It seems to me that this might be the cause of these
sporadic latency peaks which are exactly the time it takes to play
an audio fragment (time between 2 IRQs issued by the soundcard.
(see the diagrams on my page)
I noticed this on the RTC tests too:
at RTC_HZ=1024 the sporadic peaks are at about 1ms ( 1/1024)
at RTC_HZ=512 these peaks go up at about 2ms (1/512)
and I think every IRQ driven app will show up the same behaviour.
for example this was the cause why on audio apps I used 3 audio fragments
instead of 2 because one (even if it is very sporadic) lost bh might
cause audio skips.
With 3 fragments it is extremely unlikely that this peak does occur
for two consecutive IRQs therefor I'd assume that it is safe to use 3.
Let me know when there will be a fix available , so I can publish
new results.
Benno.
-- Benno Senoner E-Mail: sbenno@gardena.net Linux low-latency audio / scheduling latency benchmarks http://www.gardena.net/benno/linux/audio
- To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to majordomo@vger.rutgers.edu Please read the FAQ at http://www.tux.org/lkml/