Kernel panic on 5.3.1-arch1-1-ARCH

Hello,

loading the kernel module and starting wg-quick causes a kernel panic on Arch Linux w/ the above mentioned kernel. I have uploaded a camera based screen shot to

https://www.nico.schottelius.org/temp/IMG_20191005_221751.jpg

I tried both the wireguard-arch and wireguard-dkms approach, both led to the same result.

Does this help anyone to debug the issue?

Best regards,

Nico

--
Modern, affordable, Swiss Virtual Machines. Visit www.datacenterlight.ch
_______________________________________________
WireGuard mailing list
[email protected]
https://lists.zx2c4.com/mailman/listinfo/wireguard

This isn't WireGuard, actually. It's a line in wg-quick's bash that says ip rule add ... suppress_prefix ....

You can reproduce the kernel crash without WireGuard anywhere in the equation with this sequence:

ip netns add crash
ip -n crash link add dummy1 type dummy
ip -n crash link set dummy1 up
ip -n crash -6 route add default dev dummy1
ip -n crash -6 rule add table main suppress_prefixlength 0 ip netns exec crash ping -f -c 1000 -W 1 1234::1 || true ip -n crash -6 rule del table main suppress_prefixlength 0 ip -n crash link del dummy1
ip netns del crash

I submitted a patch for this upstream here:

https://git.kernel.org/pub/scm/linux/kernel/git/davem/net.git/commit/?id=ca7a03c4175366a92cee0ccc4fec0038c3266e26

It's been backported to stable already. Update your kernel to 5.3.4 and things should work again for you.

Jason
_______________________________________________
WireGuard mailing list
[email protected]
https://lists.zx2c4.com/mailman/listinfo/wireguard

Hey Jason,

thanks for the quick reply - I' ll upgrade as soon as a new package is released and give a status update afterwards. Thanks for tracking it down!

Best,

Nico

--
Modern, affordable, Swiss Virtual Machines. Visit www.datacenterlight.ch
_______________________________________________
WireGuard mailing list
[email protected]
https://lists.zx2c4.com/mailman/listinfo/wireguard

... and just updated via the testing repo, everything green on 5.3.4.

On this system I see one error regarding the IPv4 fib table, which might be due to IPv6 only usage:

[root@diamond ~]# wg-quick up wg0
[#] ip link add wg0 type wireguard
[#] wg setconf wg0 /dev/fd/63
[#] ip -6 address add 2a0a:e5c1:137::137/48 dev wg0
[#] ip link set mtu 1420 up dev wg0
Error: ipv4: FIB table does not exist.
Dump terminated
[#] wg set wg0 fwmark 51820
[#] ip -6 route add ::/0 dev wg0 table 51820
[#] ip -6 rule add not fwmark 51820 table 51820
[#] ip -6 rule add table main suppress_prefixlength 0
[root@diamond ~]# ip a sh dev wg0
5: wg0: <POINTOPOINT,NOARP,UP,LOWER_UP> mtu 1420 qdisc noqueue state UNKNOWN group default qlen 1000
link/none
inet6 2a0a:e5c1:137::137/48 scope global
valid_lft forever preferred_lft forever
[root@diamond ~]# pacman -Q | grep wireguard
wireguard-dkms 0.0.20190913-1
wireguard-tools 0.0.20190913-1

As everything work[tm], would it be an option to rename it to a warning instead?

Best,

Nico

--
Modern, affordable, Swiss Virtual Machines. Visit www.datacenterlight.ch
_______________________________________________
WireGuard mailing list
[email protected]
https://lists.zx2c4.com/mailman/listinfo/wireguard