On Tue, Apr 10, 2012 at 5:21 AM, Daniel Mack zonque@gmail.com wrote:
Hi Grant et al,
Again sorry for catching up so late on this.
On 16.01.2012 21:26, Grant Diffey wrote:
So I've had a chance to do some testing. I can reliably generate then following trace with daniels patch on 3.2-rt10
[ 188.561901] ------------[ cut here ]------------ [ 188.561908] WARNING: at fs/proc/generic.c:586
proc_register+0x181/0x1c7()
[ 188.561909] Hardware name: HP Compaq 6910p [ 188.561911] proc_dir_entry 'card1/pcm0p' already registered
There might have been a race condition in the shutdown path which I fixed.
Could you repeat the tests you did some months ago and report whether you still see this issue? I would also appreciate test input from other people. Any kind of USB device is good to track these issues.
Regarding the difference in reported channel numbers, I can't really point my finger to any specific line. It really puzzles me, as my patch set leaves the detection of interfaces and the registration of pcm streams untouched. It would help to know whether there is any difference in the output of "aplay -l" with and without my patches applied.
To test the new set of patches, please start off with a 3.4.0 vanilla kernel and then either
git pull git://github.com/zonque/linux.git usb-audio
or manually apply the combined patch from https://gist.github.com/1168715
I hope to *finally* get this topic finished in this round, but I need the some helpers for testing :)
Daniel
So this seems to be good it solves the clicking problem for me generating sine wave's
I need to do some further testing. which is unlikely to happen this morning.
Grant.