dashdemux: Preserve current representation on live manifest updates

When updating a manifest during live playback, preserve the current
representation for each stream.

During update_fragment_info, if the current representation changed
because it couldn't be matched, trigger a caps change and new
header download.

This reverts commit e0e1db212fd0df2239583b9099fc4361adeded05
and reapplies "dashdemux: Fix issue when manifest update sets slow start
without passing necessary header & caps changes downstream" with
changes.

Fixes https://gitlab.freedesktop.org/gstreamer/gst-plugins-bad/-/issues/507
Fixes https://gitlab.freedesktop.org/gstreamer/gst-plugins-bad/-/issues/1729

Part-of: <https://gitlab.freedesktop.org/gstreamer/gstreamer/-/merge_requests/2920>

examples: Disable bitcode support in iOS examples

For explanation, see:

https://gitlab.freedesktop.org/gstreamer/cerbero/-/merge_requests/879

Part-of: <https://gitlab.freedesktop.org/gstreamer/gstreamer/-/merge_requests/2978>

rtpjitterbuffer: Add test for crash caused by removing timers twice

Part-of: <https://gitlab.freedesktop.org/gstreamer/gstreamer/-/merge_requests/2973>

rtpjitterbuffer: Make it more explicit that update_rtx_timers() takes ownership of the passed in timer

It is not valid anymore afterwards and must not be used, otherwise an
already freed pointer might be used.

Part-of: <https://gitlab.freedesktop.org/gstreamer/gstreamer/-/merge_requests/2973>

rtpjitterbuffer: Don't shadow variable

While this didn't cause any problems in this context it is simply
confusing.

Part-of: <https://gitlab.freedesktop.org/gstreamer/gstreamer/-/merge_requests/2973>

rtpjitterbuffer: Change RTX timer availability checks to assertions

It's impossible to end up in the corresponding code without a timer for
RTX packets because otherwise it would be an unsolicited RTX packet and
we would've already returned early.

Part-of: <https://gitlab.freedesktop.org/gstreamer/gstreamer/-/merge_requests/2973>

rtpjitterbuffer: Only unschedule timers for late packets if they're not RTX packets and only once

Timers for RTX packets are dealt with later in update_rtx_timers(), and
timers for non-RTX packets would potentially also be unscheduled a
second time from there so avoid that.

Also don't shadow the timer variable from the outer scope but instead
make use of it directly.

Part-of: <https://gitlab.freedesktop.org/gstreamer/gstreamer/-/merge_requests/2973>

subprojects: update openh264 wrap to v2.3.0

.. and use tarball instead of git checkout.

Also bump ci image tags so the images cache the new tarball.

Part-of: <https://gitlab.freedesktop.org/gstreamer/gstreamer/-/merge_requests/2968>

gst-rtsp-server: Fix pushing backlog to client

Check back pressure of a stream transport before popping buffer from its backlog.

If the stream transport is not experiencing back pressure, the buffer can be popped from backlog and pushed to client.

Fixes:#1298

Part-of: <https://gitlab.freedesktop.org/gstreamer/gstreamer/-/merge_requests/2936>

meson: Subprjects can define both "plugins" and "gst_plugins"

If "gst_plugins" is defined we can ignore the value of legacy "plugins"
variable, subprojects could be using it for something else, which is the
case of gst-plugin-rs.

Part-of: <https://gitlab.freedesktop.org/gstreamer/gstreamer/-/merge_requests/2976>

rtsp-server: stream: Don't loop forever if binding to the multicast address fails

The address/port is pre-defined by the caller of the function, so
retrying is only going to loop forever.

Ideally the multicast address should be checked after allocating but
this doesn't happen currently, so it's better to error out cleanly then
to loop forever trying the same address.

Part-of: <https://gitlab.freedesktop.org/gstreamer/gstreamer/-/merge_requests/2975>

va: Fix log message when registering H264 encoder.

The log message would report an error for the H264 decoder when
registering failed, but we tried to register the H264 encoder instead.

Part-of: <https://gitlab.freedesktop.org/gstreamer/gstreamer/-/merge_requests/2974>

webrtcbin: Allow locked mlines with no caps, as the last ones

Part-of: <https://gitlab.freedesktop.org/gstreamer/gstreamer/-/merge_requests/2439>

webrtcbin: Reject creating an offer if a locked mline has no caps

This avoids getting in a bunch of corner cases. We'd have to insert
a "rejected" line from the start as a place-holder to get around this,
but the rest of the code just becomes more complicated, so just
disallow it for now.

Part-of: <https://gitlab.freedesktop.org/gstreamer/gstreamer/-/merge_requests/2439>

webrtcbin: Store pending mid to make create-offer idempotent

If the mid is not stored in the transceiver, but it is stored in
last_offer, then a further create-offer call will just ignore that
transceiver.

Also include unit test for ensure it doesn't regress.

Part-of: <https://gitlab.freedesktop.org/gstreamer/gstreamer/-/merge_requests/2439>

rtph265: keep delta unit flag

Without this patch all buffers that pass the payloader
are marked as non-delta-unit buffers.

Part-of: <https://gitlab.freedesktop.org/gstreamer/gstreamer/-/merge_requests/2969>

msdkvpp: Add va memory when fixating src caps

Part-of: <https://gitlab.freedesktop.org/gstreamer/gstreamer/-/merge_requests/2498>

msdkvpp : Use va pool at linux path and system pool for windows

We use msdkvpp's own pool to allocate buffers instead of external frame
allocator mfxFrameAllocator.

Part-of: <https://gitlab.freedesktop.org/gstreamer/gstreamer/-/merge_requests/2498>

msdkvpp: Import buffer to msdk_surface

If the buffer is not msdk_buffer, we can try to directly import the
attached memory (i.e. va mem and dmabuf mem) by applying the common
uitl function: import_to_msdk_function ().

Here add a flag "from_qdata" in GstMsdkSurface to handle the cropping case,
we should avoid updating the crop values when msdk_surface is from the
memory's qdata, because the crop info from this surface is the already
updated one.

Part-of: <https://gitlab.freedesktop.org/gstreamer/gstreamer/-/merge_requests/2498>

msdkvpp: Add va caps at sink and src pad

Part-of: <https://gitlab.freedesktop.org/gstreamer/gstreamer/-/merge_requests/2498>

msdkenc: Apply common util func to import mem as msdk_surface

Part-of: <https://gitlab.freedesktop.org/gstreamer/gstreamer/-/merge_requests/2498>

msdk: Add help functions to get mfxFrameSurface1 from GstBuffer and wrap it as GstMsdkSurface

Note that the memory abstraction for system memory is for windows path.

Part-of: <https://gitlab.freedesktop.org/gstreamer/gstreamer/-/merge_requests/2498>

msdkenc: Use va pool on linux and system pool for windows

Part-of: <https://gitlab.freedesktop.org/gstreamer/gstreamer/-/merge_requests/2498>

msdkenc: Directly import dmabuf memory as mfx surface

When input buffer is of dmabuf memory but not a msdk buffer (i.e., the
allocator is not msdk_allocator), then we can try to get fd of this mem,
create the corresponding va surface and wrap it as mfx surface.

Part-of: <https://gitlab.freedesktop.org/gstreamer/gstreamer/-/merge_requests/2498>

msdkenc: Directly import va memory as mfx surface

If input buffer is of va memory, we can directly obtain va surface from
the buffer and wrap it as mfx surface.

Part-of: <https://gitlab.freedesktop.org/gstreamer/gstreamer/-/merge_requests/2498>

msdk: Use va libs API to get VA surface from buffer

Direct apply gst_va_buffer_get_surface to get VASurface from a buffer,
so remove corresponding functions which are not used.

Part-of: <https://gitlab.freedesktop.org/gstreamer/gstreamer/-/merge_requests/2498>

meson: Call pkgconfig.generate in the loop where we declare plugins dependencies

Removing some copy pasted code

Part-of: <https://gitlab.freedesktop.org/gstreamer/gstreamer/-/merge_requests/2970>

meson: Namespace the plugins_doc_dep/libraries variables

Part-of: <https://gitlab.freedesktop.org/gstreamer/gstreamer/-/merge_requests/2970>

cuda: meson: Fix generating GstCuda gir file against gstreamer-full

Part-of: <https://gitlab.freedesktop.org/gstreamer/gstreamer/-/merge_requests/2970>

meson: Rename plugins list and make them "dependency" objects

Part-of: <https://gitlab.freedesktop.org/gstreamer/gstreamer/-/merge_requests/2970>

rtpjitterbuffer: remove lost timer for out of order packets

When receiving old packets remove the running lost timer if present.
This fixes incorrect reporting of a lost packet even if it arrived in time.

Part-of: <https://gitlab.freedesktop.org/gstreamer/gstreamer/-/merge_requests/2922>

gstaudioconvert: doc: Fix mix-matrix example

mix-matrix should contain float elements, modify the example to do so

Part-of: <https://gitlab.freedesktop.org/gstreamer/gstreamer/-/merge_requests/2966>

gstpluginloader: Don't hang on short reads/writes

If read_one or write_one was called but the stream closed before it could
read/write a whole packet, read_one/write_one would hang indefinitely,
consuming 100% CPU. This commit fixes that by treating a short read/write
as an error.

Part-of: <https://gitlab.freedesktop.org/gstreamer/gstreamer/-/merge_requests/2964>

validate: Fix gir generation

It broke in https://gitlab.freedesktop.org/gstreamer/gstreamer/-/merge_requests/2555
where we stopped parsing some .c files

Part-of: <https://gitlab.freedesktop.org/gstreamer/gstreamer/-/merge_requests/2965>

validate: don't look for the common supression file

We no longer have a common submodule

Part-of: <https://gitlab.freedesktop.org/gstreamer/gstreamer/-/merge_requests/2841>

gst_init: Initialize static plugins just before dynamic plugins

All plugins needs to be initialized after `gst_initialized = TRUE;`
otherwise they could complain that gst_init() has not been called.

Part-of: <https://gitlab.freedesktop.org/gstreamer/gstreamer/-/merge_requests/2962>

meson: Make sure devenv uses tools linked on gst-full

Part-of: <https://gitlab.freedesktop.org/gstreamer/gstreamer/-/merge_requests/2959>

rtpvp8depay: If configured to wait for keyframes after packet loss, also do that if incomplete frames are detected

This can happen if the data inside the packets is incomplete without the
seqnums being discontinuous because of ULPFEC being used.

Part-of: <https://gitlab.freedesktop.org/gstreamer/gstreamer/-/merge_requests/2947>

examples: webrtc: mp-sendrecv: Add missing semicolon

Part-of: <https://gitlab.freedesktop.org/gstreamer/gstreamer/-/merge_requests/2961>

examples/webrtc: add some missing unrefs

Part-of: <https://gitlab.freedesktop.org/gstreamer/gstreamer/-/merge_requests/2958>

d3d11memory: Fix GstAllocator::mem_copy() implementation

We were checking possible bind flags for the DXGI format
of the source texture but that's never applied to
the destination texture desc.

Just use the already configured bind (and misc) flags of source texture
for the destination texture allocation without additional check.

Part-of: <https://gitlab.freedesktop.org/gstreamer/gstreamer/-/merge_requests/2950>

audiovisualizer: fix buffer mapping to not increase refcount

Part-of: <https://gitlab.freedesktop.org/gstreamer/gstreamer/-/merge_requests/2953>

docs: disable in static build

Following gst-plugins-base, disable docs if static_build
in:
 - gstreamer
 - gst-plugins-good
 - gst-plugins-ugly
 - gst-libav
 - gstreamer-vaapi

Part-of: <https://gitlab.freedesktop.org/gstreamer/gstreamer/-/merge_requests/2529>

curlbasesink: gst_curl_base_sink_transfer_thread_close is internal

gst_curl_base_sink_transfer_thread_close is moved from external header
to be static function, as it has no users.

Part-of: <https://gitlab.freedesktop.org/gstreamer/gstreamer/-/merge_requests/2732>

curlhttpsink: Only set MIME as content-type if not set by property

Setting the content-type property shall override internally detected MIME
types, to make it possible to do as following example (where audio/basic to be
used prior to audio/x-mulaw):

gst-launch-1.0 ... ! mulawenc ! audio/x-mulaw,rate=8000,channels=1 !
  curlhttpsink location=<url> content-type=audio/basic

Part-of: <https://gitlab.freedesktop.org/gstreamer/gstreamer/-/merge_requests/2732>

d3d11: Update library doc

* Private header name is changed to gstd3d11-private.h to follow
  naming convention
* Add Since mark everywhere
* Update member variable names to be consistent with the other
  object implementations in this library
* Correct outdated documentation

Part-of: <https://gitlab.freedesktop.org/gstreamer/gstreamer/-/merge_requests/2945>

docker/windows: Pass a couple installargs to python3

From these, only PrependPath isn't enabled by default,
but its nice to explicitly enable the ones we need anyway.

Part-of: <https://gitlab.freedesktop.org/gstreamer/gstreamer/-/merge_requests/2933>

value: Use g_critical() when trying to serialize things that can't be

Part-of: <https://gitlab.freedesktop.org/gstreamer/gstreamer/-/merge_requests/2823>

gstvalue: Don't loop forever when serializing invalid flag

The serialization code would loop forever if an invalid flag was sent into it.

With unit test for this corner case.

Part-of: <https://gitlab.freedesktop.org/gstreamer/gstreamer/-/merge_requests/2823>

bin: Don't propagate state change errors of elements in locked state

Theoretically having elements in locked state should not have any effect
at all when the surrounding bin is doing state changes. However
previously a state change error of a locked element would cause the
bin's state change to also fail, which is clearly not intended.

State change failures of locked elements are to be handled by whoever
set the element to locked state. By always returning them here it is
impossible for the owner of the element to handle state change failures
gracefully without potentially affecting the whole pipeline's state
changes.

Non-failure returns are still returned as-is as the distinction between
ASYNC/NO_PREROLL/SUCCESS has big consequences on the state changes of
the bin and overall pipeline. Theoretically SUCCESS should also be
returned in all cases but I can't estimate the effects this would have
on the overall pipeline.

Part-of: <https://gitlab.freedesktop.org/gstreamer/gstreamer/-/merge_requests/2932>

x(v)imagesink: Don't create invalid GstNavigationModifierType by simply passing through X11 event states

Part-of: <https://gitlab.freedesktop.org/gstreamer/gstreamer/-/merge_requests/2931>

avfvideosrc: Fix wrong default framerate value

Current default G_MAXINT is not a correct value under any circumstances.
This creates an issue with screen capture, during which we currently do
not get any framerate info causing G_MAXINT to show up, where elements
downstream can possibly misbehave - for example, `vtenc` causes
a kernel panic.
Replace with 30/1 to avoid such scenarios.

Part-of: <https://gitlab.freedesktop.org/gstreamer/gstreamer/-/merge_requests/2944>

jpegdecoder: return the real error of decode_scan and decode_frame.

The current handle_frame() does not return the real error that happens
in decode_scan and decode_frame, which makes the pipeline continue with
the error and may trigger asserting later.

We also return the error when decode_quant_table or decode_huffman_table
fails.

Part-of: <https://gitlab.freedesktop.org/gstreamer/gstreamer/-/merge_requests/2938>

examples: d3d11videosink: Add present signal example

Add an example to show the usage of present singal.
In this example, a text overlay with alpha blended background
will be rendered on swapchain's backbuffer by using
Direct3D11, Direct2D, and DirectWrite APIs.

Part-of: <https://gitlab.freedesktop.org/gstreamer/gstreamer/-/merge_requests/2923>

d3d11videosink: Add display-format property

Make swapchian's display format configurable, since some DXGI formats
we can use for swapchain are not API interop compatible.
For instance, BGRA format should be used for Direct2D interop.

Part-of: <https://gitlab.freedesktop.org/gstreamer/gstreamer/-/merge_requests/2923>

d3d11videosink: Add present signal

The "present" signal will be emitted just before the
IDXGISwapChain::Present() call. The client can perform additional
GPU operation with given GstD3D11Device object and
ID3D11RenderTargetView handle. Or, the client can read back
the scene to be displayed on window using the signal.

Part-of: <https://gitlab.freedesktop.org/gstreamer/gstreamer/-/merge_requests/2923>

v4l2allocator: Fix invalid imported dmabuf fd

Fix a typo that set userptr to dmabuf fd. It leads to failure of
dmabuf-import io-mode.

Part-of: <https://gitlab.freedesktop.org/gstreamer/gstreamer/-/merge_requests/2935>

playback: add onvif metadata caps to raw caps

+ remove encoding from x-onvif-metadata caps output by qtdemux

Part-of: <https://gitlab.freedesktop.org/gstreamer/gstreamer/-/merge_requests/2889>

videoconvert: fix passthrough on equivalent transfer

When the input info and output info are equal, except for the
transfer functions that are *not* equivalent, we need to set up
a converter as we won't be passthrough.

Fixes an assertion in that case.

Part-of: <https://gitlab.freedesktop.org/gstreamer/gstreamer/-/merge_requests/2929>

isoff: Fix earliest pts field parse issue

earliest pts will be covered by first_offset field on version 0 case.

Part-of: <https://gitlab.freedesktop.org/gstreamer/gstreamer/-/merge_requests/2927>

dmabuf: Make it clear that the FORMAT is a DRM fourcc

Some of the formats in GstVideoFormat are a combination
of a DRM format and modifier, making them unclear.

Part-of: <https://gitlab.freedesktop.org/gstreamer/gstreamer/-/merge_requests/2616>

h265decoder: Update SpsMaxLatencyPictures properly

The SpsMaxLatencyPictures value never gets back to zero even if
it's needed. Update the value properly.

Part-of: <https://gitlab.freedesktop.org/gstreamer/gstreamer/-/merge_requests/2915>

h265decoder: Reorganize debug printing

As of the MR
https://gitlab.freedesktop.org/gstreamer/gstreamer/-/merge_requests/2575
gst_h265_decoder_process_sps() is called per slice and it causes
noisy debug message print.
Reorganize related and the other debug printings at once.

Part-of: <https://gitlab.freedesktop.org/gstreamer/gstreamer/-/merge_requests/2915>

openh264: Register debug categories earlier

Otherwise the GST_ERROR message logged in case of ABI mismatch would be done on
an uninitialized category.

Part-of: <https://gitlab.freedesktop.org/gstreamer/gstreamer/-/merge_requests/2918>

openh264enc: Fix constrained-high encoding

constrained-high is high without B-frames, there is no EProfileIdc for this, so
assume high instead of hitting an assert down the line.

Part-of: <https://gitlab.freedesktop.org/gstreamer/gstreamer/-/merge_requests/2919>

webrtc:ice: Fix candidate stats related APIs for bindings

null-terminated arrays of structures is not usable.

Part-of: <https://gitlab.freedesktop.org/gstreamer/gstreamer/-/merge_requests/2917>

va: h265dec: Enable the scc_main_4:4:4_10 profile.

We should enable this profile which is already allown in vaprofile.c
after libva 1.8 version.

Part-of: <https://gitlab.freedesktop.org/gstreamer/gstreamer/-/merge_requests/2909>

adaptivedemux2: Fix download helper with libsoup 3.0.x

libsoup 3.0.x dispatches using a single source attached when the session
is created, so we need to create the session with the same context that
our download thread is later using.

2.74 or 3.1 will dispatch a response using the context which sent the
request. However, for any context other than the one that created the
session, this will also create and destroy sources, so there's still
some slight performance benefit.

Fixes: https://gitlab.freedesktop.org/gstreamer/gstreamer/-/issues/1384
Part-of: <https://gitlab.freedesktop.org/gstreamer/gstreamer/-/merge_requests/2913>

kmssink: add fd property

This allows an application to provide their own opened DRM device
fd handle to kmssink.  For example, an application can lease
multiple fd's from a DRM master to display on different CRTC
outputs at the same time with multiple kmssink instances.

Specifying the fd property is not allowed when driver-name
and/or bus-id properties are specified.

Part-of: <https://gitlab.freedesktop.org/gstreamer/gstreamer/-/merge_requests/2807>

d3d11: Use CRITICAL_SECTION instead of GRecMutex

The GRecMutex abstraction (and heap allocation happens in GLib)
is unnecessary for this plugin.

Part-of: <https://gitlab.freedesktop.org/gstreamer/gstreamer/-/merge_requests/2914>

adaptivedemux2: Implement send_event() vfunc

Handle select-streams and seek events in an element
level send_event() vfunc, so they can be received
before any source pads are created.

This allows preferred streams to be selected before
segment downloading starts.

Part-of: <https://gitlab.freedesktop.org/gstreamer/gstreamer/-/merge_requests/2912>

webrtc: doc: Cleanup Since markers

We hide some newly public symbols when moving some objects to the library
which was not really correct in the end. Properly add the Since markers
as needed.

Part-of: <https://gitlab.freedesktop.org/gstreamer/gstreamer/-/merge_requests/2891>

webrtc:ice: Reindent header

Part-of: <https://gitlab.freedesktop.org/gstreamer/gstreamer/-/merge_requests/2891>

webrtc:ice: Mark protected fields as such

Part-of: <https://gitlab.freedesktop.org/gstreamer/gstreamer/-/merge_requests/2891>

webrtc: Fix documentaton moving symbols in the right pages

As those symbols are documented in a 'fwd' header smart indexing in
hotdoc wasn't working.

Part-of: <https://gitlab.freedesktop.org/gstreamer/gstreamer/-/merge_requests/2891>

webrtc:ice: Avoid using GArray in public API

And use plain null terminated C arrays instead.

Part-of: <https://gitlab.freedesktop.org/gstreamer/gstreamer/-/merge_requests/2891>

webrtc:ice: Fix some annotations

Part-of: <https://gitlab.freedesktop.org/gstreamer/gstreamer/-/merge_requests/2891>

webrtc: Make GstWebRTCIceCandidateStats a Boxed type

Part-of: <https://gitlab.freedesktop.org/gstreamer/gstreamer/-/merge_requests/2891>

webrtc:ice: Add padding to structures

Part-of: <https://gitlab.freedesktop.org/gstreamer/gstreamer/-/merge_requests/2891>

cuda: Hide runtime compiler related header and symbols

That's already abstracted via gst_cuda_nvrtc_compile() method
and therefore, we do not need to expose such symbols yet.

Part-of: <https://gitlab.freedesktop.org/gstreamer/gstreamer/-/merge_requests/2884>

aggregator: Improve debug output to better understand why pads are not ready or can't accept more data

Part-of: <https://gitlab.freedesktop.org/gstreamer/gstreamer/-/merge_requests/2907>

vajpegdec: Fix memory leak.

Free value content.

Part-of: <https://gitlab.freedesktop.org/gstreamer/gstreamer/-/merge_requests/2899>

vajpegdec: Fix 4:2:2 for i965.

Since i965 uses NV12 either for 4:2:0 and 4:2:2 chroma (using an
internal color conversion) the chroma validation has to be shortcut.

Part-of: <https://gitlab.freedesktop.org/gstreamer/gstreamer/-/merge_requests/2899>

va: libs: Add Y42B format.

Enable 4:2:2 JPEG decoding in iHD.

Part-of: <https://gitlab.freedesktop.org/gstreamer/gstreamer/-/merge_requests/2899>

va: Validate JPEG subsampling configurations.

There are Mesa Gallium drivers that report subsampling but without
any pixel format. So, strip out these subsamplings.

Part-of: <https://gitlab.freedesktop.org/gstreamer/gstreamer/-/merge_requests/2899>

va: gst_va_create_raw_caps_from_config() may return NULL.

This patch verifies if the function returns NULL in the caller.

Part-of: <https://gitlab.freedesktop.org/gstreamer/gstreamer/-/merge_requests/2899>

decklink: Validate video input format

Part-of: <https://gitlab.freedesktop.org/gstreamer/gstreamer/-/merge_requests/2892>

decklink: Auto-detect 10-bit YUV

Part-of: <https://gitlab.freedesktop.org/gstreamer/gstreamer/-/merge_requests/2892>

examples: gtk: Fix MSVC build

gtk-play.c(370): error C2375: 'rewind_button_clicked_cb': redefinition;
different linkage

Part-of: <https://gitlab.freedesktop.org/gstreamer/gstreamer/-/merge_requests/2871>

base: video-converter: add accessors for input and output formats

Part-of: <https://gitlab.freedesktop.org/gstreamer/gstreamer/-/merge_requests/2905>

vaapi: plugin: Handle when no encoders/decoders available.

Nouveau driver currently only exposes postproc entry. But
vaapidecodebin is registered independent if there are decoders or not,
exposing a segmentation fault.

This patch removes the encoder/decoder/codec arrays if no entries are
found, and if no decoders are found vaapidecodebin is not
registered. Also for vaapipostproc if no postproc entry is found.

Also, if general decoder, used by vaapidecodebin, doesn't have a sink
pad string, don't register the glib type.

Fixes: #1349
Part-of: <https://gitlab.freedesktop.org/gstreamer/gstreamer/-/merge_requests/2865>

doc: Fix typo

Part-of: <https://gitlab.freedesktop.org/gstreamer/gstreamer/-/merge_requests/2903>

gst-play: missing cleanup for g_autoptr

Without this change cleanup function for g_autoptr is not defined for
GstPlayMediaInfo, GstPlaySignalAdapter, GstPlayVideoRenderer,
GstPlayVideoOverlayVideoRenderer and GstPlayVisualization. Cleanup
function was defined in gstplay.h, but missing in other header files.

Part-of: <https://gitlab.freedesktop.org/gstreamer/gstreamer/-/merge_requests/2888>

splitmuxsrc: Stop pad task before cleanup

When stopping the element, make sure the pad task
is stopped before destroying the part readers.

Closes a race where the pad task might access
a freed pointer.

Also add a guard against this sort of thing
by holding a ref to the reader in the pad loop.

Part-of: <https://gitlab.freedesktop.org/gstreamer/gstreamer/-/merge_requests/2901>

qtdemux: Avoid crash on reconfiguring.

When reconfiguring a stream that never created
an output pad, don't access a NULL GstPad pointer.

Part-of: <https://gitlab.freedesktop.org/gstreamer/gstreamer/-/merge_requests/2869>

qtdemux: Set parsed=true on ONVIF Timed Metadata caps

Inside MP4 the metadata must be properly parsed into frames and in
order.

Part-of: <https://gitlab.freedesktop.org/gstreamer/gstreamer/-/merge_requests/2897>

meson: d3d11: Fix missing header install

Install gstd3d11compile.h header file

Part-of: <https://gitlab.freedesktop.org/gstreamer/gstreamer/-/merge_requests/2896>

v4l2bufferpool: Fix debug trace

The tracing of the buffer pointer was done on the secondary pointer, which
does not match with other traces of the same buffer. This made the trace
confusing and less useful.

Fixes #1379

Part-of: <https://gitlab.freedesktop.org/gstreamer/gstreamer/-/merge_requests/2893>

adaptivedemux2: Fix for period switching in live streams

When playing live, it's possible that one stream reaches
the end of the available playback window and goes to sleep
waiting for a manifest update, and the manifest update
introduces a new period. In that case, the sleeping
stream needs to wake up and go 'properly' EOS before we
can advance the input to the new period.

Accordingly, make sure that a stream's last_ret value
is not marked as EOS if it's just sleeping waiting for a live
manifest update.

Also fix the output loop to go back and re-check if it's
time to switch to the next period after dequeuing and
discarding an EOS event.

https://livesim.dashif.org/livesim/periods_20/testpic_2s/Manifest.mpd

Part-of: <https://gitlab.freedesktop.org/gstreamer/gstreamer/-/merge_requests/2895>

videoconvert,videoscale: Do conversion in videoconvert and scaling in videoscale

Keep behaving the same as before videoconvertscale port

Part-of: <https://gitlab.freedesktop.org/gstreamer/gstreamer/-/merge_requests/2778>