Uh oh!
There was an error while loading. Please reload this page.
feat: Forward ports to testnest v1.0.7 - #99
Conversation
bc8e4f1 to
bb00d3eComparebb00d3e to
00921d1Compare239b6d7 to
cd437daComparecd437da to
141d8efComparedcf3a89 to
ba77a09Compareb51cfd6 to
c915c38Comparec915c38 to
54bb361Compare11dbba3 to
fce4b10CompareThere was a problem hiding this comment.
A couple of nits below.
There will most likely need to be additional changes related to #38 (010f2af) in recursion/program/src/stark.rs and prover/src/verify.rs, specifically the checks I commented out/changed slightly (I think this PR reintroduces the checks and they're still broken and need to be removed), but I'll take care of those. I'm currently generating a couple plonk proofs with >1 shard now to check the recursive constraints and will report back after committing the necessary changes
We'll also need a companion PR updating sphinx-contracts and update the s3 bucket artifacts, I can look into that done once this is ready to merge
Uh oh!
There was an error while loading. Please reload this page.
Uh oh!
There was an error while loading. Please reload this page.
Uh oh!
There was an error while loading. Please reload this page.
Uh oh!
There was an error while loading. Please reload this page.
There was a problem hiding this comment.
Looks like all of the changes from 010f2af were already integrated, so I just removed the commented out blocks of code and TODOs, as well as removing the now-unnecessary builder.assert_var_ne(index, C::N::from_canonical_usize(EMPTY)); constraint that was added
I should be able to upload the circuit artifacts and look into the sphinx-contracts PR later, but this PR is good to merge right now
Thanks a lot for all the hard work with the ports!
huitseeker
commented
Jul 31, 2024
Thanks a lot for the review and fixes @wwared ! |
This ports us to parity with SP1 testnest v1.0.7
One leftover item to note (for @samuelburnham and I to take on): upstream tests now run on docker by default (see recursion/gnark-ffi/src/docker.rs). The default-ness of this behaviour, as well as adapting the pointers to Succinct-hosted docker images, have not been performed in this PR. See #100 for the follow-up task.