Quoting Alexandre Ratchov (a...@caoua.org): > do you know what causes these interrupts? is this the uaudio > device? (ex try "systat -s1 vmstat")
It's never the uaudio device, in fact it's not even one of the devices listed in that view. I have uhci2 which hovers around 62. re0 hovers around 300 but briefly spikes during heavy network usage. ipi hovers around 4-600, then spikes to 15000. ipi seems to be an MP thing; it's not present in GENERIC. > > Is the stuttering likely related to my hardware, is it a known problem > > or is there something else I can try? > > - switch into using the GENERIC kernel and see if > stuttering is affected > > - use trivial tools to play audio (eg. aucat -i foo.wav) during > the tests. > > - does the "-mplay" option affect stuttering? if so the > cause may be the uaudio driver. Adding -mplay to sndiod_flags and restarting sndiod didn't help. Then I rebooted into GENERIC and there seemed to be longer between stuttering (suggesting it didn't get triggered quite as easily). I wasn't kidding when I said this is my desktop system but I don't run X because this chipset is shit. I am in console most of the time. I discovered *any* output to console, even if I haven't switched to that tty could cause a small glitch in the audio. It's sporadic but more output is definately worse. Scrolling quickly through man pages is annoying and running 'make clean' made the audio unbearable. Logging out of the console and ssh'ing from another system has made the stuttering much more infrequent (once every 20-30mins instead of every 2 minutes) and it doesn't seem related to scrolling in man nor to what make is up to. Does that make any sense at all? I'll look into playing wav files with aucat and let you know. -- I prefer the dark of the night, after midnight and before four-thirty, when it's more bare, more hollow. http://a.mongers.org
