correct the calculation of wireguard header overhead

Thanks to Jason Donenfeld for the correction.

Signed-off-by: Tycho Andersen <tycho@docker.com>
This commit is contained in:
Tycho Andersen 2017-07-19 17:29:27 -06:00
parent 4fb767a480
commit b247cf5a68

View File

@ -40,7 +40,7 @@ Previous meeting notes: [2017-07-05](2017-07-05.md)
packet packet
- Kernel networking stack is the one allocating the packet buffers, - Kernel networking stack is the one allocating the packet buffers,
WireGuard doesn't allocate anything on top of this WireGuard doesn't allocate anything on top of this
- WireGuard overhead = UDP header + WireGuard = 60 bytes - WireGuard overhead: ipv4 header (20 bytes) + udp header (8 bytes) + wireguard header (4 byte type, 4 byte keyid, 8 byte nonce, 16 byte authtag = 32 bytes) = 60 bytes
- performance - performance
- fast, no copies across \*space - fast, no copies across \*space
- ChaChaPoly is fast - ChaChaPoly is fast