feat(blockdb): add lru cache for block entries

[DracoLi]

Why this should be merged

I noticed that during C-Chain block execution testing, each block accepted will trigger 3 block data fetches (header, body, and receipts). This happens when collecting unflattened logs in the acceptor queue (code).

Fetching block data should be fast but decompressing the data each time adds a lot of overhead. This overhead can be completely avoided with a cache because the data should have been just written to the database.

A cache was not originally added to blockdb because the VMs all have caches for blocks already. But they can be bypassed due to oversight and having a cache at the db level ensures these kind of issues are avoided.

How this works

• Added an lru cache with configurable size (defaults to 256).
• Cache is keyed by block height and is updated on Put and Get.
• Cache is used on Get and Has to avoid disk access.

How this was tested

Unit tests and reexecution benchmark

Need to be documented in RELEASES.md?

[Copilot]

Pull Request Overview

This PR adds an LRU cache for block entries to blockdb to improve performance by avoiding repeated decompression of recently accessed block data.

• Implements an LRU cache with configurable size (defaults to 256 entries) for block data
• Updates cache on both Put and Get operations with cloned data to prevent modification
• Adds cache lookups to Get and Has operations to avoid disk access for cached entries

Reviewed Changes

Copilot reviewed 6 out of 6 changed files in this pull request and generated 1 comment.

Show a summary per file

File	Description
x/blockdb/config.go	Adds EntryCacheSize configuration field with default value and validation
x/blockdb/database.go	Implements LRU cache integration in Database struct and Get/Has/Put operations
x/blockdb/entry_cache_test.go	Comprehensive test suite for cache functionality including cloning behavior
x/blockdb/readblock_test.go	Updates test configurations to include new EntryCacheSize parameter
x/blockdb/writeblock_test.go	Fixes test assertion to handle nil values properly and adds bytes import
x/blockdb/README.md	Updates documentation to reflect cache feature and removes TODO item

---

💡 Add Copilot custom instructions for smarter, more guided reviews. Learn how to get started.

[rrazvan1]

Suggested change

	// because concurrent writes to the same height are not an intended use case.
	// because concurrent access to the same height is not an intended use case.

[DracoLi]

concurrent Get/Has to the same height is fine and expected. its the writes to the same height that is not intended usage and will result in access inconsistencies from cached Has and Get

[rrazvan1]

Then it sounds wrong to me:

Concurrent access to the same height can result in cache inconsistencies [...] is acceptable because concurrent writes to the same height are not an intended use case.

[DracoLi]

reverted to the original comment. 9792f38
I think both are right depending on how you interpret it but I think we can just call out the writes.