vpp - Vector Packet Processing

Age	Commit message (Collapse)	Author	Files	Lines
2020-04-02	tcp: move features to separate files	Florin Coras	1	-0/+1
	Type: refactor Signed-off-by: Florin Coras <fcoras@cisco.com> Change-Id: Ia477b8dba9266f47907967e363c11048e5cd95ab
2019-10-17	tcp: Init cwnd from ssthresh.	Sergey Ivanushkin	1	-1/+31
	Set high ssthresh out of the box and make configurable Type: fix Signed-off-by: Sergey Ivanushkin <sergey.ivanushkin@enea.com> Change-Id: Iba1549b4ee55e51468ad0b28ef3d26a85fa9cae0
2019-09-25	tcp: use sacks for timer based recovery	Florin Coras	1	-6/+0
	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>
2019-09-04	tcp: cc algos handle cwnd on congestion signal	Florin Coras	1	-0/+6
	Type: refactor Change-Id: I15b10a22d0d0b83075a0eef5ef8c09cf76989866 Signed-off-by: Florin Coras <fcoras@cisco.com>
2019-07-05	tcp: add loss signal to cc algo	Florin Coras	1	-4/+12
	Type:feature Change-Id: Ibe1a4c555b55fb929d55b02599aaf099ed522cdf Signed-off-by: Florin Coras <fcoras@cisco.com>
2019-06-25	tcp: delivery rate estimator	Florin Coras	1	-2/+3
	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>
2019-04-08	host stack: update stale copyright	Florin Coras	1	-1/+1
	Change-Id: I33cd6e44d126c73c1f4c16b2041ea607b4d7f39f Signed-off-by: Florin Coras <fcoras@cisco.com>
2018-11-13	tcp: cubic fast convergence	Florin Coras	1	-0/+1
	Change-Id: I3a15960fe346763faf13e8728ce36c2f3bf7b05a Signed-off-by: Florin Coras <fcoras@cisco.com>
2018-11-09	tcp: basic cubic implementation	Florin Coras	1	-9/+1
	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>
2018-05-26	tcp: loss recovery improvements/fixes	Florin Coras	1	-1/+1
	- 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>
2018-05-23	tcp: cc improvements and fixes	Florin Coras	1	-0/+2
	Change-Id: I6615bb612bcc3f795b5f822ea55209bb30ef35b5 Signed-off-by: Florin Coras <fcoras@cisco.com>
2018-04-20	tcp: make newreno byte instead of acks dependent	Florin Coras	1	-2/+8
	Should be more resilient to ack losses Change-Id: Icec3b93c1d290dec437fcc4e6fe5171906c9ba8a Signed-off-by: Florin Coras <fcoras@cisco.com>
2017-07-11	Horizontal (nSessions) scaling draft	Dave Barach	1	-2/+2
	- Data structure preallocation. - Input state machine fixes for mid-stream 3-way handshake retries. - Batch connections in the builtin_client - Multiple private fifo segment support - Fix elog simultaneous event type registration - Fix sacks when segment hole is added after highest sacked - Add "accepting" session state for sessions pending accept - Add ssvm non-recursive locking - Estimate RTT for syn-ack - Don't init fifo pointers. We're using relative offsets for ooo segments - CLI to dump individual session Change-Id: Ie0598563fd246537bafba4feed7985478ea1d415 Signed-off-by: Dave Barach <dbarach@cisco.com> Signed-off-by: Florin Coras <fcoras@cisco.com>
2017-06-19	Overall tcp performance improvements (VPP-846)	Florin Coras	1	-2/+2
	- limit minimum rto per connection - cleanup sack scoreboard - switched svm fifo out-of-order data handling from absolute offsets to relative offsets. - improve cwnd handling when using sacks - add cc event debug stats - improved uri tcp test client/server: bugfixes and added half-duplex mode - expanded builtin client/server - updated uri socket client/server code to work in half-duplex - ensure session node unsets fifo event for empty fifo - fix session detach Change-Id: Ia446972340e32a65e0694ee2844355167d0c170d Signed-off-by: Florin Coras <fcoras@cisco.com>
2017-06-09	Implement sack based tcp loss recovery (RFC 6675)	Florin Coras	1	-3/+17
	- refactor existing congestion control code (RFC 6582/5681). Handling of ack feedback now consists of: ack parsing, cc event detection, event handling, congestion control update - extend sack scoreboard to support sack based retransmissions - basic implementation of Eifel detection algorithm (RFC 3522) for detecting spurious retransmissions - actually initialize the per-thread frame freelist hash tables - increase worker stack size to 2mb - fix session queue node out-of-buffer handling - ensure that the local buffer cache vec_len matches reality - avoid 2x spurious event requeues when short of buffers - count out-of-buffer events - make the builtin server thread-safe - fix bihash template threading issue: need to paint -1 across uninitialized working_copy_length vector elements (via rebase from master) Change-Id: I646cb9f1add9a67d08f4a87badbcb117980ebfc4 Signed-off-by: Florin Coras <fcoras@cisco.com> Signed-off-by: Dave Barach <dbarach@cisco.com>
2017-05-07	Fix TCP loss recovery, VPP-745	Florin Coras	1	-1/+1
	Allows pure loss recovery retransmits only on timeout. Change-Id: I563cdbf9e7b890a6569350bdbda4f746ace0544e Signed-off-by: Florin Coras <fcoras@cisco.com>
2017-03-01	VPP-598: tcp stack initial commit	Dave Barach	1	-0/+93
	Change-Id: I49e5ce0aae6e4ff634024387ceaf7dbc432a0351 Signed-off-by: Dave Barach <dave@barachs.net> Signed-off-by: Florin Coras <fcoras@cisco.com>

#!/usr/bin/env python3 import unittest import socket from framework import VppTestCase, VppTestRunner from vpp_ip_route import VppIpRoute, VppRoutePath from vpp_l2 import L2_PORT_TYPE, BRIDGE_FLAGS from scapy.packet import Raw from scapy.layers.l2 import Ether from scapy.layers.inet import IP, UDP NUM_PKTS = 67 class TestL2Flood(VppTestCase): """ L2-flood """ @classmethod def setUpClass(cls): super(TestL2Flood, cls).setUpClass() @classmethod def tearDownClass(cls): super(TestL2Flood, cls).tearDownClass() def setUp(self): super(TestL2Flood, self).setUp() # 12 l2 interface and one l3 self.create_pg_interfaces(range(13)) self.create_bvi_interfaces(1) for i in self.pg_interfaces: i.admin_up() for i in self.bvi_interfaces: i.admin_up() self.pg12.config_ip4() self.pg12.resolve_arp() self.bvi0.config_ip4() def tearDown(self): self.pg12.unconfig_ip4() self.bvi0.unconfig_ip4() for i in self.pg_interfaces: i.admin_down() for i in self.bvi_interfaces: i.admin_down() super(TestL2Flood, self).tearDown() def test_flood(self): """ L2 Flood Tests """ # # Create a single bridge Domain # self.vapi.bridge_domain_add_del(bd_id=1) # # add each interface to the BD. 3 interfaces per split horizon group # for i in self.pg_interfaces[0:4]: self.vapi.sw_interface_set_l2_bridge(rx_sw_if_index=i.sw_if_index, bd_id=1, shg=0) for i in self.pg_interfaces[4:8]: self.vapi.sw_interface_set_l2_bridge(rx_sw_if_index=i.sw_if_index, bd_id=1, shg=1) for i in self.pg_interfaces[8:12]: self.vapi.sw_interface_set_l2_bridge(rx_sw_if_index=i.sw_if_index, bd_id=1, shg=2) for i in self.bvi_interfaces: self.vapi.sw_interface_set_l2_bridge(rx_sw_if_index=i.sw_if_index, bd_id=1, shg=2, port_type=L2_PORT_TYPE.BVI) p = (Ether(dst="ff:ff:ff:ff:ff:ff", src="00:00:de:ad:be:ef") / IP(src="10.10.10.10", dst="1.1.1.1") / UDP(sport=1234, dport=1234) / Raw(b'\xa5' * 100)) # # input on pg0 expect copies on pg1->11 # this is in SHG=0 so its flooded to all, expect the pg0 since that's # the ingress link # self.pg0.add_stream(p*NUM_PKTS) self.pg_enable_capture(self.pg_interfaces) self.pg_start() for i in self.pg_interfaces[1:12]: rx0 = i.get_capture(NUM_PKTS, timeout=1) # # input on pg4 (SHG=1) expect copies on pg0->3 (SHG=0) # and pg8->11 (SHG=2) # self.pg4.add_stream(p*NUM_PKTS) self.pg_enable_capture(self.pg_interfaces) self.pg_start() for i in self.pg_interfaces[:4]: rx0 = i.get_capture(NUM_PKTS, timeout=1) for i in self.pg_interfaces[8:12]: rx0 = i.get_capture(NUM_PKTS, timeout=1) for i in self.pg_interfaces[4:8]: i.assert_nothing_captured(remark="Different SH group") # # An IP route so the packet that hits the BVI is sent out of pg12 # ip_route = VppIpRoute(self, "1.1.1.1", 32, [VppRoutePath(self.pg12.remote_ip4, self.pg12.sw_if_index)]) ip_route.add_vpp_config() self.logger.info(self.vapi.cli("sh bridge 1 detail")) # # input on pg0 expect copies on pg1->12 # this is in SHG=0 so its flooded to all, expect the pg0 since that's # the ingress link # self.pg0.add_stream(p*NUM_PKTS) self.pg_enable_capture(self.pg_interfaces) self.pg_start() for i in self.pg_interfaces[1:]: rx0 = i.get_capture(NUM_PKTS, timeout=1) # # input on pg4 (SHG=1) expect copies on pg0->3 (SHG=0) # and pg8->12 (SHG=2) # self.pg4.add_stream(p*NUM_PKTS) self.pg_enable_capture(self.pg_interfaces) self.pg_start() for i in self.pg_interfaces[:4]: rx0 = i.get_capture(NUM_PKTS, timeout=1) for i in self.pg_interfaces[8:13]: rx0 = i.get_capture(NUM_PKTS, timeout=1) for i in self.pg_interfaces[4:8]: i.assert_nothing_captured(remark="Different SH group") # # cleanup # for i in self.pg_interfaces[:12]: self.vapi.sw_interface_set_l2_bridge(rx_sw_if_index=i.sw_if_index, bd_id=1, enable=0) for i in self.bvi_interfaces: self.vapi.sw_interface_set_l2_bridge(rx_sw_if_index=i.sw_if_index, bd_id=1, shg=2, port_type=L2_PORT_TYPE.BVI, enable=0) self.vapi.bridge_domain_add_del(bd_id=1, is_add=0) def test_flood_one(self): """ L2 no-Flood Test """ # # Create a single bridge Domain # self.vapi.bridge_domain_add_del(bd_id=1) # # add 2 interfaces to the BD. this means a flood goes to only # one member # for i in self.pg_interfaces[:2]: self.vapi.sw_interface_set_l2_bridge(rx_sw_if_index=i.sw_if_index, bd_id=1, shg=0) p = (Ether(dst="ff:ff:ff:ff:ff:ff", src="00:00:de:ad:be:ef") / IP(src="10.10.10.10", dst="1.1.1.1") / UDP(sport=1234, dport=1234) / Raw(b'\xa5' * 100)) # # input on pg0 expect copies on pg1 # self.send_and_expect(self.pg0, p*NUM_PKTS, self.pg1) # # cleanup # for i in self.pg_interfaces[:2]: self.vapi.sw_interface_set_l2_bridge(rx_sw_if_index=i.sw_if_index, bd_id=1, enable=0) self.vapi.bridge_domain_add_del(bd_id=1, is_add=0) def test_uu_fwd(self): """ UU Flood """ # # Create a single bridge Domain # self.vapi.bridge_domain_add_del(bd_id=1, uu_flood=1) # # add each interface to the BD. 3 interfaces per split horizon group # for i in self.pg_interfaces[0:4]: self.vapi.sw_interface_set_l2_bridge(rx_sw_if_index=i.sw_if_index, bd_id=1, shg=0) # # an unknown unicast and broadcast packets # p_uu = (Ether(dst="00:00:00:c1:5c:00", src="00:00:de:ad:be:ef") / IP(src="10.10.10.10", dst="1.1.1.1") / UDP(sport=1234, dport=1234) / Raw(b'\xa5' * 100)) p_bm = (Ether(dst="ff:ff:ff:ff:ff:ff", src="00:00:de:ad:be:ef") / IP(src="10.10.10.10", dst="1.1.1.1") / UDP(sport=1234, dport=1234) / Raw(b'\xa5' * 100)) # # input on pg0, expected copies on pg1->4 # self.pg0.add_stream(p_uu*NUM_PKTS) self.pg_enable_capture(self.pg_interfaces) self.pg_start() for i in self.pg_interfaces[1:4]: rx0 = i.get_capture(NUM_PKTS, timeout=1) self.pg0.add_stream(p_bm*NUM_PKTS) self.pg_enable_capture(self.pg_interfaces) self.pg_start() for i in self.pg_interfaces[1:4]: rx0 = i.get_capture(NUM_PKTS, timeout=1) # # use pg8 as the uu-fwd interface # self.vapi.sw_interface_set_l2_bridge( rx_sw_if_index=self.pg8.sw_if_index, bd_id=1, shg=0, port_type=L2_PORT_TYPE.UU_FWD) # # expect the UU packet on the uu-fwd interface and not be flooded # self.pg0.add_stream(p_uu*NUM_PKTS) self.pg_enable_capture(self.pg_interfaces) self.pg_start() rx0 = self.pg8.get_capture(NUM_PKTS, timeout=1) for i in self.pg_interfaces[0:4]: i.assert_nothing_captured(remark="UU not flooded") self.pg0.add_stream(p_bm*NUM_PKTS) self.pg_enable_capture(self.pg_interfaces) self.pg_start() for i in self.pg_interfaces[1:4]: rx0 = i.get_capture(NUM_PKTS, timeout=1) # # remove the uu-fwd interface and expect UU to be flooded again # self.vapi.sw_interface_set_l2_bridge( rx_sw_if_index=self.pg8.sw_if_index, bd_id=1, shg=0, port_type=L2_PORT_TYPE.UU_FWD, enable=0) self.pg0.add_stream(p_uu*NUM_PKTS) self.pg_enable_capture(self.pg_interfaces) self.pg_start() for i in self.pg_interfaces[1:4]: rx0 = i.get_capture(NUM_PKTS, timeout=1) # # change the BD config to not support UU-flood # self.vapi.bridge_flags(bd_id=1, is_set=0, flags=BRIDGE_FLAGS.UU_FLOOD) self.send_and_assert_no_replies(self.pg0, p_uu) # # re-add the uu-fwd interface # self.vapi.sw_interface_set_l2_bridge( rx_sw_if_index=self.pg8.sw_if_index, bd_id=1, shg=0, port_type=L2_PORT_TYPE.UU_FWD) self.logger.info(self.vapi.cli("sh bridge 1 detail")) self.pg0.add_stream(p_uu*NUM_PKTS) self.pg_enable_capture(self.pg_interfaces) self.pg_start() rx0 = self.pg8.get_capture(NUM_PKTS, timeout=1) for i in self.pg_interfaces[0:4]: i.assert_nothing_captured(remark="UU not flooded") # # remove the uu-fwd interface # self.vapi.sw_interface_set_l2_bridge( rx_sw_if_index=self.pg8.sw_if_index, bd_id=1, shg=0, port_type=L2_PORT_TYPE.UU_FWD, enable=0) self.send_and_assert_no_replies(self.pg0, p_uu) # # cleanup # for i in self.pg_interfaces[:4]: self.vapi.sw_interface_set_l2_bridge(rx_sw_if_index=i.sw_if_index, bd_id=1, enable=0) self.vapi.bridge_domain_add_del(bd_id=1, is_add=0) if __name__ == '__main__': unittest.main(testRunner=VppTestRunner)