From owner-freebsd-hackers Fri Apr 7 01:23:24 1995 Return-Path: hackers-owner Received: (from majordom@localhost) by freefall.cdrom.com (8.6.10/8.6.6) id BAA26642 for hackers-outgoing; Fri, 7 Apr 1995 01:23:24 -0700 Received: from hq.icb.chel.su (icb-rich-gw.icb.chel.su [193.125.10.34]) by freefall.cdrom.com (8.6.10/8.6.6) with ESMTP id BAA26541 for ; Fri, 7 Apr 1995 01:18:08 -0700 Received: from localhost (babkin@localhost) by hq.icb.chel.su (8.6.5/8.6.5) id OAA00874 for freebsd-hackers@freebsd.org; Fri, 7 Apr 1995 14:18:13 -0500 From: "Serge A. Babkin" Message-Id: <199504071918.OAA00874@hq.icb.chel.su> Subject: Statistics and yet one fix for 3C509 To: freebsd-hackers@FreeBSD.org Date: Fri, 7 Apr 1995 14:18:12 -0500 (GMT-0500) X-Mailer: ELM [version 2.4 PL23] Content-Type: text Content-Length: 771 Sender: hackers-owner@FreeBSD.org Precedence: bulk I have done another fix for 3C509 driver. It gives no overruns on input now and finally it's possible to measure its throughput. I have tested it with FTP connection between my machine (Packard Bell 486DX2/66 w/512K cache) and ICL minicomputer that is able to transfer upto 1.1Mbytes per second by transferring a 10M file. Here are results (looking from my machine): ftp> get file /dev/null about 850 K/s systat gives about 85% of interrupts for processor's load ftp> get file about 640 Kbytes/s ftp> put file about 670 Kbytes/s So I think 3c509 isn't the best card for servers. Driver lies at wcarchive.cdrom.com/pub/FreeBSD/incoming/3c509c.* Serge Babkin ! (babkin@hq.icb.chel.su) ! Headquarter of Joint Stock Bank "Chelindbank" ! Chelyabinsk, Russia