[alsa-devel] 2.6.30: xruns caused by "ALSA: pcm - Safer boundary checks"

[Takashi Iwai]

At Sat, 23 May 2009 17:34:54 -0400,
Robert Krakora wrote:
> 
> On Sat, May 23, 2009 at 2:22 PM, Takashi Iwai <tiwai at suse.de> wrote:
> > At Sat, 23 May 2009 14:31:57 +0200 (CEST),
> > Jaroslav Kysela wrote:
> >>
> >> On Fri, 22 May 2009, Takashi Iwai wrote:
> >>
> >> > At Fri, 22 May 2009 16:27:09 -0400,
> >> > Chuck Ebbert wrote:
> >> >>
> >> >> Bug report: https://bugzilla.redhat.com/show_bug.cgi?id=498858
> >> >>
> >> >> To reproduce, pause a video in mplayer and then hit play:
> >> >>
> >> >> ALSA sound/core/pcm_lib.c:263: hda_codec: hw_ptr skipping! [Q] (pos=11263,
> >> >> delta=17400, period=1024, jdelta=21/362/0)
> >> >
> >> > Does the patch fix the problem?
> >>
> >> This patch does not look as a proper fix.
> >
> > Hm, at which point?
> > It just skips the first unreliable jiffies check.
> > At least, it's interesting whether it works around the problem.
> >
> >
> > thanks,
> >
> > Takashi
> > _______________________________________________
> > Alsa-devel mailing list
> > Alsa-devel at alsa-project.org
> > http://mailman.alsa-project.org/mailman/listinfo/alsa-devel
> >
> >
> 
> Takashi:
> 
> I have submitted
> https://bugtrack.alsa-project.org/alsa-bug/view.php?id=4523 in regards
> to this problem.  I can recreated the xrun messages 100% of the time
> and the routine in question never recovers.  With a workaround
> suggested by Jaroslav, the xruns go away, but the calling process
> ramps up to 100% CPU utilization and unless you are on a dual or quad
> core machine there is quite a lockup.  I revert to the 1.0.19 pcmlib.c
> in the 1.0.20 driver and I see neither problem.  Thanks for all of
> your help Jaroslav.  If there is any way that I can help debug this
> problem, please let me know.

To be sure, did you try the latest alsa-driver snapshot (not 1.0.20)?
  ftp://ftp.kernel.org/pub/linux/kernel/people/tiwai/snapshot/alsa-driver-snapshot.tar.gz
There are already many fixes since 1.0.20, so...

Your problem seems not quite as same as the problem stated in this
thread.  But, anyway testing my patch is helpful.


thanks,

Takashi