Round instant unstake fee up instead of down

[to-sh-1]

Reference issue:
code-423n4/2022-06-yieldy-findings#9

[0xean]

I think this attack is super trivial. In the current implementation if the amount = 1, the amountMinusFee will be 0 which, afaict has no ill effects really.

Just trying to confirm there are no edge cases that can break this, otherwise I think the fix LGTM

[to-sh-1]

I think this attack is super trivial. In the current implementation if the amount = 1, the amountMinusFee will be 0 which, afaict has no ill effects really.

Just trying to confirm there are no edge cases that can break this, otherwise I think the fix LGTM

I agree, I just think rounding the fee up vs down does make sense in case this does get exported to a L2 with 0 fees. We could hold off merging if we wanted as this will have no benefit on the ETH blockchain