feat: project restructuring Part1

[rakita]

Docs pending,

Roughly looks like this:

    stateDiagram-v2
        database --> db_traits
        bytecode --> primitives
        bytecode --> gas
        db_state --> bytecode
        gas --> specs
        db_traits --> db_state
        specs --> primitives
        interpreter --> bytecode
        interpreter --> gas
        context --> wiring
        context --> db_traits
        wiring --> specs
        wiring --> primitives
        wiring --> db_traits
        handler --> precompile_interface
        handler --> context
        handler --> wiring
        handler --> interpreter
        precompile_interface --> context
        revm --> database
        revm --> precompiles
        revm --> handler
        revm --> wiring
        revm --> specs
        revm --> primitives
        precompiles --> precompile_interface
        inspector --> revm
        optimism --> revm
        revme --> revm
        revme --> statetest
        statetest --> revm

Loading

remv-test removed, all crates paths are in workspace and renamed without revm_ prefix ( revm_primitives to primitives).

Crates split that look good are.

• primitives: Wrapper around primitiv types, U256/Bytes/B256/keccak256/Address etc.
• specification: Have Hardfork type and all EIP constants and Eip2930 and EIP7702 types. If needed last two can be split in separate repo.
• bytecode: Contain bytecodes.
• state: Account,AccountInfo etc.
• database: All implementations on interface crate.
  • database_interface: Database/DatabaseRef
• Wiring: Bloated but contains default Env and all Block/Tx traits.

Revm/Interpreter/Wiring is going to be tackled in Part2.

[github-actions]

Valgrind Results:

==4407== Cachegrind, a cache and branch-prediction profiler
==4407== Copyright (C) 2002-2017, and GNU GPL'd, by Nicholas Nethercote et al.
==4407== Using Valgrind-3.18.1 and LibVEX; rerun with -h for copyright info
==4407== Command: target/release/revme bench snailtracer
==4407== 
--4407-- warning: L3 cache found, using its data for the LL simulation.
Running snailtracer example!
elapsed: 1.850564014s
==4407== 
==4407== I   refs:      429,379,382
==4407== I1  misses:          4,957
==4407== LLi misses:          4,145
==4407== I1  miss rate:        0.00%
==4407== LLi miss rate:        0.00%
==4407== 
==4407== D   refs:      205,674,960  (133,128,774 rd   + 72,546,186 wr)
==4407== D1  misses:        346,379  (    203,691 rd   +    142,688 wr)
==4407== LLd misses:        142,392  (      7,265 rd   +    135,127 wr)
==4407== D1  miss rate:         0.2% (        0.2%     +        0.2%  )
==4407== LLd miss rate:         0.1% (        0.0%     +        0.2%  )
==4407== 
==4407== LL refs:           351,336  (    208,648 rd   +    142,688 wr)
==4407== LL misses:         146,537  (     11,410 rd   +    135,127 wr)
==4407== LL miss rate:          0.0% (        0.0%     +        0.2%  )