Age | Commit message (Collapse) | Author | Files | Lines |
|
Also does a bit of code cleanup.
Type: refactor
Signed-off-by: Florin Coras <fcoras@cisco.com>
Change-Id: Ifbff1e03a2e1fdca98925fdd3db7eb230ff24a37
|
|
Type: fix
Signed-off-by: Filip Tehlar <ftehlar@cisco.com>
Signed-off-by: Florin Coras <fcoras@cisco.com>
Change-Id: Ibe39bc045c3b154209a83b59ef95a37c61b32c0c
|
|
Type: improvement
Signed-off-by: Florin Coras <fcoras@cisco.com>
Change-Id: I4b3427b966f9ff1ba8895fed7db662d56650f3f5
|
|
Use autogenerated code.
Does not change API definitions.
Type: improvement
Signed-off-by: Filip Tehlar <ftehlar@cisco.com>
Change-Id: I3c06e9cb3fabdcdce9c17e93cfedfd771295f589
|
|
In the current implement, tcp would start or up an one tick
retransmit timer for that connection if vlib_buffer_alloc()
return 0. Now the tick is 0.1ms, this means that if VPP is
in a buffer shortage state, there would be a large number of
burst timer expirations.
This commit limits the minimum interval of the retransmission
timer to 100ms.
Type: fix
Signed-off-by: liuyacan <liuyacan@corp.netease.com>
Change-Id: Ia11d693fe46119c5dc16b24ca93c30c31109057a
|
|
Half-open sessions are allocated by main thread and cleaned up on main
with timers.
Type: improvement
Signed-off-by: Florin Coras <fcoras@cisco.com>
Change-Id: I37f000920a45908b62b5501ae9d54a88a9e4c609
|
|
trajectory trace has been broken for a while because we used to save the
buffer trajectory in a vector pointed to in opaque2. This does not work
well when opaque2 is copied (eg. because of a clone) as 2 buffers end up
sharing the same vector.
This dedicates a full cacheline in the buffer metadata instead when
trajectory is compiled in. No dynamic allocation, no sharing, no tears.
Type: refactor
Change-Id: I6a028ca1b48d38f393a36979e5e452c2dd48ad3f
Signed-off-by: Benoît Ganne <bganne@cisco.com>
|
|
Type: feature
Signed-off-by: Florin Coras <fcoras@cisco.com>
Change-Id: Ifdd6024daf044751895bb8d2deabad41d3a80c92
|
|
Type: improvement
Signed-off-by: Florin Coras <fcoras@cisco.com>
Change-Id: I998c0686f9f7dc556dda8b28e23bbed127d0aafc
|
|
Type: refactor
Signed-off-by: Florin Coras <fcoras@cisco.com>
Change-Id: Iae5045686c62ad9279be350cffeaa41b8297d179
|
|
Type: fix
Signed-off-by: Ryujiro Shibuya <ryujiro.shibuya@owmobility.com>
Change-Id: Ibe305b96f04ffdc95ddb07d144dc00119ae69a07
|
|
Type: improvement
Signed-off-by: Simon Zhang <yuwei1.zhang@intel.com>
Change-Id: I14de90f07d825c5c99023996a88173ee855e9a6f
|
|
- switch to using vlib_time as reference for timestamps
- use us precision ticks for tcp but keep using ms precision for
timestamps. As a result, srtt, rttvar and rto are now measured in us
instead of ms. MRTT samples from timestamps are converted from ms to
us (not accurate under ms) while high precision samples are used with us
precision, i.e., they're no longer converted to ms precision samples.
Type: improvement
Change-Id: Ibda559575d9b4fdc85b0985264f7c865ff367e34
Signed-off-by: Florin Coras <fcoras@cisco.com>
|
|
Also removes delack timer and reuses the u32 for the pending timers
list.
Type: fix
Ticket: VPP-1923
Signed-off-by: Florin Coras <fcoras@cisco.com>
Change-Id: I4edbb72d5b2aa5e14f87659f49e675af1e834aca
|
|
Reported as part of "show tcp stats"
Type: improvement
Signed-off-by: Florin Coras <fcoras@cisco.com>
Change-Id: I499ab4e41d13aaf1a0d060e37f09087e63d09567
|
|
Type: improvement
Signed-off-by: Florin Coras <fcoras@cisco.com>
Change-Id: Ieb8bb9c6deb92479fdd3e045778fe5ae4782d1ea
|
|
Type: refactor
Signed-off-by: Florin Coras <fcoras@cisco.com>
Change-Id: I06120d43c15b2fa3f437ef235bf5f83ec2beb45e
|
|
Type: refactor
Signed-off-by: Florin Coras <fcoras@cisco.com>
Change-Id: Ia477b8dba9266f47907967e363c11048e5cd95ab
|
|
Type: improvement
Signed-off-by: Florin Coras <fcoras@cisco.com>
Change-Id: I71df27049ef0193578f0c42f8f8bbd5c54e4d53e
|
|
Type: improvement
Signed-off-by: Florin Coras <fcoras@cisco.com>
Change-Id: I3b0041d72d835cbb11e803cc56ac4c68a68238a0
|
|
Type: refactor
Signed-off-by: Florin Coras <fcoras@cisco.com>
Change-Id: I5120eb321e9e2050eb64dca9e2dfb6256ff2b62d
|
|
Type: improvement
Signed-off-by: Florin Coras <fcoras@cisco.com>
Change-Id: Iae9f118f710153b6c0e390265039db7434e67ed8
|
|
Type: improvement
Use only one api to retrieve transport send parameters. Additionally,
allow transports to request postponing and descheduling of events.
With this, tcp now requests descheduling of sessions when the
connections are stuck probing for zero snd_wnd
Signed-off-by: Florin Coras <fcoras@cisco.com>
Change-Id: I722c974f3e68fa15424c519a1fffacda43af050c
|
|
Type: improvement
Signed-off-by: Florin Coras <fcoras@cisco.com>
Change-Id: Ida896e9fbe83b4c76578b089bb3fd2bef5e07fd2
|
|
Type: improvement
Signed-off-by: Florin Coras <fcoras@cisco.com>
Change-Id: If0a634c277d4f8e4d32321f52253721a990bdb1d
|
|
Type: improvement
Signed-off-by: Florin Coras <fcoras@cisco.com>
Change-Id: I93067054631d6ae2411a7b08d7b681aed7a121b2
|
|
Type: improvement
Signed-off-by: Florin Coras <fcoras@cisco.com>
Change-Id: Id931f3f308a01788d222d0d62f26f5c579321c6a
|
|
Type: feature
Signed-off-by: Florin Coras <fcoras@cisco.com>
Change-Id: I9dafe564229095d50285276a654f4983f93faff2
|
|
Type: refactor
Signed-off-by: Florin Coras <fcoras@cisco.com>
Change-Id: I06cc3f032a72eacc716eed8d6fe8856a2839def1
|
|
Type: refactor
Signed-off-by: Florin Coras <fcoras@cisco.com>
Change-Id: Id09fa54cc5657b3b5616ea6a6180c1344b0141d4
|
|
Type: feature
If set, avoid trying to cleanup the local endpoint table.
Signed-off-by: Florin Coras <fcoras@cisco.com>
Change-Id: Iaa382ae74523d608217d2d17544bf1e7d82b2728
|
|
Type: fix
Change-Id: Ib823d016c64998779fb1d00b8aad3acb5e8340be
Signed-off-by: Florin Coras <fcoras@cisco.com>
|
|
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>
|
|
Type: fix
Change-Id: Icc8cd05cb9c1844a9f8b0399797ad554f4ccf1a1
Signed-off-by: Florin Coras <fcoras@cisco.com>
|
|
Type: feature
Change-Id: I77b3ee74229f3c85e99f74bf9000cb5aedbc1760
Signed-off-by: Florin Coras <fcoras@cisco.com>
|
|
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>
|
|
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>
|
|
Type: fix
Change-Id: I6f7df0d358f57f7feadb9b7a3fcffb99558b2af8
Signed-off-by: Florin Coras <fcoras@cisco.com>
|
|
Type:feature
For cases when proxy is in use IPv6 flow label received in origin pkt needs to be added
to ipv6 header of outgoing pkts from proxy to original destination and vice versa.
Signed-off-by: Tarun Gupta <tarungup@cisco.com>
Change-Id: I143f7e67237c0f865333078628a016b50ad5e630
Signed-off-by: Tarun Gupta <tarungup@cisco.com>
|
|
Type: feature
Change-Id: I18dfac9c3d2f539edf20dba64cd2b7b541154144
Signed-off-by: Florin Coras <fcoras@cisco.com>
|
|
Type: fix
- Initialize max_seq on both transmitted and retransmitted segments
- Keep track of segments that have been sacked.
- Track new data segments sent during recovery
Change-Id: Ice55231a3da200ae6171702e54b2ce155f831143
Signed-off-by: Florin Coras <fcoras@cisco.com>
|
|
Type: feature
Add infra that allows transpors to enqueue pending buffers without the
need to build and manage their own pending frames. An important benefit
is the fact that buffer wire/tx ordering is ensured by session layer.
Change-Id: I764fd1693d610b321a1d0c84b648a314f14583db
Signed-off-by: Florin Coras <fcoras@cisco.com>
|
|
Type: fix
Change-Id: I02e2cb2d349f9ddb2bfed040a08ca448f379d565
Signed-off-by: Florin Coras <fcoras@cisco.com>
|
|
Type: refactor
Change-Id: Ib5e2ecf55732599e1322736e3662178c9da49c7d
Signed-off-by: Florin Coras <fcoras@cisco.com>
|
|
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>
|
|
Add heuristic that detects lost retransmitted segments and retries
sending them.
Type: feature
Change-Id: I34d1bb16799e1993779222eb2bfad4b40704159e
Signed-off-by: Florin Coras <fcoras@cisco.com>
|
|
Type: feature
Based on the configuration, we can disable checksum offload capability
and calculate checksum while pushing the TCP & IP header.
This saves some cycles when VPP stack is used in legacy hardware devices.
Signed-off-by: Srikanth A <srakula@cisco.com>
Change-Id: Ic1b3fcf3040917e47ee65263694ebf7437ac5668
(cherry picked from commit 3642782a2748503f5b5ccf89d1575c1d489948ef)
|
|
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>
|
|
Type: feature
Change-Id: I6bd675dc28dbc34293e895911efe986844c0564f
Signed-off-by: Florin Coras <fcoras@cisco.com>
|
|
Type: feature
Change-Id: I201155b1b92cf7e57310af726879dab039090582
Signed-off-by: Florin Coras <fcoras@cisco.com>
|