[Bug 282434] LOR: tcphash vs rib head lock

From: <bugzilla-noreply_at_freebsd.org>
Date: Wed, 30 Oct 2024 22:41:51 UTC
https://bugs.freebsd.org/bugzilla/show_bug.cgi?id=282434

            Bug ID: 282434
           Summary: LOR: tcphash vs rib head lock
           Product: Base System
           Version: 15.0-CURRENT
          Hardware: amd64
                OS: Any
            Status: New
          Severity: Affects Only Me
          Priority: ---
         Component: kern
          Assignee: bugs@FreeBSD.org
          Reporter: lexi.freebsd@le-fay.org

noticed when booting a debug kernel for an unrelated issue; the LOR doesn't
seem to cause any obvious problems.

smbus1: <System Management Bus> on ichsmb0
bridge0: Ethernet address: 58:9c:fc:10:ff:f3
epair0a: Ethernet address: 02:f4:8c:22:b3:0a
epair0b: Ethernet address: 02:f4:8c:22:b3:0b
epair0a: link state changed to UP
epair0b: link state changed to UP
epair1a: Ethernet address: 02:46:a2:57:cc:0a
epair1b: Ethernet address: 02:46:a2:57:cc:0b
epair1a: link state changed to UP
epair1b: link state changed to UP
epair2a: Ethernet address: 02:2d:f8:9d:57:0a
epair2b: Ethernet address: 02:2d:f8:9d:57:0b
epair2a: link state changed to UP
epair2b: link state changed to UP
epair3a: Ethernet address: 02:0c:1d:89:af:0a
epair3b: Ethernet address: 02:0c:1d:89:af:0b
epair3a: link state changed to UP
epair3b: link state changed to UP
igc3: link state changed to UP
debugnet_any_ifnet_update: Bad dn_init result from igc3 (ifp
0xfffff80003826000), ignoring.
lo0: link state changed to UP
debugnet_any_ifnet_update: Bad dn_init result from igc3 (ifp
0xfffff80003826000), ignoring.
igc3: link state changed to DOWN
wg0: link state changed to UP
wg1: link state changed to UP
wg2: link state changed to UP
ix0: link state changed to UP
ix0.911: link state changed to UP
ix0.3004: link state changed to UP
igc3: link state changed to UP
pflog0: promiscuous mode enabled
lo0: link state changed to UP
bridge0: Ethernet address: 42:4e:fb:0f:98:45
lo0: link state changed to UP
lock order reversal:
 1st 0xfffffe001d7e35d0 tcphash (tcphash, sleep mutex) @
/data/build/src/freebsd/lf/main/sys/netinet/tcp_usrreq.c:1515
 2nd 0xfffff8000fe6a8e0 rib head lock (rib head lock, rm) @
/data/build/src/freebsd/lf/main/sys/netinet6/in6_fib.c:164
lock order tcphash -> rib head lock attempted at:
#0 0xffffffff8052740d at witness_checkorder+0xbbd
#1 0xffffffff804ac0dc at _rm_rlock_debug+0x12c
#2 0xffffffff8065aba4 at fib6_lookup+0xc4
#3 0xffffffff80666539 at cache_route+0xe9
#4 0xffffffff80665cdc at selectroute+0x12c
#5 0xffffffff80664a83 at in6_selectsrc+0x323
#6 0xffffffff80664711 at in6_selectsrc_socket+0x41
#7 0xffffffff8066278a at in6_pcbconnect+0x25a
#8 0xffffffff806469c4 at tcp6_connect+0xa4
#9 0xffffffff80644a3f at tcp6_usr_connect+0x27f
#10 0xffffffff805558c0 at soconnectat+0xc0
#11 0xffffffff8055d7e7 at kern_connectat+0xe7
#12 0xffffffff8055d6d1 at sys_connect+0x81
#13 0xffffffff8079e798 at amd64_syscall+0x158
#14 0xffffffff8077123b at fast_syscall_common+0xf8

using src from ~22429a464a5f4f6bb5a056aae1353985db83b721.

-- 
You are receiving this mail because:
You are the assignee for the bug.