From owner-freebsd-current Sat Nov 28 14:24:35 1998 Return-Path: Received: (from majordom@localhost) by hub.freebsd.org (8.8.8/8.8.8) id OAA20942 for freebsd-current-outgoing; Sat, 28 Nov 1998 14:24:35 -0800 (PST) (envelope-from owner-freebsd-current@FreeBSD.ORG) Received: from uni-sb.de (uni-sb.de [134.96.252.33]) by hub.freebsd.org (8.8.8/8.8.8) with ESMTP id OAA20921 for ; Sat, 28 Nov 1998 14:24:31 -0800 (PST) (envelope-from rock@wurzelausix.CS.Uni-SB.DE) Received: from cs.uni-sb.de (cs.uni-sb.de [134.96.252.31]) by uni-sb.de (8.9.0/1998052000) with ESMTP id XAA19088; Sat, 28 Nov 1998 23:24:08 +0100 (CET) Received: from wurzelausix.cs.uni-sb.de (Sz40inlzoevDIllAfey9tkMz/jOonfTQ@wurzelausix.cs.uni-sb.de [134.96.247.1]) by cs.uni-sb.de (8.9.0/1998060300) with ESMTP id XAA09442; Sat, 28 Nov 1998 23:24:08 +0100 (CET) Received: (from rock@localhost) by wurzelausix.cs.uni-sb.de (8.9.1/wjp/19980821) id XAA07155; Sat, 28 Nov 1998 23:23:59 +0100 (CET) Date: Sat, 28 Nov 1998 23:23:59 +0100 (CET) From: "D. Rock" Message-Id: <199811282223.XAA07155@wurzelausix.cs.uni-sb.de> To: bde@zeta.org.au, current@FreeBSD.ORG Subject: Re: ALI Aladdin V and 2nd IDE channel Sender: owner-freebsd-current@FreeBSD.ORG Precedence: bulk X-Loop: FreeBSD.ORG >>I have an Aladdin V chipset and the 2nd IDE channel is only detected if >>I insert >>options DISABLE_PCI_IDE >>in my kernel config file. This option is nowhere documented but only with >>this option enabled my secondary IDE channel and its CDROM drive is detected >>properly. I didn't notice any performance degradation on the 1st IDE channel. >>(I get 11 MB/s with only 5% CPU usage) >> >>What does this option do exactly? > >I think it just bypasses broken unit number handling. See a recent PR. >Patches for this have been floating around for more than a year, but are >said to be broken in different ways. > >11 MB/s takes at least 11/16.6*100% CPU in PIO mode. Reports of 5% >mean that the CPU meter is broken. The CPU meter in `systat -vmstat' >should work. This was the output from systat -v! The CPU meter showed up 95% idle time, so I think, it is still using DMA. The system was also too responsive for PIO modes. To be sure I will do another test with an CPU intensive job running at the same time, so the real idle time (if it is measured incorrectly) should be scheduled to this process. Daniel To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-current" in the body of the message