fix: [Interchain Security] Add equality to validator.IsMature

[mpoke]

Closes cosmos/interchain-security#111 (see cosmos/interchain-security#111 (comment) for details).

[alexanderbez]

I think this is OK and true since it's within the context of the current executing block, but I'm not 100% certain as x/staking has the assumption of H+1 execution due to the current Tendermint model.

I see simulations failing so this makes me slightly even more worried. Regardless, let's add a changelog and then investigate sim failures.

[danwt]

Beware

• Validator.UnbondingHeight meaning is inconsistent, misused and not useful. #12486

[mpoke]

I think this is OK and true since it's within the context of the current executing block, but I'm not 100% certain as x/staking has the assumption of H+1 execution due to the current Tendermint model.

I see simulations failing so this makes me slightly even more worried. Regardless, let's add a changelog and then investigate sim failures.

@alexanderbez this PR shouldn't break the tests since the change only affects code that is called when Interchain Security (IS) is enabled, and I'm quite sure that there are no checks for that :) The function validator.IsMature() is called only in x/staking/keeper/unbonding.go when an external module (i.e., the CCV module) notifies the staking module that an unbonding operation (undelegation, redelegation, or validator unbonding) can complete from the perspective of IS (i.e., it calls UnbondingCanComplete()). Since the checks are failing also for the branch with all SDK changes needed for IS (https://github.com/cosmos/cosmos-sdk/tree/interchain-security-rebase), I assume that the problem is somewhere there. I've opened an issue on the IS repo, i.e., cosmos/interchain-security#227

[danwt]

I suggest closing this in favour of merging #12578 into #12554 and then merging #12554