This seems simple enough, since kernel_thread() returns the pid.
But what happens if someone kills nfsiod? Is there some way of
finding that out? Hmmm... In playing around, killing the
nfsiod's DOES execute MOD_DEC_USE_COUNT. So you could keep track
of someone else killing nfsiod. (Want to be sure, of course,
that you don't accidently kill off someone else's userprocess
that may have the same pid as your original nfsiod).
> >From the ps point of view you're merely exchanging insmod for mount.
Couple of points here. First, I'm a bit confused. Am I *never*
supposed to see (nfsiod) in ps if I have nfs as a module? Or
only that there is some delay in seeing it? I've notice that I
see the insmod stuff upon boot, but sometime later I see
(nfsiod).
Second, what about process memory? Running all four tasks at
once, they all share the same vm. No big deal. But, if it's
done once for each call to mount, could you end up tying up the
memory more than once? The code is no big deal (or does it lock
the executable?) The data should be eventually be paged out.
Of course, once you get the 1 nfsiod is all that's need, it's
less of a problem.
mrc
--
Mike Castle .-=NEXUS=-. Life is like a clock: You can work constantly
mcastle@cs.umr.edu and be right all the time, or not work at all
mcastle@umr.edu and be right at least twice a day. -- mrc
We are all of us living in the shadow of Manhattan. -- Watchmen