On Tue, 14 Jan 2014 15:46:43 +0100, Daniel Mack wrote: > On 01/14/2014 06:09 AM, Dennis New wrote: > > On Fri, 10 Jan 2014 10:16:07 -0500, Dennis New wrote: > >> On Sun, 05 Jan 2014 18:08:40 +0000, Daniel Mack wrote: > >>> Dennis New <dennisn@xxxxxxxxxxxxxxxxxx> wrote: > >>>> For the past many linux kernels, > >>> > >>> Can you spot the last kernel version that worked for you? > >>> > >>>> my usb bluetooth headset randomly > >>>> crashes -- the audio abruptly stops working, I cannot kill -9 the > >>>> mplayer process that was using it, and the only way to get it > >>>> working again is to reboot :|. > >>>> > >>>> The syslog only says: > >>>> kernel: ALSA sound/usb/endpoint.c:501 timeout: still 3 active > >>>> urbs on > >>>> EP #3 > >>> > >>> That's odd, and frankly is more a sign of a USB host controller > >>> issue that one with snd-usb-audio. Can you provide a full dmesg > >>> output? Do you have the same issue with that headset on a > >>> different machine? > >>> > >>>> It is very difficult to reproduce -- sometimes it can go for days > >>>> just fine, then in the middle of a song abruptly stop. > >>> > >>> Please try harder :-) Without a proper test case, it seems > >>> impossible to verify a potential fix. Quickly starting and > >>> stopping the stream might be a good start. > >>> > >>> Also, as stated above, please do test on a different machine. > >>> Ruling out the host controller would be a crucial step. > >> > >> Here's another crash log. ohci_urb_dequeue seems to show up often > >> in the tracebacks of the hung processes. > >> > >> http://dennisn.dyndns.org/guest/pubstuff/debug-usbaudio/crash2.log > > > > Here's a slightly more verbose log, after tweaking a few more kernel > > debugging options, which shows khubd and lsusb being hung/blocked. > > > > http://dennisn.dyndns.org/guest/pubstuff/debug-usbaudio/crash3.log > > Ok, thanks for the logs. > > We need to rule out either the USB host controller or the device > itself. For this, please try to reproduce the issue with the same > headset on a different PC, and also try a different USB audio device > on the machine that shows the odd behaviour. I tend to believe that > it's the USB host controller side that causes your trouble, but we > need more evidence. > > Could you do that? I actually do have 2 identical such devices, and the same problem is occurring with the other device too. So it probably is something wrong with the "usb host controller". The problem occurs on any of my laptop's 3 usb ports. I don't think my girlfriend has experienced this problem on her laptop. And when the problem occurs, none of my other usb devices work (I don't think they even get power, let alone kernel acknowledgement.) But only this usb headset (so far) seems to trigger this problem, so I'm not (yet) convinced that it's simply my (aging) laptop's hardware. Isn't there a way to for the kernel to reset the usb stuff, so at least I don't have to reboot? I can't kill the tasks (they are in a D = uninterruptible sleep state), and I can't unload the modules. ------------------------------------------------------------------------------ CenturyLink Cloud: The Leader in Enterprise Cloud Services. Learn Why More Businesses Are Choosing CenturyLink Cloud For Critical Workloads, Development Environments & Everything In Between. Get a Quote or Start a Free Trial Today. http://pubads.g.doubleclick.net/gampad/clk?id=119420431&iu=/4140/ostg.clktrk _______________________________________________ Alsa-user mailing list Alsa-user@xxxxxxxxxxxxxxxxxxxxx https://lists.sourceforge.net/lists/listinfo/alsa-user
