netlink: add netlink support
Netlinks is a communication protocol currently used in Linux kernel to modify,
read and subscribe for nearly all networking state. Interfaces, addresses, routes,
firewall, fibs, vnets, etc are controlled via netlink.
It is async, TLV-based protocol, providing 1-1 and 1-many communications.
The current implementation supports the subset of NETLINK_ROUTE
family. To be more specific, the following is supported:
* Dumps:
- routes
- nexthops / nexthop groups
- interfaces
- interface addresses
- neighbors (arp/ndp)
* Notifications:
- interface arrival/departure
- interface address arrival/departure
- route addition/deletion
* Modifications:
- adding/deleting routes
- adding/deleting nexthops/nexthops groups
- adding/deleting neghbors
- adding/deleting interfaces (basic support only)
* Rtsock interaction
- route events are bridged both ways
The implementation also supports the NETLINK_GENERIC family framework.
Implementation notes:
Netlink is implemented via loadable/unloadable kernel module,
not touching many kernel parts.
Each netlink socket uses dedicated taskqueue to support async operations
that can sleep, such as interface creation. All message processing is
performed within these taskqueues.
Compatibility:
Most of the Netlink data models specified above maps to FreeBSD concepts
nicely. Unmodified ip(8) binary correctly works with
interfaces, addresses, routes, nexthops and nexthop groups. Some
software such as net/bird require header-only modifications to compile
and work with FreeBSD netlink.
Reviewed by: imp
Differential Revision: https://reviews.freebsd.org/D36002
MFC after: 2 months
2022-01-20 21:39:21 +00:00
|
|
|
/*-
|
2023-05-10 15:40:58 +00:00
|
|
|
* SPDX-License-Identifier: BSD-2-Clause
|
netlink: add netlink support
Netlinks is a communication protocol currently used in Linux kernel to modify,
read and subscribe for nearly all networking state. Interfaces, addresses, routes,
firewall, fibs, vnets, etc are controlled via netlink.
It is async, TLV-based protocol, providing 1-1 and 1-many communications.
The current implementation supports the subset of NETLINK_ROUTE
family. To be more specific, the following is supported:
* Dumps:
- routes
- nexthops / nexthop groups
- interfaces
- interface addresses
- neighbors (arp/ndp)
* Notifications:
- interface arrival/departure
- interface address arrival/departure
- route addition/deletion
* Modifications:
- adding/deleting routes
- adding/deleting nexthops/nexthops groups
- adding/deleting neghbors
- adding/deleting interfaces (basic support only)
* Rtsock interaction
- route events are bridged both ways
The implementation also supports the NETLINK_GENERIC family framework.
Implementation notes:
Netlink is implemented via loadable/unloadable kernel module,
not touching many kernel parts.
Each netlink socket uses dedicated taskqueue to support async operations
that can sleep, such as interface creation. All message processing is
performed within these taskqueues.
Compatibility:
Most of the Netlink data models specified above maps to FreeBSD concepts
nicely. Unmodified ip(8) binary correctly works with
interfaces, addresses, routes, nexthops and nexthop groups. Some
software such as net/bird require header-only modifications to compile
and work with FreeBSD netlink.
Reviewed by: imp
Differential Revision: https://reviews.freebsd.org/D36002
MFC after: 2 months
2022-01-20 21:39:21 +00:00
|
|
|
*
|
|
|
|
* Copyright (c) 2021 Ng Peng Nam Sean
|
|
|
|
* Copyright (c) 2022 Alexander V. Chernikov <melifaro@FreeBSD.org>
|
|
|
|
*
|
|
|
|
* 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/param.h>
|
2022-10-02 01:38:55 +00:00
|
|
|
#include <sys/ck.h>
|
netlink: add netlink support
Netlinks is a communication protocol currently used in Linux kernel to modify,
read and subscribe for nearly all networking state. Interfaces, addresses, routes,
firewall, fibs, vnets, etc are controlled via netlink.
It is async, TLV-based protocol, providing 1-1 and 1-many communications.
The current implementation supports the subset of NETLINK_ROUTE
family. To be more specific, the following is supported:
* Dumps:
- routes
- nexthops / nexthop groups
- interfaces
- interface addresses
- neighbors (arp/ndp)
* Notifications:
- interface arrival/departure
- interface address arrival/departure
- route addition/deletion
* Modifications:
- adding/deleting routes
- adding/deleting nexthops/nexthops groups
- adding/deleting neghbors
- adding/deleting interfaces (basic support only)
* Rtsock interaction
- route events are bridged both ways
The implementation also supports the NETLINK_GENERIC family framework.
Implementation notes:
Netlink is implemented via loadable/unloadable kernel module,
not touching many kernel parts.
Each netlink socket uses dedicated taskqueue to support async operations
that can sleep, such as interface creation. All message processing is
performed within these taskqueues.
Compatibility:
Most of the Netlink data models specified above maps to FreeBSD concepts
nicely. Unmodified ip(8) binary correctly works with
interfaces, addresses, routes, nexthops and nexthop groups. Some
software such as net/bird require header-only modifications to compile
and work with FreeBSD netlink.
Reviewed by: imp
Differential Revision: https://reviews.freebsd.org/D36002
MFC after: 2 months
2022-01-20 21:39:21 +00:00
|
|
|
#include <sys/lock.h>
|
2022-10-02 01:38:55 +00:00
|
|
|
#include <sys/malloc.h>
|
netlink: add netlink support
Netlinks is a communication protocol currently used in Linux kernel to modify,
read and subscribe for nearly all networking state. Interfaces, addresses, routes,
firewall, fibs, vnets, etc are controlled via netlink.
It is async, TLV-based protocol, providing 1-1 and 1-many communications.
The current implementation supports the subset of NETLINK_ROUTE
family. To be more specific, the following is supported:
* Dumps:
- routes
- nexthops / nexthop groups
- interfaces
- interface addresses
- neighbors (arp/ndp)
* Notifications:
- interface arrival/departure
- interface address arrival/departure
- route addition/deletion
* Modifications:
- adding/deleting routes
- adding/deleting nexthops/nexthops groups
- adding/deleting neghbors
- adding/deleting interfaces (basic support only)
* Rtsock interaction
- route events are bridged both ways
The implementation also supports the NETLINK_GENERIC family framework.
Implementation notes:
Netlink is implemented via loadable/unloadable kernel module,
not touching many kernel parts.
Each netlink socket uses dedicated taskqueue to support async operations
that can sleep, such as interface creation. All message processing is
performed within these taskqueues.
Compatibility:
Most of the Netlink data models specified above maps to FreeBSD concepts
nicely. Unmodified ip(8) binary correctly works with
interfaces, addresses, routes, nexthops and nexthop groups. Some
software such as net/bird require header-only modifications to compile
and work with FreeBSD netlink.
Reviewed by: imp
Differential Revision: https://reviews.freebsd.org/D36002
MFC after: 2 months
2022-01-20 21:39:21 +00:00
|
|
|
#include <sys/mbuf.h>
|
2022-10-02 01:38:55 +00:00
|
|
|
#include <sys/mutex.h>
|
netlink: add netlink support
Netlinks is a communication protocol currently used in Linux kernel to modify,
read and subscribe for nearly all networking state. Interfaces, addresses, routes,
firewall, fibs, vnets, etc are controlled via netlink.
It is async, TLV-based protocol, providing 1-1 and 1-many communications.
The current implementation supports the subset of NETLINK_ROUTE
family. To be more specific, the following is supported:
* Dumps:
- routes
- nexthops / nexthop groups
- interfaces
- interface addresses
- neighbors (arp/ndp)
* Notifications:
- interface arrival/departure
- interface address arrival/departure
- route addition/deletion
* Modifications:
- adding/deleting routes
- adding/deleting nexthops/nexthops groups
- adding/deleting neghbors
- adding/deleting interfaces (basic support only)
* Rtsock interaction
- route events are bridged both ways
The implementation also supports the NETLINK_GENERIC family framework.
Implementation notes:
Netlink is implemented via loadable/unloadable kernel module,
not touching many kernel parts.
Each netlink socket uses dedicated taskqueue to support async operations
that can sleep, such as interface creation. All message processing is
performed within these taskqueues.
Compatibility:
Most of the Netlink data models specified above maps to FreeBSD concepts
nicely. Unmodified ip(8) binary correctly works with
interfaces, addresses, routes, nexthops and nexthop groups. Some
software such as net/bird require header-only modifications to compile
and work with FreeBSD netlink.
Reviewed by: imp
Differential Revision: https://reviews.freebsd.org/D36002
MFC after: 2 months
2022-01-20 21:39:21 +00:00
|
|
|
#include <sys/socket.h>
|
|
|
|
#include <sys/socketvar.h>
|
|
|
|
#include <sys/syslog.h>
|
|
|
|
|
|
|
|
#include <netlink/netlink.h>
|
|
|
|
#include <netlink/netlink_ctl.h>
|
|
|
|
#include <netlink/netlink_linux.h>
|
|
|
|
#include <netlink/netlink_var.h>
|
|
|
|
|
|
|
|
#define DEBUG_MOD_NAME nl_io
|
|
|
|
#define DEBUG_MAX_LEVEL LOG_DEBUG3
|
|
|
|
#include <netlink/netlink_debug.h>
|
2023-05-11 13:08:42 +00:00
|
|
|
_DECLARE_DEBUG(LOG_INFO);
|
netlink: add netlink support
Netlinks is a communication protocol currently used in Linux kernel to modify,
read and subscribe for nearly all networking state. Interfaces, addresses, routes,
firewall, fibs, vnets, etc are controlled via netlink.
It is async, TLV-based protocol, providing 1-1 and 1-many communications.
The current implementation supports the subset of NETLINK_ROUTE
family. To be more specific, the following is supported:
* Dumps:
- routes
- nexthops / nexthop groups
- interfaces
- interface addresses
- neighbors (arp/ndp)
* Notifications:
- interface arrival/departure
- interface address arrival/departure
- route addition/deletion
* Modifications:
- adding/deleting routes
- adding/deleting nexthops/nexthops groups
- adding/deleting neghbors
- adding/deleting interfaces (basic support only)
* Rtsock interaction
- route events are bridged both ways
The implementation also supports the NETLINK_GENERIC family framework.
Implementation notes:
Netlink is implemented via loadable/unloadable kernel module,
not touching many kernel parts.
Each netlink socket uses dedicated taskqueue to support async operations
that can sleep, such as interface creation. All message processing is
performed within these taskqueues.
Compatibility:
Most of the Netlink data models specified above maps to FreeBSD concepts
nicely. Unmodified ip(8) binary correctly works with
interfaces, addresses, routes, nexthops and nexthop groups. Some
software such as net/bird require header-only modifications to compile
and work with FreeBSD netlink.
Reviewed by: imp
Differential Revision: https://reviews.freebsd.org/D36002
MFC after: 2 months
2022-01-20 21:39:21 +00:00
|
|
|
|
|
|
|
/*
|
|
|
|
* The logic below provide a p2p interface for receiving and
|
|
|
|
* sending netlink data between the kernel and userland.
|
|
|
|
*/
|
|
|
|
|
2024-01-02 21:03:21 +00:00
|
|
|
static bool nl_process_nbuf(struct nl_buf *nb, struct nlpcb *nlp);
|
netlink: add netlink support
Netlinks is a communication protocol currently used in Linux kernel to modify,
read and subscribe for nearly all networking state. Interfaces, addresses, routes,
firewall, fibs, vnets, etc are controlled via netlink.
It is async, TLV-based protocol, providing 1-1 and 1-many communications.
The current implementation supports the subset of NETLINK_ROUTE
family. To be more specific, the following is supported:
* Dumps:
- routes
- nexthops / nexthop groups
- interfaces
- interface addresses
- neighbors (arp/ndp)
* Notifications:
- interface arrival/departure
- interface address arrival/departure
- route addition/deletion
* Modifications:
- adding/deleting routes
- adding/deleting nexthops/nexthops groups
- adding/deleting neghbors
- adding/deleting interfaces (basic support only)
* Rtsock interaction
- route events are bridged both ways
The implementation also supports the NETLINK_GENERIC family framework.
Implementation notes:
Netlink is implemented via loadable/unloadable kernel module,
not touching many kernel parts.
Each netlink socket uses dedicated taskqueue to support async operations
that can sleep, such as interface creation. All message processing is
performed within these taskqueues.
Compatibility:
Most of the Netlink data models specified above maps to FreeBSD concepts
nicely. Unmodified ip(8) binary correctly works with
interfaces, addresses, routes, nexthops and nexthop groups. Some
software such as net/bird require header-only modifications to compile
and work with FreeBSD netlink.
Reviewed by: imp
Differential Revision: https://reviews.freebsd.org/D36002
MFC after: 2 months
2022-01-20 21:39:21 +00:00
|
|
|
|
2024-01-02 21:04:01 +00:00
|
|
|
struct nl_buf *
|
|
|
|
nl_buf_alloc(size_t len, int mflag)
|
netlink: add netlink support
Netlinks is a communication protocol currently used in Linux kernel to modify,
read and subscribe for nearly all networking state. Interfaces, addresses, routes,
firewall, fibs, vnets, etc are controlled via netlink.
It is async, TLV-based protocol, providing 1-1 and 1-many communications.
The current implementation supports the subset of NETLINK_ROUTE
family. To be more specific, the following is supported:
* Dumps:
- routes
- nexthops / nexthop groups
- interfaces
- interface addresses
- neighbors (arp/ndp)
* Notifications:
- interface arrival/departure
- interface address arrival/departure
- route addition/deletion
* Modifications:
- adding/deleting routes
- adding/deleting nexthops/nexthops groups
- adding/deleting neghbors
- adding/deleting interfaces (basic support only)
* Rtsock interaction
- route events are bridged both ways
The implementation also supports the NETLINK_GENERIC family framework.
Implementation notes:
Netlink is implemented via loadable/unloadable kernel module,
not touching many kernel parts.
Each netlink socket uses dedicated taskqueue to support async operations
that can sleep, such as interface creation. All message processing is
performed within these taskqueues.
Compatibility:
Most of the Netlink data models specified above maps to FreeBSD concepts
nicely. Unmodified ip(8) binary correctly works with
interfaces, addresses, routes, nexthops and nexthop groups. Some
software such as net/bird require header-only modifications to compile
and work with FreeBSD netlink.
Reviewed by: imp
Differential Revision: https://reviews.freebsd.org/D36002
MFC after: 2 months
2022-01-20 21:39:21 +00:00
|
|
|
{
|
2024-01-02 21:04:01 +00:00
|
|
|
struct nl_buf *nb;
|
netlink: add netlink support
Netlinks is a communication protocol currently used in Linux kernel to modify,
read and subscribe for nearly all networking state. Interfaces, addresses, routes,
firewall, fibs, vnets, etc are controlled via netlink.
It is async, TLV-based protocol, providing 1-1 and 1-many communications.
The current implementation supports the subset of NETLINK_ROUTE
family. To be more specific, the following is supported:
* Dumps:
- routes
- nexthops / nexthop groups
- interfaces
- interface addresses
- neighbors (arp/ndp)
* Notifications:
- interface arrival/departure
- interface address arrival/departure
- route addition/deletion
* Modifications:
- adding/deleting routes
- adding/deleting nexthops/nexthops groups
- adding/deleting neghbors
- adding/deleting interfaces (basic support only)
* Rtsock interaction
- route events are bridged both ways
The implementation also supports the NETLINK_GENERIC family framework.
Implementation notes:
Netlink is implemented via loadable/unloadable kernel module,
not touching many kernel parts.
Each netlink socket uses dedicated taskqueue to support async operations
that can sleep, such as interface creation. All message processing is
performed within these taskqueues.
Compatibility:
Most of the Netlink data models specified above maps to FreeBSD concepts
nicely. Unmodified ip(8) binary correctly works with
interfaces, addresses, routes, nexthops and nexthop groups. Some
software such as net/bird require header-only modifications to compile
and work with FreeBSD netlink.
Reviewed by: imp
Differential Revision: https://reviews.freebsd.org/D36002
MFC after: 2 months
2022-01-20 21:39:21 +00:00
|
|
|
|
2024-01-02 21:04:01 +00:00
|
|
|
nb = malloc(sizeof(struct nl_buf) + len, M_NETLINK, mflag);
|
|
|
|
if (__predict_true(nb != NULL)) {
|
|
|
|
nb->buflen = len;
|
|
|
|
nb->datalen = nb->offset = 0;
|
netlink: add netlink support
Netlinks is a communication protocol currently used in Linux kernel to modify,
read and subscribe for nearly all networking state. Interfaces, addresses, routes,
firewall, fibs, vnets, etc are controlled via netlink.
It is async, TLV-based protocol, providing 1-1 and 1-many communications.
The current implementation supports the subset of NETLINK_ROUTE
family. To be more specific, the following is supported:
* Dumps:
- routes
- nexthops / nexthop groups
- interfaces
- interface addresses
- neighbors (arp/ndp)
* Notifications:
- interface arrival/departure
- interface address arrival/departure
- route addition/deletion
* Modifications:
- adding/deleting routes
- adding/deleting nexthops/nexthops groups
- adding/deleting neghbors
- adding/deleting interfaces (basic support only)
* Rtsock interaction
- route events are bridged both ways
The implementation also supports the NETLINK_GENERIC family framework.
Implementation notes:
Netlink is implemented via loadable/unloadable kernel module,
not touching many kernel parts.
Each netlink socket uses dedicated taskqueue to support async operations
that can sleep, such as interface creation. All message processing is
performed within these taskqueues.
Compatibility:
Most of the Netlink data models specified above maps to FreeBSD concepts
nicely. Unmodified ip(8) binary correctly works with
interfaces, addresses, routes, nexthops and nexthop groups. Some
software such as net/bird require header-only modifications to compile
and work with FreeBSD netlink.
Reviewed by: imp
Differential Revision: https://reviews.freebsd.org/D36002
MFC after: 2 months
2022-01-20 21:39:21 +00:00
|
|
|
}
|
|
|
|
|
2024-01-02 21:04:01 +00:00
|
|
|
return (nb);
|
netlink: add netlink support
Netlinks is a communication protocol currently used in Linux kernel to modify,
read and subscribe for nearly all networking state. Interfaces, addresses, routes,
firewall, fibs, vnets, etc are controlled via netlink.
It is async, TLV-based protocol, providing 1-1 and 1-many communications.
The current implementation supports the subset of NETLINK_ROUTE
family. To be more specific, the following is supported:
* Dumps:
- routes
- nexthops / nexthop groups
- interfaces
- interface addresses
- neighbors (arp/ndp)
* Notifications:
- interface arrival/departure
- interface address arrival/departure
- route addition/deletion
* Modifications:
- adding/deleting routes
- adding/deleting nexthops/nexthops groups
- adding/deleting neghbors
- adding/deleting interfaces (basic support only)
* Rtsock interaction
- route events are bridged both ways
The implementation also supports the NETLINK_GENERIC family framework.
Implementation notes:
Netlink is implemented via loadable/unloadable kernel module,
not touching many kernel parts.
Each netlink socket uses dedicated taskqueue to support async operations
that can sleep, such as interface creation. All message processing is
performed within these taskqueues.
Compatibility:
Most of the Netlink data models specified above maps to FreeBSD concepts
nicely. Unmodified ip(8) binary correctly works with
interfaces, addresses, routes, nexthops and nexthop groups. Some
software such as net/bird require header-only modifications to compile
and work with FreeBSD netlink.
Reviewed by: imp
Differential Revision: https://reviews.freebsd.org/D36002
MFC after: 2 months
2022-01-20 21:39:21 +00:00
|
|
|
}
|
|
|
|
|
2024-01-02 21:04:01 +00:00
|
|
|
void
|
|
|
|
nl_buf_free(struct nl_buf *nb)
|
netlink: add netlink support
Netlinks is a communication protocol currently used in Linux kernel to modify,
read and subscribe for nearly all networking state. Interfaces, addresses, routes,
firewall, fibs, vnets, etc are controlled via netlink.
It is async, TLV-based protocol, providing 1-1 and 1-many communications.
The current implementation supports the subset of NETLINK_ROUTE
family. To be more specific, the following is supported:
* Dumps:
- routes
- nexthops / nexthop groups
- interfaces
- interface addresses
- neighbors (arp/ndp)
* Notifications:
- interface arrival/departure
- interface address arrival/departure
- route addition/deletion
* Modifications:
- adding/deleting routes
- adding/deleting nexthops/nexthops groups
- adding/deleting neghbors
- adding/deleting interfaces (basic support only)
* Rtsock interaction
- route events are bridged both ways
The implementation also supports the NETLINK_GENERIC family framework.
Implementation notes:
Netlink is implemented via loadable/unloadable kernel module,
not touching many kernel parts.
Each netlink socket uses dedicated taskqueue to support async operations
that can sleep, such as interface creation. All message processing is
performed within these taskqueues.
Compatibility:
Most of the Netlink data models specified above maps to FreeBSD concepts
nicely. Unmodified ip(8) binary correctly works with
interfaces, addresses, routes, nexthops and nexthop groups. Some
software such as net/bird require header-only modifications to compile
and work with FreeBSD netlink.
Reviewed by: imp
Differential Revision: https://reviews.freebsd.org/D36002
MFC after: 2 months
2022-01-20 21:39:21 +00:00
|
|
|
{
|
|
|
|
|
2024-01-02 21:04:01 +00:00
|
|
|
free(nb, M_NETLINK);
|
netlink: add netlink support
Netlinks is a communication protocol currently used in Linux kernel to modify,
read and subscribe for nearly all networking state. Interfaces, addresses, routes,
firewall, fibs, vnets, etc are controlled via netlink.
It is async, TLV-based protocol, providing 1-1 and 1-many communications.
The current implementation supports the subset of NETLINK_ROUTE
family. To be more specific, the following is supported:
* Dumps:
- routes
- nexthops / nexthop groups
- interfaces
- interface addresses
- neighbors (arp/ndp)
* Notifications:
- interface arrival/departure
- interface address arrival/departure
- route addition/deletion
* Modifications:
- adding/deleting routes
- adding/deleting nexthops/nexthops groups
- adding/deleting neghbors
- adding/deleting interfaces (basic support only)
* Rtsock interaction
- route events are bridged both ways
The implementation also supports the NETLINK_GENERIC family framework.
Implementation notes:
Netlink is implemented via loadable/unloadable kernel module,
not touching many kernel parts.
Each netlink socket uses dedicated taskqueue to support async operations
that can sleep, such as interface creation. All message processing is
performed within these taskqueues.
Compatibility:
Most of the Netlink data models specified above maps to FreeBSD concepts
nicely. Unmodified ip(8) binary correctly works with
interfaces, addresses, routes, nexthops and nexthop groups. Some
software such as net/bird require header-only modifications to compile
and work with FreeBSD netlink.
Reviewed by: imp
Differential Revision: https://reviews.freebsd.org/D36002
MFC after: 2 months
2022-01-20 21:39:21 +00:00
|
|
|
}
|
|
|
|
|
2024-01-02 21:03:21 +00:00
|
|
|
void
|
netlink: add netlink support
Netlinks is a communication protocol currently used in Linux kernel to modify,
read and subscribe for nearly all networking state. Interfaces, addresses, routes,
firewall, fibs, vnets, etc are controlled via netlink.
It is async, TLV-based protocol, providing 1-1 and 1-many communications.
The current implementation supports the subset of NETLINK_ROUTE
family. To be more specific, the following is supported:
* Dumps:
- routes
- nexthops / nexthop groups
- interfaces
- interface addresses
- neighbors (arp/ndp)
* Notifications:
- interface arrival/departure
- interface address arrival/departure
- route addition/deletion
* Modifications:
- adding/deleting routes
- adding/deleting nexthops/nexthops groups
- adding/deleting neghbors
- adding/deleting interfaces (basic support only)
* Rtsock interaction
- route events are bridged both ways
The implementation also supports the NETLINK_GENERIC family framework.
Implementation notes:
Netlink is implemented via loadable/unloadable kernel module,
not touching many kernel parts.
Each netlink socket uses dedicated taskqueue to support async operations
that can sleep, such as interface creation. All message processing is
performed within these taskqueues.
Compatibility:
Most of the Netlink data models specified above maps to FreeBSD concepts
nicely. Unmodified ip(8) binary correctly works with
interfaces, addresses, routes, nexthops and nexthop groups. Some
software such as net/bird require header-only modifications to compile
and work with FreeBSD netlink.
Reviewed by: imp
Differential Revision: https://reviews.freebsd.org/D36002
MFC after: 2 months
2022-01-20 21:39:21 +00:00
|
|
|
nl_schedule_taskqueue(struct nlpcb *nlp)
|
|
|
|
{
|
|
|
|
if (!nlp->nl_task_pending) {
|
|
|
|
nlp->nl_task_pending = true;
|
|
|
|
taskqueue_enqueue(nlp->nl_taskqueue, &nlp->nl_task);
|
|
|
|
NL_LOG(LOG_DEBUG3, "taskqueue scheduled");
|
|
|
|
} else {
|
|
|
|
NL_LOG(LOG_DEBUG3, "taskqueue schedule skipped");
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
|
|
|
static bool
|
|
|
|
nl_process_received_one(struct nlpcb *nlp)
|
|
|
|
{
|
2024-01-02 21:03:21 +00:00
|
|
|
struct socket *so = nlp->nl_socket;
|
2024-01-02 21:04:01 +00:00
|
|
|
struct sockbuf *sb;
|
2024-01-02 21:03:21 +00:00
|
|
|
struct nl_buf *nb;
|
netlink: add netlink support
Netlinks is a communication protocol currently used in Linux kernel to modify,
read and subscribe for nearly all networking state. Interfaces, addresses, routes,
firewall, fibs, vnets, etc are controlled via netlink.
It is async, TLV-based protocol, providing 1-1 and 1-many communications.
The current implementation supports the subset of NETLINK_ROUTE
family. To be more specific, the following is supported:
* Dumps:
- routes
- nexthops / nexthop groups
- interfaces
- interface addresses
- neighbors (arp/ndp)
* Notifications:
- interface arrival/departure
- interface address arrival/departure
- route addition/deletion
* Modifications:
- adding/deleting routes
- adding/deleting nexthops/nexthops groups
- adding/deleting neghbors
- adding/deleting interfaces (basic support only)
* Rtsock interaction
- route events are bridged both ways
The implementation also supports the NETLINK_GENERIC family framework.
Implementation notes:
Netlink is implemented via loadable/unloadable kernel module,
not touching many kernel parts.
Each netlink socket uses dedicated taskqueue to support async operations
that can sleep, such as interface creation. All message processing is
performed within these taskqueues.
Compatibility:
Most of the Netlink data models specified above maps to FreeBSD concepts
nicely. Unmodified ip(8) binary correctly works with
interfaces, addresses, routes, nexthops and nexthop groups. Some
software such as net/bird require header-only modifications to compile
and work with FreeBSD netlink.
Reviewed by: imp
Differential Revision: https://reviews.freebsd.org/D36002
MFC after: 2 months
2022-01-20 21:39:21 +00:00
|
|
|
bool reschedule = false;
|
|
|
|
|
|
|
|
NLP_LOCK(nlp);
|
|
|
|
nlp->nl_task_pending = false;
|
2024-01-02 21:04:01 +00:00
|
|
|
NLP_UNLOCK(nlp);
|
netlink: add netlink support
Netlinks is a communication protocol currently used in Linux kernel to modify,
read and subscribe for nearly all networking state. Interfaces, addresses, routes,
firewall, fibs, vnets, etc are controlled via netlink.
It is async, TLV-based protocol, providing 1-1 and 1-many communications.
The current implementation supports the subset of NETLINK_ROUTE
family. To be more specific, the following is supported:
* Dumps:
- routes
- nexthops / nexthop groups
- interfaces
- interface addresses
- neighbors (arp/ndp)
* Notifications:
- interface arrival/departure
- interface address arrival/departure
- route addition/deletion
* Modifications:
- adding/deleting routes
- adding/deleting nexthops/nexthops groups
- adding/deleting neghbors
- adding/deleting interfaces (basic support only)
* Rtsock interaction
- route events are bridged both ways
The implementation also supports the NETLINK_GENERIC family framework.
Implementation notes:
Netlink is implemented via loadable/unloadable kernel module,
not touching many kernel parts.
Each netlink socket uses dedicated taskqueue to support async operations
that can sleep, such as interface creation. All message processing is
performed within these taskqueues.
Compatibility:
Most of the Netlink data models specified above maps to FreeBSD concepts
nicely. Unmodified ip(8) binary correctly works with
interfaces, addresses, routes, nexthops and nexthop groups. Some
software such as net/bird require header-only modifications to compile
and work with FreeBSD netlink.
Reviewed by: imp
Differential Revision: https://reviews.freebsd.org/D36002
MFC after: 2 months
2022-01-20 21:39:21 +00:00
|
|
|
|
2024-01-02 21:04:01 +00:00
|
|
|
/*
|
|
|
|
* Do not process queued up requests if there is no space to queue
|
|
|
|
* replies.
|
|
|
|
*/
|
|
|
|
sb = &so->so_rcv;
|
|
|
|
SOCK_RECVBUF_LOCK(so);
|
|
|
|
if (sb->sb_hiwat <= sb->sb_ccc) {
|
|
|
|
SOCK_RECVBUF_UNLOCK(so);
|
netlink: add netlink support
Netlinks is a communication protocol currently used in Linux kernel to modify,
read and subscribe for nearly all networking state. Interfaces, addresses, routes,
firewall, fibs, vnets, etc are controlled via netlink.
It is async, TLV-based protocol, providing 1-1 and 1-many communications.
The current implementation supports the subset of NETLINK_ROUTE
family. To be more specific, the following is supported:
* Dumps:
- routes
- nexthops / nexthop groups
- interfaces
- interface addresses
- neighbors (arp/ndp)
* Notifications:
- interface arrival/departure
- interface address arrival/departure
- route addition/deletion
* Modifications:
- adding/deleting routes
- adding/deleting nexthops/nexthops groups
- adding/deleting neghbors
- adding/deleting interfaces (basic support only)
* Rtsock interaction
- route events are bridged both ways
The implementation also supports the NETLINK_GENERIC family framework.
Implementation notes:
Netlink is implemented via loadable/unloadable kernel module,
not touching many kernel parts.
Each netlink socket uses dedicated taskqueue to support async operations
that can sleep, such as interface creation. All message processing is
performed within these taskqueues.
Compatibility:
Most of the Netlink data models specified above maps to FreeBSD concepts
nicely. Unmodified ip(8) binary correctly works with
interfaces, addresses, routes, nexthops and nexthop groups. Some
software such as net/bird require header-only modifications to compile
and work with FreeBSD netlink.
Reviewed by: imp
Differential Revision: https://reviews.freebsd.org/D36002
MFC after: 2 months
2022-01-20 21:39:21 +00:00
|
|
|
return (false);
|
|
|
|
}
|
2024-01-02 21:04:01 +00:00
|
|
|
SOCK_RECVBUF_UNLOCK(so);
|
netlink: add netlink support
Netlinks is a communication protocol currently used in Linux kernel to modify,
read and subscribe for nearly all networking state. Interfaces, addresses, routes,
firewall, fibs, vnets, etc are controlled via netlink.
It is async, TLV-based protocol, providing 1-1 and 1-many communications.
The current implementation supports the subset of NETLINK_ROUTE
family. To be more specific, the following is supported:
* Dumps:
- routes
- nexthops / nexthop groups
- interfaces
- interface addresses
- neighbors (arp/ndp)
* Notifications:
- interface arrival/departure
- interface address arrival/departure
- route addition/deletion
* Modifications:
- adding/deleting routes
- adding/deleting nexthops/nexthops groups
- adding/deleting neghbors
- adding/deleting interfaces (basic support only)
* Rtsock interaction
- route events are bridged both ways
The implementation also supports the NETLINK_GENERIC family framework.
Implementation notes:
Netlink is implemented via loadable/unloadable kernel module,
not touching many kernel parts.
Each netlink socket uses dedicated taskqueue to support async operations
that can sleep, such as interface creation. All message processing is
performed within these taskqueues.
Compatibility:
Most of the Netlink data models specified above maps to FreeBSD concepts
nicely. Unmodified ip(8) binary correctly works with
interfaces, addresses, routes, nexthops and nexthop groups. Some
software such as net/bird require header-only modifications to compile
and work with FreeBSD netlink.
Reviewed by: imp
Differential Revision: https://reviews.freebsd.org/D36002
MFC after: 2 months
2022-01-20 21:39:21 +00:00
|
|
|
|
2024-01-02 21:04:01 +00:00
|
|
|
sb = &so->so_snd;
|
2024-01-02 21:03:21 +00:00
|
|
|
SOCK_SENDBUF_LOCK(so);
|
|
|
|
while ((nb = TAILQ_FIRST(&sb->nl_queue)) != NULL) {
|
|
|
|
TAILQ_REMOVE(&sb->nl_queue, nb, tailq);
|
|
|
|
SOCK_SENDBUF_UNLOCK(so);
|
|
|
|
reschedule = nl_process_nbuf(nb, nlp);
|
|
|
|
SOCK_SENDBUF_LOCK(so);
|
|
|
|
if (reschedule) {
|
|
|
|
sb->sb_acc -= nb->datalen;
|
|
|
|
sb->sb_ccc -= nb->datalen;
|
|
|
|
/* XXXGL: potentially can reduce lock&unlock count. */
|
|
|
|
sowwakeup_locked(so);
|
2024-01-02 21:04:01 +00:00
|
|
|
nl_buf_free(nb);
|
2024-01-02 21:03:21 +00:00
|
|
|
SOCK_SENDBUF_LOCK(so);
|
|
|
|
} else {
|
|
|
|
TAILQ_INSERT_HEAD(&sb->nl_queue, nb, tailq);
|
netlink: add netlink support
Netlinks is a communication protocol currently used in Linux kernel to modify,
read and subscribe for nearly all networking state. Interfaces, addresses, routes,
firewall, fibs, vnets, etc are controlled via netlink.
It is async, TLV-based protocol, providing 1-1 and 1-many communications.
The current implementation supports the subset of NETLINK_ROUTE
family. To be more specific, the following is supported:
* Dumps:
- routes
- nexthops / nexthop groups
- interfaces
- interface addresses
- neighbors (arp/ndp)
* Notifications:
- interface arrival/departure
- interface address arrival/departure
- route addition/deletion
* Modifications:
- adding/deleting routes
- adding/deleting nexthops/nexthops groups
- adding/deleting neghbors
- adding/deleting interfaces (basic support only)
* Rtsock interaction
- route events are bridged both ways
The implementation also supports the NETLINK_GENERIC family framework.
Implementation notes:
Netlink is implemented via loadable/unloadable kernel module,
not touching many kernel parts.
Each netlink socket uses dedicated taskqueue to support async operations
that can sleep, such as interface creation. All message processing is
performed within these taskqueues.
Compatibility:
Most of the Netlink data models specified above maps to FreeBSD concepts
nicely. Unmodified ip(8) binary correctly works with
interfaces, addresses, routes, nexthops and nexthop groups. Some
software such as net/bird require header-only modifications to compile
and work with FreeBSD netlink.
Reviewed by: imp
Differential Revision: https://reviews.freebsd.org/D36002
MFC after: 2 months
2022-01-20 21:39:21 +00:00
|
|
|
break;
|
|
|
|
}
|
|
|
|
}
|
2024-01-02 21:03:21 +00:00
|
|
|
SOCK_SENDBUF_UNLOCK(so);
|
netlink: add netlink support
Netlinks is a communication protocol currently used in Linux kernel to modify,
read and subscribe for nearly all networking state. Interfaces, addresses, routes,
firewall, fibs, vnets, etc are controlled via netlink.
It is async, TLV-based protocol, providing 1-1 and 1-many communications.
The current implementation supports the subset of NETLINK_ROUTE
family. To be more specific, the following is supported:
* Dumps:
- routes
- nexthops / nexthop groups
- interfaces
- interface addresses
- neighbors (arp/ndp)
* Notifications:
- interface arrival/departure
- interface address arrival/departure
- route addition/deletion
* Modifications:
- adding/deleting routes
- adding/deleting nexthops/nexthops groups
- adding/deleting neghbors
- adding/deleting interfaces (basic support only)
* Rtsock interaction
- route events are bridged both ways
The implementation also supports the NETLINK_GENERIC family framework.
Implementation notes:
Netlink is implemented via loadable/unloadable kernel module,
not touching many kernel parts.
Each netlink socket uses dedicated taskqueue to support async operations
that can sleep, such as interface creation. All message processing is
performed within these taskqueues.
Compatibility:
Most of the Netlink data models specified above maps to FreeBSD concepts
nicely. Unmodified ip(8) binary correctly works with
interfaces, addresses, routes, nexthops and nexthop groups. Some
software such as net/bird require header-only modifications to compile
and work with FreeBSD netlink.
Reviewed by: imp
Differential Revision: https://reviews.freebsd.org/D36002
MFC after: 2 months
2022-01-20 21:39:21 +00:00
|
|
|
|
|
|
|
return (reschedule);
|
|
|
|
}
|
|
|
|
|
|
|
|
static void
|
|
|
|
nl_process_received(struct nlpcb *nlp)
|
|
|
|
{
|
|
|
|
NL_LOG(LOG_DEBUG3, "taskqueue called");
|
|
|
|
|
2023-04-28 12:44:04 +00:00
|
|
|
if (__predict_false(nlp->nl_need_thread_setup)) {
|
|
|
|
nl_set_thread_nlp(curthread, nlp);
|
|
|
|
NLP_LOCK(nlp);
|
|
|
|
nlp->nl_need_thread_setup = false;
|
|
|
|
NLP_UNLOCK(nlp);
|
|
|
|
}
|
|
|
|
|
netlink: add netlink support
Netlinks is a communication protocol currently used in Linux kernel to modify,
read and subscribe for nearly all networking state. Interfaces, addresses, routes,
firewall, fibs, vnets, etc are controlled via netlink.
It is async, TLV-based protocol, providing 1-1 and 1-many communications.
The current implementation supports the subset of NETLINK_ROUTE
family. To be more specific, the following is supported:
* Dumps:
- routes
- nexthops / nexthop groups
- interfaces
- interface addresses
- neighbors (arp/ndp)
* Notifications:
- interface arrival/departure
- interface address arrival/departure
- route addition/deletion
* Modifications:
- adding/deleting routes
- adding/deleting nexthops/nexthops groups
- adding/deleting neghbors
- adding/deleting interfaces (basic support only)
* Rtsock interaction
- route events are bridged both ways
The implementation also supports the NETLINK_GENERIC family framework.
Implementation notes:
Netlink is implemented via loadable/unloadable kernel module,
not touching many kernel parts.
Each netlink socket uses dedicated taskqueue to support async operations
that can sleep, such as interface creation. All message processing is
performed within these taskqueues.
Compatibility:
Most of the Netlink data models specified above maps to FreeBSD concepts
nicely. Unmodified ip(8) binary correctly works with
interfaces, addresses, routes, nexthops and nexthop groups. Some
software such as net/bird require header-only modifications to compile
and work with FreeBSD netlink.
Reviewed by: imp
Differential Revision: https://reviews.freebsd.org/D36002
MFC after: 2 months
2022-01-20 21:39:21 +00:00
|
|
|
while (nl_process_received_one(nlp))
|
|
|
|
;
|
|
|
|
}
|
|
|
|
|
|
|
|
/*
|
|
|
|
* Called after some data have been read from the socket.
|
|
|
|
*/
|
|
|
|
void
|
|
|
|
nl_on_transmit(struct nlpcb *nlp)
|
|
|
|
{
|
|
|
|
NLP_LOCK(nlp);
|
|
|
|
|
|
|
|
struct socket *so = nlp->nl_socket;
|
|
|
|
if (__predict_false(nlp->nl_dropped_bytes > 0 && so != NULL)) {
|
2022-10-01 21:16:31 +00:00
|
|
|
unsigned long dropped_bytes = nlp->nl_dropped_bytes;
|
|
|
|
unsigned long dropped_messages = nlp->nl_dropped_messages;
|
netlink: add netlink support
Netlinks is a communication protocol currently used in Linux kernel to modify,
read and subscribe for nearly all networking state. Interfaces, addresses, routes,
firewall, fibs, vnets, etc are controlled via netlink.
It is async, TLV-based protocol, providing 1-1 and 1-many communications.
The current implementation supports the subset of NETLINK_ROUTE
family. To be more specific, the following is supported:
* Dumps:
- routes
- nexthops / nexthop groups
- interfaces
- interface addresses
- neighbors (arp/ndp)
* Notifications:
- interface arrival/departure
- interface address arrival/departure
- route addition/deletion
* Modifications:
- adding/deleting routes
- adding/deleting nexthops/nexthops groups
- adding/deleting neghbors
- adding/deleting interfaces (basic support only)
* Rtsock interaction
- route events are bridged both ways
The implementation also supports the NETLINK_GENERIC family framework.
Implementation notes:
Netlink is implemented via loadable/unloadable kernel module,
not touching many kernel parts.
Each netlink socket uses dedicated taskqueue to support async operations
that can sleep, such as interface creation. All message processing is
performed within these taskqueues.
Compatibility:
Most of the Netlink data models specified above maps to FreeBSD concepts
nicely. Unmodified ip(8) binary correctly works with
interfaces, addresses, routes, nexthops and nexthop groups. Some
software such as net/bird require header-only modifications to compile
and work with FreeBSD netlink.
Reviewed by: imp
Differential Revision: https://reviews.freebsd.org/D36002
MFC after: 2 months
2022-01-20 21:39:21 +00:00
|
|
|
nlp->nl_dropped_bytes = 0;
|
|
|
|
nlp->nl_dropped_messages = 0;
|
|
|
|
|
|
|
|
struct sockbuf *sb = &so->so_rcv;
|
|
|
|
NLP_LOG(LOG_DEBUG, nlp,
|
|
|
|
"socket RX overflowed, %lu messages (%lu bytes) dropped. "
|
2024-01-02 21:04:01 +00:00
|
|
|
"bytes: [%u/%u]", dropped_messages, dropped_bytes,
|
|
|
|
sb->sb_ccc, sb->sb_hiwat);
|
netlink: add netlink support
Netlinks is a communication protocol currently used in Linux kernel to modify,
read and subscribe for nearly all networking state. Interfaces, addresses, routes,
firewall, fibs, vnets, etc are controlled via netlink.
It is async, TLV-based protocol, providing 1-1 and 1-many communications.
The current implementation supports the subset of NETLINK_ROUTE
family. To be more specific, the following is supported:
* Dumps:
- routes
- nexthops / nexthop groups
- interfaces
- interface addresses
- neighbors (arp/ndp)
* Notifications:
- interface arrival/departure
- interface address arrival/departure
- route addition/deletion
* Modifications:
- adding/deleting routes
- adding/deleting nexthops/nexthops groups
- adding/deleting neghbors
- adding/deleting interfaces (basic support only)
* Rtsock interaction
- route events are bridged both ways
The implementation also supports the NETLINK_GENERIC family framework.
Implementation notes:
Netlink is implemented via loadable/unloadable kernel module,
not touching many kernel parts.
Each netlink socket uses dedicated taskqueue to support async operations
that can sleep, such as interface creation. All message processing is
performed within these taskqueues.
Compatibility:
Most of the Netlink data models specified above maps to FreeBSD concepts
nicely. Unmodified ip(8) binary correctly works with
interfaces, addresses, routes, nexthops and nexthop groups. Some
software such as net/bird require header-only modifications to compile
and work with FreeBSD netlink.
Reviewed by: imp
Differential Revision: https://reviews.freebsd.org/D36002
MFC after: 2 months
2022-01-20 21:39:21 +00:00
|
|
|
/* TODO: send netlink message */
|
|
|
|
}
|
|
|
|
|
|
|
|
nl_schedule_taskqueue(nlp);
|
|
|
|
NLP_UNLOCK(nlp);
|
|
|
|
}
|
|
|
|
|
|
|
|
void
|
|
|
|
nl_taskqueue_handler(void *_arg, int pending)
|
|
|
|
{
|
|
|
|
struct nlpcb *nlp = (struct nlpcb *)_arg;
|
|
|
|
|
|
|
|
CURVNET_SET(nlp->nl_socket->so_vnet);
|
|
|
|
nl_process_received(nlp);
|
|
|
|
CURVNET_RESTORE();
|
|
|
|
}
|
|
|
|
|
|
|
|
/*
|
2024-01-02 21:04:01 +00:00
|
|
|
* Tries to send current data buffer from writer.
|
netlink: add netlink support
Netlinks is a communication protocol currently used in Linux kernel to modify,
read and subscribe for nearly all networking state. Interfaces, addresses, routes,
firewall, fibs, vnets, etc are controlled via netlink.
It is async, TLV-based protocol, providing 1-1 and 1-many communications.
The current implementation supports the subset of NETLINK_ROUTE
family. To be more specific, the following is supported:
* Dumps:
- routes
- nexthops / nexthop groups
- interfaces
- interface addresses
- neighbors (arp/ndp)
* Notifications:
- interface arrival/departure
- interface address arrival/departure
- route addition/deletion
* Modifications:
- adding/deleting routes
- adding/deleting nexthops/nexthops groups
- adding/deleting neghbors
- adding/deleting interfaces (basic support only)
* Rtsock interaction
- route events are bridged both ways
The implementation also supports the NETLINK_GENERIC family framework.
Implementation notes:
Netlink is implemented via loadable/unloadable kernel module,
not touching many kernel parts.
Each netlink socket uses dedicated taskqueue to support async operations
that can sleep, such as interface creation. All message processing is
performed within these taskqueues.
Compatibility:
Most of the Netlink data models specified above maps to FreeBSD concepts
nicely. Unmodified ip(8) binary correctly works with
interfaces, addresses, routes, nexthops and nexthop groups. Some
software such as net/bird require header-only modifications to compile
and work with FreeBSD netlink.
Reviewed by: imp
Differential Revision: https://reviews.freebsd.org/D36002
MFC after: 2 months
2022-01-20 21:39:21 +00:00
|
|
|
*
|
|
|
|
* Returns true on success.
|
|
|
|
* If no queue overrunes happened, wakes up socket owner.
|
|
|
|
*/
|
|
|
|
bool
|
2024-01-09 21:01:28 +00:00
|
|
|
nl_send(struct nl_writer *nw, struct nlpcb *nlp)
|
netlink: add netlink support
Netlinks is a communication protocol currently used in Linux kernel to modify,
read and subscribe for nearly all networking state. Interfaces, addresses, routes,
firewall, fibs, vnets, etc are controlled via netlink.
It is async, TLV-based protocol, providing 1-1 and 1-many communications.
The current implementation supports the subset of NETLINK_ROUTE
family. To be more specific, the following is supported:
* Dumps:
- routes
- nexthops / nexthop groups
- interfaces
- interface addresses
- neighbors (arp/ndp)
* Notifications:
- interface arrival/departure
- interface address arrival/departure
- route addition/deletion
* Modifications:
- adding/deleting routes
- adding/deleting nexthops/nexthops groups
- adding/deleting neghbors
- adding/deleting interfaces (basic support only)
* Rtsock interaction
- route events are bridged both ways
The implementation also supports the NETLINK_GENERIC family framework.
Implementation notes:
Netlink is implemented via loadable/unloadable kernel module,
not touching many kernel parts.
Each netlink socket uses dedicated taskqueue to support async operations
that can sleep, such as interface creation. All message processing is
performed within these taskqueues.
Compatibility:
Most of the Netlink data models specified above maps to FreeBSD concepts
nicely. Unmodified ip(8) binary correctly works with
interfaces, addresses, routes, nexthops and nexthop groups. Some
software such as net/bird require header-only modifications to compile
and work with FreeBSD netlink.
Reviewed by: imp
Differential Revision: https://reviews.freebsd.org/D36002
MFC after: 2 months
2022-01-20 21:39:21 +00:00
|
|
|
{
|
2024-01-02 21:04:01 +00:00
|
|
|
struct socket *so = nlp->nl_socket;
|
|
|
|
struct sockbuf *sb = &so->so_rcv;
|
|
|
|
struct nl_buf *nb;
|
|
|
|
|
|
|
|
MPASS(nw->hdr == NULL);
|
2024-01-11 04:51:53 +00:00
|
|
|
MPASS(nw->buf != NULL);
|
|
|
|
MPASS(nw->buf->datalen > 0);
|
netlink: add netlink support
Netlinks is a communication protocol currently used in Linux kernel to modify,
read and subscribe for nearly all networking state. Interfaces, addresses, routes,
firewall, fibs, vnets, etc are controlled via netlink.
It is async, TLV-based protocol, providing 1-1 and 1-many communications.
The current implementation supports the subset of NETLINK_ROUTE
family. To be more specific, the following is supported:
* Dumps:
- routes
- nexthops / nexthop groups
- interfaces
- interface addresses
- neighbors (arp/ndp)
* Notifications:
- interface arrival/departure
- interface address arrival/departure
- route addition/deletion
* Modifications:
- adding/deleting routes
- adding/deleting nexthops/nexthops groups
- adding/deleting neghbors
- adding/deleting interfaces (basic support only)
* Rtsock interaction
- route events are bridged both ways
The implementation also supports the NETLINK_GENERIC family framework.
Implementation notes:
Netlink is implemented via loadable/unloadable kernel module,
not touching many kernel parts.
Each netlink socket uses dedicated taskqueue to support async operations
that can sleep, such as interface creation. All message processing is
performed within these taskqueues.
Compatibility:
Most of the Netlink data models specified above maps to FreeBSD concepts
nicely. Unmodified ip(8) binary correctly works with
interfaces, addresses, routes, nexthops and nexthop groups. Some
software such as net/bird require header-only modifications to compile
and work with FreeBSD netlink.
Reviewed by: imp
Differential Revision: https://reviews.freebsd.org/D36002
MFC after: 2 months
2022-01-20 21:39:21 +00:00
|
|
|
|
|
|
|
IF_DEBUG_LEVEL(LOG_DEBUG2) {
|
2024-01-02 21:04:01 +00:00
|
|
|
struct nlmsghdr *hdr = (struct nlmsghdr *)nw->buf->data;
|
netlink: add netlink support
Netlinks is a communication protocol currently used in Linux kernel to modify,
read and subscribe for nearly all networking state. Interfaces, addresses, routes,
firewall, fibs, vnets, etc are controlled via netlink.
It is async, TLV-based protocol, providing 1-1 and 1-many communications.
The current implementation supports the subset of NETLINK_ROUTE
family. To be more specific, the following is supported:
* Dumps:
- routes
- nexthops / nexthop groups
- interfaces
- interface addresses
- neighbors (arp/ndp)
* Notifications:
- interface arrival/departure
- interface address arrival/departure
- route addition/deletion
* Modifications:
- adding/deleting routes
- adding/deleting nexthops/nexthops groups
- adding/deleting neghbors
- adding/deleting interfaces (basic support only)
* Rtsock interaction
- route events are bridged both ways
The implementation also supports the NETLINK_GENERIC family framework.
Implementation notes:
Netlink is implemented via loadable/unloadable kernel module,
not touching many kernel parts.
Each netlink socket uses dedicated taskqueue to support async operations
that can sleep, such as interface creation. All message processing is
performed within these taskqueues.
Compatibility:
Most of the Netlink data models specified above maps to FreeBSD concepts
nicely. Unmodified ip(8) binary correctly works with
interfaces, addresses, routes, nexthops and nexthop groups. Some
software such as net/bird require header-only modifications to compile
and work with FreeBSD netlink.
Reviewed by: imp
Differential Revision: https://reviews.freebsd.org/D36002
MFC after: 2 months
2022-01-20 21:39:21 +00:00
|
|
|
NLP_LOG(LOG_DEBUG2, nlp,
|
2024-01-02 21:04:01 +00:00
|
|
|
"TX len %u msgs %u msg type %d first hdrlen %u",
|
|
|
|
nw->buf->datalen, nw->num_messages, hdr->nlmsg_type,
|
|
|
|
hdr->nlmsg_len);
|
netlink: add netlink support
Netlinks is a communication protocol currently used in Linux kernel to modify,
read and subscribe for nearly all networking state. Interfaces, addresses, routes,
firewall, fibs, vnets, etc are controlled via netlink.
It is async, TLV-based protocol, providing 1-1 and 1-many communications.
The current implementation supports the subset of NETLINK_ROUTE
family. To be more specific, the following is supported:
* Dumps:
- routes
- nexthops / nexthop groups
- interfaces
- interface addresses
- neighbors (arp/ndp)
* Notifications:
- interface arrival/departure
- interface address arrival/departure
- route addition/deletion
* Modifications:
- adding/deleting routes
- adding/deleting nexthops/nexthops groups
- adding/deleting neghbors
- adding/deleting interfaces (basic support only)
* Rtsock interaction
- route events are bridged both ways
The implementation also supports the NETLINK_GENERIC family framework.
Implementation notes:
Netlink is implemented via loadable/unloadable kernel module,
not touching many kernel parts.
Each netlink socket uses dedicated taskqueue to support async operations
that can sleep, such as interface creation. All message processing is
performed within these taskqueues.
Compatibility:
Most of the Netlink data models specified above maps to FreeBSD concepts
nicely. Unmodified ip(8) binary correctly works with
interfaces, addresses, routes, nexthops and nexthop groups. Some
software such as net/bird require header-only modifications to compile
and work with FreeBSD netlink.
Reviewed by: imp
Differential Revision: https://reviews.freebsd.org/D36002
MFC after: 2 months
2022-01-20 21:39:21 +00:00
|
|
|
}
|
|
|
|
|
2024-01-02 21:04:01 +00:00
|
|
|
if (nlp->nl_linux && linux_netlink_p != NULL &&
|
|
|
|
__predict_false(!linux_netlink_p->msgs_to_linux(nw, nlp))) {
|
|
|
|
nl_buf_free(nw->buf);
|
|
|
|
nw->buf = NULL;
|
|
|
|
return (false);
|
netlink: add netlink support
Netlinks is a communication protocol currently used in Linux kernel to modify,
read and subscribe for nearly all networking state. Interfaces, addresses, routes,
firewall, fibs, vnets, etc are controlled via netlink.
It is async, TLV-based protocol, providing 1-1 and 1-many communications.
The current implementation supports the subset of NETLINK_ROUTE
family. To be more specific, the following is supported:
* Dumps:
- routes
- nexthops / nexthop groups
- interfaces
- interface addresses
- neighbors (arp/ndp)
* Notifications:
- interface arrival/departure
- interface address arrival/departure
- route addition/deletion
* Modifications:
- adding/deleting routes
- adding/deleting nexthops/nexthops groups
- adding/deleting neghbors
- adding/deleting interfaces (basic support only)
* Rtsock interaction
- route events are bridged both ways
The implementation also supports the NETLINK_GENERIC family framework.
Implementation notes:
Netlink is implemented via loadable/unloadable kernel module,
not touching many kernel parts.
Each netlink socket uses dedicated taskqueue to support async operations
that can sleep, such as interface creation. All message processing is
performed within these taskqueues.
Compatibility:
Most of the Netlink data models specified above maps to FreeBSD concepts
nicely. Unmodified ip(8) binary correctly works with
interfaces, addresses, routes, nexthops and nexthop groups. Some
software such as net/bird require header-only modifications to compile
and work with FreeBSD netlink.
Reviewed by: imp
Differential Revision: https://reviews.freebsd.org/D36002
MFC after: 2 months
2022-01-20 21:39:21 +00:00
|
|
|
}
|
|
|
|
|
2024-01-02 21:04:01 +00:00
|
|
|
nb = nw->buf;
|
|
|
|
nw->buf = NULL;
|
netlink: add netlink support
Netlinks is a communication protocol currently used in Linux kernel to modify,
read and subscribe for nearly all networking state. Interfaces, addresses, routes,
firewall, fibs, vnets, etc are controlled via netlink.
It is async, TLV-based protocol, providing 1-1 and 1-many communications.
The current implementation supports the subset of NETLINK_ROUTE
family. To be more specific, the following is supported:
* Dumps:
- routes
- nexthops / nexthop groups
- interfaces
- interface addresses
- neighbors (arp/ndp)
* Notifications:
- interface arrival/departure
- interface address arrival/departure
- route addition/deletion
* Modifications:
- adding/deleting routes
- adding/deleting nexthops/nexthops groups
- adding/deleting neghbors
- adding/deleting interfaces (basic support only)
* Rtsock interaction
- route events are bridged both ways
The implementation also supports the NETLINK_GENERIC family framework.
Implementation notes:
Netlink is implemented via loadable/unloadable kernel module,
not touching many kernel parts.
Each netlink socket uses dedicated taskqueue to support async operations
that can sleep, such as interface creation. All message processing is
performed within these taskqueues.
Compatibility:
Most of the Netlink data models specified above maps to FreeBSD concepts
nicely. Unmodified ip(8) binary correctly works with
interfaces, addresses, routes, nexthops and nexthop groups. Some
software such as net/bird require header-only modifications to compile
and work with FreeBSD netlink.
Reviewed by: imp
Differential Revision: https://reviews.freebsd.org/D36002
MFC after: 2 months
2022-01-20 21:39:21 +00:00
|
|
|
|
2024-01-02 21:04:01 +00:00
|
|
|
SOCK_RECVBUF_LOCK(so);
|
|
|
|
if (!nw->ignore_limit && __predict_false(sb->sb_hiwat <= sb->sb_ccc)) {
|
|
|
|
SOCK_RECVBUF_UNLOCK(so);
|
|
|
|
NLP_LOCK(nlp);
|
|
|
|
nlp->nl_dropped_bytes += nb->datalen;
|
|
|
|
nlp->nl_dropped_messages += nw->num_messages;
|
|
|
|
NLP_LOG(LOG_DEBUG2, nlp, "RX oveflow: %lu m (+%d), %lu b (+%d)",
|
|
|
|
(unsigned long)nlp->nl_dropped_messages, nw->num_messages,
|
|
|
|
(unsigned long)nlp->nl_dropped_bytes, nb->datalen);
|
netlink: add netlink support
Netlinks is a communication protocol currently used in Linux kernel to modify,
read and subscribe for nearly all networking state. Interfaces, addresses, routes,
firewall, fibs, vnets, etc are controlled via netlink.
It is async, TLV-based protocol, providing 1-1 and 1-many communications.
The current implementation supports the subset of NETLINK_ROUTE
family. To be more specific, the following is supported:
* Dumps:
- routes
- nexthops / nexthop groups
- interfaces
- interface addresses
- neighbors (arp/ndp)
* Notifications:
- interface arrival/departure
- interface address arrival/departure
- route addition/deletion
* Modifications:
- adding/deleting routes
- adding/deleting nexthops/nexthops groups
- adding/deleting neghbors
- adding/deleting interfaces (basic support only)
* Rtsock interaction
- route events are bridged both ways
The implementation also supports the NETLINK_GENERIC family framework.
Implementation notes:
Netlink is implemented via loadable/unloadable kernel module,
not touching many kernel parts.
Each netlink socket uses dedicated taskqueue to support async operations
that can sleep, such as interface creation. All message processing is
performed within these taskqueues.
Compatibility:
Most of the Netlink data models specified above maps to FreeBSD concepts
nicely. Unmodified ip(8) binary correctly works with
interfaces, addresses, routes, nexthops and nexthop groups. Some
software such as net/bird require header-only modifications to compile
and work with FreeBSD netlink.
Reviewed by: imp
Differential Revision: https://reviews.freebsd.org/D36002
MFC after: 2 months
2022-01-20 21:39:21 +00:00
|
|
|
NLP_UNLOCK(nlp);
|
2024-01-02 21:04:01 +00:00
|
|
|
nl_buf_free(nb);
|
netlink: add netlink support
Netlinks is a communication protocol currently used in Linux kernel to modify,
read and subscribe for nearly all networking state. Interfaces, addresses, routes,
firewall, fibs, vnets, etc are controlled via netlink.
It is async, TLV-based protocol, providing 1-1 and 1-many communications.
The current implementation supports the subset of NETLINK_ROUTE
family. To be more specific, the following is supported:
* Dumps:
- routes
- nexthops / nexthop groups
- interfaces
- interface addresses
- neighbors (arp/ndp)
* Notifications:
- interface arrival/departure
- interface address arrival/departure
- route addition/deletion
* Modifications:
- adding/deleting routes
- adding/deleting nexthops/nexthops groups
- adding/deleting neghbors
- adding/deleting interfaces (basic support only)
* Rtsock interaction
- route events are bridged both ways
The implementation also supports the NETLINK_GENERIC family framework.
Implementation notes:
Netlink is implemented via loadable/unloadable kernel module,
not touching many kernel parts.
Each netlink socket uses dedicated taskqueue to support async operations
that can sleep, such as interface creation. All message processing is
performed within these taskqueues.
Compatibility:
Most of the Netlink data models specified above maps to FreeBSD concepts
nicely. Unmodified ip(8) binary correctly works with
interfaces, addresses, routes, nexthops and nexthop groups. Some
software such as net/bird require header-only modifications to compile
and work with FreeBSD netlink.
Reviewed by: imp
Differential Revision: https://reviews.freebsd.org/D36002
MFC after: 2 months
2022-01-20 21:39:21 +00:00
|
|
|
return (false);
|
|
|
|
} else {
|
2024-01-02 21:04:01 +00:00
|
|
|
bool full;
|
|
|
|
|
|
|
|
TAILQ_INSERT_TAIL(&sb->nl_queue, nb, tailq);
|
|
|
|
sb->sb_acc += nb->datalen;
|
|
|
|
sb->sb_ccc += nb->datalen;
|
|
|
|
full = sb->sb_hiwat <= sb->sb_ccc;
|
|
|
|
sorwakeup_locked(so);
|
|
|
|
if (full) {
|
|
|
|
NLP_LOCK(nlp);
|
|
|
|
nlp->nl_tx_blocked = true;
|
|
|
|
NLP_UNLOCK(nlp);
|
netlink: add netlink support
Netlinks is a communication protocol currently used in Linux kernel to modify,
read and subscribe for nearly all networking state. Interfaces, addresses, routes,
firewall, fibs, vnets, etc are controlled via netlink.
It is async, TLV-based protocol, providing 1-1 and 1-many communications.
The current implementation supports the subset of NETLINK_ROUTE
family. To be more specific, the following is supported:
* Dumps:
- routes
- nexthops / nexthop groups
- interfaces
- interface addresses
- neighbors (arp/ndp)
* Notifications:
- interface arrival/departure
- interface address arrival/departure
- route addition/deletion
* Modifications:
- adding/deleting routes
- adding/deleting nexthops/nexthops groups
- adding/deleting neghbors
- adding/deleting interfaces (basic support only)
* Rtsock interaction
- route events are bridged both ways
The implementation also supports the NETLINK_GENERIC family framework.
Implementation notes:
Netlink is implemented via loadable/unloadable kernel module,
not touching many kernel parts.
Each netlink socket uses dedicated taskqueue to support async operations
that can sleep, such as interface creation. All message processing is
performed within these taskqueues.
Compatibility:
Most of the Netlink data models specified above maps to FreeBSD concepts
nicely. Unmodified ip(8) binary correctly works with
interfaces, addresses, routes, nexthops and nexthop groups. Some
software such as net/bird require header-only modifications to compile
and work with FreeBSD netlink.
Reviewed by: imp
Differential Revision: https://reviews.freebsd.org/D36002
MFC after: 2 months
2022-01-20 21:39:21 +00:00
|
|
|
}
|
2024-01-02 21:04:01 +00:00
|
|
|
return (true);
|
netlink: add netlink support
Netlinks is a communication protocol currently used in Linux kernel to modify,
read and subscribe for nearly all networking state. Interfaces, addresses, routes,
firewall, fibs, vnets, etc are controlled via netlink.
It is async, TLV-based protocol, providing 1-1 and 1-many communications.
The current implementation supports the subset of NETLINK_ROUTE
family. To be more specific, the following is supported:
* Dumps:
- routes
- nexthops / nexthop groups
- interfaces
- interface addresses
- neighbors (arp/ndp)
* Notifications:
- interface arrival/departure
- interface address arrival/departure
- route addition/deletion
* Modifications:
- adding/deleting routes
- adding/deleting nexthops/nexthops groups
- adding/deleting neghbors
- adding/deleting interfaces (basic support only)
* Rtsock interaction
- route events are bridged both ways
The implementation also supports the NETLINK_GENERIC family framework.
Implementation notes:
Netlink is implemented via loadable/unloadable kernel module,
not touching many kernel parts.
Each netlink socket uses dedicated taskqueue to support async operations
that can sleep, such as interface creation. All message processing is
performed within these taskqueues.
Compatibility:
Most of the Netlink data models specified above maps to FreeBSD concepts
nicely. Unmodified ip(8) binary correctly works with
interfaces, addresses, routes, nexthops and nexthop groups. Some
software such as net/bird require header-only modifications to compile
and work with FreeBSD netlink.
Reviewed by: imp
Differential Revision: https://reviews.freebsd.org/D36002
MFC after: 2 months
2022-01-20 21:39:21 +00:00
|
|
|
}
|
|
|
|
}
|
|
|
|
|
|
|
|
static int
|
|
|
|
nl_receive_message(struct nlmsghdr *hdr, int remaining_length,
|
|
|
|
struct nlpcb *nlp, struct nl_pstate *npt)
|
|
|
|
{
|
|
|
|
nl_handler_f handler = nl_handlers[nlp->nl_proto].cb;
|
|
|
|
int error = 0;
|
|
|
|
|
netlink: suppress sending NLMSG_ERROR if NLMSG_DONE is already sent
Netlink has a confirmation/error reporting mechanism for the sent
messages. Kernel explicitly acks each messages if requested (NLM_F_ACK)
or if message processing results in an error.
Similarly, for multipart messages - typically dumps, where each message
represents a single object like an interface or a route - another
message, NLMSG_DONE is used to indicate the end of dump and the
resulting status.
As a result, successfull dump ends with both NLMSG_DONE and NLMSG_ERROR
messages.
RFC 3549 does not say anything specific about such case.
Linux adopted an optimisation which suppresses NLMSG_ERROR message
when NLMSG_DONE is already sent. Certain libraries/applications like
libnl depends on such behavior.
Suppress sending NLMSG_ERROR if NLMSG_DONE is already sent, by
setting newly-added 'suppress_ack' flag in the writer and checking
this flag when generating ack.
This change restores libnl compatibility.
Before:
```
~ nl-link-list
Error: Unable to allocate link cache: Message sequence number mismatch
````
After:
```
~ nl-link-list
vtnet0 ether 52:54:00:14:e3:19 <broadcast,multicast,up,running>
lo0 ieee1394 <loopback,multicast,up,running>
```
Reviewed by: bapt,pauamma
Tested by: bapt
Differential Revision: https://reviews.freebsd.org/D37565
2022-11-30 12:15:23 +00:00
|
|
|
NLP_LOG(LOG_DEBUG2, nlp, "msg len: %u type: %d: flags: 0x%X seq: %u pid: %u",
|
|
|
|
hdr->nlmsg_len, hdr->nlmsg_type, hdr->nlmsg_flags, hdr->nlmsg_seq,
|
|
|
|
hdr->nlmsg_pid);
|
netlink: add netlink support
Netlinks is a communication protocol currently used in Linux kernel to modify,
read and subscribe for nearly all networking state. Interfaces, addresses, routes,
firewall, fibs, vnets, etc are controlled via netlink.
It is async, TLV-based protocol, providing 1-1 and 1-many communications.
The current implementation supports the subset of NETLINK_ROUTE
family. To be more specific, the following is supported:
* Dumps:
- routes
- nexthops / nexthop groups
- interfaces
- interface addresses
- neighbors (arp/ndp)
* Notifications:
- interface arrival/departure
- interface address arrival/departure
- route addition/deletion
* Modifications:
- adding/deleting routes
- adding/deleting nexthops/nexthops groups
- adding/deleting neghbors
- adding/deleting interfaces (basic support only)
* Rtsock interaction
- route events are bridged both ways
The implementation also supports the NETLINK_GENERIC family framework.
Implementation notes:
Netlink is implemented via loadable/unloadable kernel module,
not touching many kernel parts.
Each netlink socket uses dedicated taskqueue to support async operations
that can sleep, such as interface creation. All message processing is
performed within these taskqueues.
Compatibility:
Most of the Netlink data models specified above maps to FreeBSD concepts
nicely. Unmodified ip(8) binary correctly works with
interfaces, addresses, routes, nexthops and nexthop groups. Some
software such as net/bird require header-only modifications to compile
and work with FreeBSD netlink.
Reviewed by: imp
Differential Revision: https://reviews.freebsd.org/D36002
MFC after: 2 months
2022-01-20 21:39:21 +00:00
|
|
|
|
|
|
|
if (__predict_false(hdr->nlmsg_len > remaining_length)) {
|
|
|
|
NLP_LOG(LOG_DEBUG, nlp, "message is not entirely present: want %d got %d",
|
|
|
|
hdr->nlmsg_len, remaining_length);
|
|
|
|
return (EINVAL);
|
|
|
|
} else if (__predict_false(hdr->nlmsg_len < sizeof(*hdr))) {
|
|
|
|
NL_LOG(LOG_DEBUG, "message too short: %d", hdr->nlmsg_len);
|
|
|
|
return (EINVAL);
|
|
|
|
}
|
|
|
|
/* Stamp each message with sender pid */
|
|
|
|
hdr->nlmsg_pid = nlp->nl_port;
|
|
|
|
|
|
|
|
npt->hdr = hdr;
|
|
|
|
|
2024-03-29 20:35:51 +00:00
|
|
|
if (hdr->nlmsg_flags & NLM_F_REQUEST &&
|
|
|
|
hdr->nlmsg_type >= NLMSG_MIN_TYPE) {
|
netlink: add netlink support
Netlinks is a communication protocol currently used in Linux kernel to modify,
read and subscribe for nearly all networking state. Interfaces, addresses, routes,
firewall, fibs, vnets, etc are controlled via netlink.
It is async, TLV-based protocol, providing 1-1 and 1-many communications.
The current implementation supports the subset of NETLINK_ROUTE
family. To be more specific, the following is supported:
* Dumps:
- routes
- nexthops / nexthop groups
- interfaces
- interface addresses
- neighbors (arp/ndp)
* Notifications:
- interface arrival/departure
- interface address arrival/departure
- route addition/deletion
* Modifications:
- adding/deleting routes
- adding/deleting nexthops/nexthops groups
- adding/deleting neghbors
- adding/deleting interfaces (basic support only)
* Rtsock interaction
- route events are bridged both ways
The implementation also supports the NETLINK_GENERIC family framework.
Implementation notes:
Netlink is implemented via loadable/unloadable kernel module,
not touching many kernel parts.
Each netlink socket uses dedicated taskqueue to support async operations
that can sleep, such as interface creation. All message processing is
performed within these taskqueues.
Compatibility:
Most of the Netlink data models specified above maps to FreeBSD concepts
nicely. Unmodified ip(8) binary correctly works with
interfaces, addresses, routes, nexthops and nexthop groups. Some
software such as net/bird require header-only modifications to compile
and work with FreeBSD netlink.
Reviewed by: imp
Differential Revision: https://reviews.freebsd.org/D36002
MFC after: 2 months
2022-01-20 21:39:21 +00:00
|
|
|
NL_LOG(LOG_DEBUG2, "handling message with msg type: %d",
|
|
|
|
hdr->nlmsg_type);
|
2024-03-29 20:35:51 +00:00
|
|
|
if (nlp->nl_linux) {
|
|
|
|
MPASS(linux_netlink_p != NULL);
|
|
|
|
error = linux_netlink_p->msg_from_linux(nlp->nl_proto,
|
|
|
|
&hdr, npt);
|
|
|
|
if (error)
|
|
|
|
goto ack;
|
netlink: add netlink support
Netlinks is a communication protocol currently used in Linux kernel to modify,
read and subscribe for nearly all networking state. Interfaces, addresses, routes,
firewall, fibs, vnets, etc are controlled via netlink.
It is async, TLV-based protocol, providing 1-1 and 1-many communications.
The current implementation supports the subset of NETLINK_ROUTE
family. To be more specific, the following is supported:
* Dumps:
- routes
- nexthops / nexthop groups
- interfaces
- interface addresses
- neighbors (arp/ndp)
* Notifications:
- interface arrival/departure
- interface address arrival/departure
- route addition/deletion
* Modifications:
- adding/deleting routes
- adding/deleting nexthops/nexthops groups
- adding/deleting neghbors
- adding/deleting interfaces (basic support only)
* Rtsock interaction
- route events are bridged both ways
The implementation also supports the NETLINK_GENERIC family framework.
Implementation notes:
Netlink is implemented via loadable/unloadable kernel module,
not touching many kernel parts.
Each netlink socket uses dedicated taskqueue to support async operations
that can sleep, such as interface creation. All message processing is
performed within these taskqueues.
Compatibility:
Most of the Netlink data models specified above maps to FreeBSD concepts
nicely. Unmodified ip(8) binary correctly works with
interfaces, addresses, routes, nexthops and nexthop groups. Some
software such as net/bird require header-only modifications to compile
and work with FreeBSD netlink.
Reviewed by: imp
Differential Revision: https://reviews.freebsd.org/D36002
MFC after: 2 months
2022-01-20 21:39:21 +00:00
|
|
|
}
|
|
|
|
error = handler(hdr, npt);
|
|
|
|
NL_LOG(LOG_DEBUG2, "retcode: %d", error);
|
|
|
|
}
|
2024-03-29 20:35:51 +00:00
|
|
|
ack:
|
netlink: add netlink support
Netlinks is a communication protocol currently used in Linux kernel to modify,
read and subscribe for nearly all networking state. Interfaces, addresses, routes,
firewall, fibs, vnets, etc are controlled via netlink.
It is async, TLV-based protocol, providing 1-1 and 1-many communications.
The current implementation supports the subset of NETLINK_ROUTE
family. To be more specific, the following is supported:
* Dumps:
- routes
- nexthops / nexthop groups
- interfaces
- interface addresses
- neighbors (arp/ndp)
* Notifications:
- interface arrival/departure
- interface address arrival/departure
- route addition/deletion
* Modifications:
- adding/deleting routes
- adding/deleting nexthops/nexthops groups
- adding/deleting neghbors
- adding/deleting interfaces (basic support only)
* Rtsock interaction
- route events are bridged both ways
The implementation also supports the NETLINK_GENERIC family framework.
Implementation notes:
Netlink is implemented via loadable/unloadable kernel module,
not touching many kernel parts.
Each netlink socket uses dedicated taskqueue to support async operations
that can sleep, such as interface creation. All message processing is
performed within these taskqueues.
Compatibility:
Most of the Netlink data models specified above maps to FreeBSD concepts
nicely. Unmodified ip(8) binary correctly works with
interfaces, addresses, routes, nexthops and nexthop groups. Some
software such as net/bird require header-only modifications to compile
and work with FreeBSD netlink.
Reviewed by: imp
Differential Revision: https://reviews.freebsd.org/D36002
MFC after: 2 months
2022-01-20 21:39:21 +00:00
|
|
|
if ((hdr->nlmsg_flags & NLM_F_ACK) || (error != 0 && error != EINTR)) {
|
netlink: suppress sending NLMSG_ERROR if NLMSG_DONE is already sent
Netlink has a confirmation/error reporting mechanism for the sent
messages. Kernel explicitly acks each messages if requested (NLM_F_ACK)
or if message processing results in an error.
Similarly, for multipart messages - typically dumps, where each message
represents a single object like an interface or a route - another
message, NLMSG_DONE is used to indicate the end of dump and the
resulting status.
As a result, successfull dump ends with both NLMSG_DONE and NLMSG_ERROR
messages.
RFC 3549 does not say anything specific about such case.
Linux adopted an optimisation which suppresses NLMSG_ERROR message
when NLMSG_DONE is already sent. Certain libraries/applications like
libnl depends on such behavior.
Suppress sending NLMSG_ERROR if NLMSG_DONE is already sent, by
setting newly-added 'suppress_ack' flag in the writer and checking
this flag when generating ack.
This change restores libnl compatibility.
Before:
```
~ nl-link-list
Error: Unable to allocate link cache: Message sequence number mismatch
````
After:
```
~ nl-link-list
vtnet0 ether 52:54:00:14:e3:19 <broadcast,multicast,up,running>
lo0 ieee1394 <loopback,multicast,up,running>
```
Reviewed by: bapt,pauamma
Tested by: bapt
Differential Revision: https://reviews.freebsd.org/D37565
2022-11-30 12:15:23 +00:00
|
|
|
if (!npt->nw->suppress_ack) {
|
|
|
|
NL_LOG(LOG_DEBUG3, "ack");
|
|
|
|
nlmsg_ack(nlp, error, hdr, npt);
|
|
|
|
}
|
netlink: add netlink support
Netlinks is a communication protocol currently used in Linux kernel to modify,
read and subscribe for nearly all networking state. Interfaces, addresses, routes,
firewall, fibs, vnets, etc are controlled via netlink.
It is async, TLV-based protocol, providing 1-1 and 1-many communications.
The current implementation supports the subset of NETLINK_ROUTE
family. To be more specific, the following is supported:
* Dumps:
- routes
- nexthops / nexthop groups
- interfaces
- interface addresses
- neighbors (arp/ndp)
* Notifications:
- interface arrival/departure
- interface address arrival/departure
- route addition/deletion
* Modifications:
- adding/deleting routes
- adding/deleting nexthops/nexthops groups
- adding/deleting neghbors
- adding/deleting interfaces (basic support only)
* Rtsock interaction
- route events are bridged both ways
The implementation also supports the NETLINK_GENERIC family framework.
Implementation notes:
Netlink is implemented via loadable/unloadable kernel module,
not touching many kernel parts.
Each netlink socket uses dedicated taskqueue to support async operations
that can sleep, such as interface creation. All message processing is
performed within these taskqueues.
Compatibility:
Most of the Netlink data models specified above maps to FreeBSD concepts
nicely. Unmodified ip(8) binary correctly works with
interfaces, addresses, routes, nexthops and nexthop groups. Some
software such as net/bird require header-only modifications to compile
and work with FreeBSD netlink.
Reviewed by: imp
Differential Revision: https://reviews.freebsd.org/D36002
MFC after: 2 months
2022-01-20 21:39:21 +00:00
|
|
|
}
|
|
|
|
|
|
|
|
return (0);
|
|
|
|
}
|
|
|
|
|
|
|
|
static void
|
|
|
|
npt_clear(struct nl_pstate *npt)
|
|
|
|
{
|
|
|
|
lb_clear(&npt->lb);
|
|
|
|
npt->error = 0;
|
|
|
|
npt->err_msg = NULL;
|
|
|
|
npt->err_off = 0;
|
|
|
|
npt->hdr = NULL;
|
netlink: suppress sending NLMSG_ERROR if NLMSG_DONE is already sent
Netlink has a confirmation/error reporting mechanism for the sent
messages. Kernel explicitly acks each messages if requested (NLM_F_ACK)
or if message processing results in an error.
Similarly, for multipart messages - typically dumps, where each message
represents a single object like an interface or a route - another
message, NLMSG_DONE is used to indicate the end of dump and the
resulting status.
As a result, successfull dump ends with both NLMSG_DONE and NLMSG_ERROR
messages.
RFC 3549 does not say anything specific about such case.
Linux adopted an optimisation which suppresses NLMSG_ERROR message
when NLMSG_DONE is already sent. Certain libraries/applications like
libnl depends on such behavior.
Suppress sending NLMSG_ERROR if NLMSG_DONE is already sent, by
setting newly-added 'suppress_ack' flag in the writer and checking
this flag when generating ack.
This change restores libnl compatibility.
Before:
```
~ nl-link-list
Error: Unable to allocate link cache: Message sequence number mismatch
````
After:
```
~ nl-link-list
vtnet0 ether 52:54:00:14:e3:19 <broadcast,multicast,up,running>
lo0 ieee1394 <loopback,multicast,up,running>
```
Reviewed by: bapt,pauamma
Tested by: bapt
Differential Revision: https://reviews.freebsd.org/D37565
2022-11-30 12:15:23 +00:00
|
|
|
npt->nw->suppress_ack = false;
|
netlink: add netlink support
Netlinks is a communication protocol currently used in Linux kernel to modify,
read and subscribe for nearly all networking state. Interfaces, addresses, routes,
firewall, fibs, vnets, etc are controlled via netlink.
It is async, TLV-based protocol, providing 1-1 and 1-many communications.
The current implementation supports the subset of NETLINK_ROUTE
family. To be more specific, the following is supported:
* Dumps:
- routes
- nexthops / nexthop groups
- interfaces
- interface addresses
- neighbors (arp/ndp)
* Notifications:
- interface arrival/departure
- interface address arrival/departure
- route addition/deletion
* Modifications:
- adding/deleting routes
- adding/deleting nexthops/nexthops groups
- adding/deleting neghbors
- adding/deleting interfaces (basic support only)
* Rtsock interaction
- route events are bridged both ways
The implementation also supports the NETLINK_GENERIC family framework.
Implementation notes:
Netlink is implemented via loadable/unloadable kernel module,
not touching many kernel parts.
Each netlink socket uses dedicated taskqueue to support async operations
that can sleep, such as interface creation. All message processing is
performed within these taskqueues.
Compatibility:
Most of the Netlink data models specified above maps to FreeBSD concepts
nicely. Unmodified ip(8) binary correctly works with
interfaces, addresses, routes, nexthops and nexthop groups. Some
software such as net/bird require header-only modifications to compile
and work with FreeBSD netlink.
Reviewed by: imp
Differential Revision: https://reviews.freebsd.org/D36002
MFC after: 2 months
2022-01-20 21:39:21 +00:00
|
|
|
}
|
|
|
|
|
|
|
|
/*
|
|
|
|
* Processes an incoming packet, which can contain multiple netlink messages
|
|
|
|
*/
|
2024-01-02 21:03:21 +00:00
|
|
|
static bool
|
|
|
|
nl_process_nbuf(struct nl_buf *nb, struct nlpcb *nlp)
|
netlink: add netlink support
Netlinks is a communication protocol currently used in Linux kernel to modify,
read and subscribe for nearly all networking state. Interfaces, addresses, routes,
firewall, fibs, vnets, etc are controlled via netlink.
It is async, TLV-based protocol, providing 1-1 and 1-many communications.
The current implementation supports the subset of NETLINK_ROUTE
family. To be more specific, the following is supported:
* Dumps:
- routes
- nexthops / nexthop groups
- interfaces
- interface addresses
- neighbors (arp/ndp)
* Notifications:
- interface arrival/departure
- interface address arrival/departure
- route addition/deletion
* Modifications:
- adding/deleting routes
- adding/deleting nexthops/nexthops groups
- adding/deleting neghbors
- adding/deleting interfaces (basic support only)
* Rtsock interaction
- route events are bridged both ways
The implementation also supports the NETLINK_GENERIC family framework.
Implementation notes:
Netlink is implemented via loadable/unloadable kernel module,
not touching many kernel parts.
Each netlink socket uses dedicated taskqueue to support async operations
that can sleep, such as interface creation. All message processing is
performed within these taskqueues.
Compatibility:
Most of the Netlink data models specified above maps to FreeBSD concepts
nicely. Unmodified ip(8) binary correctly works with
interfaces, addresses, routes, nexthops and nexthop groups. Some
software such as net/bird require header-only modifications to compile
and work with FreeBSD netlink.
Reviewed by: imp
Differential Revision: https://reviews.freebsd.org/D36002
MFC after: 2 months
2022-01-20 21:39:21 +00:00
|
|
|
{
|
|
|
|
struct nlmsghdr *hdr;
|
|
|
|
int error;
|
|
|
|
|
2024-01-02 21:03:21 +00:00
|
|
|
NL_LOG(LOG_DEBUG3, "RX netlink buf %p on %p", nb, nlp->nl_socket);
|
netlink: add netlink support
Netlinks is a communication protocol currently used in Linux kernel to modify,
read and subscribe for nearly all networking state. Interfaces, addresses, routes,
firewall, fibs, vnets, etc are controlled via netlink.
It is async, TLV-based protocol, providing 1-1 and 1-many communications.
The current implementation supports the subset of NETLINK_ROUTE
family. To be more specific, the following is supported:
* Dumps:
- routes
- nexthops / nexthop groups
- interfaces
- interface addresses
- neighbors (arp/ndp)
* Notifications:
- interface arrival/departure
- interface address arrival/departure
- route addition/deletion
* Modifications:
- adding/deleting routes
- adding/deleting nexthops/nexthops groups
- adding/deleting neghbors
- adding/deleting interfaces (basic support only)
* Rtsock interaction
- route events are bridged both ways
The implementation also supports the NETLINK_GENERIC family framework.
Implementation notes:
Netlink is implemented via loadable/unloadable kernel module,
not touching many kernel parts.
Each netlink socket uses dedicated taskqueue to support async operations
that can sleep, such as interface creation. All message processing is
performed within these taskqueues.
Compatibility:
Most of the Netlink data models specified above maps to FreeBSD concepts
nicely. Unmodified ip(8) binary correctly works with
interfaces, addresses, routes, nexthops and nexthop groups. Some
software such as net/bird require header-only modifications to compile
and work with FreeBSD netlink.
Reviewed by: imp
Differential Revision: https://reviews.freebsd.org/D36002
MFC after: 2 months
2022-01-20 21:39:21 +00:00
|
|
|
|
|
|
|
struct nl_writer nw = {};
|
|
|
|
if (!nlmsg_get_unicast_writer(&nw, NLMSG_SMALL, nlp)) {
|
|
|
|
NL_LOG(LOG_DEBUG, "error allocating socket writer");
|
2024-01-02 21:03:21 +00:00
|
|
|
return (true);
|
netlink: add netlink support
Netlinks is a communication protocol currently used in Linux kernel to modify,
read and subscribe for nearly all networking state. Interfaces, addresses, routes,
firewall, fibs, vnets, etc are controlled via netlink.
It is async, TLV-based protocol, providing 1-1 and 1-many communications.
The current implementation supports the subset of NETLINK_ROUTE
family. To be more specific, the following is supported:
* Dumps:
- routes
- nexthops / nexthop groups
- interfaces
- interface addresses
- neighbors (arp/ndp)
* Notifications:
- interface arrival/departure
- interface address arrival/departure
- route addition/deletion
* Modifications:
- adding/deleting routes
- adding/deleting nexthops/nexthops groups
- adding/deleting neghbors
- adding/deleting interfaces (basic support only)
* Rtsock interaction
- route events are bridged both ways
The implementation also supports the NETLINK_GENERIC family framework.
Implementation notes:
Netlink is implemented via loadable/unloadable kernel module,
not touching many kernel parts.
Each netlink socket uses dedicated taskqueue to support async operations
that can sleep, such as interface creation. All message processing is
performed within these taskqueues.
Compatibility:
Most of the Netlink data models specified above maps to FreeBSD concepts
nicely. Unmodified ip(8) binary correctly works with
interfaces, addresses, routes, nexthops and nexthop groups. Some
software such as net/bird require header-only modifications to compile
and work with FreeBSD netlink.
Reviewed by: imp
Differential Revision: https://reviews.freebsd.org/D36002
MFC after: 2 months
2022-01-20 21:39:21 +00:00
|
|
|
}
|
|
|
|
|
|
|
|
nlmsg_ignore_limit(&nw);
|
|
|
|
|
|
|
|
struct nl_pstate npt = {
|
|
|
|
.nlp = nlp,
|
2024-01-02 21:03:21 +00:00
|
|
|
.lb.base = &nb->data[roundup2(nb->datalen, 8)],
|
|
|
|
.lb.size = nb->buflen - roundup2(nb->datalen, 8),
|
netlink: add netlink support
Netlinks is a communication protocol currently used in Linux kernel to modify,
read and subscribe for nearly all networking state. Interfaces, addresses, routes,
firewall, fibs, vnets, etc are controlled via netlink.
It is async, TLV-based protocol, providing 1-1 and 1-many communications.
The current implementation supports the subset of NETLINK_ROUTE
family. To be more specific, the following is supported:
* Dumps:
- routes
- nexthops / nexthop groups
- interfaces
- interface addresses
- neighbors (arp/ndp)
* Notifications:
- interface arrival/departure
- interface address arrival/departure
- route addition/deletion
* Modifications:
- adding/deleting routes
- adding/deleting nexthops/nexthops groups
- adding/deleting neghbors
- adding/deleting interfaces (basic support only)
* Rtsock interaction
- route events are bridged both ways
The implementation also supports the NETLINK_GENERIC family framework.
Implementation notes:
Netlink is implemented via loadable/unloadable kernel module,
not touching many kernel parts.
Each netlink socket uses dedicated taskqueue to support async operations
that can sleep, such as interface creation. All message processing is
performed within these taskqueues.
Compatibility:
Most of the Netlink data models specified above maps to FreeBSD concepts
nicely. Unmodified ip(8) binary correctly works with
interfaces, addresses, routes, nexthops and nexthop groups. Some
software such as net/bird require header-only modifications to compile
and work with FreeBSD netlink.
Reviewed by: imp
Differential Revision: https://reviews.freebsd.org/D36002
MFC after: 2 months
2022-01-20 21:39:21 +00:00
|
|
|
.nw = &nw,
|
|
|
|
.strict = nlp->nl_flags & NLF_STRICT,
|
|
|
|
};
|
|
|
|
|
2024-01-02 21:03:21 +00:00
|
|
|
for (; nb->offset + sizeof(struct nlmsghdr) <= nb->datalen;) {
|
|
|
|
hdr = (struct nlmsghdr *)&nb->data[nb->offset];
|
netlink: add netlink support
Netlinks is a communication protocol currently used in Linux kernel to modify,
read and subscribe for nearly all networking state. Interfaces, addresses, routes,
firewall, fibs, vnets, etc are controlled via netlink.
It is async, TLV-based protocol, providing 1-1 and 1-many communications.
The current implementation supports the subset of NETLINK_ROUTE
family. To be more specific, the following is supported:
* Dumps:
- routes
- nexthops / nexthop groups
- interfaces
- interface addresses
- neighbors (arp/ndp)
* Notifications:
- interface arrival/departure
- interface address arrival/departure
- route addition/deletion
* Modifications:
- adding/deleting routes
- adding/deleting nexthops/nexthops groups
- adding/deleting neghbors
- adding/deleting interfaces (basic support only)
* Rtsock interaction
- route events are bridged both ways
The implementation also supports the NETLINK_GENERIC family framework.
Implementation notes:
Netlink is implemented via loadable/unloadable kernel module,
not touching many kernel parts.
Each netlink socket uses dedicated taskqueue to support async operations
that can sleep, such as interface creation. All message processing is
performed within these taskqueues.
Compatibility:
Most of the Netlink data models specified above maps to FreeBSD concepts
nicely. Unmodified ip(8) binary correctly works with
interfaces, addresses, routes, nexthops and nexthop groups. Some
software such as net/bird require header-only modifications to compile
and work with FreeBSD netlink.
Reviewed by: imp
Differential Revision: https://reviews.freebsd.org/D36002
MFC after: 2 months
2022-01-20 21:39:21 +00:00
|
|
|
/* Save length prior to calling handler */
|
|
|
|
int msglen = NLMSG_ALIGN(hdr->nlmsg_len);
|
2024-01-02 21:03:21 +00:00
|
|
|
NL_LOG(LOG_DEBUG3, "parsing offset %d/%d",
|
|
|
|
nb->offset, nb->datalen);
|
netlink: add netlink support
Netlinks is a communication protocol currently used in Linux kernel to modify,
read and subscribe for nearly all networking state. Interfaces, addresses, routes,
firewall, fibs, vnets, etc are controlled via netlink.
It is async, TLV-based protocol, providing 1-1 and 1-many communications.
The current implementation supports the subset of NETLINK_ROUTE
family. To be more specific, the following is supported:
* Dumps:
- routes
- nexthops / nexthop groups
- interfaces
- interface addresses
- neighbors (arp/ndp)
* Notifications:
- interface arrival/departure
- interface address arrival/departure
- route addition/deletion
* Modifications:
- adding/deleting routes
- adding/deleting nexthops/nexthops groups
- adding/deleting neghbors
- adding/deleting interfaces (basic support only)
* Rtsock interaction
- route events are bridged both ways
The implementation also supports the NETLINK_GENERIC family framework.
Implementation notes:
Netlink is implemented via loadable/unloadable kernel module,
not touching many kernel parts.
Each netlink socket uses dedicated taskqueue to support async operations
that can sleep, such as interface creation. All message processing is
performed within these taskqueues.
Compatibility:
Most of the Netlink data models specified above maps to FreeBSD concepts
nicely. Unmodified ip(8) binary correctly works with
interfaces, addresses, routes, nexthops and nexthop groups. Some
software such as net/bird require header-only modifications to compile
and work with FreeBSD netlink.
Reviewed by: imp
Differential Revision: https://reviews.freebsd.org/D36002
MFC after: 2 months
2022-01-20 21:39:21 +00:00
|
|
|
npt_clear(&npt);
|
2024-01-02 21:03:21 +00:00
|
|
|
error = nl_receive_message(hdr, nb->datalen - nb->offset, nlp,
|
|
|
|
&npt);
|
|
|
|
nb->offset += msglen;
|
netlink: add netlink support
Netlinks is a communication protocol currently used in Linux kernel to modify,
read and subscribe for nearly all networking state. Interfaces, addresses, routes,
firewall, fibs, vnets, etc are controlled via netlink.
It is async, TLV-based protocol, providing 1-1 and 1-many communications.
The current implementation supports the subset of NETLINK_ROUTE
family. To be more specific, the following is supported:
* Dumps:
- routes
- nexthops / nexthop groups
- interfaces
- interface addresses
- neighbors (arp/ndp)
* Notifications:
- interface arrival/departure
- interface address arrival/departure
- route addition/deletion
* Modifications:
- adding/deleting routes
- adding/deleting nexthops/nexthops groups
- adding/deleting neghbors
- adding/deleting interfaces (basic support only)
* Rtsock interaction
- route events are bridged both ways
The implementation also supports the NETLINK_GENERIC family framework.
Implementation notes:
Netlink is implemented via loadable/unloadable kernel module,
not touching many kernel parts.
Each netlink socket uses dedicated taskqueue to support async operations
that can sleep, such as interface creation. All message processing is
performed within these taskqueues.
Compatibility:
Most of the Netlink data models specified above maps to FreeBSD concepts
nicely. Unmodified ip(8) binary correctly works with
interfaces, addresses, routes, nexthops and nexthop groups. Some
software such as net/bird require header-only modifications to compile
and work with FreeBSD netlink.
Reviewed by: imp
Differential Revision: https://reviews.freebsd.org/D36002
MFC after: 2 months
2022-01-20 21:39:21 +00:00
|
|
|
if (__predict_false(error != 0 || nlp->nl_tx_blocked))
|
|
|
|
break;
|
|
|
|
}
|
|
|
|
NL_LOG(LOG_DEBUG3, "packet parsing done");
|
|
|
|
nlmsg_flush(&nw);
|
|
|
|
|
|
|
|
if (nlp->nl_tx_blocked) {
|
|
|
|
NLP_LOCK(nlp);
|
|
|
|
nlp->nl_tx_blocked = false;
|
|
|
|
NLP_UNLOCK(nlp);
|
2024-01-02 21:03:21 +00:00
|
|
|
return (false);
|
|
|
|
} else
|
|
|
|
return (true);
|
netlink: add netlink support
Netlinks is a communication protocol currently used in Linux kernel to modify,
read and subscribe for nearly all networking state. Interfaces, addresses, routes,
firewall, fibs, vnets, etc are controlled via netlink.
It is async, TLV-based protocol, providing 1-1 and 1-many communications.
The current implementation supports the subset of NETLINK_ROUTE
family. To be more specific, the following is supported:
* Dumps:
- routes
- nexthops / nexthop groups
- interfaces
- interface addresses
- neighbors (arp/ndp)
* Notifications:
- interface arrival/departure
- interface address arrival/departure
- route addition/deletion
* Modifications:
- adding/deleting routes
- adding/deleting nexthops/nexthops groups
- adding/deleting neghbors
- adding/deleting interfaces (basic support only)
* Rtsock interaction
- route events are bridged both ways
The implementation also supports the NETLINK_GENERIC family framework.
Implementation notes:
Netlink is implemented via loadable/unloadable kernel module,
not touching many kernel parts.
Each netlink socket uses dedicated taskqueue to support async operations
that can sleep, such as interface creation. All message processing is
performed within these taskqueues.
Compatibility:
Most of the Netlink data models specified above maps to FreeBSD concepts
nicely. Unmodified ip(8) binary correctly works with
interfaces, addresses, routes, nexthops and nexthop groups. Some
software such as net/bird require header-only modifications to compile
and work with FreeBSD netlink.
Reviewed by: imp
Differential Revision: https://reviews.freebsd.org/D36002
MFC after: 2 months
2022-01-20 21:39:21 +00:00
|
|
|
}
|