Re: kern/43542: hdaudio at 32000Hz

["Jonathan A. Kollasch" <jakllsch%kollasch.net@localhost>]

The following reply was made to PR kern/43542; it has been noted by GNATS.

From: "Jonathan A. Kollasch" <jakllsch%kollasch.net@localhost>
To: gnats-bugs%NetBSD.org@localhost
Cc: 
Subject: Re: kern/43542: hdaudio at 32000Hz
Date: Sun, 4 Jul 2010 03:43:07 +0000

 from my dmesg:
 hdafg0: 2ch/2ch 44100Hz-96000Hz 16/16 20/32 24/32
 
 So, 32kHz isn't a rate supported directly by the codec.
 
 Further debugging shows that hdaudio_stream_param() is being fed a rate
 of 64kHz when it fails.  64kHz isn't handled by the sample_rate switch
 statement.  It appears that I can use the 48kHz base with a multiplier
 of 4 and divisor of 3 to obtain 64kHz.  However, it doesn't appear that
 64kHz is a rate in the supported rate flag bits, so this option may be
 incorrect.
 
 Anyway, shouldn't hdaudio_stream_param() fail more gracefully?  I'd
 suggest selecting the format first, and defaulting to 48kHz when an
 unsupported rate is attempted.  At least that wouldn't sound as
 horrific.  Or is there a way to EINVAL these requests at the audio(4)
 layer?