So I'm having some weird dynamic between two PulseAudio servers (using module-tunnel-sink-new). The symptoms are that the downstream server stops playing audio, but other clients for the upstream server work fine. (upstream here being the server closest to the actual audio hardware) The bug is however very fickle and I only see it about once every two weeks. However right now I'm able to semi-reliably produce it when running Super Tux Kart inside ThinLinc, so I'm trying to use this opportunity to debug it. Help is appreciated. :) Let's call the upstream server A, and the downstream server B. What I've discovered so far: a) The problem sometimes resolves itself after a few minutes. Perhaps it always resolves itself but the time varies, I haven't checked. b) Server B stops receiving requests for more data and believes the buffers are full. c) Server B still does periodic latency checks, and this is how the latency information looks from its point of view: > Breakpoint 2, stream_get_timing_info_callback (pd=pd at entry=0x7f5ce40010a0, command=command at entry=2, tag=tag at entry=107, t=t at entry=0x7f5ce4010ba0, userdata=userdata at entry=0x7f5ce400e020) at pulse/stream.c:1952 > 1952 o->stream->auto_timing_update_requested = false; > (gdb) print *i > $5 = {timestamp = {tv_sec = 1463411510, tv_usec = 297935}, synchronized_clocks = 1, sink_usec = 3026, source_usec = 0, transport_usec = 271, playing = 0, write_index_corrupt = 0, write_index = 1782548, > read_index_corrupt = 0, read_index = 1782548, configured_sink_usec = 1451, configured_source_usec = 0, since_underrun = 12642188} Note that "playing" is false! d) I tried to gdb server A and see why it is stating that it is not playing. However gdb perturbed the system enough that the bug went away. :/ I've attached the debug output from server B around the time I started the game and everything was supposed to un-suspend. Ideas? Rgds -- Pierre Ossman Software Development Cendio AB https://cendio.com Teknikringen 8 https://twitter.com/ThinLinc 583 30 Linköping https://facebook.com/ThinLinc Phone: +46-13-214600 https://plus.google.com/+CendioThinLinc A: Because it messes up the order in which people normally read text. Q: Why is top-posting such a bad thing? -------------- next part -------------- A non-text attachment was scrubbed... Name: wedge.log Type: text/x-log Size: 6321 bytes Desc: not available URL: <https://lists.freedesktop.org/archives/pulseaudio-discuss/attachments/20160516/10eb2afd/attachment.bin>