chore: update triplewz/poseidon

[magik6k]

Related Issues

triplewz/poseidon#2

Proposed Changes

Curio pulls in the new version of the Poseidon library which switched to generic fr.Element. We need that to support SnapDeal decode in pure Go, and can't use the older version.

Additional Info

Checklist

Before you mark the PR ready for review, please make sure that:

• Commits have a clear commit message.
• PR title conforms with contribution conventions
• Update CHANGELOG.md or signal that this change does not need it per contribution conventions
• New features have usage guidelines and / or documentation updates in
  • Lotus Documentation
  • Discussion Tutorials
• Tests exist for new functionality or change in behavior
• CI is green

[Stebalien]

I'd prefer a released version, but this seems reasonable. From what I can tell, this is only used in sealing, we have a separate flow to check the proofs. That means this change should be covered by the itests.

But just to double check: @magik6k have you tested the lotus-miner sealing flow and/or confirmed that it's tested (enough) in CI? I now some proof types aren't generally tested because they're too slow to test in CI.

[magik6k]

From what I can tell, this is only used in sealing, we have a separate flow to check the proofs. That means this change should be covered by the itests.

It's only used with "External PC2 binary" mode of sealing, which IIUC isn't adopted pretty much anywhere (that was used for brownfield supraseal pc2, which while faster wasn't very stable). We also use the new version of this library in Curio for many other parts of the proving system, and it definitely works fine there.

[rjan90]

It's only used with "External PC2 binary" mode of sealing, which IIUC isn't adopted pretty much anywhere (that was used for brownfield supraseal pc2, which while faster wasn't very stable).

Can confirm that the "External PC2 binary" mode of sealing is not well adopted, because it has the downside of only working with CC-sectors currently: and switching between external PC2 binary / internal PC modes depending if it is data or not in the sectors, would be a very painful experience.