PERF2GV(1)

πŸ“¦
version 0.9.4 Β· 2 versions
, in fedora 43

✨ NAME ✨

perf2gv β€” bestie, this one converts perf data into a Call Graph πŸ’…


🌸 SYNOPSIS & DESCRIPTION 🌸

Okay so the vibe is: you run it like this 🫢

perf2gv < PERF_OUTPUT > CALL_GRAPH.gv

perf2gv is a simple script β€” like, genuinely simple, not doing anything extra β€” and what it does is generate a call graph from the perf data you collected on Intel machines using the Last Branch Register (LBR) hardware. No cap, that's the whole job ✨

So you'd collect it like this:

perf record -e branches:u -j any_call executable_under_test

Iconic. Here's an actual example if you wanna see it in the wild πŸ¦‹

perf record -e branches:u -j any_call example_prog
perf report --no-demangle | perf2gv > example_prog.gv
πŸ’€ Heads up bestie: perf2gv reads from standard input and writes to standard output, so keep those < and > redirects exactly where they are or the whole thing just vibes into your terminal instead.

πŸ‘‘ AUTHOR πŸ‘‘

William Cohen


2013, William Cohen


Version 0.8 🌈 β€” dated Jul 25, 2025, living in the sediment manual, filed under PERF2GV(1). That's the whole page bestie, slay, go make some call graphs πŸ’–βœ¨πŸŽ€

πŸ“¦ where this page comes from