Inner workings of turnstiles and sleepqueues

[John Baldwin]

On Friday 19 October 2007 01:08:27 pm Attilio Rao wrote:
> 2007/10/19, John Baldwin <jhb at freebsd.org>:
> > On Friday 19 October 2007 12:56:54 am Ed Schouten wrote:
> > > * John Baldwin <jhb at freebsd.org> wrote:
> > > > The best option right now is to read the code.  There are some 
comments in
> > > > both the headers and implementation.
> > >
> > > Would it be useful to write manpages for these interfaces, or do we
> > > assume that only godlike people can use them anyway? I am willing to
> > > write manpages for them.
> >
> > They already exist, but they really are only used to implement 
higher-level
> > primitives like locks and condition variables.  The rest of the kernel 
should
> > use the higher-level primitives anyway.
> 
> Well, really turnstiles don't have manpages, but this is still OK as
> they are only used in mutex while the "real" sleeping primitive should
> be identified by sleepqueues.

Ah, there is a sleepqueue(9), but not a turnstile(9).

-- 
John Baldwin