Vxlan evpn p2mp changes for Layer2 functionality #2

dgsudharsan · 2021-06-30T18:26:05Z

What I did
VXLAN EVPN changes to support layer 2 forwarding using P2MP tunnels based on SAI capabilities.

Why I did it
Certain platforms support P2P tunnels and while few use P2MP tunnels for L2 forwarding.

How I verified it
Updated VS test cases to cover this scenario

Details if related

…nt (sonic-net#1761) * Fix mirror session ref count at acl rule attachement Signed-off-by: wenda.ni <wenda.ni@bytedance.com>

Recirc port support has been added to recent versions of DNX SAI. The earlier special handling for recirc port is not required anymore and thus reverted.

What I did Add failure handling for SAI get operations. The function allows handling failures in SAI get operations according to the orch type, SAI type, SAI status. Why I did it Enable custom failure handling for SAI get operations.

…de (sonic-net#1769) *Before puting port related configuration to APPL_DB, check autoneg mode first. If autoneg mode is enabled, "speed" and "interface_type" will not be put into APPL_DB; if autoneg mode is disabled, "adv_speeds" and "adv_interface_types" will not be put into APPL_DB; else all configuration will be put to APPL_DB for backward compatible.

…ic-net#1797) In scenario that is executed in sonic-mgmt in test_po_update.py a portchannel member is deleted from one portchannel and added to another portchannel. It is possible that requests from teamsynd will arrive in different order This reordering happens because teamsyncd has single event handler/selectable TeamSync::TeamPortSync::onChange() per team device so when two of them are ready it is swss::Select implementation detail in which order they are going to be returned. This is a fundamental issue of Producer/ConsumerStateTable, thus orchagent must be aware of this and treat it as normal situation and figure out the right order and not crash or print an errors. - What I did Check if port is already a lag member beforehand. Added an UT to cover this scenario, this UT verifies that SAI API is not called in this case. Refactored portsorch_ut.cpp by moving out Orchs creation/deletion into SetUp()/TearDown() - Why I did it To fix errors in log. - How I verified it Ran test_po_update.py test. Signed-off-by: Stepan Blyschak stepanb@nvidia.com

What I did Ignore ALREADY_EXIST error in FDB creation. Fix: sonic-net/sonic-buildimage#7798 Why I did it In FDB creation, there are scenarios where the hardware learns an FDB entry before orchagent. In such cases, the FDB SAI creation would report the status of SAI_STATUS_ITEM_ALREADY_EXISTS, and orchagent should ignore the error and treat it as entry was explicitly created.

What I did Rename XPN attributes from SAI_MACSEC_SA_ATTR_MINIMUM_XPN and SAI_MACSEC_SA_ATTR_XPN to SAI_MACSEC_SA_ATTR_CURRENT_XPN Why I did it Due to opencomputeproject/SAI#1169 that refactors the attributes about XPN, a new attributes SAI_MACSEC_SA_ATTR_CURRENT_XPN for both ingress and egress was introduced for reading the Packet number. So move the original attributes to the new one. Signed-off-by: Ze Gan <ganze718@gmail.com>

Remove Portsyncd error exit if there are no ports in config DB. Signed-off-by: liora <liora@nvidia.com>

sonic-net#1789) * [debugcounterorch] check if counter type is supported before querying object availability count Signed-off-by: Stepan Blyschak <stepanb@nvidia.com>

…pt (sonic-net#1824) **What I did** Add the option of passing custmol built libnl path to configure script **Why I did it** MPLS feature in sonic-buildimage requires a libnl patch to be applied before building libnl. Since this build is not installed in usual locations (/usr/lib/..) LGTM analysis fails. This change gives the option of passing libnl library location to 'configure' script and generate libraries to be linked. In case the options are not passed, the configure script defaults to earlier behavior where it checks for LIBNL in usual locations

*dos2unix fdbsyncd/* Signed-off-by: Sebastian Hasler <sebastian.hasler@stuvus.uni-stuttgart.de>

* Purge swss-common before installing

Signed-off-by: Neetha John <nejo@microsoft.com> By default, lossless profiles are attached to PGs 3, 4 of all ports which results in a lot of buffer wastage when most of those ports are unused (not associated with a DEVICE_NEIGHBOR). In TD2, ingress pool size comprises of the reserved space as well. Hence making use of a special cable length of '0m' to identify unused ports and skip reserving space for those ports What I did * In buffermgr, if port with cable len '0m' is identified, return immediately without creating pg lossless profile or attaching profile to the lossless pgs of that port * Listen to 'admin_status' update as well from the PORT table and update port-speed mapping. This is to handle add-rack scenario where a port is added later - The port starts of with cable length 0m - Configlet to add a port is applied. The order of operations specifc to the PORT/CABLE_LENGTH table are - port is initially set to admin down, cable length is updated for that port, port table attributes are defined and port is set to admin up - speed update might not be seen when the port is set to admin up. Hence port-speed mapping will capture the speed update whenever its seen and once the cable length is updated while the port is brought back up, profiles can be attached to the lossless pgs of the port How I verified it - Manual tests done. Verified that no space is reserved for unused ports - Verified that when a port is added using 'add-rack' scenario, profile is attached to pgs of that port - New VS test added

SAI_REDIS_SWITCH_ATTR_CONTEXT was set according to the phy_id in gearbox_config.json. This change will set it based on the context_id provided in that json file. The reason behind it is to support the case of multiple phys sharing the same context.

…or active ports without speed configured (sonic-net#1822) **What I did** Bugfix: Don't create lossless buffer profiles for active ports without speed configured Root cause: - In `handlePortTableUpdate`, `refreshPgsFromPort` is called if admin status is updated even if the speed is not configured. This is reasonable because the port can be configured as headroom override and the profile should be applied in that case. - However, as a side-effect, the port's state is set to `PORT_READY` in `refreshPgsForPort` regardless of whether the speed is configured, which is not correct. This should be avoided and `PORT_READY` should be set by the caller if necessary Fix: - Don't set the port's state to `PORT_READY` in `refreshPgsForPort` and check the port's state before calling it. - Explicitly set port's state to `PORT_READY` in `handlePortStateTable`. This has been done by all other callers. Note: - The change is covered by the existing vs tests. - The scenario where the bug was originally found can not be covered by vs test because: - The speed is always configured in vs image by default - Removing speed is not handled in buffer manager since it's not a valid flow. - A regression test will be opened to cover this case.

*Set bridge mac same as Vlan mac *Format state_db entries to have six digit precision for microseconds

What I did Refactored cfgmgr Makefile Why I did it To avoid code duplication of common libs How I verified it Build swss Signed-off-by: Nazarii Hnydyn <nazariig@nvidia.com>

*Added more appropriate log message for CRM queries of MPLS_INSEG and MPLS_NEXTHOP resources when platform does not support these resource types.

orchagent/portsorch.h

orchagent/vxlanorch.cpp

* Mclag enhacements support code changes. * Adding change to allow MCLAG remote MAC move. * Added support for adding mclag remote mac to kernel, on top of PR-1276 * Updating the change from PR1276 and PR885. * Adding new orchfiles to mock_tests * MCLAG Unique IP support changes. * Removed dependency with PR 885. * Adding observer support for mlagorch. * Fixed FDB notifiation issue * Fixing the test_mclag_fdb type attributes. * Remove as the change may not be supported on non-brcm for PortChannel settings. * Removing the isolation group handling from Mlagorch, Isolation group now will be added/updated only via mclagsyncd updates. * Added back the update function. Co-authored-by: Tapash Das <tapash.das@broadcom.com>

orchagent/vxlanorch.cpp

tests/test_evpn_fdb_p2mp.py

*Open record file (swss.rec) in append mode. Similar to sairedis.rec

* Code changes to support IPv6 Link local enchancements - Changes to handle the "ipv6_use_link_local_only" option on a interface which is added/deleted based on Global Config or Interface config. - Adds/Deletes the RIF for an interface in the absence of IPv6 global address on a interface based on "ipv6_use_link_local_only". - Allow the ipv6 link-local address as neighbors. - Allow adding the IPv6 routes with link-local nexthops. - Skip Ipv4LL neighbor add in Orchagent for RFC5549. Signed-off-by: Akhilesh Samineni <akhilesh.samineni@broadcom.com>

…c-net#1823) VOQ nexthop for remote neighbors should be created on local inband port only for the kernel purpose. SAI should use actual RIF of the remote system port interface. sonic-net#1686 seems to be break this condition and this change address it.

…p2mp

* [cbf] Added initial CBF support (#2) * Added CBF NHG support to NhgOrch * Added NhgMapOrch to handle DSCP_TO_FC and EXP_TO_FC tables * Added UT for the new NhgOrch function and NhgMapOrch Support sonic-net/SONiC#796 Co-authored-by: Alexandru Banu <v-albanu@microsoft.com>

Currently, ASAN sometimes reports the BufferOrch::m_buffer_type_maps and QosOrch::m_qos_maps as leaked. However, their lifetime is the lifetime of a process so they are not really 'leaked'. This also adds a simple way to add more suppressions later if required. Example of ASAN report: Direct leak of 48 byte(s) in 1 object(s) allocated from: #0 0x7f96aa952d30 in operator new(unsigned long) (/usr/lib/x86_64-linux-gnu/libasan.so.5+0xead30) #1 0x55ca1da9f789 in __static_initialization_and_destruction_0 /__w/2/s/orchagent/bufferorch.cpp:39 #2 0x55ca1daa02af in _GLOBAL__sub_I_bufferorch.cpp /__w/2/s/orchagent/bufferorch.cpp:1321 #3 0x55ca1e2a9cd4 (/usr/bin/orchagent+0xe89cd4) Direct leak of 48 byte(s) in 1 object(s) allocated from: #0 0x7f96aa952d30 in operator new(unsigned long) (/usr/lib/x86_64-linux-gnu/libasan.so.5+0xead30) #1 0x55ca1da6d2da in __static_initialization_and_destruction_0 /__w/2/s/orchagent/qosorch.cpp:80 #2 0x55ca1da6ecf2 in _GLOBAL__sub_I_qosorch.cpp /__w/2/s/orchagent/qosorch.cpp:2000 #3 0x55ca1e2a9cd4 (/usr/bin/orchagent+0xe89cd4) - What I did Added an lsan suppression config with static variable leak suppression - Why I did it To suppress ASAN false positives - How I verified it Run a test that produces the static variable leaks report and checked that report has these leaks suppressed. Signed-off-by: Yakiv Huryk <yhuryk@nvidia.com>

**What I did** Fix the Mem Leak by moving the raw pointers in type_maps to use smart pointers **Why I did it** ``` Indirect leak of 83776 byte(s) in 476 object(s) allocated from: #0 0x7f0a2a414647 in operator new(unsigned long) ../../../../src/libsanitizer/asan/asan_new_delete.cpp:99 #1 0x5555590cc923 in __gnu_cxx::new_allocator, std::allocator > const, referenced_object> > >::allocate(unsigned long, void const*) /usr/include/c++/10/ext/new_allocator.h:115 #2 0x5555590cc923 in std::allocator_traits, std::allocator > const, referenced_object> > > >::allocate(std::allocator, std::allocator > const, referenced_object> > >&, unsigned long) /usr/include/c++/10/bits/alloc_traits.h:460 #3 0x5555590cc923 in std::_Rb_tree, std::allocator >, std::pair, std::allocator > const, referenced_object>, std::_Select1st, std::allocator > const, referenced_object> >, std::less, std::allocator > >, std::allocator, std::allocator > const, referenced_object> > >::_M_get_node() /usr/include/c++/10/bits/stl_tree.h:584 #4 0x5555590cc923 in std::_Rb_tree_node, std::allocator > const, referenced_object> >* std::_Rb_tree, std::allocator >, std::pair, std::allocator > const, referenced_object>, std::_Select1st, std::allocator > const, referenced_object> >, std::less, std::allocator > >, std::allocator, std::allocator > const, referenced_object> > >::_M_create_node, std::allocator > const&>, std::tuple<> >(std::piecewise_construct_t const&, std::tuple, std::allocator > const&>&&, std::tuple<>&&) /usr/include/c++/10/bits/stl_tree.h:634 #5 0x5555590cc923 in std::_Rb_tree_iterator, std::allocator > const, referenced_object> > std::_Rb_tree, std::allocator >, std::pair, std::allocator > const, referenced_object>, std::_Select1st, std::allocator > const, referenced_object> >, std::less, std::allocator > >, std::allocator, std::allocator > const, referenced_object> > >::_M_emplace_hint_unique, std::allocator > const&>, std::tuple<> >(std::_Rb_tree_const_iterator, std::allocator > const, referenced_object> >, std::piecewise_construct_t const&, std::tuple, std::allocator > const&>&&, std::tuple<>&&) /usr/include/c++/10/bits/stl_tree.h:2461 #6 0x5555590e8757 in std::map, std::allocator >, referenced_object, std::less, std::allocator > >, std::allocator, std::allocator > const, referenced_object> > >::operator[](std::__cxx11::basic_string, std::allocator > const&) /usr/include/c++/10/bits/stl_map.h:501 #7 0x5555590d48b0 in Orch::setObjectReference(std::map, std::allocator >, std::map, std::allocator >, referenced_object, std::less, std::allocator > >, std::allocator, std::allocator > const, referenced_object> > >*, std::less, std::allocator > >, std::allocator, std::allocator > const, std::map, std::allocator >, referenced_object, std::less, std::allocator > >, std::allocator, std::allocator > const, referenced_object> > >*> > >&, std::__cxx11::basic_string, std::allocator > const&, std::__cxx11::basic_string, std::allocator > const&, std::__cxx11::basic_string, std::allocator > const&, std::__cxx11::basic_string, std::allocator > const&) orchagent/orch.cpp:450 #8 0x5555594ff66b in QosOrch::handleQueueTable(Consumer&, std::tuple, std::allocator >, std::__cxx11::basic_string, std::allocator >, std::vector, std::allocator >, std::__cxx11::basic_string, std::allocator > >, std::allocator, std::allocator >, std::__cxx11::basic_string, std::allocator > > > > >&) orchagent/qosorch.cpp:1763 #9 0x5555594edbd6 in QosOrch::doTask(Consumer&) orchagent/qosorch.cpp:2179 #10 0x5555590c8743 in Consumer::drain() orchagent/orch.cpp:241 #11 0x5555590c8743 in Consumer::drain() orchagent/orch.cpp:238 #12 0x5555590c8743 in Consumer::execute() orchagent/orch.cpp:235 #13 0x555559090dad in OrchDaemon::start() orchagent/orchdaemon.cpp:755 #14 0x555558e9be25 in main orchagent/main.cpp:766 #15 0x7f0a299b6d09 in __libc_start_main (/lib/x86_64-linux-gnu/libc.so.6+0x23d09) ```

dgsudharsan and others added 22 commits June 30, 2021 17:58

Vxlan evpn p2mp changes for Layer2 functionality

d9e23ad

[acl mirror action] Mirror session ref count fix at acl rule attachme…

d594c47

…nt (sonic-net#1761) * Fix mirror session ref count at acl rule attachement Signed-off-by: wenda.ni <wenda.ni@bytedance.com>

Cleanup code (sonic-net#1814)

5528ebf

Revert recirc port change (sonic-net#1813)

7c7c451

Recirc port support has been added to recent versions of DNX SAI. The earlier special handling for recirc port is not required anymore and thus reverted.

Fixing compilation

3b6620b

[swss]: Allow portsyncd to run on system without ports (sonic-net#1808)

9f0bb8d

Remove Portsyncd error exit if there are no ports in config DB. Signed-off-by: liora <liora@nvidia.com>

[debugcounterorch] check if counter type is supported before querying… (

ed6786d

sonic-net#1789) * [debugcounterorch] check if counter type is supported before querying object availability count Signed-off-by: Stepan Blyschak <stepanb@nvidia.com>

Added referencing logic

979860f

refactor(fdbsyncd): Convert files with dos2unix (sonic-net#1828)

c81e319

*dos2unix fdbsyncd/* Signed-off-by: Sebastian Hasler <sebastian.hasler@stuvus.uni-stuttgart.de>

[VS] Fix for VS test failures (sonic-net#1836)

0fe2dfe

* Purge swss-common before installing

Bridge mac setting, fix statedb time format (sonic-net#1844)

e65aec9

*Set bridge mac same as Vlan mac *Format state_db entries to have six digit precision for microseconds

[cfgmgr]: Introduce common libs. (sonic-net#1842)

fe0dba0

What I did Refactored cfgmgr Makefile Why I did it To avoid code duplication of common libs How I verified it Build swss Signed-off-by: Nazarii Hnydyn <nazariig@nvidia.com>

[crm] Fix for Issue sonic-net/sonic-buildimage#8036 (sonic-net#1829)

72a72f8

*Added more appropriate log message for CRM queries of MPLS_INSEG and MPLS_NEXTHOP resources when platform does not support these resource types.