[JNPR] Network stack as a module (RFC)

[Dheeraj Kandula]

Is there a link explaining about what changes have gone into this huge
Diff. That would help people like me to figure out what is being addressed
in the diff. I get the point that the whole network stack will be a
loadable module now. Isn't it?


On Tue, Nov 19, 2013 at 5:15 PM, Gleb Smirnoff <glebius at freebsd.org> wrote:

>   Steve,
>
> On Mon, Nov 18, 2013 at 02:35:08PM -0500, Steve Kiernan wrote:
> S> > Have you measured the performance impact on the loadable stack? I
> S> > suppose many functions, that are now called directly, with loadable
> S> > stack will be called via functon pointers.
> S>
> S> No, only a small number of functions were needed to be methods in a
> S> netstack kobj in the current implementation that we have. And some of
> S> those may need to change, as they revolve around NFS exports.
> S>
> S> Currently, we have four (4) methods that are needed:
> S> socreate
> S> vfs_export
> S> vfs_setpublicfs
> S> vfs_stdcheckexp
> S>
> S> When one is not dealing with NFS exports, only the socreate method is
> S> of interest. It is called just before mac_socket_create() would be
> S> called in socreate().
>
> Sounds okay. Thank you.
>
> S> See http://people.freebsd.org/~marcel/Juniper/netstack-v2.diff, which
> S> is the full patch set that had been previously sent out. Some things
> S> have changed a bit since then, but if you look for curnetstack usage
> S> in the patch, you should be able to see the places where netsack
> S> methods are called.
>
> The patch is quite huge to understand it as a whole. :)
>
> --
> Totus tuus, Glebius.
> _______________________________________________
> freebsd-net at freebsd.org mailing list
> http://lists.freebsd.org/mailman/listinfo/freebsd-net
> To unsubscribe, send any mail to "freebsd-net-unsubscribe at freebsd.org"
>