a/v freeze while flushing audio chain of pipeline in running state (clock and renderer sync related)

[Arun Raghavan]

On Thu, 1 Jun 2017, at 10:40 PM, Nitin Mahajan wrote:
> Please note that flush stop is passed with False so that the clock timing
> is not reset purposely as thereafter buffer with valid timestamp of
> different language is pushed. What could trigger queue thread to block /
> wait inside video renderer basesink from audio sink pulse audio. The test
> is running on x86 with gstreamer 1.4.5. Suggestion/pointer will be very
> helpful as there is not much clue from traces to establish design
> relationship between audio and video sinks and blocking wait.

It might be a good idea to look at the basesink debug output -- what is
the timestamp that you're pushing out, and what is the timestamp against
which that the sink is waiting?

-- Arun