profilelisted

# CPU and Memory Profiling Profile Go code to identify CPU hotspots and memory allocators using pprof. ## Usage - `/profile cpu ./internal/index/` - CPU profiling on index package - `/profile memory ./internal/repo/` - Memory profiling on repo package - `/profile all ./...` - Both CPU and memory on all packages ## Steps 1. **Parse arguments** - First argument: Profile type (`cpu`, `memory`, or `all`) - Second argument: Package path (defaults to `./...`) 2. **Create profile output directory** ```bash mkdir -p .profiles ``` 4. **Run profiling benchmarks** For CPU profiling: ```bash go test -cpuprofile=.profiles/cpu.prof -bench=. $PACKAGE 2>&1 ``` For memory profiling: ```bash go test -memprofile=.profiles/mem.prof -bench=. $PACKAGE 2>&1 ``` 5. **Analyze CPU profile** ```bash go tool pprof -top -cum .profiles/cpu.prof 2>&1 | head -30 ``` Identify: - Top 10 CPU consumers by cumulative time - Functions with high self time (computation hotspots) - Unexpected entries (potential optimization targets) 6. **Analyze memory profile** ```bash go tool pprof -top -alloc_space .profiles/mem.prof 2>&1 | head -30 ``` Identify: - Top allocators by total bytes - Functions with high allocation counts - Potential sources of GC pressure 7. **Generate flamegraph data** (if requested) ```bash go tool pprof -raw .profiles/cpu.prof > .profiles/cpu.raw ``` 8. **Report findings** Structure the rep