1.\" 2.\" $FreeBSD$ 3.\" 4.Dd September 28, 1998 5.Dt BRIDGE 4 6.Os 7.Sh NAME 8.Nm bridge 9.Nd bridging support 10.Sh SYNOPSIS 11.Cd "options BRIDGE" 12.Sh DESCRIPTION 13.Fx 14supports bridging on Ethernet-type interfaces. 15.Pp 16Runtime operation of the 17.Nm 18is controlled by several 19.Xr sysctl 8 20variables. 21The 22.Va net.link.ether.bridge 23variable can be set to 24.Li 1 25to enable bridging, or set to 26.Li 0 27to disable it. 28.Pp 29The 30.Va net.link.ether.bridge_ipfw 31variable can be set to 32.Li 1 33to enable 34.Xr ipfw 8 35filtering on bridged packets. 36Note that 37.Xr ipfw 8 38rules only apply 39to IP packets. 40Non-IP packets are subject to the default 41.Xr ipfw 8 42rule 43(number 65535) 44which must be an 45.Cm allow 46rule if ARP and other non-IP packets need to flow through the 47.Nm . 48.Pp 49The 50.Nm 51configuration is controlled by the 52.Va net.link.ether.bridge_cfg 53variable. 54It consists of a comma-separated list of 55.Ar interface : Ns Ar cluster 56pairs, where all interfaces with the same 57.Ar cluster 58number will 59be bridged together. 60.Pp 61Another 62variable reinitializes the 63.Nm ; 64this is required if bridged 65configurations include loadable interfaces. 66After loading new interface drivers, setting the 67.Va net.link.ether.bridge_refresh 68variable to 69.Li 1 70will cause the 71.Nm 72to reinitialize itself. 73.Sh EXAMPLES 74The following command will cause the 75.Li ep0 76and 77.Li fxp0 78interfaces to be bridged together, and the 79.Li fxp1 80and 81.Li de0 82interfaces to be bridged together: 83.Pp 84.Dl "sysctl net.link.ether.bridge_cfg ep0:0,fxp0:0,fxp1:1,de0:1" 85.Sh BUGS 86Care must be taken not to construct loops in the 87.Nm 88topology. 89The kernel supports only a primitive form of loop detection, by disabling 90some interfaces when a loop is detected. 91No support for a daemon running the 92spanning tree algorithm is currently provided. 93.Pp 94With bridging active, interfaces are in promiscuous mode, 95thus causing some load on the system to receive and filter 96out undesired traffic. 97.Pp 98Extended functionality to enable bridging selectively on clusters 99of interfaces is still in the works. 100.Pp 101Interfaces that cannot be put into promiscuous mode or that don't 102support sending packets with arbitrary Ethernet source addresses 103are not compatible with bridging. 104.Sh SEE ALSO 105.Xr ip 4 , 106.Xr ng_bridge 4 , 107.Xr ipfw 8 , 108.Xr sysctl 8 109.Sh HISTORY 110.Nm 111bridging was introduced in 112.Fx 2.2.8 113by 114.An Luigi Rizzo Aq luigi@iet.unipi.it . 115