[comp.realtime] Searching for determinism/predictability

koopman@a.gp.cs.cmu.edu (Philip Koopman) (03/02/90)

I am searching for references to previous work done on
determinism and predictability over small time intervals
(i.e. microseconds).  Apparently many of the features of
modern architectures such as scoreboards, caches, and
use of fancy DRAM access techniques can really cause some
problems.  But, I can't find very much on actual measurements
or models -- most articles just mention the problem without
any analysis or modeling.

I have found the following:
- a 1986 article by Razouk et al. on measuring '286 performance
- a 1987 Software Engineering Institute report on Ada performance
- several articles by coworkers at Harris
- several recent articles in places such as Computer Design that
    wave their hands at the problem, but do no real analysis

Any pointers to further high-quality articles are appreciated.

Thanks,
  Phil Koopman                koopman@greyhound.ece.cmu.edu   Arpanet
  2525A Wexford Run Rd.
  Wexford, PA  15090
Senior Scientist at Harris Semiconductor, adjunct professor at CMU.

jon@cs.washington.edu (Jon Jacky) (03/03/90)

> koopman@a.gp.cs.cmu.edu (Philip Koopman) asks...
>
> I am searching for references to previous work done on
> determinism and predictability over small time intervals
> (i.e. microseconds). ... I have found the following:
> 
> - a 1986 article by Razouk et al. on measuring '286 performance
> - a 1987 Software Engineering Institute report on Ada performance
> - several articles by coworkers at Harris
>
> Any pointers to further high-quality articles are appreciated.

Try:

@article{softeng:shaw89,
        author="Alan Shaw",
        title="Reasoning About Time in Higher-Level Language Software",
        journal="IEEE Transactions on Software Engineering",
        volume=15,
        number=7,
        month="July",
        year=1989,
        pages="875 -- 889"}
      
At a colloquium here at the University of Washington last November, Shaw
described an experimental tool his group had produced that 
predicts the running time of languages written in a C subset (the target was
a 68010).  That project is written up in a techical report,

@techreport{softeng:park89,
	author="Chang-Yun Park and Alan Shaw",
	title="A source-level tool for prediciting deterministic execution
		times of programs",
	institution="Department of Computer Science, University of 
		Washington",
	year=1989,
	month="Sept.",
	number="TR#89-09-12",
	address="Mail Stop FR-35, University of Washington, Seattle, WA,
			98195"}

- Jon Jacky, University of Washington, jon@gaffer.rad.washington.edu