[ETCM-746] Add support for PoW with Keccak-256 (ECIP-1049)

[leo-bogastry]

Added support for PoW with Keccak-256 (ECIP-1049)
Renamed several classes from Ethash to PoW because they will be used in both Ethash Pow and Keccak PoW

Description

Add support for PoW with Keccak-256 (ECIP-1049)
Renamed several classes from Ethash to PoW because they will be used in both Ethash Pow and Keccak PoW

Proposed Solution

Added KeccakCalculation object that computes the new mixHash and validates it against the block difficulty.
In order to control if a block should be mined using the new Keccak-Pow or current Ethash-Pow a simple if clause was added to the EthashMiner class. In ETCM-759 this will be refactored.

Added KeccakBlockHeaderValidator object that validates a block header for a block mined with the Keccak-Pow.
Added PoWBlockHeaderValidator class to controls whether (new)KeccakBlockHeaderValidator or (current) EthashBlockHeaderValidator should be used for the validation

NOTE:
Class EthashDAGManager is not new code it was extracted from EthashMiner.
Because I renamed the package, EthashMiner is marked as a new file, but only the code related with if (isKeccak(blockNumber))

The total number of files modified is big because of renaming

Testing

The branch was tested against etc network with both FastSync and RegularSync synchronization.

[bsuieric]

So we don't use it for now, just for test purposes? in ECIP-1049 I see there is an activation block

[leo-bogastry]

So we don't use it for now, just for test purposes? in ECIP-1049 I see there is an activation block

Where did you see that? I have the idea that that it is still not decided, on which block to do the switch, for mainnet. Or even if there's a final decision to really go forward with this

I don't understand what do you mean with "is this just for testing purposes?". Everything has to be tested before going to live (there's a test network called Astor that only mines with keccak, we will add support for that on ETCM-749)

[AnastasiiaL]

Looks good to me. Just minor comments about the code that was (probably) already there, but since we are touching these files - might as well improve it a bit

[AnastasiiaL]

Suggested change

	case (Some(`epoch`), Some(dag), Some(dagSize)) => (dag, dagSize)
	case (Some(_), Some(dag), Some(dagSize)) => (dag, dagSize)

[bsuieric]

So we don't use it for now, just for test purposes? in ECIP-1049 I see there is an activation block

Where did you see that? I have the idea that that it is still not decided, on which block to do the switch, for mainnet. Or even if there's a final decision to really go forward with this

I don't understand what do you mean with "is this just for testing purposes?". Everything has to be tested before going to live (there's a test network called Astor that only mines with keccak, we will add support for that on ETCM-749)

I saw isKeccak is always false based on the configuration

[leo-bogastry]

So we don't use it for now, just for test purposes? in ECIP-1049 I see there is an activation block

Where did you see that? I have the idea that that it is still not decided, on which block to do the switch, for mainnet. Or even if there's a final decision to really go forward with this
I don't understand what do you mean with "is this just for testing purposes?". Everything has to be tested before going to live (there's a test network called Astor that only mines with keccak, we will add support for that on ETCM-749)

I saw isKeccak is always false based on the configuration

Yes, at the moment yes. It will only be tested in Astor for now

[leo-bogastry]

I realized that the tests for the EthashMiner never run by default. I forced then to run locally and all passed