bad memcpy() in dhcp_socket.c, line 253 about strongswan

[tobiasbrunner]

Which technically is fine because id is an uint32_t. So it's not really a bug (our static analyzers would have complained a long time ago if that was the case).

What fix would you suggest? Setting the length to 4 or base it directly on client_hw_addr would similarly be an issue if id was smaller than 4. So maybe min(sizeof(dhcp->client_hw_addr)-2, sizeof(id))? Or even use a dynamic offset based on the size of id in relation to the size of client_hw_addr? That seems overly complicated for something that's not really an issue in the first place.

[harridu]

Sorry, you are right, I was too blind to see.

[tobiasbrunner]

No worries. It's nice the code gets looked at :)