Skip to main content

Local Execution

Transactions are initiated via a transaction execution request sent from the user to their local private execution environment (PXE). The PXE first executes the transaction locally in a simulation step, and then generates a zero-knowledge proof of correct execution. The PXE is then responsible for converting a transaction execution request into a transaction ready to be broadcasted to the network.

Execution request

A transaction execution request has the following structure. Note that, since Aztec uses full native account abstraction where every account is backed by a contract, a transaction execution request only needs to provide the contract address, function, and arguments of the initial call; nonces and signatures are arguments to the call, and thus opaque to the protocol.

originAztecAddressAddress of the contract where the transaction is initiated.
functionSelectoru32Selector (identifier) of the function to be called as entrypoint in the origin contract.
argsHashFieldHash of the arguments to be used for calling the entrypoint function.
txContextTxContextIncludes chain id, protocol version, and gas settings.
packedArgumentsPackedValues[]Preimages for argument hashes. When executing a function call with the hash of the arguments, the PXE will look for the preimage of that hash in this list, and expand the arguments to execute the call.
authWitnessesAuthWitness[]Authorization witnesses. When authorizing an action identified by a hash, the PXE will look for the authorization witness identified by that hash and provide that value to the account contract.

Simulation step

Upon receiving a transaction execution request to simulate, the PXE will locally execute the function identified by the given functionSelector in the given origin contract with the arguments committed to by argsHash. We refer to this function as the entrypoint. During execution, contracts may request authorization witnesses or expanded arguments from the execution oracle , which are answered with the packedArguments and authWitnesses from the request.

The entrypoint may enqueue additional function calls, either private or public. The simulation step will always execute all private functions in the call stack until emptied. The result of the simulation is a transaction object without an associated proof which is returned to the application that requested the simulation.

In terms of circuitry, the simulation step must execute all application circuits that correspond to private function calls, and then execute the private kernel circuit until the private call stack is empty. Note that circuits are only executed, there is no witness generation or proving involved.

Proving step

The proving step is similar to the simulation step, though witnesses are generated for all circuits and proven. Note that it is not necessary to execute the simulation step before the proving step, though it is desirable in order to provide the user with info on their transaction and catch any failed assertions early.

The output of the proving step is a transaction object with a valid proof associated, ready to be broadcasted to the network.