Commit graph

780 commits

Author SHA1 Message Date
Guillaume Desmottes
1e9d7a6a23 omxvideoenc: include vertical padding in nFilledLen when copying
According to the OMX spec (3.1.3.7.1) nFilledLen is meant to include any
padding. We use to include the horizontal one (stride) but not the
vertical one if nSliceHeight is bigger than the actual height.

The calculated nFilledLen was wrong as it didn't include the padding
between planes.

https://bugzilla.gnome.org/show_bug.cgi?id=796749
2018-07-05 16:08:03 +02:00
Guillaume Desmottes
c8969b0dbe omxvideoenc: implement decide_allocation
Increase the number of output buffers by the number of buffers requested
downstream.
Prevent buffers starvation if downstream is going to use dynamic buffer
mode on its input.

https://bugzilla.gnome.org/show_bug.cgi?id=795746
2018-06-08 09:53:01 +02:00
Guillaume Desmottes
431eac07bf omxvideodec: implement propose_allocation
Tell upstream about how many buffer we plan to use so they can adjust
their own number of buffers accordingly if needed.

Same logic as the existing gst_omx_video_enc_propose_allocation().

https://bugzilla.gnome.org/show_bug.cgi?id=795746
2018-06-08 09:53:01 +02:00
Guillaume Desmottes
84483f3d7a omxvideoenc: always signal drain cond when stopping streaming loop
Similar change as the one I just did in omxvideodec.

https://bugzilla.gnome.org/show_bug.cgi?id=796207
2018-06-08 09:47:32 +02:00
Guillaume Desmottes
fd108f4278 omxvideodec: always signal drain cond when stopping streaming loop
If for some reason something goes wrong and we stop the streaming loop
we may end up with other threads still waiting on the drain cond.
No more buffers will be produced by the component so they were waiting
forever.

Fix this by always signalling this cond when stopping the streaming
loop.

https://bugzilla.gnome.org/show_bug.cgi?id=796207
2018-06-08 09:47:32 +02:00
Guillaume Desmottes
798bbc9acf omxvideoenc: factor out gst_omx_video_enc_pause_loop()
No semantic change. I'm going to use it in more failure cases.

https://bugzilla.gnome.org/show_bug.cgi?id=796207
2018-06-08 09:47:32 +02:00
Guillaume Desmottes
9a8e863862 omxvideodec/enc: add hack updating nBufferCountActual before allocating
The OMX specs states that the nBufferCountActual of a port has to default
to its nBufferCountMin. If we don't change nBufferCountActual we purely rely
on this default. But in some cases, OMX may change nBufferCountMin before we
allocate buffers. Like for example when configuring the input ports with the
actual format, it may decrease the number of minimal buffers required.
This method checks this and update nBufferCountActual if needed so we'll use
less buffers than the worst case in such scenarios.

SetParameter() needs to be called when the port is either disabled or
the component in the Loaded state.

Don't do this for the decoder output as
gst_omx_video_dec_allocate_output_buffers() already check
nBufferCountMin when computing the number of output buffers.

On some platform, like rpi, the default nBufferCountActual is much
higher than nBufferCountMin so only enable this using a specific gst-omx
hack.

https://bugzilla.gnome.org/show_bug.cgi?id=791211
2018-06-08 09:44:49 +02:00
Guillaume Desmottes
338ff34e5e omxvidee{enc,dec}: refresh input port definition after setting format
Setting the input format and the associated encoder/decoder settings
may also affect the nBufferCountMin of the input port.
Refresh the input port so we'll use up to date values in propose/decide
allocation.

https://bugzilla.gnome.org/show_bug.cgi?id=796445
2018-06-08 09:38:57 +02:00
Guillaume Desmottes
3498f81d1f omx: always consider component in 'invalid' state when an error occured
gst_omx_component_get_state() used to early return if there was no
pending state change. So if the component raised an error it wasn't
considered in the invalid state until the next requested state change.

Fix this by checking first if we received an error.

https://bugzilla.gnome.org/show_bug.cgi?id=795874
2018-06-08 09:36:56 +02:00
Guillaume Desmottes
f706f3b73b omxvideodec: pass a GstOMXBufferMode to gst_omx_buffer_pool_new()
The output_mode is supposed to be a GstOMXBufferMode, not a boolean.
2018-05-04 13:20:55 +02:00
Guillaume Desmottes
b644699c8d omxh26{4,5}enc: don't pick default 10-bit profile
The OMX stack of the zynqultrascaleplus (the only one supporting
NV12_10LE32 and NV16_10LE32) will now pick the proper profile if none
has been requested. Best to rely on its default than hardcoding a
specific one in gst-omx.

https://bugzilla.gnome.org/show_bug.cgi?id=794319
2018-04-26 10:07:54 +02:00
Guillaume Desmottes
83e580bd6b omxh264: sync with supported profiles on zynqultrascaleplus
Add extra supported AVC profiles and remove extended and 4:4:4 profiles
which are actually not implemented.

https://bugzilla.gnome.org/show_bug.cgi?id=794177
2018-04-25 09:20:02 +02:00
Guillaume Desmottes
7242d0a2e2 omxh264: factor out gst_omx_h264_utils_get_profile_from_enum()
Move the profile <-> enum mapping to one place. Make changes easier as
I'm about to add extra profiles.

No semantic change.

https://bugzilla.gnome.org/show_bug.cgi?id=794177
2018-04-25 09:20:02 +02:00
Guillaume Desmottes
4e75bab796 omxh265: add format range extension profiles on zynqultrascaleplus
The zynqultrascaleplus OMX gained support for more format range
extensions profiles (A.3.5).

https://bugzilla.gnome.org/show_bug.cgi?id=794177
2018-04-25 09:20:02 +02:00
Guillaume Desmottes
f9922a1a56 omxh265: factor out gst_omx_h265_utils_get_profile_from_enum()
Move the profile <-> enum mapping to one place. Make changes easier as
I'm about to add some profiles.

No semantic change.

https://bugzilla.gnome.org/show_bug.cgi?id=794177
2018-04-25 09:20:02 +02:00
Guillaume Desmottes
dc0ed29a62 omxvideoenc: add NV16 support
NV16 format wasn't supported on encoder input while it was on decoder
output.

https://bugzilla.gnome.org/show_bug.cgi?id=794175
2018-04-25 09:15:38 +02:00
Guillaume Desmottes
d40c6c0449 omxvideo: display port number when listing supported formats
More convenient when debugging.

https://bugzilla.gnome.org/show_bug.cgi?id=794175
2018-04-25 09:15:38 +02:00
Guillaume Desmottes
72cb1943da omxvideoenc: restore OMX default target-bitrate if requested by user
0xffffffff is the magic number in gst-omx meaning 'the default value
defined in OMX'. This works fine with OMX parameters which are only set
once when starting the component but not with configs which can be
changed while PLAYING.
Save the actual OMX default bitrate so we can restore it later if user
sets back 0xffffffff on the property.

Added GST_OMX_PROP_OMX_DEFAULT so we stop hardcoding magic numbers
everywhere.

https://bugzilla.gnome.org/show_bug.cgi?id=794998
2018-04-23 10:41:10 +02:00
Guillaume Desmottes
542faf0f36 omxvideoenc: use gst_omx_video_enc_set_bitrate() when setting bitrate in set_format
We weren't using the usual pattern when re-setting the bitrate:
- get parameters from OMX
- update only the fields different from 0xffffffff (OMX defaults)
- set parameters

Also added a comment explaining why we re-set this param.

https://bugzilla.gnome.org/show_bug.cgi?id=794998
2018-04-23 10:35:48 +02:00
Guillaume Desmottes
256f77b621 omxvideoenc: factor out gst_omx_video_enc_set_bitrate()
No semantic change, I'm about to re-use this function in set_format().

https://bugzilla.gnome.org/show_bug.cgi?id=794998
2018-04-23 10:35:48 +02:00
Guillaume Desmottes
e1a149f3d5 omxvideoenc/dec: fix handling of component enabling failing
- Report the error from OMX if any (OMX_EventError)
- If not report the failing to the application (GST_ELEMENT_ERROR)
- return GST_FLOW_ERROR rather than FALSE
- don't leak @frame

https://bugzilla.gnome.org/show_bug.cgi?id=795352
2018-04-18 14:01:39 +02:00
Guillaume Desmottes
021b66878a log_omx_performance: convert pointers to strings
G_TYPE_POINTER are not serialized in logs.

https://bugzilla.gnome.org/show_bug.cgi?id=794331
2018-04-06 13:39:42 +02:00
Guillaume Desmottes
dec9c8c385 omxvideoenc: remove duplicated debug message
We already have the exact same message at the beginning of
gst_omx_video_enc_handle_frame(). Having it twice is confusing when
reading/grepping logs.

I kept the earlier one to keep the symetry with
gst_omx_video_dec_handle_frame().

https://bugzilla.gnome.org/show_bug.cgi?id=794897
2018-04-02 15:16:33 +02:00
Guillaume Desmottes
530a7537cf omxvideoenc: add 'roi' qp-mode on zynqultrascaleplus
New QP mode used to handle ROI metadata.

https://bugzilla.gnome.org/show_bug.cgi?id=793696
2018-03-28 10:09:20 +02:00
Guillaume Desmottes
5b55041bb4 omxh265: update 422 profile names
h265parse is gaining support for the format range extension profile
(bgo#793876).
Use the profile names defined in h265parse.

https://bugzilla.gnome.org/show_bug.cgi?id=793928
2018-03-05 13:59:22 -05:00
Nicolas Dufresne
4e9dca0761 omxvideoenc: Don't drop the frame on empty payload
This otherwise may lead to "No reference frame found" warning.
2018-03-05 13:50:08 -05:00
Nicolas Dufresne
fb455242e9 omxvideodec: Don't drop the frame on empty payload
This otherwise may lead to "No reference frame found" warning.
2018-03-05 13:50:08 -05:00
Nicolas Dufresne
32660fd294 omx: Free empty buffers list in use_dynamic_buffers
To indicate we are doing dynamic buffers importation, we pass
a list of NULL pointers, but we forgot to free that list.
2018-03-05 13:50:08 -05:00
Nicolas Dufresne
4bc3b6e567 omxvideodec: Fix CodecState leak 2018-03-05 13:46:56 -05:00
Nicolas Dufresne
5eed1a7eb2 Revert "omx: wait for flush complete and buffers being released when flushing"
This reverts commit 4211e4c29a.
2018-03-02 11:06:08 -05:00
Guillaume Desmottes
f861ad2bc5 videoenc: don't set stride padding to 0 when copying frames
Padding can be left undefined there is no point filling it with 0.

https://bugzilla.gnome.org/show_bug.cgi?id=793694
2018-02-28 08:32:26 -05:00
Guillaume Desmottes
ea2df994f3 add support for NV12_10LE32 and NV16_10LE32 on zynqultrascaleplus
The encoder and decoder on zynqultrascaleplus support these new 10 bits
format.

https://bugzilla.gnome.org/show_bug.cgi?id=793694
2018-02-28 08:32:26 -05:00
Guillaume Desmottes
d1ffc97ddd omxvideoenc: factor out gst_omx_video_enc_nv12_manual_copy()
No semantic change, I'm going to re-use it to copy the NV12_10LE32
format.

https://bugzilla.gnome.org/show_bug.cgi?id=793694
2018-02-28 08:32:26 -05:00
Guillaume Desmottes
f2570512cf omxvideoenc: display the computed buffer size when configuring input
https://bugzilla.gnome.org/show_bug.cgi?id=793694
2018-02-28 08:32:15 -05:00
Guillaume Desmottes
6921568161 videoenc: implement ROI on zynqultrascaleplus
Check input buffers for ROI meta and pass them to the encoder by using
zynqultrascaleplus's custom OMX extension. Also add a new
"default-roi-quality" in order to tell the encoder what quality level
should be applied to ROI by default.

https://bugzilla.gnome.org/show_bug.cgi?id=793696
2018-02-21 12:31:39 -05:00
Guillaume Desmottes
a8c3996317 omxvideoenc: remove GST_PARAM_MUTABLE_PLAYING from 'max-bitrate' property
This property isn't actually mutable in the PLAYING state.

https://bugzilla.gnome.org/show_bug.cgi?id=793458
2018-02-21 11:31:13 +00:00
Guillaume Desmottes
2d3816fabc omxvideoenc: protect target_bitrate with the object lock
The 'target-bitrate' property can be changed while PLAYING
(GST_PARAM_MUTABLE_PLAYING). Make it thread-safe to prevent concurrent
accesses between the application and streaming thread.

https://bugzilla.gnome.org/show_bug.cgi?id=793458
2018-02-21 11:24:52 +00:00
Guillaume Desmottes
412df8c74f omxbufferpool: add PERFORMANCE DEBUG message when copying output frames
I spent quiet some time figuring out why performance of my pipeline were
terrible. Turned out it was because of output frames being copied
because of stride/offset mismatch.

Add a PERFORMANCE DEBUG message to make it easier to spot and debug from logs.

https://bugzilla.gnome.org/show_bug.cgi?id=793637
2018-02-21 11:22:05 +00:00
Guillaume Desmottes
7cf70e650b omxh265enc: fix typo in "periodicty-idr" property name
Also fix the 'nick' of the property.

omxh265enc is based on the code from omxh264enc and suffers the same
typo as we fixed in https://bugzilla.gnome.org/show_bug.cgi?id=784370

This element isn't part of a stable release yet so it's not an API
break.

https://bugzilla.gnome.org/show_bug.cgi?id=793390
2018-02-12 15:47:19 +00:00
Guillaume Desmottes
ab181a4335 include all OMX extension headers if present
The OMX specs defines 8 headers that implementations can use to define
their custom extensions. We were checking and including 3 and ignoring
the other ones.

https://bugzilla.gnome.org/show_bug.cgi?id=792043
2018-01-30 11:54:37 +00:00
Guillaume Desmottes
13a43757f5 meson: simplify OMX extensions detection
We are now always checking which files are present or not, even when using our
internal copy of OMX, rather than hardcoding the ones present in it.

https://bugzilla.gnome.org/show_bug.cgi?id=792043
2018-01-30 11:54:37 +00:00
Nicolas Dufresne
f0469c99d2 Revert "meson: use include_directories() with external OMX headers path"
This reverts commit 9d37a92a61.
2018-01-30 11:54:24 +00:00
Guillaume Desmottes
8bd2b1f550 omxh265enc: add some encoding properties
constrained-intra-prediction and loop-filter-mode.
Those map standard OMX settings.

https://bugzilla.gnome.org/show_bug.cgi?id=792528
2018-01-30 11:51:10 +00:00
Guillaume Desmottes
74376fea2d omxh264enc: add some encoding properties
entropy-mode, constrained-intra-prediction and loop-filter-mode.
Those map standard OMX settings.

https://bugzilla.gnome.org/show_bug.cgi?id=792528
2018-01-30 11:51:10 +00:00
Guillaume Desmottes
e3396210d2 omxvideoenc: add zynqultrascaleplus specific properties
https://bugzilla.gnome.org/show_bug.cgi?id=792528
2018-01-30 11:51:10 +00:00
Guillaume Desmottes
d14b9ffa64 omxvideoenc: document unit of target-bitrate property
nTargetBitrate and nEncodeBitrate are defined in bits per second in the
OMX spec.

https://bugzilla.gnome.org/show_bug.cgi?id=792528
2018-01-30 11:51:10 +00:00
Guillaume Desmottes
52de8eaca0 omxvideodec: add internal-entropy-buffers property on zynqultrascaleplus
Custom property to control the number of internal buffers used in the
decoder to smooth out entropy decoding performance.

https://bugzilla.gnome.org/show_bug.cgi?id=792528
2018-01-30 11:51:10 +00:00
Guillaume Desmottes
9d37a92a61 meson: use include_directories() with external OMX headers path
It seems cleaner to use the proper meson tools to include this path
rather than manually tweak the build flags.

This also allows us to simplify the OMX extensions detection code. We
are now always checking which files are present, even when using our
internal copy of OMX, rather than hardcoding the ones present in it.

https://bugzilla.gnome.org/show_bug.cgi?id=792043
2018-01-30 11:10:38 +00:00
Guillaume Desmottes
6c57d06ee1 omxvideoenc: expose chroma format and bit depth in output caps
As we added in the parser (bgo#792039) expose the chroma and bit
depth information in output caps.

https://bugzilla.gnome.org/show_bug.cgi?id=792040
2018-01-30 09:23:59 +00:00
Guillaume Desmottes
1990580d79 omxvideoenc: factor out get_output_caps()
No semantic change so far.

https://bugzilla.gnome.org/show_bug.cgi?id=792040
2018-01-30 09:23:59 +00:00
Guillaume Desmottes
2ca3cf5cfa omxvideodec: add hack to pass color format from caps to OMX decoder
This hack tries to pass as much information as possible from caps to the
decoder before it receives any buffer. These information can be used by
the OMX decoder to, for example, pre-allocate its internal buffers
before starting to decode and so reduce its initial latency.

This mechanism is currently supported by the zynqultrascaleplus decoder.

https://bugzilla.gnome.org/show_bug.cgi?id=792040
2018-01-30 09:23:59 +00:00
Guillaume Desmottes
0f702d0503 log failing OMX calls as errors
I find it confusing when debugging that OMX calls returning an error
where not logged as GST_LEVEL_ERROR making them harder to spot.
Fix this by introducing simple log macros checking the return value of
the OMX call and logging failures as errors.

https://bugzilla.gnome.org/show_bug.cgi?id=791069
2018-01-29 17:47:56 +00:00
Guillaume Desmottes
65cc312aa8 add OMX_PERFORMANCE debug category
Can be used to log buffers exchange between OMX and gst-omx to profile
performances of the OMX component.
Ideally this should be done using tracer hooks but it's currently not
possible to define custom hooks outside of core.

Use GST_DEBUG="OMX_PERFORMANCE:8" to enable it.
See also
https://github.com/gdesmott/gst-log-parser/blob/master/src/bin/omx-perf.rs
as a simple program consuming those logs to generate gnuplot files and
stats.

https://bugzilla.gnome.org/show_bug.cgi?id=791093
2018-01-29 17:16:35 +00:00
Guillaume Desmottes
9f87b987ef omxvideoenc: implement dmabuf import on zynqultrascaleplus
The Zynq UltraScale+ encoder implements a custom OMX extension to
directly import dmabuf saving the need of mapping input buffers.

This can be use with either 'v4l2src io-mode=dmabuf' or an OMX video
decoder upstream.

https://bugzilla.gnome.org/show_bug.cgi?id=792361
2018-01-29 15:21:02 +00:00
Guillaume Desmottes
0c5bbbb6ca omxvideoenc: drop late input frames if QoS is enabled
Make use of the new GstVideoEncoder QoS API to drop late input frames. This may
help a live pipeline to catch up if it's being late and all frames end up
being dropped at the sink.

https://bugzilla.gnome.org/show_bug.cgi?id=792783
2018-01-22 17:04:27 +00:00
Guillaume Desmottes
5fa96cab7b omxvideoenc: early return in fill_buffer() if something goes wrong
If something goes wrong while trying to manually copy the input buffer,
the 'break' was moving us out of the 'for' loop but not out of the switch block.
So we ended up calling gst_video_frame_unmap() a second time (raising
assertions) and returning TRUE rather than FALSE.

Reproduced with a WIP zynqultrascaleplus OMX branch reporting wrong
buffer sizes and so triggering this bug.

https://bugzilla.gnome.org/show_bug.cgi?id=792167
2018-01-07 14:20:56 +00:00
Julien Isorce
6810878193 omxvideodec: ignore very little variations of the framerate
If less than 1%.

The dynamic format change should not happen when the
resolution does not change and when only the framerate
changes but very slightly, i.e. from 50000/1677=29.81
to 89/3=29.66 so a "percentage change" of less than 1%
(i.e. 100*(29.81-29.66)/29.66 = 0.50 < 1 ). In that case
just ignore it to avoid unnecessary renegotiation.

https://bugzilla.gnome.org/show_bug.cgi?id=759043
2017-12-14 09:20:49 +00:00
Guillaume Desmottes
7048134fa9 omxvideodec: use dynamic buffer mode on input if possible
Prevent from copying the input buffers between GStreamer and OMX.

Tested on zynqultrascaleplus and rpi (without dynamic buffers).

https://bugzilla.gnome.org/show_bug.cgi?id=787093
2017-12-14 09:00:35 +00:00
Guillaume Desmottes
533ee7fadc omxvideoenc: use dynamic buffer mode on input if possible
If the OMX component supports dynamic buffer mode and the input buffers
are properly aligned avoid copying each input frame between OMX and
GStreamer.

Tested on zynqultrascaleplus and rpi (without dynamic buffers).

https://bugzilla.gnome.org/show_bug.cgi?id=787093
2017-12-14 09:00:31 +00:00
Guillaume Desmottes
e55675b7ce omxvideoenc/dec: factor out input buffer allocation
No semantic change so far. I'm going to add an alternate way to allocate
input buffers.

https://bugzilla.gnome.org/show_bug.cgi?id=787093
2017-12-14 09:00:27 +00:00
Guillaume Desmottes
da07a647b8 omx: add API to implement dynamic buffers support
OMX 1.2.0 introduced a third way to manage buffers by allowing
components to only allocate buffers header during their initialization
and change their pBuffer pointer at runtime.
This new feature can save us a copy between GStreamer and OMX for each
input buffer.

This patch adds API to allocate and use such buffers.

https://bugzilla.gnome.org/show_bug.cgi?id=787093
2017-12-14 09:00:02 +00:00
Julien Isorce
6600a520a5 omxvideodec: consolidate the decision to try UseBuffer
The tee element can call gst_query_add_allocation_pool with pool as NULL.
Checking nth > 0 is not enough so we need to verify if there is a pool.

https://bugzilla.gnome.org/show_bug.cgi?id=730758
https://bugzilla.gnome.org/show_bug.cgi?id=784069
2017-12-13 12:14:48 +00:00
Julien Isorce
0d2d695eff gstomxvideodec: fix framerate overflow
Some live streams can set the framerate to 50000/1677 (=29.81).

GstVideoInfo.fps_n << 16 is wrong if the fps_n is 50000
(i.e. greater than 32767).

https://bugzilla.gnome.org/show_bug.cgi?id=759043
2017-12-13 10:12:05 +00:00
Guillaume Desmottes
56654cced5 omxvideodec: retrieve OMX_IndexParamVideoPortFormat before setting it
The usual pattern when setting OMX params is to first get the struct
param, override the values we want to set and then set the updated
param.
We were not doing this with OMX_IndexParamVideoPortFormat and so were
resetting some fields such as OMX_VIDEO_PARAM_PORTFORMATTYPE.xFramerate

https://bugzilla.gnome.org/show_bug.cgi?id=790979
2017-11-29 12:41:37 -05:00
Julien Isorce
6364463918 omxaacenc: also set 'profile' if mpegversion is 4
Like done by gst_codec_utils_aac_caps_set_level_and_profile
which is called by avenc_aac, ffaac and voaacenc.

https://bugzilla.gnome.org/show_bug.cgi?id=735208
2017-11-29 16:13:24 +00:00
Guillaume Desmottes
4211e4c29a omx: wait for flush complete and buffers being released when flushing
As stated in the existing comment, when flusing we should wait for OMX
to send the flush command complete event AND all ports being released.
We were stopping as soon as one of those condition was met.

Fix a race between FillThisBufferDone/EmptyBufferDone and the flush
EventCmdComplete messages. The OMX implementation is supposed to release
its buffers before posting the EventCmdComplete event but the ordering
isn't guaranteed as the FillThisBufferDone/EmptyBufferDone and
EventHandler callbacks can be called from different threads (cf 2.7
'Thread Safety' in the spec).

https://bugzilla.gnome.org/show_bug.cgi?id=789475
2017-11-28 13:35:07 -05:00
Guillaume Desmottes
4c0a8a6d90 gst_omx_port_set_flushing: simplify waiting loop
No semantic change so far, I just made the 'while' end condition easier
to understand as a first step before changing it.
- move error/time out checks inside the loop to make it clearer on what
we are actually waiting for.
- group port->buffers checks together with parenthesis as they are part
of the same conceptual check: waiting for all buffers to be released.

https://bugzilla.gnome.org/show_bug.cgi?id=789475
2017-11-28 13:35:07 -05:00
Guillaume Desmottes
a4c18ff243 zynqultrascaleplus: add support for extra AVC levels
The Zynqultrascaleplus has support for extra AVC levels not defined in
the OMX spec as a customer extension.

https://bugzilla.gnome.org/show_bug.cgi?id=790758
2017-11-23 11:39:33 -05:00
Guillaume Desmottes
6d44c0add5 omx: fix build on rpi
The nTimeStamp field is a OMX_TICKS struct on the rpi so use the proper
macro to set it.
Fix build on the build which has been broken by
b3173144b7

https://bugzilla.gnome.org/show_bug.cgi?id=789052
2017-10-19 16:23:01 +02:00
Guillaume Desmottes
62b96b6e14 omxvideodec: remove redundant debug message
We have already a debug message right after.

https://bugzilla.gnome.org/show_bug.cgi?id=789058
2017-10-19 16:20:10 +02:00
Guillaume Desmottes
2da61ef2bf omxh265: fix enum casting when using Allegro HEVC extensions
Allegro's HEVC implementation defines a superset of the profiles and
enums from the Android implementation.
Properly cast to fix -Wenum-conversion warnings from clang.

https://bugzilla.gnome.org/show_bug.cgi?id=789057
2017-10-19 16:19:16 +02:00
Guillaume Desmottes
cc0e3c8320 properly cast extension enums
OMX's allow 3rds party to define extensions using their own enums
(like OMX_VIDEO_CODINGEXTTYPE) and to be used as the general
ones (like OMX_VIDEO_CODINGTYPE).
Properly cast those to fix -Wenum-conversion warnings from some
compilers such as clang.

https://bugzilla.gnome.org/show_bug.cgi?id=789057
2017-10-19 16:19:16 +02:00
Guillaume Desmottes
b3173144b7 omx: also reset nTimeStamp when re-using buffers
Some OMX implementations may check if the timestamp of the output buffers
they receive is actually not set.

https://bugzilla.gnome.org/show_bug.cgi?id=788711
2017-10-11 13:51:55 -04:00
Guillaume Desmottes
514699594d omx: factor out gst_omx_buffer_reset()
https://bugzilla.gnome.org/show_bug.cgi?id=788711
2017-10-11 13:51:55 -04:00
Guillaume Desmottes
3a0e4bc35d add gst_omx_buffer_flags_to_string()
Make debug logs more readable so users don't have to manually figure out
the meaning of flags.

https://bugzilla.gnome.org/show_bug.cgi?id=788767
2017-10-11 13:49:30 -04:00
Ponnam Srinivas
8556596862 h263enc: fix caps leak in error code path
https://bugzilla.gnome.org/show_bug.cgi?id=788245
2017-09-28 13:24:07 +03:00
Ponnam Srinivas
1f36637434 omxh264enc: fix caps leak
https://bugzilla.gnome.org/show_bug.cgi?id=787711
2017-09-22 11:53:01 +03:00
Ponnam Srinivas
27cd61a21a omxh265enc: fix caps leak
https://bugzilla.gnome.org/show_bug.cgi?id=787714
2017-09-21 20:48:50 -04:00
Guillaume Desmottes
1bc6976b7b omxvideodec: log info about frame before releasing it
gst_video_decoder_release_frame() takes ownership of the frame and will
destroy it. So we should no longer use it after calling it.

https://bugzilla.gnome.org/show_bug.cgi?id=787628
2017-09-14 17:39:06 +01:00
Guillaume Desmottes
5751001ddc omxh265dec: add H265 decoder
Add HEVC decoder for the zynqultrascaleplus platform.

I used the H264 decoder code as a template.

https://bugzilla.gnome.org/show_bug.cgi?id=785434
2017-09-06 14:50:18 -04:00
Guillaume Desmottes
cf9f1903db omxh265enc: add H265 encoder
The OMX spec doesn't support HEVC but the OMX stack of the
zynqultrascaleplus adds it as a custom extension.
It uses the same API as the one of Android's OMX stack.

I used the H264 encoder code as a template.

https://bugzilla.gnome.org/show_bug.cgi?id=785434
2017-09-06 14:50:18 -04:00
Guillaume Desmottes
79df3b0e8e omxvideoenc: use caps from query in propose_allocation
Prevent crash by not deferencing a NULL pointer if self->input_state
isn't defined when propose_allocation() is called.

https://bugzilla.gnome.org/show_bug.cgi?id=786442
2017-09-06 09:15:31 -04:00
Guillaume Desmottes
318ef5b357 omx: display states as string if changing failed
Improve the error message by displaying the states in their string
representation rather than their numerical value.

https://bugzilla.gnome.org/show_bug.cgi?id=787235
2017-09-04 08:46:07 +01:00
Julien Isorce
054d5da46e omxvideodec: EGLDisplay is not indirectly defined by gstgl headers inclusion
Just use gpointer as done in GstGL to not include
EGL/egl.h just for EGLDisplay.

https://bugzilla.gnome.org/show_bug.cgi?id=784779
2017-08-22 10:29:05 +01:00
Julien Isorce
bb3778a1ac omxvideodec: remove wrong SettingsChanged ack
Partially revert 1b7d0b8:
  omxvideodec: handle IL 1.2 behavior for OMX_SetParameter

It turned out it was a problem in the decoder which was
not updating some local variables upon SetParameter.

https://bugzilla.gnome.org/show_bug.cgi?id=783976
2017-08-17 14:08:24 +01:00
Julien Isorce
eec88a9651 omx{audio,video}{dec,enc}: sequentially disable ports because buffers are not shared
For the history, the parallel disable port has been introduced by:
"00be69f omxvideodec: Disable output port when setting a new format"
and then replicated to videoenc, audiodec and audioenc.

This is only required to do 'parallel' if buffers are shared between ports.
But for decoders and encoders the input and output buffer are of different
nature by definition (bitstream vs images). So they cannot be shared.

Also starting from IL 1.2.0 it is written in the spec that the parallel
disable is not allowed and will return an error. Except when buffers are
shared.

Again here we know in advance that they are not shared so let's always
do a sequential disable.

Tested on Desktop, rpi and zynqultrascaleplus.

https://bugzilla.gnome.org/show_bug.cgi?id=786348
2017-08-17 14:01:27 +01:00
Guillaume Desmottes
6d2082003a omxvideodec: use the decoder API to set latency
https://bugzilla.gnome.org/show_bug.cgi?id=785125
2017-08-10 13:58:31 -04:00
Guillaume Desmottes
f2b0b66e39 omxvideoenc/dec: declare latency on zynqultrascaleplus
The OMX specification doesn't provide any API to expose the latency
introduced by encoders and decoders. We implemented this as a custom
extension as declaring the latency is needed for live pipelines like
video conferencing.

https://bugzilla.gnome.org/show_bug.cgi?id=785125
2017-08-09 17:01:15 -04:00
Guillaume Desmottes
0ee8213ab2 omxvideoenc: adjust stride and slice height from input
Use the stride and slice height information from the first buffer meta
data to adjust the settings of the input port.
This will ensure that the OMX input buffers match the GStreamer ones
and so will save us from having to copy line-by-line each one.

This is also the first step to allow the OMX encoder to receive dmabuf.

Tested on rpi and zynqultrascaleplus.

https://bugzilla.gnome.org/show_bug.cgi?id=785967
2017-08-08 10:40:47 -04:00
Guillaume Desmottes
07e80c9425 omxvideoenc: delay buffer configuration until component is enabled
No significant change for now. Just delay the input port configuration
of the buffer size related fields (stride, slice height, buffer size)
until the component is activated.
This will allow us to use the actual stride/height of the first input
and so avoid the buffer copying code path in most cases.

Tested on rpi and zynqultrascaleplus.

https://bugzilla.gnome.org/show_bug.cgi?id=785967
2017-08-08 10:40:47 -04:00
Guillaume Desmottes
0603e442ce omxvideodec/enc: delay allocation after the allocation query
Allocating OMX components buffers in set_format() is too early.
Doing it when receiving the first buffers will allow the element to use
the information from the allocation query and/or the first incoming
buffer to pick to best allocation mode.

Tested on raspberry pi with dynamic resolution changes on decoder and
encoder input.

https://bugzilla.gnome.org/show_bug.cgi?id=785967
2017-08-08 10:40:47 -04:00
Guillaume Desmottes
90da3c665e omxvideoenc: start src thread in handle_frame()
Makes the code simpler as we no longer need to restart the thread in
gst_omx_video_enc_flush() and It's more symetric which the omxvideodec
implementation.

I'm also going to move the enabling of the OMX component in
handle_frame() and the src pad thread needs to be started after it.

https://bugzilla.gnome.org/show_bug.cgi?id=785967
2017-08-08 10:40:47 -04:00
Guillaume Desmottes
7916bd53ab omxvideodec: earlier return if downstream_flow_ret is not OK
There is no point to (re)start the src thread if, for example, we are
flushing.

https://bugzilla.gnome.org/show_bug.cgi?id=785967
2017-08-08 10:40:47 -04:00
Guillaume Desmottes
352184dd09 omxvideodec: factor out enable and disable code
No semantic change, just factor out the code enabling and disabling the
component to their own functions.

Makes the code easier to read as the set_format() method was already
pretty big. Will also allow us to easily change the enabling logic.

https://bugzilla.gnome.org/show_bug.cgi?id=785967
2017-08-08 10:40:47 -04:00
Guillaume Desmottes
973c6e109a omxvideoenc: factor out enable and disable code
No semantic change, just factor out the code enabling and disabling the
component to their own functions.

Makes the code easier to read as the set_format() method was already
pretty big. Will also allow us to easily change the enabling logic.

https://bugzilla.gnome.org/show_bug.cgi?id=785967
2017-08-08 10:40:47 -04:00
Guillaume Desmottes
7a0949b750 omx: also reset nFilledLen before calling OMX_FillThisBuffer()
The spec states that the buffer passed to OMX_FillThisBuffer() needs to be
empty. Some implementation may check it actually is by checking its
nFilledLen field, so best to reset it as well.

https://bugzilla.gnome.org/show_bug.cgi?id=785623
2017-08-01 11:26:11 -04:00
Julien Isorce
86d9a2c81c omxvideodec: make generic the OMX_UseEGLImage code path
Will be easier to maintain and to make enhancements.

Tested with Tizonia on Desktop.
Also tested with Bellagio to make sure it does not crash when
calling OMX_UseEGLImage and indeed it returns NotImplemented.
Then gst-omx fallback to OMX_UseBuffer if it can and so on.

Also tested on rpi to make sure there is no regression.

https://bugzilla.gnome.org/show_bug.cgi?id=784365
2017-08-01 10:18:21 +01:00
Nicolas Dufresne
64f7f781d8 omxvideodec: Fix segment seek
On segment seek, unlike EOS, we drain, but we cannot expect a flush
later to reset the decoder state. As a side effect, the decoder would
remain in EOS state and ignore any new incoming buffers.

To fix this, we call _flush() inside the _drain() function, and
_finish() becomes what _drain() was before. This way, for _finish() (the
    eos case) we only drain, for _drain() triggered by segment seek or new
caps, we also reset the decoder state so it's ready to accept buffers.

https://bugzilla.gnome.org/show_bug.cgi?id=785237
2017-07-21 11:59:50 -04:00
Guillaume Desmottes
3f5d98a360 omxvideoenc: ensure enough buffers are allocated in the pool
Handle allocation query and ensure enough buffers are allocated in
the negotiated pool. This help preventing buffer starvation in the pipeline.

https://bugzilla.gnome.org/show_bug.cgi?id=785122
2017-07-19 09:37:37 -04:00
Guillaume Desmottes
136714c6ed omxvideodec: add dmabuf support for output
The zynqultrascaleplus OMX implementation has a custom extension
allowing decoders to output dmabuf and so avoid buffers copy between OMX
and GStreamer.

Make use of this extension when built on the zynqultrascaleplus. The
buffer pool code should be re-usable for other platforms as well.

https://bugzilla.gnome.org/show_bug.cgi?id=784847
2017-07-18 23:46:46 +01:00