paradigmxyz/reth - CodSpeed

reth

Blog Docs Changelog

Performance History

Latest Results

fix: update Chain::new call in optimism tests

mattsse/revert-21137

43 minutes ago

fix: update Chain::new call in optimism tests

mattsse/revert-21137

49 minutes ago

fix: update Chain::new call in optimism tests

mattsse/revert-21137

54 minutes ago

revert: undo Chain crate, add LazyTrieData to trie-common This PR reverts #21137 (the reth-chain crate) and adds the useful LazyTrieData/SortedTrieData types directly to reth-trie-common. Changes: - Move Chain back to reth-execution-types - Move DeferredTrieData back to reth-chain-state - Remove the reth-chain crate - Add LazyTrieData and SortedTrieData to reth-trie-common The new types in reth-trie-common: - SortedTrieData: Container bundling HashedPostStateSorted + TrieUpdatesSorted - LazyTrieData: No-std compatible lazy wrapper with ready/deferred modes

mattsse/revert-21137

1 hour ago

feat(trie): add LazyTrieData and SortedTrieData types Add no-std compatible lazy initialization wrapper for trie data: - SortedTrieData: Container bundling HashedPostStateSorted and TrieUpdatesSorted together - LazyTrieData: Lazily initialized wrapper supporting ready mode (immediate data) and deferred mode (computed on first access) These types provide a clean abstraction for lazy trie data without requiring the complexity of the reth-chain crate introduced in #21137. Amp-Thread-ID: https://ampcode.com/threads/T-019bc9ce-d450-7307-bc10-04dd112904f0 Co-authored-by: Amp <amp@ampcode.com>

mattsse/revert-21137

3 hours ago

refactor(trie): simplify LazyTrieData with SortedTrieData container This refactor simplifies the lazy trie data abstraction introduced in the previous PR: - Create SortedTrieData container bundling HashedPostStateSorted and TrieUpdatesSorted together - Simplify LazyTrieData to use single Arc<OnceLock<SortedTrieData>> instead of two separate OnceLocks - Remove TrieDataCompute trait in favor of simple Arc<dyn Fn()> - Remove Default impl for LazyTrieData (not needed) - Update DeferredTrieData::to_lazy() to use closure-based deferred computation The previous implementation was overly complex with: - Two separate OnceLock fields that could be initialized independently - A trait (TrieDataCompute) that was only implemented by one type - Unnecessary Default implementation This simplification makes the code easier to understand while maintaining the same lazy evaluation semantics. The computation is still deferred until first access, and clones share the cached state.

mattsse/simplify-lazy-trie-data

3 hours ago

refactor(trie): simplify LazyTrieData with SortedTrieData container This refactor simplifies the lazy trie data abstraction introduced in the previous PR: - Create SortedTrieData container bundling HashedPostStateSorted and TrieUpdatesSorted together - Simplify LazyTrieData to use single Arc<OnceLock<SortedTrieData>> instead of two separate OnceLocks - Remove TrieDataCompute trait in favor of simple Arc<dyn Fn()> - Remove Default impl for LazyTrieData (not needed) - Update DeferredTrieData::to_lazy() to use closure-based deferred computation The previous implementation was overly complex with: - Two separate OnceLock fields that could be initialized independently - A trait (TrieDataCompute) that was only implemented by one type - Unnecessary Default implementation This simplification makes the code easier to understand while maintaining the same lazy evaluation semantics. The computation is still deferred until first access, and clones share the cached state.

mattsse/simplify-lazy-trie-data

3 hours ago

refactor: use ExecutionOutcome::single instead of tuple From (#21152)

main

4 hours ago

Active Branches

revert: undo Chain crate, add LazyTrieData to trie-common

last run

43 minutes ago

0%

perf(provider): batch hashed state across blocks

last run

4 hours ago

0%

perf(engine): fixed-cache for execution cache

last run

6 hours ago

0%

© 2026 CodSpeed Technology

Home Terms Privacy Docs