free.fr>
|
|
From: Hans de Goede <j.w.r.degoede at hhs.nl>
libv4l1 is the base of the v4l1compat.so wrapper lib, which is a .so
which can be LD_PRELOAD-ed and the overrules the libc's open/close/etc,
and when opening /dev/videoX calls v4l1_open.
Signed-off-by: Hans de Goede <j.w.r.degoede at hhs.nl>
Signed-off-by: Thierry MERLE <thierry.merle@free.fr>
|
|
From: Hans de Goede <j.w.r.degoede at hhs.nl>
libv4l2 offers v4l2_ prefixed versions of
open/close/etc. The API is 100% the same as directly opening /dev/videoX
using regular open/close/etc, the big difference is that format conversion
is done if necessary when capturing. That is if you (try to) set a capture
format which is not supported by the cam, but is supported by libv4lconvert,
then the try_fmt / set_fmt will succeed as if the cam supports the format
and on dqbuf / read the data will be converted for you and returned in
the request format.
v4l2convert: open/close/ioctl/mmap/munmap library call wrapper doing format conversion
for v4l2 applications which want to be able to simply capture bgr24 / yuv420
from v4l2 devices with more exotic frame formats.
Signed-off-by: Hans de Goede <j.w.r.degoede at hhs.nl>
Signed-off-by: Thierry MERLE <thierry.merle@free.fr>
|
|
From: Hans de Goede <j.w.r.degoede at hhs.nl>
libv4lconvert: video decompression routines
Signed-off-by: Hans de Goede <j.w.r.degoede at hhs.nl>
Signed-off-by: Thierry MERLE <thierry.merle@free.fr>
|
|
From: Carl Karsten <carl@personnelware.com>
Date: Thu, 28 Feb 2008 15:21:51 -0600
Adding #include <stddef.h> solves a build problem I had on Ubnutu.
v4l-dvb/v4l2-apps/test$ make
CC driver-test.o
In file included from driver-test.c:17:
../lib/v4l2_driver.h:26: error: expected specifier-qualifier-list before 'size_t'
make: *** [driver-test.o] Error 1
Thanks to Carl Karsten <carl@personnelware.com> for pointing this issue.
CC: Carl Karsten <carl@personnelware.com>
Signed-off-by: Mauro Carvalho Chehab <mchehab@infradead.org>
|
|
From: Hans Verkuil <hverkuil@xs4all.nl>
Remove the obsolete VIDIOC_G_MPEGCOMP and VIDIOC_S_MPEGCOMP ioctls from
the V4L2 API as per the removal schedule (October 2007).
Signed-off-by: Hans Verkuil <hverkuil@xs4all.nl>
|
|
From: Mauro Carvalho Chehab <mchehab@infradead.org>
Signed-off-by: Mauro Carvalho Chehab <mchehab@infradead.org>
|
|
From: David Warman <dwarman@davidwarman.net>
v4l2_driver.c:v4l2_enum_stds: erroneously always returns failure
because errno is +ve, needs inverting for use as a v4l error code.
The diff command actually produced patched for two fixes; I separated
them manually into these two reports.
Signed-off-by: David Warman <dwarman@davidwarman.net>
Signed-off-by: Mauro Carvalho Chehab <mchehab@infradead.org>
|
|
From: Mauro Carvalho Chehab <mchehab@infradead.org>
Signed-off-by: Mauro Carvalho Chehab <mchehab@infradead.org>
|
|
From: Mauro Carvalho Chehab <mchehab@infradead.org>
Signed-off-by: Mauro Carvalho Chehab <mchehab@infradead.org>
|
|
From: Mauro Carvalho Chehab <mchehab@infradead.org>
bttv driver presents a non-compliance: When S_FMT is called, pix.sizeimage reports
a different size than when QUERYBUF is called, for the same S_FMT.
The current code can work fine with this non-compliance, but an error should be
reported.
Those are the reports from bttv driver, by running driver-test:
FMT SET: 640x480, fourcc=YVU9, 640 bytes/line, 345600 bytes/frame, colorspace=0x00000000
<snip/>
QUERYBUF: 00:00:00.00000000 index=0, type=video-cap, bytesused=0, flags=0x00000000,
field=any, sequence=0, memory=mmap, offset=0x00000000, length=348160
TIMECODE: 00:00:00 type=0, flags=0x00000000, frames=0, userbits=0x00000000
QUERYBUF: ERROR: VIDIOC_S_FMT said buffer should have 345600 size, but received 348160 from QUERYBUF!
Signed-off-by: Mauro Carvalho Chehab <mchehab@infradead.org>
|
|
From: Mauro Carvalho Chehab <mchehab@infradead.org>
Signed-off-by: Mauro Carvalho Chehab <mchehab@infradead.org>
|
|
From: Mauro Carvalho Chehab <mchehab@infradead.org>
Add capabilities to the library and to the driver to receive video streams.
Library will use a callback, called every time a new buffer is reported by
dqbuf.
Signed-off-by: Mauro Carvalho Chehab <mchehab@infradead.org>
|
|
From: Mauro Carvalho Chehab <mchehab@infradead.org>
There is a non-compliance on video-buf implementation about calling REQBUFS
with count=0. According with API, this should free all buffers, but video-buf
returns, instead, -EINVAL.
Signed-off-by: Mauro Carvalho Chehab <mchehab@infradead.org>
|
|
From: Mauro Carvalho Chehab <mchehab@infradead.org>
Using the same approach as capture_exampe.c
Signed-off-by: Mauro Carvalho Chehab <mchehab@infradead.org>
|
|
From: Mauro Carvalho Chehab <mchehab@infradead.org>
Signed-off-by: Mauro Carvalho Chehab <mchehab@infradead.org>
|
|
From: Mauro Carvalho Chehab <mchehab@infradead.org>
Signed-off-by: Mauro Carvalho Chehab <mchehab@infradead.org>
|
|
From: Mauro Carvalho Chehab <mchehab@infradead.org>
Signed-off-by: Mauro Carvalho Chehab <mchehab@infradead.org>
|
|
From: Mauro Carvalho Chehab <mchehab@infradead.org>
This method requests buffers and maps it into a mmaped memory.
Still under testing. Still need other stuff to make stream to work.
Signed-off-by: Mauro Carvalho Chehab <mchehab@infradead.org>
|
|
From: Mauro Carvalho Chehab <mchehab@infradead.org>
- Better presentation of version
- Make v4l2_enum_fmt generic for all types
- Added v4l2_gettryset_fmt_cap to GET/TRY/SET fmt on capture
- driver-test now sets a video format
Signed-off-by: Mauro Carvalho Chehab <mchehab@infradead.org>
|
|
From: Mauro Carvalho Chehab <mchehab@infradead.org>
- Created v4l2_driver.c with some basic functions to:
. Enum STD, INPUT and FORMAT
. GET/SET STD and INPUT
. GET PARAM
- ioctl-test reverted back to a simple "brute-force" all ioctl call;
- driver-test created to:
. test V4L2 kernel drivers;
. implement all functions provided by v4l2 library
- driver-test should grow following the changes at v4l2-apps/lib
Signed-off-by: Mauro Carvalho Chehab <mchehab@infradead.org>
|
|
From: Hans Verkuil <hverkuil@xs4all.nl>
Signed-off-by: Hans Verkuil <hverkuil@xs4all.nl>
|
|
From: Hans Verkuil <hverkuil@xs4all.nl>
Fix the copyright lines in the v4l2 library. Also explicitly add
the Nathan's SOB for frequency.c.
Signed-off-by: Hans Verkuil <hverkuil@xs4all.nl>
Signed-off-by: Nathan Laredo <laredo@gnu.org>
|