bulletproofs: add closure to allow increases to prover generator capacity #266
Conversation
|
|
||
| // Construct verifier generators | ||
| let pc_gens = PedersenGens::default(); | ||
| let mut bp_gens = BulletproofGens::new(128, 1); |
There was a problem hiding this comment.
Should there be an interface for the verifier to get the needed capacity for a proof (an added field in R1CSProof)?
vickiniu
changed the title
benches/r1cs.rs
Outdated
| @@ -156,7 +156,7 @@ impl KShuffleGadget { | |||
| .unzip(); | |||
|
|
|||
| Self::fill_cs(&mut prover, input_vars, output_vars)?; | |||
| let proof = prover.prove(&bp_gens)?; | |||
| let proof = prover.prove(bp_gens, |capacity, gens| gens.increase_capacity(capacity))?; | |||
There was a problem hiding this comment.
maybe instead of sending generators into prove, and then getting them into closure, instead simply provide the reference from the closure (not sure if borrowck would love this):
let gens = BulletproofGens::new(128, 1);
...
let proof = prover.prove(|capacity| gens.increase_capacity(capacity); &gens )?;?
There was a problem hiding this comment.
And the closure must return Result<&BulletproofGens, R1CSError> because the prover may decide to return Err if capacity is too high instead of blinding increasing it.
There was a problem hiding this comment.
Having the closure provide the reference makes the borrow checker unhappy 😢 says it can't infer the appropriate lifetime?
There was a problem hiding this comment.
The prover can provide a closure that doesn't resize the generators (which would result in a failure from insufficient capacity). Do we want to also have the closure return Result<(), R1CSError>? It feels maybe unnecessary to me since a no-op closure will have the same overall effect
Fixes #265