Igor Ryzhov [Fri, 8 Oct 2021 21:22:31 +0000 (00:22 +0300)]
lib: set type for newly created interfaces
Currently, the ll_type is set only in `netlink_interface` which is
executed only during startup. If the interface is created when the FRR
is already running, the type is not stored.
Renato Westphal [Fri, 8 Oct 2021 00:06:01 +0000 (21:06 -0300)]
ospfd: fix display of plain-text data on "show ... json" commands
Add a 'json' parameter to the 'show_opaque_info' callback definition,
and update all instances of that callback to not display plain-text
data when the user requested JSON data.
Donald Sharp [Fri, 8 Oct 2021 11:37:15 +0000 (07:37 -0400)]
tests: Fix ospf[6]_gr_topo1 tests to work better under load
2 things:
a) Each test was setting up for graceful restart with calls to
`graceful-restart prepare ip[v6] ospf`, then sleeping for
3 or 5 seconds. Then killing the ospf process. Under heavy
load there is no guarantee that zebra has received/processed
this signal. Write some code to ensure that this happens
b) Tests are issuing commands in this order:
1) issue gr prepare command
2) kill router
3) <ensure routes were still installed in zebra>
4) start router
5) <ensure routes were stil installed in zebra>
Imagine that the system is under some load and there is
a small amount of time before step 5 happens. In this
case ospf could have come up and started neighbor relations
and also started installing routes. If zebra receives
a new route before step 5 is issued then the route could
be in a state where it is not installed, because it is
being sent to the kernel for installation. This would
fail the test because it would only look 1 time. This
is fixed by giving time on restart for the routes to
be in the installed state.
Renato Westphal [Fri, 8 Oct 2021 02:45:31 +0000 (23:45 -0300)]
ospfd: preserve DR status across graceful restarts
RFC 3623 says:
"If the restarting router determines that it was the Designated
Router on a given segment prior to the restart, it elects
itself as the Designated Router again. The restarting router
knows that it was the Designated Router if, while the
associated interface is in Waiting state, a Hello packet is
received from a neighbor listing the router as the Designated
Router".
Implement that logic when processing Hello messages to ensure DR
interfaces will preserve their DR status across a graceful restart.
Donald Sharp [Thu, 7 Oct 2021 16:08:42 +0000 (12:08 -0400)]
zebra: Display how long zebra is expected to wait for GR
When a client sends to zebra that GR mode is being turned
on. The client also passes down the time zebra should hold
onto the routes. Display this time with the output
of the `show zebra client` command as well.
Igor Ryzhov [Thu, 7 Oct 2021 12:53:10 +0000 (15:53 +0300)]
*: don't pass pointers to a local variables to thread_add_*
We should never pass pointers to local variables to thread_add_* family.
When an event is executed, the library writes into this pointer, which
means it writes into some random memory on a stack.
ospfd: ospf nbr in full although mismatch in hello packet contents
Issue:
===================
OSPF neighbors are not going down even after 10 mins when
having a mismatch in hello and dead interval.
First neighbors are formed and then a mismatch in the interval
is created, it is observed that the neighbor is not going down.
Root Cause Analysis:
====================
The event HelloReceived defined in RFC 2328 was named as PacketReceived
and this event was scheduled whenever LS Update, LS Ack, LS Request,
DD description packet or Hello packet is received.
Although there is a mismatch in the Hello packet contents, the
event PacketReceived gets triggered due to LS Update received and the
dead timer gets reset and hence the neighbor was never going Down and
remains FULL.
Fix:
==================
As per RFC 2328, the HelloReceived needs to be triggered only when
valid OSPF Hello packet is received and not when other OSPF packets
are received. Modified the function name as well.
Donald Sharp [Tue, 5 Oct 2021 00:32:25 +0000 (20:32 -0400)]
watchfrr: Allow an integrated config to work within a namespace
Since watchfrr invokes vtysh to gather the show run output and
write the data, if we are operating inside of a namespace FRR
must also pass this in.
Yes. This seems hacky. I don't fully understand why vtysh
is invoked this way.
New output:
sharpd@eva:~/frr3$ sudo vtysh -N one
Hello, this is FRRouting (version 8.1-dev).
Copyright 1996-2005 Kunihiro Ishiguro, et al.
eva# wr mem
Note: this version of vtysh never writes vtysh.conf
% Can't open configuration file /etc/frr/one/vtysh.conf due to 'No such file or directory'.
Building Configuration...
Integrated configuration saved to /etc/frr/one/frr.conf
[OK]
eva#
Igor Ryzhov [Wed, 6 Oct 2021 14:35:07 +0000 (17:35 +0300)]
lib: fix incorrect thread management
The current code passes an address of a local variable to `thread_add_read`
which stores it into `thread->ref` by the lib. The next time the thread
callback is executed, the lib stores NULL into the `thread->ref` which
means it writes into some random memory on the stack.
To fix this, we should pass a pointer to the vector entry to the lib.
Donald Sharp [Wed, 6 Oct 2021 11:58:35 +0000 (07:58 -0400)]
bgpd: Check return from generic_set_add
Coverity found a couple of spots where FRR was
ignoring the return code of generic_set_add.
Just follow the code pattern for the rest of
the usage in the code.
In startup, zebra would dump interface information from Kernel in 3
steps w/o lock: step1, get interface information; step2, get interface
ipv4 address; step3, get interface ipv6 address.
If any interface gets added after step1, but before step2/3, zebra
would get extra interface addresses in step2/3 that has not been added
into zebra in step1. Returning error in the referenced interface lookup
would cause the startup interface retrieval to be incomplete.
Rafael Zalamena [Tue, 5 Oct 2021 15:37:34 +0000 (12:37 -0300)]
topotests: justify code sleep
Document the `sleep` statement so people know that we are sleeping
because we are waiting for the BFD down notification. If we don't
sleep here it is possible that we get outdated `show` command results.
Signed-off-by: Rafael Zalamena <rzalamena@opensourcerouting.org>
Igor Ryzhov [Tue, 5 Oct 2021 14:38:21 +0000 (17:38 +0300)]
isisd: fix redistribute CLI
Currently, it is possible to configure IPv6 protocols for IPv4
redistribution and vice versa in CLI. The YANG model doesn't allow this
so the user receives the following error:
```
nfware(config-router)# redistribute ipv4 ospf6 level-1
% Failed to edit configuration.
YANG error(s):
Invalid enumeration value "ospf6".
Invalid enumeration value "ospf6".
Invalid enumeration value "ospf6".
YANG path: Schema location /frr-isisd:isis/instance/redistribute/ipv4/protocol.
```
Let's make CLI more user-friendly and allow only supported protocols in
redistribution commands.
Rafael Zalamena [Mon, 4 Oct 2021 21:10:58 +0000 (18:10 -0300)]
lib: prevent gRPC assert on missing YANG node
`yang_dnode_get` will `assert` if no YANG node/model exist, so lets test for
its existence first before trying to access it.
This `assert` is only acceptable for internal FRR usage otherwise we
might miss typos or unmatching YANG models nodes/leaves. For gRPC usage
we should let users attempt to use non existing models without
`assert`ing.
Signed-off-by: Rafael Zalamena <rzalamena@opensourcerouting.org>
rgirada [Tue, 5 Oct 2021 07:52:36 +0000 (00:52 -0700)]
ospf6d: ospf6d is crashing upon receiving duplicated Grace LSA.
Description:
When grace lsa received, DUT is adding
the copy of the lsas to all nbrs retransmission list as part of
flooding procedure and subsequently incrementing the rmt counter in
the original the LSA. This counter is supposed to be decremented
when ack is received by nbr and the lsa will be removed from retransmission list.
But in our current scenario,
Step-1:
When GR helper is disabled, if DUT receives the grace lsa
it adds the lsa copy to nbrs retransmission list but original
LSA will be discarded since GR helper disabled.
Step-2:
GR helper enabled and DUT receives the grace lsa, as part
of flooding process all nbrs have same copy of lsa in their
corresponding rmt list which was added in step -1 due to this
the corresponding rmt counter in the original lsa is not getting
incremented.
Step-3:
If the same copy of the grace lsa received by DUT, It considers
as implicit ack from nbr if the same copy of the lsa exits in its
rmt list and subsequently decrement the rmt counter.
Since counter is zero (because of step-1 and 2) , it is asserting while decrement.
Donald Sharp [Mon, 4 Oct 2021 12:37:16 +0000 (08:37 -0400)]
ospf6d: Ensure expire thread is properly stopped
The lsa->expire thread is for keeping track of when we
are expecting to expire(remove/delete) a lsa. There
are situations where we just decide to straight up
delete the lsa, but we are not ensuring that the
lsa is not already setup for expiration.
In that case just stop the expiry thread and
do the deletion.
Additionally there was a case where ospf6d was
just dropping the fact that a thread was already
scheduled for expiration. In that case we
should just setup the timer again and it will
reset it appropriately.
Fixes: #9721 Signed-off-by: Donald Sharp <sharpd@nvidia.com>
Donald Sharp [Mon, 4 Oct 2021 13:47:29 +0000 (09:47 -0400)]
eigrpd: Ensure better `struct thread *` semantics
1) Do not explicitly set the thread pointer to NULL.
FRR should only ever use the appropriate THREAD_ON/THREAD_OFF
semantics. This is espacially true for the functions we
end up calling the thread for.
2) Fix mixup of `struct eigrp_interface` and `struct eigrp`
usage of the same thread pointer.
Donald Sharp [Mon, 4 Oct 2021 13:36:27 +0000 (09:36 -0400)]
ripd: Ensure better `struct thread *` semantics
Do not explicitly set the thread pointer to NULL.
FRR should only ever use the appropriate THREAD_ON/THREAD_OFF
semantics. This is espacially true for the functions we
end up calling the thread for.
projects / matthieu / frr.git / log