feat: Re-sign failed CCMs

[syan095]

Pull Request

Closes: PRO-263

Checklist

Please conduct a thorough self-review before opening the PR.

• I am confident that the code works.
• I have updated documentation where appropriate.

Summary

When CCM failed to broadcast, information regarding the called call is stored on-chain. This is so that the user can get the information about the call and broadcast it manually. When the chain moves to the next epoch, the call is then resigned, so the user can broadcast it in the new epoch. After the second epoch expires, the call is cleaned up from memory.

Workflow:

• Added a callback which is called when CCM message fails to be broadcasted. Egress ID, broadcast ID and other relevant information are stored.

  • IngressEgress::ccm_broadcast_failed()

• on_finalize: get failed CCM deposited during the previous epoch. For each call:

  • If the call failed during the previous epoch, schedule Re-signing:
    T::Broadcaster::threshold_resign(ccm.broadcast_id)
  • else: clean up broadcast storage for this call (expire the failed CCM)
    T::Broadcaster::clean_up_broadcast_storage(ccm.broadcast_id);

• After the threshold signing ceremony is completed, signature data is updated in the Broadcaster pallet.

• User can query failed ccm data to manually broadcast the call:

  • cf_failed_ccm_call(broadcast_id: BroadcastId): -> Option

[dandanlen]

I made some edits too because it was easier than commenting.

[dandanlen]

Keep an eye on #4207 , I imagine there might be some conflicts.

[syan095]

Keep an eye on #4207 , I imagine there might be some conflicts.

I've made a small change to the integration test harness's Polkadot signer so we can move past epoch 3 without error. This may conflict with Ramiz' change, but should be easy to resolve

[dandanlen]

Thanks Roy