> > oh? B8 of the v1.4 MPS spec is the only thing I see adressing
> > this, and it says "Operating system writers should factor in processor
> > variations, such as processor type, family, model and features, to arrive at
> > a configuration that maximizes overall system performance. At a minimum, the
> > MO OS should remain operational and should support the common features of
> > unequal processors."
i don't see what all this argument is about -- the above may be a bit
vague, but it does only confirm what is obvious anyway -- if it's not
possible to provide a constant rate monotonic TSC then it should be
disabled. [the other solution (for the constant, but different on some
cpus case) is pset(), but even then, the default config should be tsc
disabled]
> you expect to have syncronized TSC's, if the clocks are coming from
> diffrent sources?
Actually, do your tsc really drift, or were you just speculating
earlier? I'd guess only the multiplier is different, but can't
verify this as i don't have such a setup...
> In any case, blindly assuming that TSC's are the same is as bad as
> assuming that TSC is present at all, and it's not any harder to check,
It is a reasonable assumption. Expecting all apps to check for same
tsc rate on all cpus isn't.
-
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 Mar 31 2000 - 21:00:24 EST