Age | Commit message (Collapse) | Author | Files | Lines |
|
There was already a CLI command and a libvnet function
to set the keys on an existing IPsec tunnel interface.
Expose this via the API.
Change-Id: I226a9616f680fc022f04447177a2e8232690657f
Signed-off-by: Matthew Smith <mgsmith@netgate.com>
|
|
Change-Id: I5d5d4f22b6369d504455a644f73076d772fbcfb4
Signed-off-by: Neale Ranns <nranns@cisco.com>
|
|
With heavy traffic, tx code path may crash due to memory corruption
Thread 5 "vpp_wk_2" received signal SIGSEGV, Segmentation fault.
[Switching to Thread 0x7fff3995c700 (LWP 2505)]
0x00007ffff73675e8 in vhost_user_if_input (vm=0x7fffb5f5bf9c,
vum=0x7ffff7882a40 <vhost_user_main>, vui=0x7fffb65570c4, qid=0,
node=0x7fffb6577dac, mode=VNET_HW_INTERFACE_RX_MODE_POLLING)
at /home/sluong/vpp-master/vpp/build-data/../src/vnet/devices/virtio/vhost-user.c:1610
1610 bi_current = (vum->cpus[thread_index].rx_buffers)
[vum->cpus[thread_index].rx_buffers_len];
(gdb) p vum->cpus[thread_index].rx_buffers_len
$2 = 793212607
(gdb)
Apparently, some code accidentally wrote the bad value in rx_buffers_len.
rx_buffers_len should never be greater than 1024 since that is how many buffers
we request each time.
After debugging many hours, I discovered that the memory corruption happens
in the tx code path right here on line 2176.
{
vhost_copy_t *cpy = &vum->cpus[thread_index].copy[copy_len];
copy_len++;
cpy->len = bytes_left;
cpy->len = (cpy->len > buffer_len) ? buffer_len : cpy->len;
cpy->dst = buffer_map_addr;
cpy->src = (uword) vlib_buffer_get_current (current_b0) +
current_b0->current_length - bytes_left;
(gdb) p cpy
$3 = (vhost_copy_t *) 0x7fffb554077c
(gdb) p copy_len
$4 = 1025
(gdb) p &vum->cpus[3].rx_buffers_len
$8 = (u32 *) 0x7fffb5540784
copy_len is picking up the index entry 1024 before it was incremented. copy array has only
1024 members (0 - 1023 are valid).
The assignment here in cpy surely causes memory corruption. It is only discovered later
when the memory location that it corrupted is used.
The condition for the crash is to transmit jumbo frames under heavy volume. Since ring
size is 1024, with one packet taking up one index for frame size (less 2048), it does
not cause overflow. With jumbo frames, it requires multiple indices for one packet,
it can cause the overflow under heavy traffic.
The fix is to do copy out when we have 1000 entries in the array to avoid
overflow.
Change-Id: Iefbc739b8e80470f1cf13123113f8331ffcd0eb2
Signed-off-by: Steven <sluong@cisco.com>
|
|
Made Update to src/plugins/dpdk/buffer.c
Change-Id: I87bb8f38974a7be274c1b1d205f5513e7d068e48
Signed-off-by: Georgina <georgina.sheehan@intel.com>
|
|
It's way too easy to imagine leaving a mutex or a spin-lock held in
the /vpe-api shared-memory segment, or elsewhere. Set a volatile
variable and check it in a safe place...
Change-Id: I9d91c38cffeb921143c272162d055c9c24a6c312
Signed-off-by: Dave Barach <dave@barachs.net>
|
|
last i the serise of the use of the FIB table create/delete API. VPP now forces the tables to have been explicitly creted before they are used.
Change-Id: Ifde3b1bbb76697a01ab71bce4f5264e6d1725467
Signed-off-by: Neale Ranns <nranns@cisco.com>
|
|
INADDR6_ANY should be displayed as "::" instead of "0.0.0.0"(ipv4 format).
Change-Id: I24ec7b6febbfeca5db7ff894f455ecb73d954334
Signed-off-by: Steve Shin <jonshin@cisco.com>
|
|
This patch is a plausible first-cut, suitable for initial testing by
vcl (host stack client library).
Main features;
- recursive name resolution
- multiple ip4/ip6 name servers
- cache size limit enforcement
- currently limited to 65K
- ttl / aging
- static mapping support
- show / clear / debug CLI commands
Binary APIs provided for the following:
- add/delete name servers
- enable/disable the name cache
- resolve a name
To Do list:
- Respond to ip4/ip6 client DNS requests (vs. binary API requests)
- Perf / scale tuning
- map pending transaction ids to pool indices, so the cache
can (greatly) exceed 65K entries
- Security improvements
- Use unpredictable dns transaction IDs, related to previous item
- Make sure that response-packet src ip addresses match the server
- Add binary APIs
- deliver raw response data to clients
- control recursive name resolution
- Documentation
Change-Id: I48c373d5c05d7108ccd814d4055caf8c75ca10b7
Signed-off-by: Dave Barach <dave@barachs.net>
|
|
interface
From the troubleshooting perspective, it is nice to immediately know
the ACEs for the ACLs applied to an interface, so implement that.
To make the CLI more friendly, split each of the "show" variants
into an independent _cmd function with the distinct CLI path.
Change-Id: I519e4799083c04e8f0fcdf3e262a73493be4b690
Signed-off-by: Andrew Yourtchenko <ayourtch@gmail.com>
|
|
Add the subject enums to unblock LDPRELOAD. Just the enums,
no real implementation yet.
Change-Id: Ia3ec576c2779ee20956a37f0adebc06f16d1fe7f
Signed-off-by: Steven <sluong@cisco.com>
|
|
Change-Id: Ica3bc74ffbb1c0df4e198b0abff8df10cdeb2182
Signed-off-by: Sergio Gonzalez Monroy <sergio.gonzalez.monroy@intel.com>
|
|
- Management Address TLV is added as per IEEE Std 802.1AB-2009.
- Support of management ipv4/ipv6 addresses and OID.
Change-Id: I57c14741774390809ce5a829cc087947424432c7
Signed-off-by: Steve Shin <jonshin@cisco.com>
|
|
Change-Id: I011dda118f37cb31a37dda270027612d0af57ca0
Signed-off-by: Florin Coras <fcoras@cisco.com>
(cherry picked from commit 87f141172212b7568f519653ab32ebd1b5d34344)
|
|
1. Add the subject enum type
2. Fix a typo for the clib_warning
Change-Id: I59f6d15d51c66b96022d51592d65c41755c1582a
Signed-off-by: Steven <sluong@cisco.com>
|
|
Issues observed with specific kernel versions, e.g. stock Ubuntu 16.04
kernel.
This reverts commit 3eab064e3fadaf2a6a128f167ad04ca0319b4e17.
Change-Id: I24241f3b580df749fc686af3a319011ca035fb5e
Signed-off-by: Damjan Marion <damarion@cisco.com>
|
|
Change-Id: I72b0c063e89bf8299699dafec2a7e0395b86c8b9
Signed-off-by: Dave Wallace <dwallacelf@gmail.com>
|
|
Applications are now provided the option to select the namespace they
are to be attached to and the scope of their attachement. Application
namespaces are meant to:
1) constrain the scope of communication through the network by
association with source interfaces and/or fib tables that provide the
source ips to be used and limit the scope of routing
2) provide a namespace local scope to session layer communication, as
opposed to the global scope provided by 1). That is, sessions can be
established without assistance from transport and network layers.
Albeit, zero/local-host ip addresses must still be provided in session
establishment messages due to existing application idiosyncrasies. This
mode of communication uses shared-memory fifos (cut-through sessions)
exclusively.
If applications request no namespace, they are assigned to the default
one, which at its turn uses the default fib. Applications can request
access to both local and global scopes for a namespace. If no scope is
specified, session layer defaults to the global one.
When a sw_if_index is provided for a namespace, zero-ip (INADDR_ANY)
binds are converted to binds to the requested interface.
Change-Id: Ia0f660bbf7eec7f89673f75b4821fc7c3d58e3d1
Signed-off-by: Florin Coras <fcoras@cisco.com>
|
|
Change-Id: I5b48310c46ca8a2143b2132110240d7e9a52c25d
Signed-off-by: Damjan Marion <damarion@cisco.com>
|
|
1) In the previous version callbacks were generated based on
request-replay naming conventions. It turned out they were too
strict in case of events (e.g. BFD sends Details messages as
notifications). So now we generate callback for all messages,
allowing to receive any message as notification.(callback_gen.py)
2) "notification" suffix is no longer added because all messages
are treated same (dto_gen.py, jvpp_c_gen_.py)
3) name of property that holds notification/events changed in callback
facade and future apis
4) JVppNotification.java is no longer used since all events are treated
equally
Change-Id: I13f6438affc3473040d63cd4acb3984d03e97482
Signed-off-by: Matej <matej.perina@pantheon.tech>
|
|
- new IPv4 and IPv6 feature arcs on the punt and drop nodes
- new features:
- redirect punted traffic to an interface and nexthop
- police punted traffic.
Change-Id: I53be8bf4e06545add8a3619e462de5ffedd0a95c
Signed-off-by: Neale Ranns <nranns@cisco.com>
|
|
Change-Id: Ife87f9b00f918ff1bb8c91c6f13ebe53a3555a12
Signed-off-by: Ole Troan <ot@cisco.com>
|
|
Change-Id: I19fdf13a4848306ee3841d822b832cba96c5bce5
Signed-off-by: Neale Ranns <nranns@cisco.com>
|
|
Add one of these statements to foo.api:
vl_api_version 1.2.3
to generate a version tuple stanza in foo.api.h:
/****** Version tuple *****/
vl_api_version_tuple(foo, 1, 2, 3)
Change-Id: Ic514439e4677999daa8463a94f948f76b132ff15
Signed-off-by: Dave Barach <dave@barachs.net>
Signed-off-by: Ole Troan <ot@cisco.com>
|
|
Change-Id: I7c6911cd6ac366fe62675fd0ff8b0246a25ea1db
Signed-off-by: Matus Fabian <matfabia@cisco.com>
|
|
Change-Id: Ie3245b96c511cc30915e70e8c881f445291a38c2
Signed-off-by: Matus Fabian <matfabia@cisco.com>
|
|
A bug was reported where a jumbo packet would stay in vhost
queue forever or until a large enough number of other packets
arrived in the queue too.
This is due to a bug in vhost input node buffer allocation.
The fix is to make sure that vhost always allocates at least
enough buffers for one single big packet. '40' is used to
account for 65kB frames.
Change-Id: I1d293028854165083e30cd798fab9d4140230b78
Signed-off-by: Pierre Pfister <ppfister@cisco.com>
(cherry picked from commit 67700d41169ac37d21c400949a316750eabad969)
|
|
The top-level Makefile in VPP already uses the file /etc/os-release to
recognize which OS the build is being performed on.
The scripts for vagrant images instead were still using different files
for different distros and adding a new distro to be supported causes
issues with the extra-logic to deal with it.
This patch aligns the vagrant scripts to use the same approach followed
by the top-level Makefile and simplifies the steps to identify / update
the OS and build / run VPP.
Change-Id: I539f4a7c27b3fef70ed1c12b0276049bc47dc289
Signed-off-by: Marco Varlese <marco.varlese@suse.com>
|
|
- Coverity whines about a zero-length field not being initialized.
Change the struct setup to an initializer which will implicitly zero
all unused fields, and add the coverity notation that should stop
it whining. One or both of these should shut it up!
- Fix some incorrect use of ntohl that was tainting values; in these
cases htonl should have been used, and avoid a double-swap.
Change-Id: I00493a77eb23a0b8feb647165ee349e1e9d5cfdb
Signed-off-by: Chris Luke <chrisy@flirble.org>
|
|
Change-Id: Ife27795ea96919c116fb6ff33d106663b54df72d
Signed-off-by: Dave Wallace <dwallacelf@gmail.com>
|
|
https://gerrit.fd.io/r/#/c/8551/ decoupled the global variable,
namely tm->iovecs from TX and RX. However, to support multi-threads,
we have to eliminate the use of this global variable with per thread
variable. I notice that rx_buffers must also be per thread variable.
So, we introduce per thread struct to contain rx_buffers and iovecs.
Each thread will find the per thread struct with thread_index.
Change-Id: I61abf2fdace8d722525a382ac72f0d04a173b9ce
Signed-off-by: Steven <sluong@cisco.com>
|
|
Notes on this first implementation:
* First version of the implementation does NOT support GENEVE OPTIONS
HEADER: it isn't well understood what the purpose of the OPTIONS will be and/or
what content would be placed in the variable option data;
Once the IETF work will evolve and further information will be available
it could be possible to modify the frame rewrite to contemplate the
actual GENEVE OPTIONS.
Change-Id: Iddfe6f408cc45bb0800f00ce6a3e302e48a4ed52
Signed-off-by: Marco Varlese <marco.varlese@suse.com>
|
|
CP table: link_table
DP table: session_table
Change-Id: I2adbfd8f6a63d51d00d6dd291f32aebf20d13e4d
Signed-off-by: Hongjun Ni <hongjun.ni@intel.com>
|
|
This patch reworks the DPDK ipsec implementation including the cryptodev
management as well as replacing new cli commands for better usability.
For the data path:
- The dpdk-esp-encrypt-post node is not necessary anymore.
- IPv4 packets in the decrypt path are sent to ip4-input-no-checksum instead
of ip4-input.
The DPDK cryptodev cli commands are replaced by the following new commands:
- show dpdk crypto devices
- show dpdk crypto placement [verbose]
- set dpdk crypto placement (<device> <thread> | auto)
- clear dpdk crypto placement <device> [<thread>]
- show dpdk crypto pools
Change-Id: I47324517ede82d3e6e0e9f9c71c1a3433714b27b
Signed-off-by: Sergio Gonzalez Monroy <sergio.gonzalez.monroy@intel.com>
|
|
Add a primary svm_region_t pointer to the api_main_t so we can always
find the primary region, even when processing an API message from a
memfd segment.
Change-Id: I07fffe2ac1088ce44de10a34bc771ddc93af967d
Signed-off-by: Dave Barach <dave@barachs.net>
|
|
A distributed virtual router works by attmpeting to switch a packet, but on failing to find a local consumer (i.e. the packet is destined to a locally attached host) then the packet is sent unmodified 'upstream' to where the rest of the 'distributed' router is present. When L3 switching a packet this means the L2 header must not be modifed. This patch adds a 'l2-bridge' object to the L3 FIB which re-injects packets from the L3 path back into the L2 path - use with extreme caution.
Change-Id: I069724eb45956647d7980cbe40a80a788ee6ee82
Signed-off-by: Neale Ranns <nranns@cisco.com>
|
|
- always use 'va_args' as pointer in all format_* functions
- u32 for all 'indent' params as it's declaration was inconsistent
Change-Id: Ic5799309a6b104c9b50fec309cba789c8da99e79
Signed-off-by: Christophe Fontaine <christophe.fontaine@enea.com>
|
|
Change-Id: I7b175d57b85e626aab00221b6dac0498aebcbeae
Signed-off-by: Neale Ranns <nranns@cisco.com>
|
|
This allows us to have single contignuous allocation for DPDK buffers
with single mmap FD, so buffer memory can be easily shared with diffrent
process.
As a consequence dpdk socket-mem is no longer in charge for allocating
buffer memory, but still we need some space allocated for dpdk
structures so default socket-mem is reduced form 256 to 64 MB.
For a default of 16K buffers per numa node, physmem allocation is now
40MB, so basically this change reduces footprint from 256MB per socket
to 48 (64 + 40).
Change-Id: Ic8cfe83930a18411545b37a12b14aac89affd04f
Signed-off-by: Damjan Marion <damarion@cisco.com>
Signed-off-by: Sergio Gonzalez Monroy <sergio.gonzalez.monroy@intel.com>
Signed-off-by: Damjan Marion <damarion@cisco.com>
|
|
Change-Id: I58772a83e22885a9ea8a7a981d2bcb4b31a050d2
Signed-off-by: Neale Ranns <nranns@cisco.com>
|
|
Crash was seen on recent image with this BT on top of the stack
(gdb) bt full
(mif=0x7fffb6226568) at
/vpp/build-data/../src/plugins/memif/memif.c:297
ring = 0x0 <<<<<<<<<<
i = 0
j = 0
buffer_offset = 65792
r = 0x7fffb5e59f80
alloc = {flags = 1, name = 0x7fffb449f965 "memif region",
size = 4260096, numa_node = 0, addr = 0x7fff41dac000,
fd = 11,
log2_page_size = 12, n_pages = 1041}
err = 0x0
__FUNCTION__ = "memif_init_regions_and_queues"
The crash happened at this line.
ring = memif_get_ring (mif, MEMIF_RING_S2M, i);
ring=>head = ring->tail = 0; <=====
Please note that the crash is caused by dereferencing NULL rinng.
Put breakpoint into the function. I notice that
mif->regions[0].shm is not initialized.
(gdb) p mif->regions[0].shm
$8 = (void *) 0x0
It looks like we forgot to set shm after clib_mem_vm_ext_alloc().
Add the missing cide and the crash is fixed.
Change-Id: Ib722a6c241c77acfa8e33962106b57faa50e1ea7
Signed-off-by: Steven <sluong@cisco.com>
|
|
1. Add VNET headers support for checksumming - required
to operate correctly on any recent Linux
2. Bypass QDISC on transmit - improves performance by ~ 5%.
Enabled only if the macro is detected - apparently not
present on archaic distributions.
This still does not solve all issues with TSO - it can be
fixed only by going to tpacket v3 and dynamic rx ring as
well as significant changes in the TX (sendmmsg?).
Change-Id: Iea14ade12586c0a8da49e6dd1012108a08bc85b3
Signed-off-by: Anton Ivanov <anton.ivanov@cambridgegreys.com>
|
|
Add an API request message type to dump IPsec SAs. Either
all IPsec SAs can be dumped or it can be limited to a single
SA ID (numeric ID set at creation time - not an index).
Add a handler for incoming messages with the new request type.
Add an API response message type containing the data
for an IPsec SA.
Add VAT support for new message type.
Change-Id: Id7828d000efc637dee7f988a87d3f707a8b466b7
Signed-off-by: Matthew Smith <mgsmith@netgate.com>
|
|
Makes sure vlib_rp and shmem_hdr are initialized for internal
registrations. They are needed for keepalive msg exchanges.
Change-Id: I805dec2d2aa84b1efdc1fdd692fc1d94389b776e
Signed-off-by: Florin Coras <fcoras@cisco.com>
|
|
Change L2 learning path so it update stale timestamp in MAC entry
only if aging is enabled on the BD for the MAC entry.
Change-Id: I7babe986ceef3c030d8ef9185076c42b405f7b0f
Signed-off-by: John Lo <loj@cisco.com>
|
|
- add separate TIME_WAIT time constant
- fix output node for TIME_WAIT acks
- ensure snd_nxt is snd_una_max after retransmitting fin
- debugging improvements
Change-Id: Ic947153346979853f2526824b229126e47aead86
Signed-off-by: Florin Coras <fcoras@cisco.com>
|
|
Change-Id: I47f9d12d934378f18c6f841b902af2a64ee7b187
Signed-off-by: Matej Perina <mperina@cisco.com>
|
|
- Teach vpp_api_test to send/receive API messages over sockets
- Add memfd-based shared memory
- Add api messages to create memfd-based shared memory segments
- vpp_api_test supports both socket and shared memory segment connections
- vpp_api_test pivot from socket to shared memory API messaging
- add socket client support to libvlibclient.so
- dead client reaper sends ping messages, container-friendly
- dead client reaper falls back to kill (<pid>, 0) live checking
if e.g. a python app goes silent for tens of seconds
- handle ping messages in python client support code
- teach show api ring about pairwise shared-memory segments
- fix ip probing of already resolved destinations (VPP-998)
We'll need this work to implement proper host-stack client isolation
Change-Id: Ic23b65f75c854d0393d9a2e9d6b122a9551be769
Signed-off-by: Dave Barach <dave@barachs.net>
Signed-off-by: Dave Wallace <dwallacelf@gmail.com>
Signed-off-by: Florin Coras <fcoras@cisco.com>
|
|
Identify and complain when the same IP prefix is assigned
to two different SW interfaces:
vpp# set int ip address TenGigabitEthernet6/0/0 1.2.3.4/32
vpp# set int ip address TenGigabitEthernet6/0/1 1.2.3.4/32
set interface ip address: Prefix 1.2.3.4/32 already found on
interface TenGigabitEthernet6/0/0
Change-Id: I1aee1b6a7ddd00d3109a53d8e1b6ce97bf45e372
Signed-off-by: Jon Loeliger <jdl@netgate.com>
|
|
fixes an issue where events were not sent if BD doesn't enable mac aging
Change-Id: Iddc53cb5c45e560633e6c5cff2731dccfc70ad5b
Signed-off-by: Eyal Bari <ebari@cisco.com>
|
|
When making a call to vlib_packet_template_get_packet(), it
is possible to get back a NULL if the system runs out of buffer.
This can happen when there is buffer leaks. But don't crash
just because we run out of buffers, just punt.
Change-Id: Ie90ea41f3dda6e583d48959cbd18ff124158d7f8
Signed-off-by: Steven <sluong@cisco.com>
|