Well, I might have spoken too soon. As soon as a connection was established to another wireguard peer (v. 0.0.20180118), the CentOS 7.5 server kernel panic'd.

Message from syslogd@tun001 at May 10 12:53:12 ...

kernel:Kernel panic - not syncing: Fatal exception in interrupt

Seeing something along the lines of this in the console output on boot:

[ 29.418257] wireguard: loading out-of-tree module taints kernel.

[ 29.454659] wireguard: WireGuard 0.0.20180420 loaded. See www.wireguard.com for information.

[ 29.598028] ------------[ cut here ]------------

[ 29.604356] kernel BUG at mm/slub.c:3601!

[ 29.609122] invalid opcode: 0000 [#1] SMP

[ 29.614013] Modules linked in: veth vxlan wireguard(OE) bridge stp llc ip_vs ip6_udp_tunnel udp_tunnel bonding ip

On Thu, May 10, 2018 at 12:31 PM Henry Snow <henry.snow@nielsen.com> wrote:

Perfect, that patch worked for me. Thank you for the quick fix!

On Thu, May 10, 2018 at 12:14 PM Jason A. Donenfeld <Jason@zx2c4.com> wrote:
Hi again,

I've fixed that here:
https://git.zx2c4.com/WireGuard/patch/?id=94f25e47b23ddeba26120de4ee3c098438fc2251

This will be part of the next snapshot. In the meantime, you can apply
that manually to the src directory, and perhaps Joe (CCd) can add it
to the rpm.

Jason

--
Henry Snow | Director, Site Reliability Engineering
eXelate, a Nielsen company
E: henry.snow@nielsen.com | M: 347.366.2700
675 6th Ave, 3rd Floor, New York, NY 10010