man/man4/dtrace_tcp.4

.\" Copyright (c) 2015 Mark Johnston <markj@FreeBSD.org>
.\" All rights reserved.
.\"
.\" Redistribution and use in source and binary forms, with or without
.\" modification, are permitted provided that the following conditions
.\" are met:
.\" 1. Redistributions of source code must retain the above copyright
.\"    notice, this list of conditions and the following disclaimer.
.\" 2. Redistributions in binary form must reproduce the above copyright
.\"    notice, this list of conditions and the following disclaimer in the
.\"    documentation and/or other materials provided with the distribution.
.\"
.\" THIS SOFTWARE IS PROVIDED BY THE AUTHOR AND CONTRIBUTORS ``AS IS'' AND
.\" ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
.\" IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
.\" ARE DISCLAIMED.  IN NO EVENT SHALL THE AUTHOR OR CONTRIBUTORS BE LIABLE
.\" FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
.\" DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS
.\" OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
.\" HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT
.\" LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY
.\" OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF
.\" SUCH DAMAGE.
.\"
.\" $FreeBSD$
.\"
.Dd July 2, 2023
.Dt DTRACE_TCP 4
.Os
.Sh NAME
.Nm dtrace_tcp
.Nd a DTrace provider for tracing events related to the
.Xr tcp 4
protocol
.Sh SYNOPSIS
.Fn tcp:::accept-established "pktinfo_t *" "csinfo_t *" "ipinfo_t *" \
    "tcpsinfo_t *" "tcpinfo_t *"
.Fn tcp:::accept-refused "pktinfo_t *" "csinfo_t *" "ipinfo_t *" \
    "tcpsinfo_t *" "tcpinfo_t *"
.Fn tcp:::connect-established "pktinfo_t *" "csinfo_t *" "ipinfo_t *" \
    "tcpsinfo_t *" "tcpinfo_t *"
.Fn tcp:::connect-refused "pktinfo_t *" "csinfo_t *" "ipinfo_t *" \
    "tcpsinfo_t *" "tcpinfo_t *"
.Fn tcp:::connect-request "pktinfo_t *" "csinfo_t *" "ipinfo_t *" \
    "tcpsinfo_t *" "tcpinfo_t *"
.Fn tcp:::receive "pktinfo_t *" "csinfo_t *" "ipinfo_t *" "tcpsinfo_t *" \
    "tcpinfo_t *"
.Fn tcp:::send "pktinfo_t *" "csinfo_t *" "ipinfo_t *" "tcpsinfo_t *" \
    "tcpinfo_t *"
.Fn tcp:::state-change "void *" "csinfo_t *" "void *" "tcpsinfo_t *" "void *" \
    "tcplsinfo_t *"
.Fn tcp:::siftr "siftrinfo_t *"
.Sh DESCRIPTION
The DTrace
.Nm tcp
provider allows users to trace events in the
.Xr tcp 4
protocol implementation.
This provider is similar to the
.Xr dtrace_ip 4
and
.Xr dtrace_udp 4
providers, but additionally contains probes corresponding to protocol events at
a level higher than packet reception and transmission.
All
.Nm tcp
probes except for
.Fn tcp:::state-change
and
.Fn tcp:::siftr
have the same number and type of arguments.
The last three arguments are used to describe a TCP segment: the
.Vt ipinfo_t
argument exposes the version-agnostic fields of the IP header, while the
.Vt tcpinfo_t
argument exposes the TCP header, and the
.Vt tcpsinfo_t
argument describes details of the corresponding TCP connection state, if any.
Their fields are described in the ARGUMENTS section.
.Pp
The
.Fn tcp:::accept-established
probe fires when a remotely-initiated active TCP open succeeds.
At this point the new connection is in the ESTABLISHED state, and the probe
arguments expose the headers associated with the final ACK of the three-way
handshake.
The
.Fn tcp:::accept-refused
probe fires when a SYN arrives on a port without a listening socket.
The probe arguments expose the headers associated with the RST to be transmitted
to the remote host in response to the SYN segment.
.Pp
The
.Fn tcp:::connect-established ,
.Fn tcp:::connect-refused ,
and
.Fn tcp:::connect-request
probes are similar to the
.Ql accept
probes, except that they correspond to locally-initiated TCP connections.
The
.Fn tcp:::connect-established
probe fires when the SYN-ACK segment of a three-way handshake is received from
the remote host and a final ACK is prepared for transmission.
This occurs immediately after the local connection state transitions from
SYN-SENT to ESTABLISHED.
The probe arguments describe the headers associated with the received SYN-ACK
segment.
The
.Fn tcp:::connect-refused
probe fires when the local host receives a RST segment in response to a SYN
segment, indicating that the remote host refused to open a connection.
The probe arguments describe the IP and TCP headers associated with the received
RST segment.
The
.Fn tcp:::connect-request
probe fires as the kernel prepares to transmit the initial SYN segment of a
three-way handshake.
.Pp
The
.Fn tcp:::send
and
.Fn tcp:::receive
probes fire when the host sends or receives a TCP packet, respectively.
As with the
.Xr dtrace_udp 4
provider,
.Nm tcp
probes fire only for packets sent by or to the local host; forwarded packets are
handled in the IP layer and are only visible to the
.Xr dtrace_ip 4
provider.
.Pp
The
.Fn tcp:::state-change
probe fires upon local TCP connection state transitions.
Its first, third and fifth arguments are currently always
.Dv NULL .
Its last argument describes the from-state in the transition, and the to-state
can be obtained from
.Dv args[3]->tcps_state .
.Pp
The
.Fn tcp:::siftr
probe fires when a TCP segment is sent or received by the host.
For a detailed description see
.Xr siftr 4 .
The
.Vt siftrinfo_t
argument provides the information about the TCP connection.
.Sh ARGUMENTS
The
.Vt pktinfo_t
argument is currently unimplemented and is included for compatibility with other
implementations of this provider.
Its fields are:
.Bl -tag -width "uinptr_t pkt_addr" -offset indent
.It Vt uinptr_t pkt_addr
Always set to 0.
.El
.Pp
The
.Vt csinfo_t
argument is currently unimplemented and is included for compatibility with other
implementations of this provider.
Its fields are:
.Bl -tag -width "uintptr_t cs_addr" -offset indent
.It Vt uintptr_t cs_addr
Always set to 0.
.It Vt uint64_t cs_cid
A pointer to the
.Vt struct inpcb
for this packet, or
.Dv NULL .
.It Vt pid_t cs_pid
Always set to 0.
.El
.Pp
The
.Vt ipinfo_t
type is a version-agnostic representation of fields from an IP header.
Its fields are described in the
.Xr dtrace_ip 4
manual page.
.Pp
The
.Vt tcpsinfo_t
type is used to provide a stable representation of TCP connection state.
Some
.Nm tcp
probes, such as
.Fn tcp:::accept-refused ,
fire in a context where there is no TCP connection; this argument is
.Dv NULL
in that case.
Its fields are:
.Bl -tag -width "uint16_t tcps_lport" -offset indent
.It Vt uintptr_t tcps_addr
The address of the corresponding TCP control block.
This is currently a pointer to a
.Vt struct tcpcb .
.It Vt int tcps_local
A boolean indicating whether the connection is local to the host.
Currently unimplemented and always set to -1.
.It Vt int tcps_active
A boolean indicating whether the connection was initiated by the local host.
Currently unimplemented and always set to -1.
.It Vt uint16_t tcps_lport
Local TCP port.
.It Vt uint16_t tcps_rport
Remote TCP port.
.It Vt string tcps_laddr
Local address.
.It Vt string tcps_raddr
Remote address.
.It Vt int32_t tcps_state
Current TCP state.
The valid TCP state values are given by the constants prefixed with
.Ql TCPS_
in
.Pa /usr/lib/dtrace/tcp.d .
.It Vt uint32_t tcps_iss
Initial send sequence number.
.It Vt uint32_t tcps_suna
Initial sequence number of sent but unacknowledged data.
.It Vt uint32_t tcps_snxt
Next sequence number for send.
.It Vt uint32_t tcps_rack
Sequence number of received and acknowledged data.
.It Vt uint32_t tcps_rnxt
Next expected sequence number for receive.
.It Vt u_long tcps_swnd
TCP send window size.
.It Vt int32_t tcps_snd_ws
Window scaling factor for the TCP send window.
.It Vt u_long tcps_rwnd
TCP receive window size.
.It Vt int32_t tcps_rcv_ws
Window scaling factor for the TCP receive window.
.It Vt u_long tcps_cwnd
TCP congestion window size.
.It Vt u_long tcps_cwnd_ssthresh
Congestion window threshold at which slow start ends and congestion avoidance
begins.
.It Vt uint32_t tcps_sack_fack
Last sequence number selectively acknowledged by the receiver.
.It Vt uint32_t tcps_sack_snxt
Next selectively acknowledge sequence number at which to begin retransmitting.
.It Vt uint32_t tcps_rto
Round-trip timeout, in milliseconds.
.It Vt uint32_t tcps_mss
Maximum segment size.
.It Vt int tcps_retransmit
A boolean indicating that the local sender is retransmitting data.
.It Vt int tcps_srtt
Smoothed round-trip time.
.El
.Pp
The
.Vt tcpinfo_t
type exposes the fields in a TCP segment header in host order.
Its fields are:
.Bl -tag -width "struct tcphdr *tcp_hdr" -offset indent
.It Vt uint16_t tcp_sport
Source TCP port.
.It Vt uint16_t tcp_dport
Destination TCP port.
.It Vt uint32_t tcp_seq
Sequence number.
.It Vt uint32_t tcp_ack
Acknowledgement number.
.It Vt uint8_t tcp_offset
Data offset, in bytes.
.It Vt uint8_t tcp_flags
TCP flags.
.It Vt uint16_t tcp_window
TCP window size.
.It Vt uint16_t tcp_checksum
Checksum.
.It Vt uint16_t tcp_urgent
Urgent data pointer.
.It Vt struct tcphdr *tcp_hdr
A pointer to the raw TCP header.
.El
.Pp
The
.Vt tcplsinfo_t
type is used by the
.Fn tcp:::state-change
probe to provide the from-state of a transition.
Its fields are:
.Bl -tag -width "int32_t tcps_state" -offset indent
.It Vt int32_t tcps_state
A TCP state.
The valid TCP state values are given by the constants prefixed with
.Ql TCPS_
in
.Pa /usr/lib/dtrace/tcp.d .
.El
.Pp
The
.Vt siftrinfo_t
type is used by the
.Fn tcp:::siftr
probe to provide the state of the TCP connection.
Its fields are:
.Bl -tag -width "u_int sent_inflight_bytes" -offset indent
.It Vt uint8_t direction
Direction of packet that triggered the log message.
Either
.Qq 0
for in, or
.Qq 1
for out.
.It Vt uint8_t ipver
The version of the IP protocol being used.
Either
.Qq 1
for IPv4, or
.Qq 2
for IPv6.
.It Vt uint16_t lport
The TCP port that the local host is communicating via.
.It Vt uint16_t rport
The TCP port that the remote host is communicating via.
.It Vt string laddr
The IPv4 or IPv6 address of the local host.
.It Vt string raddr
The IPv4 or IPv6 address of the remote host.
.It Vt uint32_t snd_cwnd
The current congestion window (CWND) for the flow, in bytes.
.It Vt uint32_t snd_wnd
The current sending window for the flow, in bytes.
The post scaled value is reported, except during the initial handshake (first
few packets), during which time the unscaled value is reported.
.It Vt uint32_t rcv_wnd
The current receive window for the flow, in bytes.
The post scaled value is always reported.
.It Vt uint32_t t_flags2
The current value of the t_flags2 for the flow.
.It Vt uint32_t snd_ssthresh
The slow start threshold (SSTHRESH) for the flow, in bytes.
.It Vt int conn_state
A TCP state.
The valid TCP state values are given by the constants prefixed with
.Ql TCPS_
in
.Pa /usr/lib/dtrace/tcp.d .
.It Vt uint32_t mss
The maximum segment size (MSS) for the flow, in bytes.
.It Vt uint32_t srtt
The current smoothed RTT (SRTT) for the flow in microseconds.
.It Vt u_char sack_enabled
SACK enabled indicator. 1 if SACK enabled, 0 otherwise.
.It Vt u_char snd_scale
The current window scaling factor for the sending window.
.It Vt u_char rcv_scale
The current window scaling factor for the receiving window.
.It Vt u_int t_flags
The current value of the t_flags for the flow.
.It Vt uint32_t rto
The current retransmission timeout (RTO) for the flow in microseconds.
Divide by HZ to get the timeout length in seconds.
.It Vt u_int snd_buf_hiwater
The current size of the socket send buffer in bytes.
.It Vt u_int snd_buf_cc
The current number of bytes in the socket send buffer.
.It Vt u_int rcv_buf_hiwater
The current size of the socket receive buffer in bytes.
.It Vt u_int rcv_buf_cc
The current number of bytes in the socket receive buffer.
.It Vt u_int sent_inflight_bytes
The current number of unacknowledged bytes in-flight.
Bytes acknowledged via SACK are not excluded from this count.
.It Vt int t_segqlen
The current number of segments in the reassembly queue.
.It Vt u_int flowid
Flowid for the connection.
A caveat: Zero '0' either represents a valid flowid or a default value when
the flowid is not being set.
.It Vt u_int flowtype
Flow type for the connection.
Flowtype defines which protocol fields are hashed to produce the flowid.
A complete listing is available in
.Pa /usr/include/sys/mbuf.h
under
.Dv M_HASHTYPE_* .
.El
.Sh FILES
.Bl -tag -width "/usr/lib/dtrace/siftr.d" -compact
.It Pa /usr/lib/dtrace/tcp.d
DTrace type and translator definitions for all the probes of the
.Nm tcp
provider except the
.Nm siftr
probe.
.It Pa /usr/lib/dtrace/siftr.d
DTrace type and translator definitions for the
.Nm siftr
probe of the
.Nm tcp
provider.
.El
.Sh EXAMPLES
The following script logs TCP segments in real time:
.Bd -literal -offset indent
#pragma D option quiet
#pragma D option switchrate=10hz

dtrace:::BEGIN
{
        printf(" %3s %15s:%-5s      %15s:%-5s %6s  %s\\n", "CPU",
            "LADDR", "LPORT", "RADDR", "RPORT", "BYTES", "FLAGS");
}

tcp:::send
{
        this->length = args[2]->ip_plength - args[4]->tcp_offset;
        printf(" %3d %16s:%-5d -> %16s:%-5d %6d  (", cpu, args[2]->ip_saddr,
            args[4]->tcp_sport, args[2]->ip_daddr, args[4]->tcp_dport,
            this->length);
        printf("%s", args[4]->tcp_flags & TH_FIN ? "FIN|" : "");
        printf("%s", args[4]->tcp_flags & TH_SYN ? "SYN|" : "");
        printf("%s", args[4]->tcp_flags & TH_RST ? "RST|" : "");
        printf("%s", args[4]->tcp_flags & TH_PUSH ? "PUSH|" : "");
        printf("%s", args[4]->tcp_flags & TH_ACK ? "ACK|" : "");
        printf("%s", args[4]->tcp_flags & TH_URG ? "URG|" : "");
        printf("%s", args[4]->tcp_flags == 0 ? "null " : "");
        printf("\\b)\\n");
}

tcp:::receive
{
        this->length = args[2]->ip_plength - args[4]->tcp_offset;
        printf(" %3d %16s:%-5d <- %16s:%-5d %6d  (", cpu,
            args[2]->ip_daddr, args[4]->tcp_dport, args[2]->ip_saddr,
            args[4]->tcp_sport, this->length);
        printf("%s", args[4]->tcp_flags & TH_FIN ? "FIN|" : "");
        printf("%s", args[4]->tcp_flags & TH_SYN ? "SYN|" : "");
        printf("%s", args[4]->tcp_flags & TH_RST ? "RST|" : "");
        printf("%s", args[4]->tcp_flags & TH_PUSH ? "PUSH|" : "");
        printf("%s", args[4]->tcp_flags & TH_ACK ? "ACK|" : "");
        printf("%s", args[4]->tcp_flags & TH_URG ? "URG|" : "");
        printf("%s", args[4]->tcp_flags == 0 ? "null " : "");
        printf("\\b)\\n");
}
.Ed
The following script logs TCP connection state changes as they occur:
.Bd -literal -offset indent
#pragma D option quiet
#pragma D option switchrate=25hz

int last[int];

dtrace:::BEGIN
{
        printf("   %12s %-20s    %-20s %s\\n",
            "DELTA(us)", "OLD", "NEW", "TIMESTAMP");
}

tcp:::state-change
{
        this->elapsed = (timestamp - last[args[1]->cs_cid]) / 1000;
        printf("   %12d %-20s -> %-20s %d\\n", this->elapsed,
            tcp_state_string[args[5]->tcps_state],
            tcp_state_string[args[3]->tcps_state], timestamp);
        last[args[1]->cs_cid] = timestamp;
}

tcp:::state-change
/last[args[1]->cs_cid] == 0/
{
        printf("   %12s %-20s -> %-20s %d\\n", "-",
            tcp_state_string[args[5]->tcps_state],
            tcp_state_string[args[3]->tcps_state], timestamp);
        last[args[1]->cs_cid] = timestamp;
}
.Ed
The following script uses the siftr probe to show the current value of CWND
and SSTHRESH when a packet is sent or received:
.Bd -literal -offset indent
#pragma D option quiet
#pragma D option switchrate=10hz

dtrace:::BEGIN
{
        printf(" %3s %16s:%-5s %16s:%-5s %10s %10s\\n",
            "DIR", "LADDR", "LPORT", "RADDR", "RPORT", "CWND", "SSTHRESH");
}

tcp:::siftr
{
        printf(" %3s %16s:%-5d %16s:%-5d %10u %10u\\n",
            siftr_dir_string[args[0]->direction],
            args[0]->laddr, args[0]->lport, args[0]->raddr, args[0]->rport,
            args[0]->snd_cwnd, args[0]->snd_ssthresh);
}
.Ed
.Sh COMPATIBILITY
This provider is compatible with the
.Nm tcp
provider in Solaris.
.Sh SEE ALSO
.Xr dtrace 1 ,
.Xr dtrace_ip 4 ,
.Xr dtrace_sctp 4 ,
.Xr dtrace_udp 4 ,
.Xr dtrace_udplite 4 ,
.Xr siftr 4 ,
.Xr tcp 4 ,
.Xr SDT 9
.Sh HISTORY
The
.Nm tcp
provider first appeared in
.Fx
10.0.
.Sh AUTHORS
This manual page was written by
.An Mark Johnston Aq Mt markj@FreeBSD.org .
.Sh BUGS
The
.Vt tcps_local
and
.Vt tcps_active
fields of
.Vt tcpsinfo_t
are not filled in by the translator.