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
|
|
|
/*-
|
|
|
|
* SPDX-License-Identifier: BSD-2-Clause-FreeBSD
|
|
|
|
*
|
|
|
|
* 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.
|
|
|
|
*/
|
|
|
|
#ifndef _NETLINK_NETLINK_VAR_H_
|
|
|
|
#define _NETLINK_NETLINK_VAR_H_
|
|
|
|
|
2022-10-01 17:01:53 +00:00
|
|
|
#ifdef _KERNEL
|
|
|
|
|
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/ck.h>
|
|
|
|
#include <sys/epoch.h>
|
|
|
|
#include <sys/sysctl.h>
|
|
|
|
#include <sys/taskqueue.h>
|
|
|
|
#include <net/vnet.h>
|
|
|
|
|
|
|
|
#define NLSNDQ 65536 /* Default socket sendspace */
|
|
|
|
#define NLRCVQ 65536 /* Default socket recvspace */
|
|
|
|
|
|
|
|
struct ucred;
|
|
|
|
|
|
|
|
struct nl_io_queue {
|
|
|
|
STAILQ_HEAD(, mbuf) head;
|
|
|
|
int length;
|
|
|
|
int hiwat;
|
|
|
|
};
|
|
|
|
|
2022-11-03 16:44:07 +00:00
|
|
|
#define NLP_MAX_GROUPS 128
|
|
|
|
|
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 nlpcb {
|
|
|
|
struct socket *nl_socket;
|
2022-11-03 16:44:07 +00:00
|
|
|
uint64_t nl_groups[NLP_MAX_GROUPS / 64];
|
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
|
|
|
uint32_t nl_port;
|
|
|
|
uint32_t nl_flags;
|
|
|
|
uint32_t nl_process_id;
|
|
|
|
int nl_proto;
|
|
|
|
bool nl_active;
|
|
|
|
bool nl_bound;
|
|
|
|
bool nl_task_pending;
|
|
|
|
bool nl_tx_blocked; /* No new requests accepted */
|
|
|
|
bool nl_linux; /* true if running under compat */
|
2023-03-26 08:42:51 +00:00
|
|
|
bool nl_unconstrained_vnet; /* true if running under VNET jail (or without jail) */
|
2023-04-28 12:44:04 +00:00
|
|
|
bool nl_need_thread_setup;
|
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_io_queue rx_queue;
|
|
|
|
struct nl_io_queue tx_queue;
|
|
|
|
struct taskqueue *nl_taskqueue;
|
|
|
|
struct task nl_task;
|
|
|
|
struct ucred *nl_cred; /* Copy of nl_socket->so_cred */
|
|
|
|
uint64_t nl_dropped_bytes;
|
|
|
|
uint64_t nl_dropped_messages;
|
|
|
|
CK_LIST_ENTRY(nlpcb) nl_next;
|
|
|
|
CK_LIST_ENTRY(nlpcb) nl_port_next;
|
|
|
|
volatile u_int nl_refcount;
|
|
|
|
struct mtx nl_lock;
|
|
|
|
struct epoch_context nl_epoch_ctx;
|
|
|
|
};
|
|
|
|
#define sotonlpcb(so) ((struct nlpcb *)(so)->so_pcb)
|
|
|
|
|
|
|
|
#define NLP_LOCK_INIT(_nlp) mtx_init(&((_nlp)->nl_lock), "nlp mtx", NULL, MTX_DEF)
|
|
|
|
#define NLP_LOCK_DESTROY(_nlp) mtx_destroy(&((_nlp)->nl_lock))
|
|
|
|
#define NLP_LOCK(_nlp) mtx_lock(&((_nlp)->nl_lock))
|
|
|
|
#define NLP_UNLOCK(_nlp) mtx_unlock(&((_nlp)->nl_lock))
|
|
|
|
|
|
|
|
#define ALIGNED_NL_SZ(_data) roundup2((((struct nlmsghdr *)(_data))->nlmsg_len), 16)
|
|
|
|
|
|
|
|
/* nl_flags */
|
|
|
|
#define NLF_CAP_ACK 0x01 /* Do not send message body with errmsg */
|
|
|
|
#define NLF_EXT_ACK 0x02 /* Allow including extended TLVs in ack */
|
|
|
|
#define NLF_STRICT 0x04 /* Perform strict header checks */
|
2023-04-28 12:44:04 +00:00
|
|
|
#define NLF_MSG_INFO 0x08 /* Send caller info along with the notifications */
|
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
|
|
|
|
|
|
|
SYSCTL_DECL(_net_netlink);
|
netlink: add netlink KPI to the kernel by default
This change does the following:
Base Netlink KPIs (ability to register the family, parse and/or
write a Netlink message) are always present in the kernel. Specifically,
* Implementation of genetlink family/group registration/removal,
some base accessors (netlink_generic_kpi.c, 260 LoC) are compiled in
unconditionally.
* Basic TLV parser functions (netlink_message_parser.c, 507 LoC) are
compiled in unconditionally.
* Glue functions (netlink<>rtsock), malloc/core sysctl definitions
(netlink_glue.c, 259 LoC) are compiled in unconditionally.
* The rest of the KPI _functions_ are defined in the netlink_glue.c,
but their implementation calls a pointer to either the stub function
or the actual function, depending on whether the module is loaded or not.
This approach allows to have only 1k LoC out of ~3.7k LoC (current
sys/netlink implementation) in the kernel, which will not grow further.
It also allows for the generic netlink kernel customers to load
successfully without requiring Netlink module and operate correctly
once Netlink module is loaded.
Reviewed by: imp
MFC after: 2 weeks
Differential Revision: https://reviews.freebsd.org/D39269
2023-03-27 11:59:30 +00:00
|
|
|
SYSCTL_DECL(_net_netlink_debug);
|
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_io {
|
|
|
|
struct callout callout;
|
|
|
|
struct mbuf *head;
|
|
|
|
struct mbuf *last;
|
|
|
|
int64_t length;
|
|
|
|
};
|
|
|
|
|
|
|
|
struct nl_control {
|
|
|
|
CK_LIST_HEAD(nl_pid_head, nlpcb) ctl_port_head;
|
|
|
|
CK_LIST_HEAD(nlpcb_head, nlpcb) ctl_pcb_head;
|
|
|
|
CK_LIST_ENTRY(nl_control) ctl_next;
|
|
|
|
struct nl_io ctl_io;
|
|
|
|
struct rmlock ctl_lock;
|
|
|
|
};
|
|
|
|
VNET_DECLARE(struct nl_control *, nl_ctl);
|
|
|
|
#define V_nl_ctl VNET(nl_ctl)
|
|
|
|
|
|
|
|
|
|
|
|
struct sockaddr_nl;
|
|
|
|
struct sockaddr;
|
|
|
|
struct nlmsghdr;
|
|
|
|
|
|
|
|
/* netlink_module.c */
|
|
|
|
struct nl_control *vnet_nl_ctl_init(void);
|
|
|
|
|
|
|
|
int nl_verify_proto(int proto);
|
|
|
|
const char *nl_get_proto_name(int proto);
|
|
|
|
|
|
|
|
extern int netlink_unloading;
|
|
|
|
|
|
|
|
struct nl_proto_handler {
|
|
|
|
nl_handler_f cb;
|
|
|
|
const char *proto_name;
|
|
|
|
};
|
|
|
|
extern struct nl_proto_handler *nl_handlers;
|
|
|
|
|
|
|
|
/* netlink_domain.c */
|
|
|
|
void nl_send_group(struct mbuf *m, int cnt, int proto, int group_id);
|
2023-04-28 12:44:04 +00:00
|
|
|
void nl_osd_register(void);
|
|
|
|
void nl_osd_unregister(void);
|
|
|
|
void nl_set_thread_nlp(struct thread *td, 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
|
|
|
|
|
|
|
/* netlink_io.c */
|
|
|
|
#define NL_IOF_UNTRANSLATED 0x01
|
|
|
|
#define NL_IOF_IGNORE_LIMIT 0x02
|
|
|
|
bool nl_send_one(struct mbuf *m, struct nlpcb *nlp, int cnt, int io_flags);
|
|
|
|
void nlmsg_ack(struct nlpcb *nlp, int error, struct nlmsghdr *nlmsg,
|
|
|
|
struct nl_pstate *npt);
|
|
|
|
void nl_on_transmit(struct nlpcb *nlp);
|
|
|
|
void nl_init_io(struct nlpcb *nlp);
|
|
|
|
void nl_free_io(struct nlpcb *nlp);
|
|
|
|
|
|
|
|
void nl_taskqueue_handler(void *_arg, int pending);
|
|
|
|
int nl_receive_async(struct mbuf *m, struct socket *so);
|
|
|
|
void nl_process_receive_locked(struct nlpcb *nlp);
|
2023-04-28 12:44:04 +00:00
|
|
|
void nl_set_source_metadata(struct mbuf *m, int num_messages);
|
|
|
|
void nl_add_msg_info(struct mbuf *m);
|
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
|
|
|
|
netlink: add netlink KPI to the kernel by default
This change does the following:
Base Netlink KPIs (ability to register the family, parse and/or
write a Netlink message) are always present in the kernel. Specifically,
* Implementation of genetlink family/group registration/removal,
some base accessors (netlink_generic_kpi.c, 260 LoC) are compiled in
unconditionally.
* Basic TLV parser functions (netlink_message_parser.c, 507 LoC) are
compiled in unconditionally.
* Glue functions (netlink<>rtsock), malloc/core sysctl definitions
(netlink_glue.c, 259 LoC) are compiled in unconditionally.
* The rest of the KPI _functions_ are defined in the netlink_glue.c,
but their implementation calls a pointer to either the stub function
or the actual function, depending on whether the module is loaded or not.
This approach allows to have only 1k LoC out of ~3.7k LoC (current
sys/netlink implementation) in the kernel, which will not grow further.
It also allows for the generic netlink kernel customers to load
successfully without requiring Netlink module and operate correctly
once Netlink module is loaded.
Reviewed by: imp
MFC after: 2 weeks
Differential Revision: https://reviews.freebsd.org/D39269
2023-03-27 11:59:30 +00:00
|
|
|
/* netlink_generic.c */
|
|
|
|
struct genl_family {
|
|
|
|
const char *family_name;
|
|
|
|
uint16_t family_hdrsize;
|
|
|
|
uint16_t family_id;
|
|
|
|
uint16_t family_version;
|
|
|
|
uint16_t family_attr_max;
|
|
|
|
uint16_t family_cmd_size;
|
|
|
|
uint16_t family_num_groups;
|
|
|
|
struct genl_cmd *family_cmds;
|
|
|
|
};
|
|
|
|
|
|
|
|
struct genl_group {
|
|
|
|
struct genl_family *group_family;
|
|
|
|
const char *group_name;
|
|
|
|
};
|
|
|
|
|
|
|
|
struct genl_family *genl_get_family(uint32_t family_id);
|
|
|
|
struct genl_group *genl_get_group(uint32_t group_id);
|
|
|
|
|
|
|
|
#define MAX_FAMILIES 20
|
|
|
|
#define MAX_GROUPS 64
|
|
|
|
|
|
|
|
#define MIN_GROUP_NUM 48
|
|
|
|
|
|
|
|
#define CTRL_FAMILY_NAME "nlctrl"
|
|
|
|
|
2023-04-19 12:35:02 +00:00
|
|
|
struct ifnet;
|
|
|
|
struct nl_parsed_link;
|
|
|
|
struct nlattr_bmask;
|
|
|
|
struct nl_pstate;
|
|
|
|
|
netlink: add netlink KPI to the kernel by default
This change does the following:
Base Netlink KPIs (ability to register the family, parse and/or
write a Netlink message) are always present in the kernel. Specifically,
* Implementation of genetlink family/group registration/removal,
some base accessors (netlink_generic_kpi.c, 260 LoC) are compiled in
unconditionally.
* Basic TLV parser functions (netlink_message_parser.c, 507 LoC) are
compiled in unconditionally.
* Glue functions (netlink<>rtsock), malloc/core sysctl definitions
(netlink_glue.c, 259 LoC) are compiled in unconditionally.
* The rest of the KPI _functions_ are defined in the netlink_glue.c,
but their implementation calls a pointer to either the stub function
or the actual function, depending on whether the module is loaded or not.
This approach allows to have only 1k LoC out of ~3.7k LoC (current
sys/netlink implementation) in the kernel, which will not grow further.
It also allows for the generic netlink kernel customers to load
successfully without requiring Netlink module and operate correctly
once Netlink module is loaded.
Reviewed by: imp
MFC after: 2 weeks
Differential Revision: https://reviews.freebsd.org/D39269
2023-03-27 11:59:30 +00:00
|
|
|
/* Function map */
|
|
|
|
struct nl_function_wrapper {
|
|
|
|
bool (*nlmsg_add)(struct nl_writer *nw, uint32_t portid, uint32_t seq, uint16_t type,
|
|
|
|
uint16_t flags, uint32_t len);
|
|
|
|
bool (*nlmsg_refill_buffer)(struct nl_writer *nw, int required_len);
|
|
|
|
bool (*nlmsg_flush)(struct nl_writer *nw);
|
|
|
|
bool (*nlmsg_end)(struct nl_writer *nw);
|
|
|
|
void (*nlmsg_abort)(struct nl_writer *nw);
|
|
|
|
void (*nlmsg_ignore_limit)(struct nl_writer *nw);
|
|
|
|
bool (*nlmsg_get_unicast_writer)(struct nl_writer *nw, int size, struct nlpcb *nlp);
|
|
|
|
bool (*nlmsg_get_group_writer)(struct nl_writer *nw, int size, int protocol, int group_id);
|
|
|
|
bool (*nlmsg_get_chain_writer)(struct nl_writer *nw, int size, struct mbuf **pm);
|
|
|
|
bool (*nlmsg_end_dump)(struct nl_writer *nw, int error, struct nlmsghdr *hdr);
|
2023-04-19 12:35:02 +00:00
|
|
|
int (*nl_modify_ifp_generic)(struct ifnet *ifp, struct nl_parsed_link *lattrs,
|
|
|
|
const struct nlattr_bmask *bm, struct nl_pstate *npt);
|
|
|
|
void (*nl_store_ifp_cookie)(struct nl_pstate *npt, struct ifnet *ifp);
|
2023-04-28 12:44:04 +00:00
|
|
|
struct nlpcb * (*nl_get_thread_nlp)(struct thread *td);
|
netlink: add netlink KPI to the kernel by default
This change does the following:
Base Netlink KPIs (ability to register the family, parse and/or
write a Netlink message) are always present in the kernel. Specifically,
* Implementation of genetlink family/group registration/removal,
some base accessors (netlink_generic_kpi.c, 260 LoC) are compiled in
unconditionally.
* Basic TLV parser functions (netlink_message_parser.c, 507 LoC) are
compiled in unconditionally.
* Glue functions (netlink<>rtsock), malloc/core sysctl definitions
(netlink_glue.c, 259 LoC) are compiled in unconditionally.
* The rest of the KPI _functions_ are defined in the netlink_glue.c,
but their implementation calls a pointer to either the stub function
or the actual function, depending on whether the module is loaded or not.
This approach allows to have only 1k LoC out of ~3.7k LoC (current
sys/netlink implementation) in the kernel, which will not grow further.
It also allows for the generic netlink kernel customers to load
successfully without requiring Netlink module and operate correctly
once Netlink module is loaded.
Reviewed by: imp
MFC after: 2 weeks
Differential Revision: https://reviews.freebsd.org/D39269
2023-03-27 11:59:30 +00:00
|
|
|
};
|
|
|
|
void nl_set_functions(const struct nl_function_wrapper *nl);
|
|
|
|
|
2023-04-19 12:35:02 +00:00
|
|
|
|
|
|
|
|
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
|
|
|
#endif
|
2022-10-01 17:01:53 +00:00
|
|
|
#endif
|