Nicolas Dufresne [Wed, 18 Mar 2020 21:03:51 +0000 (17:03 -0400)]
v4l2codecs: pool: Create new buffer when pool is empty
This simply create an empty GstBuffer when the pool is empty. This way it's up
to the allocator to grow or wait if we ran out of memory.
Nicolas Dufresne [Wed, 18 Mar 2020 21:00:37 +0000 (17:00 -0400)]
v4l2slh264dec: Improve end_picture() robustness
Use a goto to ensure that for all cases we cleanup the current picture state.
And move the src buffer allocation higher, so we don't queue a bitstream
buffer if we don't have a picture buffer to decode into.
Nicolas Dufresne [Wed, 18 Mar 2020 20:51:11 +0000 (16:51 -0400)]
codecs: h264: Do not ignore end_picture() return value
If decoding failed because end_picture() failed, set the picture to
nonexisting, this way output_picture() will be skipped. This avoids confusing
special cases in output_picture() implementation.
Nicolas Dufresne [Mon, 17 Feb 2020 23:08:48 +0000 (18:08 -0500)]
v4l2slh264dec: Add output format negotiation
This allow negotiating the output format through caps. Some drivers can
pipeline the decoder buffer through an image processor. This only support
colorspace conversion for now.
Nicolas Dufresne [Fri, 13 Mar 2020 20:51:27 +0000 (16:51 -0400)]
v4l2format: Convert between V4L2 and GST video format
This will be needed in the output format negotiation.
Nicolas Dufresne [Fri, 13 Mar 2020 18:23:39 +0000 (14:23 -0400)]
v4l2codecs: Read driver provided stride
This implements driver stride support but only for single allocation buffers.
This code is imported from the original v4l2 plugin and adapted to the new
helper context.
Nicolas Dufresne [Thu, 12 Mar 2020 20:15:40 +0000 (16:15 -0400)]
v4l2h264dec: Copy frames when GstVideoMeta is not supported
In some case, when downstream does not support GstVideoMeta, we need to
normalize the stride and offset of the buffer so that downstream can render
properly with a GstVideoMeta. This code is not called when GstVideoMeta is
supported downstream.
Nicolas Dufresne [Wed, 11 Mar 2020 21:45:17 +0000 (17:45 -0400)]
v4l2codec: Add initial GstVideoMeta support
In this patch we strictly set the GstVideoMeta width/height to the coded width
and height. Further patches will add stride support and frame copying when
downstream does not support GstVideoMeta.
Nicolas Dufresne [Mon, 9 Mar 2020 20:44:19 +0000 (16:44 -0400)]
v4l2slh264dec: Prevent spurious renegotiation
Don't let downstream cause a renegotiation at random point in time. This would
lead to spurious renegotiation and the decoder state may not be recoverable.
Nicolas Dufresne [Wed, 4 Mar 2020 02:22:52 +0000 (21:22 -0500)]
v4l2slh264dec: Hold on picture buffers
These buffers should not be reused for decoding until they are no longer
referenced.
Nicolas Dufresne [Sun, 16 Feb 2020 22:48:12 +0000 (17:48 -0500)]
v4l2codecs: Fix bytesused value
Pass the actual amount of bytes we have copied into the bitstream buffer. Also
unmap the memory before queuing.
Nicolas Dufresne [Sun, 16 Feb 2020 19:53:08 +0000 (14:53 -0500)]
v4l2codech264dec: Implement finish frame
This enables the request to be processed by the HW (STREAMON) and to be pushed
downstream for further processing.
Nicolas Dufresne [Sun, 16 Feb 2020 19:51:37 +0000 (14:51 -0500)]
v4l2codecdecoder: Add support for dequeuing buffers
This includings polling of the request and streamon/streamoff.
Nicolas Dufresne [Sat, 15 Feb 2020 03:13:05 +0000 (22:13 -0500)]
v4l2codech264dec: Implement queuing
We now pass the controls, associated to a request, queue the bitstream, qeueue
a picture buffer to decode into and finally queue the request. This now runs
until the buffer pool is exhausted. The next step will be to dequeue.
Nicolas Dufresne [Sat, 15 Feb 2020 03:03:17 +0000 (22:03 -0500)]
v4l2decoder: Add helpers to queue buffer and requests
Nicolas Dufresne [Sat, 15 Feb 2020 02:57:34 +0000 (21:57 -0500)]
v4l2codec: Add getter for buffer index in allocator and pool
Nicolas Dufresne [Thu, 13 Feb 2020 22:46:29 +0000 (17:46 -0500)]
v4l2h264codecdec: Copy bitstream parameter and data
In this patch we fill the control structure with the bitstream paramter and
copy the bitstream data into V4L2 memory. Slice paramters are only the subset
of what Hantro needs, without any support for interlaced content.
Nicolas Dufresne [Thu, 13 Feb 2020 04:50:42 +0000 (23:50 -0500)]
v4l2slh264dec: Use the newly introduced allocator / pool
We now allocate and free the memory needed for streaming.
Nicolas Dufresne [Thu, 13 Feb 2020 04:45:14 +0000 (23:45 -0500)]
v4l2codec: Add allocator and pool implementation
This is a pooling allocator and the buffer pool does nothing other then
reusing the GstBuffer structure. Note that the pool is an internal pool, so
the start/stop/set_config virtual functions are not implemented.
Nicolas Dufresne [Thu, 13 Feb 2020 04:39:25 +0000 (23:39 -0500)]
v4l2codec: decoder: Add allocation helper
Nicolas Dufresne [Mon, 10 Feb 2020 23:05:39 +0000 (18:05 -0500)]
v4l2codecs: Implement H264 format negotiation
Nicolas Dufresne [Mon, 10 Feb 2020 23:02:37 +0000 (18:02 -0500)]
v4l2codecs: Add initial formats helpers
Nicolas Dufresne [Sun, 9 Feb 2020 16:31:11 +0000 (11:31 -0500)]
v4l2codecs: Add skeleton of H264 decoder
This introduces the skeleton of the H264 decoder. The plugin will list the
devices and register a subclass of the GstV4L2CodecH264Dec base class. The
subclass will pick the required specific information from the GstV4L2Device
stored in the subclass structure.
Nicolas Dufresne [Thu, 6 Feb 2020 02:42:56 +0000 (21:42 -0500)]
v4l2codec: Add GstV4L2Deocder helper object
This is a GstObject which will be used to hold on media and video device file
descriptor and provide abstracted ioctl calls with these descriptor. At the
moment this helper contains just enough to enumerate the supported format.
This part will be used by the plugin to register the CODEC specific elements..
Nicolas Dufresne [Thu, 6 Feb 2020 02:18:57 +0000 (21:18 -0500)]
v4l2codecs: Copy all needed Linux kernel headers
Most of the features we need are very early or not expose yet in the uAPI.
Using an internal copy ensure that we everything we need is defined avoiding
to add load of checks and conditionnal code.
Nicolas Dufresne [Tue, 4 Feb 2020 20:52:45 +0000 (15:52 -0500)]
v4l2codecs: Add device enumeration
This introduces a GstV4L2CodecDevice structure and helper to retrieve a
list of CODEC device drivers. In order to find the device driver we
enumerate all media devices with UDEV. We then get the media controller
topology and locate a entity with function encoder or decoder and make
sure it is linked to two V4L2 IO entity pointing to the same device
node.
Nicolas Dufresne [Mon, 3 Feb 2020 15:06:46 +0000 (10:06 -0500)]
v4l2codecs: Introdude a V4L2 CODECs Accelerator plugin
This plugin will support new CODECs uAPI in the Linux kernel. These
consist of an accelerator interface similar to DXVA, NVDEC, VDPAU and
VAAPI.
Seungha Yang [Mon, 30 Mar 2020 11:21:37 +0000 (20:21 +0900)]
msdkh264enc: Port to h264parser API for frame packing SEI injection
Create frame packing SEI memory only once per set_format() and
reuse it if possible.
Seungha Yang [Tue, 24 Mar 2020 10:23:46 +0000 (19:23 +0900)]
h264parse: Print all the syntax elements of frame packing for debugging
Other values might be useful for debugging
Haihao Xiang [Wed, 18 Mar 2020 02:11:19 +0000 (10:11 +0800)]
msdkh265enc: support 8-bit 422 encoding
The media driver can support HEVC 8-bit 422 encoding for non-lowpower
mode since ICL[1], so VPP is not needed for this case.
Sample pipeline:
gst-launch-1.0 videotestsrc ! video/x-raw,format=YUY2 ! msdkh265enc ! \
filesink location=output.h265
[1] https://github.com/intel/media-driver#decodingencoding-features
Haihao Xiang [Tue, 17 Mar 2020 07:18:37 +0000 (15:18 +0800)]
msdkh265enc: add support for Y210 in sink pad
Sample pipeline:
gst-launch-1.0 videotestsrc ! video/x-raw,format=Y210 !
msdkh265enc ! filesink location=output.h265
Haihao Xiang [Fri, 20 Mar 2020 01:53:28 +0000 (09:53 +0800)]
msdk: add a helper function to get codename of the platform
The features supported in MSDK vary from platform to platform. We may
support some features based on the codename of the platform in future.
Seungha Yang [Tue, 24 Mar 2020 07:00:30 +0000 (16:00 +0900)]
msdkenc: Try to find corresponding codec frame for encoded output buffer
The input and output buffers should be matched as much as possible
so that various metadata and its ordering to be preserved.
Seungha Yang [Sun, 22 Mar 2020 11:15:42 +0000 (20:15 +0900)]
msdkh264enc: Add support for CEA708 closed caption insertion
Currently supported caption format is CEA708_RAW
Seungha Yang [Wed, 25 Mar 2020 08:20:13 +0000 (17:20 +0900)]
h264parser: Add a helper method to create and inject raw SEI data
Add an API to create raw SEI nal unit. This would be useful in case
an user want to create SEI nal data and inject the SEI nal data
into bitstream.
Seungha Yang [Thu, 19 Mar 2020 09:25:18 +0000 (18:25 +0900)]
nalutils: Introduce NAL writer helper methods
Add helper methods for writing h264 and h265 NAL
Seungha Yang [Sun, 22 Mar 2020 04:59:52 +0000 (13:59 +0900)]
h264parser: Expose SEI clear function to public
gsth265parser does it already. Although corresponding API of h265parser is
gst_h265_sei_free, _clear suffix is more consistent naming for h264parser
since there are gst_h264_{sps,pps}_clear().
Seungha Yang [Sun, 29 Mar 2020 08:34:48 +0000 (17:34 +0900)]
d3d11decoder: Add support for array of texture DPB
DXVA supports two kinds of texture structure for DPB, one is
"1) texture array" and the other is "2) array of texture".
1) is a type of texture which is single ID3D11Texture2D object having
ArraySize greater than one. So the ID3D11Texture2D itself is a set of texture.
Each sub texture of this type mush have identical resolution, format and so on,
and the number of sub texture in a texture array is fixed.
2) is an array of usual ID3D11Texture2D object. That means each
ID3D11Texture2D is independent each other and might have different resolution as well.
Moreover, we can modify the number of frames of the array dynamically.
This type is more flexible than "1) texture array" in terms of dynamic
behavior and also this type of texture can be used for shader resource view
but "1) texture array" couldn't be.
If "2) array of texture" is supported by driver, DXVA spec is saying that
it's preferred format over "1) texture array" in terms of performance.
Haihao Xiang [Tue, 26 Nov 2019 07:42:03 +0000 (15:42 +0800)]
msdkh265enc: slice size control
Add max-slice-size property to set the maximum slice size
Seungha Yang [Fri, 27 Mar 2020 09:05:55 +0000 (18:05 +0900)]
Revert "d3d11decoder: Check decoder status report"
This reverts commit
418e6991c19ab5ae6bcf1890f9b947c748ca834d.
Not all drivers seem to be friendly to this API.
Revert to avoid incomprehensible crash with Nvidia.
Seungha Yang [Thu, 26 Mar 2020 12:20:20 +0000 (21:20 +0900)]
d3d11window: Always reuse swapchain
DXGI format can be updated via ResizeBuffers()
Seungha Yang [Mon, 9 Mar 2020 11:29:17 +0000 (20:29 +0900)]
d3d11window: Reorganize display colorspace selection
The set of supported color space by DXGI is not full combination of
our colorimetry. That means we should convert color space to one
of supported color space by DXGI. This commit modifies the color space
selection step so that d3d11window can find the best matching DXGI color space
first and then the selected input/output color space will be referenced
by shader and/or d3d11videoprocessor.
Seungha Yang [Fri, 27 Mar 2020 06:40:00 +0000 (15:40 +0900)]
rtmp2src: Add idle-timeout property
Add new property to signalling that there is no incoming data
from peer. This can be useful if users want to stop the streaming
when the connection is alive but no packet is arriving.
Matthew Waters [Thu, 26 Mar 2020 23:34:25 +0000 (10:34 +1100)]
dtls/connection: fix EOF handling with openssl 1.1.1e
openssl 1.1.1e does some stricker EOF handling and will throw an error
if the EOF is unexpected (like in the middle of a record). As we are
streaming data into openssl here, it is entirely possible that we push
data from multiple buffers/packets into openssl separately.
From the openssl changelog:
Changes between 1.1.1d and 1.1.1e [17 Mar 2020]
*) Properly detect EOF while reading in libssl. Previously if we hit an EOF
while reading in libssl then we would report an error back to the
application (SSL_ERROR_SYSCALL) but errno would be 0. We now add
an error to the stack (which means we instead return SSL_ERROR_SSL) and
therefore give a hint as to what went wrong.
[Matt Caswell]
We can relax the EOF signalling to only return TRUE when we have stopped
for any reason (EOS, error).
Will also remove a spurious EOF error from previous openssl version.
Matthew Waters [Wed, 25 Mar 2020 03:46:15 +0000 (14:46 +1100)]
webrtc: mark streams as active on renegotiation as well.
Otherwise when bundling, only the changed streams would be considered as
to whether the bundled transport needs to be blocked as all streams are
inactive.
Scenario is one transceiver changes direction to inactive and as that is
the only change in transciever direction, the entire bundled transport would
be blocked even if there are other active transceivers inside the same bundled
transport that are still active.
Fix by always checking the activeness of a stream regardless of if the
transceiverr has changed direction.
Guillaume Desmottes [Mon, 13 Jan 2020 08:20:26 +0000 (13:50 +0530)]
interlace: add alternate support
Allow downstream elements to negotiate the alternate interlace mode,
splitting each input buffer in two fields, each having their own buffer.
Guillaume Desmottes [Fri, 28 Feb 2020 07:01:46 +0000 (12:31 +0530)]
interlace: factor out interlace_mode_from_pattern()
Guillaume Desmottes [Thu, 23 Jan 2020 09:47:45 +0000 (15:17 +0530)]
interlace: factor out gst_interlace_push_buffer()
Guillaume Desmottes [Thu, 23 Jan 2020 09:40:36 +0000 (15:10 +0530)]
interlace: factor out gst_interlace_decorate_buffer_ts()
Guillaume Desmottes [Thu, 23 Jan 2020 09:01:02 +0000 (14:31 +0530)]
interlace: rename copy_field()
It is actually copying both fields (to a single frame/buffer).
Seungha Yang [Sat, 21 Mar 2020 11:16:04 +0000 (20:16 +0900)]
tests: h265parser: Add test for registered user data SEI
Seungha Yang [Sat, 21 Mar 2020 10:41:28 +0000 (19:41 +0900)]
h265parser: Fix registered user data SEI leak
... and add fix for the SEI data in gst_h265_sei_copy()
Seungha Yang [Sat, 21 Mar 2020 09:25:25 +0000 (18:25 +0900)]
h265parser: Do not allocate too large size of memory for registered user data SEI
Don't be confused by the unit of payload size (i.e., bits and bytes)
Also this need a documentation with Since mark
Seungha Yang [Sat, 21 Mar 2020 09:39:42 +0000 (18:39 +0900)]
h264parser: Do not allocate too large size of memory for registered user data SEI
Don't be confused by the unit of payload size (i.e., bits and bytes)
Also this need a documentation with Since mark
Philippe Normand [Mon, 23 Mar 2020 12:55:13 +0000 (12:55 +0000)]
wpe: Enable SHM support for new stable WPEBackend-FDO release
1.5.0 was the development version.
Philippe Normand [Mon, 10 Feb 2020 18:28:05 +0000 (18:28 +0000)]
wpe: Mouse scroll events support
James Westman [Sun, 1 Mar 2020 04:19:57 +0000 (22:19 -0600)]
applemedia: Add some properties to GstAvfDevice
Adds properties to the devices listed in GstDeviceMonitor by the
applemedia plugin.
These properties are:
- device.api (always set to "avf")
- avf.unique_id
- avf.model_id
- avf.manufacturer (except on iOS)
- avf.has_flash
- avf.has_torch
Everything except device.api is taken directly from the AVCaptureDevice object
provided by AVFoundation.
Seungha Yang [Fri, 20 Mar 2020 11:27:05 +0000 (20:27 +0900)]
tsdemux: Set mpegversion for AAC ADTS stream based on parsed ADTS header
Both 2 and 4 are supported version of AAC ADTS format stream.
So we need to set correct version to help negotiation
especially for non-autopluggable pipeline.
Haihao Xiang [Tue, 10 Mar 2020 07:42:48 +0000 (15:42 +0800)]
msdkh265dec: add support for main-12, main-12-intra profiles
The video format is P012_LE
Haihao Xiang [Tue, 10 Mar 2020 08:03:55 +0000 (16:03 +0800)]
msdk: map P012_LE to VA_FOURCC_P016
In media driver, VA_FOURCC_P016 is used for semi-planar 12 bits YUV
format, the corresponding RT format is VA_RT_FORMAT_YUV420_12
Haihao Xiang [Tue, 10 Mar 2020 07:41:50 +0000 (15:41 +0800)]
msdk: map MFX_FOURCC_P016 to VA_FOURCC_P016
P016 is used for semi-planar 12 bits format in the media driver, so the
RT format is VA_RT_FORMAT_YUV420_12
Haihao Xiang [Tue, 10 Mar 2020 07:32:43 +0000 (15:32 +0800)]
msdk: map P012_LE to MFX_FOURCC_P016
MFX_FOURCC_P016 is used for semi planar 12 bits YUV format in MediaSDK
Haihao Xiang [Fri, 20 Mar 2020 06:53:40 +0000 (14:53 +0800)]
msdkvp9enc: add support for 10 bits 444 input
The input format is Y410 in sink pad and the corresponding profile in
src pad is profile 3
Haihao Xiang [Fri, 20 Mar 2020 06:38:12 +0000 (14:38 +0800)]
msdkvp9enc: add support for 8 bits 444 input
The input format is VUYA in sink pad and the corresponding profile in
src pad is profile 1
Seungha Yang [Sun, 22 Mar 2020 13:30:15 +0000 (22:30 +0900)]
msdkbufferpool: Fix build warning on Windows
gstmsdkbufferpool.c(274): warning C4101: 'fd': unreferenced local variable
Guillermo Rodríguez [Wed, 22 Jan 2020 11:29:03 +0000 (11:29 +0000)]
examples/wayland: unref GstContext after use
U. Artie Eoff [Thu, 19 Mar 2020 16:28:28 +0000 (09:28 -0700)]
msdkvpp: add YV12 format to src pad
VPP YV12 output format support added since
MFX_VERSION 1032.
https://github.com/Intel-Media-SDK/MediaSDK/pull/2027
Seungha Yang [Sun, 15 Mar 2020 10:20:47 +0000 (19:20 +0900)]
nvdec: Add fallback for CUDA/OpenGL interop failure
It happens when local OpenGL context belongs to non-nvidia GPU.
Victor Manuel Jaquez Leal [Mon, 16 Mar 2020 16:19:04 +0000 (17:19 +0100)]
libs: codecparser: add missing unstable api warning
mpeg4 and vp8 parsers didn't show that compilation warning as the
others parsers.
Philippe Normand [Sun, 15 Mar 2020 12:44:31 +0000 (12:44 +0000)]
webrtcdsp: Fix documentation markup
Philippe Normand [Sun, 15 Mar 2020 10:46:24 +0000 (10:46 +0000)]
openni2: Remove spurious gtk-doc markers
Philippe Normand [Sun, 15 Mar 2020 10:45:25 +0000 (10:45 +0000)]
decklink: Remove spurious gtk-doc marker
Philippe Normand [Fri, 13 Mar 2020 17:47:46 +0000 (17:47 +0000)]
docs: Scan C++ files too
Haihao Xiang [Wed, 12 Feb 2020 03:23:18 +0000 (11:23 +0800)]
msdk: meson: check whether mfxExtVP9Param is defined
User reported MFX_VERSION is 1026 however the build is broken[1], so add
the check for mfxExtVP9Param to make sure it can be build without
compiler error.
In addtion, it fixes a stupid typo (#endif') introduced by me.
[1] https://gitlab.freedesktop.org/gstreamer/gst-plugins-bad/merge_requests/988#note_408093
rubenrua [Thu, 12 Mar 2020 11:13:10 +0000 (12:13 +0100)]
asfmux: Fix typo in property description
s/milisecs/milliseconds/g
Thibault Saunier [Thu, 12 Mar 2020 00:37:27 +0000 (21:37 -0300)]
timecodestamper: Plug a leak
Edward Hervey [Sat, 1 Feb 2020 10:52:04 +0000 (11:52 +0100)]
mpegts: Add a property to ignore broken PCR streams
Some mpeg-ts (HLS, DVB, ...) streams out there have completely broken
PCR streams on which we can't reliably recover correct timestamps.
For those, provide a property that will ignore the program PCR stream
(by faking that it's not present (0x1fff)).
Seungha Yang [Wed, 11 Mar 2020 08:18:56 +0000 (17:18 +0900)]
mpegdemux: Add ignore-scr property to ignore broken SCR
Some MPEG-PS streams might not be compliant but the SCR can be ignored
if PTS/DTS in PES header is consistently increased.
Seungha Yang [Wed, 11 Mar 2020 08:42:18 +0000 (17:42 +0900)]
mpegdemux: Remove whitespace
yychao [Tue, 10 Mar 2020 03:52:23 +0000 (11:52 +0800)]
smoothstreaming: fix H264 CodecPrivateData parsing
Do not pass SPS nal_unit_type (0x67) into gst_h264_parse_sps()
Fixes #648
Seungha Yang [Tue, 10 Mar 2020 08:40:28 +0000 (17:40 +0900)]
h265parse: In-band sps/pps update if only codec_data differs in src caps
Apply in-band sps/pps resending implementation to h265parse.
Seungha Yang [Tue, 10 Mar 2020 04:20:17 +0000 (13:20 +0900)]
h264parse: In-band sps/pps update if only codec_data differs in src caps
Initially the case "only codec_data is different" was addressed in
https://bugzilla.gnome.org/show_bug.cgi?id=705333 in order for
unusual bitstreams to be handled. That's the case where sps and pps
are placed in bitstream. When sps/pps are signalled only via caps
by upstream, however, the updated codec_data is mandatory for decoder
and therefore we shouldn't ignore them.
Dong Il Park [Tue, 10 Mar 2020 07:32:59 +0000 (16:32 +0900)]
tsdemux: Add format_identifier for AC4 codec
According to following spec document, add format_identifier for AC4 in tsdemux.
ETSI TS 103 190-2 V1.2.1 : Annex D : AC-4 in MPEG-2 transport stream
yychao [Mon, 9 Mar 2020 09:00:01 +0000 (17:00 +0800)]
tsdemux: Add support for AC4
According to following two specs, add support for AC4 in tsdemux.
1. ETSI TS 103 190-2 V1.2.1 (2018-02) : Annex D (normative): AC-4 in MPEG-2 transport streams
2. ETSI EN 300 468 V1.16.1 (2019-08) : Annex D (normative):Service information implementation of AC-3, EnhancedAC-3, and AC-4 audio in DVB systems
Sebastian Dröge [Mon, 9 Mar 2020 19:21:19 +0000 (21:21 +0200)]
webrtcbin: Use GPtrArrays or store items inline instead of using GArrays of pointers
Jan Schmidt [Mon, 9 Mar 2020 18:18:40 +0000 (05:18 +1100)]
webrtcbin: Prevent ICE gathering state reaching complete early
The ICE gathering state can transition to complete prematurely if the
underlying ICE components complete their gathering while the initial
ICE gathering state task is queued and still pending.
In that situation, the ice gathering state task will report complete
while there are still ICE candidates queued for emission.
Prevent that by storing ICE candidates in an array and checking if
there are any pending before reporting a completed ICE gathering
state.
Jan Schmidt [Mon, 9 Mar 2020 17:22:57 +0000 (04:22 +1100)]
webrtc: Protect the pending ICE candidates array
ICE candidates can be added to the array directly from the application
or from the webrtc main loop. Rename it to make it clear that it's
holding remote ICE candidates from the peer, and protect it with a
new mutex
Seungha Yang [Sat, 7 Mar 2020 10:41:53 +0000 (19:41 +0900)]
d3d11vp9dec: Add support for internal frame resizing
VP9 codec allows resizing reference frame by spec. Handling this case
is a bit tricky especially when the resizing happens on non-keyframe,
because pre-allocated decoder textures (i.e., dpb) have negotiated
resolution and to change resolution meanwhile decoding on non-keyframe,
each texture might need to be re-created, copied to new dpb somehow,
and re-negotiated with downstream.
Due to the complicated requirement of negotiation driven
resizing handling, this commit adds shader into d3d11decoder object
to resize only corresponding frames. Note that if the resolution change
is detected on keyframe, decoder will re-negotiate with downstream.
Seungha Yang [Sat, 7 Mar 2020 10:38:30 +0000 (19:38 +0900)]
d3d11decoder: Set GstVideoAlignment to downstream d3d11 buffer pool
To copy decoder output texture to another d3d11 texture, the downstream
texture needs to be aligned too.
Seungha Yang [Sat, 7 Mar 2020 07:42:14 +0000 (16:42 +0900)]
d3d11decoder: Use aligned resolution for staging texture
Not only any textures for decoder output view, any destination texture
which would be copied from decoder output texture need to be aligned too.
Otherwise driver sometimes crashed/hung (not sure why).
Seungha Yang [Sat, 7 Mar 2020 09:29:20 +0000 (18:29 +0900)]
d3d11colorconverter: Add support cropping
Vertex buffer will be updated if input texture resolution is different
from GstVideoInfo or when an user requests specific crop area.
Seungha Yang [Sat, 7 Mar 2020 06:10:43 +0000 (15:10 +0900)]
d3d11bufferpool: Add padding space for semi-planar YUV formats
Resolution of NV12, P010, and P016 formats must be multiple of two.
Otherwise texture cannot be created. Instead of doing this alignment
per API consumer side, do this in buffer pool for simplicity.
Seungha Yang [Sun, 8 Mar 2020 07:10:41 +0000 (16:10 +0900)]
h265decoder: Pass max_dpb_size to new_sequence vfunc
same as we are doing in h264decoder
Seungha Yang [Sun, 8 Mar 2020 07:03:35 +0000 (16:03 +0900)]
d3d11h264dec: Make use of max_dpb_size passed by baseclass
Most likely this modification can save GPU memory.
Seungha Yang [Wed, 26 Feb 2020 11:24:14 +0000 (20:24 +0900)]
d3d11h265dec: Fix scaling list parsing
Scaling list can be signalled via sps or pps. Decoder should check
both of them.
Seungha Yang [Wed, 26 Feb 2020 08:44:52 +0000 (17:44 +0900)]
d3d11h265dec: Properly signal the ucNumDeltaPocsOfRefRpsIdx
ucNumDeltaPocsOfRefRpsIdx should be the NumDeltaPocs[RefRpsIdx].
Seungha Yang [Wed, 26 Feb 2020 08:33:59 +0000 (17:33 +0900)]
h265parser: Store NumDeltaPocs of reference for hardware accelerators
That's the value of NumDeltaPocs[RefRpsIdx] and we might be able to derive
the value from given sps and slice header.
Because well known hardware implementations refer to the value, however,
storing the value makes things easier.
Following is the list of hardware implementations
* DXVA2: ucNumDeltaPocsOfRefRpsIdx
* NVDEC/VDPAU: NumDeltaPocsOfRefRpsIdx
Seungha Yang [Mon, 24 Feb 2020 11:53:49 +0000 (20:53 +0900)]
h265decoder: Fix for output and removal picture from DPB
See C.5.2.2 Output and removal of pictures from the DPB.
If the number of pictures in the DPB is greater than or equal to
sps_max_dec_pic_buffering_minus1[HighestTid] + 1, then the picture
should be outputted.
Seungha Yang [Fri, 6 Mar 2020 07:40:50 +0000 (16:40 +0900)]
meson: codecs: Don't install header
This library is not intended to be public. APIs are likely to
change over time and should not be disclosed to people yet.
Seungha Yang [Thu, 5 Mar 2020 05:40:28 +0000 (14:40 +0900)]
h264decoder: Rename some internals
No more a part of d3d11 implementation