We must delegate call to OPCM from superchain-ops

[blmalone]

This involves using @maurelian's new MultiCall3DelegateCall contract.

As part of the superchain-ops improvements, we must provide the ability to swap in different MultiCall3 contracts based on whether the task is using OPCM or not.

[maurelian]

FWIW the contract itself should be here on Sepolia and Mainnet: 0x95b259eae68ba96edB128eF853fFbDffe47D2Db0

[blmalone]

When creating a template it can either be a series of calls or delegate calls.

OPCM is basically on onchain script so we want to delegate call directly to that from the safe.

We would want a user level option to choose to delegate call or call.

UX Ideas:

• Recall that the upgrade function is function upgrade(OpChainConfig[] memory _opChainConfigs) this means that we will no longer be looping over L2 chains.

Larger question, who do we organize the UX here.

We need to include the absolute prestate for each chain if we're using OPCM: https://github.com/ethereum-optimism/optimism/blob/develop/packages/contracts-bedrock/src/L1/OPContractsManager.sol#L121 - this would be in the config.toml file.

Every OPCM upgrade looks a little different. It will always have the system config, proxy admin address but the struct might be larger.

Target address is an input. If it's specified, then that's who you delegate call to. If it's not specified then use the normal multicall3 address.

[blmalone]

There are other functions on OPCM that we will want to delegate call to in the future.

Two types of calls for superchain-ops 2.0:

• Loop over each chain and make calls
• Delegate call to our onchain OPCM (opcm loops - technically you can look by having an array of 1)

Maybe get rid of looping in OPCM. This might be more consistent with what we already have.

[ElliotFriedman]

Seems like this contract is going to make things more complex for upgrading using OPCM:

https://sepolia.etherscan.io/address/0x95b259eae68ba96edB128eF853fFbDffe47D2Db0#code

it only allows delegatecalls, so now we wrap a safe delegatecall to the multicall3delegatecall, which then delegatecalls the new OPCM contract.

safe -> delegatecalls multicall3delegatecall -> delegatecalls OPCM

seems like it would be simpler and easier to reason about if we just delegatecalled straight from the safe into the OPCM contract.

Maybe I'm missing something here though and we have to make a delegatecall to the multicall3delegatecall contract.

[ElliotFriedman]

I thought about it a bit more. If we can make a single delegatecall from the multisig to the OPCM contract with all of the parameters bundled up in a single call, then we can just delegatecall OPCM directly.

If we need to make multiple delegate calls to OPCM, then we need to use the multicall3delegatecall contract.

[maurelian]

We don't have to use multicall3delegatecall but we needed a way to make a delegatecall from the Safe while still leveraging all of the existing infrastructure in the superchain-ops repo infrastructure for signing and reading the input.json tx bundles.

However we handle that within the new ops repo paradigm is fine with me.

[blmalone]

Closed by: #606