From owner-freebsd-bugs Tue Aug 15 0:13:31 2000 Delivered-To: freebsd-bugs@freebsd.org Received: from vbook.express.ru (vbook.express.ru [212.24.37.106]) by hub.freebsd.org (Postfix) with ESMTP id DCBC737B887 for ; Tue, 15 Aug 2000 00:13:20 -0700 (PDT) (envelope-from vova@vbook.express.ru) Received: (from vova@localhost) by vbook.express.ru (8.9.3/8.9.3) id JAA01394; Tue, 15 Aug 2000 09:35:10 +0400 (MSD) (envelope-from vova) From: "Vladimir B. Grebenschikov" MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Message-ID: <14744.54906.18116.758951@vbook.express.ru> Date: Tue, 15 Aug 2000 09:34:50 +0400 (MSD) To: Bruce Evans Cc: freebsd-bugs@FreeBSD.ORG Subject: Re: misc/20555: 3C509 driver performance problem In-Reply-To: <200008131820.LAA24931@freefall.freebsd.org> References: <200008131820.LAA24931@freefall.freebsd.org> X-Mailer: VM 6.72 under 21.1 (patch 9) "Canyonlands" XEmacs Lucid Sender: owner-freebsd-bugs@FreeBSD.ORG Precedence: bulk X-Loop: FreeBSD.org Bruce Evans writes: > The following reply was made to PR misc/20555; it has been noted by GNATS. > > From: Bruce Evans > To: xyf@stocke.com > Cc: freebsd-gnats-submit@FreeBSD.ORG > Subject: Re: misc/20555: 3C509 driver performance problem > Date: Mon, 14 Aug 2000 04:19:48 +1000 (EST) > > On Fri, 11 Aug 2000 xyf@stocke.com wrote: > > > ab -c 10 -n 10000 http://192.168.1.27/index.html > > > > they both give me good performance, the result is almost same, but ... > > FreeBSD uses more CPU time than Linux, FreeBSD CPU idle is 40%, > > Linux CPU idle is 60%, almost 20% CPU lost in FreeBSD! > > > > I use "top" command and see FreeBSD cost 34% of CPU time in interrupt > > handling, the system time is good, only 12%, my net card is old 3COM 3C509, my question is: > > > > is this the problem of 3C509 driver in FreeBSD or just is more serious FreeBSD design problem? > > I dodn't see any problem here, except missing accounting for interrupt > time in Linux. I think Linux still counts interrupt time against > whatever process (including the idle "process") happens to be running > when the interrupt occurs. FreeBSD's idle time of 40% is actually > the idle time (modulo jitter in the measurement). Linux's idle time > of 60% may be mostly spent handling interrupts. Similarly for system > and user times. System and user times under Linux are normally larger > than under FreeBSD, because parts of them are actually interrupt times. Some times ago (with fbsd 2.2.x-3.2) I have expirience problem with old 3COM driver (ep), I don't know exactly about preformance, but often ep driver can't send full ethernet load (it was router with 2 NE2000 and 2 3C509), so ep interfaces sometimes stop sending packets and complains tomething about overloaded queue, I don't remember exactly. This time in LINT was remark about 'ep driver is bogus'. I don't try to use 3C509 after so I don't know what happens in current driver. > Bruce -- TSB Russian Express, Moscow Vladimir B. Grebenschikov, vova@express.ru To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-bugs" in the body of the message