d/freebsd-dev
1
0
Fork 0
Code Issues Pull Requests Packages Projects Releases Wiki Activity
ed2f945a39
freebsd-dev/sys/net/ieee8023ad_lacp.c

2220 lines
54 KiB
C
Raw Normal View History

Add the trunk(4) driver for providing link aggregation, failover and fault tolerance. This driver allows aggregation of multiple network interfaces as one virtual interface using a number of different protocols/algorithms. failover - Sends traffic through the secondary port if the master becomes inactive. fec - Supports Cisco Fast EtherChannel. lacp - Supports the IEEE 802.3ad Link Aggregation Control Protocol (LACP) and the Marker Protocol. loadbalance - Static loadbalancing using an outgoing hash. roundrobin - Distributes outgoing traffic using a round-robin scheduler through all active ports. This code was obtained from OpenBSD and this also includes 802.3ad LACP support from agr(4) in NetBSD.
2007-04-10 00:27:25 +00:00
/* $NetBSD: ieee8023ad_lacp.c,v 1.3 2005/12/11 12:24:54 christos Exp $ */
/*-
sys: general adoption of SPDX licensing ID tags. Mainly focus on files that use BSD 2-Clause license, however the tool I was using misidentified many licenses so this was mostly a manual - error prone - task. The Software Package Data Exchange (SPDX) group provides a specification to make it easier for automated tools to detect and summarize well known opensource licenses. We are gradually adopting the specification, noting that the tags are considered only advisory and do not, in any way, superceed or replace the license texts. No functional change intended.
2017-11-27 15:23:17 +00:00
* SPDX-License-Identifier: BSD-2-Clause-NetBSD
*
Add the trunk(4) driver for providing link aggregation, failover and fault tolerance. This driver allows aggregation of multiple network interfaces as one virtual interface using a number of different protocols/algorithms. failover - Sends traffic through the secondary port if the master becomes inactive. fec - Supports Cisco Fast EtherChannel. lacp - Supports the IEEE 802.3ad Link Aggregation Control Protocol (LACP) and the Marker Protocol. loadbalance - Static loadbalancing using an outgoing hash. roundrobin - Distributes outgoing traffic using a round-robin scheduler through all active ports. This code was obtained from OpenBSD and this also includes 802.3ad LACP support from agr(4) in NetBSD.
2007-04-10 00:27:25 +00:00
* Copyright (c)2005 YAMAMOTO Takashi,
Switch the LACP state machine over to its own mutex to protect the internals, this means that it no longer grabs the lagg rwlock. Use two port table arrays which list the active ports for Tx and switch between them with an atomic op. Now the lagg rwlock is only exclusively locked for management (ioctls) and queuing of lacp control frames isnt needed.
2008-03-16 19:25:30 +00:00
* Copyright (c)2008 Andrew Thompson <thompsa@FreeBSD.org>
Add the trunk(4) driver for providing link aggregation, failover and fault tolerance. This driver allows aggregation of multiple network interfaces as one virtual interface using a number of different protocols/algorithms. failover - Sends traffic through the secondary port if the master becomes inactive. fec - Supports Cisco Fast EtherChannel. lacp - Supports the IEEE 802.3ad Link Aggregation Control Protocol (LACP) and the Marker Protocol. loadbalance - Static loadbalancing using an outgoing hash. roundrobin - Distributes outgoing traffic using a round-robin scheduler through all active ports. This code was obtained from OpenBSD and this also includes 802.3ad LACP support from agr(4) in NetBSD.
2007-04-10 00:27:25 +00:00
* 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 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 AUTHOR 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.
*/
#include <sys/cdefs.h>
__FBSDID("$FreeBSD$");
Add kernel-side support for in-kernel TLS. KTLS adds support for in-kernel framing and encryption of Transport Layer Security (1.0-1.2) data on TCP sockets. KTLS only supports offload of TLS for transmitted data. Key negotation must still be performed in userland. Once completed, transmit session keys for a connection are provided to the kernel via a new TCP_TXTLS_ENABLE socket option. All subsequent data transmitted on the socket is placed into TLS frames and encrypted using the supplied keys. Any data written to a KTLS-enabled socket via write(2), aio_write(2), or sendfile(2) is assumed to be application data and is encoded in TLS frames with an application data type. Individual records can be sent with a custom type (e.g. handshake messages) via sendmsg(2) with a new control message (TLS_SET_RECORD_TYPE) specifying the record type. At present, rekeying is not supported though the in-kernel framework should support rekeying. KTLS makes use of the recently added unmapped mbufs to store TLS frames in the socket buffer. Each TLS frame is described by a single ext_pgs mbuf. The ext_pgs structure contains the header of the TLS record (and trailer for encrypted records) as well as references to the associated TLS session. KTLS supports two primary methods of encrypting TLS frames: software TLS and ifnet TLS. Software TLS marks mbufs holding socket data as not ready via M_NOTREADY similar to sendfile(2) when TLS framing information is added to an unmapped mbuf in ktls_frame(). ktls_enqueue() is then called to schedule TLS frames for encryption. In the case of sendfile_iodone() calls ktls_enqueue() instead of pru_ready() leaving the mbufs marked M_NOTREADY until encryption is completed. For other writes (vn_sendfile when pages are available, write(2), etc.), the PRUS_NOTREADY is set when invoking pru_send() along with invoking ktls_enqueue(). A pool of worker threads (the "KTLS" kernel process) encrypts TLS frames queued via ktls_enqueue(). Each TLS frame is temporarily mapped using the direct map and passed to a software encryption backend to perform the actual encryption. (Note: The use of PHYS_TO_DMAP could be replaced with sf_bufs if someone wished to make this work on architectures without a direct map.) KTLS supports pluggable software encryption backends. Internally, Netflix uses proprietary pure-software backends. This commit includes a simple backend in a new ktls_ocf.ko module that uses the kernel's OpenCrypto framework to provide AES-GCM encryption of TLS frames. As a result, software TLS is now a bit of a misnomer as it can make use of hardware crypto accelerators. Once software encryption has finished, the TLS frame mbufs are marked ready via pru_ready(). At this point, the encrypted data appears as regular payload to the TCP stack stored in unmapped mbufs. ifnet TLS permits a NIC to offload the TLS encryption and TCP segmentation. In this mode, a new send tag type (IF_SND_TAG_TYPE_TLS) is allocated on the interface a socket is routed over and associated with a TLS session. TLS records for a TLS session using ifnet TLS are not marked M_NOTREADY but are passed down the stack unencrypted. The ip_output_send() and ip6_output_send() helper functions that apply send tags to outbound IP packets verify that the send tag of the TLS record matches the outbound interface. If so, the packet is tagged with the TLS send tag and sent to the interface. The NIC device driver must recognize packets with the TLS send tag and schedule them for TLS encryption and TCP segmentation. If the the outbound interface does not match the interface in the TLS send tag, the packet is dropped. In addition, a task is scheduled to refresh the TLS send tag for the TLS session. If a new TLS send tag cannot be allocated, the connection is dropped. If a new TLS send tag is allocated, however, subsequent packets will be tagged with the correct TLS send tag. (This latter case has been tested by configuring both ports of a Chelsio T6 in a lagg and failing over from one port to another. As the connections migrated to the new port, new TLS send tags were allocated for the new port and connections resumed without being dropped.) ifnet TLS can be enabled and disabled on supported network interfaces via new '[-]txtls[46]' options to ifconfig(8). ifnet TLS is supported across both vlan devices and lagg interfaces using failover, lacp with flowid enabled, or lacp with flowid enabled. Applications may request the current KTLS mode of a connection via a new TCP_TXTLS_MODE socket option. They can also use this socket option to toggle between software and ifnet TLS modes. In addition, a testing tool is available in tools/tools/switch_tls. This is modeled on tcpdrop and uses similar syntax. However, instead of dropping connections, -s is used to force KTLS connections to switch to software TLS and -i is used to switch to ifnet TLS. Various sysctls and counters are available under the kern.ipc.tls sysctl node. The kern.ipc.tls.enable node must be set to true to enable KTLS (it is off by default). The use of unmapped mbufs must also be enabled via kern.ipc.mb_use_ext_pgs to enable KTLS. KTLS is enabled via the KERN_TLS kernel option. This patch is the culmination of years of work by several folks including Scott Long and Randall Stewart for the original design and implementation; Drew Gallatin for several optimizations including the use of ext_pgs mbufs, the M_NOTREADY mechanism for TLS records awaiting software encryption, and pluggable software crypto backends; and John Baldwin for modifications to support hardware TLS offload. Reviewed by: gallatin, hselasky, rrs Obtained from: Netflix Sponsored by: Netflix, Chelsio Communications Differential Revision: https://reviews.freebsd.org/D21277
2019-08-27 00:01:56 +00:00
#include "opt_kern_tls.h"
Implement kernel support for hardware rate limited sockets. - Add RATELIMIT kernel configuration keyword which must be set to enable the new functionality. - Add support for hardware driven, Receive Side Scaling, RSS aware, rate limited sendqueues and expose the functionality through the already established SO_MAX_PACING_RATE setsockopt(). The API support rates in the range from 1 to 4Gbytes/s which are suitable for regular TCP and UDP streams. The setsockopt(2) manual page has been updated. - Add rate limit function callback API to "struct ifnet" which supports the following operations: if_snd_tag_alloc(), if_snd_tag_modify(), if_snd_tag_query() and if_snd_tag_free(). - Add support to ifconfig to view, set and clear the IFCAP_TXRTLMT flag, which tells if a network driver supports rate limiting or not. - This patch also adds support for rate limiting through VLAN and LAGG intermediate network devices. - How rate limiting works: 1) The userspace application calls setsockopt() after accepting or making a new connection to set the rate which is then stored in the socket structure in the kernel. Later on when packets are transmitted a check is made in the transmit path for rate changes. A rate change implies a non-blocking ifp->if_snd_tag_alloc() call will be made to the destination network interface, which then sets up a custom sendqueue with the given rate limitation parameter. A "struct m_snd_tag" pointer is returned which serves as a "snd_tag" hint in the m_pkthdr for the subsequently transmitted mbufs. 2) When the network driver sees the "m->m_pkthdr.snd_tag" different from NULL, it will move the packets into a designated rate limited sendqueue given by the snd_tag pointer. It is up to the individual drivers how the rate limited traffic will be rate limited. 3) Route changes are detected by the NIC drivers in the ifp->if_transmit() routine when the ifnet pointer in the incoming snd_tag mismatches the one of the network interface. The network adapter frees the mbuf and returns EAGAIN which causes the ip_output() to release and clear the send tag. Upon next ip_output() a new "snd_tag" will be tried allocated. 4) When the PCB is detached the custom sendqueue will be released by a non-blocking ifp->if_snd_tag_free() call to the currently bound network interface. Reviewed by: wblock (manpages), adrian, gallatin, scottl (network) Differential Revision: https://reviews.freebsd.org/D3687 Sponsored by: Mellanox Technologies MFC after: 3 months
2017-01-18 13:31:17 +00:00
#include "opt_ratelimit.h"
Add the trunk(4) driver for providing link aggregation, failover and fault tolerance. This driver allows aggregation of multiple network interfaces as one virtual interface using a number of different protocols/algorithms. failover - Sends traffic through the secondary port if the master becomes inactive. fec - Supports Cisco Fast EtherChannel. lacp - Supports the IEEE 802.3ad Link Aggregation Control Protocol (LACP) and the Marker Protocol. loadbalance - Static loadbalancing using an outgoing hash. roundrobin - Distributes outgoing traffic using a round-robin scheduler through all active ports. This code was obtained from OpenBSD and this also includes 802.3ad LACP support from agr(4) in NetBSD.
2007-04-10 00:27:25 +00:00
#include <sys/param.h>
#include <sys/callout.h>
Include necessary headers that now are available due to pollution via if_var.h. Sponsored by: Netflix Sponsored by: Nginx, Inc.
2013-10-28 07:29:16 +00:00
#include <sys/eventhandler.h>
Add the trunk(4) driver for providing link aggregation, failover and fault tolerance. This driver allows aggregation of multiple network interfaces as one virtual interface using a number of different protocols/algorithms. failover - Sends traffic through the secondary port if the master becomes inactive. fec - Supports Cisco Fast EtherChannel. lacp - Supports the IEEE 802.3ad Link Aggregation Control Protocol (LACP) and the Marker Protocol. loadbalance - Static loadbalancing using an outgoing hash. roundrobin - Distributes outgoing traffic using a round-robin scheduler through all active ports. This code was obtained from OpenBSD and this also includes 802.3ad LACP support from agr(4) in NetBSD.
2007-04-10 00:27:25 +00:00
#include <sys/mbuf.h>
#include <sys/systm.h>
#include <sys/malloc.h>
#include <sys/kernel.h> /* hz */
#include <sys/socket.h> /* for net/if.h */
#include <sys/sockio.h>
Turn LACP debugging from a compile time option to a sysctl, it is very handy to be able to turn it on when negotiation to a switch misbehaves. Submitted by: Andrew Boyer MFC after: 3 days
2012-05-26 08:09:01 +00:00
#include <sys/sysctl.h>
Add the trunk(4) driver for providing link aggregation, failover and fault tolerance. This driver allows aggregation of multiple network interfaces as one virtual interface using a number of different protocols/algorithms. failover - Sends traffic through the secondary port if the master becomes inactive. fec - Supports Cisco Fast EtherChannel. lacp - Supports the IEEE 802.3ad Link Aggregation Control Protocol (LACP) and the Marker Protocol. loadbalance - Static loadbalancing using an outgoing hash. roundrobin - Distributes outgoing traffic using a round-robin scheduler through all active ports. This code was obtained from OpenBSD and this also includes 802.3ad LACP support from agr(4) in NetBSD.
2007-04-10 00:27:25 +00:00
#include <machine/stdarg.h>
Change from a mutex to a read/write lock. This allows the tx port to be selected simultaneously by multiple senders and transmit/receive is not serialised between aggregated interfaces.
2007-05-15 07:41:46 +00:00
#include <sys/lock.h>
#include <sys/rwlock.h>
The r48589 promised to remove implicit inclusion of if_var.h soon. Prepare to this event, adding if_var.h to files that do need it. Also, include all includes that now are included due to implicit pollution via if_var.h Sponsored by: Netflix Sponsored by: Nginx, Inc.
2013-10-26 17:58:36 +00:00
#include <sys/taskqueue.h>
Add the trunk(4) driver for providing link aggregation, failover and fault tolerance. This driver allows aggregation of multiple network interfaces as one virtual interface using a number of different protocols/algorithms. failover - Sends traffic through the secondary port if the master becomes inactive. fec - Supports Cisco Fast EtherChannel. lacp - Supports the IEEE 802.3ad Link Aggregation Control Protocol (LACP) and the Marker Protocol. loadbalance - Static loadbalancing using an outgoing hash. roundrobin - Distributes outgoing traffic using a round-robin scheduler through all active ports. This code was obtained from OpenBSD and this also includes 802.3ad LACP support from agr(4) in NetBSD.
2007-04-10 00:27:25 +00:00
#include <net/if.h>
The r48589 promised to remove implicit inclusion of if_var.h soon. Prepare to this event, adding if_var.h to files that do need it. Also, include all includes that now are included due to implicit pollution via if_var.h Sponsored by: Netflix Sponsored by: Nginx, Inc.
2013-10-26 17:58:36 +00:00
#include <net/if_var.h>
Add the trunk(4) driver for providing link aggregation, failover and fault tolerance. This driver allows aggregation of multiple network interfaces as one virtual interface using a number of different protocols/algorithms. failover - Sends traffic through the secondary port if the master becomes inactive. fec - Supports Cisco Fast EtherChannel. lacp - Supports the IEEE 802.3ad Link Aggregation Control Protocol (LACP) and the Marker Protocol. loadbalance - Static loadbalancing using an outgoing hash. roundrobin - Distributes outgoing traffic using a round-robin scheduler through all active ports. This code was obtained from OpenBSD and this also includes 802.3ad LACP support from agr(4) in NetBSD.
2007-04-10 00:27:25 +00:00
#include <net/if_dl.h>
#include <net/ethernet.h>
#include <net/if_media.h>
#include <net/if_types.h>
Rename the trunk(4) driver to lagg(4) as it is too similar to vlan trunking. The name trunk is misused as the networking term trunk means carrying multiple VLANs over a single connection. The IEEE standard for link aggregation (802.3 section 3) does not talk about 'trunk' at all while it is used throughout IEEE 802.1Q in describing vlans. The lagg(4) driver provides link aggregation, failover and fault tolerance. Discussed on: current@
2007-04-17 00:35:11 +00:00
#include <net/if_lagg.h>
Add the trunk(4) driver for providing link aggregation, failover and fault tolerance. This driver allows aggregation of multiple network interfaces as one virtual interface using a number of different protocols/algorithms. failover - Sends traffic through the secondary port if the master becomes inactive. fec - Supports Cisco Fast EtherChannel. lacp - Supports the IEEE 802.3ad Link Aggregation Control Protocol (LACP) and the Marker Protocol. loadbalance - Static loadbalancing using an outgoing hash. roundrobin - Distributes outgoing traffic using a round-robin scheduler through all active ports. This code was obtained from OpenBSD and this also includes 802.3ad LACP support from agr(4) in NetBSD.
2007-04-10 00:27:25 +00:00
#include <net/ieee8023ad_lacp.h>
/*
* actor system priority and port priority.
* XXX should be configurable.
*/
#define LACP_SYSTEM_PRIO 0x8000
#define LACP_PORT_PRIO 0x8000
const uint8_t ethermulticastaddr_slowprotocols[ETHER_ADDR_LEN] =
{ 0x01, 0x80, 0xc2, 0x00, 0x00, 0x02 };
static const struct tlv_template lacp_info_tlv_template[] = {
{ LACP_TYPE_ACTORINFO,
sizeof(struct tlvhdr) + sizeof(struct lacp_peerinfo) },
{ LACP_TYPE_PARTNERINFO,
sizeof(struct tlvhdr) + sizeof(struct lacp_peerinfo) },
{ LACP_TYPE_COLLECTORINFO,
sizeof(struct tlvhdr) + sizeof(struct lacp_collectorinfo) },
{ 0, 0 },
};
static const struct tlv_template marker_info_tlv_template[] = {
Implement the Marker Protocol. A marker frame is placed on the interface queue of each port and any further packets are blocked, when the all the marker frames have been returned to us from the remote network device then we can be sure that all interface queues are empty. This is needed when a port is added or removed from the aggregation since it will affect the hash based distribution, if the queues are not empty then a packet from an existing connection may be placed on a different interface and arrive out of order. This was previously achieved by suppressing transmission for 1 second, now that there is an active feedback this timeout as been increased to 3 seconds and used as a fallback.
2007-05-19 07:47:04 +00:00
{ MARKER_TYPE_INFO,
sizeof(struct tlvhdr) + sizeof(struct lacp_markerinfo) },
Add the trunk(4) driver for providing link aggregation, failover and fault tolerance. This driver allows aggregation of multiple network interfaces as one virtual interface using a number of different protocols/algorithms. failover - Sends traffic through the secondary port if the master becomes inactive. fec - Supports Cisco Fast EtherChannel. lacp - Supports the IEEE 802.3ad Link Aggregation Control Protocol (LACP) and the Marker Protocol. loadbalance - Static loadbalancing using an outgoing hash. roundrobin - Distributes outgoing traffic using a round-robin scheduler through all active ports. This code was obtained from OpenBSD and this also includes 802.3ad LACP support from agr(4) in NetBSD.
2007-04-10 00:27:25 +00:00
{ 0, 0 },
};
static const struct tlv_template marker_response_tlv_template[] = {
Implement the Marker Protocol. A marker frame is placed on the interface queue of each port and any further packets are blocked, when the all the marker frames have been returned to us from the remote network device then we can be sure that all interface queues are empty. This is needed when a port is added or removed from the aggregation since it will affect the hash based distribution, if the queues are not empty then a packet from an existing connection may be placed on a different interface and arrive out of order. This was previously achieved by suppressing transmission for 1 second, now that there is an active feedback this timeout as been increased to 3 seconds and used as a fallback.
2007-05-19 07:47:04 +00:00
{ MARKER_TYPE_RESPONSE,
sizeof(struct tlvhdr) + sizeof(struct lacp_markerinfo) },
Add the trunk(4) driver for providing link aggregation, failover and fault tolerance. This driver allows aggregation of multiple network interfaces as one virtual interface using a number of different protocols/algorithms. failover - Sends traffic through the secondary port if the master becomes inactive. fec - Supports Cisco Fast EtherChannel. lacp - Supports the IEEE 802.3ad Link Aggregation Control Protocol (LACP) and the Marker Protocol. loadbalance - Static loadbalancing using an outgoing hash. roundrobin - Distributes outgoing traffic using a round-robin scheduler through all active ports. This code was obtained from OpenBSD and this also includes 802.3ad LACP support from agr(4) in NetBSD.
2007-04-10 00:27:25 +00:00
{ 0, 0 },
};
Switch the LACP state machine over to its own mutex to protect the internals, this means that it no longer grabs the lagg rwlock. Use two port table arrays which list the active ports for Tx and switch between them with an atomic op. Now the lagg rwlock is only exclusively locked for management (ioctls) and queuing of lacp control frames isnt needed.
2008-03-16 19:25:30 +00:00
typedef void (*lacp_timer_func_t)(struct lacp_port *);
Add the trunk(4) driver for providing link aggregation, failover and fault tolerance. This driver allows aggregation of multiple network interfaces as one virtual interface using a number of different protocols/algorithms. failover - Sends traffic through the secondary port if the master becomes inactive. fec - Supports Cisco Fast EtherChannel. lacp - Supports the IEEE 802.3ad Link Aggregation Control Protocol (LACP) and the Marker Protocol. loadbalance - Static loadbalancing using an outgoing hash. roundrobin - Distributes outgoing traffic using a round-robin scheduler through all active ports. This code was obtained from OpenBSD and this also includes 802.3ad LACP support from agr(4) in NetBSD.
2007-04-10 00:27:25 +00:00
static void lacp_fill_actorinfo(struct lacp_port *, struct lacp_peerinfo *);
Implement the Marker Protocol. A marker frame is placed on the interface queue of each port and any further packets are blocked, when the all the marker frames have been returned to us from the remote network device then we can be sure that all interface queues are empty. This is needed when a port is added or removed from the aggregation since it will affect the hash based distribution, if the queues are not empty then a packet from an existing connection may be placed on a different interface and arrive out of order. This was previously achieved by suppressing transmission for 1 second, now that there is an active feedback this timeout as been increased to 3 seconds and used as a fallback.
2007-05-19 07:47:04 +00:00
static void lacp_fill_markerinfo(struct lacp_port *,
struct lacp_markerinfo *);
Add the trunk(4) driver for providing link aggregation, failover and fault tolerance. This driver allows aggregation of multiple network interfaces as one virtual interface using a number of different protocols/algorithms. failover - Sends traffic through the secondary port if the master becomes inactive. fec - Supports Cisco Fast EtherChannel. lacp - Supports the IEEE 802.3ad Link Aggregation Control Protocol (LACP) and the Marker Protocol. loadbalance - Static loadbalancing using an outgoing hash. roundrobin - Distributes outgoing traffic using a round-robin scheduler through all active ports. This code was obtained from OpenBSD and this also includes 802.3ad LACP support from agr(4) in NetBSD.
2007-04-10 00:27:25 +00:00
static uint64_t lacp_aggregator_bandwidth(struct lacp_aggregator *);
static void lacp_suppress_distributing(struct lacp_softc *,
struct lacp_aggregator *);
static void lacp_transit_expire(void *);
Switch the LACP state machine over to its own mutex to protect the internals, this means that it no longer grabs the lagg rwlock. Use two port table arrays which list the active ports for Tx and switch between them with an atomic op. Now the lagg rwlock is only exclusively locked for management (ioctls) and queuing of lacp control frames isnt needed.
2008-03-16 19:25:30 +00:00
static void lacp_update_portmap(struct lacp_softc *);
Add the trunk(4) driver for providing link aggregation, failover and fault tolerance. This driver allows aggregation of multiple network interfaces as one virtual interface using a number of different protocols/algorithms. failover - Sends traffic through the secondary port if the master becomes inactive. fec - Supports Cisco Fast EtherChannel. lacp - Supports the IEEE 802.3ad Link Aggregation Control Protocol (LACP) and the Marker Protocol. loadbalance - Static loadbalancing using an outgoing hash. roundrobin - Distributes outgoing traffic using a round-robin scheduler through all active ports. This code was obtained from OpenBSD and this also includes 802.3ad LACP support from agr(4) in NetBSD.
2007-04-10 00:27:25 +00:00
static void lacp_select_active_aggregator(struct lacp_softc *);
static uint16_t lacp_compose_key(struct lacp_port *);
static int tlv_check(const void *, size_t, const struct tlvhdr *,
const struct tlv_template *, boolean_t);
static void lacp_tick(void *);
static void lacp_fill_aggregator_id(struct lacp_aggregator *,
const struct lacp_port *);
static void lacp_fill_aggregator_id_peer(struct lacp_peerinfo *,
const struct lacp_peerinfo *);
static int lacp_aggregator_is_compatible(const struct lacp_aggregator *,
const struct lacp_port *);
static int lacp_peerinfo_is_compatible(const struct lacp_peerinfo *,
const struct lacp_peerinfo *);
static struct lacp_aggregator *lacp_aggregator_get(struct lacp_softc *,
struct lacp_port *);
static void lacp_aggregator_addref(struct lacp_softc *,
struct lacp_aggregator *);
static void lacp_aggregator_delref(struct lacp_softc *,
struct lacp_aggregator *);
/* receive machine */
Switch the LACP state machine over to its own mutex to protect the internals, this means that it no longer grabs the lagg rwlock. Use two port table arrays which list the active ports for Tx and switch between them with an atomic op. Now the lagg rwlock is only exclusively locked for management (ioctls) and queuing of lacp control frames isnt needed.
2008-03-16 19:25:30 +00:00
static int lacp_pdu_input(struct lacp_port *, struct mbuf *);
static int lacp_marker_input(struct lacp_port *, struct mbuf *);
Add the trunk(4) driver for providing link aggregation, failover and fault tolerance. This driver allows aggregation of multiple network interfaces as one virtual interface using a number of different protocols/algorithms. failover - Sends traffic through the secondary port if the master becomes inactive. fec - Supports Cisco Fast EtherChannel. lacp - Supports the IEEE 802.3ad Link Aggregation Control Protocol (LACP) and the Marker Protocol. loadbalance - Static loadbalancing using an outgoing hash. roundrobin - Distributes outgoing traffic using a round-robin scheduler through all active ports. This code was obtained from OpenBSD and this also includes 802.3ad LACP support from agr(4) in NetBSD.
2007-04-10 00:27:25 +00:00
static void lacp_sm_rx(struct lacp_port *, const struct lacpdu *);
static void lacp_sm_rx_timer(struct lacp_port *);
static void lacp_sm_rx_set_expired(struct lacp_port *);
static void lacp_sm_rx_update_ntt(struct lacp_port *,
const struct lacpdu *);
static void lacp_sm_rx_record_pdu(struct lacp_port *,
const struct lacpdu *);
static void lacp_sm_rx_update_selected(struct lacp_port *,
const struct lacpdu *);
static void lacp_sm_rx_record_default(struct lacp_port *);
static void lacp_sm_rx_update_default_selected(struct lacp_port *);
static void lacp_sm_rx_update_selected_from_peerinfo(struct lacp_port *,
const struct lacp_peerinfo *);
/* mux machine */
static void lacp_sm_mux(struct lacp_port *);
static void lacp_set_mux(struct lacp_port *, enum lacp_mux_state);
static void lacp_sm_mux_timer(struct lacp_port *);
/* periodic transmit machine */
static void lacp_sm_ptx_update_timeout(struct lacp_port *, uint8_t);
static void lacp_sm_ptx_tx_schedule(struct lacp_port *);
static void lacp_sm_ptx_timer(struct lacp_port *);
/* transmit machine */
static void lacp_sm_tx(struct lacp_port *);
static void lacp_sm_assert_ntt(struct lacp_port *);
static void lacp_run_timers(struct lacp_port *);
static int lacp_compare_peerinfo(const struct lacp_peerinfo *,
const struct lacp_peerinfo *);
static int lacp_compare_systemid(const struct lacp_systemid *,
const struct lacp_systemid *);
static void lacp_port_enable(struct lacp_port *);
static void lacp_port_disable(struct lacp_port *);
static void lacp_select(struct lacp_port *);
static void lacp_unselect(struct lacp_port *);
static void lacp_disable_collecting(struct lacp_port *);
static void lacp_enable_collecting(struct lacp_port *);
static void lacp_disable_distributing(struct lacp_port *);
static void lacp_enable_distributing(struct lacp_port *);
static int lacp_xmit_lacpdu(struct lacp_port *);
Implement the Marker Protocol. A marker frame is placed on the interface queue of each port and any further packets are blocked, when the all the marker frames have been returned to us from the remote network device then we can be sure that all interface queues are empty. This is needed when a port is added or removed from the aggregation since it will affect the hash based distribution, if the queues are not empty then a packet from an existing connection may be placed on a different interface and arrive out of order. This was previously achieved by suppressing transmission for 1 second, now that there is an active feedback this timeout as been increased to 3 seconds and used as a fallback.
2007-05-19 07:47:04 +00:00
static int lacp_xmit_marker(struct lacp_port *);
Add the trunk(4) driver for providing link aggregation, failover and fault tolerance. This driver allows aggregation of multiple network interfaces as one virtual interface using a number of different protocols/algorithms. failover - Sends traffic through the secondary port if the master becomes inactive. fec - Supports Cisco Fast EtherChannel. lacp - Supports the IEEE 802.3ad Link Aggregation Control Protocol (LACP) and the Marker Protocol. loadbalance - Static loadbalancing using an outgoing hash. roundrobin - Distributes outgoing traffic using a round-robin scheduler through all active ports. This code was obtained from OpenBSD and this also includes 802.3ad LACP support from agr(4) in NetBSD.
2007-04-10 00:27:25 +00:00
Turn LACP debugging from a compile time option to a sysctl, it is very handy to be able to turn it on when negotiation to a switch misbehaves. Submitted by: Andrew Boyer MFC after: 3 days
2012-05-26 08:09:01 +00:00
/* Debugging */
Add the trunk(4) driver for providing link aggregation, failover and fault tolerance. This driver allows aggregation of multiple network interfaces as one virtual interface using a number of different protocols/algorithms. failover - Sends traffic through the secondary port if the master becomes inactive. fec - Supports Cisco Fast EtherChannel. lacp - Supports the IEEE 802.3ad Link Aggregation Control Protocol (LACP) and the Marker Protocol. loadbalance - Static loadbalancing using an outgoing hash. roundrobin - Distributes outgoing traffic using a round-robin scheduler through all active ports. This code was obtained from OpenBSD and this also includes 802.3ad LACP support from agr(4) in NetBSD.
2007-04-10 00:27:25 +00:00
static void lacp_dump_lacpdu(const struct lacpdu *);
static const char *lacp_format_partner(const struct lacp_peerinfo *, char *,
size_t);
static const char *lacp_format_lagid(const struct lacp_peerinfo *,
const struct lacp_peerinfo *, char *, size_t);
static const char *lacp_format_lagid_aggregator(const struct lacp_aggregator *,
char *, size_t);
static const char *lacp_format_state(uint8_t, char *, size_t);
static const char *lacp_format_mac(const uint8_t *, char *, size_t);
static const char *lacp_format_systemid(const struct lacp_systemid *, char *,
size_t);
static const char *lacp_format_portid(const struct lacp_portid *, char *,
size_t);
static void lacp_dprintf(const struct lacp_port *, const char *, ...)
__attribute__((__format__(__printf__, 2, 3)));
Turn LACP debugging from a compile time option to a sysctl, it is very handy to be able to turn it on when negotiation to a switch misbehaves. Submitted by: Andrew Boyer MFC after: 3 days
2012-05-26 08:09:01 +00:00
Use the new VNET_DEFINE_STATIC macro when we are defining static VNET variables. Reviewed by: bz Sponsored by: DARPA, AFRL Differential Revision: https://reviews.freebsd.org/D16147
2018-07-24 16:35:52 +00:00
VNET_DEFINE_STATIC(int, lacp_debug);
Virtualize lagg(4) cloner. This change fixes a panic when tearing down if_lagg(4) interfaces which were cloned in a vnet jail. Sysctl nodes which are dynamically generated for each cloned interface (net.link.lagg.N.*) have been removed, and use_flowid and flowid_shift ifconfig(8) parameters have been added instead. Flags and per-interface statistics counters are displayed in "ifconfig -v". CR: D842
2014-10-01 21:37:32 +00:00
#define V_lacp_debug VNET(lacp_debug)
Bring over some link aggregation / LACP protocol improvements and debugging additions. * Add some new tracing events to aid in debugging. * Add in a debugging mode to drop transmit and received frames, specifically to test whether seeing or hearing heartbeats correctly cause LACP to drop the port. * Add in (and make default) a strict LACP mode, which requires the heartbeat on a port to be heard before it's used. Sometimes vendor ports will hang but the link layer stays up, resulting in hung traffic. * Add logging the number of link status flaps, again to aid in debugging badly behaving switch ports. * Calculate the lagg interface port speed as the multiple of the configured ports, rather than the largest. Obtained from: Netflix MFC after: 2 weeks
2013-07-13 04:25:03 +00:00
SYSCTL_NODE(_net_link_lagg, OID_AUTO, lacp, CTLFLAG_RD, 0, "ieee802.3ad");
Virtualize lagg(4) cloner. This change fixes a panic when tearing down if_lagg(4) interfaces which were cloned in a vnet jail. Sysctl nodes which are dynamically generated for each cloned interface (net.link.lagg.N.*) have been removed, and use_flowid and flowid_shift ifconfig(8) parameters have been added instead. Flags and per-interface statistics counters are displayed in "ifconfig -v". CR: D842
2014-10-01 21:37:32 +00:00
SYSCTL_INT(_net_link_lagg_lacp, OID_AUTO, debug, CTLFLAG_RWTUN | CTLFLAG_VNET,
&VNET_NAME(lacp_debug), 0, "Enable LACP debug logging (1=debug, 2=trace)");
Turn LACP debugging from a compile time option to a sysctl, it is very handy to be able to turn it on when negotiation to a switch misbehaves. Submitted by: Andrew Boyer MFC after: 3 days
2012-05-26 08:09:01 +00:00
Use the new VNET_DEFINE_STATIC macro when we are defining static VNET variables. Reviewed by: bz Sponsored by: DARPA, AFRL Differential Revision: https://reviews.freebsd.org/D16147
2018-07-24 16:35:52 +00:00
VNET_DEFINE_STATIC(int, lacp_default_strict_mode) = 1;
Added missing CTLFLAG_VNET to lacp default_strict_mode Added CTLFLAG_VNET to net.link.lagg.lacp.default_strict_mode which was missed in r290450. Reported by: julian@ MFC after: 1 week Sponsored by: Multiplay
2018-01-24 10:13:14 +00:00
SYSCTL_INT(_net_link_lagg_lacp, OID_AUTO, default_strict_mode,
CTLFLAG_RWTUN | CTLFLAG_VNET, &VNET_NAME(lacp_default_strict_mode), 0,
Add sysctl to control LACP strict compliance default Add net.link.lagg.lacp.default_strict_mode which defines the default value for LACP strict compliance for created lagg devices. Also: * Add lacp_strict option to ifconfig(8). * Fix lagg(4) creation examples. * Minor style(9) fix. MFC after: 1 week
2015-11-06 15:33:27 +00:00
"LACP strict protocol compliance default");
Virtualize lagg(4) cloner. This change fixes a panic when tearing down if_lagg(4) interfaces which were cloned in a vnet jail. Sysctl nodes which are dynamically generated for each cloned interface (net.link.lagg.N.*) have been removed, and use_flowid and flowid_shift ifconfig(8) parameters have been added instead. Flags and per-interface statistics counters are displayed in "ifconfig -v". CR: D842
2014-10-01 21:37:32 +00:00
#define LACP_DPRINTF(a) if (V_lacp_debug & 0x01) { lacp_dprintf a ; }
#define LACP_TRACE(a) if (V_lacp_debug & 0x02) { lacp_dprintf(a,"%s\n",__func__); }
#define LACP_TPRINTF(a) if (V_lacp_debug & 0x04) { lacp_dprintf a ; }
Add the trunk(4) driver for providing link aggregation, failover and fault tolerance. This driver allows aggregation of multiple network interfaces as one virtual interface using a number of different protocols/algorithms. failover - Sends traffic through the secondary port if the master becomes inactive. fec - Supports Cisco Fast EtherChannel. lacp - Supports the IEEE 802.3ad Link Aggregation Control Protocol (LACP) and the Marker Protocol. loadbalance - Static loadbalancing using an outgoing hash. roundrobin - Distributes outgoing traffic using a round-robin scheduler through all active ports. This code was obtained from OpenBSD and this also includes 802.3ad LACP support from agr(4) in NetBSD.
2007-04-10 00:27:25 +00:00
/*
* partner administration variables.
* XXX should be configurable.
*/
Bring over some link aggregation / LACP protocol improvements and debugging additions. * Add some new tracing events to aid in debugging. * Add in a debugging mode to drop transmit and received frames, specifically to test whether seeing or hearing heartbeats correctly cause LACP to drop the port. * Add in (and make default) a strict LACP mode, which requires the heartbeat on a port to be heard before it's used. Sometimes vendor ports will hang but the link layer stays up, resulting in hung traffic. * Add logging the number of link status flaps, again to aid in debugging badly behaving switch ports. * Calculate the lagg interface port speed as the multiple of the configured ports, rather than the largest. Obtained from: Netflix MFC after: 2 weeks
2013-07-13 04:25:03 +00:00
static const struct lacp_peerinfo lacp_partner_admin_optimistic = {
Add the trunk(4) driver for providing link aggregation, failover and fault tolerance. This driver allows aggregation of multiple network interfaces as one virtual interface using a number of different protocols/algorithms. failover - Sends traffic through the secondary port if the master becomes inactive. fec - Supports Cisco Fast EtherChannel. lacp - Supports the IEEE 802.3ad Link Aggregation Control Protocol (LACP) and the Marker Protocol. loadbalance - Static loadbalancing using an outgoing hash. roundrobin - Distributes outgoing traffic using a round-robin scheduler through all active ports. This code was obtained from OpenBSD and this also includes 802.3ad LACP support from agr(4) in NetBSD.
2007-04-10 00:27:25 +00:00
.lip_systemid = { .lsi_prio = 0xffff },
.lip_portid = { .lpi_prio = 0xffff },
.lip_state = LACP_STATE_SYNC | LACP_STATE_AGGREGATION |
LACP_STATE_COLLECTING | LACP_STATE_DISTRIBUTING,
Bring over some link aggregation / LACP protocol improvements and debugging additions. * Add some new tracing events to aid in debugging. * Add in a debugging mode to drop transmit and received frames, specifically to test whether seeing or hearing heartbeats correctly cause LACP to drop the port. * Add in (and make default) a strict LACP mode, which requires the heartbeat on a port to be heard before it's used. Sometimes vendor ports will hang but the link layer stays up, resulting in hung traffic. * Add logging the number of link status flaps, again to aid in debugging badly behaving switch ports. * Calculate the lagg interface port speed as the multiple of the configured ports, rather than the largest. Obtained from: Netflix MFC after: 2 weeks
2013-07-13 04:25:03 +00:00
};
static const struct lacp_peerinfo lacp_partner_admin_strict = {
.lip_systemid = { .lsi_prio = 0xffff },
.lip_portid = { .lpi_prio = 0xffff },
Add the trunk(4) driver for providing link aggregation, failover and fault tolerance. This driver allows aggregation of multiple network interfaces as one virtual interface using a number of different protocols/algorithms. failover - Sends traffic through the secondary port if the master becomes inactive. fec - Supports Cisco Fast EtherChannel. lacp - Supports the IEEE 802.3ad Link Aggregation Control Protocol (LACP) and the Marker Protocol. loadbalance - Static loadbalancing using an outgoing hash. roundrobin - Distributes outgoing traffic using a round-robin scheduler through all active ports. This code was obtained from OpenBSD and this also includes 802.3ad LACP support from agr(4) in NetBSD.
2007-04-10 00:27:25 +00:00
.lip_state = 0,
};
static const lacp_timer_func_t lacp_timer_funcs[LACP_NTIMER] = {
[LACP_TIMER_CURRENT_WHILE] = lacp_sm_rx_timer,
[LACP_TIMER_PERIODIC] = lacp_sm_ptx_timer,
[LACP_TIMER_WAIT_WHILE] = lacp_sm_mux_timer,
};
Pass any unmatched slowprotocols frames up the stack instead of dropping them, there are more subtypes than just LACP.
2007-12-31 01:16:35 +00:00
struct mbuf *
Change from a mutex to a read/write lock. This allows the tx port to be selected simultaneously by multiple senders and transmit/receive is not serialised between aggregated interfaces.
2007-05-15 07:41:46 +00:00
lacp_input(struct lagg_port *lgp, struct mbuf *m)
{
Switch the LACP state machine over to its own mutex to protect the internals, this means that it no longer grabs the lagg rwlock. Use two port table arrays which list the active ports for Tx and switch between them with an atomic op. Now the lagg rwlock is only exclusively locked for management (ioctls) and queuing of lacp control frames isnt needed.
2008-03-16 19:25:30 +00:00
struct lacp_port *lp = LACP_PORT(lgp);
Change from a mutex to a read/write lock. This allows the tx port to be selected simultaneously by multiple senders and transmit/receive is not serialised between aggregated interfaces.
2007-05-15 07:41:46 +00:00
uint8_t subtype;
if (m->m_pkthdr.len < sizeof(struct ether_header) + sizeof(subtype)) {
m_freem(m);
Pass any unmatched slowprotocols frames up the stack instead of dropping them, there are more subtypes than just LACP.
2007-12-31 01:16:35 +00:00
return (NULL);
Change from a mutex to a read/write lock. This allows the tx port to be selected simultaneously by multiple senders and transmit/receive is not serialised between aggregated interfaces.
2007-05-15 07:41:46 +00:00
}
m_copydata(m, sizeof(struct ether_header), sizeof(subtype), &subtype);
switch (subtype) {
case SLOWPROTOCOLS_SUBTYPE_LACP:
Switch the LACP state machine over to its own mutex to protect the internals, this means that it no longer grabs the lagg rwlock. Use two port table arrays which list the active ports for Tx and switch between them with an atomic op. Now the lagg rwlock is only exclusively locked for management (ioctls) and queuing of lacp control frames isnt needed.
2008-03-16 19:25:30 +00:00
lacp_pdu_input(lp, m);
return (NULL);
Change from a mutex to a read/write lock. This allows the tx port to be selected simultaneously by multiple senders and transmit/receive is not serialised between aggregated interfaces.
2007-05-15 07:41:46 +00:00
case SLOWPROTOCOLS_SUBTYPE_MARKER:
Switch the LACP state machine over to its own mutex to protect the internals, this means that it no longer grabs the lagg rwlock. Use two port table arrays which list the active ports for Tx and switch between them with an atomic op. Now the lagg rwlock is only exclusively locked for management (ioctls) and queuing of lacp control frames isnt needed.
2008-03-16 19:25:30 +00:00
lacp_marker_input(lp, m);
return (NULL);
Change from a mutex to a read/write lock. This allows the tx port to be selected simultaneously by multiple senders and transmit/receive is not serialised between aggregated interfaces.
2007-05-15 07:41:46 +00:00
}
Switch the LACP state machine over to its own mutex to protect the internals, this means that it no longer grabs the lagg rwlock. Use two port table arrays which list the active ports for Tx and switch between them with an atomic op. Now the lagg rwlock is only exclusively locked for management (ioctls) and queuing of lacp control frames isnt needed.
2008-03-16 19:25:30 +00:00
/* Not a subtype we are interested in */
return (m);
Change from a mutex to a read/write lock. This allows the tx port to be selected simultaneously by multiple senders and transmit/receive is not serialised between aggregated interfaces.
2007-05-15 07:41:46 +00:00
}
Add the trunk(4) driver for providing link aggregation, failover and fault tolerance. This driver allows aggregation of multiple network interfaces as one virtual interface using a number of different protocols/algorithms. failover - Sends traffic through the secondary port if the master becomes inactive. fec - Supports Cisco Fast EtherChannel. lacp - Supports the IEEE 802.3ad Link Aggregation Control Protocol (LACP) and the Marker Protocol. loadbalance - Static loadbalancing using an outgoing hash. roundrobin - Distributes outgoing traffic using a round-robin scheduler through all active ports. This code was obtained from OpenBSD and this also includes 802.3ad LACP support from agr(4) in NetBSD.
2007-04-10 00:27:25 +00:00
/*
Change from a mutex to a read/write lock. This allows the tx port to be selected simultaneously by multiple senders and transmit/receive is not serialised between aggregated interfaces.
2007-05-15 07:41:46 +00:00
* lacp_pdu_input: process lacpdu
Add the trunk(4) driver for providing link aggregation, failover and fault tolerance. This driver allows aggregation of multiple network interfaces as one virtual interface using a number of different protocols/algorithms. failover - Sends traffic through the secondary port if the master becomes inactive. fec - Supports Cisco Fast EtherChannel. lacp - Supports the IEEE 802.3ad Link Aggregation Control Protocol (LACP) and the Marker Protocol. loadbalance - Static loadbalancing using an outgoing hash. roundrobin - Distributes outgoing traffic using a round-robin scheduler through all active ports. This code was obtained from OpenBSD and this also includes 802.3ad LACP support from agr(4) in NetBSD.
2007-04-10 00:27:25 +00:00
*/
Change from a mutex to a read/write lock. This allows the tx port to be selected simultaneously by multiple senders and transmit/receive is not serialised between aggregated interfaces.
2007-05-15 07:41:46 +00:00
static int
Switch the LACP state machine over to its own mutex to protect the internals, this means that it no longer grabs the lagg rwlock. Use two port table arrays which list the active ports for Tx and switch between them with an atomic op. Now the lagg rwlock is only exclusively locked for management (ioctls) and queuing of lacp control frames isnt needed.
2008-03-16 19:25:30 +00:00
lacp_pdu_input(struct lacp_port *lp, struct mbuf *m)
Add the trunk(4) driver for providing link aggregation, failover and fault tolerance. This driver allows aggregation of multiple network interfaces as one virtual interface using a number of different protocols/algorithms. failover - Sends traffic through the secondary port if the master becomes inactive. fec - Supports Cisco Fast EtherChannel. lacp - Supports the IEEE 802.3ad Link Aggregation Control Protocol (LACP) and the Marker Protocol. loadbalance - Static loadbalancing using an outgoing hash. roundrobin - Distributes outgoing traffic using a round-robin scheduler through all active ports. This code was obtained from OpenBSD and this also includes 802.3ad LACP support from agr(4) in NetBSD.
2007-04-10 00:27:25 +00:00
{
Switch the LACP state machine over to its own mutex to protect the internals, this means that it no longer grabs the lagg rwlock. Use two port table arrays which list the active ports for Tx and switch between them with an atomic op. Now the lagg rwlock is only exclusively locked for management (ioctls) and queuing of lacp control frames isnt needed.
2008-03-16 19:25:30 +00:00
struct lacp_softc *lsc = lp->lp_lsc;
Add the trunk(4) driver for providing link aggregation, failover and fault tolerance. This driver allows aggregation of multiple network interfaces as one virtual interface using a number of different protocols/algorithms. failover - Sends traffic through the secondary port if the master becomes inactive. fec - Supports Cisco Fast EtherChannel. lacp - Supports the IEEE 802.3ad Link Aggregation Control Protocol (LACP) and the Marker Protocol. loadbalance - Static loadbalancing using an outgoing hash. roundrobin - Distributes outgoing traffic using a round-robin scheduler through all active ports. This code was obtained from OpenBSD and this also includes 802.3ad LACP support from agr(4) in NetBSD.
2007-04-10 00:27:25 +00:00
struct lacpdu *du;
int error = 0;
if (m->m_pkthdr.len != sizeof(*du)) {
goto bad;
}
if ((m->m_flags & M_MCAST) == 0) {
goto bad;
}
if (m->m_len < sizeof(*du)) {
m = m_pullup(m, sizeof(*du));
if (m == NULL) {
return (ENOMEM);
}
}
du = mtod(m, struct lacpdu *);
if (memcmp(&du->ldu_eh.ether_dhost,
&ethermulticastaddr_slowprotocols, ETHER_ADDR_LEN)) {
goto bad;
}
/*
* ignore the version for compatibility with
* the future protocol revisions.
*/
#if 0
if (du->ldu_sph.sph_version != 1) {
goto bad;
}
#endif
/*
* ignore tlv types for compatibility with
* the future protocol revisions.
*/
if (tlv_check(du, sizeof(*du), &du->ldu_tlv_actor,
lacp_info_tlv_template, FALSE)) {
goto bad;
}
Virtualize lagg(4) cloner. This change fixes a panic when tearing down if_lagg(4) interfaces which were cloned in a vnet jail. Sysctl nodes which are dynamically generated for each cloned interface (net.link.lagg.N.*) have been removed, and use_flowid and flowid_shift ifconfig(8) parameters have been added instead. Flags and per-interface statistics counters are displayed in "ifconfig -v". CR: D842
2014-10-01 21:37:32 +00:00
if (V_lacp_debug > 0) {
Turn LACP debugging from a compile time option to a sysctl, it is very handy to be able to turn it on when negotiation to a switch misbehaves. Submitted by: Andrew Boyer MFC after: 3 days
2012-05-26 08:09:01 +00:00
lacp_dprintf(lp, "lacpdu receive\n");
lacp_dump_lacpdu(du);
}
Switch the LACP state machine over to its own mutex to protect the internals, this means that it no longer grabs the lagg rwlock. Use two port table arrays which list the active ports for Tx and switch between them with an atomic op. Now the lagg rwlock is only exclusively locked for management (ioctls) and queuing of lacp control frames isnt needed.
2008-03-16 19:25:30 +00:00
Break out the static, global LACP debug options into a per-lagg unit sysctl tree. * Create a net.link.lagg.X.lacp node * Add a debug node under that for tx_test and rx_test * Add lacp_strict_mode, defaulting to 1 tx_test and rx_test are still a bitmap of unit numbers for now. At some point it would be nice to create child nodes of the lagg bundle for each sub-interface, and then populate those with various knobs and statistics. Sponsored by: Netflix
2013-07-26 19:41:13 +00:00
if ((1 << lp->lp_ifp->if_dunit) & lp->lp_lsc->lsc_debug.lsc_rx_test) {
Bring over some link aggregation / LACP protocol improvements and debugging additions. * Add some new tracing events to aid in debugging. * Add in a debugging mode to drop transmit and received frames, specifically to test whether seeing or hearing heartbeats correctly cause LACP to drop the port. * Add in (and make default) a strict LACP mode, which requires the heartbeat on a port to be heard before it's used. Sometimes vendor ports will hang but the link layer stays up, resulting in hung traffic. * Add logging the number of link status flaps, again to aid in debugging badly behaving switch ports. * Calculate the lagg interface port speed as the multiple of the configured ports, rather than the largest. Obtained from: Netflix MFC after: 2 weeks
2013-07-13 04:25:03 +00:00
LACP_TPRINTF((lp, "Dropping RX PDU\n"));
goto bad;
}
Switch the LACP state machine over to its own mutex to protect the internals, this means that it no longer grabs the lagg rwlock. Use two port table arrays which list the active ports for Tx and switch between them with an atomic op. Now the lagg rwlock is only exclusively locked for management (ioctls) and queuing of lacp control frames isnt needed.
2008-03-16 19:25:30 +00:00
LACP_LOCK(lsc);
Add the trunk(4) driver for providing link aggregation, failover and fault tolerance. This driver allows aggregation of multiple network interfaces as one virtual interface using a number of different protocols/algorithms. failover - Sends traffic through the secondary port if the master becomes inactive. fec - Supports Cisco Fast EtherChannel. lacp - Supports the IEEE 802.3ad Link Aggregation Control Protocol (LACP) and the Marker Protocol. loadbalance - Static loadbalancing using an outgoing hash. roundrobin - Distributes outgoing traffic using a round-robin scheduler through all active ports. This code was obtained from OpenBSD and this also includes 802.3ad LACP support from agr(4) in NetBSD.
2007-04-10 00:27:25 +00:00
lacp_sm_rx(lp, du);
Switch the LACP state machine over to its own mutex to protect the internals, this means that it no longer grabs the lagg rwlock. Use two port table arrays which list the active ports for Tx and switch between them with an atomic op. Now the lagg rwlock is only exclusively locked for management (ioctls) and queuing of lacp control frames isnt needed.
2008-03-16 19:25:30 +00:00
LACP_UNLOCK(lsc);
Add the trunk(4) driver for providing link aggregation, failover and fault tolerance. This driver allows aggregation of multiple network interfaces as one virtual interface using a number of different protocols/algorithms. failover - Sends traffic through the secondary port if the master becomes inactive. fec - Supports Cisco Fast EtherChannel. lacp - Supports the IEEE 802.3ad Link Aggregation Control Protocol (LACP) and the Marker Protocol. loadbalance - Static loadbalancing using an outgoing hash. roundrobin - Distributes outgoing traffic using a round-robin scheduler through all active ports. This code was obtained from OpenBSD and this also includes 802.3ad LACP support from agr(4) in NetBSD.
2007-04-10 00:27:25 +00:00
m_freem(m);
return (error);
bad:
m_freem(m);
return (EINVAL);
}
static void
lacp_fill_actorinfo(struct lacp_port *lp, struct lacp_peerinfo *info)
{
Rename the trunk(4) driver to lagg(4) as it is too similar to vlan trunking. The name trunk is misused as the networking term trunk means carrying multiple VLANs over a single connection. The IEEE standard for link aggregation (802.3 section 3) does not talk about 'trunk' at all while it is used throughout IEEE 802.1Q in describing vlans. The lagg(4) driver provides link aggregation, failover and fault tolerance. Discussed on: current@
2007-04-17 00:35:11 +00:00
struct lagg_port *lgp = lp->lp_lagg;
non-functional cleanup - remove dead code - use consistent variable names - gc unused defines - whitespace cleanup
2007-06-12 07:29:11 +00:00
struct lagg_softc *sc = lgp->lp_softc;
Add the trunk(4) driver for providing link aggregation, failover and fault tolerance. This driver allows aggregation of multiple network interfaces as one virtual interface using a number of different protocols/algorithms. failover - Sends traffic through the secondary port if the master becomes inactive. fec - Supports Cisco Fast EtherChannel. lacp - Supports the IEEE 802.3ad Link Aggregation Control Protocol (LACP) and the Marker Protocol. loadbalance - Static loadbalancing using an outgoing hash. roundrobin - Distributes outgoing traffic using a round-robin scheduler through all active ports. This code was obtained from OpenBSD and this also includes 802.3ad LACP support from agr(4) in NetBSD.
2007-04-10 00:27:25 +00:00
info->lip_systemid.lsi_prio = htons(LACP_SYSTEM_PRIO);
memcpy(&info->lip_systemid.lsi_mac,
non-functional cleanup - remove dead code - use consistent variable names - gc unused defines - whitespace cleanup
2007-06-12 07:29:11 +00:00
IF_LLADDR(sc->sc_ifp), ETHER_ADDR_LEN);
Add the trunk(4) driver for providing link aggregation, failover and fault tolerance. This driver allows aggregation of multiple network interfaces as one virtual interface using a number of different protocols/algorithms. failover - Sends traffic through the secondary port if the master becomes inactive. fec - Supports Cisco Fast EtherChannel. lacp - Supports the IEEE 802.3ad Link Aggregation Control Protocol (LACP) and the Marker Protocol. loadbalance - Static loadbalancing using an outgoing hash. roundrobin - Distributes outgoing traffic using a round-robin scheduler through all active ports. This code was obtained from OpenBSD and this also includes 802.3ad LACP support from agr(4) in NetBSD.
2007-04-10 00:27:25 +00:00
info->lip_portid.lpi_prio = htons(LACP_PORT_PRIO);
info->lip_portid.lpi_portno = htons(lp->lp_ifp->if_index);
info->lip_state = lp->lp_state;
}
Implement the Marker Protocol. A marker frame is placed on the interface queue of each port and any further packets are blocked, when the all the marker frames have been returned to us from the remote network device then we can be sure that all interface queues are empty. This is needed when a port is added or removed from the aggregation since it will affect the hash based distribution, if the queues are not empty then a packet from an existing connection may be placed on a different interface and arrive out of order. This was previously achieved by suppressing transmission for 1 second, now that there is an active feedback this timeout as been increased to 3 seconds and used as a fallback.
2007-05-19 07:47:04 +00:00
static void
lacp_fill_markerinfo(struct lacp_port *lp, struct lacp_markerinfo *info)
{
struct ifnet *ifp = lp->lp_ifp;
/* Fill in the port index and system id (encoded as the MAC) */
info->mi_rq_port = htons(ifp->if_index);
memcpy(&info->mi_rq_system, lp->lp_systemid.lsi_mac, ETHER_ADDR_LEN);
info->mi_rq_xid = htonl(0);
}
Add the trunk(4) driver for providing link aggregation, failover and fault tolerance. This driver allows aggregation of multiple network interfaces as one virtual interface using a number of different protocols/algorithms. failover - Sends traffic through the secondary port if the master becomes inactive. fec - Supports Cisco Fast EtherChannel. lacp - Supports the IEEE 802.3ad Link Aggregation Control Protocol (LACP) and the Marker Protocol. loadbalance - Static loadbalancing using an outgoing hash. roundrobin - Distributes outgoing traffic using a round-robin scheduler through all active ports. This code was obtained from OpenBSD and this also includes 802.3ad LACP support from agr(4) in NetBSD.
2007-04-10 00:27:25 +00:00
static int
lacp_xmit_lacpdu(struct lacp_port *lp)
{
Rename the trunk(4) driver to lagg(4) as it is too similar to vlan trunking. The name trunk is misused as the networking term trunk means carrying multiple VLANs over a single connection. The IEEE standard for link aggregation (802.3 section 3) does not talk about 'trunk' at all while it is used throughout IEEE 802.1Q in describing vlans. The lagg(4) driver provides link aggregation, failover and fault tolerance. Discussed on: current@
2007-04-17 00:35:11 +00:00
struct lagg_port *lgp = lp->lp_lagg;
Add the trunk(4) driver for providing link aggregation, failover and fault tolerance. This driver allows aggregation of multiple network interfaces as one virtual interface using a number of different protocols/algorithms. failover - Sends traffic through the secondary port if the master becomes inactive. fec - Supports Cisco Fast EtherChannel. lacp - Supports the IEEE 802.3ad Link Aggregation Control Protocol (LACP) and the Marker Protocol. loadbalance - Static loadbalancing using an outgoing hash. roundrobin - Distributes outgoing traffic using a round-robin scheduler through all active ports. This code was obtained from OpenBSD and this also includes 802.3ad LACP support from agr(4) in NetBSD.
2007-04-10 00:27:25 +00:00
struct mbuf *m;
struct lacpdu *du;
int error;
Switch the LACP state machine over to its own mutex to protect the internals, this means that it no longer grabs the lagg rwlock. Use two port table arrays which list the active ports for Tx and switch between them with an atomic op. Now the lagg rwlock is only exclusively locked for management (ioctls) and queuing of lacp control frames isnt needed.
2008-03-16 19:25:30 +00:00
LACP_LOCK_ASSERT(lp->lp_lsc);
Add the trunk(4) driver for providing link aggregation, failover and fault tolerance. This driver allows aggregation of multiple network interfaces as one virtual interface using a number of different protocols/algorithms. failover - Sends traffic through the secondary port if the master becomes inactive. fec - Supports Cisco Fast EtherChannel. lacp - Supports the IEEE 802.3ad Link Aggregation Control Protocol (LACP) and the Marker Protocol. loadbalance - Static loadbalancing using an outgoing hash. roundrobin - Distributes outgoing traffic using a round-robin scheduler through all active ports. This code was obtained from OpenBSD and this also includes 802.3ad LACP support from agr(4) in NetBSD.
2007-04-10 00:27:25 +00:00
Mechanically substitute flags from historic mbuf allocator with malloc(9) flags within sys. Exceptions: - sys/contrib not touched - sys/mbuf.h edited manually
2012-12-05 08:04:20 +00:00
m = m_gethdr(M_NOWAIT, MT_DATA);
Add the trunk(4) driver for providing link aggregation, failover and fault tolerance. This driver allows aggregation of multiple network interfaces as one virtual interface using a number of different protocols/algorithms. failover - Sends traffic through the secondary port if the master becomes inactive. fec - Supports Cisco Fast EtherChannel. lacp - Supports the IEEE 802.3ad Link Aggregation Control Protocol (LACP) and the Marker Protocol. loadbalance - Static loadbalancing using an outgoing hash. roundrobin - Distributes outgoing traffic using a round-robin scheduler through all active ports. This code was obtained from OpenBSD and this also includes 802.3ad LACP support from agr(4) in NetBSD.
2007-04-10 00:27:25 +00:00
if (m == NULL) {
return (ENOMEM);
}
m->m_len = m->m_pkthdr.len = sizeof(*du);
du = mtod(m, struct lacpdu *);
memset(du, 0, sizeof(*du));
memcpy(&du->ldu_eh.ether_dhost, ethermulticastaddr_slowprotocols,
ETHER_ADDR_LEN);
Rename the trunk(4) driver to lagg(4) as it is too similar to vlan trunking. The name trunk is misused as the networking term trunk means carrying multiple VLANs over a single connection. The IEEE standard for link aggregation (802.3 section 3) does not talk about 'trunk' at all while it is used throughout IEEE 802.1Q in describing vlans. The lagg(4) driver provides link aggregation, failover and fault tolerance. Discussed on: current@
2007-04-17 00:35:11 +00:00
memcpy(&du->ldu_eh.ether_shost, lgp->lp_lladdr, ETHER_ADDR_LEN);
Add the trunk(4) driver for providing link aggregation, failover and fault tolerance. This driver allows aggregation of multiple network interfaces as one virtual interface using a number of different protocols/algorithms. failover - Sends traffic through the secondary port if the master becomes inactive. fec - Supports Cisco Fast EtherChannel. lacp - Supports the IEEE 802.3ad Link Aggregation Control Protocol (LACP) and the Marker Protocol. loadbalance - Static loadbalancing using an outgoing hash. roundrobin - Distributes outgoing traffic using a round-robin scheduler through all active ports. This code was obtained from OpenBSD and this also includes 802.3ad LACP support from agr(4) in NetBSD.
2007-04-10 00:27:25 +00:00
du->ldu_eh.ether_type = htons(ETHERTYPE_SLOW);
du->ldu_sph.sph_subtype = SLOWPROTOCOLS_SUBTYPE_LACP;
du->ldu_sph.sph_version = 1;
TLV_SET(&du->ldu_tlv_actor, LACP_TYPE_ACTORINFO, sizeof(du->ldu_actor));
du->ldu_actor = lp->lp_actor;
TLV_SET(&du->ldu_tlv_partner, LACP_TYPE_PARTNERINFO,
sizeof(du->ldu_partner));
du->ldu_partner = lp->lp_partner;
TLV_SET(&du->ldu_tlv_collector, LACP_TYPE_COLLECTORINFO,
sizeof(du->ldu_collector));
du->ldu_collector.lci_maxdelay = 0;
Virtualize lagg(4) cloner. This change fixes a panic when tearing down if_lagg(4) interfaces which were cloned in a vnet jail. Sysctl nodes which are dynamically generated for each cloned interface (net.link.lagg.N.*) have been removed, and use_flowid and flowid_shift ifconfig(8) parameters have been added instead. Flags and per-interface statistics counters are displayed in "ifconfig -v". CR: D842
2014-10-01 21:37:32 +00:00
if (V_lacp_debug > 0) {
Turn LACP debugging from a compile time option to a sysctl, it is very handy to be able to turn it on when negotiation to a switch misbehaves. Submitted by: Andrew Boyer MFC after: 3 days
2012-05-26 08:09:01 +00:00
lacp_dprintf(lp, "lacpdu transmit\n");
lacp_dump_lacpdu(du);
}
Add the trunk(4) driver for providing link aggregation, failover and fault tolerance. This driver allows aggregation of multiple network interfaces as one virtual interface using a number of different protocols/algorithms. failover - Sends traffic through the secondary port if the master becomes inactive. fec - Supports Cisco Fast EtherChannel. lacp - Supports the IEEE 802.3ad Link Aggregation Control Protocol (LACP) and the Marker Protocol. loadbalance - Static loadbalancing using an outgoing hash. roundrobin - Distributes outgoing traffic using a round-robin scheduler through all active ports. This code was obtained from OpenBSD and this also includes 802.3ad LACP support from agr(4) in NetBSD.
2007-04-10 00:27:25 +00:00
m->m_flags |= M_MCAST;
/*
* XXX should use higher priority queue.
* otherwise network congestion can break aggregation.
*/
Rename the trunk(4) driver to lagg(4) as it is too similar to vlan trunking. The name trunk is misused as the networking term trunk means carrying multiple VLANs over a single connection. The IEEE standard for link aggregation (802.3 section 3) does not talk about 'trunk' at all while it is used throughout IEEE 802.1Q in describing vlans. The lagg(4) driver provides link aggregation, failover and fault tolerance. Discussed on: current@
2007-04-17 00:35:11 +00:00
error = lagg_enqueue(lp->lp_ifp, m);
Add the trunk(4) driver for providing link aggregation, failover and fault tolerance. This driver allows aggregation of multiple network interfaces as one virtual interface using a number of different protocols/algorithms. failover - Sends traffic through the secondary port if the master becomes inactive. fec - Supports Cisco Fast EtherChannel. lacp - Supports the IEEE 802.3ad Link Aggregation Control Protocol (LACP) and the Marker Protocol. loadbalance - Static loadbalancing using an outgoing hash. roundrobin - Distributes outgoing traffic using a round-robin scheduler through all active ports. This code was obtained from OpenBSD and this also includes 802.3ad LACP support from agr(4) in NetBSD.
2007-04-10 00:27:25 +00:00
return (error);
}
Implement the Marker Protocol. A marker frame is placed on the interface queue of each port and any further packets are blocked, when the all the marker frames have been returned to us from the remote network device then we can be sure that all interface queues are empty. This is needed when a port is added or removed from the aggregation since it will affect the hash based distribution, if the queues are not empty then a packet from an existing connection may be placed on a different interface and arrive out of order. This was previously achieved by suppressing transmission for 1 second, now that there is an active feedback this timeout as been increased to 3 seconds and used as a fallback.
2007-05-19 07:47:04 +00:00
static int
lacp_xmit_marker(struct lacp_port *lp)
{
struct lagg_port *lgp = lp->lp_lagg;
struct mbuf *m;
struct markerdu *mdu;
int error;
Switch the LACP state machine over to its own mutex to protect the internals, this means that it no longer grabs the lagg rwlock. Use two port table arrays which list the active ports for Tx and switch between them with an atomic op. Now the lagg rwlock is only exclusively locked for management (ioctls) and queuing of lacp control frames isnt needed.
2008-03-16 19:25:30 +00:00
LACP_LOCK_ASSERT(lp->lp_lsc);
Implement the Marker Protocol. A marker frame is placed on the interface queue of each port and any further packets are blocked, when the all the marker frames have been returned to us from the remote network device then we can be sure that all interface queues are empty. This is needed when a port is added or removed from the aggregation since it will affect the hash based distribution, if the queues are not empty then a packet from an existing connection may be placed on a different interface and arrive out of order. This was previously achieved by suppressing transmission for 1 second, now that there is an active feedback this timeout as been increased to 3 seconds and used as a fallback.
2007-05-19 07:47:04 +00:00
Mechanically substitute flags from historic mbuf allocator with malloc(9) flags within sys. Exceptions: - sys/contrib not touched - sys/mbuf.h edited manually
2012-12-05 08:04:20 +00:00
m = m_gethdr(M_NOWAIT, MT_DATA);
Implement the Marker Protocol. A marker frame is placed on the interface queue of each port and any further packets are blocked, when the all the marker frames have been returned to us from the remote network device then we can be sure that all interface queues are empty. This is needed when a port is added or removed from the aggregation since it will affect the hash based distribution, if the queues are not empty then a packet from an existing connection may be placed on a different interface and arrive out of order. This was previously achieved by suppressing transmission for 1 second, now that there is an active feedback this timeout as been increased to 3 seconds and used as a fallback.
2007-05-19 07:47:04 +00:00
if (m == NULL) {
return (ENOMEM);
}
m->m_len = m->m_pkthdr.len = sizeof(*mdu);
mdu = mtod(m, struct markerdu *);
memset(mdu, 0, sizeof(*mdu));
memcpy(&mdu->mdu_eh.ether_dhost, ethermulticastaddr_slowprotocols,
ETHER_ADDR_LEN);
memcpy(&mdu->mdu_eh.ether_shost, lgp->lp_lladdr, ETHER_ADDR_LEN);
mdu->mdu_eh.ether_type = htons(ETHERTYPE_SLOW);
mdu->mdu_sph.sph_subtype = SLOWPROTOCOLS_SUBTYPE_MARKER;
mdu->mdu_sph.sph_version = 1;
/* Bump the transaction id and copy over the marker info */
lp->lp_marker.mi_rq_xid = htonl(ntohl(lp->lp_marker.mi_rq_xid) + 1);
TLV_SET(&mdu->mdu_tlv, MARKER_TYPE_INFO, sizeof(mdu->mdu_info));
mdu->mdu_info = lp->lp_marker;
LACP_DPRINTF((lp, "marker transmit, port=%u, sys=%6D, id=%u\n",
ntohs(mdu->mdu_info.mi_rq_port), mdu->mdu_info.mi_rq_system, ":",
ntohl(mdu->mdu_info.mi_rq_xid)));
m->m_flags |= M_MCAST;
error = lagg_enqueue(lp->lp_ifp, m);
return (error);
}
Switch the LACP state machine over to its own mutex to protect the internals, this means that it no longer grabs the lagg rwlock. Use two port table arrays which list the active ports for Tx and switch between them with an atomic op. Now the lagg rwlock is only exclusively locked for management (ioctls) and queuing of lacp control frames isnt needed.
2008-03-16 19:25:30 +00:00
Add the trunk(4) driver for providing link aggregation, failover and fault tolerance. This driver allows aggregation of multiple network interfaces as one virtual interface using a number of different protocols/algorithms. failover - Sends traffic through the secondary port if the master becomes inactive. fec - Supports Cisco Fast EtherChannel. lacp - Supports the IEEE 802.3ad Link Aggregation Control Protocol (LACP) and the Marker Protocol. loadbalance - Static loadbalancing using an outgoing hash. roundrobin - Distributes outgoing traffic using a round-robin scheduler through all active ports. This code was obtained from OpenBSD and this also includes 802.3ad LACP support from agr(4) in NetBSD.
2007-04-10 00:27:25 +00:00
void
Rename the trunk(4) driver to lagg(4) as it is too similar to vlan trunking. The name trunk is misused as the networking term trunk means carrying multiple VLANs over a single connection. The IEEE standard for link aggregation (802.3 section 3) does not talk about 'trunk' at all while it is used throughout IEEE 802.1Q in describing vlans. The lagg(4) driver provides link aggregation, failover and fault tolerance. Discussed on: current@
2007-04-17 00:35:11 +00:00
lacp_linkstate(struct lagg_port *lgp)
Add the trunk(4) driver for providing link aggregation, failover and fault tolerance. This driver allows aggregation of multiple network interfaces as one virtual interface using a number of different protocols/algorithms. failover - Sends traffic through the secondary port if the master becomes inactive. fec - Supports Cisco Fast EtherChannel. lacp - Supports the IEEE 802.3ad Link Aggregation Control Protocol (LACP) and the Marker Protocol. loadbalance - Static loadbalancing using an outgoing hash. roundrobin - Distributes outgoing traffic using a round-robin scheduler through all active ports. This code was obtained from OpenBSD and this also includes 802.3ad LACP support from agr(4) in NetBSD.
2007-04-10 00:27:25 +00:00
{
Rename the trunk(4) driver to lagg(4) as it is too similar to vlan trunking. The name trunk is misused as the networking term trunk means carrying multiple VLANs over a single connection. The IEEE standard for link aggregation (802.3 section 3) does not talk about 'trunk' at all while it is used throughout IEEE 802.1Q in describing vlans. The lagg(4) driver provides link aggregation, failover and fault tolerance. Discussed on: current@
2007-04-17 00:35:11 +00:00
struct lacp_port *lp = LACP_PORT(lgp);
Switch the LACP state machine over to its own mutex to protect the internals, this means that it no longer grabs the lagg rwlock. Use two port table arrays which list the active ports for Tx and switch between them with an atomic op. Now the lagg rwlock is only exclusively locked for management (ioctls) and queuing of lacp control frames isnt needed.
2008-03-16 19:25:30 +00:00
struct lacp_softc *lsc = lp->lp_lsc;
Rename the trunk(4) driver to lagg(4) as it is too similar to vlan trunking. The name trunk is misused as the networking term trunk means carrying multiple VLANs over a single connection. The IEEE standard for link aggregation (802.3 section 3) does not talk about 'trunk' at all while it is used throughout IEEE 802.1Q in describing vlans. The lagg(4) driver provides link aggregation, failover and fault tolerance. Discussed on: current@
2007-04-17 00:35:11 +00:00
struct ifnet *ifp = lgp->lp_ifp;
Add the trunk(4) driver for providing link aggregation, failover and fault tolerance. This driver allows aggregation of multiple network interfaces as one virtual interface using a number of different protocols/algorithms. failover - Sends traffic through the secondary port if the master becomes inactive. fec - Supports Cisco Fast EtherChannel. lacp - Supports the IEEE 802.3ad Link Aggregation Control Protocol (LACP) and the Marker Protocol. loadbalance - Static loadbalancing using an outgoing hash. roundrobin - Distributes outgoing traffic using a round-robin scheduler through all active ports. This code was obtained from OpenBSD and this also includes 802.3ad LACP support from agr(4) in NetBSD.
2007-04-10 00:27:25 +00:00
struct ifmediareq ifmr;
int error = 0;
u_int media;
uint8_t old_state;
uint16_t old_key;
bzero((char *)&ifmr, sizeof(ifmr));
Make LACP based lagg work with interfaces (like 100Gbps and 25Gbps) that report extended media types. lacp_aggregator_bandwidth() uses the media to determine the speed of the interface and returns 0 for IFM_OTHER without the bits in the extended range. Reported by: kbowling@ Reviewed by: eugen_grosbein.net, mjoras@ MFC after: 1 week Sponsored by: Chelsio Communications Differential Revision: https://reviews.freebsd.org/D12188
2017-09-06 14:36:35 +00:00
error = (*ifp->if_ioctl)(ifp, SIOCGIFXMEDIA, (caddr_t)&ifmr);
if (error != 0) {
bzero((char *)&ifmr, sizeof(ifmr));
error = (*ifp->if_ioctl)(ifp, SIOCGIFMEDIA, (caddr_t)&ifmr);
}
Add the trunk(4) driver for providing link aggregation, failover and fault tolerance. This driver allows aggregation of multiple network interfaces as one virtual interface using a number of different protocols/algorithms. failover - Sends traffic through the secondary port if the master becomes inactive. fec - Supports Cisco Fast EtherChannel. lacp - Supports the IEEE 802.3ad Link Aggregation Control Protocol (LACP) and the Marker Protocol. loadbalance - Static loadbalancing using an outgoing hash. roundrobin - Distributes outgoing traffic using a round-robin scheduler through all active ports. This code was obtained from OpenBSD and this also includes 802.3ad LACP support from agr(4) in NetBSD.
2007-04-10 00:27:25 +00:00
if (error != 0)
return;
Switch the LACP state machine over to its own mutex to protect the internals, this means that it no longer grabs the lagg rwlock. Use two port table arrays which list the active ports for Tx and switch between them with an atomic op. Now the lagg rwlock is only exclusively locked for management (ioctls) and queuing of lacp control frames isnt needed.
2008-03-16 19:25:30 +00:00
LACP_LOCK(lsc);
Add the trunk(4) driver for providing link aggregation, failover and fault tolerance. This driver allows aggregation of multiple network interfaces as one virtual interface using a number of different protocols/algorithms. failover - Sends traffic through the secondary port if the master becomes inactive. fec - Supports Cisco Fast EtherChannel. lacp - Supports the IEEE 802.3ad Link Aggregation Control Protocol (LACP) and the Marker Protocol. loadbalance - Static loadbalancing using an outgoing hash. roundrobin - Distributes outgoing traffic using a round-robin scheduler through all active ports. This code was obtained from OpenBSD and this also includes 802.3ad LACP support from agr(4) in NetBSD.
2007-04-10 00:27:25 +00:00
media = ifmr.ifm_active;
- Add a disabled state for ports that can not be aggregated - Refine check for lacp links, set to disabled if not suitable
2007-05-03 08:56:20 +00:00
LACP_DPRINTF((lp, "media changed 0x%x -> 0x%x, ether = %d, fdx = %d, "
"link = %d\n", lp->lp_media, media, IFM_TYPE(media) == IFM_ETHER,
(media & IFM_FDX) != 0, ifp->if_link_state == LINK_STATE_UP));
Add the trunk(4) driver for providing link aggregation, failover and fault tolerance. This driver allows aggregation of multiple network interfaces as one virtual interface using a number of different protocols/algorithms. failover - Sends traffic through the secondary port if the master becomes inactive. fec - Supports Cisco Fast EtherChannel. lacp - Supports the IEEE 802.3ad Link Aggregation Control Protocol (LACP) and the Marker Protocol. loadbalance - Static loadbalancing using an outgoing hash. roundrobin - Distributes outgoing traffic using a round-robin scheduler through all active ports. This code was obtained from OpenBSD and this also includes 802.3ad LACP support from agr(4) in NetBSD.
2007-04-10 00:27:25 +00:00
old_state = lp->lp_state;
old_key = lp->lp_key;
lp->lp_media = media;
non-functional cleanup - remove dead code - use consistent variable names - gc unused defines - whitespace cleanup
2007-06-12 07:29:11 +00:00
/*
- Add a disabled state for ports that can not be aggregated - Refine check for lacp links, set to disabled if not suitable
2007-05-03 08:56:20 +00:00
* If the port is not an active full duplex Ethernet link then it can
* not be aggregated.
*/
if (IFM_TYPE(media) != IFM_ETHER || (media & IFM_FDX) == 0 ||
ifp->if_link_state != LINK_STATE_UP) {
Add the trunk(4) driver for providing link aggregation, failover and fault tolerance. This driver allows aggregation of multiple network interfaces as one virtual interface using a number of different protocols/algorithms. failover - Sends traffic through the secondary port if the master becomes inactive. fec - Supports Cisco Fast EtherChannel. lacp - Supports the IEEE 802.3ad Link Aggregation Control Protocol (LACP) and the Marker Protocol. loadbalance - Static loadbalancing using an outgoing hash. roundrobin - Distributes outgoing traffic using a round-robin scheduler through all active ports. This code was obtained from OpenBSD and this also includes 802.3ad LACP support from agr(4) in NetBSD.
2007-04-10 00:27:25 +00:00
lacp_port_disable(lp);
} else {
lacp_port_enable(lp);
}
lp->lp_key = lacp_compose_key(lp);
if (old_state != lp->lp_state || old_key != lp->lp_key) {
LACP_DPRINTF((lp, "-> UNSELECTED\n"));
lp->lp_selected = LACP_UNSELECTED;
}
Switch the LACP state machine over to its own mutex to protect the internals, this means that it no longer grabs the lagg rwlock. Use two port table arrays which list the active ports for Tx and switch between them with an atomic op. Now the lagg rwlock is only exclusively locked for management (ioctls) and queuing of lacp control frames isnt needed.
2008-03-16 19:25:30 +00:00
LACP_UNLOCK(lsc);
Add the trunk(4) driver for providing link aggregation, failover and fault tolerance. This driver allows aggregation of multiple network interfaces as one virtual interface using a number of different protocols/algorithms. failover - Sends traffic through the secondary port if the master becomes inactive. fec - Supports Cisco Fast EtherChannel. lacp - Supports the IEEE 802.3ad Link Aggregation Control Protocol (LACP) and the Marker Protocol. loadbalance - Static loadbalancing using an outgoing hash. roundrobin - Distributes outgoing traffic using a round-robin scheduler through all active ports. This code was obtained from OpenBSD and this also includes 802.3ad LACP support from agr(4) in NetBSD.
2007-04-10 00:27:25 +00:00
}
static void
lacp_tick(void *arg)
{
struct lacp_softc *lsc = arg;
struct lacp_port *lp;
LIST_FOREACH(lp, &lsc->lsc_ports, lp_next) {
if ((lp->lp_state & LACP_STATE_AGGREGATION) == 0)
continue;
Virtualize lagg(4) cloner. This change fixes a panic when tearing down if_lagg(4) interfaces which were cloned in a vnet jail. Sysctl nodes which are dynamically generated for each cloned interface (net.link.lagg.N.*) have been removed, and use_flowid and flowid_shift ifconfig(8) parameters have been added instead. Flags and per-interface statistics counters are displayed in "ifconfig -v". CR: D842
2014-10-01 21:37:32 +00:00
CURVNET_SET(lp->lp_ifp->if_vnet);
Add the trunk(4) driver for providing link aggregation, failover and fault tolerance. This driver allows aggregation of multiple network interfaces as one virtual interface using a number of different protocols/algorithms. failover - Sends traffic through the secondary port if the master becomes inactive. fec - Supports Cisco Fast EtherChannel. lacp - Supports the IEEE 802.3ad Link Aggregation Control Protocol (LACP) and the Marker Protocol. loadbalance - Static loadbalancing using an outgoing hash. roundrobin - Distributes outgoing traffic using a round-robin scheduler through all active ports. This code was obtained from OpenBSD and this also includes 802.3ad LACP support from agr(4) in NetBSD.
2007-04-10 00:27:25 +00:00
lacp_run_timers(lp);
lacp_select(lp);
lacp_sm_mux(lp);
lacp_sm_tx(lp);
lacp_sm_ptx_tx_schedule(lp);
Virtualize lagg(4) cloner. This change fixes a panic when tearing down if_lagg(4) interfaces which were cloned in a vnet jail. Sysctl nodes which are dynamically generated for each cloned interface (net.link.lagg.N.*) have been removed, and use_flowid and flowid_shift ifconfig(8) parameters have been added instead. Flags and per-interface statistics counters are displayed in "ifconfig -v". CR: D842
2014-10-01 21:37:32 +00:00
CURVNET_RESTORE();
Add the trunk(4) driver for providing link aggregation, failover and fault tolerance. This driver allows aggregation of multiple network interfaces as one virtual interface using a number of different protocols/algorithms. failover - Sends traffic through the secondary port if the master becomes inactive. fec - Supports Cisco Fast EtherChannel. lacp - Supports the IEEE 802.3ad Link Aggregation Control Protocol (LACP) and the Marker Protocol. loadbalance - Static loadbalancing using an outgoing hash. roundrobin - Distributes outgoing traffic using a round-robin scheduler through all active ports. This code was obtained from OpenBSD and this also includes 802.3ad LACP support from agr(4) in NetBSD.
2007-04-10 00:27:25 +00:00
}
callout_reset(&lsc->lsc_callout, hz, lacp_tick, lsc);
}
int
Rename the trunk(4) driver to lagg(4) as it is too similar to vlan trunking. The name trunk is misused as the networking term trunk means carrying multiple VLANs over a single connection. The IEEE standard for link aggregation (802.3 section 3) does not talk about 'trunk' at all while it is used throughout IEEE 802.1Q in describing vlans. The lagg(4) driver provides link aggregation, failover and fault tolerance. Discussed on: current@
2007-04-17 00:35:11 +00:00
lacp_port_create(struct lagg_port *lgp)
Add the trunk(4) driver for providing link aggregation, failover and fault tolerance. This driver allows aggregation of multiple network interfaces as one virtual interface using a number of different protocols/algorithms. failover - Sends traffic through the secondary port if the master becomes inactive. fec - Supports Cisco Fast EtherChannel. lacp - Supports the IEEE 802.3ad Link Aggregation Control Protocol (LACP) and the Marker Protocol. loadbalance - Static loadbalancing using an outgoing hash. roundrobin - Distributes outgoing traffic using a round-robin scheduler through all active ports. This code was obtained from OpenBSD and this also includes 802.3ad LACP support from agr(4) in NetBSD.
2007-04-10 00:27:25 +00:00
{
non-functional cleanup - remove dead code - use consistent variable names - gc unused defines - whitespace cleanup
2007-06-12 07:29:11 +00:00
struct lagg_softc *sc = lgp->lp_softc;
struct lacp_softc *lsc = LACP_SOFTC(sc);
Add the trunk(4) driver for providing link aggregation, failover and fault tolerance. This driver allows aggregation of multiple network interfaces as one virtual interface using a number of different protocols/algorithms. failover - Sends traffic through the secondary port if the master becomes inactive. fec - Supports Cisco Fast EtherChannel. lacp - Supports the IEEE 802.3ad Link Aggregation Control Protocol (LACP) and the Marker Protocol. loadbalance - Static loadbalancing using an outgoing hash. roundrobin - Distributes outgoing traffic using a round-robin scheduler through all active ports. This code was obtained from OpenBSD and this also includes 802.3ad LACP support from agr(4) in NetBSD.
2007-04-10 00:27:25 +00:00
struct lacp_port *lp;
Rename the trunk(4) driver to lagg(4) as it is too similar to vlan trunking. The name trunk is misused as the networking term trunk means carrying multiple VLANs over a single connection. The IEEE standard for link aggregation (802.3 section 3) does not talk about 'trunk' at all while it is used throughout IEEE 802.1Q in describing vlans. The lagg(4) driver provides link aggregation, failover and fault tolerance. Discussed on: current@
2007-04-17 00:35:11 +00:00
struct ifnet *ifp = lgp->lp_ifp;
Add the trunk(4) driver for providing link aggregation, failover and fault tolerance. This driver allows aggregation of multiple network interfaces as one virtual interface using a number of different protocols/algorithms. failover - Sends traffic through the secondary port if the master becomes inactive. fec - Supports Cisco Fast EtherChannel. lacp - Supports the IEEE 802.3ad Link Aggregation Control Protocol (LACP) and the Marker Protocol. loadbalance - Static loadbalancing using an outgoing hash. roundrobin - Distributes outgoing traffic using a round-robin scheduler through all active ports. This code was obtained from OpenBSD and this also includes 802.3ad LACP support from agr(4) in NetBSD.
2007-04-10 00:27:25 +00:00
struct sockaddr_dl sdl;
struct ifmultiaddr *rifma = NULL;
int error;
Simplify filling sockaddr_dl structure for if_resolvemulti() callback providers. link_init_sdl() function can be used to fill most of the parameters. Use caller stack instead of allocation / freing memory for each request. Do not drop support for extra-long (probably non-existing) link-layer protocols by introducing link_alloc_sdl() (used by if_resolvemulti() callback) and link_free_sdl() (used by caller). Since this change breaks KBI, MFC requires slightly different approach (link_init_sdl() auto-allocating buffer if necessary to handle cases with unmodified if_resolvemulti() callers). MFC after: 2 weeks
2014-01-18 23:24:51 +00:00
link_init_sdl(ifp, (struct sockaddr *)&sdl, IFT_ETHER);
Add the trunk(4) driver for providing link aggregation, failover and fault tolerance. This driver allows aggregation of multiple network interfaces as one virtual interface using a number of different protocols/algorithms. failover - Sends traffic through the secondary port if the master becomes inactive. fec - Supports Cisco Fast EtherChannel. lacp - Supports the IEEE 802.3ad Link Aggregation Control Protocol (LACP) and the Marker Protocol. loadbalance - Static loadbalancing using an outgoing hash. roundrobin - Distributes outgoing traffic using a round-robin scheduler through all active ports. This code was obtained from OpenBSD and this also includes 802.3ad LACP support from agr(4) in NetBSD.
2007-04-10 00:27:25 +00:00
sdl.sdl_alen = ETHER_ADDR_LEN;
bcopy(&ethermulticastaddr_slowprotocols,
LLADDR(&sdl), ETHER_ADDR_LEN);
error = if_addmulti(ifp, (struct sockaddr *)&sdl, &rifma);
if (error) {
Remove macros that hide access to struct ifnet fields.
2014-09-26 13:02:29 +00:00
printf("%s: ADDMULTI failed on %s\n", __func__,
lgp->lp_ifp->if_xname);
Add the trunk(4) driver for providing link aggregation, failover and fault tolerance. This driver allows aggregation of multiple network interfaces as one virtual interface using a number of different protocols/algorithms. failover - Sends traffic through the secondary port if the master becomes inactive. fec - Supports Cisco Fast EtherChannel. lacp - Supports the IEEE 802.3ad Link Aggregation Control Protocol (LACP) and the Marker Protocol. loadbalance - Static loadbalancing using an outgoing hash. roundrobin - Distributes outgoing traffic using a round-robin scheduler through all active ports. This code was obtained from OpenBSD and this also includes 802.3ad LACP support from agr(4) in NetBSD.
2007-04-10 00:27:25 +00:00
return (error);
}
lp = malloc(sizeof(struct lacp_port),
M_DEVBUF, M_NOWAIT|M_ZERO);
if (lp == NULL)
return (ENOMEM);
Switch the LACP state machine over to its own mutex to protect the internals, this means that it no longer grabs the lagg rwlock. Use two port table arrays which list the active ports for Tx and switch between them with an atomic op. Now the lagg rwlock is only exclusively locked for management (ioctls) and queuing of lacp control frames isnt needed.
2008-03-16 19:25:30 +00:00
LACP_LOCK(lsc);
- Whitespace. - Remove caddr_t.
2014-09-26 12:35:58 +00:00
lgp->lp_psc = lp;
Add the trunk(4) driver for providing link aggregation, failover and fault tolerance. This driver allows aggregation of multiple network interfaces as one virtual interface using a number of different protocols/algorithms. failover - Sends traffic through the secondary port if the master becomes inactive. fec - Supports Cisco Fast EtherChannel. lacp - Supports the IEEE 802.3ad Link Aggregation Control Protocol (LACP) and the Marker Protocol. loadbalance - Static loadbalancing using an outgoing hash. roundrobin - Distributes outgoing traffic using a round-robin scheduler through all active ports. This code was obtained from OpenBSD and this also includes 802.3ad LACP support from agr(4) in NetBSD.
2007-04-10 00:27:25 +00:00
lp->lp_ifp = ifp;
Rename the trunk(4) driver to lagg(4) as it is too similar to vlan trunking. The name trunk is misused as the networking term trunk means carrying multiple VLANs over a single connection. The IEEE standard for link aggregation (802.3 section 3) does not talk about 'trunk' at all while it is used throughout IEEE 802.1Q in describing vlans. The lagg(4) driver provides link aggregation, failover and fault tolerance. Discussed on: current@
2007-04-17 00:35:11 +00:00
lp->lp_lagg = lgp;
Add the trunk(4) driver for providing link aggregation, failover and fault tolerance. This driver allows aggregation of multiple network interfaces as one virtual interface using a number of different protocols/algorithms. failover - Sends traffic through the secondary port if the master becomes inactive. fec - Supports Cisco Fast EtherChannel. lacp - Supports the IEEE 802.3ad Link Aggregation Control Protocol (LACP) and the Marker Protocol. loadbalance - Static loadbalancing using an outgoing hash. roundrobin - Distributes outgoing traffic using a round-robin scheduler through all active ports. This code was obtained from OpenBSD and this also includes 802.3ad LACP support from agr(4) in NetBSD.
2007-04-10 00:27:25 +00:00
lp->lp_lsc = lsc;
Change from using if_delmulti() to if_delmulti_ifma() as it simplifies the code and is safe to use if the ifp has disappeared. Suggested by: bms
2007-05-07 00:18:56 +00:00
lp->lp_ifma = rifma;
Add the trunk(4) driver for providing link aggregation, failover and fault tolerance. This driver allows aggregation of multiple network interfaces as one virtual interface using a number of different protocols/algorithms. failover - Sends traffic through the secondary port if the master becomes inactive. fec - Supports Cisco Fast EtherChannel. lacp - Supports the IEEE 802.3ad Link Aggregation Control Protocol (LACP) and the Marker Protocol. loadbalance - Static loadbalancing using an outgoing hash. roundrobin - Distributes outgoing traffic using a round-robin scheduler through all active ports. This code was obtained from OpenBSD and this also includes 802.3ad LACP support from agr(4) in NetBSD.
2007-04-10 00:27:25 +00:00
LIST_INSERT_HEAD(&lsc->lsc_ports, lp, lp_next);
lacp_fill_actorinfo(lp, &lp->lp_actor);
Implement the Marker Protocol. A marker frame is placed on the interface queue of each port and any further packets are blocked, when the all the marker frames have been returned to us from the remote network device then we can be sure that all interface queues are empty. This is needed when a port is added or removed from the aggregation since it will affect the hash based distribution, if the queues are not empty then a packet from an existing connection may be placed on a different interface and arrive out of order. This was previously achieved by suppressing transmission for 1 second, now that there is an active feedback this timeout as been increased to 3 seconds and used as a fallback.
2007-05-19 07:47:04 +00:00
lacp_fill_markerinfo(lp, &lp->lp_marker);
Eliminate misleading comments and dead code in lacp_port_create() Variables "fast" and "active" are both constant in lacp_port_create(), but comments mispleadingly suggest that "fast" can be changed via ioctl. The constant values control the value of "lp->lp_state", so it too is constant, and the code for assigning different value to it is essentially dead. Remove both "fast" and "active", and set "lp->lp_state" unconditionally; that gets rid of the dead code and misleading comments. CID: 1305692 CID: 1305734 Reported by: asomers Reviewed by: asomers MFC after: 1 week Sponsored by: Panasas Differential Revision: https://reviews.freebsd.org/D9302
2017-01-24 01:39:40 +00:00
lp->lp_state = LACP_STATE_ACTIVITY;
Add the trunk(4) driver for providing link aggregation, failover and fault tolerance. This driver allows aggregation of multiple network interfaces as one virtual interface using a number of different protocols/algorithms. failover - Sends traffic through the secondary port if the master becomes inactive. fec - Supports Cisco Fast EtherChannel. lacp - Supports the IEEE 802.3ad Link Aggregation Control Protocol (LACP) and the Marker Protocol. loadbalance - Static loadbalancing using an outgoing hash. roundrobin - Distributes outgoing traffic using a round-robin scheduler through all active ports. This code was obtained from OpenBSD and this also includes 802.3ad LACP support from agr(4) in NetBSD.
2007-04-10 00:27:25 +00:00
lp->lp_aggregator = NULL;
lacp_sm_rx_set_expired(lp);
Switch the LACP state machine over to its own mutex to protect the internals, this means that it no longer grabs the lagg rwlock. Use two port table arrays which list the active ports for Tx and switch between them with an atomic op. Now the lagg rwlock is only exclusively locked for management (ioctls) and queuing of lacp control frames isnt needed.
2008-03-16 19:25:30 +00:00
LACP_UNLOCK(lsc);
lacp_linkstate(lgp);
Add the trunk(4) driver for providing link aggregation, failover and fault tolerance. This driver allows aggregation of multiple network interfaces as one virtual interface using a number of different protocols/algorithms. failover - Sends traffic through the secondary port if the master becomes inactive. fec - Supports Cisco Fast EtherChannel. lacp - Supports the IEEE 802.3ad Link Aggregation Control Protocol (LACP) and the Marker Protocol. loadbalance - Static loadbalancing using an outgoing hash. roundrobin - Distributes outgoing traffic using a round-robin scheduler through all active ports. This code was obtained from OpenBSD and this also includes 802.3ad LACP support from agr(4) in NetBSD.
2007-04-10 00:27:25 +00:00
return (0);
}
void
Rename the trunk(4) driver to lagg(4) as it is too similar to vlan trunking. The name trunk is misused as the networking term trunk means carrying multiple VLANs over a single connection. The IEEE standard for link aggregation (802.3 section 3) does not talk about 'trunk' at all while it is used throughout IEEE 802.1Q in describing vlans. The lagg(4) driver provides link aggregation, failover and fault tolerance. Discussed on: current@
2007-04-17 00:35:11 +00:00
lacp_port_destroy(struct lagg_port *lgp)
Add the trunk(4) driver for providing link aggregation, failover and fault tolerance. This driver allows aggregation of multiple network interfaces as one virtual interface using a number of different protocols/algorithms. failover - Sends traffic through the secondary port if the master becomes inactive. fec - Supports Cisco Fast EtherChannel. lacp - Supports the IEEE 802.3ad Link Aggregation Control Protocol (LACP) and the Marker Protocol. loadbalance - Static loadbalancing using an outgoing hash. roundrobin - Distributes outgoing traffic using a round-robin scheduler through all active ports. This code was obtained from OpenBSD and this also includes 802.3ad LACP support from agr(4) in NetBSD.
2007-04-10 00:27:25 +00:00
{
Rename the trunk(4) driver to lagg(4) as it is too similar to vlan trunking. The name trunk is misused as the networking term trunk means carrying multiple VLANs over a single connection. The IEEE standard for link aggregation (802.3 section 3) does not talk about 'trunk' at all while it is used throughout IEEE 802.1Q in describing vlans. The lagg(4) driver provides link aggregation, failover and fault tolerance. Discussed on: current@
2007-04-17 00:35:11 +00:00
struct lacp_port *lp = LACP_PORT(lgp);
Switch the LACP state machine over to its own mutex to protect the internals, this means that it no longer grabs the lagg rwlock. Use two port table arrays which list the active ports for Tx and switch between them with an atomic op. Now the lagg rwlock is only exclusively locked for management (ioctls) and queuing of lacp control frames isnt needed.
2008-03-16 19:25:30 +00:00
struct lacp_softc *lsc = lp->lp_lsc;
Change from using if_delmulti() to if_delmulti_ifma() as it simplifies the code and is safe to use if the ifp has disappeared. Suggested by: bms
2007-05-07 00:18:56 +00:00
int i;
Add the trunk(4) driver for providing link aggregation, failover and fault tolerance. This driver allows aggregation of multiple network interfaces as one virtual interface using a number of different protocols/algorithms. failover - Sends traffic through the secondary port if the master becomes inactive. fec - Supports Cisco Fast EtherChannel. lacp - Supports the IEEE 802.3ad Link Aggregation Control Protocol (LACP) and the Marker Protocol. loadbalance - Static loadbalancing using an outgoing hash. roundrobin - Distributes outgoing traffic using a round-robin scheduler through all active ports. This code was obtained from OpenBSD and this also includes 802.3ad LACP support from agr(4) in NetBSD.
2007-04-10 00:27:25 +00:00
Switch the LACP state machine over to its own mutex to protect the internals, this means that it no longer grabs the lagg rwlock. Use two port table arrays which list the active ports for Tx and switch between them with an atomic op. Now the lagg rwlock is only exclusively locked for management (ioctls) and queuing of lacp control frames isnt needed.
2008-03-16 19:25:30 +00:00
LACP_LOCK(lsc);
Add the trunk(4) driver for providing link aggregation, failover and fault tolerance. This driver allows aggregation of multiple network interfaces as one virtual interface using a number of different protocols/algorithms. failover - Sends traffic through the secondary port if the master becomes inactive. fec - Supports Cisco Fast EtherChannel. lacp - Supports the IEEE 802.3ad Link Aggregation Control Protocol (LACP) and the Marker Protocol. loadbalance - Static loadbalancing using an outgoing hash. roundrobin - Distributes outgoing traffic using a round-robin scheduler through all active ports. This code was obtained from OpenBSD and this also includes 802.3ad LACP support from agr(4) in NetBSD.
2007-04-10 00:27:25 +00:00
for (i = 0; i < LACP_NTIMER; i++) {
LACP_TIMER_DISARM(lp, i);
}
lacp_disable_collecting(lp);
lacp_disable_distributing(lp);
lacp_unselect(lp);
- Move L2 addr configuration for the primary port to a taskqueue. This fixes LOR of softc rmlock in iflladdr_event handlers. - Call if_delmulti_ifma() after LACP_UNLOCK(). This fixes another LOR. - Fix a panic in lacp_transit_expire(). - Fix a panic in lagg_input() upon shutting down a port.
2014-10-05 02:34:21 +00:00
LIST_REMOVE(lp, lp_next);
LACP_UNLOCK(lsc);
Avoid touching various unsafe parts if the interface is disappearing.
2007-05-07 00:28:55 +00:00
/* The address may have already been removed by if_purgemaddrs() */
if (!lgp->lp_detaching)
if_delmulti_ifma(lp->lp_ifma);
Add the trunk(4) driver for providing link aggregation, failover and fault tolerance. This driver allows aggregation of multiple network interfaces as one virtual interface using a number of different protocols/algorithms. failover - Sends traffic through the secondary port if the master becomes inactive. fec - Supports Cisco Fast EtherChannel. lacp - Supports the IEEE 802.3ad Link Aggregation Control Protocol (LACP) and the Marker Protocol. loadbalance - Static loadbalancing using an outgoing hash. roundrobin - Distributes outgoing traffic using a round-robin scheduler through all active ports. This code was obtained from OpenBSD and this also includes 802.3ad LACP support from agr(4) in NetBSD.
2007-04-10 00:27:25 +00:00
free(lp, M_DEVBUF);
}
Allow the LACP state to be queried from userland which at the moment is the actor and partner peer info. Print out the active aggregator and per port data in verbose mode from ifconfig. Approved by: re (mux)
2007-07-05 09:18:57 +00:00
void
- Whitespace. - Remove caddr_t.
2014-09-26 12:35:58 +00:00
lacp_req(struct lagg_softc *sc, void *data)
Allow the LACP state to be queried from userland which at the moment is the actor and partner peer info. Print out the active aggregator and per port data in verbose mode from ifconfig. Approved by: re (mux)
2007-07-05 09:18:57 +00:00
{
struct lacp_opreq *req = (struct lacp_opreq *)data;
struct lacp_softc *lsc = LACP_SOFTC(sc);
Fix a panic caused by doing "ifconfig -am" while a lagg is being destroyed. The thread that is destroying the lagg has already set sc->sc_psc=NULL when the "ifconfig -am" thread gets to lacp_req(). It tries to dereference sc->sc_psc and panics. The solution is for lacp_req() to check the value of sc->sc_psc. If NULL, harmlessly return an lacp_opreq structure full of zeros. Full details in GNATS. PR: kern/189003 Reviewed by: timeout on freebsd-net@ MFC after: 3 weeks Sponsored by: Spectra Logic Corporation
2014-05-02 16:24:09 +00:00
struct lacp_aggregator *la;
Allow the LACP state to be queried from userland which at the moment is the actor and partner peer info. Print out the active aggregator and per port data in verbose mode from ifconfig. Approved by: re (mux)
2007-07-05 09:18:57 +00:00
bzero(req, sizeof(struct lacp_opreq));
Fix a panic caused by doing "ifconfig -am" while a lagg is being destroyed. The thread that is destroying the lagg has already set sc->sc_psc=NULL when the "ifconfig -am" thread gets to lacp_req(). It tries to dereference sc->sc_psc and panics. The solution is for lacp_req() to check the value of sc->sc_psc. If NULL, harmlessly return an lacp_opreq structure full of zeros. Full details in GNATS. PR: kern/189003 Reviewed by: timeout on freebsd-net@ MFC after: 3 weeks Sponsored by: Spectra Logic Corporation
2014-05-02 16:24:09 +00:00
- Whitespace. - Remove caddr_t.
2014-09-26 12:35:58 +00:00
/*
Fix a panic caused by doing "ifconfig -am" while a lagg is being destroyed. The thread that is destroying the lagg has already set sc->sc_psc=NULL when the "ifconfig -am" thread gets to lacp_req(). It tries to dereference sc->sc_psc and panics. The solution is for lacp_req() to check the value of sc->sc_psc. If NULL, harmlessly return an lacp_opreq structure full of zeros. Full details in GNATS. PR: kern/189003 Reviewed by: timeout on freebsd-net@ MFC after: 3 weeks Sponsored by: Spectra Logic Corporation
2014-05-02 16:24:09 +00:00
* If the LACP softc is NULL, return with the opreq structure full of
* zeros. It is normal for the softc to be NULL while the lagg is
* being destroyed.
*/
if (NULL == lsc)
return;
la = lsc->lsc_active_aggregator;
LACP_LOCK(lsc);
Allow the LACP state to be queried from userland which at the moment is the actor and partner peer info. Print out the active aggregator and per port data in verbose mode from ifconfig. Approved by: re (mux)
2007-07-05 09:18:57 +00:00
if (la != NULL) {
req->actor_prio = ntohs(la->la_actor.lip_systemid.lsi_prio);
memcpy(&req->actor_mac, &la->la_actor.lip_systemid.lsi_mac,
ETHER_ADDR_LEN);
req->actor_key = ntohs(la->la_actor.lip_key);
req->actor_portprio = ntohs(la->la_actor.lip_portid.lpi_prio);
req->actor_portno = ntohs(la->la_actor.lip_portid.lpi_portno);
req->actor_state = la->la_actor.lip_state;
req->partner_prio = ntohs(la->la_partner.lip_systemid.lsi_prio);
memcpy(&req->partner_mac, &la->la_partner.lip_systemid.lsi_mac,
ETHER_ADDR_LEN);
req->partner_key = ntohs(la->la_partner.lip_key);
req->partner_portprio = ntohs(la->la_partner.lip_portid.lpi_prio);
req->partner_portno = ntohs(la->la_partner.lip_portid.lpi_portno);
req->partner_state = la->la_partner.lip_state;
}
Switch the LACP state machine over to its own mutex to protect the internals, this means that it no longer grabs the lagg rwlock. Use two port table arrays which list the active ports for Tx and switch between them with an atomic op. Now the lagg rwlock is only exclusively locked for management (ioctls) and queuing of lacp control frames isnt needed.
2008-03-16 19:25:30 +00:00
LACP_UNLOCK(lsc);
Allow the LACP state to be queried from userland which at the moment is the actor and partner peer info. Print out the active aggregator and per port data in verbose mode from ifconfig. Approved by: re (mux)
2007-07-05 09:18:57 +00:00
}
void
- Whitespace. - Remove caddr_t.
2014-09-26 12:35:58 +00:00
lacp_portreq(struct lagg_port *lgp, void *data)
Allow the LACP state to be queried from userland which at the moment is the actor and partner peer info. Print out the active aggregator and per port data in verbose mode from ifconfig. Approved by: re (mux)
2007-07-05 09:18:57 +00:00
{
struct lacp_opreq *req = (struct lacp_opreq *)data;
struct lacp_port *lp = LACP_PORT(lgp);
Switch the LACP state machine over to its own mutex to protect the internals, this means that it no longer grabs the lagg rwlock. Use two port table arrays which list the active ports for Tx and switch between them with an atomic op. Now the lagg rwlock is only exclusively locked for management (ioctls) and queuing of lacp control frames isnt needed.
2008-03-16 19:25:30 +00:00
struct lacp_softc *lsc = lp->lp_lsc;
Allow the LACP state to be queried from userland which at the moment is the actor and partner peer info. Print out the active aggregator and per port data in verbose mode from ifconfig. Approved by: re (mux)
2007-07-05 09:18:57 +00:00
Switch the LACP state machine over to its own mutex to protect the internals, this means that it no longer grabs the lagg rwlock. Use two port table arrays which list the active ports for Tx and switch between them with an atomic op. Now the lagg rwlock is only exclusively locked for management (ioctls) and queuing of lacp control frames isnt needed.
2008-03-16 19:25:30 +00:00
LACP_LOCK(lsc);
Allow the LACP state to be queried from userland which at the moment is the actor and partner peer info. Print out the active aggregator and per port data in verbose mode from ifconfig. Approved by: re (mux)
2007-07-05 09:18:57 +00:00
req->actor_prio = ntohs(lp->lp_actor.lip_systemid.lsi_prio);
memcpy(&req->actor_mac, &lp->lp_actor.lip_systemid.lsi_mac,
ETHER_ADDR_LEN);
req->actor_key = ntohs(lp->lp_actor.lip_key);
req->actor_portprio = ntohs(lp->lp_actor.lip_portid.lpi_prio);
req->actor_portno = ntohs(lp->lp_actor.lip_portid.lpi_portno);
req->actor_state = lp->lp_actor.lip_state;
req->partner_prio = ntohs(lp->lp_partner.lip_systemid.lsi_prio);
memcpy(&req->partner_mac, &lp->lp_partner.lip_systemid.lsi_mac,
ETHER_ADDR_LEN);
req->partner_key = ntohs(lp->lp_partner.lip_key);
req->partner_portprio = ntohs(lp->lp_partner.lip_portid.lpi_prio);
req->partner_portno = ntohs(lp->lp_partner.lip_portid.lpi_portno);
req->partner_state = lp->lp_partner.lip_state;
Switch the LACP state machine over to its own mutex to protect the internals, this means that it no longer grabs the lagg rwlock. Use two port table arrays which list the active ports for Tx and switch between them with an atomic op. Now the lagg rwlock is only exclusively locked for management (ioctls) and queuing of lacp control frames isnt needed.
2008-03-16 19:25:30 +00:00
LACP_UNLOCK(lsc);
Allow the LACP state to be queried from userland which at the moment is the actor and partner peer info. Print out the active aggregator and per port data in verbose mode from ifconfig. Approved by: re (mux)
2007-07-05 09:18:57 +00:00
}
Add the trunk(4) driver for providing link aggregation, failover and fault tolerance. This driver allows aggregation of multiple network interfaces as one virtual interface using a number of different protocols/algorithms. failover - Sends traffic through the secondary port if the master becomes inactive. fec - Supports Cisco Fast EtherChannel. lacp - Supports the IEEE 802.3ad Link Aggregation Control Protocol (LACP) and the Marker Protocol. loadbalance - Static loadbalancing using an outgoing hash. roundrobin - Distributes outgoing traffic using a round-robin scheduler through all active ports. This code was obtained from OpenBSD and this also includes 802.3ad LACP support from agr(4) in NetBSD.
2007-04-10 00:27:25 +00:00
static void
lacp_disable_collecting(struct lacp_port *lp)
{
LACP_DPRINTF((lp, "collecting disabled\n"));
lp->lp_state &= ~LACP_STATE_COLLECTING;
}
static void
lacp_enable_collecting(struct lacp_port *lp)
{
LACP_DPRINTF((lp, "collecting enabled\n"));
lp->lp_state |= LACP_STATE_COLLECTING;
}
static void
lacp_disable_distributing(struct lacp_port *lp)
{
struct lacp_aggregator *la = lp->lp_aggregator;
struct lacp_softc *lsc = lp->lp_lsc;
Bring over some link aggregation / LACP protocol improvements and debugging additions. * Add some new tracing events to aid in debugging. * Add in a debugging mode to drop transmit and received frames, specifically to test whether seeing or hearing heartbeats correctly cause LACP to drop the port. * Add in (and make default) a strict LACP mode, which requires the heartbeat on a port to be heard before it's used. Sometimes vendor ports will hang but the link layer stays up, resulting in hung traffic. * Add logging the number of link status flaps, again to aid in debugging badly behaving switch ports. * Calculate the lagg interface port speed as the multiple of the configured ports, rather than the largest. Obtained from: Netflix MFC after: 2 weeks
2013-07-13 04:25:03 +00:00
struct lagg_softc *sc = lsc->lsc_softc;
Add the trunk(4) driver for providing link aggregation, failover and fault tolerance. This driver allows aggregation of multiple network interfaces as one virtual interface using a number of different protocols/algorithms. failover - Sends traffic through the secondary port if the master becomes inactive. fec - Supports Cisco Fast EtherChannel. lacp - Supports the IEEE 802.3ad Link Aggregation Control Protocol (LACP) and the Marker Protocol. loadbalance - Static loadbalancing using an outgoing hash. roundrobin - Distributes outgoing traffic using a round-robin scheduler through all active ports. This code was obtained from OpenBSD and this also includes 802.3ad LACP support from agr(4) in NetBSD.
2007-04-10 00:27:25 +00:00
char buf[LACP_LAGIDSTR_MAX+1];
Switch the LACP state machine over to its own mutex to protect the internals, this means that it no longer grabs the lagg rwlock. Use two port table arrays which list the active ports for Tx and switch between them with an atomic op. Now the lagg rwlock is only exclusively locked for management (ioctls) and queuing of lacp control frames isnt needed.
2008-03-16 19:25:30 +00:00
LACP_LOCK_ASSERT(lsc);
Add the trunk(4) driver for providing link aggregation, failover and fault tolerance. This driver allows aggregation of multiple network interfaces as one virtual interface using a number of different protocols/algorithms. failover - Sends traffic through the secondary port if the master becomes inactive. fec - Supports Cisco Fast EtherChannel. lacp - Supports the IEEE 802.3ad Link Aggregation Control Protocol (LACP) and the Marker Protocol. loadbalance - Static loadbalancing using an outgoing hash. roundrobin - Distributes outgoing traffic using a round-robin scheduler through all active ports. This code was obtained from OpenBSD and this also includes 802.3ad LACP support from agr(4) in NetBSD.
2007-04-10 00:27:25 +00:00
if (la == NULL || (lp->lp_state & LACP_STATE_DISTRIBUTING) == 0) {
return;
}
KASSERT(!TAILQ_EMPTY(&la->la_ports), ("no aggregator ports"));
KASSERT(la->la_nports > 0, ("nports invalid (%d)", la->la_nports));
KASSERT(la->la_refcnt >= la->la_nports, ("aggregator refcnt invalid"));
LACP_DPRINTF((lp, "disable distributing on aggregator %s, "
"nports %d -> %d\n",
lacp_format_lagid_aggregator(la, buf, sizeof(buf)),
la->la_nports, la->la_nports - 1));
TAILQ_REMOVE(&la->la_ports, lp, lp_dist_q);
la->la_nports--;
Bring over some link aggregation / LACP protocol improvements and debugging additions. * Add some new tracing events to aid in debugging. * Add in a debugging mode to drop transmit and received frames, specifically to test whether seeing or hearing heartbeats correctly cause LACP to drop the port. * Add in (and make default) a strict LACP mode, which requires the heartbeat on a port to be heard before it's used. Sometimes vendor ports will hang but the link layer stays up, resulting in hung traffic. * Add logging the number of link status flaps, again to aid in debugging badly behaving switch ports. * Calculate the lagg interface port speed as the multiple of the configured ports, rather than the largest. Obtained from: Netflix MFC after: 2 weeks
2013-07-13 04:25:03 +00:00
sc->sc_active = la->la_nports;
Add the trunk(4) driver for providing link aggregation, failover and fault tolerance. This driver allows aggregation of multiple network interfaces as one virtual interface using a number of different protocols/algorithms. failover - Sends traffic through the secondary port if the master becomes inactive. fec - Supports Cisco Fast EtherChannel. lacp - Supports the IEEE 802.3ad Link Aggregation Control Protocol (LACP) and the Marker Protocol. loadbalance - Static loadbalancing using an outgoing hash. roundrobin - Distributes outgoing traffic using a round-robin scheduler through all active ports. This code was obtained from OpenBSD and this also includes 802.3ad LACP support from agr(4) in NetBSD.
2007-04-10 00:27:25 +00:00
if (lsc->lsc_active_aggregator == la) {
Switch the LACP state machine over to its own mutex to protect the internals, this means that it no longer grabs the lagg rwlock. Use two port table arrays which list the active ports for Tx and switch between them with an atomic op. Now the lagg rwlock is only exclusively locked for management (ioctls) and queuing of lacp control frames isnt needed.
2008-03-16 19:25:30 +00:00
lacp_suppress_distributing(lsc, la);
Add the trunk(4) driver for providing link aggregation, failover and fault tolerance. This driver allows aggregation of multiple network interfaces as one virtual interface using a number of different protocols/algorithms. failover - Sends traffic through the secondary port if the master becomes inactive. fec - Supports Cisco Fast EtherChannel. lacp - Supports the IEEE 802.3ad Link Aggregation Control Protocol (LACP) and the Marker Protocol. loadbalance - Static loadbalancing using an outgoing hash. roundrobin - Distributes outgoing traffic using a round-robin scheduler through all active ports. This code was obtained from OpenBSD and this also includes 802.3ad LACP support from agr(4) in NetBSD.
2007-04-10 00:27:25 +00:00
lacp_select_active_aggregator(lsc);
Switch the LACP state machine over to its own mutex to protect the internals, this means that it no longer grabs the lagg rwlock. Use two port table arrays which list the active ports for Tx and switch between them with an atomic op. Now the lagg rwlock is only exclusively locked for management (ioctls) and queuing of lacp control frames isnt needed.
2008-03-16 19:25:30 +00:00
/* regenerate the port map, the active aggregator has changed */
lacp_update_portmap(lsc);
Add the trunk(4) driver for providing link aggregation, failover and fault tolerance. This driver allows aggregation of multiple network interfaces as one virtual interface using a number of different protocols/algorithms. failover - Sends traffic through the secondary port if the master becomes inactive. fec - Supports Cisco Fast EtherChannel. lacp - Supports the IEEE 802.3ad Link Aggregation Control Protocol (LACP) and the Marker Protocol. loadbalance - Static loadbalancing using an outgoing hash. roundrobin - Distributes outgoing traffic using a round-robin scheduler through all active ports. This code was obtained from OpenBSD and this also includes 802.3ad LACP support from agr(4) in NetBSD.
2007-04-10 00:27:25 +00:00
}
Switch the LACP state machine over to its own mutex to protect the internals, this means that it no longer grabs the lagg rwlock. Use two port table arrays which list the active ports for Tx and switch between them with an atomic op. Now the lagg rwlock is only exclusively locked for management (ioctls) and queuing of lacp control frames isnt needed.
2008-03-16 19:25:30 +00:00
lp->lp_state &= ~LACP_STATE_DISTRIBUTING;
lagg: allow lacp to manage the link state Lacp needs to manage the link state itself. Unlike other lagg protocols, the ability of lacp to pass traffic depends not only on the lagg members having link, but also on the lacp protocol converging to a distributing state with the link partner. If we prematurely mark the link as up, then we will send a gratuitous arp (via arp_handle_ifllchange()) before the lacp interface is capable of passing traffic. When this happens, the gratuitous arp is lost, and our link partner may cache a stale mac address (eg, when the base mac address for the lagg bundle changes, due to a BIOS change re-ordering NIC unit numbers) Reviewed by: jtl, hselasky Sponsored by: Netflix
2018-08-13 14:13:25 +00:00
if_link_state_change(sc->sc_ifp,
sc->sc_active ? LINK_STATE_UP : LINK_STATE_DOWN);
Add the trunk(4) driver for providing link aggregation, failover and fault tolerance. This driver allows aggregation of multiple network interfaces as one virtual interface using a number of different protocols/algorithms. failover - Sends traffic through the secondary port if the master becomes inactive. fec - Supports Cisco Fast EtherChannel. lacp - Supports the IEEE 802.3ad Link Aggregation Control Protocol (LACP) and the Marker Protocol. loadbalance - Static loadbalancing using an outgoing hash. roundrobin - Distributes outgoing traffic using a round-robin scheduler through all active ports. This code was obtained from OpenBSD and this also includes 802.3ad LACP support from agr(4) in NetBSD.
2007-04-10 00:27:25 +00:00
}
static void
lacp_enable_distributing(struct lacp_port *lp)
{
struct lacp_aggregator *la = lp->lp_aggregator;
struct lacp_softc *lsc = lp->lp_lsc;
Bring over some link aggregation / LACP protocol improvements and debugging additions. * Add some new tracing events to aid in debugging. * Add in a debugging mode to drop transmit and received frames, specifically to test whether seeing or hearing heartbeats correctly cause LACP to drop the port. * Add in (and make default) a strict LACP mode, which requires the heartbeat on a port to be heard before it's used. Sometimes vendor ports will hang but the link layer stays up, resulting in hung traffic. * Add logging the number of link status flaps, again to aid in debugging badly behaving switch ports. * Calculate the lagg interface port speed as the multiple of the configured ports, rather than the largest. Obtained from: Netflix MFC after: 2 weeks
2013-07-13 04:25:03 +00:00
struct lagg_softc *sc = lsc->lsc_softc;
Add the trunk(4) driver for providing link aggregation, failover and fault tolerance. This driver allows aggregation of multiple network interfaces as one virtual interface using a number of different protocols/algorithms. failover - Sends traffic through the secondary port if the master becomes inactive. fec - Supports Cisco Fast EtherChannel. lacp - Supports the IEEE 802.3ad Link Aggregation Control Protocol (LACP) and the Marker Protocol. loadbalance - Static loadbalancing using an outgoing hash. roundrobin - Distributes outgoing traffic using a round-robin scheduler through all active ports. This code was obtained from OpenBSD and this also includes 802.3ad LACP support from agr(4) in NetBSD.
2007-04-10 00:27:25 +00:00
char buf[LACP_LAGIDSTR_MAX+1];
Switch the LACP state machine over to its own mutex to protect the internals, this means that it no longer grabs the lagg rwlock. Use two port table arrays which list the active ports for Tx and switch between them with an atomic op. Now the lagg rwlock is only exclusively locked for management (ioctls) and queuing of lacp control frames isnt needed.
2008-03-16 19:25:30 +00:00
LACP_LOCK_ASSERT(lsc);
Add the trunk(4) driver for providing link aggregation, failover and fault tolerance. This driver allows aggregation of multiple network interfaces as one virtual interface using a number of different protocols/algorithms. failover - Sends traffic through the secondary port if the master becomes inactive. fec - Supports Cisco Fast EtherChannel. lacp - Supports the IEEE 802.3ad Link Aggregation Control Protocol (LACP) and the Marker Protocol. loadbalance - Static loadbalancing using an outgoing hash. roundrobin - Distributes outgoing traffic using a round-robin scheduler through all active ports. This code was obtained from OpenBSD and this also includes 802.3ad LACP support from agr(4) in NetBSD.
2007-04-10 00:27:25 +00:00
if ((lp->lp_state & LACP_STATE_DISTRIBUTING) != 0) {
return;
}
LACP_DPRINTF((lp, "enable distributing on aggregator %s, "
"nports %d -> %d\n",
lacp_format_lagid_aggregator(la, buf, sizeof(buf)),
la->la_nports, la->la_nports + 1));
KASSERT(la->la_refcnt > la->la_nports, ("aggregator refcnt invalid"));
TAILQ_INSERT_HEAD(&la->la_ports, lp, lp_dist_q);
la->la_nports++;
Bring over some link aggregation / LACP protocol improvements and debugging additions. * Add some new tracing events to aid in debugging. * Add in a debugging mode to drop transmit and received frames, specifically to test whether seeing or hearing heartbeats correctly cause LACP to drop the port. * Add in (and make default) a strict LACP mode, which requires the heartbeat on a port to be heard before it's used. Sometimes vendor ports will hang but the link layer stays up, resulting in hung traffic. * Add logging the number of link status flaps, again to aid in debugging badly behaving switch ports. * Calculate the lagg interface port speed as the multiple of the configured ports, rather than the largest. Obtained from: Netflix MFC after: 2 weeks
2013-07-13 04:25:03 +00:00
sc->sc_active = la->la_nports;
Add the trunk(4) driver for providing link aggregation, failover and fault tolerance. This driver allows aggregation of multiple network interfaces as one virtual interface using a number of different protocols/algorithms. failover - Sends traffic through the secondary port if the master becomes inactive. fec - Supports Cisco Fast EtherChannel. lacp - Supports the IEEE 802.3ad Link Aggregation Control Protocol (LACP) and the Marker Protocol. loadbalance - Static loadbalancing using an outgoing hash. roundrobin - Distributes outgoing traffic using a round-robin scheduler through all active ports. This code was obtained from OpenBSD and this also includes 802.3ad LACP support from agr(4) in NetBSD.
2007-04-10 00:27:25 +00:00
lp->lp_state |= LACP_STATE_DISTRIBUTING;
Switch the LACP state machine over to its own mutex to protect the internals, this means that it no longer grabs the lagg rwlock. Use two port table arrays which list the active ports for Tx and switch between them with an atomic op. Now the lagg rwlock is only exclusively locked for management (ioctls) and queuing of lacp control frames isnt needed.
2008-03-16 19:25:30 +00:00
if (lsc->lsc_active_aggregator == la) {
lacp_suppress_distributing(lsc, la);
lacp_update_portmap(lsc);
} else
/* try to become the active aggregator */
Add the trunk(4) driver for providing link aggregation, failover and fault tolerance. This driver allows aggregation of multiple network interfaces as one virtual interface using a number of different protocols/algorithms. failover - Sends traffic through the secondary port if the master becomes inactive. fec - Supports Cisco Fast EtherChannel. lacp - Supports the IEEE 802.3ad Link Aggregation Control Protocol (LACP) and the Marker Protocol. loadbalance - Static loadbalancing using an outgoing hash. roundrobin - Distributes outgoing traffic using a round-robin scheduler through all active ports. This code was obtained from OpenBSD and this also includes 802.3ad LACP support from agr(4) in NetBSD.
2007-04-10 00:27:25 +00:00
lacp_select_active_aggregator(lsc);
lagg: allow lacp to manage the link state Lacp needs to manage the link state itself. Unlike other lagg protocols, the ability of lacp to pass traffic depends not only on the lagg members having link, but also on the lacp protocol converging to a distributing state with the link partner. If we prematurely mark the link as up, then we will send a gratuitous arp (via arp_handle_ifllchange()) before the lacp interface is capable of passing traffic. When this happens, the gratuitous arp is lost, and our link partner may cache a stale mac address (eg, when the base mac address for the lagg bundle changes, due to a BIOS change re-ordering NIC unit numbers) Reviewed by: jtl, hselasky Sponsored by: Netflix
2018-08-13 14:13:25 +00:00
if_link_state_change(sc->sc_ifp,
sc->sc_active ? LINK_STATE_UP : LINK_STATE_DOWN);
Add the trunk(4) driver for providing link aggregation, failover and fault tolerance. This driver allows aggregation of multiple network interfaces as one virtual interface using a number of different protocols/algorithms. failover - Sends traffic through the secondary port if the master becomes inactive. fec - Supports Cisco Fast EtherChannel. lacp - Supports the IEEE 802.3ad Link Aggregation Control Protocol (LACP) and the Marker Protocol. loadbalance - Static loadbalancing using an outgoing hash. roundrobin - Distributes outgoing traffic using a round-robin scheduler through all active ports. This code was obtained from OpenBSD and this also includes 802.3ad LACP support from agr(4) in NetBSD.
2007-04-10 00:27:25 +00:00
}
static void
lacp_transit_expire(void *vp)
{
struct lacp_softc *lsc = vp;
Switch the LACP state machine over to its own mutex to protect the internals, this means that it no longer grabs the lagg rwlock. Use two port table arrays which list the active ports for Tx and switch between them with an atomic op. Now the lagg rwlock is only exclusively locked for management (ioctls) and queuing of lacp control frames isnt needed.
2008-03-16 19:25:30 +00:00
LACP_LOCK_ASSERT(lsc);
- Move L2 addr configuration for the primary port to a taskqueue. This fixes LOR of softc rmlock in iflladdr_event handlers. - Call if_delmulti_ifma() after LACP_UNLOCK(). This fixes another LOR. - Fix a panic in lacp_transit_expire(). - Fix a panic in lagg_input() upon shutting down a port.
2014-10-05 02:34:21 +00:00
CURVNET_SET(lsc->lsc_softc->sc_ifp->if_vnet);
Turn LACP debugging from a compile time option to a sysctl, it is very handy to be able to turn it on when negotiation to a switch misbehaves. Submitted by: Andrew Boyer MFC after: 3 days
2012-05-26 08:09:01 +00:00
LACP_TRACE(NULL);
- Move L2 addr configuration for the primary port to a taskqueue. This fixes LOR of softc rmlock in iflladdr_event handlers. - Call if_delmulti_ifma() after LACP_UNLOCK(). This fixes another LOR. - Fix a panic in lacp_transit_expire(). - Fix a panic in lagg_input() upon shutting down a port.
2014-10-05 02:34:21 +00:00
CURVNET_RESTORE();
Turn LACP debugging from a compile time option to a sysctl, it is very handy to be able to turn it on when negotiation to a switch misbehaves. Submitted by: Andrew Boyer MFC after: 3 days
2012-05-26 08:09:01 +00:00
Add the trunk(4) driver for providing link aggregation, failover and fault tolerance. This driver allows aggregation of multiple network interfaces as one virtual interface using a number of different protocols/algorithms. failover - Sends traffic through the secondary port if the master becomes inactive. fec - Supports Cisco Fast EtherChannel. lacp - Supports the IEEE 802.3ad Link Aggregation Control Protocol (LACP) and the Marker Protocol. loadbalance - Static loadbalancing using an outgoing hash. roundrobin - Distributes outgoing traffic using a round-robin scheduler through all active ports. This code was obtained from OpenBSD and this also includes 802.3ad LACP support from agr(4) in NetBSD.
2007-04-10 00:27:25 +00:00
lsc->lsc_suppress_distributing = FALSE;
}
- When reconfiguring protocol on a lagg, first set it to LAGG_PROTO_NONE, then drop lock, run the attach routines, and then set it to specific proto. This removes tons of WITNESS warnings. - Make lagg protocol attach handlers not failing and allocate memory with M_WAITOK. Sponsored by: Netflix Sponsored by: Nginx, Inc.
2014-09-26 08:42:32 +00:00
void
non-functional cleanup - remove dead code - use consistent variable names - gc unused defines - whitespace cleanup
2007-06-12 07:29:11 +00:00
lacp_attach(struct lagg_softc *sc)
Add the trunk(4) driver for providing link aggregation, failover and fault tolerance. This driver allows aggregation of multiple network interfaces as one virtual interface using a number of different protocols/algorithms. failover - Sends traffic through the secondary port if the master becomes inactive. fec - Supports Cisco Fast EtherChannel. lacp - Supports the IEEE 802.3ad Link Aggregation Control Protocol (LACP) and the Marker Protocol. loadbalance - Static loadbalancing using an outgoing hash. roundrobin - Distributes outgoing traffic using a round-robin scheduler through all active ports. This code was obtained from OpenBSD and this also includes 802.3ad LACP support from agr(4) in NetBSD.
2007-04-10 00:27:25 +00:00
{
struct lacp_softc *lsc;
- When reconfiguring protocol on a lagg, first set it to LAGG_PROTO_NONE, then drop lock, run the attach routines, and then set it to specific proto. This removes tons of WITNESS warnings. - Make lagg protocol attach handlers not failing and allocate memory with M_WAITOK. Sponsored by: Netflix Sponsored by: Nginx, Inc.
2014-09-26 08:42:32 +00:00
lsc = malloc(sizeof(struct lacp_softc), M_DEVBUF, M_WAITOK | M_ZERO);
Add the trunk(4) driver for providing link aggregation, failover and fault tolerance. This driver allows aggregation of multiple network interfaces as one virtual interface using a number of different protocols/algorithms. failover - Sends traffic through the secondary port if the master becomes inactive. fec - Supports Cisco Fast EtherChannel. lacp - Supports the IEEE 802.3ad Link Aggregation Control Protocol (LACP) and the Marker Protocol. loadbalance - Static loadbalancing using an outgoing hash. roundrobin - Distributes outgoing traffic using a round-robin scheduler through all active ports. This code was obtained from OpenBSD and this also includes 802.3ad LACP support from agr(4) in NetBSD.
2007-04-10 00:27:25 +00:00
- Whitespace. - Remove caddr_t.
2014-09-26 12:35:58 +00:00
sc->sc_psc = lsc;
non-functional cleanup - remove dead code - use consistent variable names - gc unused defines - whitespace cleanup
2007-06-12 07:29:11 +00:00
lsc->lsc_softc = sc;
Add the trunk(4) driver for providing link aggregation, failover and fault tolerance. This driver allows aggregation of multiple network interfaces as one virtual interface using a number of different protocols/algorithms. failover - Sends traffic through the secondary port if the master becomes inactive. fec - Supports Cisco Fast EtherChannel. lacp - Supports the IEEE 802.3ad Link Aggregation Control Protocol (LACP) and the Marker Protocol. loadbalance - Static loadbalancing using an outgoing hash. roundrobin - Distributes outgoing traffic using a round-robin scheduler through all active ports. This code was obtained from OpenBSD and this also includes 802.3ad LACP support from agr(4) in NetBSD.
2007-04-10 00:27:25 +00:00
Factor out mbuf hashing code from LAGG driver so that other network drivers can use it. This avoids some code duplication. Add missing default case to all switch statements while at it. Also move the hashing of the IPv6 flow field to layer 4 because the IPv6 flow field is constant on a per L4 connection basis and not on a per L3 network. Differential Revision: https://reviews.freebsd.org/D1987 Sponsored by: Mellanox Technologies MFC after: 1 month
2015-03-11 16:02:24 +00:00
lsc->lsc_hashkey = m_ether_tcpip_hash_init();
Add the trunk(4) driver for providing link aggregation, failover and fault tolerance. This driver allows aggregation of multiple network interfaces as one virtual interface using a number of different protocols/algorithms. failover - Sends traffic through the secondary port if the master becomes inactive. fec - Supports Cisco Fast EtherChannel. lacp - Supports the IEEE 802.3ad Link Aggregation Control Protocol (LACP) and the Marker Protocol. loadbalance - Static loadbalancing using an outgoing hash. roundrobin - Distributes outgoing traffic using a round-robin scheduler through all active ports. This code was obtained from OpenBSD and this also includes 802.3ad LACP support from agr(4) in NetBSD.
2007-04-10 00:27:25 +00:00
lsc->lsc_active_aggregator = NULL;
Add sysctl to control LACP strict compliance default Add net.link.lagg.lacp.default_strict_mode which defines the default value for LACP strict compliance for created lagg devices. Also: * Add lacp_strict option to ifconfig(8). * Fix lagg(4) creation examples. * Minor style(9) fix. MFC after: 1 week
2015-11-06 15:33:27 +00:00
lsc->lsc_strict_mode = VNET(lacp_default_strict_mode);
Switch the LACP state machine over to its own mutex to protect the internals, this means that it no longer grabs the lagg rwlock. Use two port table arrays which list the active ports for Tx and switch between them with an atomic op. Now the lagg rwlock is only exclusively locked for management (ioctls) and queuing of lacp control frames isnt needed.
2008-03-16 19:25:30 +00:00
LACP_LOCK_INIT(lsc);
Add the trunk(4) driver for providing link aggregation, failover and fault tolerance. This driver allows aggregation of multiple network interfaces as one virtual interface using a number of different protocols/algorithms. failover - Sends traffic through the secondary port if the master becomes inactive. fec - Supports Cisco Fast EtherChannel. lacp - Supports the IEEE 802.3ad Link Aggregation Control Protocol (LACP) and the Marker Protocol. loadbalance - Static loadbalancing using an outgoing hash. roundrobin - Distributes outgoing traffic using a round-robin scheduler through all active ports. This code was obtained from OpenBSD and this also includes 802.3ad LACP support from agr(4) in NetBSD.
2007-04-10 00:27:25 +00:00
TAILQ_INIT(&lsc->lsc_aggregators);
LIST_INIT(&lsc->lsc_ports);
Switch the LACP state machine over to its own mutex to protect the internals, this means that it no longer grabs the lagg rwlock. Use two port table arrays which list the active ports for Tx and switch between them with an atomic op. Now the lagg rwlock is only exclusively locked for management (ioctls) and queuing of lacp control frames isnt needed.
2008-03-16 19:25:30 +00:00
callout_init_mtx(&lsc->lsc_transit_callout, &lsc->lsc_mtx, 0);
callout_init_mtx(&lsc->lsc_callout, &lsc->lsc_mtx, 0);
Add the trunk(4) driver for providing link aggregation, failover and fault tolerance. This driver allows aggregation of multiple network interfaces as one virtual interface using a number of different protocols/algorithms. failover - Sends traffic through the secondary port if the master becomes inactive. fec - Supports Cisco Fast EtherChannel. lacp - Supports the IEEE 802.3ad Link Aggregation Control Protocol (LACP) and the Marker Protocol. loadbalance - Static loadbalancing using an outgoing hash. roundrobin - Distributes outgoing traffic using a round-robin scheduler through all active ports. This code was obtained from OpenBSD and this also includes 802.3ad LACP support from agr(4) in NetBSD.
2007-04-10 00:27:25 +00:00
Rename the trunk(4) driver to lagg(4) as it is too similar to vlan trunking. The name trunk is misused as the networking term trunk means carrying multiple VLANs over a single connection. The IEEE standard for link aggregation (802.3 section 3) does not talk about 'trunk' at all while it is used throughout IEEE 802.1Q in describing vlans. The lagg(4) driver provides link aggregation, failover and fault tolerance. Discussed on: current@
2007-04-17 00:35:11 +00:00
/* if the lagg is already up then do the same */
non-functional cleanup - remove dead code - use consistent variable names - gc unused defines - whitespace cleanup
2007-06-12 07:29:11 +00:00
if (sc->sc_ifp->if_drv_flags & IFF_DRV_RUNNING)
lacp_init(sc);
Add the trunk(4) driver for providing link aggregation, failover and fault tolerance. This driver allows aggregation of multiple network interfaces as one virtual interface using a number of different protocols/algorithms. failover - Sends traffic through the secondary port if the master becomes inactive. fec - Supports Cisco Fast EtherChannel. lacp - Supports the IEEE 802.3ad Link Aggregation Control Protocol (LACP) and the Marker Protocol. loadbalance - Static loadbalancing using an outgoing hash. roundrobin - Distributes outgoing traffic using a round-robin scheduler through all active ports. This code was obtained from OpenBSD and this also includes 802.3ad LACP support from agr(4) in NetBSD.
2007-04-10 00:27:25 +00:00
}
Make lagg protocols detach methods returning void. Sponsored by: Netflix Sponsored by: Nginx, Inc.
2014-09-26 07:12:40 +00:00
void
- When reconfiguring protocol on a lagg, first set it to LAGG_PROTO_NONE, then drop lock, run the attach routines, and then set it to specific proto. This removes tons of WITNESS warnings. - Make lagg protocol attach handlers not failing and allocate memory with M_WAITOK. Sponsored by: Netflix Sponsored by: Nginx, Inc.
2014-09-26 08:42:32 +00:00
lacp_detach(void *psc)
Add the trunk(4) driver for providing link aggregation, failover and fault tolerance. This driver allows aggregation of multiple network interfaces as one virtual interface using a number of different protocols/algorithms. failover - Sends traffic through the secondary port if the master becomes inactive. fec - Supports Cisco Fast EtherChannel. lacp - Supports the IEEE 802.3ad Link Aggregation Control Protocol (LACP) and the Marker Protocol. loadbalance - Static loadbalancing using an outgoing hash. roundrobin - Distributes outgoing traffic using a round-robin scheduler through all active ports. This code was obtained from OpenBSD and this also includes 802.3ad LACP support from agr(4) in NetBSD.
2007-04-10 00:27:25 +00:00
{
- When reconfiguring protocol on a lagg, first set it to LAGG_PROTO_NONE, then drop lock, run the attach routines, and then set it to specific proto. This removes tons of WITNESS warnings. - Make lagg protocol attach handlers not failing and allocate memory with M_WAITOK. Sponsored by: Netflix Sponsored by: Nginx, Inc.
2014-09-26 08:42:32 +00:00
struct lacp_softc *lsc = (struct lacp_softc *)psc;
Add the trunk(4) driver for providing link aggregation, failover and fault tolerance. This driver allows aggregation of multiple network interfaces as one virtual interface using a number of different protocols/algorithms. failover - Sends traffic through the secondary port if the master becomes inactive. fec - Supports Cisco Fast EtherChannel. lacp - Supports the IEEE 802.3ad Link Aggregation Control Protocol (LACP) and the Marker Protocol. loadbalance - Static loadbalancing using an outgoing hash. roundrobin - Distributes outgoing traffic using a round-robin scheduler through all active ports. This code was obtained from OpenBSD and this also includes 802.3ad LACP support from agr(4) in NetBSD.
2007-04-10 00:27:25 +00:00
KASSERT(TAILQ_EMPTY(&lsc->lsc_aggregators),
("aggregators still active"));
KASSERT(lsc->lsc_active_aggregator == NULL,
("aggregator still attached"));
callout_drain(&lsc->lsc_transit_callout);
callout_drain(&lsc->lsc_callout);
Switch the LACP state machine over to its own mutex to protect the internals, this means that it no longer grabs the lagg rwlock. Use two port table arrays which list the active ports for Tx and switch between them with an atomic op. Now the lagg rwlock is only exclusively locked for management (ioctls) and queuing of lacp control frames isnt needed.
2008-03-16 19:25:30 +00:00
LACP_LOCK_DESTROY(lsc);
Add the trunk(4) driver for providing link aggregation, failover and fault tolerance. This driver allows aggregation of multiple network interfaces as one virtual interface using a number of different protocols/algorithms. failover - Sends traffic through the secondary port if the master becomes inactive. fec - Supports Cisco Fast EtherChannel. lacp - Supports the IEEE 802.3ad Link Aggregation Control Protocol (LACP) and the Marker Protocol. loadbalance - Static loadbalancing using an outgoing hash. roundrobin - Distributes outgoing traffic using a round-robin scheduler through all active ports. This code was obtained from OpenBSD and this also includes 802.3ad LACP support from agr(4) in NetBSD.
2007-04-10 00:27:25 +00:00
free(lsc, M_DEVBUF);
}
void
non-functional cleanup - remove dead code - use consistent variable names - gc unused defines - whitespace cleanup
2007-06-12 07:29:11 +00:00
lacp_init(struct lagg_softc *sc)
Add the trunk(4) driver for providing link aggregation, failover and fault tolerance. This driver allows aggregation of multiple network interfaces as one virtual interface using a number of different protocols/algorithms. failover - Sends traffic through the secondary port if the master becomes inactive. fec - Supports Cisco Fast EtherChannel. lacp - Supports the IEEE 802.3ad Link Aggregation Control Protocol (LACP) and the Marker Protocol. loadbalance - Static loadbalancing using an outgoing hash. roundrobin - Distributes outgoing traffic using a round-robin scheduler through all active ports. This code was obtained from OpenBSD and this also includes 802.3ad LACP support from agr(4) in NetBSD.
2007-04-10 00:27:25 +00:00
{
non-functional cleanup - remove dead code - use consistent variable names - gc unused defines - whitespace cleanup
2007-06-12 07:29:11 +00:00
struct lacp_softc *lsc = LACP_SOFTC(sc);
Add the trunk(4) driver for providing link aggregation, failover and fault tolerance. This driver allows aggregation of multiple network interfaces as one virtual interface using a number of different protocols/algorithms. failover - Sends traffic through the secondary port if the master becomes inactive. fec - Supports Cisco Fast EtherChannel. lacp - Supports the IEEE 802.3ad Link Aggregation Control Protocol (LACP) and the Marker Protocol. loadbalance - Static loadbalancing using an outgoing hash. roundrobin - Distributes outgoing traffic using a round-robin scheduler through all active ports. This code was obtained from OpenBSD and this also includes 802.3ad LACP support from agr(4) in NetBSD.
2007-04-10 00:27:25 +00:00
Switch the LACP state machine over to its own mutex to protect the internals, this means that it no longer grabs the lagg rwlock. Use two port table arrays which list the active ports for Tx and switch between them with an atomic op. Now the lagg rwlock is only exclusively locked for management (ioctls) and queuing of lacp control frames isnt needed.
2008-03-16 19:25:30 +00:00
LACP_LOCK(lsc);
Add the trunk(4) driver for providing link aggregation, failover and fault tolerance. This driver allows aggregation of multiple network interfaces as one virtual interface using a number of different protocols/algorithms. failover - Sends traffic through the secondary port if the master becomes inactive. fec - Supports Cisco Fast EtherChannel. lacp - Supports the IEEE 802.3ad Link Aggregation Control Protocol (LACP) and the Marker Protocol. loadbalance - Static loadbalancing using an outgoing hash. roundrobin - Distributes outgoing traffic using a round-robin scheduler through all active ports. This code was obtained from OpenBSD and this also includes 802.3ad LACP support from agr(4) in NetBSD.
2007-04-10 00:27:25 +00:00
callout_reset(&lsc->lsc_callout, hz, lacp_tick, lsc);
Switch the LACP state machine over to its own mutex to protect the internals, this means that it no longer grabs the lagg rwlock. Use two port table arrays which list the active ports for Tx and switch between them with an atomic op. Now the lagg rwlock is only exclusively locked for management (ioctls) and queuing of lacp control frames isnt needed.
2008-03-16 19:25:30 +00:00
LACP_UNLOCK(lsc);
Add the trunk(4) driver for providing link aggregation, failover and fault tolerance. This driver allows aggregation of multiple network interfaces as one virtual interface using a number of different protocols/algorithms. failover - Sends traffic through the secondary port if the master becomes inactive. fec - Supports Cisco Fast EtherChannel. lacp - Supports the IEEE 802.3ad Link Aggregation Control Protocol (LACP) and the Marker Protocol. loadbalance - Static loadbalancing using an outgoing hash. roundrobin - Distributes outgoing traffic using a round-robin scheduler through all active ports. This code was obtained from OpenBSD and this also includes 802.3ad LACP support from agr(4) in NetBSD.
2007-04-10 00:27:25 +00:00
}
void
non-functional cleanup - remove dead code - use consistent variable names - gc unused defines - whitespace cleanup
2007-06-12 07:29:11 +00:00
lacp_stop(struct lagg_softc *sc)
Add the trunk(4) driver for providing link aggregation, failover and fault tolerance. This driver allows aggregation of multiple network interfaces as one virtual interface using a number of different protocols/algorithms. failover - Sends traffic through the secondary port if the master becomes inactive. fec - Supports Cisco Fast EtherChannel. lacp - Supports the IEEE 802.3ad Link Aggregation Control Protocol (LACP) and the Marker Protocol. loadbalance - Static loadbalancing using an outgoing hash. roundrobin - Distributes outgoing traffic using a round-robin scheduler through all active ports. This code was obtained from OpenBSD and this also includes 802.3ad LACP support from agr(4) in NetBSD.
2007-04-10 00:27:25 +00:00
{
non-functional cleanup - remove dead code - use consistent variable names - gc unused defines - whitespace cleanup
2007-06-12 07:29:11 +00:00
struct lacp_softc *lsc = LACP_SOFTC(sc);
Add the trunk(4) driver for providing link aggregation, failover and fault tolerance. This driver allows aggregation of multiple network interfaces as one virtual interface using a number of different protocols/algorithms. failover - Sends traffic through the secondary port if the master becomes inactive. fec - Supports Cisco Fast EtherChannel. lacp - Supports the IEEE 802.3ad Link Aggregation Control Protocol (LACP) and the Marker Protocol. loadbalance - Static loadbalancing using an outgoing hash. roundrobin - Distributes outgoing traffic using a round-robin scheduler through all active ports. This code was obtained from OpenBSD and this also includes 802.3ad LACP support from agr(4) in NetBSD.
2007-04-10 00:27:25 +00:00
Switch the LACP state machine over to its own mutex to protect the internals, this means that it no longer grabs the lagg rwlock. Use two port table arrays which list the active ports for Tx and switch between them with an atomic op. Now the lagg rwlock is only exclusively locked for management (ioctls) and queuing of lacp control frames isnt needed.
2008-03-16 19:25:30 +00:00
LACP_LOCK(lsc);
Add the trunk(4) driver for providing link aggregation, failover and fault tolerance. This driver allows aggregation of multiple network interfaces as one virtual interface using a number of different protocols/algorithms. failover - Sends traffic through the secondary port if the master becomes inactive. fec - Supports Cisco Fast EtherChannel. lacp - Supports the IEEE 802.3ad Link Aggregation Control Protocol (LACP) and the Marker Protocol. loadbalance - Static loadbalancing using an outgoing hash. roundrobin - Distributes outgoing traffic using a round-robin scheduler through all active ports. This code was obtained from OpenBSD and this also includes 802.3ad LACP support from agr(4) in NetBSD.
2007-04-10 00:27:25 +00:00
callout_stop(&lsc->lsc_transit_callout);
callout_stop(&lsc->lsc_callout);
Switch the LACP state machine over to its own mutex to protect the internals, this means that it no longer grabs the lagg rwlock. Use two port table arrays which list the active ports for Tx and switch between them with an atomic op. Now the lagg rwlock is only exclusively locked for management (ioctls) and queuing of lacp control frames isnt needed.
2008-03-16 19:25:30 +00:00
LACP_UNLOCK(lsc);
Add the trunk(4) driver for providing link aggregation, failover and fault tolerance. This driver allows aggregation of multiple network interfaces as one virtual interface using a number of different protocols/algorithms. failover - Sends traffic through the secondary port if the master becomes inactive. fec - Supports Cisco Fast EtherChannel. lacp - Supports the IEEE 802.3ad Link Aggregation Control Protocol (LACP) and the Marker Protocol. loadbalance - Static loadbalancing using an outgoing hash. roundrobin - Distributes outgoing traffic using a round-robin scheduler through all active ports. This code was obtained from OpenBSD and this also includes 802.3ad LACP support from agr(4) in NetBSD.
2007-04-10 00:27:25 +00:00
}
Rename the trunk(4) driver to lagg(4) as it is too similar to vlan trunking. The name trunk is misused as the networking term trunk means carrying multiple VLANs over a single connection. The IEEE standard for link aggregation (802.3 section 3) does not talk about 'trunk' at all while it is used throughout IEEE 802.1Q in describing vlans. The lagg(4) driver provides link aggregation, failover and fault tolerance. Discussed on: current@
2007-04-17 00:35:11 +00:00
struct lagg_port *
non-functional cleanup - remove dead code - use consistent variable names - gc unused defines - whitespace cleanup
2007-06-12 07:29:11 +00:00
lacp_select_tx_port(struct lagg_softc *sc, struct mbuf *m)
Add the trunk(4) driver for providing link aggregation, failover and fault tolerance. This driver allows aggregation of multiple network interfaces as one virtual interface using a number of different protocols/algorithms. failover - Sends traffic through the secondary port if the master becomes inactive. fec - Supports Cisco Fast EtherChannel. lacp - Supports the IEEE 802.3ad Link Aggregation Control Protocol (LACP) and the Marker Protocol. loadbalance - Static loadbalancing using an outgoing hash. roundrobin - Distributes outgoing traffic using a round-robin scheduler through all active ports. This code was obtained from OpenBSD and this also includes 802.3ad LACP support from agr(4) in NetBSD.
2007-04-10 00:27:25 +00:00
{
non-functional cleanup - remove dead code - use consistent variable names - gc unused defines - whitespace cleanup
2007-06-12 07:29:11 +00:00
struct lacp_softc *lsc = LACP_SOFTC(sc);
Switch the LACP state machine over to its own mutex to protect the internals, this means that it no longer grabs the lagg rwlock. Use two port table arrays which list the active ports for Tx and switch between them with an atomic op. Now the lagg rwlock is only exclusively locked for management (ioctls) and queuing of lacp control frames isnt needed.
2008-03-16 19:25:30 +00:00
struct lacp_portmap *pm;
Add the trunk(4) driver for providing link aggregation, failover and fault tolerance. This driver allows aggregation of multiple network interfaces as one virtual interface using a number of different protocols/algorithms. failover - Sends traffic through the secondary port if the master becomes inactive. fec - Supports Cisco Fast EtherChannel. lacp - Supports the IEEE 802.3ad Link Aggregation Control Protocol (LACP) and the Marker Protocol. loadbalance - Static loadbalancing using an outgoing hash. roundrobin - Distributes outgoing traffic using a round-robin scheduler through all active ports. This code was obtained from OpenBSD and this also includes 802.3ad LACP support from agr(4) in NetBSD.
2007-04-10 00:27:25 +00:00
struct lacp_port *lp;
Select lacp egress ports based on NUMA domain This change creates an array of port maps indexed by numa domain for lacp port selection. If we have lacp interfaces in more than one domain, then we select the egress port by indexing into the numa port maps and picking a port on the appropriate numa domain. This is behavior is controlled by the new ifconfig use_numa flag and net.link.lagg.use_numa sysctl/tunable (both modeled after the existing use_flowid), which default to enabled. Reviewed by: bz, hselasky, markj (and scottl, earlier version) Sponsored by: Netflix Differential Revision: https://reviews.freebsd.org/D20060
2019-05-03 14:43:21 +00:00
struct lacp_port **map;
Add the trunk(4) driver for providing link aggregation, failover and fault tolerance. This driver allows aggregation of multiple network interfaces as one virtual interface using a number of different protocols/algorithms. failover - Sends traffic through the secondary port if the master becomes inactive. fec - Supports Cisco Fast EtherChannel. lacp - Supports the IEEE 802.3ad Link Aggregation Control Protocol (LACP) and the Marker Protocol. loadbalance - Static loadbalancing using an outgoing hash. roundrobin - Distributes outgoing traffic using a round-robin scheduler through all active ports. This code was obtained from OpenBSD and this also includes 802.3ad LACP support from agr(4) in NetBSD.
2007-04-10 00:27:25 +00:00
uint32_t hash;
Select lacp egress ports based on NUMA domain This change creates an array of port maps indexed by numa domain for lacp port selection. If we have lacp interfaces in more than one domain, then we select the egress port by indexing into the numa port maps and picking a port on the appropriate numa domain. This is behavior is controlled by the new ifconfig use_numa flag and net.link.lagg.use_numa sysctl/tunable (both modeled after the existing use_flowid), which default to enabled. Reviewed by: bz, hselasky, markj (and scottl, earlier version) Sponsored by: Netflix Differential Revision: https://reviews.freebsd.org/D20060
2019-05-03 14:43:21 +00:00
int count;
Add the trunk(4) driver for providing link aggregation, failover and fault tolerance. This driver allows aggregation of multiple network interfaces as one virtual interface using a number of different protocols/algorithms. failover - Sends traffic through the secondary port if the master becomes inactive. fec - Supports Cisco Fast EtherChannel. lacp - Supports the IEEE 802.3ad Link Aggregation Control Protocol (LACP) and the Marker Protocol. loadbalance - Static loadbalancing using an outgoing hash. roundrobin - Distributes outgoing traffic using a round-robin scheduler through all active ports. This code was obtained from OpenBSD and this also includes 802.3ad LACP support from agr(4) in NetBSD.
2007-04-10 00:27:25 +00:00
if (__predict_false(lsc->lsc_suppress_distributing)) {
LACP_DPRINTF((NULL, "%s: waiting transit\n", __func__));
return (NULL);
}
Switch the LACP state machine over to its own mutex to protect the internals, this means that it no longer grabs the lagg rwlock. Use two port table arrays which list the active ports for Tx and switch between them with an atomic op. Now the lagg rwlock is only exclusively locked for management (ioctls) and queuing of lacp control frames isnt needed.
2008-03-16 19:25:30 +00:00
pm = &lsc->lsc_pmap[lsc->lsc_activemap];
if (pm->pm_count == 0) {
Add the trunk(4) driver for providing link aggregation, failover and fault tolerance. This driver allows aggregation of multiple network interfaces as one virtual interface using a number of different protocols/algorithms. failover - Sends traffic through the secondary port if the master becomes inactive. fec - Supports Cisco Fast EtherChannel. lacp - Supports the IEEE 802.3ad Link Aggregation Control Protocol (LACP) and the Marker Protocol. loadbalance - Static loadbalancing using an outgoing hash. roundrobin - Distributes outgoing traffic using a round-robin scheduler through all active ports. This code was obtained from OpenBSD and this also includes 802.3ad LACP support from agr(4) in NetBSD.
2007-04-10 00:27:25 +00:00
LACP_DPRINTF((NULL, "%s: no active aggregator\n", __func__));
return (NULL);
}
Select lacp egress ports based on NUMA domain This change creates an array of port maps indexed by numa domain for lacp port selection. If we have lacp interfaces in more than one domain, then we select the egress port by indexing into the numa port maps and picking a port on the appropriate numa domain. This is behavior is controlled by the new ifconfig use_numa flag and net.link.lagg.use_numa sysctl/tunable (both modeled after the existing use_flowid), which default to enabled. Reviewed by: bz, hselasky, markj (and scottl, earlier version) Sponsored by: Netflix Differential Revision: https://reviews.freebsd.org/D20060
2019-05-03 14:43:21 +00:00
#ifdef NUMA
if ((sc->sc_opts & LAGG_OPT_USE_NUMA) &&
pm->pm_num_dom > 1 && m->m_pkthdr.numa_domain < MAXMEMDOM) {
count = pm->pm_numa[m->m_pkthdr.numa_domain].count;
if (count > 0) {
map = pm->pm_numa[m->m_pkthdr.numa_domain].map;
} else {
/* No ports on this domain; use global hash. */
map = pm->pm_map;
count = pm->pm_count;
}
} else
#endif
{
map = pm->pm_map;
count = pm->pm_count;
}
Start process of removing the use of the deprecated "M_FLOWID" flag from the FreeBSD network code. The flag is still kept around in the "sys/mbuf.h" header file, but does no longer have any users. Instead the "m_pkthdr.rsstype" field in the mbuf structure is now used to decide the meaning of the "m_pkthdr.flowid" field. To modify the "m_pkthdr.rsstype" field please use the existing "M_HASHTYPE_XXX" macros as defined in the "sys/mbuf.h" header file. This patch introduces new behaviour in the transmit direction. Previously network drivers checked if "M_FLOWID" was set in "m_flags" before using the "m_pkthdr.flowid" field. This check has now now been replaced by checking if "M_HASHTYPE_GET(m)" is different from "M_HASHTYPE_NONE". In the future more hashtypes will be added, for example hashtypes for hardware dedicated flows. "M_HASHTYPE_OPAQUE" indicates that the "m_pkthdr.flowid" value is valid and has no particular type. This change removes the need for an "if" statement in TCP transmit code checking for the presence of a valid flowid value. The "if" statement mentioned above is now a direct variable assignment which is then later checked by the respective network drivers like before. Additional notes: - The SCTP code changes will be committed as a separate patch. - Removal of the "M_FLOWID" flag will also be done separately. - The FreeBSD version has been bumped. MFC after: 1 month Sponsored by: Mellanox Technologies
2014-12-01 11:45:24 +00:00
if ((sc->sc_opts & LAGG_OPT_USE_FLOWID) &&
M_HASHTYPE_GET(m) != M_HASHTYPE_NONE)
Multi-queue NIC drivers and multi-port lagg tend to use the same lower bits of the flowid as each other, resulting in a poor distribution of packets among queues in certain cases. Work around this by adding a set of sysctls for controlling a bit-shift on the flowid when doing multi-port aggrigation in lagg and lacp. By default, lagg/lacp will now use bits 16 and higher instead of 0 and higher. Reviewed by: max Obtained from: Netflix MFC after: 3 days
2013-12-30 01:32:17 +00:00
hash = m->m_pkthdr.flowid >> sc->flowid_shift;
Use the flowid if its available for selecting the tx port.
2009-04-30 14:25:44 +00:00
else
Factor out mbuf hashing code from LAGG driver so that other network drivers can use it. This avoids some code duplication. Add missing default case to all switch statements while at it. Also move the hashing of the IPv6 flow field to layer 4 because the IPv6 flow field is constant on a per L4 connection basis and not on a per L3 network. Differential Revision: https://reviews.freebsd.org/D1987 Sponsored by: Mellanox Technologies MFC after: 1 month
2015-03-11 16:02:24 +00:00
hash = m_ether_tcpip_hash(sc->sc_flags, m, lsc->lsc_hashkey);
Select lacp egress ports based on NUMA domain This change creates an array of port maps indexed by numa domain for lacp port selection. If we have lacp interfaces in more than one domain, then we select the egress port by indexing into the numa port maps and picking a port on the appropriate numa domain. This is behavior is controlled by the new ifconfig use_numa flag and net.link.lagg.use_numa sysctl/tunable (both modeled after the existing use_flowid), which default to enabled. Reviewed by: bz, hselasky, markj (and scottl, earlier version) Sponsored by: Netflix Differential Revision: https://reviews.freebsd.org/D20060
2019-05-03 14:43:21 +00:00
hash %= count;
lp = map[hash];
Add the trunk(4) driver for providing link aggregation, failover and fault tolerance. This driver allows aggregation of multiple network interfaces as one virtual interface using a number of different protocols/algorithms. failover - Sends traffic through the secondary port if the master becomes inactive. fec - Supports Cisco Fast EtherChannel. lacp - Supports the IEEE 802.3ad Link Aggregation Control Protocol (LACP) and the Marker Protocol. loadbalance - Static loadbalancing using an outgoing hash. roundrobin - Distributes outgoing traffic using a round-robin scheduler through all active ports. This code was obtained from OpenBSD and this also includes 802.3ad LACP support from agr(4) in NetBSD.
2007-04-10 00:27:25 +00:00
KASSERT((lp->lp_state & LACP_STATE_DISTRIBUTING) != 0,
("aggregated port is not distributing"));
Rename the trunk(4) driver to lagg(4) as it is too similar to vlan trunking. The name trunk is misused as the networking term trunk means carrying multiple VLANs over a single connection. The IEEE standard for link aggregation (802.3 section 3) does not talk about 'trunk' at all while it is used throughout IEEE 802.1Q in describing vlans. The lagg(4) driver provides link aggregation, failover and fault tolerance. Discussed on: current@
2007-04-17 00:35:11 +00:00
return (lp->lp_lagg);
Add the trunk(4) driver for providing link aggregation, failover and fault tolerance. This driver allows aggregation of multiple network interfaces as one virtual interface using a number of different protocols/algorithms. failover - Sends traffic through the secondary port if the master becomes inactive. fec - Supports Cisco Fast EtherChannel. lacp - Supports the IEEE 802.3ad Link Aggregation Control Protocol (LACP) and the Marker Protocol. loadbalance - Static loadbalancing using an outgoing hash. roundrobin - Distributes outgoing traffic using a round-robin scheduler through all active ports. This code was obtained from OpenBSD and this also includes 802.3ad LACP support from agr(4) in NetBSD.
2007-04-10 00:27:25 +00:00
}
Implement kernel support for hardware rate limited sockets. - Add RATELIMIT kernel configuration keyword which must be set to enable the new functionality. - Add support for hardware driven, Receive Side Scaling, RSS aware, rate limited sendqueues and expose the functionality through the already established SO_MAX_PACING_RATE setsockopt(). The API support rates in the range from 1 to 4Gbytes/s which are suitable for regular TCP and UDP streams. The setsockopt(2) manual page has been updated. - Add rate limit function callback API to "struct ifnet" which supports the following operations: if_snd_tag_alloc(), if_snd_tag_modify(), if_snd_tag_query() and if_snd_tag_free(). - Add support to ifconfig to view, set and clear the IFCAP_TXRTLMT flag, which tells if a network driver supports rate limiting or not. - This patch also adds support for rate limiting through VLAN and LAGG intermediate network devices. - How rate limiting works: 1) The userspace application calls setsockopt() after accepting or making a new connection to set the rate which is then stored in the socket structure in the kernel. Later on when packets are transmitted a check is made in the transmit path for rate changes. A rate change implies a non-blocking ifp->if_snd_tag_alloc() call will be made to the destination network interface, which then sets up a custom sendqueue with the given rate limitation parameter. A "struct m_snd_tag" pointer is returned which serves as a "snd_tag" hint in the m_pkthdr for the subsequently transmitted mbufs. 2) When the network driver sees the "m->m_pkthdr.snd_tag" different from NULL, it will move the packets into a designated rate limited sendqueue given by the snd_tag pointer. It is up to the individual drivers how the rate limited traffic will be rate limited. 3) Route changes are detected by the NIC drivers in the ifp->if_transmit() routine when the ifnet pointer in the incoming snd_tag mismatches the one of the network interface. The network adapter frees the mbuf and returns EAGAIN which causes the ip_output() to release and clear the send tag. Upon next ip_output() a new "snd_tag" will be tried allocated. 4) When the PCB is detached the custom sendqueue will be released by a non-blocking ifp->if_snd_tag_free() call to the currently bound network interface. Reviewed by: wblock (manpages), adrian, gallatin, scottl (network) Differential Revision: https://reviews.freebsd.org/D3687 Sponsored by: Mellanox Technologies MFC after: 3 months
2017-01-18 13:31:17 +00:00
Add kernel-side support for in-kernel TLS. KTLS adds support for in-kernel framing and encryption of Transport Layer Security (1.0-1.2) data on TCP sockets. KTLS only supports offload of TLS for transmitted data. Key negotation must still be performed in userland. Once completed, transmit session keys for a connection are provided to the kernel via a new TCP_TXTLS_ENABLE socket option. All subsequent data transmitted on the socket is placed into TLS frames and encrypted using the supplied keys. Any data written to a KTLS-enabled socket via write(2), aio_write(2), or sendfile(2) is assumed to be application data and is encoded in TLS frames with an application data type. Individual records can be sent with a custom type (e.g. handshake messages) via sendmsg(2) with a new control message (TLS_SET_RECORD_TYPE) specifying the record type. At present, rekeying is not supported though the in-kernel framework should support rekeying. KTLS makes use of the recently added unmapped mbufs to store TLS frames in the socket buffer. Each TLS frame is described by a single ext_pgs mbuf. The ext_pgs structure contains the header of the TLS record (and trailer for encrypted records) as well as references to the associated TLS session. KTLS supports two primary methods of encrypting TLS frames: software TLS and ifnet TLS. Software TLS marks mbufs holding socket data as not ready via M_NOTREADY similar to sendfile(2) when TLS framing information is added to an unmapped mbuf in ktls_frame(). ktls_enqueue() is then called to schedule TLS frames for encryption. In the case of sendfile_iodone() calls ktls_enqueue() instead of pru_ready() leaving the mbufs marked M_NOTREADY until encryption is completed. For other writes (vn_sendfile when pages are available, write(2), etc.), the PRUS_NOTREADY is set when invoking pru_send() along with invoking ktls_enqueue(). A pool of worker threads (the "KTLS" kernel process) encrypts TLS frames queued via ktls_enqueue(). Each TLS frame is temporarily mapped using the direct map and passed to a software encryption backend to perform the actual encryption. (Note: The use of PHYS_TO_DMAP could be replaced with sf_bufs if someone wished to make this work on architectures without a direct map.) KTLS supports pluggable software encryption backends. Internally, Netflix uses proprietary pure-software backends. This commit includes a simple backend in a new ktls_ocf.ko module that uses the kernel's OpenCrypto framework to provide AES-GCM encryption of TLS frames. As a result, software TLS is now a bit of a misnomer as it can make use of hardware crypto accelerators. Once software encryption has finished, the TLS frame mbufs are marked ready via pru_ready(). At this point, the encrypted data appears as regular payload to the TCP stack stored in unmapped mbufs. ifnet TLS permits a NIC to offload the TLS encryption and TCP segmentation. In this mode, a new send tag type (IF_SND_TAG_TYPE_TLS) is allocated on the interface a socket is routed over and associated with a TLS session. TLS records for a TLS session using ifnet TLS are not marked M_NOTREADY but are passed down the stack unencrypted. The ip_output_send() and ip6_output_send() helper functions that apply send tags to outbound IP packets verify that the send tag of the TLS record matches the outbound interface. If so, the packet is tagged with the TLS send tag and sent to the interface. The NIC device driver must recognize packets with the TLS send tag and schedule them for TLS encryption and TCP segmentation. If the the outbound interface does not match the interface in the TLS send tag, the packet is dropped. In addition, a task is scheduled to refresh the TLS send tag for the TLS session. If a new TLS send tag cannot be allocated, the connection is dropped. If a new TLS send tag is allocated, however, subsequent packets will be tagged with the correct TLS send tag. (This latter case has been tested by configuring both ports of a Chelsio T6 in a lagg and failing over from one port to another. As the connections migrated to the new port, new TLS send tags were allocated for the new port and connections resumed without being dropped.) ifnet TLS can be enabled and disabled on supported network interfaces via new '[-]txtls[46]' options to ifconfig(8). ifnet TLS is supported across both vlan devices and lagg interfaces using failover, lacp with flowid enabled, or lacp with flowid enabled. Applications may request the current KTLS mode of a connection via a new TCP_TXTLS_MODE socket option. They can also use this socket option to toggle between software and ifnet TLS modes. In addition, a testing tool is available in tools/tools/switch_tls. This is modeled on tcpdrop and uses similar syntax. However, instead of dropping connections, -s is used to force KTLS connections to switch to software TLS and -i is used to switch to ifnet TLS. Various sysctls and counters are available under the kern.ipc.tls sysctl node. The kern.ipc.tls.enable node must be set to true to enable KTLS (it is off by default). The use of unmapped mbufs must also be enabled via kern.ipc.mb_use_ext_pgs to enable KTLS. KTLS is enabled via the KERN_TLS kernel option. This patch is the culmination of years of work by several folks including Scott Long and Randall Stewart for the original design and implementation; Drew Gallatin for several optimizations including the use of ext_pgs mbufs, the M_NOTREADY mechanism for TLS records awaiting software encryption, and pluggable software crypto backends; and John Baldwin for modifications to support hardware TLS offload. Reviewed by: gallatin, hselasky, rrs Obtained from: Netflix Sponsored by: Netflix, Chelsio Communications Differential Revision: https://reviews.freebsd.org/D21277
2019-08-27 00:01:56 +00:00
#if defined(RATELIMIT) || defined(KERN_TLS)
Implement kernel support for hardware rate limited sockets. - Add RATELIMIT kernel configuration keyword which must be set to enable the new functionality. - Add support for hardware driven, Receive Side Scaling, RSS aware, rate limited sendqueues and expose the functionality through the already established SO_MAX_PACING_RATE setsockopt(). The API support rates in the range from 1 to 4Gbytes/s which are suitable for regular TCP and UDP streams. The setsockopt(2) manual page has been updated. - Add rate limit function callback API to "struct ifnet" which supports the following operations: if_snd_tag_alloc(), if_snd_tag_modify(), if_snd_tag_query() and if_snd_tag_free(). - Add support to ifconfig to view, set and clear the IFCAP_TXRTLMT flag, which tells if a network driver supports rate limiting or not. - This patch also adds support for rate limiting through VLAN and LAGG intermediate network devices. - How rate limiting works: 1) The userspace application calls setsockopt() after accepting or making a new connection to set the rate which is then stored in the socket structure in the kernel. Later on when packets are transmitted a check is made in the transmit path for rate changes. A rate change implies a non-blocking ifp->if_snd_tag_alloc() call will be made to the destination network interface, which then sets up a custom sendqueue with the given rate limitation parameter. A "struct m_snd_tag" pointer is returned which serves as a "snd_tag" hint in the m_pkthdr for the subsequently transmitted mbufs. 2) When the network driver sees the "m->m_pkthdr.snd_tag" different from NULL, it will move the packets into a designated rate limited sendqueue given by the snd_tag pointer. It is up to the individual drivers how the rate limited traffic will be rate limited. 3) Route changes are detected by the NIC drivers in the ifp->if_transmit() routine when the ifnet pointer in the incoming snd_tag mismatches the one of the network interface. The network adapter frees the mbuf and returns EAGAIN which causes the ip_output() to release and clear the send tag. Upon next ip_output() a new "snd_tag" will be tried allocated. 4) When the PCB is detached the custom sendqueue will be released by a non-blocking ifp->if_snd_tag_free() call to the currently bound network interface. Reviewed by: wblock (manpages), adrian, gallatin, scottl (network) Differential Revision: https://reviews.freebsd.org/D3687 Sponsored by: Mellanox Technologies MFC after: 3 months
2017-01-18 13:31:17 +00:00
struct lagg_port *
lacp_select_tx_port_by_hash(struct lagg_softc *sc, uint32_t flowid)
{
struct lacp_softc *lsc = LACP_SOFTC(sc);
struct lacp_portmap *pm;
struct lacp_port *lp;
uint32_t hash;
if (__predict_false(lsc->lsc_suppress_distributing)) {
LACP_DPRINTF((NULL, "%s: waiting transit\n", __func__));
return (NULL);
}
pm = &lsc->lsc_pmap[lsc->lsc_activemap];
if (pm->pm_count == 0) {
LACP_DPRINTF((NULL, "%s: no active aggregator\n", __func__));
return (NULL);
}
hash = flowid >> sc->flowid_shift;
hash %= pm->pm_count;
lp = pm->pm_map[hash];
return (lp->lp_lagg);
}
#endif
Add the trunk(4) driver for providing link aggregation, failover and fault tolerance. This driver allows aggregation of multiple network interfaces as one virtual interface using a number of different protocols/algorithms. failover - Sends traffic through the secondary port if the master becomes inactive. fec - Supports Cisco Fast EtherChannel. lacp - Supports the IEEE 802.3ad Link Aggregation Control Protocol (LACP) and the Marker Protocol. loadbalance - Static loadbalancing using an outgoing hash. roundrobin - Distributes outgoing traffic using a round-robin scheduler through all active ports. This code was obtained from OpenBSD and this also includes 802.3ad LACP support from agr(4) in NetBSD.
2007-04-10 00:27:25 +00:00
/*
* lacp_suppress_distributing: drop transmit packets for a while
* to preserve packet ordering.
*/
static void
lacp_suppress_distributing(struct lacp_softc *lsc, struct lacp_aggregator *la)
{
Implement the Marker Protocol. A marker frame is placed on the interface queue of each port and any further packets are blocked, when the all the marker frames have been returned to us from the remote network device then we can be sure that all interface queues are empty. This is needed when a port is added or removed from the aggregation since it will affect the hash based distribution, if the queues are not empty then a packet from an existing connection may be placed on a different interface and arrive out of order. This was previously achieved by suppressing transmission for 1 second, now that there is an active feedback this timeout as been increased to 3 seconds and used as a fallback.
2007-05-19 07:47:04 +00:00
struct lacp_port *lp;
Add the trunk(4) driver for providing link aggregation, failover and fault tolerance. This driver allows aggregation of multiple network interfaces as one virtual interface using a number of different protocols/algorithms. failover - Sends traffic through the secondary port if the master becomes inactive. fec - Supports Cisco Fast EtherChannel. lacp - Supports the IEEE 802.3ad Link Aggregation Control Protocol (LACP) and the Marker Protocol. loadbalance - Static loadbalancing using an outgoing hash. roundrobin - Distributes outgoing traffic using a round-robin scheduler through all active ports. This code was obtained from OpenBSD and this also includes 802.3ad LACP support from agr(4) in NetBSD.
2007-04-10 00:27:25 +00:00
if (lsc->lsc_active_aggregator != la) {
return;
}
Turn LACP debugging from a compile time option to a sysctl, it is very handy to be able to turn it on when negotiation to a switch misbehaves. Submitted by: Andrew Boyer MFC after: 3 days
2012-05-26 08:09:01 +00:00
LACP_TRACE(NULL);
Add the trunk(4) driver for providing link aggregation, failover and fault tolerance. This driver allows aggregation of multiple network interfaces as one virtual interface using a number of different protocols/algorithms. failover - Sends traffic through the secondary port if the master becomes inactive. fec - Supports Cisco Fast EtherChannel. lacp - Supports the IEEE 802.3ad Link Aggregation Control Protocol (LACP) and the Marker Protocol. loadbalance - Static loadbalancing using an outgoing hash. roundrobin - Distributes outgoing traffic using a round-robin scheduler through all active ports. This code was obtained from OpenBSD and this also includes 802.3ad LACP support from agr(4) in NetBSD.
2007-04-10 00:27:25 +00:00
lsc->lsc_suppress_distributing = TRUE;
Implement the Marker Protocol. A marker frame is placed on the interface queue of each port and any further packets are blocked, when the all the marker frames have been returned to us from the remote network device then we can be sure that all interface queues are empty. This is needed when a port is added or removed from the aggregation since it will affect the hash based distribution, if the queues are not empty then a packet from an existing connection may be placed on a different interface and arrive out of order. This was previously achieved by suppressing transmission for 1 second, now that there is an active feedback this timeout as been increased to 3 seconds and used as a fallback.
2007-05-19 07:47:04 +00:00
/* send a marker frame down each port to verify the queues are empty */
LIST_FOREACH(lp, &lsc->lsc_ports, lp_next) {
lp->lp_flags |= LACP_PORT_MARK;
lacp_xmit_marker(lp);
}
/* set a timeout for the marker frames */
Add the trunk(4) driver for providing link aggregation, failover and fault tolerance. This driver allows aggregation of multiple network interfaces as one virtual interface using a number of different protocols/algorithms. failover - Sends traffic through the secondary port if the master becomes inactive. fec - Supports Cisco Fast EtherChannel. lacp - Supports the IEEE 802.3ad Link Aggregation Control Protocol (LACP) and the Marker Protocol. loadbalance - Static loadbalancing using an outgoing hash. roundrobin - Distributes outgoing traffic using a round-robin scheduler through all active ports. This code was obtained from OpenBSD and this also includes 802.3ad LACP support from agr(4) in NetBSD.
2007-04-10 00:27:25 +00:00
callout_reset(&lsc->lsc_transit_callout,
LACP_TRANSIT_DELAY * hz / 1000, lacp_transit_expire, lsc);
}
static int
lacp_compare_peerinfo(const struct lacp_peerinfo *a,
const struct lacp_peerinfo *b)
{
return (memcmp(a, b, offsetof(struct lacp_peerinfo, lip_state)));
}
static int
lacp_compare_systemid(const struct lacp_systemid *a,
const struct lacp_systemid *b)
{
return (memcmp(a, b, sizeof(*a)));
}
#if 0 /* unused */
static int
lacp_compare_portid(const struct lacp_portid *a,
const struct lacp_portid *b)
{
return (memcmp(a, b, sizeof(*a)));
}
#endif
static uint64_t
lacp_aggregator_bandwidth(struct lacp_aggregator *la)
{
struct lacp_port *lp;
uint64_t speed;
lp = TAILQ_FIRST(&la->la_ports);
if (lp == NULL) {
return (0);
}
speed = ifmedia_baudrate(lp->lp_media);
speed *= la->la_nports;
if (speed == 0) {
LACP_DPRINTF((lp, "speed 0? media=0x%x nports=%d\n",
lp->lp_media, la->la_nports));
}
return (speed);
}
/*
* lacp_select_active_aggregator: select an aggregator to be used to transmit
Rename the trunk(4) driver to lagg(4) as it is too similar to vlan trunking. The name trunk is misused as the networking term trunk means carrying multiple VLANs over a single connection. The IEEE standard for link aggregation (802.3 section 3) does not talk about 'trunk' at all while it is used throughout IEEE 802.1Q in describing vlans. The lagg(4) driver provides link aggregation, failover and fault tolerance. Discussed on: current@
2007-04-17 00:35:11 +00:00
* packets from lagg(4) interface.
Add the trunk(4) driver for providing link aggregation, failover and fault tolerance. This driver allows aggregation of multiple network interfaces as one virtual interface using a number of different protocols/algorithms. failover - Sends traffic through the secondary port if the master becomes inactive. fec - Supports Cisco Fast EtherChannel. lacp - Supports the IEEE 802.3ad Link Aggregation Control Protocol (LACP) and the Marker Protocol. loadbalance - Static loadbalancing using an outgoing hash. roundrobin - Distributes outgoing traffic using a round-robin scheduler through all active ports. This code was obtained from OpenBSD and this also includes 802.3ad LACP support from agr(4) in NetBSD.
2007-04-10 00:27:25 +00:00
*/
static void
lacp_select_active_aggregator(struct lacp_softc *lsc)
{
struct lacp_aggregator *la;
struct lacp_aggregator *best_la = NULL;
uint64_t best_speed = 0;
char buf[LACP_LAGIDSTR_MAX+1];
Turn LACP debugging from a compile time option to a sysctl, it is very handy to be able to turn it on when negotiation to a switch misbehaves. Submitted by: Andrew Boyer MFC after: 3 days
2012-05-26 08:09:01 +00:00
LACP_TRACE(NULL);
Add the trunk(4) driver for providing link aggregation, failover and fault tolerance. This driver allows aggregation of multiple network interfaces as one virtual interface using a number of different protocols/algorithms. failover - Sends traffic through the secondary port if the master becomes inactive. fec - Supports Cisco Fast EtherChannel. lacp - Supports the IEEE 802.3ad Link Aggregation Control Protocol (LACP) and the Marker Protocol. loadbalance - Static loadbalancing using an outgoing hash. roundrobin - Distributes outgoing traffic using a round-robin scheduler through all active ports. This code was obtained from OpenBSD and this also includes 802.3ad LACP support from agr(4) in NetBSD.
2007-04-10 00:27:25 +00:00
TAILQ_FOREACH(la, &lsc->lsc_aggregators, la_q) {
uint64_t speed;
if (la->la_nports == 0) {
continue;
}
speed = lacp_aggregator_bandwidth(la);
LACP_DPRINTF((NULL, "%s, speed=%jd, nports=%d\n",
lacp_format_lagid_aggregator(la, buf, sizeof(buf)),
speed, la->la_nports));
Compare the partner system priority when choosing the aggregator.
2007-05-19 09:37:04 +00:00
- Whitespace. - Remove caddr_t.
2014-09-26 12:35:58 +00:00
/*
* This aggregator is chosen if the partner has a better
* system priority or, the total aggregated speed is higher
* or, it is already the chosen aggregator
Compare the partner system priority when choosing the aggregator.
2007-05-19 09:37:04 +00:00
*/
if ((best_la != NULL && LACP_SYS_PRI(la->la_partner) <
- Whitespace. - Remove caddr_t.
2014-09-26 12:35:58 +00:00
LACP_SYS_PRI(best_la->la_partner)) ||
Compare the partner system priority when choosing the aggregator.
2007-05-19 09:37:04 +00:00
speed > best_speed ||
Add the trunk(4) driver for providing link aggregation, failover and fault tolerance. This driver allows aggregation of multiple network interfaces as one virtual interface using a number of different protocols/algorithms. failover - Sends traffic through the secondary port if the master becomes inactive. fec - Supports Cisco Fast EtherChannel. lacp - Supports the IEEE 802.3ad Link Aggregation Control Protocol (LACP) and the Marker Protocol. loadbalance - Static loadbalancing using an outgoing hash. roundrobin - Distributes outgoing traffic using a round-robin scheduler through all active ports. This code was obtained from OpenBSD and this also includes 802.3ad LACP support from agr(4) in NetBSD.
2007-04-10 00:27:25 +00:00
(speed == best_speed &&
la == lsc->lsc_active_aggregator)) {
best_la = la;
best_speed = speed;
}
}
KASSERT(best_la == NULL || best_la->la_nports > 0,
("invalid aggregator refcnt"));
KASSERT(best_la == NULL || !TAILQ_EMPTY(&best_la->la_ports),
("invalid aggregator list"));
if (lsc->lsc_active_aggregator != best_la) {
LACP_DPRINTF((NULL, "active aggregator changed\n"));
LACP_DPRINTF((NULL, "old %s\n",
lacp_format_lagid_aggregator(lsc->lsc_active_aggregator,
buf, sizeof(buf))));
} else {
LACP_DPRINTF((NULL, "active aggregator not changed\n"));
}
LACP_DPRINTF((NULL, "new %s\n",
lacp_format_lagid_aggregator(best_la, buf, sizeof(buf))));
if (lsc->lsc_active_aggregator != best_la) {
lsc->lsc_active_aggregator = best_la;
Switch the LACP state machine over to its own mutex to protect the internals, this means that it no longer grabs the lagg rwlock. Use two port table arrays which list the active ports for Tx and switch between them with an atomic op. Now the lagg rwlock is only exclusively locked for management (ioctls) and queuing of lacp control frames isnt needed.
2008-03-16 19:25:30 +00:00
lacp_update_portmap(lsc);
Add the trunk(4) driver for providing link aggregation, failover and fault tolerance. This driver allows aggregation of multiple network interfaces as one virtual interface using a number of different protocols/algorithms. failover - Sends traffic through the secondary port if the master becomes inactive. fec - Supports Cisco Fast EtherChannel. lacp - Supports the IEEE 802.3ad Link Aggregation Control Protocol (LACP) and the Marker Protocol. loadbalance - Static loadbalancing using an outgoing hash. roundrobin - Distributes outgoing traffic using a round-robin scheduler through all active ports. This code was obtained from OpenBSD and this also includes 802.3ad LACP support from agr(4) in NetBSD.
2007-04-10 00:27:25 +00:00
if (best_la) {
lacp_suppress_distributing(lsc, best_la);
}
}
}
Switch the LACP state machine over to its own mutex to protect the internals, this means that it no longer grabs the lagg rwlock. Use two port table arrays which list the active ports for Tx and switch between them with an atomic op. Now the lagg rwlock is only exclusively locked for management (ioctls) and queuing of lacp control frames isnt needed.
2008-03-16 19:25:30 +00:00
/*
* Updated the inactive portmap array with the new list of ports and
* make it live.
*/
static void
lacp_update_portmap(struct lacp_softc *lsc)
{
Bring over some link aggregation / LACP protocol improvements and debugging additions. * Add some new tracing events to aid in debugging. * Add in a debugging mode to drop transmit and received frames, specifically to test whether seeing or hearing heartbeats correctly cause LACP to drop the port. * Add in (and make default) a strict LACP mode, which requires the heartbeat on a port to be heard before it's used. Sometimes vendor ports will hang but the link layer stays up, resulting in hung traffic. * Add logging the number of link status flaps, again to aid in debugging badly behaving switch ports. * Calculate the lagg interface port speed as the multiple of the configured ports, rather than the largest. Obtained from: Netflix MFC after: 2 weeks
2013-07-13 04:25:03 +00:00
struct lagg_softc *sc = lsc->lsc_softc;
Switch the LACP state machine over to its own mutex to protect the internals, this means that it no longer grabs the lagg rwlock. Use two port table arrays which list the active ports for Tx and switch between them with an atomic op. Now the lagg rwlock is only exclusively locked for management (ioctls) and queuing of lacp control frames isnt needed.
2008-03-16 19:25:30 +00:00
struct lacp_aggregator *la;
struct lacp_portmap *p;
struct lacp_port *lp;
Bring over some link aggregation / LACP protocol improvements and debugging additions. * Add some new tracing events to aid in debugging. * Add in a debugging mode to drop transmit and received frames, specifically to test whether seeing or hearing heartbeats correctly cause LACP to drop the port. * Add in (and make default) a strict LACP mode, which requires the heartbeat on a port to be heard before it's used. Sometimes vendor ports will hang but the link layer stays up, resulting in hung traffic. * Add logging the number of link status flaps, again to aid in debugging badly behaving switch ports. * Calculate the lagg interface port speed as the multiple of the configured ports, rather than the largest. Obtained from: Netflix MFC after: 2 weeks
2013-07-13 04:25:03 +00:00
uint64_t speed;
Switch the LACP state machine over to its own mutex to protect the internals, this means that it no longer grabs the lagg rwlock. Use two port table arrays which list the active ports for Tx and switch between them with an atomic op. Now the lagg rwlock is only exclusively locked for management (ioctls) and queuing of lacp control frames isnt needed.
2008-03-16 19:25:30 +00:00
u_int newmap;
int i;
Select lacp egress ports based on NUMA domain This change creates an array of port maps indexed by numa domain for lacp port selection. If we have lacp interfaces in more than one domain, then we select the egress port by indexing into the numa port maps and picking a port on the appropriate numa domain. This is behavior is controlled by the new ifconfig use_numa flag and net.link.lagg.use_numa sysctl/tunable (both modeled after the existing use_flowid), which default to enabled. Reviewed by: bz, hselasky, markj (and scottl, earlier version) Sponsored by: Netflix Differential Revision: https://reviews.freebsd.org/D20060
2019-05-03 14:43:21 +00:00
#ifdef NUMA
int count;
uint8_t domain;
#endif
Switch the LACP state machine over to its own mutex to protect the internals, this means that it no longer grabs the lagg rwlock. Use two port table arrays which list the active ports for Tx and switch between them with an atomic op. Now the lagg rwlock is only exclusively locked for management (ioctls) and queuing of lacp control frames isnt needed.
2008-03-16 19:25:30 +00:00
newmap = lsc->lsc_activemap == 0 ? 1 : 0;
p = &lsc->lsc_pmap[newmap];
la = lsc->lsc_active_aggregator;
Bring over some link aggregation / LACP protocol improvements and debugging additions. * Add some new tracing events to aid in debugging. * Add in a debugging mode to drop transmit and received frames, specifically to test whether seeing or hearing heartbeats correctly cause LACP to drop the port. * Add in (and make default) a strict LACP mode, which requires the heartbeat on a port to be heard before it's used. Sometimes vendor ports will hang but the link layer stays up, resulting in hung traffic. * Add logging the number of link status flaps, again to aid in debugging badly behaving switch ports. * Calculate the lagg interface port speed as the multiple of the configured ports, rather than the largest. Obtained from: Netflix MFC after: 2 weeks
2013-07-13 04:25:03 +00:00
speed = 0;
Switch the LACP state machine over to its own mutex to protect the internals, this means that it no longer grabs the lagg rwlock. Use two port table arrays which list the active ports for Tx and switch between them with an atomic op. Now the lagg rwlock is only exclusively locked for management (ioctls) and queuing of lacp control frames isnt needed.
2008-03-16 19:25:30 +00:00
bzero(p, sizeof(struct lacp_portmap));
if (la != NULL && la->la_nports > 0) {
p->pm_count = la->la_nports;
i = 0;
Select lacp egress ports based on NUMA domain This change creates an array of port maps indexed by numa domain for lacp port selection. If we have lacp interfaces in more than one domain, then we select the egress port by indexing into the numa port maps and picking a port on the appropriate numa domain. This is behavior is controlled by the new ifconfig use_numa flag and net.link.lagg.use_numa sysctl/tunable (both modeled after the existing use_flowid), which default to enabled. Reviewed by: bz, hselasky, markj (and scottl, earlier version) Sponsored by: Netflix Differential Revision: https://reviews.freebsd.org/D20060
2019-05-03 14:43:21 +00:00
TAILQ_FOREACH(lp, &la->la_ports, lp_dist_q) {
Switch the LACP state machine over to its own mutex to protect the internals, this means that it no longer grabs the lagg rwlock. Use two port table arrays which list the active ports for Tx and switch between them with an atomic op. Now the lagg rwlock is only exclusively locked for management (ioctls) and queuing of lacp control frames isnt needed.
2008-03-16 19:25:30 +00:00
p->pm_map[i++] = lp;
Select lacp egress ports based on NUMA domain This change creates an array of port maps indexed by numa domain for lacp port selection. If we have lacp interfaces in more than one domain, then we select the egress port by indexing into the numa port maps and picking a port on the appropriate numa domain. This is behavior is controlled by the new ifconfig use_numa flag and net.link.lagg.use_numa sysctl/tunable (both modeled after the existing use_flowid), which default to enabled. Reviewed by: bz, hselasky, markj (and scottl, earlier version) Sponsored by: Netflix Differential Revision: https://reviews.freebsd.org/D20060
2019-05-03 14:43:21 +00:00
#ifdef NUMA
domain = lp->lp_ifp->if_numa_domain;
if (domain >= MAXMEMDOM)
continue;
count = p->pm_numa[domain].count;
p->pm_numa[domain].map[count] = lp;
p->pm_numa[domain].count++;
#endif
}
Switch the LACP state machine over to its own mutex to protect the internals, this means that it no longer grabs the lagg rwlock. Use two port table arrays which list the active ports for Tx and switch between them with an atomic op. Now the lagg rwlock is only exclusively locked for management (ioctls) and queuing of lacp control frames isnt needed.
2008-03-16 19:25:30 +00:00
KASSERT(i == p->pm_count, ("Invalid port count"));
Select lacp egress ports based on NUMA domain This change creates an array of port maps indexed by numa domain for lacp port selection. If we have lacp interfaces in more than one domain, then we select the egress port by indexing into the numa port maps and picking a port on the appropriate numa domain. This is behavior is controlled by the new ifconfig use_numa flag and net.link.lagg.use_numa sysctl/tunable (both modeled after the existing use_flowid), which default to enabled. Reviewed by: bz, hselasky, markj (and scottl, earlier version) Sponsored by: Netflix Differential Revision: https://reviews.freebsd.org/D20060
2019-05-03 14:43:21 +00:00
#ifdef NUMA
for (i = 0; i < MAXMEMDOM; i++) {
if (p->pm_numa[i].count != 0)
p->pm_num_dom++;
}
#endif
Bring over some link aggregation / LACP protocol improvements and debugging additions. * Add some new tracing events to aid in debugging. * Add in a debugging mode to drop transmit and received frames, specifically to test whether seeing or hearing heartbeats correctly cause LACP to drop the port. * Add in (and make default) a strict LACP mode, which requires the heartbeat on a port to be heard before it's used. Sometimes vendor ports will hang but the link layer stays up, resulting in hung traffic. * Add logging the number of link status flaps, again to aid in debugging badly behaving switch ports. * Calculate the lagg interface port speed as the multiple of the configured ports, rather than the largest. Obtained from: Netflix MFC after: 2 weeks
2013-07-13 04:25:03 +00:00
speed = lacp_aggregator_bandwidth(la);
Switch the LACP state machine over to its own mutex to protect the internals, this means that it no longer grabs the lagg rwlock. Use two port table arrays which list the active ports for Tx and switch between them with an atomic op. Now the lagg rwlock is only exclusively locked for management (ioctls) and queuing of lacp control frames isnt needed.
2008-03-16 19:25:30 +00:00
}
Bring over some link aggregation / LACP protocol improvements and debugging additions. * Add some new tracing events to aid in debugging. * Add in a debugging mode to drop transmit and received frames, specifically to test whether seeing or hearing heartbeats correctly cause LACP to drop the port. * Add in (and make default) a strict LACP mode, which requires the heartbeat on a port to be heard before it's used. Sometimes vendor ports will hang but the link layer stays up, resulting in hung traffic. * Add logging the number of link status flaps, again to aid in debugging badly behaving switch ports. * Calculate the lagg interface port speed as the multiple of the configured ports, rather than the largest. Obtained from: Netflix MFC after: 2 weeks
2013-07-13 04:25:03 +00:00
sc->sc_ifp->if_baudrate = speed;
Switch the LACP state machine over to its own mutex to protect the internals, this means that it no longer grabs the lagg rwlock. Use two port table arrays which list the active ports for Tx and switch between them with an atomic op. Now the lagg rwlock is only exclusively locked for management (ioctls) and queuing of lacp control frames isnt needed.
2008-03-16 19:25:30 +00:00
/* switch the active portmap over */
atomic_store_rel_int(&lsc->lsc_activemap, newmap);
LACP_DPRINTF((NULL, "Set table %d with %d ports\n",
lsc->lsc_activemap,
lsc->lsc_pmap[lsc->lsc_activemap].pm_count));
}
Add the trunk(4) driver for providing link aggregation, failover and fault tolerance. This driver allows aggregation of multiple network interfaces as one virtual interface using a number of different protocols/algorithms. failover - Sends traffic through the secondary port if the master becomes inactive. fec - Supports Cisco Fast EtherChannel. lacp - Supports the IEEE 802.3ad Link Aggregation Control Protocol (LACP) and the Marker Protocol. loadbalance - Static loadbalancing using an outgoing hash. roundrobin - Distributes outgoing traffic using a round-robin scheduler through all active ports. This code was obtained from OpenBSD and this also includes 802.3ad LACP support from agr(4) in NetBSD.
2007-04-10 00:27:25 +00:00
static uint16_t
lacp_compose_key(struct lacp_port *lp)
{
Rename the trunk(4) driver to lagg(4) as it is too similar to vlan trunking. The name trunk is misused as the networking term trunk means carrying multiple VLANs over a single connection. The IEEE standard for link aggregation (802.3 section 3) does not talk about 'trunk' at all while it is used throughout IEEE 802.1Q in describing vlans. The lagg(4) driver provides link aggregation, failover and fault tolerance. Discussed on: current@
2007-04-17 00:35:11 +00:00
struct lagg_port *lgp = lp->lp_lagg;
non-functional cleanup - remove dead code - use consistent variable names - gc unused defines - whitespace cleanup
2007-06-12 07:29:11 +00:00
struct lagg_softc *sc = lgp->lp_softc;
Add the trunk(4) driver for providing link aggregation, failover and fault tolerance. This driver allows aggregation of multiple network interfaces as one virtual interface using a number of different protocols/algorithms. failover - Sends traffic through the secondary port if the master becomes inactive. fec - Supports Cisco Fast EtherChannel. lacp - Supports the IEEE 802.3ad Link Aggregation Control Protocol (LACP) and the Marker Protocol. loadbalance - Static loadbalancing using an outgoing hash. roundrobin - Distributes outgoing traffic using a round-robin scheduler through all active ports. This code was obtained from OpenBSD and this also includes 802.3ad LACP support from agr(4) in NetBSD.
2007-04-10 00:27:25 +00:00
u_int media = lp->lp_media;
uint16_t key;
if ((lp->lp_state & LACP_STATE_AGGREGATION) == 0) {
/*
* non-aggregatable links should have unique keys.
*
* XXX this isn't really unique as if_index is 16 bit.
*/
/* bit 0..14: (some bits of) if_index of this port */
key = lp->lp_ifp->if_index;
/* bit 15: 1 */
key |= 0x8000;
} else {
u_int subtype = IFM_SUBTYPE(media);
- Add a disabled state for ports that can not be aggregated - Refine check for lacp links, set to disabled if not suitable
2007-05-03 08:56:20 +00:00
KASSERT(IFM_TYPE(media) == IFM_ETHER, ("invalid media type"));
KASSERT((media & IFM_FDX) != 0, ("aggregating HDX interface"));
Add the trunk(4) driver for providing link aggregation, failover and fault tolerance. This driver allows aggregation of multiple network interfaces as one virtual interface using a number of different protocols/algorithms. failover - Sends traffic through the secondary port if the master becomes inactive. fec - Supports Cisco Fast EtherChannel. lacp - Supports the IEEE 802.3ad Link Aggregation Control Protocol (LACP) and the Marker Protocol. loadbalance - Static loadbalancing using an outgoing hash. roundrobin - Distributes outgoing traffic using a round-robin scheduler through all active ports. This code was obtained from OpenBSD and this also includes 802.3ad LACP support from agr(4) in NetBSD.
2007-04-10 00:27:25 +00:00
Use the same actor key for media types of the same speed. PR: 176097 MFC after: 2 weeks
2013-10-17 15:14:58 +00:00
/* bit 0..4: IFM_SUBTYPE modulo speed */
switch (subtype) {
case IFM_10_T:
case IFM_10_2:
case IFM_10_5:
case IFM_10_STP:
case IFM_10_FL:
key = IFM_10_T;
break;
case IFM_100_TX:
case IFM_100_FX:
case IFM_100_T4:
case IFM_100_VG:
case IFM_100_T2:
ifmedia changes: - Extend the number of available subtypes for Ethernet media by using some of the ifmedia word's option bits to help denote subtypes. As a result, the number of possible Ethernet subtype values increases from 31 to 511. - Use some of those new values to define new media types. - lacp_compose_key() recgonizes the new Ethernet media types added. (Change made as required by a comment in if_media.h) - New ioctl, SIOGIFXMEDIA, to handle getting the new extended media types. SIOCGIFMEDIA is retained for backwards compatibility. - Changes to ifconfig to allow it to handle the new extended media types. Submitted by: mike@karels.net (original), hselasky Reviewed by: jfvogel, gnn, hselasky Approved by: jfvogel (mentor), gnn (mentor) Differential Revision: http://reviews.freebsd.org/D1965
2015-04-07 21:31:17 +00:00
case IFM_100_T:
if_media: Add new 2.5G/5G/25G/40G/50G/100G/200G/400G media types Upcoming Ethernet hardware will support new media types that aren't in the kernel yet, so they are added here. These mostly include new 25G/50G/100G media types; and this commit introduces new 200G/400G speeds and media. Reviewed by: hselasky@, jhb@ MFC after: 1 week Sponsored by: Intel Corporation Differential Revision: https://reviews.freebsd.org/D16731
2018-08-22 18:19:56 +00:00
case IFM_100_SGMII:
Use the same actor key for media types of the same speed. PR: 176097 MFC after: 2 weeks
2013-10-17 15:14:58 +00:00
key = IFM_100_TX;
break;
case IFM_1000_SX:
case IFM_1000_LX:
case IFM_1000_CX:
case IFM_1000_T:
ifmedia changes: - Extend the number of available subtypes for Ethernet media by using some of the ifmedia word's option bits to help denote subtypes. As a result, the number of possible Ethernet subtype values increases from 31 to 511. - Use some of those new values to define new media types. - lacp_compose_key() recgonizes the new Ethernet media types added. (Change made as required by a comment in if_media.h) - New ioctl, SIOGIFXMEDIA, to handle getting the new extended media types. SIOCGIFMEDIA is retained for backwards compatibility. - Changes to ifconfig to allow it to handle the new extended media types. Submitted by: mike@karels.net (original), hselasky Reviewed by: jfvogel, gnn, hselasky Approved by: jfvogel (mentor), gnn (mentor) Differential Revision: http://reviews.freebsd.org/D1965
2015-04-07 21:31:17 +00:00
case IFM_1000_KX:
case IFM_1000_SGMII:
case IFM_1000_CX_SGMII:
Use the same actor key for media types of the same speed. PR: 176097 MFC after: 2 weeks
2013-10-17 15:14:58 +00:00
key = IFM_1000_SX;
break;
case IFM_10G_LR:
case IFM_10G_SR:
case IFM_10G_CX4:
case IFM_10G_TWINAX:
case IFM_10G_TWINAX_LONG:
case IFM_10G_LRM:
case IFM_10G_T:
ifmedia changes: - Extend the number of available subtypes for Ethernet media by using some of the ifmedia word's option bits to help denote subtypes. As a result, the number of possible Ethernet subtype values increases from 31 to 511. - Use some of those new values to define new media types. - lacp_compose_key() recgonizes the new Ethernet media types added. (Change made as required by a comment in if_media.h) - New ioctl, SIOGIFXMEDIA, to handle getting the new extended media types. SIOCGIFMEDIA is retained for backwards compatibility. - Changes to ifconfig to allow it to handle the new extended media types. Submitted by: mike@karels.net (original), hselasky Reviewed by: jfvogel, gnn, hselasky Approved by: jfvogel (mentor), gnn (mentor) Differential Revision: http://reviews.freebsd.org/D1965
2015-04-07 21:31:17 +00:00
case IFM_10G_KX4:
case IFM_10G_KR:
case IFM_10G_CR1:
case IFM_10G_ER:
case IFM_10G_SFI:
Add several new media types to if_media.h These include several 25G types (for active direct attach cables and LR modules), and a missing type for 10G active direct attach. Differential Revision: https://reviews.freebsd.org/D10425 Reviewed by: smh, imp MFC after: 3 days Sponsored by: Intel Corporation
2017-05-10 18:33:40 +00:00
case IFM_10G_AOC:
Use the same actor key for media types of the same speed. PR: 176097 MFC after: 2 weeks
2013-10-17 15:14:58 +00:00
key = IFM_10G_LR;
break;
ifmedia changes: - Extend the number of available subtypes for Ethernet media by using some of the ifmedia word's option bits to help denote subtypes. As a result, the number of possible Ethernet subtype values increases from 31 to 511. - Use some of those new values to define new media types. - lacp_compose_key() recgonizes the new Ethernet media types added. (Change made as required by a comment in if_media.h) - New ioctl, SIOGIFXMEDIA, to handle getting the new extended media types. SIOCGIFMEDIA is retained for backwards compatibility. - Changes to ifconfig to allow it to handle the new extended media types. Submitted by: mike@karels.net (original), hselasky Reviewed by: jfvogel, gnn, hselasky Approved by: jfvogel (mentor), gnn (mentor) Differential Revision: http://reviews.freebsd.org/D1965
2015-04-07 21:31:17 +00:00
case IFM_20G_KR2:
key = IFM_20G_KR2;
break;
case IFM_2500_KX:
case IFM_2500_T:
if_media: Add new 2.5G/5G/25G/40G/50G/100G/200G/400G media types Upcoming Ethernet hardware will support new media types that aren't in the kernel yet, so they are added here. These mostly include new 25G/50G/100G media types; and this commit introduces new 200G/400G speeds and media. Reviewed by: hselasky@, jhb@ MFC after: 1 week Sponsored by: Intel Corporation Differential Revision: https://reviews.freebsd.org/D16731
2018-08-22 18:19:56 +00:00
case IFM_2500_X:
ifmedia changes: - Extend the number of available subtypes for Ethernet media by using some of the ifmedia word's option bits to help denote subtypes. As a result, the number of possible Ethernet subtype values increases from 31 to 511. - Use some of those new values to define new media types. - lacp_compose_key() recgonizes the new Ethernet media types added. (Change made as required by a comment in if_media.h) - New ioctl, SIOGIFXMEDIA, to handle getting the new extended media types. SIOCGIFMEDIA is retained for backwards compatibility. - Changes to ifconfig to allow it to handle the new extended media types. Submitted by: mike@karels.net (original), hselasky Reviewed by: jfvogel, gnn, hselasky Approved by: jfvogel (mentor), gnn (mentor) Differential Revision: http://reviews.freebsd.org/D1965
2015-04-07 21:31:17 +00:00
key = IFM_2500_KX;
break;
case IFM_5000_T:
if_media: Add new 2.5G/5G/25G/40G/50G/100G/200G/400G media types Upcoming Ethernet hardware will support new media types that aren't in the kernel yet, so they are added here. These mostly include new 25G/50G/100G media types; and this commit introduces new 200G/400G speeds and media. Reviewed by: hselasky@, jhb@ MFC after: 1 week Sponsored by: Intel Corporation Differential Revision: https://reviews.freebsd.org/D16731
2018-08-22 18:19:56 +00:00
case IFM_5000_KR:
case IFM_5000_KR_S:
case IFM_5000_KR1:
ifmedia changes: - Extend the number of available subtypes for Ethernet media by using some of the ifmedia word's option bits to help denote subtypes. As a result, the number of possible Ethernet subtype values increases from 31 to 511. - Use some of those new values to define new media types. - lacp_compose_key() recgonizes the new Ethernet media types added. (Change made as required by a comment in if_media.h) - New ioctl, SIOGIFXMEDIA, to handle getting the new extended media types. SIOCGIFMEDIA is retained for backwards compatibility. - Changes to ifconfig to allow it to handle the new extended media types. Submitted by: mike@karels.net (original), hselasky Reviewed by: jfvogel, gnn, hselasky Approved by: jfvogel (mentor), gnn (mentor) Differential Revision: http://reviews.freebsd.org/D1965
2015-04-07 21:31:17 +00:00
key = IFM_5000_T;
break;
case IFM_50G_PCIE:
case IFM_50G_CR2:
case IFM_50G_KR2:
if_media: Add new 2.5G/5G/25G/40G/50G/100G/200G/400G media types Upcoming Ethernet hardware will support new media types that aren't in the kernel yet, so they are added here. These mostly include new 25G/50G/100G media types; and this commit introduces new 200G/400G speeds and media. Reviewed by: hselasky@, jhb@ MFC after: 1 week Sponsored by: Intel Corporation Differential Revision: https://reviews.freebsd.org/D16731
2018-08-22 18:19:56 +00:00
case IFM_50G_SR2:
case IFM_50G_LR2:
case IFM_50G_LAUI2_AC:
case IFM_50G_LAUI2:
case IFM_50G_AUI2_AC:
case IFM_50G_AUI2:
case IFM_50G_CP:
case IFM_50G_SR:
case IFM_50G_LR:
case IFM_50G_FR:
case IFM_50G_KR_PAM4:
case IFM_50G_AUI1_AC:
case IFM_50G_AUI1:
ifmedia changes: - Extend the number of available subtypes for Ethernet media by using some of the ifmedia word's option bits to help denote subtypes. As a result, the number of possible Ethernet subtype values increases from 31 to 511. - Use some of those new values to define new media types. - lacp_compose_key() recgonizes the new Ethernet media types added. (Change made as required by a comment in if_media.h) - New ioctl, SIOGIFXMEDIA, to handle getting the new extended media types. SIOCGIFMEDIA is retained for backwards compatibility. - Changes to ifconfig to allow it to handle the new extended media types. Submitted by: mike@karels.net (original), hselasky Reviewed by: jfvogel, gnn, hselasky Approved by: jfvogel (mentor), gnn (mentor) Differential Revision: http://reviews.freebsd.org/D1965
2015-04-07 21:31:17 +00:00
key = IFM_50G_PCIE;
break;
case IFM_56G_R4:
key = IFM_56G_R4;
break;
case IFM_25G_PCIE:
case IFM_25G_CR:
case IFM_25G_KR:
case IFM_25G_SR:
Add several new media types to if_media.h These include several 25G types (for active direct attach cables and LR modules), and a missing type for 10G active direct attach. Differential Revision: https://reviews.freebsd.org/D10425 Reviewed by: smh, imp MFC after: 3 days Sponsored by: Intel Corporation
2017-05-10 18:33:40 +00:00
case IFM_25G_LR:
case IFM_25G_ACC:
case IFM_25G_AOC:
if_media: Add new 2.5G/5G/25G/40G/50G/100G/200G/400G media types Upcoming Ethernet hardware will support new media types that aren't in the kernel yet, so they are added here. These mostly include new 25G/50G/100G media types; and this commit introduces new 200G/400G speeds and media. Reviewed by: hselasky@, jhb@ MFC after: 1 week Sponsored by: Intel Corporation Differential Revision: https://reviews.freebsd.org/D16731
2018-08-22 18:19:56 +00:00
case IFM_25G_T:
case IFM_25G_CR_S:
case IFM_25G_CR1:
case IFM_25G_KR_S:
case IFM_25G_AUI:
case IFM_25G_KR1:
ifmedia changes: - Extend the number of available subtypes for Ethernet media by using some of the ifmedia word's option bits to help denote subtypes. As a result, the number of possible Ethernet subtype values increases from 31 to 511. - Use some of those new values to define new media types. - lacp_compose_key() recgonizes the new Ethernet media types added. (Change made as required by a comment in if_media.h) - New ioctl, SIOGIFXMEDIA, to handle getting the new extended media types. SIOCGIFMEDIA is retained for backwards compatibility. - Changes to ifconfig to allow it to handle the new extended media types. Submitted by: mike@karels.net (original), hselasky Reviewed by: jfvogel, gnn, hselasky Approved by: jfvogel (mentor), gnn (mentor) Differential Revision: http://reviews.freebsd.org/D1965
2015-04-07 21:31:17 +00:00
key = IFM_25G_PCIE;
break;
Use the same actor key for media types of the same speed. PR: 176097 MFC after: 2 weeks
2013-10-17 15:14:58 +00:00
case IFM_40G_CR4:
case IFM_40G_SR4:
case IFM_40G_LR4:
ifmedia changes: - Extend the number of available subtypes for Ethernet media by using some of the ifmedia word's option bits to help denote subtypes. As a result, the number of possible Ethernet subtype values increases from 31 to 511. - Use some of those new values to define new media types. - lacp_compose_key() recgonizes the new Ethernet media types added. (Change made as required by a comment in if_media.h) - New ioctl, SIOGIFXMEDIA, to handle getting the new extended media types. SIOCGIFMEDIA is retained for backwards compatibility. - Changes to ifconfig to allow it to handle the new extended media types. Submitted by: mike@karels.net (original), hselasky Reviewed by: jfvogel, gnn, hselasky Approved by: jfvogel (mentor), gnn (mentor) Differential Revision: http://reviews.freebsd.org/D1965
2015-04-07 21:31:17 +00:00
case IFM_40G_XLPPI:
case IFM_40G_KR4:
if_media: Add new 2.5G/5G/25G/40G/50G/100G/200G/400G media types Upcoming Ethernet hardware will support new media types that aren't in the kernel yet, so they are added here. These mostly include new 25G/50G/100G media types; and this commit introduces new 200G/400G speeds and media. Reviewed by: hselasky@, jhb@ MFC after: 1 week Sponsored by: Intel Corporation Differential Revision: https://reviews.freebsd.org/D16731
2018-08-22 18:19:56 +00:00
case IFM_40G_XLAUI:
case IFM_40G_XLAUI_AC:
case IFM_40G_ER4:
Use the same actor key for media types of the same speed. PR: 176097 MFC after: 2 weeks
2013-10-17 15:14:58 +00:00
key = IFM_40G_CR4;
break;
ifmedia changes: - Extend the number of available subtypes for Ethernet media by using some of the ifmedia word's option bits to help denote subtypes. As a result, the number of possible Ethernet subtype values increases from 31 to 511. - Use some of those new values to define new media types. - lacp_compose_key() recgonizes the new Ethernet media types added. (Change made as required by a comment in if_media.h) - New ioctl, SIOGIFXMEDIA, to handle getting the new extended media types. SIOCGIFMEDIA is retained for backwards compatibility. - Changes to ifconfig to allow it to handle the new extended media types. Submitted by: mike@karels.net (original), hselasky Reviewed by: jfvogel, gnn, hselasky Approved by: jfvogel (mentor), gnn (mentor) Differential Revision: http://reviews.freebsd.org/D1965
2015-04-07 21:31:17 +00:00
case IFM_100G_CR4:
case IFM_100G_SR4:
case IFM_100G_KR4:
case IFM_100G_LR4:
if_media: Add new 2.5G/5G/25G/40G/50G/100G/200G/400G media types Upcoming Ethernet hardware will support new media types that aren't in the kernel yet, so they are added here. These mostly include new 25G/50G/100G media types; and this commit introduces new 200G/400G speeds and media. Reviewed by: hselasky@, jhb@ MFC after: 1 week Sponsored by: Intel Corporation Differential Revision: https://reviews.freebsd.org/D16731
2018-08-22 18:19:56 +00:00
case IFM_100G_CAUI4_AC:
case IFM_100G_CAUI4:
case IFM_100G_AUI4_AC:
case IFM_100G_AUI4:
case IFM_100G_CR_PAM4:
case IFM_100G_KR_PAM4:
case IFM_100G_CP2:
case IFM_100G_SR2:
case IFM_100G_DR:
case IFM_100G_KR2_PAM4:
case IFM_100G_CAUI2_AC:
case IFM_100G_CAUI2:
case IFM_100G_AUI2_AC:
case IFM_100G_AUI2:
ifmedia changes: - Extend the number of available subtypes for Ethernet media by using some of the ifmedia word's option bits to help denote subtypes. As a result, the number of possible Ethernet subtype values increases from 31 to 511. - Use some of those new values to define new media types. - lacp_compose_key() recgonizes the new Ethernet media types added. (Change made as required by a comment in if_media.h) - New ioctl, SIOGIFXMEDIA, to handle getting the new extended media types. SIOCGIFMEDIA is retained for backwards compatibility. - Changes to ifconfig to allow it to handle the new extended media types. Submitted by: mike@karels.net (original), hselasky Reviewed by: jfvogel, gnn, hselasky Approved by: jfvogel (mentor), gnn (mentor) Differential Revision: http://reviews.freebsd.org/D1965
2015-04-07 21:31:17 +00:00
key = IFM_100G_CR4;
break;
if_media: Add new 2.5G/5G/25G/40G/50G/100G/200G/400G media types Upcoming Ethernet hardware will support new media types that aren't in the kernel yet, so they are added here. These mostly include new 25G/50G/100G media types; and this commit introduces new 200G/400G speeds and media. Reviewed by: hselasky@, jhb@ MFC after: 1 week Sponsored by: Intel Corporation Differential Revision: https://reviews.freebsd.org/D16731
2018-08-22 18:19:56 +00:00
case IFM_200G_CR4_PAM4:
case IFM_200G_SR4:
case IFM_200G_FR4:
case IFM_200G_LR4:
case IFM_200G_DR4:
case IFM_200G_KR4_PAM4:
case IFM_200G_AUI4_AC:
case IFM_200G_AUI4:
case IFM_200G_AUI8_AC:
case IFM_200G_AUI8:
key = IFM_200G_CR4_PAM4;
break;
case IFM_400G_FR8:
case IFM_400G_LR8:
case IFM_400G_DR4:
case IFM_400G_AUI8_AC:
case IFM_400G_AUI8:
key = IFM_400G_FR8;
break;
Use the same actor key for media types of the same speed. PR: 176097 MFC after: 2 weeks
2013-10-17 15:14:58 +00:00
default:
key = subtype;
ifmedia changes: - Extend the number of available subtypes for Ethernet media by using some of the ifmedia word's option bits to help denote subtypes. As a result, the number of possible Ethernet subtype values increases from 31 to 511. - Use some of those new values to define new media types. - lacp_compose_key() recgonizes the new Ethernet media types added. (Change made as required by a comment in if_media.h) - New ioctl, SIOGIFXMEDIA, to handle getting the new extended media types. SIOCGIFMEDIA is retained for backwards compatibility. - Changes to ifconfig to allow it to handle the new extended media types. Submitted by: mike@karels.net (original), hselasky Reviewed by: jfvogel, gnn, hselasky Approved by: jfvogel (mentor), gnn (mentor) Differential Revision: http://reviews.freebsd.org/D1965
2015-04-07 21:31:17 +00:00
break;
Use the same actor key for media types of the same speed. PR: 176097 MFC after: 2 weeks
2013-10-17 15:14:58 +00:00
}
Rename the trunk(4) driver to lagg(4) as it is too similar to vlan trunking. The name trunk is misused as the networking term trunk means carrying multiple VLANs over a single connection. The IEEE standard for link aggregation (802.3 section 3) does not talk about 'trunk' at all while it is used throughout IEEE 802.1Q in describing vlans. The lagg(4) driver provides link aggregation, failover and fault tolerance. Discussed on: current@
2007-04-17 00:35:11 +00:00
/* bit 5..14: (some bits of) if_index of lagg device */
non-functional cleanup - remove dead code - use consistent variable names - gc unused defines - whitespace cleanup
2007-06-12 07:29:11 +00:00
key |= 0x7fe0 & ((sc->sc_ifp->if_index) << 5);
Add the trunk(4) driver for providing link aggregation, failover and fault tolerance. This driver allows aggregation of multiple network interfaces as one virtual interface using a number of different protocols/algorithms. failover - Sends traffic through the secondary port if the master becomes inactive. fec - Supports Cisco Fast EtherChannel. lacp - Supports the IEEE 802.3ad Link Aggregation Control Protocol (LACP) and the Marker Protocol. loadbalance - Static loadbalancing using an outgoing hash. roundrobin - Distributes outgoing traffic using a round-robin scheduler through all active ports. This code was obtained from OpenBSD and this also includes 802.3ad LACP support from agr(4) in NetBSD.
2007-04-10 00:27:25 +00:00
/* bit 15: 0 */
}
return (htons(key));
}
static void
lacp_aggregator_addref(struct lacp_softc *lsc, struct lacp_aggregator *la)
{
char buf[LACP_LAGIDSTR_MAX+1];
LACP_DPRINTF((NULL, "%s: lagid=%s, refcnt %d -> %d\n",
__func__,
lacp_format_lagid(&la->la_actor, &la->la_partner,
buf, sizeof(buf)),
la->la_refcnt, la->la_refcnt + 1));
KASSERT(la->la_refcnt > 0, ("refcount <= 0"));
la->la_refcnt++;
KASSERT(la->la_refcnt > la->la_nports, ("invalid refcount"));
}
static void
lacp_aggregator_delref(struct lacp_softc *lsc, struct lacp_aggregator *la)
{
char buf[LACP_LAGIDSTR_MAX+1];
LACP_DPRINTF((NULL, "%s: lagid=%s, refcnt %d -> %d\n",
__func__,
lacp_format_lagid(&la->la_actor, &la->la_partner,
buf, sizeof(buf)),
la->la_refcnt, la->la_refcnt - 1));
KASSERT(la->la_refcnt > la->la_nports, ("invalid refcnt"));
la->la_refcnt--;
if (la->la_refcnt > 0) {
return;
}
KASSERT(la->la_refcnt == 0, ("refcount not zero"));
KASSERT(lsc->lsc_active_aggregator != la, ("aggregator active"));
TAILQ_REMOVE(&lsc->lsc_aggregators, la, la_q);
free(la, M_DEVBUF);
}
/*
* lacp_aggregator_get: allocate an aggregator.
*/
static struct lacp_aggregator *
lacp_aggregator_get(struct lacp_softc *lsc, struct lacp_port *lp)
{
struct lacp_aggregator *la;
la = malloc(sizeof(*la), M_DEVBUF, M_NOWAIT);
if (la) {
la->la_refcnt = 1;
la->la_nports = 0;
TAILQ_INIT(&la->la_ports);
la->la_pending = 0;
TAILQ_INSERT_TAIL(&lsc->lsc_aggregators, la, la_q);
}
return (la);
}
/*
* lacp_fill_aggregator_id: setup a newly allocated aggregator from a port.
*/
static void
lacp_fill_aggregator_id(struct lacp_aggregator *la, const struct lacp_port *lp)
{
lacp_fill_aggregator_id_peer(&la->la_partner, &lp->lp_partner);
lacp_fill_aggregator_id_peer(&la->la_actor, &lp->lp_actor);
la->la_actor.lip_state = lp->lp_state & LACP_STATE_AGGREGATION;
}
static void
lacp_fill_aggregator_id_peer(struct lacp_peerinfo *lpi_aggr,
const struct lacp_peerinfo *lpi_port)
{
memset(lpi_aggr, 0, sizeof(*lpi_aggr));
lpi_aggr->lip_systemid = lpi_port->lip_systemid;
lpi_aggr->lip_key = lpi_port->lip_key;
}
/*
* lacp_aggregator_is_compatible: check if a port can join to an aggregator.
*/
static int
lacp_aggregator_is_compatible(const struct lacp_aggregator *la,
const struct lacp_port *lp)
{
if (!(lp->lp_state & LACP_STATE_AGGREGATION) ||
!(lp->lp_partner.lip_state & LACP_STATE_AGGREGATION)) {
return (0);
}
if (!(la->la_actor.lip_state & LACP_STATE_AGGREGATION)) {
return (0);
}
if (!lacp_peerinfo_is_compatible(&la->la_partner, &lp->lp_partner)) {
return (0);
}
if (!lacp_peerinfo_is_compatible(&la->la_actor, &lp->lp_actor)) {
return (0);
}
return (1);
}
static int
lacp_peerinfo_is_compatible(const struct lacp_peerinfo *a,
const struct lacp_peerinfo *b)
{
if (memcmp(&a->lip_systemid, &b->lip_systemid,
sizeof(a->lip_systemid))) {
return (0);
}
if (memcmp(&a->lip_key, &b->lip_key, sizeof(a->lip_key))) {
return (0);
}
return (1);
}
static void
lacp_port_enable(struct lacp_port *lp)
{
lp->lp_state |= LACP_STATE_AGGREGATION;
}
static void
lacp_port_disable(struct lacp_port *lp)
{
lacp_set_mux(lp, LACP_MUX_DETACHED);
lp->lp_state &= ~LACP_STATE_AGGREGATION;
lp->lp_selected = LACP_UNSELECTED;
lacp_sm_rx_record_default(lp);
lp->lp_partner.lip_state &= ~LACP_STATE_AGGREGATION;
lp->lp_state &= ~LACP_STATE_EXPIRED;
}
/*
* lacp_select: select an aggregator. create one if necessary.
*/
static void
lacp_select(struct lacp_port *lp)
{
struct lacp_softc *lsc = lp->lp_lsc;
struct lacp_aggregator *la;
char buf[LACP_LAGIDSTR_MAX+1];
if (lp->lp_aggregator) {
return;
}
Do some minimal work to better conform to the 802.3ad (LACP) standard. In particular, don't set the synchronized bit for the peer unless it truly appears to be synchronized to us. Also, don't set our own synchronized bit unless we have actually seen a remote system. Prior to this change, we were seeing some strange behavior, such as: 1. We send an advertisement with the Activity, Aggregation, and Default flags, followed by an advertisement with the Activity, Aggregation, Synchronization, and Default flags. However, we hadn't seen an advertisement from another peer and were still advertising the default (NULL) peer. A closer examination of the in-kernel data structures (using kgdb) showed that the system had added the default (NULL) peer as a valid aggregator for the segment. 2. We were receiving an advertisement from a peer that included the default (NULL) peer instead of including our system information. However, we responded with an advertisement that included the Synchronization flag for both our system and the peer. (Since the peer's advertisement did not include our system information, we shouldn't add the synchronization bit for the peer.) This patch corrects those two items. Reviewed by: smh MFC after: 2 weeks Sponsored by: Netflix Differential Revision: https://reviews.freebsd.org/D9485
2017-02-26 00:19:02 +00:00
/* If we haven't heard from our peer, skip this step. */
if (lp->lp_state & LACP_STATE_DEFAULTED)
return;
Add the trunk(4) driver for providing link aggregation, failover and fault tolerance. This driver allows aggregation of multiple network interfaces as one virtual interface using a number of different protocols/algorithms. failover - Sends traffic through the secondary port if the master becomes inactive. fec - Supports Cisco Fast EtherChannel. lacp - Supports the IEEE 802.3ad Link Aggregation Control Protocol (LACP) and the Marker Protocol. loadbalance - Static loadbalancing using an outgoing hash. roundrobin - Distributes outgoing traffic using a round-robin scheduler through all active ports. This code was obtained from OpenBSD and this also includes 802.3ad LACP support from agr(4) in NetBSD.
2007-04-10 00:27:25 +00:00
KASSERT(!LACP_TIMER_ISARMED(lp, LACP_TIMER_WAIT_WHILE),
("timer_wait_while still active"));
LACP_DPRINTF((lp, "port lagid=%s\n",
lacp_format_lagid(&lp->lp_actor, &lp->lp_partner,
buf, sizeof(buf))));
TAILQ_FOREACH(la, &lsc->lsc_aggregators, la_q) {
if (lacp_aggregator_is_compatible(la, lp)) {
break;
}
}
if (la == NULL) {
la = lacp_aggregator_get(lsc, lp);
if (la == NULL) {
LACP_DPRINTF((lp, "aggregator creation failed\n"));
/*
* will retry on the next tick.
*/
return;
}
lacp_fill_aggregator_id(la, lp);
LACP_DPRINTF((lp, "aggregator created\n"));
} else {
LACP_DPRINTF((lp, "compatible aggregator found\n"));
Switch the LACP state machine over to its own mutex to protect the internals, this means that it no longer grabs the lagg rwlock. Use two port table arrays which list the active ports for Tx and switch between them with an atomic op. Now the lagg rwlock is only exclusively locked for management (ioctls) and queuing of lacp control frames isnt needed.
2008-03-16 19:25:30 +00:00
if (la->la_refcnt == LACP_MAX_PORTS)
return;
Add the trunk(4) driver for providing link aggregation, failover and fault tolerance. This driver allows aggregation of multiple network interfaces as one virtual interface using a number of different protocols/algorithms. failover - Sends traffic through the secondary port if the master becomes inactive. fec - Supports Cisco Fast EtherChannel. lacp - Supports the IEEE 802.3ad Link Aggregation Control Protocol (LACP) and the Marker Protocol. loadbalance - Static loadbalancing using an outgoing hash. roundrobin - Distributes outgoing traffic using a round-robin scheduler through all active ports. This code was obtained from OpenBSD and this also includes 802.3ad LACP support from agr(4) in NetBSD.
2007-04-10 00:27:25 +00:00
lacp_aggregator_addref(lsc, la);
}
LACP_DPRINTF((lp, "aggregator lagid=%s\n",
lacp_format_lagid(&la->la_actor, &la->la_partner,
buf, sizeof(buf))));
lp->lp_aggregator = la;
lp->lp_selected = LACP_SELECTED;
}
/*
* lacp_unselect: finish unselect/detach process.
*/
static void
lacp_unselect(struct lacp_port *lp)
{
struct lacp_softc *lsc = lp->lp_lsc;
struct lacp_aggregator *la = lp->lp_aggregator;
KASSERT(!LACP_TIMER_ISARMED(lp, LACP_TIMER_WAIT_WHILE),
("timer_wait_while still active"));
if (la == NULL) {
return;
}
lp->lp_aggregator = NULL;
lacp_aggregator_delref(lsc, la);
}
/* mux machine */
static void
lacp_sm_mux(struct lacp_port *lp)
{
Bring over some link aggregation / LACP protocol improvements and debugging additions. * Add some new tracing events to aid in debugging. * Add in a debugging mode to drop transmit and received frames, specifically to test whether seeing or hearing heartbeats correctly cause LACP to drop the port. * Add in (and make default) a strict LACP mode, which requires the heartbeat on a port to be heard before it's used. Sometimes vendor ports will hang but the link layer stays up, resulting in hung traffic. * Add logging the number of link status flaps, again to aid in debugging badly behaving switch ports. * Calculate the lagg interface port speed as the multiple of the configured ports, rather than the largest. Obtained from: Netflix MFC after: 2 weeks
2013-07-13 04:25:03 +00:00
struct lagg_port *lgp = lp->lp_lagg;
struct lagg_softc *sc = lgp->lp_softc;
Add the trunk(4) driver for providing link aggregation, failover and fault tolerance. This driver allows aggregation of multiple network interfaces as one virtual interface using a number of different protocols/algorithms. failover - Sends traffic through the secondary port if the master becomes inactive. fec - Supports Cisco Fast EtherChannel. lacp - Supports the IEEE 802.3ad Link Aggregation Control Protocol (LACP) and the Marker Protocol. loadbalance - Static loadbalancing using an outgoing hash. roundrobin - Distributes outgoing traffic using a round-robin scheduler through all active ports. This code was obtained from OpenBSD and this also includes 802.3ad LACP support from agr(4) in NetBSD.
2007-04-10 00:27:25 +00:00
enum lacp_mux_state new_state;
boolean_t p_sync =
(lp->lp_partner.lip_state & LACP_STATE_SYNC) != 0;
boolean_t p_collecting =
(lp->lp_partner.lip_state & LACP_STATE_COLLECTING) != 0;
enum lacp_selected selected = lp->lp_selected;
struct lacp_aggregator *la;
Virtualize lagg(4) cloner. This change fixes a panic when tearing down if_lagg(4) interfaces which were cloned in a vnet jail. Sysctl nodes which are dynamically generated for each cloned interface (net.link.lagg.N.*) have been removed, and use_flowid and flowid_shift ifconfig(8) parameters have been added instead. Flags and per-interface statistics counters are displayed in "ifconfig -v". CR: D842
2014-10-01 21:37:32 +00:00
if (V_lacp_debug > 1)
Bring over some link aggregation / LACP protocol improvements and debugging additions. * Add some new tracing events to aid in debugging. * Add in a debugging mode to drop transmit and received frames, specifically to test whether seeing or hearing heartbeats correctly cause LACP to drop the port. * Add in (and make default) a strict LACP mode, which requires the heartbeat on a port to be heard before it's used. Sometimes vendor ports will hang but the link layer stays up, resulting in hung traffic. * Add logging the number of link status flaps, again to aid in debugging badly behaving switch ports. * Calculate the lagg interface port speed as the multiple of the configured ports, rather than the largest. Obtained from: Netflix MFC after: 2 weeks
2013-07-13 04:25:03 +00:00
lacp_dprintf(lp, "%s: state= 0x%x, selected= 0x%x, "
"p_sync= 0x%x, p_collecting= 0x%x\n", __func__,
lp->lp_mux_state, selected, p_sync, p_collecting);
Add the trunk(4) driver for providing link aggregation, failover and fault tolerance. This driver allows aggregation of multiple network interfaces as one virtual interface using a number of different protocols/algorithms. failover - Sends traffic through the secondary port if the master becomes inactive. fec - Supports Cisco Fast EtherChannel. lacp - Supports the IEEE 802.3ad Link Aggregation Control Protocol (LACP) and the Marker Protocol. loadbalance - Static loadbalancing using an outgoing hash. roundrobin - Distributes outgoing traffic using a round-robin scheduler through all active ports. This code was obtained from OpenBSD and this also includes 802.3ad LACP support from agr(4) in NetBSD.
2007-04-10 00:27:25 +00:00
re_eval:
la = lp->lp_aggregator;
KASSERT(lp->lp_mux_state == LACP_MUX_DETACHED || la != NULL,
("MUX not detached"));
new_state = lp->lp_mux_state;
switch (lp->lp_mux_state) {
case LACP_MUX_DETACHED:
if (selected != LACP_UNSELECTED) {
new_state = LACP_MUX_WAITING;
}
break;
case LACP_MUX_WAITING:
KASSERT(la->la_pending > 0 ||
!LACP_TIMER_ISARMED(lp, LACP_TIMER_WAIT_WHILE),
("timer_wait_while still active"));
if (selected == LACP_SELECTED && la->la_pending == 0) {
new_state = LACP_MUX_ATTACHED;
} else if (selected == LACP_UNSELECTED) {
new_state = LACP_MUX_DETACHED;
}
break;
case LACP_MUX_ATTACHED:
if (selected == LACP_SELECTED && p_sync) {
new_state = LACP_MUX_COLLECTING;
} else if (selected != LACP_SELECTED) {
new_state = LACP_MUX_DETACHED;
}
break;
case LACP_MUX_COLLECTING:
if (selected == LACP_SELECTED && p_sync && p_collecting) {
new_state = LACP_MUX_DISTRIBUTING;
} else if (selected != LACP_SELECTED || !p_sync) {
new_state = LACP_MUX_ATTACHED;
}
break;
case LACP_MUX_DISTRIBUTING:
if (selected != LACP_SELECTED || !p_sync || !p_collecting) {
new_state = LACP_MUX_COLLECTING;
Fix typo. Sponsored by: Netflix
2013-07-25 19:10:23 +00:00
lacp_dprintf(lp, "Interface stopped DISTRIBUTING, possible flapping\n");
Bring over some link aggregation / LACP protocol improvements and debugging additions. * Add some new tracing events to aid in debugging. * Add in a debugging mode to drop transmit and received frames, specifically to test whether seeing or hearing heartbeats correctly cause LACP to drop the port. * Add in (and make default) a strict LACP mode, which requires the heartbeat on a port to be heard before it's used. Sometimes vendor ports will hang but the link layer stays up, resulting in hung traffic. * Add logging the number of link status flaps, again to aid in debugging badly behaving switch ports. * Calculate the lagg interface port speed as the multiple of the configured ports, rather than the largest. Obtained from: Netflix MFC after: 2 weeks
2013-07-13 04:25:03 +00:00
sc->sc_flapping++;
Add the trunk(4) driver for providing link aggregation, failover and fault tolerance. This driver allows aggregation of multiple network interfaces as one virtual interface using a number of different protocols/algorithms. failover - Sends traffic through the secondary port if the master becomes inactive. fec - Supports Cisco Fast EtherChannel. lacp - Supports the IEEE 802.3ad Link Aggregation Control Protocol (LACP) and the Marker Protocol. loadbalance - Static loadbalancing using an outgoing hash. roundrobin - Distributes outgoing traffic using a round-robin scheduler through all active ports. This code was obtained from OpenBSD and this also includes 802.3ad LACP support from agr(4) in NetBSD.
2007-04-10 00:27:25 +00:00
}
break;
default:
panic("%s: unknown state", __func__);
}
if (lp->lp_mux_state == new_state) {
return;
}
lacp_set_mux(lp, new_state);
goto re_eval;
}
static void
lacp_set_mux(struct lacp_port *lp, enum lacp_mux_state new_state)
{
struct lacp_aggregator *la = lp->lp_aggregator;
if (lp->lp_mux_state == new_state) {
return;
}
switch (new_state) {
case LACP_MUX_DETACHED:
lp->lp_state &= ~LACP_STATE_SYNC;
lacp_disable_distributing(lp);
lacp_disable_collecting(lp);
lacp_sm_assert_ntt(lp);
/* cancel timer */
if (LACP_TIMER_ISARMED(lp, LACP_TIMER_WAIT_WHILE)) {
KASSERT(la->la_pending > 0,
("timer_wait_while not active"));
la->la_pending--;
}
LACP_TIMER_DISARM(lp, LACP_TIMER_WAIT_WHILE);
lacp_unselect(lp);
break;
case LACP_MUX_WAITING:
LACP_TIMER_ARM(lp, LACP_TIMER_WAIT_WHILE,
LACP_AGGREGATE_WAIT_TIME);
la->la_pending++;
break;
case LACP_MUX_ATTACHED:
lp->lp_state |= LACP_STATE_SYNC;
lacp_disable_collecting(lp);
lacp_sm_assert_ntt(lp);
break;
case LACP_MUX_COLLECTING:
lacp_enable_collecting(lp);
lacp_disable_distributing(lp);
lacp_sm_assert_ntt(lp);
break;
case LACP_MUX_DISTRIBUTING:
lacp_enable_distributing(lp);
break;
default:
panic("%s: unknown state", __func__);
}
LACP_DPRINTF((lp, "mux_state %d -> %d\n", lp->lp_mux_state, new_state));
lp->lp_mux_state = new_state;
}
static void
lacp_sm_mux_timer(struct lacp_port *lp)
{
struct lacp_aggregator *la = lp->lp_aggregator;
char buf[LACP_LAGIDSTR_MAX+1];
KASSERT(la->la_pending > 0, ("no pending event"));
LACP_DPRINTF((lp, "%s: aggregator %s, pending %d -> %d\n", __func__,
lacp_format_lagid(&la->la_actor, &la->la_partner,
buf, sizeof(buf)),
la->la_pending, la->la_pending - 1));
la->la_pending--;
}
/* periodic transmit machine */
static void
lacp_sm_ptx_update_timeout(struct lacp_port *lp, uint8_t oldpstate)
{
if (LACP_STATE_EQ(oldpstate, lp->lp_partner.lip_state,
LACP_STATE_TIMEOUT)) {
return;
}
LACP_DPRINTF((lp, "partner timeout changed\n"));
/*
* FAST_PERIODIC -> SLOW_PERIODIC
* or
* SLOW_PERIODIC (-> PERIODIC_TX) -> FAST_PERIODIC
*
* let lacp_sm_ptx_tx_schedule to update timeout.
*/
LACP_TIMER_DISARM(lp, LACP_TIMER_PERIODIC);
/*
* if timeout has been shortened, assert NTT.
*/
if ((lp->lp_partner.lip_state & LACP_STATE_TIMEOUT)) {
lacp_sm_assert_ntt(lp);
}
}
static void
lacp_sm_ptx_tx_schedule(struct lacp_port *lp)
{
int timeout;
if (!(lp->lp_state & LACP_STATE_ACTIVITY) &&
!(lp->lp_partner.lip_state & LACP_STATE_ACTIVITY)) {
/*
* NO_PERIODIC
*/
LACP_TIMER_DISARM(lp, LACP_TIMER_PERIODIC);
return;
}
if (LACP_TIMER_ISARMED(lp, LACP_TIMER_PERIODIC)) {
return;
}
timeout = (lp->lp_partner.lip_state & LACP_STATE_TIMEOUT) ?
LACP_FAST_PERIODIC_TIME : LACP_SLOW_PERIODIC_TIME;
LACP_TIMER_ARM(lp, LACP_TIMER_PERIODIC, timeout);
}
static void
lacp_sm_ptx_timer(struct lacp_port *lp)
{
lacp_sm_assert_ntt(lp);
}
static void
lacp_sm_rx(struct lacp_port *lp, const struct lacpdu *du)
{
int timeout;
/*
* check LACP_DISABLED first
*/
if (!(lp->lp_state & LACP_STATE_AGGREGATION)) {
return;
}
/*
* check loopback condition.
*/
if (!lacp_compare_systemid(&du->ldu_actor.lip_systemid,
&lp->lp_actor.lip_systemid)) {
return;
}
/*
* EXPIRED, DEFAULTED, CURRENT -> CURRENT
*/
lacp_sm_rx_update_selected(lp, du);
lacp_sm_rx_update_ntt(lp, du);
lacp_sm_rx_record_pdu(lp, du);
timeout = (lp->lp_state & LACP_STATE_TIMEOUT) ?
LACP_SHORT_TIMEOUT_TIME : LACP_LONG_TIMEOUT_TIME;
LACP_TIMER_ARM(lp, LACP_TIMER_CURRENT_WHILE, timeout);
lp->lp_state &= ~LACP_STATE_EXPIRED;
/*
* kick transmit machine without waiting the next tick.
*/
lacp_sm_tx(lp);
}
static void
lacp_sm_rx_set_expired(struct lacp_port *lp)
{
lp->lp_partner.lip_state &= ~LACP_STATE_SYNC;
lp->lp_partner.lip_state |= LACP_STATE_TIMEOUT;
LACP_TIMER_ARM(lp, LACP_TIMER_CURRENT_WHILE, LACP_SHORT_TIMEOUT_TIME);
lp->lp_state |= LACP_STATE_EXPIRED;
}
static void
lacp_sm_rx_timer(struct lacp_port *lp)
{
if ((lp->lp_state & LACP_STATE_EXPIRED) == 0) {
/* CURRENT -> EXPIRED */
LACP_DPRINTF((lp, "%s: CURRENT -> EXPIRED\n", __func__));
lacp_sm_rx_set_expired(lp);
} else {
/* EXPIRED -> DEFAULTED */
LACP_DPRINTF((lp, "%s: EXPIRED -> DEFAULTED\n", __func__));
lacp_sm_rx_update_default_selected(lp);
lacp_sm_rx_record_default(lp);
lp->lp_state &= ~LACP_STATE_EXPIRED;
}
}
static void
lacp_sm_rx_record_pdu(struct lacp_port *lp, const struct lacpdu *du)
{
boolean_t active;
uint8_t oldpstate;
char buf[LACP_STATESTR_MAX+1];
Turn LACP debugging from a compile time option to a sysctl, it is very handy to be able to turn it on when negotiation to a switch misbehaves. Submitted by: Andrew Boyer MFC after: 3 days
2012-05-26 08:09:01 +00:00
LACP_TRACE(lp);
Add the trunk(4) driver for providing link aggregation, failover and fault tolerance. This driver allows aggregation of multiple network interfaces as one virtual interface using a number of different protocols/algorithms. failover - Sends traffic through the secondary port if the master becomes inactive. fec - Supports Cisco Fast EtherChannel. lacp - Supports the IEEE 802.3ad Link Aggregation Control Protocol (LACP) and the Marker Protocol. loadbalance - Static loadbalancing using an outgoing hash. roundrobin - Distributes outgoing traffic using a round-robin scheduler through all active ports. This code was obtained from OpenBSD and this also includes 802.3ad LACP support from agr(4) in NetBSD.
2007-04-10 00:27:25 +00:00
oldpstate = lp->lp_partner.lip_state;
active = (du->ldu_actor.lip_state & LACP_STATE_ACTIVITY)
|| ((lp->lp_state & LACP_STATE_ACTIVITY) &&
(du->ldu_partner.lip_state & LACP_STATE_ACTIVITY));
lp->lp_partner = du->ldu_actor;
if (active &&
((LACP_STATE_EQ(lp->lp_state, du->ldu_partner.lip_state,
LACP_STATE_AGGREGATION) &&
!lacp_compare_peerinfo(&lp->lp_actor, &du->ldu_partner))
|| (du->ldu_partner.lip_state & LACP_STATE_AGGREGATION) == 0)) {
Do some minimal work to better conform to the 802.3ad (LACP) standard. In particular, don't set the synchronized bit for the peer unless it truly appears to be synchronized to us. Also, don't set our own synchronized bit unless we have actually seen a remote system. Prior to this change, we were seeing some strange behavior, such as: 1. We send an advertisement with the Activity, Aggregation, and Default flags, followed by an advertisement with the Activity, Aggregation, Synchronization, and Default flags. However, we hadn't seen an advertisement from another peer and were still advertising the default (NULL) peer. A closer examination of the in-kernel data structures (using kgdb) showed that the system had added the default (NULL) peer as a valid aggregator for the segment. 2. We were receiving an advertisement from a peer that included the default (NULL) peer instead of including our system information. However, we responded with an advertisement that included the Synchronization flag for both our system and the peer. (Since the peer's advertisement did not include our system information, we shouldn't add the synchronization bit for the peer.) This patch corrects those two items. Reviewed by: smh MFC after: 2 weeks Sponsored by: Netflix Differential Revision: https://reviews.freebsd.org/D9485
2017-02-26 00:19:02 +00:00
/*
* XXX Maintain legacy behavior of leaving the
* LACP_STATE_SYNC bit unchanged from the partner's
* advertisement if lsc_strict_mode is false.
* TODO: We should re-examine the concept of the "strict mode"
* to ensure it makes sense to maintain a non-strict mode.
*/
if (lp->lp_lsc->lsc_strict_mode)
lp->lp_partner.lip_state |= LACP_STATE_SYNC;
Add the trunk(4) driver for providing link aggregation, failover and fault tolerance. This driver allows aggregation of multiple network interfaces as one virtual interface using a number of different protocols/algorithms. failover - Sends traffic through the secondary port if the master becomes inactive. fec - Supports Cisco Fast EtherChannel. lacp - Supports the IEEE 802.3ad Link Aggregation Control Protocol (LACP) and the Marker Protocol. loadbalance - Static loadbalancing using an outgoing hash. roundrobin - Distributes outgoing traffic using a round-robin scheduler through all active ports. This code was obtained from OpenBSD and this also includes 802.3ad LACP support from agr(4) in NetBSD.
2007-04-10 00:27:25 +00:00
} else {
lp->lp_partner.lip_state &= ~LACP_STATE_SYNC;
}
lp->lp_state &= ~LACP_STATE_DEFAULTED;
if (oldpstate != lp->lp_partner.lip_state) {
LACP_DPRINTF((lp, "old pstate %s\n",
lacp_format_state(oldpstate, buf, sizeof(buf))));
LACP_DPRINTF((lp, "new pstate %s\n",
lacp_format_state(lp->lp_partner.lip_state, buf,
sizeof(buf))));
}
lacp_sm_ptx_update_timeout(lp, oldpstate);
}
static void
lacp_sm_rx_update_ntt(struct lacp_port *lp, const struct lacpdu *du)
{
Turn LACP debugging from a compile time option to a sysctl, it is very handy to be able to turn it on when negotiation to a switch misbehaves. Submitted by: Andrew Boyer MFC after: 3 days
2012-05-26 08:09:01 +00:00
LACP_TRACE(lp);
Add the trunk(4) driver for providing link aggregation, failover and fault tolerance. This driver allows aggregation of multiple network interfaces as one virtual interface using a number of different protocols/algorithms. failover - Sends traffic through the secondary port if the master becomes inactive. fec - Supports Cisco Fast EtherChannel. lacp - Supports the IEEE 802.3ad Link Aggregation Control Protocol (LACP) and the Marker Protocol. loadbalance - Static loadbalancing using an outgoing hash. roundrobin - Distributes outgoing traffic using a round-robin scheduler through all active ports. This code was obtained from OpenBSD and this also includes 802.3ad LACP support from agr(4) in NetBSD.
2007-04-10 00:27:25 +00:00
if (lacp_compare_peerinfo(&lp->lp_actor, &du->ldu_partner) ||
!LACP_STATE_EQ(lp->lp_state, du->ldu_partner.lip_state,
LACP_STATE_ACTIVITY | LACP_STATE_SYNC | LACP_STATE_AGGREGATION)) {
LACP_DPRINTF((lp, "%s: assert ntt\n", __func__));
lacp_sm_assert_ntt(lp);
}
}
static void
lacp_sm_rx_record_default(struct lacp_port *lp)
{
uint8_t oldpstate;
Turn LACP debugging from a compile time option to a sysctl, it is very handy to be able to turn it on when negotiation to a switch misbehaves. Submitted by: Andrew Boyer MFC after: 3 days
2012-05-26 08:09:01 +00:00
LACP_TRACE(lp);
Add the trunk(4) driver for providing link aggregation, failover and fault tolerance. This driver allows aggregation of multiple network interfaces as one virtual interface using a number of different protocols/algorithms. failover - Sends traffic through the secondary port if the master becomes inactive. fec - Supports Cisco Fast EtherChannel. lacp - Supports the IEEE 802.3ad Link Aggregation Control Protocol (LACP) and the Marker Protocol. loadbalance - Static loadbalancing using an outgoing hash. roundrobin - Distributes outgoing traffic using a round-robin scheduler through all active ports. This code was obtained from OpenBSD and this also includes 802.3ad LACP support from agr(4) in NetBSD.
2007-04-10 00:27:25 +00:00
oldpstate = lp->lp_partner.lip_state;
Break out the static, global LACP debug options into a per-lagg unit sysctl tree. * Create a net.link.lagg.X.lacp node * Add a debug node under that for tx_test and rx_test * Add lacp_strict_mode, defaulting to 1 tx_test and rx_test are still a bitmap of unit numbers for now. At some point it would be nice to create child nodes of the lagg bundle for each sub-interface, and then populate those with various knobs and statistics. Sponsored by: Netflix
2013-07-26 19:41:13 +00:00
if (lp->lp_lsc->lsc_strict_mode)
Bring over some link aggregation / LACP protocol improvements and debugging additions. * Add some new tracing events to aid in debugging. * Add in a debugging mode to drop transmit and received frames, specifically to test whether seeing or hearing heartbeats correctly cause LACP to drop the port. * Add in (and make default) a strict LACP mode, which requires the heartbeat on a port to be heard before it's used. Sometimes vendor ports will hang but the link layer stays up, resulting in hung traffic. * Add logging the number of link status flaps, again to aid in debugging badly behaving switch ports. * Calculate the lagg interface port speed as the multiple of the configured ports, rather than the largest. Obtained from: Netflix MFC after: 2 weeks
2013-07-13 04:25:03 +00:00
lp->lp_partner = lacp_partner_admin_strict;
else
Add sysctl to control LACP strict compliance default Add net.link.lagg.lacp.default_strict_mode which defines the default value for LACP strict compliance for created lagg devices. Also: * Add lacp_strict option to ifconfig(8). * Fix lagg(4) creation examples. * Minor style(9) fix. MFC after: 1 week
2015-11-06 15:33:27 +00:00
lp->lp_partner = lacp_partner_admin_optimistic;
Add the trunk(4) driver for providing link aggregation, failover and fault tolerance. This driver allows aggregation of multiple network interfaces as one virtual interface using a number of different protocols/algorithms. failover - Sends traffic through the secondary port if the master becomes inactive. fec - Supports Cisco Fast EtherChannel. lacp - Supports the IEEE 802.3ad Link Aggregation Control Protocol (LACP) and the Marker Protocol. loadbalance - Static loadbalancing using an outgoing hash. roundrobin - Distributes outgoing traffic using a round-robin scheduler through all active ports. This code was obtained from OpenBSD and this also includes 802.3ad LACP support from agr(4) in NetBSD.
2007-04-10 00:27:25 +00:00
lp->lp_state |= LACP_STATE_DEFAULTED;
lacp_sm_ptx_update_timeout(lp, oldpstate);
}
static void
lacp_sm_rx_update_selected_from_peerinfo(struct lacp_port *lp,
const struct lacp_peerinfo *info)
{
Turn LACP debugging from a compile time option to a sysctl, it is very handy to be able to turn it on when negotiation to a switch misbehaves. Submitted by: Andrew Boyer MFC after: 3 days
2012-05-26 08:09:01 +00:00
LACP_TRACE(lp);
Add the trunk(4) driver for providing link aggregation, failover and fault tolerance. This driver allows aggregation of multiple network interfaces as one virtual interface using a number of different protocols/algorithms. failover - Sends traffic through the secondary port if the master becomes inactive. fec - Supports Cisco Fast EtherChannel. lacp - Supports the IEEE 802.3ad Link Aggregation Control Protocol (LACP) and the Marker Protocol. loadbalance - Static loadbalancing using an outgoing hash. roundrobin - Distributes outgoing traffic using a round-robin scheduler through all active ports. This code was obtained from OpenBSD and this also includes 802.3ad LACP support from agr(4) in NetBSD.
2007-04-10 00:27:25 +00:00
if (lacp_compare_peerinfo(&lp->lp_partner, info) ||
!LACP_STATE_EQ(lp->lp_partner.lip_state, info->lip_state,
LACP_STATE_AGGREGATION)) {
lp->lp_selected = LACP_UNSELECTED;
/* mux machine will clean up lp->lp_aggregator */
}
}
static void
lacp_sm_rx_update_selected(struct lacp_port *lp, const struct lacpdu *du)
{
Turn LACP debugging from a compile time option to a sysctl, it is very handy to be able to turn it on when negotiation to a switch misbehaves. Submitted by: Andrew Boyer MFC after: 3 days
2012-05-26 08:09:01 +00:00
LACP_TRACE(lp);
Add the trunk(4) driver for providing link aggregation, failover and fault tolerance. This driver allows aggregation of multiple network interfaces as one virtual interface using a number of different protocols/algorithms. failover - Sends traffic through the secondary port if the master becomes inactive. fec - Supports Cisco Fast EtherChannel. lacp - Supports the IEEE 802.3ad Link Aggregation Control Protocol (LACP) and the Marker Protocol. loadbalance - Static loadbalancing using an outgoing hash. roundrobin - Distributes outgoing traffic using a round-robin scheduler through all active ports. This code was obtained from OpenBSD and this also includes 802.3ad LACP support from agr(4) in NetBSD.
2007-04-10 00:27:25 +00:00
lacp_sm_rx_update_selected_from_peerinfo(lp, &du->ldu_actor);
}
static void
lacp_sm_rx_update_default_selected(struct lacp_port *lp)
{
Turn LACP debugging from a compile time option to a sysctl, it is very handy to be able to turn it on when negotiation to a switch misbehaves. Submitted by: Andrew Boyer MFC after: 3 days
2012-05-26 08:09:01 +00:00
LACP_TRACE(lp);
Add the trunk(4) driver for providing link aggregation, failover and fault tolerance. This driver allows aggregation of multiple network interfaces as one virtual interface using a number of different protocols/algorithms. failover - Sends traffic through the secondary port if the master becomes inactive. fec - Supports Cisco Fast EtherChannel. lacp - Supports the IEEE 802.3ad Link Aggregation Control Protocol (LACP) and the Marker Protocol. loadbalance - Static loadbalancing using an outgoing hash. roundrobin - Distributes outgoing traffic using a round-robin scheduler through all active ports. This code was obtained from OpenBSD and this also includes 802.3ad LACP support from agr(4) in NetBSD.
2007-04-10 00:27:25 +00:00
Break out the static, global LACP debug options into a per-lagg unit sysctl tree. * Create a net.link.lagg.X.lacp node * Add a debug node under that for tx_test and rx_test * Add lacp_strict_mode, defaulting to 1 tx_test and rx_test are still a bitmap of unit numbers for now. At some point it would be nice to create child nodes of the lagg bundle for each sub-interface, and then populate those with various knobs and statistics. Sponsored by: Netflix
2013-07-26 19:41:13 +00:00
if (lp->lp_lsc->lsc_strict_mode)
Bring over some link aggregation / LACP protocol improvements and debugging additions. * Add some new tracing events to aid in debugging. * Add in a debugging mode to drop transmit and received frames, specifically to test whether seeing or hearing heartbeats correctly cause LACP to drop the port. * Add in (and make default) a strict LACP mode, which requires the heartbeat on a port to be heard before it's used. Sometimes vendor ports will hang but the link layer stays up, resulting in hung traffic. * Add logging the number of link status flaps, again to aid in debugging badly behaving switch ports. * Calculate the lagg interface port speed as the multiple of the configured ports, rather than the largest. Obtained from: Netflix MFC after: 2 weeks
2013-07-13 04:25:03 +00:00
lacp_sm_rx_update_selected_from_peerinfo(lp,
&lacp_partner_admin_strict);
else
lacp_sm_rx_update_selected_from_peerinfo(lp,
&lacp_partner_admin_optimistic);
Add the trunk(4) driver for providing link aggregation, failover and fault tolerance. This driver allows aggregation of multiple network interfaces as one virtual interface using a number of different protocols/algorithms. failover - Sends traffic through the secondary port if the master becomes inactive. fec - Supports Cisco Fast EtherChannel. lacp - Supports the IEEE 802.3ad Link Aggregation Control Protocol (LACP) and the Marker Protocol. loadbalance - Static loadbalancing using an outgoing hash. roundrobin - Distributes outgoing traffic using a round-robin scheduler through all active ports. This code was obtained from OpenBSD and this also includes 802.3ad LACP support from agr(4) in NetBSD.
2007-04-10 00:27:25 +00:00
}
/* transmit machine */
static void
lacp_sm_tx(struct lacp_port *lp)
{
Bring over some link aggregation / LACP protocol improvements and debugging additions. * Add some new tracing events to aid in debugging. * Add in a debugging mode to drop transmit and received frames, specifically to test whether seeing or hearing heartbeats correctly cause LACP to drop the port. * Add in (and make default) a strict LACP mode, which requires the heartbeat on a port to be heard before it's used. Sometimes vendor ports will hang but the link layer stays up, resulting in hung traffic. * Add logging the number of link status flaps, again to aid in debugging badly behaving switch ports. * Calculate the lagg interface port speed as the multiple of the configured ports, rather than the largest. Obtained from: Netflix MFC after: 2 weeks
2013-07-13 04:25:03 +00:00
int error = 0;
Add the trunk(4) driver for providing link aggregation, failover and fault tolerance. This driver allows aggregation of multiple network interfaces as one virtual interface using a number of different protocols/algorithms. failover - Sends traffic through the secondary port if the master becomes inactive. fec - Supports Cisco Fast EtherChannel. lacp - Supports the IEEE 802.3ad Link Aggregation Control Protocol (LACP) and the Marker Protocol. loadbalance - Static loadbalancing using an outgoing hash. roundrobin - Distributes outgoing traffic using a round-robin scheduler through all active ports. This code was obtained from OpenBSD and this also includes 802.3ad LACP support from agr(4) in NetBSD.
2007-04-10 00:27:25 +00:00
if (!(lp->lp_state & LACP_STATE_AGGREGATION)
#if 1
|| (!(lp->lp_state & LACP_STATE_ACTIVITY)
&& !(lp->lp_partner.lip_state & LACP_STATE_ACTIVITY))
#endif
) {
lp->lp_flags &= ~LACP_PORT_NTT;
}
if (!(lp->lp_flags & LACP_PORT_NTT)) {
return;
}
/* Rate limit to 3 PDUs per LACP_FAST_PERIODIC_TIME */
if (ppsratecheck(&lp->lp_last_lacpdu, &lp->lp_lacpdu_sent,
(3 / LACP_FAST_PERIODIC_TIME)) == 0) {
LACP_DPRINTF((lp, "rate limited pdu\n"));
return;
}
Break out the static, global LACP debug options into a per-lagg unit sysctl tree. * Create a net.link.lagg.X.lacp node * Add a debug node under that for tx_test and rx_test * Add lacp_strict_mode, defaulting to 1 tx_test and rx_test are still a bitmap of unit numbers for now. At some point it would be nice to create child nodes of the lagg bundle for each sub-interface, and then populate those with various knobs and statistics. Sponsored by: Netflix
2013-07-26 19:41:13 +00:00
if (((1 << lp->lp_ifp->if_dunit) & lp->lp_lsc->lsc_debug.lsc_tx_test) == 0) {
Bring over some link aggregation / LACP protocol improvements and debugging additions. * Add some new tracing events to aid in debugging. * Add in a debugging mode to drop transmit and received frames, specifically to test whether seeing or hearing heartbeats correctly cause LACP to drop the port. * Add in (and make default) a strict LACP mode, which requires the heartbeat on a port to be heard before it's used. Sometimes vendor ports will hang but the link layer stays up, resulting in hung traffic. * Add logging the number of link status flaps, again to aid in debugging badly behaving switch ports. * Calculate the lagg interface port speed as the multiple of the configured ports, rather than the largest. Obtained from: Netflix MFC after: 2 weeks
2013-07-13 04:25:03 +00:00
error = lacp_xmit_lacpdu(lp);
Break out the static, global LACP debug options into a per-lagg unit sysctl tree. * Create a net.link.lagg.X.lacp node * Add a debug node under that for tx_test and rx_test * Add lacp_strict_mode, defaulting to 1 tx_test and rx_test are still a bitmap of unit numbers for now. At some point it would be nice to create child nodes of the lagg bundle for each sub-interface, and then populate those with various knobs and statistics. Sponsored by: Netflix
2013-07-26 19:41:13 +00:00
} else {
Bring over some link aggregation / LACP protocol improvements and debugging additions. * Add some new tracing events to aid in debugging. * Add in a debugging mode to drop transmit and received frames, specifically to test whether seeing or hearing heartbeats correctly cause LACP to drop the port. * Add in (and make default) a strict LACP mode, which requires the heartbeat on a port to be heard before it's used. Sometimes vendor ports will hang but the link layer stays up, resulting in hung traffic. * Add logging the number of link status flaps, again to aid in debugging badly behaving switch ports. * Calculate the lagg interface port speed as the multiple of the configured ports, rather than the largest. Obtained from: Netflix MFC after: 2 weeks
2013-07-13 04:25:03 +00:00
LACP_TPRINTF((lp, "Dropping TX PDU\n"));
Break out the static, global LACP debug options into a per-lagg unit sysctl tree. * Create a net.link.lagg.X.lacp node * Add a debug node under that for tx_test and rx_test * Add lacp_strict_mode, defaulting to 1 tx_test and rx_test are still a bitmap of unit numbers for now. At some point it would be nice to create child nodes of the lagg bundle for each sub-interface, and then populate those with various knobs and statistics. Sponsored by: Netflix
2013-07-26 19:41:13 +00:00
}
Add the trunk(4) driver for providing link aggregation, failover and fault tolerance. This driver allows aggregation of multiple network interfaces as one virtual interface using a number of different protocols/algorithms. failover - Sends traffic through the secondary port if the master becomes inactive. fec - Supports Cisco Fast EtherChannel. lacp - Supports the IEEE 802.3ad Link Aggregation Control Protocol (LACP) and the Marker Protocol. loadbalance - Static loadbalancing using an outgoing hash. roundrobin - Distributes outgoing traffic using a round-robin scheduler through all active ports. This code was obtained from OpenBSD and this also includes 802.3ad LACP support from agr(4) in NetBSD.
2007-04-10 00:27:25 +00:00
if (error == 0) {
lp->lp_flags &= ~LACP_PORT_NTT;
} else {
LACP_DPRINTF((lp, "lacpdu transmit failure, error %d\n",
error));
}
}
static void
lacp_sm_assert_ntt(struct lacp_port *lp)
{
lp->lp_flags |= LACP_PORT_NTT;
}
static void
lacp_run_timers(struct lacp_port *lp)
{
int i;
for (i = 0; i < LACP_NTIMER; i++) {
KASSERT(lp->lp_timer[i] >= 0,
("invalid timer value %d", lp->lp_timer[i]));
if (lp->lp_timer[i] == 0) {
continue;
} else if (--lp->lp_timer[i] <= 0) {
if (lacp_timer_funcs[i]) {
(*lacp_timer_funcs[i])(lp);
}
}
}
}
int
Switch the LACP state machine over to its own mutex to protect the internals, this means that it no longer grabs the lagg rwlock. Use two port table arrays which list the active ports for Tx and switch between them with an atomic op. Now the lagg rwlock is only exclusively locked for management (ioctls) and queuing of lacp control frames isnt needed.
2008-03-16 19:25:30 +00:00
lacp_marker_input(struct lacp_port *lp, struct mbuf *m)
Add the trunk(4) driver for providing link aggregation, failover and fault tolerance. This driver allows aggregation of multiple network interfaces as one virtual interface using a number of different protocols/algorithms. failover - Sends traffic through the secondary port if the master becomes inactive. fec - Supports Cisco Fast EtherChannel. lacp - Supports the IEEE 802.3ad Link Aggregation Control Protocol (LACP) and the Marker Protocol. loadbalance - Static loadbalancing using an outgoing hash. roundrobin - Distributes outgoing traffic using a round-robin scheduler through all active ports. This code was obtained from OpenBSD and this also includes 802.3ad LACP support from agr(4) in NetBSD.
2007-04-10 00:27:25 +00:00
{
Implement the Marker Protocol. A marker frame is placed on the interface queue of each port and any further packets are blocked, when the all the marker frames have been returned to us from the remote network device then we can be sure that all interface queues are empty. This is needed when a port is added or removed from the aggregation since it will affect the hash based distribution, if the queues are not empty then a packet from an existing connection may be placed on a different interface and arrive out of order. This was previously achieved by suppressing transmission for 1 second, now that there is an active feedback this timeout as been increased to 3 seconds and used as a fallback.
2007-05-19 07:47:04 +00:00
struct lacp_softc *lsc = lp->lp_lsc;
Switch the LACP state machine over to its own mutex to protect the internals, this means that it no longer grabs the lagg rwlock. Use two port table arrays which list the active ports for Tx and switch between them with an atomic op. Now the lagg rwlock is only exclusively locked for management (ioctls) and queuing of lacp control frames isnt needed.
2008-03-16 19:25:30 +00:00
struct lagg_port *lgp = lp->lp_lagg;
struct lacp_port *lp2;
Add the trunk(4) driver for providing link aggregation, failover and fault tolerance. This driver allows aggregation of multiple network interfaces as one virtual interface using a number of different protocols/algorithms. failover - Sends traffic through the secondary port if the master becomes inactive. fec - Supports Cisco Fast EtherChannel. lacp - Supports the IEEE 802.3ad Link Aggregation Control Protocol (LACP) and the Marker Protocol. loadbalance - Static loadbalancing using an outgoing hash. roundrobin - Distributes outgoing traffic using a round-robin scheduler through all active ports. This code was obtained from OpenBSD and this also includes 802.3ad LACP support from agr(4) in NetBSD.
2007-04-10 00:27:25 +00:00
struct markerdu *mdu;
int error = 0;
Implement the Marker Protocol. A marker frame is placed on the interface queue of each port and any further packets are blocked, when the all the marker frames have been returned to us from the remote network device then we can be sure that all interface queues are empty. This is needed when a port is added or removed from the aggregation since it will affect the hash based distribution, if the queues are not empty then a packet from an existing connection may be placed on a different interface and arrive out of order. This was previously achieved by suppressing transmission for 1 second, now that there is an active feedback this timeout as been increased to 3 seconds and used as a fallback.
2007-05-19 07:47:04 +00:00
int pending = 0;
Add the trunk(4) driver for providing link aggregation, failover and fault tolerance. This driver allows aggregation of multiple network interfaces as one virtual interface using a number of different protocols/algorithms. failover - Sends traffic through the secondary port if the master becomes inactive. fec - Supports Cisco Fast EtherChannel. lacp - Supports the IEEE 802.3ad Link Aggregation Control Protocol (LACP) and the Marker Protocol. loadbalance - Static loadbalancing using an outgoing hash. roundrobin - Distributes outgoing traffic using a round-robin scheduler through all active ports. This code was obtained from OpenBSD and this also includes 802.3ad LACP support from agr(4) in NetBSD.
2007-04-10 00:27:25 +00:00
if (m->m_pkthdr.len != sizeof(*mdu)) {
goto bad;
}
if ((m->m_flags & M_MCAST) == 0) {
goto bad;
}
if (m->m_len < sizeof(*mdu)) {
m = m_pullup(m, sizeof(*mdu));
if (m == NULL) {
return (ENOMEM);
}
}
mdu = mtod(m, struct markerdu *);
if (memcmp(&mdu->mdu_eh.ether_dhost,
&ethermulticastaddr_slowprotocols, ETHER_ADDR_LEN)) {
goto bad;
}
if (mdu->mdu_sph.sph_version != 1) {
goto bad;
}
switch (mdu->mdu_tlv.tlv_type) {
case MARKER_TYPE_INFO:
if (tlv_check(mdu, sizeof(*mdu), &mdu->mdu_tlv,
marker_info_tlv_template, TRUE)) {
goto bad;
}
mdu->mdu_tlv.tlv_type = MARKER_TYPE_RESPONSE;
memcpy(&mdu->mdu_eh.ether_dhost,
&ethermulticastaddr_slowprotocols, ETHER_ADDR_LEN);
memcpy(&mdu->mdu_eh.ether_shost,
Rename the trunk(4) driver to lagg(4) as it is too similar to vlan trunking. The name trunk is misused as the networking term trunk means carrying multiple VLANs over a single connection. The IEEE standard for link aggregation (802.3 section 3) does not talk about 'trunk' at all while it is used throughout IEEE 802.1Q in describing vlans. The lagg(4) driver provides link aggregation, failover and fault tolerance. Discussed on: current@
2007-04-17 00:35:11 +00:00
lgp->lp_lladdr, ETHER_ADDR_LEN);
error = lagg_enqueue(lp->lp_ifp, m);
Add the trunk(4) driver for providing link aggregation, failover and fault tolerance. This driver allows aggregation of multiple network interfaces as one virtual interface using a number of different protocols/algorithms. failover - Sends traffic through the secondary port if the master becomes inactive. fec - Supports Cisco Fast EtherChannel. lacp - Supports the IEEE 802.3ad Link Aggregation Control Protocol (LACP) and the Marker Protocol. loadbalance - Static loadbalancing using an outgoing hash. roundrobin - Distributes outgoing traffic using a round-robin scheduler through all active ports. This code was obtained from OpenBSD and this also includes 802.3ad LACP support from agr(4) in NetBSD.
2007-04-10 00:27:25 +00:00
break;
case MARKER_TYPE_RESPONSE:
if (tlv_check(mdu, sizeof(*mdu), &mdu->mdu_tlv,
marker_response_tlv_template, TRUE)) {
goto bad;
}
Implement the Marker Protocol. A marker frame is placed on the interface queue of each port and any further packets are blocked, when the all the marker frames have been returned to us from the remote network device then we can be sure that all interface queues are empty. This is needed when a port is added or removed from the aggregation since it will affect the hash based distribution, if the queues are not empty then a packet from an existing connection may be placed on a different interface and arrive out of order. This was previously achieved by suppressing transmission for 1 second, now that there is an active feedback this timeout as been increased to 3 seconds and used as a fallback.
2007-05-19 07:47:04 +00:00
LACP_DPRINTF((lp, "marker response, port=%u, sys=%6D, id=%u\n",
ntohs(mdu->mdu_info.mi_rq_port), mdu->mdu_info.mi_rq_system,
":", ntohl(mdu->mdu_info.mi_rq_xid)));
/* Verify that it is the last marker we sent out */
if (memcmp(&mdu->mdu_info, &lp->lp_marker,
sizeof(struct lacp_markerinfo)))
goto bad;
Switch the LACP state machine over to its own mutex to protect the internals, this means that it no longer grabs the lagg rwlock. Use two port table arrays which list the active ports for Tx and switch between them with an atomic op. Now the lagg rwlock is only exclusively locked for management (ioctls) and queuing of lacp control frames isnt needed.
2008-03-16 19:25:30 +00:00
LACP_LOCK(lsc);
Implement the Marker Protocol. A marker frame is placed on the interface queue of each port and any further packets are blocked, when the all the marker frames have been returned to us from the remote network device then we can be sure that all interface queues are empty. This is needed when a port is added or removed from the aggregation since it will affect the hash based distribution, if the queues are not empty then a packet from an existing connection may be placed on a different interface and arrive out of order. This was previously achieved by suppressing transmission for 1 second, now that there is an active feedback this timeout as been increased to 3 seconds and used as a fallback.
2007-05-19 07:47:04 +00:00
lp->lp_flags &= ~LACP_PORT_MARK;
if (lsc->lsc_suppress_distributing) {
/* Check if any ports are waiting for a response */
LIST_FOREACH(lp2, &lsc->lsc_ports, lp_next) {
if (lp2->lp_flags & LACP_PORT_MARK) {
pending = 1;
break;
}
}
if (pending == 0) {
/* All interface queues are clear */
LACP_DPRINTF((NULL, "queue flush complete\n"));
lsc->lsc_suppress_distributing = FALSE;
}
}
Switch the LACP state machine over to its own mutex to protect the internals, this means that it no longer grabs the lagg rwlock. Use two port table arrays which list the active ports for Tx and switch between them with an atomic op. Now the lagg rwlock is only exclusively locked for management (ioctls) and queuing of lacp control frames isnt needed.
2008-03-16 19:25:30 +00:00
LACP_UNLOCK(lsc);
Implement the Marker Protocol. A marker frame is placed on the interface queue of each port and any further packets are blocked, when the all the marker frames have been returned to us from the remote network device then we can be sure that all interface queues are empty. This is needed when a port is added or removed from the aggregation since it will affect the hash based distribution, if the queues are not empty then a packet from an existing connection may be placed on a different interface and arrive out of order. This was previously achieved by suppressing transmission for 1 second, now that there is an active feedback this timeout as been increased to 3 seconds and used as a fallback.
2007-05-19 07:47:04 +00:00
m_freem(m);
break;
Add the trunk(4) driver for providing link aggregation, failover and fault tolerance. This driver allows aggregation of multiple network interfaces as one virtual interface using a number of different protocols/algorithms. failover - Sends traffic through the secondary port if the master becomes inactive. fec - Supports Cisco Fast EtherChannel. lacp - Supports the IEEE 802.3ad Link Aggregation Control Protocol (LACP) and the Marker Protocol. loadbalance - Static loadbalancing using an outgoing hash. roundrobin - Distributes outgoing traffic using a round-robin scheduler through all active ports. This code was obtained from OpenBSD and this also includes 802.3ad LACP support from agr(4) in NetBSD.
2007-04-10 00:27:25 +00:00
default:
goto bad;
}
return (error);
bad:
Implement the Marker Protocol. A marker frame is placed on the interface queue of each port and any further packets are blocked, when the all the marker frames have been returned to us from the remote network device then we can be sure that all interface queues are empty. This is needed when a port is added or removed from the aggregation since it will affect the hash based distribution, if the queues are not empty then a packet from an existing connection may be placed on a different interface and arrive out of order. This was previously achieved by suppressing transmission for 1 second, now that there is an active feedback this timeout as been increased to 3 seconds and used as a fallback.
2007-05-19 07:47:04 +00:00
LACP_DPRINTF((lp, "bad marker frame\n"));
Add the trunk(4) driver for providing link aggregation, failover and fault tolerance. This driver allows aggregation of multiple network interfaces as one virtual interface using a number of different protocols/algorithms. failover - Sends traffic through the secondary port if the master becomes inactive. fec - Supports Cisco Fast EtherChannel. lacp - Supports the IEEE 802.3ad Link Aggregation Control Protocol (LACP) and the Marker Protocol. loadbalance - Static loadbalancing using an outgoing hash. roundrobin - Distributes outgoing traffic using a round-robin scheduler through all active ports. This code was obtained from OpenBSD and this also includes 802.3ad LACP support from agr(4) in NetBSD.
2007-04-10 00:27:25 +00:00
m_freem(m);
return (EINVAL);
}
static int
tlv_check(const void *p, size_t size, const struct tlvhdr *tlv,
const struct tlv_template *tmpl, boolean_t check_type)
{
while (/* CONSTCOND */ 1) {
if ((const char *)tlv - (const char *)p + sizeof(*tlv) > size) {
return (EINVAL);
}
if ((check_type && tlv->tlv_type != tmpl->tmpl_type) ||
tlv->tlv_length != tmpl->tmpl_length) {
return (EINVAL);
}
if (tmpl->tmpl_type == 0) {
break;
}
tlv = (const struct tlvhdr *)
((const char *)tlv + tlv->tlv_length);
tmpl++;
}
return (0);
}
Turn LACP debugging from a compile time option to a sysctl, it is very handy to be able to turn it on when negotiation to a switch misbehaves. Submitted by: Andrew Boyer MFC after: 3 days
2012-05-26 08:09:01 +00:00
/* Debugging */
Add the trunk(4) driver for providing link aggregation, failover and fault tolerance. This driver allows aggregation of multiple network interfaces as one virtual interface using a number of different protocols/algorithms. failover - Sends traffic through the secondary port if the master becomes inactive. fec - Supports Cisco Fast EtherChannel. lacp - Supports the IEEE 802.3ad Link Aggregation Control Protocol (LACP) and the Marker Protocol. loadbalance - Static loadbalancing using an outgoing hash. roundrobin - Distributes outgoing traffic using a round-robin scheduler through all active ports. This code was obtained from OpenBSD and this also includes 802.3ad LACP support from agr(4) in NetBSD.
2007-04-10 00:27:25 +00:00
const char *
lacp_format_mac(const uint8_t *mac, char *buf, size_t buflen)
{
snprintf(buf, buflen, "%02X-%02X-%02X-%02X-%02X-%02X",
(int)mac[0],
(int)mac[1],
(int)mac[2],
(int)mac[3],
(int)mac[4],
(int)mac[5]);
return (buf);
}
const char *
lacp_format_systemid(const struct lacp_systemid *sysid,
char *buf, size_t buflen)
{
char macbuf[LACP_MACSTR_MAX+1];
snprintf(buf, buflen, "%04X,%s",
ntohs(sysid->lsi_prio),
lacp_format_mac(sysid->lsi_mac, macbuf, sizeof(macbuf)));
return (buf);
}
const char *
lacp_format_portid(const struct lacp_portid *portid, char *buf, size_t buflen)
{
snprintf(buf, buflen, "%04X,%04X",
ntohs(portid->lpi_prio),
ntohs(portid->lpi_portno));
return (buf);
}
const char *
lacp_format_partner(const struct lacp_peerinfo *peer, char *buf, size_t buflen)
{
char sysid[LACP_SYSTEMIDSTR_MAX+1];
char portid[LACP_PORTIDSTR_MAX+1];
snprintf(buf, buflen, "(%s,%04X,%s)",
lacp_format_systemid(&peer->lip_systemid, sysid, sizeof(sysid)),
ntohs(peer->lip_key),
lacp_format_portid(&peer->lip_portid, portid, sizeof(portid)));
return (buf);
}
const char *
lacp_format_lagid(const struct lacp_peerinfo *a,
const struct lacp_peerinfo *b, char *buf, size_t buflen)
{
char astr[LACP_PARTNERSTR_MAX+1];
char bstr[LACP_PARTNERSTR_MAX+1];
#if 0
/*
* there's a convention to display small numbered peer
* in the left.
*/
if (lacp_compare_peerinfo(a, b) > 0) {
const struct lacp_peerinfo *t;
t = a;
a = b;
b = t;
}
#endif
snprintf(buf, buflen, "[%s,%s]",
lacp_format_partner(a, astr, sizeof(astr)),
lacp_format_partner(b, bstr, sizeof(bstr)));
return (buf);
}
const char *
lacp_format_lagid_aggregator(const struct lacp_aggregator *la,
char *buf, size_t buflen)
{
if (la == NULL) {
return ("(none)");
}
return (lacp_format_lagid(&la->la_actor, &la->la_partner, buf, buflen));
}
const char *
lacp_format_state(uint8_t state, char *buf, size_t buflen)
{
snprintf(buf, buflen, "%b", state, LACP_STATE_BITS);
return (buf);
}
static void
lacp_dump_lacpdu(const struct lacpdu *du)
{
char buf[LACP_PARTNERSTR_MAX+1];
char buf2[LACP_STATESTR_MAX+1];
printf("actor=%s\n",
lacp_format_partner(&du->ldu_actor, buf, sizeof(buf)));
printf("actor.state=%s\n",
lacp_format_state(du->ldu_actor.lip_state, buf2, sizeof(buf2)));
printf("partner=%s\n",
lacp_format_partner(&du->ldu_partner, buf, sizeof(buf)));
printf("partner.state=%s\n",
lacp_format_state(du->ldu_partner.lip_state, buf2, sizeof(buf2)));
printf("maxdelay=%d\n", ntohs(du->ldu_collector.lci_maxdelay));
}
static void
lacp_dprintf(const struct lacp_port *lp, const char *fmt, ...)
{
va_list va;
if (lp) {
printf("%s: ", lp->lp_ifp->if_xname);
}
va_start(va, fmt);
vprintf(fmt, va);
va_end(va);
}
Reference in New Issue Copy Permalink