From owner-freebsd-scsi Fri Mar 19 14:13:51 1999 Delivered-To: freebsd-scsi@freebsd.org Received: from arjun.niksun.com (gw.niksun.com [206.20.52.122]) by hub.freebsd.org (Postfix) with ESMTP id E543A1579F for ; Fri, 19 Mar 1999 14:13:44 -0800 (PST) (envelope-from ath@niksun.com) Received: from stiegl.niksun.com (stiegl.niksun.com [10.0.0.44]) by arjun.niksun.com (8.8.8/8.8.8) with ESMTP id RAA25299; Fri, 19 Mar 1999 17:13:24 -0500 (EST) Received: from stiegl.niksun.com (localhost.niksun.com [127.0.0.1]) by stiegl.niksun.com (8.8.8/8.8.7) with ESMTP id RAA15833; Fri, 19 Mar 1999 17:13:22 -0500 (EST) (envelope-from ath@stiegl.niksun.com) Message-Id: <199903192213.RAA15833@stiegl.niksun.com> From: Andrew Heybey To: Gerard Roudier Cc: freebsd-scsi@freebsd.org Subject: Re: NCR 895 performance In-reply-to: Your message of Fri, 19 Mar 1999 22:06:00 +0100. Mime-Version: 1.0 (generated by tm-edit 7.108) Content-Type: text/plain; charset=US-ASCII Date: Fri, 19 Mar 1999 17:13:22 -0500 Sender: owner-freebsd-scsi@FreeBSD.ORG Precedence: bulk X-Loop: FreeBSD.org > Things that can explain at least part of the difference may be, in my > opinion: > > 1 - Some important features of the chip not being enabled. > 2 - The different number of tags. > 3 - The way the FreeBSD ncr driver walk the CCB list on reselection. > 4 - Also the number of interrupts. > > Could you make the same benchmark using a small number of tags, for > example 8 and report the interrupts and IO statistics of both drivers for > that benchmark. Thanks for the reply. When I get a little time, I will try the same benchmark with the same number of tags for each controller. > It would also be interesting to know which of the following features: > - Prefetch > - Burst-op > - Cache Line size > - Write and Invalidate > - Bursting (and burst size max) > - On-chip RAM > were being used in your evaluation of the 895. I used whatever sys/pci/ncr.c gives me. Do you have any suggestions as to which features should be set how? andrew To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-scsi" in the body of the message