Regression: secp256k1 consensus support

[joe-bowman]

Summary of Bug

Through tendermint/tendermint#5016 and tendermint/tendermint#5280 secp256k1 was removed (intentionally, but incorrectly) from Tendermint Core.

As a result, #7268 made changes to hard code consensus pubkey to ed25519 types.

Clearly, if the Tendermint change was - while good intentioned - a mistake, then this change must also be reverted.

This currently blocks networks using secp256k1 consensus (most notably Band) from migrating to Stargate.

Version

0.40

Steps to Reproduce

Try to use secp256k1 consensus key on a 0.40 network (with tendermint patched to revert secp256k1 removal).

PR

• remove totmPubkey #7652

---

For Admin Use

• Not duplicate issue
• Appropriate labels applied
• Appropriate contributors tagged
• Contributor assigned/self-assigned

[joe-bowman]

Tracking issue in Tendermint: tendermint/tendermint#5495

[tessr]

Thanks for reporting this! We're looking into reintroducing support for secp256k1 today.

[clevinson]

This seems important to get addressed before a final published Stargate release (or at the very least by v0.40.1)

@marbar3778 Are you tackling the SDK side of this or should we have someone from our team look at it?

[tac0turtle]

Oh I didn't know I was assigned to this, I'll take a look in the morning.

[tessr]

Ah, my apologies, everyone - I think I'd meant to assign Marko to the Tendermint side and I was confused; I didn't mean to be prescriptive about who or how this gets fixed on the SDK side! But Marko, thanks for taking this on.