| Age | Commit message (Collapse) | Author |
|---|
|
Without this patch we always see a double-reset, e.g.:
```
2024/11/04 12:42:43.010 BGP: [VQY9X-CQZKG] bgp_peer_bfd_update_source: address [0.0.0.0->172.18.0.3] to [172.18.0.2->172.18.0.3]
2024/11/04 12:42:43.010 BGP: [X8BD9-8RKN4] bgp_peer_bfd_update_source: interface none to eth0
2024/11/04 12:42:43.010 BFD: [MSVDW-Y8Z5Q] ptm-del-dest: deregister peer [mhop:no peer:172.18.0.3 local:0.0.0.0 vrf:default cbit:0x00 minimum-ttl:255]
2024/11/04 12:42:43.010 BFD: [NYF5K-SE3NS] ptm-del-session: [mhop:no peer:172.18.0.3 local:0.0.0.0 vrf:default] refcount=0
2024/11/04 12:42:43.010 BFD: [NW21R-MRYNT] session-delete: mhop:no peer:172.18.0.3 local:0.0.0.0 vrf:default
2024/11/04 12:42:43.010 BGP: [P3D3N-3277A] 172.18.0.3 [FSM] Timer (routeadv timer expire)
2024/11/04 12:42:43.010 BFD: [YA0Q5-C0BPV] control-packet: no session found [mhop:no peer:172.18.0.3 local:172.18.0.2 port:11]
2024/11/04 12:42:43.010 BFD: [MSVDW-Y8Z5Q] ptm-add-dest: register peer [mhop:no peer:172.18.0.3 local:172.18.0.2 vrf:default cbit:0x00 minimum-ttl:255]
2024/11/04 12:42:43.011 BFD: [PSB4R-8T1TJ] session-new: mhop:no peer:172.18.0.3 local:172.18.0.2 vrf:default ifname:eth0
2024/11/04 12:42:43.011 BGP: [Q4BCV-6FHZ5] zclient_bfd_session_update: 172.18.0.2/32 -> 172.18.0.3/32 (interface eth0) VRF default(0) (CPI bit no): Down
2024/11/04 12:42:43.011 BGP: [MKVHZ-7MS3V] bfd_session_status_update: neighbor 172.18.0.3 vrf default(0) bfd state Up -> Down
2024/11/04 12:42:43.011 BGP: [HZN6M-XRM1G] %NOTIFICATION: sent to neighbor 172.18.0.3 6/10 (Cease/BFD Down) 0 bytes
2024/11/04 12:42:43.011 BGP: [QFMSE-NPSNN] zclient_bfd_session_update: sessions updated: 1
2024/11/04 12:42:43.011 BGP: [ZWCSR-M7FG9] 172.18.0.3 [FSM] BGP_Stop (Established->Clearing), fd 22
```
Reset is due to the source address change.
With this patch, we reset the session only if it's a _REAL_ BFD down event, which
means we trigger session reset if BFD session is established earlier than BGP.
Signed-off-by: Donatas Abraitis <donatas@opensourcerouting.org>
|
|
In bgp_show_table_rd(), the is_last argument is determined using the
expression "next == NULL" to check if the RD table is the last one. This
helps ensure proper JSON formatting.
However, if next is not NULL but is no longer associated with a BGP
table, the JSON output becomes malformed.
Updates the condition to also verify the existence of the next bgp_dest
table.
Fixes: 1ae44dfcba ("bgpd: unify 'show bgp' with RD with normal unicast bgp show")
Signed-off-by: Louis Scalbert <louis.scalbert@6wind.com>
(cherry picked from commit cf0269649cdd09b8d3f2dd8815caf6ecf9cdeef9)
|
|
pimd: Fix for FHR mroute taking longer to age out (backport #14105)
|
|
pimd: fix DR election race on startup (backport #18048)
|
|
zebra: include resolving nexthops in nhg hash (backport #17935)
|
|
nhrpd: fix dont consider incomplete L2 entry (backport #18078)
|
|
Sometimes, NHRP receives L2 information on a cache entry with the
0.0.0.0 IP address. NHRP considers it as valid and updates the binding
with the new IP address.
> Feb 09 20:09:54 aws-sin-vpn01 nhrpd[2695]: [QQ0NK-1H449] Netlink: new-neigh 10.2.114.238 dev dmvpn1 lladdr 162.251.180.10 nud 0x2 cache used 0 type 4
> Feb 09 20:10:35 aws-sin-vpn01 nhrpd[2695]: [QQ0NK-1H449] Netlink: new-neigh 10.2.114.238 dev dmvpn1 lladdr 162.251.180.10 nud 0x4 cache used 1 type 4
> Feb 09 20:10:48 aws-sin-vpn01 nhrpd[2695]: [QQ0NK-1H449] Netlink: del-neigh 10.2.114.238 dev dmvpn1 lladdr 162.251.180.10 nud 0x4 cache used 1 type 4
> Feb 09 20:10:49 aws-sin-vpn01 nhrpd[2695]: [QQ0NK-1H449] Netlink: who-has 10.2.114.238 dev dmvpn1 lladdr (unspec) nud 0x1 cache used 1 type 4
> Feb 09 20:10:49 aws-sin-vpn01 nhrpd[2695]: [QVXNM-NVHEQ] Netlink: update binding for 10.2.114.238 dev dmvpn1 from c 162.251.180.10 peer.vc.nbma 162.251.180.10 to lladdr (unspec)
> Feb 09 20:10:49 aws-sin-vpn01 nhrpd[2695]: [QQ0NK-1H449] Netlink: new-neigh 10.2.114.238 dev dmvpn1 lladdr 0.0.0.0 nud 0x2 cache used 1 type 4
> Feb 09 20:11:30 aws-sin-vpn01 nhrpd[2695]: [QQ0NK-1H449] Netlink: new-neigh 10.2.114.238 dev dmvpn1 lladdr 0.0.0.0 nud 0x4 cache used 1 type 4
Actually, the 0.0.0.0 IP addressed mentiones in the 'who-has' message is
wrong because the nud state value means that value is incomplete and
should not be handled as a valid entry. Instead of considering it, fix
this by by invalidating the current binding. This step is necessary in
order to permit NHRP to trigger resolution requests again.
Signed-off-by: Philippe Guibert <philippe.guibert@6wind.com>
(cherry picked from commit 3202323052485d8138a3440e9c9907594ad99c57)
|
|
lib: crash handlers must be allowed on threads (backport #18060)
|
|
lib: actually hash all 16 bytes of IPv6 addresses, not just 4 (backport #17901)
|
|
Blocking all signals on non-main threads is not the way to go, at least
the handlers for SIGSEGV, SIGBUS, SIGILL, SIGABRT and SIGFPE need to run
so we get backtraces. Otherwise the process just exits.
Signed-off-by: David Lamparter <equinox@opensourcerouting.org>
(cherry picked from commit 13a6ac5b4ca8fc08b348f64de64a787982f24250)
|
|
Currently FRR does not handle v6 recurisive resolution properly
when the route being recursed through changes and the most
significant bits of the route are not changed.
Signed-off-by: Donald Sharp <sharpd@nvidia.com>
(cherry picked from commit 73ab6a46c51db91df297774221053ab8fc4d12ae)
|
|
Ensure that the nhg hash comparison function includes all
nexthops, including recursive-resolving nexthops.
Signed-off-by: Mark Stapp <mjs@cisco.com>
(cherry picked from commit cb7cf73992847cfd4af796085bf14f2fdc4fa8db)
|
|
We were hashing 4 bytes of the address. Even for IPv6 addresses.
Oops.
The reason this was done was to try to make it faster, but made a
complex maze out of everything. Time for a refactor.
Signed-off-by: David Lamparter <equinox@opensourcerouting.org>
(cherry picked from commit 001fcfa1dd9f7dc2639b4f5c7a52ab59cc425452)
|
|
When reading in a nexthop from ZAPI, only set the fields that actually
have meaning. While it shouldn't happen to begin with, we can otherwise
carry padding garbage into the unused leftover union bytes.
Signed-off-by: David Lamparter <equinox@opensourcerouting.org>
(cherry picked from commit 4a0e1419a69d07496c7adfb744beecd00e1efef2)
|
|
rmap_src wasn't initialized, so for IPv4 the unused 12 bytes would
contain whatever junk is on the stack on function entry. Also move
the IPv4 parse before the IPv6 parse so if it's successful we can be
sure the other bytes haven't been touched.
Signed-off-by: David Lamparter <equinox@opensourcerouting.org>
(cherry picked from commit b666ee510eb480da50476b1bbc84bdf8365df95c)
|
|
Doesn't seem to break anything but really poor style to pass potentially
uninitialized data to hash_lookup.
Signed-off-by: David Lamparter <equinox@opensourcerouting.org>
(cherry picked from commit c88589f5e9351654c04322eb395003297656989d)
|
|
Zero out the 12 unused bytes (for the IPv6 address) when reading in an
IPv4 address.
Signed-off-by: David Lamparter <equinox@opensourcerouting.org>
(cherry picked from commit 95cf0b227980999e2af22a2c171e5237e5ffca8e)
|
|
While the loop is currently exited in all cases after using nexthop, it
is a footgun to have "nh" around to be reused in another iteration of
the loop. This would leave nexthop with partial data from the previous
use. Make it local where needed instead.
Signed-off-by: David Lamparter <equinox@opensourcerouting.org>
(cherry picked from commit ce7f5b21221f0b3557d1f4a40793230d8bc4cf02)
|
|
In case interface address is learnt during configuration, make sure to
run DR election when configuring PIM/PIM passive on interface.
Signed-off-by: Rafael Zalamena <rzalamena@opensourcerouting.org>
(cherry picked from commit 86445246062583197d4a6dff7b8c74003cd8049d)
|
|
Issue:
When there is no traffic for a group, the LHR and RP take the default KAT+Join timer expiry of
a maximum of 480 seconds to clear the S,G . However, in the FHR, we update the state from JOINED
to NOT Joined, downstream state from PPto NOINFO. This restarts the ET timer, causing S,G on FHR to
take more than 10 minutes to age out.
In other words,
Consider a case where (S,G) is in Join state. When the traffic stops and the KAT (210) expires,
the Join expiry timer restarts. At this time, if we receive a prune, the expectation is to set
PPT to 0 (RFC 4601 sec 4.5.2).
When the PPT expires, we move to the noinfo state and restart the expiry timer one more time. We remove the
(S,G) entry only after ~10 minutes when there is no active traffic.
Summary:
KAT Join ET 210 + PP ET 210 + NOINFO ET 210.
Solution:
Delete the ifchannel when in noinfo state, and KAT is not running.
Ticket: #13703
Signed-off-by: Rajesh Varatharaj <rvaratharaj@nvidia.com>
(cherry picked from commit afed39ea2be25bf30d50ac49b4edf424deadcb17)
|
|
Stabilize 10.1 branch
|
|
This reverts commit 0a923af56dbe43fdb4e9184c3525d0537740aef9.
|
|
This reverts commit aba588dd09aa098a88ba1355798c0e784e91ebc8.
|
|
This reverts commit 8c187fb4f838d8d8a21f8608c3a510136764b122.
|
|
This reverts commit daa68852a2a78acf103e8ae1127953b2870c6772.
|
|
This reverts commit 3abd84ef5be1ef56b66f0e7617f8afab6da6c5cc.
|
|
bgpd: Recent failed backports for 10.1
|
|
When running the bgp_evpn_rt5 setup with unified config, memory leak
about a non deleted BGP instance happens.
> root@ubuntu2204hwe:~/frr/tests/topotests/bgp_evpn_rt5# cat /tmp/topotests/bgp_evpn_rt5.test_bgp_evpn/r1.asan.bgpd.1164105
>
> =================================================================
> ==1164105==ERROR: LeakSanitizer: detected memory leaks
>
> Indirect leak of 12496 byte(s) in 1 object(s) allocated from:
> #0 0x7f358eeb4a57 in __interceptor_calloc ../../../../src/libsanitizer/asan/asan_malloc_linux.cpp:154
> #1 0x7f358e877233 in qcalloc lib/memory.c:106
> #2 0x55d06c95680a in bgp_create bgpd/bgpd.c:3405
> #3 0x55d06c95a7b3 in bgp_get bgpd/bgpd.c:3805
> #4 0x55d06c87a9b5 in bgp_get_vty bgpd/bgp_vty.c:603
> #5 0x55d06c68dc71 in bgp_evpn_local_l3vni_add bgpd/bgp_evpn.c:7032
> #6 0x55d06c92989b in bgp_zebra_process_local_l3vni bgpd/bgp_zebra.c:3204
> #7 0x7f358e9e3feb in zclient_read lib/zclient.c:4626
> #8 0x7f358e98082d in event_call lib/event.c:1996
> #9 0x7f358e848931 in frr_run lib/libfrr.c:1232
> #10 0x55d06c60eae1 in main bgpd/bgp_main.c:557
> #11 0x7f358e229d8f in __libc_start_call_main ../sysdeps/nptl/libc_start_call_main.h:58
Actually, a BGP VRF Instance is created in auto mode when creating the
global BGP instance for the L3 VNI. And again, an other BGP VRF instance
is created. Fix this by ensuring that a non existing BGP instance is not
present. If it is present, and with auto mode or in hidden mode, then
override the AS value.
Fixes: f153b9a9b636 ("bgpd: Ignore auto created VRF BGP instances")
Signed-off-by: Philippe Guibert <philippe.guibert@6wind.com>
Signed-off-by: Donatas Abraitis <donatas@opensourcerouting.org>
|
|
This reverts commit aba588dd09aa098a88ba1355798c0e784e91ebc8.
|
|
When peering with an EVPN device from other vendor, FRR acting as route
reflector is not able to read nor transmit the label value.
Actually, EVPN AD routes completely ignore the label value in the code,
whereas in some functionalities like evpn-vpws, it is authorised to
carry and propagate label value.
Fix this by handling the label value.
Signed-off-by: Philippe Guibert <philippe.guibert@6wind.com>
|
|
If we have IPv6-only network and no IPv4 addresses at all, then by default
0.0.0.0 is created which is treated as malformed according to RFC 6286.
Signed-off-by: Donatas Abraitis <donatas@opensourcerouting.org>
|
|
zebra: fix evpn svd hash avoid double free (backport #17991)
|
|
bgpd: fix route-distinguisher in vrf leak json cmd (backport #17992)
|
|
bgpd: Auto vrf instance (backports)
|
|
For auto configured value RD value comes as NULL,
switching back to original change will ensure to cover
for both auto and user configured RD value in JSON.
tor-11# show bgp vrf blue ipv4 unicast route-leak json
{
"vrf":"blue",
"afiSafi":"ipv4Unicast",
"importFromVrfs":[
"purple"
],
"importRts":"10.10.3.11:6",
"exportToVrfs":[
"purple"
],
"routeDistinguisher":"(null)", <<<<<
"exportRts":"10.10.3.11:10"
}
Signed-off-by: Chirag Shah <chirag@nvidia.com>
(cherry picked from commit 892704d07f5286464728720648ad392b485a9966)
|
|
Upon zebra shutdown hash_clean_and_free is called
where user free function is passed,
The free function should not call hash_release
which lead to double free of hash bucket.
Fix:
The fix is to avoid calling hash_release from
free function if its called from hash_clean_and_free
path.
10 0x00007f0422b7df1f in free () from /lib/x86_64-linux-gnu/libc.so.6
11 0x00007f0422edd779 in qfree (mt=0x7f0423047ca0 <MTYPE_HASH_BUCKET>,
ptr=0x55fc8bc81980) at ../lib/memory.c:130
12 0x00007f0422eb97e2 in hash_clean (hash=0x55fc8b979a60,
free_func=0x55fc8a529478 <svd_nh_del_terminate>) at
../lib/hash.c:290
13 0x00007f0422eb98a1 in hash_clean_and_free (hash=0x55fc8a675920
<svd_nh_table>, free_func=0x55fc8a529478 <svd_nh_del_terminate>) at
../lib/hash.c:305
14 0x000055fc8a5323a5 in zebra_vxlan_terminate () at
../zebra/zebra_vxlan.c:6099
15 0x000055fc8a4c9227 in zebra_router_terminate () at
../zebra/zebra_router.c:276
16 0x000055fc8a4413b3 in zebra_finalize (dummy=0x7fffb881c1d0) at
../zebra/main.c:269
17 0x00007f0422f44387 in event_call (thread=0x7fffb881c1d0) at
../lib/event.c:2011
18 0x00007f0422ecb6fa in frr_run (master=0x55fc8b733cb0) at
../lib/libfrr.c:1243
19 0x000055fc8a441987 in main (argc=14, argv=0x7fffb881c4a8) at
../zebra/main.c:584
Signed-off-by: Chirag Shah <chirag@nvidia.com>
(cherry picked from commit 1d4f5b9b19588d77d3eaf06440c26a8c974831a3)
|
|
When VRF instance is going to be deleted inside bgp_vrf_disable(), it uses
a helper method that skips auto created VRF instances and that leads to STALE
issue.
When creating a VNI for a particular VRF vrfX with e.g. `advertise-all-vni`,
auto VRF instance is created, and then we do `router bgp ASN vrf vrfX`.
But when we do a reload bgp_vrf_disable() is called, and we miss previously
created auto instance.
Signed-off-by: Donatas Abraitis <donatas@opensourcerouting.org>
|
|
The more the vrf green is referenced in the import bgp command, the more
there are instances created. The below configuration shows that the vrf
green is referenced twice, and two BGP instances of vrf green are
created.
The below configuration:
> router bgp 99
> [..]
> import vrf green
> exit
> router bgp 99 vrf blue
> [..]
> import vrf green
> exit
> router bgp 99 vrf green
> [..]
> exit
>
> r4# show bgp vrfs
> Type Id routerId #PeersCfg #PeersEstb Name
> L3-VNI RouterMAC Interface
> DFLT 0 10.0.3.4 0 0 default
> 0 00:00:00:00:00:00 unknown
> VRF 5 10.0.40.4 0 0 blue
> 0 00:00:00:00:00:00 unknown
> VRF 6 0.0.0.0 0 0 green
> 0 00:00:00:00:00:00 unknown
> VRF 6 10.0.94.4 0 0 green
> 0 00:00:00:00:00:00 unknown
Fix this at import command, by looking at an already present bgp
instance.
Signed-off-by: Philippe Guibert <philippe.guibert@6wind.com>
|
|
When running the bgp_evpn_rt5 setup with unified config, memory leak
about a non deleted BGP instance happens.
> root@ubuntu2204hwe:~/frr/tests/topotests/bgp_evpn_rt5# cat /tmp/topotests/bgp_evpn_rt5.test_bgp_evpn/r1.asan.bgpd.1164105
>
> =================================================================
> ==1164105==ERROR: LeakSanitizer: detected memory leaks
>
> Indirect leak of 12496 byte(s) in 1 object(s) allocated from:
> #0 0x7f358eeb4a57 in __interceptor_calloc ../../../../src/libsanitizer/asan/asan_malloc_linux.cpp:154
> #1 0x7f358e877233 in qcalloc lib/memory.c:106
> #2 0x55d06c95680a in bgp_create bgpd/bgpd.c:3405
> #3 0x55d06c95a7b3 in bgp_get bgpd/bgpd.c:3805
> #4 0x55d06c87a9b5 in bgp_get_vty bgpd/bgp_vty.c:603
> #5 0x55d06c68dc71 in bgp_evpn_local_l3vni_add bgpd/bgp_evpn.c:7032
> #6 0x55d06c92989b in bgp_zebra_process_local_l3vni bgpd/bgp_zebra.c:3204
> #7 0x7f358e9e3feb in zclient_read lib/zclient.c:4626
> #8 0x7f358e98082d in event_call lib/event.c:1996
> #9 0x7f358e848931 in frr_run lib/libfrr.c:1232
> #10 0x55d06c60eae1 in main bgpd/bgp_main.c:557
> #11 0x7f358e229d8f in __libc_start_call_main ../sysdeps/nptl/libc_start_call_main.h:58
Actually, a BGP VRF Instance is created in auto mode when creating the
global BGP instance for the L3 VNI. And again, an other BGP VRF instance
is created. Fix this by ensuring that a non existing BGP instance is not
present. If it is present, and with auto mode or in hidden mode, then
override the AS value.
Fixes: f153b9a9b636 ("bgpd: Ignore auto created VRF BGP instances")
Signed-off-by: Philippe Guibert <philippe.guibert@6wind.com>
Signed-off-by: Donatas Abraitis <donatas@opensourcerouting.org>
|
|
bgpd: With suppress-fib-pending ensure withdrawal is sent (backport #17971)
|
|
When you have suppress-fib-pending turned on it is possible
to end up in a situation where the prefix is not withdrawn
from downstream peers.
Here is the timing that I believe is happening:
a) have 2 paths to a peer.
b) receive a withdrawal from 1 path, set BGP_NODE_FIB_INSTALL_PENDING
and send the route install to zebra.
c) receive a withdrawal from the other path.
d) At this point we have a dest->flags set BGP_NODE_FIB_INSTALL_PENDING
old_select the path_info going away, new_select is NULL
e) A bit further down we call group_announce_route() which calls
the code to see if we should advertise the path. It sees the
BGP_NODE_FIB_INSTALL_PENDING flag and says, nope.
f) the route is sent to zebra to withdraw, which unsets the
BGP_NODE_FIB_INSTALL_PENDING.
g) This function winds up and deletes the path_info. Dest now
has no path infos.
h) BGP receives the route install(from step b) and unsets the
BGP_NODE_FIB_INSTALL_PENDING flag
i) BGP receives the route removed from zebra (from step f) and
unsets the flag again.
We know if there is no new_select, let's go ahead and just
unset the PENDING flag to allow the withdrawal to go out
at the time when the second withdrawal is received.
Signed-off-by: Donald Sharp <sharpd@nvidia.com>
(cherry picked from commit 4e8eda74ec7d30ba84e7f53f077f4b896728505a)
|
|
tools: Fix frr-reload for ebgp-multihop TTL reconfiguration. (backport #17946)
|
|
In ebgp-multihop, there is a difference in reload behavior when TTL is
unspecified (meaning default 255) and when 255 is explicitly specified.
For example, when reloading with 'neighbor <neighbor> ebgp-multihop
255' in the config, the following difference is created. This commit
fixes that.
Lines To Delete
===============
router bgp 65001
no neighbor 10.0.0.4 ebgp-multihop
exit
Lines To Add
============
router bgp 65001
neighbor 10.0.0.4 ebgp-multihop 255
exit
The commit 767aaa3a8048 is not sufficient and frr-reload needs to be
fixed to handle both unspecified and specified cases.
Signed-off-by: Nobuhiro MIKI <nob@bobuhiro11.net>
(cherry picked from commit 594e917656da5502b302309aed3cf596df24713f)
|
|
opensourcerouting/fix/revert_4338e21aa2feba57ea7004c36362e5d8186340b8_10.1
Revert "bgpd: Handle Addpath capability using dynamic capabilities" (backport)
|
|
This reverts commit 05cf9d03b345393b8d63ffe9345c42debd8362b6.
TL;DR; Handling BGP AddPath capability is not trivial (possible) dynamically.
When the sender is AddPath-capable and sends NLRIs encoded with AddPath ID,
and at the same time the receiver sends AddPath capability "disable-addpath-rx"
(flag update) via dynamic capabilities, both peers are out of sync about the
AddPath state. The receiver thinks already he's not AddPath-capable anymore,
hence it tries to parse NLRIs as non-AddPath, while they are actually encoded
as AddPath.
AddPath capability itself does not provide (in RFC) any mechanism on backward
compatible way to handle NLRIs if they come mixed (AddPath + non-AddPath).
This explains why we have failures in our CI periodically.
Signed-off-by: Donatas Abraitis <donatas@opensourcerouting.org>
|
|
Backport 17229 some to 10 1
|
|
0 __pthread_kill_implementation (no_tid=0, signo=6, threadid=130719886083648) at ./nptl/pthread_kill.c:44
1 __pthread_kill_internal (signo=6, threadid=130719886083648) at ./nptl/pthread_kill.c:78
2 __GI___pthread_kill (threadid=130719886083648, signo=signo@entry=6) at ./nptl/pthread_kill.c:89
3 0x000076e399e42476 in __GI_raise (sig=6) at ../sysdeps/posix/raise.c:26
4 0x000076e39a34f950 in core_handler (signo=6, siginfo=0x76e3985fca30, context=0x76e3985fc900) at lib/sigevent.c:258
5 <signal handler called>
6 __pthread_kill_implementation (no_tid=0, signo=6, threadid=130719886083648) at ./nptl/pthread_kill.c:44
7 __pthread_kill_internal (signo=6, threadid=130719886083648) at ./nptl/pthread_kill.c:78
8 __GI___pthread_kill (threadid=130719886083648, signo=signo@entry=6) at ./nptl/pthread_kill.c:89
9 0x000076e399e42476 in __GI_raise (sig=sig@entry=6) at ../sysdeps/posix/raise.c:26
10 0x000076e399e287f3 in __GI_abort () at ./stdlib/abort.c:79
11 0x000076e39a39874b in _zlog_assert_failed (xref=0x76e39a46cca0 <_xref.27>, extra=0x0) at lib/zlog.c:789
12 0x000076e39a369dde in cancel_event_helper (m=0x5eda32df5e40, arg=0x5eda33afeed0, flags=1) at lib/event.c:1428
13 0x000076e39a369ef6 in event_cancel_event_ready (m=0x5eda32df5e40, arg=0x5eda33afeed0) at lib/event.c:1470
14 0x00005eda0a94a5b3 in bgp_stop (connection=0x5eda33afeed0) at bgpd/bgp_fsm.c:1355
15 0x00005eda0a94b4ae in bgp_stop_with_notify (connection=0x5eda33afeed0, code=8 '\b', sub_code=0 '\000') at bgpd/bgp_fsm.c:1610
16 0x00005eda0a979498 in bgp_packet_add (connection=0x5eda33afeed0, peer=0x5eda33b11800, s=0x76e3880daf90) at bgpd/bgp_packet.c:152
17 0x00005eda0a97a80f in bgp_keepalive_send (peer=0x5eda33b11800) at bgpd/bgp_packet.c:639
18 0x00005eda0a9511fd in peer_process (hb=0x5eda33c9ab80, arg=0x76e3985ffaf0) at bgpd/bgp_keepalives.c:111
19 0x000076e39a2cd8e6 in hash_iterate (hash=0x76e388000be0, func=0x5eda0a95105e <peer_process>, arg=0x76e3985ffaf0) at lib/hash.c:252
20 0x00005eda0a951679 in bgp_keepalives_start (arg=0x5eda3306af80) at bgpd/bgp_keepalives.c:214
21 0x000076e39a2c9932 in frr_pthread_inner (arg=0x5eda3306af80) at lib/frr_pthread.c:180
22 0x000076e399e94ac3 in start_thread (arg=<optimized out>) at ./nptl/pthread_create.c:442
23 0x000076e399f26850 in clone3 () at ../sysdeps/unix/sysv/linux/x86_64/clone3.S:81
(gdb) f 12
12 0x000076e39a369dde in cancel_event_helper (m=0x5eda32df5e40, arg=0x5eda33afeed0, flags=1) at lib/event.c:1428
1428 assert(m->owner == pthread_self());
In this decode the attempt to cancel the connection's events from
the wrong thread is causing the crash. Modify the code to create an
event on the bm->master to cancel the events for the connection.
Signed-off-by: Donald Sharp <sharpd@nvidia.com>
|
|
(gdb) bt
0 futex_wait (private=0, expected=2, futex_word=0x5c438e9a98d8) at ../sysdeps/nptl/futex-internal.h:146
1 __GI___lll_lock_wait (futex=futex@entry=0x5c438e9a98d8, private=0) at ./nptl/lowlevellock.c:49
2 0x00007af16d698002 in lll_mutex_lock_optimized (mutex=0x5c438e9a98d8) at ./nptl/pthread_mutex_lock.c:48
3 ___pthread_mutex_lock (mutex=0x5c438e9a98d8) at ./nptl/pthread_mutex_lock.c:93
4 0x00005c4369c17e70 in _frr_mtx_lock (mutex=0x5c438e9a98d8, func=0x5c4369dc2750 <__func__.265> "bgp_notify_send_internal") at ./lib/frr_pthread.h:258
5 0x00005c4369c1a07a in bgp_notify_send_internal (connection=0x5c438e9a98c0, code=8 '\b', sub_code=0 '\000', data=0x0, datalen=0, use_curr=true) at bgpd/bgp_packet.c:928
6 0x00005c4369c1a707 in bgp_notify_send (connection=0x5c438e9a98c0, code=8 '\b', sub_code=0 '\000') at bgpd/bgp_packet.c:1069
7 0x00005c4369bea422 in bgp_stop_with_notify (connection=0x5c438e9a98c0, code=8 '\b', sub_code=0 '\000') at bgpd/bgp_fsm.c:1597
8 0x00005c4369c18480 in bgp_packet_add (connection=0x5c438e9a98c0, peer=0x5c438e9b6010, s=0x7af15c06bf70) at bgpd/bgp_packet.c:151
9 0x00005c4369c19816 in bgp_keepalive_send (peer=0x5c438e9b6010) at bgpd/bgp_packet.c:639
10 0x00005c4369bf01fd in peer_process (hb=0x5c438ed05520, arg=0x7af16bdffaf0) at bgpd/bgp_keepalives.c:111
11 0x00007af16dacd8e6 in hash_iterate (hash=0x7af15c000be0, func=0x5c4369bf005e <peer_process>, arg=0x7af16bdffaf0) at lib/hash.c:252
12 0x00005c4369bf0679 in bgp_keepalives_start (arg=0x5c438e0db110) at bgpd/bgp_keepalives.c:214
13 0x00007af16dac9932 in frr_pthread_inner (arg=0x5c438e0db110) at lib/frr_pthread.c:180
14 0x00007af16d694ac3 in start_thread (arg=<optimized out>) at ./nptl/pthread_create.c:442
15 0x00007af16d726850 in clone3 () at ../sysdeps/unix/sysv/linux/x86_64/clone3.S:81
(gdb)
The bgp keepalive pthread gets deadlocked with itself and consequently
the bgp master pthread gets locked when it attempts to lock
the peerhash_mtx, since it is also locked by the keepalive_pthread
The keepalive pthread is locking the peerhash_mtx in
bgp_keepalives_start. Next the connection->io_mtx mutex in
bgp_keepalives_send is locked and then when it notices a problem it invokes
bgp_stop_with_notify which relocks the same mutex ( and of course
the relock causes it to get stuck on itself ). This generates a
deadlock condition.
Modify the code to only hold the connection->io_mtx as short as
possible.
Signed-off-by: Donald Sharp <sharpd@nvidia.com>
|
|
bgpd: Fix for local interface MAC cache issue in 'bgp mac hash' table (backport #17888)
|
|
Issue:
During FRR restart, we fail to add some of the local interface's MAC
to the 'bgp mac hash'. Not having local MAC in the hash table can cause
lookup issues while receiving EVPN RT-2.
Currently, we have code to add local MAC(bgp_mac_add_mac_entry) while handling
interface add/up events in BGP(bgp_ifp_up/bgp_ifp_create). But the code
'bgp_mac_add_mac_entry' in bgp_ifp_create is not getting invoked as it
is placed under a specific check(vrf->bgp link check).
Fix:
We can skip this check 'vrf->bgp link existence' as the tenant VRF might
not have BGP instance but still we want to cache the tenant VRF local
MACs. So keeping this check in bgp_ifp_create inline with bgp_ifp_up.
Ticket: #4204154
Signed-off-by: Krishnasamy R <krishnasamyr@nvidia.com>
(cherry picked from commit 016528364e686fb3b23a688707bd6ae6c5ea5f41)
|
