From owner-freebsd-stable Tue Mar 26 8:14:14 2002 Delivered-To: freebsd-stable@freebsd.org Received: from nexus.npqr.net (nexus.npqr.net [63.147.19.34]) by hub.freebsd.org (Postfix) with ESMTP id 222D537B41A for ; Tue, 26 Mar 2002 08:14:09 -0800 (PST) Received: from cat (cat.npqr.net [63.147.19.40]) by nexus.npqr.net (8.11.6/8.11.6) with ESMTP id g2QGE0c36967 for ; Tue, 26 Mar 2002 08:14:00 -0800 (PST) (envelope-from pan@npqr.net) Message-ID: <03fc01c1d4e1$3d1b1640$2813933f@cat> From: "pan" To: Subject: more => 4.5r -> 4.5s panic: breaks raid array Date: Tue, 26 Mar 2002 08:13:59 -0800 Organization: happy dogs MIME-Version: 1.0 Content-Type: text/plain; charset="Windows-1252" Content-Transfer-Encoding: 7bit X-Priority: 3 X-MSMail-Priority: Normal X-Mailer: Microsoft Outlook Express 6.00.2600.0000 X-MimeOLE: Produced By Microsoft MimeOLE V6.00.2600.0000 Sender: owner-freebsd-stable@FreeBSD.ORG Precedence: bulk List-ID: List-Archive: (Web Archive) List-Help: (List Instructions) List-Subscribe: List-Unsubscribe: X-Loop: FreeBSD.ORG additional data re previous posting For this mb there are several bios flashes available from Abit and several htp372 flashes available from Via (the KR7A-RAID uses a mb mounted hpt 372 not a card so flashes from Highpoint are irrelevant) Tested combinations of bios/hpt flashes with these results Abit 6g + Via 2.0.1024 -- 4.5r(CDROM) installs; -- cvsup to 4.5s breaks ar0 Abit 6g + Via 2.0.1203 -- 4.5r(CDROM) installs -- cvsup to 4.5s breaks ar0 Abit 6n + Via 2.0.1024 -- 4.5r(CDROM) installs -- cvsup to 4.5s breaks ar0 Abit 6n + Via 2.3.1 !! 4.5r(CDROM) install fails All cases of attempted dangerously dedicated 4.5r install failed. Only default disklabel and fdisk with standard mbr sysinstall choices allowed install. Possible clue: Even though in-core and raw disklabels are identical (both before and after invoking post-install sysinstall), invoking sysinstall after install produces a write to console referring to "slice != partition", i.e. the old "slice extends beyond end of disk" error. At no time during operation of any other process besides sysinstall does this message appear - must be related to ata driver not handling hpt372 correctly? Possible clue: When cvsup to 4.5s reports what it believes to be a broken ar0 raid0 array and subsequent reboot to kernel.old, ar0 operates just fine - the actual raid array is never broken - only fbsd 4.5s thinks it is. Tested both with custom kernel and GENERIC; nothing extra-ordinary Besides the bios flash code issues this has to be an issue with the ata driver in 4.5s What other data can I try to provide? Is there any way to preserve console messages when 4.5s fails? When 4.5s thinks ar0 is broken there isn't any place for it to write boot messages to ... can fd0 have dmesg streamed to it during boot? - probably not as boot never completes with a 4.5s kernel. Abit KR7A-RAID w HPT372 Athlon XP 1900+ 1024 mgB Raid 0 with 2 x Maxtor 6L040J2 UDMA 133 hdd {ar0: ad4/ad5} Pan To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-stable" in the body of the message