From owner-freebsd-bugs Wed Mar 20 14:23:52 2002 Delivered-To: freebsd-bugs@freebsd.org Received: from mail.libertysurf.net (mail.libertysurf.net [213.36.80.91]) by hub.freebsd.org (Postfix) with ESMTP id 62FE437B417 for ; Wed, 20 Mar 2002 14:23:50 -0800 (PST) Received: from [192.168.1.129] (212.129.44.220) by mail.libertysurf.net (5.1.053) id 3C939F17000E2A17; Wed, 20 Mar 2002 23:23:50 +0100 Date: Thu, 21 Mar 2002 00:31:42 +0100 (CET) From: =?ISO-8859-1?Q?G=E9rard_Roudier?= X-X-Sender: groudier@localhost.my.domain To: George Mitchell Cc: freebsd-bugs@freeBSD.org Subject: Re: misc/35350: Can't boot on ASUS TXP In-Reply-To: <200203202149.g2KLngR0000449@m5p.com> Message-ID: <20020321001405.L2263-100000@localhost.my.domain> MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=ISO-8859-1 Content-Transfer-Encoding: QUOTED-PRINTABLE Sender: owner-freebsd-bugs@FreeBSD.ORG Precedence: bulk List-ID: List-Archive: (Web Archive) List-Help: (List Instructions) List-Subscribe: List-Unsubscribe: X-Loop: FreeBSD.org On Wed, 20 Mar 2002, George Mitchell wrote: > More verbosely, here is what the kernel reported while trying to bring > up sysinstall: > > sym0: <810> port 0xd400-0xd4ff mem 0xe600000000-0xe60000ff irq 12 at devi= ce 11.0 on pci0 > sym0: no NVRAM, ID 7, fast-10, SE, parity checking > sym0: CACHE TEST FAILED: DMA error (dstat=3D0xa0).sym0: CACHE INCORRECTLY= CONFIGURED. > device_probe_and_attach: sym0 attach returned 6. > > Here's what it says during a normal boot under FreeBSD 3.4: > > ncr0: rev 0x01 int a irq 12 on pci0.11.0 > > Thanks for your attention. -- George The problem is due to a PCI BUS fault detected in a mini-script executed by the PCI chip to check accessibilty to memory from the PCI BUS. In fact sym kept with such checking performed initially by ncr, but just traps the condition. Such problem has been reported multiple times and never had been due to a driver problem, neither in sym nor in ncr. In such situation, ncr will not detect the problem, but may just repeatitively fail with such an error each time it restarts the SCSI SCRIPTS, as you reported in your previous mail. The cause could be that BUS addresses used from SCRIPTS don't point to valid memory, for example. Another cause, given that your chip does not support LOAD/STORE, could be that the MMIO addresses used from SCRIPTS to access IO registers are wrong. For the latter possible cause, it can be interesting to check if the MMIO address base used under 3.4 matches the one used on the O/S version that fails (0xe60000000 - your report has an additionnal zero, but I guess you wrote it by hand). Regards, G=E9rard. To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-bugs" in the body of the message