kamalmarhubi / linux-ipc-benchmarks

Linux IPC benchmarks

This is a collection of benchmarks of Linux interprocess communication mechanisms.

At present, it benchmarks round-trip latency for pipes and eventfd.

To build the benchmarks, simply run make. The pipes and eventfd binaries both take the following flags:

  -i, -n, --iters=COUNT      number of iterations to measure; default: 100000
  -w, --warmup-iters=COUNT   number of iterations before measurement; default:
                             1000
  -?, --help                 Give this help list
      --usage                Give a short usage message

linux-ipc-benchmarks's Issues

af_unix benchmark

Your af_unix implementation is actually doing both the reading and writing in the same thread, hence the absurdly small latency. You actually need 2 sets of socketpairs (parent->child & child->parent), which should give similar performance to pipes.

If you strace the binary you will see the child process blocks on the first read while the parent process does all the work.

Add licensing information

Having a license (and no, public domain would not do) is a must for commercial users, even if it's something trivial that might just be for including results in a presentation's slides.