In ALSA, there is possible temporary failures that may require a retry,
though in certain situation, this may leak to the write() function
holding on a lock forever preventing the pipeline from going to pause
or stop. Fix this by shortly dropping the lock between retries.
Part-of: <https://gitlab.freedesktop.org/gstreamer/gst-plugins-base/-/merge_requests/1261>
Below fallback paths were introduced in
9759810d82
if setting period time after buffer time failed :
1) Set period time and then buffer time if it doesn't work
2) Set only buffer time
3) Set only period time
These all were not functioning properly since they were using old
copy of snd_pcm_hw_params_t which already had some fields set
as per previous try and this was causing issues as driver was
referring to that old value while trying to set them again in
fallback paths.
So now we always use the initial copy of snd_pcm_hw_params_t
for every fallback and same is also being done at
557c429510
Also we change the sequence to set period time earlier than
buffer time since period bytes being the smaller unit, most of the times
if underlying alsa device has a dependency then it is of period bytes
to be a multiple of some value (as per underlying DMA constraint)
and rest of the parameters like buffer bytes need to be adjusted
as per period bytes.
The same sequence is also followed in alsa-utils at
9b621eeac4
Fix 2) and 3) scenarios by returning success if the exclusive setting is passed
and not doing any further setting for buffer time or period time.
Add new fallback path of not setting any buffer time and period time
if all above fallback paths fail. The same is also being
followed at aforementioned pulseaudio commit.
In case of alsasink, remove the retry goto label, since it is not
required anymore as fallback paths take care of setting default
values if driver is not accepting any of the fallback paths.
Use separate label for exit to free params structs and return err
code. This also fixes leak in no_rate goto path in alsasink
Part-of: <https://gitlab.freedesktop.org/gstreamer/gst-plugins-base/-/merge_requests/1212>
There may be two or more threads involved here however the important
interaction is the use of ogg->seeK_event_drop_till value that was only
set in the push-mode seek-event thread and could race with upstream
sending e.g. and EOS (or data).
Scenario is this:
1. oggdemux performs a seek to near the end of the file to try and find
the duration. ogg->push_state is set to PUSH_DURATION.
2. Seek is picked up by the dedicated seek event thread and sets
ogg->seek_event_drop_till to the seek event's seqnum.
3. Most operations are blocked or dropped waiting on the duration to
be determined and processing continues until a duration is found.
4. Two branching options for how this ultimately plays out
4a. The source is too fast and we receive an EOS event which is dropped
because ogg->push_state == PUSH_DURATION. In this case everything
works.
4b. We hit our 'almost at the end' check in
gst_ogg_pad_handle_push_mode_state() and attempt to seek back to the
beginning (or to a user-provided seek). This seek is marshalled to
the seek event thread without setting ogg->seek_event_drop_till but
with change ogg->push_state = PUSH_PLAYING. If an EOS event or
e.g. buffers arrive from upstream before the seek event thread has
picked up the seek event, then the EOS/data is processed as if it
came as a result of the seek event. This is the case that fails.
The fix is two-fold:
1. Preemptively set ogg->seek_event_drop_till when setting the seek
event so that data and other events can be dropped correctly.
2. In addition to dropping and EOS events while ogg->push_state ==
PUSH_DURATION, also drop any EOS events that are received before the
seek event has been processed by also tracking the seqnum of the seek.
Part-of: <https://gitlab.freedesktop.org/gstreamer/gst-plugins-base/-/merge_requests/1196>
The issue can be reproduced on a computer with a Radeon graphics card
when trying to force GStreamer Editing Services to use GL for video
mixing in GESSmartMixer, instead of the GstCompositor that smart mixer
would normally use. This change causes the resulting video stream to
have "video/x-raw(memory:GLMemory) ... texture-target: 2D" caps (instead
of "video/x-raw ..." caps). At the PlaySink stage of the pipeline, a
GstGLImageSinkBin is plugged, with a GstGLColorBalance on it. For some
reason that is still to be debugged (and out of the scope of this
patch), gst_gl_filter_set_caps() is never called on that color balance
element, leaving filter->in_texture_target set to its default
GST_GL_TEXTURE_TARGET_NONE value. The incomplete _create_shader() logic
does the rest and silently generates a shader code that doesn't build.
This is the command I use to reproduce the issue (I'm not sure if I
would be able to isolate the issue in a simple pipeline, though):
GST_PLUGIN_FEATURE_RANK=vaapih265enc:NONE,vaapih264enc:NONE,vaapisink:NONE,vaapidecodebin:NONE,vaapipostproc:NONE,vaapih265dec:NONE,vaapivc1dec:NONE,vaapih264dec:NONE,vaapimpeg2dec:NONE,vaapijpegdec:NONE,glvideomixer:260
ges-launch-1.0 +clip /tmp/video.mp4
Part-of: <https://gitlab.freedesktop.org/gstreamer/gst-plugins-base/-/merge_requests/1159>
Like other foobarA variant APIs on Windows, formatted string
by strftime() is ANSI string, not unicode encoded one.
It would be problematic for non-english locale systems.
We should use unicode version API (wcsftime in this case)
whenever it's possible on Windows.
Part-of: <https://gitlab.freedesktop.org/gstreamer/gst-plugins-base/-/merge_requests/1093>
If the alsasink thread starts the write loop but another thread pauses
the underlying alsa device, the sink thread will endlessly loop.
snd_pcm_writei() will return 0 if the state is SND_PCM_STATE_PAUSED
and the loop will never make any progress.
Part-of: <https://gitlab.freedesktop.org/gstreamer/gst-plugins-base/-/merge_requests/1097>
This because underlying driver may have constraint on
buffer size to be dependent on period size, so period
time needs to be set first.
For e.g. Xilinx ASoC driver requires
buffer size to be multiple of period size for it's DMA
operation.
alsa-utils also set period time first as seen in below commit :
9b621eeac4
Tested it on zcu106 board with HDMI based record and playback.
Also tested on Intel PC using Logitech C920 Webcam mic and ALC887-VD
Analog for playback.
Part-of: <https://gitlab.freedesktop.org/gstreamer/gst-plugins-base/-/merge_requests/1040>
The current manner in transform_caps() for src direction is not very correct. For example,
when the src caps is:
video/x-raw(memory:DMABuf); video/x-raw; video/x-raw(memory:GLMemory)
this function returns:
video/x-raw(memory:DMABuf); video/x-raw; video/x-raw(memory:GLMemory)
as the sink caps. This is not correct, because DMABuf feature is not even in the sink pad's
caps template. The correct answer should be:
video/x-raw(memory:GLMemory); video/x-raw
only.
Part-of: <https://gitlab.freedesktop.org/gstreamer/gst-plugins-base/-/merge_requests/1058>
These parameters are incorrectly regarded as mutable in G-IR making them
"incompatible" with languages that are explicit about mutability like
Rust. In order to clean up the code and expected API there, update the
signatures here, right at the source (instead of overriding them in
Gir.toml and hoping for the best).
Part-of: <https://gitlab.freedesktop.org/gstreamer/gst-plugins-base/-/merge_requests/1005>
libvisual api expects a priv data pointer to be passed, though we know its
going to be `GstDebugLevel`.
```
../subprojects/gst-plugins-base/ext/libvisual/plugin.c:33:39: error: cast to smaller integer type 'GstDebugLevel' from 'void *' [-Werror,-Wvoid-pointer-to-enum-cast]
GST_CAT_LEVEL_LOG (libvisual_debug, (GstDebugLevel) (priv), NULL, "%s - %s",
```
Part-of: <https://gitlab.freedesktop.org/gstreamer/gst-plugins-base/-/merge_requests/975>
The sink_query just uses context, other_context and display to query info.
But all these objects can be changed or distroyed in state_change() func
and other places.
This patch is not very perfect. The condition race still exists in other
places in this element. All the functions directly access these objects
without protection. Most of them are executed when the data is pushing and
draw context/window have already been established, so they should not have
problems. But the sink_query and propose_allocation functions are the query
-like functions and executed in query context, which can be called in any
state of the element. So it can cause some crash issues because of destroyed
context object.
Part-of: <https://gitlab.freedesktop.org/gstreamer/gst-plugins-base/-/merge_requests/922>
The operations for the inside GstGLUploadElement->upload have race
condition. The _transform_caps() will creates this object if it does
not exist, while the _stop() and change_state() can destroy this object.
The _transform_caps() is called by the gst_base_transform_query(),
so it does not hold the stream lock. It may use the upload while the
_stop() and change_state() has already destroy that object, and then
crash.
Fix: #645
Part-of: <https://gitlab.freedesktop.org/gstreamer/gst-plugins-base/-/merge_requests/913>
Setting telemetry options, even to zero, causes libtheora to enable an expensive code path. For large enough videos (e.g. 1920x1080) this can increase the time to decode each frame by 30-40 ms, which can be enough to cause noticeable stutter.
Part-of: <https://gitlab.freedesktop.org/gstreamer/gst-plugins-base/-/merge_requests/887>
