Re: autobridge patch



On Mon, 16 Jan 2006 10:03:33 -0800
Brooks Davis <brooks@xxxxxxxxxxxxxxxxxx> wrote:

> On Mon, Jan 16, 2006 at 11:25:04AM +0000, Marcin Jessa wrote:
> > On Mon, 16 Jan 2006 22:16:05 +1300
> > Andrew Thompson <thompsa@xxxxxxxxxxx> wrote:
> >
> > > Hi,
> > >
> > >
> > > I have a patch here that adds the ability to automatically add an
> > > interface to a bridge when its attached. This is aimed towards
> > > apps like qemu or vmware that open a tap interface and need it
> > > bridged with the network adapter, the user can set up a glob for
> > > interfaces to be automatically added (eg tap*). It may also be
> > > useful for Xen dom0 support.
> > >
> > > This patch includes a big change to how interfaces are configured
> > > in userland. Before only physical Ethernet cards were handled by
> > > devd, now _ALL_ interfaces are (vlans, pflog, pfsync, tap, tun,
> > > etc..). This has the added bonus that the pseudo interfaces can
> > > be configured after boottime in rc.conf, ifconfig_xxx="".
> > >
> > > Please test this patch, even if you dont use a bridge. Im not in
> > > a hurry to commit it.
> >
> > I just rebuild and installed my world and kernel to test the
> > patches. Seems like things did not work as expected.
> > The name part of ifconfig_vlan0 left me with renamed but not
> > configured device. The IP etc of VLAN01 was not set at boot. Am I
> > doning something wrong?
> >
> > cloned_interfaces="bridge0 vlan0"
> > ifconfig_sk0="up"
> > ifconfig_vlan0="name VLAN01 inet 10.137.99.4 vlan 1 vlandev sk0"
>
> If you want to rename the vlan interface use:
>
> ifconfig_vlan0_name="name VLAN01"
> ifconfig_VLAN01="inet 10.137.99.4 vlan 1 vlandev sk0"
>
> not that this has anything to do with the Andrew's bridging script.

Would be great if Andrew also wrote an example of how the patchset
works. I read the patches now and it seems it works for bridge
interfaces only adding ability to automatically add new interfaces to
the bridge with the new "autobridge_*" options.
The original post confused me making me think the patches made it
possible to use a feature I was looking for and that I did know existed
- operate on pseudo interfaces the same way one does with the normal
ones.
Your last comment in the "Automatic VLANS" thread on the same
mailinglist made me assume I could not rename pseudo devices at boot
with the rc.conf script, which I can see now was fully possible.

Thanks,
Marcin.

_______________________________________________
freebsd-net@xxxxxxxxxxx mailing list
http://lists.freebsd.org/mailman/listinfo/freebsd-net
To unsubscribe, send any mail to "freebsd-net-unsubscribe@xxxxxxxxxxx"



Relevant Pages

  • Re: autobridge patch
    ... >> like qemu or vmware that open a tap interface and need it bridged ... the user can set up a glob for interfaces ... >> This patch includes a big change to how interfaces are configured in ... even if you dont use a bridge. ...
    (freebsd-net)
  • Re: distributed bridging [script]
    ... >> I'm trying to create a configuration that would bridge interfaces on ... Each host would have at least two interfaces, ... here's a completely untested script. ...
    (freebsd-net)
  • Re: autobridge patch
    ... > like qemu or vmware that open a tap interface and need it bridged ... the user can set up a glob for interfaces ... > This patch includes a big change to how interfaces are configured in ... even if you dont use a bridge. ...
    (freebsd-net)
  • Re: bridge with access on both interfaces
    ... >>'inside' and 'outside' IPs separately on each bridge interface, ... Some tcpdumps on both in and outside interfaces suggest an ARP ... > If you want to use IP while bridging, ... and net4801 is that the 3 ethernet controllers now share one interrupt. ...
    (freebsd-net)
  • Re: autobridge patch
    ... > I have a patch here that adds the ability to automatically add an interface to a ... > open a tap interface and need it bridged with the network adapter, ... > set up a glob for interfaces to be automatically added. ... even if you dont use a bridge. ...
    (freebsd-net)