From owner-freebsd-hackers Fri Nov 3 15: 9: 7 2000 Delivered-To: freebsd-hackers@freebsd.org Received: from rover.village.org (unknown [204.144.255.49]) by hub.freebsd.org (Postfix) with ESMTP id B760A37B4CF for ; Fri, 3 Nov 2000 15:09:03 -0800 (PST) Received: from harmony.village.org (harmony.village.org [10.0.0.6]) by rover.village.org (8.11.0/8.11.0) with ESMTP id eA3N90n47603; Fri, 3 Nov 2000 16:09:00 -0700 (MST) (envelope-from imp@harmony.village.org) Received: from harmony.village.org (localhost.village.org [127.0.0.1]) by harmony.village.org (8.9.3/8.8.3) with ESMTP id QAA21297; Fri, 3 Nov 2000 16:08:59 -0700 (MST) Message-Id: <200011032308.QAA21297@harmony.village.org> To: Hans Ottevanger Subject: Re: Broken PCI-IDE RZ1000 & ata Cc: John Summerfield , freebsd-hackers@FreeBSD.ORG In-reply-to: Your message of "Fri, 03 Nov 2000 22:29:06 +0100." <3A032E22.1E8EFE84@iae.nl> References: <3A032E22.1E8EFE84@iae.nl> <200011030901.eA391mt30196@possum.os2.ami.com.au> Date: Fri, 03 Nov 2000 16:08:59 -0700 From: Warner Losh Sender: owner-freebsd-hackers@FreeBSD.ORG Precedence: bulk X-Loop: FreeBSD.ORG In message <3A032E22.1E8EFE84@iae.nl> Hans Ottevanger writes: : http://www.intel.com/procs/support/rz1000/ According to this, disabling readahead fixes the problem. It also says you need to mask interrupts so that you don't access the register file for the device. Looking at this I'm not sure if you need to implement both workarounds, or just one of them. One workaround is to disable readahead. The other is to make sure that interrupts are masked during disk I/O. That one is likely to be hard if not impossible to implement in FreeBSD. There's also some interaction with the floppy drive as well. However, Soren says that there are other causes for data corruption. I'm inclined to believe him. These chips really suck. Warner To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-hackers" in the body of the message