Re: a joint letter on low latency and Linux

On Fri, Jun 30, 2000 at 04:17:11PM +0200, Roger Larsson wrote:
> > I want to try this trick on TCP/IP sometime. 
> 
> Sorry, this will not help...
> 
> It is in fact very close to what we are doing with low latency tests.
> The test thread runs AT HIGHEST SCHED_FIFO (there is no higher prio...)
> 
> The driver sets need_resched of currently running process and moves
> the SCHED_FIFO process to the run queue.
> 
> The interrupt ends and the current process should be rescheduled - but it
> can't since it is running in the kernel and may remain there for over 100ms.

The trick with the RTL module changes the scheduling parameters
bypasses some problems and permits a type of fine grained scheduling
that would be tough to do directly in Linux .
> 
> Even RTLinux needs a low latency kernel to be able to USE standard
> processes.

Absolutely correct. In fact, improvements in Linux latency are critical
to RTLinux over the long haul. If we had a lower latency kernel, the
pressure to do some sort of user mode RTL would decrease and I really
don't want to have that around as it leads, quite rapidly, to
making RTLinux/Linux start looking like Chorus on top of Linux -- a
sickening prospect.

> 
> /RogerL

-- 
---------------------------------------------------------
Victor Yodaiken 
FSMLabs:  www.fsmlabs.com  www.rtlinux.com
FSMLabs is a servicemark and a service of 
VJY Associates L.L.C, New Mexico.