[perf]: performance tuning

[popcnt1]

Objective:

1. current goal: analyze basic execution flow and fix known issues
2. Next stage: run the rooch process independently and simulate performance data by means of external calls. Rich parameter support will be provided

The following steps address the current goal:

Analyze bottlenecks in the basic execution flow:

1. optimize bench related code to help narrow the problem domain for pprof to locate the problem [perf] switch to enable pure memory/disk file system for bench testing #1395
2. use pprof-rs to get bottleneck (enough to get rough bottleneck information）

Known issues:

1. DA and executor share tx buffer, resulting in blocking when DA submits batch executor append tx: DA server by DA control concurrency, tx buffer use read/write cache to reduce blocking
2. bitcoin block as a parameter causes the transaction output json is too large: when converting RoochTransaction to MoveTransaction, fetcher is used to get the content of the block, and pass it to the VM. Move provides a native method to get the block of bitcoin through hash.

[popcnt1]

too old data, move to #3031