lkml.org 
[lkml]   [1999]   [Apr]   [12]   [last100]   RSS Feed
Views: [wrap][no wrap]   [headers]  [forward] 
 
Messages in this thread
/
Date
From
SubjectRe: CPU affinity
Andrea Arcangeli wrote:

> On Sat, 10 Apr 1999, Ben Bridgwater wrote:
>
> >I don't have an SMP box yet, but I'm working on a real-time signal
> >processing application that is eventually going to need the ability to
> >hard assign processes to specified CPUs. I see this an an SMP extension
>
> Why? What do you want to achieve? Performances? I think it make no sense
> as far as the IOAPIC will run as now (and the current setting of the
> IO-APIC is sure the best for general purposes). I think that to give it a
> _sense_ you should at least also drive all irqs of your machine to the
> other CPU (it's possible setting the IOAPIC in a different way, and in
> such different way the IOAPCI will also allow you to select different
> priority for irqs).

I need deterministic performance. I can control things on a single processor
by using SCHED_FIFO, but want to carry this level of control over to SMP.

I'm processing video (bttv - DMA) and 8KHz sound data, so interrupts arn't
really an issue. But I am doing *heavy* processing, and my application
requires to be as time sensitive as possible, so I really can't afford to have
the scheduler 2nd guessing me when I move to SMP.

Ben


-
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/

\
 
 \ /
  Last update: 2005-03-22 13:51    [W:0.058 / U:0.028 seconds]
©2003-2020 Jasper Spaans|hosted at Digital Ocean and TransIP|Read the blog|Advertise on this site