Combine implementer, part, variant, and revision into one cpu
description line.
For example : ARM (Cortex-A57 PASS 1.2)
* get infos from /proc/cpuinfo
* only recognize armv8 processors
* add all given cavium processors
* Cavium starts counting variants from 1 instead of 0
Change-Id: I4f3820fb13a6bd2a0dc59e28fbe6f48a5b0ceb25
Signed-off-by: Gabriel Ganne <gabriel.ganne@enea.com>
- WIP: taking a LARGE patch and breaking into smaller chunks
- large LDP/VCL refactor in progress. This minimizes rebase.
- LARGE patch removed clib_warning()s erroneously
- VPPCOM_DEBUG checks duplicated deliberately as clib_warning()
usage will be refactored.
Change-Id: I5905ed074187fe22c9f1320fd07ffcaab3a0f63c
Signed-off-by: Keith Burns (alagalah) <alagalah@gmail.com>
The code path in src/vnet/gre/interface.c uses fib_id as a
component to generate hash key for GRE tunnel. This should be
fib_index as the GRE rx/decap data path will be using fib_index
to generate the hash key to lookup the GRE tunnel.
Change-Id: Ia7f0892d84f9dac79223a6e016775892b61eb5fb
Signed-off-by: John Lo <loj@cisco.com>
(cherry picked from commit 72247c803250894834d15952a6fddcd8f4a39a9c)
- add function to sock client that bootstraps shm api
- allow sock clients to request custom shm ring configs
Change-Id: Iabc1dd4f0dc8bbf8ba24de37f4966339fcf86107
Signed-off-by: Florin Coras <fcoras@cisco.com>
Move the functions hash_set_key_copy() and hash_unset_key_free()
which are dupilicated in various tunnel support code modules to
hash.h as hash_set_mem_alloc() and hash_unset_mem_free() to be
used by all.
Change-Id: I40723cabe29072ab7feb1804c221f28606d8e4fe
Signed-off-by: John Lo <loj@cisco.com>
IP table mapping is set using 'set int ip table X Y"
Change-Id: I2adec40015f9281c9b00c55506000b322f42d91a
Signed-off-by: Neale Ranns <neale.ranns@cisco.com>
Even though the trace now prints the hex as well as human readable format for acl plugin,
it can be handy to have a separate function which allows to decode the hex. So add this debug CLI.
Change-Id: I1db133a043374817ea9e94ae3736b8a98630669d
Signed-off-by: Andrew Yourtchenko <ayourtch@gmail.com>
L2 Emulation is a feautre that is applied to L2 ports to 'extract'
IP packets from the L2 path and inject them into the L3 path (i.e.
into the appropriate ip[4|6]_input node).
L3 routes in the table_id for that interface should then be configured
as DVR routes, therefore the forwarded packet has the L2 header
preserved and togehter the L3 routed system behaves like an L2 bridge.
Change-Id: I8effd7e2f4c67ee277b73c7bc79aa3e5a3e34d03
Signed-off-by: Neale Ranns <nranns@cisco.com>
* NAT46: fix test cleanup, missing del keyword
* NAT66: fix kube-proxy vip, is ipv6
* add some missing kp_put_writer_lock
* wipe flowtable after each unit test
* Add new cli api: "test kube-proxy flowtable flush" to flushes everything
* Call this new cli function after the end of each kube-proxy unit test.
* same as commit b3d1b203579226ca5136b9d6a2744577d07cfcc6 for the lb plugin
Change-Id: I4146f44841328ec96eb66729e3bae3d40f33e4aa
Signed-off-by: Gabriel Ganne <gabriel.ganne@enea.com>
Add API function which enables forwarding of packets not matching
existing translation or static mapping instead of dropping them.
When forwarding is enabled matching packets will be translated
while non-matching packets will be forwarded without translation.
Change-Id: Ic13040cbad16d3a1ecdc3e02a497171bef6aa413
Signed-off-by: Juraj Sloboda <jsloboda@cisco.com>
CLI parsing code (unformat_fib_route_path) does not
recognise 'resolve-via-connected'.
However it understands 'resolve-via-attached'.
Change-Id: I57e5eb75199cb8ae72c0fc8642a41f042b022201
Signed-off-by: Marek Gradzki <mgradzki@cisco.com>
Translation of both source and destination addresses and ports for 1:1 NAT
session initiated from outside network (ExternalIP K8 use case).
Change-Id: Ic0000497cf71619aac996d6d580844f0ea0edc14
Signed-off-by: Matus Fabian <matfabia@cisco.com>
Added Radu Nicolau as maintainer for
src/devices/dpdk/ipsec/ and src/vnet/ipsec/
Change-Id: If354344f5685ce0e431757768b77c4d5e8e409f9
Signed-off-by: Radu Nicolau <radu.nicolau@intel.com>
Set l_addr to the interface address if the interface address is known
when the identity mapping is created.
Change-Id: I61af0f5248c9d86d23a24457b342b2e1fb4ac726
Signed-off-by: Dave Barach <dave@barachs.net>
This plugin provides kube-proxy data plane on user space,
which is used to replace linux kernal's kube-proxy based on iptables.
The idea is largely inspired from VPP LB plugin.
Currently, kube-proxy plugin supports three service types:
1) Cluster IP plus Port: support any protocols, including TCP, UDP.
2) Node IP plus Node Port: currently only support UDP.
3) External Load Balancer.
Please refer to kp_plugin_doc.md for details.
Change-Id: I36690e417dd26ad5ec1bd77c7ea4b8100416cac6
Signed-off-by: Hongjun Ni <hongjun.ni@intel.com>
- update of CSIT operational branch to be used for VPP-patch test
Change-Id: I5ad0e36fa40095cc107a31b32dd2bbf936636dc4
Signed-off-by: Jan Gelety <jgelety@cisco.com>
- GRE tunnels with the same src,dst addresses are not the same tunnel
- Two data-plane improvements:
- the cached key was never updated and so useless
- no need to dereference the tunnel's HW interface to get the sw_if_index
Change-Id: I2f2ea6e08c759a810b753cec22c497e921a2ca01
Signed-off-by: Neale Ranns <nranns@cisco.com>
This allows arm platforms to also take advantage of crc32 hardware
acceleration.
* add a wrapper for crc32_u64. It's the only one really used. Using it
instead of a call to clib_crc32c() eases building symmetrical hash
functions.
* replace #ifdef on SSE4 by a test on clib_crc32c_uses_intrinsics.
Note: keep the test on i386
* fix typo in lb test log
Change-Id: I03a0897b70f6c1717e6901d93cf0fe024d5facb5
Signed-off-by: Gabriel Ganne <gabriel.ganne@enea.com>
Queue RPC calls and send them from the main dispatch loop. As things stood,
if the vpp main input queue filled, worker threads could enter a
barrier-sync spin-wait in the middle of processing a frame. If thread
0 decided to recreate worker thread data structures, the worker thread(s)
could easily crash.
Legislate the problem out of existence by enqueueing RPC messages only
from the main dispatch loop. At that point, doing a barrier-sync wait
is perfectly OK.
Change-Id: I18da3e44bb1f29a63fe5f30cf11de732ecfd5bf7
Signed-off-by: Dave Barach <dave@barachs.net>
