Add Ibc sample contract #714
Conversation
ethanfrey
force-pushed
the
ibc-sample-contract
branch
from
df092b4 to
ebb09be
Compare
ethanfrey
force-pushed
the
ibc-sample-contract
branch
from
35d07ab to
911187c
Compare
|
|
||
| #[entry_point] | ||
| /// we do nothing | ||
| /// TODO: remove the account from the lookup? |
There was a problem hiding this comment.
If you would propagated this event to the "reflect" contract then it would be able to do some "cleanup" as well. For example if the contract owns tokens for whatever reasons it could return them and would not lock them.
There was a problem hiding this comment.
I think this would never actually be called.
The flow I have is "to-be-defined-sender-contract on chain A" sends a packet to "ibc-reflect on chain B", which dispatches it.
ibc-reflect never sends a packet itself, thus should never get these two callbacks (we could error here I guess?)
| // and we are golden | ||
| Ok(IbcReceiveResponse { | ||
| acknowledgement, | ||
| messages: vec![wasm_msg.into()], |
There was a problem hiding this comment.
How can be ensure that only "authorized" messages are passed?
There was a problem hiding this comment.
When we create a channel, there is exactly one contract/module on the other side that can send packets on it, correct? (Please double-check this - this is key to all my authorization reasoning).
When they connect, we give them a fresh address that is theirs. Any messages they send get reflected by that custom address. Thus the channel -> address lookup.
If sending packets on the channel is not as limited as I believe, I need to revisit this.
There was a problem hiding this comment.
Note: We really need local (port_id, channel_id) pairs for uniqueness.
However, the receiving contract only has one port, so by fixing the channel_id, we fix the pair.
ethanfrey
force-pushed
the
ibc-sample-contract
branch
4 times, most recently
from
5ce3369 to
b87385f
Compare
|
Okay, I think this is done (and a bit of refactoring for future contracts as well). Happy for a final review, then I can merge it and use as a demo contract for |
ethanfrey
force-pushed
the
ibc-sample-contract
branch
from
4983720 to
9cbc21d
Compare
|
This should be ready for a merge now. I will start integrating the last commit (62b9fc7) into CosmWasm/wasmvm#167 |
Addresses #692
It builds on some other items:
#711 to define the entry points we use (MERGED)
#716 to add stargate support and ibc entry points to the vm (MERGED)
ibc-reflectcontractNote: a unit test brought up CosmWasm/serde-json-wasm#27 but this is not blocked on that (it parses correctly in the vm code, and we never parse this type in contracts).