From owner-freebsd-stable@FreeBSD.ORG Mon Oct 16 03:16:18 2006 Return-Path: X-Original-To: freebsd-stable@freebsd.org Delivered-To: freebsd-stable@freebsd.org Received: from mx1.FreeBSD.org (mx1.freebsd.org [216.136.204.125]) by hub.freebsd.org (Postfix) with ESMTP id C04BB16A403; Mon, 16 Oct 2006 03:16:18 +0000 (UTC) (envelope-from kris@obsecurity.org) Received: from elvis.mu.org (elvis.mu.org [192.203.228.196]) by mx1.FreeBSD.org (Postfix) with ESMTP id 8043943D53; Mon, 16 Oct 2006 03:16:18 +0000 (GMT) (envelope-from kris@obsecurity.org) Received: from obsecurity.dyndns.org (elvis.mu.org [192.203.228.196]) by elvis.mu.org (Postfix) with ESMTP id 62E7B1A3C20; Sun, 15 Oct 2006 20:16:18 -0700 (PDT) Received: by obsecurity.dyndns.org (Postfix, from userid 1000) id B29C9514B1; Sun, 15 Oct 2006 23:16:17 -0400 (EDT) Date: Sun, 15 Oct 2006 23:16:17 -0400 From: Kris Kennaway To: Adrian Chadd Message-ID: <20061016031617.GB1303@xor.obsecurity.org> References: Mime-Version: 1.0 Content-Type: multipart/signed; micalg=pgp-sha1; protocol="application/pgp-signature"; boundary="GID0FwUMdk1T2AWN" Content-Disposition: inline In-Reply-To: User-Agent: Mutt/1.4.2.2i Cc: FreeBSD Stable Subject: Re: kernel profiling? X-BeenThere: freebsd-stable@freebsd.org X-Mailman-Version: 2.1.5 Precedence: list List-Id: Production branch of FreeBSD source code List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , X-List-Received-Date: Mon, 16 Oct 2006 03:16:18 -0000 --GID0FwUMdk1T2AWN Content-Type: text/plain; charset=us-ascii Content-Disposition: inline Content-Transfer-Encoding: quoted-printable On Mon, Oct 16, 2006 at 10:53:36AM +0800, Adrian Chadd wrote: > Hiya, >=20 > Whats the "right" way to grab kernel profiling data these days? >=20 > I've tried using the kernel profiling w/ kgmon and gprof but the top CPU > wasters are the profiling functions themselves, quickly followed by > write_eflags(). I'm not sure this is valid at all. >=20 > I'm running 6-stable on an Athlon 1800XP, so its uniprocessor and > (relatively) slow. I'm hitting the server rather hard with a few thousand > TCP connections a second; I'm trying to figure out where my ~60% of kernel > time and ~35% of interrupt time is going. pmc can be extremely useful although it doesn't do call graphs. Kris --GID0FwUMdk1T2AWN Content-Type: application/pgp-signature Content-Disposition: inline -----BEGIN PGP SIGNATURE----- Version: GnuPG v1.4.5 (FreeBSD) iD8DBQFFMvmBWry0BWjoQKURAuCjAKDHos6pk6N+XKmS4EyvDC1OdAz08QCgv8EK 3KKoKLIvXNPlYSw93RcQcW4= =jxxX -----END PGP SIGNATURE----- --GID0FwUMdk1T2AWN--