This patch prepares code for bumping DPDK version to 22.11, but the DPDK version of this patch keeps at 22.07 for compatibility.
the "no-dsa" parameter in DPDK configuration is removed, the "blacklist" parameter can be used to block the related DSA devices.
Type: feature
Signed-off-by: Xinyao Cai <xinyao.cai@intel.com>
Change-Id: I08787c6584bba66383fc0a784963f33171196910
Apache ab sometimes fails during extensive performace testing.
This patch makes sure hs-test perf tests always pass.
Type: test
Signed-off-by: Filip Tehlar <ftehlar@cisco.com>
Change-Id: I0921682f0f07df3af45b342b9a7ddfa1af037ceb
Set the mask of calculating the next cqe index to the corresponding CQ
size instead of rxq size.
Type: fix
Signed-off-by: Jieqiang Wang <jieqiang.wang@arm.com>
Change-Id: I67494f029967af64051f51452eba1fd699984cd9
With 'api-trace { on }' in startup.conf, running 'api trace dump' in
vppctl was causing VPP to seg fault. vl_msg_print_trace() was calling
m->endian_handler() without checking whether its null.
Checking if its non-null prevents a crash, but the trace dump prints
the message IDs for trace_plugin_msg_ids in network byte order. There is
an auto-generated endian function for that message. Set it on the call
to vl_msg_api_config() for trace_plugin_msg_ids so the IDs will be
printed in host byte order in trace dump output.
Type: fix
Fixes: fe45f8f5
Signed-off-by: Matthew Smith <mgsmith@netgate.com>
Change-Id: I0ab463985e9a983155feba13ac4eb99ab883ace6
Previously, .src_ip_sticky may have been left uninitialized.
Type: fix
Fixes: 613e6dc0bf928def5d337312d522e1a15df87b00
Change-Id: Ifd866d6322fe9ff723f92b7ab3fd77e720a3cfa4
Signed-off-by: Vratko Polak <vrpolak@cisco.com>
To make out of tree plugins require a particular version, they need
access to the version they are built with. Install version.h.
Type: fix
Change-Id: I5916d0a16aed7e054ede452af956fee56cd078f0
Signed-off-by: Ole Troan <ot@cisco.com>
When a stats entry is removed it is marked empty.
The stats client did not check for that and returned an empty string.
This resulted in blank lines in vpp_get_stats. Fix by returning null instead
and checking value.
Type: fix
Signed-off-by: Ole Troan <ot@cisco.com>
Change-Id: I08a39ba3ef4421bf275747a6300f97fe36791b50
for testing purposes, disabled by default
Type: improvement
Signed-off-by: Damjan Marion <damarion@cisco.com>
Change-Id: Id616e2b3b21ae0f0b44e2b55ecefd501afacc7f2
RTA_VIA allows routes to have a next-hop in a different address family.
This commit makes linux-cp import those types of routes correctly,
instead of importing the routes without a gateway.
This uses rtnl_route_nh_get_gateway, which is available since libnl
3.4.0 (Oct. 9, 2017). Even Debian Stretch has it via backports.
Type: fix
Change-Id: I06297c700461ba7874eb8baf9355bd40990b3121
Signed-off-by: Adrian Pistol <vifino@posteo.net>
Nat in2out sessions are distributing among workers by client
addresses. In case there's multiple client vrfs with very
similar client addresses (usually from rfc1918), session
distribution/load can be unfair just due similar hash.
Let's take dynamic client fib_index into account, it'll affect
external port range only, outside address picking has own
address-based hash therefore not affected.
Type: improvement
Change-Id: I56ab2e1ce8dd27f2b1f9e7f22839ccf7774bfb82
Signed-off-by: Vladislav Grishenko <themiron@yandex-team.ru>
The unformat type for "%d" should be u32 or int.
Type: fix
Signed-off-by: Ted Chen <znscnchen@gmail.com>
Change-Id: I2483df6259ed8d3c7648c8db6345e5063ac8b57e
Adding api nat44_ed_vrf_tables_v2_dump which may replace
nat44_ed_vrf_tables_dump in the future.
- fixing endianess
Type: improvement
Signed-off-by: Daniel Béreš <daniel.beres@pantheon.tech>
Change-Id: I40d09ea3252589bdcb61db9f1629dacd87f69978
Some components, like dhcp, log constantly changing strings which in
turn forces elog string table to grow unbound.
To avoid this, as a workaround, only turn on elog logging if requested.
Actual fix that adds configuration for logging subclasses should come in
a later patch.
Type: improvement
Signed-off-by: Florin Coras <fcoras@cisco.com>
Change-Id: Ie8b26251fb7115d866c2bd65353daa33cdab1ab6
The punt socket code rewinds the current_data pointer by sizeof (ethernet_header_t),
which is incorrect if the header is tagged - resulting in truncated destination MAC
address. Use ethernet_buffer_header_size() instead, which takes tags into account.
Also add the unittest that verifies the issue and the fix.
Type: fix
Change-Id: I6352a174df144ca1e4230390c126f4b698724ebc
Signed-off-by: Andrew Yourtchenko <ayourtch@gmail.com>
Creation of lcp tap for non-ethernet interfaces can potentially lead to a crash, so avoid it.
Type: fix
Change-Id: I76ded8a08ea38a2c31d0215804af023207d4d3e1
Signed-off-by: Stanislav Zaikin <stanislav.zaikin@46labs.com>
Previously we encountered the issue of failing to create completion
queues on some Arm platforms because DPDK may set MLX5_CQE_SIZE to 128
if DPDK MLX PMDs are built and DPDK plugin is loaded, which does not
satisfy the requirement of 64B size CQE by RDMA plugin.
We fixed this issue in 844a0e8b0("always use 64 byte CQEs for MLX5"),
but some of CSIT test cases failed due to this code change. It turns out
that we don't need to specify compressed CQE mode for txq CQ because
RDMA tx doesn't have the code logic to handle compressed CQEs, which
might cause unexpected behavior if it is enabled.
Type: fix
Fixes: 844a0e8b0 ("always use 64 byte CQEs for MLX5")
Signed-off-by: Jieqiang Wang <jieqiang.wang@arm.com>
Change-Id: I7909a6d44b15bcf39c15dfac9377b65520a0cbfb
The time wheel should not be started in the loop while processing expired events.
can be set p->stop_timer_handle = ~0 to solve.
Type: fix
Signed-off-by: jinsh <jinsh11@chinatelecom.cn>
Change-Id: Ie9a4293f39f981f50d280b39a5d958d319ee2300
Signed-off-by: Matthew Smith <mgsmith@netgate.com>
Type: fix
The main loop populates a vector of suspended process nodes to dispatch
by calling TW (tw_timer_expire_timers_vec), which identifies expired
timers and appends the user handle for each one to the vector.
Subsequently, the vector is iterated and the process node corresponding
to each handle is dispatched. The vast majority of the time, the process
node will end up suspending itself again to wait for a new timer or
event.
Given a process node A whose timer has expired, between the point when
the timer expired and the point when A is dispatched and suspends itself
again, its stop_timer_handle contains a stale value.
If another process node B is dispatched before A is dispatched, it may
end up using the timer ID that A formerly used. If another process node
C is dispatched after B and before A and calls
vlib_process_signal_event() to signal A, the timer started by B can be
deleted by vlib_process_signal_event_helper().
After getting the vector of process node IDs for expired timers, reset
the stop_timer_handle on each of those nodes.
Change-Id: I266da438e76e1fc356016da0b9b4941efac1c28a
Signed-off-by: Matthew Smith <mgsmith@netgate.com>
Change of enums used in REPLY_MACRO() to appropriate one
for handlers:
-vl_api_nat44_ed_add_del_vrf_table_t_handler
-vl_api_nat44_ed_add_del_vrf_route_t_handler
Type: fix
Change-Id: I58e97817b1678da7c025c0d03a8b938a4e0f7b6c
Signed-off-by: Daniel Béreš <daniel.beres@pantheon.tech>
support with GTPv1 TEID added to the flow hash.
This can able to ECMP to PGW and parallelization.
Type: feature
Change-Id: I6f758579027caf6123831ef2db7afe17e424a6eb
Signed-off-by: Takeru Hayasaka <hayatake396@gmail.com>
Since connects can be done without a worker barrier, first
worker should flush connects to destination workers only
after session layer has a chance to fully initialize the
half-open session.
Type: fix
Signed-off-by: Florin Coras <fcoras@cisco.com>
Change-Id: I82fe0f0c7e520baa72fd380d0a43a76ebbd5f548
Vlib panic uses longjmp to exit main loop, but workers don't set main_loop_exit
field on initialization, so this jump corrupts registers and causes segfault.
There I add clib_warning and abort if longjmp context hasn't been set.
Type: fix
Signed-off-by: Mikhail Sokolovskiy <sokolmish@gmail.com>
Change-Id: I0d705f1f139c4083af75066aeb525964ed0aa202
Originally the name for each session pool is incorrectly prepared.
It doesn't have right length. It is not null terminated.
The fix corrects the name formatting for each session pool.
Type: fix
Signed-off-by: Gabriel Oginski <gabrielx.oginski@intel.com>
Change-Id: I67da3d64702ccb27a5907825528f8c95d91040bb
In case of UDP length errors in udp_local node, these errors are
being lost and incomplete header may be advanced by wrong offset.
Fix it with only full packets processing and explicit error set
otherwise. Also, optimize two buffer loop perfomance into fast
path with both buffers are ok and slow path with one or none.
Type: fix
Change-Id: I6b7edc3eb5593981e55d7ae20d753c0fd1549d86
Signed-off-by: Vladislav Grishenko <themiron@yandex-team.ru>
Originally the name for each session pool can be incorrect prepared.
The fix changes formatting for name for each session pool.
Type: fix
Signed-off-by: Gabriel Oginski <gabrielx.oginski@intel.com>
Change-Id: I42e0752f9f46c5a42524ec7b863a7c9dd3c23110
wrk->event_elts has 5 elements if no user events
Type: fix
Signed-off-by: Xiaoming Jiang <jiangxiaoming@outlook.com>
Change-Id: Ib38fab422304efc470e20ccb7121442f05bf8bf3
