Cached db handles access may panic (call get_dbi on more than one thread)

[AtomicAzzaz]

Describe the bug

#6750 Replaces a Mutex by OnceLock but as it is the OnceLock doesn't guarantee it's not set before the following unwrap is called.
I believe it may be set by another thread before as I got panics from here in my code.

Steps to reproduce

Make many db access in parallel I guess

Node logs

No response

Platform(s)

Linux (ARM)

What version/commit are you on?

7d6fb37

What database version are you on?

1

What type of node are you running?

Pruned with custom reth.toml config

What prune config do you use, if any?

No response

If you've built Reth from source, provide the full command you used

No response

Code of Conduct

• I agree to follow the Code of Conduct