Fix merkle tree hashing for poseidon

[Leonard-Pat]

Pull Request type

Please check the type of change your PR introduces:

• Bugfix
• Feature
• Code style update (formatting, renaming)
• Refactoring (no functional changes, no API changes)
• Build-related changes
• Documentation content changes
• Other (please describe):

What is the current behavior?

Issue Number: N/A

What is the new behavior?

The current Poseidon hashing implementation for the merkle tree is incorrect. It uses

            let mut state = PoseidonTrait::new();
            state = state.update(data1);
            state = state.update(data2);
            state.finalize()

Under the hood this actually hashes the elements twice (once from the state update and once from finalization, see corelib), the intended behaviours is that only one round of hashing is done (when hashing two elements). This is noted in https://docs.starknet.io/documentation/architecture_and_concepts/Cryptography/hash-functions/#poseidon_hash where by the hash of two elements is simply
.
Also please see POC https://github.com/Leonard-Pat/merkle-tree-failing

• Fixes Poseidon hashing for merkle tree poseidon impl

Does this introduce a breaking change?

• Yes
• No

Other information

Updated tests to reflect this change - note the gas limit was exceeded and therefore needed to be udpated as well

[0xLucqs]

can you also fix pedersen plz

[Leonard-Pat]

can you also fix pedersen plz

I believe Pedersen is correct as the finalization step does not add another round of hashing

[0xLucqs]

ah indeed but i think that using directly the pedersen function would reduce a bit the step cost

[Leonard-Pat]

ah indeed but i think that using directly the pedersen function would reduce a bit the step cost

okay done :)