Date: Wed, 1 Sep 1999 18:46:57 +0300 From: Ruslan Ermilov <ru@ucb.crimea.ua> To: David Kelly <dkelly@HiWAAY.net> Cc: freebsd-stable@FreeBSD.org Subject: Re: wiring down SCSI devices, and the root device Message-ID: <19990901184657.A77510@relay.ucb.crimea.ua> In-Reply-To: <199909010317.WAA00426@nospam.hiwaay.net>; from David Kelly on Tue, Aug 31, 1999 at 10:17:33PM -0500 References: <199909010317.WAA00426@nospam.hiwaay.net>
next in thread | previous in thread | raw e-mail | index | archive | help
On Tue, Aug 31, 1999 at 10:17:33PM -0500, David Kelly wrote: > Messing around with SCSI drives and two SCSI controllers. One is a > first generation Adaptec 2940A, fast, narrow. The other is a much newer > Sybios 875-based card. Apparently I have them in the wrong PCI slots to > boot the way I wish if a HD is on the Adaptec. "Normally" only tape > drives are on the Adaptec. And only one HD is on ncr0. > > So rather than solve my problem in hardware, thought I'd do it in > software. Altered my config file to wire down my SCSI devices like this > (thought I'd throw in the "kernel root on" for completness): > > config kernel root on da0 > > # old way: > #controller scbus0 > #device da0 > # > # wired down: > controller scbus0 at ncr0 > controller scbus1 at ahc0 > device da0 at scbus0 target 0 > > The comments in LINT suggest "device" and "disk" are the same thing so > I commented out the old "device da0" and used the LINT disk line as my > example. > > Is the "controller scbus1" line needed? Based on the LINT example I > didn't think it could hurt. > > Built and installed new kernel. Shut down. Added a 2G HD to the ahc0. > During boot BIOS says this new HD is the C: drive. Fromerly the 9G > drive on ncr0 was so identified. > > My boot block code says something about version 0.7. > > F5 to boot second disk. Then the normal F2 to boot BSD. Eventually > followed by a fatal error. This part is typed from written notes: > > da1 at ahc0 bus 0 target 1 lun 0 > ... (normal da1 stuff for the new HD) > changing root device to da1s1a > error 22: panic: cannot mount root(2) > (2 blank lines) > syncing disks... done > da0 at ncr0 bus 0 target 0 lun 0 > ... (normal da0 stuff for the old HD) > Automatic reboot in 15 seconds - press a key on the console to abort > > For some reason my kernel reassigned root contrary to my instructions > in config. And after deciding to sync and reboot apparently a process > or thread or something completed the probe of da0, which was supposed > to be root. And finally the shutdown completed. > > System is 3.2-STABLE, cvsup'ed the evening of 8/30/99. Matching kernel. > But using my config. I don't see anything in GENERIC addressing > possible solutions. > Normally, the system uses the device from which it was booted as the root device if possible. (See reboot(2) and boot(8) for details). There are a lot of possible solutions: 1) Hit any key while the boot blocks are loading (thus skipping an invokation of the third-stage loader(8)), and type `-r' to boot blocks prompt. Refer to the boot(8) manpage. 2) Inside loader(8), interrupt the autoboot sequence by pressing <Space> and type `boot -r' on the command prompt. Refer to the loader(8) manpage. 3) Inside loader(8), set `rootdev' variable to the appropriate value and execute `boot'. Or better yet put this into /boot/loader.conf. Refer to the loader(8) and loader.conf(8) manpages. Cheers, -- Ruslan Ermilov Sysadmin and DBA of the ru@ucb.crimea.ua United Commercial Bank, ru@FreeBSD.org FreeBSD committer, +380.652.247.647 Simferopol, Ukraine http://www.FreeBSD.org The Power To Serve http://www.oracle.com Enabling The Information Age To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-stable" in the body of the message
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?19990901184657.A77510>