In some upload implementations the out buffer has more than one references,
turning the buffer not writable, so it won't be possible to modify its
meta-data.
This patch moves the meta-data copy before increasing the reference of the out
buffer.
https://bugzilla.gnome.org/show_bug.cgi?id=746173
glupload ! glcolorconvert ! sink
Some properties are manually forwarded. The rest are available using
GstChildProxy.
The two signals are forwarded as well.
It encapsulates a confiurable GL processing element in the
upload/colorconvert/download dance required to transparently process
the majority of GstBuffer's.
GLImage does not use any kind of internal pool. There was some
remaining code and comment stating that it was managing the
pool, and it was in fact setting the active state when doing
to ready state.
* Only create the pool if requested and in propose_allocation
* Cache the pool to avoid reallocation on spurious reconfigure
* Don't try to deactivate the pool (we don't own it)
https://bugzilla.gnome.org/show_bug.cgi?id=745705
LibJPEG uses macroblock of 8x8 sample. In this element we use RGB and
Y444, two 24bit formats that are stored in 32bit pixels. This mean we
have 32x32 bytes macroblocks. For this reason, we need to allocate
our buffer slightly larger. We also need to pass the line pointer in
the right order, otherwise the image endup upside-down.
https://bugzilla.gnome.org/show_bug.cgi?id=745109
This implements support for GstAllocationParams and memory alignments.
The parameters where simply ignored which could lead to crash on
certain platform when used with libav and no luck.
https://bugzilla.gnome.org/show_bug.cgi?id=744246
+ Split headers from source
+ Remove uneeded AM_CFLAGS, AM_LDFLAGS
+ Always set OBJCFLAGS
Due to the presence of a .m and regardless of the conditional values,
automake will promote the link command to OBJC using OBJCFLAGS. Only
the basic flags (like warnings and optimization) are going to make a
difference though.
This cleanup builds up the makefile with less specific files first
toward more specific file. FLAGS are built with the basic that unused
flags will have empty variable.
i686-apple-darwin11-llvm-gcc-4.2
gstglmixer.h:43: error: redefinition of typedef ‘GstGLMixer’
gstglmixerpad.h:32: error: previous declaration of ‘GstGLMixer’ was here
gstglmixer.h:46: error: redefinition of typedef ‘GstGLMixerFrameData’
gstglmixerpad.h:33: error: previous declaration of ‘GstGLMixerFrameData’ was here
The graphene-1.0 part should not be in the source code. This directory
is part of the cflags include. This is similar to gstreamer-1.0/
directory. This break compilation if the include directory where
graphene is installed is not in your include path.
Allows finer grain decisions about formats and features at each
stage of the pipeline.
Also provide propose_allocation for glupload besed on the supported
methods.
Make GstGLMemory hold the texture target (tex_target) the texture it represents
(tex_id) is bound to. Modify gst_gl_memory_wrapped_texture and
gst_gl_download_perform_with_data to take the texture target as an argument.
This change is needed to support wrapping textures created outside libgstgl,
which might be bound to a target other than GL_TEXTURE_2D. For example on OSX
textures coming from VideoToolbox have target GL_TEXTURE_RECTANGLE.
With this change we still keep (and sometimes imply) GL_TEXTURE_2D as the
target of textures created with libgstgl.
API: modify GstGLMemory
API: modify gst_gl_memory_wrapped_texture
API: gst_gl_download_perform_with_data
Depending on the platform, it was only ever implemented to 1) set a
default surface size, 2) resize based on the video frame or 3) nothing.
Instead, provide a set_preferred_size () that elements/applications
can use to request a certain size which may be ignored for
videooverlay/other cases.
A context can create a GLsync object that can be waited on in order
to ensure that GL resources created in one context are able to be
used in another shared context without any chance of reading invalid
data.
This meta would be placed on buffers that are known to cross from
one context to another. The receiving element would then wait
on the sync object to ensure that the data to be used is complete.
This gives more flexibility to the subclasses and permits to remove the
GstVideoAggregatorClass->disable_frame_conversion ugly API.
WARNING: This breaks the API as it removes the disable_frame_conversion
field
API:
+ GstVideoAggregatorClass->find_best_format
+ GstVideoAggregatorPadClass->set_format
+ GstVideoAggregatorPadClass->prepare_frame
+ GstVideoAggregatorPadClass->clean_frame
- GstVideoAggregatorClass->disable_frame_conversion
https://bugzilla.gnome.org/show_bug.cgi?id=740768
If we seek when media is in stop state, playback-test gives
critical error, since context of glimagesink is destroyed during stop.
But since context is not present, we need not handle send_event in glimagesink
Hence adding a condition to check if context is valid.
https://bugzilla.gnome.org/show_bug.cgi?id=740305
gst_glimage_sink_handle_events can be called from the overlay interface and from
the main thread before GL is setup. Before this change, that would call
_ensure_gl_setup() and deadlock on OSX.
Change things so that it's always safe to call gst_glimage_sink_handle_events()
without stuff deadlocking.
Remove gst_glimage_sink_handle_events call in gst_glimage_sink_init. It was
unnecessary and when the element was instantiated from the main thread, caused a
deadlock in OSX creating the context (thread).
with force-aspect-ratio=true, if the width or height changed, the
viewport wasn't being updated to respect the new video width and height
until a resize occured.
Otherwise, it is only possible for the sink pads and the src pads to
have the exact same caps features. We can convert from any feature
to another feature so support that.
Otherwise, it is only possible for the sink pads and the src pads to
have the exact same caps features. We can convert from any feature
to another feature so support that.
- update for shaders
- add alpha property
- image placement properties shamelessly borrowed from gdkpixbufoverlay
- image placement properties are GstController able
- use GstGLMemory for the overlay image data
- add support for gles2
Otherwise we could pass on a RGBA formatted buffer and downstream would
misinterpret that as some other video format.
Fixes pipelines of the form
gleffects ! tee ! xvimagesink
Allows callers to properly reference count the buffers used for
rendering.
Fixes a redraw race in glimagesink where the previous buffer
(the one used for redraw operations) is freed as soon as the next
buffer is uploaded.
1. glimagesink uploads in _prepare() to texture n
1.1 glupload holds buffer n
2. glimagesink _render()s texture n
3. glimagesink uploads texture n+1
3.1 glupload free previous buffer which deletes texture n
3.2 glupload holds buffer n+1
4. glwindow resize/expose
5. glimagesink redraws with texture n
The race is that the buffer n (the one used for redrawing) is freed as soon as
the buffer n+1 arrives. There could be any amount of time and number of
redraws between this event and when buffer n+1 is actually rendered and thus
replaces buffer n as the redraw source.
https://bugzilla.gnome.org/show_bug.cgi?id=736740
* aspect should not be 0 on init
* rename fovy to fov
* add mvp to properties as boxed graphene type
* fix transformation order. scale first
* clear color with 1.0 alpha
https://bugzilla.gnome.org/show_bug.cgi?id=734223
Dynamic pipelines that get and release the sink pads will finalize
the pad without going through gst_gl_mixer_stop() which is where the
upload object is usually freed. Don't leak objects in such case.
If window is resized, GstStructure pointer values have to be rescaled to
original geometry. A get_surface_dimensions GLWindow class method is added for
this purpose and used in the navigation send_event function.
https://bugzilla.gnome.org/show_bug.cgi?id=703486
The expected default behaviour for video sink is to maintain the
aspect ratio. Fix the default value to reflect this. The property
default was already TRUE, but the value was not initially TRUE.
Allows automatic negotiation of the size in the following case:
gst-launch-1.0 glvideomixer name=m sink_0::xpos=0 sink_1::xpos=320 ! glimagesink \
videotestsrc ! m. \
videotestsrc pattern=1 ! m.
https://bugzilla.gnome.org/show_bug.cgi?id=731878
This is too allow gst-launch debugging with multiple GL contexts as
well as avoiding segfaulting innocent gtk+ apps that have not called
XInitThreads.
https://bugzilla.gnome.org/show_bug.cgi?id=731525
The reshape property was never used.
Replace the draw property with a signal.
Based on patch by Mathieu Duponchelle <mathieu.duponchelle@epitech.eu>
https://bugzilla.gnome.org/show_bug.cgi?id=704507
This patch provides the basic infrastructure required for this.
Upload and Download has been ported to this.
Has the nice effect of allowing GstGLMemory to be our
refcounted texture object for any texture type (not just RGBA).
Should not lose any features/video formats.
gst_gl_context_create() might need to dispatch some operations to the
application's main thread, and calling this in the change_state function
can cause deadlocks.
* picked from old libgstegl:
- GstEGLImageMemory
- GstEGLImageAllocator
- last_buffer management from removed GstEGLImageBufferPool
* add-ons:
- GstEGLImageMemory now old a reference on GstGLContext
so that it can delete the EGLImage and its gltexture source
while having the associated gl context being current.
- add EGLImage support for GstVideoGLTextureUploadMeta which
mainly call EGLImageTargetTexture2D
- GstGLBufferPool now supports GstEGLImageAllocator
- glimagesink / glfilters / etc.. now propose GstEGLImageAllocator
to upstream
https://bugzilla.gnome.org/show_bug.cgi?id=703343
Make buffer timestamps more accurate and, more importantly, actually
representative of the MIDI events timing.
Previously, buffers were only sent with timetamps aligned at a 10ms
boundary which was just wrong, now the buffer timestamp represents the
real time of the MIDI event.
Conveniently, the ALSA sequencer API supports scheduling events in the
future so the sequencer infrastructure can be used to have the tick
delivered at the right time, avoiding any custom scheduling mechanism.
The ticks scheduling starts on the first transition to PLAYING, and the
delay is also calculated when the pipeline goes into PLAYING.
https://bugzilla.gnome.org/show_bug.cgi?id=787683
When setting the "ports" property the value is duplicated but it's not
freed when the elements stops.
Reported by Valgrind (example run with "alsamidisrc ports=128:0"):
6 bytes in 1 blocks are definitely lost in loss record 30 of 1,911
at 0x4C2BBEF: malloc (in /usr/lib/valgrind/vgpreload_memcheck-amd64-linux.so)
by 0x5411528: g_malloc (gmem.c:94)
by 0x542A9FE: g_strdup (gstrfuncs.c:363)
by 0x775211E: gst_alsa_midi_src_set_property (gstalsamidisrc.c:284)
by 0x5184A4D: object_set_property (gobject.c:1439)
by 0x5184A4D: g_object_setv (gobject.c:2245)
by 0x51859DD: g_object_set_property (gobject.c:2529)
by 0x4F0474C: ??? (in /usr/lib/x86_64-linux-gnu/libgstreamer-1.0.so.0.1203.0)
by 0x4F065C8: ??? (in /usr/lib/x86_64-linux-gnu/libgstreamer-1.0.so.0.1203.0)
by 0x4F07557: ??? (in /usr/lib/x86_64-linux-gnu/libgstreamer-1.0.so.0.1203.0)
by 0x4EFE3EE: gst_parse_launch_full (in /usr/lib/x86_64-linux-gnu/libgstreamer-1.0.so.0.1203.0)
by 0x4EFE673: gst_parse_launchv_full (in /usr/lib/x86_64-linux-gnu/libgstreamer-1.0.so.0.1203.0)
https://bugzilla.gnome.org/show_bug.cgi?id=787683
We only need to initialize the mutex/cond once when creating the
element and then release them when we are done with the element.
Avoids weird "mutex_clear called when still locked" issues
There were still some races going on where seeking events wouldn't
be properly intercepted/executed by this thread.
* Instead of always waiting for the GCond to be emitted, first just
check if there is an event available
* Take ownership of the event *while* the lock is taken and not
after releasing/reacquiring it
* Finally acquire lock at the very top and release it at the end
to make it a bit more streamlined
This removes the remaining issues with seeks not being executed
The previous branch will release the lock in the call to
gst_ogg_demux_seek_back_after_push_duration_check_unlock()
Only unlock it if we didn't call that function
When calculating duration in push-mode we seek to a certain position
and discard any data until we get data from that requested position.
The problem is that basing ourselves solely on offset to determine
whether we reached the target offset is wrong since the source might
be fast enough to send us that target position *before* it processed
the requested seek.
This would end up in a situation where:
* We think we're done with duration estimate
* We fire a seek back to "0" in the loop thread
* We resume normal processing
* ... except that we're still getting data from too far ahead which
we decide to process.
* And we start doing totally wrong granule/time/duration calculation
and pushing wrong data.
Instead of this confusion, wait until we receive data from the requested
seek. We do that by using the fact that the seqnum in
seek_event_drop_til will be non-zero until the SEGMENT corresponding
to the requested SEEK has been received.
Bonus: makes startup slightly faster
Code using the push_loop_thread (using for sending seeks) assumes
that the thread was properly started, except that this isn't always
true and the thread might not have completely started.
Instead wait for the thread to properly start before doing anything
else.
If we are going to return a (potentially) 64bit integer, don't use
a 32bit one for calculation, otherwise we could end up exceeding
the maximum size of a 32bit int.
For stream mappers that don't set a specific granuleshift, it will
have the default value of -1.
Protect the code for that and return the granule value as-is
Since the default value of a GstOggPad.map.map was 0 ... we would
end up using wrong functions from mappers() if the stream wasn't
initialized yet.
Instead of that, use a default blank/empty first entry.
In some corner cases we end up with the building chain not being
properly tracked (and therefore not properly freed).
Add a FIXME so it can later be fixed, but for now just fix the leak
... as expected later on when end time is used to determine end running time.
Otherwise the latter is determined as NONE and the resulting text buffer is
then used indefinitely.
Fix various issues with reverse playback by clearing tracking
vars when working in reverse, and where possible using the
timestamp interpolation code to generate timestamps for
outgoing buffers. Make sure to mark things as discontinuous
only when looping backward to a new position and fix seeking
to the next page when starting.
In gst_ogg_demux_do_seek() when calculating the
keyframe time, account for a non-zero start-time
Handle a discontinuous first packet in
gst_ogg_demux_setup_first_granule() because that's pretty
normal after a seek. Also differentiate between a genuinely
truncated first packet and just bailing out early, by not using
granule = -1 as an error code.
Make the debug output logs clearer about which timestamps
are stream times (PTS) and which are ogg timestamps.
This is a followup commit to b95725c37e
* Resetting the decoder should only happen when we get a new initialization
header (0x01) and not on the other headers
* The initialized variable only gets set to TRUE once all headers have
been parsed. Also check if the vorbis_info struct has been properly resetted
also. Failure to do that would cause vorbisdec to error if it got
two initialization header in a row (the first would configure the underlying
library and the second one would error out because it's already initialized)
https://bugzilla.gnome.org/show_bug.cgi?id=779515
This fixes missing audio when we get buffers with zero
duration, denoting unknown duration. When several such
buffers are received in a row, they're all at the same
timestamp, with zero duration.
https://bugzilla.gnome.org/show_bug.cgi?id=771723
Always supply a buffer with max size to the decoder, as we
can't really decide how many samples will be in the lost packet
based on the timestamps we get.
https://bugzilla.gnome.org/show_bug.cgi?id=771723
If we can't find a valid granule near the end of the file, we
disable seeking. This guards against the whole file being then
read and never going to PLAYING.
https://bugzilla.gnome.org/show_bug.cgi?id=770314
https://github.com/mesonbuild/meson
With contributions from:
Tim-Philipp Müller <tim@centricular.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.
This workaround tried to avoid an EOS event when seeking to the
end of an Ogg stream in order to find its duration. At some point,
an EOS event there would cause any queue2 upstream to pause and
not restart on a seek back to the beginning. This now appears to
not be the case anymore, and so the workaround can be removed.
https://bugzilla.gnome.org/show_bug.cgi?id=767689
This reverts commit a16cd5d2a5.
Setting the stop time on the segment breaks reconfiguration, as the
encoder signals an EOS, but we reconfigure it an continue to produce
buffers.
This information should not be required via the segment downstream
since we already have the sample count being used to generate buffer
durations.
https://bugzilla.gnome.org/show_bug.cgi?id=768763
If the duration is not known from the chain, it might be known
by the startup seek.
This fixes failure to seek.
Merged with a patch from Tim-Philipp Müller <tim@centricular.com>
https://bugzilla.gnome.org/show_bug.cgi?id=768991
Dropping a buffer because we have a seek pending is normal,
and will now happen when we trigger a seek while going through
the packets in a page. So this should not be an error.
A low bitrate stream which can pack more than 2 seconds of audio
in a page would cause the stream's position to be updated not
often enough, and would trigger a spurious "jump" via a GAP
event. Instead, we update the stream position after calculating
the new overall segment position.
https://bugzilla.gnome.org/show_bug.cgi?id=764966
The only way for ALSA to expose a position-less multi channels is to
return an array full of SND_CHMAP_MONO. Converting this to a
GST_AUDIO_CHANNEL_POSITION_MONO array would be invalid as
GST_AUDIO_CHANNEL_POSITION_MONO is meant to be used only with one
channel.
Fix this by using GST_AUDIO_CHANNEL_POSITION_NONE which is meant to be
used for position-less channels.
https://bugzilla.gnome.org/show_bug.cgi?id=763799
Introduces [x-absolute, y-absolute] properties
for positioning in +/- MAX_DOUBLE range.
Adds new (h/v)alignment type "absolute" where coordinates
map the text area to be exactly inside of video canvas for [0, 0] - [1, 1]:
[0, 0]: Top-Lefts of video and text are aligned
[0.5, 0.5]: Centers are aligned
[1, 1]: Bottom-Rights are aligned
https://bugzilla.gnome.org/show_bug.cgi?id=761251
libgstreamer currently exports some debug category
symbols GST_CAT_*, but those are not declared in any
public headers.
Some plugins and libgstvideo just use GST_DEBUG_CATEGORY_EXTERN()
to declare and use those, but that's just not right at
all, and it won't work on Windows with MSVC. Instead look
up the categories via the API.
In order to detect graphical user input on the
textoverlay, the resulting rendering properties
need to be exposed to applications.
Fixes delayx property declaration.
https://bugzilla.gnome.org/show_bug.cgi?id=761251
The current position property is limited to X,Y positions
in the range of [0, 1]. This patch allows full control
over the overlay position, including partially outside
of the video area.
https://bugzilla.gnome.org/show_bug.cgi?id=761251
FEC may only be used when PLC is enabled on the audio decoder,
as it relies on empty buffers to generate audio from the next
buffer. Hooking to the gap events doesn't work as the audio
decoder does not like more buffers output than it sends.
The length of data to generate using FEC from the next packet
is determined by rounding the gap duration to nearest. This
ensures that duration imprecision does not cause quantization
to 2.5 milliseconds less than available. Doing so causes the
Opus API to fail decoding. Such duration imprecision is common
in live cases.
The buffer to consider when determining the length of audio
to be decoded is the previous buffer when using FEC, and the
new buffer otherwise. In the FEC case, this means we determine
the amount of audio from the previous buffer, whether it was
missing or not (and get the data either from this buffer, or
the current one if the previous one was missing).
rename gst-launch --> gst-launch-1.0
replace old elements with new elements(ffmpegcolorspace -> videoconvert, ffenc_** -> avenc_**)
fix caps in examples
https://bugzilla.gnome.org/show_bug.cgi?id=759432
The result of the two expressions will be promoted to guint64 anyway,
perform all the arithmetic in 64 bits to avoid potential overflows.
CID 1338690, CID 1338691
We always require the channel-mapping-field. If it's 0 we require nothing
else, otherwise we need channels, stream-count and coupled count to be
available.
oggdemux is outputting the meta now, and only outputs if it should really
apply to the current buffer. Previously we would skip N samples also if we
started the decoder in the middle of the stream.
https://bugzilla.gnome.org/show_bug.cgi?id=757153
It is doing the wrong thing because of the Opus pre-skip: while the timestamps
are shifted by the pre-skip, the granule positions are not shifted.
oggmux is doing the right thing here already.
https://bugzilla.gnome.org/show_bug.cgi?id=757153
The first frame has lookahead less samples, the last frame might have some
padding or we might have to encode another frame of silence to get all our
input into the encoded data.
This is because of a) the lookahead at the beginning of the encoding, which
shifts all data by that amount of samples and b) the padding needed to fill
the very last frame completely.
Ideally we would use LPC to calculate something better than silence for the
padding to make the encoding as smooth as possible.
With this we get exactly the same amount of samples again in an
opusenc ! opusdec pipeline.
https://bugzilla.gnome.org/show_bug.cgi?id=757153
... instead of relying on the segment. For the clipping at the start we assume
a proper value in the OpusHead, as generated by opusparse or opusenc.
Transmuxing in general is not guaranteed to produce the correct values, or
even have a OpusHead (e.g. when having RTP input).
https://bugzilla.gnome.org/show_bug.cgi?id=757153
The granulepos does not have the pre-skip subtracted while timestamps do,
and the last granulepos will be shorter by the number of samples that should
be dropped because of padding in the end.
As such, extrapolating the granule of the beginning of the first frame will
lead to a negative value, which is not a problem but intentional.
https://bugzilla.gnome.org/show_bug.cgi?id=757153
Usually these loops only run once, so there's no problem here. But sometimes
they run twice, and by adding the number of bytes to a 16 bit pointer type we
would advance twice as much as we should.
Also use snd_pcm_frames_to_bytes() in alsasrc to calculate
the number of bytes to skip, same as we do in alsasink.
Thanks to Lucio A. Hernandez <lucio.a.hernandez@gmail.com> for reporting.
The alsamidisrc element allows to get input event from ALSA MIDI
sequencer devices, and possibly convert them to sound using some
downstream element like fluiddec.
Fixes#738687
It is faster than doing a query that propagates downstream and
should be enough
Elements: faac, gsmenc, opusenc, sbcenc, voamrwbenc, adpcmenc, sirenenc
Removes the need for custom caps query handling and makes it more
correct from the beginning on the template. It is a bit uglier
to read because there is 1 entry per channel but makes code easier
to maintain.
accept-caps is not recursive and might stop at the next downstream element,
while caps queries are generally recursive. The next element might accept any
capsfeatures we want, but that doesn't mean that further downstream it will
also work.
Additionally for the future:
We should probably check if downstream *prefers* the
overlay meta, and only enforce usage of it if we can't handle
the format ourselves and thus would have to drop the overlays.
Otherwise we should prefer what downstream wants here.
the extents rectangle is what you need to know to properly position
a buffer that has been rendered in a surface of the ink rectangle
size. This patch make the placement on par with the placement we had
before without having to over allocate.
This patch also enable placement for vertical rendering. Note that
the halginement, valighment and line-alignment default are set to
the previous default when this property is set. This is for backward
compatibility, you can change the value after setting vertical render.
https://bugzilla.gnome.org/show_bug.cgi?id=728636
This patch uses the ink rectangle in order to compute the size
of the surface require to render. It also correctly compute the
transformation matrix as the ink_rect position might not be at
0, 0. Additionally, shadow_offset and outline_offset (which is
in fact the diameter of a dot, not a really an offset) is now
taken into account. Redundant matrix operation has been removed
for the vertical rendering.
Take note that the matrix operation in cairo are excuted in
reverse order.
https://bugzilla.gnome.org/show_bug.cgi?id=728636
* Only send the caps event once if the query had support for the
overlay composition meta.
* Only do the allocation query if it is supported through caps.
* Send overlay_caps before doing allocation query rather then normal
caps
https://bugzilla.gnome.org/show_bug.cgi?id=751157
