[pulseaudio-discuss] [PATCH 00/12] A new ringbuffer based protocol for PulseAudio

[David Henningsson]

On 2014-06-17 16:10, Peter Meerwald wrote:
> Hello David,
>
>>>> Third iteration.
>>>
>>> paplay --latency-msec XXX audio_s16_2ch_44k1.wav
>>> on beagleboard-xm @ 1 GHz, kernel 3.15
>
> I want to re-benchmark with SHM off, at first sight it appears that load
> is LOWER (for low-latency)?!

Iochannel + no SHM could very well be faster than iochannel + SHM, 
because SHM means some extra code.

I would be surprised if iochannel + no SHM was significantly faster than 
srbchannel + SHM, because eventfds are supposed to be faster than pipes [1].

-- 
David Henningsson, Canonical Ltd.
https://launchpad.net/~diwic

[1] http://man7.org/linux/man-pages/man2/eventfd.2.html - "The kernel 
overhead of an eventfd file descriptor is much lower than that of a pipe".