22.05.b.20220513.0600 UPnP Crash

stephenw10

Can we see the backtrace from ddb.txt?

The output between db:0:kdb.enter.default> bt and db:0:kdb.enter.default> ps.

Steve

Bob.Dig

db:0:kdb.enter.default>  bt
Tracing pid 27958 tid 100558 td 0xfffff8003af7c740
kdb_enter() at kdb_enter+0x37/frame 0xfffffe003a45e070
vpanic() at vpanic+0x194/frame 0xfffffe003a45e0c0
panic() at panic+0x43/frame 0xfffffe003a45e120
trap_fatal() at trap_fatal+0x38f/frame 0xfffffe003a45e180
trap_pfault() at trap_pfault+0x4f/frame 0xfffffe003a45e1e0
calltrap() at calltrap+0x8/frame 0xfffffe003a45e1e0
--- trap 0xc, rip = 0xffffffff810a8fef, rsp = 0xfffffe003a45e2b0, rbp = 0xfffffe003a45e2e0 ---
pf_krule_to_nvrule() at pf_krule_to_nvrule+0x55f/frame 0xfffffe003a45e2e0
pfioctl() at pfioctl+0x79cf/frame 0xfffffe003a45e7d0
devfs_ioctl() at devfs_ioctl+0xb0/frame 0xfffffe003a45e820
VOP_IOCTL_APV() at VOP_IOCTL_APV+0x7b/frame 0xfffffe003a45e850
vn_ioctl() at vn_ioctl+0x16c/frame 0xfffffe003a45e960
devfs_ioctl_f() at devfs_ioctl_f+0x1e/frame 0xfffffe003a45e980
kern_ioctl() at kern_ioctl+0x298/frame 0xfffffe003a45e9f0
sys_ioctl() at sys_ioctl+0x100/frame 0xfffffe003a45eac0
amd64_syscall() at amd64_syscall+0x387/frame 0xfffffe003a45ebf0
fast_syscall_common() at fast_syscall_common+0xf8/frame 0xfffffe003a45ebf0
--- syscall (54, FreeBSD ELF64, sys_ioctl), rip = 0x8004960da, rsp = 0x7fffffffcc78, rbp = 0x7fffffffcce0 ---
db:0:kdb.enter.default>  ps

stephenw10

Do you know what Tixati is trying to open that triggers this?

Bob.Dig

@stephenw10 No, I only know that it is doing this right at the startup, no DLs needed. I would guess it is connecting to some DHT network and also checks if your port is open and UPnP is working.
And it has worked before.
But also I am not a regular user of its UPnP features, mostly using it in a VM together with a privacy VPN, so in general no UPnP for me.
Edit: Just to make sure I disabled UPnP in Pfsense and started tixati, no problem, so it is UPnP related.

stephenw10

Ok, we replicated that: https://redmine.pfsense.org/issues/13176

Nothing special is required, a single port mapping will hit it.

Steve

stephenw10

This should be fixed in the new snapshot. Looks good in my local testing.
Please retest and let us know.

Steve

Bob.Dig

@stephenw10 Working great now, thx guys!

havastamas

@stephenw10

Im facing with this issue again, lastest on lastest dev build. Any suggestions to solve this problem?

Thank you!

stephenw10

Same crash? Same backtrace? Todays snapshot?

w0w

@stephenw10
https://forum.netgate.com/topic/172502/fatal-trap-12-since-22-05-b-20220526-0600/6

stephenw10

@w0w said in 22.05.b.20220513.0600 UPnP Crash:

https://forum.netgate.com/topic/172502/fatal-trap-12-since-22-05-b-20220526-0600/6

Ah, OK. Continuing there...