AudioSampleBufferList::zeroABL() takes a byte-count, not a sample-count.
Created attachment 302232 [details] Patch
Comment on attachment 302232 [details] Patch View in context: https://bugs.webkit.org/attachment.cgi?id=302232&action=review > Source/WebCore/ChangeLog:10 > + (WebCore::AudioSampleDataSource::pullAvalaibleSamplesAsChunks): Weird that it didn't pick up pullSamples instead. And also, terrible spelling of Available.
Created attachment 302264 [details] Patch for landing
(In reply to comment #2) > Comment on attachment 302232 [details] > Patch > > View in context: > https://bugs.webkit.org/attachment.cgi?id=302232&action=review > > > Source/WebCore/ChangeLog:10 > > + (WebCore::AudioSampleDataSource::pullAvalaibleSamplesAsChunks): > > Weird that it didn't pick up pullSamples instead. And also, terrible > spelling of Available. Yeah, I think that when a function is too short, the function detector freaks out and picks the previous one.
Comment on attachment 302264 [details] Patch for landing Clearing flags on attachment: 302264 Committed r212714: <http://trac.webkit.org/changeset/212714>
Why is there no test?
Comment on attachment 302264 [details] Patch for landing View in context: https://bugs.webkit.org/attachment.cgi?id=302264&action=review > Source/WebCore/ChangeLog:10 > + (WebCore::AudioSampleDataSource::pullAvalaibleSamplesAsChunks): Speling.
(In reply to comment #6) > Why is there no test? We don't have any test infrastructure for inspecting rendered audio.
Doesn't web audio have something?
https://github.com/WebKit/WebKit/commit/661e7be6621f71f2f55632903c62ad69591dd11d Landed and didn't backed out. Marking this as "RESOLVED FIXED".
<rdar://problem/100997360>