fix: Make rechecking a tx check the sequence number

[ValarDragon]

Upstream of my change to Osmosis that makes recheck still check sequence numbers.

This has been incorporated into Terra and Secret Network.

osmosis-labs#58

The antehandler check updated (despite its name) is responsible for checking two things:

1. Sequence Numbers
2. Signatures

Previously on recheck, it skipped the entire check, and therefore skipped both of its responsibilities.
Now it only skips the signature verification component, which was the purpose of recheck. (To save the heavy comptutional burden)

Every node with a default mempool, runs this ante handler check in "recheck" mode, on every single tx in their mempool upon each block being committed. Any tx with an invalidated sequence number will then get evicted from the mempool, ensuring that bad sequence numbers no longer enter in blocks produced by nodes with the default mempool.

You only really run into this as a noticeable problem once two conditions are satisfied:

1. At least one validator runs a modified mempool (Sentinel does this)
2. There are bots on the network, creating many txs from the same addr

which is the case for Osmosis, Terra, Secret, Evmos

---

Author Checklist

All items are required. Please add a note to the item if the item is not applicable and
please add links to any relevant follow up issues.

I have...

• included the correct type prefix in the PR title
• added ! to the type prefix if API or client breaking change
• targeted the correct branch (see PR Targeting)
• provided a link to the relevant issue or specification
• followed the guidelines for building modules
• included the necessary unit and integration tests
• added a changelog entry to CHANGELOG.md
• included comments for documenting Go code
• updated the relevant documentation or specification
• reviewed "Files changed" and left comments if necessary
• confirmed all CI checks have passed

Reviewers Checklist

All items are required. Please add a note if the item is not applicable and please add
your handle next to the items reviewed if you only reviewed selected items.

I have...

• confirmed the correct type prefix in the PR title
• confirmed ! in the type prefix if API or client breaking change
• confirmed all author checklist items have been addressed
• reviewed state machine logic
• reviewed API design and naming
• reviewed documentation is accurate
• reviewed tests and test coverage
• manually tested (if applicable)

[alexanderbez]

❤️

[robert-zaremba]

This is to clean the mempool from the transactions with the same sequence number, right?

[WhiteNoise1989]

@ValarDragon Does this mean that all user transactions that were not included in a block, will get evicted from mempool?

[WhiteNoise1989]

For example -

Account 'A' with expected sequence 's'. User sends a tx with sequence 's'. CheckTx runs, and passes(assuming all other checks were passed). CheckTx bumps account sequence to s+1. For some reason tx was not included in block. At this point account seq is s+1, but tx seq remains s. Now at end of block(since tx was not included in block), RecheckTx runs on above tx(with above pr this will check the account sequence), and fail bcz of account seq mismatch. Tx is evicted from mempool.

[tac0turtle]

This is to clean the mempool from the transactions with the same sequence number, right?

correct

[ValarDragon]

@ValarDragon Does this mean that all user transactions that were not included in a block, will get evicted from mempool?

No, the Tendermint mempool maintains an ordering s.t. a tx with sequence number s + 1 must come after a tx w/ sequence number s (or last sequence number on chain was s)

Therefore if tx w/ seq num s gets in a block, then s+1 will remain in the mempool. If competing txs w/ sequence number s, s+1, s+2 get on-chain, the tx with w/ seq s+1 in your mempool will get removed on the recheck.