Re: [alsa-devel] [PATCH 0/6] sound/pci/hda: CX20549 cleanup/fixes
At Fri, 06 Apr 2012 15:01:47 +0200, Michael Karcher wrote:
On Thu, 2012-04-05 at 16:28 +0200, Takashi Iwai wrote:
BTW, did you try model=auto option to enable the auto-parser? If it doesn't work well for you, let me know.
I tried the model=auto option, too. It works mostly, but the mixer node seems to not be utilized correctly/at all. The capture source list does not contain "mixer" anymore, and there are no playback controls for Internal Mic/Mic/Line.
Right, the analog loopback is dropped in the auto-parser for simplicity.
The Capture setup seems to be built correctly (except for the missing mixer). The auto setup adds two extra input devices for jack sense.
In anyway, give alsa-info.sh output (run with --no-upload option) for better checks.
alsa-info for the revised patch-set (mailed soon) with and without model=auto is attached. Information I consider irrelevant (non-sound modules and dmesg context not relating to sound) has been edited out.
Thanks.
BTW: I seem to have encountered an ALSA user-space bug: The codec provides a "master capture switch", i.e. the "mute" of the adc in amp. It also provides a "capture source selection enum", i.e. the "Connection Select" of the adc widget. The enum provides mutually exclusive capture sources that have no volume control, whilst the master capture switch is coupled with the capture volume control, and should not be related to the source enumeration. But if I play around with capture sources in alsamixer, the "Capture Switch" is turned off as soon as I switch the source from the first item in the Mux list to something else.
This sounds like a bug in alsa-lib mixer abstraction. There is a known shortcomings in the "capture source" enum, e.g. it can't take more than one ADCs.
This happens with both the patched hard-coded "BenQ" configuration and the auto configuration.
OK, I'll check it in the next week.
thanks,
Takashi
participants (1)
-
Takashi Iwai