No black video sent out if track is muted before MediaPipeline is activated
Categories
(Core :: WebRTC: Audio/Video, defect, P2)
Tracking
()
People
(Reporter: pehrsons, Assigned: pehrsons)
References
(Regression)
Details
(Keywords: regression)
Attachments
(4 files)
(deleted),
text/x-phabricator-request
|
RyanVM
:
approval-mozilla-beta+
RyanVM
:
approval-mozilla-esr68+
|
Details |
(deleted),
text/x-phabricator-request
|
RyanVM
:
approval-mozilla-beta+
RyanVM
:
approval-mozilla-esr68+
|
Details |
(deleted),
text/x-phabricator-request
|
RyanVM
:
approval-mozilla-beta+
RyanVM
:
approval-mozilla-esr68+
|
Details |
(deleted),
patch
|
pehrsons
:
review+
|
Details | Diff | Splinter Review |
Discovered from https://groups.google.com/forum/?utm_medium=email&utm_source=footer#!msg/discuss-webrtc/Gqo2MfnQWkw/sOZlb2EqDgAJ
There is a pernosco recording at https://pernos.co/debug/dxLVskE-28Jb11gvrw4iSA/index.html
The black frame reaches MediaPipeline but is dropped at https://searchfox.org/mozilla-central/rev/b38e3beb658b80e1ed03e0fdf64d225bd4a40327/media/webrtc/signaling/src/mediapipeline/MediaPipeline.cpp#1128.
We might want to propagate the active state into the VideoFrameConverter.
Updated•5 years ago
|
Assignee | ||
Comment 1•5 years ago
|
||
There's a second bug here where frames from SameFrameTimer don't get sent due to bug 1522238. Frames broadcast by the timer have the same timestamp as the last frame and get dropped upstream.
This didn't seem to impact my manual test of the original STR, but it impacts the mochitest I wrote.
Assignee | ||
Comment 2•5 years ago
|
||
Assignee | ||
Comment 3•5 years ago
|
||
Depends on D40597
Assignee | ||
Comment 4•5 years ago
|
||
Depends on D40598
Assignee | ||
Updated•5 years ago
|
Updated•5 years ago
|
Comment 6•5 years ago
|
||
bugherder |
https://hg.mozilla.org/mozilla-central/rev/d2a030d452cb
https://hg.mozilla.org/mozilla-central/rev/91d822ac2647
https://hg.mozilla.org/mozilla-central/rev/acd8b80087e0
Assignee | ||
Updated•5 years ago
|
Assignee | ||
Updated•5 years ago
|
Assignee | ||
Comment 7•5 years ago
|
||
Comment on attachment 9082952 [details]
Bug 1570673 - Add an active state to VideoFrameConverter and propagate it from MediaPipeline. r?bwc
Beta/Release Uplift Approval Request
- User impact if declined: Chrome and safari (possibly also Edge) will not be able to receive audio from Firefox in video calls in some cases
- Is this code covered by automated tests?: Yes
- Has the fix been verified in Nightly?: No
- Needs manual test from QE?: Yes
- If yes, steps to reproduce: Join a room on https://appr.tc with Firefox
Mute the camera in Firefox
Join the same room (same URL) with Chrome or Safari
Mute the microphone in Chrome or Safari
Expected: you can hear yourself
With bug: you can only hear yourself after unmuting the camera in Firefox - List of other uplifts needed: None
- Risk to taking this patch: Low
- Why is the change risky/not risky? (and alternatives if risky): Well covered with unittests and mochitests
- String changes made/needed:
ESR Uplift Approval Request
- If this is not a sec:{high,crit} bug, please state case for ESR consideration: 68 regression that got public attention: https://groups.google.com/d/msg/discuss-webrtc/Gqo2MfnQWkw/XZYGhjHvDQAJ
- User impact if declined: Chrome and safari (possibly also Edge) will not be able to receive audio from Firefox in video calls in some cases
- Fix Landed on Version: 70
- Risk to taking this patch: Low
- Why is the change risky/not risky? (and alternatives if risky): Well covered with unittests and mochitests
- String or UUID changes made by this patch:
Assignee | ||
Updated•5 years ago
|
Assignee | ||
Updated•5 years ago
|
Assignee | ||
Updated•5 years ago
|
Updated•5 years ago
|
Comment 8•5 years ago
|
||
Verified fixed on the latest Firefox Nightly (20190808214929) on Windows 10 and Mac OS 10.14.6
Updated•5 years ago
|
Updated•5 years ago
|
Comment 9•5 years ago
|
||
Comment on attachment 9082951 [details]
Bug 1570673 - Add mochitest. r?bwc
Fixes a WebRTC interop bug which can cause users of other browsers to not receive audio from Firefox users. Approved for 69.0b13 and 68.1esr.
Updated•5 years ago
|
Updated•5 years ago
|
Comment 10•5 years ago
|
||
bugherder uplift |
Comment 11•5 years ago
|
||
At least the second patch needs rebasing for ESR68.
Comment 12•5 years ago
|
||
Verified fixed on Windows 10, Mac OS 10.14.6 and Ubuntu 18.04 using Firefox Beta 69.0b13 (20190812173625)
Updated•5 years ago
|
Assignee | ||
Comment 13•5 years ago
|
||
Slight collision with bug 1556696. I'll have the 2nd patch up rebased for esr68 in a bit.
Assignee | ||
Comment 14•5 years ago
|
||
Rebased attachment 9082952 [details] on top of esr68. Carries forward the r+.
Comment 15•5 years ago
|
||
bugherder uplift |
Comment 16•5 years ago
|
||
Verified fixed on Windows 10, Mac OS 10.14.6 and Ubuntu 18.04 using Firefox 68.1.0 ESR (20190814031227) downloaded from https://treeherder.mozilla.org/#/jobs?repo=mozilla-esr68
Updated•3 years ago
|
Description
•