From owner-freebsd-amd64@FreeBSD.ORG Tue Feb 1 19:44:19 2011 Return-Path: Delivered-To: freebsd-amd64@freebsd.org Received: from mx1.freebsd.org (mx1.freebsd.org [IPv6:2001:4f8:fff6::34]) by hub.freebsd.org (Postfix) with ESMTP id CEDBC1065693 for ; Tue, 1 Feb 2011 19:44:19 +0000 (UTC) (envelope-from ralf.folkerts@gmx.de) Received: from mailout-de.gmx.net (mailout-de.gmx.net [213.165.64.22]) by mx1.freebsd.org (Postfix) with SMTP id 248708FC1C for ; Tue, 1 Feb 2011 19:44:18 +0000 (UTC) Received: (qmail invoked by alias); 01 Feb 2011 19:17:37 -0000 Received: from dslb-088-067-041-178.pools.arcor-ip.net (EHLO beaster.home.folkerts-net.de) [88.67.41.178] by mail.gmx.net (mp038) with SMTP; 01 Feb 2011 20:17:37 +0100 X-Authenticated: #18511927 X-Provags-ID: V01U2FsdGVkX1+FUm05JOmFAol6ADSrw6qcHogMDPccKnopBY5lXX mbla+LVIMOdAVb Message-ID: <4D485C50.3040600@gmx.de> Date: Tue, 01 Feb 2011 20:17:36 +0100 From: Ralf Folkerts User-Agent: Mozilla/5.0 (X11; U; FreeBSD amd64; de-DE; rv:1.9.2.13) Gecko/20101212 Lightning/1.0b2 Thunderbird/3.1.7 MIME-Version: 1.0 To: freebsd-amd64@freebsd.org Content-Type: text/plain; charset=ISO-8859-15; format=flowed Content-Transfer-Encoding: quoted-printable X-Y-GMX-Trusted: 0 Subject: Check 64 bit DMA from Disk-Controller (amd64/135040) X-BeenThere: freebsd-amd64@freebsd.org X-Mailman-Version: 2.1.5 Precedence: list List-Id: Porting FreeBSD to the AMD64 platform List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , X-List-Received-Date: Tue, 01 Feb 2011 19:44:19 -0000 Hi, I asked this a while ago but didn't get a reply - as, however, no one=20 replied that such a check is impossible I'd like to ask again: My SB600 ATA-Controller used to cause massive Problems (from disk not=20 being detected at all to disk being detected but later "trashed" during=20 buildworld) with DMA with 4G RAM. I circumvented that by setting=20 "hw.physmem=3D4g", as recommended by John Baldwin. Now I'd like to perform a test, if this DMA-issue was *really* fixed by=20 a recent BIOS-Update. How can such a test being performed? Or is it impossible at all? I already ran several buildworld/buildkernel instalworld/installkernel=20 to a "fresh" disk in a row after the BIOS-Update - and didn't see any=20 Problems - unlike previously where the disk was at latest trashed during = the first run. However, I have no idea whether these runs were successful because of=20 the DMA issue being fixed or simply because by some coincidence no DMA=20 was being performed beyond 32 bit addressability. Would really be cool to get some confidence - as I'd like to run ZFS but = don't think it'll be much fun with the hw.physmem being set to 4g :-( Many thanks in advance! Btw: System is now "8.1-STABLE FreeBSD 8.1-STABLE #0: Fri Nov 12=20 19:45:42 CET 2010 amd64" I ran my Tests against the 8.1 RELEASE (booted off USB Stick, then=20 installed to a "fresh" disk. Cheers, _ralf_