fix: incorrect tracking of num instantiations

[czarcas7ic]

Every time we instantiate a contract the classic way, we increment types.KeySequenceInstanceID: https://github.com/CosmWasm/wasmd/blob/e0bfaa52317a3d1f73bf26c5b9d61a777f63bd88/x/wasm/keeper/addresses.go#L17-L22
Every time we instantiate a contract with a predictable address (instantiate2), we do not increment this ID: https://github.com/CosmWasm/wasmd/blob/e0bfaa52317a3d1f73bf26c5b9d61a777f63bd88/x/wasm/keeper/addresses.go#L24-L32
In ImportGenesis, maxContractID takes into consideration ALL contracts: https://github.com/CosmWasm/wasmd/blob/e0bfaa52317a3d1f73bf26c5b9d61a777f63bd88/x/wasm/keeper/genesis.go#L45-L55
ImportGenesis expects the types.KeySequenceInstanceID to be greater than the number of contracts, but its not since we don’t always increment this value: https://github.com/CosmWasm/wasmd/blob/e0bfaa52317a3d1f73bf26c5b9d61a777f63bd88/x/wasm/keeper/genesis.go#L69-L72

This means that the discrepancy between the expected value and the actual value should be all contracts that were instantiated with predictable addresses.

This fix manually increments the KeyLastInstanceID to what we expect it to be. This way the check in InitGenesis passes. This also increments the KeyLastInstanceID on all instantiate2 calls. If chains manually set this key to the correct value in their upgrade handler, they can remove the manual setting of the KeyLastInstanceID in export genesis (but it's probably not a big deal if this isn't touched).

[alpe]

Thanks for the PR. It would be great if you can share bug reports on wasmd repo so that we can address them upstream.

[alpe]

Please see CosmWasm#1630

[czarcas7ic]

@alpe I was actually intending to upstream this myself, we were just internally testing (and still are testing) this with state exported testnets prior to opening a PR!