Comments inline. At 11:58 PM 11/18/02 +1300, cr wrote:
Perfectly understandable. I hope you will be equally understanding ... the long gap between my last message and your reply today means that I do not recall the details of the prior discussion very well. So please forgive me if I suggest here something I've already suggested (or you've previously tried).Sorry for the long delay. Real Life (TM) interfered with my investigations, and also a point you mentioned brought up a possible complication which I'll mention at the end...
As I check my setup, it appears that sr_mod needs an argument that tells it what device to attach to. See if adding that helps. Here, as an example, is what I do (in /etc/modules)On Friday 01 November 2002 05:05, Ray Olszewski wrote: > At 02:29 AM 11/1/02 +1300, cr wrote: > >[...] > > > >I don't have enough info about your setup (kernel details, > > > > > mostly) to know why you are getting ghe "unresolved symbol" errors, but > > > the usual reasons are kernel-version mismatch or missing dependency. > > > See what effect "modprobe sr_mod" has. > > > >Same list of four "unresolved symbol" errors, then > >"insmod /lib/modules/2.4.7-10/kernel/drivers/scsi/sr_mod.o failed" > > > >The module's there in /lib/.... , I looked. > > Yes, of course it is. The messages you are seeing say that some external > function calls *in* the module cannot be resolved. > > Usually this occurs because some other module that the one you are trying > to load provides the missing functionality; modprobe is supposed to handle > this for you (by consulting modules.dep and loading anything that is > needed). After the unsuccessful modprobe, what does "lsmod" tell you about > what is loaded? Is the module scsi_mod loaded, and what does the output say > about what other modules are using it (the stuff in [] after the size)?OK. /sbin/lsmod gives: Module Size Used by esssolo1 25376 0 (autoclean) gameport 1840 0 (autoclean) [esssolo1] soundcore 4208 4 (autoclean) [esssolo1] binfmt_misc 6064 1 autofs 11232 0 (autoclean) (unused) ipchains 36000 0 nls_iso8859-1 2800 1 (autoclean) nls_cp437 4320 1 (autoclean) umsdos 24832 1 (autoclean) msdos 5104 0 (autoclean) [umsdos] fat 31392 0 (autoclean) [umsdos msdos] usb-ohci 17936 0 (unused) usbcore 49792 1 [usb-ohci] ext3 61936 2 jbd 38976 2 [ext3] Doing /sbin/modprobe sr_mod brings a string of 'unresolved symbol' messages and "insmod sr_mod failed", and lsmod indicates no change. Trying to run X-CD-Roast gives an error message 'Failed to scan SCSI-bus' *but* lsmod indicates it's loaded a couple of modules: Module Size Used by sg 27488 0 (autoclean) (unused) scsi_mod 92208 1 (autoclean) [sg] esssolo1 25376 0 (autoclean) .. .. Then (but not before), /sbin/insmod ide-scsi works, after which X-CD-Roast works.
# the stuff to enable ide-scsi emulation for the CD-ROM
scsi_mod
sr_mod scd0
sg
ide-scsi
This is where my hazy memory of our prior discussion raises problems. Most lilely, this is the direct result of sr_mod not loading. So address that first off.However, trying to mount /mnt/cdrom still brings up a 'wrong fs type' message.
If that's not it ... is /etc/cdrom properly symlinked, probably to /dev/scd0? (I can't check my system for the right answer, because it has both a CD-R and a DVD drve, and /dev/cdrom symlinks to the DVD.) Does the same thing happen if you try to mount /dev/scd0 someplace.
No. That's just a different practice with respect to the device's group; it's the kind of distro-to-distro difference that is commonplace. As long as you are doing your tests as root, it will be of no significance.> The other reason this occurs is because the module was compiled for a > different kernel than the one loaded. Did you compile (or otherwise change) > your kernel? If so, did you also compile (or otherwise change) modules, > including sr_mod specifically? No, I'm using the standard RedHat 7.2 distribution. I'm not sufficiently confident to do alarming things to the kernel. Or modules. ;) It looks maybe like sr_mod is the cause of the problem, since I can't seem to make it load. Referring back to a previous message, you said > If you do "ls -l /dev/scd0", the result should be approximately this (this > one is from my workstation): > > brw-rw---- 1 root cdrom 11, 0 Jun 13 2001 /dev/scd0 My system gives brw-rw---- 1 root disk 11, 0 Aug 31 2001 scd0 I'm not sure if the 'disk' indicates a source of problems?
> >Is it possible to 'switch off' modules once installed? It occurs to me
> > I could switch off ide-scsi and sg and see if I can get my IDE CD-ROM
> > reading back.
>
> Well, you can remove modules with "rmmod". I don't know how that interacts
> with the lilo.conf settings that tell the kernel to use ide-scsi emulation
> for an IDE device.
Actually it's grub.conf, since I'm booting with Grub - or trying to.
/boot/grub/grub.conf says
# grub.conf generated by anaconda
#
# Note that you do not have to rerun grub after making changes to this file
# NOTICE: You have a /boot partition. This means that
# all kernel and initrd paths are relative to /boot/, eg.
# root (hd0,0)
# kernel /vmlinuz-version ro root=/dev/hda2
# initrd /initrd-version.img
#boot=/dev/hda
default=0
timeout=10
splashimage=(hd0,0)/grub/splash.xpm.gz
title Red Hat Linux (2.4.7-10)
root (hd0,0)
kernel /vmlinuz-2.4.7-10 ro root=/dev/hda2 hdb=ide-scsi
initrd /initrd-2.4.7-10.img
HOWEVER, I'm not booting successfully off the hard drive, when I start up it
says 'GRUB' and hangs. (and the previous install where I was using lilo did
the same). So I'm booting off the floppy made at install time. I presume
that reads grub.conf, but I'm not sure of that.
(Trying to fix this was next on my To Do list after the CD-ROM drive).
grub may be different, though. Certainly this comment in grub.conf -- "Note that you do not have to rerun grub after making changes to this file" -- suggests that it is; the analogous statement about lilo would NOT be true.
Since you can burn CDs, enough of the ide-scsi stuff is being loaded to allow the burning application to find the drive on the pseudo-scsi bus. This means that problems at this level ae unlikely to exist. My best guess is that you are simply not loading sr_mod correctly.I did wonder if the booting problem had any relevance to the cd-rom, but everything else on the system seems to work OK. I hope it isn't something vitally relevant I should have mentioned.
--
-------------------------------------------"Never tell me the odds!"--------
Ray Olszewski -- Han Solo
Palo Alto, California, USA [EMAIL PROTECTED]
-------------------------------------------------------------------------------
-
To unsubscribe from this list: send the line "unsubscribe linux-newbie" in
the body of a message to [EMAIL PROTECTED]
More majordomo info at http://vger.kernel.org/majordomo-info.html
Please read the FAQ at http://www.linux-learn.org/faqs
