Add eth_simulate , support array of eth_call for simulation across multiple blocks

[OlegJakushkin]

This is a PR in progress; changes to it will happen.
In this PR we are implementing eth_simulate proposal.

Changes to be made:

• Support for RPC types
• Unit testing of the main algorithm (in progress)
• Implementation (account states, blocks, precompiled codes)
• Implementation unit testing
• Implementation precompile redirects
• Start from not head block (shall be tested shall be real-life - state reversal)
• Implementation of RPC pipeline
• Implementation of Configuration pipeline (added a basic multiplier to eth_call)
• Have block numbers independent from chain head
• Add Transfer log events on eth transfer (added to tracers shall be real life tested - tags correctness)
• test out insufficient funds errors like others have on eth_call
• Error codes from protocol
• Protocol response serialisation checks
• Gas limits
• eth_call like behavior (no validation)
• 100000+ blocks performance test

Types of changes

What types of changes does your code introduce?

• Bugfix (a non-breaking change that fixes an issue)
• New feature (a non-breaking change that adds functionality)
• Breaking change (a change that causes existing functionality not to work as expected)
• Optimization
• Refactoring
• Documentation update
• Build-related changes
• Other: Description

Testing

Requires testing

Also requires comparison with Geth implementation

• Yes
• No

If yes, did you write tests?

• Yes
• No

Remarks

Work in progress.

Docker image from commit 2684fba created at nethermindeth/nethermind:eth_multicall (08.06.2023)

Important configuration options:

• in JsonRpcConfig
  • GasCap (default: 100000000)
  • GasCapMultiplier, new (default: 2) applied as GasCap *GasCapMultiplier to limit totall max gas consumption over all blocks on a single eth_multicall

Important protocol expansions:

• Added option to turn off additional Logs of eth transfers to function signature
• Added to AccountOverride
  • MoveToAddress field for original code/precompiles redirection

Example call:

import requests
import json

# Define the JSON-RPC request payload
payload = {
    "jsonrpc": "2.0",
    "id": 1,
    "method": "eth_multicall",
    "params": [
        "0x0",
        [
            {
                "stateOverrides": [
                    {
                        "nonce": "0x1",
                        "address": "0xc02aaa39b223fe8d0a0e5c4f27ead9083c756cc2",
                        "balance": "0xde0b6b3a7640000",
                        "state": {
                            "0x2292f0db49e1af24fbcac7f32b7537f334244455ad0ed0b46a78202982e7b70d": "0xde0b6b3a7640000",
                            "0x0x877bd4632ef8c8ddc43b67e5a4511d939eadb612553c8a060670e05ebb1bb83c": "0xde0b6b3a7640000"
                        }
                    }
                ],
                "calls": [
                    {
                        "from": "0xaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaa",
                        "to": "0x6b175474e89094c44da98b954eedeac495271d0f",
                        "data": "0x18160ddd"
                    }
                ]
            }
        ],
        "latest"
    ]
}

response = requests.post("http://localhost:8547", json=payload)

[smartprogrammer93]

Wanted to merge this so i can use the new processingEnv, but the spec is not finalized yet. :(

[LukaszRozmej]

Hm... I see potential need for ReportEvent due to injecting artificial events in tracer. But would love to avoid it, maybe we can inject those events directly in MultiCallVirtualMachine to original collection? Need to think about it.

Apart of that low code hygene, lot of unused stuff, potential exceptions, nulls handled wrong, some bugs.

[LukaszRozmej]

Hm... I see potential need for ReportEvent due to injecting artificial events in tracer. But would love to avoid it, maybe we can inject those events directly in MultiCallVirtualMachine to original collection? Need to think about it.

Apart of that low code hygene, lot of unused stuff, potential exceptions, nulls handled wrong, some bugs.

[LukaszRozmej]

I think we don't have multiple block gas constraints enforced either

[flcl42]

Can I simulate invalid chain?

[ak88]

Very nice work!

[LukaszRozmej]

Can we pass the ErrorCode explicitly rather than parsing string?

[OlegJakushkin]

Sadly "InvalidBlockException" happens for multiple cases here and differs only by its text currently...

[OlegJakushkin]

Merging for ease of core development.
Future to-do items will be shipped in alternative PRs.