Age | Commit message (Collapse) | Author | Files | Lines |
|
- fix issue caused by assumption that ethertype starts at 0
- intoduce buffer templates to speed-up vlib_buffer_t metadata
initialization
- avoid check for multiseg buffers inside loop if multiseg is disabled
- interleave prefetches to reduce load on L1 cache
Change-Id: I3b76e6d3e1e15ed28f01625edb7fbe9f38112e03
Signed-off-by: Damjan Marion <damarion@cisco.com>
|
|
Change-Id: I1a8edaf1e0fca89e5d24d57dbb186ee46801f661
Signed-off-by: Neale Ranns <nranns@cisco.com>
|
|
Add operation_mode for create_vhost_user_if, modify_vhost_user_if,
and sw_interface_vhost_user_details. Only polling mode is supported
for these APIs. Other mode is rejected and gets
VNET_API_ERROR_UNIMPLEMENTED error.
Change-Id: I0596f4e2c087aa2b6f78eb3e0b63910b1859641e
Signed-off-by: Steven <sluong@cisco.com>
|
|
Change-Id: I94e8737cad9222d24602db4ad03bf43c12c62b79
Signed-off-by: Dave Barach <dave@barachs.net>
|
|
Change-Id: If9c4a5da06daf78000ea8637b95048f19b5019d0
Signed-off-by: Ole Troan <ot@cisco.com>
|
|
Change-Id: I9ff64b6c583e1c513f91f90798298b55d0793f06
Signed-off-by: Radu Nicolau <radu.nicolau@intel.com>
|
|
Change-Id: I080b90a4bc53c2595ade696c592e86790c7ca939
Signed-off-by: Filip Tehlar <ftehlar@cisco.com>
|
|
Change-Id: I70e32b2196bfc8712d270f9a4365daca2504d52d
Signed-off-by: Neale Ranns <nranns@cisco.com>
|
|
Change-Id: I0ffab147c3218a75b7c3bb829983f538c7b637ee
Signed-off-by: Martin <magalik@cisco.com>
|
|
bridging. That is hosts in one sub-net reachable via differenet interfaces.
Introducate a new API command:
ip6 nd proxy <host-address> <interface>
this indicates 2 things;
1) that host <host-address> is reachable out of interface <interface>. VPP will thus install that route.
2) NS requests sent to <host-address> will be responeded to (i.e. proxied).
Change-Id: I863f967fdb5097ab3b574769c70afdbfc8d5478a
Signed-off-by: Neale Ranns <nranns@cisco.com>
|
|
Change-Id: I5c5b4d680359cf3635c2d7b0088ea2ba6a428f93
Signed-off-by: Filip Tehlar <ftehlar@cisco.com>
|
|
Change-Id: I5143328b2da62ce4d6bb2915e2a51855696d87fc
Signed-off-by: Matus Fabian <matfabia@cisco.com>
|
|
Change-Id: I73154179e78aeae5f879125237bce593d0978fae
Signed-off-by: Florin Coras <fcoras@cisco.com>
|
|
expression as the destination
warning translates as an invalid write :
sizeof(u8* b_dmac) == 8 != sizeof(eth_hdr->dst_address) == 6
~/vpp/build-data/../src/vnet/l2/l2_vtr.c: In function 'l2pbb_get':
~/vpp/build-data/../src/vnet/l2/l2_vtr.c:734:63: error: argument to 'sizeof' in 'memcpy' call is the same expression as the destination;
did you mean to provide an explicit length? [-Werror=sizeof-pointer-memaccess]
~/vpp/build-data/../src/vnet/l2/l2_vtr.c:736:63: error: argument to 'sizeof' in 'memcpy' call is the same expression as the destination;
did you mean to provide an explicit length? [-Werror=sizeof-pointer-memaccess]
update l2pbb_get to take an ethernet header instead of two u8* pointers
for source and dest mac addresses.
Change-Id: Ifcf1319a9e22614d57682f940e10f0420dc6fb8c
Signed-off-by: Gabriel Ganne <gabriel.ganne@enea.com>
|
|
Create hash tables before loading plugins. Previous init sequence
wiped out most if not all of them.
Change-Id: I5dd628895f68f740d525951511a9fef1822d39da
Signed-off-by: Dave Barach <dave@barachs.net>
|
|
Multiple DHCP (4 and/or 6) servers can be added and removed through multiple calls to the 'set dhcp server' API.
All 4/6/ discover/solicit messages will then be replicated to all servers in the list. The expectation is that the servers/system is configured in such a way that this is viable.
If VSS information is providied for the clinet VRF which also has multiple servers configured, then the same VSS information is sent to each server. Likewise the source address of packets sent to from VPP to each server is the same.
Change-Id: I3287cb084c84b3f612b78bc69cfcb5b9c1f8934d
Signed-off-by: Neale Ranns <nranns@cisco.com>
|
|
ARP Event reply sw_if_index was getting passed in host byte order.
Change-Id: Ifae8673906ac2c4233f146786a3d02c38280809b
Signed-off-by: Wojciech Dec <wdec@cisco.com>
|
|
Added a new interface flag - HIDDEN
Indicates that the interface does not appear in CLI/API.
Added three new interface functions:
vnet_sw_interface_is_api_visible - indicates if the sw_if_index should be displayed
vnet_swif_is_api_visible - variant for sw_interface
vnet_sw_interface_is_api_valid - tests if the given if_index exists and is visible
for future use by api functions
Changed the unformat function to only accept visible interfaces
Changed vxlan to add the HIDDEN flag to deleted interfaces
This is the first part in a series to hide deleted interfaces from the API
Change-Id: Ib43cc5cf1c450856560faf4e84126eb3671038e2
Signed-off-by: Eyal Bari <ebari@cisco.com>
|
|
Change-Id: Icf0ddf76ba1c8b588c79387284cd0349ebc6e45f
Signed-off-by: AkshayaNadahalli <anadahal@cisco.com>
|
|
Inside user is statically mapped to a set of outside ports. Support endpoint
dependent mapping to deal with overloading of the outside ports.
Change-Id: I8014438744597a976f8ae459283e8b91f63b7f72
Signed-off-by: Matus Fabian <matfabia@cisco.com>
|
|
Change-Id: I9c0bb35ba16e04206ac481495f6638d3763754a1
Signed-off-by: Juraj Sloboda <jsloboda@cisco.com>
|
|
fix for functions:
- stream_session_lookup_listener6
- stream_session_lookup6
- stream_session_lookup_transport6
- stream_session_connect_notify
eg.
~/vpp/build-data/../src/vnet/session/session.c: In function 'stream_session_lookup_listener6':
~/vpp/build-data/../src/vnet/session/session.c:222:17: error: 'kv6.key[3]' is used uninitialized in this function [-Werror=uninitialized]
session_kv6_t kv6;
^
~/vpp/build-data/../src/vnet/session/session.c:222:17: error: 'kv6.key[2]' is used uninitialized in this function [-Werror=uninitialized]
~/vpp/build-data/../src/vnet/session/session.c:222:17: error: 'kv6.key[5]' is used uninitialized in this function [-Werror=uninitialized]
~/vpp/build-data/../src/vnet/session/session.c:222:17: error: 'kv6.key[4]' is used uninitialized in this function [-Werror=uninitialized]
Change-Id: I221cafbaf4effa43662ffcc7bf5f9581e28f9db9
Signed-off-by: Gabriel Ganne <gabriel.ganne@enea.com>
|
|
It does not look like it actually can be used uninitialized ... but gcc
complains, so with Werror this prevents from compiling.
sample warning output:
In file included from /home/gannega/export/vpp/build-data/../src/vnet/handoff.h:21:0,
from /home/gannega/export/vpp/build-data/../src/plugins/snat/in2out.c:19:
/home/gannega/export/vpp/build-data/../src/plugins/snat/in2out.c: In function 'snat_in2out_node_fn_inline':
/home/gannega/export/vpp/build-data/../src/vnet/ip/ip4_packet.h:244:15: error: 'inner_ip0' may be used uninitialized in this function [-Werror=maybe-uninitialized]
i->checksum = 0;
^
/home/gannega/export/vpp/build-data/../src/plugins/snat/in2out.c:433:17: note: 'inner_ip0' was declared here
ip4_header_t *inner_ip0;
Change-Id: If7c35dfb528861fd4ec6374aa1343b48f0f93057
Signed-off-by: Gabriel Ganne <gabriel.ganne@enea.com>
|
|
Change-Id: Ic8f186dbb35bb4e2e191d311cab51315a88a2d81
Signed-off-by: Ole Troan <ot@cisco.com>
|
|
Change-Id: Ifdd2b204ecf7d855f1269c11224b9c825311904d
Signed-off-by: Neale Ranns <nranns@cisco.com>
|
|
Change-Id: Ibe1734aeb94bc17cd8d8bc6f35ca7b780aaa9599
Signed-off-by: AkshayaNadahalli <anadahal@cisco.com>
|
|
Change-Id: I6d2afda991d771fb4a89fc3f6544f8e940a9b9f0
Signed-off-by: Shwetha Bhandari <shwethab@cisco.com>
|
|
Change-Id: I06a23d24340c5527f3848177d2178bf3e55f7614
Signed-off-by: Klement Sekera <ksekera@cisco.com>
|
|
Change-Id: I8db55443c0beb33a89520ab3167f703f8646ec9e
Signed-off-by: Neale Ranns <nranns@cisco.com>
|
|
Change-Id: Ifac7d9134d03d79164ce6f06ae9413279bbaadb3
Signed-off-by: Damjan Marion <damarion@cisco.com>
|
|
In a very unlikely scenario, during which a session is removed
and replaced with a different session sharing the same session
index, while a packet is in-flight during RPC call, drop that packet.
Change-Id: If1c4a77fc2ab460bae2435db066a133185b98747
Signed-off-by: Klement Sekera <ksekera@cisco.com>
|
|
Change-Id: I0371e8e3c94bb793e3c64d5f51aaebf19dddc4b4
Signed-off-by: Klement Sekera <ksekera@cisco.com>
|
|
Change-Id: If0e30837e07a21f3912676f5147cb242d3d2b235
Signed-off-by: Klement Sekera <ksekera@cisco.com>
|
|
Unless overridden by COREDUMP_SIZE env variable, tell VPP to set
coredump size to unlimited, otherwise use $COREDUMP_SIZE as
the argument.
Change-Id: Ia2a6508207c66a171b33d272c820b1deb4a83e82
Signed-off-by: Klement Sekera <ksekera@cisco.com>
|
|
Use setrlimit to set the core size limit if the argument is passed to
vpp.
Change-Id: Ie76c082b2af81337310fcb1925af915a42067f15
Signed-off-by: Klement Sekera <ksekera@cisco.com>
|
|
Change-Id: I9bea397e50fff4211c6e1e31503829e4dd16be42
Signed-off-by: Pavel Kotucek <pkotucek@cisco.com>
|
|
Change-Id: I0c24d9af90241083ae56b1d52239d2d55e70b65e
Signed-off-by: Damjan Marion <damarion@cisco.com>
|
|
Change-Id: I80831cee062a38a0f5ab1f1e56c2dc6dcd512b9d
Signed-off-by: Damjan Marion <damarion@cisco.com>
|
|
Check if there are existing vpp processes before running the test suite
and refuse to run if there are. This prevents the removal of other test
suite temporary files and also makes sure that if the machine is loaded
by (zombie) vpp processes, interactive tests (like bfd) won't fail.
Change-Id: I88a74098188cb3f51966de5db19d7f80f39e51e2
Signed-off-by: Klement Sekera <ksekera@cisco.com>
|
|
Change-Id: I3323f7d4bb5da4bc2b19637964e5467ac92680cd
Signed-off-by: Martin <magalik@cisco.com>
|
|
Changed vat_api_hookup(...) to <plugin-name>_api_hookup, change to
static functions. Fixed the related emacs-lisp plugin skeleton.
Change-Id: Id14f8fc3138751f469d48fecb26175e938f5f028
Signed-off-by: Dave Barach <dave@barachs.net>
|
|
session.api.h is included in src/vnet/vnet_all_api_h.h
since 68b0fb0c620, as such it should be added to the 'devel' package
and installed for out-of-tree builds which use binary VPP API
Change-Id: Ie8ad03d8df2a96cbc6913f767500d5ff34dcc915
Signed-off-by: Alexander Popovsky (apopovsk) <apopovsk@cisco.com>
|
|
DHCPv6. For the same table ID, the unicast-FIB index is not necessarily the same value as the multicast-FIB index, since features (like LISP, SR) can create unicast-tables, and thus affect only the index of the unicast FIBs
Change-Id: Ibfa334d7eda822f742c241b7ce69a6271b4753a9
Signed-off-by: Neale Ranns <nranns@cisco.com>
|
|
- Add CLI/API to enable session layer, by default it's disabled
- Improve rcv wnd computation
- Improvements to tx path
- URI code cleanup
- Builtin test tcp server
- Improve src port allocation
Change-Id: I2ace498e76a0771d4c31a8075cc14fe33d7dfa38
Signed-off-by: Florin Coras <fcoras@cisco.com>
|
|
Implements:
1.- SR Policies with several (weighted) SID lists
2.- Binding SID
3.- SR LocalSIDs with support for the following functions
- End
- End.X
- End.DX6
- End.DX4
- End.DX2
- End.DT6
- End.DT2
- End.B6
- End.B6.Encaps
4.- SR Steering policies (to steer a traffic through an SR Policy)
- Support for IPv6 traffic (IPv6 Encapsulation / SRH insertion)
- Support for IPv4 traffic (IPv6 Encapsulation)
- Support for L2 traffic
(T.Insert / T.Encaps)
5.- Doxygen documentation
6.- Framework (APIs) to allow the definition of new SR LocalSID behaviors by means of plugins
7.- Sample SRv6 LocalSID plugin
Change-Id: I2de3d126699d4f11f54c0f7f3b71420ea41fd389
Signed-off-by: Pablo Camarillo <pcamaril@cisco.com>
|
|
expiry value
This commit addresses two issues:
1) Avoid refilling the timing wheel with stale timers in rare circumstances.
The timing_wheel_advance() may call advance_cpu_time_base() to update the cpu_time_base,
which is used as a starting point for 32-bit offsets of events on the timer wheel.
If the timing_wheel_advance() is not called for a longer period of time,
then advance_cpu_time_base() is called multiple times in a loop.
advance_cpu_time_base() has two parts - the first part adjusting
the base for the existing event, and the second part trying to fill
with the new events from the overflow queue, which now fit into
the 32-bit-sized time window off the new cpu_time_base.
In doing so this second part incorrectly considers the timers which
have just expired (have the time index == w->current_time_index)
to still be unexpired and places them onto the wheel instead of returning
them as expired.
For quick successive executions of timing_wheel_advance() these events
result in a relatively benign late expiry - the newly placed events expire
during the next call to timing_wheel_advance().
If the successive executions of timing_wheel_advance() result in multiple
invocations of advance_cpu_time_base(), the Nth iteration of it may place a stale
event on the timer wheel if the event time index equals to the current time index
(which has been previously purged), while the N+1th iteration of it will trigger
an assert violation on this stale event, resulting in a reboot.
As part of the testing, two test runs were done before and after the change.
Each of the test runs consisted of the following command:
for i in `seq 1 300`; do ./test_timing_wheel validate events 10000 synthetic-time verbose seed $i iter 10000 wait-time 2 max-time 300; done
The test runs completed identically, however they uncovered the following assert failure:
vpp/src/vppinfra/test_timing_wheel.c:225 (test_timing_wheel_main) assertion `min_next_time[0] <= tm->events[i]' fails
This assert is the second issue covered by this commit:
2) Inserting a new element may result in incorrect cached expiry value
The w->cached_min_cpu_time_on_wheel is being updated within timing_wheel_advance() every time
the elements are expired.
However, it is not touched if the new elements are inserted. Assuming current time is "T"
and the cached min cpu time is "T+X", if a new element is being inserted whose expiry time is "T+Y",
and Y is such that Y < X, then the value w->cached_min_cpu_time_on_wheel becomes incorrect
until the next expiry event, during which it is updated. The test catches this transient condition
which results in the asserts seen in the runs above.
The solution is to update the w->cached_min_cpu_time_on_wheel within timing_wheel_insert_helper()
as necessary.
Change-Id: I56a65a9a11cc2a1e0b36937a9c6d5ad10233a731
Signed-off-by: Andrew Yourtchenko <ayourtch@gmail.com>
|
|
To support creating loopback interfaces with a specific
instance number, a new CREATE_LOOPBACK_INSTANCE API call
with flag is_specified and value user_instance is introduced.
Presumably the existing CREATE_LOOPBACK API message will be
obsoleted and revmoved.
The VAT cli commands can now mention and format
the new field as 'instance %d' data. If no instance
number is named, the old call CREATE_LOOPBACK is used
to maintain backward compatibility. However, if the
instance is named, the new CREATE_LOOPBACK_INSTANCE
message will be used.
Both the dynamically allocated and user-requested instance
number are tracked in a bitvector. If is_specified is 0,
the next free instance will be used.. A request for a specific
instance number will be granted if it is available. On error,
the value ~0 is returned.
Change-Id: I849815563a5da736dcd6bccd262ef49b963f6643
Signed-off-by: Jon Loeliger <jdl@netgate.com>
|
|
Change-Id: I7a0ff25a8f74e6eda3a44f2cd7d2e022683accd9
Signed-off-by: Pablo Camarillo <pcamaril@cisco.com>
|
|
Remove the duplcate VXLAN related API handlers from api.c and keep
the proper ones in ./src/vnet/vxlan/vxlan_api.c.
Change-Id: I3b17e17d735bb453d080243bfa2783ce0de64885
Signed-off-by: John Lo <loj@cisco.com>
|
|
via l2output_main.next_nodes
Before this commit, several output features that happen to be the
last in the list of features to be executed, send the packets directly
to <interfaceName>-output. To do this, they use l2_output_dispatch,
which builds a list of sw_if_index to next index mappings.
When interfaces are deleted and the new interfaces are created,
these mappings become stale, and cause the packets being sent to wrong
interface output nodes.
This patch (thanks John Lo for the brilliant idea!) adds a feature node "output",
whose sole purpose is dispatching the packets to the correct interface output
nodes. To do that, it uses the l2output_main.next_nodes, which is already
taken care of for the case of the sw_if_index reuse, so this makes the dependent
features all work correctly.
Since this changes the packet path, for the features that were always the last ones
it has triggered a side problem of the output feat_next_node_index not being properly
initalized. These two users are l2-output-classify node and the output nodes belonging
to the acl-plugin.
For the first one the less invasive fix is just to initialize that field.
For the acl-plugin nodes, rewrite the affected part of the code to use
feat_bitmap_get_next_node_index since this is essentially what the conditional
in l2_output_dispatch does, and fix the compiler warnings generated.
This fix was first made in stable/1701 under commit e7dcee4027854b0ad076101471afdfff67eb9011.
Change-Id: I32e876ab1e1d498cf0854c19c6318dcf59a93805
Signed-off-by: Andrew Yourtchenko <ayourtch@gmail.com>
|