add data in genesis

[juniuszhou]

No description provided.

[codecov]

Codecov Report

Merging #3062 (cef49c5) into future/peggy2 (ef044b7) will increase coverage by 0.22%.
The diff coverage is 90.18%.

@@                Coverage Diff                @@
##           future/peggy2    #3062      +/-   ##
=================================================
+ Coverage          47.96%   48.18%   +0.22%     
=================================================
  Files                169      169              
  Lines              15024    15120      +96     
=================================================
+ Hits                7206     7286      +80     
- Misses              7385     7395      +10     
- Partials             433      439       +6     

Impacted Files	Coverage Δ
x/ethbridge/keeper/ethereumLockBurnSequence.go	90.32% <86.04%> (-9.68%)	⬇️
x/ethbridge/keeper/globalSequence.go	91.66% <91.66%> (ø)

[banshee]

Can we switch

func (k Keeper) getKeyViaNetworkDescriptorGlobalNonce(networkDescriptor types.NetworkDescriptor,
	globalSequence uint64) []byte {
	bs1 := make([]byte, 4)
	binary.LittleEndian.PutUint32(bs1, uint32(networkDescriptor))

	bs2 := make([]byte, 8)
	binary.LittleEndian.PutUint64(bs2, globalSequence)

	storeKey := append(append(types.GlobalNonceProphecyIDPrefix, bs1[:]...), bs2[:]...)
	return storeKey
}

to also use bigendian? If we use bigendian here, there's no littleendian in our code at all.

[juniuszhou]

Can we switch

func (k Keeper) getKeyViaNetworkDescriptorGlobalNonce(networkDescriptor types.NetworkDescriptor,
	globalSequence uint64) []byte {
	bs1 := make([]byte, 4)
	binary.LittleEndian.PutUint32(bs1, uint32(networkDescriptor))

	bs2 := make([]byte, 8)
	binary.LittleEndian.PutUint64(bs2, globalSequence)

	storeKey := append(append(types.GlobalNonceProphecyIDPrefix, bs1[:]...), bs2[:]...)
	return storeKey
}

to also use bigendian? If we use bigendian here, there's no littleendian in our code at all.

yes, it is better to use bigendian. will update it.

[smartyalgo]

Curious, why is bigEndian better?

[juniuszhou]

Curious, why is bigEndian better?

just use the same endian in the code, I have other PR #2748 to use the protobuf. I think it is better solution, can avoid some mistake for different endian usage.

[smartyalgo]

Is it necessary to include ethereum_lock_burn_sequence in genesis?
If the entry doesn't exist in keeper, keeper returns 0, which is our desired behavior

[juniuszhou]

1. if ethereum_lock_burn_sequence not set before, then no data will be included in genesis, it doesn't generate any empty/default value
2. 0 is expected if no vallue exists, to avoid such confusion, we set the lockBurnSequence starting from 1, also in contract

[smartyalgo]

Need to update x/ethbridge/genesis.go::ValidateGenesis function

[juniuszhou]

CrosschainFeeReceiveAccount already checked, for other three items, they are map, so that they are empty is correct if no any data before in peggy1.0. Do you mean we should check each entry in these maps?

[smartyalgo]

This looks very error prone.
I suggest write a function to extract NetworkDescriptor and ValidatorAddress, and use those to call SetEthereumLockBurnSequence

[juniuszhou]

Ok, will add extract method.

[banshee]

Are these comments still current?

[juniuszhou]

I implemented a new method DecodeKey to extract NetworkDescriptor and ValidatorAddress according to comment. please check.

[juniuszhou]

need deal with map issue first.