KSE system scope vs non system scope threads

Christopher Sedore cmsedore at maxwell.syr.edu
Fri Dec 12 15:52:14 PST 2003



On Mon, 8 Dec 2003, Daniel Eischen wrote:

> On Sun, 7 Dec 2003, Christopher Sedore wrote:
> >
> > On Sun, 30 Nov 2003, Daniel Eischen wrote:
> >
> > > On Sun, 30 Nov 2003, Christopher M. Sedore wrote:
> > > >
> > [...]
> >
> > v1.18 of the thr_spinlock.c seemed to help.  I'm going to try to do more
> > testing this week.  As I noted in my other message, 5.2Beta won't run this
> > code on SMP machines (all my code is userland code, too), so it took a
> > while to actually get set up to try it out.  Now I need to set up kernel
> > debugging to see what's going on.
>
> Are you using gethostby* or anything else in libc that isn't
> thread-safe without using mutexes to protect them?

I had a couple stray uses of strtok left over from an earlier non-threaded
version--I cleaned them up, but that didn't help.  Whatever it was, it was
fixed in between 5.2Beta and 5.2RC1.  I'm not having any performance
problems with system scope threads at all, and SMP works.  Overall
performance is dramatically better than it had been previously.  I'll try
process-scope kse threads again soon.  I'm just happy I don't need to move
to a different platform.

Now, I just have to track down why I'm getting EAGAIN from sendfile()
when I'm not using non-blocking sockets...

-Chris



More information about the freebsd-threads mailing list