Richard B. Johnson <root@chaos.analogic.com> write:
[ ... early history of digital controllers elided ... ]
>This leads to the requirement of a "Real-time Operating System".
>Real-time does not mean "fast". Real-time means "fast enough", but
>"precise".
>
>Many systems can be controlled using a relatively slow sample and
>control rate. [ ..... ]
>The sampling of the speed at precise intervals, over some previous
>history, allows mathematics to predict the time-interval at which
>the speed will be within specification. [ .... ]
This is all perfectly fine, but is completely irrelevant for hardware
that generates interrupts rather than expecting to be polled.
For such hardware, "real time" does, in fact, mean "fast enough",
since what matters is how quickly after the interrupt is generated the
relevant action (if any) is taken. This is a function of several things:
* the delay in the CPU noticing the IRQ (h/w dependent, mostly)
* the delay between the CPU noticing the IRQ and the execution of
the IRQ handler.
* the delay between the execution of the IRQ handler
and the execution of the relevant code path.
* the number of CPU cycles needed to execute the relevant code path
directly.
* the number of CPU cycles "stolen" by other code while the relevant
code path is running (e.g. preemption driven by other IRQ's)
--p (whose "real time" control system for processing film canisters
sorts about 15% of the rolls of consumer 35mm film in the USA)
-
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 : Sun Jul 23 2000 - 21:00:14 EST