Crossfading is a bit more complex than just having two pads with the
right keyframes as the blending is not exactly the same.
The difference is in the way we compute the alpha channel, in the case
of crossfading, we have to compute an additive operation between
the destination and the source (factored by the alpha property of both
the input pad alpha property and the crossfading ratio) basically so
that the crossfade result of 2 opaque frames is also fully opaque at any
time in the crossfading process, avoid bleeding through the layer
blending.
Some rationnal can be found in https://phabricator.freedesktop.org/T7773.
https://bugzilla.gnome.org/show_bug.cgi?id=784827
Found on rpi when gpu_mem is too low so there is not enough memory to
create the eglimage. But still gst_buffer_pool_acquire_buffer succeeded.
And it leads to a CRITICAL assert:
gst_egl_image_get_image: assertion 'GST_IS_EGL_IMAGE (image)' failed
https://bugzilla.gnome.org/show_bug.cgi?id=785518
Otherwise check_events() will not remove the GAP event (as the queue
tail is not the event anymore but the GAP buffer), then the GAP buffer
is handled, then the GAP event is handled again, ... forever.
Avoids dereferencing dead objects
What happens in the autovideosink case is that context 1 is created and
destroyed before all the async operations hae executed on the associated
window. When the delayed operations execute, they then reference dead
objects and crash.
We fix this by keeping refs over all async operations so the object
cannot be deleted while async operations are in flight.
https://bugzilla.gnome.org/show_bug.cgi?id=782379
This is a workaround for a regression introduced by
f4190a49c0
( adaptivedemux: Check live seeking range more often )
The goal of the previous commit was to be able to cope with non-1.0
rates on live streams which have a "seeking window" (i.e. the server
keeps around quite a bit of the live stream so you can seek back into
it).
Without that commit, two different kind of issues would happen:
* When doing reverse playback, you would never check whether you
are outside of the seekable region. And would then continuously
try to download fragments that are no longer present.
* When doing fast forward, you would end up requesting fragments
which are not present yet.
In order to determine whether one was *really* outside of the seekable
window, we check whether the current stream position is still
within the seekable region.
The *problem* though with that commit is that it assumes that subclasses
will return continuously updated seeking ranges (i.e. dependent on the
current time), which is *NOT* the case.
For example:
* dashdemux does use the current UTC to determine the seekable region
* hlsdemux uses the values from the last updated manifest
Therefore if one downloads fragments faster than realtime, for HLS
we would end up at the end of the last manifest seekable range, and
the previous commit would consider the stream as being ended... which
is not the case.
In the long run, we need to figure out a way to cope with non-1.0
rates on live streams for all types of stream (including HLS).
https://bugzilla.gnome.org/show_bug.cgi?id=783075
This is a race that was exposed by the {hls|dash}.scrub_forward_seeking
validate test.
The "race" is that a subclass might want to change format, causing
a new stream to be created (but not exposed/switched yet) and put on the
prepared_streams list. That stream will have values (including pending
segment) from the pre-seek state.
Before the stream is exposed/switched, a new seek comes in and the stream
values get updated ... but the ones that will be changed don't get updated
causing them to push out wrong segments once they are exposed.
https://bugzilla.gnome.org/show_bug.cgi?id=773159
Add a function to install the default RGBA pad templates,
but don't make them required so that there can be
GstGLFilter sub-classes with different input/output
caps if they want. Remove the hard-coded RGBA restriction in
the set_caps_features call, as it will be taken care
of by intersecting with the pad templates.
Update all the sub-classes to match
Adaptive demuxers are special demuxers that runs their own sources
internally. In this patch we flag the demuxer as being a source in order
to receive the downstream events. We then handle the EOS event by
resetting the internal state and pushing EOS on all pads. This handling
is done asynchronously to avoid blocking user thread.
https://bugzilla.gnome.org/show_bug.cgi?id=723868
When the pad has received EOS, its buffer may still be mixed
any number of times, when the pad's framerate is inferior
to the output framerate.
This was introduced by my patch in
https://bugzilla.gnome.org/show_bug.cgi?id=782962, this patch
also correctly addresses the initial issue.
On the raspberry pi no pkg-config file is provided for the bcm_host
library. We are using AC_CHECK_LIB to detect this lib with autotools,
cc.find_library() library is a closer meson equivalent.
https://bugzilla.gnome.org/show_bug.cgi?id=784026
We have to pass the "height" as height = vmeta->offset[1] / width to the
API, which of course does not work well for formats with only a single
plane. Use the whole memory size instead of the offset in that case.
Previous commit let demux call gst_uri_downloader_cancel() on _demux_reset().
Note that, _demux_reset() called during PAUSED_TO_READY and READY_TO_PAUSED.
And, it will set "cancelled" on uridownloader which blocks the use of
uridownloader. The issue is that, subclass can use the uridownloader not only
live streaming for manifest update, but also for fetching another manifests
such as variant and rendition m3u8 of hls streaming. So to unblock it,
demux should clear "cancelled" before processing initial manifest.
https://bugzilla.gnome.org/show_bug.cgi?id=783401
before broadcasting preroll.
The deadlock was as follows:
-> The subclass pushes a buffer on a newly-created stream in T1
-> We take the preroll lock in T1, to handle_preroll
-> The demuxer is stopped in T2, we take the MANIFEST_LOCK
-> T1 starts blocking because it received a reconfigure event
and needs to take the MANIFEST_LOCK
-> T2 deadlocks because it now wants the preroll_lock.
https://bugzilla.gnome.org/show_bug.cgi?id=783255
Make sure the manifest update loop is stopped before proceeding with the
resetting of the manifest data. Otherwise, the updates loop will try to
use it and it leads to a segfault
https://bugzilla.gnome.org/show_bug.cgi?id=783028
As we release the MANIFEST_LOCK in stop_tasks,
demux->priv->old_streams can be set, we need to free these
otherwise we may end up trying to dispose elements in the
READY state.
https://bugzilla.gnome.org/show_bug.cgi?id=783256
When an accurate seek is requested on a live stream, only requests the
exact value for the "starting position" (i.e. start in forward playback
and stop in reverse playback).
https://bugzilla.gnome.org/show_bug.cgi?id=782698
And to make that 100% obvious, only use variables declared within the
switch cases instead of function-wide ones.
Also remove useless one-time-use-only variable.
CID #1409857
The live seeking range was only checked when doing actual seeks. This was
assuming that the rate would always be 1.0 (i.e. the playback would
advance in realtime, and therefore fragments would always be available
since the seeking window moves at the same rate).
With non-1.0 rates, this no longer becomes valid, and therefore we need
to check whether we are still within the live seeking range when advancing.
https://bugzilla.gnome.org/show_bug.cgi?id=783075
What we want is to retry downloading the fragment on 4xx/5xx errors
however returning EOS will cause waiting for a manifest update for live
(which may be a really long time) or stop everything for non-live.
Change that to only return EOS/ERROR once we've reached the error limit.
https://bugzilla.gnome.org/show_bug.cgi?id=776609
GL_RGB565 is sized internal glformat, the corresponding glformat
should be GL_RGB and type is GL_UNSIGNED_SHORT_565. Otherwise will
return GL_INVALID_ENUM when creating texture.
https://bugzilla.gnome.org/show_bug.cgi?id=783066
This ensures that they really get processed in order with
buffers. Just waiting for the queue to be empty is sometimes not
enough as the buffers are dropped from the pad before the result is
pushed to the next element, sometimes resulting in surprising
re-ordering.
In the case an aggregator is created and pads are requested but only
linked later, we end up never updating the upstream latency.
This was because latency queries on pads that are not linked succeed,
so we never did a new query once a live source has been linked, so the
thread was never started.
https://bugzilla.gnome.org/show_bug.cgi?id=757548
When caps changes while streaming, the new caps was getting processed
immediately in videoaggregator, but the next buffer in the queue that
corresponds to this new caps was not necessarily being used immediately,
which resulted sometimes in using an old buffer with new caps. Of course
there used to be a separate buffer_vinfo for mapping the buffer with its
own caps, but in compositor the GstVideoConverter was still using wrong
info and resulted in invalid reads and corrupt output.
This approach here is more safe. We delay using the new caps
until we actually select the next buffer in the queue for use.
This way we also eliminate the need for buffer_vinfo, since the
pad->info is always in sync with the format of the selected buffer.
https://bugzilla.gnome.org/show_bug.cgi?id=780682
The function needs to be unlocked if any data is received, but only
end the first buffer processing on an actual buffer, synchronized events
don't matter on the first buffer processing.
https://bugzilla.gnome.org/show_bug.cgi?id=781673
With the macOS/iOS implementations, the active thread can change
multiple times over the life of a pipeline which would expose a race in
the thread tracking.
Fix by taking a ref on the active thread while the context is active.
https://bugzilla.gnome.org/show_bug.cgi?id=779202
Otherwise fall back to glDrawBuffers. Also check if glReadBuffer exists
before using it.
glDrawBuffer does not exist for GLES, only glDrawBuffers does.
https://bugzilla.gnome.org/show_bug.cgi?id=782376
This commit fixes the following assumptions with live seeking:
1) start was always valid and of type GST_SEEK_TYPE_SET
2) direction was always forward
3) stop should be offsetted when handling non-accurate seeks before
the range start position.
In order to handle more live seeking use-cases (including reverse playback),
only do non-accurate start/stop value clamping for GST_SEEK_TYPE_SET values.
Also add a bit more debugging lines for issues
https://bugzilla.gnome.org/show_bug.cgi?id=782330
When dealing with live streams, we can't rely on GstSegment calculation
since it uses the segment duration to calculate the absolute values.
But since we are dealing with live *and* we know the ranges, we can
compute the absolute seeking values using the range stop (i.e. "now")
as the END position.
Allows seeking back to "live" by using start_type:GST_SEEK_TYPE_END
and start:0
https://bugzilla.gnome.org/show_bug.cgi?id=782228
The allowed live seek ranges returned by subclasses are "inclusive", that is
to say that the "range_stop" value they return is the highest acceptable position
one can seek to (i.e. "now").
Allow seeking to exactly that value
Rationale is to allow the manifest update task to continue running while
seeks are occurring. Otherwise, if the user reliably performs a seek
before the manifest is updated, then as the manifest task is reset on
seeks (and thus the time to wait between manifest updates), the manifest
would never be updated.
This fix makes the manifest update task free-running and continously
update even during seeks.
Some actions (Qos, reconfigure, ...) might take place before we finish pushing out flush_start.
One problem would be that:
1) The QOS handling in adaptivedemux takes the MANIFEST LOCK
That QOS event comes from basesink with its PREROLL_LOCK taken
2) FLUSH_START is sent from adaptivedemux with the MANIFEST_LOCK taken and the basesink flushing handler needs to take the PREROLL_LOCK
=> deadlock
https://bugzilla.gnome.org/show_bug.cgi?id=781320
meson's configure_file emits only a comment like /* #undef ... */
for values which are unset in the configuration_data. For
gstglconfig.h, this differs from the autotools build where the
preprocessor definitions are always either 0 or 1. So loop over a
list of variables to set to zero as default.
Also sync up the gstglconfig.h.meson file with the additional
macros defined by the autotools build.
https://bugzilla.gnome.org/show_bug.cgi?id=781043
Windows aren't always removed in time, and it turns out to be
very, very hard to remove a window in a way that's not racy and
not deadlocky. Since the window itself doesn't leak, freeing
the list on object destruction is enough.
https://bugzilla.gnome.org/show_bug.cgi?id=781018
At the moment that demux is waiting manifest update, the target sequence
of fragment was advanced already. So, checking stream_has_next_fragment()
means looking for the next fragment of target fragment.
This might cause unexpected buffering if each fragment has
large duration and manifest is listing only limited number of fragments.
https://bugzilla.gnome.org/show_bug.cgi?id=780494
The GstGLFramebufferClass struct is typedeffed in
gstgl_fwd.h, and having a duplicate elsewhere is
breaking the cerbero build on my OSX machine,
even though it seems to be working in CI.
When entering this code path, we know that:
We received EOS on this pad.
We consumed all its buffers.
In any case, we want to replace vaggpad->buffer with NULL,
otherwise we will end up mixing the same buffer twice.
https://bugzilla.gnome.org/show_bug.cgi?id=781037
If set, the parent is used to proxy need-context messages from
uridownloader's http source in order to get cookies/headers
from the pipeline.
Based on a patch from Philippe Normand
https://bugzilla.gnome.org/show_bug.cgi?id=726314
When there are new pads pending for a bitrate switch, don't allow
EOS through from the old streams. It will be sent when the new pads are
ready, just before the old streams are removed.
This fixes racy bitrate switching with hlsdemux in urisourcebin
where old pads EOS before new pads appear and the entire pipeline can
EOS if those propagate fast enough
libgstgl has a dependency on libgstallocators starting with:
ed1e4c1 glupload: Add support for Vivante DirectTexture uploads
The dependency wasn't stated in the Makefile, so the build could
occasionally fail if you got unlucky with the ordering of build
steps.
https://bugzilla.gnome.org/show_bug.cgi?id=780330
In commit
> 956c4d0 gl/format: use our own GL format enum's instead of gstvideo's
the name and return type of gst_gl_format_from_video_info changed,
but some returns of the old type were missed. Here they are
updated to the correct type.
https://bugzilla.gnome.org/show_bug.cgi?id=780064
gst_player_set_uri_internal shouldn't free suburi which maybe set
by user to load external subtitle before start play. It just need
reset playbin's subutri property to NULL no matter if there was a
previous one or not.
https://bugzilla.gnome.org/show_bug.cgi?id=779453
All code interacting with Objective-C objects should now use Automated
Reference Counting rather than manual memory management or Garbage
Collection. Because ARC prohibits C-structs from containing
references to Objective-C objects, all such fields are now typed
'gpointer'. Setting and gettings Objective-C fields on such a
struct now uses explicit __bridge_* calls to tell ARC about
object lifetimes.
https://bugzilla.gnome.org/show_bug.cgi?id=777847
For duration queries on live streams, adaptivedemux ignores the query.
The problem then is that the query is answered by the downstream
qtdemux element, with the duration of the currently passing fragment.
This commit changes the behaviour of adaptivedemux to answer the duration
queries for live streams, returning GST_CLOCK_TIME_NONE.
https://bugzilla.gnome.org/show_bug.cgi?id=753879
If we need to send EOS on a pad that hasn't prerolled, generate
an error on the bus instead, otherwise the app will have no idea.
Fixes the HLS testFragmentNotFound test, which is waiting
for either EOS or an error.
To ensure that pads have caps when they are exposed, do
the exposing when all pending streams have prerolled an
output buffer, and only then EOS and remove any old pads.
Improves the switching sequence by making caps available
as soon as a pad appears.
With fixes from Seungha Yang <sh.yang@lge.com>
https://bugzilla.gnome.org/show_bug.cgi?id=758257
send_event() of parent class (i.e., GstBinClass) iterates srcpads
to send SEEK event. And performing it per srcpad is inefficient.
So, let's drop duplicated SEEK event by checking seqnum
https://bugzilla.gnome.org/show_bug.cgi?id=776612
- Add overlay video renderer "video-sink" property, so that can be set
- In create_video_sink, it returns video sink instead of always NULL
- Add new renderer_new_with_sink() API to set video sink
https://bugzilla.gnome.org/show_bug.cgi?id=776490
'extern inline' was added in 2fb76c89 for MSVC (it was just
'inline' before), but all of this doesn't really make sense,
the functions are not going to be inlined anyway, and what
'extern inline' means exactly also appears to depend on the
Cxx standard targetted. Let's just remove the 'extern inline'
entirely. At least gcc6 still emits the exact same code as
before anyway. Fixes compilation/linking with gcc 4.8 as
used on L4T on the TK1.
The reason we previously used queue2 was to calculate the download rate,
but that wasn't entirely correct and we therefore calculate it before
queue2. We therefore now just need a simple queue.
../../../../gst-libs/gst/gl/gl.h:57:45: fatal error: gst/gl/gstglcontrolbindingproxy.h: No such file or directory
#include <gst/gl/gstglcontrolbindingproxy.h>
^
No-one's using/depending on it (it would have criticalled and not worked)
and it's causing more problems than it's solving. Store the GMainContext
in the public struct instead for subclasses to optionally use instead of
relying on the push/pop state to be correct.
https://bugzilla.gnome.org/show_bug.cgi?id=775970
This is potentially racy (in the unlikely scenario that we get two
first-time calls to gst_player_error_quark() at the same time). This
should not impact anything in terms of performance since it's only on
the error path.
The call itself could just be inlined by making GST_PLAYER_ERROR be
defined to the g_quark_from_static_string() call, but this feels ugly
from an API perspective.
If a sub class of GstGLContext does not create a group
then it currently crashes:
0 g_atomic_int_get (&share->refcount)
1 _context_share_group_is_shared (context->priv->sharegroup)
2 gst_gl_context_is_shared
3 _default_set_sync_gl
https://bugzilla.gnome.org/show_bug.cgi?id=774518
This is a subblass of VideoFilter but yet does not use any of it's
features. This also fixes issue in case the incoming images have custom
strides as the VideoMeta is no longer ignored.
https://bugzilla.gnome.org/show_bug.cgi?id=775288
When a MSS server hosts a live stream the fragments listed in the
manifest usually don't have accurate timestamps and duration, except
for the first fragment, which additionally stores timing information
for the few upcoming fragments. In this scenario it is useless to
periodically fetch and update the manifest and the fragments list can
be incrementally built by parsing the first/current fragment.
https://bugzilla.gnome.org/show_bug.cgi?id=755036
Calling g_main_context_push_thread and then g_main_context_invoke()
(used by gst_gl_window_send_message_async()) in the same thread will
cause the invoked function to run immediately instead of being delayed.
This had implications for the creation of the OpenGL context not waiting
until the main loop had completely started up and as a result would
sometimes deadlock in short create/destroy scenarios.
https://bugzilla.gnome.org/show_bug.cgi?id=775171
626bcccff9 removed some locks that
allowed the main loop quit to occur before the context was fully
created.
2776cef25d attempted to readd them but
missed the scop of the quit() call.
Also remove the use of g_thread_join() as that's not safe to use when
it's possible to lose the last reference from the GL thread.
https://bugzilla.gnome.org/show_bug.cgi?id=775171
The smallest section ever needs to be at least 3 bytes (i.e. just the short
header).
Non-short headers need to be at least 11 bytes long (3 for the minimum header,
5 for the non-short header, and 4 for the CRC).
https://bugzilla.gnome.org/show_bug.cgi?id=775048
It's been removed and thus compiling anything against GstGLMemoryEGL
would error with:
In file included from gstomxvideodec.c:41:0:
usr/include/gstreamer-1.0/gst/gl/egl/gstglmemoryegl.h:32:41: fatal error: gst/gl/egl/gstglcontext_egl.h: No such file or directory
#include <gst/gl/egl/gstglcontext_egl.h>
^
https://bugzilla.gnome.org/show_bug.cgi?id=774886
Otherwise, when the application reuses the same UIView, we were getting
draw notifications on the previous view/layer's which weren't valid anymore
and were referencing pointers that had been freed.
https://bugzilla.gnome.org/show_bug.cgi?id=753003
This changes the failure case to require a consecutive number of
failures rather than being spread out over the entire stream.
Fixes the case where fetching the manifest was intermittent.
https://bugzilla.gnome.org/show_bug.cgi?id=774177
For formats that need to update the manifest to know about new
fragments as they're being written by the server would never receive an
updated fragment list after a seek event
https://bugzilla.gnome.org/show_bug.cgi?id=774177
- xcb is supposedly thread-safe!
videotestsrc ! glimagesink now doesn't spuriously result in a
'call XInitThreads()' error however if anybody else is using X11,
then XInitThreads() still needs to be called and multiple glimagesink's
still need XInitThreads().
Everything still takes libX11 handles as they are compatible with the xcb
variants. Unfortunately we cannot move fully over to xcb due to GLX being
entirely based on Xlib. It's also impossible to transform a xcb_connection
to a Display which means we require X11 handles.
The spec allows the core/compatibility profiles to be used
with #version 150.
Also tighten up the tests to check for default profiles being chosen
correctly.
The change to use GST_EXPORT for symbols under Windows requires
GST_EXPORTS for internal use, and that is also needed under Autotools.
The same thing is done for gstreamer-1.0.dll in -core.
The calling convention may be deprecated, but we still need it for
OpenGL. The build issue was caused by an incorrect syntax being used for
the WINAPI (__stdcall) prototype in function pointers which was accepted
by GCC but is rejected by MSVC.
With MSVC, this gives the following warning:
warning C4305: 'function': truncation from 'double' to 'gfloat'
Apparently, MSVC does not figure out what type to use for constants
based on the assignment. This warning is very spammy, so let's try to
fix it.
At minimum, we only need to glFlush() if we are in a shared GL context
environment. Move the glFinish() to when the actual wait is requested
which may be never. Improves the throughput on older GL systems without
GL3/GLES3 and/or fence sync objects.
In order to calculate the *actual* bitrate for downloading a fragment
we need to take into account the time since we requested the fragment.
Without this, the bitrate calculations (previously reported by queue2)
would be biased since they wouldn't take into account the request latency
(that is the time between the moment we request a specific URI and the
moment we receive the first byte of that request).
Such examples were it would be biased would be high-bandwith but high-latency
networks. If you download 5MB in 500ms, but it takes 200ms to get the first
byte, queue2 would report 80Mbit/s (5Mb in 500ms) , but taking the request
into account it is only 57Mbit/s (5Mb in 700ms).
While this would not cause too much issues if the above fragment represented
a much longer duration (5s of content), it would cause issues with short
ones (say 1s, or when doing keyframe-only requests which are even shorter)
where the code would expect to be able to download up to 80Mbit/s ... whereas
if we take the request time into account it's much lower (and we would
therefore end up doing late requests).
Also calculate the request latency for debugging purposes and further
usage (it could allow us to figure out the maximum request rate for
example).
https://bugzilla.gnome.org/show_bug.cgi?id=733959https://bugzilla.gnome.org/show_bug.cgi?id=772330
Using g_thread_join() in _finalize() handlers may result in a deadlock
joining the current thread when the last reference is held by a signal
handler.
e.g.:
error 'Resource deadlock avoided' during 'pthread_join (pt->system_thread, NULL)'
The backtrace looks like this:
[...]
g_thread_join ()
gst_gl_window_finalize ()
gst_gl_window_x11_finalize ()
g_object_unref ()
g_value_unset ()
g_signal_emit_valist ()
g_signal_emit ()
gst_gl_window_send_mouse_event ()
gst_gl_window_mouse_event_cb ()
g_main_dispatch ()
[..]
g_main_loop_run ()
gst_gl_window_navigation_thread ()
g_thread_proxy ()
start_thread ()
clone ()
We cannot set the x, y coordinate of the video frame at the dispmanx at
this point. We need to teach dispmanx backend to understand about
set_render_rectangle API to draw a video with other UI.
This patch keeps the current behavior which places video frame at the
center of the display if there is no set_render_rectangle call to the
dispmanx window.
https://bugzilla.gnome.org/show_bug.cgi?id=766018
e.g. passing with_gl_api=gles2 would still build the glx code but not be
linking against the libGL library which is where the glX* functions are
located and would result in a linker error.
Solved by checking for the libGL library if either opengl or glx may be
needed and then disabling the corresponding deps as requested.
Allowing us to tell GstPad why we are failing an event, which might
be because we are 'flushing' even if the sinkpad is not in flush state
at that point.
The tests were broken since 91fea30, which changed glupload to return
GST_GL_UPLOAD_RECONFIGURE if the texture target in the input buffers doesn't
match the texture-target configured in the output caps.
This commit fixes that and adds more checks for the new behaviour.
Now when used with video/x-raw as input, the GLMemoryUpload method checks for
->tex_target in input GLMemory(es) and sets the output texture-target
accordingly.
Fixes video corruption with a pipeline like avfvideosrc ! video/x-raw !
glimagesink where on macos avfvideosrc pushes RECTANGLE textures but glupload
was configuring texture-target=2D as output.
The headers passed as parametter are relative to the build dir
basically "../subproject/gst-plugins-bad/gst-libs/gst/mpegts/XXX.h"
but that does not match what is needed at build time when building as
subproject, also we always add current dir as include_dir so we are
safe including directly.
And link mpegtsdemux against the 'math' library as it is needed.
Don't set the chosen texture-target into the wrong structure.
The input caps may not be writable, and in any case - the
intention was to configure the othercaps. Also, remove an
extra unref - the othercaps ref is consumed by
gst_caps_make_writable already.
And scale the bitrate with the absolute rate (if it's bigger than 1.0) to get
to the real bitrate due to faster playback.
This allowed in my tests to play a stream with 10x speed without buffering as
the lowest bitrate is chosen, instead of staying/selecting the highest bitrate
and then buffering all the time.
It was previously disabled for not very well specified reasons, which seem to
be not valid anymore nowadays.
Prevent the manifest update loop from looping endlessly
after a seek event, by clearing the variable that tells
the task function not to immediately exit.
The new streams should not be exposed until all streams are done with the
current fragment. The old code is incorrect and actually only checked the
current stream. Fix this by properly checking all streams.
Also, ignore the current stream. The code is only reached when the current
stream finished downloading and since
07f49f15b1 ("adaptivedemux: On EOS, handle it
before waking download loop") download_finished is set after
gst_adaptive_demux_stream_advance_fragment_unlocked() is called.
Without this HLS playback with multiple streams is broken, because the new
streams are never exposed.
https://bugzilla.gnome.org/show_bug.cgi?id=770075
https://github.com/mesonbuild/meson
With contributions from:
Tim-Philipp Müller <tim@centricular.com>
Matej Knopp <matej.knopp@gmail.com>
Jussi Pakkanen <jpakkane@gmail.com> (original port)
Highlights of the features provided are:
* Faster builds on Linux (~40-50% faster)
* The ability to build with MSVC on Windows
* Generate Visual Studio project files
* Generate XCode project files
* Much faster builds on Windows (on-par with Linux)
* Seriously fast configure and building on embedded
... and many more. For more details see:
http://blog.nirbheek.in/2016/05/gstreamer-and-meson-new-hope.htmlhttp://blog.nirbheek.in/2016/07/building-and-developing-gstreamer-using.html
Building with Meson should work on both Linux and Windows, but may
need a few more tweaks on other operating systems.
Multiple threads may be accessing the wayland fd at the same time which
requires the use of special wayland API to deal with to ensure nobody
will steal reads and cause a stall for anyone else.
This allows to gradually download part of a fragment when the final size is
not known and only a part of it should be downloaded. For example when only
the moof should be parsed and/or a single keyframe should be downloaded.
https://bugzilla.gnome.org/show_bug.cgi?id=741104
When connect to qmlglsrc, x11 event loop will be replace by qt event loop
which will cause the window cannot receive event from xserver, such as resize
https://bugzilla.gnome.org/show_bug.cgi?id=768160
Makes infinitely more sense and implementation were expecting that behaviour
anyway and would enter a resize, draw, resize, draw, ... cycle instead of only
resizing once.
This helps catch those 404 server errors in live streams when
seeking to the very beginning, as the server will handle a
request with some delay, which can cause it to drop the fragment
before sending it.
https://bugzilla.gnome.org/show_bug.cgi?id=753751
To allow adaptivedemux to make retry decisions, it needs to know what
sort of HTTP error has occurred. For example, the retry logic for a
410 error is different from a 504 error.
https://bugzilla.gnome.org/show_bug.cgi?id=753751
Some derived classes (at least dashdemux) expose a seeking range
based on wall clock. This means that a subsequent seek to the start
of this range will be before the allowed range.
To solve this, seeks without the ACCURATE flag are allowed to seek
before the start for live streams, in which case the segment is
shifted to start at the start of the new seek range. If there is
an end position, is is shifted too, to keep the duration constant.
https://bugzilla.gnome.org/show_bug.cgi?id=753751
There's no need for the jump to an extra thread in most cases, especially
when relying solely on a shader to render. We can use the provided
render_to_target() functions to simplify filter writing.
Facilities are given to create fbo's and attach GL memory (renderbuffers
or textures). It also keeps track of the renderable size for use with
effective use with glViewport().
Make state changes of internal elements more reliable by locking
their state, and ensuring that they aren't blocked pushing data
downstream before trying to set their state.
Add a boolean to avoid starting tasks when the main
thread is busy trying to shut the element down.
Try harder to make switching pads work better by
making sure concurrent downloads are finished before exposing
a new set of pads.
Release the manifest lock when signalling no-more-pads, as
that can call back into adaptivedemux again
If other stream fragments are still downloading but new streams
have been scheduled, don't expose them yet - wait until the last
one finishes. Otherwise, we can cancel a partially downloaded
auxilliary stream and cause a gap.
Drop the manifest lock when performing actions that might
call back into adaptivedemux and trigger deadlocks, such
as adding/removing pads or sending in-band events (EOS).
Unlock the manifest lock when changing the child bin state to
NULL, as it might call back to acquire the manifest lock when
shutting down pads.
Drop the manifest lock while pushing events.
In the case of KEY_UNIT and TRICKMODE_KEY_UNITS seeks, we want to
"snap" to the closest fragment.
Without this, we end up pushing out a segment which does not match
the first fragment timestamp being pushed out, resulting in one or
more buffers being eventually dropped because they are out of segment.
Calling glUniformMatrix before the shader is bound is invalid and
would result in errors like:
GL_INVALID_OPERATION in glUniformMatrix(program not linked)
Move glUniformMatrix() to after the gst_gl_shader_use() call.
Rather than assuming something. e.g. zerocopy on iOS with GLES3 requires
the use of Luminance/Luminance Alpha formats and does not work with
Red/RG textures.
Take the used texture type from the memory instead.
Fixes conversion from multi-planar YUV formats with two components per plane
(NV12, NV21, YUY2, UYVY, GRAY16_*, etc) with Luminance Alpha input textures.
This is also needed for zerocopy decoding on iOS with GLES 3.x.
The intention was to assert if both maj and min were NULL (as there would be no
point calling the function). Instead if either maj or min were NULL, the assert
would occur.
Fix that.
Newer devices require using a different GLSL extension for accessing
external-oes textures in a shader using the texture() functions.
While the GL_OES_EGL_image_external_essl3 should supposedly be supported
on a any GLES3 android device, the extension was defined after a lot of the
older drivers were built so they will not know about it. Thus there are two
possible interpretations of which of texture[2D]() should be supported for
external-oes textures. Strict adherence to the GL_OES_EGL_image_external
extension spec which uses texture2D() or following GLES3's pattern, also
allowing texture() as a function for accessing external-oes textures
This adds another mangling pass to convert
#extension GL_OES_EGL_image_external : ...
into
#extension GL_OES_EGL_image_external_essl3 : ...
on GLES3 and when the GL_OES_EGL_image_external_essl3 extension is supported.
Only uses texture() when the GLES3 and the GL_OES_EGL_image_external_essl3
extension is supported for external-oes textures.
Uses GLES2 + texture2D() + GL_OES_EGL_image_external in all other external-oes
cases.
https://bugzilla.gnome.org/show_bug.cgi?id=766993
Otherwise we will leak GstGLContext's when adding the same context more than
once.
Fixes a regression caused by 5f9d10f603 in the
gstglcontext unit test that failed with:
Assertion 'tmp == NULL' failed
Until now we would start the task when the pad is activated. Part of the
activiation concist of testing if the pipeline is live or not.
Unfortunatly, this is often too soon, as it's likely that the pad get
activated before it is fully linked in dynamic pipeline.
Instead, start the task when the first serialized event arrive. This is
a safe moment as we know that the upstream chain is complete and just
like the pad activation, the pads are locked, hence cannot change.
https://bugzilla.gnome.org/show_bug.cgi?id=757548
If the input buffer is after the end of the output buffer, then waiting
for more data won't help. We will never get an input buffer for this point.
This fixes compositing of streams from rtspsrc.
https://bugzilla.gnome.org/show_bug.cgi?id=766422
Provide a function to get the affine matrix in the meta in terms of NDC
coordinates and use as a standard opengl matrix.
Also advertise support for the affine transformation meta in the allocation
query.
The URI downloader is creating the source element with
gst_element_factory_make() that returns a floating reference that nobody
is consuming. This is causing problems in WebKit, where the smart
pointers used to take references of the source elment get confused and
end up consuming the floating reference and then releasing the element,
which usually crashes because the URI downloader still tries to use its
src element. See https://bugs.webkit.org/show_bug.cgi?id=144040.
This commit adds two helper functions to ensure and destroy the source
element, to make the code simpler and less error prone. The ensure
method takes care of checking if we can reuse the existing one or we
need to create a new one, taking always its ownership. The destroy
method simply avoids duplicated code to set the source to NULL state and
then unref it.
https://bugzilla.gnome.org/show_bug.cgi?id=766053
The gst_adaptive_demux_wait_until() function can be woken up either
by its end_time being reached, or from other threads that want to
interrupt the waiting thread.
If the thread is interrupted, it needs to cancel its async clock callback
by unscheduling the clock callback. However, the callback task might already
have been activated, but is waiting for the mutex to become available. In this
case, the call to unschedule does not stop the callback from executing.
The solution to this second issue is to use a reference counted object that
is decremented by both the gst_adaptive_demux_wait_until() function and the
call to gst_clock_id_wait_async (). In this way, the GstAdaptiveDemuxTimer
object is only deleted when both the gst_adaptive_demux_wait_until() function
and the async callback are finished with the object.
https://bugzilla.gnome.org/show_bug.cgi?id=765728
Because current GstEGLImageMemory does not inherit GstGLMemory, GLUpload
allocates additional GLMemory and upload the decoded contents from the decoder
which uses EGLImage (e.g. gst-omx in RPi).
This work adds GstGLMemoryEGL to avoid this overhead. Decoders allocate
GstGLMemoryEGL and decode its contents to the EGLImage of GstGLMemoryEGL. And
GLUpload uses this memory without allocation of additional textures and blit
operations.
[Matthew Waters]: gst-indent the sources and fix a critical retreiving the egl
display from the memory.
https://bugzilla.gnome.org/show_bug.cgi?id=760916
Allows creating wrapped memories with GstGLAllocationParams.
The wrapped pointers will be set in the parameters before being passed
to the memory allocation function.
Some platforms provide an old version of GLES2/gl2.h and GLES2/gl2ext.h that
will fail when including GLES3/gl3.h due to missing typedef's.
Seen on the RPi.
Adds a new function to mpegts lib to create a iso639 language
descriptor from a language and use it in mpegtsmux to add
a language descriptor to audio streams that have a language set.
https://bugzilla.gnome.org/show_bug.cgi?id=763647
This fixes a race where we check if there is a clock, then it get
removed and we endup calling gst_clock_new_single_shot_id() with a NULL
pointer instead of a valid clock and also calling gst_object_unref()
with a NULL pointer later.
https://bugzilla.gnome.org/show_bug.cgi?id=757548
There are several places in adaptivedemux where it waits for
time to pass, for example to wait until it should next download
a fragment. The problem with this approach is that it means that
unit tests are forced to execute in realtime.
This commit replaces the use of g_cond_wait_until() with single
shot GstClockID that signals the condition variable. Under normal
usage, this behaves exactly as before. A unit test can replace the
system clock with a GstTestClock, allowing the test to control the
timing in adaptivedemux.
https://bugzilla.gnome.org/show_bug.cgi?id=762147
A realtime clock is used in many places, such as deciding which
fragment to select at start up and deciding how long to sleep
before a fragment becomes available. For example dashdemux needs
sample the client's estimate of UTC when selecting where to start
in a live DASH stream.
The problem with dashdemux calculating the client's idea of UTC is
that it makes it difficult to create unit tests, because the passage
of time is a factor in the test.
This commit changes dashdemux and adaptivedemux to use the
GstSystemClock, so that a unit test can replace the system clock when
it needs to be able to control the clock.
This commit makes no change to the behaviour under normal usage, as
GstSystemClock is based upon the system time.
https://bugzilla.gnome.org/show_bug.cgi?id=762147
Don't create many short lived locks/conds in gst_gl_window_send_message. This is
a micro optimization to save a bunch of pthread_* calls which are expensive on
OSX/iOS and possibly other platforms.
We weren't using the result of find_best_format at all.
Also, move the find_best_format usage to the default update_caps() to make
sure that it is also overridable.
https://bugzilla.gnome.org/show_bug.cgi?id=764363
The subsampling_x, subsampling_y, bit_depth, color_space and color_range
fileds are moved from GstVp9FrameHdr to the global GstVp9Parser structure.
These fields are only present in keyframe or intra-only frame, no need to
duplicate them for inter-frames. This is an ABI change.
https://bugzilla.gnome.org/show_bug.cgi?id=764370
while calling eglCreateImage without a GL context current in the executing
thread works on the RPi, some other implementations will return errors.
Marshall the eglCreateImage to the GL thread to appease these implementations.
There are numerous slight differences required between Desktop GL and GLES3 for
multiple render targets.
1. gl_FragData doesn't exist at all and one is required to use
'layout (location = ?) out ...' instead.
2. gl_FragColor doesn't exist, same as 1
3. texture2D() has been deprecated
Fortunately most of these have been taken care of with GL3 and the shader
mangling already exists so just expand the conditions they are used in. The
gl_FragData issue requires a new mangle pass though. We also use this new
pass on desktop GL for consistency.
By default, reading GL_RED or GL_RG us unsupported by glReadPixels unless
exposed through GL_COLOR_READ_IMPLEMENTATION_FORMAT/TYPE. This allows
downloading multiple-planar video frames where possible.
There are some cases where it's needed for binding in/out variables in shaders.
e.g. glsl 150 (gl 3.2) doesn't support the 'layout (location = ?)' specifiers in
the shader source so we have to bind them ourselves.
Happens e.g. if a RECONFIGURE event is sent from downstream while we're
switching pads at this very moment. The old pad is gone and the stream has a
new pad.
https://bugzilla.gnome.org/show_bug.cgi?id=764404
Previously, while allocating the pad number for a new pad, aggregator was
maintaining an interesting relationship between the pad count and the pad
number.
If you requested a sink pad called "sink_6", padcount (which is badly named and
actually means number-of-pads-minus-one) would be set to 6. Which means that if
you then requested a sink pad called "sink_0", it would be assigned the name
"sink_6" again, which fails the non-uniqueness test inside gstelement.c.
This can be fixed by instead setting padcount to be 7 in that case, but this
breaks manual management of pad names by the application since it then becomes
impossible to request a pad called "sink_2". Instead, we fix this by always
directly using the requested name as the sink pad name. Uniqueness of the pad
name is tested separately inside gstreamer core. If no name is requested, we use
the next available pad number.
Note that this is important since the sinkpad numbering in aggregator is not
meaningless. Videoaggregator uses it to decide the Z-order of video frames.
Without the GST_GL_API_GLES2 bit set, we will not even attempt to look
for the function pointers in the core library and will fallback to
glFlush/glFinish.
If the user uploads their own texture without setting the unpack length, then
then the result will have the appearance of stride mismanagement due to
an incorrect row length.
If we are given caps with extra features (like the overlay composition
features), we can only deal with that when we are in passthrough mode.
Previously we were bailing entirely and not allowing passthrough filter elements
with things like textoverlay.
Fixes the following pipeline (assuming glfilter supports passthrough):
gl ! textoverlay ! glfilter ! ... ! glimagesinkelement
https://bugzilla.gnome.org/show_bug.cgi?id=763756
When transforming, xplode it out into the necessary caps features both
with and without the passthough features.
Fixes negotiation in the following class of pipelines:
gl ! textoverlay ! glupload ! glimagesinkelement
https://bugzilla.gnome.org/show_bug.cgi?id=763756
GL 1.4 (with GL_ARB_shader_objects) doesn't have glIsProgram or glIsShader
equivalents. As they are simply assertions, skip them when there isn't a
valid function pointer.
This is a regression from since mpegvideoparser was switched to
use the codecparsing library.
The problem is that the high bit of the profile_and_level is used
to specify non-hierarchical profiles and levels. Unfortunately we
were discarding that information.
Expose that escape bit, and use it in the element
https://bugzilla.gnome.org/show_bug.cgi?id=763220
With e38af23044 returning the correct contexts,
gst_gl_display_add_context() was susceptible to causing infinte loops when
adding the same GstGLContext more than once. Fix and add a test for
gst_gl_display_add_context().
Fixes glvideomixer gst-validate tests.
Execute GL calls without marshalling them to the context thread. In the cocoa
and eagl backends calling gst_gl_context_activate is cheap and therefore calling
it on the current thread and serializing GL calls with a per-context lock is
more efficient (faster and has less overhead) than marshalling everything to the
context thread.
This optimization cuts a large overhead in g_poll (continuously waking up the
context thread) and in g_mutex_*/g_cond_* (waiting for results from the context
thread).
When requesting a glcontext (regardless of thread), the result was correct.
However, when requesting current glcontext on a specific thread, it could
come up with a glcontext active on another thread.
https://bugzilla.gnome.org/show_bug.cgi?id=763168
Maxsize is initialized once and should never change. Allocating data
should have no impact on the selected max size for this memory. This
causing memory map failure as the maxsize would become smaller then
size. This happened when using direct rendering in avviddec on GL that
does not support PBO transfer.
https://bugzilla.gnome.org/show_bug.cgi?id=763045
When the start_type is GST_SEEK_TYPE_NONE for a forward seek
(or stop_type for a reverse) is not set on a snap seeking operation,
the element should use the current position and then snap as requested.
Also fixes uninitialized variable complaint by clang about
'ts' variable.
Intended for use with wrapped contexts that are created shared with gst's
gl contexts in order to manage the internal sharegroup state correctly.
e.g. with caopengllayer (which is used in glimagesink and caopengllayersink
on OS X), we create a CGL context from the gst context and the sharing state
was not being correctly set on either GL context and gst_gl_context_is_shared()
was always returning FALSE.
With 11fb4fff80 only flushing with multiple
shared contexts, the required flush was not occuring causing screen
corruption or stuttering.
Note: this didn't affect GST_GL_API=opengl pipelines
https://bugzilla.gnome.org/show_bug.cgi?id=762620
When caps are already negotiated it should be possible to
select formats other than the one that was negotiated. If downstream
allows alpha video caps and it has already negotiated to a non-alpha
format, caps queries should still return the alpha caps as a possible
format as caps renegotiation can happen.
Includes tests (for compositor) to check that caps queries done after
a caps has been negotiated returns complete results
https://bugzilla.gnome.org/show_bug.cgi?id=757610
Expose the expose() and set_render_rectangle() methods. These are useful for
proper functioning of the video overlay in various situations and toolkits.
H.265 7.4.7.1 says:
> When slice_deblocking_filter_disabled_flag is not present, it is
> inferred to be equal to pps_deblocking_filter_disabled_flag.
https://bugzilla.gnome.org/show_bug.cgi?id=762351
Usually gl debug is initialized in gst_gl_context_create_thread.
But this function is not used when using the GstGLContextGPUProcess
from ChromiumGStreamerBackend.
Received signal 11 SEGV_MAPERR 000000000000
gst_debug_category_get_threshold
gst_gl_insert_debug_marker
gst_gl_base_filter_gl_start
CC libgstgl_x11_la-gstglcontext_glx.lo
In file included from gstglcontext_glx.c:39:0:
../utils/opengl_versions.h:52:43: error: ‘gles2_versions’ defined but not used [-Werror=unused-const-variable]
static const struct { int major, minor; } gles2_versions[] = {
^~~~~~~~~~~~~~
Invoke the callback right away when called on the context thread. Removes
overhead when nesting libgstgl calls (for example when working with the sync
meta).
CPU waits are more expensive and are only required if the CPU is ever going to
access the data. GPU waits perform inter-context synchronisation and are cheaper
as they don't require CPU intervention.
Adaptive demuxers need to start downloading from specific positions
(fragments) for every stream, this means that all streams can snap-seek
to a different position when requested. Snap seeking in this case will
be done in 2 steps:
1) do the snap seeking on the pad that received the seek event and
get the final position
2) use this position to do a regular seek on the other streams to
make sure they all start from the same position
More arguments were added to the stream_seek function, allowing better control
of how seeking is done. Knowing the flags and the playback direction allows
subclasses to handle snap-seeking.
And also adds a new return parameter to inform of the final
selected seeking position that is used to align the other streams.
https://bugzilla.gnome.org/show_bug.cgi?id=759158
All uses of query->context in gstglquery.c assume it exists. We can assume
this as well before unrefing it. Furthermore, gst_object_unref() will just
silently return if it ever were to not exist.
Improves the responsiveness of the pipeline when resources are close/above the
limitations of the hardware.
Any subclass that wishes not to enable qos can do so themselves.
https://bugzilla.gnome.org/show_bug.cgi?id=761519
This reverts commit 96b9666d59.
This reverts commit d11385d167.
This breaks the texture sharing with the applemedia elements as
CVOpenGLESTextureCache seems to have an arbitrary restriction on GLES2 only.
We may need them to transform into a different set of formats.
Fixes YUV->YUV with two glcolorconverts, e.g:
format=I420 ! glcolorconvert ! glcolorconvert ! format=NV12
1.0 / width does not offset by one pixel in rectangular textures (which use
unnormalized coordinates).
Provide the actual pixel offset as a uniform to the shader.
1. Correctly describe what we can caps we can transform to/from.
i.e. no YUV->YUV or GRAY->YUV or YUV->GRAY (except for passthrough).
2. Prefer similar formats and ignore incompatible formats on fixation.
Useful when gst_gl_window.c::gst_gl_window_new is not used.
This is the case when using a custom GstGLWindow.
(ex: GstGLWindowGPUProcess from Chromium)
Allows the subclass to completely override the chosen src caps.
This is needed as videoaggregator generally has no idea exactly
what operation is being performed.
- Adds a fixate_caps vfunc for fixation
- Merges gst_video_aggregator_update_converters() into
gst_videoaggregator_update_src_caps() as we need some of its info
for proper caps handling.
- Pass the downstream caps to the update_caps vfunc
https://bugzilla.gnome.org/show_bug.cgi?id=756207
The function gst_adaptive_demux_stream_update_source() function creates
a new GstPad called internal_pad. This pad is not freed when releasing
the stream.
The solution is to set GST_PAD_FLAG_NEED_PARENT so that the chain
functions do not get called when the pad has no parent and then
remove the parent in the gst_adaptive_demux_stream_free() function. This
causes the refcount of the pad to be set to zero.
https://bugzilla.gnome.org/show_bug.cgi?id=760982
It's useful enough already to be used in other elements for audio aggregation,
let's give people the opportunity to use it and give it some API testing.
https://bugzilla.gnome.org/show_bug.cgi?id=760733
Handling the ghostpad and its internal pad was causing more issues
than helping because of their coupled activation/deactivation
actions.
As we have to install custom chain,event and query functions it is
better to use a floating sink pad internally in the demuxer and just
use those pad functions to push through a standard pad in the demuxer
https://bugzilla.gnome.org/show_bug.cgi?id=757951
The messages are stored by gst_gl_async_debug_store_log_msg() and output later
by a corresponding store(), output() or an unset()/free().
Some wrapper macros are provided to avoid callers explicitly using __FILE__,
GST_FUNCTION and __LINE__
This makes a pipeline like:
... ! video/x-raw(memory:GLMemory),format=UYVY ! glcolorconvert !
video/x-raw(memory:GLMemory),format={UYVY, NV12} ! ...
passthrough instead of converting UYVY => NV12. The conversion would happen
before this change since the element (and basetransform) transform the src caps
to format={NV12, UYVY} (since NV12 comes first in the glcolorconvert:src
template) and then the default caps fixate func would fixate to NV12. Blah.
Also there's no need to intersect against the template caps in ::transform_caps
since basetransform does that right after calling the vfunc.
The optimistic download_transfer was not setting the required flag to not
perform glReadPixels on subsequent map (READ). resulting in glReadPixels
happening twice.
Fixed adaptivedemux seeking without flushing that just wants
to update stop position. This required protecting the segment
variables with a new mutex so that the seeking thread and the
download threads could safely manipulate the segment and
events related to it.
This contention is only locked/unlocked when starting a new
download, when the first fragment of a segment is received and
when seeking so, hopefully, it won't damage performance.
Some operations are unnecessary when running with only a single GL
context.
e.g. glFlush when setting a fence object as the flush happens on wait.
API: gst_gl_context_is_shared
Avoids downloading and pushing a full segment just to get 1 nanosecond
of data. This happens frequently when seeking is done with flags
that adjust to boundaries or when the start is aligned with segment
starts. The later is common when segment durations is a multiple of
a second.
For reverse, set position to segment.stop when starting and also
don't set the position to fragment end timestamp when it finishes,
just leave it at the fragment start.
1. Various elements/base classes only perform a subset check on accept-caps
2. Some GL elements have texture-target in their pad template
3. When checking subsets, only the caps to check are allowed to contain extra
fields. If the 'template' caps have extra fields, the subset fails.
Thus without texture-target on the caps, various accept-caps implementations
were failing.
Also, add some convenience functions for setting and retrieving
texture targets to/from GValue.
https://bugzilla.gnome.org/show_bug.cgi?id=759860
In very few cases the simple version was actually needed and having the
parameters hidden by a _full() version caused application that actually needed
it to not use it.
GL 2.1 only supports pbo upload.
The wrapped data pointer was only being set on the pbo memory and on the
glmemory so when a download was requested (in GL 2.1), glmemory was
allocating a new data pointer and thus not returning the wrapped data.
