Age | Commit message (Collapse) | Author | Files | Lines |
|
Change-Id: Iab943988d1c714fe315e1dd13bd5d21f6bebdca1
Signed-off-by: Chris Luke <chrisy@flirble.org>
|
|
calculation
Change-Id: I62f625a93e5d818caef382316035cd5447bd8fef
Signed-off-by: Jakub Grajciar <Jakub.Grajciar@pantheon.tech>
|
|
Change-Id: I1326f21f0a00a201d2bdb55b73af14fca6ba8888
Signed-off-by: Klement Sekera <ksekera@cisco.com>
|
|
clib_mem_unaligned + zap64 casts its input as u64, computes a mask
according to the input length, and returns the casted maked value.
Therefore all the 8 Bytes of the u64 are systematically read, and
the invalid ones are discarded.
Since they are discarded correctly, this invalid read can safely be
ignored.
Revert "fix clib_mem_unaligned() invalid read"
This reverts commit 0ed3d81a5fa274283ae69b69a405c385189897d3.
Change-Id: I5cc33ad36063c414085636debe93707d9a75157a
Signed-off-by: Gabriel Ganne <gabriel.ganne@enea.com>
|
|
l2fib_make_key() casts and reads the input 6-Bytes mac_address as u64,
therefore if the mac_address is declared with 6 Bytes only, address-sanitizer
rightly triggers an invalid read on the last two Bytes.
However, l2fib_make_key() does a 16 bits shift to discard those 2 values,
therefore, this invalid read is of no consequence (and so can be silenced safely).
Change-Id: I38646fe60073093d25cdf135185d4c96136d55d0
Signed-off-by: Gabriel Ganne <gabriel.ganne@enea.com>
|
|
Change-Id: Ibcc20c24f6feb2b91245b0d88830a6c730d704e6
Signed-off-by: Florin Coras <fcoras@cisco.com>
|
|
Change-Id: Ided2980373ed5329c68f958f61be893428bccd31
Signed-off-by: Florin Coras <fcoras@cisco.com>
|
|
Change-Id: Id01a363bf2b574376651de7bc8f3f7b2bb58c615
Signed-off-by: Hongjun Ni <hongjun.ni@intel.com>
|
|
Change-Id: I68152d7338ce0d7805e50ccf9e9046de02cfd206
Signed-off-by: Florin Coras <fcoras@cisco.com>
|
|
- Add support to configure VCL to set
application proxy transport types
and app_is_proxy flag via vcl
config file or env vars.
- Clean up debug output.
- Rename session_scope -> app_scope.
Change-Id: I6e35f3cdd84daf0112133af8f33acd0152f87ca3
Signed-off-by: Dave Wallace <dwallacelf@gmail.com>
|
|
Change-Id: Ie57b81f8743f14182813558887d84d6667c81d43
Signed-off-by: Filip Tehlar <ftehlar@cisco.com>
|
|
Change-Id: I0b437ac5fecc81c7762d9cad0f33e977fcf3aa27
Signed-off-by: Neale Ranns <nranns@cisco.com>
|
|
Change-Id: I9f4d1c7ee7b460a93198930a5a935fa90177cdad
Signed-off-by: Florin Coras <fcoras@cisco.com>
|
|
Change-Id: I8d233d2301fb0da0fe27fa36ba870484bb290b7b
Signed-off-by: Florin Coras <fcoras@cisco.com>
|
|
Change-Id: I49b998644b8b79c778c1186fc09831b1cd8fc015
Signed-off-by: John DeNisco <jdenisco@cisco.com>
|
|
Change-Id: I3d3e5dff5b22fca58a50da6a9d0aaf1182e736dd
Signed-off-by: Ole Troan <ot@cisco.com>
|
|
Change-Id: I0b5806dd1d8cb45f40354cfe6cae7f4e76309f92
Signed-off-by: Damjan Marion <damarion@cisco.com>
|
|
Change-Id: I4b03a4f86a7e0e47874715398ca9f8ff0f5386ee
Signed-off-by: Neale Ranns <neale.ranns@cisco.com>
|
|
clib_mem_unaligned + zap64 casts its input as u64, computes a mask
according to the input length, and returns the casted maked value.
Therefore all the 8 Bytes of the u64 are systematically read, and
the invalid ones are discarded.
For example, for a 5-Bytes string, we will do an invalid read of size 3,
even though those 3 Bytes are never used.
This patch proposes to only read what we have at the cost of reading as
a u64 in one call, but that way, we do not trigger an invalid read
error.
Change-Id: I3e0b31c4113d9c8e53aa5fa3d3d396ec80f06a27
Signed-off-by: Gabriel Ganne <gabriel.ganne@enea.com>
|
|
warning found by clang:
warning: logical not is only applied to the left hand side of
this bitwise operator [-Wlogical-not-parentheses]
Change-Id: I964651a4444b11da145edc329da83675cd830f78
Signed-off-by: Gabriel Ganne <gabriel.ganne@enea.com>
|
|
Change-Id: I5d36a8a54fa6f73ea2b3761413098867dae1df74
Signed-off-by: Dave Barach <dave@barachs.net>
|
|
To enable this, applications set the proxy flag in their attach requests
and pass the transport protocols they want to act as proxies for as part
of the attach options.
When proxy is enabled, session rules that point incoming packets to the
proxy app are addedd to the local and global session tables, if these
scopes are accessible to the app. In particular, in case of the former,
the rule accepts packets from all sources and all ports destined to the
namespace's supporting interface address on any port. While in case of
the latter, a generic any destination and any port rule is addedd.
Change-Id: I791f8c1cc083350f02e26a2ac3bdbbfbfa19ece3
Signed-off-by: Florin Coras <fcoras@cisco.com>
|
|
Change-Id: I2feb3e07c3070e8a525c539dd2feffa0dd1bca21
Signed-off-by: Sergio Gonzalez Monroy <sergio.gonzalez.monroy@intel.com>
|
|
VPP-1034
Change-Id: I02b4db9e52446ab8578df1f011dd27f39de64c70
Signed-off-by: Sergio Gonzalez Monroy <sergio.gonzalez.monroy@intel.com>
|
|
The VOM is a C++ library for use by clients/agents of VPP for programming
state. It uses the binary APIs to do so. Various other common client side
functions are also provided. Please see om.hpp for a more detailed description.
Change-Id: Ib756bfe99817093815a9e26ccf464aa5583fc523
Signed-off-by: Neale Ranns <neale.ranns@cisco.com>
Co-authored-by: Mohsin Kazmi <sykazmi@cisco.com>
|
|
When adding a filter MAC entry, the default sw_if_index of -1
was incorrectly validated and rejected.
Change-Id: Id7f122b6269ea7c299a4335b05b748afaf01383c
Signed-off-by: John Lo <loj@cisco.com>
|
|
Store and pass MAC address as 6 byte u8 array instead of u64 to
make MAC address handling in set interface MAC endian neutral.
The previous API handler only works for little endian.
Change-Id: Ie4ec33a840bc5122ab1f17e25977e58f3466253b
Signed-off-by: John Lo <loj@cisco.com>
|
|
A bug in the decoder of messages when there was a non-array compound type.
The typical result was an error message from the struct library:
"error:unpack_from requires a buffer of at least 4 bytes"
Change-Id: Ie30fec6fc39b9f4177b54fa4adc4fc69674f0e12
Signed-off-by: Ole Troan <ot@cisco.com>
|
|
- update of CSIT operational branch to be used for VPP-patch test
Change-Id: I818bb7f1778cd92716feeaeb9fc1af59c62bdbc2
Signed-off-by: Jan Gelety <jgelety@cisco.com>
|
|
Change-Id: I25937cd7470c826d1e833e65530ae959c39139d8
Signed-off-by: Filip Tehlar <ftehlar@cisco.com>
|
|
As per proposal on the mailing list, this patch fixes
the represntation of MAC address in VPP API calls for
· L2fib_add_del
· L2_fib_table_details
Change-Id: I31e17efd1a6314cded69666e693cb8fc33158d02
Signed-off-by: Mohsin Kazmi <sykazmi@cisco.com>
|
|
Change-Id: If4a128186a732266eb27b6aa3caf6f1feff449bf
Signed-off-by: Dave Barach <dave@barachs.net>
|
|
VFIO file descriptor is extracted in a hackish way, as DPDK doesn't
provide imethod to retrieve it.
This fixes issue with DPDK drivers not working correctly when IOMMU is
enabled and external buffer memory is used.
Change-Id: I5eaa8e78741c50504d87d44e706b5997c8189554
Signed-off-by: Damjan Marion <damarion@cisco.com>
|
|
Change-Id: I7aafdecd6f370411138e6ab67b2ff72cda6e0666
Signed-off-by: Neale Ranns <nranns@cisco.com>
|
|
Add new cli api: "test lb flowtable flush" which flushes everything.
Call this new cli function after the end of each lb unit test.
Change-Id: I71d04a7bfba398f7d4dd9cc3ed24bba786943663
Signed-off-by: Gabriel Ganne <gabriel.ganne@enea.com>
|
|
Change-Id: I7fe34e99f566dd2e0a36a9ba38f894973989ca8a
Signed-off-by: Jon Loeliger <jdl@netgate.com>
|
|
This is an all-purpose octet-vector memory hash, intended as a
thread-safe replacement for hash_create_mem / hash_create_string. All
u8 * key vectors are memorized by the hash table.
Change-Id: I22944daea8fda07dde8ba118a6529a6d382491f9
Signed-off-by: Dave Barach <dave@barachs.net>
|
|
- The binary API method for executing commands was superceded
by the unix-socket CLI approach used by the C vppctl utility.
The API call it makes will soon be deprecated.
- This Python vppctl script was still being copied to the install
directory, and then later overwritten by the binary vppctl.
Change-Id: I0a4b09b7c733049d0545b4c83b9ac5ac499a7300
Signed-off-by: Chris Luke <chrisy@flirble.org>
|
|
Change-Id: Ifc0296834e25ddbdd0ad8283c061f309801b053c
Signed-off-by: Filip Tehlar <ftehlar@cisco.com>
|
|
- message sending method inside synchronization blocks causes
deadlock between sending and receiving part
- breaking atomicity of sending message and putting future with
corresponding id to map needs additional handling by writer and receiver,
regardless which part get access to sync block first will create
new future and second one will complete it and remove from map,
in case of dump calls where control ping reply is required
as confirmation that all information were send, if ping reply is
received before writer put future in map, reader will create
regular control ping future instead and writer needs to made association
between these two futures
Change-Id: Id29a19be7a5319291a5e07cf931080610178f00c
Signed-off-by: Matej <matej.perina@pantheon.tech>
|
|
Change-Id: Ie3bd75e0e98c7ecb4c66dba701131363e5ec89a2
Signed-off-by: Jakub Grajciar <Jakub.Grajciar@pantheon.tech>
|
|
Build fails on ubuntu 17.10...
Change-Id: I98dfffa3f8f5f9bca2ba97a856709714fbeec49c
Signed-off-by: Damjan Marion <damarion@cisco.com>
|
|
Also add vat test code to test the subject API. The format is
sw_interface_set_rx_mode sw_if_index <index> [queue <id>]
<polling|nterrupt|adaptive>
Change-Id: Ib810d85d430077865bead8cc08a070f8ae478225
Signed-off-by: Steven <sluong@cisco.com>
|
|
Change-Id: Ib87eccb853cafceea5f5513f6bb51c2364449afa
Signed-off-by: Florin Coras <fcoras@cisco.com>
|
|
This introduces 5-tuple lookup tables that may be used to implement
custom session layer actions at connection establishment time (session
layer perspective).
The rules table build mask-match-action lookup trees that for a given
5-tuple key return the action for the first longest match. If rules
overlap, ordering is established by tuple longest match with the
following descending priority: remote ip, local ip, remote port, local
port.
At this time, the only match action supported is to forward packets to
the application identified by the action.
Change-Id: Icbade6fac720fa3979820d50cd7d6137f8b635c3
Signed-off-by: Florin Coras <fcoras@cisco.com>
|
|
- Interactive commands like "ping" read extra input from the
input stream.
- In the case of "ping" it is simply a signal to cease the current
operation.
- "vppctl", in non-interactive mode, will issue a "quit" immediately
after the requested command to queue up closing of the session.
- This resulted in "ping" thinking a keypress was seen and returning
control to the CLI; the "quit" command however is consumed by the
keypress event handler and thus the session does not close.
- This patch reworks vppctl slightly to only issue "quit" after the
command has completed. In particular it uses the fact that VPP issues
NUL bytes as a surrogate prompt between output of commands to signal
acknowledgement that the command has completed; vppctl now flags
that the quit should be issued after the next such acknowledgement.
- Since input it still accepted, the user can still terminate the
"ping" early, if desired.
Change-Id: I7e3dbe767f32f8e364ccb5f81799759b311585df
Signed-off-by: Chris Luke <chrisy@flirble.org>
|
|
- Pointer to local outside scope
Change-Id: I84a7958e10dbc9369ffa5172cbad5dbade9b2bc9
Signed-off-by: Dave Wallace <dwallacelf@gmail.com>
|
|
Change-Id: I39d634b7691a524e5221c28997a737102298c281
Signed-off-by: Florin Coras <fcoras@cisco.com>
|
|
>>> CID 178271: Memory - illegal accesses (RETURN_LOCAL)
>>> Using "ep", which points to an out-of-scope variable "_ep".
Change-Id: Icf21e602bc2718bfedd9480e9da41aff8693048e
Signed-off-by: Steven <sluong@cisco.com>
|
|
MACIP ACLs
The classifier tables upper bound of memory was just big enough
to cause the unittests pass most of the time but not always.
Increase the amount of space and run several hundred iterations
of unittests to ensure they always pass.
Change-Id: Ieb7876c6ebdde1f8c5273dbb9b090f12f2c38915
Signed-off-by: Andrew Yourtchenko <ayourtch@gmail.com>
|