update kzgbn254 function to work with eigenDA encoded blobs #2
Conversation
|
Awesome work @afkbyte. Can we rebase and clean up comments etc? Would love to get this merged. |
arbitrator/prover/src/kzgbn254.rs
Outdated
|
|
||
| // blob header is the first 32 bytes of the blob bytes | ||
| let blob_header = blob_bytes[..32].to_vec(); | ||
| println!("blob header {:?}", blob_header); |
arbitrator/prover/src/kzgbn254.rs
Outdated
|
|
||
| temp_buffer.reverse(); | ||
| buffer.extend(temp_buffer); | ||
| } |
There was a problem hiding this comment.
Doesn't to_byte_array from the lib provide something similar or is this something more specific than that ?
There was a problem hiding this comment.
so it's a bit different, to_byte_array deserializes them in big endian but I need them in little endian
arbitrator/prover/src/utils.rs
Outdated
| @@ -297,26 +297,18 @@ pub fn hash_preimage(preimage: &[u8], ty: PreimageType) -> Result<[u8; 32]> { | |||
| 3000 | |||
| ).unwrap(); | |||
There was a problem hiding this comment.
You might want to use Prod numbers for these. In the lib, there is mainnet-data folder under tests. You can use those files to specify the input files.
|
Also looks like we're introducing cargo.lock dependency changes which are causing jit compilations to fail in CI - lets fix that too before merging https://github.com/Layr-Labs/nitro-private/actions/runs/9370515667/job/25797665060?pr=2 |
… to ignore kzg-bn254
works with blobs encoded by Layr-Labs/eigenda#583
also uses some modified functions from the kzg library, diff can be found here:
https://github.com/Layr-Labs/rust-kzg-bn254/compare/master...afkbyte:rust-kzg-bn254:master?expand=1
when we recieve the data from the preimage it's already in evaluation formation not coefficient (has been ifftd before being dispersed) so we don't need to do a
g1_iffton the g1 SRS bc kzgcommitment = <C, IFFT(SRS)> = <IFFT(C), SRS>