Commit graph

1763 commits

Author SHA1 Message Date
Guillaume Desmottes
f7e280bf0d v4l2: fix division by 0 for complex video formats
So complex video formats have 0 as pstride. Don't try to divide the
stride in such cases.

https://bugzilla.gnome.org/show_bug.cgi?id=792596
2018-01-17 13:32:58 -05:00
Guillaume Desmottes
3bbfd15047 v4l2: display stride and width values if stride is too small
https://bugzilla.gnome.org/show_bug.cgi?id=792596
2018-01-17 13:32:58 -05:00
Peter Seiderer
d2f9040e1d v4l2videodec: add property set/get PROP_CAPTURE_IO_MODE error handling
https://bugzilla.gnome.org/show_bug.cgi?id=791841
2018-01-12 09:21:18 -05:00
Peter Seiderer
2cd772b97a v4l2videodec: fold property set/get PROP_OUTPUT_IO_MODE case into default
https://bugzilla.gnome.org/show_bug.cgi?id=791841
2018-01-12 09:21:18 -05:00
Peter Seiderer
f579ece0fb v4l2videoenc: add property set/get PROP_CAPTURE_IO_MODE error handling
https://bugzilla.gnome.org/show_bug.cgi?id=791841
2018-01-12 09:21:18 -05:00
Peter Seiderer
5c0cf56193 v4l2videoenc: fold property set/get PROP_OUTPUT_IO_MODE case into default
https://bugzilla.gnome.org/show_bug.cgi?id=791841
2018-01-12 09:21:18 -05:00
Peter Seiderer
189822350b v4l2videoenc: fix capture-io-mode property get
https://bugzilla.gnome.org/show_bug.cgi?id=791841
2018-01-12 09:21:18 -05:00
Nicolas Dufresne
21b01acef1 v4l2src: Maintain downstream caps order
The g_list_insert_sorted() will behave like prepend when the compare
function returns 0. In our case, we want to maintain the order hence
append. This fixes this issue and improve the sorting algorithm to make
a 10x10 prefered over 10x200 with a preference of 10x8 (and similar
cases which was badly handled). This fixes generally fixes issue were a
sub-optimal format / size is picked.

https://bugzilla.gnome.org/show_bug.cgi?id=792435
2018-01-11 17:51:13 -05:00
Nicolas Dufresne
b3afa0dc1d v4l2videoenc: Also re-enabled paused task
When we only run _finish(), the task is never stopped externally,
instead it's only paused from the inside. We still want to restart
it in this case.
2018-01-11 11:27:38 -05:00
Nicolas Dufresne
d90b6ec459 v4l2videoenc: Call stop on object before renegotiation
Otherwise renegotiation fails as we are still streaming.

https://bugzilla.gnome.org/show_bug.cgi?id=791338
2018-01-08 17:20:52 -05:00
Nicolas Dufresne
a46756bd56 v4l2videoenc: Remove dead code
gst_v4l2_object_stop() will free and nullify the pool, so the
following if will never be true.

https://bugzilla.gnome.org/show_bug.cgi?id=791338
2018-01-08 17:20:52 -05:00
Nicolas Dufresne
4aa5298f56 v4l2videoenc: Delay capture pool activation
This is support CODA driver which prevents setting the output format if
the capture is streaming.

https://bugzilla.gnome.org/show_bug.cgi?id=791338
2018-01-08 17:20:52 -05:00
Nicolas Dufresne
97985a335c v4l2videodec: Add dynamic resolution change support
This implements a "big hammer" reallocation method. We effectively
drain and stop both side of the decoder and restart. This though is
the most generic method. This change should enable on most drivers
adaptive streaming.

https://bugzilla.gnome.org/show_bug.cgi?id=752962
2018-01-08 17:16:22 -05:00
Nicolas Dufresne
4632f33334 v4l2deviceprovider: Don't do slow probes
This is problematic in the current design at it seriously slow down
startup of applications. As of now, no known application uses the
colorimetry and the interlace-modes for anything (the two fields that
won't be probed). So let's disable it, in the long term we'll try and
find a way to interact with the provider so applicaiton could opt-in
these slow probing methods for more advance configuration.
2017-12-22 10:30:49 -05:00
Nicolas Dufresne
80815c292a v4l2object: Don't redefine mmap64
On Linux, there exist a case where mmap64 is already a define to mmap,
so avoid the redefine warning here.
2017-12-22 10:30:49 -05:00
Ting-Wei Lan
4026211154 v4l2object: Don't use mmap64 if off_t is 64-bit
The difference between mmap and mmap64 is the type of 'offset' argument.
mmap64 always uses a 64-bit interger as offset, while mmap uses off_t,
whose size can vary on different operating systems or architectures.

However, not all operating systems support mmap64. Fortunately, although
FreeBSD only has mmap, its off_t is always 64-bit regardless of
architectures, so we can simply use mmap when sizeof(off_t) == 8.

https://bugzilla.gnome.org/show_bug.cgi?id=791779
2017-12-22 10:30:49 -05:00
Nicolas Dufresne
24f01944ee Revert "v4l2object: Use mmap64 to match libv4l2 signature"
This reverts commit b61bba4848.
2017-12-22 10:30:13 -05:00
Ting-Wei Lan
784348235a v4l2object: Check for mmap64 before using it
mmap64 is not available on FreeBSD.

https://bugzilla.gnome.org/show_bug.cgi?id=791779
2017-12-21 21:58:29 -05:00
Nicolas Dufresne
e575e4eda8 v4l2object: Use a debug object for tracing
This way we can pass the pad name instead of the element for tracing
which helps identifying which v4l2object is used withing M2M element
like decoder, encoder and transform. For the reference, pads are name
<parent-name>:<pad-name>.
2017-12-13 14:42:25 -05:00
Philipp Zabel
5e05cadb17 v4l2videodec: Handle drivers that only round up height
Commit 1f31715c98 ("v4l2videodec: use visible size, not coded size,
for downstream negotiation filter") added support for removing the
padding obtained as the difference between width/height from G_FMT and
visible width/height from G_SELECTION from the probed caps obtained
via TRY_FMT.
This patch fixes the padding removal for drivers that only round up
height, but not width, to the padded frame size. This might happen
because horizontal padding can be handled by line stride (bytesperline),
but there is no such thing as plane stride in the V4L2 API for
single-buffer planar formats.

https://bugzilla.gnome.org/show_bug.cgi?id=791271
2017-12-06 14:01:02 -05:00
Florent Thiéry
a315810640 v4l2object: Fix dmabuf support detection
This resulted in improper selection of dmabuf on unsupported drivers.
The checked ioctl errno was not correct.

https://bugzilla.gnome.org/show_bug.cgi?id=790940
2017-11-29 10:13:44 -05:00
Jan Alexander Steffens (heftig)
7e4bc6d418 v4l2deviceprovider: Ignore touch sensing devices
With GST_V4L2_USE_LIBV4L2=1, my laptop's touchpad shows up as a video
source device in gst-device-monitor, but attempting to stream from it
fails because the device doesn't actually support any video formats.

name  : Synaptics RMI4 Touch Sensor
class : Video/Source
caps  : video/x-raw, format=(string)I420, framerate=(fraction)[ 0/1, 2147483647/1 ], width=(int)0, height=(int)0, interlace-mode=(string)progressive, pixel-aspect-ratio=(fraction)1/1;
        video/x-raw, format=(string)YV12, framerate=(fraction)[ 0/1, 2147483647/1 ], width=(int)0, height=(int)0, interlace-mode=(string)progressive, pixel-aspect-ratio=(fraction)1/1;
        video/x-raw, format=(string)BGR, framerate=(fraction)[ 0/1, 2147483647/1 ], width=(int)0, height=(int)0, interlace-mode=(string)progressive, pixel-aspect-ratio=(fraction)1/1;
        video/x-raw, format=(string)RGB, framerate=(fraction)[ 0/1, 2147483647/1 ], width=(int)0, height=(int)0, interlace-mode=(string)progressive, pixel-aspect-ratio=(fraction)1/1;
properties:
        udev-probed = true
        device.bus_path = /sys/devices/rmi4-00/rmi4-00.fn54/video4linux/v4l-touch0
        sysfs.path = /sys/devices/rmi4-00/rmi4-00.fn54/video4linux/v4l-touch0
        device.subsystem = video4linux
        device.product.name = "Synaptics\ RMI4\ Touch\ Sensor"
        device.capabilities = :capture:
        device.api = v4l2
        device.path = /dev/v4l-touch0
        v4l2.device.driver = rmi4_f54
        v4l2.device.card = "Synaptics\ RMI4\ Touch\ Sensor"
        v4l2.device.bus_info = rmi4:rmi4-00.fn54
        v4l2.device.version = 265480 (0x00040d08)
        v4l2.device.capabilities = 2501902337 (0x95200001)
        v4l2.device.device_caps = 354418689 (0x15200001)
gst-launch-1.0 v4l2src device=/dev/v4l-touch0 ! ...

v4l2-ctl -d /dev/v4l-touch0 --list-formats reports:

ioctl: VIDIOC_ENUM_FMT
        Index       : 0
        Type        : Video Capture
        Pixel Format: 'TD16'
        Name        : 16-bit signed deltas

        Index       : 1
        Type        : Video Capture
        Pixel Format: 'TD08'
        Name        : 8-bit signed deltas

        Index       : 2
        Type        : Video Capture
        Pixel Format: 'TU16'
        Name        : 16-bit unsigned touch data

https://bugzilla.gnome.org/show_bug.cgi?id=789197
2017-11-06 19:34:22 -05:00
Nicolas Dufresne
57d7ae8496 v4l2allocator: Add support for data_offset
In MPLANE mode, the driver may set data_offset, which represent some
padding at the start of the buffer used internally. This portion of the
data need to be skipped, though it is included in bytesused.

This patch removes frame size sanity check as the method used will no
longer work. This check was simply there to help detect broken kernel
drivers. It would be re-implement by estimating the plane size, which is
not totally trivial and may be too much work for a simple debug check.

https://bugzilla.gnome.org/show_bug.cgi?id=733501
2017-11-03 15:47:11 -04:00
Philippe Renon
09af01a088 directsoundsink: simplify how DirecSoundBuffer is cleared
we always want to clear the whole buffer so no need to
start from offset even if the offset is always zero.

https://bugzilla.gnome.org/show_bug.cgi?id=788847
2017-10-12 10:32:34 +03:00
Philippe Renon
24561309f9 directsoundsink: fix comment
https://bugzilla.gnome.org/show_bug.cgi?id=788847
2017-10-12 10:32:34 +03:00
Philippe Renon
73993149ec directsoundsink: don't call set_volume with private scaled volume
use get_volume() instead to get unscaled volume

https://bugzilla.gnome.org/show_bug.cgi?id=788847
2017-10-12 10:32:34 +03:00
Philippe Renon
d519d14ad9 directsoundsink: remove duplicate volume initialization
https://bugzilla.gnome.org/show_bug.cgi?id=788847
2017-10-12 10:32:34 +03:00
Nicolas Dufresne
33e595b41c v4l2videoenc: Fix driver capability dectection
Use the right set of caps when checking if caps intersect. That makes
the check only select the supported devices.
2017-10-07 17:33:41 -04:00
Nicolas Dufresne
cebb5baaa6 v4l2videoenc/dec: Don't leak template caps 2017-10-07 17:33:41 -04:00
Nicolas Dufresne
9ad90f3bf2 v4l2videodec: Protect against null pool in _stop
This may happen if the negotiation fails, as we will have never
created the pools.
2017-10-07 17:33:41 -04:00
Tim-Philipp Müller
97285753fc v4l2: fix build without libv4l
https://bugzilla.gnome.org/show_bug.cgi?id=779466
2017-10-07 15:39:47 +01:00
Enrico Jorns
e9c00c083b v4l2object: auto-detect dmabuf export for V4L2_IO_AUTO on capture side
Issue an invalid VIDIOC_EXPBUF ioctl to the driver to check if the
driver supports dmabuf export. If the driver does not implement the
IOCTL, the error is ENOTTY. Any other error codes mean that the driver
implements VIDIOC_EXPBUF.

https://bugzilla.gnome.org/show_bug.cgi?id=779466
2017-10-05 21:20:19 -04:00
Nicolas Dufresne
fb26575b73 v4l2videodec: Don't leak codec name 2017-09-19 17:55:58 -04:00
Nicolas Dufresne
1f902e2f6e v4l2bufferpool: Don't stop streaming when pool is flushing
The purpose of being able to flush the buffer pool is only to
unlock any blocked operation. Doing streamoff/streamon had the
side effect of turning off and on the camera. As we do a flush_start
/ flush_stop sequence when shutting down, that would cause a really
quick sequence of streamoff/streamon/streamoff/close which was
causing some cameras to stop working.

https://bugzilla.gnome.org/show_bug.cgi?id=783945
2017-09-19 17:33:35 -04:00
Nicolas Dufresne
a08d7cdef5 v4l2src: Ensure all caps a fixated
The code relied on the list compare function to fixate the caps
but if the caps only has one structure, the compare function will
never get called. Capture device for which there is only one
structure in the caps would then get some assertion and later
fail badly.

Instead, fixate before inserting into the list and split the reading
and the fixation of the structures.
2017-09-13 16:05:08 -04:00
Nicolas Dufresne
c56eb21a82 v4l2object: Don't leak the par value 2017-09-13 11:52:09 -04:00
Ian Jamison
85ec6430d9 v4l2object: Handle BT2020 for colorspace and transfer
This was not fully handled in switches and
ub gst_v4l2_object_get_colorspace();

https://bugzilla.gnome.org/show_bug.cgi?id=787313
2017-09-06 11:33:53 -04:00
Ian Jamison
fc5d9c34a7 v4l2object: Fix colorimetry transfer lookup for 4K video
https://bugzilla.gnome.org/show_bug.cgi?id=787160
2017-09-06 11:33:53 -04:00
Nicolas Dufresne
05a5ab0e55 v4l2src: Check if caps have changed after try_fmt
try_fmt will update the caps colorimetry and interlace-mode. Before this
call, those field are missing. The caps equality check was always
failing when a spurious reconfigure event was received.
2017-09-06 11:33:53 -04:00
Cyril Lashkevich
f2372b46bc v4l2bufferpool: Don't mark jpeg frames as deltas
JPEG formats are encoded, but they never have keyframe flag. But in
fact they are keyframes

https://bugzilla.gnome.org/show_bug.cgi?id=785990
2017-08-08 11:09:52 -04:00
Philippe Normand
acb50ce82e osxvideo: rename library according to the plugin name
https://bugzilla.gnome.org/show_bug.cgi?id=785880
2017-08-06 13:26:01 +01:00
Nicolas Dufresne
d8b18f6b8f v4l2bufferpool: Don't drop buffer ref on qbuf
This function no longer take ownership of the buffer.

CID 1414800
2017-08-02 17:18:43 -04:00
Nicolas Dufresne
0e9fe263ca v4l2: Enable VP9 format
This was missing, preventing the encoder and decoder to work
properly. This also adds support for camera that would produce
VP9 (if that exists).
2017-08-02 17:13:55 -04:00
Nicolas Dufresne
a13c24dc8c v4l2: Remove spurious CATEGORY_EXTERN
These have been copy pasted all over the place and are not used anymore.
All object have it's own category now. This fixes build warning since
the VP9 decoder had vp8 category declared.
2017-08-02 12:28:38 -04:00
Nicolas Dufresne
783ac9a9f9 v4l2videoenc: Move the profile/level negotation in the base class
This removes duplicated code across different codec.
2017-08-02 11:28:23 -04:00
Nicolas Dufresne
b0fb95b956 v4l2videoenc: Turn gst_v4l2_is_video_enc into a helper
This reduces the amount of code needed in each codec class.
2017-08-02 09:36:08 -04:00
Nicolas Dufresne
040e25218a v4l2: Add VP8/9 encoder support 2017-08-01 16:01:11 -04:00
Nicolas Dufresne
b61bba4848 v4l2object: Use mmap64 to match libv4l2 signature
https://bugzilla.gnome.org/show_bug.cgi?id=785628
2017-08-01 15:05:35 -04:00
Nicolas Dufresne
ba8aa440bb v4l2bufferpool: Copy flags and timestamp when importing
Whenever we import from downstream pool (userptr or dmabuf-import), we
should copy over the flags and timestamp, otherwise downstream will not
get proper synchronization or will not be able to notice frames that has
corruption in it.

https://bugzilla.gnome.org/show_bug.cgi?id=785680
2017-08-01 09:48:46 -04:00
Nicolas Dufresne
2c7dfff047 v4l2: Add H263 Encoder support 2017-07-31 16:46:25 -04:00