woods@weird.com (Greg A. Woods) writes:
> True enough but on a well tuned multi-user machine I generally find
> everyone gets their fair share and if someone wants to burn his share on
> a bunch of simple cmps then so be it!  ;-)

In other words:

I'm running a compile in one window, and running a bunch of 'cmp's in
another.

they each get their fair share of time, great.  but they also finish
(potentially much) more slowly than they otherwise would.

it's great to trade off execution time to e.g. fairness by
implementing a time-sliced scheduler.  It's simply stupid to throw
cycles away just because you can.


> Yes, except that in this case the time spent in moving the bits about is
> accounted to the process in question and while it may not be nice to
> waste cycles on a multi-user system, there's not much you can really do
> to prevent people from doing even more stupid things!  ;-)

if you're going to talk about fairness issues -- and that's largely
irrelevant here -- in fact, no, it does hose down fairness, too, but
not fairness that's typically tracked by the OS very carefully.

process A kicks process B's lines out of the cache by a bunch of
unnecessary copying.  process B then has to spend a bunch of extra
time getting those lines back into cache, before it can execute them.
process B gets charged for that time.

With regard to cache usage, just about every OS i've seen is in fact
cooperative (because it'd be fairly hard to do otherwise).  But in
fact bad behaviour by one process _can_ negatively impact other
unrelated processes execution times.


> If I understand correctly [ ... ]

as i said before, _go read the literature_.



> No, what I mean is that you or I are not able to take it for free
> (i.e. under that license) and then charge a fee for it.  We can't make
> it un-free.  We must give it away for free if we make it available.

I didn't see anything that said that in the SFIO license.



cgd