Gas optimizations

[0xbok]

I think the following will give gas savings in the decreasing order of savings:

• self.depth is read several times in a function and in for loops, which can be avoided by caching it in memory first. It avoids expensive SLOADs. An example here: https://github.com/privacy-scaling-explorations/zk-kit/blob/db0af5cc0bc1c50691a306e951ddca345a99702e/packages/incremental-merkle-tree.sol/contracts/IncrementalBinaryTree.sol#L162-L168

• depth can be made uint256 instead of uint8. This can give gas savings as while reading self.depth, it won't need to unpack to uint8 since storage reads the entire slot. https://github.com/privacy-scaling-explorations/zk-kit/blob/db0af5cc0bc1c50691a306e951ddca345a99702e/packages/incremental-merkle-tree.sol/contracts/IncrementalBinaryTree.sol#L9

• i++ in the for loops can be instead incremented through:

  unchecked {
  ++i;
  }

I tried verifying that these steps actually give savings, but didn't find a way to generate gas report since it uses jest, which I'm not familiar with.

[cedoor]

Hey @0xbok, thank you very much! The ZK-kit solidity packages don't use Jest, they uses Mocha. You can run yarn test:report-gas to generate a report and check your performance improvements, from the packages/incremental-merkle-tree.sol folder. Let me know if it works!

[0xbok]

Thanks @cedoor, I checked locally and there were significant gas savings, created the PR.