<html>
<head>
<style><!--
.hmmessage P
{
margin:0px;
padding:0px
}
body.hmmessage
{
font-size: 10pt;
font-family:Verdana
}
--></style>
</head>
<body class='hmmessage'>
uncontended pthread_mutex_lock/unlock:<BR>
syscall on FreeBSD? Solaris? NetBSD? Darwin?<BR>
Not on Linux.<BR>
I really doubt on Solaris.<BR>
<BR>
Ditto pthread_self, pthread_getspecific, pthread_setspecific?<BR>
(Though I don't believe we call them frequently on Solaris, and I have a very rough plan to reduce them *later* -- libunwind.)<BR>
<BR>
- Jay<BR><BR>> From: dragisha@m3w.org<BR>> To: Highjinks@gmx.com<BR>> Date: Thu, 8 Apr 2010 00:06:29 +0200<BR>> CC: m3devel@elegosoft.com<BR>> Subject: Re: [M3devel] lock performance, random thoughts<BR>> <BR>> It's all about synchronization in kernel space, ie part of it when it<BR>> comes to kernel space through syscalls.<BR>> <BR>> Context switching is lesser problem as it is what happens a lot less<BR>> often than synchronization.<BR>> <BR>> Synchronization problem is with systems which are, unlike Linux, doing<BR>> syscalls on every pthread_mutex_lock and friends.<BR>> <BR>> Hybridization of thread system (mixing user and kernel threads) is done<BR>> talk and I hope nobody takes such thing seriously anymore.<BR>> <BR>> On Wed, 2010-04-07 at 22:09 +0200, Chris wrote:<BR>> > As I've said before, this really has NOTHING to do with user/kernel<BR>> > level threading. It's all about synchronization.<BR>> -- <BR>> Dragiša Durić <dragisha@m3w.org><BR>> <BR> </body>
</html>