performance analysis and optimization

⚡ Performance Analysis and Optimization

Version: 2.0.0 Last Updated: 2025-11-22 Priority: ⭐⭐⭐ (P2 Level) Purpose: Performance Analysis and Optimization for Rust Applications

---

📋 Overview

We provide identification and optimization of performance issues through CPU profiling, benchmarking, and memory analysis.

---

🎯 P0: Invocation Triggers

Trigger	Example
Slow	"this is slow"
Memory usage	"why is memory usage so high?"
Optimization	"optimize this function"
Profiling	"profile this code"
Benchmark	"benchmark performance"

---

🔧 P1: Analysis Tools List

Tool Priority

Tool	Purpose	Target	Command
criterion	Benchmark	Function	cargo bench
flamegraph	CPU profile	Process	cargo flamegraph
perf	Detailed profile	Linux	perf record
valgrind	Memory	Heap	valgrind --tool=massif
heaptrack	Heap tracking	Allocation	heaptrack ./binary
cargo-bloat	Binary size	Size	cargo bloat
tokio-console	Async	Task	tokio-console

---

🚀 P2: Analysis Patterns

Pattern 1: Benchmarking (criterion)

// benches/my_benchmark.rs
use criterion::{black_box, criterion_group, criterion_main, Criterion};

fn bench_function(c: &mut Criterion) {
    c.bench_function("my_function", |b| {
        b.iter(|| my_function(black_box(input)))
    });
}

criterion_group!(benches, bench_function);
criterion_main!(benches);

cargo bench

Pattern 2: Flamegraph

# フレームグラフ生成
cargo flamegraph --bin miyabi -- --issue 270

# 出力: flamegraph.svg

Pattern 3: Memory Profiling

# valgrind massif
valgrind --tool=massif ./target/release/miyabi
ms_print massif.out.*

# heaptrack（推奨）
heaptrack ./target/release/miyabi
heaptrack_gui heaptrack.miyabi.*

Pattern 4: Binary Size Analysis

# サイズ分析
cargo bloat --release --crates

# シンボル別
cargo bloat --release -n 20

---

⚡ P3: Optimization Strategies

Optimization Priority

Priority	Strategy	Effect	Difficulty
1	Algorithm improvement	High	Medium
2	Data structure change	High	Medium
3	Memory allocation reduction	Medium	Low
4	Parallelization	Medium	High
5	Cache utilization	Medium	Medium
6	SIMD/Low-level	Low	High

Common Optimizations

// ❌ 毎回allocate
for item in items {
    let s = item.to_string();
    // ...
}

// ✅ 事前allocate
let mut buf = String::with_capacity(1024);
for item in items {
    buf.clear();
    write!(&mut buf, "{}", item).unwrap();
    // ...
}

// ❌ Clone多用
fn process(data: Vec<T>) -> Vec<T> {
    data.clone()
}

// ✅ 参照で渡す
fn process(data: &[T]) -> Vec<T> {
    // ...
}

---

📊 Performance Goals

Metric	Goal	Measurement Method
Build time	<5 minutes	CI measurement
Test time	<2 minutes	cargo test
Binary size	<50MB	cargo bloat
Memory usage	<500MB	runtime measurement

---

🛡️ Precautions

Measure with release builds

# ❌ デバッグビルド（遅い）
cargo run

# ✅ リリースビルド
cargo run --release

PGO（Profile-Guided Optimization）

# Step 1: インストルメント
RUSTFLAGS="-Cprofile-generate=/tmp/pgo" cargo build --release

# Step 2: プロファイル収集
./target/release/miyabi [typical workload]

# Step 3: 最適化ビルド
llvm-profdata merge -o /tmp/pgo/merged.profdata /tmp/pgo
RUSTFLAGS="-Cprofile-use=/tmp/pgo/merged.profdata" cargo build --release

---

✅ Success Criteria

Check Item	Criteria
Bottleneck identification	Top 3 locations
Benchmark	Before/after comparison
Memory leak	None
Regression test	Performance maintained

---

🔗 Related Skills

• Rust Development: Build optimization
• Debugging: Problem location identification
• Security Audit: Trade-off with security