The suggestion of NAT to allow the same subnet to be used at both locations is interesting. But if I am understanding the original post correctly the objective is to have a L2 connection between sites. This suggests that a device at one site should be able to arp for a device at the other site. And NAT does not accomplish that.
The basic problem here is that GRE tunnels generally operate as L3 links, with the tunnel having its own IP subnet while devices at both ends have their own IP subnets. There is one option that you could consider, if your platforms support it. You might be able to enable bridging on the GRE interfaces by configuring Concurrent Routing and Bridging and putting both the GRE tunnel and the Lan interface into the bridge group. Note that this is officially "not supported". This means that it might work. But if something does not work as you expect Cisco has no responsibility to respond to that. It is fine for a lab and for learning purposes, but I would be very reluctant to do this in a live environment.
HTH
Rick
