Module limits
Linea uses module limits to keep proof complexity manageable and ensure that the prover can generate proofs efficiently without overwhelming the infrastructure.
Linea's arithmetization specification is divided into modules, and each module represents a collection of EVM or Linea zkEVM operations (opcodes). Modules can:
- Process transaction data.
- Manage zkEVM memory.
- Coordinate interactions between modules.
Each module has a specific limit on the number of lines of data its operations can generate, regardless of its purpose. You can view these limits in the Linea source code. Additional lines are generated each time unique instructions are provided to a module's operations, but repeat calls with the same arguments do not generate new lines.
High-complexity transactions that repeatedly use a large volume of operations may generate too many lines of data and exceed a module's limit. Transactions that reach this point are rejected by the sequencer to ensure that the trace data passed to the prover is of a manageable scope, and that a proof can be generated.
Transactions rejected for exceeding module limits can be identified through the following methods, which return an error:
Use
linea_getTransactionExclusionStatusV1
to get the rejection status of a transaction hash, if it was sent and didn't get
rejected directly, nor included in a block.
This API method can be used to query the reason for transaction rejection in
edge cases where linea_estimateGas
or eth_sendRawTransaction
did not return
an error. It's intended to be used only in cases the transaction execution
differs between an RPC node and the sequencer.