[nSLUG] mounting CDROM

Donald Teed dteed at artistic.ca
Tue Jun 3 09:53:28 ADT 2003


Let's look at the situation from this angle...

The drive works OK for Linux while using the stock kernel
that is part of the Redhat installer.  Therefore it should
be able to work while running Redhat booting from your
disk.

Redhat does not provide or require ide device support in
kernel *modules*.  The stock Redhat kernel has IDE support compiled
in to the kernel binary.  I know many people think Redhat does
strange things, but providing basic IDE support is not a problem area.
The only way a kernel problem could come to be, is if you have been
compiling your own kernel, or you have changed
what the CDROM is connected to since installing Redhat
(e.g. you added an IDE controller card, you changed
something in the hardware or CMOS setup).

My guess is that this is something related to the hardware
involved.  For example, some IDE CDROMs can be set to
use DMA in the CMOS setup, but many CDROM drives do not like it.
Perhaps the installer will actually ignore that, but once
the system is booting from disk, it tries to set DMA on the
device and this is not making it happy.  This may not
be exactly what is wrong, but I mention this as an
example of the angles that need to be thought of.

I don't think we have a full dmesg report from you,
and I don't think we have enough information on the
hardware connections.  We need details, details, details
until you are ready to puke.

Post a full dmesg dump to us.  Never mind whether it
looks relevant, it will be.  There will be a clue in
there as to what is loading at boot time.  It will tell
us the Kernel command line, the chipsets involved, etc.

Open up your case and tell us what motherboard you have,
what connection the IDE cable goes to, and what the jumper
is set to on the CDROM for master/slave.  Some motherboards
have 3 or 4 IDE connection points : 1 or 2 are set up for
UDMA, and the connectors are blue in color.  If you have this,
do not plug your CDROM into the blue ones, use a black one.
If you have anything in your case related to IDE connections
other than 2 black connectors, go into your CMOS set up and
see what it says regarding disabling/enabling the
additional or PCI based IDE controller, or setting (U)DMA on 
anything.  Tell us what you see there.

I noticed one odd thing in the recent report.  50 Mhz
for the PCI bus sounds unusual.  It should normally
be 33 Mhz.  Have you been tweaking any jumpers recently?

--Donald Teed


On Mon, 2 Jun 2003, Diane Derrick wrote:

> dmesg | less says
> 
> Uniform Multi-Platform E-IDE driver Revision: 6.31
> ide: Assuming 50MHz PCI bus speed for PIO modes; override with idebus=xx
> hda: CD-ROM CDU76E, ATAPI CD/DVD-ROM drive
> ide0 at 0x1f0-0x1f7,0x3f6 on irq 14
> 
> looks ok i think (the system is EIDE) but at the very bottom it says
> 
> hda: driver not present
> 
> A problem I think.
> The CD is connected to the only ide controller on the motherboard and I know
> it all works because thats what I used to install redhat.
> 
> [root at netserver dev]# less -f /dev/hda   gives me
> /dev/hda: No such device or address
> 
> sorry to keep posting on this but I would like to get this old box set up so
> I can learn some stuff. (apache, MYsql, PHP, etc)  Thanks again for all the
> help.
> 
> ----- Original Message -----
> From: "Peter Cordes" <peter at llama.nslug.ns.ca>
> To: <nslug at nslug.ns.ca>
> Sent: Friday, May 30, 2003 9:56 AM
> Subject: Re: [nSLUG] mounting CDROM
> 
> 
> > On Wed, May 28, 2003 at 08:07:42PM -0300, Diane Derrick wrote:
> > > Sorry.  Didn't realize it was not all in the email.  Ok a new installof
> > > redhat 7.2.  1 SCSI hdd, 1 ide CDROM.  After install I am unable to
> mount
> > > CDROM.  I think the install did not install ide support.  Where does
> redhat
> > > put the module files.  Do they end in .o .  I have read the man on
> insmod
> > > and will try to add the module if the output from lsmod is telling that
> > > there is no ide support.
> >
> >  Ok, right, good summary.  You said that all files under /proc/ide are
> > empty.  That implies you _have_ files there.  The show zero size in the
> > directory listing because they're only generated on the fly by the kernel
> > when you read from them, and it doesn't bother remembering how big they
> > are.  My /proc/ide looks like:
> > llama]~$ ll /proc/ide
> > total 0
> > -r--r--r--    1 root     root            0 May 30 09:31 drivers
> > lrwxrwxrwx    1 root     root            8 May 30 09:31 hda -> ide0/hda/
> > lrwxrwxrwx    1 root     root            8 May 30 09:31 hdb -> ide0/hdb/
> > lrwxrwxrwx    1 root     root            8 May 30 09:31 hdd -> ide1/hdd/
> > dr-xr-xr-x    4 root     root            0 May 30 09:31 ide0/
> > dr-xr-xr-x    3 root     root            0 May 30 09:31 ide1/
> > -r--r--r--    1 root     root            0 May 30 09:31 piix
> >
> >  There are three IDE drives in my computer (and the CD writer is
> temporarily
> > not connected).  Reading from the files works just fine:
> > llama]~$ cat /proc/ide/piix
> >
> > Controller: 0
> >
> >                                 Intel PIIX4 Ultra 33 Chipset.
> > --------------- Primary Channel ---------------- Secondary
> Channel -------------
> >                  enabled                          enabled
> > --------------- drive0 --------- drive1 -------- drive0 ----------
> drive1 ------
> > DMA enabled:    yes              yes             no                yes
> > UDMA enabled:   yes              yes             no                yes
> > UDMA enabled:   2                2               X                 2
> > UDMA
> > DMA
> > PIO
> >
> >
> >  The kernel log messages (view with dmesg | less) also have information
> > about IDE stuff:
> > ...
> > Uniform Multi-Platform E-IDE driver Revision: 7.00beta-2.4
> > ide: Assuming 33MHz system bus speed for PIO modes; override with
> idebus=xx
> > PIIX4: IDE controller at PCI slot 00:04.1
> > ...
> > hda: QUANTUM FIREBALL CR13.0A, ATA DISK drive
> > hdb: Maxtor 5T030H3, ATA DISK drive
> > blk: queue c02d6460, I/O limit 4095Mb (mask 0xffffffff)
> > blk: queue c02d659c, I/O limit 4095Mb (mask 0xffffffff)
> > hdd: WDC WD600BB-00CAA1, ATA DISK drive
> > blk: queue c02d69e8, I/O limit 4095Mb (mask 0xffffffff)
> > ide0 at 0x1f0-0x1f7,0x3f6 on irq 14
> > ide1 at 0x170-0x177,0x376 on irq 15
> > hda: host protected area => 1
> > hda: 25429824 sectors (13020 MB) w/418KiB Cache, CHS=1582/255/63, UDMA(33)
> > hdb: host protected area => 1
> > hdb: 60030432 sectors (30736 MB) w/2048KiB Cache, CHS=3736/255/63,
> UDMA(33)
> > hdd: host protected area => 1
> > hdd: 117231408 sectors (60022 MB) w/2048KiB Cache, CHS=116301/16/63,
> UDMA(33)
> > Partition check:
> >  hda: hda1 hda2 hda3
> >  hdb: hdb1 hdb2
> >  hdd: hdd1 hdd2 hdd3 hdd4 < hdd5 hdd6 >
> > ...
> >
> >  The IDE stuff comes between PCI and agpgart stuff and the md driver,
> > followed by the network stack (NET4, and all that).  If there's nothing in
> > between PCI/AGP and RAID or TCP/IP, then IDE is not built into the kernel,
> > but is probably present as a module (but not loaded yet, if lsmod doesn't
> > list it).  (If you haven't compiled your own kernel, and there wasn't a
> > special option for installing a kernel without IDE support, you're
> probably
> > using the stock kernel, which almost certainly has IDE support.  If you
> > think you did choose an option to omit IDE support, run uname -r to see
> what
> > your kernel is called.  If it's 2.4.something-scsi or something like that,
> > it might not have IDE support, but look for a module.)  Donald explained
> > what to do about all that quite well.
> >
> >  Once you've ruled out lack of IDE support as the cause of your CDROM
> > problems, check that you can get data from the cdrom at all.  Figure out
> > which IDE channel the CDROM is on, and run (as root) less -f /dev/hda (or
> b,
> > c, or d).  You should see a bunch of binary-looking data, depending on
> what
> > kind of CDROM you've got in the drive.  (Be warned that less's search
> > funtion doesn't work right on binary data.) :(  If that works, then try:
> > mkdir /mnt/cdrom
> > mount /dev/hda /mnt/cdrom
> >
> >  That should work...
> >
> >
> > >
> > > > > So here's
> > > > > lsmod results.
> > > > > Module                  Size  Used by
> > > > > autofs                 11232   0  (autoclean) (unused)
> > > > > ne                      7040   1
> > > > > 8390                    6208   0  [ne]
> > > > > ipchains               36000   0
> > > > > ext3                   61936   1
> > > > > jbd                    38976   1  [ext3]
> > > > > aha152x                29712   2
> > > > > sd_mod                 11552   2
> > > > > scsi_mod               92176   2  [aha152x sd_mod]
> > > > > Should I be seeing something about IDE in here?  If so how would I
> go
> > > about
> > > > > installing the module(s) I need?  All files under /proc/ide are
> empty.
> > > > > iso9660 shows up in /proc/filesystems without the nodev infront of
> it.
> > > Am I
> > > > > wroong or is my kernal not supporting ide?
> > >
> > >
> > > ----- Original Message -----
> > > From: "Peter Cordes" <peter at llama.nslug.ns.ca>
> > > To: <nslug at nslug.ns.ca>
> > > Sent: Tuesday, May 27, 2003 12:00 PM
> > > Subject: Re: [nSLUG] mounting CDROM
> > >
> > >
> > > > On Fri, May 23, 2003 at 10:07:03PM -0300, Diane Derrick wrote:
> > > > > Sorry it took so long to get back to this but I've been away.
> > > >
> > > >  Maybe you could remind us what sort of problems you're having;  Just
> a
> > > quick
> > > > summary to trigger the memory as to which of the recent parade of
> problems
> > > > we're talking about here :)
> > > >
> > > >
> > > >  You have a /proc/ide, so your kernel supports IDE.  If it didn't find
> any
> > > > IDE controllers it supported, then nothing would show up there, I
> > > suppose...
> > > > If this is a "standard" kernel compiled by the organization that
> supplied
> > > > the distribution you're using, it's pretty much guaranteed to include
> IDE
> > > > built-in (_not_ as modules), because a lot of people want to boot from
> IDE
> > > > devices, and that only works when IDE support is compiled in[1].
> > > >
> > > >  You should run
> > > > dmesg | less
> > > >  and look for anything about IDE.  Also, lspci  (or lspci -v) might be
> > > > useful, to confirm that there is a (PCI) IDE controller in the system,
> > > > whether Linux is talking to it or not.  I've seen a mobo where the IDE
> > > > controller was fried by a power surge, but the computer booted and ran
> > > fine
> > > > (off a floppy).
> > > >
> > > > > Thanks again for the help.
> > > > >
> > > > > ----- Original Message -----
> > > > > From: <bdavidso at supercity.ns.ca>
> > > > > To: "NSLUG Maillist" <nslug at nslug.ns.ca>
> > > > > Sent: Thursday, May 08, 2003 9:42 PM
> > > > > Subject: Re: [nSLUG] mounting CDROM
> > > > >
> > > > >
> > > > > > Hi:
> > > > > >
> > > > > > On Thu, 8 May 2003, Diane Derrick wrote:
> > > > > >
> > > > > > > No offence taken.  I had the install cd for redhat 7.2 in when I
> > > tried
> > > > > to
> > > > > > > mount.  As for kernel modules, does redhat install those
> > > automatically?
> > > > > If
> > > > > > > not how would I check for their presence (newbie question)?
> > > > > >
> > > > > > Try '/sbin/lsmod' to list installed modules.  You could also try
> 'cat
> > > > > > /proc/filesystems' to see what filesystems are currently supported
> by
> > > your
> > > > > > kernel.  And 'cat /proc/partitions' will show you what partitions,
> > > even
> > > > > > unmounted, are present.  And there is a bunch of stuff under
> /proc/ide
> > > > > > that can tell you more about configured ide devices.
> > > > > >
> > > > > > --
> > > > > > Bill Davidson
> > > > > > bdavidso at supercity.ns.ca
> > > >
> > > > [1] Actually, the loaded-by-LILO ramdisk (initrd) lets you have even
> IDE
> > > as
> > > > modules, if you want, since the kernel boots from the initrd, loads
> > > modules
> > > > out of it, then switches the root device from the initrd to a hard
> drive
> > > > partition.
> > > >
> > > > --
> > > > #define X(x,y) x##y
> > > > Peter Cordes ;  e-mail: X(peter at llama.nslug.n , s.ca)
> > > >
> > > > "The gods confound the man who first found out how to distinguish the
> > > hours!
> > > >  Confound him, too, who in this place set up a sundial, to cut and
> hack
> > > >  my day so wretchedly into small pieces!" -- Plautus, 200 BC
> > > >
> > > > _______________________________________________
> > > > nSLUG mailing list
> > > > nSLUG at nslug.ns.ca
> > > > http://nslug.ns.ca/cgi-bin/mailman/listinfo/nslug
> > > >
> > >
> > > _______________________________________________
> > > nSLUG mailing list
> > > nSLUG at nslug.ns.ca
> > > http://nslug.ns.ca/cgi-bin/mailman/listinfo/nslug
> >
> > --
> > #define X(x,y) x##y
> > Peter Cordes ;  e-mail: X(peter at llama.nslug.n , s.ca)
> >
> > "The gods confound the man who first found out how to distinguish the
> hours!
> >  Confound him, too, who in this place set up a sundial, to cut and hack
> >  my day so wretchedly into small pieces!" -- Plautus, 200 BC
> >
> > _______________________________________________
> > nSLUG mailing list
> > nSLUG at nslug.ns.ca
> > http://nslug.ns.ca/cgi-bin/mailman/listinfo/nslug
> 
> _______________________________________________
> nSLUG mailing list
> nSLUG at nslug.ns.ca
> http://nslug.ns.ca/cgi-bin/mailman/listinfo/nslug
> 




More information about the nSLUG mailing list