diff options
author | Glenn Kasten <gkasten@google.com> | 2014-05-12 11:06:26 -0700 |
---|---|---|
committer | Glenn Kasten <gkasten@google.com> | 2014-05-22 16:19:05 -0700 |
commit | 29b703eec27b305e7b5b2343bf257643e38f6b68 (patch) | |
tree | 0612b8e650dac1f4f7c2e51afa187b4601f113ca /services/audioflinger | |
parent | 6181ffd90a436f333c43a7f812391eee2c35005a (diff) | |
download | frameworks_av-29b703eec27b305e7b5b2343bf257643e38f6b68.zip frameworks_av-29b703eec27b305e7b5b2343bf257643e38f6b68.tar.gz frameworks_av-29b703eec27b305e7b5b2343bf257643e38f6b68.tar.bz2 |
Move validation of frameCount from set to openRecord_l
This move is needed because frameCount is validated on server side for
fast tracks (as should be done for normal tracks too).
Change-Id: I6d99e80869fd90fab373cf60ef348c01f075fbca
Diffstat (limited to 'services/audioflinger')
-rw-r--r-- | services/audioflinger/Threads.cpp | 1 |
1 files changed, 1 insertions, 0 deletions
diff --git a/services/audioflinger/Threads.cpp b/services/audioflinger/Threads.cpp index cdc9a19..b6782a9 100644 --- a/services/audioflinger/Threads.cpp +++ b/services/audioflinger/Threads.cpp @@ -5128,6 +5128,7 @@ sp<AudioFlinger::RecordThread::RecordTrack> AudioFlinger::RecordThread::createRe // to be at least 2 x the record thread frame count and cover audio hardware latency. // This is probably too conservative, but legacy application code may depend on it. // If you change this calculation, also review the start threshold which is related. + // FIXME It's not clear how input latency actually matters. Perhaps this should be 0. uint32_t latencyMs = 50; // FIXME mInput->stream->get_latency(mInput->stream); size_t mNormalFrameCount = 2048; // FIXME uint32_t minBufCount = latencyMs / ((1000 * mNormalFrameCount) / mSampleRate); |