summaryrefslogtreecommitdiffstats
path: root/src/vnet/tcp/tcp_input.c
AgeCommit message (Collapse)AuthorFilesLines
2019-12-27tcp: validate port reuseFlorin Coras1-3/+8
Type: fix Make sure existing connection is in time-wait Change-Id: I8e8bef151f81bcd589b4da0d4bf63cc59f9f451b Signed-off-by: Florin Coras <fcoras@cisco.com>
2019-12-27tcp: Enable TCP timewait port useYu Ping1-1/+40
Improve host stack CPS test, and it help improve 10x performance Type: feature Change-Id: I6af61e0bad7c16ee2d30a1422cc46bb89f1cedb4 Signed-off-by: Yu Ping <ping.yu@intel.com> Signed-off-by: Yuwei Zhang <yuwei1.zhang@intel.com>
2019-12-23tcp: accept sack reneging as a cc eventFlorin Coras1-4/+4
Type: fix Change-Id: Iead1303ca3dec7593eb3ce54f291b82d94c821a4 Signed-off-by: Florin Coras <fcoras@cisco.com>
2019-11-20session tcp: support pacer idle timeoutsFlorin Coras1-40/+19
Type: feature To avoid excessive bursts, pacer must be provided with an estimated rtt for the connection. That's used to compute an idle timeout, i.e., time after which the bucket is reset to 1 mtu due to inactivity. For now, idle timeout is computed as 5% of the rtt. Change-Id: Ia0b752fe7b4ad0ce97b477fb886b0133a2321541 Signed-off-by: Florin Coras <fcoras@cisco.com>
2019-11-11tcp: improve lost rxt heuristicFlorin Coras1-32/+42
Type: feature - retransmit first unacked segment if newer retransmitted packets are acked - avoid spurious retransmits if recovery ends with sacked bytes Change-Id: Ic1b56d22e025822edb7609afb136e47440ea6032 Signed-off-by: Florin Coras <fcoras@cisco.com>
2019-11-08tcp: fix ip check in lookup validationFlorin Coras1-11/+17
Type: fix Change-Id: Ia18632c8fe22bdcfdf3cb48a4234f8703a7ac1d7 Signed-off-by: Florin Coras <fcoras@cisco.com>
2019-11-07tcp: fix retransmit with no sacksFlorin Coras1-2/+5
Type: fix Change-Id: I6f7df0d358f57f7feadb9b7a3fcffb99558b2af8 Signed-off-by: Florin Coras <fcoras@cisco.com>
2019-11-06tcp: validate the IP address while checking TCP connectionSrikanth Akula1-9/+31
Type: feature Along with the port information, we need to validate the IP address details as well. This is very useful in the case port re-use scenario Signed-off-by: Srikanth Akula <srakula@cisco.com> Change-Id: I11e1ebcd3e56aae47ac235a89606a83c928aa6bb
2019-11-05session tcp: refactor pacer timeFlorin Coras1-3/+1
Type: refactor Change-Id: Ic1c3e1f7987702cd88972acc34849dc1f585d5fe Signed-off-by: Florin Coras <fcoras@cisco.com>
2019-11-05tcp: correct validity check return valueRyujiro Shibuya1-0/+1
Type: fix Signed-off-by: Ryujiro Shibuya <ryujiro.shibuya@owmobility.com> Change-Id: Ib59bf664d8da20516d8f16d716c5e8698675da4b
2019-10-29tcp: correct tcp connection lookup conditionRyujiro Shibuya1-7/+10
Type: fix Signed-off-by: Ryujiro Shibuya <ryujiro.shibuya@owmobility.com> Change-Id: Ibdbac568d5f765e6b5c959cab5e32dc33335064d
2019-10-23tcp: fix sack retransmit beyond snd_nxtFlorin Coras1-0/+6
Type: fix Ensure that sack retransmit logic does not try to inadvertently send new data. Change-Id: Idfda19643577d9c1b58e2af8d8283cabfbaf98e6 Signed-off-by: Florin Coras <fcoras@cisco.com>
2019-10-17tcp: avoid sending acks when data availableFlorin Coras1-1/+4
Type: fix Change-Id: I02e2cb2d349f9ddb2bfed040a08ca448f379d565 Signed-off-by: Florin Coras <fcoras@cisco.com>
2019-10-17tcp: fix mss flag in option parsingFlorin Coras1-1/+1
Type: fix Change-Id: Ib03151a097caf15e4e8285afea83d1b930a874b0 Signed-off-by: Florin Coras <fcoras@cisco.com>
2019-10-16tcp: avoid head retransmit if scoreboard has no holesFlorin Coras1-2/+0
Type: fix Change-Id: I6164ad0d97cd91b60c85302eff691836d12a0a70 Signed-off-by: Florin Coras <fcoras@cisco.com>
2019-10-11tcp: refactor flags into cfg and conn flagsFlorin Coras1-7/+9
Type: refactor Change-Id: Ib5e2ecf55732599e1322736e3662178c9da49c7d Signed-off-by: Florin Coras <fcoras@cisco.com>
2019-10-10tcp: improve pacing after idle send periodsFlorin Coras1-22/+39
Rest pacer on ack reception if we haven't recently sent anything. Type: feature Change-Id: I820bacd81b65130052dfafbfcbe6ca4553069fbc Signed-off-by: Florin Coras <fcoras@cisco.com>
2019-10-10tcp: retry lost retransmitsFlorin Coras1-132/+118
Add heuristic that detects lost retransmitted segments and retries sending them. Type: feature Change-Id: I34d1bb16799e1993779222eb2bfad4b40704159e Signed-off-by: Florin Coras <fcoras@cisco.com>
2019-09-25tcp: use sacks for timer based recoveryFlorin Coras1-145/+141
Type: feature If available, reuse sack scoreboard in timer triggered retransmit to minimize spurious retransmits. Additional changes/refactoring: - limited transmit updates - add sacked rxt count to scoreboard - prr pacing of fast retransmits - startup pacing updates - changed loss window to flight + mss Change-Id: I057de6a9d6401698bd1031d5cf5cfbb62f2bdf61 Signed-off-by: Florin Coras <fcoras@cisco.com>
2019-09-20tcp: refactor congestion event handlingFlorin Coras1-40/+36
Type: refactor Minor cleanup to congestion event handling. Change-Id: I7179ff9e21c1853382ba814652c9024f3e5b8a37 Signed-off-by: Florin Coras <fcoras@cisco.com>
2019-09-16tcp: use rate sample rtt in recovery if possibleFlorin Coras1-6/+14
If in recovery and rate samples are taken for burts that have not been retransmitted, use the rtt estimate. Type: feature Change-Id: I95028f492008457c959157aa4ee4c3435fa9c3f0 Signed-off-by: Florin Coras <fcoras@cisco.com>
2019-09-12tcp: cleanup scoreboard after recoveryFlorin Coras1-0/+7
Type: fix When recovering from congestion, remove last hole in scoreboard if it covers all bytes sent above snd_congestion. Change-Id: I752accd4590cd9bf553c35a1bb40a38c308c04f5 Signed-off-by: Florin Coras <fcoras@cisco.com>
2019-09-11tcp: handle sack renegingFlorin Coras1-88/+120
Type: feature Change-Id: I201155b1b92cf7e57310af726879dab039090582 Signed-off-by: Florin Coras <fcoras@cisco.com>
2019-09-10tcp: enable gso in tcp hoststackSimon Zhang1-0/+37
Type: feature Change-Id: If68d07fbe8c6f7fffd2f93c7e854367082927e4f Signed-off-by: Simon Zhang <yuwei1.zhang@intel.com>
2019-09-04tcp: cc algos handle cwnd on congestion signalFlorin Coras1-12/+1
Type: refactor Change-Id: I15b10a22d0d0b83075a0eef5ef8c09cf76989866 Signed-off-by: Florin Coras <fcoras@cisco.com>
2019-08-30tcp: track zero rwnd errorsFlorin Coras1-0/+7
Type: feature Distinguish between rcv window errors and errors after we advertised a zero rcv window, i.e., potential window probes. Change-Id: I6cb453c7aaae456c0a05a8328cfaa55eaca10bf7 Signed-off-by: Florin Coras <fcoras@cisco.com>
2019-08-21tcp: add option for always on event loggingFlorin Coras1-33/+31
Type: feature Add option to always compile in debugging infra. Debug verbosity can be controlled via configuration updatable via cli. Compile time inclusion and configuration of event logging verbosity is still available in order to minimize event logging impact. Change-Id: I9d946efe9fa4204fdace1adb1d6588c97b5ae758 Signed-off-by: Florin Coras <fcoras@cisco.com>
2019-08-19tcp: set cc_algo on connection allocFlorin Coras1-5/+2
Type: feature Change-Id: Ie00329b11e26cab5db64b9c006171d283854c8ef Signed-off-by: Florin Coras <fcoras@cisco.com>
2019-08-14tcp: extend protocol configurationFlorin Coras1-15/+18
Type: feature Expose more configuration parameters and refactor some of the existing ones. Change-Id: If44c31ff77ce3d7e8da67d39a4ff61346bdf5ccc Signed-off-by: Florin Coras <fcoras@cisco.com>
2019-08-12tcp: handle fin+rst+syn in closing stateFlorin Coras1-0/+2
Type: fix Change-Id: I32d4d4a5de04e89087f4ab4a5c425eda572932a8 Signed-off-by: Florin Coras <fcoras@cisco.com>
2019-08-08tcp: cleanup timersFlorin Coras1-3/+0
Type:refactor Change-Id: I37dbc8b55827d66c2578d6ab8b86ed7e18198aa6 Signed-off-by: Florin Coras <fcoras@cisco.com>
2019-08-02tcp: add more connection statsFlorin Coras1-0/+14
Type:feature Change-Id: If02884d0f1f26bfe31ec609ea9611cb27b699868 Signed-off-by: Florin Coras <fcoras@cisco.com>
2019-07-26session tcp: handle rxt and acks as custom eventsFlorin Coras1-92/+17
Type: feature Control ack generation and retransmissions with session layer scheduler. Change-Id: Iacdf9f84ab81f44851980aa45a83e75f29be2b7b Signed-off-by: Florin Coras <fcoras@cisco.com>
2019-07-23tcp: generate closed notifications whenever tcp closesFlorin Coras1-0/+2
Type: fix This gives the session layer a chance to to cleanup events. Change-Id: I7499e94acf06340e2b19a1d91a4c61a63cd66c52 Signed-off-by: Florin Coras <fcoras@cisco.com>
2019-07-17tcp: add node with no 6-tuple lookupVladimir Kropylev1-22/+96
Type: feature Add new node in TCP stack where TCP 6 tuple lookup is not required. In new node, packet metadata contains connection-index which can be used to retrieve the TCP connection. The new node will be used by proxy. Change-Id: I3aa0268946898912f4176d5c8c5903e06657479d Signed-off-by: Vladimir Kropylev <vladimir.kropylev@enea.com>
2019-07-15session: allow transports to generate closed notificationsFlorin Coras1-0/+2
In contrast to the closing notification, whereby a transport informs the session layer that is beginning the closing procedure, this allows transports to notify the session layer of the fact that the transport is "fully" closed, i.e., it expects no more data. Also: - adds app closed state for sessions - changes tcp to have it notify when an active close has finished Type: feature Change-Id: I13c738006c03f85015e05ab82843a33a69382aaf Signed-off-by: Florin Coras <fcoras@cisco.com>
2019-07-09tcp: track last lost bytes in sack sbFlorin Coras1-3/+9
Type:feature Change-Id: I687809ebcc759cec8cb1d5c3b2b7e6bc995a7985 Signed-off-by: Florin Coras <fcoras@cisco.com>
2019-07-05tcp: timestamp adjustmentVladimir Kropylev1-1/+1
Type: feature Change-Id: Icb3c574100cde95ab5be4923c8739889cf7e48c6 Signed-off-by: Vladimir Kropylev <vladimir.kropylev@enea.com>
2019-07-05tcp: add cc algo undo recovery notificationFlorin Coras1-18/+9
Type:feature Change-Id: Iedefe87555f0a0033abed8569bc2995c8f523d7e Signed-off-by: Florin Coras <fcoras@cisco.com>
2019-07-05tcp: add loss signal to cc algoFlorin Coras1-5/+16
Type:feature Change-Id: Ibe1a4c555b55fb929d55b02599aaf099ed522cdf Signed-off-by: Florin Coras <fcoras@cisco.com>
2019-06-28tcp: reject out-of-order finsFlorin Coras1-0/+4
Type:fix Change-Id: Iab2c308739f7733dbf70953e0ea87dcc404c60da Signed-off-by: Florin Coras <fcoras@cisco.com>
2019-06-28tcp: fix close-waiting check for outstanding tx dataFlorin Coras1-1/+2
Type: fix Change-Id: I9e0e1bb933c38b17ac9fb8bf11b81e2fed021bf8 Signed-off-by: Florin Coras <fcoras@cisco.com>
2019-06-25tcp: delivery rate estimatorFlorin Coras1-12/+17
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>
2019-06-17session: use listener_handle instead of listener_indexNathan Skrzypczak1-1/+1
Type: refactor Change-Id: I97fa59a0ba0b6b7a98698926020ffffcf6ae6ba3 Signed-off-by: Nathan Skrzypczak <nathan.skrzypczak@gmail.com>
2019-06-13tcp: fix tcp_do_fastretransmits failure.Guoao Sun1-0/+2
Type: fix vpp would fail 'tcp_in_fastrecovery (tc)' if tcp_connection_get returns 0. Change-Id: I512cba5234b3777f1737bec9451940e7e7975f15 Signed-off-by: Guoao Sun <guoao.sun@intel.com>
2019-06-12tcp: better estimate snd space for rxt with no sacksFlorin Coras1-2/+0
Type: refactor Use tcp_available_cc_snd_space instead of sack scoreboard last_delivered_bytes to estimate available space when retransmitting with no sack support Change-Id: I938c637279eaf8c5634c4e97f0633fa2d0054fac Signed-off-by: Florin Coras <fcoras@cisco.com>
2019-06-08tcp: send challenge ack for in wnd synFlorin Coras1-1/+2
Type: fix Per rfc793, in window syns for established connections should lead to connection resets. As a mitigation for blind reset attacks, rfc5961 requests that such syns be replied to with challange acks. Change-Id: I75e4972bbb515e48d9cf1bda32ea5d9891d670f0 Signed-off-by: Florin Coras <fcoras@cisco.com>
2019-06-07tcp: log connect notify errorsFlorin Coras1-1/+2
Type: fix Change-Id: Idf0b2e16b2e7d126940bb38c7983d6784b5bfdc8 Signed-off-by: Florin Coras <fcoras@cisco.com>
2019-05-27tcp: handle disconnects in rcv-processFlorin Coras1-0/+1
Change-Id: I005f96480e81f3e750c18261e78d0e401da7528e Signed-off-by: Florin Coras <fcoras@cisco.com>
2019-04-16svm_fifo rework to avoid contention on cursizeSirshak Das1-1/+1
Problems Addressed: - Contention of cursize by producer and consumer. - Reduce the no of modulo operations. Changes: - Synchronization between producer and consumer changed from cursize to head and tail indexes Implications: reduces the usable size of fifo by 1. - Using weaker memory ordering C++11 atomics to access head and tail based on producer and consumer role. - Head and tail indexes are unsigned 32 bit integers. Additions and subtraction on them are implicit 32 bit Modulo operation. - Adding weaker memory ordering variants of max_enq, max_deq, is_empty and is_full Using them appropriately in all places. Perfomance improvement (iperf3 via Hoststack): iperf3 Server: Marvell ThunderX2(AArch64) - iperf3 Client: Skylake(x86) ~6%(256 rxd/txd) - ~11%(2048 rxd/txd) Change-Id: I1d484e000e437430fdd5a819657d1c6b62443018 Signed-off-by: Sirshak Das <sirshak.das@arm.com> Reviewed-by: Honnappa Nagarahalli <honnappa.nagarahalli@arm.com>