graphene/LibOS/shim/test/apps/lmbench/lmbench-2.5/doc/bw_unix.8

.\" $Id$
.TH BW_UNIX 8 "$Date$" "(c)1994 Larry McVoy" "LMBENCH"
.SH NAME
bw_unix \- time data movement through Unix stream sockets
.SH SYNOPSIS
.B bw_unix
.SH DESCRIPTION
.B bw_unix
creates a Unix stream socket between two processes and moves 10MB 
through the stream in 64KB chunks (note that sockets are typically 
sized smaller than that).
.SH OUTPUT
Output format is \f(CB"AF_UNIX sock stream bandwidth: %0.2f MB/sec\\n", megabytes_per_second\fP, i.e.,
.sp
.ft CB
AF_UNIX sock stream bandwidth: 4.87 MB/sec
.ft
.SH MEMORY UTILIZATION
This benchmark can move up to six times the requested memory per process.
There are two processes, the sender and the receiver.
Most Unix systems implement the read/write system calls as a bcopy 
from/to kernel space to/from user space.  
Bcopy will use 2-3 times as much memory bandwidth: there is one read 
from the source and a write to the destionation.  
The write usually results in a cache line read and then a write back of
the cache line at some later point.  
Memory utilization might be reduced by 1/3 if the processor
architecture implemented "load cache line" and "store cache line"
instructions (as well as getcachelinesize). 
.SH ACKNOWLEDGEMENT
Funding for the development of
this tool was provided by Sun Microsystems Computer Corporation.
.SH "SEE ALSO"
lmbench(8).