aboutsummaryrefslogtreecommitdiffstats
path: root/src/plugins/quic/quic.h
AgeCommit message (Collapse)AuthorFilesLines
2021-08-03quic: do not update crypto keys from workersAloys Augustin1-2/+1
The vnet_crypto_key_add should only be called from the main thread. This patch works around this limitation by allocating one key per worker and updating it on the fly everytime we need to do a crypto operation. This solution is far from ideal, but quicly has a strong assumption that it can use a key immediately after determining it, so making the key creation asynchronous is not a possibility. Signed-off-by: Aloys Augustin <aloaugus@cisco.com> Change-Id: I19fc5814195156003c36a73bb616738ba9d828f7 Type: fix
2021-01-22quic: quicly v0.1.2 updateMathias Raoul1-23/+9
- update quic plugin with new quicly/picotls API - remove packet allocator - remove crypto batching - update picotls plugin - add cli for quicly congestion control configuration Type: feature Change-Id: If76ef31c43b430eea2f7674539b2112aee0f351e Signed-off-by: Mathias Raoul <mathias.raoul@gmail.com>
2020-04-10quic: fix clean stream closeAloys Augustin1-0/+1
Properly close the stream sending side instead of resetting the entire stream. Change-Id: I2daca7e507daa7bf9cd6a27fa8828a9ea7d175f9 Type: fix Signed-off-by: Aloys Augustin <aloaugus@cisco.com>
2020-03-24quic: disable vnet_crypto and batching if no crypto engines are loadedMathiasRaoul1-0/+2
Type: fix Change-Id: I95d3f8431b468cefc8777526dd3b988a299f0687 Signed-off-by: MathiasRaoul <mathias.raoul@gmail.com>
2020-01-31quic: quicly crypto offloadingMathiasRaoul1-0/+41
- Implement our own quic packet allocator to allocate more memory at the end of the packet to store crypto offloading related data - 1RTT packets offloading encryption/decryption using vnet crypto - Add cli to change max packet per key Type: feature Change-Id: I7557fd457d7ba492329d5d8ed192509cbd727f9c Signed-off-by: MathiasRaoul <mathias.raoul@gmail.com>
2019-12-17quic: Implement crypto contextsNathan Skrzypczak1-8/+12
Type: feature Make quic use the crypto contexts. This introduces a crypto context pool backed by a hashtable giving ctx indexes by connect params (ckpair, engine, rx & tx fifo sizes). Applications keep the initialization vector common. Change-Id: I22ed6711196cd70a2f2f74240f12113c7af8dfcd Signed-off-by: Nathan Skrzypczak <nathan.skrzypczak@gmail.com>
2019-12-11quic: Use per thread next_cidNathan Skrzypczak1-1/+1
Type: fix Change-Id: Ia89427257b78f6f7ba62c9d1076bf0593ee2dca7 Signed-off-by: Nathan Skrzypczak <nathan.skrzypczak@gmail.com>
2019-12-04quic: fix stream tx_fifo race conditionNathan Skrzypczak1-0/+13
Type: fix There is a race condition in when receiving TX from a client application : As egress_emit writes as much data as possible to the stream, if during egress_emit the app writes to the fifo, the data will be directly passed to quicly. Then TX callback happens and triggers a scheduler update telling quilcy the stream has data to send. When the next egress_emit is called and no more data has come, we have nothing to write, we return len = 0 to quicly which breaks an assert if a loss happens later on. Change-Id: I47e00a14dfc9068b5dac7b5c090a89124aea004f Signed-off-by: Nathan Skrzypczak <nathan.skrzypczak@gmail.com>
2019-12-04quic: add conn-timeout config optionNathan Skrzypczak1-0/+2
Type: feature Change-Id: Ia9cb57b9f7c2d14a137556d6918763f1cb11850d Signed-off-by: Nathan Skrzypczak <nathan.skrzypczak@gmail.com>
2019-11-26quic: Refactor for crypto contextsNathan Skrzypczak1-11/+6
Type: refactor Change-Id: I5ec7079d34826edd7a3048ae1d44037386f5d3ff Signed-off-by: Nathan Skrzypczak <nathan.skrzypczak@gmail.com>
2019-11-22quic: Refactor quic rx callback for offloadNathan Skrzypczak1-0/+20
Type: refactor Change-Id: I1d90c8807e7eea4a77403bafc0628ec0d393c572 Signed-off-by: Nathan Skrzypczak <nathan.skrzypczak@gmail.com>
2019-11-22quic: Remove qctx opening poolNathan Skrzypczak1-1/+0
Type: fix We don't need the pool anymore as we store the ctx index in the underlying UDP session opaque Change-Id: Iec9be9c84a1f929d94b084176a143d304b0acb83 Signed-off-by: Nathan Skrzypczak <nathan.skrzypczak@gmail.com>
2019-11-07quic: Add support for ckpair & crypto engineNathan Skrzypczak1-1/+3
quic choice used ckpair is now the one passed to connect or listen via mq. The crypto engine is chosen with the value passed to connect or listen via mq: * If NONE(0) is provided, we default to quic_main. default_crypto_engine (picotls at init, can be changed via debug cli : quic set crypto api [crypto engine]) * If PICTOLS/VPP is provided, use this one * Other values return an error Type: feature Change-Id: Ifab893d6d03c83f202e6c7e7a9936f546a4b1530 Signed-off-by: Nathan Skrzypczak <nathan.skrzypczak@gmail.com>
2019-11-04quic: enhance error reportingDave Wallace1-0/+5
Type: refactor Signed-off-by: Dave Wallace <dwallacelf@gmail.com> Change-Id: I710d00e4a6c4356d0f00f7592bf14b55596ec6ae
2019-11-01quic: fifo size is u32Dave Wallace1-2/+2
- Fix cli / config fifo size to only accept u32 size input. - Make cli / config fifo-size input type handling to be the same as vpp hoststack - Update external transfer tests to use new syntax with different fifo sizes for vpp_echo client/server and vpp. Type: fix Change-Id: Ia5ddb2b8d3d9908ab502352819eebeec8ac0971d Signed-off-by: Dave Wallace <dwallacelf@gmail.com>
2019-10-09session: Add certificate storeNathan Skrzypczak1-1/+2
Type: feature This changes the behavior of both API calls APPLICATION_TLS_CERT_ADD & APPLICATION_TLS_KEY_ADD certificates and keys aren't bound to an app, they are passed to it via connect / listen using the message queue. This should be followed by a per protocol (QUIC/TLS) crypto_context store to save devrived structs Change-Id: I36873bc8b63b5c72776c69e8cd9febc9cae31882 Signed-off-by: Nathan Skrzypczak <nathan.skrzypczak@gmail.com>
2019-10-08hsa: make vpp_echo use mq instead of bapiNathan Skrzypczak1-6/+0
Type: refactor Change-Id: I4f370b09e22dbbc8920272df9a042dae04825bfc Signed-off-by: Nathan Skrzypczak <nathan.skrzypczak@gmail.com>
2019-10-08session: add is_migrating flagNathan Skrzypczak1-0/+1
Type: fix Add SESSION_F_IS_MIGRATING flag for session. It is set by the session layer before poking the transport for migration. It's the transport responsibility to unset the flag & act apropriatly if RX happens on a migrating session. Change-Id: Ie722917f1cf9344d8f041cad4ed8b064fb5853b6 Signed-off-by: Nathan Skrzypczak <nathan.skrzypczak@gmail.com>
2019-10-07quic: Refactor naming & clibsNathan Skrzypczak1-0/+15
Type: refactor Change-Id: I067fc7dbcc0985277df01384d9dcdb0bdf62fba5 Signed-off-by: Nathan Skrzypczak <nathan.skrzypczak@gmail.com>
2019-10-04quic: make quic fifo size configurable via cliMathiasRaoul1-1/+4
Type: feature Change-Id: I6d6ac649094ef4bee2aed311915dd58f11972e79 Signed-off-by: MathiasRaoul <mathias.raoul@gmail.com>
2019-10-01quic: Add Tx, Rx and packet drop countersMathiasRaoul1-0/+10
Type: feature Change-Id: I25aeeed49fc569315296a73c5595c2e2e302434f Signed-off-by: MathiasRaoul <mathias.raoul@gmail.com>
2019-09-06quic: fix server opening stream immediatelyAloys Augustin1-1/+1
Previously, if the first thing a server did was to open a stream and send data, this could trigger a crash on the clilent side VPP as the quic session wouldn't be allocated. Change-Id: I43990ce2a71217d6719ecae4da60111d05fbcfc0 Type: fix Signed-off-by: Aloys Augustin <aloaugus@cisco.com>
2019-08-01quic: Improve quicly_ctx handling & crypto ctxNathan Skrzypczak1-0/+17
Type: fix Change-Id: I898de67c017c3a45bed123d81041b32b43f749d0 Signed-off-by: Nathan Skrzypczak <nathan.skrzypczak@gmail.com>
2019-07-30quic: cleanup and refactorisationAloys Augustin1-19/+23
Change-Id: I031a60ac010b55110f32f0a08e19b1156aeda268 Type: refactor Signed-off-by: Aloys Augustin <aloaugus@cisco.com>
2019-07-23quic: Refactor connections closing and deletionAloys Augustin1-0/+3
This code should handle the 3 following cases: - Active close quic_proto_on_close sets state to ACTIVE_CLOSING send packets eventually returns an error, calling quic_connection_closed which deletes the connection - Passive close quic_on_closed_by_peer -> set state to PASSIVE_CLOSING "race" between app confirmation (calling quic_proto_on_close) and quicly signalling that it's done (triggers call to quic_connection_closed). If quic_connection_closed is called first, it sets the state to PASSIVE CLOSING QUIC CLOSED, then when quic_proto_on_close is called it frees the connection. If quic_proto_on_close is called first, it sets the state to PASSIVE CLOSING APP CLOSED, then when quic_connection_closed is called it frees the connection - Error close (reset) quic_connection_closed is called in state READY. This means a timeout or protocol error happened. This calls session_transport_reset_notify, the app should confirm the deletion and quic_proto_on_close will be called to delete the connection. Change-Id: I3acbf9b079ed2439bdbb447197c428c78915d8c0 Signed-off-by: Aloys Augustin <aloaugus@cisco.com> Type: feature
2019-07-09quic: fix passive disconnect handling & refactorNathan Skrzypczak1-24/+51
Type: refactor Change-Id: Ie54a77252e9f58a90f9e1f9595b9ede354952f70 Signed-off-by: Nathan Skrzypczak <nathan.skrzypczak@gmail.com>
2019-06-28quic : Use TX event for app read notificationNathan Skrzypczak1-2/+1
Type: feature Change-Id: I1846cdeb35f079249f66a0351aa244c540923a43 Signed-off-by: Nathan Skrzypczak <nathan.skrzypczak@gmail.com>
2019-06-18quic: fix handling of stream reset & closeNathan Skrzypczak1-2/+2
Type: refactor Change-Id: I4981704e3c886d90d482a1deba42633e92626743 Signed-off-by: Nathan Skrzypczak <nathan.skrzypczak@gmail.com>
2019-06-17session: use listener_handle instead of listener_indexNathan Skrzypczak1-1/+0
Type: refactor Change-Id: I97fa59a0ba0b6b7a98698926020ffffcf6ae6ba3 Signed-off-by: Nathan Skrzypczak <nathan.skrzypczak@gmail.com>
2019-05-14QUIC multi thread updateAloys Augustin1-5/+9
* Add support for multiple threads * Replace quicly buffers with fifos * Fix cleanup of sessions * Update quicly release version Change-Id: I551f936bbec05a15703f043ee85c8e1ba0ab9723 Signed-off-by: Nathan Skrzypczak <nathan.skrzypczak@gmail.com>
2019-05-06Add QUIC multistream supportNathan Skrzypczak1-21/+15
Change-Id: I9dc746b8c62c3e7ee2f65e34a1a1dca243ed4bd9 Signed-off-by: Nathan Skrzypczak <nathan.skrzypczak@gmail.com>
2019-04-16QUIC: Initial multi stream supportAloys Augustin1-19/+38
To connect a stream, apps should call connect while passing the id of the QUIC connection in the new transport_opts field in session_endpoint_cfg_t. Apps are notified of new streams with their accept callback, which is called each time a peer opens a stream. Change-Id: I0f82ec344db58008d54641553eddec2973768435 Signed-off-by: Aloys Augustin <aloaugus@cisco.com>
2019-04-10quic: add dummy accepted cbFlorin Coras1-1/+1
Add dummy function to avoid session layer warning. Also, decrease default debug level to avoid startup quic init notifications. Change-Id: I2964fc15a8d93d78c976ed7c333e0cc5746c069e Signed-off-by: Florin Coras <fcoras@cisco.com>
2019-03-29Integrate first QUIC protocol implementationNathan Skrzypczak1-0/+118
Currently supports on single stream exposed through standard internal APIs Based on libquicly & picotls by h2o Change-Id: I7bc1ec0e399d1fb02bfd1da91aa7410076d08d14 Signed-off-by: Nathan Skrzypczak <nathan.skrzypczak@gmail.com>