AM62A7: Wave5 Encoder not closing properly in SDK 11.01

Sathiya Narayanan

Part Number: AM62A7

Hi,

I am using an AM62A7 with TI Processor SDK 11.01 and have encountered an issue with the Wave5 hardware encoder for both H.264 and H.265 encoding.

Issue

Whenever v4l2h264enc or v4l2h265enc is used, the GStreamer EOS event does not propagate beyond the hardware encoder.

As a result, when stopping a recording pipeline, qtmux / mp4mux does not receive EOS and cannot finalize the MP4 file. The resulting MP4 file is therefore unplayable.

H.265 reproduction

The following pipeline reproduces the issue:

GST_DEBUG=3 gst-launch-1.0 -e \
videotestsrc ! \
video/x-raw,width=1920,height=1080,format=NV12,framerate=30/1 ! \
v4l2h265enc ! \
h265parse ! \
qtmux ! \
filesink location=test-hw.mp4

The pipeline starts and encodes normally.When I press Ctrl+C, I get:

Handling interrupt.
Interrupt: Stopping pipeline ...
EOS on shutdown enabled -- Forcing EOS on the pipeline
Waiting for EOS...

The pipeline remains indefinitely at:

Waiting for EOS...

A second Ctrl+C is required to terminate gst-launch-1.0.

The generated MP4 is not playable because the EOS/finalization sequence does not complete.

H.265 test without qtmux

I also removed qtmux completely to determine whether the problem was related to MP4 muxing. The same behavior occurs.
Therefore, the issue occurs even without qtmux. This suggests that the EOS/drain operation is getting stuck at or below v4l2h265enc.

H.264

I observe the same behavior with the H.264 hardware encoder.

1 day ago

0 Suren Porwar 1 day ago

TI__Mastermind 36000 points

Hi Sathiya,

How is the behavior when you don't specify the -e option?

Also instead of specifying the -e option in the pipeline, providing num-buffers in videotestsrc is recommended. This will ensure a clean shutdown by sending an EOS when num-buffers is reached.

Example pipeline:

gst-launch-1.0 videotestsrc num-buffers=90 ! \

video/x-raw,width=1920,height=1080,framerate=30/1,format=NV12 ! \

v4l2h264enc ! \

h264parse ! \

qtmux ! \

filesink location=test.mp4

Hope this helps

Best Regards

Suren

0 Sathiya Narayanan 16 hours ago in reply to Suren Porwar

Intellectual 840 points

Hi Suren,

Without -e option, the stream is ending but the mp4 file is not playable.

Here is the pipeline:

root@am62axx-evm:~# GST_DEBUG=3 gst-launch-1.0 videotestsrc ! video/x-raw,width=1920,height=1080,format=NV12,framerate=30/1 ! v4l2h264enc ! h264parse ! qtmux ! filesink location=test-qtmux-h264.mp4
Setting pipeline to PAUSED ...
0:00:00.110647495 754 0x2347d90 WARN v4l2 gstv4l2object.c:4867:gst_v4l2_object_probe_caps:<v4l2h264enc0:src> Failed to probe pixel aspect ratio with VIDIOC_CROPCAP: Invalid argument
0:00:00.112102525 754 0xffff94000b70 WARN aggregator gstaggregator.c:2159:gst_aggregator_query_latency_unlocked:<qtmux0> Latency query failed
Pipeline is PREROLLING ...
0:00:00.113236800 754 0xffff94000d00 FIXME default gstutils.c:4036:gst_pad_create_stream_id_internal:<videotestsrc0:src> Creating random stream-id, consider implementing a deterministic way of creating a stream-id
0:00:00.116853610 754 0xffff94000d00 WARN v4l2 gstv4l2object.c:3408:gst_v4l2_object_reset_compose_region:<v4l2h264enc0:src> Failed to get default compose rectangle with VIDIOC_G_SELECTION: Invalid argument
Redistribute latency...
0:00:00.117669570 754 0xffff94000d00 WARN v4l2 gstv4l2object.c:3408:gst_v4l2_object_reset_compose_region:<v4l2h264enc0:sink> Failed to get default compose rectangle with VIDIOC_G_SELECTION: Invalid argument
0:00:00.152777365 754 0xffff94000d00 WARN v4l2bufferpool gstv4l2bufferpool.c:850:gst_v4l2_buffer_pool_start:<v4l2h264enc0:pool0:src> Uncertain or not enough buffers, enabling copy threshold
0:00:00.174852670 754 0xffff94000e90 WARN h264parse gsth264parse.c:2219:gst_h264_parse_update_src_caps:<h264parse0> VUI framerate 30.0 exceeds allowed maximum 0.2
Redistribute latency...
0:00:00.176165090 754 0xffff94000b70 FIXME basesink gstbasesink.c:3399:gst_base_sink_default_event:<filesink0> stream-start event without group-id. Consider implementing group-id handling in the upstream elements
Pipeline is PREROLLED ...
Setting pipeline to PLAYING ...
0:00:00.177271125 754 0xffff94000b70 FIXME aggregator gstaggregator.c:1448:gst_aggrRedistribute latency...
egator_aggregate_func:<qtmux0> Subclass should call gst_aggregator_selected_samples() from its aggregate implementation.
New clock: GstSystemClock
^Chandling interrupt.
Interrupt: Stopping pipeline ...
Execution ended after 0:00:20.717739645
Setting pipeline to NULL ...
0:00:20.894882970 754 0xffff94000b70 WARN qtmux gstqtmux.c:1921:gst_qt_mux_send_buffer:<qtmux0> Failed to send buffer (0xffff880136e0) size 68554
0:00:20.897362905 754 0x2347d90 WARN bufferpool gstbufferpool.c:1423:gst_buffer_pool_set_flushing:<v4l2h264enc0:pool0:sink> can't change flushing state of inactive pool
0:00:20.907078610 754 0x2347d90 WARN bufferpool gstbufferpool.c:1423:gst_buffer_pool_set_flushing:<v4l2h264enc0:pool0:sink> can't change flushing state of inactive pool
Freeing pipeline ...
root@am62axx-evm:~#

I tried playing the video in ubuntu. Here is the ffplay log.

snarayanan@SLS:~$ ffplay test-qtmux-h264.mp4 
ffplay version 4.4.2-0ubuntu0.22.04.1 Copyright (c) 2003-2021 the FFmpeg developers
  built with gcc 11 (Ubuntu 11.2.0-19ubuntu1)
  configuration: --prefix=/usr --extra-version=0ubuntu0.22.04.1 --toolchain=hardened --libdir=/usr/lib/x86_64-linux-gnu --incdir=/usr/include/x86_64-linux-gnu --arch=amd64 --enable-gpl --disable-stripping --enable-gnutls --enable-ladspa --enable-libaom --enable-libass --enable-libbluray --enable-libbs2b --enable-libcaca --enable-libcdio --enable-libcodec2 --enable-libdav1d --enable-libflite --enable-libfontconfig --enable-libfreetype --enable-libfribidi --enable-libgme --enable-libgsm --enable-libjack --enable-libmp3lame --enable-libmysofa --enable-libopenjpeg --enable-libopenmpt --enable-libopus --enable-libpulse --enable-librabbitmq --enable-librubberband --enable-libshine --enable-libsnappy --enable-libsoxr --enable-libspeex --enable-libsrt --enable-libssh --enable-libtheora --enable-libtwolame --enable-libvidstab --enable-libvorbis --enable-libvpx --enable-libwebp --enable-libx265 --enable-libxml2 --enable-libxvid --enable-libzimg --enable-libzmq --enable-libzvbi --enable-lv2 --enable-omx --enable-openal --enable-opencl --enable-opengl --enable-sdl2 --enable-pocketsphinx --enable-librsvg --enable-libmfx --enable-libdc1394 --enable-libdrm --enable-libiec61883 --enable-chromaprint --enable-frei0r --enable-libx264 --enable-shared
  libavutil      56. 70.100 / 56. 70.100
  libavcodec     58.134.100 / 58.134.100
  libavformat    58. 76.100 / 58. 76.100
  libavdevice    58. 13.100 / 58. 13.100
  libavfilter     7.110.100 /  7.110.100
  libswscale      5.  9.100 /  5.  9.100
  libswresample   3.  9.100 /  3.  9.100
  libpostproc    55.  9.100 / 55.  9.100
[mov,mp4,m4a,3gp,3g2,mj2 @ 0x7a8654000cc0] moov atom not found
test-qtmux-h264.mp4: Invalid data found when processing input
    nan    :  0.000 fd=   0 aq=    0KB vq=    0KB sq=    0B f=0/0   
snarayanan@SLS:~$

MP4 Muxers need EOS signal to finalize the file. Otherwise, stream will not be playable.

I also wanted to add that I getting this error / warning in dmesg:

[   79.411396] vdec 30210000.video-codec: error -ENXIO: IRQ index 0 not found
[   79.411422] vdec 30210000.video-codec: failed to get irq resource, falling back to polling
[   79.411585] vdec 30210000.video-codec: OPP table not found in device tree
[   79.414228] vdec 30210000.video-codec: Added wave5 driver with caps: 'ENCODE' 'DECODE'
[   79.414247] vdec 30210000.video-codec: Product Code:      0x521c
[   79.414254] vdec 30210000.video-codec: Firmware Revision: 291549

Processors

Processors forum

AM62A7: Wave5 Encoder not closing properly in SDK 11.01

Issue

H.265 reproduction

H.265 test without qtmux

H.264