Re: More PCI races...

From: George Anzinger (george@pioneer.net)
Date: Fri Jun 16 2000 - 15:57:40 EST


Well, maybe some do, but the scheduler checks for bottom half stuff each
time thru, thus there at least is the expectation that some of them run
outside of interrupt context. They do, however, always run before any
other task that the scheduler might want to switch to.

In the HPRT system (a clone of lynx OS), I set up a timer thread that
priority tracked the highest priority task that needed timer completion
work. The time tick that caused a timer to "pop" just put the structure
on the threads queue of things to complete, made sure it had the correct
priority and put it in the schedule list if needed. The idea was that
the timer thread could run at a lower priority than other more important
tasks, thus keep alive timers for lan traffic, for example, did not
impact high priority real time tasks. This was and still is the way
HPRT handles ALL timer events, except the profile timer(s).

George

David Hinds wrote:
>
> On Fri, Jun 16, 2000 at 03:50:13PM -0400, Olivier Galibert wrote:
> >
> > Maybe I'm dense, but isn't that exactly what bottom halves are for?
>
> My understanding is that bottom halves, softirqs, and tasklets all run
> in interrupt context.
>
> -- Dave
>
> -
> 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/

-
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/



This archive was generated by hypermail 2b29 : Fri Jun 23 2000 - 21:00:13 EST