Skip to content
Latency benchmarks of Unix IPC mechanisms
Branch: master
Clone or download
#17 Compare This branch is 240 commits ahead, 19 commits behind rigtorp:master.
Fetching latest commit…
Cannot retrieve the latest commit at this time.
Permalink
Type Name Latest commit message Commit time
Failed to load latest commit information.
resolve
results
web/processing correct typo in html output script Aug 2, 2012
xen/mini-os
.gitignore
Makefile
README
all_lat.sh
all_thr.py Remove -t option to save disk space, and stop checking the return cod… Jan 13, 2012
atomicio.c
atomicio.h
futex.h
futex_lat.c
mempipe_lat.c
mempipe_thr.c
pipe_lat.c
pipe_thr.c
run.py
shm.c
shmem_pipe_thr.c
stats.c
summarise_tsc_counters.c Move summarise_tsc_counters.c back into the main test pipeline, so Jan 31, 2012
tcp_lat.c
tcp_local_lat.c cleanup of the various tests and more structured output for graphing Oct 31, 2011
tcp_remote_lat.c
tcp_thr.c
test.c
test.h
unix_lat.c
unix_thr.c
vmsplice_pipe_thr.c
xutil.c
xutil.h

README

ipc-bench
=========

Some very crude IPC benchmarks.

ping-pong latency benchmarks:
* pipes
* unix domain sockets
* tcp sockets

throughput benchmarks:
* pipes
* unix doamin sockets
* tcp sockets

one way latency benchmarks:
* shm

The shared memory benchmark is a kind of "control". If run under a
real-time OS it will give you the intra core communication
latency.

This software is distributed under the MIT License.
You can’t perform that action at this time.