At Fri, 22 Sep 2006 16:09:23 +0100, Liam Girdwood wrote: > > On Fri, 2006-09-22 at 12:44 +0200, Takashi Iwai wrote: > > > > the patch below is an experimental patch to achieve more aggressive > > power-saving on HD-audio. With the patch, the device is disabled > > after a certain timeout when no apps are using the PCM and no loopback > > channels are being used. (According to Pavel Machek, sound chips > > consume significant powers.) > > > > For some reasons, the patch is a bit intrusive and touches many > > places. But the change is pretty straightforward: add power-up/down > > calls in appropriate place (e.g. mixer changes, PCM open/close). > > > > For the power-up/down control, a damn simple way is chosen in this > > version. The driver does basically a same job like suspend/resume. > > We can tune up this later if the perfomance matters. > > > > The default timeout value is 30 seconds. This is adjustable via > > power_save module option of snd-hda-intel driver. You can change this > > value dynamically via writing to > > /sys/modules/snd_hda_intel/parameters/power_save. The value changed > > isn't propagated immediately, and you have to do something (like > > change a mixer value or open/close PCM) to activate the new value, > > though. > > I'd be interested to know if there are any pops or clicks associated > with the HD-audio power up/down events (I don't have an HD PC or laptop > to test). Pops and clicks are an unwanted side effect of audio power > management atm and occur when the audio subsystems analog circuits are > power switched. Yeah it's possible although a laptop I've tested doesn't cause it. We'd need mute/unmute appropriately before power-switching. But the muting didn't help much on my intel8x0 laptop, though. > If the pops are bad then it may be more desirable to disable the digital > components in the audio subsystem (e.g. PLL's, DAI, digital filters, > core) and selectively (per machine/codec) disable certain analog > components. Basically, the power-down shouldn't occur so frequently unless you set the timeout value too short. So, I don't think it's a critical problem if the power-saving is enough. (Of course, we should fix the thing that is fixable :) Takashi ------------------------------------------------------------------------- Take Surveys. Earn Cash. Influence the Future of IT Join SourceForge.net's Techsay panel and you'll get the chance to share your opinions on IT & business topics through brief surveys -- and earn cash http://www.techsay.com/default.php?page=join.php&p=sourceforge&CID=DEVDEV _______________________________________________ Alsa-devel mailing list Alsa-devel@xxxxxxxxxxxxxxxxxxxxx https://lists.sourceforge.net/lists/listinfo/alsa-devel
