maramihali

Results 34 comments of maramihali

sys/linux: things to test

Working on FQ_PIE from include/uapi/linux/pkt_sched.h

tools/syz-verifier: cross-checking two kernels

Edit to this issue: `syz-verifier` aims to provide a utility for cross-checking any two kernels (with one important milestone being cross-checking gVisor against linux)

tools/syz-verifier: cross-checking two kernels

Also, #2609 and #2649 and #2637

`*RecursiveVerifier.verify_proof` should receive `StdlibProof` rather than `HonkProof` as argument

+ GoblinProof passed to the recursive verifier should also be a stdlib proof and translation evaluations should be constrained prior entering the `verify` function

Optimize batching MSM in Shplemini verifier

Edited as we will possibly do this for Shplemini not ZM

Pairing-input aggregation in Honk recursive verifier

This is an issue relevant for the RollupIVC milestone as well https://github.com/AztecProtocol/barretenberg/milestone/10 but apparently I cannot add two milestones :-? Another thing that occurred to me is that we (might)...

ECCVM: properly handle transcript polynomial univariate evaluations

Benchmark for native IPA ![image](https://github.com/AztecProtocol/barretenberg/assets/33263352/f060a8d0-2161-4e96-b205-c8eee81f78c6)

ECCVM: properly handle transcript polynomial univariate evaluations

Idea to avoid two IPA openings without reviving Shplonk: https://hackmd.io/_dcQgfPjTPutOopViZLv-Q

ECCVM: properly handle transcript polynomial univariate evaluations

https://github.com/AztecProtocol/aztec-packages/pull/7164 removes the need for a second opening but we still need the `hack_polynomial` for IPA to work.

ECCVM: Make ProverPolynomials/ProvingKey/Flavor model consistent with rest of codebase

In this arc of work - the flavor should also be cleaned up