freebsd-dev/sys/net/if_var.h

800 lines
30 KiB
C
Raw Normal View History

/*-
* SPDX-License-Identifier: BSD-3-Clause
*
* Copyright (c) 1982, 1986, 1989, 1993
* The Regents of the University of California. All rights reserved.
*
* Redistribution and use in source and binary forms, with or without
* modification, are permitted provided that the following conditions
* are met:
* 1. Redistributions of source code must retain the above copyright
* notice, this list of conditions and the following disclaimer.
* 2. Redistributions in binary form must reproduce the above copyright
* notice, this list of conditions and the following disclaimer in the
* documentation and/or other materials provided with the distribution.
* 3. Neither the name of the University nor the names of its contributors
* may be used to endorse or promote products derived from this software
* without specific prior written permission.
*
* THIS SOFTWARE IS PROVIDED BY THE REGENTS AND CONTRIBUTORS ``AS IS'' AND
* ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
* IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
* ARE DISCLAIMED. IN NO EVENT SHALL THE REGENTS OR CONTRIBUTORS BE LIABLE
* FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
* DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS
* OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
* HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT
* LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY
* OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF
* SUCH DAMAGE.
*
* From: @(#)if.h 8.1 (Berkeley) 6/10/93
1999-08-28 01:08:13 +00:00
* $FreeBSD$
*/
#ifndef _NET_IF_VAR_H_
#define _NET_IF_VAR_H_
/*
* Structures defining a network interface, providing a packet
* transport mechanism (ala level 0 of the PUP protocols).
*
* Each interface accepts output datagrams of a specified maximum
* length, and provides higher level routines with input datagrams
* received from its medium.
*
* Output occurs when the routine if_output is called, with three parameters:
* (*ifp->if_output)(ifp, m, dst, ro)
* Here m is the mbuf chain to be sent and dst is the destination address.
* The output routine encapsulates the supplied datagram if necessary,
* and then transmits it on its medium.
*
* On input, each interface unwraps the data received by it, and either
* places it on the input queue of an internetwork datagram routine
* and posts the associated software interrupt, or passes the datagram to a raw
* packet input routine.
*
* Routines exist for locating interfaces by their addresses
* or for locating an interface on a certain network, as well as more general
* routing and gateway routines maintaining information used to locate
* interfaces. These routines live in the files if.c and route.c
*/
struct rtentry; /* ifa_rtrequest */
struct socket;
struct carp_if;
struct carp_softc;
2006-01-30 13:45:15 +00:00
struct ifvlantrunk;
struct route; /* if_output */
struct vnet;
Introduce a procedural interface to the ifnet structure. The new interface allows the ifnet structure to be defined as an opaque type in NIC drivers. This then allows the ifnet structure to be changed without a need to change or recompile NIC drivers. Put differently, NIC drivers can be written and compiled once and be used with different network stack implementations, provided of course that those network stack implementations have an API and ABI compatible interface. This commit introduces the 'if_t' type to replace 'struct ifnet *' as the type of a network interface. The 'if_t' type is defined as 'void *' to enable the compiler to perform type conversion to 'struct ifnet *' and vice versa where needed and without warnings. The functions that implement the API are the only functions that need to have an explicit cast. The MII code has been converted to use the driver API to avoid unnecessary code churn. Code churn comes from having to work with both converted and unconverted drivers in correlation with having callback functions that take an interface. By converting the MII code first, the callback functions can be defined so that the compiler will perform the typecasts automatically. As soon as all drivers have been converted, the if_t type can be redefined as needed and the API functions can be fix to not need an explicit cast. The immediate benefactors of this change are: 1. Juniper Networks - The network stack implementation in Junos is entirely different from FreeBSD's one and this change allows Juniper to build "stock" NIC drivers that can be used in combination with both the FreeBSD and Junos stacks. 2. FreeBSD - This change opens the door towards changing ifnet and implementing new features and optimizations in the network stack without it requiring a change in the many NIC drivers FreeBSD has. Submitted by: Anuranjan Shukla <anshukla@juniper.net> Reviewed by: glebius@ Obtained from: Juniper Networks, Inc.
2014-06-02 17:54:39 +00:00
struct ifmedia;
struct netmap_adapter;
Split out a more generic debugnet(4) from netdump(4) Debugnet is a simplistic and specialized panic- or debug-time reliable datagram transport. It can drive a single connection at a time and is currently unidirectional (debug/panic machine transmit to remote server only). It is mostly a verbatim code lift from netdump(4). Netdump(4) remains the only consumer (until the rest of this patch series lands). The INET-specific logic has been extracted somewhat more thoroughly than previously in netdump(4), into debugnet_inet.c. UDP-layer logic and up, as much as possible as is protocol-independent, remains in debugnet.c. The separation is not perfect and future improvement is welcome. Supporting INET6 is a long-term goal. Much of the diff is "gratuitous" renaming from 'netdump_' or 'nd_' to 'debugnet_' or 'dn_' -- sorry. I thought keeping the netdump name on the generic module would be more confusing than the refactoring. The only functional change here is the mbuf allocation / tracking. Instead of initiating solely on netdump-configured interface(s) at dumpon(8) configuration time, we watch for any debugnet-enabled NIC for link activation and query it for mbuf parameters at that time. If they exceed the existing high-water mark allocation, we re-allocate and track the new high-water mark. Otherwise, we leave the pre-panic mbuf allocation alone. In a future patch in this series, this will allow initiating netdump from panic ddb(4) without pre-panic configuration. No other functional change intended. Reviewed by: markj (earlier version) Some discussion with: emaste, jhb Objection from: marius Differential Revision: https://reviews.freebsd.org/D21421
2019-10-17 16:23:03 +00:00
struct debugnet_methods;
#ifdef _KERNEL
#include <sys/_eventhandler.h>
#include <sys/mbuf.h> /* ifqueue only? */
#include <sys/buf_ring.h>
Build on Jeff Roberson's linker-set based dynamic per-CPU allocator (DPCPU), as suggested by Peter Wemm, and implement a new per-virtual network stack memory allocator. Modify vnet to use the allocator instead of monolithic global container structures (vinet, ...). This change solves many binary compatibility problems associated with VIMAGE, and restores ELF symbols for virtualized global variables. Each virtualized global variable exists as a "reference copy", and also once per virtual network stack. Virtualized global variables are tagged at compile-time, placing the in a special linker set, which is loaded into a contiguous region of kernel memory. Virtualized global variables in the base kernel are linked as normal, but those in modules are copied and relocated to a reserved portion of the kernel's vnet region with the help of a the kernel linker. Virtualized global variables exist in per-vnet memory set up when the network stack instance is created, and are initialized statically from the reference copy. Run-time access occurs via an accessor macro, which converts from the current vnet and requested symbol to a per-vnet address. When "options VIMAGE" is not compiled into the kernel, normal global ELF symbols will be used instead and indirection is avoided. This change restores static initialization for network stack global variables, restores support for non-global symbols and types, eliminates the need for many subsystem constructors, eliminates large per-subsystem structures that caused many binary compatibility issues both for monitoring applications (netstat) and kernel modules, removes the per-function INIT_VNET_*() macros throughout the stack, eliminates the need for vnet_symmap ksym(2) munging, and eliminates duplicate definitions of virtualized globals under VIMAGE_GLOBALS. Bump __FreeBSD_version and update UPDATING. Portions submitted by: bz Reviewed by: bz, zec Discussed with: gnn, jamie, jeff, jhb, julian, sam Suggested by: peter Approved by: re (kensmith)
2009-07-14 22:48:30 +00:00
#include <net/vnet.h>
#endif /* _KERNEL */
#include <sys/ck.h>
#include <sys/counter.h>
#include <sys/epoch.h>
#include <sys/lock.h> /* XXX */
#include <sys/mutex.h> /* struct ifqueue */
#include <sys/rwlock.h> /* XXX */
#include <sys/sx.h> /* XXX */
#include <sys/_task.h> /* if_link_task */
#define IF_DUNIT_NONE -1
#include <net/altq/if_altq.h>
CK_STAILQ_HEAD(ifnethead, ifnet); /* we use TAILQs so that the order of */
ifnet: Replace if_addr_lock rwlock with epoch + mutex Run on LLNW canaries and tested by pho@ gallatin: Using a 14-core, 28-HTT single socket E5-2697 v3 with a 40GbE MLX5 based ConnectX 4-LX NIC, I see an almost 12% improvement in received packet rate, and a larger improvement in bytes delivered all the way to userspace. When the host receiving 64 streams of netperf -H $DUT -t UDP_STREAM -- -m 1, I see, using nstat -I mce0 1 before the patch: InMpps OMpps InGbs OGbs err TCP Est %CPU syscalls csw irq GBfree 4.98 0.00 4.42 0.00 4235592 33 83.80 4720653 2149771 1235 247.32 4.73 0.00 4.20 0.00 4025260 33 82.99 4724900 2139833 1204 247.32 4.72 0.00 4.20 0.00 4035252 33 82.14 4719162 2132023 1264 247.32 4.71 0.00 4.21 0.00 4073206 33 83.68 4744973 2123317 1347 247.32 4.72 0.00 4.21 0.00 4061118 33 80.82 4713615 2188091 1490 247.32 4.72 0.00 4.21 0.00 4051675 33 85.29 4727399 2109011 1205 247.32 4.73 0.00 4.21 0.00 4039056 33 84.65 4724735 2102603 1053 247.32 After the patch InMpps OMpps InGbs OGbs err TCP Est %CPU syscalls csw irq GBfree 5.43 0.00 4.20 0.00 3313143 33 84.96 5434214 1900162 2656 245.51 5.43 0.00 4.20 0.00 3308527 33 85.24 5439695 1809382 2521 245.51 5.42 0.00 4.19 0.00 3316778 33 87.54 5416028 1805835 2256 245.51 5.42 0.00 4.19 0.00 3317673 33 90.44 5426044 1763056 2332 245.51 5.42 0.00 4.19 0.00 3314839 33 88.11 5435732 1792218 2499 245.52 5.44 0.00 4.19 0.00 3293228 33 91.84 5426301 1668597 2121 245.52 Similarly, netperf reports 230Mb/s before the patch, and 270Mb/s after the patch Reviewed by: gallatin Sponsored by: Limelight Networks Differential Revision: https://reviews.freebsd.org/D15366
2018-05-18 20:13:34 +00:00
CK_STAILQ_HEAD(ifaddrhead, ifaddr); /* instantiation is preserved in the list */
CK_STAILQ_HEAD(ifmultihead, ifmultiaddr);
CK_STAILQ_HEAD(ifgrouphead, ifg_group);
#ifdef _KERNEL
New pfil(9) KPI together with newborn pfil API and control utility. The KPI have been reviewed and cleansed of features that were planned back 20 years ago and never implemented. The pfil(9) internals have been made opaque to protocols with only returned types and function declarations exposed. The KPI is made more strict, but at the same time more extensible, as kernel uses same command structures that userland ioctl uses. In nutshell [KA]PI is about declaring filtering points, declaring filters and linking and unlinking them together. New [KA]PI makes it possible to reconfigure pfil(9) configuration: change order of hooks, rehook filter from one filtering point to a different one, disconnect a hook on output leaving it on input only, prepend/append a filter to existing list of filters. Now it possible for a single packet filter to provide multiple rulesets that may be linked to different points. Think of per-interface ACLs in Cisco or Juniper. None of existing packet filters yet support that, however limited usage is already possible, e.g. default ruleset can be moved to single interface, as soon as interface would pride their filtering points. Another future feature is possiblity to create pfil heads, that provide not an mbuf pointer but just a memory pointer with length. That would allow filtering at very early stages of a packet lifecycle, e.g. when packet has just been received by a NIC and no mbuf was yet allocated. Differential Revision: https://reviews.freebsd.org/D18951
2019-01-31 23:01:03 +00:00
VNET_DECLARE(struct pfil_head *, link_pfil_head);
#define V_link_pfil_head VNET(link_pfil_head)
#define PFIL_ETHER_NAME "ethernet"
#define HHOOK_IPSEC_INET 0
#define HHOOK_IPSEC_INET6 1
#define HHOOK_IPSEC_COUNT 2
VNET_DECLARE(struct hhook_head *, ipsec_hhh_in[HHOOK_IPSEC_COUNT]);
VNET_DECLARE(struct hhook_head *, ipsec_hhh_out[HHOOK_IPSEC_COUNT]);
#define V_ipsec_hhh_in VNET(ipsec_hhh_in)
#define V_ipsec_hhh_out VNET(ipsec_hhh_out)
#endif /* _KERNEL */
typedef enum {
IFCOUNTER_IPACKETS = 0,
IFCOUNTER_IERRORS,
IFCOUNTER_OPACKETS,
IFCOUNTER_OERRORS,
IFCOUNTER_COLLISIONS,
IFCOUNTER_IBYTES,
IFCOUNTER_OBYTES,
IFCOUNTER_IMCASTS,
IFCOUNTER_OMCASTS,
IFCOUNTER_IQDROPS,
IFCOUNTER_OQDROPS,
IFCOUNTER_NOPROTO,
IFCOUNTERS /* Array size. */
} ift_counter;
typedef struct ifnet * if_t;
Introduce a procedural interface to the ifnet structure. The new interface allows the ifnet structure to be defined as an opaque type in NIC drivers. This then allows the ifnet structure to be changed without a need to change or recompile NIC drivers. Put differently, NIC drivers can be written and compiled once and be used with different network stack implementations, provided of course that those network stack implementations have an API and ABI compatible interface. This commit introduces the 'if_t' type to replace 'struct ifnet *' as the type of a network interface. The 'if_t' type is defined as 'void *' to enable the compiler to perform type conversion to 'struct ifnet *' and vice versa where needed and without warnings. The functions that implement the API are the only functions that need to have an explicit cast. The MII code has been converted to use the driver API to avoid unnecessary code churn. Code churn comes from having to work with both converted and unconverted drivers in correlation with having callback functions that take an interface. By converting the MII code first, the callback functions can be defined so that the compiler will perform the typecasts automatically. As soon as all drivers have been converted, the if_t type can be redefined as needed and the API functions can be fix to not need an explicit cast. The immediate benefactors of this change are: 1. Juniper Networks - The network stack implementation in Junos is entirely different from FreeBSD's one and this change allows Juniper to build "stock" NIC drivers that can be used in combination with both the FreeBSD and Junos stacks. 2. FreeBSD - This change opens the door towards changing ifnet and implementing new features and optimizations in the network stack without it requiring a change in the many NIC drivers FreeBSD has. Submitted by: Anuranjan Shukla <anshukla@juniper.net> Reviewed by: glebius@ Obtained from: Juniper Networks, Inc.
2014-06-02 17:54:39 +00:00
typedef void (*if_start_fn_t)(if_t);
typedef int (*if_ioctl_fn_t)(if_t, u_long, caddr_t);
typedef void (*if_init_fn_t)(void *);
typedef void (*if_qflush_fn_t)(if_t);
typedef int (*if_transmit_fn_t)(if_t, struct mbuf *);
typedef uint64_t (*if_get_counter_t)(if_t, ift_counter);
Introduce a procedural interface to the ifnet structure. The new interface allows the ifnet structure to be defined as an opaque type in NIC drivers. This then allows the ifnet structure to be changed without a need to change or recompile NIC drivers. Put differently, NIC drivers can be written and compiled once and be used with different network stack implementations, provided of course that those network stack implementations have an API and ABI compatible interface. This commit introduces the 'if_t' type to replace 'struct ifnet *' as the type of a network interface. The 'if_t' type is defined as 'void *' to enable the compiler to perform type conversion to 'struct ifnet *' and vice versa where needed and without warnings. The functions that implement the API are the only functions that need to have an explicit cast. The MII code has been converted to use the driver API to avoid unnecessary code churn. Code churn comes from having to work with both converted and unconverted drivers in correlation with having callback functions that take an interface. By converting the MII code first, the callback functions can be defined so that the compiler will perform the typecasts automatically. As soon as all drivers have been converted, the if_t type can be redefined as needed and the API functions can be fix to not need an explicit cast. The immediate benefactors of this change are: 1. Juniper Networks - The network stack implementation in Junos is entirely different from FreeBSD's one and this change allows Juniper to build "stock" NIC drivers that can be used in combination with both the FreeBSD and Junos stacks. 2. FreeBSD - This change opens the door towards changing ifnet and implementing new features and optimizations in the network stack without it requiring a change in the many NIC drivers FreeBSD has. Submitted by: Anuranjan Shukla <anshukla@juniper.net> Reviewed by: glebius@ Obtained from: Juniper Networks, Inc.
2014-06-02 17:54:39 +00:00
struct ifnet_hw_tsomax {
u_int tsomaxbytes; /* TSO total burst length limit in bytes */
u_int tsomaxsegcount; /* TSO maximum segment count */
u_int tsomaxsegsize; /* TSO maximum segment size in bytes */
};
Implement interface link header precomputation API. Add if_requestencap() interface method which is capable of calculating various link headers for given interface. Right now there is support for INET/INET6/ARP llheader calculation (IFENCAP_LL type request). Other types are planned to support more complex calculation (L2 multipath lagg nexthops, tunnel encap nexthops, etc..). Reshape 'struct route' to be able to pass additional data (with is length) to prepend to mbuf. These two changes permits routing code to pass pre-calculated nexthop data (like L2 header for route w/gateway) down to the stack eliminating the need for other lookups. It also brings us closer to more complex scenarios like transparently handling MPLS nexthops and tunnel interfaces. Last, but not least, it removes layering violation introduced by flowtable code (ro_lle) and simplifies handling of existing if_output consumers. ARP/ND changes: Make arp/ndp stack pre-calculate link header upon installing/updating lle record. Interface link address change are handled by re-calculating headers for all lles based on if_lladdr event. After these changes, arpresolve()/nd6_resolve() returns full pre-calculated header for supported interfaces thus simplifying if_output(). Move these lookups to separate ether_resolve_addr() function which ether returs error or fully-prepared link header. Add <arp|nd6_>resolve_addr() compat versions to return link addresses instead of pre-calculated data. BPF changes: Raw bpf writes occupied _two_ cases: AF_UNSPEC and pseudo_AF_HDRCMPLT. Despite the naming, both of there have ther header "complete". The only difference is that interface source mac has to be filled by OS for AF_UNSPEC (controlled via BIOCGHDRCMPLT). This logic has to stay inside BPF and not pollute if_output() routines. Convert BPF to pass prepend data via new 'struct route' mechanism. Note that it does not change non-optimized if_output(): ro_prepend handling is purely optional. Side note: hackish pseudo_AF_HDRCMPLT is supported for ethernet and FDDI. It is not needed for ethernet anymore. The only remaining FDDI user is dev/pdq mostly untouched since 2007. FDDI support was eliminated from OpenBSD in 2013 (sys/net/if_fddisubr.c rev 1.65). Flowtable changes: Flowtable violates layering by saving (and not correctly managing) rtes/lles. Instead of passing lle pointer, pass pointer to pre-calculated header data from that lle. Differential Revision: https://reviews.freebsd.org/D4102
2015-12-31 05:03:27 +00:00
/* Interface encap request types */
typedef enum {
IFENCAP_LL = 1 /* pre-calculate link-layer header */
} ife_type;
/*
* The structure below allows to request various pre-calculated L2/L3 headers
* for different media. Requests varies by type (rtype field).
*
* IFENCAP_LL type: pre-calculates link header based on address family
* and destination lladdr.
*
* Input data fields:
* buf: pointer to destination buffer
* bufsize: buffer size
* flags: IFENCAP_FLAG_BROADCAST if destination is broadcast
* family: address family defined by AF_ constant.
* lladdr: pointer to link-layer address
* lladdr_len: length of link-layer address
* hdata: pointer to L3 header (optional, used for ARP requests).
* Output data fields:
* buf: encap data is stored here
* bufsize: resulting encap length is stored here
* lladdr_off: offset of link-layer address from encap hdr start
* hdata: L3 header may be altered if necessary
*/
struct if_encap_req {
u_char *buf; /* Destination buffer (w) */
size_t bufsize; /* size of provided buffer (r) */
ife_type rtype; /* request type (r) */
uint32_t flags; /* Request flags (r) */
int family; /* Address family AF_* (r) */
int lladdr_off; /* offset from header start (w) */
int lladdr_len; /* lladdr length (r) */
char *lladdr; /* link-level address pointer (r) */
char *hdata; /* Upper layer header data (rw) */
};
#define IFENCAP_FLAG_BROADCAST 0x02 /* Destination is broadcast */
Implement kernel support for hardware rate limited sockets. - Add RATELIMIT kernel configuration keyword which must be set to enable the new functionality. - Add support for hardware driven, Receive Side Scaling, RSS aware, rate limited sendqueues and expose the functionality through the already established SO_MAX_PACING_RATE setsockopt(). The API support rates in the range from 1 to 4Gbytes/s which are suitable for regular TCP and UDP streams. The setsockopt(2) manual page has been updated. - Add rate limit function callback API to "struct ifnet" which supports the following operations: if_snd_tag_alloc(), if_snd_tag_modify(), if_snd_tag_query() and if_snd_tag_free(). - Add support to ifconfig to view, set and clear the IFCAP_TXRTLMT flag, which tells if a network driver supports rate limiting or not. - This patch also adds support for rate limiting through VLAN and LAGG intermediate network devices. - How rate limiting works: 1) The userspace application calls setsockopt() after accepting or making a new connection to set the rate which is then stored in the socket structure in the kernel. Later on when packets are transmitted a check is made in the transmit path for rate changes. A rate change implies a non-blocking ifp->if_snd_tag_alloc() call will be made to the destination network interface, which then sets up a custom sendqueue with the given rate limitation parameter. A "struct m_snd_tag" pointer is returned which serves as a "snd_tag" hint in the m_pkthdr for the subsequently transmitted mbufs. 2) When the network driver sees the "m->m_pkthdr.snd_tag" different from NULL, it will move the packets into a designated rate limited sendqueue given by the snd_tag pointer. It is up to the individual drivers how the rate limited traffic will be rate limited. 3) Route changes are detected by the NIC drivers in the ifp->if_transmit() routine when the ifnet pointer in the incoming snd_tag mismatches the one of the network interface. The network adapter frees the mbuf and returns EAGAIN which causes the ip_output() to release and clear the send tag. Upon next ip_output() a new "snd_tag" will be tried allocated. 4) When the PCB is detached the custom sendqueue will be released by a non-blocking ifp->if_snd_tag_free() call to the currently bound network interface. Reviewed by: wblock (manpages), adrian, gallatin, scottl (network) Differential Revision: https://reviews.freebsd.org/D3687 Sponsored by: Mellanox Technologies MFC after: 3 months
2017-01-18 13:31:17 +00:00
/*
* Network interface send tag support. The storage of "struct
* m_snd_tag" comes from the network driver and it is free to allocate
* as much additional space as it wants for its own use.
*/
Add kernel-side support for in-kernel TLS. KTLS adds support for in-kernel framing and encryption of Transport Layer Security (1.0-1.2) data on TCP sockets. KTLS only supports offload of TLS for transmitted data. Key negotation must still be performed in userland. Once completed, transmit session keys for a connection are provided to the kernel via a new TCP_TXTLS_ENABLE socket option. All subsequent data transmitted on the socket is placed into TLS frames and encrypted using the supplied keys. Any data written to a KTLS-enabled socket via write(2), aio_write(2), or sendfile(2) is assumed to be application data and is encoded in TLS frames with an application data type. Individual records can be sent with a custom type (e.g. handshake messages) via sendmsg(2) with a new control message (TLS_SET_RECORD_TYPE) specifying the record type. At present, rekeying is not supported though the in-kernel framework should support rekeying. KTLS makes use of the recently added unmapped mbufs to store TLS frames in the socket buffer. Each TLS frame is described by a single ext_pgs mbuf. The ext_pgs structure contains the header of the TLS record (and trailer for encrypted records) as well as references to the associated TLS session. KTLS supports two primary methods of encrypting TLS frames: software TLS and ifnet TLS. Software TLS marks mbufs holding socket data as not ready via M_NOTREADY similar to sendfile(2) when TLS framing information is added to an unmapped mbuf in ktls_frame(). ktls_enqueue() is then called to schedule TLS frames for encryption. In the case of sendfile_iodone() calls ktls_enqueue() instead of pru_ready() leaving the mbufs marked M_NOTREADY until encryption is completed. For other writes (vn_sendfile when pages are available, write(2), etc.), the PRUS_NOTREADY is set when invoking pru_send() along with invoking ktls_enqueue(). A pool of worker threads (the "KTLS" kernel process) encrypts TLS frames queued via ktls_enqueue(). Each TLS frame is temporarily mapped using the direct map and passed to a software encryption backend to perform the actual encryption. (Note: The use of PHYS_TO_DMAP could be replaced with sf_bufs if someone wished to make this work on architectures without a direct map.) KTLS supports pluggable software encryption backends. Internally, Netflix uses proprietary pure-software backends. This commit includes a simple backend in a new ktls_ocf.ko module that uses the kernel's OpenCrypto framework to provide AES-GCM encryption of TLS frames. As a result, software TLS is now a bit of a misnomer as it can make use of hardware crypto accelerators. Once software encryption has finished, the TLS frame mbufs are marked ready via pru_ready(). At this point, the encrypted data appears as regular payload to the TCP stack stored in unmapped mbufs. ifnet TLS permits a NIC to offload the TLS encryption and TCP segmentation. In this mode, a new send tag type (IF_SND_TAG_TYPE_TLS) is allocated on the interface a socket is routed over and associated with a TLS session. TLS records for a TLS session using ifnet TLS are not marked M_NOTREADY but are passed down the stack unencrypted. The ip_output_send() and ip6_output_send() helper functions that apply send tags to outbound IP packets verify that the send tag of the TLS record matches the outbound interface. If so, the packet is tagged with the TLS send tag and sent to the interface. The NIC device driver must recognize packets with the TLS send tag and schedule them for TLS encryption and TCP segmentation. If the the outbound interface does not match the interface in the TLS send tag, the packet is dropped. In addition, a task is scheduled to refresh the TLS send tag for the TLS session. If a new TLS send tag cannot be allocated, the connection is dropped. If a new TLS send tag is allocated, however, subsequent packets will be tagged with the correct TLS send tag. (This latter case has been tested by configuring both ports of a Chelsio T6 in a lagg and failing over from one port to another. As the connections migrated to the new port, new TLS send tags were allocated for the new port and connections resumed without being dropped.) ifnet TLS can be enabled and disabled on supported network interfaces via new '[-]txtls[46]' options to ifconfig(8). ifnet TLS is supported across both vlan devices and lagg interfaces using failover, lacp with flowid enabled, or lacp with flowid enabled. Applications may request the current KTLS mode of a connection via a new TCP_TXTLS_MODE socket option. They can also use this socket option to toggle between software and ifnet TLS modes. In addition, a testing tool is available in tools/tools/switch_tls. This is modeled on tcpdrop and uses similar syntax. However, instead of dropping connections, -s is used to force KTLS connections to switch to software TLS and -i is used to switch to ifnet TLS. Various sysctls and counters are available under the kern.ipc.tls sysctl node. The kern.ipc.tls.enable node must be set to true to enable KTLS (it is off by default). The use of unmapped mbufs must also be enabled via kern.ipc.mb_use_ext_pgs to enable KTLS. KTLS is enabled via the KERN_TLS kernel option. This patch is the culmination of years of work by several folks including Scott Long and Randall Stewart for the original design and implementation; Drew Gallatin for several optimizations including the use of ext_pgs mbufs, the M_NOTREADY mechanism for TLS records awaiting software encryption, and pluggable software crypto backends; and John Baldwin for modifications to support hardware TLS offload. Reviewed by: gallatin, hselasky, rrs Obtained from: Netflix Sponsored by: Netflix, Chelsio Communications Differential Revision: https://reviews.freebsd.org/D21277
2019-08-27 00:01:56 +00:00
struct ktls_session;
Implement kernel support for hardware rate limited sockets. - Add RATELIMIT kernel configuration keyword which must be set to enable the new functionality. - Add support for hardware driven, Receive Side Scaling, RSS aware, rate limited sendqueues and expose the functionality through the already established SO_MAX_PACING_RATE setsockopt(). The API support rates in the range from 1 to 4Gbytes/s which are suitable for regular TCP and UDP streams. The setsockopt(2) manual page has been updated. - Add rate limit function callback API to "struct ifnet" which supports the following operations: if_snd_tag_alloc(), if_snd_tag_modify(), if_snd_tag_query() and if_snd_tag_free(). - Add support to ifconfig to view, set and clear the IFCAP_TXRTLMT flag, which tells if a network driver supports rate limiting or not. - This patch also adds support for rate limiting through VLAN and LAGG intermediate network devices. - How rate limiting works: 1) The userspace application calls setsockopt() after accepting or making a new connection to set the rate which is then stored in the socket structure in the kernel. Later on when packets are transmitted a check is made in the transmit path for rate changes. A rate change implies a non-blocking ifp->if_snd_tag_alloc() call will be made to the destination network interface, which then sets up a custom sendqueue with the given rate limitation parameter. A "struct m_snd_tag" pointer is returned which serves as a "snd_tag" hint in the m_pkthdr for the subsequently transmitted mbufs. 2) When the network driver sees the "m->m_pkthdr.snd_tag" different from NULL, it will move the packets into a designated rate limited sendqueue given by the snd_tag pointer. It is up to the individual drivers how the rate limited traffic will be rate limited. 3) Route changes are detected by the NIC drivers in the ifp->if_transmit() routine when the ifnet pointer in the incoming snd_tag mismatches the one of the network interface. The network adapter frees the mbuf and returns EAGAIN which causes the ip_output() to release and clear the send tag. Upon next ip_output() a new "snd_tag" will be tried allocated. 4) When the PCB is detached the custom sendqueue will be released by a non-blocking ifp->if_snd_tag_free() call to the currently bound network interface. Reviewed by: wblock (manpages), adrian, gallatin, scottl (network) Differential Revision: https://reviews.freebsd.org/D3687 Sponsored by: Mellanox Technologies MFC after: 3 months
2017-01-18 13:31:17 +00:00
struct m_snd_tag;
#define IF_SND_TAG_TYPE_RATE_LIMIT 0
#define IF_SND_TAG_TYPE_UNLIMITED 1
Add kernel-side support for in-kernel TLS. KTLS adds support for in-kernel framing and encryption of Transport Layer Security (1.0-1.2) data on TCP sockets. KTLS only supports offload of TLS for transmitted data. Key negotation must still be performed in userland. Once completed, transmit session keys for a connection are provided to the kernel via a new TCP_TXTLS_ENABLE socket option. All subsequent data transmitted on the socket is placed into TLS frames and encrypted using the supplied keys. Any data written to a KTLS-enabled socket via write(2), aio_write(2), or sendfile(2) is assumed to be application data and is encoded in TLS frames with an application data type. Individual records can be sent with a custom type (e.g. handshake messages) via sendmsg(2) with a new control message (TLS_SET_RECORD_TYPE) specifying the record type. At present, rekeying is not supported though the in-kernel framework should support rekeying. KTLS makes use of the recently added unmapped mbufs to store TLS frames in the socket buffer. Each TLS frame is described by a single ext_pgs mbuf. The ext_pgs structure contains the header of the TLS record (and trailer for encrypted records) as well as references to the associated TLS session. KTLS supports two primary methods of encrypting TLS frames: software TLS and ifnet TLS. Software TLS marks mbufs holding socket data as not ready via M_NOTREADY similar to sendfile(2) when TLS framing information is added to an unmapped mbuf in ktls_frame(). ktls_enqueue() is then called to schedule TLS frames for encryption. In the case of sendfile_iodone() calls ktls_enqueue() instead of pru_ready() leaving the mbufs marked M_NOTREADY until encryption is completed. For other writes (vn_sendfile when pages are available, write(2), etc.), the PRUS_NOTREADY is set when invoking pru_send() along with invoking ktls_enqueue(). A pool of worker threads (the "KTLS" kernel process) encrypts TLS frames queued via ktls_enqueue(). Each TLS frame is temporarily mapped using the direct map and passed to a software encryption backend to perform the actual encryption. (Note: The use of PHYS_TO_DMAP could be replaced with sf_bufs if someone wished to make this work on architectures without a direct map.) KTLS supports pluggable software encryption backends. Internally, Netflix uses proprietary pure-software backends. This commit includes a simple backend in a new ktls_ocf.ko module that uses the kernel's OpenCrypto framework to provide AES-GCM encryption of TLS frames. As a result, software TLS is now a bit of a misnomer as it can make use of hardware crypto accelerators. Once software encryption has finished, the TLS frame mbufs are marked ready via pru_ready(). At this point, the encrypted data appears as regular payload to the TCP stack stored in unmapped mbufs. ifnet TLS permits a NIC to offload the TLS encryption and TCP segmentation. In this mode, a new send tag type (IF_SND_TAG_TYPE_TLS) is allocated on the interface a socket is routed over and associated with a TLS session. TLS records for a TLS session using ifnet TLS are not marked M_NOTREADY but are passed down the stack unencrypted. The ip_output_send() and ip6_output_send() helper functions that apply send tags to outbound IP packets verify that the send tag of the TLS record matches the outbound interface. If so, the packet is tagged with the TLS send tag and sent to the interface. The NIC device driver must recognize packets with the TLS send tag and schedule them for TLS encryption and TCP segmentation. If the the outbound interface does not match the interface in the TLS send tag, the packet is dropped. In addition, a task is scheduled to refresh the TLS send tag for the TLS session. If a new TLS send tag cannot be allocated, the connection is dropped. If a new TLS send tag is allocated, however, subsequent packets will be tagged with the correct TLS send tag. (This latter case has been tested by configuring both ports of a Chelsio T6 in a lagg and failing over from one port to another. As the connections migrated to the new port, new TLS send tags were allocated for the new port and connections resumed without being dropped.) ifnet TLS can be enabled and disabled on supported network interfaces via new '[-]txtls[46]' options to ifconfig(8). ifnet TLS is supported across both vlan devices and lagg interfaces using failover, lacp with flowid enabled, or lacp with flowid enabled. Applications may request the current KTLS mode of a connection via a new TCP_TXTLS_MODE socket option. They can also use this socket option to toggle between software and ifnet TLS modes. In addition, a testing tool is available in tools/tools/switch_tls. This is modeled on tcpdrop and uses similar syntax. However, instead of dropping connections, -s is used to force KTLS connections to switch to software TLS and -i is used to switch to ifnet TLS. Various sysctls and counters are available under the kern.ipc.tls sysctl node. The kern.ipc.tls.enable node must be set to true to enable KTLS (it is off by default). The use of unmapped mbufs must also be enabled via kern.ipc.mb_use_ext_pgs to enable KTLS. KTLS is enabled via the KERN_TLS kernel option. This patch is the culmination of years of work by several folks including Scott Long and Randall Stewart for the original design and implementation; Drew Gallatin for several optimizations including the use of ext_pgs mbufs, the M_NOTREADY mechanism for TLS records awaiting software encryption, and pluggable software crypto backends; and John Baldwin for modifications to support hardware TLS offload. Reviewed by: gallatin, hselasky, rrs Obtained from: Netflix Sponsored by: Netflix, Chelsio Communications Differential Revision: https://reviews.freebsd.org/D21277
2019-08-27 00:01:56 +00:00
#define IF_SND_TAG_TYPE_TLS 2
Support hardware rate limiting (pacing) with TLS offload. - Add a new send tag type for a send tag that supports both rate limiting (packet pacing) and TLS offload (mostly similar to D22669 but adds a separate structure when allocating the new tag type). - When allocating a send tag for TLS offload, check to see if the connection already has a pacing rate. If so, allocate a tag that supports both rate limiting and TLS offload rather than a plain TLS offload tag. - When setting an initial rate on an existing ifnet KTLS connection, set the rate in the TCP control block inp and then reset the TLS send tag (via ktls_output_eagain) to reallocate a TLS + ratelimit send tag. This allocates the TLS send tag asynchronously from a task queue, so the TLS rate limit tag alloc is always sleepable. - When modifying a rate on a connection using KTLS, look for a TLS send tag. If the send tag is only a plain TLS send tag, assume we failed to allocate a TLS ratelimit tag (either during the TCP_TXTLS_ENABLE socket option, or during the send tag reset triggered by ktls_output_eagain) and ignore the new rate. If the send tag is a ratelimit TLS send tag, change the rate on the TLS tag and leave the inp tag alone. - Lock the inp lock when setting sb_tls_info for a socket send buffer so that the routines in tcp_ratelimit can safely dereference the pointer without needing to grab the socket buffer lock. - Add an IFCAP_TXTLS_RTLMT capability flag and associated administrative controls in ifconfig(8). TLS rate limit tags are only allocated if this capability is enabled. Note that TLS offload (whether unlimited or rate limited) always requires IFCAP_TXTLS[46]. Reviewed by: gallatin, hselasky Relnotes: yes Sponsored by: Netflix Differential Revision: https://reviews.freebsd.org/D26691
2020-10-29 00:23:16 +00:00
#define IF_SND_TAG_TYPE_TLS_RATE_LIMIT 3
#define IF_SND_TAG_TYPE_MAX 4
Implement kernel support for hardware rate limited sockets. - Add RATELIMIT kernel configuration keyword which must be set to enable the new functionality. - Add support for hardware driven, Receive Side Scaling, RSS aware, rate limited sendqueues and expose the functionality through the already established SO_MAX_PACING_RATE setsockopt(). The API support rates in the range from 1 to 4Gbytes/s which are suitable for regular TCP and UDP streams. The setsockopt(2) manual page has been updated. - Add rate limit function callback API to "struct ifnet" which supports the following operations: if_snd_tag_alloc(), if_snd_tag_modify(), if_snd_tag_query() and if_snd_tag_free(). - Add support to ifconfig to view, set and clear the IFCAP_TXRTLMT flag, which tells if a network driver supports rate limiting or not. - This patch also adds support for rate limiting through VLAN and LAGG intermediate network devices. - How rate limiting works: 1) The userspace application calls setsockopt() after accepting or making a new connection to set the rate which is then stored in the socket structure in the kernel. Later on when packets are transmitted a check is made in the transmit path for rate changes. A rate change implies a non-blocking ifp->if_snd_tag_alloc() call will be made to the destination network interface, which then sets up a custom sendqueue with the given rate limitation parameter. A "struct m_snd_tag" pointer is returned which serves as a "snd_tag" hint in the m_pkthdr for the subsequently transmitted mbufs. 2) When the network driver sees the "m->m_pkthdr.snd_tag" different from NULL, it will move the packets into a designated rate limited sendqueue given by the snd_tag pointer. It is up to the individual drivers how the rate limited traffic will be rate limited. 3) Route changes are detected by the NIC drivers in the ifp->if_transmit() routine when the ifnet pointer in the incoming snd_tag mismatches the one of the network interface. The network adapter frees the mbuf and returns EAGAIN which causes the ip_output() to release and clear the send tag. Upon next ip_output() a new "snd_tag" will be tried allocated. 4) When the PCB is detached the custom sendqueue will be released by a non-blocking ifp->if_snd_tag_free() call to the currently bound network interface. Reviewed by: wblock (manpages), adrian, gallatin, scottl (network) Differential Revision: https://reviews.freebsd.org/D3687 Sponsored by: Mellanox Technologies MFC after: 3 months
2017-01-18 13:31:17 +00:00
struct if_snd_tag_alloc_header {
uint32_t type; /* send tag type, see IF_SND_TAG_XXX */
uint32_t flowid; /* mbuf hash value */
uint32_t flowtype; /* mbuf hash type */
uint8_t numa_domain; /* numa domain of associated inp */
Implement kernel support for hardware rate limited sockets. - Add RATELIMIT kernel configuration keyword which must be set to enable the new functionality. - Add support for hardware driven, Receive Side Scaling, RSS aware, rate limited sendqueues and expose the functionality through the already established SO_MAX_PACING_RATE setsockopt(). The API support rates in the range from 1 to 4Gbytes/s which are suitable for regular TCP and UDP streams. The setsockopt(2) manual page has been updated. - Add rate limit function callback API to "struct ifnet" which supports the following operations: if_snd_tag_alloc(), if_snd_tag_modify(), if_snd_tag_query() and if_snd_tag_free(). - Add support to ifconfig to view, set and clear the IFCAP_TXRTLMT flag, which tells if a network driver supports rate limiting or not. - This patch also adds support for rate limiting through VLAN and LAGG intermediate network devices. - How rate limiting works: 1) The userspace application calls setsockopt() after accepting or making a new connection to set the rate which is then stored in the socket structure in the kernel. Later on when packets are transmitted a check is made in the transmit path for rate changes. A rate change implies a non-blocking ifp->if_snd_tag_alloc() call will be made to the destination network interface, which then sets up a custom sendqueue with the given rate limitation parameter. A "struct m_snd_tag" pointer is returned which serves as a "snd_tag" hint in the m_pkthdr for the subsequently transmitted mbufs. 2) When the network driver sees the "m->m_pkthdr.snd_tag" different from NULL, it will move the packets into a designated rate limited sendqueue given by the snd_tag pointer. It is up to the individual drivers how the rate limited traffic will be rate limited. 3) Route changes are detected by the NIC drivers in the ifp->if_transmit() routine when the ifnet pointer in the incoming snd_tag mismatches the one of the network interface. The network adapter frees the mbuf and returns EAGAIN which causes the ip_output() to release and clear the send tag. Upon next ip_output() a new "snd_tag" will be tried allocated. 4) When the PCB is detached the custom sendqueue will be released by a non-blocking ifp->if_snd_tag_free() call to the currently bound network interface. Reviewed by: wblock (manpages), adrian, gallatin, scottl (network) Differential Revision: https://reviews.freebsd.org/D3687 Sponsored by: Mellanox Technologies MFC after: 3 months
2017-01-18 13:31:17 +00:00
};
struct if_snd_tag_alloc_rate_limit {
struct if_snd_tag_alloc_header hdr;
uint64_t max_rate; /* in bytes/s */
uint32_t flags; /* M_NOWAIT or M_WAITOK */
uint32_t reserved; /* alignment */
Implement kernel support for hardware rate limited sockets. - Add RATELIMIT kernel configuration keyword which must be set to enable the new functionality. - Add support for hardware driven, Receive Side Scaling, RSS aware, rate limited sendqueues and expose the functionality through the already established SO_MAX_PACING_RATE setsockopt(). The API support rates in the range from 1 to 4Gbytes/s which are suitable for regular TCP and UDP streams. The setsockopt(2) manual page has been updated. - Add rate limit function callback API to "struct ifnet" which supports the following operations: if_snd_tag_alloc(), if_snd_tag_modify(), if_snd_tag_query() and if_snd_tag_free(). - Add support to ifconfig to view, set and clear the IFCAP_TXRTLMT flag, which tells if a network driver supports rate limiting or not. - This patch also adds support for rate limiting through VLAN and LAGG intermediate network devices. - How rate limiting works: 1) The userspace application calls setsockopt() after accepting or making a new connection to set the rate which is then stored in the socket structure in the kernel. Later on when packets are transmitted a check is made in the transmit path for rate changes. A rate change implies a non-blocking ifp->if_snd_tag_alloc() call will be made to the destination network interface, which then sets up a custom sendqueue with the given rate limitation parameter. A "struct m_snd_tag" pointer is returned which serves as a "snd_tag" hint in the m_pkthdr for the subsequently transmitted mbufs. 2) When the network driver sees the "m->m_pkthdr.snd_tag" different from NULL, it will move the packets into a designated rate limited sendqueue given by the snd_tag pointer. It is up to the individual drivers how the rate limited traffic will be rate limited. 3) Route changes are detected by the NIC drivers in the ifp->if_transmit() routine when the ifnet pointer in the incoming snd_tag mismatches the one of the network interface. The network adapter frees the mbuf and returns EAGAIN which causes the ip_output() to release and clear the send tag. Upon next ip_output() a new "snd_tag" will be tried allocated. 4) When the PCB is detached the custom sendqueue will be released by a non-blocking ifp->if_snd_tag_free() call to the currently bound network interface. Reviewed by: wblock (manpages), adrian, gallatin, scottl (network) Differential Revision: https://reviews.freebsd.org/D3687 Sponsored by: Mellanox Technologies MFC after: 3 months
2017-01-18 13:31:17 +00:00
};
Add kernel-side support for in-kernel TLS. KTLS adds support for in-kernel framing and encryption of Transport Layer Security (1.0-1.2) data on TCP sockets. KTLS only supports offload of TLS for transmitted data. Key negotation must still be performed in userland. Once completed, transmit session keys for a connection are provided to the kernel via a new TCP_TXTLS_ENABLE socket option. All subsequent data transmitted on the socket is placed into TLS frames and encrypted using the supplied keys. Any data written to a KTLS-enabled socket via write(2), aio_write(2), or sendfile(2) is assumed to be application data and is encoded in TLS frames with an application data type. Individual records can be sent with a custom type (e.g. handshake messages) via sendmsg(2) with a new control message (TLS_SET_RECORD_TYPE) specifying the record type. At present, rekeying is not supported though the in-kernel framework should support rekeying. KTLS makes use of the recently added unmapped mbufs to store TLS frames in the socket buffer. Each TLS frame is described by a single ext_pgs mbuf. The ext_pgs structure contains the header of the TLS record (and trailer for encrypted records) as well as references to the associated TLS session. KTLS supports two primary methods of encrypting TLS frames: software TLS and ifnet TLS. Software TLS marks mbufs holding socket data as not ready via M_NOTREADY similar to sendfile(2) when TLS framing information is added to an unmapped mbuf in ktls_frame(). ktls_enqueue() is then called to schedule TLS frames for encryption. In the case of sendfile_iodone() calls ktls_enqueue() instead of pru_ready() leaving the mbufs marked M_NOTREADY until encryption is completed. For other writes (vn_sendfile when pages are available, write(2), etc.), the PRUS_NOTREADY is set when invoking pru_send() along with invoking ktls_enqueue(). A pool of worker threads (the "KTLS" kernel process) encrypts TLS frames queued via ktls_enqueue(). Each TLS frame is temporarily mapped using the direct map and passed to a software encryption backend to perform the actual encryption. (Note: The use of PHYS_TO_DMAP could be replaced with sf_bufs if someone wished to make this work on architectures without a direct map.) KTLS supports pluggable software encryption backends. Internally, Netflix uses proprietary pure-software backends. This commit includes a simple backend in a new ktls_ocf.ko module that uses the kernel's OpenCrypto framework to provide AES-GCM encryption of TLS frames. As a result, software TLS is now a bit of a misnomer as it can make use of hardware crypto accelerators. Once software encryption has finished, the TLS frame mbufs are marked ready via pru_ready(). At this point, the encrypted data appears as regular payload to the TCP stack stored in unmapped mbufs. ifnet TLS permits a NIC to offload the TLS encryption and TCP segmentation. In this mode, a new send tag type (IF_SND_TAG_TYPE_TLS) is allocated on the interface a socket is routed over and associated with a TLS session. TLS records for a TLS session using ifnet TLS are not marked M_NOTREADY but are passed down the stack unencrypted. The ip_output_send() and ip6_output_send() helper functions that apply send tags to outbound IP packets verify that the send tag of the TLS record matches the outbound interface. If so, the packet is tagged with the TLS send tag and sent to the interface. The NIC device driver must recognize packets with the TLS send tag and schedule them for TLS encryption and TCP segmentation. If the the outbound interface does not match the interface in the TLS send tag, the packet is dropped. In addition, a task is scheduled to refresh the TLS send tag for the TLS session. If a new TLS send tag cannot be allocated, the connection is dropped. If a new TLS send tag is allocated, however, subsequent packets will be tagged with the correct TLS send tag. (This latter case has been tested by configuring both ports of a Chelsio T6 in a lagg and failing over from one port to another. As the connections migrated to the new port, new TLS send tags were allocated for the new port and connections resumed without being dropped.) ifnet TLS can be enabled and disabled on supported network interfaces via new '[-]txtls[46]' options to ifconfig(8). ifnet TLS is supported across both vlan devices and lagg interfaces using failover, lacp with flowid enabled, or lacp with flowid enabled. Applications may request the current KTLS mode of a connection via a new TCP_TXTLS_MODE socket option. They can also use this socket option to toggle between software and ifnet TLS modes. In addition, a testing tool is available in tools/tools/switch_tls. This is modeled on tcpdrop and uses similar syntax. However, instead of dropping connections, -s is used to force KTLS connections to switch to software TLS and -i is used to switch to ifnet TLS. Various sysctls and counters are available under the kern.ipc.tls sysctl node. The kern.ipc.tls.enable node must be set to true to enable KTLS (it is off by default). The use of unmapped mbufs must also be enabled via kern.ipc.mb_use_ext_pgs to enable KTLS. KTLS is enabled via the KERN_TLS kernel option. This patch is the culmination of years of work by several folks including Scott Long and Randall Stewart for the original design and implementation; Drew Gallatin for several optimizations including the use of ext_pgs mbufs, the M_NOTREADY mechanism for TLS records awaiting software encryption, and pluggable software crypto backends; and John Baldwin for modifications to support hardware TLS offload. Reviewed by: gallatin, hselasky, rrs Obtained from: Netflix Sponsored by: Netflix, Chelsio Communications Differential Revision: https://reviews.freebsd.org/D21277
2019-08-27 00:01:56 +00:00
struct if_snd_tag_alloc_tls {
struct if_snd_tag_alloc_header hdr;
struct inpcb *inp;
const struct ktls_session *tls;
};
Support hardware rate limiting (pacing) with TLS offload. - Add a new send tag type for a send tag that supports both rate limiting (packet pacing) and TLS offload (mostly similar to D22669 but adds a separate structure when allocating the new tag type). - When allocating a send tag for TLS offload, check to see if the connection already has a pacing rate. If so, allocate a tag that supports both rate limiting and TLS offload rather than a plain TLS offload tag. - When setting an initial rate on an existing ifnet KTLS connection, set the rate in the TCP control block inp and then reset the TLS send tag (via ktls_output_eagain) to reallocate a TLS + ratelimit send tag. This allocates the TLS send tag asynchronously from a task queue, so the TLS rate limit tag alloc is always sleepable. - When modifying a rate on a connection using KTLS, look for a TLS send tag. If the send tag is only a plain TLS send tag, assume we failed to allocate a TLS ratelimit tag (either during the TCP_TXTLS_ENABLE socket option, or during the send tag reset triggered by ktls_output_eagain) and ignore the new rate. If the send tag is a ratelimit TLS send tag, change the rate on the TLS tag and leave the inp tag alone. - Lock the inp lock when setting sb_tls_info for a socket send buffer so that the routines in tcp_ratelimit can safely dereference the pointer without needing to grab the socket buffer lock. - Add an IFCAP_TXTLS_RTLMT capability flag and associated administrative controls in ifconfig(8). TLS rate limit tags are only allocated if this capability is enabled. Note that TLS offload (whether unlimited or rate limited) always requires IFCAP_TXTLS[46]. Reviewed by: gallatin, hselasky Relnotes: yes Sponsored by: Netflix Differential Revision: https://reviews.freebsd.org/D26691
2020-10-29 00:23:16 +00:00
struct if_snd_tag_alloc_tls_rate_limit {
struct if_snd_tag_alloc_header hdr;
struct inpcb *inp;
const struct ktls_session *tls;
uint64_t max_rate; /* in bytes/s */
};
Implement kernel support for hardware rate limited sockets. - Add RATELIMIT kernel configuration keyword which must be set to enable the new functionality. - Add support for hardware driven, Receive Side Scaling, RSS aware, rate limited sendqueues and expose the functionality through the already established SO_MAX_PACING_RATE setsockopt(). The API support rates in the range from 1 to 4Gbytes/s which are suitable for regular TCP and UDP streams. The setsockopt(2) manual page has been updated. - Add rate limit function callback API to "struct ifnet" which supports the following operations: if_snd_tag_alloc(), if_snd_tag_modify(), if_snd_tag_query() and if_snd_tag_free(). - Add support to ifconfig to view, set and clear the IFCAP_TXRTLMT flag, which tells if a network driver supports rate limiting or not. - This patch also adds support for rate limiting through VLAN and LAGG intermediate network devices. - How rate limiting works: 1) The userspace application calls setsockopt() after accepting or making a new connection to set the rate which is then stored in the socket structure in the kernel. Later on when packets are transmitted a check is made in the transmit path for rate changes. A rate change implies a non-blocking ifp->if_snd_tag_alloc() call will be made to the destination network interface, which then sets up a custom sendqueue with the given rate limitation parameter. A "struct m_snd_tag" pointer is returned which serves as a "snd_tag" hint in the m_pkthdr for the subsequently transmitted mbufs. 2) When the network driver sees the "m->m_pkthdr.snd_tag" different from NULL, it will move the packets into a designated rate limited sendqueue given by the snd_tag pointer. It is up to the individual drivers how the rate limited traffic will be rate limited. 3) Route changes are detected by the NIC drivers in the ifp->if_transmit() routine when the ifnet pointer in the incoming snd_tag mismatches the one of the network interface. The network adapter frees the mbuf and returns EAGAIN which causes the ip_output() to release and clear the send tag. Upon next ip_output() a new "snd_tag" will be tried allocated. 4) When the PCB is detached the custom sendqueue will be released by a non-blocking ifp->if_snd_tag_free() call to the currently bound network interface. Reviewed by: wblock (manpages), adrian, gallatin, scottl (network) Differential Revision: https://reviews.freebsd.org/D3687 Sponsored by: Mellanox Technologies MFC after: 3 months
2017-01-18 13:31:17 +00:00
struct if_snd_tag_rate_limit_params {
uint64_t max_rate; /* in bytes/s */
uint32_t queue_level; /* 0 (empty) .. 65535 (full) */
#define IF_SND_QUEUE_LEVEL_MIN 0
#define IF_SND_QUEUE_LEVEL_MAX 65535
uint32_t flags; /* M_NOWAIT or M_WAITOK */
Implement kernel support for hardware rate limited sockets. - Add RATELIMIT kernel configuration keyword which must be set to enable the new functionality. - Add support for hardware driven, Receive Side Scaling, RSS aware, rate limited sendqueues and expose the functionality through the already established SO_MAX_PACING_RATE setsockopt(). The API support rates in the range from 1 to 4Gbytes/s which are suitable for regular TCP and UDP streams. The setsockopt(2) manual page has been updated. - Add rate limit function callback API to "struct ifnet" which supports the following operations: if_snd_tag_alloc(), if_snd_tag_modify(), if_snd_tag_query() and if_snd_tag_free(). - Add support to ifconfig to view, set and clear the IFCAP_TXRTLMT flag, which tells if a network driver supports rate limiting or not. - This patch also adds support for rate limiting through VLAN and LAGG intermediate network devices. - How rate limiting works: 1) The userspace application calls setsockopt() after accepting or making a new connection to set the rate which is then stored in the socket structure in the kernel. Later on when packets are transmitted a check is made in the transmit path for rate changes. A rate change implies a non-blocking ifp->if_snd_tag_alloc() call will be made to the destination network interface, which then sets up a custom sendqueue with the given rate limitation parameter. A "struct m_snd_tag" pointer is returned which serves as a "snd_tag" hint in the m_pkthdr for the subsequently transmitted mbufs. 2) When the network driver sees the "m->m_pkthdr.snd_tag" different from NULL, it will move the packets into a designated rate limited sendqueue given by the snd_tag pointer. It is up to the individual drivers how the rate limited traffic will be rate limited. 3) Route changes are detected by the NIC drivers in the ifp->if_transmit() routine when the ifnet pointer in the incoming snd_tag mismatches the one of the network interface. The network adapter frees the mbuf and returns EAGAIN which causes the ip_output() to release and clear the send tag. Upon next ip_output() a new "snd_tag" will be tried allocated. 4) When the PCB is detached the custom sendqueue will be released by a non-blocking ifp->if_snd_tag_free() call to the currently bound network interface. Reviewed by: wblock (manpages), adrian, gallatin, scottl (network) Differential Revision: https://reviews.freebsd.org/D3687 Sponsored by: Mellanox Technologies MFC after: 3 months
2017-01-18 13:31:17 +00:00
};
union if_snd_tag_alloc_params {
struct if_snd_tag_alloc_header hdr;
struct if_snd_tag_alloc_rate_limit rate_limit;
struct if_snd_tag_alloc_rate_limit unlimited;
Add kernel-side support for in-kernel TLS. KTLS adds support for in-kernel framing and encryption of Transport Layer Security (1.0-1.2) data on TCP sockets. KTLS only supports offload of TLS for transmitted data. Key negotation must still be performed in userland. Once completed, transmit session keys for a connection are provided to the kernel via a new TCP_TXTLS_ENABLE socket option. All subsequent data transmitted on the socket is placed into TLS frames and encrypted using the supplied keys. Any data written to a KTLS-enabled socket via write(2), aio_write(2), or sendfile(2) is assumed to be application data and is encoded in TLS frames with an application data type. Individual records can be sent with a custom type (e.g. handshake messages) via sendmsg(2) with a new control message (TLS_SET_RECORD_TYPE) specifying the record type. At present, rekeying is not supported though the in-kernel framework should support rekeying. KTLS makes use of the recently added unmapped mbufs to store TLS frames in the socket buffer. Each TLS frame is described by a single ext_pgs mbuf. The ext_pgs structure contains the header of the TLS record (and trailer for encrypted records) as well as references to the associated TLS session. KTLS supports two primary methods of encrypting TLS frames: software TLS and ifnet TLS. Software TLS marks mbufs holding socket data as not ready via M_NOTREADY similar to sendfile(2) when TLS framing information is added to an unmapped mbuf in ktls_frame(). ktls_enqueue() is then called to schedule TLS frames for encryption. In the case of sendfile_iodone() calls ktls_enqueue() instead of pru_ready() leaving the mbufs marked M_NOTREADY until encryption is completed. For other writes (vn_sendfile when pages are available, write(2), etc.), the PRUS_NOTREADY is set when invoking pru_send() along with invoking ktls_enqueue(). A pool of worker threads (the "KTLS" kernel process) encrypts TLS frames queued via ktls_enqueue(). Each TLS frame is temporarily mapped using the direct map and passed to a software encryption backend to perform the actual encryption. (Note: The use of PHYS_TO_DMAP could be replaced with sf_bufs if someone wished to make this work on architectures without a direct map.) KTLS supports pluggable software encryption backends. Internally, Netflix uses proprietary pure-software backends. This commit includes a simple backend in a new ktls_ocf.ko module that uses the kernel's OpenCrypto framework to provide AES-GCM encryption of TLS frames. As a result, software TLS is now a bit of a misnomer as it can make use of hardware crypto accelerators. Once software encryption has finished, the TLS frame mbufs are marked ready via pru_ready(). At this point, the encrypted data appears as regular payload to the TCP stack stored in unmapped mbufs. ifnet TLS permits a NIC to offload the TLS encryption and TCP segmentation. In this mode, a new send tag type (IF_SND_TAG_TYPE_TLS) is allocated on the interface a socket is routed over and associated with a TLS session. TLS records for a TLS session using ifnet TLS are not marked M_NOTREADY but are passed down the stack unencrypted. The ip_output_send() and ip6_output_send() helper functions that apply send tags to outbound IP packets verify that the send tag of the TLS record matches the outbound interface. If so, the packet is tagged with the TLS send tag and sent to the interface. The NIC device driver must recognize packets with the TLS send tag and schedule them for TLS encryption and TCP segmentation. If the the outbound interface does not match the interface in the TLS send tag, the packet is dropped. In addition, a task is scheduled to refresh the TLS send tag for the TLS session. If a new TLS send tag cannot be allocated, the connection is dropped. If a new TLS send tag is allocated, however, subsequent packets will be tagged with the correct TLS send tag. (This latter case has been tested by configuring both ports of a Chelsio T6 in a lagg and failing over from one port to another. As the connections migrated to the new port, new TLS send tags were allocated for the new port and connections resumed without being dropped.) ifnet TLS can be enabled and disabled on supported network interfaces via new '[-]txtls[46]' options to ifconfig(8). ifnet TLS is supported across both vlan devices and lagg interfaces using failover, lacp with flowid enabled, or lacp with flowid enabled. Applications may request the current KTLS mode of a connection via a new TCP_TXTLS_MODE socket option. They can also use this socket option to toggle between software and ifnet TLS modes. In addition, a testing tool is available in tools/tools/switch_tls. This is modeled on tcpdrop and uses similar syntax. However, instead of dropping connections, -s is used to force KTLS connections to switch to software TLS and -i is used to switch to ifnet TLS. Various sysctls and counters are available under the kern.ipc.tls sysctl node. The kern.ipc.tls.enable node must be set to true to enable KTLS (it is off by default). The use of unmapped mbufs must also be enabled via kern.ipc.mb_use_ext_pgs to enable KTLS. KTLS is enabled via the KERN_TLS kernel option. This patch is the culmination of years of work by several folks including Scott Long and Randall Stewart for the original design and implementation; Drew Gallatin for several optimizations including the use of ext_pgs mbufs, the M_NOTREADY mechanism for TLS records awaiting software encryption, and pluggable software crypto backends; and John Baldwin for modifications to support hardware TLS offload. Reviewed by: gallatin, hselasky, rrs Obtained from: Netflix Sponsored by: Netflix, Chelsio Communications Differential Revision: https://reviews.freebsd.org/D21277
2019-08-27 00:01:56 +00:00
struct if_snd_tag_alloc_tls tls;
Support hardware rate limiting (pacing) with TLS offload. - Add a new send tag type for a send tag that supports both rate limiting (packet pacing) and TLS offload (mostly similar to D22669 but adds a separate structure when allocating the new tag type). - When allocating a send tag for TLS offload, check to see if the connection already has a pacing rate. If so, allocate a tag that supports both rate limiting and TLS offload rather than a plain TLS offload tag. - When setting an initial rate on an existing ifnet KTLS connection, set the rate in the TCP control block inp and then reset the TLS send tag (via ktls_output_eagain) to reallocate a TLS + ratelimit send tag. This allocates the TLS send tag asynchronously from a task queue, so the TLS rate limit tag alloc is always sleepable. - When modifying a rate on a connection using KTLS, look for a TLS send tag. If the send tag is only a plain TLS send tag, assume we failed to allocate a TLS ratelimit tag (either during the TCP_TXTLS_ENABLE socket option, or during the send tag reset triggered by ktls_output_eagain) and ignore the new rate. If the send tag is a ratelimit TLS send tag, change the rate on the TLS tag and leave the inp tag alone. - Lock the inp lock when setting sb_tls_info for a socket send buffer so that the routines in tcp_ratelimit can safely dereference the pointer without needing to grab the socket buffer lock. - Add an IFCAP_TXTLS_RTLMT capability flag and associated administrative controls in ifconfig(8). TLS rate limit tags are only allocated if this capability is enabled. Note that TLS offload (whether unlimited or rate limited) always requires IFCAP_TXTLS[46]. Reviewed by: gallatin, hselasky Relnotes: yes Sponsored by: Netflix Differential Revision: https://reviews.freebsd.org/D26691
2020-10-29 00:23:16 +00:00
struct if_snd_tag_alloc_tls_rate_limit tls_rate_limit;
Implement kernel support for hardware rate limited sockets. - Add RATELIMIT kernel configuration keyword which must be set to enable the new functionality. - Add support for hardware driven, Receive Side Scaling, RSS aware, rate limited sendqueues and expose the functionality through the already established SO_MAX_PACING_RATE setsockopt(). The API support rates in the range from 1 to 4Gbytes/s which are suitable for regular TCP and UDP streams. The setsockopt(2) manual page has been updated. - Add rate limit function callback API to "struct ifnet" which supports the following operations: if_snd_tag_alloc(), if_snd_tag_modify(), if_snd_tag_query() and if_snd_tag_free(). - Add support to ifconfig to view, set and clear the IFCAP_TXRTLMT flag, which tells if a network driver supports rate limiting or not. - This patch also adds support for rate limiting through VLAN and LAGG intermediate network devices. - How rate limiting works: 1) The userspace application calls setsockopt() after accepting or making a new connection to set the rate which is then stored in the socket structure in the kernel. Later on when packets are transmitted a check is made in the transmit path for rate changes. A rate change implies a non-blocking ifp->if_snd_tag_alloc() call will be made to the destination network interface, which then sets up a custom sendqueue with the given rate limitation parameter. A "struct m_snd_tag" pointer is returned which serves as a "snd_tag" hint in the m_pkthdr for the subsequently transmitted mbufs. 2) When the network driver sees the "m->m_pkthdr.snd_tag" different from NULL, it will move the packets into a designated rate limited sendqueue given by the snd_tag pointer. It is up to the individual drivers how the rate limited traffic will be rate limited. 3) Route changes are detected by the NIC drivers in the ifp->if_transmit() routine when the ifnet pointer in the incoming snd_tag mismatches the one of the network interface. The network adapter frees the mbuf and returns EAGAIN which causes the ip_output() to release and clear the send tag. Upon next ip_output() a new "snd_tag" will be tried allocated. 4) When the PCB is detached the custom sendqueue will be released by a non-blocking ifp->if_snd_tag_free() call to the currently bound network interface. Reviewed by: wblock (manpages), adrian, gallatin, scottl (network) Differential Revision: https://reviews.freebsd.org/D3687 Sponsored by: Mellanox Technologies MFC after: 3 months
2017-01-18 13:31:17 +00:00
};
union if_snd_tag_modify_params {
struct if_snd_tag_rate_limit_params rate_limit;
struct if_snd_tag_rate_limit_params unlimited;
Support hardware rate limiting (pacing) with TLS offload. - Add a new send tag type for a send tag that supports both rate limiting (packet pacing) and TLS offload (mostly similar to D22669 but adds a separate structure when allocating the new tag type). - When allocating a send tag for TLS offload, check to see if the connection already has a pacing rate. If so, allocate a tag that supports both rate limiting and TLS offload rather than a plain TLS offload tag. - When setting an initial rate on an existing ifnet KTLS connection, set the rate in the TCP control block inp and then reset the TLS send tag (via ktls_output_eagain) to reallocate a TLS + ratelimit send tag. This allocates the TLS send tag asynchronously from a task queue, so the TLS rate limit tag alloc is always sleepable. - When modifying a rate on a connection using KTLS, look for a TLS send tag. If the send tag is only a plain TLS send tag, assume we failed to allocate a TLS ratelimit tag (either during the TCP_TXTLS_ENABLE socket option, or during the send tag reset triggered by ktls_output_eagain) and ignore the new rate. If the send tag is a ratelimit TLS send tag, change the rate on the TLS tag and leave the inp tag alone. - Lock the inp lock when setting sb_tls_info for a socket send buffer so that the routines in tcp_ratelimit can safely dereference the pointer without needing to grab the socket buffer lock. - Add an IFCAP_TXTLS_RTLMT capability flag and associated administrative controls in ifconfig(8). TLS rate limit tags are only allocated if this capability is enabled. Note that TLS offload (whether unlimited or rate limited) always requires IFCAP_TXTLS[46]. Reviewed by: gallatin, hselasky Relnotes: yes Sponsored by: Netflix Differential Revision: https://reviews.freebsd.org/D26691
2020-10-29 00:23:16 +00:00
struct if_snd_tag_rate_limit_params tls_rate_limit;
Implement kernel support for hardware rate limited sockets. - Add RATELIMIT kernel configuration keyword which must be set to enable the new functionality. - Add support for hardware driven, Receive Side Scaling, RSS aware, rate limited sendqueues and expose the functionality through the already established SO_MAX_PACING_RATE setsockopt(). The API support rates in the range from 1 to 4Gbytes/s which are suitable for regular TCP and UDP streams. The setsockopt(2) manual page has been updated. - Add rate limit function callback API to "struct ifnet" which supports the following operations: if_snd_tag_alloc(), if_snd_tag_modify(), if_snd_tag_query() and if_snd_tag_free(). - Add support to ifconfig to view, set and clear the IFCAP_TXRTLMT flag, which tells if a network driver supports rate limiting or not. - This patch also adds support for rate limiting through VLAN and LAGG intermediate network devices. - How rate limiting works: 1) The userspace application calls setsockopt() after accepting or making a new connection to set the rate which is then stored in the socket structure in the kernel. Later on when packets are transmitted a check is made in the transmit path for rate changes. A rate change implies a non-blocking ifp->if_snd_tag_alloc() call will be made to the destination network interface, which then sets up a custom sendqueue with the given rate limitation parameter. A "struct m_snd_tag" pointer is returned which serves as a "snd_tag" hint in the m_pkthdr for the subsequently transmitted mbufs. 2) When the network driver sees the "m->m_pkthdr.snd_tag" different from NULL, it will move the packets into a designated rate limited sendqueue given by the snd_tag pointer. It is up to the individual drivers how the rate limited traffic will be rate limited. 3) Route changes are detected by the NIC drivers in the ifp->if_transmit() routine when the ifnet pointer in the incoming snd_tag mismatches the one of the network interface. The network adapter frees the mbuf and returns EAGAIN which causes the ip_output() to release and clear the send tag. Upon next ip_output() a new "snd_tag" will be tried allocated. 4) When the PCB is detached the custom sendqueue will be released by a non-blocking ifp->if_snd_tag_free() call to the currently bound network interface. Reviewed by: wblock (manpages), adrian, gallatin, scottl (network) Differential Revision: https://reviews.freebsd.org/D3687 Sponsored by: Mellanox Technologies MFC after: 3 months
2017-01-18 13:31:17 +00:00
};
union if_snd_tag_query_params {
struct if_snd_tag_rate_limit_params rate_limit;
struct if_snd_tag_rate_limit_params unlimited;
Support hardware rate limiting (pacing) with TLS offload. - Add a new send tag type for a send tag that supports both rate limiting (packet pacing) and TLS offload (mostly similar to D22669 but adds a separate structure when allocating the new tag type). - When allocating a send tag for TLS offload, check to see if the connection already has a pacing rate. If so, allocate a tag that supports both rate limiting and TLS offload rather than a plain TLS offload tag. - When setting an initial rate on an existing ifnet KTLS connection, set the rate in the TCP control block inp and then reset the TLS send tag (via ktls_output_eagain) to reallocate a TLS + ratelimit send tag. This allocates the TLS send tag asynchronously from a task queue, so the TLS rate limit tag alloc is always sleepable. - When modifying a rate on a connection using KTLS, look for a TLS send tag. If the send tag is only a plain TLS send tag, assume we failed to allocate a TLS ratelimit tag (either during the TCP_TXTLS_ENABLE socket option, or during the send tag reset triggered by ktls_output_eagain) and ignore the new rate. If the send tag is a ratelimit TLS send tag, change the rate on the TLS tag and leave the inp tag alone. - Lock the inp lock when setting sb_tls_info for a socket send buffer so that the routines in tcp_ratelimit can safely dereference the pointer without needing to grab the socket buffer lock. - Add an IFCAP_TXTLS_RTLMT capability flag and associated administrative controls in ifconfig(8). TLS rate limit tags are only allocated if this capability is enabled. Note that TLS offload (whether unlimited or rate limited) always requires IFCAP_TXTLS[46]. Reviewed by: gallatin, hselasky Relnotes: yes Sponsored by: Netflix Differential Revision: https://reviews.freebsd.org/D26691
2020-10-29 00:23:16 +00:00
struct if_snd_tag_rate_limit_params tls_rate_limit;
Implement kernel support for hardware rate limited sockets. - Add RATELIMIT kernel configuration keyword which must be set to enable the new functionality. - Add support for hardware driven, Receive Side Scaling, RSS aware, rate limited sendqueues and expose the functionality through the already established SO_MAX_PACING_RATE setsockopt(). The API support rates in the range from 1 to 4Gbytes/s which are suitable for regular TCP and UDP streams. The setsockopt(2) manual page has been updated. - Add rate limit function callback API to "struct ifnet" which supports the following operations: if_snd_tag_alloc(), if_snd_tag_modify(), if_snd_tag_query() and if_snd_tag_free(). - Add support to ifconfig to view, set and clear the IFCAP_TXRTLMT flag, which tells if a network driver supports rate limiting or not. - This patch also adds support for rate limiting through VLAN and LAGG intermediate network devices. - How rate limiting works: 1) The userspace application calls setsockopt() after accepting or making a new connection to set the rate which is then stored in the socket structure in the kernel. Later on when packets are transmitted a check is made in the transmit path for rate changes. A rate change implies a non-blocking ifp->if_snd_tag_alloc() call will be made to the destination network interface, which then sets up a custom sendqueue with the given rate limitation parameter. A "struct m_snd_tag" pointer is returned which serves as a "snd_tag" hint in the m_pkthdr for the subsequently transmitted mbufs. 2) When the network driver sees the "m->m_pkthdr.snd_tag" different from NULL, it will move the packets into a designated rate limited sendqueue given by the snd_tag pointer. It is up to the individual drivers how the rate limited traffic will be rate limited. 3) Route changes are detected by the NIC drivers in the ifp->if_transmit() routine when the ifnet pointer in the incoming snd_tag mismatches the one of the network interface. The network adapter frees the mbuf and returns EAGAIN which causes the ip_output() to release and clear the send tag. Upon next ip_output() a new "snd_tag" will be tried allocated. 4) When the PCB is detached the custom sendqueue will be released by a non-blocking ifp->if_snd_tag_free() call to the currently bound network interface. Reviewed by: wblock (manpages), adrian, gallatin, scottl (network) Differential Revision: https://reviews.freebsd.org/D3687 Sponsored by: Mellanox Technologies MFC after: 3 months
2017-01-18 13:31:17 +00:00
};
typedef int (if_snd_tag_alloc_t)(struct ifnet *, union if_snd_tag_alloc_params *,
struct m_snd_tag **);
typedef int (if_snd_tag_modify_t)(struct m_snd_tag *, union if_snd_tag_modify_params *);
typedef int (if_snd_tag_query_t)(struct m_snd_tag *, union if_snd_tag_query_params *);
typedef void (if_snd_tag_free_t)(struct m_snd_tag *);
typedef struct m_snd_tag *(if_next_send_tag_t)(struct m_snd_tag *);
struct if_snd_tag_sw {
if_snd_tag_modify_t *snd_tag_modify;
if_snd_tag_query_t *snd_tag_query;
if_snd_tag_free_t *snd_tag_free;
if_next_send_tag_t *next_snd_tag;
u_int type; /* One of IF_SND_TAG_TYPE_*. */
};
/* Query return flags */
#define RT_NOSUPPORT 0x00000000 /* Not supported */
#define RT_IS_INDIRECT 0x00000001 /*
* Interface like a lagg, select
* the actual interface for
* capabilities.
*/
#define RT_IS_SELECTABLE 0x00000002 /*
* No rate table, you select
* rates and the first
* number_of_rates are created.
*/
#define RT_IS_FIXED_TABLE 0x00000004 /* A fixed table is attached */
#define RT_IS_UNUSABLE 0x00000008 /* It is not usable for this */
#define RT_IS_SETUP_REQ 0x00000010 /* The interface setup must be called before use */
struct if_ratelimit_query_results {
const uint64_t *rate_table; /* Pointer to table if present */
uint32_t flags; /* Flags indicating results */
uint32_t max_flows; /* Max flows using, 0=unlimited */
uint32_t number_of_rates; /* How many unique rates can be created */
uint32_t min_segment_burst; /* The amount the adapter bursts at each send */
};
typedef void (if_ratelimit_query_t)(struct ifnet *,
struct if_ratelimit_query_results *);
typedef int (if_ratelimit_setup_t)(struct ifnet *, uint64_t, uint32_t);
Implement interface link header precomputation API. Add if_requestencap() interface method which is capable of calculating various link headers for given interface. Right now there is support for INET/INET6/ARP llheader calculation (IFENCAP_LL type request). Other types are planned to support more complex calculation (L2 multipath lagg nexthops, tunnel encap nexthops, etc..). Reshape 'struct route' to be able to pass additional data (with is length) to prepend to mbuf. These two changes permits routing code to pass pre-calculated nexthop data (like L2 header for route w/gateway) down to the stack eliminating the need for other lookups. It also brings us closer to more complex scenarios like transparently handling MPLS nexthops and tunnel interfaces. Last, but not least, it removes layering violation introduced by flowtable code (ro_lle) and simplifies handling of existing if_output consumers. ARP/ND changes: Make arp/ndp stack pre-calculate link header upon installing/updating lle record. Interface link address change are handled by re-calculating headers for all lles based on if_lladdr event. After these changes, arpresolve()/nd6_resolve() returns full pre-calculated header for supported interfaces thus simplifying if_output(). Move these lookups to separate ether_resolve_addr() function which ether returs error or fully-prepared link header. Add <arp|nd6_>resolve_addr() compat versions to return link addresses instead of pre-calculated data. BPF changes: Raw bpf writes occupied _two_ cases: AF_UNSPEC and pseudo_AF_HDRCMPLT. Despite the naming, both of there have ther header "complete". The only difference is that interface source mac has to be filled by OS for AF_UNSPEC (controlled via BIOCGHDRCMPLT). This logic has to stay inside BPF and not pollute if_output() routines. Convert BPF to pass prepend data via new 'struct route' mechanism. Note that it does not change non-optimized if_output(): ro_prepend handling is purely optional. Side note: hackish pseudo_AF_HDRCMPLT is supported for ethernet and FDDI. It is not needed for ethernet anymore. The only remaining FDDI user is dev/pdq mostly untouched since 2007. FDDI support was eliminated from OpenBSD in 2013 (sys/net/if_fddisubr.c rev 1.65). Flowtable changes: Flowtable violates layering by saving (and not correctly managing) rtes/lles. Instead of passing lle pointer, pass pointer to pre-calculated header data from that lle. Differential Revision: https://reviews.freebsd.org/D4102
2015-12-31 05:03:27 +00:00
/*
* Structure defining a network interface.
*/
struct ifnet {
/* General book keeping of interface lists. */
CK_STAILQ_ENTRY(ifnet) if_link; /* all struct ifnets are chained (CK_) */
LIST_ENTRY(ifnet) if_clones; /* interfaces of a cloner */
CK_STAILQ_HEAD(, ifg_list) if_groups; /* linked list of groups per if (CK_) */
/* protected by if_addr_lock */
u_char if_alloctype; /* if_type at time of allocation */
uint8_t if_numa_domain; /* NUMA domain of device */
/* Driver and protocol specific information that remains stable. */
void *if_softc; /* pointer to driver state */
void *if_llsoftc; /* link layer softc */
void *if_l2com; /* pointer to protocol bits */
const char *if_dname; /* driver name */
int if_dunit; /* unit or IF_DUNIT_NONE */
u_short if_index; /* numeric abbreviation for this if */
short if_index_reserved; /* spare space to grow if_index */
char if_xname[IFNAMSIZ]; /* external name (name + unit) */
char *if_description; /* interface description */
/* Variable fields that are touched by the stack and drivers. */
int if_flags; /* up/down, broadcast, etc. */
int if_drv_flags; /* driver-managed status flags */
int if_capabilities; /* interface features & capabilities */
int if_capenable; /* enabled features & capabilities */
void *if_linkmib; /* link-type-specific MIB data */
size_t if_linkmiblen; /* length of above data */
Start to address a number of races relating to use of ifnet pointers after the corresponding interface has been destroyed: (1) Add an ifnet refcount, ifp->if_refcount. Initialize it to 1 in if_alloc(), and modify if_free_type() to decrement and check the refcount. (2) Add new if_ref() and if_rele() interfaces to allow kernel code walking global interface lists to release IFNET_[RW]LOCK() yet keep the ifnet stable. Currently, if_rele() is a no-op wrapper around if_free(), but this may change in the future. (3) Add new ifnet field, if_alloctype, which caches the type passed to if_alloc(), but unlike if_type, won't be changed by drivers. This allows asynchronous free's of the interface after the driver has released it to still use the right type. Use that instead of the type passed to if_free_type(), but assert that they are the same (might have to rethink this if that doesn't work out). (4) Add a new ifnet_byindex_ref(), which looks up an interface by index and returns a reference rather than a pointer to it. (5) Fix if_alloc() to fully initialize the if_addr_mtx before hooking up the ifnet to global lists. (6) Modify sysctls in if_mib.c to use ifnet_byindex_ref() and release the ifnet when done. When this change is MFC'd, it will need to replace if_ispare fields rather than adding new fields in order to avoid breaking the binary interface. Once this change is MFC'd, if_free_type() should be removed, as its 'type' argument is now optional. This refcount is not appropriate for counting mbuf pkthdr references, and also not for counting entry into the device driver via ifnet function pointers. An rmlock may be appropriate for the latter. Rather, this is about ensuring data structure stability when reaching an ifnet via global ifnet lists and tables followed by copy in or out of userspace. MFC after: 3 weeks Reported by: mdtancsa Reviewed by: brooks
2009-04-21 22:43:32 +00:00
u_int if_refcount; /* reference count */
/* These fields are shared with struct if_data. */
uint8_t if_type; /* ethernet, tokenring, etc */
uint8_t if_addrlen; /* media address length */
uint8_t if_hdrlen; /* media header length */
uint8_t if_link_state; /* current link state */
uint32_t if_mtu; /* maximum transmission unit */
uint32_t if_metric; /* routing metric (external only) */
uint64_t if_baudrate; /* linespeed */
uint64_t if_hwassist; /* HW offload capabilities, see IFCAP */
time_t if_epoch; /* uptime at attach or stat reset */
struct timeval if_lastchange; /* time of last administrative change */
struct ifaltq if_snd; /* output queue (includes altq) */
struct task if_linktask; /* task for link change events */
struct task if_addmultitask; /* task for SIOCADDMULTI */
/* Addresses of different protocol families assigned to this if. */
ifnet: Replace if_addr_lock rwlock with epoch + mutex Run on LLNW canaries and tested by pho@ gallatin: Using a 14-core, 28-HTT single socket E5-2697 v3 with a 40GbE MLX5 based ConnectX 4-LX NIC, I see an almost 12% improvement in received packet rate, and a larger improvement in bytes delivered all the way to userspace. When the host receiving 64 streams of netperf -H $DUT -t UDP_STREAM -- -m 1, I see, using nstat -I mce0 1 before the patch: InMpps OMpps InGbs OGbs err TCP Est %CPU syscalls csw irq GBfree 4.98 0.00 4.42 0.00 4235592 33 83.80 4720653 2149771 1235 247.32 4.73 0.00 4.20 0.00 4025260 33 82.99 4724900 2139833 1204 247.32 4.72 0.00 4.20 0.00 4035252 33 82.14 4719162 2132023 1264 247.32 4.71 0.00 4.21 0.00 4073206 33 83.68 4744973 2123317 1347 247.32 4.72 0.00 4.21 0.00 4061118 33 80.82 4713615 2188091 1490 247.32 4.72 0.00 4.21 0.00 4051675 33 85.29 4727399 2109011 1205 247.32 4.73 0.00 4.21 0.00 4039056 33 84.65 4724735 2102603 1053 247.32 After the patch InMpps OMpps InGbs OGbs err TCP Est %CPU syscalls csw irq GBfree 5.43 0.00 4.20 0.00 3313143 33 84.96 5434214 1900162 2656 245.51 5.43 0.00 4.20 0.00 3308527 33 85.24 5439695 1809382 2521 245.51 5.42 0.00 4.19 0.00 3316778 33 87.54 5416028 1805835 2256 245.51 5.42 0.00 4.19 0.00 3317673 33 90.44 5426044 1763056 2332 245.51 5.42 0.00 4.19 0.00 3314839 33 88.11 5435732 1792218 2499 245.52 5.44 0.00 4.19 0.00 3293228 33 91.84 5426301 1668597 2121 245.52 Similarly, netperf reports 230Mb/s before the patch, and 270Mb/s after the patch Reviewed by: gallatin Sponsored by: Limelight Networks Differential Revision: https://reviews.freebsd.org/D15366
2018-05-18 20:13:34 +00:00
struct mtx if_addr_lock; /* lock to protect address lists */
/*
* if_addrhead is the list of all addresses associated to
* an interface.
* Some code in the kernel assumes that first element
* of the list has type AF_LINK, and contains sockaddr_dl
* addresses which store the link-level address and the name
* of the interface.
* However, access to the AF_LINK address through this
* field is deprecated. Use if_addr or ifaddr_byindex() instead.
*/
struct ifaddrhead if_addrhead; /* linked list of addresses per if */
struct ifmultihead if_multiaddrs; /* multicast addresses configured */
int if_amcount; /* number of all-multicast requests */
struct ifaddr *if_addr; /* pointer to link-level address */
void *if_hw_addr; /* hardware link-level address */
const u_int8_t *if_broadcastaddr; /* linklevel broadcast bytestring */
struct mtx if_afdata_lock;
void *if_afdata[AF_MAX];
int if_afdata_initialized;
/* Additional features hung off the interface. */
u_int if_fib; /* interface FIB */
struct vnet *if_vnet; /* pointer to network stack instance */
struct vnet *if_home_vnet; /* where this ifnet originates from */
struct ifvlantrunk *if_vlantrunk; /* pointer to 802.1q data */
struct bpf_if *if_bpf; /* packet filter structure */
int if_pcount; /* number of promiscuous listeners */
void *if_bridge; /* bridge glue */
void *if_lagg; /* lagg glue */
void *if_pf_kif; /* pf glue */
struct carp_if *if_carp; /* carp interface structure */
struct label *if_label; /* interface MAC label */
struct netmap_adapter *if_netmap; /* netmap(4) softc */
/* Various procedures of the layer2 encapsulation and drivers. */
int (*if_output) /* output routine (enqueue) */
(struct ifnet *, struct mbuf *, const struct sockaddr *,
struct route *);
void (*if_input) /* input routine (from h/w driver) */
(struct ifnet *, struct mbuf *);
struct mbuf *(*if_bridge_input)(struct ifnet *, struct mbuf *);
int (*if_bridge_output)(struct ifnet *, struct mbuf *, struct sockaddr *,
struct rtentry *);
void (*if_bridge_linkstate)(struct ifnet *ifp);
Introduce a procedural interface to the ifnet structure. The new interface allows the ifnet structure to be defined as an opaque type in NIC drivers. This then allows the ifnet structure to be changed without a need to change or recompile NIC drivers. Put differently, NIC drivers can be written and compiled once and be used with different network stack implementations, provided of course that those network stack implementations have an API and ABI compatible interface. This commit introduces the 'if_t' type to replace 'struct ifnet *' as the type of a network interface. The 'if_t' type is defined as 'void *' to enable the compiler to perform type conversion to 'struct ifnet *' and vice versa where needed and without warnings. The functions that implement the API are the only functions that need to have an explicit cast. The MII code has been converted to use the driver API to avoid unnecessary code churn. Code churn comes from having to work with both converted and unconverted drivers in correlation with having callback functions that take an interface. By converting the MII code first, the callback functions can be defined so that the compiler will perform the typecasts automatically. As soon as all drivers have been converted, the if_t type can be redefined as needed and the API functions can be fix to not need an explicit cast. The immediate benefactors of this change are: 1. Juniper Networks - The network stack implementation in Junos is entirely different from FreeBSD's one and this change allows Juniper to build "stock" NIC drivers that can be used in combination with both the FreeBSD and Junos stacks. 2. FreeBSD - This change opens the door towards changing ifnet and implementing new features and optimizations in the network stack without it requiring a change in the many NIC drivers FreeBSD has. Submitted by: Anuranjan Shukla <anshukla@juniper.net> Reviewed by: glebius@ Obtained from: Juniper Networks, Inc.
2014-06-02 17:54:39 +00:00
if_start_fn_t if_start; /* initiate output routine */
if_ioctl_fn_t if_ioctl; /* ioctl routine */
if_init_fn_t if_init; /* Init routine */
int (*if_resolvemulti) /* validate/resolve multicast */
2002-03-19 21:54:18 +00:00
(struct ifnet *, struct sockaddr **, struct sockaddr *);
if_qflush_fn_t if_qflush; /* flush any queue */
Introduce a procedural interface to the ifnet structure. The new interface allows the ifnet structure to be defined as an opaque type in NIC drivers. This then allows the ifnet structure to be changed without a need to change or recompile NIC drivers. Put differently, NIC drivers can be written and compiled once and be used with different network stack implementations, provided of course that those network stack implementations have an API and ABI compatible interface. This commit introduces the 'if_t' type to replace 'struct ifnet *' as the type of a network interface. The 'if_t' type is defined as 'void *' to enable the compiler to perform type conversion to 'struct ifnet *' and vice versa where needed and without warnings. The functions that implement the API are the only functions that need to have an explicit cast. The MII code has been converted to use the driver API to avoid unnecessary code churn. Code churn comes from having to work with both converted and unconverted drivers in correlation with having callback functions that take an interface. By converting the MII code first, the callback functions can be defined so that the compiler will perform the typecasts automatically. As soon as all drivers have been converted, the if_t type can be redefined as needed and the API functions can be fix to not need an explicit cast. The immediate benefactors of this change are: 1. Juniper Networks - The network stack implementation in Junos is entirely different from FreeBSD's one and this change allows Juniper to build "stock" NIC drivers that can be used in combination with both the FreeBSD and Junos stacks. 2. FreeBSD - This change opens the door towards changing ifnet and implementing new features and optimizations in the network stack without it requiring a change in the many NIC drivers FreeBSD has. Submitted by: Anuranjan Shukla <anshukla@juniper.net> Reviewed by: glebius@ Obtained from: Juniper Networks, Inc.
2014-06-02 17:54:39 +00:00
if_transmit_fn_t if_transmit; /* initiate output routine */
void (*if_reassign) /* reassign to vnet routine */
(struct ifnet *, struct vnet *, char *);
if_get_counter_t if_get_counter; /* get counter values */
Implement interface link header precomputation API. Add if_requestencap() interface method which is capable of calculating various link headers for given interface. Right now there is support for INET/INET6/ARP llheader calculation (IFENCAP_LL type request). Other types are planned to support more complex calculation (L2 multipath lagg nexthops, tunnel encap nexthops, etc..). Reshape 'struct route' to be able to pass additional data (with is length) to prepend to mbuf. These two changes permits routing code to pass pre-calculated nexthop data (like L2 header for route w/gateway) down to the stack eliminating the need for other lookups. It also brings us closer to more complex scenarios like transparently handling MPLS nexthops and tunnel interfaces. Last, but not least, it removes layering violation introduced by flowtable code (ro_lle) and simplifies handling of existing if_output consumers. ARP/ND changes: Make arp/ndp stack pre-calculate link header upon installing/updating lle record. Interface link address change are handled by re-calculating headers for all lles based on if_lladdr event. After these changes, arpresolve()/nd6_resolve() returns full pre-calculated header for supported interfaces thus simplifying if_output(). Move these lookups to separate ether_resolve_addr() function which ether returs error or fully-prepared link header. Add <arp|nd6_>resolve_addr() compat versions to return link addresses instead of pre-calculated data. BPF changes: Raw bpf writes occupied _two_ cases: AF_UNSPEC and pseudo_AF_HDRCMPLT. Despite the naming, both of there have ther header "complete". The only difference is that interface source mac has to be filled by OS for AF_UNSPEC (controlled via BIOCGHDRCMPLT). This logic has to stay inside BPF and not pollute if_output() routines. Convert BPF to pass prepend data via new 'struct route' mechanism. Note that it does not change non-optimized if_output(): ro_prepend handling is purely optional. Side note: hackish pseudo_AF_HDRCMPLT is supported for ethernet and FDDI. It is not needed for ethernet anymore. The only remaining FDDI user is dev/pdq mostly untouched since 2007. FDDI support was eliminated from OpenBSD in 2013 (sys/net/if_fddisubr.c rev 1.65). Flowtable changes: Flowtable violates layering by saving (and not correctly managing) rtes/lles. Instead of passing lle pointer, pass pointer to pre-calculated header data from that lle. Differential Revision: https://reviews.freebsd.org/D4102
2015-12-31 05:03:27 +00:00
int (*if_requestencap) /* make link header from request */
(struct ifnet *, struct if_encap_req *);
/* Statistics. */
counter_u64_t if_counters[IFCOUNTERS];
/* Stuff that's only temporary and doesn't belong here. */
/*
* Network adapter TSO limits:
* ===========================
*
* If the "if_hw_tsomax" field is zero the maximum segment
* length limit does not apply. If the "if_hw_tsomaxsegcount"
* or the "if_hw_tsomaxsegsize" field is zero the TSO segment
* count limit does not apply. If all three fields are zero,
* there is no TSO limit.
*
Update TSO limits to include all headers. To make driver programming easier the TSO limits are changed to reflect the values used in the BUSDMA tag a network adapter driver is using. The TCP/IP network stack will subtract space for all linklevel and protocol level headers and ensure that the full mbuf chain passed to the network adapter fits within the given limits. Implementation notes: If a network adapter driver needs to fixup the first mbuf in order to support VLAN tag insertion, the size of the VLAN tag should be subtracted from the TSO limit. Else not. Network adapters which typically inline the complete header mbuf could technically transmit one more segment. This patch does not implement a mechanism to recover the last segment for data transmission. It is believed when sufficiently large mbuf clusters are used, the segment limit will not be reached and recovering the last segment will not have any effect. The current TSO algorithm tries to send MTU-sized packets, where the MTU typically is 1500 bytes, which gives 1448 bytes of TCP data payload per packet for IPv4. That means if the TSO length limitiation is set to 65536 bytes, there will be a data payload remainder of (65536 - 1500) mod 1448 bytes which is equal to 324 bytes. Trying to recover total TSO length due to inlining mbuf header data will not have any effect, because adding or removing the ETH/IP/TCP headers to or from 324 bytes will not cause more or less TCP payload to be TSO'ed. Existing network adapter limits will be updated separately. Differential Revision: https://reviews.freebsd.org/D3458 Reviewed by: rmacklem MFC after: 2 weeks
2015-09-14 08:36:22 +00:00
* NOTE: The TSO limits should reflect the values used in the
* BUSDMA tag a network adapter is using to load a mbuf chain
* for transmission. The TCP/IP network stack will subtract
* space for all linklevel and protocol level headers and
* ensure that the full mbuf chain passed to the network
* adapter fits within the given limits.
*/
u_int if_hw_tsomax; /* TSO maximum size in bytes */
u_int if_hw_tsomaxsegcount; /* TSO maximum segment count */
u_int if_hw_tsomaxsegsize; /* TSO maximum segment size in bytes */
Implement kernel support for hardware rate limited sockets. - Add RATELIMIT kernel configuration keyword which must be set to enable the new functionality. - Add support for hardware driven, Receive Side Scaling, RSS aware, rate limited sendqueues and expose the functionality through the already established SO_MAX_PACING_RATE setsockopt(). The API support rates in the range from 1 to 4Gbytes/s which are suitable for regular TCP and UDP streams. The setsockopt(2) manual page has been updated. - Add rate limit function callback API to "struct ifnet" which supports the following operations: if_snd_tag_alloc(), if_snd_tag_modify(), if_snd_tag_query() and if_snd_tag_free(). - Add support to ifconfig to view, set and clear the IFCAP_TXRTLMT flag, which tells if a network driver supports rate limiting or not. - This patch also adds support for rate limiting through VLAN and LAGG intermediate network devices. - How rate limiting works: 1) The userspace application calls setsockopt() after accepting or making a new connection to set the rate which is then stored in the socket structure in the kernel. Later on when packets are transmitted a check is made in the transmit path for rate changes. A rate change implies a non-blocking ifp->if_snd_tag_alloc() call will be made to the destination network interface, which then sets up a custom sendqueue with the given rate limitation parameter. A "struct m_snd_tag" pointer is returned which serves as a "snd_tag" hint in the m_pkthdr for the subsequently transmitted mbufs. 2) When the network driver sees the "m->m_pkthdr.snd_tag" different from NULL, it will move the packets into a designated rate limited sendqueue given by the snd_tag pointer. It is up to the individual drivers how the rate limited traffic will be rate limited. 3) Route changes are detected by the NIC drivers in the ifp->if_transmit() routine when the ifnet pointer in the incoming snd_tag mismatches the one of the network interface. The network adapter frees the mbuf and returns EAGAIN which causes the ip_output() to release and clear the send tag. Upon next ip_output() a new "snd_tag" will be tried allocated. 4) When the PCB is detached the custom sendqueue will be released by a non-blocking ifp->if_snd_tag_free() call to the currently bound network interface. Reviewed by: wblock (manpages), adrian, gallatin, scottl (network) Differential Revision: https://reviews.freebsd.org/D3687 Sponsored by: Mellanox Technologies MFC after: 3 months
2017-01-18 13:31:17 +00:00
/*
* Network adapter send tag support:
*/
if_snd_tag_alloc_t *if_snd_tag_alloc;
/* Ratelimit (packet pacing) */
if_ratelimit_query_t *if_ratelimit_query;
if_ratelimit_setup_t *if_ratelimit_setup;
Implement kernel support for hardware rate limited sockets. - Add RATELIMIT kernel configuration keyword which must be set to enable the new functionality. - Add support for hardware driven, Receive Side Scaling, RSS aware, rate limited sendqueues and expose the functionality through the already established SO_MAX_PACING_RATE setsockopt(). The API support rates in the range from 1 to 4Gbytes/s which are suitable for regular TCP and UDP streams. The setsockopt(2) manual page has been updated. - Add rate limit function callback API to "struct ifnet" which supports the following operations: if_snd_tag_alloc(), if_snd_tag_modify(), if_snd_tag_query() and if_snd_tag_free(). - Add support to ifconfig to view, set and clear the IFCAP_TXRTLMT flag, which tells if a network driver supports rate limiting or not. - This patch also adds support for rate limiting through VLAN and LAGG intermediate network devices. - How rate limiting works: 1) The userspace application calls setsockopt() after accepting or making a new connection to set the rate which is then stored in the socket structure in the kernel. Later on when packets are transmitted a check is made in the transmit path for rate changes. A rate change implies a non-blocking ifp->if_snd_tag_alloc() call will be made to the destination network interface, which then sets up a custom sendqueue with the given rate limitation parameter. A "struct m_snd_tag" pointer is returned which serves as a "snd_tag" hint in the m_pkthdr for the subsequently transmitted mbufs. 2) When the network driver sees the "m->m_pkthdr.snd_tag" different from NULL, it will move the packets into a designated rate limited sendqueue given by the snd_tag pointer. It is up to the individual drivers how the rate limited traffic will be rate limited. 3) Route changes are detected by the NIC drivers in the ifp->if_transmit() routine when the ifnet pointer in the incoming snd_tag mismatches the one of the network interface. The network adapter frees the mbuf and returns EAGAIN which causes the ip_output() to release and clear the send tag. Upon next ip_output() a new "snd_tag" will be tried allocated. 4) When the PCB is detached the custom sendqueue will be released by a non-blocking ifp->if_snd_tag_free() call to the currently bound network interface. Reviewed by: wblock (manpages), adrian, gallatin, scottl (network) Differential Revision: https://reviews.freebsd.org/D3687 Sponsored by: Mellanox Technologies MFC after: 3 months
2017-01-18 13:31:17 +00:00
/* Ethernet PCP */
uint8_t if_pcp;
/*
Split out a more generic debugnet(4) from netdump(4) Debugnet is a simplistic and specialized panic- or debug-time reliable datagram transport. It can drive a single connection at a time and is currently unidirectional (debug/panic machine transmit to remote server only). It is mostly a verbatim code lift from netdump(4). Netdump(4) remains the only consumer (until the rest of this patch series lands). The INET-specific logic has been extracted somewhat more thoroughly than previously in netdump(4), into debugnet_inet.c. UDP-layer logic and up, as much as possible as is protocol-independent, remains in debugnet.c. The separation is not perfect and future improvement is welcome. Supporting INET6 is a long-term goal. Much of the diff is "gratuitous" renaming from 'netdump_' or 'nd_' to 'debugnet_' or 'dn_' -- sorry. I thought keeping the netdump name on the generic module would be more confusing than the refactoring. The only functional change here is the mbuf allocation / tracking. Instead of initiating solely on netdump-configured interface(s) at dumpon(8) configuration time, we watch for any debugnet-enabled NIC for link activation and query it for mbuf parameters at that time. If they exceed the existing high-water mark allocation, we re-allocate and track the new high-water mark. Otherwise, we leave the pre-panic mbuf allocation alone. In a future patch in this series, this will allow initiating netdump from panic ddb(4) without pre-panic configuration. No other functional change intended. Reviewed by: markj (earlier version) Some discussion with: emaste, jhb Objection from: marius Differential Revision: https://reviews.freebsd.org/D21421
2019-10-17 16:23:03 +00:00
* Debugnet (Netdump) hooks to be called while in db/panic.
*/
Split out a more generic debugnet(4) from netdump(4) Debugnet is a simplistic and specialized panic- or debug-time reliable datagram transport. It can drive a single connection at a time and is currently unidirectional (debug/panic machine transmit to remote server only). It is mostly a verbatim code lift from netdump(4). Netdump(4) remains the only consumer (until the rest of this patch series lands). The INET-specific logic has been extracted somewhat more thoroughly than previously in netdump(4), into debugnet_inet.c. UDP-layer logic and up, as much as possible as is protocol-independent, remains in debugnet.c. The separation is not perfect and future improvement is welcome. Supporting INET6 is a long-term goal. Much of the diff is "gratuitous" renaming from 'netdump_' or 'nd_' to 'debugnet_' or 'dn_' -- sorry. I thought keeping the netdump name on the generic module would be more confusing than the refactoring. The only functional change here is the mbuf allocation / tracking. Instead of initiating solely on netdump-configured interface(s) at dumpon(8) configuration time, we watch for any debugnet-enabled NIC for link activation and query it for mbuf parameters at that time. If they exceed the existing high-water mark allocation, we re-allocate and track the new high-water mark. Otherwise, we leave the pre-panic mbuf allocation alone. In a future patch in this series, this will allow initiating netdump from panic ddb(4) without pre-panic configuration. No other functional change intended. Reviewed by: markj (earlier version) Some discussion with: emaste, jhb Objection from: marius Differential Revision: https://reviews.freebsd.org/D21421
2019-10-17 16:23:03 +00:00
struct debugnet_methods *if_debugnet_methods;
struct epoch_context if_epoch_ctx;
/*
* Spare fields to be added before branching a stable branch, so
* that structure can be enhanced without changing the kernel
* binary interface.
*/
Implement kernel support for hardware rate limited sockets. - Add RATELIMIT kernel configuration keyword which must be set to enable the new functionality. - Add support for hardware driven, Receive Side Scaling, RSS aware, rate limited sendqueues and expose the functionality through the already established SO_MAX_PACING_RATE setsockopt(). The API support rates in the range from 1 to 4Gbytes/s which are suitable for regular TCP and UDP streams. The setsockopt(2) manual page has been updated. - Add rate limit function callback API to "struct ifnet" which supports the following operations: if_snd_tag_alloc(), if_snd_tag_modify(), if_snd_tag_query() and if_snd_tag_free(). - Add support to ifconfig to view, set and clear the IFCAP_TXRTLMT flag, which tells if a network driver supports rate limiting or not. - This patch also adds support for rate limiting through VLAN and LAGG intermediate network devices. - How rate limiting works: 1) The userspace application calls setsockopt() after accepting or making a new connection to set the rate which is then stored in the socket structure in the kernel. Later on when packets are transmitted a check is made in the transmit path for rate changes. A rate change implies a non-blocking ifp->if_snd_tag_alloc() call will be made to the destination network interface, which then sets up a custom sendqueue with the given rate limitation parameter. A "struct m_snd_tag" pointer is returned which serves as a "snd_tag" hint in the m_pkthdr for the subsequently transmitted mbufs. 2) When the network driver sees the "m->m_pkthdr.snd_tag" different from NULL, it will move the packets into a designated rate limited sendqueue given by the snd_tag pointer. It is up to the individual drivers how the rate limited traffic will be rate limited. 3) Route changes are detected by the NIC drivers in the ifp->if_transmit() routine when the ifnet pointer in the incoming snd_tag mismatches the one of the network interface. The network adapter frees the mbuf and returns EAGAIN which causes the ip_output() to release and clear the send tag. Upon next ip_output() a new "snd_tag" will be tried allocated. 4) When the PCB is detached the custom sendqueue will be released by a non-blocking ifp->if_snd_tag_free() call to the currently bound network interface. Reviewed by: wblock (manpages), adrian, gallatin, scottl (network) Differential Revision: https://reviews.freebsd.org/D3687 Sponsored by: Mellanox Technologies MFC after: 3 months
2017-01-18 13:31:17 +00:00
int if_ispare[4]; /* general use */
};
/* for compatibility with other BSDs */
#define if_name(ifp) ((ifp)->if_xname)
#define IF_NODOM 255
/*
* Locks for address lists on the network interface.
*/
ifnet: Replace if_addr_lock rwlock with epoch + mutex Run on LLNW canaries and tested by pho@ gallatin: Using a 14-core, 28-HTT single socket E5-2697 v3 with a 40GbE MLX5 based ConnectX 4-LX NIC, I see an almost 12% improvement in received packet rate, and a larger improvement in bytes delivered all the way to userspace. When the host receiving 64 streams of netperf -H $DUT -t UDP_STREAM -- -m 1, I see, using nstat -I mce0 1 before the patch: InMpps OMpps InGbs OGbs err TCP Est %CPU syscalls csw irq GBfree 4.98 0.00 4.42 0.00 4235592 33 83.80 4720653 2149771 1235 247.32 4.73 0.00 4.20 0.00 4025260 33 82.99 4724900 2139833 1204 247.32 4.72 0.00 4.20 0.00 4035252 33 82.14 4719162 2132023 1264 247.32 4.71 0.00 4.21 0.00 4073206 33 83.68 4744973 2123317 1347 247.32 4.72 0.00 4.21 0.00 4061118 33 80.82 4713615 2188091 1490 247.32 4.72 0.00 4.21 0.00 4051675 33 85.29 4727399 2109011 1205 247.32 4.73 0.00 4.21 0.00 4039056 33 84.65 4724735 2102603 1053 247.32 After the patch InMpps OMpps InGbs OGbs err TCP Est %CPU syscalls csw irq GBfree 5.43 0.00 4.20 0.00 3313143 33 84.96 5434214 1900162 2656 245.51 5.43 0.00 4.20 0.00 3308527 33 85.24 5439695 1809382 2521 245.51 5.42 0.00 4.19 0.00 3316778 33 87.54 5416028 1805835 2256 245.51 5.42 0.00 4.19 0.00 3317673 33 90.44 5426044 1763056 2332 245.51 5.42 0.00 4.19 0.00 3314839 33 88.11 5435732 1792218 2499 245.52 5.44 0.00 4.19 0.00 3293228 33 91.84 5426301 1668597 2121 245.52 Similarly, netperf reports 230Mb/s before the patch, and 270Mb/s after the patch Reviewed by: gallatin Sponsored by: Limelight Networks Differential Revision: https://reviews.freebsd.org/D15366
2018-05-18 20:13:34 +00:00
#define IF_ADDR_LOCK_INIT(if) mtx_init(&(if)->if_addr_lock, "if_addr_lock", NULL, MTX_DEF)
#define IF_ADDR_LOCK_DESTROY(if) mtx_destroy(&(if)->if_addr_lock)
#define IF_ADDR_WLOCK(if) mtx_lock(&(if)->if_addr_lock)
#define IF_ADDR_WUNLOCK(if) mtx_unlock(&(if)->if_addr_lock)
#define IF_ADDR_LOCK_ASSERT(if) MPASS(in_epoch(net_epoch_preempt) || mtx_owned(&(if)->if_addr_lock))
ifnet: Replace if_addr_lock rwlock with epoch + mutex Run on LLNW canaries and tested by pho@ gallatin: Using a 14-core, 28-HTT single socket E5-2697 v3 with a 40GbE MLX5 based ConnectX 4-LX NIC, I see an almost 12% improvement in received packet rate, and a larger improvement in bytes delivered all the way to userspace. When the host receiving 64 streams of netperf -H $DUT -t UDP_STREAM -- -m 1, I see, using nstat -I mce0 1 before the patch: InMpps OMpps InGbs OGbs err TCP Est %CPU syscalls csw irq GBfree 4.98 0.00 4.42 0.00 4235592 33 83.80 4720653 2149771 1235 247.32 4.73 0.00 4.20 0.00 4025260 33 82.99 4724900 2139833 1204 247.32 4.72 0.00 4.20 0.00 4035252 33 82.14 4719162 2132023 1264 247.32 4.71 0.00 4.21 0.00 4073206 33 83.68 4744973 2123317 1347 247.32 4.72 0.00 4.21 0.00 4061118 33 80.82 4713615 2188091 1490 247.32 4.72 0.00 4.21 0.00 4051675 33 85.29 4727399 2109011 1205 247.32 4.73 0.00 4.21 0.00 4039056 33 84.65 4724735 2102603 1053 247.32 After the patch InMpps OMpps InGbs OGbs err TCP Est %CPU syscalls csw irq GBfree 5.43 0.00 4.20 0.00 3313143 33 84.96 5434214 1900162 2656 245.51 5.43 0.00 4.20 0.00 3308527 33 85.24 5439695 1809382 2521 245.51 5.42 0.00 4.19 0.00 3316778 33 87.54 5416028 1805835 2256 245.51 5.42 0.00 4.19 0.00 3317673 33 90.44 5426044 1763056 2332 245.51 5.42 0.00 4.19 0.00 3314839 33 88.11 5435732 1792218 2499 245.52 5.44 0.00 4.19 0.00 3293228 33 91.84 5426301 1668597 2121 245.52 Similarly, netperf reports 230Mb/s before the patch, and 270Mb/s after the patch Reviewed by: gallatin Sponsored by: Limelight Networks Differential Revision: https://reviews.freebsd.org/D15366
2018-05-18 20:13:34 +00:00
#define IF_ADDR_WLOCK_ASSERT(if) mtx_assert(&(if)->if_addr_lock, MA_OWNED)
#ifdef _KERNEL
/* interface link layer address change event */
typedef void (*iflladdr_event_handler_t)(void *, struct ifnet *);
EVENTHANDLER_DECLARE(iflladdr_event, iflladdr_event_handler_t);
/* interface address change event */
typedef void (*ifaddr_event_handler_t)(void *, struct ifnet *);
EVENTHANDLER_DECLARE(ifaddr_event, ifaddr_event_handler_t);
typedef void (*ifaddr_event_ext_handler_t)(void *, struct ifnet *,
struct ifaddr *, int);
EVENTHANDLER_DECLARE(ifaddr_event_ext, ifaddr_event_ext_handler_t);
#define IFADDR_EVENT_ADD 0
#define IFADDR_EVENT_DEL 1
/* new interface arrival event */
typedef void (*ifnet_arrival_event_handler_t)(void *, struct ifnet *);
EVENTHANDLER_DECLARE(ifnet_arrival_event, ifnet_arrival_event_handler_t);
/* interface departure event */
typedef void (*ifnet_departure_event_handler_t)(void *, struct ifnet *);
EVENTHANDLER_DECLARE(ifnet_departure_event, ifnet_departure_event_handler_t);
/* Interface link state change event */
typedef void (*ifnet_link_event_handler_t)(void *, struct ifnet *, int);
EVENTHANDLER_DECLARE(ifnet_link_event, ifnet_link_event_handler_t);
/* Interface up/down event */
#define IFNET_EVENT_UP 0
#define IFNET_EVENT_DOWN 1
#define IFNET_EVENT_PCP 2 /* priority code point, PCP */
typedef void (*ifnet_event_fn)(void *, struct ifnet *ifp, int event);
EVENTHANDLER_DECLARE(ifnet_event, ifnet_event_fn);
/*
* interface groups
*/
struct ifg_group {
char ifg_group[IFNAMSIZ];
u_int ifg_refcnt;
void *ifg_pf_kif;
CK_STAILQ_HEAD(, ifg_member) ifg_members; /* (CK_) */
CK_STAILQ_ENTRY(ifg_group) ifg_next; /* (CK_) */
};
struct ifg_member {
CK_STAILQ_ENTRY(ifg_member) ifgm_next; /* (CK_) */
struct ifnet *ifgm_ifp;
};
struct ifg_list {
struct ifg_group *ifgl_group;
CK_STAILQ_ENTRY(ifg_list) ifgl_next; /* (CK_) */
};
#ifdef _SYS_EVENTHANDLER_H_
/* group attach event */
typedef void (*group_attach_event_handler_t)(void *, struct ifg_group *);
EVENTHANDLER_DECLARE(group_attach_event, group_attach_event_handler_t);
/* group detach event */
typedef void (*group_detach_event_handler_t)(void *, struct ifg_group *);
EVENTHANDLER_DECLARE(group_detach_event, group_detach_event_handler_t);
/* group change event */
typedef void (*group_change_event_handler_t)(void *, const char *);
EVENTHANDLER_DECLARE(group_change_event, group_change_event_handler_t);
#endif /* _SYS_EVENTHANDLER_H_ */
#define IF_AFDATA_LOCK_INIT(ifp) \
mtx_init(&(ifp)->if_afdata_lock, "if_afdata", NULL, MTX_DEF)
#define IF_AFDATA_WLOCK(ifp) mtx_lock(&(ifp)->if_afdata_lock)
#define IF_AFDATA_WUNLOCK(ifp) mtx_unlock(&(ifp)->if_afdata_lock)
#define IF_AFDATA_LOCK(ifp) IF_AFDATA_WLOCK(ifp)
#define IF_AFDATA_UNLOCK(ifp) IF_AFDATA_WUNLOCK(ifp)
#define IF_AFDATA_TRYLOCK(ifp) mtx_trylock(&(ifp)->if_afdata_lock)
#define IF_AFDATA_DESTROY(ifp) mtx_destroy(&(ifp)->if_afdata_lock)
#define IF_AFDATA_LOCK_ASSERT(ifp) MPASS(in_epoch(net_epoch_preempt) || mtx_owned(&(ifp)->if_afdata_lock))
#define IF_AFDATA_WLOCK_ASSERT(ifp) mtx_assert(&(ifp)->if_afdata_lock, MA_OWNED)
#define IF_AFDATA_UNLOCK_ASSERT(ifp) mtx_assert(&(ifp)->if_afdata_lock, MA_NOTOWNED)
/*
* 72 was chosen below because it is the size of a TCP/IP
* header (40) + the minimum mss (32).
*/
#define IF_MINMTU 72
#define IF_MAXMTU 65535
#define TOEDEV(ifp) ((ifp)->if_llsoftc)
/*
* The ifaddr structure contains information about one address
* of an interface. They are maintained by the different address families,
* are allocated and attached when an address is set, and are linked
* together so all addresses for an interface can be located.
*
* NOTE: a 'struct ifaddr' is always at the beginning of a larger
* chunk of malloc'ed memory, where we store the three addresses
* (ifa_addr, ifa_dstaddr and ifa_netmask) referenced here.
*/
struct ifaddr {
struct sockaddr *ifa_addr; /* address of interface */
struct sockaddr *ifa_dstaddr; /* other end of p-to-p link */
#define ifa_broadaddr ifa_dstaddr /* broadcast address interface */
struct sockaddr *ifa_netmask; /* used to determine subnet */
struct ifnet *ifa_ifp; /* back-pointer to interface */
struct carp_softc *ifa_carp; /* pointer to CARP data */
CK_STAILQ_ENTRY(ifaddr) ifa_link; /* queue macro glue */
u_short ifa_flags; /* mostly rt_flags for cloning */
#define IFA_ROUTE RTF_UP /* route installed */
#define IFA_RTSELF RTF_HOST /* loopback route to self installed */
u_int ifa_refcnt; /* references to this structure */
counter_u64_t ifa_ipackets;
counter_u64_t ifa_opackets;
counter_u64_t ifa_ibytes;
counter_u64_t ifa_obytes;
ifnet: Replace if_addr_lock rwlock with epoch + mutex Run on LLNW canaries and tested by pho@ gallatin: Using a 14-core, 28-HTT single socket E5-2697 v3 with a 40GbE MLX5 based ConnectX 4-LX NIC, I see an almost 12% improvement in received packet rate, and a larger improvement in bytes delivered all the way to userspace. When the host receiving 64 streams of netperf -H $DUT -t UDP_STREAM -- -m 1, I see, using nstat -I mce0 1 before the patch: InMpps OMpps InGbs OGbs err TCP Est %CPU syscalls csw irq GBfree 4.98 0.00 4.42 0.00 4235592 33 83.80 4720653 2149771 1235 247.32 4.73 0.00 4.20 0.00 4025260 33 82.99 4724900 2139833 1204 247.32 4.72 0.00 4.20 0.00 4035252 33 82.14 4719162 2132023 1264 247.32 4.71 0.00 4.21 0.00 4073206 33 83.68 4744973 2123317 1347 247.32 4.72 0.00 4.21 0.00 4061118 33 80.82 4713615 2188091 1490 247.32 4.72 0.00 4.21 0.00 4051675 33 85.29 4727399 2109011 1205 247.32 4.73 0.00 4.21 0.00 4039056 33 84.65 4724735 2102603 1053 247.32 After the patch InMpps OMpps InGbs OGbs err TCP Est %CPU syscalls csw irq GBfree 5.43 0.00 4.20 0.00 3313143 33 84.96 5434214 1900162 2656 245.51 5.43 0.00 4.20 0.00 3308527 33 85.24 5439695 1809382 2521 245.51 5.42 0.00 4.19 0.00 3316778 33 87.54 5416028 1805835 2256 245.51 5.42 0.00 4.19 0.00 3317673 33 90.44 5426044 1763056 2332 245.51 5.42 0.00 4.19 0.00 3314839 33 88.11 5435732 1792218 2499 245.52 5.44 0.00 4.19 0.00 3293228 33 91.84 5426301 1668597 2121 245.52 Similarly, netperf reports 230Mb/s before the patch, and 270Mb/s after the patch Reviewed by: gallatin Sponsored by: Limelight Networks Differential Revision: https://reviews.freebsd.org/D15366
2018-05-18 20:13:34 +00:00
struct epoch_context ifa_epoch_ctx;
};
struct ifaddr * ifa_alloc(size_t size, int flags);
void ifa_free(struct ifaddr *ifa);
void ifa_ref(struct ifaddr *ifa);
int __result_use_check ifa_try_ref(struct ifaddr *ifa);
2002-12-18 11:46:59 +00:00
/*
* Multicast address structure. This is analogous to the ifaddr
* structure except that it keeps track of multicast addresses.
*/
#define IFMA_F_ENQUEUED 0x1
struct ifmultiaddr {
CK_STAILQ_ENTRY(ifmultiaddr) ifma_link; /* queue macro glue */
struct sockaddr *ifma_addr; /* address this membership is for */
struct sockaddr *ifma_lladdr; /* link-layer translation, if any */
struct ifnet *ifma_ifp; /* back-pointer to interface */
u_int ifma_refcount; /* reference count */
int ifma_flags;
void *ifma_protospec; /* protocol-specific state, if any */
struct ifmultiaddr *ifma_llifma; /* pointer to ifma for ifma_lladdr */
ifnet: Replace if_addr_lock rwlock with epoch + mutex Run on LLNW canaries and tested by pho@ gallatin: Using a 14-core, 28-HTT single socket E5-2697 v3 with a 40GbE MLX5 based ConnectX 4-LX NIC, I see an almost 12% improvement in received packet rate, and a larger improvement in bytes delivered all the way to userspace. When the host receiving 64 streams of netperf -H $DUT -t UDP_STREAM -- -m 1, I see, using nstat -I mce0 1 before the patch: InMpps OMpps InGbs OGbs err TCP Est %CPU syscalls csw irq GBfree 4.98 0.00 4.42 0.00 4235592 33 83.80 4720653 2149771 1235 247.32 4.73 0.00 4.20 0.00 4025260 33 82.99 4724900 2139833 1204 247.32 4.72 0.00 4.20 0.00 4035252 33 82.14 4719162 2132023 1264 247.32 4.71 0.00 4.21 0.00 4073206 33 83.68 4744973 2123317 1347 247.32 4.72 0.00 4.21 0.00 4061118 33 80.82 4713615 2188091 1490 247.32 4.72 0.00 4.21 0.00 4051675 33 85.29 4727399 2109011 1205 247.32 4.73 0.00 4.21 0.00 4039056 33 84.65 4724735 2102603 1053 247.32 After the patch InMpps OMpps InGbs OGbs err TCP Est %CPU syscalls csw irq GBfree 5.43 0.00 4.20 0.00 3313143 33 84.96 5434214 1900162 2656 245.51 5.43 0.00 4.20 0.00 3308527 33 85.24 5439695 1809382 2521 245.51 5.42 0.00 4.19 0.00 3316778 33 87.54 5416028 1805835 2256 245.51 5.42 0.00 4.19 0.00 3317673 33 90.44 5426044 1763056 2332 245.51 5.42 0.00 4.19 0.00 3314839 33 88.11 5435732 1792218 2499 245.52 5.44 0.00 4.19 0.00 3293228 33 91.84 5426301 1668597 2121 245.52 Similarly, netperf reports 230Mb/s before the patch, and 270Mb/s after the patch Reviewed by: gallatin Sponsored by: Limelight Networks Differential Revision: https://reviews.freebsd.org/D15366
2018-05-18 20:13:34 +00:00
struct epoch_context ifma_epoch_ctx;
};
extern struct sx ifnet_sxlock;
#define IFNET_WLOCK() sx_xlock(&ifnet_sxlock)
#define IFNET_WUNLOCK() sx_xunlock(&ifnet_sxlock)
#define IFNET_RLOCK_ASSERT() sx_assert(&ifnet_sxlock, SA_SLOCKED)
#define IFNET_WLOCK_ASSERT() sx_assert(&ifnet_sxlock, SA_XLOCKED)
#define IFNET_RLOCK() sx_slock(&ifnet_sxlock)
#define IFNET_RUNLOCK() sx_sunlock(&ifnet_sxlock)
2002-12-22 05:35:03 +00:00
Start to address a number of races relating to use of ifnet pointers after the corresponding interface has been destroyed: (1) Add an ifnet refcount, ifp->if_refcount. Initialize it to 1 in if_alloc(), and modify if_free_type() to decrement and check the refcount. (2) Add new if_ref() and if_rele() interfaces to allow kernel code walking global interface lists to release IFNET_[RW]LOCK() yet keep the ifnet stable. Currently, if_rele() is a no-op wrapper around if_free(), but this may change in the future. (3) Add new ifnet field, if_alloctype, which caches the type passed to if_alloc(), but unlike if_type, won't be changed by drivers. This allows asynchronous free's of the interface after the driver has released it to still use the right type. Use that instead of the type passed to if_free_type(), but assert that they are the same (might have to rethink this if that doesn't work out). (4) Add a new ifnet_byindex_ref(), which looks up an interface by index and returns a reference rather than a pointer to it. (5) Fix if_alloc() to fully initialize the if_addr_mtx before hooking up the ifnet to global lists. (6) Modify sysctls in if_mib.c to use ifnet_byindex_ref() and release the ifnet when done. When this change is MFC'd, it will need to replace if_ispare fields rather than adding new fields in order to avoid breaking the binary interface. Once this change is MFC'd, if_free_type() should be removed, as its 'type' argument is now optional. This refcount is not appropriate for counting mbuf pkthdr references, and also not for counting entry into the device driver via ifnet function pointers. An rmlock may be appropriate for the latter. Rather, this is about ensuring data structure stability when reaching an ifnet via global ifnet lists and tables followed by copy in or out of userspace. MFC after: 3 weeks Reported by: mdtancsa Reviewed by: brooks
2009-04-21 22:43:32 +00:00
/*
* Look up an ifnet given its index. The returned value protected from
* being freed by the network epoch. The _ref variant also acquires a
* reference that must be freed using if_rele().
Start to address a number of races relating to use of ifnet pointers after the corresponding interface has been destroyed: (1) Add an ifnet refcount, ifp->if_refcount. Initialize it to 1 in if_alloc(), and modify if_free_type() to decrement and check the refcount. (2) Add new if_ref() and if_rele() interfaces to allow kernel code walking global interface lists to release IFNET_[RW]LOCK() yet keep the ifnet stable. Currently, if_rele() is a no-op wrapper around if_free(), but this may change in the future. (3) Add new ifnet field, if_alloctype, which caches the type passed to if_alloc(), but unlike if_type, won't be changed by drivers. This allows asynchronous free's of the interface after the driver has released it to still use the right type. Use that instead of the type passed to if_free_type(), but assert that they are the same (might have to rethink this if that doesn't work out). (4) Add a new ifnet_byindex_ref(), which looks up an interface by index and returns a reference rather than a pointer to it. (5) Fix if_alloc() to fully initialize the if_addr_mtx before hooking up the ifnet to global lists. (6) Modify sysctls in if_mib.c to use ifnet_byindex_ref() and release the ifnet when done. When this change is MFC'd, it will need to replace if_ispare fields rather than adding new fields in order to avoid breaking the binary interface. Once this change is MFC'd, if_free_type() should be removed, as its 'type' argument is now optional. This refcount is not appropriate for counting mbuf pkthdr references, and also not for counting entry into the device driver via ifnet function pointers. An rmlock may be appropriate for the latter. Rather, this is about ensuring data structure stability when reaching an ifnet via global ifnet lists and tables followed by copy in or out of userspace. MFC after: 3 weeks Reported by: mdtancsa Reviewed by: brooks
2009-04-21 22:43:32 +00:00
*/
struct ifnet *ifnet_byindex(u_int);
struct ifnet *ifnet_byindex_ref(u_int);
2008-08-20 03:14:48 +00:00
/*
* Given the index, ifaddr_byindex() returns the one and only
* link-level ifaddr for the interface. You are not supposed to use
* it to traverse the list of addresses associated to the interface.
*/
struct ifaddr *ifaddr_byindex(u_short idx);
Build on Jeff Roberson's linker-set based dynamic per-CPU allocator (DPCPU), as suggested by Peter Wemm, and implement a new per-virtual network stack memory allocator. Modify vnet to use the allocator instead of monolithic global container structures (vinet, ...). This change solves many binary compatibility problems associated with VIMAGE, and restores ELF symbols for virtualized global variables. Each virtualized global variable exists as a "reference copy", and also once per virtual network stack. Virtualized global variables are tagged at compile-time, placing the in a special linker set, which is loaded into a contiguous region of kernel memory. Virtualized global variables in the base kernel are linked as normal, but those in modules are copied and relocated to a reserved portion of the kernel's vnet region with the help of a the kernel linker. Virtualized global variables exist in per-vnet memory set up when the network stack instance is created, and are initialized statically from the reference copy. Run-time access occurs via an accessor macro, which converts from the current vnet and requested symbol to a per-vnet address. When "options VIMAGE" is not compiled into the kernel, normal global ELF symbols will be used instead and indirection is avoided. This change restores static initialization for network stack global variables, restores support for non-global symbols and types, eliminates the need for many subsystem constructors, eliminates large per-subsystem structures that caused many binary compatibility issues both for monitoring applications (netstat) and kernel modules, removes the per-function INIT_VNET_*() macros throughout the stack, eliminates the need for vnet_symmap ksym(2) munging, and eliminates duplicate definitions of virtualized globals under VIMAGE_GLOBALS. Bump __FreeBSD_version and update UPDATING. Portions submitted by: bz Reviewed by: bz, zec Discussed with: gnn, jamie, jeff, jhb, julian, sam Suggested by: peter Approved by: re (kensmith)
2009-07-14 22:48:30 +00:00
VNET_DECLARE(struct ifnethead, ifnet);
VNET_DECLARE(struct ifgrouphead, ifg_head);
VNET_DECLARE(struct ifnet *, loif); /* first loopback interface */
#define V_ifnet VNET(ifnet)
#define V_ifg_head VNET(ifg_head)
#define V_loif VNET(loif)
Build on Jeff Roberson's linker-set based dynamic per-CPU allocator (DPCPU), as suggested by Peter Wemm, and implement a new per-virtual network stack memory allocator. Modify vnet to use the allocator instead of monolithic global container structures (vinet, ...). This change solves many binary compatibility problems associated with VIMAGE, and restores ELF symbols for virtualized global variables. Each virtualized global variable exists as a "reference copy", and also once per virtual network stack. Virtualized global variables are tagged at compile-time, placing the in a special linker set, which is loaded into a contiguous region of kernel memory. Virtualized global variables in the base kernel are linked as normal, but those in modules are copied and relocated to a reserved portion of the kernel's vnet region with the help of a the kernel linker. Virtualized global variables exist in per-vnet memory set up when the network stack instance is created, and are initialized statically from the reference copy. Run-time access occurs via an accessor macro, which converts from the current vnet and requested symbol to a per-vnet address. When "options VIMAGE" is not compiled into the kernel, normal global ELF symbols will be used instead and indirection is avoided. This change restores static initialization for network stack global variables, restores support for non-global symbols and types, eliminates the need for many subsystem constructors, eliminates large per-subsystem structures that caused many binary compatibility issues both for monitoring applications (netstat) and kernel modules, removes the per-function INIT_VNET_*() macros throughout the stack, eliminates the need for vnet_symmap ksym(2) munging, and eliminates duplicate definitions of virtualized globals under VIMAGE_GLOBALS. Bump __FreeBSD_version and update UPDATING. Portions submitted by: bz Reviewed by: bz, zec Discussed with: gnn, jamie, jeff, jhb, julian, sam Suggested by: peter Approved by: re (kensmith)
2009-07-14 22:48:30 +00:00
#ifdef MCAST_VERBOSE
#define MCDPRINTF printf
#else
#define MCDPRINTF(...)
#endif
int if_addgroup(struct ifnet *, const char *);
int if_delgroup(struct ifnet *, const char *);
2002-03-19 21:54:18 +00:00
int if_addmulti(struct ifnet *, struct sockaddr *, struct ifmultiaddr **);
int if_allmulti(struct ifnet *, int);
struct ifnet* if_alloc(u_char);
struct ifnet* if_alloc_dev(u_char, device_t dev);
2002-03-19 21:54:18 +00:00
void if_attach(struct ifnet *);
void if_dead(struct ifnet *);
2002-03-19 21:54:18 +00:00
int if_delmulti(struct ifnet *, struct sockaddr *);
void if_delmulti_ifma(struct ifmultiaddr *);
void if_delmulti_ifma_flags(struct ifmultiaddr *, int flags);
2002-03-19 21:54:18 +00:00
void if_detach(struct ifnet *);
void if_purgeaddrs(struct ifnet *);
void if_delallmulti(struct ifnet *);
2002-03-19 21:54:18 +00:00
void if_down(struct ifnet *);
struct ifmultiaddr *
if_findmulti(struct ifnet *, const struct sockaddr *);
void if_freemulti(struct ifmultiaddr *ifma);
void if_free(struct ifnet *);
void if_initname(struct ifnet *, const char *, int);
void if_link_state_change(struct ifnet *, int);
int if_printf(struct ifnet *, const char *, ...) __printflike(2, 3);
int if_log(struct ifnet *, int, const char *, ...) __printflike(3, 4);
Start to address a number of races relating to use of ifnet pointers after the corresponding interface has been destroyed: (1) Add an ifnet refcount, ifp->if_refcount. Initialize it to 1 in if_alloc(), and modify if_free_type() to decrement and check the refcount. (2) Add new if_ref() and if_rele() interfaces to allow kernel code walking global interface lists to release IFNET_[RW]LOCK() yet keep the ifnet stable. Currently, if_rele() is a no-op wrapper around if_free(), but this may change in the future. (3) Add new ifnet field, if_alloctype, which caches the type passed to if_alloc(), but unlike if_type, won't be changed by drivers. This allows asynchronous free's of the interface after the driver has released it to still use the right type. Use that instead of the type passed to if_free_type(), but assert that they are the same (might have to rethink this if that doesn't work out). (4) Add a new ifnet_byindex_ref(), which looks up an interface by index and returns a reference rather than a pointer to it. (5) Fix if_alloc() to fully initialize the if_addr_mtx before hooking up the ifnet to global lists. (6) Modify sysctls in if_mib.c to use ifnet_byindex_ref() and release the ifnet when done. When this change is MFC'd, it will need to replace if_ispare fields rather than adding new fields in order to avoid breaking the binary interface. Once this change is MFC'd, if_free_type() should be removed, as its 'type' argument is now optional. This refcount is not appropriate for counting mbuf pkthdr references, and also not for counting entry into the device driver via ifnet function pointers. An rmlock may be appropriate for the latter. Rather, this is about ensuring data structure stability when reaching an ifnet via global ifnet lists and tables followed by copy in or out of userspace. MFC after: 3 weeks Reported by: mdtancsa Reviewed by: brooks
2009-04-21 22:43:32 +00:00
void if_ref(struct ifnet *);
void if_rele(struct ifnet *);
bool __result_use_check if_try_ref(struct ifnet *);
2002-03-19 21:54:18 +00:00
int if_setlladdr(struct ifnet *, const u_char *, int);
int if_tunnel_check_nesting(struct ifnet *, struct mbuf *, uint32_t, int);
2002-03-19 21:54:18 +00:00
void if_up(struct ifnet *);
int ifioctl(struct socket *, u_long, caddr_t, struct thread *);
int ifpromisc(struct ifnet *, int);
struct ifnet *ifunit(const char *);
struct ifnet *ifunit_ref(const char *);
2002-03-19 21:54:18 +00:00
int ifa_add_loopback_route(struct ifaddr *, struct sockaddr *);
int ifa_del_loopback_route(struct ifaddr *, struct sockaddr *);
int ifa_switch_loopback_route(struct ifaddr *, struct sockaddr *);
struct ifaddr *ifa_ifwithaddr(const struct sockaddr *);
int ifa_ifwithaddr_check(const struct sockaddr *);
struct ifaddr *ifa_ifwithbroadaddr(const struct sockaddr *, int);
struct ifaddr *ifa_ifwithdstaddr(const struct sockaddr *, int);
struct ifaddr *ifa_ifwithnet(const struct sockaddr *, int, int);
struct ifaddr *ifa_ifwithroute(int, const struct sockaddr *,
const struct sockaddr *, u_int);
struct ifaddr *ifaof_ifpforaddr(const struct sockaddr *, struct ifnet *);
int ifa_preferred(struct ifaddr *, struct ifaddr *);
2002-03-19 21:54:18 +00:00
int if_simloop(struct ifnet *ifp, struct mbuf *m, int af, int hlen);
typedef void *if_com_alloc_t(u_char type, struct ifnet *ifp);
typedef void if_com_free_t(void *com, u_char type);
void if_register_com_alloc(u_char type, if_com_alloc_t *a, if_com_free_t *f);
void if_deregister_com_alloc(u_char type);
void if_data_copy(struct ifnet *, struct if_data *);
uint64_t if_get_counter_default(struct ifnet *, ift_counter);
void if_inc_counter(struct ifnet *, ift_counter, int64_t);
#define IF_LLADDR(ifp) \
LLADDR((struct sockaddr_dl *)((ifp)->if_addr->ifa_addr))
Introduce a procedural interface to the ifnet structure. The new interface allows the ifnet structure to be defined as an opaque type in NIC drivers. This then allows the ifnet structure to be changed without a need to change or recompile NIC drivers. Put differently, NIC drivers can be written and compiled once and be used with different network stack implementations, provided of course that those network stack implementations have an API and ABI compatible interface. This commit introduces the 'if_t' type to replace 'struct ifnet *' as the type of a network interface. The 'if_t' type is defined as 'void *' to enable the compiler to perform type conversion to 'struct ifnet *' and vice versa where needed and without warnings. The functions that implement the API are the only functions that need to have an explicit cast. The MII code has been converted to use the driver API to avoid unnecessary code churn. Code churn comes from having to work with both converted and unconverted drivers in correlation with having callback functions that take an interface. By converting the MII code first, the callback functions can be defined so that the compiler will perform the typecasts automatically. As soon as all drivers have been converted, the if_t type can be redefined as needed and the API functions can be fix to not need an explicit cast. The immediate benefactors of this change are: 1. Juniper Networks - The network stack implementation in Junos is entirely different from FreeBSD's one and this change allows Juniper to build "stock" NIC drivers that can be used in combination with both the FreeBSD and Junos stacks. 2. FreeBSD - This change opens the door towards changing ifnet and implementing new features and optimizations in the network stack without it requiring a change in the many NIC drivers FreeBSD has. Submitted by: Anuranjan Shukla <anshukla@juniper.net> Reviewed by: glebius@ Obtained from: Juniper Networks, Inc.
2014-06-02 17:54:39 +00:00
uint64_t if_setbaudrate(if_t ifp, uint64_t baudrate);
uint64_t if_getbaudrate(if_t ifp);
int if_setcapabilities(if_t ifp, int capabilities);
int if_setcapabilitiesbit(if_t ifp, int setbit, int clearbit);
int if_getcapabilities(if_t ifp);
int if_togglecapenable(if_t ifp, int togglecap);
int if_setcapenable(if_t ifp, int capenable);
int if_setcapenablebit(if_t ifp, int setcap, int clearcap);
int if_getcapenable(if_t ifp);
const char *if_getdname(if_t ifp);
int if_setdev(if_t ifp, void *dev);
int if_setdrvflagbits(if_t ifp, int if_setflags, int clear_flags);
int if_getdrvflags(if_t ifp);
int if_setdrvflags(if_t ifp, int flags);
int if_clearhwassist(if_t ifp);
int if_sethwassistbits(if_t ifp, int toset, int toclear);
int if_sethwassist(if_t ifp, int hwassist_bit);
int if_gethwassist(if_t ifp);
int if_setsoftc(if_t ifp, void *softc);
void *if_getsoftc(if_t ifp);
int if_setflags(if_t ifp, int flags);
int if_gethwaddr(if_t ifp, struct ifreq *);
Introduce a procedural interface to the ifnet structure. The new interface allows the ifnet structure to be defined as an opaque type in NIC drivers. This then allows the ifnet structure to be changed without a need to change or recompile NIC drivers. Put differently, NIC drivers can be written and compiled once and be used with different network stack implementations, provided of course that those network stack implementations have an API and ABI compatible interface. This commit introduces the 'if_t' type to replace 'struct ifnet *' as the type of a network interface. The 'if_t' type is defined as 'void *' to enable the compiler to perform type conversion to 'struct ifnet *' and vice versa where needed and without warnings. The functions that implement the API are the only functions that need to have an explicit cast. The MII code has been converted to use the driver API to avoid unnecessary code churn. Code churn comes from having to work with both converted and unconverted drivers in correlation with having callback functions that take an interface. By converting the MII code first, the callback functions can be defined so that the compiler will perform the typecasts automatically. As soon as all drivers have been converted, the if_t type can be redefined as needed and the API functions can be fix to not need an explicit cast. The immediate benefactors of this change are: 1. Juniper Networks - The network stack implementation in Junos is entirely different from FreeBSD's one and this change allows Juniper to build "stock" NIC drivers that can be used in combination with both the FreeBSD and Junos stacks. 2. FreeBSD - This change opens the door towards changing ifnet and implementing new features and optimizations in the network stack without it requiring a change in the many NIC drivers FreeBSD has. Submitted by: Anuranjan Shukla <anshukla@juniper.net> Reviewed by: glebius@ Obtained from: Juniper Networks, Inc.
2014-06-02 17:54:39 +00:00
int if_setmtu(if_t ifp, int mtu);
int if_getmtu(if_t ifp);
Make checks for rt_mtu generic: Some virtual if drivers has (ab)used ifa ifa_rtrequest hook to enforce route MTU to be not bigger that interface MTU. While ifa_rtrequest hooking might be an option in some situation, it is not feasible to do MTU checks there: generic (or per-domain) routing code is perfectly capable of doing this. We currrently have 3 places where MTU is altered: 1) route addition. In this case domain overrides radix _addroute callback (in[6]_addroute) and all necessary checks/fixes are/can be done there. 2) route change (especially, GW change). In this case, there are no explicit per-domain calls, but one can override rte by setting ifa_rtrequest hook to domain handler (inet6 does this). 3) ifconfig ifaceX mtu YYYY In this case, we have no callbacks, but ip[6]_output performes runtime checks and decreases rt_mtu if necessary. Generally, the goals are to be able to handle all MTU changes in control plane, not in runtime part, and properly deal with increased interface MTU. This commit changes the following: * removes hooks setting MTU from drivers side * adds proper per-doman MTU checks for case 1) * adds generic MTU check for case 2) * The latter is done by using new dom_ifmtu callback since if_mtu denotes L3 interface MTU, e.g. maximum trasmitted _packet_ size. However, IPv6 mtu might be different from if_mtu one (e.g. default 1280) for some cases, so we need an abstract way to know maximum MTU size for given interface and domain. * moves rt_setmetrics() before MTU/ifa_rtrequest hooks since it copies user-supplied data which must be checked. * removes RT_LOCK_ASSERT() from other ifa_rtrequest hooks to be able to use this functions on new non-inserted rte. More changes will follow soon. MFC after: 1 month Sponsored by: Yandex LLC
2014-11-06 13:13:09 +00:00
int if_getmtu_family(if_t ifp, int family);
Introduce a procedural interface to the ifnet structure. The new interface allows the ifnet structure to be defined as an opaque type in NIC drivers. This then allows the ifnet structure to be changed without a need to change or recompile NIC drivers. Put differently, NIC drivers can be written and compiled once and be used with different network stack implementations, provided of course that those network stack implementations have an API and ABI compatible interface. This commit introduces the 'if_t' type to replace 'struct ifnet *' as the type of a network interface. The 'if_t' type is defined as 'void *' to enable the compiler to perform type conversion to 'struct ifnet *' and vice versa where needed and without warnings. The functions that implement the API are the only functions that need to have an explicit cast. The MII code has been converted to use the driver API to avoid unnecessary code churn. Code churn comes from having to work with both converted and unconverted drivers in correlation with having callback functions that take an interface. By converting the MII code first, the callback functions can be defined so that the compiler will perform the typecasts automatically. As soon as all drivers have been converted, the if_t type can be redefined as needed and the API functions can be fix to not need an explicit cast. The immediate benefactors of this change are: 1. Juniper Networks - The network stack implementation in Junos is entirely different from FreeBSD's one and this change allows Juniper to build "stock" NIC drivers that can be used in combination with both the FreeBSD and Junos stacks. 2. FreeBSD - This change opens the door towards changing ifnet and implementing new features and optimizations in the network stack without it requiring a change in the many NIC drivers FreeBSD has. Submitted by: Anuranjan Shukla <anshukla@juniper.net> Reviewed by: glebius@ Obtained from: Juniper Networks, Inc.
2014-06-02 17:54:39 +00:00
int if_setflagbits(if_t ifp, int set, int clear);
int if_getflags(if_t ifp);
int if_sendq_empty(if_t ifp);
int if_setsendqready(if_t ifp);
int if_setsendqlen(if_t ifp, int tx_desc_count);
int if_sethwtsomax(if_t ifp, u_int if_hw_tsomax);
int if_sethwtsomaxsegcount(if_t ifp, u_int if_hw_tsomaxsegcount);
int if_sethwtsomaxsegsize(if_t ifp, u_int if_hw_tsomaxsegsize);
u_int if_gethwtsomax(if_t ifp);
u_int if_gethwtsomaxsegcount(if_t ifp);
u_int if_gethwtsomaxsegsize(if_t ifp);
Introduce a procedural interface to the ifnet structure. The new interface allows the ifnet structure to be defined as an opaque type in NIC drivers. This then allows the ifnet structure to be changed without a need to change or recompile NIC drivers. Put differently, NIC drivers can be written and compiled once and be used with different network stack implementations, provided of course that those network stack implementations have an API and ABI compatible interface. This commit introduces the 'if_t' type to replace 'struct ifnet *' as the type of a network interface. The 'if_t' type is defined as 'void *' to enable the compiler to perform type conversion to 'struct ifnet *' and vice versa where needed and without warnings. The functions that implement the API are the only functions that need to have an explicit cast. The MII code has been converted to use the driver API to avoid unnecessary code churn. Code churn comes from having to work with both converted and unconverted drivers in correlation with having callback functions that take an interface. By converting the MII code first, the callback functions can be defined so that the compiler will perform the typecasts automatically. As soon as all drivers have been converted, the if_t type can be redefined as needed and the API functions can be fix to not need an explicit cast. The immediate benefactors of this change are: 1. Juniper Networks - The network stack implementation in Junos is entirely different from FreeBSD's one and this change allows Juniper to build "stock" NIC drivers that can be used in combination with both the FreeBSD and Junos stacks. 2. FreeBSD - This change opens the door towards changing ifnet and implementing new features and optimizations in the network stack without it requiring a change in the many NIC drivers FreeBSD has. Submitted by: Anuranjan Shukla <anshukla@juniper.net> Reviewed by: glebius@ Obtained from: Juniper Networks, Inc.
2014-06-02 17:54:39 +00:00
int if_input(if_t ifp, struct mbuf* sendmp);
int if_sendq_prepend(if_t ifp, struct mbuf *m);
struct mbuf *if_dequeue(if_t ifp);
int if_setifheaderlen(if_t ifp, int len);
void if_setrcvif(struct mbuf *m, if_t ifp);
void if_setvtag(struct mbuf *m, u_int16_t tag);
u_int16_t if_getvtag(struct mbuf *m);
int if_vlantrunkinuse(if_t ifp);
caddr_t if_getlladdr(if_t ifp);
void *if_gethandle(u_char);
void if_bpfmtap(if_t ifp, struct mbuf *m);
void if_etherbpfmtap(if_t ifp, struct mbuf *m);
void if_vlancap(if_t ifp);
/*
* Traversing through interface address lists.
*/
struct sockaddr_dl;
typedef u_int iflladdr_cb_t(void *, struct sockaddr_dl *, u_int);
u_int if_foreach_lladdr(if_t, iflladdr_cb_t, void *);
u_int if_foreach_llmaddr(if_t, iflladdr_cb_t, void *);
u_int if_lladdr_count(if_t);
u_int if_llmaddr_count(if_t);
Introduce a procedural interface to the ifnet structure. The new interface allows the ifnet structure to be defined as an opaque type in NIC drivers. This then allows the ifnet structure to be changed without a need to change or recompile NIC drivers. Put differently, NIC drivers can be written and compiled once and be used with different network stack implementations, provided of course that those network stack implementations have an API and ABI compatible interface. This commit introduces the 'if_t' type to replace 'struct ifnet *' as the type of a network interface. The 'if_t' type is defined as 'void *' to enable the compiler to perform type conversion to 'struct ifnet *' and vice versa where needed and without warnings. The functions that implement the API are the only functions that need to have an explicit cast. The MII code has been converted to use the driver API to avoid unnecessary code churn. Code churn comes from having to work with both converted and unconverted drivers in correlation with having callback functions that take an interface. By converting the MII code first, the callback functions can be defined so that the compiler will perform the typecasts automatically. As soon as all drivers have been converted, the if_t type can be redefined as needed and the API functions can be fix to not need an explicit cast. The immediate benefactors of this change are: 1. Juniper Networks - The network stack implementation in Junos is entirely different from FreeBSD's one and this change allows Juniper to build "stock" NIC drivers that can be used in combination with both the FreeBSD and Junos stacks. 2. FreeBSD - This change opens the door towards changing ifnet and implementing new features and optimizations in the network stack without it requiring a change in the many NIC drivers FreeBSD has. Submitted by: Anuranjan Shukla <anshukla@juniper.net> Reviewed by: glebius@ Obtained from: Juniper Networks, Inc.
2014-06-02 17:54:39 +00:00
int if_getamcount(if_t ifp);
struct ifaddr * if_getifaddr(if_t ifp);
/* Functions */
void if_setinitfn(if_t ifp, void (*)(void *));
void if_setioctlfn(if_t ifp, int (*)(if_t, u_long, caddr_t));
void if_setstartfn(if_t ifp, void (*)(if_t));
Introduce a procedural interface to the ifnet structure. The new interface allows the ifnet structure to be defined as an opaque type in NIC drivers. This then allows the ifnet structure to be changed without a need to change or recompile NIC drivers. Put differently, NIC drivers can be written and compiled once and be used with different network stack implementations, provided of course that those network stack implementations have an API and ABI compatible interface. This commit introduces the 'if_t' type to replace 'struct ifnet *' as the type of a network interface. The 'if_t' type is defined as 'void *' to enable the compiler to perform type conversion to 'struct ifnet *' and vice versa where needed and without warnings. The functions that implement the API are the only functions that need to have an explicit cast. The MII code has been converted to use the driver API to avoid unnecessary code churn. Code churn comes from having to work with both converted and unconverted drivers in correlation with having callback functions that take an interface. By converting the MII code first, the callback functions can be defined so that the compiler will perform the typecasts automatically. As soon as all drivers have been converted, the if_t type can be redefined as needed and the API functions can be fix to not need an explicit cast. The immediate benefactors of this change are: 1. Juniper Networks - The network stack implementation in Junos is entirely different from FreeBSD's one and this change allows Juniper to build "stock" NIC drivers that can be used in combination with both the FreeBSD and Junos stacks. 2. FreeBSD - This change opens the door towards changing ifnet and implementing new features and optimizations in the network stack without it requiring a change in the many NIC drivers FreeBSD has. Submitted by: Anuranjan Shukla <anshukla@juniper.net> Reviewed by: glebius@ Obtained from: Juniper Networks, Inc.
2014-06-02 17:54:39 +00:00
void if_settransmitfn(if_t ifp, if_transmit_fn_t);
void if_setqflushfn(if_t ifp, if_qflush_fn_t);
void if_setgetcounterfn(if_t ifp, if_get_counter_t);
/* TSO */
void if_hw_tsomax_common(if_t ifp, struct ifnet_hw_tsomax *);
int if_hw_tsomax_update(if_t ifp, struct ifnet_hw_tsomax *);
/* accessors for struct ifreq */
void *ifr_data_get_ptr(void *ifrp);
void *ifr_buffer_get_buffer(void *data);
size_t ifr_buffer_get_length(void *data);
int ifhwioctl(u_long, struct ifnet *, caddr_t, struct thread *);
#ifdef DEVICE_POLLING
enum poll_cmd { POLL_ONLY, POLL_AND_CHECK_STATUS };
typedef int poll_handler_t(if_t ifp, enum poll_cmd cmd, int count);
int ether_poll_register(poll_handler_t *h, if_t ifp);
int ether_poll_deregister(if_t ifp);
#endif /* DEVICE_POLLING */
#endif /* _KERNEL */
#include <net/ifq.h> /* XXXAO: temporary unconditional include */
#endif /* !_NET_IF_VAR_H_ */