lkml.org 
[lkml]   [2002]   [Jan]   [22]   [last100]   RSS Feed
Views: [wrap][no wrap]   [headers]  [forward] 
 
Messages in this thread
/
Date
From
SubjectRe: [2.4.17/18pre] VM and swap - it's really unusable
Mark Hahn schrieb:
>
> > > > To me the benefit is clear enough: ASAP scheduling of IO threads, a
> > > > simple heuristic that improves both throughput and latency.
> > >
> > > I think of "benefit", perhaps naiively, in terms of something that can
> > > be measured or demonstrated rather than just announced.
> >
> > But you see why asap scheduling improves latency/throughput *in theory*,
> > don't you?
>
> NO, IT DOES NOT. why can't you preempt-ophiles get that through your heads?
>
> eager scheduling is NOT optimal in general.
>
> for instance, suppose my disk can only read a sector at a time.
> scheduling my sequentially-reading process to wake eagerly
> is most definitly PESSIMAL. laziness is a cardinal virtue!
> this doesn't preclude heuristics to sometimes short-cut the laziness.
>

Do you think there are no other benefits besides the scheduling latency in
a realtime system?

In a realtime system you want your event handling code (outside of the
interrupt handler [on Linux: bottom halves/tasklets/sorftirq?) get running
on the CPU as fast as possible. Therefore a realtime kernel is often fully
preemptible (well, there are always critical sections that has to disable
interrupts).

So the time between the interrupt handler wanting to schedule a specific
task/thread and the next scheduling decision is crucial, right?

I have no hard numbers, but I can imagine that this can also lead to
better IO (in terms of latency AND IO throughput but with the cost of
cpu cycles [user space CPU throughput]).

I don't know the Linux kernel good enough right now, but if you shorten
the scheduling latency: that could be a win for faster IO. But there's always
a tradeoff: if you spent too much time in scheduling decisions/preparations
the overhead eats the lower latency (especially if your mutexes have to deal
with priority inversion, giving a lock holder at least the same priority as
the lock contender for the period it holds the lock).
-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at http://vger.kernel.org/majordomo-info.html
Please read the FAQ at http://www.tux.org/lkml/

\
 
 \ /
  Last update: 2005-03-22 13:15    [W:0.202 / U:0.720 seconds]
©2003-2020 Jasper Spaans|hosted at Digital Ocean and TransIP|Read the blog|Advertise on this site