Comment 25 for bug 267913

I'd like to add some more information.

I sometimes get this message close to login on 8.10:

[ 1815.360302] canberra-gtk-pl[28014]: segfault at ae64790 ip b718884d sp b6725190 error 6 in libpulse.so.0.4.1[b7149000+4e000]
[ 1885.644970] hda-intel: IRQ timing workaround is activated for card #0. Suggest a bigger bdl_pos_adj.

This message does /not/ result in the impaired audio functionality I describe below. In other circumstances, the below occurs:

My hda-intel driver sometimes shares an interrupt with sata_nv, and when it does (after some time, maybe 4-5 hours of using the computer) I end up with one of my internal hard drives reporting many SATA error codes and bus read errors, and eventually the kernel unmounts the partition to save it from data loss. Any use of audio is severely impaired -- any app that plays audio hangs while looping the first bit of the audio. For example, `aplay /usr/share/sounds/question.wav' will result in aplay hanging, and the first half second of question.wav being played repeatedly.

Shutting down pulseaudio (pulseaudio --kill), and then removing/reloading the hda-intel driver (rmmod snd_hda_intel / modprobe snd_hda_intel) does not fix the problem.

Hope this helps -- feel free to ask for any more info.

Thanks,
RJ Ryan