HomeEIPsNewsletter
EIPsEIP-7545
EIP-7545

Verkle proof verification precompile

Add a precompile to help dapps verify verkle proofs
StagnantStandards Track: Core
Created: 2023-10-13
Guillaume Ballet (@gballet), Diederik Loerakker (@protolambda)
Discussions ForumOriginal Proposal LinkEdit
1 min read
Anyone may contribute to propose contents.
Go propose
Video
Anyone may contribute to propose contents.
Go propose
Original

Abstract

This EIP proposes the addition of a precompiled contract to provide up-to-date state proof verification capabilities to smart contracts in a stateless Ethereum context.

Motivation

The proposed proof systems for stateless Ethereum require an upgrade to many tools and applications, that need a simple path to keep their proving systems up-to-date, without having to develop and deploy new proving libraries each time another proof format must be supported.

Specification

The key words "MUST", "MUST NOT", "REQUIRED", "SHALL", "SHALL NOT", "SHOULD", "SHOULD NOT", "RECOMMENDED", "NOT RECOMMENDED", "MAY", and "OPTIONAL" in this document are to be interpreted as described in RFC 2119 and RFC 8174.

A precompiled contract is added at address 0x21, wrapping the stateless ethereum proof verification function.

The precompile's input is the tightly packed concatenation of the following fields:

  • version (1 byte) specifies which version of the stateless proof verification function should be used. Version 0 is used for an MPT and version 1 is used for the polynomial commitment scheme multiproof used in EIP-6800.
  • state_root (32 bytes) specifies the state root that the proof is proving against.
  • proof_data (arbitrary long) is the proof data.

Pseudo-code behavior of the precompile:

def proof_verification_precompile(input): version = input[0] state_root = input[1:33] proof_data = input[33:33+proof_data_size] if version == 0: proof = deserialize_proof(state_root, proof_data) return verify_mpt_multiproof(proof) if version == 1: proof = deserialize_proof(state_root, proof_data) return verify_pcs_multiproof(proof) return 0

If version is 0 then the proof is expected to follow the SSZ format described in "the verge" proposal in the consensus spec.

The precompile returns 1 if it was able to verify the proof, and 0 otherwise.

Gas costs

Constant namecost
POINT_COSTTBD
POLY_EVAL_COSTTBD

The precompile cost is:

cost = (POINT_COST + 1)*len(get_commitments(input)) + POLY_EVAL_COST * [leaf_depth(key, get_tree(input)) for key in get_keys(input))]

where:

  • get_commitments extracts the list of commitments in the proof, as encoded in input
  • get_keys extracts the list of keys in the proof, as encoded in input
  • leaf_depth returns the depth of the leaf in the tree
  • get tree reconstruct a stateless view of the tree from input

Rationale

Stateless Ethereum relies on proofs using advanced mathematical concepts and tools from a fast-moving area of cryptography. As a result, a soft-fork approach is currently favored in the choice of the proof format: proofs are going to be distributed outside of consensus, and in the future, stateless clients will be able to chose their favorite proof format.

This introduces a burden on several application, e.g. bridges, as they will potentially need to support proof formats designed after the release of the bridge contract.

Delegating the proof verification burden to a version-aware precompile will ensure that these applications can support newer proving primitives without having to upgrade their contracts.

Backwards Compatibility

No backward compatibility issues found.

Test Cases

TODO

Reference Implementation

WIP

  • First implementation in Optimism, pull request #192 of ethereum-optimism/op-geth by @protolambda

Security Considerations

Needs discussion.

Copyright and related rights waived via CC0.

Further reading
Anyone may contribute to propose contents.
Go propose

Not miss a beat of EIPs' update?

Subscribe EIPs Fun to receive the latest updates of EIPs Good for Buidlers to follow up.

View all
Serve Ethereum Builders, Scale the Community.
Resources
GitHub
Supported by