- added address field to Source struct definition to contain the destination address. If this field is left empty the default multicast address is used.
- if a unicast address is provided when creating a new Source the daemon looks in the ARP cache to retrieve the corresponding MAC address.
If this is not found the daemon tries to connect to echo TCP service of the destination host and looks into the ARP cache up to 3 times and then returns and error.
- added RTP address field in the Source creating form of the WebUI. If this field is left empty the default multicast address is used instead.
- updated documentation and regression test suite
- set current daemon version to bondagit-1.0
- added HTTP API to return current daemon version and updated documentation
- added diplay of current daemon version in WebUI Config page
This repo is a fork of the original one with the bondagit patches applied to the aes67-daemon branch
- removed ravenna-alsa-lkm driver patches from the repo
Removed driver copy_user(), copy_kernel() and fill_silence() ops.
Removed user space buffer transfer handling mode from playback and capture copy_internal() functions.
The problem arises from pre-buffering performed after ALSA prepare and before ALSA start trigger-
The patch enables the early startup of the audio playback interrupt and disables the cleanup (mute) of the playback buffer.
This issue can be reproduced using the test suite developed to investigate #17 and by removing the (-M) option from aplay in run_test.sh script.
To run the test suite:
./run_test.sh sample_format sample_rate channels duration
Where:
sample_format can be one of S16_LE, S24_3LE, S32_LE
sample_rate can be one of 44100, 48000, 96000
channels can be one of 1, 2, 4
duration is in the range 1 to 10 minutes
The test suite creates a raw file with the specified parameters, runs a loopback test where the file gets played and recorded using the loopback network interface and checks that the recorded file, after the initial silence, contains the expected samples sequence.
This test was developed to further investigate the issue #17.
* daemon:
- do not assume filesystem hierarchy in CMakeLists.txt
- do not replace default CXX compilation flags to avoid breaking
builds with external toolchain (e.g. yocto)
- allow for tests to be enabled/disabled
- allow for cpp-httplib dir to be provided by hand; otherwise search
for it in system locations
- allow for the location of the ravenna alsa lkm to be provided by
hand, otherwise look for it in system locations
Issue is in the mr_alsa_audio_pcm_playback_copy_internal() function of audio_driver.c that transfers the PCM samples from the ALSA buffer to the Ravenna buffer using the get_user() function.
In case of memory access mode the ALSA buffer can be directly accessed by the kernel and the samples must be copied.
mr_alsa_audio_pcm_playback_copy_internal() has been modified to handle the transfer of the ALSA buffer from user space (used with read/write access mode) and from kernel space (used by memory access mode).
The issue is common to all the platform but impacting ARM only.
Additional changes:
- changed type for dma_playback_offset and dma_capture_offset variables to uint32_t instead of atomic_t since the pointer callback mr_alsa_audio_pcm_pointer() is now atomic
- added variables to the mr_alsa_audio_chip scructure to decrease the number of calls to snd_pcm_lib_xxx() functions. These get assigned in the prepare callback mr_alsa_audio_pcm_prepare()
- changed mr_alsa_audio_get_playback_buffer_size_in_frames() to return configured size of Ravenna playback buffer (same as mr_alsa_audio_get_capture_buffer_size_in_frames())
- some rework of the existing patches
This fixes issue #20
Issue is in the mr_alsa_audio_pcm_playback_copy_internal() function of audio_driver.c that transfers the PCM samples from the ALSA buffer to the Ravenna buffer using the get_user() function.
In case of memory access mode the ALSA buffer can be directly accessed by the kernel and the samples must be copied.
mr_alsa_audio_pcm_playback_copy_internal() has been modified can to handle the transfer of the ALSA buffer from user space (used with read/write access mode) and from kernel space (used by memory access mode).
The issue is common to all the platform but impacting ARM only.
Additional changes:
- changed type for dma_playback_offset and dma_capture_offset variables to uint32_t instead of atomic_t since the pointer callback mr_alsa_audio_pcm_pointer() is now atomic
- added variables to the mr_alsa_audio_chip scructure to decrease the number of calls to snd_pcm_lib_xxx() functions. These get assigned in the prepare callback mr_alsa_audio_pcm_prepare()
- changed mr_alsa_audio_get_playback_buffer_size_in_frames() to return configured size of Ravenna playback buffer (same as mr_alsa_audio_get_capture_buffer_size_in_frames())
- some rework of the existing patches
This fixes issue #20wq
