Commit graph

79 commits

Author SHA1 Message Date
Mathieu Duponchelle
72118b9db4 rtptwcc: complete bufferlist fix
When dealing with bufferlists, we need to store one "SentPacket"
structure per buffer, not one per buffer list!

Part-of: <https://gitlab.freedesktop.org/gstreamer/gstreamer/-/merge_requests/1278>
2021-11-12 18:40:32 +00:00
Sebastian Dröge
efb2b6d478 qtdemux: Log cslg_shift that was determined
Part-of: <https://gitlab.freedesktop.org/gstreamer/gstreamer/-/merge_requests/1294>
2021-11-12 17:51:04 +00:00
Sebastian Dröge
12e918428a qtdemux: Use a composition time offset of 0 for "no decode samples" for the time being
This needs codec-specific handling, but using 0 instead of G_MININT32 at
least gives somewhat reasonable behaviour.

See https://gitlab.freedesktop.org/gstreamer/gstreamer/-/issues/883

Part-of: <https://gitlab.freedesktop.org/gstreamer/gstreamer/-/merge_requests/1294>
2021-11-12 17:51:03 +00:00
Sebastian Dröge
ad412d257b qtdemux: Always check ctts for unreasonably large offsets
If this happens then ignore the whole ctts. Previously we only did this
if the PTS/DTS shift was determined from the ctts instead of the cslg.

Part-of: <https://gitlab.freedesktop.org/gstreamer/gstreamer/-/merge_requests/1294>
2021-11-12 17:51:03 +00:00
Sebastian Dröge
93a10a4ba1 qtdemux: Dump composition time offsets in trun as signed integers
Just like we do for ctts without regard of the version of the box.
Huge offsets are interpreted as negative offsets by qtdemux so this
works.

Part-of: <https://gitlab.freedesktop.org/gstreamer/gstreamer/-/merge_requests/1294>
2021-11-12 17:51:03 +00:00
Sebastian Dröge
a6f3391c81 qtdemux: Add a comment why only positive cslg shifts are considered
Part-of: <https://gitlab.freedesktop.org/gstreamer/gstreamer/-/merge_requests/1294>
2021-11-12 17:51:03 +00:00
Sebastian Dröge
a33e30cfc4 qtdemux: Only adjust segment.stop by cslg_shift if stop is not -1
Part-of: <https://gitlab.freedesktop.org/gstreamer/gstreamer/-/merge_requests/1294>
2021-11-12 17:51:03 +00:00
Sebastian Dröge
bb5a5ae8a8 qtdemux: Handle negative composition offsets in the trun box the same way as for non-fragmented streams
Part-of: <https://gitlab.freedesktop.org/gstreamer/gstreamer/-/merge_requests/1294>
2021-11-12 17:51:03 +00:00
Sebastian Dröge
767e8bf668 qtdemux: Parse ctts version
Negative composition time offsets are only allowed with version 1 of the
box, however we parse it as a signed value also for version 0 boxes as
unfortunately there are such files out there and it's unlikely to have
(valid) huge composition offsets.

Part-of: <https://gitlab.freedesktop.org/gstreamer/gstreamer/-/merge_requests/1294>
2021-11-12 17:51:03 +00:00
Sebastian Dröge
284dd5443f qtdemux: Add support for version 1 cslg boxes
They use 64 bit fields instead of 32 bit.

Also parse offset as a signed integer (in both versions) and clamp it to
a positive value as negative values don't really interest us here.

Part-of: <https://gitlab.freedesktop.org/gstreamer/gstreamer/-/merge_requests/1294>
2021-11-12 17:51:03 +00:00
Sebastian Dröge
7f105a919a qtdemux: Don't free cslg data that we don't own on corrupt files
Part-of: <https://gitlab.freedesktop.org/gstreamer/gstreamer/-/merge_requests/1294>
2021-11-12 17:51:03 +00:00
Rafał Dzięgiel
41385ab6f7 matroska: Ref index table when updating track info
Track index table array was being lost during track info update.
Ref it over to updated info, so it can be used for finding
nearest seek points.

Part-of: <https://gitlab.freedesktop.org/gstreamer/gstreamer/-/merge_requests/1203>
2021-11-12 12:28:40 +00:00
Rafał Dzięgiel
478f94edc7 matroska: Use g_array_unref everywhere
Instead of using g_array_free which is not thread safe use g_array_unref instead

Part-of: <https://gitlab.freedesktop.org/gstreamer/gstreamer/-/merge_requests/1203>
2021-11-12 12:28:40 +00:00
Tim-Philipp Müller
972615cf22 docs: fix unnecessary ampersand, < and > escaping in code blocks
Part-of: <https://gitlab.freedesktop.org/gstreamer/gstreamer/-/merge_requests/1340>
2021-11-12 11:39:19 +00:00
Mathieu Duponchelle
792fb05cec st2022-1-fecdec: fix packet trimming
g_sequence_remove_range's end iter is exclusive, so if one
wants to remove that item as well, it should be called with
the next iter.

This could in theory fix an issue where:

* The sequence isn't entirely trimmed, with an old item lingering

* Following FEC packets are immediately discarded because they
  arrived later than corresponding media packets, long enough for
  seqnums to wrap around

* We now try to reconstruct a media packet with a completely obsolete
  FEC packet, chaos ensues.

Part-of: <https://gitlab.freedesktop.org/gstreamer/gstreamer/-/merge_requests/1341>
2021-11-12 08:15:28 +00:00
Matthew Waters
d2fd5f1534 qmlsink: support caps changes better
We need to hold onto the last buffer until the next buffer arrives.
Before, if a caps change comes we would remove the currently rendering
buffer.  if Qt asks use to render something, we would render the dummy
black texture.

Fixes a period of black output when upstream is e.g. changing resolution
as in hls adaptive bitrate scenarios.

Part-of: <https://gitlab.freedesktop.org/gstreamer/gstreamer/-/merge_requests/1338>
2021-11-12 07:17:17 +00:00
James Cowgill
1e27ea63af v4l2: Record buffer states in pool to fix dequeue race
The `gst_v4l2_buffer_pool_dqbuf` function contains this ominous comment:

    /* get our GstBuffer with that index from the pool, if the buffer was
     * outstanding we have a serious problem.
     */
    outbuf = pool->buffers[group->buffer.index];

Unfortunately it is common for buffers in _output_ buffer pools to be
both queued and outstanding at the same time. This can happen if the
upstream element keeps a reference to the buffer, or in an encoder
element itself when it keeps a reference to the input buffer for each
frame.

Since the current code doesn't handle this case properly we can end up
with crashes in other elements such as:

    (gst-launch-1.0:32559): CRITICAL **: 17:33:35.740: gst_video_frame_map_id: assertion 'GST_IS_BUFFER (buffer)' failed

and:

    (gst-launch-1.0:231): GStreamer-CRITICAL **: 00:16:20.882: write map requested on non-writable buffer

Both these crashes are caused by a race condition related to releasing
the same buffer twice from two different threads. If a buffer is queued
and outstanding this situation is possible:

**Thread 1**
- Calls `gst_buffer_unref` decrementing the reference count to zero.
- The core GstBufferPool object marks the buffer non-outstanding.
- Calls the V4L2 release buffer function.
- If the buffer is _not_ queued:
  - Release it back to the free pool (containing non-queued buffers).

**Thread 2**
- Dequeues the queued output buffer.
  - Marks the buffer as not queued.
- If the buffer is _not_ outstanding:
  - Calls the V4L2 release buffer function.
  - Release it back to the free pool (containing non-queued buffers).

If both of these threads run at exactly the same time there is a small
window where the buffer is marked both not outstanding and not queued
but before it has been released. In this case the buffer will be freed
twice causing the above crashes.

Unfortunately the variable recording whether a buffer is outstanding is
part of the core `GstBuffer` object and is managed by `GstBufferPool` so
it's not as straightforward as adding a mutex. Instead we can fix this
by additionally recording the buffer state in `GstV4l2BufferPool`, and
handle "internal" and "external" buffer release separately so we can
detect when a buffer becomes not outstanding.

In the new solution:
- The "external" buffer pool release and the "dqbuf" functions
  atomically update the buffer state and determine if a buffer is still
  queued or outstanding.
- Subsequent code and a new
  `gst_v4l2_buffer_pool_complete_release_buffer` function can proceed to
  release (or not) a buffer knowing that it's not racing with another
  thread.

Part-of: <https://gitlab.freedesktop.org/gstreamer/gstreamer/-/merge_requests/1010>
2021-11-11 22:30:31 +00:00
Matthew Waters
71dd47516c rtpbin: separate out the two fec decoder locations
The pipeline flow for receiving looks like this:

rtpsession ! rtpssrcdemux ! session_fec_decoder ! rtpjitterbuffer ! \
  rtpptdemux ! stream_fec_decoder ! ...

There are two places where a fec decoder could be placed.
1. As requested from the 'request-fec-decoder' signal: after rtpptdemux
   for each ssrc/pt produced
2. after rtpssrcdemux but before rtpjitterbuffer: added for the
   rtpst2022-1-fecenc/dec elements,

However, there was some cross-contamination of the elements involved and
the request-fec-decoder signal was also being used to request the fec
decoder for the session_fec_decoder which would then be cached and
re-used for subsequent fec decoder requests.  This would cause the same
element to be attempted to be linked to multiple elements in different
places in the pipeline.  This would fail and cause all kinds of havoc
usually resulting in a not-linked error being returned upstream and an
error message being posted by the source.

Fix by not using the request-fec-decoder signal for requesting the
session_fec_decoder and instead solely rely on the added properties for
that case.

Part-of: <https://gitlab.freedesktop.org/gstreamer/gstreamer/-/merge_requests/1300>
2021-11-10 10:38:26 +00:00
Zhao, Gang
6cad2a7150 qtdemux: Fix can not demux Opus track made by qtmux
Opus stream info is read from dOps box [1]. The offset of dOps box in Opus box is different in mp4a version 1 and 0 [2]. Calculate the offset of dOps box according to mp4a version.

[1] https://opus-codec.org/docs/opus_in_isobmff.html

[2] subprojects/gst-plugins-good/gst/isomp4/atoms.c:sample_entry_mp4a_copy_data:2146

Fixed: https://gitlab.freedesktop.org/gstreamer/gst-plugins-good/-/issues/918
Part-of: <https://gitlab.freedesktop.org/gstreamer/gstreamer/-/merge_requests/1329>
2021-11-09 17:57:49 +00:00
Haihua Hu
a66124a79c v4l2bufferpool: set video alignment of video meta
need apply video alignment info on video meta, downstream
element can get buffer alignment from video meta

Part-of: <https://gitlab.freedesktop.org/gstreamer/gstreamer/-/merge_requests/1326>
2021-11-09 13:32:46 +00:00
Zhao, Gang
5ac1a12927 vpxdec: Skip check of key frame when open GOP
Valid stream [1] might has no key frame.

Fixed: https://gitlab.freedesktop.org/gstreamer/gst-plugins-good/-/issues/890

[1] https://storage.googleapis.com/downloads.webmproject.org/test_data/libvpx/vp90-2-16-intra-only.webm

Part-of: <https://gitlab.freedesktop.org/gstreamer/gstreamer/-/merge_requests/1316>
2021-11-08 19:42:47 +00:00
Joe Todd
a0df14f36b osxaudio: Increase max channels to 64
Part-of: <https://gitlab.freedesktop.org/gstreamer/gstreamer/-/merge_requests/1059>
2021-11-08 18:09:52 +00:00
Sanchayan Maity
3f49b21b80 imagefreeze: Only set caps if they do not match current caps
Part-of: <https://gitlab.freedesktop.org/gstreamer/gstreamer/-/merge_requests/1311>
2021-11-05 15:34:47 +05:30
Sebastian Dröge
d2f66106a8 v4l2: Support bools for setting extra-controls
They're just mapped to 0 / 1 and can already be set as integers, but
being able to set them as booleans seems useful.

Part-of: <https://gitlab.freedesktop.org/gstreamer/gstreamer/-/merge_requests/1308>
2021-11-04 15:21:40 +00:00
Tim-Philipp Müller
a7b376011b Back to development 2021-11-03 19:31:23 +00:00
Tim-Philipp Müller
f513c289b0 Release 1.19.3 2021-11-03 15:43:43 +00:00
Tim-Philipp Müller
d51b091cd9 Update ChangeLogs for 1.19.3 2021-11-03 15:43:32 +00:00
Sebastian Dröge
dac82a8932 multifilesink: Make minimum distance between keyframes in next-file=key-frame mode configurable
Previously this was hardcoded to 10s, which is not necessarily the
desired behaviour.

Part-of: <https://gitlab.freedesktop.org/gstreamer/gstreamer/-/merge_requests/1292>
2021-11-02 20:21:10 +00:00
Erlend Eriksen
0805ffdce9 qtmux: Fix deadlock in gst_qt_mux_prepare_moov_recovery
Regression from 5766731bd4

Part-of: <https://gitlab.freedesktop.org/gstreamer/gstreamer/-/merge_requests/1288>
2021-11-02 19:26:57 +00:00
Sebastian Dröge
577cdcafe0 video: Fix order of new video formats
Part-of: <https://gitlab.freedesktop.org/gstreamer/gstreamer/-/merge_requests/1282>
2021-10-31 06:38:36 +00:00
Tim-Philipp Müller
ea8dc0c737 Couple more g_memdup() -> g_memdup2() fixes
Fixes deprecation warnings with newer GLib versions.

Part-of: <https://gitlab.freedesktop.org/gstreamer/gstreamer/-/merge_requests/1279>
2021-10-30 10:37:37 +01:00
Piotrek Brzeziński
428b4104b0 video-format: Add support for ARGB64 LE/BE and similar variants
Co-authored-by: Sebastian Dröge <sebastian@centricular.com>
Part-of: <https://gitlab.freedesktop.org/gstreamer/gstreamer/-/merge_requests/1247>
2021-10-29 14:57:58 +00:00
Mathieu Duponchelle
c414f9560a rtptwcc: don't assume all PacketInfo->data are buffers
They can also be buffer lists

Part-of: <https://gitlab.freedesktop.org/gstreamer/gstreamer/-/merge_requests/1268>
2021-10-29 03:46:25 +02:00
Tim-Philipp Müller
f034ec3e14 meson: require matching GStreamer dep versions for unstable development releases
Fixes https://gitlab.freedesktop.org/gstreamer/gst-plugins-good/-/issues/929

Part-of: <https://gitlab.freedesktop.org/gstreamer/gstreamer/-/merge_requests/1244>
2021-10-28 23:29:27 +00:00
Jan Schmidt
f6ed40c93a qtdemux: Fix text and closed-caption handling.
https://gitlab.freedesktop.org/gstreamer/gstreamer/-/merge_requests/1182
broke text and and closed caption extraction when introducing WebVTT
support, by making the output buffers not have timestamps any more.

Fix that by making the process functions copy buffer metadata
when generating new output buffers.

Part-of: <https://gitlab.freedesktop.org/gstreamer/gstreamer/-/merge_requests/1262>
2021-10-28 21:45:33 +11:00
Tim-Philipp Müller
97d35912d9 gst-plugins-good: update translations
Fixes #656

Part-of: <https://gitlab.freedesktop.org/gstreamer/gstreamer/-/merge_requests/1240>
2021-10-25 01:11:29 +01:00
Sebastian Dröge
2853c085c7 qtdemux: Add pasp box to the list of known boxes
Part-of: <https://gitlab.freedesktop.org/gstreamer/gstreamer/-/merge_requests/1235>
2021-10-23 11:39:36 +00:00
Rafał Dzięgiel
b57a7c3de7 matroska: Set image/attachment structure mimetype
Set structure mimetype to fix data detection by mimetype in other plugins.

Part-of: <https://gitlab.freedesktop.org/gstreamer/gstreamer/-/merge_requests/1205>
2021-10-21 21:42:00 +00:00
Rafał Dzięgiel
0ba65a00c9 matroska: Treat non-image structure as attachment
Otherwise each structure is named as GstTagImageInfo even if
it does not contain any images which is misleading.

Part-of: <https://gitlab.freedesktop.org/gstreamer/gstreamer/-/merge_requests/1205>
2021-10-21 21:42:00 +00:00
Matthew Waters
8c35850f23 rtpbin: fix leak of pad when a fec encoder and aux sender a created
The ghost sink pad retrieved by rtpbin from the aux sender was not freed
when there was a previous element (fec encoder) in the chain.

Part-of: <https://gitlab.freedesktop.org/gstreamer/gstreamer/-/merge_requests/1222>
2021-10-21 13:46:03 +00:00
Tim-Philipp Müller
8dfab0b08c meson: update for meson.build_root() and .build_source() deprecation
-> use meson.project_build_root() or .global_build_root() instead.

Part-of: <https://gitlab.freedesktop.org/gstreamer/gstreamer/-/merge_requests/1183>
2021-10-20 11:20:44 +00:00
Tim-Philipp Müller
f1bb2c76c6 meson: update for dep.get_pkgconfig_variable() deprecation
... in favour of dep.get_variable('foo', ..) which in some
cases allows for further cleanups in future since we can
extract variables from pkg-config dependencies as well as
internal dependencies using this mechanism.

Part-of: <https://gitlab.freedesktop.org/gstreamer/gstreamer/-/merge_requests/1183>
2021-10-20 11:20:44 +00:00
Tim-Philipp Müller
531aa4064a meson: clean up conditional paths after version bump
Part-of: <https://gitlab.freedesktop.org/gstreamer/gstreamer/-/merge_requests/1183>
2021-10-20 11:20:44 +00:00
Olivier Crête
0dbe0e21fe rtphdrext-clientaudiolevel: Rename RFC 6464 element
Multiplying elements named after RFC numbers is confusing,
so let's give them meaningful names.

Part-of: <https://gitlab.freedesktop.org/gstreamer/gstreamer/-/merge_requests/1125>
2021-10-20 00:03:09 +00:00
Jan Schmidt
6cada5b064 qtdemux: Add support for wvtt (WebVTT) subtitles.
WebVTT in ISO MP4 is specified in ISO 14496-30,
and needed for DASH support. It's stored in an
mp4 specific format. To handle it compatibly,
the wvtt boxes are converted back into WebVTT text
and pushed as application/x-subtitle-vtt

Part-of: <https://gitlab.freedesktop.org/gstreamer/gstreamer/-/merge_requests/1182>
2021-10-19 08:56:58 +00:00
Jan Schmidt
41d6f47f2b isomp4: Split buffer process functions.
Split the different handling for closed captions, VobSub subpicture
and timed text samples into separate simplified process functions.

Part-of: <https://gitlab.freedesktop.org/gstreamer/gstreamer/-/merge_requests/1182>
2021-10-19 08:56:58 +00:00
Jan Schmidt
c47c6b4c2b isomp4: Use a function pointer for buffer splitting.
Swap the `need_process` boolean check on qtdemux streams
for a direct function pointer to the splitting function,
so we can stop adding extra cases to the single growing
`gst_qtdemux_process_buffer()` function.

Part-of: <https://gitlab.freedesktop.org/gstreamer/gstreamer/-/merge_requests/1182>
2021-10-19 08:56:58 +00:00
Tim-Philipp Müller
0a25f22f1b meson: bump meson requirement to >= 0.59
For monorepo build and ugly/bad, for advanced feature
option API like get_option('xyz').required(..) which
we use in combination with the 'gpl' option.

For rest of modules for consistency (people will likely
use newer features based on the top-level requirement).

Part-of: <https://gitlab.freedesktop.org/gstreamer/gstreamer/-/merge_requests/1084>
2021-10-18 18:03:19 +01:00
Thibault Saunier
20f6af7e3c qtmux: Allow more fields changes renegotiating h264
And consider interlace-mode=progress as equivalent to the field not
being specified.

Part-of: <https://gitlab.freedesktop.org/gstreamer/gstreamer/-/merge_requests/1062>
2021-10-18 10:10:24 -03:00
Thibault Saunier
32a88fe16f matroskamux: Ignore some fields when renegotiating
Those values are never used in the muxer so we should not fail the
negotiation if those are changing.

Part-of: <https://gitlab.freedesktop.org/gstreamer/gstreamer/-/merge_requests/1062>
2021-10-18 10:10:24 -03:00