Donald Sharp [Tue, 14 Nov 2023 16:00:54 +0000 (11:00 -0500)]
lib: Prevent infinite loop in ospf
For some series of calls in FREEBSD setting the SO_RCVBUF size will
always fail under freebsd. This is no bueno since the
setsockopt_so_recvbuf call goes into an infinite loop.
(gdb) bt
0 setsockopt () at setsockopt.S:4
1 0x0000000083065870 in setsockopt_so_recvbuf (sock=15, size=0) at lib/sockopt.c:26
2 0x00000000002bd200 in ospf_ifp_sock_init (ifp=<optimized out>, ifp@entry=0x8d1dd500) at ospfd/ospf_network.c:290
3 0x00000000002ad1e0 in ospf_if_new (ospf=0x8eefc000, ifp=0x8d1dd500, p=0x8eecf1c0) at ospfd/ospf_interface.c:276
4 0x0000000000304ee0 in add_ospf_interface (co=0x8eecbe10, area=0x8d192100) at ospfd/ospfd.c:1115
5 0x00000000003050fc in ospf_network_run_interface (ospf=0x8eefc000, ifp=0x8d1dd500, p=0x80ff63f8, given_area=0x8d192100)
at ospfd/ospfd.c:1460
6 ospf_network_run (p=0x80ff63f8, area=0x8d192100) at ospfd/ospfd.c:1474
7 ospf_network_set (ospf=ospf@entry=0x8eefc000, p=p@entry=0x80ff63f8, area_id=..., df=<optimized out>) at ospfd/ospfd.c:1247
8 0x00000000002e876c in ospf_network_area (self=<optimized out>, vty=0x8eef3180, argc=<optimized out>, argv=<optimized out>)
at ospfd/ospf_vty.c:560
9 0x0000000083006f24 in cmd_execute_command_real (vline=vline@entry=0x8eee9100, vty=vty@entry=0x8eef3180, cmd=<optimized out>,
cmd@entry=0x0, up_level=<optimized out>) at lib/command.c:978
10 0x0000000083006b30 in cmd_execute_command (vline=0x8eee9100, vty=vty@entry=0x8eef3180, cmd=cmd@entry=0x0, vtysh=vtysh@entry=0)
at lib/command.c:1037
11 0x0000000083007044 in cmd_execute (vty=vty@entry=0x8eef3180, cmd=cmd@entry=0x8eefb000 "network 192.168.64.0/24 area 0.0.0.0",
matched=0x0, vtysh=0) at lib/command.c:1203
12 0x000000008307e9cc in vty_command (vty=0x8eef3180, buf=0x8eefb000 "network 192.168.64.0/24 area 0.0.0.0") at lib/vty.c:594
13 vty_execute (vty=vty@entry=0x8eef3180) at lib/vty.c:1357
14 0x000000008307ce40 in vtysh_read (thread=<optimized out>) at lib/vty.c:2365
15 0x0000000083073db0 in event_call (thread=thread@entry=0x80ff88a0) at lib/event.c:1965
16 0x000000008302c604 in frr_run (master=0x8d188140) at lib/libfrr.c:1214
17 0x000000000029c330 in main (argc=6, argv=<optimized out>) at ospfd/ospf_main.c:252
(gdb)
Force the setsockopt function to quit when the value we are passing no
longer makes any sense.
Fixes: #14790 Signed-off-by: Donald Sharp <sharpd@nvidia.com>
CID 1570969 Overrun
/bgpd/bgp_snmp_bgp4v2.c: 534 in bgp4v2PathAttrLookup()
/bgpd/bgp_snmp_bgp4v2.c: 575 in bgp4v2PathAttrLookup()
/bgpd/bgp_snmp_bgp4v2.c: 514 in bgp4v2PathAttrLookup()
>>> CID 1570969: (OVERRUN)
>>> Overrunning array "bgp->rib" of 4 64-byte elements at element index 4 (byte offset 319) using index "afi" (which evaluates to 4).
Donald Sharp [Mon, 13 Nov 2023 15:26:48 +0000 (10:26 -0500)]
bgpd: Used %pBD instead of %pRN
Let's use the natural data structure in bgp for the prefix display
instead of a bunch of places where we call a translator function.
The %pBD does this and actually ensures data is correct.
Also fix a few spots in bgp_zebra.c where the cast to a NULL
pointer causes the catcher functionality to not work and fix
the resulting crash that resulted.
Igor Ryzhov [Sat, 11 Nov 2023 00:13:17 +0000 (02:13 +0200)]
lib, mgmtd: respect base xpath in mgmtd
`nb_cli_apply_changes` can be called with base xpath which should be
prepended to xpaths of every change in a transaction. This base xpath is
respected by regular northbound CLI but not by mgmtd. This commit fixes
the problem.
Igor Ryzhov [Sun, 12 Nov 2023 00:57:25 +0000 (02:57 +0200)]
bgpd: fix build error
I recieve the following error with GCC 9.4.0:
```
In file included from /usr/include/string.h:495,
from ./lib/zebra.h:23,
from bgpd/bgp_snmp_bgp4v2.c:7:
In function ‘memset’,
inlined from ‘bgp4v2PathAttrLookup’ at bgpd/bgp_snmp_bgp4v2.c:605:3,
inlined from ‘bgp4v2PathAttrTable’ at bgpd/bgp_snmp_bgp4v2.c:747:9:
/usr/include/x86_64-linux-gnu/bits/string_fortified.h:71:10: error: ‘__builtin_memset’ offset [9, 20] from the object at ‘paddr’ is out of the bounds of referenced subobject ‘_v4_addr’ with type ‘struct in_addr’ at offset 4 [-Werror=array-bounds]
71 | return __builtin___memset_chk (__dest, __ch, __len, __bos0 (__dest));
| ^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
```
Christian Hopps [Sat, 11 Nov 2023 19:56:26 +0000 (20:56 +0100)]
tools: suppress some reachable external lib "leaks"
- These are just normal reachable allocs from inside external library code;
however, when running valgrind with memleak types "all" in order to find FRR
specific leaks (from memory.h) these are the only other reported leaks. Makes
easier checking and fixing.
Christian Hopps [Sat, 11 Nov 2023 18:21:17 +0000 (19:21 +0100)]
lib: mgmtd: fix debug cli commands and memleaks
- Cannot have 2 cmd_node's with same .node number. Install the mgmtd
client library debug nodes (client frontend and client backend) using
new unique node numbers. Fixes memleaks.
- Fix "debug mgmt client backend" to generate correct config (and not
for frontend).
Igor Ryzhov [Sat, 11 Nov 2023 00:06:11 +0000 (02:06 +0200)]
lib: fix possible freeing of libyang data
mgmtd frees all non-NULL change->value variables at the end of every
commit. We shouldn't assign change->value with data returned by libyang
to prevent freeing of library-allocated memory.
The walk stopped because the index used in the NlriTable entries is
decrementing, and this is against the snmp specifications. Also, the
computed index is wrong, and does not match the provided
draft-ietf-idr-bgp4-mibv2-1 specification.
Fix this by computing a valid index, and by finding out the next
consecutive prefix.
The resulting changes do not break the walk, and the output is changed:
tests: Check received prefixes before immediately sending dynamic capabilities
If we send capabilities immediately, before receiving an UPDATE message, we end up
with a notification received from the neighbor. Let's wait until we have the fully
converged topology and do the stuff.
Tested locally and can't replicate the failure, let's see how happy is the CI this time.
bgpd: Send software version capability for OAD EBGP peers
We send this capability for iBGP peers by default. Recently OAD support was
merged, and we should adopt sending the capability according to OAD as well.
Donald Sharp [Sat, 4 Nov 2023 08:15:54 +0000 (08:15 +0000)]
*: Remove netlink headers from lib/zebra.h
The headers associated with netlink code
really only belong in those that need it.
Move these headers out of lib/zebra.h and
into more appropriate places. bgp's usage
of the RT_TABLE_XXX defines are probably not
appropriate and will be cleaned up in future
commits.
Christian Hopps [Tue, 24 Oct 2023 18:58:11 +0000 (14:58 -0400)]
tools: checkpatch updated with darr_foreach iter macros
Also update `checkpatch.sh` so it runs `checkpatch.pl` from the same directory
it resides in. This allows copying them both somewhere else to use a specific
version.
Donald Sharp [Mon, 6 Nov 2023 13:56:50 +0000 (08:56 -0500)]
zebra: When using Nexthop groups, use v6 RR semantics
The nexthop group route replace operation was made consistent
across all versions of the kernel. A v6 route replacement
does not need to do a delete than add when using nexthop
groups
Donald Sharp [Fri, 3 Nov 2023 16:35:45 +0000 (16:35 +0000)]
zebra: Add v6_rr_semantics status to `show zebra`
The v6_rr_semantics variable was being set but never
reported and had to be inferred from watching netlink
messages to the kernel. Let's add a bit of code
to `show zebra` so that we can know how it is being
used.
David Lamparter [Wed, 30 Nov 2022 10:19:14 +0000 (11:19 +0100)]
lib: add a transparent union for sockaddrs
As with "prefixptr"/"prefixconstptr", this allows function prototypes to
say "just give me any sockaddr" typing-wise, without having the compiler
complain that a "struct sockaddr_in" is not a "struct sockaddr".
(Split off from other changes that start using this.)
Signed-off-by: David Lamparter <equinox@opensourcerouting.org>
Maxence Younsi [Mon, 6 Feb 2023 16:30:28 +0000 (17:30 +0100)]
bgpd: skip bmp messages when vrf id is unknown
changed result type of bmp_get_peer_distinguisher to int
added result pointer parameter to bmp_get_peer_distinguisher
bmp_get_peer_distinguisher returns 0 means the result is valid else
error occured do not use result
bgpd: loc-rib uptime moved to bgp_path_info_extra and set in header
moved loc-rib uptime field "bgp_rib_uptime" to struct bgp_path_info_extra for memory concerns
moved logic into bgp_route_update's callback bmp_route_update
written timestamp in per peer header
mxyns [Fri, 12 Aug 2022 16:00:38 +0000 (18:00 +0200)]
bgpd: applied styling and fixed warnings
frrbot found style &/| linter errors
fixed bmp_process_one return value warnings and added safety checks
fixed const modifier warning in bmp_put_vrftablename_info_tlv
added unused attribute to bmp_put_vrftablename_info_tlv
remove unused variables in bmp_process_one and bmp_route_update
Maxou [Mon, 8 Aug 2022 12:32:38 +0000 (14:32 +0200)]
bgpd: refactored bmp_route_update & cleanup TODOs
TODOs that are done/un-necessary now deleted
refactored bmp_route_update to use a modified bmp_process_one function call instead of duplicating similar code
bgpd: temporary set timestamp to 0 for loc rib monitoring messages
set timestamp to 0 for loc rib monitoring messages as path selection time is not available atm
this is temporary and tv is meant to be set to the path selection/install time at some point
vrf_id_to_name is used for display values only and returns "Unknown" when the vrf is not found
doing a manual lookup and not providing any tlv when the vrf is not found should be better
set field peer bgp id to the peer's remote id in every case except loc-rib (RFC9069 case) in which we put the bgp instance's router-id if available or 0-filled if not available
set field peer asn to local primary bgp asn in case of loc-rib instance (RFC9069) else it's set to the peer's asn
set field peer address to 0 in loc-rib instance (RFC9069 case) and to the peer's address in other cases
had to pass struct bgp reference to bmp_per_peer_hdr to access router-id and such, but it's always safely accessed when used
added afi/safi monitoring synchronisation for loc-rib
added peer_type_flag to bmp_eor signature, only set to BMP_PEER_TYPE_LOC_RIB and to 0 in other cases like it was before
updated tracelog to include peer_type_flag value
bgpd: peer flag set for loc-rib monitoring (left set to 0 in other cases)
set peer type flag to 3 for loc rib monitoring
leave to 0 in other cases like before, even though RFC7854 tells us to set it to 0 1 or 2 depending on the case global/rd/local instance
projects / mirror / frr.git / log