[ecasound] Re: [linux-audio-dev] huge latency peaks with 2.4.10-smp kernel

New Message Reply About this list Date view Thread view Subject view Author view Other groups

Subject: [ecasound] Re: [linux-audio-dev] huge latency peaks with 2.4.10-smp kernel
From: Paul Davis (pbd_AT_op.net)
Date: Thu Oct 18 2001 - 21:21:37 EEST

>This is something I've been complaining about for a long time and since
>given up because kernel developers seem to be only interested about

kernel developers like ingo molnar? like andrew morton? like richard
love? yes, the majority are focused on throughput, but you know what -
that will actually help us too when it comes to streaming huge amounts
of media data around. and besides, why shouldn't they be - they are
working on their own problems and domains.

in the meantime, a small but focused group has provided ways to avoid
the behaviour that Kai reported. linus has indicated that he's happy
with the goal of low latency, but he doesn't want "hacks" to fix
it. despite the fact that it was "hacks" that created it (code that
steals the CPU for too long), one can understand his point. there's no
point fixing a mistake with another mistake.

the preemptive kernel patch seems to have generated a great deal of
interest in this area, and mr. love has been a very effective
spokesperson for the work. its hard to tell right now which is better,
andrew's lowish patches or the PE patch.

also, AFAIK, the BSD family generally has much worse average latency
characteristics than Linux does, even though its worst case scenarios
are not quite as bad as the regular linux kernel.

>No one has been willing to add device specific bandwidth limits /
>requirements... I hate when some non-critical device causes dropouts.

any device driver will always be able to create dropouts if its badly
written. all it has to is to block interrupts for too long, and boom!
there is absolutely *nothing* the kernel can do about it. this is true
on every operating system, including RTLinux, QNX and other similar
systems. you have to be able to trust device drivers, and in general,
you can, but alas, there are some that are not worthy of it.


To unsubscribe send message 'unsubscribe' in the body of the
message to <ecasound-list-request_AT_wakkanet.fi>.

New Message Reply About this list Date view Thread view Subject view Author view Other groups

This archive was generated by hypermail 2b28 : Thu Oct 18 2001 - 21:19:21 EEST