From owner-freebsd-java Sun Mar 16 23:28:39 2003 Delivered-To: freebsd-java@freebsd.org Received: from mx1.FreeBSD.org (mx1.freebsd.org [216.136.204.125]) by hub.freebsd.org (Postfix) with ESMTP id 9AB3937B405 for ; Sun, 16 Mar 2003 23:28:36 -0800 (PST) Received: from mta06-svc.ntlworld.com (mta06-svc.ntlworld.com [62.253.162.46]) by mx1.FreeBSD.org (Postfix) with ESMTP id 548D043F75 for ; Sun, 16 Mar 2003 23:28:35 -0800 (PST) (envelope-from antony.t.curtis@ntlworld.com) Received: from pcgem.rdg.cyberkinetica.com ([81.98.110.96]) by mta06-svc.ntlworld.com (InterMail vM.4.01.03.37 201-229-121-137-20020806) with ESMTP id <20030317072833.GFMC20605.mta06-svc.ntlworld.com@pcgem.rdg.cyberkinetica.com> for ; Mon, 17 Mar 2003 07:28:33 +0000 Content-Type: text/plain; charset="us-ascii" From: Antony T Curtis To: freebsd-java@FreeBSD.ORG Subject: JDK 1.3.1 performance Date: Mon, 17 Mar 2003 07:28:31 +0000 User-Agent: KMail/1.4.2 MIME-Version: 1.0 Content-Transfer-Encoding: quoted-printable Message-Id: <200303170728.33168.antony.t.curtis@ntlworld.com> Sender: owner-freebsd-java@FreeBSD.ORG Precedence: bulk List-ID: List-Archive: (Web Archive) List-Help: (List Instructions) List-Subscribe: List-Unsubscribe: X-Loop: FreeBSD.org -----BEGIN PGP SIGNED MESSAGE----- Hash: SHA1 Java benchmark using Volano client/server Results are scaled, bigger is better. classic=09=091.000 (1094 messages per second) +tya=09=091.107 +shujit=091.114 +OpenJIT=091.154 native=09=090.675 +tya=09=090.868 +shujit=090.833 +OpenJIT=090.847 HotSpot -Xint=090.983 client=091.427 server=090.746 jdk1.4 -Xint=090.896 client=091.301 server=091.234 These benchmarks should be important to people running networked java=20 programs. Please note, I only ran the tests once, and both client and ser= ver=20 are sitting on the same box. The JDK1.3.1 is with my own patches (includes some maybe bad fixes for=20 - -native). I expected -native to have a much lower performance than -classic as Free= BSD#s=20 userland threads have much heavier context switching than Java's green=20 threads (which by nature, know about what it's running). What is surprisi= ng=20 is that -server performs so badly whereas on 1.4, it performs reasonably. These results tell me that we need KSE's for native threading to be truel= y=20 effective and that we should not implement Java's threading on top of a=20 KSEified pthread but directly on FreeBSD-5's KSE api. I'm tempted to roll a different threading for native on FreeBSD 4 based u= pon=20 rfork()... Just my $0.02 worth... - --=20 Antony T Curtis BSc Unix Analyst Programmer http://homepage.ntlworld.com/antony.t.curtis/ -----BEGIN PGP SIGNATURE----- Version: GnuPG v1.2.1 (FreeBSD) iD8DBQE+dXkhql7dp2cddmIRAks7AJsFL37Oq9EhCi2dEZjueynkGSfhoQCff/7G dhgy8klgwXt873e3DEGJB00=3D =3DF5XB -----END PGP SIGNATURE----- To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-java" in the body of the message