diff options
author | xians@chromium.org <xians@chromium.org@0039d316-1c4b-4281-b951-d872f2087c98> | 2013-11-22 14:12:41 +0000 |
---|---|---|
committer | xians@chromium.org <xians@chromium.org@0039d316-1c4b-4281-b951-d872f2087c98> | 2013-11-22 14:12:41 +0000 |
commit | f269dc3c19468e46e36de8a7fd91ecd7f6de1f9d (patch) | |
tree | 39f86eb5575a2e5322b6b6d14b26fb0868bae040 /gin/test | |
parent | b2cad65c03de86e326cf0e62869a2bf41426347b (diff) | |
download | chromium_src-f269dc3c19468e46e36de8a7fd91ecd7f6de1f9d.zip chromium_src-f269dc3c19468e46e36de8a7fd91ecd7f6de1f9d.tar.gz chromium_src-f269dc3c19468e46e36de8a7fd91ecd7f6de1f9d.tar.bz2 |
Move the SetCaptureFormat to the capture audio thread.
Before this CL, the thread model of our capture class is that they will be updated with the format in the main render thread by SetCaptureFormat() and getting the audio callback in the capture audio thread.
And this complicates our code since it is required that the audio format in the data callback in the capture audio thread has to be the same as the cached format we get from the main render thread.
This CL simplies the thread model by moving the SetCaptureFormat() to the capture audio thread, so the format change callback will be in the same thread as the data callback.
This CL fixed the problems of https://codereview.chromium.org/76293004/
(crash when the webrtcLocalSourceProvider has not been getting the format callback before the data callback)
https://codereview.chromium.org/59273005/
(WebRtcLocalAudioRenderer might get a SetCaptureFormat() after Start(), in such case we should start the sink_ in WebRtcLocalAudioRenderer to adapt to the new format)
TBR=joi@chromium.org
BUG=320808
TEST=content_unittests
Review URL: https://codereview.chromium.org/83133002
git-svn-id: svn://svn.chromium.org/chrome/trunk/src@236765 0039d316-1c4b-4281-b951-d872f2087c98
Diffstat (limited to 'gin/test')
0 files changed, 0 insertions, 0 deletions