statedb.go: fix NoTries mode snapshot storage delete

[zlacfzy]

Description

In NoTries mode (--tries-verify-mode none), storage data is not properly deleted from the snapshot when handleDestruction is called during SELFDESTRUCT.

Rationale

In NoTries mode, stateObject.updateRoot() always sets Root to EmptyRootHash regardless of actual storage content:

if s.db.db.NoTries() {
    s.data.Root = types.EmptyRootHash
}

This causes two issues in the SELFDESTRUCT flow:

1. handleDestruction skips storage deletion: The check prev.Root == types.EmptyRootHash incorrectly assumes no storage exists, bypassing deleteStorage().
2. fastDeleteStorage hash verification fails: Even if deleteStorage is called, the hash check stack.Hash() != root fails because root is EmptyRootHash but the actual storage hash is not.

Example

add an example CLI or API response...

Changes

Notable changes:

• add each change in a bullet point here
• ...

[NathanBSC]

this bug exists more than 2 years, no later than PR https://github.com/bnb-chain/bsc/pull/1840/files

[NathanBSC]

after cancun, with EIP-6780, no wiping out old storages then.
so people download fastnode snapshots and run, this issue will not happen

[buddh0]

fix issue #3487