1TCP protocol
2============
3
4Last updated: 3 June 2017
5
6Contents
7========
8
9- Congestion control
10- How the new TCP output machine [nyi] works
11
12Congestion control
13==================
14
15The following variables are used in the tcp_sock for congestion control:
16snd_cwnd		The size of the congestion window
17snd_ssthresh		Slow start threshold. We are in slow start if
18			snd_cwnd is less than this.
19snd_cwnd_cnt		A counter used to slow down the rate of increase
20			once we exceed slow start threshold.
21snd_cwnd_clamp		This is the maximum size that snd_cwnd can grow to.
22snd_cwnd_stamp		Timestamp for when congestion window last validated.
23snd_cwnd_used		Used as a highwater mark for how much of the
24			congestion window is in use. It is used to adjust
25			snd_cwnd down when the link is limited by the
26			application rather than the network.
27
28As of 2.6.13, Linux supports pluggable congestion control algorithms.
29A congestion control mechanism can be registered through functions in
30tcp_cong.c. The functions used by the congestion control mechanism are
31registered via passing a tcp_congestion_ops struct to
32tcp_register_congestion_control. As a minimum, the congestion control
33mechanism must provide a valid name and must implement either ssthresh,
34cong_avoid and undo_cwnd hooks or the "omnipotent" cong_control hook.
35
36Private data for a congestion control mechanism is stored in tp->ca_priv.
37tcp_ca(tp) returns a pointer to this space.  This is preallocated space - it
38is important to check the size of your private data will fit this space, or
39alternatively, space could be allocated elsewhere and a pointer to it could
40be stored here.
41
42There are three kinds of congestion control algorithms currently: The
43simplest ones are derived from TCP reno (highspeed, scalable) and just
44provide an alternative congestion window calculation. More complex
45ones like BIC try to look at other events to provide better
46heuristics.  There are also round trip time based algorithms like
47Vegas and Westwood+.
48
49Good TCP congestion control is a complex problem because the algorithm
50needs to maintain fairness and performance. Please review current
51research and RFC's before developing new modules.
52
53The default congestion control mechanism is chosen based on the
54DEFAULT_TCP_CONG Kconfig parameter. If you really want a particular default
55value then you can set it using sysctl net.ipv4.tcp_congestion_control. The
56module will be autoloaded if needed and you will get the expected protocol. If
57you ask for an unknown congestion method, then the sysctl attempt will fail.
58
59If you remove a TCP congestion control module, then you will get the next
60available one. Since reno cannot be built as a module, and cannot be
61removed, it will always be available.
62
63How the new TCP output machine [nyi] works.
64===========================================
65
66Data is kept on a single queue. The skb->users flag tells us if the frame is
67one that has been queued already. To add a frame we throw it on the end. Ack
68walks down the list from the start.
69
70We keep a set of control flags
71
72
73	sk->tcp_pend_event
74
75		TCP_PEND_ACK			Ack needed
76		TCP_ACK_NOW			Needed now
77		TCP_WINDOW			Window update check
78		TCP_WINZERO			Zero probing
79
80
81	sk->transmit_queue		The transmission frame begin
82	sk->transmit_new		First new frame pointer
83	sk->transmit_end		Where to add frames
84
85	sk->tcp_last_tx_ack		Last ack seen
86	sk->tcp_dup_ack			Dup ack count for fast retransmit
87
88
89Frames are queued for output by tcp_write. We do our best to send the frames
90off immediately if possible, but otherwise queue and compute the body
91checksum in the copy.
92
93When a write is done we try to clear any pending events and piggy back them.
94If the window is full we queue full sized frames. On the first timeout in
95zero window we split this.
96
97On a timer we walk the retransmit list to send any retransmits, update the
98backoff timers etc. A change of route table stamp causes a change of header
99and recompute. We add any new tcp level headers and refinish the checksum
100before sending.
101
102