Age | Commit message (Collapse) | Author | Files | Lines |
|
Type: fix
Signed-off-by: Florin Coras <fcoras@cisco.com>
Change-Id: Iec1fe8315a057214901250d5fb06d1c1e33dda46
|
|
Type: fix
Signed-off-by: Florin Coras <fcoras@cisco.com>
Change-Id: I575068df56623f24c0290fa42b8d301548a089d4
|
|
Type: feature
Signed-off-by: Florin Coras <fcoras@cisco.com>
Change-Id: I2d5039cd35edd02ffa2439bcac119d0647234385
|
|
Type: fix
Signed-off-by: Stanislav Zaikin <zstaseg@gmail.com>
Change-Id: Ib4957bfc03b0dfc90fd28689d5b32cc3d82e7b74
|
|
Type: fix
Currently, is_interface_addr() tests if a given IPv4 address belongs to
an interface by a given sw_if_index. However, there are several issues:
* only the first found address on the interface is actually tested,
* sw_if_index is always cached even if the interface hasn't been
assigned any addresses yet.
With this change, is_interface_addr() tests all IPv4 addresses on an
interface by a given sw_if_index and caches sw_if_index only if there
are addresses present.
Signed-off-by: Alexander Chernavin <achernavin@netgate.com>
Change-Id: If1acc4a534647a5f0ce8e9b565b867c92a016dc3
|
|
Type: fix
Currently, there might be a crash in NAT ED mode:
* if a session for an unknown proto packet cannot be created in2out,
* if a session for an unknown proto packet cannot be found out2in.
With this change, translate packets only if a session is given in NAT
ED mode.
Signed-off-by: Alexander Chernavin <achernavin@netgate.com>
Change-Id: Iafb332db1ae9a3e76435964ad636037d1b8a51e8
|
|
sm_directory_vector_by_name is a C-string hashtable, make sure the names
are NULL-terminated.
Type: fix
Change-Id: If44456ef9da69ed012f2b981b58ce732663af67d
Signed-off-by: Benoît Ganne <bganne@cisco.com>
|
|
Type: improvement
Change-Id: I4b9b2be8817be10e46accc19219deb2b544f266b
Signed-off-by: Filip Tehlar <ftehlar@cisco.com>
|
|
Length check must also take current_data into account.
Type: fix
Change-Id: I7a1b1752868892d40f59490d05452ef24565cca6
Signed-off-by: Benoît Ganne <bganne@cisco.com>
|
|
In the current implement, tcp would start or up an one tick
retransmit timer for that connection if vlib_buffer_alloc()
return 0. Now the tick is 0.1ms, this means that if VPP is
in a buffer shortage state, there would be a large number of
burst timer expirations.
This commit limits the minimum interval of the retransmission
timer to 100ms.
Type: fix
Signed-off-by: liuyacan <liuyacan@corp.netease.com>
Change-Id: Ia11d693fe46119c5dc16b24ca93c30c31109057a
|
|
use autogenerated code
Type: improvement
Signed-off-by: Filip Tehlar <ftehlar@cisco.com>
Change-Id: I91a55412f96b138c3c00cbb8943d271c8a6452c9
|
|
This commit does following:
- Change the behavior of shutdown() with SHUT_RDWR flag.
- Check SHUT_RD flag when read()
- Change the errno when write() after SHUT_WR
- Remove unused code
All the above modification passed the packetdrill test.
Type: improvement
Signed-off-by: liuyacan <liuyacan@corp.netease.com>
Change-Id: I0c81f52e563562e58580d70976526b898e65e915
|
|
Type: test
Signed-off-by: Neale Ranns <neale@graphiant.com>
Change-Id: Iec69d8624b15766ed65e7d09777819d2242dee17
|
|
Type: feature
this allows VPP to simulate linux tun devices.
Signed-off-by: Neale Ranns <neale@graphiant.com>
Change-Id: I3adf38b49a254804370f78edd5d275d192fd00a6
|
|
When recycling a debug CLI process node, unix_cli_file_add() needs to
delete and recreate the related node_by_name hash table entry.
Type: fix
Signed-off-by: Dave Barach <dave@barachs.net>
Change-Id: I635da4918509d5b22eae37627c2d9b3608380ca6
|
|
Type: fix
Signed-off-by: Florin Coras <fcoras@cisco.com>
Change-Id: Ied2608e7a28c59c908803ca676abbe93072fadb8
|
|
Type: fix
Signed-off-by: Florin Coras <fcoras@cisco.com>
Change-Id: I338e61654a62ed6308ecd8bb15e1a8b13cd859b9
|
|
TCP and (D)TLS clean up half-opens on main without a lock/barrier so
cleanup initiated from first worker, e.g., cut-throughs, can corrupt the
session pool.
Type: fix
Signed-off-by: Florin Coras <fcoras@cisco.com>
Change-Id: I2e5162831c0e201b22454f17fe55bfac44b85fa9
|
|
Type: fix
This prevents reordering when a push flag is received.
GRO appends the segment with the push flag to the existing
flow and flushes it immediately.
Change-Id: I61b36209b3381f340594a9cb3ed816d43b02bdff
Signed-off-by: Aloys Augustin <aloaugus@cisco.com>
Signed-off-by: Mohsin Kazmi <sykazmi@cisco.com>
|
|
Type: fix
Change-Id: I456cc0b0a6f2dc32b14791baf9d4a7f67279e8df
Signed-off-by: Damjan Marion <damarion@cisco.com>
|
|
Type: fix
Signed-off-by: Florin Coras <fcoras@cisco.com>
Change-Id: I0971c18a183319d09486459b0da08d44ce38acef
|
|
when the configuration of tx queues is larger than the worker-threads,
the clib_spinlock_lock_if_init will not be executed, and then this function
will executed the clib_spinlock_unlock_if_init, so this may caused the issue.
Type: fix
Signed-off-by: fangtong <fangtong2007@163.com>
Change-Id: I3ce244cd5e1f410e9f14bd060b929238f069b9fa
|
|
VRRP prefix length is 5 bytes, doesn't make sense
to compare with 6 bytes mac address
Type: fix
Signed-off-by: Tianyu Li <tianyu.li@arm.com>
Change-Id: I70a9c9ca769f34a81cd568f9e6b9a3538e6ec710
|
|
Type: fix
Change-Id: Id6406702061ada54e51ebcf5c367328605de2b6b
Signed-off-by: Nathan Skrzypczak <nathan.skrzypczak@gmail.com>
|
|
The path pool can expand during in fib_path_attached_next_hop_get_adj()
when calling adj_nbr_add_or_lock(). If dpo points to a path->fp_dpo, its
reference becomes stale.
Use a temporary copy instead.
Type: fix
Change-Id: Ie966cb5f3f7b416425964dca12f1f586bfc2010c
Signed-off-by: Benoît Ganne <bganne@cisco.com>
|
|
The adj can be deleted during fib_walk_sync(), make sure it can happen
only after clearing the SYNC_WALK_ACTIVE flag.
Type: fix
Change-Id: I68be00e9602e2783d9dced71c51547c38b7e8a00
Signed-off-by: Benoît Ganne <bganne@cisco.com>
|
|
Type: fix
If an async crypto frame is allocated during ESP encrypt/decrypt but
a buffer/op is not subsequently added to the frame, the frame leaks. It
is not submitted if the count of async ops is zero nor is it
returned to the frame pool. This happens frequently if >= 2 worker
threads are configured and a vector of buffers all have to be handed
off to other threads.
Wait until it is almost certain that the buffer will be added to the
frame before allocating the frame to make it more unlikely that an
allocated frame will not have any operations added to it.
For encrypt this is sufficient to ressolve the leak. For decrypt there
is still a chance that the buffer will fail to be added to the frame, so
remove the counter of async ops and ensure that all frames that were
allocated get either submitted or freed at the end.
Change-Id: I4778c3265359b192d8a88ab9f8c53519d46285a2
Signed-off-by: Matthew Smith <mgsmith@netgate.com>
|
|
Test added to the unittest plugin / test_vlib.py
Type: improvement
Signed-off-by: Dave Barach <dave@barachs.net>
Change-Id: I73445e57918347c102ff6f5e8c9ddb9bd96f1407
|
|
Type: fix
Signed-off-by: Florin Coras <fcoras@cisco.com>
Change-Id: I84985f59274e77219a38ea6ac865fc73ac635d72
|
|
Only try once and return what was found.
Type: fix
Signed-off-by: Florin Coras <fcoras@cisco.com>
Change-Id: I50b1d26babf1b7431d36f6b7472a1fb01475bb35
|
|
Type: fix
Change-Id: I58931e235535be7d596ca578790f389b64a4fbd2
Signed-off-by: Simon Zhang <yuwei1.zhang@intel.com>
|
|
Type: style
reduce the number of files recompiled after changing pg.h from 1110 to
102.
Signed-off-by: Neale Ranns <neale@graphiant.com>
Change-Id: I50611eba818eeb3a2dffd437a3c72c77766bed80
|
|
- return VPPCOM_EEXIST if attempting to re-add a session
- return VPPCOM_ENOENT if the session to be removed is not epolled
- generate EPOLLIN if adding it through a mod operation on a session
that has data and did not have the event previously set.
Type: fix
Signed-off-by: Florin Coras <fcoras@cisco.com>
Change-Id: I728a06b8cf84af8d8c1dea7406e284de8886dffc
|
|
When both chained and non-chained buffers are processed in the same
vector, make sure the non-chained buffers are processed as non-chained
crypto ops.
Type: fix
Change-Id: I19fc02c25a0d5e2e8a1342e2b88bbae3fe92862f
Signed-off-by: Benoît Ganne <bganne@cisco.com>
|
|
Type: fix
Signed-off-by: Piotr Bronowski <piotrx.bronowski@intel.com>
Change-Id: I76923ad5035498aae821db4fd42a127617476fbb
|
|
handle
an exception
Type: fix
Signed-off-by: jinshaohui <jinsh11@chinatelecom.cn>
Change-Id: I67b7d0b52c33a5b13ace8fe2d918139d2820e9bf
|
|
Type: fix
Signed-off-by: Florin Coras <fcoras@cisco.com>
Change-Id: I7f5887865aa29c9881b5063be1d7d42369a493c1
|
|
Type: fix
Signed-off-by: Florin Coras <fcoras@cisco.com>
Signed-off-by: Dave Wallace <dwallacelf@gmail.com>
Change-Id: Ifad679f46abd6e9c18a3eaf7e55800a09f3791ab
|
|
Type: fix
Signed-off-by: Florin Coras <fcoras@cisco.com>
Change-Id: I99af136ecab9be1f9e00de6d197b8f1c74ab4b20
|
|
Type: improvement
Signed-off-by: Florin Coras <fcoras@cisco.com>
Change-Id: Ia46b0b8afed30f84b244c06f0457303f9e8832cd
|
|
Type: fix
Outbound packets from the host have an adjacency lookup performed so
buffer metadata can be filled in and output features can be applied.
If no adjacency is found for a packet, it gets dropped. This breaks
DHCP and possibly other things since the DHCP reply to a discover
request is sent to a destination MAC address that is determined by
the contents of the request packet rather than any existing neighbor
table entry.
If adjacency lookup for outbound packets does not find an entry, use
the default adjacency for the interface & address family.
Change-Id: Ia53a4df3a5bad2991768cfe4a84c560b879e656f
Signed-off-by: Matthew Smith <mgsmith@netgate.com>
|
|
When adding a route via a udp encap instance, FIB_ENTRY_FLAG_IMPORT
should not be set. In particular, fib_route_attached_cross_table should
always return false for such paths.
Modified test_udp_encap to leverage the bug that needed to be fixed.
Type: fix
Signed-off-by: Arthur de Kerhor <arthurdekerhor@gmail.com>
Change-Id: Iaa9489e96d1cff09751f92c62caf7999d924fd7f
|
|
Make sure we don't return error if app tries to setsockopt it.
Type: improvement
Signed-off-by: Florin Coras <fcoras@cisco.com>
Change-Id: I1a9a6a8da03380b0b6a6d01efed9b672a3f0b1f9
|
|
Type: fix
Signed-off-by: Florin Coras <fcoras@cisco.com>
Change-Id: I62fb56257445a05105e556d1ea6cc6280b5eeccc
|
|
We do not want to encap headers containing wrong checksums.
Additionnally, this clears the checksums offlads flags, which
was something missing since the outer headers checksums were
calculated during the encap. Hence, those should not be
recalculated afterwards.
Type: fix
Change-Id: I7fd07987b4f13f76c6990a1c08dc2f960bdd8de1
Signed-off-by: Arthur de Kerhor <arthurdekerhor@gmail.com>
|
|
snat_set_frame_queue_nelts has been replaced with
nat44_ed_set_frame_queue_nelts.
Type: fix
Signed-off-by: Ruslan Babayev <ruslan@babayev.com>
Change-Id: I8d970be71376fdbb2bfd383d4d5824a8def93bb3
|
|
Unless a software interface is actually unnumbered, do not set
ip[46]_main.lookup_main.if_address_pool_index_by_sw_if_index [sw_if_index]
to ~0
Fixes this scenario:
loop create
set int state loop0 up
create sub-interface loop0 1
set interface ip addr loop0.1 192.168.1.1/24
delete sub-interface loop0.1
set int ip addr loop0 192.168.1.1/24
Type: fix
Signed-off-by: Dave Barach <dave@barachs.net>
Change-Id: I46141d862fa57d70b93d7bb0c105403708165264
|
|
Type: fix
Signed-off-by: Florin Coras <fcoras@cisco.com>
Change-Id: Ifbf43dd69aa07be485fe7fa01c917512e423036e
|
|
Possibility to register a port via CLI or API to decap incoming UDP
packets:
- For CLI, a user needs to specify the inner protocol (only MPLS
supported for now)
- For API, the protocol is specified by index
Added unittests
Type: feature
Change-Id: Ifedd86d8db2e355b7618472554fd67d77a13a4aa
Signed-off-by: Arthur de Kerhor <arthurdekerhor@gmail.com>
|
|
Type: fix
Signed-off-by: Florin Coras <fcoras@cisco.com>
Change-Id: I59218f784140ef5ecb8468af6aebfe25fd7703a4
|