2 Sep
2008
2 Sep
'08
11:36 a.m.
Tim Blechmann wrote:
i am lurking on the linux-rt-users list for a while and i haven't read
any references, why 2.6.26 (patched or unpatched) is not suitable for low-
latency audio ...
still 2.6.26-rt3 is unusable for me and, as paul mentioned, also for
others ... i'd be curious about the reasons, though ...
is not that .26 is that bad for audio, but *it is* seriously broken for
midi. afaict that relates to the -rt only. vanilla might be ok.
it has been suggested that the preemptive-BKL changes made after .26-rc2
are the suspect subject. good news are that .27 will have this fixed or
at least goes a few steps into an acceptable solution regarding -rt
byee
--
rncbc aka Rui Nuno Capela
rncbc(a)rncbc.org
while sharing this experience, i am curious, whether
there are any
insights, why this is the case?
did anyone come up with an analysis of the problem source? maybe
bisecting the kernel source?
I haven't been following the RT list that closely
over the last couple
of weeks, but I think you'll find that version is WAY ahead of the
stable RT patches.
I got my RT working about 4 weeks ago and was advised by the RT
developers to go for the latest stable - 2.6.24.7-rt17