Wow, my system got wrecked (exaggeration) during this latest stretch... Pulseaudio was stretched to the limit and beyond and was forced to restart. Anything that was producing audio had to be restarted to get it back. This time was much like the first time and went from timestamp 573100.060927 (line 1) to 572506.604155 (line 11069), where 100% (literally) of it was that event 37 in the journal, no other kernel log entries except for the systemd-hostnamed audit before it all went down. And as usual, it was my USB TV tuner (tvheadend really) giving the Poll Timeout log entries. Those same uploaded trace files will be updated with the latest bugout. On Wed, Jan 2, 2019 at 5:32 AM Mathias Nyman <mathias.nyman@xxxxxxxxxxxxxxx> wrote: > > The event type 37 is a host controller event, most likely a event ring full error. > > So there are probably so many events that we fill the event ring before we can handle them. > > Could you take traces of this? > Note that the trace file will be huge. > > mount -t debugfs none /sys/kernel/debug > echo 81920 > /sys/kernel/debug/tracing/buffer_size_kb > echo 1 > /sys/kernel/debug/tracing/events/xhci-hcd/enable > > copy the traces somewhere safe once the error is triggered: > cp /sys/kernel/debug/tracing/trace /<somewhere> > > -Mathias
