From owner-freebsd-amd64@FreeBSD.ORG Thu Aug 6 18:22:24 2009 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 69E341065676 for ; Thu, 6 Aug 2009 18:22:24 +0000 (UTC) (envelope-from stickybit@gmx.net) Received: from mail.gmx.net (mail.gmx.net [213.165.64.20]) by mx1.freebsd.org (Postfix) with SMTP id B4F408FC1C for ; Thu, 6 Aug 2009 18:22:23 +0000 (UTC) Received: (qmail invoked by alias); 06 Aug 2009 18:22:21 -0000 Received: from port-92-195-32-176.dynamic.qsc.de (EHLO localhost) [92.195.32.176] by mail.gmx.net (mp011) with SMTP; 06 Aug 2009 20:22:21 +0200 X-Authenticated: #23197544 X-Provags-ID: V01U2FsdGVkX1+CT4AOoKCCi15ezcXYJde2Rp0+Vy2KGmheE1Nzj1 nYcMm9DERqTSkT From: Sticky Bit To: Florian Smeets Date: Thu, 6 Aug 2009 20:21:56 +0200 User-Agent: KMail (FreeBSD) References: <200908052217.06935.stickybit@gmx.net> <4A7A863A.7080608@kasimir.com> In-Reply-To: <4A7A863A.7080608@kasimir.com> MIME-Version: 1.0 Content-Disposition: inline Message-Id: <200908062021.56672.stickybit@gmx.net> Content-Type: Text/Plain; charset="iso-8859-1" Content-Transfer-Encoding: 7bit X-Y-GMX-Trusted: 0 X-FuHaFi: 0.65 Cc: freebsd-current@freebsd.org, freebsd-amd64@freebsd.org Subject: Re: [regression] 8.0-CURRENT amd64: SATA disks not attaching, MCP55 controller X-BeenThere: freebsd-amd64@freebsd.org X-Mailman-Version: 2.1.5 Precedence: list Reply-To: stickybit@gmx.net List-Id: Porting FreeBSD to the AMD64 platform List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , X-List-Received-Date: Thu, 06 Aug 2009 18:22:24 -0000 On Thu August 6 2009 09:28:58 Florian Smeets wrote: > i had the same problem with MCP55 on amd64, i tried the new ahci module > but to no avail. > > However if you set hw.pci.mcfg=0 the disks are found by the normal ata Thanks for the hint! I tried it again with 'set hw.pci.mcfg=0' from boot loader prompt and now it actually finds and attaches the disks as usual. But I do not know if I need 'support for PCI-e memory mapped config access', where it is needed and if it can cause problems later with an RELENG_8 installation if it is disabled like in RELENG_7. I only have one PCIe device, a PCIe graphics card, which is working well with drm.ko / radeon.ko modules and radeonhd driver under RELENG_7. Is it safe to stay with 'hw.pci.mcfg=0' under RELENG_8 for the future? And if so, should the default setting be changed from enabled to disabled for 8.0-RELEASE if it is not yet mature enough and the cause of problems? Looks like several people are affected by this regression. -- Regards