This a very pathological situation: when we have a HEVC encoder but not a HEVC
decoder.
The encoder needs functions that are only available when the decoder is
enabled.
This patch moves the utils functions into the generic sources, such as the
rest of the utils.
Signed-off-by: Víctor Manuel Jáquez Leal <victorx.jaquez@intel.com>
As gstreamer-vaapi now only supports from GStreamer 1.6, this patch removes
all the old GStreamer version guards.
Signed-off-by: Víctor Manuel Jáquez Leal <victorx.jaquez@intel.com>
Since we don't install libraries anymore, it makes no sense to keep
versioning them according to the gstreamer's version.
Signed-off-by: Víctor Manuel Jáquez Leal <victorx.jaquez@intel.com>
This basically reverts 62c3888b76 (wayland:
decouple wl_buffer from frame).
Otherwise the frame may be overwritten while it is still used by the
compositer:
The frame done callback (frame_done_callback()) is called, when the
compositor is done processing the frame and hands it to the hardware.
The buffer release callback (frame_release_callback()) is called when the
buffer memory is no longer used.
This can be quite some time later: E.g. if weston (with the DRM backend)
puts the buffer on a hardware plane, then then buffer release callback is
called when the kernel is done with the buffer. This is usually when the
next frame is shown, so most likely after the frame done callback for the
next frame!
Since 70eff01d36 "wayland: sync() when
destroy()" the mentioned possible leak should no longer be a problem, so
reverting this change should cause no leaking buffers.
Signed-off-by: Víctor Manuel Jáquez Leal <victorx.jaquez@intel.com>
https://bugzilla.gnome.org/show_bug.cgi?id=758848
Signed-off-by: Víctor Manuel Jáquez Leal <victorx.jaquez@intel.com>
Only supporting vaapidecode ! vaapisink combination for now.
Missing dependencies:
1: No support for P010 video format in GStreamer
2: No support for P010 vaGetImage()/vaPutimage() in vaapi-intel-driver
3: As a result of 1&2 , we have no support for Vaapi Video memory mapping
through GstVideoMeta.
Right now we only set chroma format (YUV420 with more than 8 bits per channel)
for surface pool and keeping GST_VIDEO_FORMAT as ENCODED. The underlying format
of the surfaces is implementation (driver) defined, which is P010.
This new API gst_vaapi_surface_pool_new_with_chroma_type() is for
creating a new GstVaapiVideoPool of GstVaapiSurfaces with the specified
chroam type and dimensions. The underlying format of the surfaces is
implementation (driver) defined.
When receiving the texture from the application or the video sink, we must
know it size and border. To query the texture the API has changed according to
the OpenGL version used in the GL context of the application/vsink.
This patch checks the current context API type and queries the texture
according to this detected API.
Signed-off-by: Víctor Manuel Jáquez Leal <victorx.jaquez@intel.com>
https://bugzilla.gnome.org/show_bug.cgi?id=753099
gst_vaapi_texture_glx_new_wrapped() only handles a GL_TEXTURE_2D target and
formats GL_RGBA or GL_BGRA.
This patch adds a debugging verification of those values.
Signed-off-by: Víctor Manuel Jáquez Leal <victorx.jaquez@intel.com>
https://bugzilla.gnome.org/show_bug.cgi?id=753099
In order to know which OpenGL API use, we must detect the API type of current
context. This patch adds the function gl_get_current_api() which returns the
OpenGL API type.
This function is an adaptation of gst_gl_context_get_current_gl_api() from
GstGL.
Signed-off-by: Víctor Manuel Jáquez Leal <victorx.jaquez@intel.com>
https://bugzilla.gnome.org/show_bug.cgi?id=753099
We always set GST_VAAPI_PICTURE_FLAG_SKIPPED for DECODE_ONLY frames and the
gstvaapidecoder base calss is reponsible for handling those frames later on.
No need for explicit verification of frame header's show_frame in order to
do picture outputing.
Set crop rectange if:
There is display_width and display_height which is different from actual width/height
or
The changed resolution is less than the actual configured dimension of surfaces
Unlike other decoders, vp9 decoder doesn't need to reset the
whole context and surfaces for each resolution change. Context
reset only needed if resolution of any frame is greater than
what actullay configured. There are streams where a bigger
resolution set in ivf header or webm header but actual resolution
of all frames are less. Also it is possible to have inter-prediction
between these multi resolution frames.
clang complains about a couple variables and one label which were not
used. This patch removes them.
Signed-off-by: Víctor Manuel Jáquez Leal <victorx.jaquez@intel.com>
https://bugzilla.gnome.org/show_bug.cgi?id=757958
Since gstvaapidisplay_glx.h do not expose gl.h/glx.h structures, it is not
required to include them in the header. It is not also required to include
them in gstvaapidisplay_glx.c, since gstvaapiutils_glx.h includes them and
exposes their structures (e.g. GLXPixmap).
Nonetheless, glext.h neither glxext.h are required to include, they are
already included conditionally by gl.h and glx.h, respectively.
Signed-off-by: Víctor Manuel Jáquez Leal <victorx.jaquez@intel.com>
https://bugzilla.gnome.org/show_bug.cgi?id=757577
When the GstVaapiParserInfoH264 is allocated, the memory is not initialized,
so it contains random data.
When gst_h264_parser_parse_pps() fails, the PPS structure keeps slice_group_id
pointer uninitialized, leading to a segmentation fault when the memory is
freed.
This patch prevents this by initializing the slice_group_id before the PPS
parsing.
Signed-off-by: Víctor Manuel Jáquez Leal <victorx.jaquez@intel.com>
https://bugzilla.gnome.org/show_bug.cgi?id=754845
Explicit flushing of dpb for EOS and EOB nal decoding is wrong,
the dpb_add() itself will handle the flusing(if needed) of dpb
for end of sequence and end of bitstream.
https://bugzilla.gnome.org/show_bug.cgi?id=754010
This fix is based on the V3 vesion of spec which was missing in older versions.
When the current picture has PicOutputFlag equal to 1, for each picture in the
DPB that is marked as "needed for output" and follows the current picture in output order,
the associated variable PicLatencyCount is set equal to PicLatencyCount + 1 (C.5.2.3).
https://bugzilla.gnome.org/show_bug.cgi?id=754010
The default scan order of scaling lists are up-right-diagonal
as per hevc specification. Use the newly implemented
uprightdiagonal_to_raster conversion codecparser APIs to
get the the scaling_list values in raster order, which is
what the VA intel driver requires.
Setting the sink to flushing causes gst_vaapi_window_wayland_sync() to
return FALSE which makes gst_vaapi_window_wayland_render() return
FALSE which ends up posting an ERROR message in
gst_vaapisink_show_frame_unlocked(). Solution is to just return TRUE
in the EBUSY case.
https://bugzilla.gnome.org/show_bug.cgi?id=753598
First added the function gst_vaapi_video_format_get_best_native(), which
returns the best native format that matches a particular chroma type:
YUV 4:2:0 -> NV12, YUV 4:2:2 -> YUY2, YUV 4:0:0 -> Y800
RGB32 chroma and encoded format map to NV12 too.
That format is used to configure, initially, the surface's pool for the
allocator.
Signed-off-by: Víctor Manuel Jáquez Leal <victorx.jaquez@intel.com>
https://bugzilla.gnome.org/show_bug.cgi?id=744042
Before, only YUV420 color space where supported. With this patch, the
encoder is queried to know the supported formats and admits YUV422
color space if its available.
Signed-off-by: Víctor Manuel Jáquez Leal <victorx.jaquez@intel.com>
https://bugzilla.gnome.org/show_bug.cgi?id=744042
This is necessary for finding ChromaOffsetL0/ChromaOffsetL1
prediction weight table values with out using any hard coding.
Fixme: We don't have parser API for sps_range_extension, so
assumed zero value for high_precision_offsets_enabled_flag.
Signed-off-by: Sreerenj Balachandran <sreerenj.balachandran@intel.com>
Based on ITU-T rec H265(4/2015): 7-56
This was a wrong equation in rec H265 (4/2013): 7-44...
Signed-off-by: Sreerenj Balachandran <sreerenj.balachandran@intel.com>
HACK: This is a work-around to identify some main profile streams having wrong profile_idc.
There are some wrongly encoded main profile streams(eg: ENTP_C_LG_3.bin) which doesn't
have any of the profile_idc values mentioned in Annex-A, instead general_profile_idc
has been set as zero and having general_profile_compatibility_flag[general_profile_idc]
is TRUE. Assuming them as MAIN profile for now.
https://bugzilla.gnome.org/show_bug.cgi?id=753226
Signed-off-by: Sreerenj Balachandran <sreerenj.balachandran@intel.com>
We are calculating the dpb size based on max_dec_pic_buffering.
But if there are more than one temporal sublayers, we are supposed
to use the max_dec_pic_buffering[max_sub_layers_minus] for dpb
size calculation (Assuming HighestTid as max_sub_layers_minus).
Sample streams: TSCL_A_VIDYO_5.bin, TSCL_B_VIDYO_4.bin
https://bugzilla.gnome.org/show_bug.cgi?id=753226
Signed-off-by: Sreerenj Balachandran <sreerenj.balachandran@intel.com>
Decoding process for reference picture list construction needs to be
invoked only for P and B slice and the value for slice_type of dependent slice
segment should be taken from the previous independent slice segment header
of the same pic.
https://bugzilla.gnome.org/show_bug.cgi?id=753226
Signed-off-by: Sreerenj Balachandran <sreerenj.balachandran@intel.com>
When we query for the VAConfigAttribEncJPEG, we get a value which packs the
VAConfigAttribValEncJPEG structure, but we did not assign it. This patch
assigns the returned value to the attribute.
Signed-off-by: Víctor Manuel Jáquez Leal <victorx.jaquez@intel.com>
https://bugzilla.gnome.org/show_bug.cgi?id=744042
Framerate 0/1 is valid, and it is particularly useful for picture
encoding, such as jpeg. This patch makes the encoder to admit that
framerate.
Signed-off-by: Víctor Manuel Jáquez Leal <victorx.jaquez@intel.com>
https://bugzilla.gnome.org/show_bug.cgi?id=744042
Improve closure of gaps in frame_num by strictly following and trying
to fill them with previous reference frames. So, they are now tracked
thus avoiding insertion of dummy ("greenish") frames.
If the new picture to be added to the DPB is not a first field, then
it shall be the second field of the previous picture that was added
before.
This removes the need for dpb_find_picture() now that we track the
immediately preceding decoded picture, in decode order.
When a dummy "other-field" is inserted, it is assumed to inherit the
reference flags from the first field, and the sliding window decoded
reference picture marking process is also executed so that corrupted
frames are moved out as early as possible.
While doing so, we also try to output frames that now contain a single
valid field picture, prior to inserting any other picture into the DPB.
Note: this may be superfluous currently based on the fact that dpb_add()
combines the two most recent pairable fields, but this process would be
further simplified later on.
Mark the picture as "corrupted" if it is reconstructed from corrupted
references or if those references are fake, e.g. resulting from lost
frames.
This is useful for notifying the upper layer, or downstream elements,
that the decoded frame may contain artefacts.
https://bugzilla.gnome.org/show_bug.cgi?id=703921
Add initial infrastructure in core codec library and vaapidecode to mark
corrupted frames as such. A corrupted frame is such a frame that was
reconstructed from invalid references for instance.
https://bugzilla.gnome.org/show_bug.cgi?id=751434
Signed-off-by: Gwenole Beauchesne <gwenole.beauchesne@intel.com>
This patch fix the auto-plugging problem in gstreamer 1.2 and gstreamer 1.4
Right now there is not a primary ranked parser for vc1 and the demuxers
delivers caps without specifying the profile. This situation is not an issue
for avdec_vc1 but for vaapidecode it is, which refuses to negotiate without a
explicit profile defined in the negotiated caps.
Nonetheless, in gstreamer 1.5 it seems not to be a problem since the
negotiation admits caps subsets try outs.
This patch solves the issue ignoring the profile negotiation in the caps. For
gstreamer < 1.5 the profile string is not handled, so the auto-plugging get
done without the vc1 parser, such as happens in gstreamer 1.5.
Signed-off-by: Víctor Manuel Jáquez Leal <victorx.jaquez@intel.com>
Fix the regression introduced in commit eb465fb.
VAProcFilterSkinToneEnhancement is avaialbe from VA >= 0.36.
Signed-off-by: Sreerenj Balachandran <sreerenj.balachandran@intel.com>
One buffering_period() SEI message shall be present in every IDR access unit
when NalHrdBpPresentFlag is inferred to be equal to 1. This is the case when we
use a non-CQP mode, e.g. CBR. In other words, when
nal_hrd_parameters_present_flag is set to 1.
One picture_timing() SEI messages shall be present in every access unit
if CpbDpbDelaysPresentFlag is equal to 1 or pic_struct_present_flag is equal to 1
https://bugzilla.gnome.org/show_bug.cgi?id=722734https://bugzilla.gnome.org/show_bug.cgi?id=751831
Signed-off-by: Sreerenj Balachandran <sreerenj.balachandran@intel.com>
This is a work-around to satisfy the va-intel-driver.
Normalize the quality factor and scale QM values (only for packed header
generation) similar to what VA-Intel driver is doing . Otherwise the
generated packed headers will be wrong, since the driver itself is
scaling the QM values using the normalized quality factor.
https://bugzilla.gnome.org/show_bug.cgi?id=748335
Signed-off-by: Sreerenj Balachandran <sreerenj.balachandran@intel.com>
Fix uninitialized variables when decoding SPS and PPS NAL units from
"codec-data" buffers. This is particularly important when seeking ops
are involved, and the new persistent states are used more often.
https://bugzilla.gnome.org/show_bug.cgi?id=750094
The approximation of 6 times compression ratio migh not
work in all cases. Especially when enabling I frames.
Provide large enough size for coded-buffer creation.
The approximation of 4 times compression ratio will not
work in all cases. Especially when enabling I frames.
Provide large enough size for coded-buffer creation.
Implement decoding process for gaps in frame_num (8.5.2). This
also somewhat supports unintentional loss of pictures.
https://bugzilla.gnome.org/show_bug.cgi?id=745048https://bugzilla.gnome.org/show_bug.cgi?id=703921
Original-patch-by: Wind Yuan <feng.yuan@intel.com>
[fixed derivation of POC, ensured clone is valid for reference,
actually fixed detection of gaps in FrameNum by PrevRefFrameNum]
Signed-off-by: Gwenole Beauchesne <gwenole.beauchesne@intel.com>
Try to identify missing first fields too, thus disregarding any
intermediate gaps in frames. We also assume that we keep the same
field sequence, i.e. if previous frames were in top-field-first
(TFF) order, then so are subsequent frames.
Note that insertion of dummy first fields need to operate in two
steps: (i) create the original first field that the current field
will inherit from, and (ii) submit that field into the DPB prior
to initializing the current (other) field POC values but after any
reference flag was set. i.e. copy reference flags from the child
(other field) to the parent (first field).
https://bugzilla.gnome.org/show_bug.cgi?id=745048
Interlaced H.264 video frames always have two fields to decode and
display. However, in some cases, e.g. packet loss, one of the field
can be missing. This perturbs the reference picture marking process,
whereby the number of references available in DPB no longer matches
the expected value.
This patch adds initial support for missing field within a decoded
frame. The current strategy taken is to find out the nearest field,
by POC value, and with the same parity.
https://bugzilla.gnome.org/show_bug.cgi?id=745048
Try to maintain a "top-field-first" (TFF) flag, even if the H.264 standard
does not mandate it. This will be useful for tracking missing fields, and
also for more correct _split_fields() implementation for frames in the DPB.
Don't try to decode pictures until the first I-frame is received within
the currently active sequence. There is no point is decoding and then
displaying frames with artifacts.
Fix decoding of end_of_seq() NAL unit so that to not submit the current
picture for decoding again. This is pretty vintage code that dates back
before the existing of the whole decoder units machinery.
One issue that could be arising if that code was kept is that we could
have submitted a picture, and subsequently a GstVideoCodec frame, twice.
Once without the decode_only flag set, and once with that flag set. The
end result is that the GstVideoDecoder would release the codec frame
twice, thus releasing stale data.
In short, the piece of code that is removed by this patch is for once
completely obsolete for a while, and secondly error-prone in corner
cases.
If the GOP temporal sequence number (TSN) is interpolated from a valid
PTS, then we need to compensate that PTS corresponding to the start of
GOP with the next picture to be decoded, which shall be an I-frame,
based on its sequence number.
https://bugzilla.gnome.org/show_bug.cgi?id=748676
Before pushing a the new frame, the render() method calls sync() to flush the
pending frames. Nonetheless, the last pushed frame never gets rendered, leading
to a memory leak too.
This patch calls sync() in the destroy() to flush the pending frames before
destroying the window.
Also a is_cancelled flag is added. This flag tells to not flush the event
queue again since the method failed previously or were cancelled by the user.
https://bugzilla.gnome.org/show_bug.cgi?id=749078
Otherwise wl_display_dispatch_queue() might prevent the pipeline from
shutting down. This can happen e.g. if the wayland compositor exits while
the pipeline is running.
Changes:
* renamed unlock()/unlock_stop() to unblock()/unblock_cancel() in gstvaapiwindow
* splitted the patch removing wl_display_dispatch_queue()
Signed-off-by: Víctor Manuel Jáquez Leal <victorx.jaquez@intel.com>
https://bugzilla.gnome.org/show_bug.cgi?id=747492https://bugzilla.gnome.org/show_bug.cgi?id=749078
wl_display_dispatch_queue() might prevent the pipeline from shutting
down. This can happen e.g. if the wayland compositor exits while the
pipeline is running.
This patch replaces it with these steps:
- With wl_display_prepare_read() all threads announce their intention
to read.
- wl_display_read_events() is thread save. On threads reads, the other
wait for it to finish.
- With wl_display_dispatch_queue_pending() each thread dispatches its
own events.
wl_display_dispatch_queue_pending() was defined since wayland 1.0.2
Original-patch-by: Michael Olbrich <m.olbrich@pengutronix.de>
* stripped out the unlock() unlock_stop() logic
* stripped out the poll handling
Signed-off-by: Víctor Manuel Jáquez Leal <victorx.jaquez@intel.com>
https://bugzilla.gnome.org/show_bug.cgi?id=749078https://bugzilla.gnome.org/show_bug.cgi?id=747492
Since frame in the private data means the last frame sent, it would
semantically better use last_frame.
Also, this patch makes use of g_atomic_pointer_{compare_and_exchange, set}()
functions.
https://bugzilla.gnome.org/show_bug.cgi?id=749078
Wayland window has a pointer to the last pushed frame and use it to set the
flag for stopping the queue dispatch loop. This may lead to memory leaks,
since we are not keeping track of all the queued frames structures.
This patch removes the last pushed frame pointer and change the binary flag
for an atomic counter, keeping track of number of queued frames and use it for
the queue dispatch loop.
https://bugzilla.gnome.org/show_bug.cgi?id=749078
This patch takes out the wayland's buffer from the the frame structure. The
buffer is queued to wayland and destroyed in the "release" callback. The
frame is freed in the surface's "done" callback.
In this way a buffer may be leaked but not the whole frame structure.
- surface 'done' callback is used to throttle the rendering operation and to
unallocate the frame, but not the buffer.
- buffer 'release' callback is used to destroy wl_buffer.
Original-patch-by: Zhao Halley <halley.zhao@intel.com>
* code rebase
* kept the the event_queue for buffer's proxy
Signed-off-by: Víctor Manuel Jáquez Leal <victorx.jaquez@intel.com>
https://bugzilla.gnome.org/show_bug.cgi?id=749078
This patch fixes several issues found when running the `make distcheck`
target:
- In commit c561b8da, the update of gstcompat.h in Makefile.am was
forgotten.
- In commit c5756a91 add the simple_encoder_source_h in EXTRA_DIST was
forgotten.
- vpx.build.stamp is not generated at all, only vpx.configure.stamp.
- The make target distcleancheck failed because some autogenerated files
were not handled with the DISTCLEANFILES variable.
Note: `make distcheck -jXX` is not currently supported.
GST_VAAPI_ENCODER_STATUS_NO_SURFACE and GST_VAAPI_ENCODER_STATUS_NO_BUFFER
are not errors, so they do not have the ERROR namespace.
This patch fixes this typo in documentation.
On s390x, guintptr and GstVaapiID are not compatible types. The
implementation of gst_vaapi_window_new_internal() and all its callers
seem to assume that its third argument is a GstVaapiID, while the
header gives it guintptr type.
https://bugzilla.gnome.org/show_bug.cgi?id=744559
Since bug #745728 was fixed the oldest supported version of GStreamer is
1.2. That GStreamer release requires glib 2.32, so we can upgrade our
requirement too.
This patch changes the required version of glib in configure.ac and removes
the hacks in glibcompat.h
https://bugzilla.gnome.org/show_bug.cgi?id=748698
This patch only intends to improve readability: in the method
gst_vaapi_window_wayland_sync() the if/do instructions are squashed into a
single while loop.
Also renames the frame_redraw_callback() callback into frame_done_callback(),
which is a bit more aligned to Wayland API.
