Jaroslav Kysela wrote:
>> Which gives threshold = 6552 from buffer=6553, period=3276...
>> exactly the same value as the code I was already using gives!  What
>> else am I missing from the garish red and yellow blocks of WinDiff?
> Then it really appears like a bug. Could you add snd_pcm_dump() call
> just before the block occurs?

Sure... but it doesn't seem to show any anomalies... I've also included the
matching status report from /proc...


state: PREPARED
trigger_time: 1081246741.324684000
tstamp      : 1081246792.298694000
delay       : 0
avail       : 8
avail_max   : 7229
-----
hw_ptr      : 0
appl_ptr    : 6545

--------------

Hardware PCM card 0 'Hoontech SoundTrack Audio DSP24' device 0 subdevice 0

Its setup is:
stream       : PLAYBACK
access       : RW_INTERLEAVED
format       : S32_LE
subformat    : STD
channels     : 10
rate         : 88200
exact rate   : 88200 (88200/1)
msbits       : 24
buffer_size  : 6553
period_size  : 3276
period_time  : 37142
tick_time    : 1000
tstamp_mode  : NONE
period_step  : 1
sleep_min    : 0
avail_min    : 3276
xfer_align   : 1
start_threshold  : 6552
stop_threshold   : 6553
silence_threshold: 0
silence_size : 0
boundary     : 1717829632


>> Why does changing to mmap writes make the blocking problem go away?
>> And as I said, the problem only has been made to occur at 88.2 and
>> 96K sampling rates...
> mmap does never block

That'd explain why it doesn't halt like snd_pcm_writei, then... :)

... but if it was having to throw away data, surely wouldn't it drop out
when I swap in mmap?  Because it doesn't... snd_pcm_mmap_writei *seems* to
work perfectly...

I'm suspecting an "off by one" calculation error somewhere.

- Tom.



-------------------------------------------------------
This SF.Net email is sponsored by: IBM Linux Tutorials
Free Linux tutorial presented by Daniel Robbins, President and CEO of
GenToo technologies. Learn everything from fundamentals to system
administration.http://ads.osdn.com/?ad_id=1470&alloc_id=3638&op=click
_______________________________________________
Alsa-devel mailing list
[EMAIL PROTECTED]
https://lists.sourceforge.net/lists/listinfo/alsa-devel

Reply via email to