60x speed-up of Linux “perf”

• linux

  980 170,074 10.0 C

  Linux kernel source tree

“Linux releases as GPLv2 or later for example”

Linux doesn’t use “or later”. https://github.com/torvalds/linux/blob/master/COPYING:

The Linux Kernel is provided under:

SPDX-License-Identifier: GPL-2.0 WITH Linux-syscall-note

Being under the terms of the GNU General Public License version 2 only,

• gimli

  2 793 7.6 Rust

  A library for reading and writing the DWARF debugging format (by gimli-rs)

  

I haven't blogged about this yet, but we saw a 1000x fold speed-up doing several things around symbolication. The more optimal approach we found was to use the gimli crate[1] directly & carefully optimize it to read in the data structures for the executable(s) you are symbolicating upfront & then issuing in-process queries. They also have a drop-in replacement of addr2line that outperforms it (both in symbolication speed & memory usage).

[1] https://github.com/gimli-rs/gimli

• WorkOS

  workos.com sponsored

  The modern identity platform for B2B SaaS. The APIs are flexible and easy-to-use, supporting authentication, user identity, and complex enterprise features like SSO and SCIM provisioning.

  

• addr2line

  2 261 5.4 Rust

  A cross-platform `addr2line` clone written in Rust, using `gimli`

  

• honggfuzz

  5 2,974 7.1 C

  Security oriented software fuzzer. Supports evolutionary, feedback-driven fuzzing based on code coverage (SW and HW based)

  

Here's the libbfd-based code in case anyone would like to port it, it's basically addr2line.

https://github.com/google/honggfuzz/blob/master/linux/bfd.c

• AFLplusplus

  16 4,637 9.7 C

  The fuzzer afl++ is afl with community patches, qemu 5.1 upgrade, collision-free coverage, enhanced laf-intel & redqueen, AFLfast++ power schedules, MOpt mutators, unicorn_mode, and a lot more!

  

With AFL++ you can even determine exactly where the fork happens:

https://github.com/AFLplusplus/AFLplusplus/blob/stable/instr...

• stacktrace

  1 396 7.2 C++

  C++ library for storing and printing backtraces. (by boostorg)

  

I do know Boost.Stacktrace calls addr2line too. From the code(https://github.com/boostorg/stacktrace/blob/develop/include/...), it seems Boost.Stacktrace also shells out to addr2line for every address. But in practice, I found the overhead of boost::stacktrace::stacktrace() is not as horrendous as my crappy implementation, which calls addr2line, too.

Suggest a related project