Massive migrations data availability

[krlosMata]

Feature

Compute data availability for all processed transactions, including L1, transactions In order to provide massive migration feature. So far, data availability for L1 transactions was forced to be 0. Therefore it was not possible to proof that a certain forceTransfer transactions was done to a specific merkle tree index. By including L1 transactions into data availability, it is ensured that any transaction type is represented in smart contract calldata and be used for afterwards for any purpose such as massive migrations

L1 data availability

L1 data availability would be computed as L2 data availability with a slightly difference in amount field:

L1TxData:  [ NLevels  bits ] fromIdx
           [ NLevels  bits ] toIdx
           [    16  bits   ] effectiveAmountF
           [     8  bits   ] fee

• effectiveAmountF = amountF * (1 - nullifyAmount)
  • amountF is the amount used in the L1 transaction
  • nullifyAmount is computed and set to 1 if the L1 transaction is not valid

Note that this could not be for L2 transactions since L2 transactions must be valid in order to be processed

Steps

• [x] update L2DataAvailablity for L1Tx
• [x] add isAmountNullified in circuit

[krlosMata]

• circuit computes new signal isAmountNullified which is 0 for L2 transactions and 1 for L1 invalid transactions
• isAmountNullified is used to set to 0 the amountF field in data-availability before they are hashed into hash-inputs circuits
• decode-tx sets to 0 the userFee for L1 transactions