freebsd-nq/sys/netinet/tcp_var.h

1182 lines
47 KiB
C
Raw Normal View History

/*-
* SPDX-License-Identifier: BSD-3-Clause
*
* Copyright (c) 1982, 1986, 1993, 1994, 1995
1994-05-24 10:09:53 +00:00
* 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
1994-05-24 10:09:53 +00:00
* 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.
*
* @(#)tcp_var.h 8.4 (Berkeley) 5/24/95
1999-08-28 01:08:13 +00:00
* $FreeBSD$
1994-05-24 10:09:53 +00:00
*/
1994-08-21 05:27:42 +00:00
#ifndef _NETINET_TCP_VAR_H_
#define _NETINET_TCP_VAR_H_
#include <netinet/tcp.h>
#include <netinet/tcp_fsm.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
#ifdef _KERNEL
#include "opt_kern_tls.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>
There are times when it would be really nice to have a record of the last few packets and/or state transitions from each TCP socket. That would help with narrowing down certain problems we see in the field that are hard to reproduce without understanding the history of how we got into a certain state. This change provides just that. It saves copies of the last N packets in a list in the tcpcb. When the tcpcb is destroyed, the list is freed. I thought this was likely to be more performance-friendly than saving copies of the tcpcb. Plus, with the packets, you should be able to reverse-engineer what happened to the tcpcb. To enable the feature, you will need to compile a kernel with the TCPPCAP option. Even then, the feature defaults to being deactivated. You can activate it by setting a positive value for the number of captured packets. You can do that on either a global basis or on a per-socket basis (via a setsockopt call). There is no way to get the packets out of the kernel other than using kmem or getting a coredump. I thought that would help some of the legal/privacy concerns regarding such a feature. However, it should be possible to add a future effort to export them in PCAP format. I tested this at low scale, and found that there were no mbuf leaks and the peak mbuf usage appeared to be unchanged with and without the feature. The main performance concern I can envision is the number of mbufs that would be used on systems with a large number of sockets. If you save five packets per direction per socket and have 3,000 sockets, that will consume at least 30,000 mbufs just to keep these packets. I tried to reduce the concerns associated with this by limiting the number of clusters (not mbufs) that could be used for this feature. Again, in my testing, that appears to work correctly. Differential Revision: D3100 Submitted by: Jonathan Looney <jlooney at juniper dot net> Reviewed by: gnn, hiren
2015-10-14 00:35:37 +00:00
#include <sys/mbuf.h>
#include <sys/ktls.h>
#endif
Permit buiding kernels with options VIMAGE, restricted to only a single active network stack instance. Turning on options VIMAGE at compile time yields the following changes relative to default kernel build: 1) V_ accessor macros for virtualized variables resolve to structure fields via base pointers, instead of being resolved as fields in global structs or plain global variables. As an example, V_ifnet becomes: options VIMAGE: ((struct vnet_net *) vnet_net)->_ifnet default build: vnet_net_0._ifnet options VIMAGE_GLOBALS: ifnet 2) INIT_VNET_* macros will declare and set up base pointers to be used by V_ accessor macros, instead of resolving to whitespace: INIT_VNET_NET(ifp->if_vnet); becomes struct vnet_net *vnet_net = (ifp->if_vnet)->mod_data[VNET_MOD_NET]; 3) Memory for vnet modules registered via vnet_mod_register() is now allocated at run time in sys/kern/kern_vimage.c, instead of per vnet module structs being declared as globals. If required, vnet modules can now request the framework to provide them with allocated bzeroed memory by filling in the vmi_size field in their vmi_modinfo structures. 4) structs socket, ifnet, inpcbinfo, tcpcb and syncache_head are extended to hold a pointer to the parent vnet. options VIMAGE builds will fill in those fields as required. 5) curvnet is introduced as a new global variable in options VIMAGE builds, always pointing to the default and only struct vnet. 6) struct sysctl_oid has been extended with additional two fields to store major and minor virtualization module identifiers, oid_v_subs and oid_v_mod. SYSCTL_V_* family of macros will fill in those fields accordingly, and store the offset in the appropriate vnet container struct in oid_arg1. In sysctl handlers dealing with virtualized sysctls, the SYSCTL_RESOLVE_V_ARG1() macro will compute the address of the target variable and make it available in arg1 variable for further processing. Unused fields in structs vnet_inet, vnet_inet6 and vnet_ipfw have been deleted. Reviewed by: bz, rwatson Approved by: julian (mentor)
2009-04-30 13:36:26 +00:00
#define TCP_END_BYTE_INFO 8 /* Bytes that makeup the "end information array" */
/* Types of ending byte info */
#define TCP_EI_EMPTY_SLOT 0
#define TCP_EI_STATUS_CLIENT_FIN 0x1
#define TCP_EI_STATUS_CLIENT_RST 0x2
#define TCP_EI_STATUS_SERVER_FIN 0x3
#define TCP_EI_STATUS_SERVER_RST 0x4
#define TCP_EI_STATUS_RETRAN 0x5
#define TCP_EI_STATUS_PROGRESS 0x6
#define TCP_EI_STATUS_PERSIST_MAX 0x7
#define TCP_EI_STATUS_KEEP_MAX 0x8
#define TCP_EI_STATUS_DATA_A_CLOSE 0x9
#define TCP_EI_STATUS_RST_IN_FRONT 0xa
#define TCP_EI_STATUS_2MSL 0xb
#define TCP_EI_STATUS_MAX_VALUE 0xb
/************************************************/
/* Status bits we track to assure no duplicates,
* the bits here are not used by the code but
* for human representation. To check a bit we
* take and shift over by 1 minus the value (1-8).
*/
/************************************************/
#define TCP_EI_BITS_CLIENT_FIN 0x001
#define TCP_EI_BITS_CLIENT_RST 0x002
#define TCP_EI_BITS_SERVER_FIN 0x004
#define TCP_EI_BITS_SERVER_RST 0x008
#define TCP_EI_BITS_RETRAN 0x010
#define TCP_EI_BITS_PROGRESS 0x020
#define TCP_EI_BITS_PRESIST_MAX 0x040
#define TCP_EI_BITS_KEEP_MAX 0x080
#define TCP_EI_BITS_DATA_A_CLO 0x100
#define TCP_EI_BITS_RST_IN_FR 0x200 /* a front state reset */
#define TCP_EI_BITS_2MS_TIMER 0x400 /* 2 MSL timer expired */
#if defined(_KERNEL) || defined(_WANT_TCPCB)
/* TCP segment queue entry */
struct tseg_qent {
TAILQ_ENTRY(tseg_qent) tqe_q;
struct mbuf *tqe_m; /* mbuf contains packet */
struct mbuf *tqe_last; /* last mbuf in chain */
tcp_seq tqe_start; /* TCP Sequence number start */
int tqe_len; /* TCP segment data length */
uint32_t tqe_flags; /* The flags from the th->th_flags */
uint32_t tqe_mbuf_cnt; /* Count of mbuf overhead */
};
TAILQ_HEAD(tsegqe_head, tseg_qent);
struct sackblk {
tcp_seq start; /* start seq no. of sack block */
2004-06-25 02:29:58 +00:00
tcp_seq end; /* end seq no. */
};
struct sackhole {
tcp_seq start; /* start seq no. of hole */
tcp_seq end; /* end seq no. */
tcp_seq rxmit; /* next seq. no in hole to be retransmitted */
TAILQ_ENTRY(sackhole) scblink; /* scoreboard linkage */
};
2004-06-25 02:29:58 +00:00
struct sackhint {
struct sackhole *nexthole;
int32_t sack_bytes_rexmit;
tcp_seq last_sack_ack; /* Most recent/largest sacked ack */
int32_t delivered_data; /* Newly acked data from last SACK */
int32_t sacked_bytes; /* Total sacked bytes reported by the
Calculate the correct amount of bytes that are in-flight for a connection as suggested by RFC 6675. Currently differnt places in the stack tries to guess this in suboptimal ways. The main problem is that current calculations don't take sacked bytes into account. Sacked bytes are the bytes receiver acked via SACK option. This is suboptimal because it assumes that network has more outstanding (unacked) bytes than the actual value and thus sends less data by setting congestion window lower than what's possible which in turn may cause slower recovery from losses. As an example, one of the current calculations looks something like this: snd_nxt - snd_fack + sackhint.sack_bytes_rexmit New proposal from RFC 6675 is: snd_max - snd_una - sackhint.sacked_bytes + sackhint.sack_bytes_rexmit which takes sacked bytes into account which is a new addition to the sackhint struct. Only thing we are missing from RFC 6675 is isLost() i.e. segment being considered lost and thus adjusting pipe based on that which makes this calculation a bit on conservative side. The approach is very simple. We already process each ack with sack info in tcp_sack_doack() and extract sack blocks/holes out of it. We'd now also track this new variable sacked_bytes which keeps track of total sacked bytes reported. One downside to this approach is that we may get incorrect count of sacked_bytes if the other end decides to drop sack info in the ack because of memory pressure or some other reasons. But in this (not very likely) case also the pipe calculation would be conservative which is okay as opposed to being aggressive in sending packets into the network. Next step is to use this more accurate pipe estimation to drive congestion window adjustments. In collaboration with: rrs Reviewed by: jason_eggnet dot com, rrs MFC after: 2 weeks Sponsored by: Limelight Networks Differential Revision: https://reviews.freebsd.org/D3971
2015-10-28 22:57:51 +00:00
* receiver via sack option
*/
uint32_t recover_fs; /* Flight Size at the start of Loss recovery */
uint32_t prr_delivered; /* Total bytes delivered using PRR */
uint32_t prr_out; /* Bytes sent during IN_RECOVERY */
};
#define SEGQ_EMPTY(tp) TAILQ_EMPTY(&(tp)->t_segq)
Add the "TCP Blackbox Recorder" which we discussed at the developer summits at BSDCan and BSDCam in 2017. The TCP Blackbox Recorder allows you to capture events on a TCP connection in a ring buffer. It stores metadata with the event. It optionally stores the TCP header associated with an event (if the event is associated with a packet) and also optionally stores information on the sockets. It supports setting a log ID on a TCP connection and using this to correlate multiple connections that share a common log ID. You can log connections in different modes. If you are doing a coordinated test with a particular connection, you may tell the system to put it in mode 4 (continuous dump). Or, if you just want to monitor for errors, you can put it in mode 1 (ring buffer) and dump all the ring buffers associated with the connection ID when we receive an error signal for that connection ID. You can set a default mode that will be applied to a particular ratio of incoming connections. You can also manually set a mode using a socket option. This commit includes only basic probes. rrs@ has added quite an abundance of probes in his TCP development work. He plans to commit those soon. There are user-space programs which we plan to commit as ports. These read the data from the log device and output pcapng files, and then let you analyze the data (and metadata) in the pcapng files. Reviewed by: gnn (previous version) Obtained from: Netflix, Inc. Relnotes: yes Differential Revision: https://reviews.freebsd.org/D11085
2018-03-22 09:40:08 +00:00
STAILQ_HEAD(tcp_log_stailq, tcp_log_mem);
1994-05-24 10:09:53 +00:00
/*
* Tcp control block, one per tcp; fields:
* Organized for 64 byte cacheline efficiency based
* on common tcp_input/tcp_output processing.
1994-05-24 10:09:53 +00:00
*/
struct tcpcb {
/* Cache line 1 */
Improved connection establishment performance by doing local port lookups via a hashed port list. In the new scheme, in_pcblookup() goes away and is replaced by a new routine, in_pcblookup_local() for doing the local port check. Note that this implementation is space inefficient in that the PCB struct is now too large to fit into 128 bytes. I might deal with this in the future by using the new zone allocator, but I wanted these changes to be extensively tested in their current form first. Also: 1) Fixed off-by-one errors in the port lookup loops in in_pcbbind(). 2) Got rid of some unneeded rehashing. Adding a new routine, in_pcbinshash() to do the initialial hash insertion. 3) Renamed in_pcblookuphash() to in_pcblookup_hash() for easier readability. 4) Added a new routine, in_pcbremlists() to remove the PCB from the various hash lists. 5) Added/deleted comments where appropriate. 6) Removed unnecessary splnet() locking. In general, the PCB functions should be called at splnet()...there are unfortunately a few exceptions, however. 7) Reorganized a few structs for better cache line behavior. 8) Killed my TCP_ACK_HACK kludge. It may come back in a different form in the future, however. These changes have been tested on wcarchive for more than a month. In tests done here, connection establishment overhead is reduced by more than 50 times, thus getting rid of one of the major networking scalability problems. Still to do: make tcp_fastimo/tcp_slowtimo scale well for systems with a large number of connections. tcp_fastimo is easy; tcp_slowtimo is difficult. WARNING: Anything that knows about inpcb and tcpcb structs will have to be recompiled; at the very least, this includes netstat(1).
1998-01-27 09:15:13 +00:00
struct inpcb *t_inpcb; /* back pointer to internet pcb */
struct tcp_function_block *t_fb;/* TCP function call block */
void *t_fb_ptr; /* Pointer to t_fb specific data */
uint32_t t_maxseg:24, /* maximum segment size */
t_logstate:8; /* State of "black box" logging */
uint32_t t_port:16, /* Tunneling (over udp) port */
t_state:4, /* state of this connection */
t_idle_reduce : 1,
t_delayed_ack: 7, /* Delayed ack variable */
t_fin_is_rst: 1, /* Are fin's treated as resets */
t_log_state_set: 1,
bits_spare : 2;
u_int t_flags;
tcp_seq snd_una; /* sent but unacknowledged */
Improved connection establishment performance by doing local port lookups via a hashed port list. In the new scheme, in_pcblookup() goes away and is replaced by a new routine, in_pcblookup_local() for doing the local port check. Note that this implementation is space inefficient in that the PCB struct is now too large to fit into 128 bytes. I might deal with this in the future by using the new zone allocator, but I wanted these changes to be extensively tested in their current form first. Also: 1) Fixed off-by-one errors in the port lookup loops in in_pcbbind(). 2) Got rid of some unneeded rehashing. Adding a new routine, in_pcbinshash() to do the initialial hash insertion. 3) Renamed in_pcblookuphash() to in_pcblookup_hash() for easier readability. 4) Added a new routine, in_pcbremlists() to remove the PCB from the various hash lists. 5) Added/deleted comments where appropriate. 6) Removed unnecessary splnet() locking. In general, the PCB functions should be called at splnet()...there are unfortunately a few exceptions, however. 7) Reorganized a few structs for better cache line behavior. 8) Killed my TCP_ACK_HACK kludge. It may come back in a different form in the future, however. These changes have been tested on wcarchive for more than a month. In tests done here, connection establishment overhead is reduced by more than 50 times, thus getting rid of one of the major networking scalability problems. Still to do: make tcp_fastimo/tcp_slowtimo scale well for systems with a large number of connections. tcp_fastimo is easy; tcp_slowtimo is difficult. WARNING: Anything that knows about inpcb and tcpcb structs will have to be recompiled; at the very least, this includes netstat(1).
1998-01-27 09:15:13 +00:00
tcp_seq snd_max; /* highest sequence number sent;
* used to recognize retransmits
*/
1994-05-24 10:09:53 +00:00
tcp_seq snd_nxt; /* send next */
tcp_seq snd_up; /* send urgent pointer */
uint32_t snd_wnd; /* send window */
uint32_t snd_cwnd; /* congestion-controlled window */
uint32_t t_peakrate_thr; /* pre-calculated peak rate threshold */
/* Cache line 2 */
u_int32_t ts_offset; /* our timestamp offset */
u_int32_t rfbuf_ts; /* recv buffer autoscaling timestamp */
int rcv_numsacks; /* # distinct sack blks present */
u_int t_tsomax; /* TSO total burst length limit in bytes */
u_int t_tsomaxsegcount; /* TSO maximum segment count */
u_int t_tsomaxsegsize; /* TSO maximum segment size in bytes */
Improved connection establishment performance by doing local port lookups via a hashed port list. In the new scheme, in_pcblookup() goes away and is replaced by a new routine, in_pcblookup_local() for doing the local port check. Note that this implementation is space inefficient in that the PCB struct is now too large to fit into 128 bytes. I might deal with this in the future by using the new zone allocator, but I wanted these changes to be extensively tested in their current form first. Also: 1) Fixed off-by-one errors in the port lookup loops in in_pcbbind(). 2) Got rid of some unneeded rehashing. Adding a new routine, in_pcbinshash() to do the initialial hash insertion. 3) Renamed in_pcblookuphash() to in_pcblookup_hash() for easier readability. 4) Added a new routine, in_pcbremlists() to remove the PCB from the various hash lists. 5) Added/deleted comments where appropriate. 6) Removed unnecessary splnet() locking. In general, the PCB functions should be called at splnet()...there are unfortunately a few exceptions, however. 7) Reorganized a few structs for better cache line behavior. 8) Killed my TCP_ACK_HACK kludge. It may come back in a different form in the future, however. These changes have been tested on wcarchive for more than a month. In tests done here, connection establishment overhead is reduced by more than 50 times, thus getting rid of one of the major networking scalability problems. Still to do: make tcp_fastimo/tcp_slowtimo scale well for systems with a large number of connections. tcp_fastimo is easy; tcp_slowtimo is difficult. WARNING: Anything that knows about inpcb and tcpcb structs will have to be recompiled; at the very least, this includes netstat(1).
1998-01-27 09:15:13 +00:00
tcp_seq rcv_nxt; /* receive next */
1994-05-24 10:09:53 +00:00
tcp_seq rcv_adv; /* advertised window */
uint32_t rcv_wnd; /* receive window */
u_int t_flags2; /* More tcpcb flags storage */
int t_srtt; /* smoothed round-trip time */
int t_rttvar; /* variance in round-trip time */
u_int32_t ts_recent; /* timestamp echo data */
u_char snd_scale; /* window scaling for send window */
u_char rcv_scale; /* window scaling for recv window */
u_char snd_limited; /* segments limited transmitted */
u_char request_r_scale; /* pending window scaling */
tcp_seq last_ack_sent;
u_int t_rcvtime; /* inactivity time */
/* Cache line 3 */
Improved connection establishment performance by doing local port lookups via a hashed port list. In the new scheme, in_pcblookup() goes away and is replaced by a new routine, in_pcblookup_local() for doing the local port check. Note that this implementation is space inefficient in that the PCB struct is now too large to fit into 128 bytes. I might deal with this in the future by using the new zone allocator, but I wanted these changes to be extensively tested in their current form first. Also: 1) Fixed off-by-one errors in the port lookup loops in in_pcbbind(). 2) Got rid of some unneeded rehashing. Adding a new routine, in_pcbinshash() to do the initialial hash insertion. 3) Renamed in_pcblookuphash() to in_pcblookup_hash() for easier readability. 4) Added a new routine, in_pcbremlists() to remove the PCB from the various hash lists. 5) Added/deleted comments where appropriate. 6) Removed unnecessary splnet() locking. In general, the PCB functions should be called at splnet()...there are unfortunately a few exceptions, however. 7) Reorganized a few structs for better cache line behavior. 8) Killed my TCP_ACK_HACK kludge. It may come back in a different form in the future, however. These changes have been tested on wcarchive for more than a month. In tests done here, connection establishment overhead is reduced by more than 50 times, thus getting rid of one of the major networking scalability problems. Still to do: make tcp_fastimo/tcp_slowtimo scale well for systems with a large number of connections. tcp_fastimo is easy; tcp_slowtimo is difficult. WARNING: Anything that knows about inpcb and tcpcb structs will have to be recompiled; at the very least, this includes netstat(1).
1998-01-27 09:15:13 +00:00
tcp_seq rcv_up; /* receive urgent pointer */
int t_segqlen; /* segment reassembly queue length */
uint32_t t_segqmbuflen; /* Count of bytes mbufs on all entries */
struct tsegqe_head t_segq; /* segment reassembly queue */
struct mbuf *t_in_pkt;
struct mbuf *t_tail_pkt;
struct tcp_timer *t_timers; /* All the TCP timers in one struct */
struct vnet *t_vnet; /* back pointer to parent vnet */
uint32_t snd_ssthresh; /* snd_cwnd size threshold for
1994-05-24 10:09:53 +00:00
* for slow start exponential to
* linear switch
*/
tcp_seq snd_wl1; /* window update seg seq number */
/* Cache line 4 */
tcp_seq snd_wl2; /* window update seg ack number */
tcp_seq irs; /* initial receive sequence number */
tcp_seq iss; /* initial send sequence number */
u_int t_acktime; /* RACK and BBR incoming new data was acked */
u_int t_sndtime; /* time last data was sent */
u_int ts_recent_age; /* when last updated */
tcp_seq snd_recover; /* for use in NewReno Fast Recovery */
uint16_t cl4_spare; /* Spare to adjust CL 4 */
char t_oobflags; /* have some */
char t_iobc; /* input character */
int t_rxtcur; /* current retransmit value (ticks) */
int t_rxtshift; /* log(2) of rexmt exp. backoff */
u_int t_rtttime; /* RTT measurement start time */
1994-05-24 10:09:53 +00:00
tcp_seq t_rtseq; /* sequence number being timed */
u_int t_starttime; /* time connection was established */
u_int t_fbyte_in; /* ticks time when first byte queued in */
u_int t_fbyte_out; /* ticks time when first byte queued out */
Improved connection establishment performance by doing local port lookups via a hashed port list. In the new scheme, in_pcblookup() goes away and is replaced by a new routine, in_pcblookup_local() for doing the local port check. Note that this implementation is space inefficient in that the PCB struct is now too large to fit into 128 bytes. I might deal with this in the future by using the new zone allocator, but I wanted these changes to be extensively tested in their current form first. Also: 1) Fixed off-by-one errors in the port lookup loops in in_pcbbind(). 2) Got rid of some unneeded rehashing. Adding a new routine, in_pcbinshash() to do the initialial hash insertion. 3) Renamed in_pcblookuphash() to in_pcblookup_hash() for easier readability. 4) Added a new routine, in_pcbremlists() to remove the PCB from the various hash lists. 5) Added/deleted comments where appropriate. 6) Removed unnecessary splnet() locking. In general, the PCB functions should be called at splnet()...there are unfortunately a few exceptions, however. 7) Reorganized a few structs for better cache line behavior. 8) Killed my TCP_ACK_HACK kludge. It may come back in a different form in the future, however. These changes have been tested on wcarchive for more than a month. In tests done here, connection establishment overhead is reduced by more than 50 times, thus getting rid of one of the major networking scalability problems. Still to do: make tcp_fastimo/tcp_slowtimo scale well for systems with a large number of connections. tcp_fastimo is easy; tcp_slowtimo is difficult. WARNING: Anything that knows about inpcb and tcpcb structs will have to be recompiled; at the very least, this includes netstat(1).
1998-01-27 09:15:13 +00:00
u_int t_pmtud_saved_maxseg; /* pre-blackhole MSS */
int t_blackhole_enter; /* when to enter blackhole detection */
int t_blackhole_exit; /* when to exit blackhole detection */
u_int t_rttmin; /* minimum rtt allowed */
u_int t_rttbest; /* best rtt we've seen */
1994-05-24 10:09:53 +00:00
Improved connection establishment performance by doing local port lookups via a hashed port list. In the new scheme, in_pcblookup() goes away and is replaced by a new routine, in_pcblookup_local() for doing the local port check. Note that this implementation is space inefficient in that the PCB struct is now too large to fit into 128 bytes. I might deal with this in the future by using the new zone allocator, but I wanted these changes to be extensively tested in their current form first. Also: 1) Fixed off-by-one errors in the port lookup loops in in_pcbbind(). 2) Got rid of some unneeded rehashing. Adding a new routine, in_pcbinshash() to do the initialial hash insertion. 3) Renamed in_pcblookuphash() to in_pcblookup_hash() for easier readability. 4) Added a new routine, in_pcbremlists() to remove the PCB from the various hash lists. 5) Added/deleted comments where appropriate. 6) Removed unnecessary splnet() locking. In general, the PCB functions should be called at splnet()...there are unfortunately a few exceptions, however. 7) Reorganized a few structs for better cache line behavior. 8) Killed my TCP_ACK_HACK kludge. It may come back in a different form in the future, however. These changes have been tested on wcarchive for more than a month. In tests done here, connection establishment overhead is reduced by more than 50 times, thus getting rid of one of the major networking scalability problems. Still to do: make tcp_fastimo/tcp_slowtimo scale well for systems with a large number of connections. tcp_fastimo is easy; tcp_slowtimo is difficult. WARNING: Anything that knows about inpcb and tcpcb structs will have to be recompiled; at the very least, this includes netstat(1).
1998-01-27 09:15:13 +00:00
int t_softerror; /* possible error not yet reported */
uint32_t max_sndwnd; /* largest window peer has offered */
/* Cache line 5 */
uint32_t snd_cwnd_prev; /* cwnd prior to retransmit */
uint32_t snd_ssthresh_prev; /* ssthresh prior to retransmit */
tcp_seq snd_recover_prev; /* snd_recover prior to retransmit */
int t_sndzerowin; /* zero-window updates sent */
u_long t_rttupdated; /* number of times rtt sampled */
int snd_numholes; /* number of holes seen by sender */
u_int t_badrxtwin; /* window for retransmit recovery */
TAILQ_HEAD(sackhole_head, sackhole) snd_holes;
/* SACK scoreboard (sorted) */
tcp_seq snd_fack; /* last seq number(+1) sack'd by rcv'r*/
struct sackblk sackblks[MAX_SACK_BLKS]; /* seq nos. of sack blocks */
struct sackhint sackhint; /* SACK scoreboard hint */
int t_rttlow; /* smallest observerved RTT */
int rfbuf_cnt; /* recv buffer autoscaling byte count */
struct toedev *tod; /* toedev handling this connection */
int t_sndrexmitpack; /* retransmit packets sent */
int t_rcvoopack; /* out-of-order packets received */
void *t_toe; /* TOE pcb pointer */
This commit marks the first formal contribution of the "Five New TCP Congestion Control Algorithms for FreeBSD" FreeBSD Foundation funded project. More details about the project are available at: http://caia.swin.edu.au/freebsd/5cc/ - Add a KPI and supporting infrastructure to allow modular congestion control algorithms to be used in the net stack. Algorithms can maintain per-connection state if required, and connections maintain their own algorithm pointer, which allows different connections to concurrently use different algorithms. The TCP_CONGESTION socket option can be used with getsockopt()/setsockopt() to programmatically query or change the congestion control algorithm respectively from within an application at runtime. - Integrate the framework with the TCP stack in as least intrusive a manner as possible. Care was also taken to develop the framework in a way that should allow integration with other congestion aware transport protocols (e.g. SCTP) in the future. The hope is that we will one day be able to share a single set of congestion control algorithm modules between all congestion aware transport protocols. - Introduce a new congestion recovery (TF_CONGRECOVERY) state into the TCP stack and use it to decouple the meaning of recovery from a congestion event and recovery from packet loss (TF_FASTRECOVERY) a la RFC2581. ECN and delay based congestion control protocols don't generally need to recover from packet loss and need a different way to note a congestion recovery episode within the stack. - Remove the net.inet.tcp.newreno sysctl, which simplifies some portions of code and ensures the stack always uses the appropriate mechanisms for recovering from packet loss during a congestion recovery episode. - Extract the NewReno congestion control algorithm from the TCP stack and massage it into module form. NewReno is always built into the kernel and will remain the default algorithm for the forseeable future. Implementations of additional different algorithms will become available in the near future. - Bump __FreeBSD_version to 900025 and note in UPDATING that rebuilding code that relies on the size of "struct tcpcb" is required. Many thanks go to the Cisco University Research Program Fund at Community Foundation Silicon Valley and the FreeBSD Foundation. Their support of our work at the Centre for Advanced Internet Architectures, Swinburne University of Technology is greatly appreciated. In collaboration with: David Hayes <dahayes at swin edu au> and Grenville Armitage <garmitage at swin edu au> Sponsored by: Cisco URP, FreeBSD Foundation Reviewed by: rpaulo Tested by: David Hayes (and many others over the years) MFC after: 3 months
2010-11-12 06:41:55 +00:00
struct cc_algo *cc_algo; /* congestion control algorithm */
struct cc_var *ccv; /* congestion control specific vars */
struct osd *osd; /* storage for Khelp module data */
int t_bytes_acked; /* # bytes acked during current RTT */
u_int t_maxunacktime;
u_int t_keepinit; /* time to establish connection */
u_int t_keepidle; /* time before keepalive probes begin */
u_int t_keepintvl; /* interval between keepalives */
u_int t_keepcnt; /* number of keepalives before close */
int t_dupacks; /* consecutive dup acks recd */
Add the "TCP Blackbox Recorder" which we discussed at the developer summits at BSDCan and BSDCam in 2017. The TCP Blackbox Recorder allows you to capture events on a TCP connection in a ring buffer. It stores metadata with the event. It optionally stores the TCP header associated with an event (if the event is associated with a packet) and also optionally stores information on the sockets. It supports setting a log ID on a TCP connection and using this to correlate multiple connections that share a common log ID. You can log connections in different modes. If you are doing a coordinated test with a particular connection, you may tell the system to put it in mode 4 (continuous dump). Or, if you just want to monitor for errors, you can put it in mode 1 (ring buffer) and dump all the ring buffers associated with the connection ID when we receive an error signal for that connection ID. You can set a default mode that will be applied to a particular ratio of incoming connections. You can also manually set a mode using a socket option. This commit includes only basic probes. rrs@ has added quite an abundance of probes in his TCP development work. He plans to commit those soon. There are user-space programs which we plan to commit as ports. These read the data from the log device and output pcapng files, and then let you analyze the data (and metadata) in the pcapng files. Reviewed by: gnn (previous version) Obtained from: Netflix, Inc. Relnotes: yes Differential Revision: https://reviews.freebsd.org/D11085
2018-03-22 09:40:08 +00:00
int t_lognum; /* Number of log entries */
int t_loglimit; /* Maximum number of log entries */
int64_t t_pacing_rate; /* bytes / sec, -1 => unlimited */
struct tcp_log_stailq t_logs; /* Log buffer */
Add the "TCP Blackbox Recorder" which we discussed at the developer summits at BSDCan and BSDCam in 2017. The TCP Blackbox Recorder allows you to capture events on a TCP connection in a ring buffer. It stores metadata with the event. It optionally stores the TCP header associated with an event (if the event is associated with a packet) and also optionally stores information on the sockets. It supports setting a log ID on a TCP connection and using this to correlate multiple connections that share a common log ID. You can log connections in different modes. If you are doing a coordinated test with a particular connection, you may tell the system to put it in mode 4 (continuous dump). Or, if you just want to monitor for errors, you can put it in mode 1 (ring buffer) and dump all the ring buffers associated with the connection ID when we receive an error signal for that connection ID. You can set a default mode that will be applied to a particular ratio of incoming connections. You can also manually set a mode using a socket option. This commit includes only basic probes. rrs@ has added quite an abundance of probes in his TCP development work. He plans to commit those soon. There are user-space programs which we plan to commit as ports. These read the data from the log device and output pcapng files, and then let you analyze the data (and metadata) in the pcapng files. Reviewed by: gnn (previous version) Obtained from: Netflix, Inc. Relnotes: yes Differential Revision: https://reviews.freebsd.org/D11085
2018-03-22 09:40:08 +00:00
struct tcp_log_id_node *t_lin;
struct tcp_log_id_bucket *t_lib;
const char *t_output_caller; /* Function that called tcp_output */
struct statsblob *t_stats; /* Per-connection stats */
uint32_t t_logsn; /* Log "serial number" */
uint32_t gput_ts; /* Time goodput measurement started */
tcp_seq gput_seq; /* Outbound measurement seq */
tcp_seq gput_ack; /* Inbound measurement ack */
int32_t t_stats_gput_prev; /* XXXLAS: Prev gput measurement */
uint32_t t_maxpeakrate; /* max peak rate set by user, in bytes/s */
uint32_t t_sndtlppack; /* tail loss probe packets sent */
uint64_t t_sndtlpbyte; /* total tail loss probe bytes sent */
uint64_t t_sndbytes; /* total bytes sent */
uint64_t t_snd_rxt_bytes; /* total bytes retransmitted */
uint32_t t_dsack_bytes; /* Total number of dsack bytes we have received */
uint32_t t_dsack_tlp_bytes; /* Total number of dsack bytes we have received for TLPs sent */
uint32_t t_dsack_pack; /* Total dsack packets we have recieved */
uint8_t t_tfo_client_cookie_len; /* TCP Fast Open client cookie length */
uint32_t t_end_info_status; /* Status flag of end info */
unsigned int *t_tfo_pending; /* TCP Fast Open server pending counter */
union {
uint8_t client[TCP_FASTOPEN_MAX_COOKIE_LEN];
uint64_t server;
} t_tfo_cookie; /* TCP Fast Open cookie to send */
union {
uint8_t t_end_info_bytes[TCP_END_BYTE_INFO];
uint64_t t_end_info;
};
#ifdef TCPPCAP
There are times when it would be really nice to have a record of the last few packets and/or state transitions from each TCP socket. That would help with narrowing down certain problems we see in the field that are hard to reproduce without understanding the history of how we got into a certain state. This change provides just that. It saves copies of the last N packets in a list in the tcpcb. When the tcpcb is destroyed, the list is freed. I thought this was likely to be more performance-friendly than saving copies of the tcpcb. Plus, with the packets, you should be able to reverse-engineer what happened to the tcpcb. To enable the feature, you will need to compile a kernel with the TCPPCAP option. Even then, the feature defaults to being deactivated. You can activate it by setting a positive value for the number of captured packets. You can do that on either a global basis or on a per-socket basis (via a setsockopt call). There is no way to get the packets out of the kernel other than using kmem or getting a coredump. I thought that would help some of the legal/privacy concerns regarding such a feature. However, it should be possible to add a future effort to export them in PCAP format. I tested this at low scale, and found that there were no mbuf leaks and the peak mbuf usage appeared to be unchanged with and without the feature. The main performance concern I can envision is the number of mbufs that would be used on systems with a large number of sockets. If you save five packets per direction per socket and have 3,000 sockets, that will consume at least 30,000 mbufs just to keep these packets. I tried to reduce the concerns associated with this by limiting the number of clusters (not mbufs) that could be used for this feature. Again, in my testing, that appears to work correctly. Differential Revision: D3100 Submitted by: Jonathan Looney <jlooney at juniper dot net> Reviewed by: gnn, hiren
2015-10-14 00:35:37 +00:00
struct mbufq t_inpkts; /* List of saved input packets. */
struct mbufq t_outpkts; /* List of saved output packets. */
#endif
};
#endif /* _KERNEL || _WANT_TCPCB */
#ifdef _KERNEL
struct tcptemp {
u_char tt_ipgen[40]; /* the size must be of max ip header, now IPv6 */
struct tcphdr tt_t;
1994-05-24 10:09:53 +00:00
};
/* Enable TCP/UDP tunneling port */
#define TCP_TUNNELING_PORT_MIN 0
#define TCP_TUNNELING_PORT_MAX 65535
#define TCP_TUNNELING_PORT_DEFAULT 0
/* Enable TCP/UDP tunneling port */
#define TCP_TUNNELING_OVERHEAD_MIN sizeof(struct udphdr)
#define TCP_TUNNELING_OVERHEAD_MAX 1024
#define TCP_TUNNELING_OVERHEAD_DEFAULT TCP_TUNNELING_OVERHEAD_MIN
/* Minimum map entries limit value, if set */
#define TCP_MIN_MAP_ENTRIES_LIMIT 128
/*
* TODO: We yet need to brave plowing in
* to tcp_input() and the pru_usrreq() block.
* Right now these go to the old standards which
* are somewhat ok, but in the long term may
* need to be changed. If we do tackle tcp_input()
* then we need to get rid of the tcp_do_segment()
* function below.
*/
/* Flags for tcp functions */
#define TCP_FUNC_BEING_REMOVED 0x01 /* Can no longer be referenced */
/*
* If defining the optional tcp_timers, in the
* tfb_tcp_timer_stop call you must use the
* callout_async_drain() function with the
* tcp_timer_discard callback. You should check
* the return of callout_async_drain() and if 0
* increment tt_draincnt. Since the timer sub-system
* does not know your callbacks you must provide a
* stop_all function that loops through and calls
* tcp_timer_stop() with each of your defined timers.
* Adding a tfb_tcp_handoff_ok function allows the socket
* option to change stacks to query you even if the
* connection is in a later stage. You return 0 to
* say you can take over and run your stack, you return
* non-zero (an error number) to say no you can't.
* If the function is undefined you can only change
* in the early states (before connect or listen).
* tfb_tcp_fb_fini is changed to add a flag to tell
* the old stack if the tcb is being destroyed or
* not. A one in the flag means the TCB is being
* destroyed, a zero indicates its transitioning to
* another stack (via socket option).
*/
struct tcp_function_block {
char tfb_tcp_block_name[TCP_FUNCTION_NAME_LEN_MAX];
int (*tfb_tcp_output)(struct tcpcb *);
int (*tfb_tcp_output_wtime)(struct tcpcb *, const struct timeval *);
void (*tfb_tcp_do_segment)(struct mbuf *, struct tcphdr *,
struct socket *, struct tcpcb *,
int, int, uint8_t);
int (*tfb_do_queued_segments)(struct socket *, struct tcpcb *, int);
int (*tfb_do_segment_nounlock)(struct mbuf *, struct tcphdr *,
struct socket *, struct tcpcb *,
int, int, uint8_t,
int, struct timeval *);
void (*tfb_tcp_hpts_do_segment)(struct mbuf *, struct tcphdr *,
struct socket *, struct tcpcb *,
int, int, uint8_t,
int, struct timeval *);
int (*tfb_tcp_ctloutput)(struct socket *so, struct sockopt *sopt,
struct inpcb *inp, struct tcpcb *tp);
/* Optional memory allocation/free routine */
int (*tfb_tcp_fb_init)(struct tcpcb *);
void (*tfb_tcp_fb_fini)(struct tcpcb *, int);
/* Optional timers, must define all if you define one */
int (*tfb_tcp_timer_stop_all)(struct tcpcb *);
void (*tfb_tcp_timer_activate)(struct tcpcb *,
uint32_t, u_int);
int (*tfb_tcp_timer_active)(struct tcpcb *, uint32_t);
void (*tfb_tcp_timer_stop)(struct tcpcb *, uint32_t);
void (*tfb_tcp_rexmit_tmr)(struct tcpcb *);
int (*tfb_tcp_handoff_ok)(struct tcpcb *);
void (*tfb_tcp_mtu_chg)(struct tcpcb *);
int (*tfb_pru_options)(struct tcpcb *, int);
void (*tfb_hwtls_change)(struct tcpcb *, int);
volatile uint32_t tfb_refcnt;
uint32_t tfb_flags;
Add the "TCP Blackbox Recorder" which we discussed at the developer summits at BSDCan and BSDCam in 2017. The TCP Blackbox Recorder allows you to capture events on a TCP connection in a ring buffer. It stores metadata with the event. It optionally stores the TCP header associated with an event (if the event is associated with a packet) and also optionally stores information on the sockets. It supports setting a log ID on a TCP connection and using this to correlate multiple connections that share a common log ID. You can log connections in different modes. If you are doing a coordinated test with a particular connection, you may tell the system to put it in mode 4 (continuous dump). Or, if you just want to monitor for errors, you can put it in mode 1 (ring buffer) and dump all the ring buffers associated with the connection ID when we receive an error signal for that connection ID. You can set a default mode that will be applied to a particular ratio of incoming connections. You can also manually set a mode using a socket option. This commit includes only basic probes. rrs@ has added quite an abundance of probes in his TCP development work. He plans to commit those soon. There are user-space programs which we plan to commit as ports. These read the data from the log device and output pcapng files, and then let you analyze the data (and metadata) in the pcapng files. Reviewed by: gnn (previous version) Obtained from: Netflix, Inc. Relnotes: yes Differential Revision: https://reviews.freebsd.org/D11085
2018-03-22 09:40:08 +00:00
uint8_t tfb_id;
};
struct tcp_function {
TAILQ_ENTRY(tcp_function) tf_next;
char tf_name[TCP_FUNCTION_NAME_LEN_MAX];
struct tcp_function_block *tf_fb;
};
TAILQ_HEAD(tcp_funchead, tcp_function);
#endif /* _KERNEL */
/*
* Flags and utility macros for the t_flags field.
*/
#define TF_ACKNOW 0x00000001 /* ack peer immediately */
#define TF_DELACK 0x00000002 /* ack, but try to delay it */
#define TF_NODELAY 0x00000004 /* don't delay packets to coalesce */
#define TF_NOOPT 0x00000008 /* don't use tcp options */
#define TF_SENTFIN 0x00000010 /* have sent FIN */
#define TF_REQ_SCALE 0x00000020 /* have/will request window scaling */
#define TF_RCVD_SCALE 0x00000040 /* other side has requested scaling */
#define TF_REQ_TSTMP 0x00000080 /* have/will request timestamps */
#define TF_RCVD_TSTMP 0x00000100 /* a timestamp was received in SYN */
#define TF_SACK_PERMIT 0x00000200 /* other side said I could SACK */
#define TF_NEEDSYN 0x00000400 /* send SYN (implicit state) */
#define TF_NEEDFIN 0x00000800 /* send FIN (implicit state) */
#define TF_NOPUSH 0x00001000 /* don't push */
#define TF_PREVVALID 0x00002000 /* saved values for bad rxmit valid */
#define TF_WAKESOR 0x00004000 /* wake up receive socket */
#define TF_GPUTINPROG 0x00008000 /* Goodput measurement in progress */
#define TF_MORETOCOME 0x00010000 /* More data to be appended to sock */
#define TF_LQ_OVERFLOW 0x00020000 /* listen queue overflow */
#define TF_LASTIDLE 0x00040000 /* connection was previously idle */
#define TF_RXWIN0SENT 0x00080000 /* sent a receiver win 0 in response */
#define TF_FASTRECOVERY 0x00100000 /* in NewReno Fast Recovery */
#define TF_WASFRECOVERY 0x00200000 /* was in NewReno Fast Recovery */
#define TF_SIGNATURE 0x00400000 /* require MD5 digests (RFC2385) */
#define TF_FORCEDATA 0x00800000 /* force out a byte */
#define TF_TSO 0x01000000 /* TSO enabled on this connection */
#define TF_TOE 0x02000000 /* this connection is offloaded */
#define TF_UNUSED0 0x04000000 /* unused */
#define TF_UNUSED1 0x08000000 /* unused */
#define TF_LRD 0x10000000 /* Lost Retransmission Detection */
This commit marks the first formal contribution of the "Five New TCP Congestion Control Algorithms for FreeBSD" FreeBSD Foundation funded project. More details about the project are available at: http://caia.swin.edu.au/freebsd/5cc/ - Add a KPI and supporting infrastructure to allow modular congestion control algorithms to be used in the net stack. Algorithms can maintain per-connection state if required, and connections maintain their own algorithm pointer, which allows different connections to concurrently use different algorithms. The TCP_CONGESTION socket option can be used with getsockopt()/setsockopt() to programmatically query or change the congestion control algorithm respectively from within an application at runtime. - Integrate the framework with the TCP stack in as least intrusive a manner as possible. Care was also taken to develop the framework in a way that should allow integration with other congestion aware transport protocols (e.g. SCTP) in the future. The hope is that we will one day be able to share a single set of congestion control algorithm modules between all congestion aware transport protocols. - Introduce a new congestion recovery (TF_CONGRECOVERY) state into the TCP stack and use it to decouple the meaning of recovery from a congestion event and recovery from packet loss (TF_FASTRECOVERY) a la RFC2581. ECN and delay based congestion control protocols don't generally need to recover from packet loss and need a different way to note a congestion recovery episode within the stack. - Remove the net.inet.tcp.newreno sysctl, which simplifies some portions of code and ensures the stack always uses the appropriate mechanisms for recovering from packet loss during a congestion recovery episode. - Extract the NewReno congestion control algorithm from the TCP stack and massage it into module form. NewReno is always built into the kernel and will remain the default algorithm for the forseeable future. Implementations of additional different algorithms will become available in the near future. - Bump __FreeBSD_version to 900025 and note in UPDATING that rebuilding code that relies on the size of "struct tcpcb" is required. Many thanks go to the Cisco University Research Program Fund at Community Foundation Silicon Valley and the FreeBSD Foundation. Their support of our work at the Centre for Advanced Internet Architectures, Swinburne University of Technology is greatly appreciated. In collaboration with: David Hayes <dahayes at swin edu au> and Grenville Armitage <garmitage at swin edu au> Sponsored by: Cisco URP, FreeBSD Foundation Reviewed by: rpaulo Tested by: David Hayes (and many others over the years) MFC after: 3 months
2010-11-12 06:41:55 +00:00
#define TF_CONGRECOVERY 0x20000000 /* congestion recovery mode */
#define TF_WASCRECOVERY 0x40000000 /* was in congestion recovery */
#define TF_FASTOPEN 0x80000000 /* TCP Fast Open indication */
This commit marks the first formal contribution of the "Five New TCP Congestion Control Algorithms for FreeBSD" FreeBSD Foundation funded project. More details about the project are available at: http://caia.swin.edu.au/freebsd/5cc/ - Add a KPI and supporting infrastructure to allow modular congestion control algorithms to be used in the net stack. Algorithms can maintain per-connection state if required, and connections maintain their own algorithm pointer, which allows different connections to concurrently use different algorithms. The TCP_CONGESTION socket option can be used with getsockopt()/setsockopt() to programmatically query or change the congestion control algorithm respectively from within an application at runtime. - Integrate the framework with the TCP stack in as least intrusive a manner as possible. Care was also taken to develop the framework in a way that should allow integration with other congestion aware transport protocols (e.g. SCTP) in the future. The hope is that we will one day be able to share a single set of congestion control algorithm modules between all congestion aware transport protocols. - Introduce a new congestion recovery (TF_CONGRECOVERY) state into the TCP stack and use it to decouple the meaning of recovery from a congestion event and recovery from packet loss (TF_FASTRECOVERY) a la RFC2581. ECN and delay based congestion control protocols don't generally need to recover from packet loss and need a different way to note a congestion recovery episode within the stack. - Remove the net.inet.tcp.newreno sysctl, which simplifies some portions of code and ensures the stack always uses the appropriate mechanisms for recovering from packet loss during a congestion recovery episode. - Extract the NewReno congestion control algorithm from the TCP stack and massage it into module form. NewReno is always built into the kernel and will remain the default algorithm for the forseeable future. Implementations of additional different algorithms will become available in the near future. - Bump __FreeBSD_version to 900025 and note in UPDATING that rebuilding code that relies on the size of "struct tcpcb" is required. Many thanks go to the Cisco University Research Program Fund at Community Foundation Silicon Valley and the FreeBSD Foundation. Their support of our work at the Centre for Advanced Internet Architectures, Swinburne University of Technology is greatly appreciated. In collaboration with: David Hayes <dahayes at swin edu au> and Grenville Armitage <garmitage at swin edu au> Sponsored by: Cisco URP, FreeBSD Foundation Reviewed by: rpaulo Tested by: David Hayes (and many others over the years) MFC after: 3 months
2010-11-12 06:41:55 +00:00
#define IN_FASTRECOVERY(t_flags) (t_flags & TF_FASTRECOVERY)
#define ENTER_FASTRECOVERY(t_flags) t_flags |= TF_FASTRECOVERY
#define EXIT_FASTRECOVERY(t_flags) t_flags &= ~TF_FASTRECOVERY
#define IN_CONGRECOVERY(t_flags) (t_flags & TF_CONGRECOVERY)
#define ENTER_CONGRECOVERY(t_flags) t_flags |= TF_CONGRECOVERY
#define EXIT_CONGRECOVERY(t_flags) t_flags &= ~TF_CONGRECOVERY
#define IN_RECOVERY(t_flags) (t_flags & (TF_CONGRECOVERY | TF_FASTRECOVERY))
#define ENTER_RECOVERY(t_flags) t_flags |= (TF_CONGRECOVERY | TF_FASTRECOVERY)
#define EXIT_RECOVERY(t_flags) t_flags &= ~(TF_CONGRECOVERY | TF_FASTRECOVERY)
#if defined(_KERNEL) && !defined(TCP_RFC7413)
#define IS_FASTOPEN(t_flags) (false)
#else
#define IS_FASTOPEN(t_flags) (t_flags & TF_FASTOPEN)
#endif
This commit marks the first formal contribution of the "Five New TCP Congestion Control Algorithms for FreeBSD" FreeBSD Foundation funded project. More details about the project are available at: http://caia.swin.edu.au/freebsd/5cc/ - Add a KPI and supporting infrastructure to allow modular congestion control algorithms to be used in the net stack. Algorithms can maintain per-connection state if required, and connections maintain their own algorithm pointer, which allows different connections to concurrently use different algorithms. The TCP_CONGESTION socket option can be used with getsockopt()/setsockopt() to programmatically query or change the congestion control algorithm respectively from within an application at runtime. - Integrate the framework with the TCP stack in as least intrusive a manner as possible. Care was also taken to develop the framework in a way that should allow integration with other congestion aware transport protocols (e.g. SCTP) in the future. The hope is that we will one day be able to share a single set of congestion control algorithm modules between all congestion aware transport protocols. - Introduce a new congestion recovery (TF_CONGRECOVERY) state into the TCP stack and use it to decouple the meaning of recovery from a congestion event and recovery from packet loss (TF_FASTRECOVERY) a la RFC2581. ECN and delay based congestion control protocols don't generally need to recover from packet loss and need a different way to note a congestion recovery episode within the stack. - Remove the net.inet.tcp.newreno sysctl, which simplifies some portions of code and ensures the stack always uses the appropriate mechanisms for recovering from packet loss during a congestion recovery episode. - Extract the NewReno congestion control algorithm from the TCP stack and massage it into module form. NewReno is always built into the kernel and will remain the default algorithm for the forseeable future. Implementations of additional different algorithms will become available in the near future. - Bump __FreeBSD_version to 900025 and note in UPDATING that rebuilding code that relies on the size of "struct tcpcb" is required. Many thanks go to the Cisco University Research Program Fund at Community Foundation Silicon Valley and the FreeBSD Foundation. Their support of our work at the Centre for Advanced Internet Architectures, Swinburne University of Technology is greatly appreciated. In collaboration with: David Hayes <dahayes at swin edu au> and Grenville Armitage <garmitage at swin edu au> Sponsored by: Cisco URP, FreeBSD Foundation Reviewed by: rpaulo Tested by: David Hayes (and many others over the years) MFC after: 3 months
2010-11-12 06:41:55 +00:00
#define BYTES_THIS_ACK(tp, th) (th->th_ack - tp->snd_una)
/*
* Flags for the t_oobflags field.
*/
#define TCPOOB_HAVEDATA 0x01
#define TCPOOB_HADDATA 0x02
/*
Add the "TCP Blackbox Recorder" which we discussed at the developer summits at BSDCan and BSDCam in 2017. The TCP Blackbox Recorder allows you to capture events on a TCP connection in a ring buffer. It stores metadata with the event. It optionally stores the TCP header associated with an event (if the event is associated with a packet) and also optionally stores information on the sockets. It supports setting a log ID on a TCP connection and using this to correlate multiple connections that share a common log ID. You can log connections in different modes. If you are doing a coordinated test with a particular connection, you may tell the system to put it in mode 4 (continuous dump). Or, if you just want to monitor for errors, you can put it in mode 1 (ring buffer) and dump all the ring buffers associated with the connection ID when we receive an error signal for that connection ID. You can set a default mode that will be applied to a particular ratio of incoming connections. You can also manually set a mode using a socket option. This commit includes only basic probes. rrs@ has added quite an abundance of probes in his TCP development work. He plans to commit those soon. There are user-space programs which we plan to commit as ports. These read the data from the log device and output pcapng files, and then let you analyze the data (and metadata) in the pcapng files. Reviewed by: gnn (previous version) Obtained from: Netflix, Inc. Relnotes: yes Differential Revision: https://reviews.freebsd.org/D11085
2018-03-22 09:40:08 +00:00
* Flags for the extended TCP flags field, t_flags2
*/
#define TF2_PLPMTU_BLACKHOLE 0x00000001 /* Possible PLPMTUD Black Hole. */
#define TF2_PLPMTU_PMTUD 0x00000002 /* Allowed to attempt PLPMTUD. */
#define TF2_PLPMTU_MAXSEGSNT 0x00000004 /* Last seg sent was full seg. */
Add the "TCP Blackbox Recorder" which we discussed at the developer summits at BSDCan and BSDCam in 2017. The TCP Blackbox Recorder allows you to capture events on a TCP connection in a ring buffer. It stores metadata with the event. It optionally stores the TCP header associated with an event (if the event is associated with a packet) and also optionally stores information on the sockets. It supports setting a log ID on a TCP connection and using this to correlate multiple connections that share a common log ID. You can log connections in different modes. If you are doing a coordinated test with a particular connection, you may tell the system to put it in mode 4 (continuous dump). Or, if you just want to monitor for errors, you can put it in mode 1 (ring buffer) and dump all the ring buffers associated with the connection ID when we receive an error signal for that connection ID. You can set a default mode that will be applied to a particular ratio of incoming connections. You can also manually set a mode using a socket option. This commit includes only basic probes. rrs@ has added quite an abundance of probes in his TCP development work. He plans to commit those soon. There are user-space programs which we plan to commit as ports. These read the data from the log device and output pcapng files, and then let you analyze the data (and metadata) in the pcapng files. Reviewed by: gnn (previous version) Obtained from: Netflix, Inc. Relnotes: yes Differential Revision: https://reviews.freebsd.org/D11085
2018-03-22 09:40:08 +00:00
#define TF2_LOG_AUTO 0x00000008 /* Session is auto-logging. */
#define TF2_DROP_AF_DATA 0x00000010 /* Drop after all data ack'd */
#define TF2_ECN_PERMIT 0x00000020 /* connection ECN-ready */
#define TF2_ECN_SND_CWR 0x00000040 /* ECN CWR in queue */
#define TF2_ECN_SND_ECE 0x00000080 /* ECN ECE in queue */
#define TF2_ACE_PERMIT 0x00000100 /* Accurate ECN mode */
#define TF2_FBYTES_COMPLETE 0x00000400 /* We have first bytes in and out */
/*
* Structure to hold TCP options that are only used during segment
* processing (in tcp_input), but not held in the tcpcb.
* It's basically used to reduce the number of parameters
* to tcp_dooptions and tcp_addoptions.
* The binary order of the to_flags is relevant for packing of the
* options in tcp_addoptions.
*/
struct tcpopt {
u_int32_t to_flags; /* which options are present */
#define TOF_MSS 0x0001 /* maximum segment size */
#define TOF_SCALE 0x0002 /* window scaling */
#define TOF_SACKPERM 0x0004 /* SACK permitted */
#define TOF_TS 0x0010 /* timestamp */
#define TOF_SIGNATURE 0x0040 /* TCP-MD5 signature option (RFC2385) */
#define TOF_SACK 0x0080 /* Peer sent SACK option */
#define TOF_FASTOPEN 0x0100 /* TCP Fast Open (TFO) cookie */
#define TOF_MAXOPT 0x0200
u_int32_t to_tsval; /* new timestamp */
u_int32_t to_tsecr; /* reflected timestamp */
u_char *to_sacks; /* pointer to the first SACK blocks */
u_char *to_signature; /* pointer to the TCP-MD5 signature */
u_int8_t *to_tfo_cookie; /* pointer to the TFO cookie */
u_int16_t to_mss; /* maximum segment size */
u_int8_t to_wscale; /* window scaling */
u_int8_t to_nsacks; /* number of SACK blocks */
u_int8_t to_tfo_len; /* TFO cookie length */
u_int32_t to_spare; /* UTO */
};
/*
* Flags for tcp_dooptions.
*/
#define TO_SYN 0x01 /* parse SYN-only options */
struct hc_metrics_lite { /* must stay in sync with hc_metrics */
uint32_t rmx_mtu; /* MTU for this path */
uint32_t rmx_ssthresh; /* outbound gateway buffer limit */
uint32_t rmx_rtt; /* estimated round trip time */
uint32_t rmx_rttvar; /* estimated rtt variance */
uint32_t rmx_cwnd; /* congestion window */
uint32_t rmx_sendpipe; /* outbound delay-bandwidth product */
uint32_t rmx_recvpipe; /* inbound delay-bandwidth product */
};
/*
* Used by tcp_maxmtu() to communicate interface specific features
* and limits at the time of connection setup.
*/
struct tcp_ifcap {
int ifcap;
u_int tsomax;
u_int tsomaxsegcount;
u_int tsomaxsegsize;
};
#ifndef _NETINET_IN_PCB_H_
struct in_conninfo;
#endif /* _NETINET_IN_PCB_H_ */
struct tcptw {
struct inpcb *tw_inpcb; /* XXX back pointer to internet pcb */
uint32_t t_port:16, /* UDP port number if TCPoUDP */
t_unused:16;
tcp_seq snd_nxt;
tcp_seq rcv_nxt;
tcp_seq iss;
tcp_seq irs;
u_short last_win; /* cached window value */
short tw_so_options; /* copy of so_options */
struct ucred *tw_cred; /* user credentials */
u_int32_t t_recent;
u_int32_t ts_offset; /* our timestamp offset */
u_int t_starttime;
int tw_time;
TAILQ_ENTRY(tcptw) tw_2msl;
void *tw_pspare; /* TCP_SIGNATURE */
u_int *tw_spare; /* TCP_SIGNATURE */
};
1994-05-24 10:09:53 +00:00
#define intotcpcb(ip) ((struct tcpcb *)(ip)->inp_ppcb)
#define intotw(ip) ((struct tcptw *)(ip)->inp_ppcb)
1994-05-24 10:09:53 +00:00
#define sototcpcb(so) (intotcpcb(sotoinpcb(so)))
/*
* The smoothed round-trip time and estimated variance
* are stored as fixed point numbers scaled by the values below.
* For convenience, these scales are also used in smoothing the average
* (smoothed = (1/scale)sample + ((scale-1)/scale)smoothed).
* With these scales, srtt has 3 bits to the right of the binary point,
* and thus an "ALPHA" of 0.875. rttvar has 2 bits to the right of the
* binary point, and is smoothed with an ALPHA of 0.75.
*/
#define TCP_RTT_SCALE 32 /* multiplier for srtt; 3 bits frac. */
#define TCP_RTT_SHIFT 5 /* shift for srtt; 3 bits frac. */
#define TCP_RTTVAR_SCALE 16 /* multiplier for rttvar; 2 bits */
#define TCP_RTTVAR_SHIFT 4 /* shift for rttvar; 2 bits */
#define TCP_DELTA_SHIFT 2 /* see tcp_input.c */
1994-05-24 10:09:53 +00:00
/*
* The initial retransmission should happen at rtt + 4 * rttvar.
* Because of the way we do the smoothing, srtt and rttvar
* will each average +1/2 tick of bias. When we compute
* the retransmit timer, we want 1/2 tick of rounding and
* 1 extra tick because of +-1/2 tick uncertainty in the
* firing of the timer. The bias will give us exactly the
* 1.5 tick we need. But, because the bias is
* statistical, we have to test that we don't drop below
* the minimum feasible timer (which is 2 ticks).
* This version of the macro adapted from a paper by Lawrence
* Brakmo and Larry Peterson which outlines a problem caused
* by insufficient precision in the original implementation,
* which results in inappropriately large RTO values for very
* fast networks.
1994-05-24 10:09:53 +00:00
*/
#define TCP_REXMTVAL(tp) \
max((tp)->t_rttmin, (((tp)->t_srtt >> (TCP_RTT_SHIFT - TCP_DELTA_SHIFT)) \
+ (tp)->t_rttvar) >> TCP_DELTA_SHIFT)
1994-05-24 10:09:53 +00:00
/*
* TCP statistics.
* Many of these should be kept per connection,
* but that's inconvenient at the moment.
*/
struct tcpstat {
uint64_t tcps_connattempt; /* connections initiated */
uint64_t tcps_accepts; /* connections accepted */
uint64_t tcps_connects; /* connections established */
uint64_t tcps_drops; /* connections dropped */
uint64_t tcps_conndrops; /* embryonic connections dropped */
uint64_t tcps_minmssdrops; /* average minmss too low drops */
uint64_t tcps_closed; /* conn. closed (includes drops) */
uint64_t tcps_segstimed; /* segs where we tried to get rtt */
uint64_t tcps_rttupdated; /* times we succeeded */
uint64_t tcps_delack; /* delayed acks sent */
uint64_t tcps_timeoutdrop; /* conn. dropped in rxmt timeout */
uint64_t tcps_rexmttimeo; /* retransmit timeouts */
uint64_t tcps_persisttimeo; /* persist timeouts */
uint64_t tcps_keeptimeo; /* keepalive timeouts */
uint64_t tcps_keepprobe; /* keepalive probes sent */
uint64_t tcps_keepdrops; /* connections dropped in keepalive */
uint64_t tcps_sndtotal; /* total packets sent */
uint64_t tcps_sndpack; /* data packets sent */
uint64_t tcps_sndbyte; /* data bytes sent */
uint64_t tcps_sndrexmitpack; /* data packets retransmitted */
uint64_t tcps_sndrexmitbyte; /* data bytes retransmitted */
uint64_t tcps_sndrexmitbad; /* unnecessary packet retransmissions */
uint64_t tcps_sndacks; /* ack-only packets sent */
uint64_t tcps_sndprobe; /* window probes sent */
uint64_t tcps_sndurg; /* packets sent with URG only */
uint64_t tcps_sndwinup; /* window update-only packets sent */
uint64_t tcps_sndctrl; /* control (SYN|FIN|RST) packets sent */
uint64_t tcps_rcvtotal; /* total packets received */
uint64_t tcps_rcvpack; /* packets received in sequence */
uint64_t tcps_rcvbyte; /* bytes received in sequence */
uint64_t tcps_rcvbadsum; /* packets received with ccksum errs */
uint64_t tcps_rcvbadoff; /* packets received with bad offset */
uint64_t tcps_rcvreassfull; /* packets dropped for no reass space */
uint64_t tcps_rcvshort; /* packets received too short */
uint64_t tcps_rcvduppack; /* duplicate-only packets received */
uint64_t tcps_rcvdupbyte; /* duplicate-only bytes received */
uint64_t tcps_rcvpartduppack; /* packets with some duplicate data */
uint64_t tcps_rcvpartdupbyte; /* dup. bytes in part-dup. packets */
uint64_t tcps_rcvoopack; /* out-of-order packets received */
uint64_t tcps_rcvoobyte; /* out-of-order bytes received */
uint64_t tcps_rcvpackafterwin; /* packets with data after window */
uint64_t tcps_rcvbyteafterwin; /* bytes rcvd after window */
uint64_t tcps_rcvafterclose; /* packets rcvd after "close" */
uint64_t tcps_rcvwinprobe; /* rcvd window probe packets */
uint64_t tcps_rcvdupack; /* rcvd duplicate acks */
uint64_t tcps_rcvacktoomuch; /* rcvd acks for unsent data */
uint64_t tcps_rcvackpack; /* rcvd ack packets */
uint64_t tcps_rcvackbyte; /* bytes acked by rcvd acks */
uint64_t tcps_rcvwinupd; /* rcvd window update packets */
uint64_t tcps_pawsdrop; /* segments dropped due to PAWS */
uint64_t tcps_predack; /* times hdr predict ok for acks */
uint64_t tcps_preddat; /* times hdr predict ok for data pkts */
uint64_t tcps_pcbcachemiss;
uint64_t tcps_cachedrtt; /* times cached RTT in route updated */
uint64_t tcps_cachedrttvar; /* times cached rttvar updated */
uint64_t tcps_cachedssthresh; /* times cached ssthresh updated */
uint64_t tcps_usedrtt; /* times RTT initialized from route */
uint64_t tcps_usedrttvar; /* times RTTVAR initialized from rt */
uint64_t tcps_usedssthresh; /* times ssthresh initialized from rt*/
uint64_t tcps_persistdrop; /* timeout in persist state */
uint64_t tcps_badsyn; /* bogus SYN, e.g. premature ACK */
uint64_t tcps_mturesent; /* resends due to MTU discovery */
uint64_t tcps_listendrop; /* listen queue overflows */
uint64_t tcps_badrst; /* ignored RSTs in the window */
uint64_t tcps_sc_added; /* entry added to syncache */
uint64_t tcps_sc_retransmitted; /* syncache entry was retransmitted */
uint64_t tcps_sc_dupsyn; /* duplicate SYN packet */
uint64_t tcps_sc_dropped; /* could not reply to packet */
uint64_t tcps_sc_completed; /* successful extraction of entry */
uint64_t tcps_sc_bucketoverflow;/* syncache per-bucket limit hit */
uint64_t tcps_sc_cacheoverflow; /* syncache cache limit hit */
uint64_t tcps_sc_reset; /* RST removed entry from syncache */
uint64_t tcps_sc_stale; /* timed out or listen socket gone */
uint64_t tcps_sc_aborted; /* syncache entry aborted */
uint64_t tcps_sc_badack; /* removed due to bad ACK */
uint64_t tcps_sc_unreach; /* ICMP unreachable received */
uint64_t tcps_sc_zonefail; /* zalloc() failed */
uint64_t tcps_sc_sendcookie; /* SYN cookie sent */
uint64_t tcps_sc_recvcookie; /* SYN cookie received */
uint64_t tcps_hc_added; /* entry added to hostcache */
uint64_t tcps_hc_bucketoverflow;/* hostcache per bucket limit hit */
uint64_t tcps_finwait2_drops; /* Drop FIN_WAIT_2 connection after time limit */
/* SACK related stats */
uint64_t tcps_sack_recovery_episode; /* SACK recovery episodes */
uint64_t tcps_sack_rexmits; /* SACK rexmit segments */
uint64_t tcps_sack_rexmit_bytes; /* SACK rexmit bytes */
uint64_t tcps_sack_rcv_blocks; /* SACK blocks (options) received */
uint64_t tcps_sack_send_blocks; /* SACK blocks (options) sent */
uint64_t tcps_sack_lostrexmt; /* SACK lost retransmission recovered */
uint64_t tcps_sack_sboverflow; /* times scoreboard overflowed */
/* ECN related stats */
uint64_t tcps_ecn_ce; /* ECN Congestion Experienced */
uint64_t tcps_ecn_ect0; /* ECN Capable Transport */
uint64_t tcps_ecn_ect1; /* ECN Capable Transport */
uint64_t tcps_ecn_shs; /* ECN successful handshakes */
uint64_t tcps_ecn_rcwnd; /* # times ECN reduced the cwnd */
/* TCP_SIGNATURE related stats */
uint64_t tcps_sig_rcvgoodsig; /* Total matching signature received */
uint64_t tcps_sig_rcvbadsig; /* Total bad signature received */
Merge projects/ipsec into head/. Small summary ------------- o Almost all IPsec releated code was moved into sys/netipsec. o New kernel modules added: ipsec.ko and tcpmd5.ko. New kernel option IPSEC_SUPPORT added. It enables support for loading and unloading of ipsec.ko and tcpmd5.ko kernel modules. o IPSEC_NAT_T option was removed. Now NAT-T support is enabled by default. The UDP_ENCAP_ESPINUDP_NON_IKE encapsulation type support was removed. Added TCP/UDP checksum handling for inbound packets that were decapsulated by transport mode SAs. setkey(8) modified to show run-time NAT-T configuration of SA. o New network pseudo interface if_ipsec(4) added. For now it is build as part of ipsec.ko module (or with IPSEC kernel). It implements IPsec virtual tunnels to create route-based VPNs. o The network stack now invokes IPsec functions using special methods. The only one header file <netipsec/ipsec_support.h> should be included to declare all the needed things to work with IPsec. o All IPsec protocols handlers (ESP/AH/IPCOMP protosw) were removed. Now these protocols are handled directly via IPsec methods. o TCP_SIGNATURE support was reworked to be more close to RFC. o PF_KEY SADB was reworked: - now all security associations stored in the single SPI namespace, and all SAs MUST have unique SPI. - several hash tables added to speed up lookups in SADB. - SADB now uses rmlock to protect access, and concurrent threads can do SA lookups in the same time. - many PF_KEY message handlers were reworked to reflect changes in SADB. - SADB_UPDATE message was extended to support new PF_KEY headers: SADB_X_EXT_NEW_ADDRESS_SRC and SADB_X_EXT_NEW_ADDRESS_DST. They can be used by IKE daemon to change SA addresses. o ipsecrequest and secpolicy structures were cardinally changed to avoid locking protection for ipsecrequest. Now we support only limited number (4) of bundled SAs, but they are supported for both INET and INET6. o INPCB security policy cache was introduced. Each PCB now caches used security policies to avoid SP lookup for each packet. o For inbound security policies added the mode, when the kernel does check for full history of applied IPsec transforms. o References counting rules for security policies and security associations were changed. The proper SA locking added into xform code. o xform code was also changed. Now it is possible to unregister xforms. tdb_xxx structures were changed and renamed to reflect changes in SADB/SPDB, and changed rules for locking and refcounting. Reviewed by: gnn, wblock Obtained from: Yandex LLC Relnotes: yes Sponsored by: Yandex LLC Differential Revision: https://reviews.freebsd.org/D9352
2017-02-06 08:49:57 +00:00
uint64_t tcps_sig_err_buildsig; /* Failed to make signature */
uint64_t tcps_sig_err_sigopt; /* No signature expected by socket */
uint64_t tcps_sig_err_nosigopt; /* No signature provided by segment */
/* Path MTU Discovery Black Hole Detection related stats */
uint64_t tcps_pmtud_blackhole_activated; /* Black Hole Count */
uint64_t tcps_pmtud_blackhole_activated_min_mss; /* BH at min MSS Count */
uint64_t tcps_pmtud_blackhole_failed; /* Black Hole Failure Count */
uint64_t tcps_tunneled_pkts; /* Packets encap's in UDP received */
uint64_t tcps_tunneled_errs; /* Packets that had errors that were UDP encaped */
/* Dsack related stats */
uint64_t tcps_dsack_count; /* Number of ACKs arriving with DSACKs */
uint64_t tcps_dsack_bytes; /* Number of bytes DSACK'ed no TLP */
uint64_t tcps_dsack_tlp_bytes; /* Number of bytes DSACK'ed due to TLPs */
uint64_t _pad[6]; /* 3 UTO, 3 TBD */
1994-05-24 10:09:53 +00:00
};
#define tcps_rcvmemdrop tcps_rcvreassfull /* compat */
#ifdef _KERNEL
#define TI_UNLOCKED 1
#define TI_RLOCKED 2
#include <sys/counter.h>
VNET_PCPUSTAT_DECLARE(struct tcpstat, tcpstat); /* tcp statistics */
/*
* In-kernel consumers can use these accessor macros directly to update
* stats.
*/
#define TCPSTAT_ADD(name, val) \
VNET_PCPUSTAT_ADD(struct tcpstat, tcpstat, name, (val))
#define TCPSTAT_INC(name) TCPSTAT_ADD(name, 1)
/*
* Kernel module consumers must use this accessor macro.
*/
void kmod_tcpstat_add(int statnum, int val);
#define KMOD_TCPSTAT_ADD(name, val) \
kmod_tcpstat_add(offsetof(struct tcpstat, name) / sizeof(uint64_t), val)
#define KMOD_TCPSTAT_INC(name) KMOD_TCPSTAT_ADD(name, 1)
/*
* Running TCP connection count by state.
*/
VNET_DECLARE(counter_u64_t, tcps_states[TCP_NSTATES]);
#define V_tcps_states VNET(tcps_states)
#define TCPSTATES_INC(state) counter_u64_add(V_tcps_states[state], 1)
#define TCPSTATES_DEC(state) counter_u64_add(V_tcps_states[state], -1)
/*
* TCP specific helper hook point identifiers.
*/
#define HHOOK_TCP_EST_IN 0
#define HHOOK_TCP_EST_OUT 1
#define HHOOK_TCP_LAST HHOOK_TCP_EST_OUT
struct tcp_hhook_data {
struct tcpcb *tp;
struct tcphdr *th;
struct tcpopt *to;
uint32_t len;
int tso;
tcp_seq curack;
};
#ifdef TCP_HHOOK
void hhook_run_tcp_est_out(struct tcpcb *tp,
struct tcphdr *th, struct tcpopt *to,
uint32_t len, int tso);
#endif
#endif
/*
* TCB structure exported to user-land via sysctl(3).
*
* Fields prefixed with "xt_" are unique to the export structure, and fields
* with "t_" or other prefixes match corresponding fields of 'struct tcpcb'.
*
* Legend:
* (s) - used by userland utilities in src
* (p) - used by utilities in ports
* (3) - is known to be used by third party software not in ports
* (n) - no known usage
*
* Evil hack: declare only if in_pcb.h and sys/socketvar.h have been
* included. Not all of our clients do.
*/
#if defined(_NETINET_IN_PCB_H_) && defined(_SYS_SOCKETVAR_H_)
struct xtcpcb {
ksize_t xt_len; /* length of this structure */
struct xinpcb xt_inp;
char xt_stack[TCP_FUNCTION_NAME_LEN_MAX]; /* (s) */
Add the "TCP Blackbox Recorder" which we discussed at the developer summits at BSDCan and BSDCam in 2017. The TCP Blackbox Recorder allows you to capture events on a TCP connection in a ring buffer. It stores metadata with the event. It optionally stores the TCP header associated with an event (if the event is associated with a packet) and also optionally stores information on the sockets. It supports setting a log ID on a TCP connection and using this to correlate multiple connections that share a common log ID. You can log connections in different modes. If you are doing a coordinated test with a particular connection, you may tell the system to put it in mode 4 (continuous dump). Or, if you just want to monitor for errors, you can put it in mode 1 (ring buffer) and dump all the ring buffers associated with the connection ID when we receive an error signal for that connection ID. You can set a default mode that will be applied to a particular ratio of incoming connections. You can also manually set a mode using a socket option. This commit includes only basic probes. rrs@ has added quite an abundance of probes in his TCP development work. He plans to commit those soon. There are user-space programs which we plan to commit as ports. These read the data from the log device and output pcapng files, and then let you analyze the data (and metadata) in the pcapng files. Reviewed by: gnn (previous version) Obtained from: Netflix, Inc. Relnotes: yes Differential Revision: https://reviews.freebsd.org/D11085
2018-03-22 09:40:08 +00:00
char xt_logid[TCP_LOG_ID_LEN]; /* (s) */
char xt_cc[TCP_CA_NAME_MAX]; /* (s) */
int64_t spare64[6];
int32_t t_state; /* (s,p) */
uint32_t t_flags; /* (s,p) */
int32_t t_sndzerowin; /* (s) */
int32_t t_sndrexmitpack; /* (s) */
int32_t t_rcvoopack; /* (s) */
int32_t t_rcvtime; /* (s) */
int32_t tt_rexmt; /* (s) */
int32_t tt_persist; /* (s) */
int32_t tt_keep; /* (s) */
int32_t tt_2msl; /* (s) */
int32_t tt_delack; /* (s) */
Add the "TCP Blackbox Recorder" which we discussed at the developer summits at BSDCan and BSDCam in 2017. The TCP Blackbox Recorder allows you to capture events on a TCP connection in a ring buffer. It stores metadata with the event. It optionally stores the TCP header associated with an event (if the event is associated with a packet) and also optionally stores information on the sockets. It supports setting a log ID on a TCP connection and using this to correlate multiple connections that share a common log ID. You can log connections in different modes. If you are doing a coordinated test with a particular connection, you may tell the system to put it in mode 4 (continuous dump). Or, if you just want to monitor for errors, you can put it in mode 1 (ring buffer) and dump all the ring buffers associated with the connection ID when we receive an error signal for that connection ID. You can set a default mode that will be applied to a particular ratio of incoming connections. You can also manually set a mode using a socket option. This commit includes only basic probes. rrs@ has added quite an abundance of probes in his TCP development work. He plans to commit those soon. There are user-space programs which we plan to commit as ports. These read the data from the log device and output pcapng files, and then let you analyze the data (and metadata) in the pcapng files. Reviewed by: gnn (previous version) Obtained from: Netflix, Inc. Relnotes: yes Differential Revision: https://reviews.freebsd.org/D11085
2018-03-22 09:40:08 +00:00
int32_t t_logstate; /* (3) */
uint32_t t_snd_cwnd; /* (s) */
uint32_t t_snd_ssthresh; /* (s) */
uint32_t t_maxseg; /* (s) */
uint32_t t_rcv_wnd; /* (s) */
uint32_t t_snd_wnd; /* (s) */
uint32_t xt_ecn; /* (s) */
uint32_t t_dsack_bytes; /* (n) */
uint32_t t_dsack_tlp_bytes; /* (n) */
uint32_t t_dsack_pack; /* (n) */
uint16_t xt_encaps_port; /* (s) */
int16_t spare16;
int32_t spare32[22];
} __aligned(8);
Add the "TCP Blackbox Recorder" which we discussed at the developer summits at BSDCan and BSDCam in 2017. The TCP Blackbox Recorder allows you to capture events on a TCP connection in a ring buffer. It stores metadata with the event. It optionally stores the TCP header associated with an event (if the event is associated with a packet) and also optionally stores information on the sockets. It supports setting a log ID on a TCP connection and using this to correlate multiple connections that share a common log ID. You can log connections in different modes. If you are doing a coordinated test with a particular connection, you may tell the system to put it in mode 4 (continuous dump). Or, if you just want to monitor for errors, you can put it in mode 1 (ring buffer) and dump all the ring buffers associated with the connection ID when we receive an error signal for that connection ID. You can set a default mode that will be applied to a particular ratio of incoming connections. You can also manually set a mode using a socket option. This commit includes only basic probes. rrs@ has added quite an abundance of probes in his TCP development work. He plans to commit those soon. There are user-space programs which we plan to commit as ports. These read the data from the log device and output pcapng files, and then let you analyze the data (and metadata) in the pcapng files. Reviewed by: gnn (previous version) Obtained from: Netflix, Inc. Relnotes: yes Differential Revision: https://reviews.freebsd.org/D11085
2018-03-22 09:40:08 +00:00
#ifdef _KERNEL
void tcp_inptoxtp(const struct inpcb *, struct xtcpcb *);
#endif
#endif
Add the "TCP Blackbox Recorder" which we discussed at the developer summits at BSDCan and BSDCam in 2017. The TCP Blackbox Recorder allows you to capture events on a TCP connection in a ring buffer. It stores metadata with the event. It optionally stores the TCP header associated with an event (if the event is associated with a packet) and also optionally stores information on the sockets. It supports setting a log ID on a TCP connection and using this to correlate multiple connections that share a common log ID. You can log connections in different modes. If you are doing a coordinated test with a particular connection, you may tell the system to put it in mode 4 (continuous dump). Or, if you just want to monitor for errors, you can put it in mode 1 (ring buffer) and dump all the ring buffers associated with the connection ID when we receive an error signal for that connection ID. You can set a default mode that will be applied to a particular ratio of incoming connections. You can also manually set a mode using a socket option. This commit includes only basic probes. rrs@ has added quite an abundance of probes in his TCP development work. He plans to commit those soon. There are user-space programs which we plan to commit as ports. These read the data from the log device and output pcapng files, and then let you analyze the data (and metadata) in the pcapng files. Reviewed by: gnn (previous version) Obtained from: Netflix, Inc. Relnotes: yes Differential Revision: https://reviews.freebsd.org/D11085
2018-03-22 09:40:08 +00:00
/*
* TCP function information (name-to-id mapping, aliases, and refcnt)
* exported to user-land via sysctl(3).
Add the "TCP Blackbox Recorder" which we discussed at the developer summits at BSDCan and BSDCam in 2017. The TCP Blackbox Recorder allows you to capture events on a TCP connection in a ring buffer. It stores metadata with the event. It optionally stores the TCP header associated with an event (if the event is associated with a packet) and also optionally stores information on the sockets. It supports setting a log ID on a TCP connection and using this to correlate multiple connections that share a common log ID. You can log connections in different modes. If you are doing a coordinated test with a particular connection, you may tell the system to put it in mode 4 (continuous dump). Or, if you just want to monitor for errors, you can put it in mode 1 (ring buffer) and dump all the ring buffers associated with the connection ID when we receive an error signal for that connection ID. You can set a default mode that will be applied to a particular ratio of incoming connections. You can also manually set a mode using a socket option. This commit includes only basic probes. rrs@ has added quite an abundance of probes in his TCP development work. He plans to commit those soon. There are user-space programs which we plan to commit as ports. These read the data from the log device and output pcapng files, and then let you analyze the data (and metadata) in the pcapng files. Reviewed by: gnn (previous version) Obtained from: Netflix, Inc. Relnotes: yes Differential Revision: https://reviews.freebsd.org/D11085
2018-03-22 09:40:08 +00:00
*/
struct tcp_function_info {
uint32_t tfi_refcnt;
Add the "TCP Blackbox Recorder" which we discussed at the developer summits at BSDCan and BSDCam in 2017. The TCP Blackbox Recorder allows you to capture events on a TCP connection in a ring buffer. It stores metadata with the event. It optionally stores the TCP header associated with an event (if the event is associated with a packet) and also optionally stores information on the sockets. It supports setting a log ID on a TCP connection and using this to correlate multiple connections that share a common log ID. You can log connections in different modes. If you are doing a coordinated test with a particular connection, you may tell the system to put it in mode 4 (continuous dump). Or, if you just want to monitor for errors, you can put it in mode 1 (ring buffer) and dump all the ring buffers associated with the connection ID when we receive an error signal for that connection ID. You can set a default mode that will be applied to a particular ratio of incoming connections. You can also manually set a mode using a socket option. This commit includes only basic probes. rrs@ has added quite an abundance of probes in his TCP development work. He plans to commit those soon. There are user-space programs which we plan to commit as ports. These read the data from the log device and output pcapng files, and then let you analyze the data (and metadata) in the pcapng files. Reviewed by: gnn (previous version) Obtained from: Netflix, Inc. Relnotes: yes Differential Revision: https://reviews.freebsd.org/D11085
2018-03-22 09:40:08 +00:00
uint8_t tfi_id;
char tfi_name[TCP_FUNCTION_NAME_LEN_MAX];
char tfi_alias[TCP_FUNCTION_NAME_LEN_MAX];
Add the "TCP Blackbox Recorder" which we discussed at the developer summits at BSDCan and BSDCam in 2017. The TCP Blackbox Recorder allows you to capture events on a TCP connection in a ring buffer. It stores metadata with the event. It optionally stores the TCP header associated with an event (if the event is associated with a packet) and also optionally stores information on the sockets. It supports setting a log ID on a TCP connection and using this to correlate multiple connections that share a common log ID. You can log connections in different modes. If you are doing a coordinated test with a particular connection, you may tell the system to put it in mode 4 (continuous dump). Or, if you just want to monitor for errors, you can put it in mode 1 (ring buffer) and dump all the ring buffers associated with the connection ID when we receive an error signal for that connection ID. You can set a default mode that will be applied to a particular ratio of incoming connections. You can also manually set a mode using a socket option. This commit includes only basic probes. rrs@ has added quite an abundance of probes in his TCP development work. He plans to commit those soon. There are user-space programs which we plan to commit as ports. These read the data from the log device and output pcapng files, and then let you analyze the data (and metadata) in the pcapng files. Reviewed by: gnn (previous version) Obtained from: Netflix, Inc. Relnotes: yes Differential Revision: https://reviews.freebsd.org/D11085
2018-03-22 09:40:08 +00:00
};
/*
* Identifiers for TCP sysctl nodes
*/
#define TCPCTL_DO_RFC1323 1 /* use RFC-1323 extensions */
#define TCPCTL_MSSDFLT 3 /* MSS default */
#define TCPCTL_STATS 4 /* statistics */
#define TCPCTL_RTTDFLT 5 /* default RTT estimate */
#define TCPCTL_KEEPIDLE 6 /* keepalive idle timer */
#define TCPCTL_KEEPINTVL 7 /* interval to send keepalives */
#define TCPCTL_SENDSPACE 8 /* send buffer space */
#define TCPCTL_RECVSPACE 9 /* receive buffer space */
#define TCPCTL_KEEPINIT 10 /* timeout for establishing syn */
#define TCPCTL_PCBLIST 11 /* list of all outstanding PCBs */
#define TCPCTL_DELACKTIME 12 /* time before sending delayed ACK */
#define TCPCTL_V6MSSDFLT 13 /* MSS default for IPv6 */
#define TCPCTL_SACK 14 /* Selective Acknowledgement,rfc 2018 */
#define TCPCTL_DROP 15 /* drop tcp connection */
#define TCPCTL_STATES 16 /* connection counts by TCP state */
#ifdef _KERNEL
#ifdef SYSCTL_DECL
SYSCTL_DECL(_net_inet_tcp);
SYSCTL_DECL(_net_inet_tcp_sack);
MALLOC_DECLARE(M_TCPLOG);
#endif
VNET_DECLARE(int, tcp_log_in_vain);
#define V_tcp_log_in_vain VNET(tcp_log_in_vain)
/*
* Global TCP tunables shared between different stacks.
* Please keep the list sorted.
*/
VNET_DECLARE(int, drop_synfin);
VNET_DECLARE(int, path_mtu_discovery);
VNET_DECLARE(int, tcp_abc_l_var);
VNET_DECLARE(int, tcp_autorcvbuf_max);
VNET_DECLARE(int, tcp_autosndbuf_inc);
VNET_DECLARE(int, tcp_autosndbuf_max);
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(int, tcp_delack_enabled);
VNET_DECLARE(int, tcp_do_autorcvbuf);
VNET_DECLARE(int, tcp_do_autosndbuf);
VNET_DECLARE(int, tcp_do_ecn);
VNET_DECLARE(int, tcp_do_lrd);
VNET_DECLARE(int, tcp_do_prr);
VNET_DECLARE(int, tcp_do_prr_conservative);
VNET_DECLARE(int, tcp_do_newcwv);
VNET_DECLARE(int, tcp_do_rfc1323);
VNET_DECLARE(int, tcp_tolerate_missing_ts);
VNET_DECLARE(int, tcp_do_rfc3042);
VNET_DECLARE(int, tcp_do_rfc3390);
This commit marks the first formal contribution of the "Five New TCP Congestion Control Algorithms for FreeBSD" FreeBSD Foundation funded project. More details about the project are available at: http://caia.swin.edu.au/freebsd/5cc/ - Add a KPI and supporting infrastructure to allow modular congestion control algorithms to be used in the net stack. Algorithms can maintain per-connection state if required, and connections maintain their own algorithm pointer, which allows different connections to concurrently use different algorithms. The TCP_CONGESTION socket option can be used with getsockopt()/setsockopt() to programmatically query or change the congestion control algorithm respectively from within an application at runtime. - Integrate the framework with the TCP stack in as least intrusive a manner as possible. Care was also taken to develop the framework in a way that should allow integration with other congestion aware transport protocols (e.g. SCTP) in the future. The hope is that we will one day be able to share a single set of congestion control algorithm modules between all congestion aware transport protocols. - Introduce a new congestion recovery (TF_CONGRECOVERY) state into the TCP stack and use it to decouple the meaning of recovery from a congestion event and recovery from packet loss (TF_FASTRECOVERY) a la RFC2581. ECN and delay based congestion control protocols don't generally need to recover from packet loss and need a different way to note a congestion recovery episode within the stack. - Remove the net.inet.tcp.newreno sysctl, which simplifies some portions of code and ensures the stack always uses the appropriate mechanisms for recovering from packet loss during a congestion recovery episode. - Extract the NewReno congestion control algorithm from the TCP stack and massage it into module form. NewReno is always built into the kernel and will remain the default algorithm for the forseeable future. Implementations of additional different algorithms will become available in the near future. - Bump __FreeBSD_version to 900025 and note in UPDATING that rebuilding code that relies on the size of "struct tcpcb" is required. Many thanks go to the Cisco University Research Program Fund at Community Foundation Silicon Valley and the FreeBSD Foundation. Their support of our work at the Centre for Advanced Internet Architectures, Swinburne University of Technology is greatly appreciated. In collaboration with: David Hayes <dahayes at swin edu au> and Grenville Armitage <garmitage at swin edu au> Sponsored by: Cisco URP, FreeBSD Foundation Reviewed by: rpaulo Tested by: David Hayes (and many others over the years) MFC after: 3 months
2010-11-12 06:41:55 +00:00
VNET_DECLARE(int, tcp_do_rfc3465);
VNET_DECLARE(int, tcp_do_newsack);
VNET_DECLARE(int, tcp_do_sack);
VNET_DECLARE(int, tcp_do_tso);
VNET_DECLARE(int, tcp_ecn_maxretries);
VNET_DECLARE(int, tcp_initcwnd_segments);
VNET_DECLARE(int, tcp_insecure_rst);
VNET_DECLARE(int, tcp_insecure_syn);
VNET_DECLARE(uint32_t, tcp_map_entries_limit);
VNET_DECLARE(uint32_t, tcp_map_split_limit);
VNET_DECLARE(int, tcp_minmss);
VNET_DECLARE(int, tcp_mssdflt);
#ifdef STATS
VNET_DECLARE(int, tcp_perconn_stats_dflt_tpl);
VNET_DECLARE(int, tcp_perconn_stats_enable);
#endif /* STATS */
VNET_DECLARE(int, tcp_recvspace);
VNET_DECLARE(int, tcp_sack_globalholes);
VNET_DECLARE(int, tcp_sack_globalmaxholes);
VNET_DECLARE(int, tcp_sack_maxholes);
VNET_DECLARE(int, tcp_sc_rst_sock_fail);
VNET_DECLARE(int, tcp_sendspace);
VNET_DECLARE(int, tcp_udp_tunneling_overhead);
VNET_DECLARE(int, tcp_udp_tunneling_port);
VNET_DECLARE(struct inpcbinfo, tcbinfo);
#define V_tcp_do_lrd VNET(tcp_do_lrd)
#define V_tcp_do_prr VNET(tcp_do_prr)
#define V_tcp_do_prr_conservative VNET(tcp_do_prr_conservative)
#define V_tcp_do_newcwv VNET(tcp_do_newcwv)
#define V_drop_synfin VNET(drop_synfin)
#define V_path_mtu_discovery VNET(path_mtu_discovery)
#define V_tcbinfo VNET(tcbinfo)
#define V_tcp_abc_l_var VNET(tcp_abc_l_var)
#define V_tcp_autorcvbuf_max VNET(tcp_autorcvbuf_max)
#define V_tcp_autosndbuf_inc VNET(tcp_autosndbuf_inc)
#define V_tcp_autosndbuf_max VNET(tcp_autosndbuf_max)
#define V_tcp_delack_enabled VNET(tcp_delack_enabled)
#define V_tcp_do_autorcvbuf VNET(tcp_do_autorcvbuf)
#define V_tcp_do_autosndbuf VNET(tcp_do_autosndbuf)
#define V_tcp_do_ecn VNET(tcp_do_ecn)
#define V_tcp_do_rfc1323 VNET(tcp_do_rfc1323)
#define V_tcp_tolerate_missing_ts VNET(tcp_tolerate_missing_ts)
#define V_tcp_ts_offset_per_conn VNET(tcp_ts_offset_per_conn)
#define V_tcp_do_rfc3042 VNET(tcp_do_rfc3042)
#define V_tcp_do_rfc3390 VNET(tcp_do_rfc3390)
#define V_tcp_do_rfc3465 VNET(tcp_do_rfc3465)
#define V_tcp_do_newsack VNET(tcp_do_newsack)
#define V_tcp_do_sack VNET(tcp_do_sack)
#define V_tcp_do_tso VNET(tcp_do_tso)
#define V_tcp_ecn_maxretries VNET(tcp_ecn_maxretries)
#define V_tcp_initcwnd_segments VNET(tcp_initcwnd_segments)
#define V_tcp_insecure_rst VNET(tcp_insecure_rst)
#define V_tcp_insecure_syn VNET(tcp_insecure_syn)
#define V_tcp_map_entries_limit VNET(tcp_map_entries_limit)
#define V_tcp_map_split_limit VNET(tcp_map_split_limit)
#define V_tcp_minmss VNET(tcp_minmss)
#define V_tcp_mssdflt VNET(tcp_mssdflt)
#ifdef STATS
#define V_tcp_perconn_stats_dflt_tpl VNET(tcp_perconn_stats_dflt_tpl)
#define V_tcp_perconn_stats_enable VNET(tcp_perconn_stats_enable)
#endif /* STATS */
#define V_tcp_recvspace VNET(tcp_recvspace)
#define V_tcp_sack_globalholes VNET(tcp_sack_globalholes)
#define V_tcp_sack_globalmaxholes VNET(tcp_sack_globalmaxholes)
#define V_tcp_sack_maxholes VNET(tcp_sack_maxholes)
#define V_tcp_sc_rst_sock_fail VNET(tcp_sc_rst_sock_fail)
#define V_tcp_sendspace VNET(tcp_sendspace)
#define V_tcp_udp_tunneling_overhead VNET(tcp_udp_tunneling_overhead)
#define V_tcp_udp_tunneling_port VNET(tcp_udp_tunneling_port)
#ifdef TCP_HHOOK
VNET_DECLARE(struct hhook_head *, tcp_hhh[HHOOK_TCP_LAST + 1]);
#define V_tcp_hhh VNET(tcp_hhh)
#endif
int tcp_addoptions(struct tcpopt *, u_char *);
int tcp_ccalgounload(struct cc_algo *unload_algo);
1994-05-24 10:09:53 +00:00
struct tcpcb *
2002-03-19 21:25:46 +00:00
tcp_close(struct tcpcb *);
Update TCP for infrastructural changes to the socket/pcb refcount model, pru_abort(), pru_detach(), and in_pcbdetach(): - Universally support and enforce the invariant that so_pcb is never NULL, converting dozens of unnecessary NULL checks into assertions, and eliminating dozens of unnecessary error handling cases in protocol code. - In some cases, eliminate unnecessary pcbinfo locking, as it is no longer required to ensure so_pcb != NULL. For example, the receive code no longer requires the pcbinfo lock, and the send code only requires it if building a new connection on an otherwise unconnected socket triggered via sendto() with an address. This should significnatly reduce tcbinfo lock contention in the receive and send cases. - In order to support the invariant that so_pcb != NULL, it is now necessary for the TCP code to not discard the tcpcb any time a connection is dropped, but instead leave the tcpcb until the socket is shutdown. This case is handled by setting INP_DROPPED, to substitute for using a NULL so_pcb to indicate that the connection has been dropped. This requires the inpcb lock, but not the pcbinfo lock. - Unlike all other protocols in the tree, TCP may need to retain access to the socket after the file descriptor has been closed. Set SS_PROTOREF in tcp_detach() in order to prevent the socket from being freed, and add a flag, INP_SOCKREF, so that the TCP code knows whether or not it needs to free the socket when the connection finally does close. The typical case where this occurs is if close() is called on a TCP socket before all sent data in the send socket buffer has been transmitted or acknowledged. If INP_SOCKREF is found when the connection is dropped, we release the inpcb, tcpcb, and socket instead of flagging INP_DROPPED. - Abort and detach protocol switch methods no longer return failures, nor attempt to free sockets, as the socket layer does this. - Annotate the existence of a long-standing race in the TCP timer code, in which timers are stopped but not drained when the socket is freed, as waiting for drain may lead to deadlocks, or have to occur in a context where waiting is not permitted. This race has been handled by testing to see if the tcpcb pointer in the inpcb is NULL (and vice versa), which is not normally permitted, but may be true of a inpcb and tcpcb have been freed. Add a counter to test how often this race has actually occurred, and a large comment for each instance where we compare potentially freed memory with NULL. This will have to be fixed in the near future, but requires is to further address how to handle the timer shutdown shutdown issue. - Several TCP calls no longer potentially free the passed inpcb/tcpcb, so no longer need to return a pointer to indicate whether the argument passed in is still valid. - Un-macroize debugging and locking setup for various protocol switch methods for TCP, as it lead to more obscurity, and as locking becomes more customized to the methods, offers less benefit. - Assert copyright on tcp_usrreq.c due to significant modifications that have been made as part of this work. These changes significantly modify the memory management and connection logic of our TCP implementation, and are (as such) High Risk Changes, and likely to contain serious bugs. Please report problems to the current@ mailing list ASAP, ideally with simple test cases, and optionally, packet traces. MFC after: 3 months
2006-04-01 16:36:36 +00:00
void tcp_discardcb(struct tcpcb *);
bool tcp_freecb(struct tcpcb *);
void tcp_twstart(struct tcpcb *);
void tcp_twclose(struct tcptw *, int);
2002-03-19 21:25:46 +00:00
void tcp_ctlinput(int, struct sockaddr *, void *);
int tcp_ctloutput(struct socket *, struct sockopt *);
void tcp_ctlinput_viaudp(int, struct sockaddr *, void *, void *);
1994-05-24 10:09:53 +00:00
struct tcpcb *
2002-03-19 21:25:46 +00:00
tcp_drop(struct tcpcb *, int);
void tcp_drain(void);
void tcp_init(void);
void tcp_fini(void *);
2011-01-07 21:40:34 +00:00
char *tcp_log_addrs(struct in_conninfo *, struct tcphdr *, void *,
const void *);
char *tcp_log_vain(struct in_conninfo *, struct tcphdr *, void *,
const void *);
int tcp_reass(struct tcpcb *, struct tcphdr *, tcp_seq *, int *,
struct mbuf *);
void tcp_reass_global_init(void);
void tcp_reass_flush(struct tcpcb *);
void tcp_dooptions(struct tcpopt *, u_char *, int, int);
void tcp_dropwithreset(struct mbuf *, struct tcphdr *,
struct tcpcb *, int, int);
void tcp_pulloutofband(struct socket *,
struct tcphdr *, struct mbuf *, int);
void tcp_xmit_timer(struct tcpcb *, int);
void tcp_newreno_partial_ack(struct tcpcb *, struct tcphdr *);
void cc_ack_received(struct tcpcb *tp, struct tcphdr *th,
uint16_t nsegs, uint16_t type);
void cc_conn_init(struct tcpcb *tp);
void cc_post_recovery(struct tcpcb *tp, struct tcphdr *th);
void cc_ecnpkt_handler(struct tcpcb *tp, struct tcphdr *th, uint8_t iptos);
void cc_ecnpkt_handler_flags(struct tcpcb *tp, uint16_t flags, uint8_t iptos);
void cc_cong_signal(struct tcpcb *tp, struct tcphdr *th, uint32_t type);
#ifdef TCP_HHOOK
void hhook_run_tcp_est_in(struct tcpcb *tp,
struct tcphdr *th, struct tcpopt *to);
#endif
int tcp_input(struct mbuf **, int *, int);
int tcp_autorcvbuf(struct mbuf *, struct tcphdr *, struct socket *,
struct tcpcb *, int);
int tcp_input_with_port(struct mbuf **, int *, int, uint16_t);
void tcp_handle_wakeup(struct tcpcb *, struct socket *);
void tcp_do_segment(struct mbuf *, struct tcphdr *,
struct socket *, struct tcpcb *, int, int, uint8_t);
int register_tcp_functions(struct tcp_function_block *blk, int wait);
int register_tcp_functions_as_names(struct tcp_function_block *blk,
int wait, const char *names[], int *num_names);
int register_tcp_functions_as_name(struct tcp_function_block *blk,
const char *name, int wait);
int deregister_tcp_functions(struct tcp_function_block *blk, bool quiesce,
bool force);
struct tcp_function_block *find_and_ref_tcp_functions(struct tcp_function_set *fs);
int find_tcp_function_alias(struct tcp_function_block *blk, struct tcp_function_set *fs);
void tcp_switch_back_to_default(struct tcpcb *tp);
struct tcp_function_block *
find_and_ref_tcp_fb(struct tcp_function_block *fs);
int tcp_default_ctloutput(struct socket *so, struct sockopt *sopt, struct inpcb *inp, struct tcpcb *tp);
extern counter_u64_t tcp_inp_lro_direct_queue;
extern counter_u64_t tcp_inp_lro_wokeup_queue;
extern counter_u64_t tcp_inp_lro_compressed;
extern counter_u64_t tcp_inp_lro_locks_taken;
extern counter_u64_t tcp_extra_mbuf;
extern counter_u64_t tcp_would_have_but;
extern counter_u64_t tcp_comp_total;
extern counter_u64_t tcp_uncomp_total;
extern counter_u64_t tcp_bad_csums;
#ifdef NETFLIX_EXP_DETECTION
/* Various SACK attack thresholds */
extern int32_t tcp_force_detection;
extern int32_t tcp_sack_to_ack_thresh;
extern int32_t tcp_sack_to_move_thresh;
extern int32_t tcp_restoral_thresh;
extern int32_t tcp_sad_decay_val;
extern int32_t tcp_sad_pacing_interval;
extern int32_t tcp_sad_low_pps;
extern int32_t tcp_map_minimum;
extern int32_t tcp_attack_on_turns_on_logging;
#endif
extern uint32_t tcp_ack_war_time_window;
extern uint32_t tcp_ack_war_cnt;
uint32_t tcp_maxmtu(struct in_conninfo *, struct tcp_ifcap *);
uint32_t tcp_maxmtu6(struct in_conninfo *, struct tcp_ifcap *);
void tcp6_use_min_mtu(struct tcpcb *);
u_int tcp_maxseg(const struct tcpcb *);
u_int tcp_fixed_maxseg(const struct tcpcb *);
void tcp_mss_update(struct tcpcb *, int, int, struct hc_metrics_lite *,
struct tcp_ifcap *);
2002-03-19 21:25:46 +00:00
void tcp_mss(struct tcpcb *, int);
int tcp_mssopt(struct in_conninfo *);
struct inpcb *
tcp_drop_syn_sent(struct inpcb *, int);
1994-05-24 10:09:53 +00:00
struct tcpcb *
2002-03-19 21:25:46 +00:00
tcp_newtcpcb(struct inpcb *);
int tcp_output(struct tcpcb *);
void tcp_state_change(struct tcpcb *, int);
2002-03-19 21:25:46 +00:00
void tcp_respond(struct tcpcb *, void *,
struct tcphdr *, struct mbuf *, tcp_seq, tcp_seq, int);
void tcp_tw_init(void);
#ifdef VIMAGE
void tcp_tw_destroy(void);
#endif
void tcp_tw_zone_change(void);
int tcp_twcheck(struct inpcb *, struct tcpopt *, struct tcphdr *,
struct mbuf *, int);
2002-03-19 21:25:46 +00:00
void tcp_setpersist(struct tcpcb *);
void tcp_slowtimo(void);
void tcp_record_dsack(struct tcpcb *tp, tcp_seq start, tcp_seq end, int tlp);
struct tcptemp *
tcpip_maketemplate(struct inpcb *);
void tcpip_fillheaders(struct inpcb *, uint16_t, void *, void *);
void tcp_timer_activate(struct tcpcb *, uint32_t, u_int);
int tcp_timer_suspend(struct tcpcb *, uint32_t);
void tcp_timers_unsuspend(struct tcpcb *, uint32_t);
int tcp_timer_active(struct tcpcb *, uint32_t);
void tcp_timer_stop(struct tcpcb *, uint32_t);
void tcp_trace(short, short, struct tcpcb *, void *, struct tcphdr *, int);
int inp_to_cpuid(struct inpcb *inp);
/*
* All tcp_hc_* functions are IPv4 and IPv6 (via in_conninfo)
*/
void tcp_hc_init(void);
#ifdef VIMAGE
void tcp_hc_destroy(void);
#endif
void tcp_hc_get(struct in_conninfo *, struct hc_metrics_lite *);
uint32_t tcp_hc_getmtu(struct in_conninfo *);
void tcp_hc_updatemtu(struct in_conninfo *, uint32_t);
void tcp_hc_update(struct in_conninfo *, struct hc_metrics_lite *);
extern struct pr_usrreqs tcp_usrreqs;
uint32_t tcp_new_ts_offset(struct in_conninfo *);
tcp_seq tcp_new_isn(struct in_conninfo *);
One of the ways to detect loss is to count duplicate acks coming back from the other end till it reaches predetermined threshold which is 3 for us right now. Once that happens, we trigger fast-retransmit to do loss recovery. Main problem with the current implementation is that we don't honor SACK information well to detect whether an incoming ack is a dupack or not. RFC6675 has latest recommendations for that. According to it, dupack is a segment that arrives carrying a SACK block that identifies previously unknown information between snd_una and snd_max even if it carries new data, changes the advertised window, or moves the cumulative acknowledgment point. With the prevalence of Selective ACK (SACK) these days, improper handling can lead to delayed loss recovery. With the fix, new behavior looks like following: 0) th_ack < snd_una --> ignore Old acks are ignored. 1) th_ack == snd_una, !sack_changed --> ignore Acks with SACK enabled but without any new SACK info in them are ignored. 2) th_ack == snd_una, window == old_window --> increment Increment on a good dupack. 3) th_ack == snd_una, window != old_window, sack_changed --> increment When SACK enabled, it's okay to have advertized window changed if the ack has new SACK info. 4) th_ack > snd_una --> reset to 0 Reset to 0 when left edge moves. 5) th_ack > snd_una, sack_changed --> increment Increment if left edge moves but there is new SACK info. Here, sack_changed is the indicator that incoming ack has previously unknown SACK info in it. Note: This fix is not fully compliant to RFC6675. That may require a few changes to current implementation in order to keep per-sackhole dupack counter and change to the way we mark/handle sack holes. PR: 203663 Reviewed by: jtl MFC after: 3 weeks Sponsored by: Limelight Networks Differential Revision: https://reviews.freebsd.org/D4225
2015-12-08 21:21:48 +00:00
int tcp_sack_doack(struct tcpcb *, struct tcpopt *, tcp_seq);
int tcp_dsack_block_exists(struct tcpcb *);
void tcp_update_dsack_list(struct tcpcb *, tcp_seq, tcp_seq);
void tcp_update_sack_list(struct tcpcb *tp, tcp_seq rcv_laststart, tcp_seq rcv_lastend);
void tcp_clean_dsack_blocks(struct tcpcb *tp);
void tcp_clean_sackreport(struct tcpcb *tp);
void tcp_sack_adjust(struct tcpcb *tp);
struct sackhole *tcp_sack_output(struct tcpcb *tp, int *sack_bytes_rexmt);
void tcp_do_prr_ack(struct tcpcb *, struct tcphdr *, struct tcpopt *);
void tcp_lost_retransmission(struct tcpcb *, struct tcphdr *);
void tcp_sack_partialack(struct tcpcb *, struct tcphdr *);
void tcp_free_sackholes(struct tcpcb *tp);
void tcp_sack_lost_retransmission(struct tcpcb *, struct tcphdr *);
int tcp_newreno(struct tcpcb *, struct tcphdr *);
Calculate the correct amount of bytes that are in-flight for a connection as suggested by RFC 6675. Currently differnt places in the stack tries to guess this in suboptimal ways. The main problem is that current calculations don't take sacked bytes into account. Sacked bytes are the bytes receiver acked via SACK option. This is suboptimal because it assumes that network has more outstanding (unacked) bytes than the actual value and thus sends less data by setting congestion window lower than what's possible which in turn may cause slower recovery from losses. As an example, one of the current calculations looks something like this: snd_nxt - snd_fack + sackhint.sack_bytes_rexmit New proposal from RFC 6675 is: snd_max - snd_una - sackhint.sacked_bytes + sackhint.sack_bytes_rexmit which takes sacked bytes into account which is a new addition to the sackhint struct. Only thing we are missing from RFC 6675 is isLost() i.e. segment being considered lost and thus adjusting pipe based on that which makes this calculation a bit on conservative side. The approach is very simple. We already process each ack with sack info in tcp_sack_doack() and extract sack blocks/holes out of it. We'd now also track this new variable sacked_bytes which keeps track of total sacked bytes reported. One downside to this approach is that we may get incorrect count of sacked_bytes if the other end decides to drop sack info in the ack because of memory pressure or some other reasons. But in this (not very likely) case also the pipe calculation would be conservative which is okay as opposed to being aggressive in sending packets into the network. Next step is to use this more accurate pipe estimation to drive congestion window adjustments. In collaboration with: rrs Reviewed by: jason_eggnet dot com, rrs MFC after: 2 weeks Sponsored by: Limelight Networks Differential Revision: https://reviews.freebsd.org/D3971
2015-10-28 22:57:51 +00:00
int tcp_compute_pipe(struct tcpcb *);
uint32_t tcp_compute_initwnd(uint32_t);
void tcp_sndbuf_autoscale(struct tcpcb *, struct socket *, uint32_t);
int tcp_stats_sample_rollthedice(struct tcpcb *tp, void *seed_bytes,
size_t seed_len);
int tcp_can_enable_pacing(void);
void tcp_decrement_paced_conn(void);
struct mbuf *
tcp_m_copym(struct mbuf *m, int32_t off0, int32_t *plen,
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
int32_t seglimit, int32_t segsize, struct sockbuf *sb, bool hw_tls);
int tcp_stats_init(void);
void tcp_log_end_status(struct tcpcb *tp, uint8_t status);
static inline void
tcp_fields_to_host(struct tcphdr *th)
{
th->th_seq = ntohl(th->th_seq);
th->th_ack = ntohl(th->th_ack);
th->th_win = ntohs(th->th_win);
th->th_urp = ntohs(th->th_urp);
}
static inline void
tcp_fields_to_net(struct tcphdr *th)
{
th->th_seq = htonl(th->th_seq);
th->th_ack = htonl(th->th_ack);
th->th_win = htons(th->th_win);
th->th_urp = htons(th->th_urp);
}
static inline void
tcp_account_for_send(struct tcpcb *tp, uint32_t len, uint8_t is_rxt,
uint8_t is_tlp, int hw_tls)
{
if (is_tlp) {
tp->t_sndtlppack++;
tp->t_sndtlpbyte += len;
}
/* To get total bytes sent you must add t_snd_rxt_bytes to t_sndbytes */
if (is_rxt)
tp->t_snd_rxt_bytes += len;
else
tp->t_sndbytes += len;
#ifdef KERN_TLS
if (hw_tls && is_rxt && len != 0) {
uint64_t rexmit_percent = (1000ULL * tp->t_snd_rxt_bytes) / (10ULL * (tp->t_snd_rxt_bytes + tp->t_sndbytes));
if (rexmit_percent > ktls_ifnet_max_rexmit_pct)
ktls_disable_ifnet(tp);
}
#endif
}
#endif /* _KERNEL */
1994-08-21 05:27:42 +00:00
#endif /* _NETINET_TCP_VAR_H_ */