Repeated kernel panics since update to 22.1.10

[Patrick M. Hausen]

Hi all,

this can be a coincidence, because I also set up 3 additional VLANs with IPv4 and IPv6 at the same time and the panic seems to occur somewhere in the IPv6 filtering. Anyway, here's the panic to get this discussion started.

Hardware is an unmodified DEC690.

Any ideas? 

Thanks,
Patrick

Code: [Select]

Fatal trap 12: page fault while in kernel mode
cpuid = 0; apic id = 00
fault virtual address = 0x10
fault code = supervisor read data, page not present
instruction pointer = 0x20:0xffffffff80eb0b8d
stack pointer         = 0x28:0xfffffe000e1ca580
frame pointer         = 0x28:0xfffffe000e1ca6a0
code segment = base 0x0, limit 0xfffff, type 0x1b
= DPL 0, pres 1, long 1, def32 0, gran 1
processor eflags = interrupt enabled, resume, IOPL = 0
current process = 0 (if_io_tqg_0)
trap number = 12
panic: page fault
cpuid = 0
time = 1658741268
KDB: stack backtrace:
db_trace_self_wrapper() at db_trace_self_wrapper+0x2b/frame 0xfffffe000e1ca340
vpanic() at vpanic+0x17f/frame 0xfffffe000e1ca390
panic() at panic+0x43/frame 0xfffffe000e1ca3f0
trap_fatal() at trap_fatal+0x385/frame 0xfffffe000e1ca450
trap_pfault() at trap_pfault+0x4f/frame 0xfffffe000e1ca4b0
calltrap() at calltrap+0x8/frame 0xfffffe000e1ca4b0
--- trap 0xc, rip = 0xffffffff80eb0b8d, rsp = 0xfffffe000e1ca580, rbp = 0xfffffe000e1ca6a0 ---
ip6_forward() at ip6_forward+0x62d/frame 0xfffffe000e1ca6a0
pf_refragment6() at pf_refragment6+0x164/frame 0xfffffe000e1ca6f0
pf_test6() at pf_test6+0xfdb/frame 0xfffffe000e1ca860
pf_check6_out() at pf_check6_out+0x40/frame 0xfffffe000e1ca890
pfil_run_hooks() at pfil_run_hooks+0x97/frame 0xfffffe000e1ca8d0
ip6_tryforward() at ip6_tryforward+0x2ce/frame 0xfffffe000e1ca950
ip6_input() at ip6_input+0x60f/frame 0xfffffe000e1caa30
netisr_dispatch_src() at netisr_dispatch_src+0xb9/frame 0xfffffe000e1caa80
ether_demux() at ether_demux+0x138/frame 0xfffffe000e1caab0
ether_nh_input() at ether_nh_input+0x355/frame 0xfffffe000e1cab10
netisr_dispatch_src() at netisr_dispatch_src+0xb9/frame 0xfffffe000e1cab60
ether_input() at ether_input+0x69/frame 0xfffffe000e1cabc0
ether_demux() at ether_demux+0x121/frame 0xfffffe000e1cabf0
ether_nh_input() at ether_nh_input+0x355/frame 0xfffffe000e1cac50
netisr_dispatch_src() at netisr_dispatch_src+0xb9/frame 0xfffffe000e1caca0
ether_input() at ether_input+0x69/frame 0xfffffe000e1cad00
iflib_rxeof() at iflib_rxeof+0xc27/frame 0xfffffe000e1cae00
_task_fn_rx() at _task_fn_rx+0x72/frame 0xfffffe000e1cae40
gtaskqueue_run_locked() at gtaskqueue_run_locked+0x15d/frame 0xfffffe000e1caec0
gtaskqueue_thread_loop() at gtaskqueue_thread_loop+0xc2/frame 0xfffffe000e1caef0
fork_exit() at fork_exit+0x7e/frame 0xfffffe000e1caf30
fork_trampoline() at fork_trampoline+0xe/frame 0xfffffe000e1caf30
--- trap 0, rip = 0xffffffff80c2b90f, rsp = 0, rbp = 0 ---
mi_startup() at mi_startup+0xdf
KDB: enter: panic
panic.txt0600001214267461024  7137 ustarrootwheelpage faultversion.txt0600007014267461024  7536 ustarrootwheelFreeBSD 13.0-STABLE stable/22.1-n248088-47fa565344c SMP

[franco]

Known issue, hard to work on this for me still (IPv6 fragments over production box with gateway set / route-to use).

As a workaround set tunable "net.pf.share_forward6" with value "0".


Cheers,
Franco

[Patrick M. Hausen]

Known issue, hard to work on this for me still (IPv6 fragments over production box with gateway set / route-to use).

Motivation: in our office LAN we do fancy things with NPT and VPNs etc. I added a guest VLAN over the weekend. So to make sure guests are only able to use "the Internet" I explicitly set the gateway to the DSL uplink in addition to firewall rules.

As a workaround set tunable "net.pf.share_forward6" with value "0".

Hero! Thanks!