diff --git a/share/man/man4/bridge.4 b/share/man/man4/bridge.4 index b4dabcbb79bc..2671bfaf73cc 100644 --- a/share/man/man4/bridge.4 +++ b/share/man/man4/bridge.4 @@ -1,526 +1,550 @@ .\" $NetBSD: bridge.4,v 1.5 2004/01/31 20:14:11 jdc Exp $ .\" .\" Copyright 2001 Wasabi Systems, Inc. .\" All rights reserved. .\" .\" Written by Jason R. Thorpe for Wasabi Systems, Inc. .\" .\" Redistribution and use in source and binary forms, with or without .\" modification, are permitted provided that the following conditions .\" are met: .\" 1. Redistributions of source code must retain the above copyright .\" notice, this list of conditions and the following disclaimer. .\" 2. Redistributions in binary form must reproduce the above copyright .\" notice, this list of conditions and the following disclaimer in the .\" documentation and/or other materials provided with the distribution. .\" 3. All advertising materials mentioning features or use of this software .\" must display the following acknowledgement: .\" This product includes software developed for the NetBSD Project by .\" Wasabi Systems, Inc. .\" 4. The name of Wasabi Systems, Inc. may not be used to endorse .\" or promote products derived from this software without specific prior .\" written permission. .\" .\" THIS SOFTWARE IS PROVIDED BY WASABI SYSTEMS, INC. ``AS IS'' AND .\" ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED .\" TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR .\" PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL WASABI SYSTEMS, INC .\" BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR .\" CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF .\" SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS .\" INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN .\" CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) .\" ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE .\" POSSIBILITY OF SUCH DAMAGE. .\" .\" $FreeBSD$ .\" .Dd April 10, 2023 .Dt IF_BRIDGE 4 .Os .Sh NAME .Nm if_bridge .Nd network bridge device .Sh SYNOPSIS To compile this driver into the kernel, place the following line in your kernel configuration file: .Bd -ragged -offset indent .Cd "device if_bridge" .Ed .Pp Alternatively, to load the driver as a module at boot time, place the following lines in .Xr loader.conf 5 : .Bd -literal -offset indent if_bridge_load="YES" bridgestp_load="YES" .Ed .Sh DESCRIPTION The .Nm driver creates a logical link between two or more IEEE 802 networks that use the same (or .Dq "similar enough" ) framing format. For example, it is possible to bridge Ethernet and 802.11 networks together, but it is not possible to bridge Ethernet and Token Ring together. .Pp Each .Nm interface is created at runtime using interface cloning. This is most easily done with the .Xr ifconfig 8 .Cm create command or using the .Va cloned_interfaces variable in .Xr rc.conf 5 . .Pp The .Nm interface randomly chooses a link (MAC) address in the range reserved for locally administered addresses when it is created. This address is guaranteed to be unique .Em only across all .Nm interfaces on the local machine. Thus you can theoretically have two bridges on different machines with the same link addresses. The address can be changed by assigning the desired link address using .Xr ifconfig 8 . .Pp If .Xr sysctl 8 node .Va net.link.bridge.inherit_mac has a non-zero value, the newly created bridge will inherit the MAC address from its first member instead of choosing a random link-level address. This will provide more predictable bridge MAC addresses without any additional configuration, but currently this feature is known to break some L2 protocols, for example PPPoE that is provided by .Xr ng_pppoe 4 and .Xr ppp 8 . Currently this feature is considered as experimental and is turned off by default. .Pp A bridge can be used to provide several services, such as a simple 802.11-to-Ethernet bridge for wireless hosts, or traffic isolation. .Pp A bridge works like a switch, forwarding traffic from one interface to another. Multicast and broadcast packets are always forwarded to all interfaces that are part of the bridge. For unicast traffic, the bridge learns which MAC addresses are associated with which interfaces and will forward the traffic selectively. .Pp By default the bridge logs MAC address port flapping to .Xr syslog 3 . This behavior can be disabled by setting the .Xr sysctl 8 variable .Va net.link.bridge.log_mac_flap to .Li 0 . .Pp All the bridged member interfaces need to be up in order to pass network traffic. These can be enabled using .Xr ifconfig 8 or .Va ifconfig_ Ns Ao Ar interface Ac Ns Li ="up" in .Xr rc.conf 5 . .Pp The MTU of the first member interface to be added is used as the bridge MTU. All additional members will have their MTU changed to match. If the MTU of a bridge is changed after its creation, the MTU of all member interfaces is also changed to match. .Pp The TOE, TSO, TXCSUM and TXCSUM6 capabilities on all interfaces added to the bridge are disabled if any of the interfaces do not support/enable them. The LRO capability is always disabled. All the capabilities are restored when the interface is removed from the bridge. Changing capabilities at run-time may cause NIC reinit and a link flap. .Pp The bridge supports .Dq monitor mode , where the packets are discarded after .Xr bpf 4 processing, and are not processed or forwarded further. This can be used to multiplex the input of two or more interfaces into a single .Xr bpf 4 stream. This is useful for reconstructing the traffic for network taps that transmit the RX/TX signals out through two separate interfaces. .Sh IPV6 SUPPORT .Nm supports the .Li AF_INET6 address family on bridge interfaces. The following .Xr rc.conf 5 variable configures an IPv6 link-local address on .Li bridge0 interface: .Bd -literal -offset indent ifconfig_bridge0_ipv6="up" .Ed .Pp or in a more explicit manner: .Bd -literal -offset indent ifconfig_bridge0_ipv6="inet6 auto_linklocal" .Ed .Pp However, the .Li AF_INET6 address family has a concept of scope zone. Bridging multiple interfaces changes the zone configuration because multiple links are merged to each other and form a new single link while the member interfaces still work individually. This means each member interface still has a separate link-local scope zone and the .Nm interface has another single, aggregated link-local scope zone at the same time. This situation is clearly against the description .Qq zones of the same scope cannot overlap in Section 5, RFC 4007. Although it works in most cases, it can cause some counterintuitive or undesirable behavior in some edge cases when both, the .Nm interface and one of the member interfaces, have an IPv6 address and applications use both of them. .Pp To prevent this situation, .Nm checks whether a link-local scoped IPv6 address is configured on a member interface to be added and the .Nm interface. When the .Nm interface has IPv6 addresses, IPv6 addresses on the member interface will be automatically removed before the interface is added. .Pp This behavior can be disabled by setting .Xr sysctl 8 variable .Va net.link.bridge.allow_llz_overlap to .Li 1 . .Pp Note that .Li ACCEPT_RTADV and .Li AUTO_LINKLOCAL interface flags are not enabled by default on .Nm interfaces even when .Va net.inet6.ip6.accept_rtadv and/or .Va net.inet6.ip6.auto_linklocal is set to .Li 1 . .Sh SPANNING TREE The .Nm driver implements the Rapid Spanning Tree Protocol (RSTP or 802.1w) with backwards compatibility with the legacy Spanning Tree Protocol (STP). Spanning Tree is used to detect and remove loops in a network topology. .Pp RSTP provides faster spanning tree convergence than legacy STP, the protocol will exchange information with neighbouring switches to quickly transition to forwarding without creating loops. .Pp The code will default to RSTP mode but will downgrade any port connected to a legacy STP network so is fully backward compatible. A bridge can be forced to operate in STP mode without rapid state transitions via the .Va proto command in .Xr ifconfig 8 . .Pp The bridge can log STP port changes to .Xr syslog 3 by setting the .Va net.link.bridge.log_stp node using .Xr sysctl 8 . .Sh PACKET FILTERING Packet filtering can be used with any firewall package that hooks in via the .Xr pfil 9 framework. When filtering is enabled, bridged packets will pass through the filter inbound on the originating interface, on the bridge interface and outbound on the appropriate interfaces. Either stage can be disabled. The filtering behavior can be controlled using .Xr sysctl 8 : .Bl -tag -width ".Va net.link.bridge.pfil_onlyip" .It Va net.link.bridge.pfil_onlyip Controls the handling of non-IP packets which are not passed to .Xr pfil 9 . Set to .Li 1 to only allow IP packets to pass (subject to firewall rules), set to .Li 0 to unconditionally pass all non-IP Ethernet frames. .It Va net.link.bridge.pfil_member Set to .Li 1 to enable filtering on the incoming and outgoing member interfaces, set to .Li 0 to disable it. .It Va net.link.bridge.pfil_bridge Set to .Li 1 to enable filtering on the bridge interface, set to .Li 0 to disable it. .It Va net.link.bridge.pfil_local_phys Set to .Li 1 to additionally filter on the physical interface for locally destined packets. Set to .Li 0 to disable this feature. .It Va net.link.bridge.ipfw Set to .Li 1 to enable layer2 filtering with .Xr ipfirewall 4 , set to .Li 0 to disable it. This needs to be enabled for .Xr dummynet 4 support. When .Va ipfw is enabled, .Va pfil_bridge and .Va pfil_member will be disabled so that IPFW is not run twice; these can be re-enabled if desired. .It Va net.link.bridge.ipfw_arp Set to .Li 1 to enable layer2 ARP filtering with .Xr ipfirewall 4 , set to .Li 0 to disable it. Requires .Va ipfw to be enabled. .El .Pp ARP and REVARP packets are forwarded without being filtered and others that are not IP nor IPv6 packets are not forwarded when .Va pfil_onlyip is enabled. IPFW can filter Ethernet types using .Cm mac-type so all packets are passed to the filter for processing. .Pp The packets originating from the bridging host will be seen by the filter on the interface that is looked up in the routing table. .Pp The packets destined to the bridging host will be seen by the filter on the interface with the MAC address equal to the packet's destination MAC. There are situations when some of the bridge members are sharing the same MAC address (for example the .Xr vlan 4 interfaces: they are currently sharing the MAC address of the parent physical interface). It is not possible to distinguish between these interfaces using their MAC address, excluding the case when the packet's destination MAC address is equal to the MAC address of the interface on which the packet was entered to the system. In this case the filter will see the incoming packet on this interface. In all other cases the interface seen by the packet filter is chosen from the list of bridge members with the same MAC address and the result strongly depends on the member addition sequence and the actual implementation of .Nm . It is not recommended to rely on the order chosen by the current .Nm implementation since it may change in the future. .Pp The previous paragraph is best illustrated with the following pictures. Let .Bl -bullet .It the MAC address of the incoming packet's destination is .Nm nn:nn:nn:nn:nn:nn , .It the interface on which packet entered the system is .Nm ifX , .It .Nm ifX MAC address is .Nm xx:xx:xx:xx:xx:xx , .It there are possibly other bridge members with the same MAC address .Nm xx:xx:xx:xx:xx:xx , .It the bridge has more than one interface that are sharing the same MAC address .Nm yy:yy:yy:yy:yy:yy ; we will call them .Nm vlanY1 , .Nm vlanY2 , etc. .El .Pp If the MAC address .Nm nn:nn:nn:nn:nn:nn is equal to .Nm xx:xx:xx:xx:xx:xx the filter will see the packet on interface .Nm ifX no matter if there are any other bridge members carrying the same MAC address. But if the MAC address .Nm nn:nn:nn:nn:nn:nn is equal to .Nm yy:yy:yy:yy:yy:yy then the interface that will be seen by the filter is one of the .Nm vlanYn . It is not possible to predict the name of the actual interface without the knowledge of the system state and the .Nm implementation details. .Pp This problem arises for any bridge members that are sharing the same MAC address, not only to the .Xr vlan 4 ones: they were taken just as an example of such a situation. So if one wants to filter the locally destined packets based on their interface name, one should be aware of this implication. The described situation will appear at least on the filtering bridges that are doing IP-forwarding; in some of such cases it is better to assign the IP address only to the .Nm interface and not to the bridge members. Enabling .Va net.link.bridge.pfil_local_phys will let you do the additional filtering on the physical interface. +.Sh NETMAP +.Xr netmap 4 +applications may open a bridge interface in emulated mode. +The netmap application will receive all packets which arrive from member +interfaces. +In particular, packets which would otherwise be forwarded to another +member interface will be received by the netmap application. +.Pp +When the +.Xr netmap 4 +application transmits a packet to the host stack via the bridge interface, +.Nm +receive it and attempts to determine its +.Ql source +interface by looking up the source MAC address in the interface's learning +tables. +Packets for which no matching source interface is found are dropped and the +input error counter is incremented. +If a matching source interface is found, +.Nm +treats the packet as though it was received from the corresponding interface +and handles it normally without passing the packet back to +.Xr netmap 4 . .Sh EXAMPLES The following when placed in the file .Pa /etc/rc.conf will cause a bridge called .Dq Li bridge0 to be created, and will add the interfaces .Dq Li wlan0 and .Dq Li fxp0 to the bridge, and then enable packet forwarding. Such a configuration could be used to implement a simple 802.11-to-Ethernet bridge (assuming the 802.11 interface is in ad-hoc mode). .Bd -literal -offset indent cloned_interfaces="bridge0" ifconfig_bridge0="addm wlan0 addm fxp0 up" .Ed .Pp For the bridge to forward packets, all member interfaces and the bridge need to be up. The above example would also require: .Bd -literal -offset indent create_args_wlan0="wlanmode hostap" ifconfig_wlan0="up ssid my_ap mode 11g" ifconfig_fxp0="up" .Ed .Pp Consider a system with two 4-port Ethernet boards. The following will cause a bridge consisting of all 8 ports with Rapid Spanning Tree enabled to be created: .Bd -literal -offset indent ifconfig bridge0 create ifconfig bridge0 \e addm fxp0 stp fxp0 \e addm fxp1 stp fxp1 \e addm fxp2 stp fxp2 \e addm fxp3 stp fxp3 \e addm fxp4 stp fxp4 \e addm fxp5 stp fxp5 \e addm fxp6 stp fxp6 \e addm fxp7 stp fxp7 \e up .Ed .Pp The bridge can be used as a regular host interface at the same time as bridging between its member ports. In this example, the bridge connects em0 and em1, and will receive its IP address through DHCP: .Bd -literal -offset indent cloned_interfaces="bridge0" ifconfig_bridge0="addm em0 addm em1 DHCP" ifconfig_em0="up" ifconfig_em1="up" .Ed .Pp The bridge can tunnel Ethernet across an IP internet using the EtherIP protocol. This can be combined with .Xr ipsec 4 to provide an encrypted connection. Create a .Xr gif 4 interface and set the local and remote IP addresses for the tunnel, these are reversed on the remote bridge. .Bd -literal -offset indent ifconfig gif0 create ifconfig gif0 tunnel 1.2.3.4 5.6.7.8 up ifconfig bridge0 create ifconfig bridge0 addm fxp0 addm gif0 up .Ed .Sh SEE ALSO .Xr gif 4 , .Xr ipf 4 , .Xr ipfw 4 , +.Xr netmap 4 , .Xr pf 4 , .Xr ifconfig 8 .Sh HISTORY The .Nm driver first appeared in .Fx 6.0 . .Sh AUTHORS .An -nosplit The .Nm bridge driver was originally written by .An Jason L. Wright Aq Mt jason@thought.net as part of an undergraduate independent study at the University of North Carolina at Greensboro. .Pp This version of the .Nm driver has been heavily modified from the original version by .An Jason R. Thorpe Aq Mt thorpej@wasabisystems.com . .Pp Rapid Spanning Tree Protocol (RSTP) support was added by .An Andrew Thompson Aq Mt thompsa@FreeBSD.org . .Sh BUGS The .Nm driver currently supports only Ethernet and Ethernet-like (e.g., 802.11) network devices, which can be configured with the same MTU size as the bridge device. diff --git a/sys/net/ethernet.h b/sys/net/ethernet.h index 98b02d71da50..e259ab53e77e 100644 --- a/sys/net/ethernet.h +++ b/sys/net/ethernet.h @@ -1,482 +1,483 @@ /* * Fundamental constants relating to ethernet. * * $FreeBSD$ * */ #ifndef _NET_ETHERNET_H_ #define _NET_ETHERNET_H_ #include /* * Some basic Ethernet constants. */ #define ETHER_ADDR_LEN 6 /* length of an Ethernet address */ #define ETHER_TYPE_LEN 2 /* length of the Ethernet type field */ #define ETHER_CRC_LEN 4 /* length of the Ethernet CRC */ #define ETHER_HDR_LEN (ETHER_ADDR_LEN*2+ETHER_TYPE_LEN) #define ETHER_MIN_LEN 64 /* minimum frame len, including CRC */ #define ETHER_MAX_LEN 1518 /* maximum frame len, including CRC */ #define ETHER_MAX_LEN_JUMBO 9018 /* max jumbo frame len, including CRC */ #define ETHER_VLAN_ENCAP_LEN 4 /* len of 802.1Q VLAN encapsulation */ /* * Mbuf adjust factor to force 32-bit alignment of IP header. * Drivers should do m_adj(m, ETHER_ALIGN) when setting up a * receive so the upper layers get the IP header properly aligned * past the 14-byte Ethernet header. */ #define ETHER_ALIGN 2 /* driver adjust for IP hdr alignment */ /* * Compute the maximum frame size based on ethertype (i.e. possible * encapsulation) and whether or not an FCS is present. */ #define ETHER_MAX_FRAME(ifp, etype, hasfcs) \ (if_getmtu(ifp) + ETHER_HDR_LEN + \ ((hasfcs) ? ETHER_CRC_LEN : 0) + \ (((etype) == ETHERTYPE_VLAN) ? ETHER_VLAN_ENCAP_LEN : 0)) /* * Ethernet-specific mbuf flags. */ #define M_HASFCS M_PROTO5 /* FCS included at end of frame */ +#define M_BRIDGE_INJECT M_PROTO6 /* if_bridge-injected frame */ /* * Ethernet CRC32 polynomials (big- and little-endian versions). */ #define ETHER_CRC_POLY_LE 0xedb88320 #define ETHER_CRC_POLY_BE 0x04c11db6 /* * A macro to validate a length with */ #define ETHER_IS_VALID_LEN(foo) \ ((foo) >= ETHER_MIN_LEN && (foo) <= ETHER_MAX_LEN) /* * Structure of a 10Mb/s Ethernet header. */ struct ether_header { u_char ether_dhost[ETHER_ADDR_LEN]; u_char ether_shost[ETHER_ADDR_LEN]; u_short ether_type; } __packed; /* * Structure of a 48-bit Ethernet address. */ struct ether_addr { u_char octet[ETHER_ADDR_LEN]; } __packed; #define ETHER_IS_MULTICAST(addr) (*(addr) & 0x01) /* is address mcast/bcast? */ #define ETHER_IS_IPV6_MULTICAST(addr) \ (((addr)[0] == 0x33) && ((addr)[1] == 0x33)) #define ETHER_IS_BROADCAST(addr) \ (((addr)[0] & (addr)[1] & (addr)[2] & \ (addr)[3] & (addr)[4] & (addr)[5]) == 0xff) #define ETHER_IS_ZERO(addr) \ (((addr)[0] | (addr)[1] | (addr)[2] | \ (addr)[3] | (addr)[4] | (addr)[5]) == 0x00) /* * 802.1q Virtual LAN header. */ struct ether_vlan_header { uint8_t evl_dhost[ETHER_ADDR_LEN]; uint8_t evl_shost[ETHER_ADDR_LEN]; uint16_t evl_encap_proto; uint16_t evl_tag; uint16_t evl_proto; } __packed; #define EVL_VLID_MASK 0x0FFF #define EVL_PRI_MASK 0xE000 #define EVL_VLANOFTAG(tag) ((tag) & EVL_VLID_MASK) #define EVL_PRIOFTAG(tag) (((tag) >> 13) & 7) #define EVL_CFIOFTAG(tag) (((tag) >> 12) & 1) #define EVL_MAKETAG(vlid, pri, cfi) \ ((((((pri) & 7) << 1) | ((cfi) & 1)) << 12) | ((vlid) & EVL_VLID_MASK)) /* * Ethernet protocol types. * * A public list is available from the IEEE Registration Authority: * https://standards.ieee.org/products-services/regauth/ * * NOTE: 0x0000-0x05DC (0..1500) are generally IEEE 802.3 length fields. * However, there are some conflicts. */ #define ETHERTYPE_8023 0x0004 /* IEEE 802.3 packet */ /* 0x0101 .. 0x1FF Experimental */ #define ETHERTYPE_PUP 0x0200 /* Xerox PUP protocol - see 0A00 */ #define ETHERTYPE_PUPAT 0x0200 /* PUP Address Translation - see 0A01 */ #define ETHERTYPE_SPRITE 0x0500 /* ??? */ /* 0x0400 Nixdorf */ #define ETHERTYPE_NS 0x0600 /* XNS */ #define ETHERTYPE_NSAT 0x0601 /* XNS Address Translation (3Mb only) */ #define ETHERTYPE_DLOG1 0x0660 /* DLOG (?) */ #define ETHERTYPE_DLOG2 0x0661 /* DLOG (?) */ #define ETHERTYPE_IP 0x0800 /* IP protocol */ #define ETHERTYPE_X75 0x0801 /* X.75 Internet */ #define ETHERTYPE_NBS 0x0802 /* NBS Internet */ #define ETHERTYPE_ECMA 0x0803 /* ECMA Internet */ #define ETHERTYPE_CHAOS 0x0804 /* CHAOSnet */ #define ETHERTYPE_X25 0x0805 /* X.25 Level 3 */ #define ETHERTYPE_ARP 0x0806 /* Address resolution protocol */ #define ETHERTYPE_NSCOMPAT 0x0807 /* XNS Compatibility */ #define ETHERTYPE_FRARP 0x0808 /* Frame Relay ARP (RFC1701) */ /* 0x081C Symbolics Private */ /* 0x0888 - 0x088A Xyplex */ #define ETHERTYPE_UBDEBUG 0x0900 /* Ungermann-Bass network debugger */ #define ETHERTYPE_IEEEPUP 0x0A00 /* Xerox IEEE802.3 PUP */ #define ETHERTYPE_IEEEPUPAT 0x0A01 /* Xerox IEEE802.3 PUP Address Translation */ #define ETHERTYPE_VINES 0x0BAD /* Banyan VINES */ #define ETHERTYPE_VINESLOOP 0x0BAE /* Banyan VINES Loopback */ #define ETHERTYPE_VINESECHO 0x0BAF /* Banyan VINES Echo */ /* 0x1000 - 0x100F Berkeley Trailer */ /* * The ETHERTYPE_NTRAILER packet types starting at ETHERTYPE_TRAIL have * (type-ETHERTYPE_TRAIL)*512 bytes of data followed * by an ETHER type (as given above) and then the (variable-length) header. */ #define ETHERTYPE_TRAIL 0x1000 /* Trailer packet */ #define ETHERTYPE_NTRAILER 16 #define ETHERTYPE_DCA 0x1234 /* DCA - Multicast */ #define ETHERTYPE_VALID 0x1600 /* VALID system protocol */ #define ETHERTYPE_DOGFIGHT 0x1989 /* Artificial Horizons ("Aviator" dogfight simulator [on Sun]) */ #define ETHERTYPE_RCL 0x1995 /* Datapoint Corporation (RCL lan protocol) */ /* The following 3C0x types are unregistered: */ #define ETHERTYPE_NBPVCD 0x3C00 /* 3Com NBP virtual circuit datagram (like XNS SPP) not registered */ #define ETHERTYPE_NBPSCD 0x3C01 /* 3Com NBP System control datagram not registered */ #define ETHERTYPE_NBPCREQ 0x3C02 /* 3Com NBP Connect request (virtual cct) not registered */ #define ETHERTYPE_NBPCRSP 0x3C03 /* 3Com NBP Connect response not registered */ #define ETHERTYPE_NBPCC 0x3C04 /* 3Com NBP Connect complete not registered */ #define ETHERTYPE_NBPCLREQ 0x3C05 /* 3Com NBP Close request (virtual cct) not registered */ #define ETHERTYPE_NBPCLRSP 0x3C06 /* 3Com NBP Close response not registered */ #define ETHERTYPE_NBPDG 0x3C07 /* 3Com NBP Datagram (like XNS IDP) not registered */ #define ETHERTYPE_NBPDGB 0x3C08 /* 3Com NBP Datagram broadcast not registered */ #define ETHERTYPE_NBPCLAIM 0x3C09 /* 3Com NBP Claim NetBIOS name not registered */ #define ETHERTYPE_NBPDLTE 0x3C0A /* 3Com NBP Delete NetBIOS name not registered */ #define ETHERTYPE_NBPRAS 0x3C0B /* 3Com NBP Remote adaptor status request not registered */ #define ETHERTYPE_NBPRAR 0x3C0C /* 3Com NBP Remote adaptor response not registered */ #define ETHERTYPE_NBPRST 0x3C0D /* 3Com NBP Reset not registered */ #define ETHERTYPE_PCS 0x4242 /* PCS Basic Block Protocol */ #define ETHERTYPE_IMLBLDIAG 0x424C /* Information Modes Little Big LAN diagnostic */ #define ETHERTYPE_DIDDLE 0x4321 /* THD - Diddle */ #define ETHERTYPE_IMLBL 0x4C42 /* Information Modes Little Big LAN */ #define ETHERTYPE_SIMNET 0x5208 /* BBN Simnet Private */ #define ETHERTYPE_DECEXPER 0x6000 /* DEC Unassigned, experimental */ #define ETHERTYPE_MOPDL 0x6001 /* DEC MOP dump/load */ #define ETHERTYPE_MOPRC 0x6002 /* DEC MOP remote console */ #define ETHERTYPE_DECnet 0x6003 /* DEC DECNET Phase IV route */ #define ETHERTYPE_DN ETHERTYPE_DECnet /* libpcap, tcpdump */ #define ETHERTYPE_LAT 0x6004 /* DEC LAT */ #define ETHERTYPE_DECDIAG 0x6005 /* DEC diagnostic protocol (at interface initialization?) */ #define ETHERTYPE_DECCUST 0x6006 /* DEC customer protocol */ #define ETHERTYPE_SCA 0x6007 /* DEC LAVC, SCA */ #define ETHERTYPE_AMBER 0x6008 /* DEC AMBER */ #define ETHERTYPE_DECMUMPS 0x6009 /* DEC MUMPS */ /* 0x6010 - 0x6014 3Com Corporation */ #define ETHERTYPE_TRANSETHER 0x6558 /* Trans Ether Bridging (RFC1701)*/ #define ETHERTYPE_RAWFR 0x6559 /* Raw Frame Relay (RFC1701) */ #define ETHERTYPE_UBDL 0x7000 /* Ungermann-Bass download */ #define ETHERTYPE_UBNIU 0x7001 /* Ungermann-Bass NIUs */ #define ETHERTYPE_UBDIAGLOOP 0x7002 /* Ungermann-Bass diagnostic/loopback */ #define ETHERTYPE_UBNMC 0x7003 /* Ungermann-Bass ??? (NMC to/from UB Bridge) */ #define ETHERTYPE_UBBST 0x7005 /* Ungermann-Bass Bridge Spanning Tree */ #define ETHERTYPE_OS9 0x7007 /* OS/9 Microware */ #define ETHERTYPE_OS9NET 0x7009 /* OS/9 Net? */ /* 0x7020 - 0x7029 LRT (England) (now Sintrom) */ #define ETHERTYPE_RACAL 0x7030 /* Racal-Interlan */ #define ETHERTYPE_PRIMENTS 0x7031 /* Prime NTS (Network Terminal Service) */ #define ETHERTYPE_CABLETRON 0x7034 /* Cabletron */ #define ETHERTYPE_CRONUSVLN 0x8003 /* Cronus VLN */ #define ETHERTYPE_CRONUS 0x8004 /* Cronus Direct */ #define ETHERTYPE_HP 0x8005 /* HP Probe */ #define ETHERTYPE_NESTAR 0x8006 /* Nestar */ #define ETHERTYPE_ATTSTANFORD 0x8008 /* AT&T/Stanford (local use) */ #define ETHERTYPE_EXCELAN 0x8010 /* Excelan */ #define ETHERTYPE_SG_DIAG 0x8013 /* SGI diagnostic type */ #define ETHERTYPE_SG_NETGAMES 0x8014 /* SGI network games */ #define ETHERTYPE_SG_RESV 0x8015 /* SGI reserved type */ #define ETHERTYPE_SG_BOUNCE 0x8016 /* SGI bounce server */ #define ETHERTYPE_APOLLODOMAIN 0x8019 /* Apollo DOMAIN */ #define ETHERTYPE_TYMSHARE 0x802E /* Tymeshare */ #define ETHERTYPE_TIGAN 0x802F /* Tigan, Inc. */ #define ETHERTYPE_REVARP 0x8035 /* Reverse addr resolution protocol */ #define ETHERTYPE_AEONIC 0x8036 /* Aeonic Systems */ #define ETHERTYPE_IPXNEW 0x8037 /* IPX (Novell Netware?) */ #define ETHERTYPE_LANBRIDGE 0x8038 /* DEC LANBridge */ #define ETHERTYPE_DSMD 0x8039 /* DEC DSM/DDP */ #define ETHERTYPE_ARGONAUT 0x803A /* DEC Argonaut Console */ #define ETHERTYPE_VAXELN 0x803B /* DEC VAXELN */ #define ETHERTYPE_DECDNS 0x803C /* DEC DNS Naming Service */ #define ETHERTYPE_ENCRYPT 0x803D /* DEC Ethernet Encryption */ #define ETHERTYPE_DECDTS 0x803E /* DEC Distributed Time Service */ #define ETHERTYPE_DECLTM 0x803F /* DEC LAN Traffic Monitor */ #define ETHERTYPE_DECNETBIOS 0x8040 /* DEC PATHWORKS DECnet NETBIOS Emulation */ #define ETHERTYPE_DECLAST 0x8041 /* DEC Local Area System Transport */ /* 0x8042 DEC Unassigned */ #define ETHERTYPE_PLANNING 0x8044 /* Planning Research Corp. */ /* 0x8046 - 0x8047 AT&T */ #define ETHERTYPE_DECAM 0x8048 /* DEC Availability Manager for Distributed Systems DECamds (but someone at DEC says not) */ #define ETHERTYPE_EXPERDATA 0x8049 /* ExperData */ #define ETHERTYPE_VEXP 0x805B /* Stanford V Kernel exp. */ #define ETHERTYPE_VPROD 0x805C /* Stanford V Kernel prod. */ #define ETHERTYPE_ES 0x805D /* Evans & Sutherland */ #define ETHERTYPE_LITTLE 0x8060 /* Little Machines */ #define ETHERTYPE_COUNTERPOINT 0x8062 /* Counterpoint Computers */ /* 0x8065 - 0x8066 Univ. of Mass @ Amherst */ #define ETHERTYPE_VEECO 0x8067 /* Veeco Integrated Auto. */ #define ETHERTYPE_GENDYN 0x8068 /* General Dynamics */ #define ETHERTYPE_ATT 0x8069 /* AT&T */ #define ETHERTYPE_AUTOPHON 0x806A /* Autophon */ #define ETHERTYPE_COMDESIGN 0x806C /* ComDesign */ #define ETHERTYPE_COMPUGRAPHIC 0x806D /* Compugraphic Corporation */ /* 0x806E - 0x8077 Landmark Graphics Corp. */ #define ETHERTYPE_MATRA 0x807A /* Matra */ #define ETHERTYPE_DDE 0x807B /* Dansk Data Elektronik */ #define ETHERTYPE_MERIT 0x807C /* Merit Internodal (or Univ of Michigan?) */ /* 0x807D - 0x807F Vitalink Communications */ #define ETHERTYPE_VLTLMAN 0x8080 /* Vitalink TransLAN III Management */ /* 0x8081 - 0x8083 Counterpoint Computers */ /* 0x8088 - 0x808A Xyplex */ #define ETHERTYPE_ATALK 0x809B /* AppleTalk */ #define ETHERTYPE_AT ETHERTYPE_ATALK /* old NetBSD */ #define ETHERTYPE_APPLETALK ETHERTYPE_ATALK /* HP-UX */ /* 0x809C - 0x809E Datability */ #define ETHERTYPE_SPIDER 0x809F /* Spider Systems Ltd. */ /* 0x80A3 Nixdorf */ /* 0x80A4 - 0x80B3 Siemens Gammasonics Inc. */ /* 0x80C0 - 0x80C3 DCA (Digital Comm. Assoc.) Data Exchange Cluster */ /* 0x80C4 - 0x80C5 Banyan Systems */ #define ETHERTYPE_PACER 0x80C6 /* Pacer Software */ #define ETHERTYPE_APPLITEK 0x80C7 /* Applitek Corporation */ /* 0x80C8 - 0x80CC Intergraph Corporation */ /* 0x80CD - 0x80CE Harris Corporation */ /* 0x80CF - 0x80D2 Taylor Instrument */ /* 0x80D3 - 0x80D4 Rosemount Corporation */ #define ETHERTYPE_SNA 0x80D5 /* IBM SNA Services over Ethernet */ #define ETHERTYPE_VARIAN 0x80DD /* Varian Associates */ /* 0x80DE - 0x80DF TRFS (Integrated Solutions Transparent Remote File System) */ /* 0x80E0 - 0x80E3 Allen-Bradley */ /* 0x80E4 - 0x80F0 Datability */ #define ETHERTYPE_RETIX 0x80F2 /* Retix */ #define ETHERTYPE_AARP 0x80F3 /* AppleTalk AARP */ /* 0x80F4 - 0x80F5 Kinetics */ #define ETHERTYPE_APOLLO 0x80F7 /* Apollo Computer */ #define ETHERTYPE_VLAN 0x8100 /* IEEE 802.1Q VLAN tagging (XXX conflicts) */ /* 0x80FF - 0x8101 Wellfleet Communications (XXX conflicts) */ #define ETHERTYPE_BOFL 0x8102 /* Wellfleet; BOFL (Breath OF Life) pkts [every 5-10 secs.] */ #define ETHERTYPE_WELLFLEET 0x8103 /* Wellfleet Communications */ /* 0x8107 - 0x8109 Symbolics Private */ #define ETHERTYPE_TALARIS 0x812B /* Talaris */ #define ETHERTYPE_WATERLOO 0x8130 /* Waterloo Microsystems Inc. (XXX which?) */ #define ETHERTYPE_HAYES 0x8130 /* Hayes Microcomputers (XXX which?) */ #define ETHERTYPE_VGLAB 0x8131 /* VG Laboratory Systems */ /* 0x8132 - 0x8137 Bridge Communications */ #define ETHERTYPE_IPX 0x8137 /* Novell (old) NetWare IPX (ECONFIG E option) */ #define ETHERTYPE_NOVELL 0x8138 /* Novell, Inc. */ /* 0x8139 - 0x813D KTI */ #define ETHERTYPE_MUMPS 0x813F /* M/MUMPS data sharing */ #define ETHERTYPE_AMOEBA 0x8145 /* Vrije Universiteit (NL) Amoeba 4 RPC (obsolete) */ #define ETHERTYPE_FLIP 0x8146 /* Vrije Universiteit (NL) FLIP (Fast Local Internet Protocol) */ #define ETHERTYPE_VURESERVED 0x8147 /* Vrije Universiteit (NL) [reserved] */ #define ETHERTYPE_LOGICRAFT 0x8148 /* Logicraft */ #define ETHERTYPE_NCD 0x8149 /* Network Computing Devices */ #define ETHERTYPE_ALPHA 0x814A /* Alpha Micro */ #define ETHERTYPE_SNMP 0x814C /* SNMP over Ethernet (see RFC1089) */ /* 0x814D - 0x814E BIIN */ #define ETHERTYPE_TEC 0x814F /* Technically Elite Concepts */ #define ETHERTYPE_RATIONAL 0x8150 /* Rational Corp */ /* 0x8151 - 0x8153 Qualcomm */ /* 0x815C - 0x815E Computer Protocol Pty Ltd */ /* 0x8164 - 0x8166 Charles River Data Systems */ #define ETHERTYPE_XTP 0x817D /* Protocol Engines XTP */ #define ETHERTYPE_SGITW 0x817E /* SGI/Time Warner prop. */ #define ETHERTYPE_HIPPI_FP 0x8180 /* HIPPI-FP encapsulation */ #define ETHERTYPE_STP 0x8181 /* Scheduled Transfer STP, HIPPI-ST */ /* 0x8182 - 0x8183 Reserved for HIPPI-6400 */ /* 0x8184 - 0x818C SGI prop. */ #define ETHERTYPE_MOTOROLA 0x818D /* Motorola */ #define ETHERTYPE_NETBEUI 0x8191 /* PowerLAN NetBIOS/NetBEUI (PC) */ /* 0x819A - 0x81A3 RAD Network Devices */ /* 0x81B7 - 0x81B9 Xyplex */ /* 0x81CC - 0x81D5 Apricot Computers */ /* 0x81D6 - 0x81DD Artisoft Lantastic */ /* 0x81E6 - 0x81EF Polygon */ /* 0x81F0 - 0x81F2 Comsat Labs */ /* 0x81F3 - 0x81F5 SAIC */ /* 0x81F6 - 0x81F8 VG Analytical */ /* 0x8203 - 0x8205 QNX Software Systems Ltd. */ /* 0x8221 - 0x8222 Ascom Banking Systems */ /* 0x823E - 0x8240 Advanced Encryption Systems */ /* 0x8263 - 0x826A Charles River Data Systems */ /* 0x827F - 0x8282 Athena Programming */ /* 0x829A - 0x829B Inst Ind Info Tech */ /* 0x829C - 0x82AB Taurus Controls */ /* 0x82AC - 0x8693 Walker Richer & Quinn */ #define ETHERTYPE_ACCTON 0x8390 /* Accton Technologies (unregistered) */ #define ETHERTYPE_TALARISMC 0x852B /* Talaris multicast */ #define ETHERTYPE_KALPANA 0x8582 /* Kalpana */ /* 0x8694 - 0x869D Idea Courier */ /* 0x869E - 0x86A1 Computer Network Tech */ /* 0x86A3 - 0x86AC Gateway Communications */ #define ETHERTYPE_SECTRA 0x86DB /* SECTRA */ #define ETHERTYPE_IPV6 0x86DD /* IP protocol version 6 */ #define ETHERTYPE_DELTACON 0x86DE /* Delta Controls */ #define ETHERTYPE_ATOMIC 0x86DF /* ATOMIC */ /* 0x86E0 - 0x86EF Landis & Gyr Powers */ /* 0x8700 - 0x8710 Motorola */ #define ETHERTYPE_RDP 0x8739 /* Control Technology Inc. RDP Without IP */ #define ETHERTYPE_MICP 0x873A /* Control Technology Inc. Mcast Industrial Ctrl Proto. */ /* 0x873B - 0x873C Control Technology Inc. Proprietary */ #define ETHERTYPE_TCPCOMP 0x876B /* TCP/IP Compression (RFC1701) */ #define ETHERTYPE_IPAS 0x876C /* IP Autonomous Systems (RFC1701) */ #define ETHERTYPE_SECUREDATA 0x876D /* Secure Data (RFC1701) */ #define ETHERTYPE_FLOWCONTROL 0x8808 /* 802.3x flow control packet */ #define ETHERTYPE_SLOW 0x8809 /* 802.3ad link aggregation (LACP) */ #define ETHERTYPE_PPP 0x880B /* PPP (obsolete by PPPoE) */ #define ETHERTYPE_HITACHI 0x8820 /* Hitachi Cable (Optoelectronic Systems Laboratory) */ #define ETHERTYPE_TEST 0x8822 /* Network Conformance Testing */ #define ETHERTYPE_MPLS 0x8847 /* MPLS Unicast */ #define ETHERTYPE_MPLS_MCAST 0x8848 /* MPLS Multicast */ #define ETHERTYPE_AXIS 0x8856 /* Axis Communications AB proprietary bootstrap/config */ #define ETHERTYPE_PPPOEDISC 0x8863 /* PPP Over Ethernet Discovery Stage */ #define ETHERTYPE_PPPOE 0x8864 /* PPP Over Ethernet Session Stage */ #define ETHERTYPE_LANPROBE 0x8888 /* HP LanProbe test? */ #define ETHERTYPE_PAE 0x888E /* EAPOL PAE/802.1x */ #define ETHERTYPE_PROFINET 0x8892 /* PROFINET RT Protocol */ #define ETHERTYPE_AOE 0x88A2 /* ATA Protocol */ #define ETHERTYPE_ETHERCAT 0x88A4 /* EtherCat Protocol */ #define ETHERTYPE_QINQ 0x88A8 /* 802.1ad VLAN stacking */ #define ETHERTYPE_POWERLINK 0x88AB /* Ethernet Powerlink (EPL) */ #define ETHERTYPE_LLDP 0x88CC /* Link Layer Discovery Protocol */ #define ETHERTYPE_SERCOS 0x88CD /* SERCOS III Protocol */ #define ETHERTYPE_MACSEC 0x88E5 /* 802.1AE MAC Security */ #define ETHERTYPE_PBB 0x88E7 /* 802.1Q Provider Backbone Bridges */ #define ETHERTYPE_FCOE 0x8906 /* Fibre Channel over Ethernet */ #define ETHERTYPE_LOOPBACK 0x9000 /* Loopback: used to test interfaces */ #define ETHERTYPE_8021Q9100 0x9100 /* IEEE 802.1Q stacking (proprietary) */ #define ETHERTYPE_LBACK ETHERTYPE_LOOPBACK /* DEC MOP loopback */ #define ETHERTYPE_XNSSM 0x9001 /* 3Com (Formerly Bridge Communications), XNS Systems Management */ #define ETHERTYPE_TCPSM 0x9002 /* 3Com (Formerly Bridge Communications), TCP/IP Systems Management */ #define ETHERTYPE_BCLOOP 0x9003 /* 3Com (Formerly Bridge Communications), loopback detection */ #define ETHERTYPE_8021Q9200 0x9200 /* IEEE 802.1Q stacking (proprietary) */ #define ETHERTYPE_8021Q9300 0x9300 /* IEEE 802.1Q stacking (proprietary) */ #define ETHERTYPE_DEBNI 0xAAAA /* DECNET? Used by VAX 6220 DEBNI */ #define ETHERTYPE_SONIX 0xFAF5 /* Sonix Arpeggio */ #define ETHERTYPE_VITAL 0xFF00 /* BBN VITAL-LanBridge cache wakeups */ /* 0xFF00 - 0xFFOF ISC Bunker Ramo */ #define ETHERTYPE_MAX 0xFFFF /* Maximum valid ethernet type, reserved */ /* * The ETHERTYPE_NTRAILER packet types starting at ETHERTYPE_TRAIL have * (type-ETHERTYPE_TRAIL)*512 bytes of data followed * by an ETHER type (as given above) and then the (variable-length) header. */ #define ETHERTYPE_TRAIL 0x1000 /* Trailer packet */ #define ETHERTYPE_NTRAILER 16 #define ETHERMTU (ETHER_MAX_LEN-ETHER_HDR_LEN-ETHER_CRC_LEN) #define ETHERMIN (ETHER_MIN_LEN-ETHER_HDR_LEN-ETHER_CRC_LEN) #define ETHERMTU_JUMBO (ETHER_MAX_LEN_JUMBO - ETHER_HDR_LEN - ETHER_CRC_LEN) /* * The ETHER_BPF_MTAP macro should be used by drivers which support hardware * offload for VLAN tag processing. It will check the mbuf to see if it has * M_VLANTAG set, and if it does, will pass the packet along to * ether_vlan_mtap. This function will re-insert VLAN tags for the duration * of the tap, so they show up properly for network analyzers. */ struct ifnet; struct mbuf; void ether_bpf_mtap_if(struct ifnet *ifp, struct mbuf *m); #define ETHER_BPF_MTAP(_ifp, _m) ether_bpf_mtap_if((_ifp), (_m)) /* * Names for 802.1q priorities ("802.1p"). Notice that in this scheme, * (0 < 1), allowing default 0-tagged traffic to take priority over background * tagged traffic. */ #define IEEE8021Q_PCP_BK 1 /* Background (lowest) */ #define IEEE8021Q_PCP_BE 0 /* Best effort (default) */ #define IEEE8021Q_PCP_EE 2 /* Excellent effort */ #define IEEE8021Q_PCP_CA 3 /* Critical applications */ #define IEEE8021Q_PCP_VI 4 /* Video, < 100ms latency and jitter */ #define IEEE8021Q_PCP_VO 5 /* Voice, < 10ms latency and jitter */ #define IEEE8021Q_PCP_IC 6 /* Internetwork control */ #define IEEE8021Q_PCP_NC 7 /* Network control (highest) */ #ifdef _KERNEL #include struct ifnet; struct mbuf; struct route; struct sockaddr; struct bpf_if; struct ether_8021q_tag; extern uint32_t ether_crc32_le(const uint8_t *, size_t); extern uint32_t ether_crc32_be(const uint8_t *, size_t); extern void ether_demux(struct ifnet *, struct mbuf *); extern void ether_ifattach(struct ifnet *, const u_int8_t *); extern void ether_ifdetach(struct ifnet *); #ifdef VIMAGE struct vnet; extern void ether_reassign(struct ifnet *, struct vnet *, char *); #endif extern int ether_ioctl(struct ifnet *, u_long, caddr_t); extern int ether_output(struct ifnet *, struct mbuf *, const struct sockaddr *, struct route *); extern int ether_output_frame(struct ifnet *, struct mbuf *); extern char *ether_sprintf(const u_int8_t *); void ether_vlan_mtap(struct bpf_if *, struct mbuf *, void *, u_int); struct mbuf *ether_vlanencap_proto(struct mbuf *, uint16_t, uint16_t); bool ether_8021q_frame(struct mbuf **mp, struct ifnet *ife, struct ifnet *p, struct ether_8021q_tag *); void ether_gen_addr(struct ifnet *ifp, struct ether_addr *hwaddr); static __inline struct mbuf *ether_vlanencap(struct mbuf *m, uint16_t tag) { return ether_vlanencap_proto(m, tag, ETHERTYPE_VLAN); } /* new ethernet interface attached event */ typedef void (*ether_ifattach_event_handler_t)(void *, struct ifnet *); EVENTHANDLER_DECLARE(ether_ifattach_event, ether_ifattach_event_handler_t); #else /* _KERNEL */ #include /* * Ethernet address conversion/parsing routines. */ __BEGIN_DECLS struct ether_addr *ether_aton(const char *); struct ether_addr *ether_aton_r(const char *, struct ether_addr *); int ether_hostton(const char *, struct ether_addr *); int ether_line(const char *, struct ether_addr *, char *); char *ether_ntoa(const struct ether_addr *); char *ether_ntoa_r(const struct ether_addr *, char *); int ether_ntohost(char *, const struct ether_addr *); __END_DECLS #endif /* !_KERNEL */ #endif /* !_NET_ETHERNET_H_ */ diff --git a/sys/net/if_bridge.c b/sys/net/if_bridge.c index 78be987d681b..f71d7c0cdfd1 100644 --- a/sys/net/if_bridge.c +++ b/sys/net/if_bridge.c @@ -1,3815 +1,3907 @@ /* $NetBSD: if_bridge.c,v 1.31 2005/06/01 19:45:34 jdc Exp $ */ /*- * SPDX-License-Identifier: BSD-4-Clause * * Copyright 2001 Wasabi Systems, Inc. * All rights reserved. * * Written by Jason R. Thorpe for Wasabi Systems, Inc. * * Redistribution and use in source and binary forms, with or without * modification, are permitted provided that the following conditions * are met: * 1. Redistributions of source code must retain the above copyright * notice, this list of conditions and the following disclaimer. * 2. Redistributions in binary form must reproduce the above copyright * notice, this list of conditions and the following disclaimer in the * documentation and/or other materials provided with the distribution. * 3. All advertising materials mentioning features or use of this software * must display the following acknowledgement: * This product includes software developed for the NetBSD Project by * Wasabi Systems, Inc. * 4. The name of Wasabi Systems, Inc. may not be used to endorse * or promote products derived from this software without specific prior * written permission. * * THIS SOFTWARE IS PROVIDED BY WASABI SYSTEMS, INC. ``AS IS'' AND * ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED * TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR * PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL WASABI SYSTEMS, INC * BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR * CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF * SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS * INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN * CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) * ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE * POSSIBILITY OF SUCH DAMAGE. */ /* * Copyright (c) 1999, 2000 Jason L. Wright (jason@thought.net) * All rights reserved. * * Redistribution and use in source and binary forms, with or without * modification, are permitted provided that the following conditions * are met: * 1. Redistributions of source code must retain the above copyright * notice, this list of conditions and the following disclaimer. * 2. Redistributions in binary form must reproduce the above copyright * notice, this list of conditions and the following disclaimer in the * documentation and/or other materials provided with the distribution. * * THIS SOFTWARE IS PROVIDED BY THE AUTHOR ``AS IS'' AND ANY EXPRESS OR * IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE IMPLIED * WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE ARE * DISCLAIMED. IN NO EVENT SHALL THE AUTHOR BE LIABLE FOR ANY DIRECT, * INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES * (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR * SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, * STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN * ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE * POSSIBILITY OF SUCH DAMAGE. * * OpenBSD: if_bridge.c,v 1.60 2001/06/15 03:38:33 itojun Exp */ /* * Network interface bridge support. * * TODO: * * - Currently only supports Ethernet-like interfaces (Ethernet, * 802.11, VLANs on Ethernet, etc.) Figure out a nice way * to bridge other types of interfaces (maybe consider * heterogeneous bridges). */ #include __FBSDID("$FreeBSD$"); #include "opt_inet.h" #include "opt_inet6.h" #include #include #include #include #include #include #include #include #include /* for net/if.h */ #include #include /* string functions */ #include #include #include #include #include #include #include #include #include #include #include #include #include #include #include #include #include #include #include #include #include #include #include #include #ifdef INET6 #include #include #include #endif #if defined(INET) || defined(INET6) #include #endif #include #include #include #include #include #include #include /* * Size of the route hash table. Must be a power of two. */ #ifndef BRIDGE_RTHASH_SIZE #define BRIDGE_RTHASH_SIZE 1024 #endif #define BRIDGE_RTHASH_MASK (BRIDGE_RTHASH_SIZE - 1) /* * Default maximum number of addresses to cache. */ #ifndef BRIDGE_RTABLE_MAX #define BRIDGE_RTABLE_MAX 2000 #endif /* * Timeout (in seconds) for entries learned dynamically. */ #ifndef BRIDGE_RTABLE_TIMEOUT #define BRIDGE_RTABLE_TIMEOUT (20 * 60) /* same as ARP */ #endif /* * Number of seconds between walks of the route list. */ #ifndef BRIDGE_RTABLE_PRUNE_PERIOD #define BRIDGE_RTABLE_PRUNE_PERIOD (5 * 60) #endif /* * List of capabilities to possibly mask on the member interface. */ #define BRIDGE_IFCAPS_MASK (IFCAP_TOE|IFCAP_TSO|IFCAP_TXCSUM|\ IFCAP_TXCSUM_IPV6) /* * List of capabilities to strip */ #define BRIDGE_IFCAPS_STRIP IFCAP_LRO /* * Bridge locking * * The bridge relies heavily on the epoch(9) system to protect its data * structures. This means we can safely use CK_LISTs while in NET_EPOCH, but we * must ensure there is only one writer at a time. * * That is: for read accesses we only need to be in NET_EPOCH, but for write * accesses we must hold: * * - BRIDGE_RT_LOCK, for any change to bridge_rtnodes * - BRIDGE_LOCK, for any other change * * The BRIDGE_LOCK is a sleepable lock, because it is held across ioctl() * calls to bridge member interfaces and these ioctl()s can sleep. * The BRIDGE_RT_LOCK is a non-sleepable mutex, because it is sometimes * required while we're in NET_EPOCH and then we're not allowed to sleep. */ #define BRIDGE_LOCK_INIT(_sc) do { \ sx_init(&(_sc)->sc_sx, "if_bridge"); \ mtx_init(&(_sc)->sc_rt_mtx, "if_bridge rt", NULL, MTX_DEF); \ } while (0) #define BRIDGE_LOCK_DESTROY(_sc) do { \ sx_destroy(&(_sc)->sc_sx); \ mtx_destroy(&(_sc)->sc_rt_mtx); \ } while (0) #define BRIDGE_LOCK(_sc) sx_xlock(&(_sc)->sc_sx) #define BRIDGE_UNLOCK(_sc) sx_xunlock(&(_sc)->sc_sx) #define BRIDGE_LOCK_ASSERT(_sc) sx_assert(&(_sc)->sc_sx, SX_XLOCKED) #define BRIDGE_LOCK_OR_NET_EPOCH_ASSERT(_sc) \ MPASS(in_epoch(net_epoch_preempt) || sx_xlocked(&(_sc)->sc_sx)) #define BRIDGE_UNLOCK_ASSERT(_sc) sx_assert(&(_sc)->sc_sx, SX_UNLOCKED) #define BRIDGE_RT_LOCK(_sc) mtx_lock(&(_sc)->sc_rt_mtx) #define BRIDGE_RT_UNLOCK(_sc) mtx_unlock(&(_sc)->sc_rt_mtx) #define BRIDGE_RT_LOCK_ASSERT(_sc) mtx_assert(&(_sc)->sc_rt_mtx, MA_OWNED) #define BRIDGE_RT_LOCK_OR_NET_EPOCH_ASSERT(_sc) \ MPASS(in_epoch(net_epoch_preempt) || mtx_owned(&(_sc)->sc_rt_mtx)) /* * Bridge interface list entry. */ struct bridge_iflist { CK_LIST_ENTRY(bridge_iflist) bif_next; struct ifnet *bif_ifp; /* member if */ struct bstp_port bif_stp; /* STP state */ uint32_t bif_flags; /* member if flags */ int bif_savedcaps; /* saved capabilities */ uint32_t bif_addrmax; /* max # of addresses */ uint32_t bif_addrcnt; /* cur. # of addresses */ uint32_t bif_addrexceeded;/* # of address violations */ struct epoch_context bif_epoch_ctx; }; /* * Bridge route node. */ struct bridge_rtnode { CK_LIST_ENTRY(bridge_rtnode) brt_hash; /* hash table linkage */ CK_LIST_ENTRY(bridge_rtnode) brt_list; /* list linkage */ struct bridge_iflist *brt_dst; /* destination if */ unsigned long brt_expire; /* expiration time */ uint8_t brt_flags; /* address flags */ uint8_t brt_addr[ETHER_ADDR_LEN]; uint16_t brt_vlan; /* vlan id */ struct vnet *brt_vnet; struct epoch_context brt_epoch_ctx; }; #define brt_ifp brt_dst->bif_ifp /* * Software state for each bridge. */ struct bridge_softc { struct ifnet *sc_ifp; /* make this an interface */ LIST_ENTRY(bridge_softc) sc_list; struct sx sc_sx; struct mtx sc_rt_mtx; uint32_t sc_brtmax; /* max # of addresses */ uint32_t sc_brtcnt; /* cur. # of addresses */ uint32_t sc_brttimeout; /* rt timeout in seconds */ struct callout sc_brcallout; /* bridge callout */ CK_LIST_HEAD(, bridge_iflist) sc_iflist; /* member interface list */ CK_LIST_HEAD(, bridge_rtnode) *sc_rthash; /* our forwarding table */ CK_LIST_HEAD(, bridge_rtnode) sc_rtlist; /* list version of above */ uint32_t sc_rthash_key; /* key for hash */ CK_LIST_HEAD(, bridge_iflist) sc_spanlist; /* span ports list */ struct bstp_state sc_stp; /* STP state */ uint32_t sc_brtexceeded; /* # of cache drops */ struct ifnet *sc_ifaddr; /* member mac copied from */ struct ether_addr sc_defaddr; /* Default MAC address */ + if_input_fn_t sc_if_input; /* Saved copy of if_input */ struct epoch_context sc_epoch_ctx; }; VNET_DEFINE_STATIC(struct sx, bridge_list_sx); #define V_bridge_list_sx VNET(bridge_list_sx) static eventhandler_tag bridge_detach_cookie; int bridge_rtable_prune_period = BRIDGE_RTABLE_PRUNE_PERIOD; VNET_DEFINE_STATIC(uma_zone_t, bridge_rtnode_zone); #define V_bridge_rtnode_zone VNET(bridge_rtnode_zone) static int bridge_clone_create(struct if_clone *, char *, size_t, struct ifc_data *, struct ifnet **); static int bridge_clone_destroy(struct if_clone *, struct ifnet *, uint32_t); static int bridge_ioctl(struct ifnet *, u_long, caddr_t); static void bridge_mutecaps(struct bridge_softc *); static void bridge_set_ifcap(struct bridge_softc *, struct bridge_iflist *, int); static void bridge_ifdetach(void *arg __unused, struct ifnet *); static void bridge_init(void *); static void bridge_dummynet(struct mbuf *, struct ifnet *); static void bridge_stop(struct ifnet *, int); static int bridge_transmit(struct ifnet *, struct mbuf *); #ifdef ALTQ static void bridge_altq_start(if_t); static int bridge_altq_transmit(if_t, struct mbuf *); #endif static void bridge_qflush(struct ifnet *); static struct mbuf *bridge_input(struct ifnet *, struct mbuf *); +static void bridge_inject(struct ifnet *, struct mbuf *); static int bridge_output(struct ifnet *, struct mbuf *, struct sockaddr *, struct rtentry *); static int bridge_enqueue(struct bridge_softc *, struct ifnet *, struct mbuf *); static void bridge_rtdelete(struct bridge_softc *, struct ifnet *ifp, int); static void bridge_forward(struct bridge_softc *, struct bridge_iflist *, struct mbuf *m); static void bridge_timer(void *); static void bridge_broadcast(struct bridge_softc *, struct ifnet *, struct mbuf *, int); static void bridge_span(struct bridge_softc *, struct mbuf *); static int bridge_rtupdate(struct bridge_softc *, const uint8_t *, uint16_t, struct bridge_iflist *, int, uint8_t); static struct ifnet *bridge_rtlookup(struct bridge_softc *, const uint8_t *, uint16_t); static void bridge_rttrim(struct bridge_softc *); static void bridge_rtage(struct bridge_softc *); static void bridge_rtflush(struct bridge_softc *, int); static int bridge_rtdaddr(struct bridge_softc *, const uint8_t *, uint16_t); static void bridge_rtable_init(struct bridge_softc *); static void bridge_rtable_fini(struct bridge_softc *); static int bridge_rtnode_addr_cmp(const uint8_t *, const uint8_t *); static struct bridge_rtnode *bridge_rtnode_lookup(struct bridge_softc *, const uint8_t *, uint16_t); static int bridge_rtnode_insert(struct bridge_softc *, struct bridge_rtnode *); static void bridge_rtnode_destroy(struct bridge_softc *, struct bridge_rtnode *); static void bridge_rtable_expire(struct ifnet *, int); static void bridge_state_change(struct ifnet *, int); static struct bridge_iflist *bridge_lookup_member(struct bridge_softc *, const char *name); static struct bridge_iflist *bridge_lookup_member_if(struct bridge_softc *, struct ifnet *ifp); static void bridge_delete_member(struct bridge_softc *, struct bridge_iflist *, int); static void bridge_delete_span(struct bridge_softc *, struct bridge_iflist *); static int bridge_ioctl_add(struct bridge_softc *, void *); static int bridge_ioctl_del(struct bridge_softc *, void *); static int bridge_ioctl_gifflags(struct bridge_softc *, void *); static int bridge_ioctl_sifflags(struct bridge_softc *, void *); static int bridge_ioctl_scache(struct bridge_softc *, void *); static int bridge_ioctl_gcache(struct bridge_softc *, void *); static int bridge_ioctl_gifs(struct bridge_softc *, void *); static int bridge_ioctl_rts(struct bridge_softc *, void *); static int bridge_ioctl_saddr(struct bridge_softc *, void *); static int bridge_ioctl_sto(struct bridge_softc *, void *); static int bridge_ioctl_gto(struct bridge_softc *, void *); static int bridge_ioctl_daddr(struct bridge_softc *, void *); static int bridge_ioctl_flush(struct bridge_softc *, void *); static int bridge_ioctl_gpri(struct bridge_softc *, void *); static int bridge_ioctl_spri(struct bridge_softc *, void *); static int bridge_ioctl_ght(struct bridge_softc *, void *); static int bridge_ioctl_sht(struct bridge_softc *, void *); static int bridge_ioctl_gfd(struct bridge_softc *, void *); static int bridge_ioctl_sfd(struct bridge_softc *, void *); static int bridge_ioctl_gma(struct bridge_softc *, void *); static int bridge_ioctl_sma(struct bridge_softc *, void *); static int bridge_ioctl_sifprio(struct bridge_softc *, void *); static int bridge_ioctl_sifcost(struct bridge_softc *, void *); static int bridge_ioctl_sifmaxaddr(struct bridge_softc *, void *); static int bridge_ioctl_addspan(struct bridge_softc *, void *); static int bridge_ioctl_delspan(struct bridge_softc *, void *); static int bridge_ioctl_gbparam(struct bridge_softc *, void *); static int bridge_ioctl_grte(struct bridge_softc *, void *); static int bridge_ioctl_gifsstp(struct bridge_softc *, void *); static int bridge_ioctl_sproto(struct bridge_softc *, void *); static int bridge_ioctl_stxhc(struct bridge_softc *, void *); static int bridge_pfil(struct mbuf **, struct ifnet *, struct ifnet *, int); static int bridge_ip_checkbasic(struct mbuf **mp); #ifdef INET6 static int bridge_ip6_checkbasic(struct mbuf **mp); #endif /* INET6 */ static int bridge_fragment(struct ifnet *, struct mbuf **mp, struct ether_header *, int, struct llc *); static void bridge_linkstate(struct ifnet *ifp); static void bridge_linkcheck(struct bridge_softc *sc); /* The default bridge vlan is 1 (IEEE 802.1Q-2003 Table 9-2) */ #define VLANTAGOF(_m) \ (_m->m_flags & M_VLANTAG) ? EVL_VLANOFTAG(_m->m_pkthdr.ether_vtag) : 1 static struct bstp_cb_ops bridge_ops = { .bcb_state = bridge_state_change, .bcb_rtage = bridge_rtable_expire }; SYSCTL_DECL(_net_link); static SYSCTL_NODE(_net_link, IFT_BRIDGE, bridge, CTLFLAG_RW | CTLFLAG_MPSAFE, 0, "Bridge"); /* only pass IP[46] packets when pfil is enabled */ VNET_DEFINE_STATIC(int, pfil_onlyip) = 1; #define V_pfil_onlyip VNET(pfil_onlyip) SYSCTL_INT(_net_link_bridge, OID_AUTO, pfil_onlyip, CTLFLAG_RWTUN | CTLFLAG_VNET, &VNET_NAME(pfil_onlyip), 0, "Only pass IP packets when pfil is enabled"); /* run pfil hooks on the bridge interface */ VNET_DEFINE_STATIC(int, pfil_bridge) = 0; #define V_pfil_bridge VNET(pfil_bridge) SYSCTL_INT(_net_link_bridge, OID_AUTO, pfil_bridge, CTLFLAG_RWTUN | CTLFLAG_VNET, &VNET_NAME(pfil_bridge), 0, "Packet filter on the bridge interface"); /* layer2 filter with ipfw */ VNET_DEFINE_STATIC(int, pfil_ipfw); #define V_pfil_ipfw VNET(pfil_ipfw) /* layer2 ARP filter with ipfw */ VNET_DEFINE_STATIC(int, pfil_ipfw_arp); #define V_pfil_ipfw_arp VNET(pfil_ipfw_arp) SYSCTL_INT(_net_link_bridge, OID_AUTO, ipfw_arp, CTLFLAG_RWTUN | CTLFLAG_VNET, &VNET_NAME(pfil_ipfw_arp), 0, "Filter ARP packets through IPFW layer2"); /* run pfil hooks on the member interface */ VNET_DEFINE_STATIC(int, pfil_member) = 0; #define V_pfil_member VNET(pfil_member) SYSCTL_INT(_net_link_bridge, OID_AUTO, pfil_member, CTLFLAG_RWTUN | CTLFLAG_VNET, &VNET_NAME(pfil_member), 0, "Packet filter on the member interface"); /* run pfil hooks on the physical interface for locally destined packets */ VNET_DEFINE_STATIC(int, pfil_local_phys); #define V_pfil_local_phys VNET(pfil_local_phys) SYSCTL_INT(_net_link_bridge, OID_AUTO, pfil_local_phys, CTLFLAG_RWTUN | CTLFLAG_VNET, &VNET_NAME(pfil_local_phys), 0, "Packet filter on the physical interface for locally destined packets"); /* log STP state changes */ VNET_DEFINE_STATIC(int, log_stp); #define V_log_stp VNET(log_stp) SYSCTL_INT(_net_link_bridge, OID_AUTO, log_stp, CTLFLAG_RWTUN | CTLFLAG_VNET, &VNET_NAME(log_stp), 0, "Log STP state changes"); /* share MAC with first bridge member */ VNET_DEFINE_STATIC(int, bridge_inherit_mac); #define V_bridge_inherit_mac VNET(bridge_inherit_mac) SYSCTL_INT(_net_link_bridge, OID_AUTO, inherit_mac, CTLFLAG_RWTUN | CTLFLAG_VNET, &VNET_NAME(bridge_inherit_mac), 0, "Inherit MAC address from the first bridge member"); VNET_DEFINE_STATIC(int, allow_llz_overlap) = 0; #define V_allow_llz_overlap VNET(allow_llz_overlap) SYSCTL_INT(_net_link_bridge, OID_AUTO, allow_llz_overlap, CTLFLAG_RW | CTLFLAG_VNET, &VNET_NAME(allow_llz_overlap), 0, "Allow overlap of link-local scope " "zones of a bridge interface and the member interfaces"); /* log MAC address port flapping */ VNET_DEFINE_STATIC(bool, log_mac_flap) = true; #define V_log_mac_flap VNET(log_mac_flap) SYSCTL_BOOL(_net_link_bridge, OID_AUTO, log_mac_flap, CTLFLAG_RW | CTLFLAG_VNET, &VNET_NAME(log_mac_flap), true, "Log MAC address port flapping"); VNET_DEFINE_STATIC(int, log_interval) = 5; VNET_DEFINE_STATIC(int, log_count) = 0; VNET_DEFINE_STATIC(struct timeval, log_last) = { 0 }; #define V_log_interval VNET(log_interval) #define V_log_count VNET(log_count) #define V_log_last VNET(log_last) struct bridge_control { int (*bc_func)(struct bridge_softc *, void *); int bc_argsize; int bc_flags; }; #define BC_F_COPYIN 0x01 /* copy arguments in */ #define BC_F_COPYOUT 0x02 /* copy arguments out */ #define BC_F_SUSER 0x04 /* do super-user check */ static const struct bridge_control bridge_control_table[] = { { bridge_ioctl_add, sizeof(struct ifbreq), BC_F_COPYIN|BC_F_SUSER }, { bridge_ioctl_del, sizeof(struct ifbreq), BC_F_COPYIN|BC_F_SUSER }, { bridge_ioctl_gifflags, sizeof(struct ifbreq), BC_F_COPYIN|BC_F_COPYOUT }, { bridge_ioctl_sifflags, sizeof(struct ifbreq), BC_F_COPYIN|BC_F_SUSER }, { bridge_ioctl_scache, sizeof(struct ifbrparam), BC_F_COPYIN|BC_F_SUSER }, { bridge_ioctl_gcache, sizeof(struct ifbrparam), BC_F_COPYOUT }, { bridge_ioctl_gifs, sizeof(struct ifbifconf), BC_F_COPYIN|BC_F_COPYOUT }, { bridge_ioctl_rts, sizeof(struct ifbaconf), BC_F_COPYIN|BC_F_COPYOUT }, { bridge_ioctl_saddr, sizeof(struct ifbareq), BC_F_COPYIN|BC_F_SUSER }, { bridge_ioctl_sto, sizeof(struct ifbrparam), BC_F_COPYIN|BC_F_SUSER }, { bridge_ioctl_gto, sizeof(struct ifbrparam), BC_F_COPYOUT }, { bridge_ioctl_daddr, sizeof(struct ifbareq), BC_F_COPYIN|BC_F_SUSER }, { bridge_ioctl_flush, sizeof(struct ifbreq), BC_F_COPYIN|BC_F_SUSER }, { bridge_ioctl_gpri, sizeof(struct ifbrparam), BC_F_COPYOUT }, { bridge_ioctl_spri, sizeof(struct ifbrparam), BC_F_COPYIN|BC_F_SUSER }, { bridge_ioctl_ght, sizeof(struct ifbrparam), BC_F_COPYOUT }, { bridge_ioctl_sht, sizeof(struct ifbrparam), BC_F_COPYIN|BC_F_SUSER }, { bridge_ioctl_gfd, sizeof(struct ifbrparam), BC_F_COPYOUT }, { bridge_ioctl_sfd, sizeof(struct ifbrparam), BC_F_COPYIN|BC_F_SUSER }, { bridge_ioctl_gma, sizeof(struct ifbrparam), BC_F_COPYOUT }, { bridge_ioctl_sma, sizeof(struct ifbrparam), BC_F_COPYIN|BC_F_SUSER }, { bridge_ioctl_sifprio, sizeof(struct ifbreq), BC_F_COPYIN|BC_F_SUSER }, { bridge_ioctl_sifcost, sizeof(struct ifbreq), BC_F_COPYIN|BC_F_SUSER }, { bridge_ioctl_addspan, sizeof(struct ifbreq), BC_F_COPYIN|BC_F_SUSER }, { bridge_ioctl_delspan, sizeof(struct ifbreq), BC_F_COPYIN|BC_F_SUSER }, { bridge_ioctl_gbparam, sizeof(struct ifbropreq), BC_F_COPYOUT }, { bridge_ioctl_grte, sizeof(struct ifbrparam), BC_F_COPYOUT }, { bridge_ioctl_gifsstp, sizeof(struct ifbpstpconf), BC_F_COPYIN|BC_F_COPYOUT }, { bridge_ioctl_sproto, sizeof(struct ifbrparam), BC_F_COPYIN|BC_F_SUSER }, { bridge_ioctl_stxhc, sizeof(struct ifbrparam), BC_F_COPYIN|BC_F_SUSER }, { bridge_ioctl_sifmaxaddr, sizeof(struct ifbreq), BC_F_COPYIN|BC_F_SUSER }, }; static const int bridge_control_table_size = nitems(bridge_control_table); VNET_DEFINE_STATIC(LIST_HEAD(, bridge_softc), bridge_list); #define V_bridge_list VNET(bridge_list) #define BRIDGE_LIST_LOCK_INIT(x) sx_init(&V_bridge_list_sx, \ "if_bridge list") #define BRIDGE_LIST_LOCK_DESTROY(x) sx_destroy(&V_bridge_list_sx) #define BRIDGE_LIST_LOCK(x) sx_xlock(&V_bridge_list_sx) #define BRIDGE_LIST_UNLOCK(x) sx_xunlock(&V_bridge_list_sx) VNET_DEFINE_STATIC(struct if_clone *, bridge_cloner); #define V_bridge_cloner VNET(bridge_cloner) static const char bridge_name[] = "bridge"; static void vnet_bridge_init(const void *unused __unused) { V_bridge_rtnode_zone = uma_zcreate("bridge_rtnode", sizeof(struct bridge_rtnode), NULL, NULL, NULL, NULL, UMA_ALIGN_PTR, 0); BRIDGE_LIST_LOCK_INIT(); LIST_INIT(&V_bridge_list); struct if_clone_addreq req = { .create_f = bridge_clone_create, .destroy_f = bridge_clone_destroy, .flags = IFC_F_AUTOUNIT, }; V_bridge_cloner = ifc_attach_cloner(bridge_name, &req); } VNET_SYSINIT(vnet_bridge_init, SI_SUB_PROTO_IFATTACHDOMAIN, SI_ORDER_ANY, vnet_bridge_init, NULL); static void vnet_bridge_uninit(const void *unused __unused) { ifc_detach_cloner(V_bridge_cloner); V_bridge_cloner = NULL; BRIDGE_LIST_LOCK_DESTROY(); /* Callbacks may use the UMA zone. */ NET_EPOCH_DRAIN_CALLBACKS(); uma_zdestroy(V_bridge_rtnode_zone); } VNET_SYSUNINIT(vnet_bridge_uninit, SI_SUB_PSEUDO, SI_ORDER_ANY, vnet_bridge_uninit, NULL); static int bridge_modevent(module_t mod, int type, void *data) { switch (type) { case MOD_LOAD: bridge_dn_p = bridge_dummynet; bridge_detach_cookie = EVENTHANDLER_REGISTER( ifnet_departure_event, bridge_ifdetach, NULL, EVENTHANDLER_PRI_ANY); break; case MOD_UNLOAD: EVENTHANDLER_DEREGISTER(ifnet_departure_event, bridge_detach_cookie); bridge_dn_p = NULL; break; default: return (EOPNOTSUPP); } return (0); } static moduledata_t bridge_mod = { "if_bridge", bridge_modevent, 0 }; DECLARE_MODULE(if_bridge, bridge_mod, SI_SUB_PSEUDO, SI_ORDER_ANY); MODULE_VERSION(if_bridge, 1); MODULE_DEPEND(if_bridge, bridgestp, 1, 1, 1); /* * handler for net.link.bridge.ipfw */ static int sysctl_pfil_ipfw(SYSCTL_HANDLER_ARGS) { int enable = V_pfil_ipfw; int error; error = sysctl_handle_int(oidp, &enable, 0, req); enable &= 1; if (enable != V_pfil_ipfw) { V_pfil_ipfw = enable; /* * Disable pfil so that ipfw doesnt run twice, if the user * really wants both then they can re-enable pfil_bridge and/or * pfil_member. Also allow non-ip packets as ipfw can filter by * layer2 type. */ if (V_pfil_ipfw) { V_pfil_onlyip = 0; V_pfil_bridge = 0; V_pfil_member = 0; } } return (error); } SYSCTL_PROC(_net_link_bridge, OID_AUTO, ipfw, CTLTYPE_INT | CTLFLAG_RW | CTLFLAG_VNET | CTLFLAG_NEEDGIANT, &VNET_NAME(pfil_ipfw), 0, &sysctl_pfil_ipfw, "I", "Layer2 filter with IPFW"); #ifdef VIMAGE static void bridge_reassign(struct ifnet *ifp, struct vnet *newvnet, char *arg) { struct bridge_softc *sc = ifp->if_softc; struct bridge_iflist *bif; BRIDGE_LOCK(sc); while ((bif = CK_LIST_FIRST(&sc->sc_iflist)) != NULL) bridge_delete_member(sc, bif, 0); while ((bif = CK_LIST_FIRST(&sc->sc_spanlist)) != NULL) { bridge_delete_span(sc, bif); } BRIDGE_UNLOCK(sc); ether_reassign(ifp, newvnet, arg); } #endif /* * bridge_clone_create: * * Create a new bridge instance. */ static int bridge_clone_create(struct if_clone *ifc, char *name, size_t len, struct ifc_data *ifd, struct ifnet **ifpp) { struct bridge_softc *sc; struct ifnet *ifp; sc = malloc(sizeof(*sc), M_DEVBUF, M_WAITOK|M_ZERO); ifp = sc->sc_ifp = if_alloc(IFT_ETHER); if (ifp == NULL) { free(sc, M_DEVBUF); return (ENOSPC); } BRIDGE_LOCK_INIT(sc); sc->sc_brtmax = BRIDGE_RTABLE_MAX; sc->sc_brttimeout = BRIDGE_RTABLE_TIMEOUT; /* Initialize our routing table. */ bridge_rtable_init(sc); callout_init_mtx(&sc->sc_brcallout, &sc->sc_rt_mtx, 0); CK_LIST_INIT(&sc->sc_iflist); CK_LIST_INIT(&sc->sc_spanlist); ifp->if_softc = sc; if_initname(ifp, bridge_name, ifd->unit); ifp->if_flags = IFF_BROADCAST | IFF_SIMPLEX | IFF_MULTICAST; ifp->if_ioctl = bridge_ioctl; #ifdef ALTQ ifp->if_start = bridge_altq_start; ifp->if_transmit = bridge_altq_transmit; IFQ_SET_MAXLEN(&ifp->if_snd, ifqmaxlen); ifp->if_snd.ifq_drv_maxlen = 0; IFQ_SET_READY(&ifp->if_snd); #else ifp->if_transmit = bridge_transmit; #endif ifp->if_qflush = bridge_qflush; ifp->if_init = bridge_init; ifp->if_type = IFT_BRIDGE; ether_gen_addr(ifp, &sc->sc_defaddr); bstp_attach(&sc->sc_stp, &bridge_ops); ether_ifattach(ifp, sc->sc_defaddr.octet); /* Now undo some of the damage... */ ifp->if_baudrate = 0; ifp->if_type = IFT_BRIDGE; #ifdef VIMAGE ifp->if_reassign = bridge_reassign; #endif + sc->sc_if_input = ifp->if_input; /* ether_input */ + ifp->if_input = bridge_inject; + + /* + * Allow BRIDGE_INPUT() to pass in packets originating from the bridge + * itself via bridge_inject(). This is required for netmap but + * otherwise has no effect. + */ + ifp->if_bridge_input = bridge_input; BRIDGE_LIST_LOCK(); LIST_INSERT_HEAD(&V_bridge_list, sc, sc_list); BRIDGE_LIST_UNLOCK(); *ifpp = ifp; return (0); } static void bridge_clone_destroy_cb(struct epoch_context *ctx) { struct bridge_softc *sc; sc = __containerof(ctx, struct bridge_softc, sc_epoch_ctx); BRIDGE_LOCK_DESTROY(sc); free(sc, M_DEVBUF); } /* * bridge_clone_destroy: * * Destroy a bridge instance. */ static int bridge_clone_destroy(struct if_clone *ifc, struct ifnet *ifp, uint32_t flags) { struct bridge_softc *sc = ifp->if_softc; struct bridge_iflist *bif; struct epoch_tracker et; BRIDGE_LOCK(sc); bridge_stop(ifp, 1); ifp->if_flags &= ~IFF_UP; while ((bif = CK_LIST_FIRST(&sc->sc_iflist)) != NULL) bridge_delete_member(sc, bif, 0); while ((bif = CK_LIST_FIRST(&sc->sc_spanlist)) != NULL) { bridge_delete_span(sc, bif); } /* Tear down the routing table. */ bridge_rtable_fini(sc); BRIDGE_UNLOCK(sc); NET_EPOCH_ENTER(et); callout_drain(&sc->sc_brcallout); BRIDGE_LIST_LOCK(); LIST_REMOVE(sc, sc_list); BRIDGE_LIST_UNLOCK(); bstp_detach(&sc->sc_stp); #ifdef ALTQ IFQ_PURGE(&ifp->if_snd); #endif NET_EPOCH_EXIT(et); ether_ifdetach(ifp); if_free(ifp); NET_EPOCH_CALL(bridge_clone_destroy_cb, &sc->sc_epoch_ctx); return (0); } /* * bridge_ioctl: * * Handle a control request from the operator. */ static int bridge_ioctl(struct ifnet *ifp, u_long cmd, caddr_t data) { struct bridge_softc *sc = ifp->if_softc; struct ifreq *ifr = (struct ifreq *)data; struct bridge_iflist *bif; struct thread *td = curthread; union { struct ifbreq ifbreq; struct ifbifconf ifbifconf; struct ifbareq ifbareq; struct ifbaconf ifbaconf; struct ifbrparam ifbrparam; struct ifbropreq ifbropreq; } args; struct ifdrv *ifd = (struct ifdrv *) data; const struct bridge_control *bc; int error = 0, oldmtu; BRIDGE_LOCK(sc); switch (cmd) { case SIOCADDMULTI: case SIOCDELMULTI: break; case SIOCGDRVSPEC: case SIOCSDRVSPEC: if (ifd->ifd_cmd >= bridge_control_table_size) { error = EINVAL; break; } bc = &bridge_control_table[ifd->ifd_cmd]; if (cmd == SIOCGDRVSPEC && (bc->bc_flags & BC_F_COPYOUT) == 0) { error = EINVAL; break; } else if (cmd == SIOCSDRVSPEC && (bc->bc_flags & BC_F_COPYOUT) != 0) { error = EINVAL; break; } if (bc->bc_flags & BC_F_SUSER) { error = priv_check(td, PRIV_NET_BRIDGE); if (error) break; } if (ifd->ifd_len != bc->bc_argsize || ifd->ifd_len > sizeof(args)) { error = EINVAL; break; } bzero(&args, sizeof(args)); if (bc->bc_flags & BC_F_COPYIN) { error = copyin(ifd->ifd_data, &args, ifd->ifd_len); if (error) break; } oldmtu = ifp->if_mtu; error = (*bc->bc_func)(sc, &args); if (error) break; /* * Bridge MTU may change during addition of the first port. * If it did, do network layer specific procedure. */ if (ifp->if_mtu != oldmtu) if_notifymtu(ifp); if (bc->bc_flags & BC_F_COPYOUT) error = copyout(&args, ifd->ifd_data, ifd->ifd_len); break; case SIOCSIFFLAGS: if (!(ifp->if_flags & IFF_UP) && (ifp->if_drv_flags & IFF_DRV_RUNNING)) { /* * If interface is marked down and it is running, * then stop and disable it. */ bridge_stop(ifp, 1); } else if ((ifp->if_flags & IFF_UP) && !(ifp->if_drv_flags & IFF_DRV_RUNNING)) { /* * If interface is marked up and it is stopped, then * start it. */ BRIDGE_UNLOCK(sc); (*ifp->if_init)(sc); BRIDGE_LOCK(sc); } break; case SIOCSIFMTU: oldmtu = sc->sc_ifp->if_mtu; if (ifr->ifr_mtu < 576) { error = EINVAL; break; } if (CK_LIST_EMPTY(&sc->sc_iflist)) { sc->sc_ifp->if_mtu = ifr->ifr_mtu; break; } CK_LIST_FOREACH(bif, &sc->sc_iflist, bif_next) { error = (*bif->bif_ifp->if_ioctl)(bif->bif_ifp, SIOCSIFMTU, (caddr_t)ifr); if (error != 0) { log(LOG_NOTICE, "%s: invalid MTU: %u for" " member %s\n", sc->sc_ifp->if_xname, ifr->ifr_mtu, bif->bif_ifp->if_xname); error = EINVAL; break; } } if (error) { /* Restore the previous MTU on all member interfaces. */ ifr->ifr_mtu = oldmtu; CK_LIST_FOREACH(bif, &sc->sc_iflist, bif_next) { (*bif->bif_ifp->if_ioctl)(bif->bif_ifp, SIOCSIFMTU, (caddr_t)ifr); } } else { sc->sc_ifp->if_mtu = ifr->ifr_mtu; } break; default: /* * drop the lock as ether_ioctl() will call bridge_start() and * cause the lock to be recursed. */ BRIDGE_UNLOCK(sc); error = ether_ioctl(ifp, cmd, data); BRIDGE_LOCK(sc); break; } BRIDGE_UNLOCK(sc); return (error); } /* * bridge_mutecaps: * * Clear or restore unwanted capabilities on the member interface */ static void bridge_mutecaps(struct bridge_softc *sc) { struct bridge_iflist *bif; int enabled, mask; BRIDGE_LOCK_ASSERT(sc); /* Initial bitmask of capabilities to test */ mask = BRIDGE_IFCAPS_MASK; CK_LIST_FOREACH(bif, &sc->sc_iflist, bif_next) { /* Every member must support it or its disabled */ mask &= bif->bif_savedcaps; } CK_LIST_FOREACH(bif, &sc->sc_iflist, bif_next) { enabled = bif->bif_ifp->if_capenable; enabled &= ~BRIDGE_IFCAPS_STRIP; /* strip off mask bits and enable them again if allowed */ enabled &= ~BRIDGE_IFCAPS_MASK; enabled |= mask; bridge_set_ifcap(sc, bif, enabled); } } static void bridge_set_ifcap(struct bridge_softc *sc, struct bridge_iflist *bif, int set) { struct ifnet *ifp = bif->bif_ifp; struct ifreq ifr; int error, mask, stuck; bzero(&ifr, sizeof(ifr)); ifr.ifr_reqcap = set; if (ifp->if_capenable != set) { error = (*ifp->if_ioctl)(ifp, SIOCSIFCAP, (caddr_t)&ifr); if (error) if_printf(sc->sc_ifp, "error setting capabilities on %s: %d\n", ifp->if_xname, error); mask = BRIDGE_IFCAPS_MASK | BRIDGE_IFCAPS_STRIP; stuck = ifp->if_capenable & mask & ~set; if (stuck != 0) if_printf(sc->sc_ifp, "can't disable some capabilities on %s: 0x%x\n", ifp->if_xname, stuck); } } /* * bridge_lookup_member: * * Lookup a bridge member interface. */ static struct bridge_iflist * bridge_lookup_member(struct bridge_softc *sc, const char *name) { struct bridge_iflist *bif; struct ifnet *ifp; BRIDGE_LOCK_OR_NET_EPOCH_ASSERT(sc); CK_LIST_FOREACH(bif, &sc->sc_iflist, bif_next) { ifp = bif->bif_ifp; if (strcmp(ifp->if_xname, name) == 0) return (bif); } return (NULL); } /* * bridge_lookup_member_if: * * Lookup a bridge member interface by ifnet*. */ static struct bridge_iflist * bridge_lookup_member_if(struct bridge_softc *sc, struct ifnet *member_ifp) { struct bridge_iflist *bif; BRIDGE_LOCK_OR_NET_EPOCH_ASSERT(sc); CK_LIST_FOREACH(bif, &sc->sc_iflist, bif_next) { if (bif->bif_ifp == member_ifp) return (bif); } return (NULL); } static void bridge_delete_member_cb(struct epoch_context *ctx) { struct bridge_iflist *bif; bif = __containerof(ctx, struct bridge_iflist, bif_epoch_ctx); free(bif, M_DEVBUF); } /* * bridge_delete_member: * * Delete the specified member interface. */ static void bridge_delete_member(struct bridge_softc *sc, struct bridge_iflist *bif, int gone) { struct ifnet *ifs = bif->bif_ifp; struct ifnet *fif = NULL; struct bridge_iflist *bifl; BRIDGE_LOCK_ASSERT(sc); if (bif->bif_flags & IFBIF_STP) bstp_disable(&bif->bif_stp); ifs->if_bridge = NULL; CK_LIST_REMOVE(bif, bif_next); /* * If removing the interface that gave the bridge its mac address, set * the mac address of the bridge to the address of the next member, or * to its default address if no members are left. */ if (V_bridge_inherit_mac && sc->sc_ifaddr == ifs) { if (CK_LIST_EMPTY(&sc->sc_iflist)) { bcopy(&sc->sc_defaddr, IF_LLADDR(sc->sc_ifp), ETHER_ADDR_LEN); sc->sc_ifaddr = NULL; } else { bifl = CK_LIST_FIRST(&sc->sc_iflist); fif = bifl->bif_ifp; bcopy(IF_LLADDR(fif), IF_LLADDR(sc->sc_ifp), ETHER_ADDR_LEN); sc->sc_ifaddr = fif; } EVENTHANDLER_INVOKE(iflladdr_event, sc->sc_ifp); } bridge_linkcheck(sc); bridge_mutecaps(sc); /* recalcuate now this interface is removed */ BRIDGE_RT_LOCK(sc); bridge_rtdelete(sc, ifs, IFBF_FLUSHALL); BRIDGE_RT_UNLOCK(sc); KASSERT(bif->bif_addrcnt == 0, ("%s: %d bridge routes referenced", __func__, bif->bif_addrcnt)); ifs->if_bridge_output = NULL; ifs->if_bridge_input = NULL; ifs->if_bridge_linkstate = NULL; if (!gone) { switch (ifs->if_type) { case IFT_ETHER: case IFT_L2VLAN: /* * Take the interface out of promiscuous mode, but only * if it was promiscuous in the first place. It might * not be if we're in the bridge_ioctl_add() error path. */ if (ifs->if_flags & IFF_PROMISC) (void) ifpromisc(ifs, 0); break; case IFT_GIF: break; default: #ifdef DIAGNOSTIC panic("bridge_delete_member: impossible"); #endif break; } /* reneable any interface capabilities */ bridge_set_ifcap(sc, bif, bif->bif_savedcaps); } bstp_destroy(&bif->bif_stp); /* prepare to free */ NET_EPOCH_CALL(bridge_delete_member_cb, &bif->bif_epoch_ctx); } /* * bridge_delete_span: * * Delete the specified span interface. */ static void bridge_delete_span(struct bridge_softc *sc, struct bridge_iflist *bif) { BRIDGE_LOCK_ASSERT(sc); KASSERT(bif->bif_ifp->if_bridge == NULL, ("%s: not a span interface", __func__)); CK_LIST_REMOVE(bif, bif_next); NET_EPOCH_CALL(bridge_delete_member_cb, &bif->bif_epoch_ctx); } static int bridge_ioctl_add(struct bridge_softc *sc, void *arg) { struct ifbreq *req = arg; struct bridge_iflist *bif = NULL; struct ifnet *ifs; int error = 0; ifs = ifunit(req->ifbr_ifsname); if (ifs == NULL) return (ENOENT); if (ifs->if_ioctl == NULL) /* must be supported */ return (EINVAL); /* If it's in the span list, it can't be a member. */ CK_LIST_FOREACH(bif, &sc->sc_spanlist, bif_next) if (ifs == bif->bif_ifp) return (EBUSY); if (ifs->if_bridge == sc) return (EEXIST); if (ifs->if_bridge != NULL) return (EBUSY); switch (ifs->if_type) { case IFT_ETHER: case IFT_L2VLAN: case IFT_GIF: /* permitted interface types */ break; default: return (EINVAL); } #ifdef INET6 /* * Two valid inet6 addresses with link-local scope must not be * on the parent interface and the member interfaces at the * same time. This restriction is needed to prevent violation * of link-local scope zone. Attempts to add a member * interface which has inet6 addresses when the parent has * inet6 triggers removal of all inet6 addresses on the member * interface. */ /* Check if the parent interface has a link-local scope addr. */ if (V_allow_llz_overlap == 0 && in6ifa_llaonifp(sc->sc_ifp) != NULL) { /* * If any, remove all inet6 addresses from the member * interfaces. */ CK_LIST_FOREACH(bif, &sc->sc_iflist, bif_next) { if (in6ifa_llaonifp(bif->bif_ifp)) { in6_ifdetach(bif->bif_ifp); if_printf(sc->sc_ifp, "IPv6 addresses on %s have been removed " "before adding it as a member to prevent " "IPv6 address scope violation.\n", bif->bif_ifp->if_xname); } } if (in6ifa_llaonifp(ifs)) { in6_ifdetach(ifs); if_printf(sc->sc_ifp, "IPv6 addresses on %s have been removed " "before adding it as a member to prevent " "IPv6 address scope violation.\n", ifs->if_xname); } } #endif /* Allow the first Ethernet member to define the MTU */ if (CK_LIST_EMPTY(&sc->sc_iflist)) sc->sc_ifp->if_mtu = ifs->if_mtu; else if (sc->sc_ifp->if_mtu != ifs->if_mtu) { struct ifreq ifr; snprintf(ifr.ifr_name, sizeof(ifr.ifr_name), "%s", ifs->if_xname); ifr.ifr_mtu = sc->sc_ifp->if_mtu; error = (*ifs->if_ioctl)(ifs, SIOCSIFMTU, (caddr_t)&ifr); if (error != 0) { log(LOG_NOTICE, "%s: invalid MTU: %u for" " new member %s\n", sc->sc_ifp->if_xname, ifr.ifr_mtu, ifs->if_xname); return (EINVAL); } } bif = malloc(sizeof(*bif), M_DEVBUF, M_NOWAIT|M_ZERO); if (bif == NULL) return (ENOMEM); bif->bif_ifp = ifs; bif->bif_flags = IFBIF_LEARNING | IFBIF_DISCOVER; bif->bif_savedcaps = ifs->if_capenable; /* * Assign the interface's MAC address to the bridge if it's the first * member and the MAC address of the bridge has not been changed from * the default randomly generated one. */ if (V_bridge_inherit_mac && CK_LIST_EMPTY(&sc->sc_iflist) && !memcmp(IF_LLADDR(sc->sc_ifp), sc->sc_defaddr.octet, ETHER_ADDR_LEN)) { bcopy(IF_LLADDR(ifs), IF_LLADDR(sc->sc_ifp), ETHER_ADDR_LEN); sc->sc_ifaddr = ifs; EVENTHANDLER_INVOKE(iflladdr_event, sc->sc_ifp); } ifs->if_bridge = sc; ifs->if_bridge_output = bridge_output; ifs->if_bridge_input = bridge_input; ifs->if_bridge_linkstate = bridge_linkstate; bstp_create(&sc->sc_stp, &bif->bif_stp, bif->bif_ifp); /* * XXX: XLOCK HERE!?! * * NOTE: insert_***HEAD*** should be safe for the traversals. */ CK_LIST_INSERT_HEAD(&sc->sc_iflist, bif, bif_next); /* Set interface capabilities to the intersection set of all members */ bridge_mutecaps(sc); bridge_linkcheck(sc); /* Place the interface into promiscuous mode */ switch (ifs->if_type) { case IFT_ETHER: case IFT_L2VLAN: error = ifpromisc(ifs, 1); break; } if (error) bridge_delete_member(sc, bif, 0); return (error); } static int bridge_ioctl_del(struct bridge_softc *sc, void *arg) { struct ifbreq *req = arg; struct bridge_iflist *bif; bif = bridge_lookup_member(sc, req->ifbr_ifsname); if (bif == NULL) return (ENOENT); bridge_delete_member(sc, bif, 0); return (0); } static int bridge_ioctl_gifflags(struct bridge_softc *sc, void *arg) { struct ifbreq *req = arg; struct bridge_iflist *bif; struct bstp_port *bp; bif = bridge_lookup_member(sc, req->ifbr_ifsname); if (bif == NULL) return (ENOENT); bp = &bif->bif_stp; req->ifbr_ifsflags = bif->bif_flags; req->ifbr_state = bp->bp_state; req->ifbr_priority = bp->bp_priority; req->ifbr_path_cost = bp->bp_path_cost; req->ifbr_portno = bif->bif_ifp->if_index & 0xfff; req->ifbr_proto = bp->bp_protover; req->ifbr_role = bp->bp_role; req->ifbr_stpflags = bp->bp_flags; req->ifbr_addrcnt = bif->bif_addrcnt; req->ifbr_addrmax = bif->bif_addrmax; req->ifbr_addrexceeded = bif->bif_addrexceeded; /* Copy STP state options as flags */ if (bp->bp_operedge) req->ifbr_ifsflags |= IFBIF_BSTP_EDGE; if (bp->bp_flags & BSTP_PORT_AUTOEDGE) req->ifbr_ifsflags |= IFBIF_BSTP_AUTOEDGE; if (bp->bp_ptp_link) req->ifbr_ifsflags |= IFBIF_BSTP_PTP; if (bp->bp_flags & BSTP_PORT_AUTOPTP) req->ifbr_ifsflags |= IFBIF_BSTP_AUTOPTP; if (bp->bp_flags & BSTP_PORT_ADMEDGE) req->ifbr_ifsflags |= IFBIF_BSTP_ADMEDGE; if (bp->bp_flags & BSTP_PORT_ADMCOST) req->ifbr_ifsflags |= IFBIF_BSTP_ADMCOST; return (0); } static int bridge_ioctl_sifflags(struct bridge_softc *sc, void *arg) { struct epoch_tracker et; struct ifbreq *req = arg; struct bridge_iflist *bif; struct bstp_port *bp; int error; bif = bridge_lookup_member(sc, req->ifbr_ifsname); if (bif == NULL) return (ENOENT); bp = &bif->bif_stp; if (req->ifbr_ifsflags & IFBIF_SPAN) /* SPAN is readonly */ return (EINVAL); NET_EPOCH_ENTER(et); if (req->ifbr_ifsflags & IFBIF_STP) { if ((bif->bif_flags & IFBIF_STP) == 0) { error = bstp_enable(&bif->bif_stp); if (error) { NET_EPOCH_EXIT(et); return (error); } } } else { if ((bif->bif_flags & IFBIF_STP) != 0) bstp_disable(&bif->bif_stp); } /* Pass on STP flags */ bstp_set_edge(bp, req->ifbr_ifsflags & IFBIF_BSTP_EDGE ? 1 : 0); bstp_set_autoedge(bp, req->ifbr_ifsflags & IFBIF_BSTP_AUTOEDGE ? 1 : 0); bstp_set_ptp(bp, req->ifbr_ifsflags & IFBIF_BSTP_PTP ? 1 : 0); bstp_set_autoptp(bp, req->ifbr_ifsflags & IFBIF_BSTP_AUTOPTP ? 1 : 0); /* Save the bits relating to the bridge */ bif->bif_flags = req->ifbr_ifsflags & IFBIFMASK; NET_EPOCH_EXIT(et); return (0); } static int bridge_ioctl_scache(struct bridge_softc *sc, void *arg) { struct ifbrparam *param = arg; sc->sc_brtmax = param->ifbrp_csize; bridge_rttrim(sc); return (0); } static int bridge_ioctl_gcache(struct bridge_softc *sc, void *arg) { struct ifbrparam *param = arg; param->ifbrp_csize = sc->sc_brtmax; return (0); } static int bridge_ioctl_gifs(struct bridge_softc *sc, void *arg) { struct ifbifconf *bifc = arg; struct bridge_iflist *bif; struct ifbreq breq; char *buf, *outbuf; int count, buflen, len, error = 0; count = 0; CK_LIST_FOREACH(bif, &sc->sc_iflist, bif_next) count++; CK_LIST_FOREACH(bif, &sc->sc_spanlist, bif_next) count++; buflen = sizeof(breq) * count; if (bifc->ifbic_len == 0) { bifc->ifbic_len = buflen; return (0); } outbuf = malloc(buflen, M_TEMP, M_NOWAIT | M_ZERO); if (outbuf == NULL) return (ENOMEM); count = 0; buf = outbuf; len = min(bifc->ifbic_len, buflen); bzero(&breq, sizeof(breq)); CK_LIST_FOREACH(bif, &sc->sc_iflist, bif_next) { if (len < sizeof(breq)) break; strlcpy(breq.ifbr_ifsname, bif->bif_ifp->if_xname, sizeof(breq.ifbr_ifsname)); /* Fill in the ifbreq structure */ error = bridge_ioctl_gifflags(sc, &breq); if (error) break; memcpy(buf, &breq, sizeof(breq)); count++; buf += sizeof(breq); len -= sizeof(breq); } CK_LIST_FOREACH(bif, &sc->sc_spanlist, bif_next) { if (len < sizeof(breq)) break; strlcpy(breq.ifbr_ifsname, bif->bif_ifp->if_xname, sizeof(breq.ifbr_ifsname)); breq.ifbr_ifsflags = bif->bif_flags; breq.ifbr_portno = bif->bif_ifp->if_index & 0xfff; memcpy(buf, &breq, sizeof(breq)); count++; buf += sizeof(breq); len -= sizeof(breq); } bifc->ifbic_len = sizeof(breq) * count; error = copyout(outbuf, bifc->ifbic_req, bifc->ifbic_len); free(outbuf, M_TEMP); return (error); } static int bridge_ioctl_rts(struct bridge_softc *sc, void *arg) { struct ifbaconf *bac = arg; struct bridge_rtnode *brt; struct ifbareq bareq; char *buf, *outbuf; int count, buflen, len, error = 0; if (bac->ifbac_len == 0) return (0); count = 0; CK_LIST_FOREACH(brt, &sc->sc_rtlist, brt_list) count++; buflen = sizeof(bareq) * count; outbuf = malloc(buflen, M_TEMP, M_NOWAIT | M_ZERO); if (outbuf == NULL) return (ENOMEM); count = 0; buf = outbuf; len = min(bac->ifbac_len, buflen); bzero(&bareq, sizeof(bareq)); CK_LIST_FOREACH(brt, &sc->sc_rtlist, brt_list) { if (len < sizeof(bareq)) goto out; strlcpy(bareq.ifba_ifsname, brt->brt_ifp->if_xname, sizeof(bareq.ifba_ifsname)); memcpy(bareq.ifba_dst, brt->brt_addr, sizeof(brt->brt_addr)); bareq.ifba_vlan = brt->brt_vlan; if ((brt->brt_flags & IFBAF_TYPEMASK) == IFBAF_DYNAMIC && time_uptime < brt->brt_expire) bareq.ifba_expire = brt->brt_expire - time_uptime; else bareq.ifba_expire = 0; bareq.ifba_flags = brt->brt_flags; memcpy(buf, &bareq, sizeof(bareq)); count++; buf += sizeof(bareq); len -= sizeof(bareq); } out: bac->ifbac_len = sizeof(bareq) * count; error = copyout(outbuf, bac->ifbac_req, bac->ifbac_len); free(outbuf, M_TEMP); return (error); } static int bridge_ioctl_saddr(struct bridge_softc *sc, void *arg) { struct ifbareq *req = arg; struct bridge_iflist *bif; struct epoch_tracker et; int error; NET_EPOCH_ENTER(et); bif = bridge_lookup_member(sc, req->ifba_ifsname); if (bif == NULL) { NET_EPOCH_EXIT(et); return (ENOENT); } /* bridge_rtupdate() may acquire the lock. */ error = bridge_rtupdate(sc, req->ifba_dst, req->ifba_vlan, bif, 1, req->ifba_flags); NET_EPOCH_EXIT(et); return (error); } static int bridge_ioctl_sto(struct bridge_softc *sc, void *arg) { struct ifbrparam *param = arg; sc->sc_brttimeout = param->ifbrp_ctime; return (0); } static int bridge_ioctl_gto(struct bridge_softc *sc, void *arg) { struct ifbrparam *param = arg; param->ifbrp_ctime = sc->sc_brttimeout; return (0); } static int bridge_ioctl_daddr(struct bridge_softc *sc, void *arg) { struct ifbareq *req = arg; return (bridge_rtdaddr(sc, req->ifba_dst, req->ifba_vlan)); } static int bridge_ioctl_flush(struct bridge_softc *sc, void *arg) { struct ifbreq *req = arg; BRIDGE_RT_LOCK(sc); bridge_rtflush(sc, req->ifbr_ifsflags); BRIDGE_RT_UNLOCK(sc); return (0); } static int bridge_ioctl_gpri(struct bridge_softc *sc, void *arg) { struct ifbrparam *param = arg; struct bstp_state *bs = &sc->sc_stp; param->ifbrp_prio = bs->bs_bridge_priority; return (0); } static int bridge_ioctl_spri(struct bridge_softc *sc, void *arg) { struct ifbrparam *param = arg; return (bstp_set_priority(&sc->sc_stp, param->ifbrp_prio)); } static int bridge_ioctl_ght(struct bridge_softc *sc, void *arg) { struct ifbrparam *param = arg; struct bstp_state *bs = &sc->sc_stp; param->ifbrp_hellotime = bs->bs_bridge_htime >> 8; return (0); } static int bridge_ioctl_sht(struct bridge_softc *sc, void *arg) { struct ifbrparam *param = arg; return (bstp_set_htime(&sc->sc_stp, param->ifbrp_hellotime)); } static int bridge_ioctl_gfd(struct bridge_softc *sc, void *arg) { struct ifbrparam *param = arg; struct bstp_state *bs = &sc->sc_stp; param->ifbrp_fwddelay = bs->bs_bridge_fdelay >> 8; return (0); } static int bridge_ioctl_sfd(struct bridge_softc *sc, void *arg) { struct ifbrparam *param = arg; return (bstp_set_fdelay(&sc->sc_stp, param->ifbrp_fwddelay)); } static int bridge_ioctl_gma(struct bridge_softc *sc, void *arg) { struct ifbrparam *param = arg; struct bstp_state *bs = &sc->sc_stp; param->ifbrp_maxage = bs->bs_bridge_max_age >> 8; return (0); } static int bridge_ioctl_sma(struct bridge_softc *sc, void *arg) { struct ifbrparam *param = arg; return (bstp_set_maxage(&sc->sc_stp, param->ifbrp_maxage)); } static int bridge_ioctl_sifprio(struct bridge_softc *sc, void *arg) { struct ifbreq *req = arg; struct bridge_iflist *bif; bif = bridge_lookup_member(sc, req->ifbr_ifsname); if (bif == NULL) return (ENOENT); return (bstp_set_port_priority(&bif->bif_stp, req->ifbr_priority)); } static int bridge_ioctl_sifcost(struct bridge_softc *sc, void *arg) { struct ifbreq *req = arg; struct bridge_iflist *bif; bif = bridge_lookup_member(sc, req->ifbr_ifsname); if (bif == NULL) return (ENOENT); return (bstp_set_path_cost(&bif->bif_stp, req->ifbr_path_cost)); } static int bridge_ioctl_sifmaxaddr(struct bridge_softc *sc, void *arg) { struct ifbreq *req = arg; struct bridge_iflist *bif; bif = bridge_lookup_member(sc, req->ifbr_ifsname); if (bif == NULL) return (ENOENT); bif->bif_addrmax = req->ifbr_addrmax; return (0); } static int bridge_ioctl_addspan(struct bridge_softc *sc, void *arg) { struct ifbreq *req = arg; struct bridge_iflist *bif = NULL; struct ifnet *ifs; ifs = ifunit(req->ifbr_ifsname); if (ifs == NULL) return (ENOENT); CK_LIST_FOREACH(bif, &sc->sc_spanlist, bif_next) if (ifs == bif->bif_ifp) return (EBUSY); if (ifs->if_bridge != NULL) return (EBUSY); switch (ifs->if_type) { case IFT_ETHER: case IFT_GIF: case IFT_L2VLAN: break; default: return (EINVAL); } bif = malloc(sizeof(*bif), M_DEVBUF, M_NOWAIT|M_ZERO); if (bif == NULL) return (ENOMEM); bif->bif_ifp = ifs; bif->bif_flags = IFBIF_SPAN; CK_LIST_INSERT_HEAD(&sc->sc_spanlist, bif, bif_next); return (0); } static int bridge_ioctl_delspan(struct bridge_softc *sc, void *arg) { struct ifbreq *req = arg; struct bridge_iflist *bif; struct ifnet *ifs; ifs = ifunit(req->ifbr_ifsname); if (ifs == NULL) return (ENOENT); CK_LIST_FOREACH(bif, &sc->sc_spanlist, bif_next) if (ifs == bif->bif_ifp) break; if (bif == NULL) return (ENOENT); bridge_delete_span(sc, bif); return (0); } static int bridge_ioctl_gbparam(struct bridge_softc *sc, void *arg) { struct ifbropreq *req = arg; struct bstp_state *bs = &sc->sc_stp; struct bstp_port *root_port; req->ifbop_maxage = bs->bs_bridge_max_age >> 8; req->ifbop_hellotime = bs->bs_bridge_htime >> 8; req->ifbop_fwddelay = bs->bs_bridge_fdelay >> 8; root_port = bs->bs_root_port; if (root_port == NULL) req->ifbop_root_port = 0; else req->ifbop_root_port = root_port->bp_ifp->if_index; req->ifbop_holdcount = bs->bs_txholdcount; req->ifbop_priority = bs->bs_bridge_priority; req->ifbop_protocol = bs->bs_protover; req->ifbop_root_path_cost = bs->bs_root_pv.pv_cost; req->ifbop_bridgeid = bs->bs_bridge_pv.pv_dbridge_id; req->ifbop_designated_root = bs->bs_root_pv.pv_root_id; req->ifbop_designated_bridge = bs->bs_root_pv.pv_dbridge_id; req->ifbop_last_tc_time.tv_sec = bs->bs_last_tc_time.tv_sec; req->ifbop_last_tc_time.tv_usec = bs->bs_last_tc_time.tv_usec; return (0); } static int bridge_ioctl_grte(struct bridge_softc *sc, void *arg) { struct ifbrparam *param = arg; param->ifbrp_cexceeded = sc->sc_brtexceeded; return (0); } static int bridge_ioctl_gifsstp(struct bridge_softc *sc, void *arg) { struct ifbpstpconf *bifstp = arg; struct bridge_iflist *bif; struct bstp_port *bp; struct ifbpstpreq bpreq; char *buf, *outbuf; int count, buflen, len, error = 0; count = 0; CK_LIST_FOREACH(bif, &sc->sc_iflist, bif_next) { if ((bif->bif_flags & IFBIF_STP) != 0) count++; } buflen = sizeof(bpreq) * count; if (bifstp->ifbpstp_len == 0) { bifstp->ifbpstp_len = buflen; return (0); } outbuf = malloc(buflen, M_TEMP, M_NOWAIT | M_ZERO); if (outbuf == NULL) return (ENOMEM); count = 0; buf = outbuf; len = min(bifstp->ifbpstp_len, buflen); bzero(&bpreq, sizeof(bpreq)); CK_LIST_FOREACH(bif, &sc->sc_iflist, bif_next) { if (len < sizeof(bpreq)) break; if ((bif->bif_flags & IFBIF_STP) == 0) continue; bp = &bif->bif_stp; bpreq.ifbp_portno = bif->bif_ifp->if_index & 0xfff; bpreq.ifbp_fwd_trans = bp->bp_forward_transitions; bpreq.ifbp_design_cost = bp->bp_desg_pv.pv_cost; bpreq.ifbp_design_port = bp->bp_desg_pv.pv_port_id; bpreq.ifbp_design_bridge = bp->bp_desg_pv.pv_dbridge_id; bpreq.ifbp_design_root = bp->bp_desg_pv.pv_root_id; memcpy(buf, &bpreq, sizeof(bpreq)); count++; buf += sizeof(bpreq); len -= sizeof(bpreq); } bifstp->ifbpstp_len = sizeof(bpreq) * count; error = copyout(outbuf, bifstp->ifbpstp_req, bifstp->ifbpstp_len); free(outbuf, M_TEMP); return (error); } static int bridge_ioctl_sproto(struct bridge_softc *sc, void *arg) { struct ifbrparam *param = arg; return (bstp_set_protocol(&sc->sc_stp, param->ifbrp_proto)); } static int bridge_ioctl_stxhc(struct bridge_softc *sc, void *arg) { struct ifbrparam *param = arg; return (bstp_set_holdcount(&sc->sc_stp, param->ifbrp_txhc)); } /* * bridge_ifdetach: * * Detach an interface from a bridge. Called when a member * interface is detaching. */ static void bridge_ifdetach(void *arg __unused, struct ifnet *ifp) { struct bridge_softc *sc = ifp->if_bridge; struct bridge_iflist *bif; if (ifp->if_flags & IFF_RENAMING) return; if (V_bridge_cloner == NULL) { /* * This detach handler can be called after * vnet_bridge_uninit(). Just return in that case. */ return; } /* Check if the interface is a bridge member */ if (sc != NULL) { BRIDGE_LOCK(sc); bif = bridge_lookup_member_if(sc, ifp); if (bif != NULL) bridge_delete_member(sc, bif, 1); BRIDGE_UNLOCK(sc); return; } /* Check if the interface is a span port */ BRIDGE_LIST_LOCK(); LIST_FOREACH(sc, &V_bridge_list, sc_list) { BRIDGE_LOCK(sc); CK_LIST_FOREACH(bif, &sc->sc_spanlist, bif_next) if (ifp == bif->bif_ifp) { bridge_delete_span(sc, bif); break; } BRIDGE_UNLOCK(sc); } BRIDGE_LIST_UNLOCK(); } /* * bridge_init: * * Initialize a bridge interface. */ static void bridge_init(void *xsc) { struct bridge_softc *sc = (struct bridge_softc *)xsc; struct ifnet *ifp = sc->sc_ifp; if (ifp->if_drv_flags & IFF_DRV_RUNNING) return; BRIDGE_LOCK(sc); callout_reset(&sc->sc_brcallout, bridge_rtable_prune_period * hz, bridge_timer, sc); ifp->if_drv_flags |= IFF_DRV_RUNNING; bstp_init(&sc->sc_stp); /* Initialize Spanning Tree */ BRIDGE_UNLOCK(sc); } /* * bridge_stop: * * Stop the bridge interface. */ static void bridge_stop(struct ifnet *ifp, int disable) { struct bridge_softc *sc = ifp->if_softc; BRIDGE_LOCK_ASSERT(sc); if ((ifp->if_drv_flags & IFF_DRV_RUNNING) == 0) return; BRIDGE_RT_LOCK(sc); callout_stop(&sc->sc_brcallout); bstp_stop(&sc->sc_stp); bridge_rtflush(sc, IFBF_FLUSHDYN); BRIDGE_RT_UNLOCK(sc); ifp->if_drv_flags &= ~IFF_DRV_RUNNING; } /* * bridge_enqueue: * * Enqueue a packet on a bridge member interface. * */ static int bridge_enqueue(struct bridge_softc *sc, struct ifnet *dst_ifp, struct mbuf *m) { int len, err = 0; short mflags; struct mbuf *m0; /* We may be sending a fragment so traverse the mbuf */ for (; m; m = m0) { m0 = m->m_nextpkt; m->m_nextpkt = NULL; len = m->m_pkthdr.len; mflags = m->m_flags; /* * If underlying interface can not do VLAN tag insertion itself * then attach a packet tag that holds it. */ if ((m->m_flags & M_VLANTAG) && (dst_ifp->if_capenable & IFCAP_VLAN_HWTAGGING) == 0) { m = ether_vlanencap(m, m->m_pkthdr.ether_vtag); if (m == NULL) { if_printf(dst_ifp, "unable to prepend VLAN header\n"); if_inc_counter(dst_ifp, IFCOUNTER_OERRORS, 1); continue; } m->m_flags &= ~M_VLANTAG; } M_ASSERTPKTHDR(m); /* We shouldn't transmit mbuf without pkthdr */ if ((err = dst_ifp->if_transmit(dst_ifp, m))) { int n; for (m = m0, n = 1; m != NULL; m = m0, n++) { m0 = m->m_nextpkt; m_freem(m); } if_inc_counter(sc->sc_ifp, IFCOUNTER_OERRORS, n); break; } if_inc_counter(sc->sc_ifp, IFCOUNTER_OPACKETS, 1); if_inc_counter(sc->sc_ifp, IFCOUNTER_OBYTES, len); if (mflags & M_MCAST) if_inc_counter(sc->sc_ifp, IFCOUNTER_OMCASTS, 1); } return (err); } /* * bridge_dummynet: * * Receive a queued packet from dummynet and pass it on to the output * interface. * * The mbuf has the Ethernet header already attached. */ static void bridge_dummynet(struct mbuf *m, struct ifnet *ifp) { struct bridge_softc *sc; sc = ifp->if_bridge; /* * The packet didnt originate from a member interface. This should only * ever happen if a member interface is removed while packets are * queued for it. */ if (sc == NULL) { m_freem(m); return; } if (PFIL_HOOKED_OUT(V_inet_pfil_head) #ifdef INET6 || PFIL_HOOKED_OUT(V_inet6_pfil_head) #endif ) { if (bridge_pfil(&m, sc->sc_ifp, ifp, PFIL_OUT) != 0) return; if (m == NULL) return; } bridge_enqueue(sc, ifp, m); } /* * bridge_output: * * Send output from a bridge member interface. This * performs the bridging function for locally originated * packets. * * The mbuf has the Ethernet header already attached. We must * enqueue or free the mbuf before returning. */ static int bridge_output(struct ifnet *ifp, struct mbuf *m, struct sockaddr *sa, struct rtentry *rt) { struct ether_header *eh; struct ifnet *bifp, *dst_if; struct bridge_softc *sc; uint16_t vlan; NET_EPOCH_ASSERT(); if (m->m_len < ETHER_HDR_LEN) { m = m_pullup(m, ETHER_HDR_LEN); if (m == NULL) return (0); } eh = mtod(m, struct ether_header *); sc = ifp->if_bridge; vlan = VLANTAGOF(m); bifp = sc->sc_ifp; /* * If bridge is down, but the original output interface is up, * go ahead and send out that interface. Otherwise, the packet * is dropped below. */ if ((bifp->if_drv_flags & IFF_DRV_RUNNING) == 0) { dst_if = ifp; goto sendunicast; } /* * If the packet is a multicast, or we don't know a better way to * get there, send to all interfaces. */ if (ETHER_IS_MULTICAST(eh->ether_dhost)) dst_if = NULL; else dst_if = bridge_rtlookup(sc, eh->ether_dhost, vlan); /* Tap any traffic not passing back out the originating interface */ if (dst_if != ifp) ETHER_BPF_MTAP(bifp, m); if (dst_if == NULL) { struct bridge_iflist *bif; struct mbuf *mc; int used = 0; bridge_span(sc, m); CK_LIST_FOREACH(bif, &sc->sc_iflist, bif_next) { dst_if = bif->bif_ifp; if (dst_if->if_type == IFT_GIF) continue; if ((dst_if->if_drv_flags & IFF_DRV_RUNNING) == 0) continue; /* * If this is not the original output interface, * and the interface is participating in spanning * tree, make sure the port is in a state that * allows forwarding. */ if (dst_if != ifp && (bif->bif_flags & IFBIF_STP) && bif->bif_stp.bp_state == BSTP_IFSTATE_DISCARDING) continue; if (CK_LIST_NEXT(bif, bif_next) == NULL) { used = 1; mc = m; } else { mc = m_dup(m, M_NOWAIT); if (mc == NULL) { if_inc_counter(bifp, IFCOUNTER_OERRORS, 1); continue; } } bridge_enqueue(sc, dst_if, mc); } if (used == 0) m_freem(m); return (0); } sendunicast: /* * XXX Spanning tree consideration here? */ bridge_span(sc, m); if ((dst_if->if_drv_flags & IFF_DRV_RUNNING) == 0) { m_freem(m); return (0); } bridge_enqueue(sc, dst_if, m); return (0); } /* * bridge_transmit: * * Do output on a bridge. * */ static int bridge_transmit(struct ifnet *ifp, struct mbuf *m) { struct bridge_softc *sc; struct ether_header *eh; struct ifnet *dst_if; int error = 0; sc = ifp->if_softc; ETHER_BPF_MTAP(ifp, m); eh = mtod(m, struct ether_header *); if (((m->m_flags & (M_BCAST|M_MCAST)) == 0) && (dst_if = bridge_rtlookup(sc, eh->ether_dhost, 1)) != NULL) { error = bridge_enqueue(sc, dst_if, m); } else bridge_broadcast(sc, ifp, m, 0); return (error); } #ifdef ALTQ static void bridge_altq_start(if_t ifp) { struct ifaltq *ifq = &ifp->if_snd; struct mbuf *m; IFQ_LOCK(ifq); IFQ_DEQUEUE_NOLOCK(ifq, m); while (m != NULL) { bridge_transmit(ifp, m); IFQ_DEQUEUE_NOLOCK(ifq, m); } IFQ_UNLOCK(ifq); } static int bridge_altq_transmit(if_t ifp, struct mbuf *m) { int err; if (ALTQ_IS_ENABLED(&ifp->if_snd)) { IFQ_ENQUEUE(&ifp->if_snd, m, err); if (err == 0) bridge_altq_start(ifp); } else err = bridge_transmit(ifp, m); return (err); } #endif /* ALTQ */ /* * The ifp->if_qflush entry point for if_bridge(4) is no-op. */ static void bridge_qflush(struct ifnet *ifp __unused) { } /* * bridge_forward: * * The forwarding function of the bridge. * * NOTE: Releases the lock on return. */ static void bridge_forward(struct bridge_softc *sc, struct bridge_iflist *sbif, struct mbuf *m) { struct bridge_iflist *dbif; struct ifnet *src_if, *dst_if, *ifp; struct ether_header *eh; uint16_t vlan; uint8_t *dst; int error; NET_EPOCH_ASSERT(); src_if = m->m_pkthdr.rcvif; ifp = sc->sc_ifp; if_inc_counter(ifp, IFCOUNTER_IPACKETS, 1); if_inc_counter(ifp, IFCOUNTER_IBYTES, m->m_pkthdr.len); vlan = VLANTAGOF(m); if ((sbif->bif_flags & IFBIF_STP) && sbif->bif_stp.bp_state == BSTP_IFSTATE_DISCARDING) goto drop; eh = mtod(m, struct ether_header *); dst = eh->ether_dhost; /* If the interface is learning, record the address. */ if (sbif->bif_flags & IFBIF_LEARNING) { error = bridge_rtupdate(sc, eh->ether_shost, vlan, sbif, 0, IFBAF_DYNAMIC); /* * If the interface has addresses limits then deny any source * that is not in the cache. */ if (error && sbif->bif_addrmax) goto drop; } if ((sbif->bif_flags & IFBIF_STP) != 0 && sbif->bif_stp.bp_state == BSTP_IFSTATE_LEARNING) goto drop; +#ifdef DEV_NETMAP + /* + * Hand the packet to netmap only if it wasn't injected by netmap + * itself. + */ + if ((m->m_flags & M_BRIDGE_INJECT) == 0 && + (if_getcapenable(ifp) & IFCAP_NETMAP) != 0) { + ifp->if_input(ifp, m); + return; + } + m->m_flags &= ~M_BRIDGE_INJECT; +#endif + /* * At this point, the port either doesn't participate * in spanning tree or it is in the forwarding state. */ /* * If the packet is unicast, destined for someone on * "this" side of the bridge, drop it. */ if ((m->m_flags & (M_BCAST|M_MCAST)) == 0) { dst_if = bridge_rtlookup(sc, dst, vlan); if (src_if == dst_if) goto drop; } else { /* * Check if its a reserved multicast address, any address * listed in 802.1D section 7.12.6 may not be forwarded by the * bridge. * This is currently 01-80-C2-00-00-00 to 01-80-C2-00-00-0F */ if (dst[0] == 0x01 && dst[1] == 0x80 && dst[2] == 0xc2 && dst[3] == 0x00 && dst[4] == 0x00 && dst[5] <= 0x0f) goto drop; /* ...forward it to all interfaces. */ if_inc_counter(ifp, IFCOUNTER_IMCASTS, 1); dst_if = NULL; } /* * If we have a destination interface which is a member of our bridge, * OR this is a unicast packet, push it through the bpf(4) machinery. * For broadcast or multicast packets, don't bother because it will * be reinjected into ether_input. We do this before we pass the packets * through the pfil(9) framework, as it is possible that pfil(9) will * drop the packet, or possibly modify it, making it difficult to debug * firewall issues on the bridge. */ if (dst_if != NULL || (m->m_flags & (M_BCAST | M_MCAST)) == 0) ETHER_BPF_MTAP(ifp, m); /* run the packet filter */ if (PFIL_HOOKED_IN(V_inet_pfil_head) #ifdef INET6 || PFIL_HOOKED_IN(V_inet6_pfil_head) #endif ) { if (bridge_pfil(&m, ifp, src_if, PFIL_IN) != 0) return; if (m == NULL) return; } if (dst_if == NULL) { bridge_broadcast(sc, src_if, m, 1); return; } /* * At this point, we're dealing with a unicast frame * going to a different interface. */ if ((dst_if->if_drv_flags & IFF_DRV_RUNNING) == 0) goto drop; dbif = bridge_lookup_member_if(sc, dst_if); if (dbif == NULL) /* Not a member of the bridge (anymore?) */ goto drop; /* Private segments can not talk to each other */ if (sbif->bif_flags & dbif->bif_flags & IFBIF_PRIVATE) goto drop; if ((dbif->bif_flags & IFBIF_STP) && dbif->bif_stp.bp_state == BSTP_IFSTATE_DISCARDING) goto drop; if (PFIL_HOOKED_OUT(V_inet_pfil_head) #ifdef INET6 || PFIL_HOOKED_OUT(V_inet6_pfil_head) #endif ) { if (bridge_pfil(&m, ifp, dst_if, PFIL_OUT) != 0) return; if (m == NULL) return; } bridge_enqueue(sc, dst_if, m); return; drop: m_freem(m); } /* * bridge_input: * * Receive input from a member interface. Queue the packet for * bridging if it is not for us. */ static struct mbuf * bridge_input(struct ifnet *ifp, struct mbuf *m) { - struct bridge_softc *sc = ifp->if_bridge; + struct bridge_softc *sc; struct bridge_iflist *bif, *bif2; struct ifnet *bifp; struct ether_header *eh; struct mbuf *mc, *mc2; uint16_t vlan; int error; NET_EPOCH_ASSERT(); - if ((sc->sc_ifp->if_drv_flags & IFF_DRV_RUNNING) == 0) - return (m); + eh = mtod(m, struct ether_header *); + vlan = VLANTAGOF(m); + sc = ifp->if_bridge; + if (sc == NULL) { + /* + * This packet originated from the bridge itself, so it must + * have been transmitted by netmap. Derive the "source" + * interface from the source address and drop the packet if the + * source address isn't known. + */ + KASSERT((m->m_flags & M_BRIDGE_INJECT) != 0, + ("%s: ifnet %p missing a bridge softc", __func__, ifp)); + sc = if_getsoftc(ifp); + ifp = bridge_rtlookup(sc, eh->ether_shost, vlan); + if (ifp == NULL) { + if_inc_counter(sc->sc_ifp, IFCOUNTER_IERRORS, 1); + m_freem(m); + return (NULL); + } + m->m_pkthdr.rcvif = ifp; + } bifp = sc->sc_ifp; - vlan = VLANTAGOF(m); + if ((bifp->if_drv_flags & IFF_DRV_RUNNING) == 0) + return (m); /* * Implement support for bridge monitoring. If this flag has been * set on this interface, discard the packet once we push it through * the bpf(4) machinery, but before we do, increment the byte and * packet counters associated with this interface. */ if ((bifp->if_flags & IFF_MONITOR) != 0) { m->m_pkthdr.rcvif = bifp; ETHER_BPF_MTAP(bifp, m); if_inc_counter(bifp, IFCOUNTER_IPACKETS, 1); if_inc_counter(bifp, IFCOUNTER_IBYTES, m->m_pkthdr.len); m_freem(m); return (NULL); } bif = bridge_lookup_member_if(sc, ifp); if (bif == NULL) { return (m); } - eh = mtod(m, struct ether_header *); - bridge_span(sc, m); if (m->m_flags & (M_BCAST|M_MCAST)) { /* Tap off 802.1D packets; they do not get forwarded. */ if (memcmp(eh->ether_dhost, bstp_etheraddr, ETHER_ADDR_LEN) == 0) { bstp_input(&bif->bif_stp, ifp, m); /* consumes mbuf */ return (NULL); } if ((bif->bif_flags & IFBIF_STP) && bif->bif_stp.bp_state == BSTP_IFSTATE_DISCARDING) { return (m); } /* * Make a deep copy of the packet and enqueue the copy * for bridge processing; return the original packet for * local processing. */ mc = m_dup(m, M_NOWAIT); if (mc == NULL) { return (m); } /* Perform the bridge forwarding function with the copy. */ bridge_forward(sc, bif, mc); +#ifdef DEV_NETMAP + /* + * If netmap is enabled and has not already seen this packet, + * then it will be consumed by bridge_forward(). + */ + if ((if_getcapenable(bifp) & IFCAP_NETMAP) != 0 && + (m->m_flags & M_BRIDGE_INJECT) == 0) { + m_freem(m); + return (NULL); + } +#endif + /* * Reinject the mbuf as arriving on the bridge so we have a * chance at claiming multicast packets. We can not loop back * here from ether_input as a bridge is never a member of a * bridge. */ KASSERT(bifp->if_bridge == NULL, ("loop created in bridge_input")); mc2 = m_dup(m, M_NOWAIT); if (mc2 != NULL) { /* Keep the layer3 header aligned */ int i = min(mc2->m_pkthdr.len, max_protohdr); mc2 = m_copyup(mc2, i, ETHER_ALIGN); } if (mc2 != NULL) { mc2->m_pkthdr.rcvif = bifp; - (*bifp->if_input)(bifp, mc2); + mc2->m_flags &= ~M_BRIDGE_INJECT; + sc->sc_if_input(bifp, mc2); } /* Return the original packet for local processing. */ return (m); } if ((bif->bif_flags & IFBIF_STP) && bif->bif_stp.bp_state == BSTP_IFSTATE_DISCARDING) { return (m); } #if defined(INET) || defined(INET6) #define CARP_CHECK_WE_ARE_DST(iface) \ ((iface)->if_carp && (*carp_forus_p)((iface), eh->ether_dhost)) #define CARP_CHECK_WE_ARE_SRC(iface) \ ((iface)->if_carp && (*carp_forus_p)((iface), eh->ether_shost)) #else #define CARP_CHECK_WE_ARE_DST(iface) false #define CARP_CHECK_WE_ARE_SRC(iface) false #endif #ifdef INET6 #define PFIL_HOOKED_INET6 PFIL_HOOKED_IN(V_inet6_pfil_head) #else #define PFIL_HOOKED_INET6 false #endif +#ifdef DEV_NETMAP +#define GRAB_FOR_NETMAP(ifp, m) do { \ + if ((if_getcapenable(ifp) & IFCAP_NETMAP) != 0 && \ + ((m)->m_flags & M_BRIDGE_INJECT) == 0) { \ + (ifp)->if_input(ifp, m); \ + return (NULL); \ + } \ +} while (0) +#else +#define GRAB_FOR_NETMAP(ifp, m) +#endif + #define GRAB_OUR_PACKETS(iface) \ if ((iface)->if_type == IFT_GIF) \ continue; \ /* It is destined for us. */ \ if (memcmp(IF_LLADDR(iface), eh->ether_dhost, ETHER_ADDR_LEN) == 0 || \ CARP_CHECK_WE_ARE_DST(iface)) { \ if (bif->bif_flags & IFBIF_LEARNING) { \ error = bridge_rtupdate(sc, eh->ether_shost, \ vlan, bif, 0, IFBAF_DYNAMIC); \ if (error && bif->bif_addrmax) { \ m_freem(m); \ return (NULL); \ } \ } \ m->m_pkthdr.rcvif = iface; \ if ((iface) == ifp) { \ /* Skip bridge processing... src == dest */ \ return (m); \ } \ /* It's passing over or to the bridge, locally. */ \ ETHER_BPF_MTAP(bifp, m); \ if_inc_counter(bifp, IFCOUNTER_IPACKETS, 1); \ - if_inc_counter(bifp, IFCOUNTER_IBYTES, m->m_pkthdr.len); \ + if_inc_counter(bifp, IFCOUNTER_IBYTES, m->m_pkthdr.len);\ + /* Hand the packet over to netmap if necessary. */ \ + GRAB_FOR_NETMAP(bifp, m); \ /* Filter on the physical interface. */ \ if (V_pfil_local_phys && (PFIL_HOOKED_IN(V_inet_pfil_head) || \ PFIL_HOOKED_INET6)) { \ if (bridge_pfil(&m, NULL, ifp, \ PFIL_IN) != 0 || m == NULL) { \ return (NULL); \ } \ } \ if ((iface) != bifp) \ ETHER_BPF_MTAP(iface, m); \ return (m); \ } \ \ /* We just received a packet that we sent out. */ \ if (memcmp(IF_LLADDR(iface), eh->ether_shost, ETHER_ADDR_LEN) == 0 || \ CARP_CHECK_WE_ARE_SRC(iface)) { \ m_freem(m); \ return (NULL); \ } /* * Unicast. Make sure it's not for the bridge. */ do { GRAB_OUR_PACKETS(bifp) } while (0); /* * Give a chance for ifp at first priority. This will help when the * packet comes through the interface like VLAN's with the same MACs * on several interfaces from the same bridge. This also will save * some CPU cycles in case the destination interface and the input * interface (eq ifp) are the same. */ do { GRAB_OUR_PACKETS(ifp) } while (0); /* Now check the all bridge members. */ CK_LIST_FOREACH(bif2, &sc->sc_iflist, bif_next) { GRAB_OUR_PACKETS(bif2->bif_ifp) } #undef CARP_CHECK_WE_ARE_DST #undef CARP_CHECK_WE_ARE_SRC #undef PFIL_HOOKED_INET6 +#undef GRAB_FOR_NETMAP #undef GRAB_OUR_PACKETS /* Perform the bridge forwarding function. */ bridge_forward(sc, bif, m); return (NULL); } +/* + * Inject a packet back into the host ethernet stack. This will generally only + * be used by netmap when an application writes to the host TX ring. The + * M_BRIDGE_INJECT flag ensures that the packet is re-routed to the bridge + * interface after ethernet processing. + */ +static void +bridge_inject(struct ifnet *ifp, struct mbuf *m) +{ + struct bridge_softc *sc; + + KASSERT((if_getcapenable(ifp) & IFCAP_NETMAP) != 0, + ("%s: iface %s is not running in netmap mode", + __func__, if_name(ifp))); + KASSERT((m->m_flags & M_BRIDGE_INJECT) == 0, + ("%s: mbuf %p has M_BRIDGE_INJECT set", __func__, m)); + + m->m_flags |= M_BRIDGE_INJECT; + sc = if_getsoftc(ifp); + sc->sc_if_input(ifp, m); +} + /* * bridge_broadcast: * * Send a frame to all interfaces that are members of * the bridge, except for the one on which the packet * arrived. * * NOTE: Releases the lock on return. */ static void bridge_broadcast(struct bridge_softc *sc, struct ifnet *src_if, struct mbuf *m, int runfilt) { struct bridge_iflist *dbif, *sbif; struct mbuf *mc; struct ifnet *dst_if; int used = 0, i; NET_EPOCH_ASSERT(); sbif = bridge_lookup_member_if(sc, src_if); /* Filter on the bridge interface before broadcasting */ if (runfilt && (PFIL_HOOKED_OUT(V_inet_pfil_head) #ifdef INET6 || PFIL_HOOKED_OUT(V_inet6_pfil_head) #endif )) { if (bridge_pfil(&m, sc->sc_ifp, NULL, PFIL_OUT) != 0) return; if (m == NULL) return; } CK_LIST_FOREACH(dbif, &sc->sc_iflist, bif_next) { dst_if = dbif->bif_ifp; if (dst_if == src_if) continue; /* Private segments can not talk to each other */ if (sbif && (sbif->bif_flags & dbif->bif_flags & IFBIF_PRIVATE)) continue; if ((dbif->bif_flags & IFBIF_STP) && dbif->bif_stp.bp_state == BSTP_IFSTATE_DISCARDING) continue; if ((dbif->bif_flags & IFBIF_DISCOVER) == 0 && (m->m_flags & (M_BCAST|M_MCAST)) == 0) continue; if ((dst_if->if_drv_flags & IFF_DRV_RUNNING) == 0) continue; if (CK_LIST_NEXT(dbif, bif_next) == NULL) { mc = m; used = 1; } else { mc = m_dup(m, M_NOWAIT); if (mc == NULL) { if_inc_counter(sc->sc_ifp, IFCOUNTER_OERRORS, 1); continue; } } /* * Filter on the output interface. Pass a NULL bridge interface * pointer so we do not redundantly filter on the bridge for * each interface we broadcast on. */ if (runfilt && (PFIL_HOOKED_OUT(V_inet_pfil_head) #ifdef INET6 || PFIL_HOOKED_OUT(V_inet6_pfil_head) #endif )) { if (used == 0) { /* Keep the layer3 header aligned */ i = min(mc->m_pkthdr.len, max_protohdr); mc = m_copyup(mc, i, ETHER_ALIGN); if (mc == NULL) { if_inc_counter(sc->sc_ifp, IFCOUNTER_OERRORS, 1); continue; } } if (bridge_pfil(&mc, NULL, dst_if, PFIL_OUT) != 0) continue; if (mc == NULL) continue; } bridge_enqueue(sc, dst_if, mc); } if (used == 0) m_freem(m); } /* * bridge_span: * * Duplicate a packet out one or more interfaces that are in span mode, * the original mbuf is unmodified. */ static void bridge_span(struct bridge_softc *sc, struct mbuf *m) { struct bridge_iflist *bif; struct ifnet *dst_if; struct mbuf *mc; NET_EPOCH_ASSERT(); if (CK_LIST_EMPTY(&sc->sc_spanlist)) return; CK_LIST_FOREACH(bif, &sc->sc_spanlist, bif_next) { dst_if = bif->bif_ifp; if ((dst_if->if_drv_flags & IFF_DRV_RUNNING) == 0) continue; mc = m_dup(m, M_NOWAIT); if (mc == NULL) { if_inc_counter(sc->sc_ifp, IFCOUNTER_OERRORS, 1); continue; } bridge_enqueue(sc, dst_if, mc); } } /* * bridge_rtupdate: * * Add a bridge routing entry. */ static int bridge_rtupdate(struct bridge_softc *sc, const uint8_t *dst, uint16_t vlan, struct bridge_iflist *bif, int setflags, uint8_t flags) { struct bridge_rtnode *brt; struct bridge_iflist *obif; int error; BRIDGE_LOCK_OR_NET_EPOCH_ASSERT(sc); /* Check the source address is valid and not multicast. */ if (ETHER_IS_MULTICAST(dst) || (dst[0] == 0 && dst[1] == 0 && dst[2] == 0 && dst[3] == 0 && dst[4] == 0 && dst[5] == 0) != 0) return (EINVAL); /* 802.1p frames map to vlan 1 */ if (vlan == 0) vlan = 1; /* * A route for this destination might already exist. If so, * update it, otherwise create a new one. */ if ((brt = bridge_rtnode_lookup(sc, dst, vlan)) == NULL) { BRIDGE_RT_LOCK(sc); /* Check again, now that we have the lock. There could have * been a race and we only want to insert this once. */ if (bridge_rtnode_lookup(sc, dst, vlan) != NULL) { BRIDGE_RT_UNLOCK(sc); return (0); } if (sc->sc_brtcnt >= sc->sc_brtmax) { sc->sc_brtexceeded++; BRIDGE_RT_UNLOCK(sc); return (ENOSPC); } /* Check per interface address limits (if enabled) */ if (bif->bif_addrmax && bif->bif_addrcnt >= bif->bif_addrmax) { bif->bif_addrexceeded++; BRIDGE_RT_UNLOCK(sc); return (ENOSPC); } /* * Allocate a new bridge forwarding node, and * initialize the expiration time and Ethernet * address. */ brt = uma_zalloc(V_bridge_rtnode_zone, M_NOWAIT | M_ZERO); if (brt == NULL) { BRIDGE_RT_UNLOCK(sc); return (ENOMEM); } brt->brt_vnet = curvnet; if (bif->bif_flags & IFBIF_STICKY) brt->brt_flags = IFBAF_STICKY; else brt->brt_flags = IFBAF_DYNAMIC; memcpy(brt->brt_addr, dst, ETHER_ADDR_LEN); brt->brt_vlan = vlan; if ((error = bridge_rtnode_insert(sc, brt)) != 0) { uma_zfree(V_bridge_rtnode_zone, brt); BRIDGE_RT_UNLOCK(sc); return (error); } brt->brt_dst = bif; bif->bif_addrcnt++; BRIDGE_RT_UNLOCK(sc); } if ((brt->brt_flags & IFBAF_TYPEMASK) == IFBAF_DYNAMIC && (obif = brt->brt_dst) != bif) { BRIDGE_RT_LOCK(sc); brt->brt_dst->bif_addrcnt--; brt->brt_dst = bif; brt->brt_dst->bif_addrcnt++; BRIDGE_RT_UNLOCK(sc); if (V_log_mac_flap && ppsratecheck(&V_log_last, &V_log_count, V_log_interval)) { uint8_t *addr = &brt->brt_addr[0]; log(LOG_NOTICE, "%s: mac address %02x:%02x:%02x:%02x:%02x:%02x vlan %d moved from %s to %s\n", sc->sc_ifp->if_xname, addr[0], addr[1], addr[2], addr[3], addr[4], addr[5], brt->brt_vlan, obif->bif_ifp->if_xname, bif->bif_ifp->if_xname); } } if ((flags & IFBAF_TYPEMASK) == IFBAF_DYNAMIC) brt->brt_expire = time_uptime + sc->sc_brttimeout; if (setflags) brt->brt_flags = flags; return (0); } /* * bridge_rtlookup: * * Lookup the destination interface for an address. */ static struct ifnet * bridge_rtlookup(struct bridge_softc *sc, const uint8_t *addr, uint16_t vlan) { struct bridge_rtnode *brt; NET_EPOCH_ASSERT(); if ((brt = bridge_rtnode_lookup(sc, addr, vlan)) == NULL) return (NULL); return (brt->brt_ifp); } /* * bridge_rttrim: * * Trim the routine table so that we have a number * of routing entries less than or equal to the * maximum number. */ static void bridge_rttrim(struct bridge_softc *sc) { struct bridge_rtnode *brt, *nbrt; NET_EPOCH_ASSERT(); BRIDGE_RT_LOCK_ASSERT(sc); /* Make sure we actually need to do this. */ if (sc->sc_brtcnt <= sc->sc_brtmax) return; /* Force an aging cycle; this might trim enough addresses. */ bridge_rtage(sc); if (sc->sc_brtcnt <= sc->sc_brtmax) return; CK_LIST_FOREACH_SAFE(brt, &sc->sc_rtlist, brt_list, nbrt) { if ((brt->brt_flags & IFBAF_TYPEMASK) == IFBAF_DYNAMIC) { bridge_rtnode_destroy(sc, brt); if (sc->sc_brtcnt <= sc->sc_brtmax) return; } } } /* * bridge_timer: * * Aging timer for the bridge. */ static void bridge_timer(void *arg) { struct bridge_softc *sc = arg; BRIDGE_RT_LOCK_ASSERT(sc); /* Destruction of rtnodes requires a proper vnet context */ CURVNET_SET(sc->sc_ifp->if_vnet); bridge_rtage(sc); if (sc->sc_ifp->if_drv_flags & IFF_DRV_RUNNING) callout_reset(&sc->sc_brcallout, bridge_rtable_prune_period * hz, bridge_timer, sc); CURVNET_RESTORE(); } /* * bridge_rtage: * * Perform an aging cycle. */ static void bridge_rtage(struct bridge_softc *sc) { struct bridge_rtnode *brt, *nbrt; BRIDGE_RT_LOCK_ASSERT(sc); CK_LIST_FOREACH_SAFE(brt, &sc->sc_rtlist, brt_list, nbrt) { if ((brt->brt_flags & IFBAF_TYPEMASK) == IFBAF_DYNAMIC) { if (time_uptime >= brt->brt_expire) bridge_rtnode_destroy(sc, brt); } } } /* * bridge_rtflush: * * Remove all dynamic addresses from the bridge. */ static void bridge_rtflush(struct bridge_softc *sc, int full) { struct bridge_rtnode *brt, *nbrt; BRIDGE_RT_LOCK_ASSERT(sc); CK_LIST_FOREACH_SAFE(brt, &sc->sc_rtlist, brt_list, nbrt) { if (full || (brt->brt_flags & IFBAF_TYPEMASK) == IFBAF_DYNAMIC) bridge_rtnode_destroy(sc, brt); } } /* * bridge_rtdaddr: * * Remove an address from the table. */ static int bridge_rtdaddr(struct bridge_softc *sc, const uint8_t *addr, uint16_t vlan) { struct bridge_rtnode *brt; int found = 0; BRIDGE_RT_LOCK(sc); /* * If vlan is zero then we want to delete for all vlans so the lookup * may return more than one. */ while ((brt = bridge_rtnode_lookup(sc, addr, vlan)) != NULL) { bridge_rtnode_destroy(sc, brt); found = 1; } BRIDGE_RT_UNLOCK(sc); return (found ? 0 : ENOENT); } /* * bridge_rtdelete: * * Delete routes to a speicifc member interface. */ static void bridge_rtdelete(struct bridge_softc *sc, struct ifnet *ifp, int full) { struct bridge_rtnode *brt, *nbrt; BRIDGE_RT_LOCK_ASSERT(sc); CK_LIST_FOREACH_SAFE(brt, &sc->sc_rtlist, brt_list, nbrt) { if (brt->brt_ifp == ifp && (full || (brt->brt_flags & IFBAF_TYPEMASK) == IFBAF_DYNAMIC)) bridge_rtnode_destroy(sc, brt); } } /* * bridge_rtable_init: * * Initialize the route table for this bridge. */ static void bridge_rtable_init(struct bridge_softc *sc) { int i; sc->sc_rthash = malloc(sizeof(*sc->sc_rthash) * BRIDGE_RTHASH_SIZE, M_DEVBUF, M_WAITOK); for (i = 0; i < BRIDGE_RTHASH_SIZE; i++) CK_LIST_INIT(&sc->sc_rthash[i]); sc->sc_rthash_key = arc4random(); CK_LIST_INIT(&sc->sc_rtlist); } /* * bridge_rtable_fini: * * Deconstruct the route table for this bridge. */ static void bridge_rtable_fini(struct bridge_softc *sc) { KASSERT(sc->sc_brtcnt == 0, ("%s: %d bridge routes referenced", __func__, sc->sc_brtcnt)); free(sc->sc_rthash, M_DEVBUF); } /* * The following hash function is adapted from "Hash Functions" by Bob Jenkins * ("Algorithm Alley", Dr. Dobbs Journal, September 1997). */ #define mix(a, b, c) \ do { \ a -= b; a -= c; a ^= (c >> 13); \ b -= c; b -= a; b ^= (a << 8); \ c -= a; c -= b; c ^= (b >> 13); \ a -= b; a -= c; a ^= (c >> 12); \ b -= c; b -= a; b ^= (a << 16); \ c -= a; c -= b; c ^= (b >> 5); \ a -= b; a -= c; a ^= (c >> 3); \ b -= c; b -= a; b ^= (a << 10); \ c -= a; c -= b; c ^= (b >> 15); \ } while (/*CONSTCOND*/0) static __inline uint32_t bridge_rthash(struct bridge_softc *sc, const uint8_t *addr) { uint32_t a = 0x9e3779b9, b = 0x9e3779b9, c = sc->sc_rthash_key; b += addr[5] << 8; b += addr[4]; a += addr[3] << 24; a += addr[2] << 16; a += addr[1] << 8; a += addr[0]; mix(a, b, c); return (c & BRIDGE_RTHASH_MASK); } #undef mix static int bridge_rtnode_addr_cmp(const uint8_t *a, const uint8_t *b) { int i, d; for (i = 0, d = 0; i < ETHER_ADDR_LEN && d == 0; i++) { d = ((int)a[i]) - ((int)b[i]); } return (d); } /* * bridge_rtnode_lookup: * * Look up a bridge route node for the specified destination. Compare the * vlan id or if zero then just return the first match. */ static struct bridge_rtnode * bridge_rtnode_lookup(struct bridge_softc *sc, const uint8_t *addr, uint16_t vlan) { struct bridge_rtnode *brt; uint32_t hash; int dir; BRIDGE_RT_LOCK_OR_NET_EPOCH_ASSERT(sc); hash = bridge_rthash(sc, addr); CK_LIST_FOREACH(brt, &sc->sc_rthash[hash], brt_hash) { dir = bridge_rtnode_addr_cmp(addr, brt->brt_addr); if (dir == 0 && (brt->brt_vlan == vlan || vlan == 0)) return (brt); if (dir > 0) return (NULL); } return (NULL); } /* * bridge_rtnode_insert: * * Insert the specified bridge node into the route table. We * assume the entry is not already in the table. */ static int bridge_rtnode_insert(struct bridge_softc *sc, struct bridge_rtnode *brt) { struct bridge_rtnode *lbrt; uint32_t hash; int dir; BRIDGE_RT_LOCK_ASSERT(sc); hash = bridge_rthash(sc, brt->brt_addr); lbrt = CK_LIST_FIRST(&sc->sc_rthash[hash]); if (lbrt == NULL) { CK_LIST_INSERT_HEAD(&sc->sc_rthash[hash], brt, brt_hash); goto out; } do { dir = bridge_rtnode_addr_cmp(brt->brt_addr, lbrt->brt_addr); if (dir == 0 && brt->brt_vlan == lbrt->brt_vlan) return (EEXIST); if (dir > 0) { CK_LIST_INSERT_BEFORE(lbrt, brt, brt_hash); goto out; } if (CK_LIST_NEXT(lbrt, brt_hash) == NULL) { CK_LIST_INSERT_AFTER(lbrt, brt, brt_hash); goto out; } lbrt = CK_LIST_NEXT(lbrt, brt_hash); } while (lbrt != NULL); #ifdef DIAGNOSTIC panic("bridge_rtnode_insert: impossible"); #endif out: CK_LIST_INSERT_HEAD(&sc->sc_rtlist, brt, brt_list); sc->sc_brtcnt++; return (0); } static void bridge_rtnode_destroy_cb(struct epoch_context *ctx) { struct bridge_rtnode *brt; brt = __containerof(ctx, struct bridge_rtnode, brt_epoch_ctx); CURVNET_SET(brt->brt_vnet); uma_zfree(V_bridge_rtnode_zone, brt); CURVNET_RESTORE(); } /* * bridge_rtnode_destroy: * * Destroy a bridge rtnode. */ static void bridge_rtnode_destroy(struct bridge_softc *sc, struct bridge_rtnode *brt) { BRIDGE_RT_LOCK_ASSERT(sc); CK_LIST_REMOVE(brt, brt_hash); CK_LIST_REMOVE(brt, brt_list); sc->sc_brtcnt--; brt->brt_dst->bif_addrcnt--; NET_EPOCH_CALL(bridge_rtnode_destroy_cb, &brt->brt_epoch_ctx); } /* * bridge_rtable_expire: * * Set the expiry time for all routes on an interface. */ static void bridge_rtable_expire(struct ifnet *ifp, int age) { struct bridge_softc *sc = ifp->if_bridge; struct bridge_rtnode *brt; CURVNET_SET(ifp->if_vnet); BRIDGE_RT_LOCK(sc); /* * If the age is zero then flush, otherwise set all the expiry times to * age for the interface */ if (age == 0) bridge_rtdelete(sc, ifp, IFBF_FLUSHDYN); else { CK_LIST_FOREACH(brt, &sc->sc_rtlist, brt_list) { /* Cap the expiry time to 'age' */ if (brt->brt_ifp == ifp && brt->brt_expire > time_uptime + age && (brt->brt_flags & IFBAF_TYPEMASK) == IFBAF_DYNAMIC) brt->brt_expire = time_uptime + age; } } BRIDGE_RT_UNLOCK(sc); CURVNET_RESTORE(); } /* * bridge_state_change: * * Callback from the bridgestp code when a port changes states. */ static void bridge_state_change(struct ifnet *ifp, int state) { struct bridge_softc *sc = ifp->if_bridge; static const char *stpstates[] = { "disabled", "listening", "learning", "forwarding", "blocking", "discarding" }; CURVNET_SET(ifp->if_vnet); if (V_log_stp) log(LOG_NOTICE, "%s: state changed to %s on %s\n", sc->sc_ifp->if_xname, stpstates[state], ifp->if_xname); CURVNET_RESTORE(); } /* * Send bridge packets through pfil if they are one of the types pfil can deal * with, or if they are ARP or REVARP. (pfil will pass ARP and REVARP without * question.) If *bifp or *ifp are NULL then packet filtering is skipped for * that interface. */ static int bridge_pfil(struct mbuf **mp, struct ifnet *bifp, struct ifnet *ifp, int dir) { int snap, error, i, hlen; struct ether_header *eh1, eh2; struct ip *ip; struct llc llc1; u_int16_t ether_type; pfil_return_t rv; snap = 0; error = -1; /* Default error if not error == 0 */ #if 0 /* we may return with the IP fields swapped, ensure its not shared */ KASSERT(M_WRITABLE(*mp), ("%s: modifying a shared mbuf", __func__)); #endif if (V_pfil_bridge == 0 && V_pfil_member == 0 && V_pfil_ipfw == 0) return (0); /* filtering is disabled */ i = min((*mp)->m_pkthdr.len, max_protohdr); if ((*mp)->m_len < i) { *mp = m_pullup(*mp, i); if (*mp == NULL) { printf("%s: m_pullup failed\n", __func__); return (-1); } } eh1 = mtod(*mp, struct ether_header *); ether_type = ntohs(eh1->ether_type); /* * Check for SNAP/LLC. */ if (ether_type < ETHERMTU) { struct llc *llc2 = (struct llc *)(eh1 + 1); if ((*mp)->m_len >= ETHER_HDR_LEN + 8 && llc2->llc_dsap == LLC_SNAP_LSAP && llc2->llc_ssap == LLC_SNAP_LSAP && llc2->llc_control == LLC_UI) { ether_type = htons(llc2->llc_un.type_snap.ether_type); snap = 1; } } /* * If we're trying to filter bridge traffic, don't look at anything * other than IP and ARP traffic. If the filter doesn't understand * IPv6, don't allow IPv6 through the bridge either. This is lame * since if we really wanted, say, an AppleTalk filter, we are hosed, * but of course we don't have an AppleTalk filter to begin with. * (Note that since pfil doesn't understand ARP it will pass *ALL* * ARP traffic.) */ switch (ether_type) { case ETHERTYPE_ARP: case ETHERTYPE_REVARP: if (V_pfil_ipfw_arp == 0) return (0); /* Automatically pass */ break; case ETHERTYPE_IP: #ifdef INET6 case ETHERTYPE_IPV6: #endif /* INET6 */ break; default: /* * Check to see if the user wants to pass non-ip * packets, these will not be checked by pfil(9) and * passed unconditionally so the default is to drop. */ if (V_pfil_onlyip) goto bad; } /* Run the packet through pfil before stripping link headers */ if (PFIL_HOOKED_OUT(V_link_pfil_head) && V_pfil_ipfw != 0 && dir == PFIL_OUT && ifp != NULL) { switch (pfil_mbuf_out(V_link_pfil_head, mp, ifp, NULL)) { case PFIL_DROPPED: return (EACCES); case PFIL_CONSUMED: return (0); } } /* Strip off the Ethernet header and keep a copy. */ m_copydata(*mp, 0, ETHER_HDR_LEN, (caddr_t) &eh2); m_adj(*mp, ETHER_HDR_LEN); /* Strip off snap header, if present */ if (snap) { m_copydata(*mp, 0, sizeof(struct llc), (caddr_t) &llc1); m_adj(*mp, sizeof(struct llc)); } /* * Check the IP header for alignment and errors */ if (dir == PFIL_IN) { switch (ether_type) { case ETHERTYPE_IP: error = bridge_ip_checkbasic(mp); break; #ifdef INET6 case ETHERTYPE_IPV6: error = bridge_ip6_checkbasic(mp); break; #endif /* INET6 */ default: error = 0; } if (error) goto bad; } error = 0; /* * Run the packet through pfil */ rv = PFIL_PASS; switch (ether_type) { case ETHERTYPE_IP: /* * Run pfil on the member interface and the bridge, both can * be skipped by clearing pfil_member or pfil_bridge. * * Keep the order: * in_if -> bridge_if -> out_if */ if (V_pfil_bridge && dir == PFIL_OUT && bifp != NULL && (rv = pfil_mbuf_out(V_inet_pfil_head, mp, bifp, NULL)) != PFIL_PASS) break; if (V_pfil_member && ifp != NULL) { rv = (dir == PFIL_OUT) ? pfil_mbuf_out(V_inet_pfil_head, mp, ifp, NULL) : pfil_mbuf_in(V_inet_pfil_head, mp, ifp, NULL); if (rv != PFIL_PASS) break; } if (V_pfil_bridge && dir == PFIL_IN && bifp != NULL && (rv = pfil_mbuf_in(V_inet_pfil_head, mp, bifp, NULL)) != PFIL_PASS) break; /* check if we need to fragment the packet */ /* bridge_fragment generates a mbuf chain of packets */ /* that already include eth headers */ if (V_pfil_member && ifp != NULL && dir == PFIL_OUT) { i = (*mp)->m_pkthdr.len; if (i > ifp->if_mtu) { error = bridge_fragment(ifp, mp, &eh2, snap, &llc1); return (error); } } /* Recalculate the ip checksum. */ ip = mtod(*mp, struct ip *); hlen = ip->ip_hl << 2; if (hlen < sizeof(struct ip)) goto bad; if (hlen > (*mp)->m_len) { if ((*mp = m_pullup(*mp, hlen)) == NULL) goto bad; ip = mtod(*mp, struct ip *); if (ip == NULL) goto bad; } ip->ip_sum = 0; if (hlen == sizeof(struct ip)) ip->ip_sum = in_cksum_hdr(ip); else ip->ip_sum = in_cksum(*mp, hlen); break; #ifdef INET6 case ETHERTYPE_IPV6: if (V_pfil_bridge && dir == PFIL_OUT && bifp != NULL && (rv = pfil_mbuf_out(V_inet6_pfil_head, mp, bifp, NULL)) != PFIL_PASS) break; if (V_pfil_member && ifp != NULL) { rv = (dir == PFIL_OUT) ? pfil_mbuf_out(V_inet6_pfil_head, mp, ifp, NULL) : pfil_mbuf_in(V_inet6_pfil_head, mp, ifp, NULL); if (rv != PFIL_PASS) break; } if (V_pfil_bridge && dir == PFIL_IN && bifp != NULL && (rv = pfil_mbuf_in(V_inet6_pfil_head, mp, bifp, NULL)) != PFIL_PASS) break; break; #endif } switch (rv) { case PFIL_CONSUMED: return (0); case PFIL_DROPPED: return (EACCES); default: break; } error = -1; /* * Finally, put everything back the way it was and return */ if (snap) { M_PREPEND(*mp, sizeof(struct llc), M_NOWAIT); if (*mp == NULL) return (error); bcopy(&llc1, mtod(*mp, caddr_t), sizeof(struct llc)); } M_PREPEND(*mp, ETHER_HDR_LEN, M_NOWAIT); if (*mp == NULL) return (error); bcopy(&eh2, mtod(*mp, caddr_t), ETHER_HDR_LEN); return (0); bad: m_freem(*mp); *mp = NULL; return (error); } /* * Perform basic checks on header size since * pfil assumes ip_input has already processed * it for it. Cut-and-pasted from ip_input.c. * Given how simple the IPv6 version is, * does the IPv4 version really need to be * this complicated? * * XXX Should we update ipstat here, or not? * XXX Right now we update ipstat but not * XXX csum_counter. */ static int bridge_ip_checkbasic(struct mbuf **mp) { struct mbuf *m = *mp; struct ip *ip; int len, hlen; u_short sum; if (*mp == NULL) return (-1); if (IP_HDR_ALIGNED_P(mtod(m, caddr_t)) == 0) { if ((m = m_copyup(m, sizeof(struct ip), (max_linkhdr + 3) & ~3)) == NULL) { /* XXXJRT new stat, please */ KMOD_IPSTAT_INC(ips_toosmall); goto bad; } } else if (__predict_false(m->m_len < sizeof (struct ip))) { if ((m = m_pullup(m, sizeof (struct ip))) == NULL) { KMOD_IPSTAT_INC(ips_toosmall); goto bad; } } ip = mtod(m, struct ip *); if (ip == NULL) goto bad; if (ip->ip_v != IPVERSION) { KMOD_IPSTAT_INC(ips_badvers); goto bad; } hlen = ip->ip_hl << 2; if (hlen < sizeof(struct ip)) { /* minimum header length */ KMOD_IPSTAT_INC(ips_badhlen); goto bad; } if (hlen > m->m_len) { if ((m = m_pullup(m, hlen)) == NULL) { KMOD_IPSTAT_INC(ips_badhlen); goto bad; } ip = mtod(m, struct ip *); if (ip == NULL) goto bad; } if (m->m_pkthdr.csum_flags & CSUM_IP_CHECKED) { sum = !(m->m_pkthdr.csum_flags & CSUM_IP_VALID); } else { if (hlen == sizeof(struct ip)) { sum = in_cksum_hdr(ip); } else { sum = in_cksum(m, hlen); } } if (sum) { KMOD_IPSTAT_INC(ips_badsum); goto bad; } /* Retrieve the packet length. */ len = ntohs(ip->ip_len); /* * Check for additional length bogosity */ if (len < hlen) { KMOD_IPSTAT_INC(ips_badlen); goto bad; } /* * Check that the amount of data in the buffers * is as at least much as the IP header would have us expect. * Drop packet if shorter than we expect. */ if (m->m_pkthdr.len < len) { KMOD_IPSTAT_INC(ips_tooshort); goto bad; } /* Checks out, proceed */ *mp = m; return (0); bad: *mp = m; return (-1); } #ifdef INET6 /* * Same as above, but for IPv6. * Cut-and-pasted from ip6_input.c. * XXX Should we update ip6stat, or not? */ static int bridge_ip6_checkbasic(struct mbuf **mp) { struct mbuf *m = *mp; struct ip6_hdr *ip6; /* * If the IPv6 header is not aligned, slurp it up into a new * mbuf with space for link headers, in the event we forward * it. Otherwise, if it is aligned, make sure the entire base * IPv6 header is in the first mbuf of the chain. */ if (IP6_HDR_ALIGNED_P(mtod(m, caddr_t)) == 0) { struct ifnet *inifp = m->m_pkthdr.rcvif; if ((m = m_copyup(m, sizeof(struct ip6_hdr), (max_linkhdr + 3) & ~3)) == NULL) { /* XXXJRT new stat, please */ IP6STAT_INC(ip6s_toosmall); in6_ifstat_inc(inifp, ifs6_in_hdrerr); goto bad; } } else if (__predict_false(m->m_len < sizeof(struct ip6_hdr))) { struct ifnet *inifp = m->m_pkthdr.rcvif; if ((m = m_pullup(m, sizeof(struct ip6_hdr))) == NULL) { IP6STAT_INC(ip6s_toosmall); in6_ifstat_inc(inifp, ifs6_in_hdrerr); goto bad; } } ip6 = mtod(m, struct ip6_hdr *); if ((ip6->ip6_vfc & IPV6_VERSION_MASK) != IPV6_VERSION) { IP6STAT_INC(ip6s_badvers); in6_ifstat_inc(m->m_pkthdr.rcvif, ifs6_in_hdrerr); goto bad; } /* Checks out, proceed */ *mp = m; return (0); bad: *mp = m; return (-1); } #endif /* INET6 */ /* * bridge_fragment: * * Fragment mbuf chain in multiple packets and prepend ethernet header. */ static int bridge_fragment(struct ifnet *ifp, struct mbuf **mp, struct ether_header *eh, int snap, struct llc *llc) { struct mbuf *m = *mp, *nextpkt = NULL, *mprev = NULL, *mcur = NULL; struct ip *ip; int error = -1; if (m->m_len < sizeof(struct ip) && (m = m_pullup(m, sizeof(struct ip))) == NULL) goto dropit; ip = mtod(m, struct ip *); m->m_pkthdr.csum_flags |= CSUM_IP; error = ip_fragment(ip, &m, ifp->if_mtu, ifp->if_hwassist); if (error) goto dropit; /* * Walk the chain and re-add the Ethernet header for * each mbuf packet. */ for (mcur = m; mcur; mcur = mcur->m_nextpkt) { nextpkt = mcur->m_nextpkt; mcur->m_nextpkt = NULL; if (snap) { M_PREPEND(mcur, sizeof(struct llc), M_NOWAIT); if (mcur == NULL) { error = ENOBUFS; if (mprev != NULL) mprev->m_nextpkt = nextpkt; goto dropit; } bcopy(llc, mtod(mcur, caddr_t),sizeof(struct llc)); } M_PREPEND(mcur, ETHER_HDR_LEN, M_NOWAIT); if (mcur == NULL) { error = ENOBUFS; if (mprev != NULL) mprev->m_nextpkt = nextpkt; goto dropit; } bcopy(eh, mtod(mcur, caddr_t), ETHER_HDR_LEN); /* * The previous two M_PREPEND could have inserted one or two * mbufs in front so we have to update the previous packet's * m_nextpkt. */ mcur->m_nextpkt = nextpkt; if (mprev != NULL) mprev->m_nextpkt = mcur; else { /* The first mbuf in the original chain needs to be * updated. */ *mp = mcur; } mprev = mcur; } KMOD_IPSTAT_INC(ips_fragmented); return (error); dropit: for (mcur = *mp; mcur; mcur = m) { /* droping the full packet chain */ m = mcur->m_nextpkt; m_freem(mcur); } return (error); } static void bridge_linkstate(struct ifnet *ifp) { struct bridge_softc *sc = ifp->if_bridge; struct bridge_iflist *bif; struct epoch_tracker et; NET_EPOCH_ENTER(et); bif = bridge_lookup_member_if(sc, ifp); if (bif == NULL) { NET_EPOCH_EXIT(et); return; } bridge_linkcheck(sc); bstp_linkstate(&bif->bif_stp); NET_EPOCH_EXIT(et); } static void bridge_linkcheck(struct bridge_softc *sc) { struct bridge_iflist *bif; int new_link, hasls; BRIDGE_LOCK_OR_NET_EPOCH_ASSERT(sc); new_link = LINK_STATE_DOWN; hasls = 0; /* Our link is considered up if at least one of our ports is active */ CK_LIST_FOREACH(bif, &sc->sc_iflist, bif_next) { if (bif->bif_ifp->if_capabilities & IFCAP_LINKSTATE) hasls++; if (bif->bif_ifp->if_link_state == LINK_STATE_UP) { new_link = LINK_STATE_UP; break; } } if (!CK_LIST_EMPTY(&sc->sc_iflist) && !hasls) { /* If no interfaces support link-state then we default to up */ new_link = LINK_STATE_UP; } if_link_state_change(sc->sc_ifp, new_link); } diff --git a/sys/net/if_bridgevar.h b/sys/net/if_bridgevar.h index 69b9aa33f133..5696e8a62ae0 100644 --- a/sys/net/if_bridgevar.h +++ b/sys/net/if_bridgevar.h @@ -1,324 +1,326 @@ /* $NetBSD: if_bridgevar.h,v 1.4 2003/07/08 07:13:50 itojun Exp $ */ /* * SPDX-License-Identifier: BSD-4-Clause * * Copyright 2001 Wasabi Systems, Inc. * All rights reserved. * * Written by Jason R. Thorpe for Wasabi Systems, Inc. * * Redistribution and use in source and binary forms, with or without * modification, are permitted provided that the following conditions * are met: * 1. Redistributions of source code must retain the above copyright * notice, this list of conditions and the following disclaimer. * 2. Redistributions in binary form must reproduce the above copyright * notice, this list of conditions and the following disclaimer in the * documentation and/or other materials provided with the distribution. * 3. All advertising materials mentioning features or use of this software * must display the following acknowledgement: * This product includes software developed for the NetBSD Project by * Wasabi Systems, Inc. * 4. The name of Wasabi Systems, Inc. may not be used to endorse * or promote products derived from this software without specific prior * written permission. * * THIS SOFTWARE IS PROVIDED BY WASABI SYSTEMS, INC. ``AS IS'' AND * ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED * TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR * PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL WASABI SYSTEMS, INC * BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR * CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF * SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS * INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN * CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) * ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE * POSSIBILITY OF SUCH DAMAGE. */ /* * Copyright (c) 1999, 2000 Jason L. Wright (jason@thought.net) * All rights reserved. * * Redistribution and use in source and binary forms, with or without * modification, are permitted provided that the following conditions * are met: * 1. Redistributions of source code must retain the above copyright * notice, this list of conditions and the following disclaimer. * 2. Redistributions in binary form must reproduce the above copyright * notice, this list of conditions and the following disclaimer in the * documentation and/or other materials provided with the distribution. * 3. All advertising materials mentioning features or use of this software * must display the following acknowledgement: * This product includes software developed by Jason L. Wright * 4. The name of the author may not be used to endorse or promote products * derived from this software without specific prior written permission. * * THIS SOFTWARE IS PROVIDED BY THE AUTHOR ``AS IS'' AND ANY EXPRESS OR * IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE IMPLIED * WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE ARE * DISCLAIMED. IN NO EVENT SHALL THE AUTHOR BE LIABLE FOR ANY DIRECT, * INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES * (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR * SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, * STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN * ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE * POSSIBILITY OF SUCH DAMAGE. * * OpenBSD: if_bridge.h,v 1.14 2001/03/22 03:48:29 jason Exp * * $FreeBSD$ */ /* * Data structure and control definitions for bridge interfaces. */ #include #include #include #include #include #include /* * Commands used in the SIOCSDRVSPEC ioctl. Note the lookup of the * bridge interface itself is keyed off the ifdrv structure. */ #define BRDGADD 0 /* add bridge member (ifbreq) */ #define BRDGDEL 1 /* delete bridge member (ifbreq) */ #define BRDGGIFFLGS 2 /* get member if flags (ifbreq) */ #define BRDGSIFFLGS 3 /* set member if flags (ifbreq) */ #define BRDGSCACHE 4 /* set cache size (ifbrparam) */ #define BRDGGCACHE 5 /* get cache size (ifbrparam) */ #define BRDGGIFS 6 /* get member list (ifbifconf) */ #define BRDGRTS 7 /* get address list (ifbaconf) */ #define BRDGSADDR 8 /* set static address (ifbareq) */ #define BRDGSTO 9 /* set cache timeout (ifbrparam) */ #define BRDGGTO 10 /* get cache timeout (ifbrparam) */ #define BRDGDADDR 11 /* delete address (ifbareq) */ #define BRDGFLUSH 12 /* flush address cache (ifbreq) */ #define BRDGGPRI 13 /* get priority (ifbrparam) */ #define BRDGSPRI 14 /* set priority (ifbrparam) */ #define BRDGGHT 15 /* get hello time (ifbrparam) */ #define BRDGSHT 16 /* set hello time (ifbrparam) */ #define BRDGGFD 17 /* get forward delay (ifbrparam) */ #define BRDGSFD 18 /* set forward delay (ifbrparam) */ #define BRDGGMA 19 /* get max age (ifbrparam) */ #define BRDGSMA 20 /* set max age (ifbrparam) */ #define BRDGSIFPRIO 21 /* set if priority (ifbreq) */ #define BRDGSIFCOST 22 /* set if path cost (ifbreq) */ #define BRDGADDS 23 /* add bridge span member (ifbreq) */ #define BRDGDELS 24 /* delete bridge span member (ifbreq) */ #define BRDGPARAM 25 /* get bridge STP params (ifbropreq) */ #define BRDGGRTE 26 /* get cache drops (ifbrparam) */ #define BRDGGIFSSTP 27 /* get member STP params list * (ifbpstpconf) */ #define BRDGSPROTO 28 /* set protocol (ifbrparam) */ #define BRDGSTXHC 29 /* set tx hold count (ifbrparam) */ #define BRDGSIFAMAX 30 /* set max interface addrs (ifbreq) */ /* * Generic bridge control request. */ struct ifbreq { char ifbr_ifsname[IFNAMSIZ]; /* member if name */ uint32_t ifbr_ifsflags; /* member if flags */ uint32_t ifbr_stpflags; /* member if STP flags */ uint32_t ifbr_path_cost; /* member if STP cost */ uint8_t ifbr_portno; /* member if port number */ uint8_t ifbr_priority; /* member if STP priority */ uint8_t ifbr_proto; /* member if STP protocol */ uint8_t ifbr_role; /* member if STP role */ uint8_t ifbr_state; /* member if STP state */ uint32_t ifbr_addrcnt; /* member if addr number */ uint32_t ifbr_addrmax; /* member if addr max */ uint32_t ifbr_addrexceeded; /* member if addr violations */ uint8_t pad[32]; }; /* BRDGGIFFLAGS, BRDGSIFFLAGS */ #define IFBIF_LEARNING 0x0001 /* if can learn */ #define IFBIF_DISCOVER 0x0002 /* if sends packets w/ unknown dest. */ #define IFBIF_STP 0x0004 /* if participates in spanning tree */ #define IFBIF_SPAN 0x0008 /* if is a span port */ #define IFBIF_STICKY 0x0010 /* if learned addresses stick */ #define IFBIF_BSTP_EDGE 0x0020 /* member stp edge port */ #define IFBIF_BSTP_AUTOEDGE 0x0040 /* member stp autoedge enabled */ #define IFBIF_BSTP_PTP 0x0080 /* member stp point to point */ #define IFBIF_BSTP_AUTOPTP 0x0100 /* member stp autoptp enabled */ #define IFBIF_BSTP_ADMEDGE 0x0200 /* member stp admin edge enabled */ #define IFBIF_BSTP_ADMCOST 0x0400 /* member stp admin path cost */ #define IFBIF_PRIVATE 0x0800 /* if is a private segment */ #define IFBIFBITS "\020\001LEARNING\002DISCOVER\003STP\004SPAN" \ "\005STICKY\014PRIVATE\006EDGE\007AUTOEDGE\010PTP" \ "\011AUTOPTP" #define IFBIFMASK ~(IFBIF_BSTP_EDGE|IFBIF_BSTP_AUTOEDGE|IFBIF_BSTP_PTP| \ IFBIF_BSTP_AUTOPTP|IFBIF_BSTP_ADMEDGE| \ IFBIF_BSTP_ADMCOST) /* not saved */ /* BRDGFLUSH */ #define IFBF_FLUSHDYN 0x00 /* flush learned addresses only */ #define IFBF_FLUSHALL 0x01 /* flush all addresses */ /* * Interface list structure. */ struct ifbifconf { uint32_t ifbic_len; /* buffer size */ union { caddr_t ifbicu_buf; struct ifbreq *ifbicu_req; } ifbic_ifbicu; #define ifbic_buf ifbic_ifbicu.ifbicu_buf #define ifbic_req ifbic_ifbicu.ifbicu_req }; /* * Bridge address request. */ struct ifbareq { char ifba_ifsname[IFNAMSIZ]; /* member if name */ unsigned long ifba_expire; /* address expire time */ uint8_t ifba_flags; /* address flags */ uint8_t ifba_dst[ETHER_ADDR_LEN];/* destination address */ uint16_t ifba_vlan; /* vlan id */ }; #define IFBAF_TYPEMASK 0x03 /* address type mask */ #define IFBAF_DYNAMIC 0x00 /* dynamically learned address */ #define IFBAF_STATIC 0x01 /* static address */ #define IFBAF_STICKY 0x02 /* sticky address */ #define IFBAFBITS "\020\1STATIC\2STICKY" /* * Address list structure. */ struct ifbaconf { uint32_t ifbac_len; /* buffer size */ union { caddr_t ifbacu_buf; struct ifbareq *ifbacu_req; } ifbac_ifbacu; #define ifbac_buf ifbac_ifbacu.ifbacu_buf #define ifbac_req ifbac_ifbacu.ifbacu_req }; /* * Bridge parameter structure. */ struct ifbrparam { union { uint32_t ifbrpu_int32; uint16_t ifbrpu_int16; uint8_t ifbrpu_int8; } ifbrp_ifbrpu; }; #define ifbrp_csize ifbrp_ifbrpu.ifbrpu_int32 /* cache size */ #define ifbrp_ctime ifbrp_ifbrpu.ifbrpu_int32 /* cache time (sec) */ #define ifbrp_prio ifbrp_ifbrpu.ifbrpu_int16 /* bridge priority */ #define ifbrp_proto ifbrp_ifbrpu.ifbrpu_int8 /* bridge protocol */ #define ifbrp_txhc ifbrp_ifbrpu.ifbrpu_int8 /* bpdu tx holdcount */ #define ifbrp_hellotime ifbrp_ifbrpu.ifbrpu_int8 /* hello time (sec) */ #define ifbrp_fwddelay ifbrp_ifbrpu.ifbrpu_int8 /* fwd time (sec) */ #define ifbrp_maxage ifbrp_ifbrpu.ifbrpu_int8 /* max age (sec) */ #define ifbrp_cexceeded ifbrp_ifbrpu.ifbrpu_int32 /* # of cache dropped * adresses */ /* * Bridge current operational parameters structure. */ struct ifbropreq { uint8_t ifbop_holdcount; uint8_t ifbop_maxage; uint8_t ifbop_hellotime; uint8_t ifbop_fwddelay; uint8_t ifbop_protocol; uint16_t ifbop_priority; uint16_t ifbop_root_port; uint32_t ifbop_root_path_cost; uint64_t ifbop_bridgeid; uint64_t ifbop_designated_root; uint64_t ifbop_designated_bridge; struct timeval ifbop_last_tc_time; }; /* * Bridge member operational STP params structure. */ struct ifbpstpreq { uint8_t ifbp_portno; /* bp STP port number */ uint32_t ifbp_fwd_trans; /* bp STP fwd transitions */ uint32_t ifbp_design_cost; /* bp STP designated cost */ uint32_t ifbp_design_port; /* bp STP designated port */ uint64_t ifbp_design_bridge; /* bp STP designated bridge */ uint64_t ifbp_design_root; /* bp STP designated root */ }; /* * Bridge STP ports list structure. */ struct ifbpstpconf { uint32_t ifbpstp_len; /* buffer size */ union { caddr_t ifbpstpu_buf; struct ifbpstpreq *ifbpstpu_req; } ifbpstp_ifbpstpu; #define ifbpstp_buf ifbpstp_ifbpstpu.ifbpstpu_buf #define ifbpstp_req ifbpstp_ifbpstpu.ifbpstpu_req }; #define STP_STATES \ "disabled", \ "listening", \ "learning", \ "forwarding", \ "blocking", \ "discarding" #define STP_PROTOS \ "stp", \ "-", \ "rstp" #define STP_ROLES \ "disabled", \ "root", \ "designated", \ "alternate", \ "backup" #define PV2ID(pv, epri, eaddr) do { \ epri = pv >> 48; \ eaddr[0] = pv >> 40; \ eaddr[1] = pv >> 32; \ eaddr[2] = pv >> 24; \ eaddr[3] = pv >> 16; \ eaddr[4] = pv >> 8; \ eaddr[5] = pv >> 0; \ } while (0) #ifdef _KERNEL #define BRIDGE_INPUT(_ifp, _m) do { \ KASSERT((_ifp)->if_bridge_input != NULL, \ ("%s: if_bridge not loaded!", __func__)); \ _m = (*(_ifp)->if_bridge_input)(_ifp, _m); \ - if (_m != NULL) \ + if (_m != NULL) { \ _ifp = _m->m_pkthdr.rcvif; \ + m->m_flags &= ~M_BRIDGE_INJECT; \ + } \ } while (0) #define BRIDGE_OUTPUT(_ifp, _m, _err) do { \ KASSERT((_ifp)->if_bridge_output != NULL, \ ("%s: if_bridge not loaded!", __func__)); \ _err = (*(_ifp)->if_bridge_output)(_ifp, _m, NULL, NULL); \ } while (0) extern void (*bridge_dn_p)(struct mbuf *, struct ifnet *); #endif /* _KERNEL */ diff --git a/sys/net/if_ethersubr.c b/sys/net/if_ethersubr.c index 34ff4ac22e7f..839bae8e9d43 100644 --- a/sys/net/if_ethersubr.c +++ b/sys/net/if_ethersubr.c @@ -1,1499 +1,1504 @@ /*- * SPDX-License-Identifier: BSD-3-Clause * * Copyright (c) 1982, 1989, 1993 * The Regents of the University of California. All rights reserved. * * Redistribution and use in source and binary forms, with or without * modification, are permitted provided that the following conditions * are met: * 1. Redistributions of source code must retain the above copyright * notice, this list of conditions and the following disclaimer. * 2. Redistributions in binary form must reproduce the above copyright * notice, this list of conditions and the following disclaimer in the * documentation and/or other materials provided with the distribution. * 3. Neither the name of the University nor the names of its contributors * may be used to endorse or promote products derived from this software * without specific prior written permission. * * THIS SOFTWARE IS PROVIDED BY THE REGENTS AND CONTRIBUTORS ``AS IS'' AND * ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE * ARE DISCLAIMED. IN NO EVENT SHALL THE REGENTS OR CONTRIBUTORS BE LIABLE * FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL * DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS * OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT * LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY * OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF * SUCH DAMAGE. * * @(#)if_ethersubr.c 8.1 (Berkeley) 6/10/93 * $FreeBSD$ */ #include "opt_inet.h" #include "opt_inet6.h" #include "opt_netgraph.h" #include "opt_mbuf_profiling.h" #include "opt_rss.h" #include #include #include #include #include #include #include #include #include #include #include #include #include #include #include #include #include #include #include #include #include #include #include #include #include #include #include #include #include #include #include #include #include #include #include #include #include #if defined(INET) || defined(INET6) #include #include #include #include #include #endif #ifdef INET6 #include #endif #include #include #ifdef CTASSERT CTASSERT(sizeof (struct ether_header) == ETHER_ADDR_LEN * 2 + 2); CTASSERT(sizeof (struct ether_addr) == ETHER_ADDR_LEN); #endif VNET_DEFINE(pfil_head_t, link_pfil_head); /* Packet filter hooks */ /* netgraph node hooks for ng_ether(4) */ void (*ng_ether_input_p)(struct ifnet *ifp, struct mbuf **mp); void (*ng_ether_input_orphan_p)(struct ifnet *ifp, struct mbuf *m); int (*ng_ether_output_p)(struct ifnet *ifp, struct mbuf **mp); void (*ng_ether_attach_p)(struct ifnet *ifp); void (*ng_ether_detach_p)(struct ifnet *ifp); void (*vlan_input_p)(struct ifnet *, struct mbuf *); /* if_bridge(4) support */ void (*bridge_dn_p)(struct mbuf *, struct ifnet *); /* if_lagg(4) support */ struct mbuf *(*lagg_input_ethernet_p)(struct ifnet *, struct mbuf *); static const u_char etherbroadcastaddr[ETHER_ADDR_LEN] = { 0xff, 0xff, 0xff, 0xff, 0xff, 0xff }; static int ether_resolvemulti(struct ifnet *, struct sockaddr **, struct sockaddr *); static int ether_requestencap(struct ifnet *, struct if_encap_req *); #define senderr(e) do { error = (e); goto bad;} while (0) static void update_mbuf_csumflags(struct mbuf *src, struct mbuf *dst) { int csum_flags = 0; if (src->m_pkthdr.csum_flags & CSUM_IP) csum_flags |= (CSUM_IP_CHECKED|CSUM_IP_VALID); if (src->m_pkthdr.csum_flags & CSUM_DELAY_DATA) csum_flags |= (CSUM_DATA_VALID|CSUM_PSEUDO_HDR); if (src->m_pkthdr.csum_flags & CSUM_SCTP) csum_flags |= CSUM_SCTP_VALID; dst->m_pkthdr.csum_flags |= csum_flags; if (csum_flags & CSUM_DATA_VALID) dst->m_pkthdr.csum_data = 0xffff; } /* * Handle link-layer encapsulation requests. */ static int ether_requestencap(struct ifnet *ifp, struct if_encap_req *req) { struct ether_header *eh; struct arphdr *ah; uint16_t etype; const u_char *lladdr; if (req->rtype != IFENCAP_LL) return (EOPNOTSUPP); if (req->bufsize < ETHER_HDR_LEN) return (ENOMEM); eh = (struct ether_header *)req->buf; lladdr = req->lladdr; req->lladdr_off = 0; switch (req->family) { case AF_INET: etype = htons(ETHERTYPE_IP); break; case AF_INET6: etype = htons(ETHERTYPE_IPV6); break; case AF_ARP: ah = (struct arphdr *)req->hdata; ah->ar_hrd = htons(ARPHRD_ETHER); switch(ntohs(ah->ar_op)) { case ARPOP_REVREQUEST: case ARPOP_REVREPLY: etype = htons(ETHERTYPE_REVARP); break; case ARPOP_REQUEST: case ARPOP_REPLY: default: etype = htons(ETHERTYPE_ARP); break; } if (req->flags & IFENCAP_FLAG_BROADCAST) lladdr = ifp->if_broadcastaddr; break; default: return (EAFNOSUPPORT); } memcpy(&eh->ether_type, &etype, sizeof(eh->ether_type)); memcpy(eh->ether_dhost, lladdr, ETHER_ADDR_LEN); memcpy(eh->ether_shost, IF_LLADDR(ifp), ETHER_ADDR_LEN); req->bufsize = sizeof(struct ether_header); return (0); } static int ether_resolve_addr(struct ifnet *ifp, struct mbuf *m, const struct sockaddr *dst, struct route *ro, u_char *phdr, uint32_t *pflags, struct llentry **plle) { uint32_t lleflags = 0; int error = 0; #if defined(INET) || defined(INET6) struct ether_header *eh = (struct ether_header *)phdr; uint16_t etype; #endif if (plle) *plle = NULL; switch (dst->sa_family) { #ifdef INET case AF_INET: if ((m->m_flags & (M_BCAST | M_MCAST)) == 0) error = arpresolve(ifp, 0, m, dst, phdr, &lleflags, plle); else { if (m->m_flags & M_BCAST) memcpy(eh->ether_dhost, ifp->if_broadcastaddr, ETHER_ADDR_LEN); else { const struct in_addr *a; a = &(((const struct sockaddr_in *)dst)->sin_addr); ETHER_MAP_IP_MULTICAST(a, eh->ether_dhost); } etype = htons(ETHERTYPE_IP); memcpy(&eh->ether_type, &etype, sizeof(etype)); memcpy(eh->ether_shost, IF_LLADDR(ifp), ETHER_ADDR_LEN); } break; #endif #ifdef INET6 case AF_INET6: if ((m->m_flags & M_MCAST) == 0) { int af = RO_GET_FAMILY(ro, dst); error = nd6_resolve(ifp, LLE_SF(af, 0), m, dst, phdr, &lleflags, plle); } else { const struct in6_addr *a6; a6 = &(((const struct sockaddr_in6 *)dst)->sin6_addr); ETHER_MAP_IPV6_MULTICAST(a6, eh->ether_dhost); etype = htons(ETHERTYPE_IPV6); memcpy(&eh->ether_type, &etype, sizeof(etype)); memcpy(eh->ether_shost, IF_LLADDR(ifp), ETHER_ADDR_LEN); } break; #endif default: if_printf(ifp, "can't handle af%d\n", dst->sa_family); if (m != NULL) m_freem(m); return (EAFNOSUPPORT); } if (error == EHOSTDOWN) { if (ro != NULL && (ro->ro_flags & RT_HAS_GW) != 0) error = EHOSTUNREACH; } if (error != 0) return (error); *pflags = RT_MAY_LOOP; if (lleflags & LLE_IFADDR) *pflags |= RT_L2_ME; return (0); } /* * Ethernet output routine. * Encapsulate a packet of type family for the local net. * Use trailer local net encapsulation if enough data in first * packet leaves a multiple of 512 bytes of data in remainder. */ int ether_output(struct ifnet *ifp, struct mbuf *m, const struct sockaddr *dst, struct route *ro) { int error = 0; char linkhdr[ETHER_HDR_LEN], *phdr; struct ether_header *eh; struct pf_mtag *t; bool loop_copy; int hlen; /* link layer header length */ uint32_t pflags; struct llentry *lle = NULL; int addref = 0; phdr = NULL; pflags = 0; if (ro != NULL) { /* XXX BPF uses ro_prepend */ if (ro->ro_prepend != NULL) { phdr = ro->ro_prepend; hlen = ro->ro_plen; } else if (!(m->m_flags & (M_BCAST | M_MCAST))) { if ((ro->ro_flags & RT_LLE_CACHE) != 0) { lle = ro->ro_lle; if (lle != NULL && (lle->la_flags & LLE_VALID) == 0) { LLE_FREE(lle); lle = NULL; /* redundant */ ro->ro_lle = NULL; } if (lle == NULL) { /* if we lookup, keep cache */ addref = 1; } else /* * Notify LLE code that * the entry was used * by datapath. */ llentry_provide_feedback(lle); } if (lle != NULL) { phdr = lle->r_linkdata; hlen = lle->r_hdrlen; pflags = lle->r_flags; } } } #ifdef MAC error = mac_ifnet_check_transmit(ifp, m); if (error) senderr(error); #endif M_PROFILE(m); if (ifp->if_flags & IFF_MONITOR) senderr(ENETDOWN); if (!((ifp->if_flags & IFF_UP) && (ifp->if_drv_flags & IFF_DRV_RUNNING))) senderr(ENETDOWN); if (phdr == NULL) { /* No prepend data supplied. Try to calculate ourselves. */ phdr = linkhdr; hlen = ETHER_HDR_LEN; error = ether_resolve_addr(ifp, m, dst, ro, phdr, &pflags, addref ? &lle : NULL); if (addref && lle != NULL) ro->ro_lle = lle; if (error != 0) return (error == EWOULDBLOCK ? 0 : error); } if ((pflags & RT_L2_ME) != 0) { update_mbuf_csumflags(m, m); return (if_simloop(ifp, m, RO_GET_FAMILY(ro, dst), 0)); } loop_copy = (pflags & RT_MAY_LOOP) != 0; /* * Add local net header. If no space in first mbuf, * allocate another. * * Note that we do prepend regardless of RT_HAS_HEADER flag. * This is done because BPF code shifts m_data pointer * to the end of ethernet header prior to calling if_output(). */ M_PREPEND(m, hlen, M_NOWAIT); if (m == NULL) senderr(ENOBUFS); if ((pflags & RT_HAS_HEADER) == 0) { eh = mtod(m, struct ether_header *); memcpy(eh, phdr, hlen); } /* * If a simplex interface, and the packet is being sent to our * Ethernet address or a broadcast address, loopback a copy. * XXX To make a simplex device behave exactly like a duplex * device, we should copy in the case of sending to our own * ethernet address (thus letting the original actually appear * on the wire). However, we don't do that here for security * reasons and compatibility with the original behavior. */ if ((m->m_flags & M_BCAST) && loop_copy && (ifp->if_flags & IFF_SIMPLEX) && ((t = pf_find_mtag(m)) == NULL || !t->routed)) { struct mbuf *n; /* * Because if_simloop() modifies the packet, we need a * writable copy through m_dup() instead of a readonly * one as m_copy[m] would give us. The alternative would * be to modify if_simloop() to handle the readonly mbuf, * but performancewise it is mostly equivalent (trading * extra data copying vs. extra locking). * * XXX This is a local workaround. A number of less * often used kernel parts suffer from the same bug. * See PR kern/105943 for a proposed general solution. */ if ((n = m_dup(m, M_NOWAIT)) != NULL) { update_mbuf_csumflags(m, n); (void)if_simloop(ifp, n, RO_GET_FAMILY(ro, dst), hlen); } else if_inc_counter(ifp, IFCOUNTER_IQDROPS, 1); } /* * Bridges require special output handling. */ if (ifp->if_bridge) { BRIDGE_OUTPUT(ifp, m, error); return (error); } #if defined(INET) || defined(INET6) if (ifp->if_carp && (error = (*carp_output_p)(ifp, m, dst))) goto bad; #endif /* Handle ng_ether(4) processing, if any */ if (ifp->if_l2com != NULL) { KASSERT(ng_ether_output_p != NULL, ("ng_ether_output_p is NULL")); if ((error = (*ng_ether_output_p)(ifp, &m)) != 0) { bad: if (m != NULL) m_freem(m); return (error); } if (m == NULL) return (0); } /* Continue with link-layer output */ return ether_output_frame(ifp, m); } static bool ether_set_pcp(struct mbuf **mp, struct ifnet *ifp, uint8_t pcp) { struct ether_8021q_tag qtag; struct ether_header *eh; eh = mtod(*mp, struct ether_header *); if (ntohs(eh->ether_type) == ETHERTYPE_VLAN || ntohs(eh->ether_type) == ETHERTYPE_QINQ) return (true); qtag.vid = 0; qtag.pcp = pcp; qtag.proto = ETHERTYPE_VLAN; if (ether_8021q_frame(mp, ifp, ifp, &qtag)) return (true); if_inc_counter(ifp, IFCOUNTER_OERRORS, 1); return (false); } /* * Ethernet link layer output routine to send a raw frame to the device. * * This assumes that the 14 byte Ethernet header is present and contiguous * in the first mbuf (if BRIDGE'ing). */ int ether_output_frame(struct ifnet *ifp, struct mbuf *m) { uint8_t pcp; pcp = ifp->if_pcp; if (pcp != IFNET_PCP_NONE && ifp->if_type != IFT_L2VLAN && !ether_set_pcp(&m, ifp, pcp)) return (0); if (PFIL_HOOKED_OUT(V_link_pfil_head)) switch (pfil_mbuf_out(V_link_pfil_head, &m, ifp, NULL)) { case PFIL_DROPPED: return (EACCES); case PFIL_CONSUMED: return (0); } #ifdef EXPERIMENTAL #if defined(INET6) && defined(INET) /* draft-ietf-6man-ipv6only-flag */ /* Catch ETHERTYPE_IP, and ETHERTYPE_[REV]ARP if we are v6-only. */ if ((ND_IFINFO(ifp)->flags & ND6_IFF_IPV6_ONLY_MASK) != 0) { struct ether_header *eh; eh = mtod(m, struct ether_header *); switch (ntohs(eh->ether_type)) { case ETHERTYPE_IP: case ETHERTYPE_ARP: case ETHERTYPE_REVARP: m_freem(m); return (EAFNOSUPPORT); /* NOTREACHED */ break; }; } #endif #endif /* * Queue message on interface, update output statistics if successful, * and start output if interface not yet active. * * If KMSAN is enabled, use it to verify that the data does not contain * any uninitialized bytes. */ kmsan_check_mbuf(m, "ether_output"); return ((ifp->if_transmit)(ifp, m)); } /* * Process a received Ethernet packet; the packet is in the * mbuf chain m with the ethernet header at the front. */ static void ether_input_internal(struct ifnet *ifp, struct mbuf *m) { struct ether_header *eh; u_short etype; if ((ifp->if_flags & IFF_UP) == 0) { m_freem(m); return; } #ifdef DIAGNOSTIC if ((ifp->if_drv_flags & IFF_DRV_RUNNING) == 0) { if_printf(ifp, "discard frame at !IFF_DRV_RUNNING\n"); m_freem(m); return; } #endif if (m->m_len < ETHER_HDR_LEN) { /* XXX maybe should pullup? */ if_printf(ifp, "discard frame w/o leading ethernet " "header (len %u pkt len %u)\n", m->m_len, m->m_pkthdr.len); if_inc_counter(ifp, IFCOUNTER_IERRORS, 1); m_freem(m); return; } eh = mtod(m, struct ether_header *); etype = ntohs(eh->ether_type); random_harvest_queue_ether(m, sizeof(*m)); #ifdef EXPERIMENTAL #if defined(INET6) && defined(INET) /* draft-ietf-6man-ipv6only-flag */ /* Catch ETHERTYPE_IP, and ETHERTYPE_[REV]ARP if we are v6-only. */ if ((ND_IFINFO(ifp)->flags & ND6_IFF_IPV6_ONLY_MASK) != 0) { switch (etype) { case ETHERTYPE_IP: case ETHERTYPE_ARP: case ETHERTYPE_REVARP: m_freem(m); return; /* NOTREACHED */ break; }; } #endif #endif CURVNET_SET_QUIET(ifp->if_vnet); if (ETHER_IS_MULTICAST(eh->ether_dhost)) { if (ETHER_IS_BROADCAST(eh->ether_dhost)) m->m_flags |= M_BCAST; else m->m_flags |= M_MCAST; if_inc_counter(ifp, IFCOUNTER_IMCASTS, 1); } #ifdef MAC /* * Tag the mbuf with an appropriate MAC label before any other * consumers can get to it. */ mac_ifnet_create_mbuf(ifp, m); #endif /* * Give bpf a chance at the packet. */ ETHER_BPF_MTAP(ifp, m); /* * If the CRC is still on the packet, trim it off. We do this once * and once only in case we are re-entered. Nothing else on the * Ethernet receive path expects to see the FCS. */ if (m->m_flags & M_HASFCS) { m_adj(m, -ETHER_CRC_LEN); m->m_flags &= ~M_HASFCS; } if (!(ifp->if_capenable & IFCAP_HWSTATS)) if_inc_counter(ifp, IFCOUNTER_IBYTES, m->m_pkthdr.len); /* Allow monitor mode to claim this frame, after stats are updated. */ if (ifp->if_flags & IFF_MONITOR) { m_freem(m); CURVNET_RESTORE(); return; } /* Handle input from a lagg(4) port */ if (ifp->if_type == IFT_IEEE8023ADLAG) { KASSERT(lagg_input_ethernet_p != NULL, ("%s: if_lagg not loaded!", __func__)); m = (*lagg_input_ethernet_p)(ifp, m); if (m != NULL) ifp = m->m_pkthdr.rcvif; else { CURVNET_RESTORE(); return; } } /* * If the hardware did not process an 802.1Q tag, do this now, * to allow 802.1P priority frames to be passed to the main input * path correctly. */ if ((m->m_flags & M_VLANTAG) == 0 && ((etype == ETHERTYPE_VLAN) || (etype == ETHERTYPE_QINQ))) { struct ether_vlan_header *evl; if (m->m_len < sizeof(*evl) && (m = m_pullup(m, sizeof(*evl))) == NULL) { #ifdef DIAGNOSTIC if_printf(ifp, "cannot pullup VLAN header\n"); #endif if_inc_counter(ifp, IFCOUNTER_IERRORS, 1); CURVNET_RESTORE(); return; } evl = mtod(m, struct ether_vlan_header *); m->m_pkthdr.ether_vtag = ntohs(evl->evl_tag); m->m_flags |= M_VLANTAG; bcopy((char *)evl, (char *)evl + ETHER_VLAN_ENCAP_LEN, ETHER_HDR_LEN - ETHER_TYPE_LEN); m_adj(m, ETHER_VLAN_ENCAP_LEN); eh = mtod(m, struct ether_header *); } M_SETFIB(m, ifp->if_fib); /* Allow ng_ether(4) to claim this frame. */ if (ifp->if_l2com != NULL) { KASSERT(ng_ether_input_p != NULL, ("%s: ng_ether_input_p is NULL", __func__)); m->m_flags &= ~M_PROMISC; (*ng_ether_input_p)(ifp, &m); if (m == NULL) { CURVNET_RESTORE(); return; } eh = mtod(m, struct ether_header *); } /* * Allow if_bridge(4) to claim this frame. + * * The BRIDGE_INPUT() macro will update ifp if the bridge changed it * and the frame should be delivered locally. + * + * If M_BRIDGE_INJECT is set, the packet was received directly by the + * bridge via netmap, so "ifp" is the bridge itself and the packet + * should be re-examined. */ - if (ifp->if_bridge != NULL) { + if (ifp->if_bridge != NULL || (m->m_flags & M_BRIDGE_INJECT) != 0) { m->m_flags &= ~M_PROMISC; BRIDGE_INPUT(ifp, m); if (m == NULL) { CURVNET_RESTORE(); return; } eh = mtod(m, struct ether_header *); } #if defined(INET) || defined(INET6) /* * Clear M_PROMISC on frame so that carp(4) will see it when the * mbuf flows up to Layer 3. * FreeBSD's implementation of carp(4) uses the inprotosw * to dispatch IPPROTO_CARP. carp(4) also allocates its own * Ethernet addresses of the form 00:00:5e:00:01:xx, which * is outside the scope of the M_PROMISC test below. * TODO: Maintain a hash table of ethernet addresses other than * ether_dhost which may be active on this ifp. */ if (ifp->if_carp && (*carp_forus_p)(ifp, eh->ether_dhost)) { m->m_flags &= ~M_PROMISC; } else #endif { /* * If the frame received was not for our MAC address, set the * M_PROMISC flag on the mbuf chain. The frame may need to * be seen by the rest of the Ethernet input path in case of * re-entry (e.g. bridge, vlan, netgraph) but should not be * seen by upper protocol layers. */ if (!ETHER_IS_MULTICAST(eh->ether_dhost) && bcmp(IF_LLADDR(ifp), eh->ether_dhost, ETHER_ADDR_LEN) != 0) m->m_flags |= M_PROMISC; } ether_demux(ifp, m); CURVNET_RESTORE(); } /* * Ethernet input dispatch; by default, direct dispatch here regardless of * global configuration. However, if RSS is enabled, hook up RSS affinity * so that when deferred or hybrid dispatch is enabled, we can redistribute * load based on RSS. * * XXXRW: Would be nice if the ifnet passed up a flag indicating whether or * not it had already done work distribution via multi-queue. Then we could * direct dispatch in the event load balancing was already complete and * handle the case of interfaces with different capabilities better. * * XXXRW: Sort of want an M_DISTRIBUTED flag to avoid multiple distributions * at multiple layers? * * XXXRW: For now, enable all this only if RSS is compiled in, although it * works fine without RSS. Need to characterise the performance overhead * of the detour through the netisr code in the event the result is always * direct dispatch. */ static void ether_nh_input(struct mbuf *m) { M_ASSERTPKTHDR(m); KASSERT(m->m_pkthdr.rcvif != NULL, ("%s: NULL interface pointer", __func__)); ether_input_internal(m->m_pkthdr.rcvif, m); } static struct netisr_handler ether_nh = { .nh_name = "ether", .nh_handler = ether_nh_input, .nh_proto = NETISR_ETHER, #ifdef RSS .nh_policy = NETISR_POLICY_CPU, .nh_dispatch = NETISR_DISPATCH_DIRECT, .nh_m2cpuid = rss_m2cpuid, #else .nh_policy = NETISR_POLICY_SOURCE, .nh_dispatch = NETISR_DISPATCH_DIRECT, #endif }; static void ether_init(__unused void *arg) { netisr_register(ðer_nh); } SYSINIT(ether, SI_SUB_INIT_IF, SI_ORDER_ANY, ether_init, NULL); static void vnet_ether_init(__unused void *arg) { struct pfil_head_args args; args.pa_version = PFIL_VERSION; args.pa_flags = PFIL_IN | PFIL_OUT; args.pa_type = PFIL_TYPE_ETHERNET; args.pa_headname = PFIL_ETHER_NAME; V_link_pfil_head = pfil_head_register(&args); #ifdef VIMAGE netisr_register_vnet(ðer_nh); #endif } VNET_SYSINIT(vnet_ether_init, SI_SUB_PROTO_IF, SI_ORDER_ANY, vnet_ether_init, NULL); #ifdef VIMAGE static void vnet_ether_pfil_destroy(__unused void *arg) { pfil_head_unregister(V_link_pfil_head); } VNET_SYSUNINIT(vnet_ether_pfil_uninit, SI_SUB_PROTO_PFIL, SI_ORDER_ANY, vnet_ether_pfil_destroy, NULL); static void vnet_ether_destroy(__unused void *arg) { netisr_unregister_vnet(ðer_nh); } VNET_SYSUNINIT(vnet_ether_uninit, SI_SUB_PROTO_IF, SI_ORDER_ANY, vnet_ether_destroy, NULL); #endif static void ether_input(struct ifnet *ifp, struct mbuf *m) { struct epoch_tracker et; struct mbuf *mn; bool needs_epoch; needs_epoch = !(ifp->if_flags & IFF_KNOWSEPOCH); /* * The drivers are allowed to pass in a chain of packets linked with * m_nextpkt. We split them up into separate packets here and pass * them up. This allows the drivers to amortize the receive lock. */ CURVNET_SET_QUIET(ifp->if_vnet); if (__predict_false(needs_epoch)) NET_EPOCH_ENTER(et); while (m) { mn = m->m_nextpkt; m->m_nextpkt = NULL; /* * We will rely on rcvif being set properly in the deferred * context, so assert it is correct here. */ MPASS((m->m_pkthdr.csum_flags & CSUM_SND_TAG) == 0); KASSERT(m->m_pkthdr.rcvif == ifp, ("%s: ifnet mismatch m %p " "rcvif %p ifp %p", __func__, m, m->m_pkthdr.rcvif, ifp)); netisr_dispatch(NETISR_ETHER, m); m = mn; } if (__predict_false(needs_epoch)) NET_EPOCH_EXIT(et); CURVNET_RESTORE(); } /* * Upper layer processing for a received Ethernet packet. */ void ether_demux(struct ifnet *ifp, struct mbuf *m) { struct ether_header *eh; int i, isr; u_short ether_type; NET_EPOCH_ASSERT(); KASSERT(ifp != NULL, ("%s: NULL interface pointer", __func__)); /* Do not grab PROMISC frames in case we are re-entered. */ if (PFIL_HOOKED_IN(V_link_pfil_head) && !(m->m_flags & M_PROMISC)) { i = pfil_mbuf_in(V_link_pfil_head, &m, ifp, NULL); if (i != 0 || m == NULL) return; } eh = mtod(m, struct ether_header *); ether_type = ntohs(eh->ether_type); /* * If this frame has a VLAN tag other than 0, call vlan_input() * if its module is loaded. Otherwise, drop. */ if ((m->m_flags & M_VLANTAG) && EVL_VLANOFTAG(m->m_pkthdr.ether_vtag) != 0) { if (ifp->if_vlantrunk == NULL) { if_inc_counter(ifp, IFCOUNTER_NOPROTO, 1); m_freem(m); return; } KASSERT(vlan_input_p != NULL,("%s: VLAN not loaded!", __func__)); /* Clear before possibly re-entering ether_input(). */ m->m_flags &= ~M_PROMISC; (*vlan_input_p)(ifp, m); return; } /* * Pass promiscuously received frames to the upper layer if the user * requested this by setting IFF_PPROMISC. Otherwise, drop them. */ if ((ifp->if_flags & IFF_PPROMISC) == 0 && (m->m_flags & M_PROMISC)) { m_freem(m); return; } /* * Reset layer specific mbuf flags to avoid confusing upper layers. */ m->m_flags &= ~M_VLANTAG; m_clrprotoflags(m); /* * Dispatch frame to upper layer. */ switch (ether_type) { #ifdef INET case ETHERTYPE_IP: isr = NETISR_IP; break; case ETHERTYPE_ARP: if (ifp->if_flags & IFF_NOARP) { /* Discard packet if ARP is disabled on interface */ m_freem(m); return; } isr = NETISR_ARP; break; #endif #ifdef INET6 case ETHERTYPE_IPV6: isr = NETISR_IPV6; break; #endif default: goto discard; } /* Strip off Ethernet header. */ m_adj(m, ETHER_HDR_LEN); netisr_dispatch(isr, m); return; discard: /* * Packet is to be discarded. If netgraph is present, * hand the packet to it for last chance processing; * otherwise dispose of it. */ if (ifp->if_l2com != NULL) { KASSERT(ng_ether_input_orphan_p != NULL, ("ng_ether_input_orphan_p is NULL")); (*ng_ether_input_orphan_p)(ifp, m); return; } m_freem(m); } /* * Convert Ethernet address to printable (loggable) representation. * This routine is for compatibility; it's better to just use * * printf("%6D", , ":"); * * since there's no static buffer involved. */ char * ether_sprintf(const u_char *ap) { static char etherbuf[18]; snprintf(etherbuf, sizeof (etherbuf), "%6D", ap, ":"); return (etherbuf); } /* * Perform common duties while attaching to interface list */ void ether_ifattach(struct ifnet *ifp, const u_int8_t *lla) { int i; struct ifaddr *ifa; struct sockaddr_dl *sdl; ifp->if_addrlen = ETHER_ADDR_LEN; ifp->if_hdrlen = ETHER_HDR_LEN; ifp->if_mtu = ETHERMTU; if_attach(ifp); ifp->if_output = ether_output; ifp->if_input = ether_input; ifp->if_resolvemulti = ether_resolvemulti; ifp->if_requestencap = ether_requestencap; #ifdef VIMAGE ifp->if_reassign = ether_reassign; #endif if (ifp->if_baudrate == 0) ifp->if_baudrate = IF_Mbps(10); /* just a default */ ifp->if_broadcastaddr = etherbroadcastaddr; ifa = ifp->if_addr; KASSERT(ifa != NULL, ("%s: no lladdr!\n", __func__)); sdl = (struct sockaddr_dl *)ifa->ifa_addr; sdl->sdl_type = IFT_ETHER; sdl->sdl_alen = ifp->if_addrlen; bcopy(lla, LLADDR(sdl), ifp->if_addrlen); if (ifp->if_hw_addr != NULL) bcopy(lla, ifp->if_hw_addr, ifp->if_addrlen); bpfattach(ifp, DLT_EN10MB, ETHER_HDR_LEN); if (ng_ether_attach_p != NULL) (*ng_ether_attach_p)(ifp); /* Announce Ethernet MAC address if non-zero. */ for (i = 0; i < ifp->if_addrlen; i++) if (lla[i] != 0) break; if (i != ifp->if_addrlen) if_printf(ifp, "Ethernet address: %6D\n", lla, ":"); uuid_ether_add(LLADDR(sdl)); /* Add necessary bits are setup; announce it now. */ EVENTHANDLER_INVOKE(ether_ifattach_event, ifp); if (IS_DEFAULT_VNET(curvnet)) devctl_notify("ETHERNET", ifp->if_xname, "IFATTACH", NULL); } /* * Perform common duties while detaching an Ethernet interface */ void ether_ifdetach(struct ifnet *ifp) { struct sockaddr_dl *sdl; sdl = (struct sockaddr_dl *)(ifp->if_addr->ifa_addr); uuid_ether_del(LLADDR(sdl)); if (ifp->if_l2com != NULL) { KASSERT(ng_ether_detach_p != NULL, ("ng_ether_detach_p is NULL")); (*ng_ether_detach_p)(ifp); } bpfdetach(ifp); if_detach(ifp); } #ifdef VIMAGE void ether_reassign(struct ifnet *ifp, struct vnet *new_vnet, char *unused __unused) { if (ifp->if_l2com != NULL) { KASSERT(ng_ether_detach_p != NULL, ("ng_ether_detach_p is NULL")); (*ng_ether_detach_p)(ifp); } if (ng_ether_attach_p != NULL) { CURVNET_SET_QUIET(new_vnet); (*ng_ether_attach_p)(ifp); CURVNET_RESTORE(); } } #endif SYSCTL_DECL(_net_link); SYSCTL_NODE(_net_link, IFT_ETHER, ether, CTLFLAG_RW | CTLFLAG_MPSAFE, 0, "Ethernet"); #if 0 /* * This is for reference. We have a table-driven version * of the little-endian crc32 generator, which is faster * than the double-loop. */ uint32_t ether_crc32_le(const uint8_t *buf, size_t len) { size_t i; uint32_t crc; int bit; uint8_t data; crc = 0xffffffff; /* initial value */ for (i = 0; i < len; i++) { for (data = *buf++, bit = 0; bit < 8; bit++, data >>= 1) { carry = (crc ^ data) & 1; crc >>= 1; if (carry) crc = (crc ^ ETHER_CRC_POLY_LE); } } return (crc); } #else uint32_t ether_crc32_le(const uint8_t *buf, size_t len) { static const uint32_t crctab[] = { 0x00000000, 0x1db71064, 0x3b6e20c8, 0x26d930ac, 0x76dc4190, 0x6b6b51f4, 0x4db26158, 0x5005713c, 0xedb88320, 0xf00f9344, 0xd6d6a3e8, 0xcb61b38c, 0x9b64c2b0, 0x86d3d2d4, 0xa00ae278, 0xbdbdf21c }; size_t i; uint32_t crc; crc = 0xffffffff; /* initial value */ for (i = 0; i < len; i++) { crc ^= buf[i]; crc = (crc >> 4) ^ crctab[crc & 0xf]; crc = (crc >> 4) ^ crctab[crc & 0xf]; } return (crc); } #endif uint32_t ether_crc32_be(const uint8_t *buf, size_t len) { size_t i; uint32_t crc, carry; int bit; uint8_t data; crc = 0xffffffff; /* initial value */ for (i = 0; i < len; i++) { for (data = *buf++, bit = 0; bit < 8; bit++, data >>= 1) { carry = ((crc & 0x80000000) ? 1 : 0) ^ (data & 0x01); crc <<= 1; if (carry) crc = (crc ^ ETHER_CRC_POLY_BE) | carry; } } return (crc); } int ether_ioctl(struct ifnet *ifp, u_long command, caddr_t data) { struct ifaddr *ifa = (struct ifaddr *) data; struct ifreq *ifr = (struct ifreq *) data; int error = 0; switch (command) { case SIOCSIFADDR: ifp->if_flags |= IFF_UP; switch (ifa->ifa_addr->sa_family) { #ifdef INET case AF_INET: ifp->if_init(ifp->if_softc); /* before arpwhohas */ arp_ifinit(ifp, ifa); break; #endif default: ifp->if_init(ifp->if_softc); break; } break; case SIOCGIFADDR: bcopy(IF_LLADDR(ifp), &ifr->ifr_addr.sa_data[0], ETHER_ADDR_LEN); break; case SIOCSIFMTU: /* * Set the interface MTU. */ if (ifr->ifr_mtu > ETHERMTU) { error = EINVAL; } else { ifp->if_mtu = ifr->ifr_mtu; } break; case SIOCSLANPCP: error = priv_check(curthread, PRIV_NET_SETLANPCP); if (error != 0) break; if (ifr->ifr_lan_pcp > 7 && ifr->ifr_lan_pcp != IFNET_PCP_NONE) { error = EINVAL; } else { ifp->if_pcp = ifr->ifr_lan_pcp; /* broadcast event about PCP change */ EVENTHANDLER_INVOKE(ifnet_event, ifp, IFNET_EVENT_PCP); } break; case SIOCGLANPCP: ifr->ifr_lan_pcp = ifp->if_pcp; break; default: error = EINVAL; /* XXX netbsd has ENOTTY??? */ break; } return (error); } static int ether_resolvemulti(struct ifnet *ifp, struct sockaddr **llsa, struct sockaddr *sa) { struct sockaddr_dl *sdl; #ifdef INET struct sockaddr_in *sin; #endif #ifdef INET6 struct sockaddr_in6 *sin6; #endif u_char *e_addr; switch(sa->sa_family) { case AF_LINK: /* * No mapping needed. Just check that it's a valid MC address. */ sdl = (struct sockaddr_dl *)sa; e_addr = LLADDR(sdl); if (!ETHER_IS_MULTICAST(e_addr)) return EADDRNOTAVAIL; *llsa = NULL; return 0; #ifdef INET case AF_INET: sin = (struct sockaddr_in *)sa; if (!IN_MULTICAST(ntohl(sin->sin_addr.s_addr))) return EADDRNOTAVAIL; sdl = link_init_sdl(ifp, *llsa, IFT_ETHER); sdl->sdl_alen = ETHER_ADDR_LEN; e_addr = LLADDR(sdl); ETHER_MAP_IP_MULTICAST(&sin->sin_addr, e_addr); *llsa = (struct sockaddr *)sdl; return 0; #endif #ifdef INET6 case AF_INET6: sin6 = (struct sockaddr_in6 *)sa; if (IN6_IS_ADDR_UNSPECIFIED(&sin6->sin6_addr)) { /* * An IP6 address of 0 means listen to all * of the Ethernet multicast address used for IP6. * (This is used for multicast routers.) */ ifp->if_flags |= IFF_ALLMULTI; *llsa = NULL; return 0; } if (!IN6_IS_ADDR_MULTICAST(&sin6->sin6_addr)) return EADDRNOTAVAIL; sdl = link_init_sdl(ifp, *llsa, IFT_ETHER); sdl->sdl_alen = ETHER_ADDR_LEN; e_addr = LLADDR(sdl); ETHER_MAP_IPV6_MULTICAST(&sin6->sin6_addr, e_addr); *llsa = (struct sockaddr *)sdl; return 0; #endif default: /* * Well, the text isn't quite right, but it's the name * that counts... */ return EAFNOSUPPORT; } } static moduledata_t ether_mod = { .name = "ether", }; void ether_vlan_mtap(struct bpf_if *bp, struct mbuf *m, void *data, u_int dlen) { struct ether_vlan_header vlan; struct mbuf mv, mb; KASSERT((m->m_flags & M_VLANTAG) != 0, ("%s: vlan information not present", __func__)); KASSERT(m->m_len >= sizeof(struct ether_header), ("%s: mbuf not large enough for header", __func__)); bcopy(mtod(m, char *), &vlan, sizeof(struct ether_header)); vlan.evl_proto = vlan.evl_encap_proto; vlan.evl_encap_proto = htons(ETHERTYPE_VLAN); vlan.evl_tag = htons(m->m_pkthdr.ether_vtag); m->m_len -= sizeof(struct ether_header); m->m_data += sizeof(struct ether_header); /* * If a data link has been supplied by the caller, then we will need to * re-create a stack allocated mbuf chain with the following structure: * * (1) mbuf #1 will contain the supplied data link * (2) mbuf #2 will contain the vlan header * (3) mbuf #3 will contain the original mbuf's packet data * * Otherwise, submit the packet and vlan header via bpf_mtap2(). */ if (data != NULL) { mv.m_next = m; mv.m_data = (caddr_t)&vlan; mv.m_len = sizeof(vlan); mb.m_next = &mv; mb.m_data = data; mb.m_len = dlen; bpf_mtap(bp, &mb); } else bpf_mtap2(bp, &vlan, sizeof(vlan), m); m->m_len += sizeof(struct ether_header); m->m_data -= sizeof(struct ether_header); } struct mbuf * ether_vlanencap_proto(struct mbuf *m, uint16_t tag, uint16_t proto) { struct ether_vlan_header *evl; M_PREPEND(m, ETHER_VLAN_ENCAP_LEN, M_NOWAIT); if (m == NULL) return (NULL); /* M_PREPEND takes care of m_len, m_pkthdr.len for us */ if (m->m_len < sizeof(*evl)) { m = m_pullup(m, sizeof(*evl)); if (m == NULL) return (NULL); } /* * Transform the Ethernet header into an Ethernet header * with 802.1Q encapsulation. */ evl = mtod(m, struct ether_vlan_header *); bcopy((char *)evl + ETHER_VLAN_ENCAP_LEN, (char *)evl, ETHER_HDR_LEN - ETHER_TYPE_LEN); evl->evl_encap_proto = htons(proto); evl->evl_tag = htons(tag); return (m); } void ether_bpf_mtap_if(struct ifnet *ifp, struct mbuf *m) { if (bpf_peers_present(ifp->if_bpf)) { M_ASSERTVALID(m); if ((m->m_flags & M_VLANTAG) != 0) ether_vlan_mtap(ifp->if_bpf, m, NULL, 0); else bpf_mtap(ifp->if_bpf, m); } } static SYSCTL_NODE(_net_link, IFT_L2VLAN, vlan, CTLFLAG_RW | CTLFLAG_MPSAFE, 0, "IEEE 802.1Q VLAN"); static SYSCTL_NODE(_net_link_vlan, PF_LINK, link, CTLFLAG_RW | CTLFLAG_MPSAFE, 0, "for consistency"); VNET_DEFINE_STATIC(int, soft_pad); #define V_soft_pad VNET(soft_pad) SYSCTL_INT(_net_link_vlan, OID_AUTO, soft_pad, CTLFLAG_RW | CTLFLAG_VNET, &VNET_NAME(soft_pad), 0, "pad short frames before tagging"); /* * For now, make preserving PCP via an mbuf tag optional, as it increases * per-packet memory allocations and frees. In the future, it would be * preferable to reuse ether_vtag for this, or similar. */ VNET_DEFINE(int, vlan_mtag_pcp) = 0; #define V_vlan_mtag_pcp VNET(vlan_mtag_pcp) SYSCTL_INT(_net_link_vlan, OID_AUTO, mtag_pcp, CTLFLAG_RW | CTLFLAG_VNET, &VNET_NAME(vlan_mtag_pcp), 0, "Retain VLAN PCP information as packets are passed up the stack"); bool ether_8021q_frame(struct mbuf **mp, struct ifnet *ife, struct ifnet *p, struct ether_8021q_tag *qtag) { struct m_tag *mtag; int n; uint16_t tag; static const char pad[8]; /* just zeros */ /* * Pad the frame to the minimum size allowed if told to. * This option is in accord with IEEE Std 802.1Q, 2003 Ed., * paragraph C.4.4.3.b. It can help to work around buggy * bridges that violate paragraph C.4.4.3.a from the same * document, i.e., fail to pad short frames after untagging. * E.g., a tagged frame 66 bytes long (incl. FCS) is OK, but * untagging it will produce a 62-byte frame, which is a runt * and requires padding. There are VLAN-enabled network * devices that just discard such runts instead or mishandle * them somehow. */ if (V_soft_pad && p->if_type == IFT_ETHER) { for (n = ETHERMIN + ETHER_HDR_LEN - (*mp)->m_pkthdr.len; n > 0; n -= sizeof(pad)) { if (!m_append(*mp, min(n, sizeof(pad)), pad)) break; } if (n > 0) { m_freem(*mp); *mp = NULL; if_printf(ife, "cannot pad short frame"); return (false); } } /* * If PCP is set in mbuf, use it */ if ((*mp)->m_flags & M_VLANTAG) { qtag->pcp = EVL_PRIOFTAG((*mp)->m_pkthdr.ether_vtag); } /* * If underlying interface can do VLAN tag insertion itself, * just pass the packet along. However, we need some way to * tell the interface where the packet came from so that it * knows how to find the VLAN tag to use, so we attach a * packet tag that holds it. */ if (V_vlan_mtag_pcp && (mtag = m_tag_locate(*mp, MTAG_8021Q, MTAG_8021Q_PCP_OUT, NULL)) != NULL) tag = EVL_MAKETAG(qtag->vid, *(uint8_t *)(mtag + 1), 0); else tag = EVL_MAKETAG(qtag->vid, qtag->pcp, 0); if ((p->if_capenable & IFCAP_VLAN_HWTAGGING) && (qtag->proto == ETHERTYPE_VLAN)) { (*mp)->m_pkthdr.ether_vtag = tag; (*mp)->m_flags |= M_VLANTAG; } else { *mp = ether_vlanencap_proto(*mp, tag, qtag->proto); if (*mp == NULL) { if_printf(ife, "unable to prepend 802.1Q header"); return (false); } } return (true); } /* * Allocate an address from the FreeBSD Foundation OUI. This uses a * cryptographic hash function on the containing jail's name, UUID and the * interface name to attempt to provide a unique but stable address. * Pseudo-interfaces which require a MAC address should use this function to * allocate non-locally-administered addresses. */ void ether_gen_addr(struct ifnet *ifp, struct ether_addr *hwaddr) { SHA1_CTX ctx; char *buf; char uuid[HOSTUUIDLEN + 1]; uint64_t addr; int i, sz; char digest[SHA1_RESULTLEN]; char jailname[MAXHOSTNAMELEN]; getcredhostuuid(curthread->td_ucred, uuid, sizeof(uuid)); if (strncmp(uuid, DEFAULT_HOSTUUID, sizeof(uuid)) == 0) { /* Fall back to a random mac address. */ goto rando; } /* If each (vnet) jail would also have a unique hostuuid this would not * be necessary. */ getjailname(curthread->td_ucred, jailname, sizeof(jailname)); sz = asprintf(&buf, M_TEMP, "%s-%s-%s", uuid, if_name(ifp), jailname); if (sz < 0) { /* Fall back to a random mac address. */ goto rando; } SHA1Init(&ctx); SHA1Update(&ctx, buf, sz); SHA1Final(digest, &ctx); free(buf, M_TEMP); addr = ((digest[0] << 16) | (digest[1] << 8) | digest[2]) & OUI_FREEBSD_GENERATED_MASK; addr = OUI_FREEBSD(addr); for (i = 0; i < ETHER_ADDR_LEN; ++i) { hwaddr->octet[i] = addr >> ((ETHER_ADDR_LEN - i - 1) * 8) & 0xFF; } return; rando: arc4rand(hwaddr, sizeof(*hwaddr), 0); /* Unicast */ hwaddr->octet[0] &= 0xFE; /* Locally administered. */ hwaddr->octet[0] |= 0x02; } DECLARE_MODULE(ether, ether_mod, SI_SUB_INIT_IF, SI_ORDER_ANY); MODULE_VERSION(ether, 1);