summaryrefslogtreecommitdiffstats
path: root/src/vnet/tcp/tcp.h
AgeCommit message (Collapse)AuthorFilesLines
2019-06-25tcp: window update ackVladimir Kropylev1-0/+6
Type: feature Provide interface for sending Window Update ACK, ensuring it will be sent only once, if RWND became non-zero, after zero RWND has been advertised before. Change-Id: I7f0d8af76d7763208625df68ab4ac3727fdaf449 Signed-off-by: Vladimir Kropylev <vladimir.kropylev@enea.com>
2019-06-25tcp: delivery rate estimatorFlorin Coras1-12/+127
Type: feature First cut implementation with limited testing. The feature is not enabled by default and the expectation is that cc algorithms will enable it on demand. Change-Id: I92b70cb4dabcff0e9ccd1d725952c4880af394da Signed-off-by: Florin Coras <fcoras@cisco.com>
2019-06-21tcp: add option to pass opaque to next nodeFlorin Coras1-1/+2
Type:feature Change-Id: I0b72954a6ae6a05abe0761cb4f227072863f127b Signed-off-by: Florin Coras <fcoras@cisco.com>
2019-06-18tcp: allow custom output next nodeFlorin Coras1-0/+1
Type: feature Change-Id: If9804d3685d2a52efa06e412ae1ebb39c6a44b8b Signed-off-by: Florin Coras <fcoras@cisco.com>
2019-06-11tcp: allow registration of new cc algo typesFlorin Coras1-4/+14
Type: feature Change-Id: I83e21b508a19df1beb207d961c8f2b52347deca0 Signed-off-by: Florin Coras <fcoras@cisco.com>
2019-06-08tcp: close connection if not recovered after max triesFlorin Coras1-0/+1
Type: feature Change-Id: Ia8af6a62a2be2265bc42955d90e8c2222bdb8f50 Signed-off-by: Florin Coras <fcoras@cisco.com>
2019-06-07tcp: add cc cleanup functionFlorin Coras1-0/+1
Type: feature Change-Id: I54521078cf96e459d041c86297c6ca80045bf0a3 Signed-off-by: Florin Coras <fcoras@cisco.com>
2019-04-02tcp: improve rcv process ack processingFlorin Coras1-0/+1
- Avoid doing cc in closing states. - Rest connections closed with unread data Change-Id: I97d46b0459f03ea5439eeb0f233b6c17d3e06dfd Signed-off-by: Florin Coras <fcoras@cisco.com>
2019-03-29tcp: remove sent rcv wnd 0 flagFlorin Coras1-1/+0
Change-Id: If6c672d1caa8884eb5d819311606a79a3de81200 Signed-off-by: Florin Coras <fcoras@cisco.com>
2019-03-28tcp: remove unused buffer flagsFlorin Coras1-20/+0
Change-Id: I86e71f32dee3cc48f680e6432a96caef0a7f66a1 Signed-off-by: Florin Coras <fcoras@cisco.com>
2019-03-27tcp: add cc algo lookup tableFlorin Coras1-0/+3
Change-Id: Ie7be0136c182cdc35193e47dd3249153c2f8d65e Signed-off-by: Florin Coras <fcoras@cisco.com>
2019-03-23tcp: make default mtu configurableFlorin Coras1-0/+3
Change-Id: I56d8d8d67d5590e24c1ddb54b0c63a2cb03798e1 Signed-off-by: Florin Coras <fcoras@cisco.com>
2019-03-22tcp: improve handling of snd_nxtFlorin Coras1-2/+2
- avoid changing snd_nxt when doing fast retransmits - use snd_una_max only to keep track of the max seq number sent - simplify future ack testing Change-Id: I3580ad3aefe30128486c3375d0ac3f3f62c04c5e Signed-off-by: Florin Coras <fcoras@cisco.com>
2019-03-08session/tcp/vcl: fixes and optimizationsFlorin Coras1-1/+2
Change-Id: Idc7dfe743399dd8dee0f6b3ec83f194f3fca580b Signed-off-by: Florin Coras <fcoras@cisco.com>
2019-03-06tcp: migrate old MULTIARCH macros to VLIB_NODE_FNFilip Tehlar1-0/+8
Change-Id: Ifd9fa30eed343e2c5d40582b3e3aa589b070637d Signed-off-by: Filip Tehlar <ftehlar@cisco.com>
2019-03-02session: cleanup/rename functionsFlorin Coras1-1/+1
- remove unused functions - rename old "stream" functions Change-Id: Icbb03daa9f9f1c58b5be5c38aa8a9cbcf9159b47 Signed-off-by: Florin Coras <fcoras@cisco.com>
2019-02-22tcp: keep snd sack block free listFlorin Coras1-0/+1
Instead of constantly reallocating the new sack block list, keep the old one as a reusable free list. Change-Id: Iad79a72204f97b96352c1c6eea66c2839a35cfe6 Signed-off-by: Florin Coras <fcoras@cisco.com>
2019-02-22tcp: send enough dupacks to cover all sack holesFlorin Coras1-0/+1
Make sure we send enough dupacks to cover all the holes created in the last frame received. Also make sure we send all the blocks, not just the first. Change-Id: I9597a34ac14473d1cc3ad07d65bc37043e3d0582 Signed-off-by: Florin Coras <fcoras@cisco.com>
2019-02-18tcp: harden for high scale scenariosFlorin Coras1-1/+1
- Better handle buffer starvation scenarios - Handle case when both peers enter recovery due to packet loss. - Fix passive open establish cleanup Change-Id: I2f28baa2ff0383bb8f5f6d2452b49aa38ce69bce Signed-off-by: Florin Coras <fcoras@cisco.com>
2019-02-07tcp: fix reset sending in syn-sentFlorin Coras1-1/+1
Change-Id: I468b1015bca5f95b162fb49e904aa4e4d38db75e Signed-off-by: Florin Coras <fcoras@cisco.com>
2019-02-02tcp: remove internal buffer cacheFlorin Coras1-1/+0
Change-Id: I3051d49aa9b9c38dcd8069ea1748f74ab20ac0f6 Signed-off-by: Florin Coras <fcoras@cisco.com>
2019-01-03tcp: better randomize issFlorin Coras1-0/+9
Change-Id: I3638221e59024d6b7d82499d57e25b8e609f73cb Signed-off-by: Florin Coras <fcoras@cisco.com>
2018-12-26tcp: separate active and passive establish timersFlorin Coras1-1/+2
Change-Id: Ia2241e963cf45765d8d17c65eea781edbf74d4f9 Signed-off-by: Florin Coras <fcoras@cisco.com>
2018-12-22tcp: improve waitclose in closing statesFlorin Coras1-0/+7
Change-Id: I90056176194cb2a144d49a3cb283653d8d30f051 Signed-off-by: Florin Coras <fcoras@cisco.com>
2018-12-19tcp/session: notify transport of close when tx fifo is not emptyFlorin Coras1-2/+3
Disconnect transport even if tx fifo is not empty and have transport deal with the problem. In case of tcp, add timer to fin_wait_1. If it expires and we're still in established state, cleanup but only after waiting for session tx events to cleanup. Change-Id: I45759a3c43dd096bb2c03daf5372416c30678d62 Signed-off-by: Florin Coras <fcoras@cisco.com>
2018-12-17tcp: fix handling of broken syn optionsFlorin Coras1-1/+2
Change-Id: Ia8b2a077ba4897ddd15cf33221b191cd7a3f1d33 Signed-off-by: Florin Coras <fcoras@cisco.com>
2018-12-13session/tcp: support tx flush markFlorin Coras1-0/+3
For tcp this means that the last enqueued data goes out with a psh bit set. Change-Id: I29d357ecae6f02e748b59a7b799150ec73d14ba2 Signed-off-by: Florin Coras <fcoras@cisco.com>
2018-11-13tcp: cubic fast convergenceFlorin Coras1-2/+8
Change-Id: I3a15960fe346763faf13e8728ce36c2f3bf7b05a Signed-off-by: Florin Coras <fcoras@cisco.com>
2018-11-12tcp: handle disconnects after enq notificationsFlorin Coras1-21/+40
Make sure that we notify the app of the data enqueued in the burst before notifying of disconnect. Change-Id: I7747a5cbb4c6bc9132007f849c24ce04b7841273 Signed-off-by: Florin Coras <fcoras@cisco.com>
2018-11-09tcp: basic cubic implementationFlorin Coras1-0/+34
Because the code is not optimized, newreno is still the default congestion control algorithm. Change-Id: I7061cc80c5a75fa8e8265901fae4ea2888e35173 Signed-off-by: Florin Coras <fcoras@cisco.com>
2018-11-08tcp: pacer and mrtt estimation improvementsFlorin Coras1-1/+0
- update pacer once per burst - better estimate initial rtt - compute smoothed average for higher precision rtt estimate Change-Id: I06d41a98784cdf861bedfbee2e7d0afc0d0154ef Signed-off-by: Florin Coras <fcoras@cisco.com>
2018-11-07tcp: consume incoming buffers instead of reusingFlorin Coras1-9/+7
Instead of reusing buffers for acking, consume all buffers and program output for (dup)ack generation. This implicitly fixes the drop counters that were artificially inflated by both data and feedback traffic. Moreover, the patch also significantly reduces the ack traffic as we now only generate an ack per frame, unless duplicate acks need to be sent. Because of the reduced feedback traffic, a sender's rx path and a receiver's tx path are now significantly less loaded. In particular, a sender can overwhelm a 40Gbps NIC and generate tx drop bursts for low rtts. Consequently, tx pacing is now enforced by default. Change-Id: I619c29a8945bf26c093f8f9e197e3c6d5d43868e Signed-off-by: Florin Coras <fcoras@cisco.com>
2018-11-06tcp: dequeue acked only once per burstFlorin Coras1-0/+3
Avoid dequeuing acked bytes more than once per burst for a connection. Although the fifos do not use locks, size decrements are atomic, so they rely on locked instructions. Change-Id: Id65f4ea40b2c10057461402dfd0393034e6472d5 Signed-off-by: Florin Coras <fcoras@cisco.com>
2018-11-02session: measure dispatch period only if under loadFlorin Coras1-2/+4
Also reset pacer on tcp retransmit timeout Change-Id: I5a9edee4c00d1d169248d79587a9b10437c2bd87 Signed-off-by: Florin Coras <fcoras@cisco.com>
2018-11-02tcp: minimize use of tlsFlorin Coras1-12/+28
Also propagate tcp worker context instead of retrieving it multiple times. Change-Id: I7b273b981826b37783566d0172a64cd6957f3b33 Signed-off-by: Florin Coras <fcoras@cisco.com>
2018-11-01tcp: fast retransmit pacingFlorin Coras1-2/+5
Force pacing for fast retransmit to avoid bursts of retransmitted packets. Change-Id: I2ff42c328899b36322c4de557b1f7d853dba8fe2 Signed-off-by: Florin Coras <fcoras@cisco.com>
2018-10-25tcp/session: add tx pacerFlorin Coras1-2/+21
Adds tx pacing infrastructure for transport protocols that want to use it. Particularly useful for connections with non-negligible rtt and constrained network throughput as it avoids large tx bursts that lead to local interface tx or network drops. By default the pacer is disabled. To enabled it for tcp, add tx-pacing to tcp's startup conf. We are still slightly inefficient in the handling of incoming packets in established state so the pacer slightly affect maximum throughput in low lacency scenarios. Change-Id: Id445b2ffcd64cce015f75b773f7d722faa0f7ca9 Signed-off-by: Florin Coras <fcoras@cisco.com>
2018-10-23tcp: fast retransmit improvementsFlorin Coras1-4/+18
Patch is too large to be ported to 18.10 just days before release. - handle fast retransmits outside of established node and limit the retransmit burst size to avoid tx losses and worsening congestion. - in the absance of a tx pacer, use slow start after fast retransmit exists - add fast retransmit heuristic that re-retries sending the first segment if everything else fails - fine tuning Change-Id: I84a2ab8fbba8b97f1d2b26584dc11a1e2c33c8d2 Signed-off-by: Florin Coras <fcoras@cisco.com>
2018-09-27tcp: use scaled window for new connectsFlorin Coras1-0/+1
Change-Id: Idf83fce8ca176e57b323e3741034e3223f1d195a Signed-off-by: Florin Coras <fcoras@cisco.com>
2018-09-25tcp: add option to cfg max rx fifo sizeFlorin Coras1-4/+12
Change-Id: Icff3d688506e7658330db004c58bcfcac273fcec Signed-off-by: Florin Coras <fcoras@cisco.com>
2018-06-26tcp/session: tx optimizationsFlorin Coras1-1/+4
- cache and reuse tcp options and rcv_wnd for session layer tx bursts - avoid reading/setting total_length_not_including_first_buffer. It's part of a buffer's second cache line so it comes at a "cost". Change-Id: Id18219c2f7e07cf4c63ee74f9cdd9e5918904036 Signed-off-by: Florin Coras <fcoras@cisco.com>
2018-06-26tcp: cleanup functionsFlorin Coras1-139/+29
- sprinkle statics for functions - move some inlines from header files to corresponding .c files - replace some always_inlines with statics where inlining is not performance critical Change-Id: I371dbf63431ce7e27e4ebbbdd844a9546a1f1849 Signed-off-by: Florin Coras <fcoras@cisco.com>
2018-06-25tw: add light weight timer update functionFlorin Coras1-7/+8
Because it avoids pool putting/getting the timer, this function is somewhat faster than stopping and restarting a timer. Change-Id: Id99ed9d356b0b1f7e12facfe8da193e1cd30b3ec Signed-off-by: Florin Coras <fcoras@cisco.com>
2018-06-20tcp: add per worker ctx structureFlorin Coras1-19/+27
Change-Id: I28d3c31bdc4255a4ca223d80bcf44709fb39f4ed Signed-off-by: Florin Coras <fcoras@cisco.com>
2018-06-19tcp: optimize tcp outputFlorin Coras1-1/+2
Change-Id: Idf17a0633a1618b12c22b1119e40c2e9d3192df9 Signed-off-by: Florin Coras <fcoras@cisco.com>
2018-06-11tcp: cleanup connection/session fixesFlorin Coras1-1/+1
- Cleanup session state after last ack and avoid using a cleanup timer. - Change session cleanup to free the session as opposed to waiting for delete notify. - When in close-wait, postpone sending the fin on close until all outstanding data has been sent. - Don't flush rx fifo unless in closed state Change-Id: Ic2a4f0d5568b65c83f4b55b6c469a7b24b947f39 Signed-off-by: Florin Coras <fcoras@cisco.com>
2018-05-26tcp: loss recovery improvements/fixesFlorin Coras1-3/+3
- fix newreno cwnd computation - reset snd_una_max on entering recovery - accept acks beyond snd_nxt but less than snd_congestion when in recovery - avoid entering fast recovery multiple times when using sacks - avoid as much as possible sending small segments when doing fast retransmit - more event logging Change-Id: I19dd151d7704e39d4eae06de3a26f5e124875366 Signed-off-by: Florin Coras <fcoras@cisco.com>
2018-05-17tcp: handle link-local addressesFlorin Coras1-0/+1
Change-Id: I9ede6bc861350c7d9e78fa4d96cd584c2816d06f Signed-off-by: Florin Coras <fcoras@cisco.com>
2018-04-20tcp: make newreno byte instead of acks dependentFlorin Coras1-0/+1
Should be more resilient to ack losses Change-Id: Icec3b93c1d290dec437fcc4e6fe5171906c9ba8a Signed-off-by: Florin Coras <fcoras@cisco.com>
2017-12-11session: generalize handling of network transportsFlorin Coras1-9/+0
- compute session type out of transport and network protos - make session, session lookup and session queue code network protocol agnostic This does not update the session layer to support non-ip network layer protocols Change-Id: Ifc2f92845e158b649d59462eb7d51c12af536691 Signed-off-by: Florin Coras <fcoras@cisco.com>