* David E. Cross <[EMAIL PROTECTED]> [000212 00:35] wrote:
> > Well, I'd first be very interested to know if anyone has even seen
> > this work. :)
> >
> > I have seen regrettably little feedback about it so far.
> >
> > - Jordan
> I can understand part of the reason for this... 4.0-RELEASE is right
> arround the corner, and people are focusing on delivering a stable product,
> not introducing alpha code into the system at the last second. the current
> rpc.lockd is a known value, placing a "maybe" version into the stream at
> this point would do no one any good.
Unfortunatly it doesn't even ack NLM4 locks.
> > > I'm suprised no one has gotten back to you on this, since it seems to
> > > need no kernel changes and the fact that the existing lockd just
> > > doesn't really do anything...
> > >
> > > But then again we are 'frozen'. (hello Jordan :) )
> > >
> > > I think that having "we now serve locks" in 4.0-release as a real
> > > step forward. I would like to see it go in before the release is
> > > cut.
> Again, I feel that this isn't woth it at this point... it should be something
> we strive for in 4.1-RELEASE.
>
> > >
> > > Can anyone comment?
> > >
> > > More importantly can any of the people that have been _begging_
> > > for this please take a moment and test or at least give David the
> > > courtisy of a responce?
> > >
> > > You guys do want this right?
> > >
> > > You guys don't want to have to run -current for another year just to
> > > have lockd right? (the reason I cc'd -stable)
> > >
> > > *hello? is this thing on?*
>
> Mmmm.... feedback would be very good. Also, what are the chanches of
> getting the 3 required functions MFC-ed? fh{open|stat|statfs} ?
Not very likely, they require a split of VFS functionality, that
although somewhat trivial, it would break third party filesystem
KLDs and the work would touch every single filesystem. You can
avoid it by hacking at the netexport stuff, but it gets kinda ugly
afaik.
It is a shame that no one who's begged for locks has come forward,
I have no good test env for stressing the code, and I'd be nice to
get some preliminary feedback on its stability.
--
-Alfred Perlstein - [[EMAIL PROTECTED]|[EMAIL PROTECTED]]
To Unsubscribe: send mail to [EMAIL PROTECTED]
with "unsubscribe freebsd-current" in the body of the message
