Re: CVS commit: src/sys/dev/hdaudio

[Robert Elz <kre%munnari.OZ.AU@localhost>]

    Date:        Fri, 04 Aug 2017 17:14:28 +1000
    From:        matthew green <mrg%eterna.com.au@localhost>
    Message-ID:  <24600.1501830868%splode.eterna.com.au@localhost>

  | this code is not functional and only complains when it fails
  | to parse stuff, that probably is quite valid.

That is most likely true.

  | all it does is whine, not actually do take any useful action.

as is that.

  | ie, it's not useful and actively spams dmesg.

but that is not.   That is, it happened to me, and that "whine" saved me
a whole heap of time attempting to debug why the hdmi audio was not
attaching, or why I was failing to config or use it properly.   The message
was easy to grep for in the sources, and it became immediately obvious that
the audio device was not going to be attached.

Now it is very likely incomplete/buggy code that causes this, but whenever
code thinks that something is broken, and so gives up on what it is doing,
it us useful to know why.   Regardless of whether it is the code that is
broken, or the (in this case) data it is attempting to decode (and it
is much more likely the code here) doesn't really matter - it fails, and
it is useful to know why.

We are talking about just 1 message at boot, compared with the rest of
what is there, that is hardly "spam" - once one of the errors occurs, the
device fails to attach, and that's it (I don't think, or at least I am not
aware of, any way to re-attempt the attach without a reboot, not that it
would do any good.)

kre