// the find
davecheney/high-performance-go-workshop
A set of workshop materials (asciidoc slides plus runnable example code) by Dave Cheney covering Go performance: profiling, benchmarking, compiler optimizations, the execution tracer, and memory/GC behavior. It's for Go developers who want to learn to diagnose and fix performance problems, not a library to import.
Every technique is backed by actual runnable code in examples/ rather than just slide text - you can clone it and run pprof, go test -bench, and the execution tracer against real programs like the mandelbrot renderer. It covers the real toolchain developers use day to day (pprof, benchmarking, -gcflags escape analysis output, go tool trace) instead of abstract theory. The author has deep first-hand knowledge of the Go runtime, so the explanations of GC and scheduler behavior are accurate rather than guessed at from blog posts.
Last commit is from 2020, predating generics and several Go releases' worth of scheduler and GC changes, so some tracer and memory-profiling specifics may no longer match current Go behavior. There's no real README - just a licensing note and a bit.ly link to find the actual slide deck, so there's no table of contents for someone trying to navigate the material cold. The repo has generated build artifacts committed directly (pprof files, trace.out files, compiled binaries like esc/sum, and dozens of SVGs) which bloats the clone and clutters the directory tree. No CI and no sign of ongoing maintenance, so don't expect fixes if an example breaks on a newer Go toolchain.