This repository has been archived by the owner on Jul 5, 2024. It is now read-only.
Refactor Tx circuit and use challenge api #803
Merged
Conversation
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
github-actions
bot
added
crate-circuit-benchmarks
han0110
force-pushed
the
feature/challenge-api-for-tx-circuit
branch
from
e716130 to
fe61a14
Compare
han0110
requested review from
CPerezz and
ed255
and removed request for
a team and
CPerezz
ed255
approved these changes
Sep 28, 2022
There was a problem hiding this comment.
LGTM! Amazing work! Not only you updated the circuit to use the challenge API, but also optimized it removing many rows, and also cleaned up some stuff :D
I've left some questions (so that I make sure i understood everything) and a suggestion, please take a look :)
han0110
force-pushed
the
feature/challenge-api-for-tx-circuit
branch
from
885034b to
ab7a796
Compare
han0110
force-pushed
the
feature/challenge-api-for-tx-circuit
branch
from
ab7a796 to
bf7e44d
Compare
ChihChengLiang
approved these changes
Oct 4, 2022
han0110
force-pushed
the
feature/challenge-api-for-tx-circuit
branch
from
bf7e44d to
bbb5921
Compare
han0110
force-pushed
the
feature/challenge-api-for-tx-circuit
branch
from
bbb5921 to
7f123a9
Compare
|
Although the spec has not been merged, but since privacy-scaling-explorations/zkevm-specs#292 only adds some more description to Tx circuit, so I think we can merge this first to unblock #816. |
StefanosChaliasos
referenced
this pull request
in Veridise/zkevm-circuits
Oct 7, 2022
* feat: refactor tx circuit and use challenge api for it * feat: remove `#[ignore]` for tests of tx circuit * fix: apply suggestion and clean up * fix: apply suggestion from @ChihChengLiang
GopherJ
pushed a commit
to dompute/zkevm-circuits
that referenced
this pull request
Sep 13, 2023
* support max snark upto 16 * more tests * fix bug in data input len * merge required features from max_snark=29 * clean up
Sign up for free
to subscribe to this conversation on GitHub.
Already have an account?
Sign in.
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
This PR aims to refactor Tx circuit with challenge API.
It reuses
MainGate's advice columns to create an extra random linear combination gate to get rid of 130 columns inSignVerifyChip, with the cost of few more rows to doSignVerifyChip::assign_signature_verifythan previous approach. So now tests and benchmark of Tx circuit can be ran locally.In respect to more rows (76 rows precisely) to verify a single signature, I think it should be negligible cause ecdsa verification already costs ~70k, and is still the bottleneck. Also even we have reached our bottleneck on ecdsa verification, we can actually just allocate another
MainGateto double our capacity (eachMainGatewithRangeChipcosts ~60 msm) without too much change to current code.There is also a caveat about multi-phase: The
AssignedCellwill havevalue: Value::unknownin any phase that is not equal to current one, because we need to follow the principle that the functiontoofassign_{fixed,advice}will only be called once (it's aFnMutso we need to guarantee that). So it's a little bit awkward to calculate later phase witness with previous phase'sAssignedCell, we need to pass theCellandValueseparately to make sure we can do the calculation and at the same time enforce the copy constraint.And currently this behavior is not shown in
MockProver, so it takes me some time to find this issue, I have also created an issue privacy-scaling-explorations/halo2#102 to propose to makeMockProverto do multi-phsae synthesis just like real prover.Todo