mirror of
https://github.com/mollyim/webrtc.git
synced 2025-05-13 05:40:42 +01:00
![]() This should mostly be a noop, but in a follow up cl we will insert all packets after splitting, which will allow for adapting the delay to FEC (both RED and codec inband) that is useful for decoding (i.e. not already covered by primary packets). A slight behavior change is that reordered packets are no longer included in max delay calculation. Implementation details: - A map ordered by RTP timestamp is used to store the arrivals. - When inserting new packets, we check if the timestamp is too old, already exists or if the packet is fully covered by another packet (based on timestamp and packet duration). - Separate deques are used to keep track of "min" and "max" arrivals (as defined by ordering operators). The queues maintain a strictly increasing/decreasing order so that min/max is always at begin(). Bug: webrtc:13322 Change-Id: I8b6cf5afff77b4adc3c29745b95627e955715b5a Reviewed-on: https://webrtc-review.googlesource.com/c/src/+/337184 Reviewed-by: Henrik Lundin <henrik.lundin@webrtc.org> Commit-Queue: Jakob Ivarsson <jakobi@webrtc.org> Cr-Commit-Position: refs/heads/main@{#41656} |
||
---|---|---|
.. | ||
async_audio_processing | ||
audio_coding | ||
audio_device | ||
audio_mixer | ||
audio_processing | ||
congestion_controller | ||
desktop_capture | ||
include | ||
pacing | ||
portal | ||
remote_bitrate_estimator | ||
rtp_rtcp | ||
third_party | ||
utility | ||
video_capture | ||
video_coding | ||
BUILD.gn | ||
module_common_types_unittest.cc |