Re: [LAU] hdsp not happy with recent kernels - Linux-audio-user

26 Sep 2012

...
  On Mon, 2012-09-24 at 16:19 -0400, Peter Lutek wrote:
> hi again, folks.... i've got some possibly interesting new results:
> 
> i've been compiling some kernels myself, to see if i can glean
> anything useful. first approach is to use the config from the 3.1
> kernel which works well with my hdsp multiface as a starting point.
> then, doing "make oldconfig" and accepting all the defaults for
> options that need answers, my self-rolled 3.0 *and* 3.2 kernels both
> fail with the hdsp, while appearing to run well otherwise.
> unfortunately, i don't see a 3.1 tarball at kernel.org to try.
> 
> anyways, since i know i've had good success with 3.0 kernels in the
> past, perhaps this is simply a config issue? or perhaps there are
> things being back-ported into the older stable kernels that are
> causing the problem.
> 
> i'm gonna see if i can make any sense out of the config options for
> which "make oldconfig" requires answers, but i'm getting out of my
> depth here. if anyone has any helpful thoughts, or might be willing to
> look at .config diffs and make some suggestions, i'd be most grateful!
> 
> cheers!
> .pltk. 
so, i've now done some comparisons of .config files, with help from kccmp
(http://stoopidsimple.com/kccmp)! here's what i've got:

** NOT in 3.1(good) but in BOTH 3.0(bad) and 3.2(bad)
   EXTRA_FIRMWARE=""
   SND_CS46XX_NEW_DSP=y
   USB_ENESTORAGE=n
   WLAGS49_H2=n
   WLAGS49_H25=n
** IN 3.1(good) but NOT in EITHER 3.0(bad) or 3.2(bad)
   ACERHK=m
   AUFS_BDEV_LOOP=y
   AUFS_BRANCH_MAX_1023=n
   AUFS_BRANCH_MAX_127=y
   AUFS_BRANCH_MAX_32767=n
   AUFS_BRANCH_MAX_511=n
   AUFS_BR_FUSE=n
   AUFS_BR_HFSPLUS=n
   AUFS_BR_RAMFS=n
   AUFS_BR_DEBUG=n
   AUFS_EXPORT=n
   AUFS_FS=m
   AUFS_HNOTIFY=n
   AUFS_PROC_MAP=n
   AUFS_RDU=n
   AUFS_SBILIST=y
   AUFS_SP_IATTR=y
   LLIST=y

there are, of course, other differences between the three configs, but these are the ones
which are consistent across "good" or "bad" kernels in the three
i've compiled. now, i really don't know if this sort of analysis has any bearing
on the problem... to any of you who might know more about these things than i do: does
this look useful at all? and do any of these configs trigger any alarm bells for anybody?

btw, there aren't any config settings which are PRESENT but SET DIFFERENTLY between
"good" and "bad" kernels.

thanks again.... cheers!
.pltk.