From owner-freebsd-hackers Tue Sep 26 13:10:30 2000 Delivered-To: freebsd-hackers@freebsd.org Received: from front5.grolier.fr (front5.grolier.fr [194.158.96.55]) by hub.freebsd.org (Postfix) with ESMTP id 06CBC37B422 for ; Tue, 26 Sep 2000 13:10:24 -0700 (PDT) Received: from nas26-239.vlt.club-internet.fr (nas26-239.vlt.club-internet.fr [195.36.223.239]) by front5.grolier.fr (8.9.3/No_Relay+No_Spam_MGC990224) with ESMTP id WAA04608; Tue, 26 Sep 2000 22:10:17 +0200 (MET DST) Date: Tue, 26 Sep 2000 20:11:23 +0200 (CEST) From: =?ISO-8859-1?Q?G=E9rard_Roudier?= X-Sender: groudier@linux.local To: Joe McGuckin Cc: freebsd-hackers@FreeBSD.ORG Subject: Re: What's the difference between the ncr0 and sym0 drivers? In-Reply-To: <200009260216.TAA25565@monk.via.net> Message-ID: MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=ISO-8859-1 Content-Transfer-Encoding: QUOTED-PRINTABLE Sender: owner-freebsd-hackers@FreeBSD.ORG Precedence: bulk X-Loop: FreeBSD.ORG On Mon, 25 Sep 2000, Joe McGuckin wrote: > Is one preferable? Here's the history: BSD ncr -> Linux ncr53c8xx -> Linux sym53c8xx -> FreeBSD sym The ncr is minimally maintained mainly against O/S changes since the latest real improvement that has been the support of 875/895/896 Ultra chips: - Ultra, Ultra 2 - On-chip RAM These changes came from the Linux ncr53c8xx driver, as I am using both Linux and FreeBSD since 1995. Note that Linux ncr53c8xx is now also minimally maintained. Only `sym53c8xx' and `sym' support phase mismatch handling from SCRIPTS, Ultra3 chips (SYM53C1010), residual calculation, completion queue (rather than walking CCB lists), etc ... (would be too long :) ). Speaking about Linux + FreeBSD + FreeEtc..., my main project at the moment is `sym' for all. Btw, this is not an original strategy. :-) It seems that a new driver is developped from scratch in the NetBSD project. That's a courageous effort. Just the `from scratch' approach has been a bad idea, in my opinion. About the `sym' driver, I try to maintain it up-to-date regarding both O/S features and chips features. It seems extremally stable in practice and is both very fast and scale perfectly with the number of simulatenous IOs, at least in theory :). Just a number: I have measured a bit more that 16000 small IOs per second (512 bytes READs) using a single old SYM53C895 chip, which let me think that theory should match practice here. :) G=E9rard. PS: LSILogic supports actively sym53c8xx and sym drivers development by providing me with controllers, documentations and hardware upgrades. > Thanks, >=20 > Joe >=20 >=20 >=20 > -- >=20 > Joe McGuckin >=20 > ViaNet Communications > 994 San Antonio Road > Palo Alto, CA 94303 >=20 > Phone: 650-969-2203 > Cell: 650-207-0372 > Fax: 650-969-2124 To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-hackers" in the body of the message