Commit graph

120 commits

Author SHA1 Message Date
Shengqi Yu
1875c178cd v4l2object: use v4l2 reported width for padded_width when complex video formats
Stride means bytes per line, and padded_width means pixels. Here,
padded_width shoule be pix width reported by v4l2 instead of stride.

Part-of: <https://gitlab.freedesktop.org/gstreamer/gstreamer/-/merge_requests/7076>
2024-07-01 18:53:04 +00:00
Jakub Vaněk
0b65f667af v4l2src: Interpret V4L2 report of sync loss as video signal loss
Certain V4L2 drivers can report that a video receiver is seeing
some signal, but that it is unable to synchronize to it. IOW: the driver
can sometimes report V4L2_IN_ST_NO_SYNC and not report V4L2_IN_ST_NO_SIGNAL.

In particular, I've seen the tc358743 (HDMI-to-CSI2 converter) driver
sometimes report this when deployed to a fleet of embedded Raspberry Pis.
The relevant kernel code is in [1]. The video output is not practically
usable when V4L2_IN_ST_NO_SYNC is reported (only visually corrupted frames,
sometimes with random "snow", are received). I assume that this happens when
either the HDMI cable is poorly plugged in or damaged or when a CSI2 FFC
cable is used and is damaged.

The change in this commit is useful for detecting this working-but-not-really
condition in application code. Applications already listening for the "Signal lost"
message will gain the ability to handle this condition.

There seem to be more V4L2 error flags like this, see [2]. However, I do not
have practical experience with them and adding only V4L2_IN_ST_NO_SYNC seems
like a safer option.

[1]: https://github.com/raspberrypi/linux/blob/be8498ee21aa/drivers/media/i2c/tc358743.c#L1534
[2]: https://www.kernel.org/doc/html/v6.6/userspace-api/media/v4l/vidioc-enuminput.html

Part-of: <https://gitlab.freedesktop.org/gstreamer/gstreamer/-/merge_requests/7021>
2024-06-12 17:26:48 +00:00
Philipp Zabel
46a41667a3 v4l2bufferpool: Ensure freshly created buffers are not marked as queued
Otherwise, if we run in to the copy case, this can cause these
groups to stay around with queued flag set, but never actually
queued, until gst_v4l2_allocator_flush() is called, which then
erroneously frees the associated memories, causing the release
function to decrement the allocator refcount where it was never
incremented, resulting in early allocator disposal, and either
deadlock or use after free.

Part-of: <https://gitlab.freedesktop.org/gstreamer/gstreamer/-/merge_requests/6552>
2024-04-18 16:42:43 +00:00
Qian Hu (胡骞)
8d003f00e9 v4l2: add multiplane y42b(yuv422m)
for some jpg file, mediatek v4l2 jpeg decoder
hardware produce multi plane YUV 4:2:2 data

Part-of: <https://gitlab.freedesktop.org/gstreamer/gstreamer/-/merge_requests/6617>
2024-04-16 09:03:47 +00:00
Hou Qi
105d232fde v4l2bufferpool: queue back the buffer flagged LAST but empty
Some decoder drivers need to wait enough capture buffers before
starting to decode. But the dequeued buffer flag LAST but empty
has no chance to queue back to driver, which makes decode hang
after seek. So need to queue back such kind of buffer to driver.

Part-of: <https://gitlab.freedesktop.org/gstreamer/gstreamer/-/merge_requests/6579>
2024-04-15 18:07:17 +00:00
Philipp Zabel
e1f5bacf8d v4l2: bufferpool: Drop writable check on output pool process
Output buffers don't have to be writable. Accepting read-only buffers
from the V4L2 buffer pool allows upstream elements to write directly
into the V4L2 buffers without triggering a CPU copy into a new buffer
from the same V4L2 buffer pool every time.

Tested with the vivid output device:

  GST_DEBUG=GST_PERFORMANCE:7 gst-launch-1.0 videotestsrc ! v4l2sink device=/dev/video5

With this change, gst_v4l2_buffer_pool_dqbuf() must be allowed to not
resize read-only memories of output buffers.

Part-of: <https://gitlab.freedesktop.org/gstreamer/gstreamer/-/merge_requests/6572>
2024-04-15 17:11:00 +00:00
Tim Blechmann
1c9fe19b23 v4l2: enforce a pixel aspect ratio of 1/1 if no data are available
Part-of: <https://gitlab.freedesktop.org/gstreamer/gstreamer/-/merge_requests/6242>
2024-04-07 10:14:18 +00:00
Philipp Zabel
6f9872cb56 v4l2: allocator: Fix unref log/trace on memory release
Use gst_object_unref() instead of g_object_unref() in
gst_v4l2_allocator_release(), so refcounting log and
tracer get to know about this unref.

Part-of: <https://gitlab.freedesktop.org/gstreamer/gstreamer/-/merge_requests/6551>
2024-04-06 11:44:27 +00:00
Elliot Chen
e4ee4ca716 v4l2: fix error in calculating padding bottom for tile format
This is a regression while porting to arbitrary tile dimensions
introduced in !3424.

Part-of: <https://gitlab.freedesktop.org/gstreamer/gstreamer/-/merge_requests/6480>
2024-04-05 13:28:47 +00:00
Haihua Hu
37e3a38ba9 v4l2src: need maintain the caps order in caps compare when fixate
if the calculated "distance" of caps A and B from the preference are
equal, need to keep the original order instead of swap them

Part-of: <https://gitlab.freedesktop.org/gstreamer/gstreamer/-/merge_requests/6451>
2024-03-28 12:53:01 +00:00
Hou Qi
024d3ab051 v4l2: Also set max_width/max_height if enum framesize fail
Some driver doesn't implement enum_framesize. The maximum supported
size can be got by trying format with a very large size. Also need
to set max_width/max_height for this case, otherwise default encoded
buffer size 256kB is too small.

Part-of: <https://gitlab.freedesktop.org/gstreamer/gstreamer/-/merge_requests/6416>
2024-03-22 16:02:51 +00:00
Chao Guo
2e75b8c8e9 v4l2object: clear old fds in poll when closing v4l2object
When reopening a v4l2 device, the v4l2object->poll will include some old fds,
which was assigned to this device before. If the pipeline opens multiple v4l2
devices, the old fd may been assigned to other v4l2 devices when reopening
devices.

This will cause the timing of the pipeline become confusing when polling devices,
leading functional abnormalities.

Therefore, when closing v4l2object, remove the old fds in poll to ensure that the
pipeline timing is normal.

Signed-off-by: Chao Guo <chao.guo@nxp.com>
Part-of: <https://gitlab.freedesktop.org/gstreamer/gstreamer/-/merge_requests/5820>
2023-12-19 15:23:23 +00:00
Xavier Claessens
b80f4a1fa4 v4l2src: Consider framerate during caps selection
This simplifies the way it picks the closest caps to preference and take into
consideration the framerate to avoid picking high resolution at 5fps or so.
Simply calculate a "distance" of caps A and B from the preference and put
closest first, sorting by framerate first.

Part-of: <https://gitlab.freedesktop.org/gstreamer/gstreamer/-/merge_requests/5777>
2023-12-08 21:05:46 +00:00
Shengqi Yu
25c00b5ba2 v4l2object: scale the encoded sizeimage based on maximum resolution
The default 2MB ENCODED_BUFFER_SIZE can't support some 4K video playback. We now
detect the driver reported maximum resolution and choose an appropriate
default bitstream size accordingly. For 4K video these results in around 4MB
buffer instead of 2MB.

Part-of: <https://gitlab.freedesktop.org/gstreamer/gstreamer/-/merge_requests/4549>
2023-10-23 14:10:56 +00:00
Michael Tretter
0563a25494 v4l2videoenc: unconditionally activate the OUTPUT pool
If the v4l2videoenc receives an QUERY_ALLOCATION, it must not propose a
currently used pool, because it cannot be sure that the allocation query came
from exactly the same upstream element. The QUERY_ALLOCATION will not contain
the internal OUTPUT pool.

The upstream element (the basesrc) detects that the newly proposed pool differs
from the old pool. It deactivates the old pool and switches to the new pool.

If there was a format change, a new OUTPUT buffer pool will be allocated in
gst_v4l2_object_set_format_full() and the CAPTURE task will be stopped to switch
the format. If there hasn't been a format change,
gst_v4l2_object_set_format_full() will not be called. The old pool will be kept
and reused.

Without a format change, the processing task continues running.

This leads to the situation that the processing task is running, but the OUTPUT
buffer pool (the old pool) is deactivated. Therefore, the encoder is not able to
get buffers from the OUTPUT pool and encoding cannot continue.

This situation can be triggered by sending a RECONFIGURE event without a format
change.

Resolve this situation by ensuring that the OUTPUT buffer pool is always
activated when frames arrive at the encoder.

Part-of: <https://gitlab.freedesktop.org/gstreamer/gstreamer/-/merge_requests/4235>
2023-10-11 19:35:54 +00:00
Michael Tretter
41ce99ebab v4l2videoenc: fix activation of internal pool
Fix the buffer pool activation if the driver does not support VIDIOC_CREATE_BUFS
the same way as it was fixed for the v4l2videodec.

Part-of: <https://gitlab.freedesktop.org/gstreamer/gstreamer/-/merge_requests/4235>
2023-10-11 19:35:54 +00:00
Michael Tretter
5e72e1985a v4l2videoenc: rename OUTPUT pool to opool
There is a CAPTURE pool in the same function. While the CAPTURE pool is called
cpool, using pool for the OUTPUT pool is confusing.

Using opool for the OUTPUT pool makes it more obvious, which pool is used.

Part-of: <https://gitlab.freedesktop.org/gstreamer/gstreamer/-/merge_requests/4235>
2023-10-11 19:35:54 +00:00
Nicolas Dufresne
bcfbdfbbca v4l2: Fix tiled formats stride conversion
While adding arbitrary tile support, a round up operation was badly
converter. This caused the Y component of the stride to be 0. This
eventually lead to a crash in glupoad preceded by the following
assertion.

  gst_gl_buffer_allocation_params_new: assertion 'alloc_size > 0' failed

Part-of: <https://gitlab.freedesktop.org/gstreamer/gstreamer/-/merge_requests/5458>
2023-10-11 14:13:53 +00:00
Hou Qi
be9d9371b7 v4l2videodec: Correctly free caps to avoid memory leak
Part-of: <https://gitlab.freedesktop.org/gstreamer/gstreamer/-/merge_requests/5379>
2023-09-24 12:50:01 +00:00
Nicolas Dufresne
c1e03081c0 v4l2: object: Handle video helper return value
gst_video_info_set_interlaced_format() can return an error if the
width/height causes integer overflow. Handle this case, so that we can
fail cleanly. This has been experienced while testing an in-progress
driver.

Part-of: <https://gitlab.freedesktop.org/gstreamer/gstreamer/-/merge_requests/5286>
2023-09-11 15:05:34 -04:00
Nicolas Dufresne
353cb2da92 v4l2: bufferpool: Avoid warnings on empty last buffer
Some drivers will push an buffer flagged LAST but empty. In decoder
case, this results in an "producing too many buffer" warning, even
though the result is entirely correct. Detect this case in order to
signal EOS earlier and avoid this warning.

Part-of: <https://gitlab.freedesktop.org/gstreamer/gstreamer/-/merge_requests/5286>
2023-09-11 18:08:21 +00:00
Nicolas Dufresne
65350b601e v4l2: bufferpool: Do not resize compressed buffer
Avoid resizing compressed buffer to their maximum size. This fixes a
regression that caused valid but very large streams to be generated.

Fixes #2953

Part-of: <https://gitlab.freedesktop.org/gstreamer/gstreamer/-/merge_requests/5286>
2023-09-11 18:08:21 +00:00
Nicolas Dufresne
1e7ff1ac45 gstv4l2object: fix TODO comment about HDR configure
add following todo list
- Missing capture (v4l2src) HDR10 configuration and/or reporting
- The API is not capable of HDR to HDR conversion as controls are
      not specific to queues

Part-of: <https://gitlab.freedesktop.org/gstreamer/gstreamer/-/merge_requests/4403>
2023-08-22 21:20:11 +00:00
HuQian
fc7b776387 gstv4l2object: passing HDR10 information
when playing some codec such as matroska with vp9 codec,
demuxer will save information like video_mastering_display_info
and video_content_light_level in caps that decoder need,
v4l2videodecoder can use it by calling V4L2_CTRL_CLASS_COLORIMETRY
ioctl.

Part-of: <https://gitlab.freedesktop.org/gstreamer/gstreamer/-/merge_requests/4403>
2023-08-22 21:20:11 +00:00
Ming Qian
fd720fbf64 v4l2object: clear format lists if source change event is received
If decoder notify a source change event when the capture format is
changed, not the resolution changed.

then gst_v4l2_object_acquire_format will retuen false due to
unsupported format.

we need to clear the format lists in the source change flow,
and reenumerate format list

Part-of: <https://gitlab.freedesktop.org/gstreamer/gstreamer/-/merge_requests/5218>
2023-08-22 19:26:22 +00:00
Nicolas Dufresne
54ae2fcf77 v4l2: allocator: Don't close foreign dmabuf
Imported dmabuf are not being duped, so they should never be closed. Instead,
we ensure their live time by having strong reference on their original
buffer. This should fix potential flickering due to dmabuf being closed
too early.

Part-of: <https://gitlab.freedesktop.org/gstreamer/gstreamer/-/merge_requests/5101>
2023-08-21 20:45:14 +00:00
Nicolas Dufresne
8974318003 v4l2: bufferpool: Fix hang when splitting buffer
Now that we can split GStreamer buffers over multiple v4l2 buffer, we may
endup waiting for these buffers to be processed. Avoid waiting for any of
the parts being processed. As a side effect, the pool will now try to
grow if the number of buffers is not sufficient, and will fail
otherwise.

This fixes a hang if the very first frame did not fit. In this case, the
driver will retrain that buffer until the capture is setup, but
GStreamer won't setup the capture until process() function have
returned.

Fixes https://gitlab.freedesktop.org/gstreamer/gstreamer/-/merge_requests/5100

Part-of: <https://gitlab.freedesktop.org/gstreamer/gstreamer/-/merge_requests/5143>
2023-08-21 20:01:39 +00:00
Nicolas Dufresne
76fbc79494 v4l2: bufferpool: Keep processing bitstream buffer
Bitstream buffers may no fit a single v4l2 buffer, following spec
recommendation, keep processing the buffer until all the data has been
queued.

Part-of: <https://gitlab.freedesktop.org/gstreamer/gstreamer/-/merge_requests/5100>
2023-08-01 11:02:12 +00:00
Nicolas Dufresne
de5e1e334e v4l2: bufferpool: Fix buffer resize asserstion
When we fill a bitstream buffer the buffer might be too small to hold
the entire frame. Only resize to the filled size, preventing the
following assertion to happen.

  gst_buffer_resize_range: assertion 'bufmax >= bufoffs + offset + size' failed

Part-of: <https://gitlab.freedesktop.org/gstreamer/gstreamer/-/merge_requests/5100>
2023-08-01 11:02:12 +00:00
Michael Tretter
a92a64ae67 v4l2videoenc: remove empty sink_query
The sink_query() function simply calls the sink_query() function of the parent
videoencoder class. Remove the override to simply directly call the parent's
function.

Part-of: <https://gitlab.freedesktop.org/gstreamer/gstreamer/-/merge_requests/5034>
2023-07-13 16:25:06 +00:00
Michael Tretter
a28b81fb4f v4l2videoenc: replace custom QUERY_CAPS handling with getcaps callback
The videoencoder base class uses getcaps() to ask a subclass for the caps in its
sink_query_default() implementation.

Replace the custom handling of the QUERY_CAPS in the v4l2videoenc with an
implementation of getcaps() that returns the caps that are supported by the
v4l2videoenc to return these caps in the query.

This getcaps() implementation also calls the provided proxy_getcaps(), which
sends a caps query to downstream. This fixes the v4l2videoenc element to respect
limits of downstream elements in a sink query.

Part-of: <https://gitlab.freedesktop.org/gstreamer/gstreamer/-/merge_requests/5034>
2023-07-13 16:25:06 +00:00
Michael Tretter
5de27e0620 v4l2videoenc: always allocate CAPTURE buffer from our pool
The videoencoder base class always uses the negotiated allocator for allocating
coded buffers and ignores the negotiated buffer pool. Therefore, the
v4l2videoenc always has to copy buffers from the pool into the allocated
output buffers.

This breaks downstream elements that want to import the CAPTURE buffers of the
v4l2videoenc, since the v4l2videoenc copies the exported CAPTURE buffers and
sends the copies downstream.

Always use the CAPTURE buffer pool for acquiring CAPTURE buffers instead of
allocating the buffers in the base class.

Part-of: <https://gitlab.freedesktop.org/gstreamer/gstreamer/-/merge_requests/4230>
2023-07-12 16:15:06 +00:00
Nicolas Dufresne
bc294bd89d v4l2: videodec: Don't wait for src_ch if active
If the capture pool is already active, like when handling gaps at the
start of a stream, do not setup the decoder to wait for src_ch event.
Otherwise the decoder will endup waiting for that at the wrong moment
and exit the decoding thread unexpectedly.

Part-of: <https://gitlab.freedesktop.org/gstreamer/gstreamer/-/merge_requests/4590>
2023-07-07 16:23:48 -04:00
Nicolas Dufresne
c293ebc039 v4l2: videodec: Move pool setup inside negotiate()
Move all the pool configuration inside the negotiate() virtual function.
This allow settting up a pool with default format whenever the base
class wants to start without input data, like gaps.

Part-of: <https://gitlab.freedesktop.org/gstreamer/gstreamer/-/merge_requests/4590>
2023-07-07 16:23:48 -04:00
Hou Qi
8230c927f0 v4l2videodec: correctly register v4l2mpeg2dec
Part-of: <https://gitlab.freedesktop.org/gstreamer/gstreamer/-/merge_requests/4977>
2023-07-07 13:49:27 +00:00
Michael Olbrich
c6a7c88fd9 v4l2src: handle resolution change when buffers are copied
When buffers are copied then GST_V4L2_FLOW_RESOLUTION_CHANGE is returned by
gst_v4l2_buffer_pool_process() so do renegotiation here as well.

Part-of: <https://gitlab.freedesktop.org/gstreamer/gstreamer/-/merge_requests/4980>
2023-07-06 14:59:22 +02:00
Nicolas Dufresne
170dcd58db v4l2: Fix support for left and top padding
In the current implementation, we support for most pixel format left
and top padding by changing the offset in the video meta. Though, to
align driver bytesused to the offset, we recalculate the offset, which
removed the modification we did before.

Instead, save the plane size, and truncate the driver reported bytesused
to the expected size, which ensures that the offsets still match. This
should also fix issues were the buffer size ended up bigger then the
pool size due to driver introduced padding.

Part-of: <https://gitlab.freedesktop.org/gstreamer/gstreamer/-/merge_requests/4920>
2023-06-28 01:56:05 +00:00
Daniel Morin
6f43bdce49 v4l2src: adding support for bayer 10,12,14,16
- Adding bayer 10,12,14,16 bits components with 16 bits storage. These
  changes only adds capabilities. Capability format string is a complete
  description of the frame and pixels layout. Only mapping LE bayer
  formats as v4l2 only define LE bayer formats.

Part-of: <https://gitlab.freedesktop.org/gstreamer/gstreamer/-/merge_requests/4852>
2023-06-15 18:41:42 +00:00
Nicolas Dufresne
aea74db1a2 v4l2: Sync headers to current media_stage
commit d78b9d6671decdaedb539635b1d0a34f8f5934f8

Part-of: <https://gitlab.freedesktop.org/gstreamer/gstreamer/-/merge_requests/3447>
2023-06-15 14:32:32 +00:00
Daniel Morin
6ece5f3b90 v4l2src: fix support for bayer format
- Define a new function that identify if the v4l2object is raw based
on pixel format
- Only consider setting delta flag on buffer if the video is not raw.

Sponsored by Living Optics Ltd.

Part-of: <https://gitlab.freedesktop.org/gstreamer/gstreamer/-/merge_requests/4848>
2023-06-14 18:32:45 +00:00
ekwange
e7cfc1f5bd v4l2: Change to query only up to V4L2_CID_PRIVATE_BASE+V4L2_CID_MAX_CTRLS
Fix to prevent infinite querying.
There are devices that exceed V4L2_CID_PRIVATE_BASE+V4L2_CID_MAX_CTRLS
but do not return EINVAL.

Part-of: <https://gitlab.freedesktop.org/gstreamer/gstreamer/-/merge_requests/4790>
2023-06-13 13:04:37 +00:00
Hou Qi
fe21b750f9 v4l2videodec: treat MPEG 1 format as MPEG 2
Part-of: <https://gitlab.freedesktop.org/gstreamer/gstreamer/-/merge_requests/4770>
2023-06-06 15:41:47 +00:00
YURI FEDOSEEV
8dd51501d0 v4l2videoenc: support force keyframe event in v4l2 encoder
Part-of: <https://gitlab.freedesktop.org/gstreamer/gstreamer/-/merge_requests/4663>
2023-05-24 12:42:24 +00:00
Nicolas Dufresne
0c9ab49579 v4l2: videodec: Fix stalls on empty buffer
Drivers may signal end of sequence using an empty buffer and LAST buffer
set, or just an empty buffer on certain legacy implementation. When this
occured, we'd send GST_V4L2_FLOW_LAST_BUFFER were the code expected
GST_FLOW_EOS. Stop abusing GST_FLOW_EOS and port all the code to the new
GST_V4L2_FLOW_LAST_BUFFER.

Part-of: <https://gitlab.freedesktop.org/gstreamer/gstreamer/-/merge_requests/4669>
2023-05-19 23:06:06 +00:00
Shengqi Yu
5da9a8e2f4 v4l2object: fix some errors in probe_caps_for_fromat
1, there is a mistake when print stepwise.max_height, fix it
2, modify the calculation of width and height under the step wise
condition

Part-of: <https://gitlab.freedesktop.org/gstreamer/gstreamer/-/merge_requests/4562>
2023-05-18 13:45:11 +00:00
Hou Qi
783ebbeecb v4l2videoenc: fix set format failure when needs reset encoder
In cases that encoder needs to reset format, there is race while draining.
v4l2videoenc finish() sends CMD_STOP command to driver, and desire to return
GST_FLOW_OK. But at this time, encoder CAPTURE may have dequeued the last
buffer and got eos. finish() return value changes to be GST_FLOW_EOS which
causes set format fail. So there is no need to check return value for finish()
when set format.

Also need to flush encoder after draining to make sure flush is finished.

Part-of: <https://gitlab.freedesktop.org/gstreamer/gstreamer/-/merge_requests/4495>
2023-05-17 17:59:29 +00:00
Nicolas Dufresne
3bd43672ec v4l2: device provider: Fix GMainLoop leak
On very quick start/stop, the mainloop may never be run. As a side
effect, our idle stop function is not really being ran, so we can't rely
on that to free the main loop. Simply unref the mainloop when the
thread have completely stop.

Part-of: <https://gitlab.freedesktop.org/gstreamer/gstreamer/-/merge_requests/4521>
2023-05-03 10:04:58 +00:00
Nicolas Dufresne
51fa6a2656 v4l2: pool: Flush events on capture queue
Unfortunately streamoff does not flush the events, and this can cause all
sort of issues. Flush events on capture queue. We also return
GST_V4L2_FLOW_RESOLUTION_CHANGE in case a resolution change was seen.
This allow skipping streamon(capture) on flush, which could lead to a
configuration miss-match, or failure if the buffers aren't of the right
size.

Part-of: <https://gitlab.freedesktop.org/gstreamer/gstreamer/-/merge_requests/4437>
2023-05-01 15:08:10 -04:00
Nicolas Dufresne
00492234bd v4l2: videodec: Detect flushes while setting up the capture
As we missed the fact we were flushing, we could create and activate
that buffer pool, and wait on it, causing a hang. We detect that we
are flushing by checking the related pad state.

Part-of: <https://gitlab.freedesktop.org/gstreamer/gstreamer/-/merge_requests/4437>
2023-05-01 13:45:39 -04:00
Nicolas Dufresne
c9841a5383 v4l2: bufferpool: Don't copy buffer when flushing
Threshold handling can race with flushing operation. This can lead to
avoidable buffer copies. Simply check and return the flushing status.

Part-of: <https://gitlab.freedesktop.org/gstreamer/gstreamer/-/merge_requests/4437>
2023-05-01 13:45:16 -04:00