Upgrade ethereumjs components for VM v5

[cgewecke]

This is some initial work migrating hardhat to the new ethereumjs stack. Very WIP 😄 .

Main areas of (real) test failure are forking related, also some TxPool and mining stuff. However, atm almost all provider tests fail due to issue with proposed fix at ethereumjs/ethereumjs-monorepo#1153.

(If you want to see what the actual state of this PR is, you'd need to implement that change locally in your node_modules.)

---

Summary:

• Replaces ethereumjs vm, tx, block, util, merkle-patricia-tree, and account with their latest versions.
  • ethereumjs-account no longer exists, has been merged into ethereumjs-util
• Replaces PStatemanager with ethereumjs vm's statemanager
• Implements various data type changes - notably addresses have their own type and are no longer Buffer
• Adds a FakeTransaction class (this no longer exists in ethereumjs/tx)
• Most packages in the new stack use static generators instead of new to create objects, and their internal properties are immutable. Have left comments in the code where this is presenting difficulties. One place is here.

Resources

• Upcoming Berlin release PR (enabling EIP 2718/2930 (Transaction Types / Gas AccessLists))
• Release notes for V5
• ethereumjs is tracking this PR in issue 1154

TODO

• Remove extraneous types
  • hardhat-network/provider/types/StateManager
  • hardhat-network/provider/types/PStateManager
  • config/typescripts/types/ethereumjs-block
• Resolve ethereumjs-util@7.x incompatibility w/ nomiclabs/trufflev4/v5 packages. (See lint error in CI)
• Bump vm / tx / block packages on ethereumjs Berlin release (est. March 23)
• Investigate / understand callbackified Blockchain (changes there are v. likely wrong)
• Integrate EIP selection into network config, so users can toggle EIPs on/off at will. (We won't do this right now)

Major areas of test failure

• node#mineBlock (transaction is not signed)
• TxPool (transaction is not signed)
• ForkedBlockchain (6 failing )
• ForkedProvider: ( 1 failing )
• All stack traces tests (This transaction is not signed)

NB: There are scattered single failures in other parts of the test suite (usually just one or two). FSM tests all pass.

Have left some comments in the code describing rationale for misc changes - apologies for doing this directly in the files. I wanted to keep notes as I went through rather than forget to leave notes in GH when opening the PR. Most of these will need to be removed.

Obviously there's still a lot to do (plus some things are done incorrectly) Am just hoping to get the "chore" part of this over the fence so the trickier parts can be fixed by smart people here :)

Please lmk/advise at your leisure about how to proceed and if you have questions.

[alcuadrado]

Just posting the progress so far

[cgewecke]

Leaving this for tonight... there are still 3 test failures that am unsure about the cause of or remedy for:

• From node L1601 - 20 where the gasLimit adjustment is made, throws: [EDIT fixed in c075f43]

• This, which seems pretty bad...

1) Local accounts provider
        Should, given two identical tx, return send the same raw tansaction:
 
 |       AssertionError: 
 |       + expected - actual
 | 
 |       -0xf86280830a5c008094b5bc06d4548a3ac17d72b372ae1e416bf65b8ead018082011aa052ab42ab3e2e74ca24df12dbd3539197028316216087c1a333c971377b3cc3dba068a5be657ade9e4476f4c54c3f4984e894d82548df4405a024a8f7c37bbc6de6
 |       +0xf86480830a5c0082520894b5bc06d4548a3ac17d72b372ae1e416bf65b8ead018082011aa0614471b82c6ffedd4722ca5faa7f9b309a923661a4b2adc1a53a3ebe8c4d1f0aa06aebf2fbbe82703e5075965c65c776a9caeeff4b637f203d65383e1ed2e22654

• ...and this, on the forked provider ...

1) Forked provider
       Using Alchemy
         eth_getTransactionByHash
           supports remote transactions:

      AssertionError: expected '0x2d3b7f324aca71436c5c43ccd615bdbe86661c6f' to equal '0x4e87582f5e48f3e505b7d3b544972399ad9f2e5f'
      + expected - actual

      -0x2d3b7f324aca71436c5c43ccd615bdbe86661c6f
      +0x4e87582f5e48f3e505b7d3b544972399ad9f2e5f
      
      at Context.<anonymous> (test/internal/hardhat-network/provider/forked-provider.ts:298:18)
      at process._tickCallback (internal/process/next_tick.js:68:7)

Something's not right.

[alcuadrado]

I'll continue this later today

[alcuadrado]

Oops, I forgot to submit these comments yesterday

[cgewecke]

Thanks so much @fvictorio for all those fixes!!! (Sorry this needed so much additional work)

Thanks @alcuadrado 😄