core: add an end-to-end verkle test

[gballet]

In order to reduce the merging complexity of verkle, this PR adds a test meant to serve as a sentinel for changes that break verkle execution. It is a subset of a test that is used extensively in the verkle branch.

This requires #29084 as the current code hashes the node, which isn't really practical to do in a verkle way (on top of using a different hashing algorithm).

Co-authored-by: Ignacio Hagopian jsign.uy@gmail.com

[holiman]

This change only adds defensive nil-checks. Why is this change needed?

[gballet]

because it panics in tests if the genesis isn't configured

[holiman]

Shouldn't it?

[holiman]

I mean

nil.IsVerkle() == false, according to this PR. I think nil deref is right and proper here

[holiman]

It looks a bit odd to me, to have ( ..., config ConfigType, isFoo bool). I mean, IsVerkle, isn't that a config option? If so, why not just put it inside the config struct?

[rjl493456442]

Actually we have the multi-level config structure:

For triedb, we have Preimage and IsVerkle to configure if preimage is enabled and triedb is opened for verkle tree or merkle tree.

And HashDB and PathDB config is to select which concrete db implementation is used and the configuration for the internal db backend. (In theory, HashDB can also be used for storing verkle tree, although we choose to not as it's not storage efficient)

// Config defines all necessary options for database.
type Config struct {
	Preimages bool           // Flag whether the preimage of node key is recorded
	IsVerkle  bool           // Flag whether the db is holding a verkle tree
	HashDB    *hashdb.Config // Configs for hash-based scheme
	PathDB    *pathdb.Config // Configs for experimental path-based scheme
}

So I don't think IsVerkle should be put into either HashDB or PathDB.

[holiman]

LGTM