Rust Optimise Best Practices

Performance optimization guide for Rust applications. Contains 42 rules across 8 categories, prioritized by impact from critical (memory allocation, ownership) to incremental (micro-optimizations).

When to Apply

• Optimizing Rust code for performance or reducing allocations
• Choosing data structures for optimal algorithmic complexity
• Working with iterators to avoid unnecessary intermediate allocations
• Writing async code with Tokio or other runtimes
• Profiling hot paths and eliminating performance bottlenecks
• Reviewing code for performance anti-patterns

Rule Categories by Priority

Priority	Category	Impact	Prefix
1	Memory Allocation	CRITICAL	mem-
2	Ownership & Borrowing	CRITICAL	own-
3	Data Structure Selection	HIGH	ds-
4	Iterator & Collection Patterns	HIGH	iter-
5	Async & Concurrency	MEDIUM-HIGH	async-
6	Algorithm Complexity	MEDIUM	algo-
7	Compile-Time Optimization	MEDIUM	comp-
8	Micro-optimizations	LOW	micro-

Quick Reference

1. Memory Allocation (CRITICAL)

• mem-avoid-unnecessary-clone - Avoid unnecessary clone calls
• mem-preallocate-vec-capacity - Preallocate Vec capacity
• mem-use-cow-for-conditional-ownership - Use Cow for conditional ownership
• mem-use-arc-for-shared-immutable-data - Use Arc for shared immutable data
• mem-avoid-format-for-simple-concatenation - Avoid format! for simple concatenation
• mem-use-smallvec-for-small-collections - Use SmallVec for small collections

2. Ownership & Borrowing (CRITICAL)

• own-accept-str-slice-not-string - Accept &str instead of &String
• own-accept-slice-not-vec - Accept &[T] instead of &Vec<T>
• own-use-into-for-flexible-ownership - Use Into<T> for flexible ownership
• own-return-borrowed-when-possible - Return borrowed data when possible
• own-use-asref-for-generic-borrows - Use AsRef<T> for generic borrows

3. Data Structure Selection (HIGH)

• ds-use-hashset-for-membership - Use HashSet for membership tests
• ds-use-hashmap-for-key-lookup - Use HashMap for key-value lookups
• ds-use-btreemap-for-sorted-iteration - Use BTreeMap for sorted iteration
• ds-use-vecdeque-for-queue-operations - Use VecDeque for queue operations
• ds-use-entry-api-for-conditional-insert - Use Entry API for conditional insert

4. Iterator & Collection Patterns (HIGH)

• iter-chain-instead-of-intermediate-collect - Chain iterators instead of intermediate collect
• iter-use-iter-over-into-iter-when-borrowing - Use iter() over into_iter() when borrowing
• iter-use-filter-map-for-combined-operations - Use filter_map for combined operations
• iter-use-flat-map-for-nested-iteration - Use flat_map for nested iteration
• iter-use-extend-for-bulk-append - Use extend() for bulk append
• iter-use-fold-for-accumulation - Use fold() for complex accumulation

5. Async & Concurrency (MEDIUM-HIGH)

• async-avoid-blocking-in-async-context - Avoid blocking in async context
• async-use-join-for-concurrent-futures - Use join! for concurrent futures
• async-use-rwlock-over-mutex-for-read-heavy - Use RwLock over Mutex for read-heavy
• async-minimize-lock-scope - Minimize lock scope
• async-use-buffered-for-bounded-concurrency - Use buffered() for bounded concurrency
• async-avoid-holding-lock-across-await - Avoid holding lock across await

6. Algorithm Complexity (MEDIUM)

• algo-avoid-nested-loops-for-lookup - Avoid nested loops for lookups
• algo-use-binary-search-for-sorted-data - Use binary search for sorted data
• algo-use-sort-unstable-when-order-irrelevant - Use sort_unstable when order irrelevant
• algo-use-select-nth-unstable-for-partial-sort - Use select_nth_unstable for partial sort
• algo-use-chunks-for-batch-processing - Use chunks() for batch processing

7. Compile-Time Optimization (MEDIUM)

• comp-use-const-for-compile-time-computation - Use const for compile-time computation
• comp-prefer-static-dispatch - Prefer static dispatch over dynamic
• comp-reduce-monomorphization-bloat - Reduce monomorphization bloat
• comp-use-const-generics-for-array-sizes - Use const generics for array sizes
• comp-avoid-repeated-parsing-of-static-data - Avoid repeated parsing of static data

8. Micro-optimizations (LOW)

• micro-use-inline-for-small-functions - Apply inline attribute to small hot functions
• micro-avoid-bounds-checks-in-hot-loops - Avoid bounds checks in hot loops
• micro-use-wrapping-arithmetic-when-safe - Use wrapping arithmetic when safe
• micro-use-byte-literals-for-ascii - Use byte literals for ASCII

References

1. https://nnethercote.github.io/perf-book/
2. https://rust-lang.github.io/api-guidelines/
3. https://doc.rust-lang.org/nomicon/
4. https://tokio.rs/tokio/tutorial

Related Skills

• For idiomatic patterns, architecture, and code organization, see rust-refactor skill