chore: Add dummy handler of MsgInjectedCheckpoint

[gitferry]

Closes #393

[Lazar955]

lgtm

[KonradStaniec]

Looks good !

Though one note here, we can't backport it / deploy it on phase-2 testnet without additional work.

The checkpoint transactions which were previously marked as failed i.e had code different than 0, now they will have code 0, and transaction results influence hash of transactions which is in block headers. https://github.com/cometbft/cometbft/blob/v0.38.16/types/results.go#L47 shows that code is taken into computing tx hash in header.

This leaves us with two options:

1. Schedule it for babylon v2.
2. Do coordinated binary upgrade on testnet

Tbh given this is not critical issue, I would leave it for Babylon v2

[gitferry]

@KonradStaniec MsgInjectedCheckpoint is currently been handled directly in PreBlocker. Why introducing this dummy handler would affect apphash?

[KonradStaniec]

@KonradStaniec MsgInjectedCheckpoint is currently been handled directly in PreBlocker. Why introducing this dummy handler would affect apphash?

Even though message is handled in PreBlocker, it is still transaction in the Block which means it gets executed during DeliverTx here - https://github.com/cosmos/cosmos-sdk/blob/8cc13ed2e2b4fca118b27cfaaaef71c2420aff06/baseapp/baseapp.go#L771. Thats why we see it as failed in explorers as its execution code is different than 0.

This is not part of the appHash but it is part of the LastResultsHash - https://github.com/cometbft/cometbft/blob/0d4d40f19dbf7e724b9a65b598b5431e733eee0c/types/block.go#L352

This means the node with with this change will produce and expects different headers that are in the current testnet, which will lead to chain split.

Actually , the easiest way to test this pr would be to try to sync the node build with this version with current testnet, this would verify:

1. Whether there is risk of chain split or not
2. Whether transactions which checkpoints are now considered successfull i.e have Code == 0

[gitferry]

@KonradStaniec Thanks for the explanation. It all makes sense. I agree that this pr is not a urgent fix can can leave it for v2 but will test it out soon