Linux Kernel Upwards-Compatibility Deficiencies

GERALD HOOPER (ghphpr@tesco.net)
Thu, 15 Jul 1999 07:16:40 -0700

This is a multi-part message in MIME format.

------=_NextPart_000_0007_01BECE91.FC1D21E0
Content-Type: multipart/alternative;
boundary="----=_NextPart_001_0008_01BECE91.FC1D21E0"

------=_NextPart_001_0008_01BECE91.FC1D21E0
Content-Type: text/plain;
charset="iso-8859-1"
Content-Transfer-Encoding: quoted-printable

Hi Larry

I am not an expert in the internal of kernel device drivers but I need =
to contact someone who is and who can help me. The help desk staff at =
RedHat do not seem to have the necessary levels of expertise to sort out =
what is a very real problem with the linux kernels and in particular the =
differences between 2.0.36-0.7 (Linux 5.2) and 2.2.5-15 (Linux 6.0). My =
problem is that the kernels do not seem to be upwardly compatible.

I have a system which uses a modern BX board with built in AIC 7895 scsi =
controllers on which I have three scci 4.3 scci disks mounted. I also =
have an LS120 drive with a standard floppy. The reason for my upgrading =
from 5.2 to 6.0 was because of timeout problems on the scci bus where =
the 2.0.36 kernel keeps resetting after a timeout. The 6.0 system =
appears to be better in this respect but using the 2.2.5-15 kernel gives =
rise to more serious problems namely that the parallel port does not =
appear to have been configured in the kernel and I cannot gain access to =
my LS120 drive because it is not prperly detected by the system.

The essential fault appears to lie in the specification of the =
/proc/devices directory which is a virtual file system which forms part =
of the kernel. If you examine this directory on both the 2.0.36-07 =
(Linux Version 5.2) and 2.2.5-15 (Linux Version 6.0) kernels you will =
see that they are very different and that the key 'lp' item is missing =
from the 2.2.5-15 kernel. The label for the 'lp' item is 6 which I =
believe to be the major port number in the 'major minor' number port =
identifier sequence. The way in which the /proc directory is used by the =
rest of the system is critical but I have yet to find any documentation =
on this. To put matters right you would need to rebuild the kernel but =
Redhat do not provide the means to do this or of copies of the make =
files from which the kernels were originally constructed, so that their =
construction can be verified against their specification.=20

Doing a cat on the /proc/devices virtual directory for kernel 2.2.5-15=20
gave the following results

Character devices: Block devices:
1 mem 1 ramdisk
2 pty 2 fd
3 ttyp 3 ide0
4 ttyS 8 sd
5 cua 9 md
7 vcs 22 ide1
10 misc
29 fb
36 netlink
128 ptm
136 pts

The corresponding values for the 2.0.36-07 kernel were

Character devices: Block devices:
1 mem 1 ramdisk
2 pty 2 fd
3 ttyp 3 ide0
4 ttyp 8 sd
5 cua 9 md
6 lp 22 ide1
7 vcs
10 misc

While the Block devices are the same the Character Devices differ =
significantly

I attach a copy of my correspondence with the RedHat Help desk

Thanks

Gerry Hooper

ghphpr@tesco.net

------=_NextPart_001_0008_01BECE91.FC1D21E0
Content-Type: text/html;
charset="iso-8859-1"
Content-Transfer-Encoding: quoted-printable

<!DOCTYPE HTML PUBLIC "-//W3C//DTD HTML 4.0 Transitional//EN">

Hi Larry
 
I am not an expert in the internal of kernel device drivers but I = need to=20 contact someone who is and who can help me. The help desk staff at = RedHat do not=20 seem to have the necessary levels of expertise to sort out what is a = very real=20 problem with the linux kernels and in particular the differences between = 2.0.36-0.7 (Linux 5.2) and 2.2.5-15 (Linux 6.0). My problem is that the = kernels=20 do not seem to be upwardly compatible.
 
I have a system which uses a modern BX board with built in AIC 7895 = scsi=20 controllers on which I have three scci 4.3 scci disks mounted. I also = have an=20 LS120 drive with a standard floppy. The reason for my upgrading from 5.2 = to 6.0=20 was because of timeout problems on the scci bus where the 2.0.36 kernel = keeps=20 resetting after a timeout. The 6.0 system appears to be better in this = respect=20 but using the 2.2.5-15 kernel gives rise to more serious problems namely = that=20 the parallel port does not appear to have been configured in the kernel = and I=20 cannot gain access to my LS120 drive because it is not prperly detected = by the=20 system.
 
The essential fault appears to lie = in the=20 specification of the /proc/devices directory which is a virtual file = system=20 which forms part of the kernel. If you examine this directory on both = the=20 2.0.36-07 (Linux Version 5.2) and 2.2.5-15 (Linux Version 6.0) kernels = you will=20 see that they are very different and that the key 'lp' item is = missing from=20 the 2.2.5-15 kernel. The label for the 'lp' item is 6 which I believe to = be the=20 major port number in the 'major minor' number port identifier sequence. = The way=20 in which the /proc directory is used by the rest of the system is = critical but I=20 have yet to find any documentation on this. To put matters right you = would need=20 to rebuild the kernel but Redhat do not provide the means to do this or = of=20 copies of the make files from which the kernels were originally = constructed, so=20 that their construction can be verified against their specification.=20
 
Doing a cat on the /proc/devices = virtual directory=20 for kernel 2.2.5-15
gave the following = results

Character=20 devices:    Block devices:
  1=20 mem           &nbs= p;=20         1 ramdisk
  2=20 pty           &nbs= p;           =20 2 fd
  3=20 ttyp           &nb= sp;          =20 3 ide0
  4=20 ttyS           &nb= sp;  =20         8 sd
  5=20 cua           &nbs= p;   =20         9 md
  7=20 vcs           &nbs= p;           =20 22 ide1
 10 misc
 29 fb
 36 netlink
128 = ptm
136=20 pts


The corresponding values for the 2.0.36-07 kernel=20 were

Character devices:    Block = devices:
  1=20 mem           &nbs= p;     =20 1 ramdisk
  2=20 pty           &nbs= p;        =20 2 fd
  3=20 ttyp           &nb= sp;       =20 3 ide0
  4=20 ttyp           &nb= sp;       =20 8 sd
  5=20 cua           &nbs= p;        9=20 md
  6=20 lp            = ;           22=20 ide1
  7 vcs
 10 misc

 
While the Block devices are the same=20 the Character Devices differ significantly
 
I attach a copy of my correspondence with the RedHat Help = desk
 
Thanks
 
Gerry Hooper
 
ghphpr@tesco.net
 
 
 
 
------=_NextPart_001_0008_01BECE91.FC1D21E0-- ------=_NextPart_000_0007_01BECE91.FC1D21E0 Content-Type: text/html; name="Update Incident2.htm" Content-Transfer-Encoding: quoted-printable Content-Disposition: attachment; filename="Update Incident2.htm" =0A= =0A= Update Incident=0A= =0A= =0A= =0A= RedHat Support=0A= =0A= =0A= =0A= =0A= =0A= =0A= =0A= =0A= =0A= =0A= =0A= =0A= =0A= =0A= =0A= =0A= =0A= =0A= =0A= =0A= =0A= =0A= =0A= =0A= =0A= =0A=
3D""3D"redhat.com"
3D"Buy"3D"Support"3D"Download"3D"Redhat"3D""
3D"Linux3D"Knowledgebase"3D"Developer3D"User3D"Application3D"Documentation"
=0A= 3D"Search3D"on"   redhat.com Linux sites     powered by = Google
=0A= =0A=
=0A= =0A= =0A= =0A= =0A= =0A= =0A= =0A= =0A= =0A= =0A= =0A= =0A= =0A= =0A= =0A= =0A= =0A= =0A= =0A= =0A= =0A= =0A= =0A= =0A= =0A= =0A= =0A= =0A= =0A= =0A= =0A= =0A= =0A= =0A= =0A=
=0A=  
=0A=  
=0A= 3D"Support
=0A= 3D"FAQ"
=0A= 3D"Search"
=0A=
3D"Log
=0A= 3D"Register
=0A= 3D"New
=0A=
=0A= 3D"My
=0A= 3D"My
=0A= 3D"My
=0A= =0A= =0A= 		3D""3D""

=0A= =0A= =0A=
=0A=
=0A= Update = Incident
=0A= Logged in as Ghphpr@tesco.net

=0A=
=0A= =0A= =0A= =0A= =0A= =0A= =0A= =0A= =0A= =0A= =0A= =0A= =0A= =0A= =0A= =0A= =0A= =0A=
=0A= Contact Information:
=0A= Ghphpr@tesco.net=0A= =0A=
Registration:
=0A=
=0A= =0A= =0A= =0A= =0A=
91b0460a956ce482Red Hat = Linux 6.0/Intel
=0A=
Product:
=0A= =0A= Red Hat Linux 6.0
Category:
=0A= =0A= Bugs and other Problems
Summary:
=0A= =0A= Mounting LS120 drives using 6.0 kernel
Your Previous = Description:
=0A= On v5.2 I mount my LS120 drive OK
=0A= This uses an msdos filesystem supported by the 2.0.36 kernel
=0A= e.g. mount /dev/hdd /mnt/floppy2
=0A= works OK using an msdos filesystem
=0A=
=0A=
=0A= On v6.0 I cannot mount my LS120 drive because the 2.2.5-15 does not = support the msdos filesystem. On boot the hdd device is linked to the = LS120 drive as shown by the screen text but non of the filesystems can = be used to mount the drive.
=0A=
=0A= e.g non of these commands work
=0A= mount /dev/hdd /mnt/floppy2
=0A= mount -t ext2 /dev/hdd /mnt/floppy2
=0A= mount -t msdos /dev/hdd /mnt/floppy2
=0A= mount -t umsdos /dev/hdd /mnt/floppy2
=0A=
=0A= How would you suggest that I proceed to mount my LS120 under the = 2.2.5-15 kernel?
=0A=
=0A= Thank You
=0A=
=0A= Gerry Hooper
=0A=
=0A= ghphpr@tesco.net
=0A=
=0A= At 6/21/99 5:28 PM you wrote - Thanks for your response Marty
=0A= My apologies if you get this twice throgh a newer incident report
=0A=
=0A= I have only just found the update botton on your screen
=0A=
=0A=
=0A= I have tried your suggestions as per original response
=0A=
=0A= Both the msdos.o and vfat.o files are contained in the
=0A= /lib/modules/2.2.5-15/fs directory as you suggested.
=0A=
=0A= However when I tried
=0A=
=0A= mount -t vfat /dev/hdd /dev/floppy2
=0A= I get the message
=0A= fs type vfat not supported by kernel
=0A=
=0A= also when I try
=0A= mount -t msdos /dev/hdd /dev/floppy2
=0A= I get the corresponding message
=0A= fs type msdos not supported by kernel
=0A=
=0A= We are therefore back to square one.
=0A= How do you suggest that I proceed ?
=0A=
=0A= Gerry Hooper
=0A=
=0A= ghphpr@tesco.net
=0A=
=0A=
=0A= At 6/22/99 2:53 AM you wrote - I tried your suggestion to use
=0A= #insmod vfat
=0A= as root login: It did not work
=0A= I attach file showing result.
=0A= This file obtained by
=0A= #insmod vfat 2>> /tmp/gherr
=0A= Icould not mount a floppy under 2.2.5-15 kernel
=0A= Neither could I use my printer using the lpr command
=0A= I also could not produce printer output using
=0A= cat /tmp/gherr >> ttyS1
=0A= which is my printer port
=0A=
=0A= I believe that the 2.2.5-15 kernel is a very sick beast
=0A= On boot up it does not recognise my printer by getting the hardware to = respond
=0A= Also on boot up it does not perform a partition check on hdd
=0A=
=0A= All of these things work ok on my original kernel 2.0.36
=0A=
=0A= The file attached has been obtained by mounting the 2.0.36 kernel and = then copying the file gherr to a mounted floppy.
=0A=
=0A= This actual communication comes to you through my dual boot Windows 98 = facilities
=0A=
=0A= If I had to put my finger on the problems that exist with the 2.2.5-15 = kernel I would look at the bits which mount the harware and check out = the system during initialisation i.e at the init process
=0A=
=0A= Please find attached the gherr file contents
=0A= The file attachment did not appear to work on your system so I am = resending this again
=0A= At 6/22/99 2:58 PM you wrote - Hi Marty
=0A= I could not get the' add file attachment' mechanism attached to this = system to work.It generated a message saying contact supervisor. I = therefore sent you the error files through the RedHat sup-manager Email = connection.
=0A= Basically the 2.2.5-15 Kernel only generates error messages from a = command such as
=0A= #insmod vfat
=0A= (see Gherr file)
=0A=
=0A= or
=0A= #insmod msdos
=0A= (see Gherr2 file)
=0A=
=0A= My lpr utility also does not work
=0A=
=0A= During boot up the kernel does not perform the same check actions that = are carried out by the 2.0.36-37 kernel
=0A=
=0A= As stated above I believe that the errors in the 2.2.5-15 kernel are in = its use of the init process and that they are fatal !!
=0A=
=0A= Perhaps you could get your development team to try them out on a similar = architecture to my own?
=0A=
=0A= Thanks
=0A=
=0A= GerryAt 6/23/99 0:30 AM you wrote - Hi Marty
=0A= Thanks for your latest advice
=0A= I will try them out later to-doday
=0A= I did send you the error file through your sup-manager by Email
=0A= I have hand copied the vfat one below
=0A= The error file for msdos is very similar
=0A=
=0A= /lib/modules/preferred/fs/vfat.o: unresolved symbol = fat_read_super_R7101628c
=0A= /lib/modules/preferred/fs/vfat.o: unresolved symbol = fat_date_unix2dos_R83fb36a1
=0A= /lib/modules/preferred/fs/vfat.o: unresolved symbol = fat_read_inode_Rf23f0483
=0A= /lib/modules/preferred/fs/vfat.o: unresolved symbol = fat_put_inode_R00b60105
=0A= /lib/modules/preferred/fs/vfat.o: unresolved symbol = fat_dir_operations_R060b8452
=0A= /lib/modules/preferred/fs/vfat.o: unresolved symbol = fat_put_super_R6cc696f2
=0A= /lib/modules/preferred/fs/vfat.o: unresolved symbol = fat_fs_panic_R240a9fc6
=0A= /lib/modules/preferred/fs/vfat.o: unresolved symbol = fat_unlock_creation_R9e32cc5b
=0A= /lib/modules/preferred/fs/vfat.o: unresolved symbol = fat_get_entry_R57ed075c
=0A= /lib/modules/preferred/fs/vfat.o: unresolved symbol = fat_brelse_R2190b516
=0A= /lib/modules/preferred/fs/vfat.o: unresolved symbol = fat_add_cluster_R71aedf1a
=0A= /lib/modules/preferred/fs/vfat.o: unresolved symbol = fat_delete_inode_R6499d474
=0A= /lib/modules/preferred/fs/vfat.o: unresolved symbol = fat_mark_buffer_dirty_R3a53e527
=0A= /lib/modules/preferred/fs/vfat.o: unresolved symbol = fat_notify_change_Re08c2a5f
=0A= /lib/modules/preferred/fs/vfat.o: unresolved symbol = fat_esc2uni_Rdf8b5a1e
=0A= /lib/modules/preferred/fs/vfat.o: unresolved symbol = fat_write_inode_Rc75806f3
=0A= /lib/modules/preferred/fs/vfat.o: unresolved symbol = fat_lock_creation_R4d85e995
=0A= /lib/modules/preferred/fs/vfat.o: unresolved symbol = fat_scan_R84c7a6b4
=0A= /lib/modules/preferred/fs/vfat.o: unresolved symbol = fat_cache_inval_inode_R173fe278
=0A= /lib/modules/preferred/fs/vfat.o: unresolved symbol = fat_statfs_Rf5ee95c8
=0A= /lib/modules/preferred/fs/vfat.o: unresolved symbol = fat_readdirx_Rd5f12b92
=0A=
=0A= Hope this helps you to sort out the problems with the 2.2.5-15 kernel
=0A=
=0A= Gerry
=0A=
=0A= At 6/23/99 6:10 AM you wrote - HI MARTY
=0A= You were correct to pick up my typo error in my last input to you
=0A= /dev/ttyS1 should have read /dev/lp1
=0A= My apologies for the typing mistake
=0A=
=0A= I have tried out what you have suggested and the problems remain
=0A=
=0A= When using kernel 2.2.5
=0A=
=0A= cat file > /dev/lp0
=0A= cat file > /dev/lp1
=0A= cat file > /dev/lp2
=0A= cat file | lpr
=0A= all generate the same error message which is
=0A= bash: ........ Operation not supported by device
=0A=
=0A= When using kernel 2.0.36
=0A= cat file > /dev/lp1
=0A= cat file | lpr
=0A=
=0A= work ok.
=0A=
=0A= Now to the hdd problem
=0A= I copy now my /etc/fstab file for you to look at
=0A= I believe that this is ok
=0A=
=0A=
=0A= /dev/sdb3 / ext2 defaults = 1 1
=0A= /dev/sda2 /boot ext2 defaults = 1 2
=0A= /dev/sdb5 /home ext2 defaults = 1 2
=0A= /dev/sdc3 /hoop ext2 defaults = 1 2
=0A= /dev/sdb6 /opt ext2 defaults = 1 2
=0A= /dev/sdb2 /usr ext2 defaults = 1 2
=0A= /dev/sdc2 /usr/src ext2 defaults = 1 2
=0A= /dev/sdc1 /var ext2 defaults = 1 2
=0A= /dev/sdc4 /work ext2 defaults = 1 2
=0A= /dev/sda3 swap swap defaults = 0 0
=0A= /dev/fd0 /mnt/floppy ext2 noauto = 0 0
=0A= /dev/hdd/mnt/floppy2ext2noauto,user0 0
=0A= /dev/cdrom /mnt/cdrom iso9660 noauto,ro = 0 0
=0A= none /proc proc defaults = 0 0
=0A=
=0A=
=0A= There are however significant difference in the /var/dmesg files which = are created by the two kernels. I therefore attach them now for your = inspection and comments
=0A=
=0A= The first file is for kernel 2.0.36
=0A=
=0A= Memory: sized by int13 0e801h
=0A= Console: 16 point font, 400 scans
=0A= Console: colour VGA+ 80x25, 1 virtual console (max 63)
=0A= pcibios_init : BIOS32 Service Directory structure at 0x000fab20
=0A= pcibios_init : BIOS32 Service Directory entry at 0xfafa0
=0A= pcibios_init : PCI BIOS revision 2.10 entry at 0xfafd0
=0A= Probing PCI hardware.
=0A= Calibrating delay loop.. ok - 331.78 BogoMIPS
=0A= Memory: 127568k/131072k available (748k kernel code, 384k reserved, = 2164k data)
=0A= Swansea University Computer Society NET3.035 for Linux 2.0
=0A= NET3: Unix domain sockets 0.13 for Linux NET3.035.
=0A= Swansea University Computer Society TCP/IP for NET3.034
=0A= IP Protocols: IGMP, ICMP, UDP, TCP
=0A= Linux IP multicast router 0.07.
=0A= VFS: Diskquotas version dquot_5.6.0 initialized
=0A= Checking 386/387 coupling... Ok, fpu using exception 16 error = reporting.
=0A= Checking 'hlt' instruction... Ok.
=0A= Linux version 2.0.36 (root@porky.redhat.com) (gcc version 2.7.2.3) #1 = Tue Oct 13 22:17:11 EDT 1998
=0A= Starting kswapd v 1.4.2.2
=0A= Serial driver version 4.13 with no serial options enabled
=0A= tty00 at 0x03f8 (irq =3D 4) is a 16550A
=0A= tty01 at 0x02f8 (irq =3D 3) is a 16550A
=0A= PS/2 auxiliary pointing device detected -- driver installed.
=0A= Real Time Clock Driver v1.09
=0A= Ramdisk driver initialized : 16 ramdisks of 4096K size
=0A= ide: i82371 PIIX (Triton) on PCI bus 0 function 57
=0A= ide0: BM-DMA at 0xf000-0xf007
=0A= ide1: BM-DMA at 0xf008-0xf00f
=0A= hdb: ATAPI CDROM, ATAPI CDROM drive
=0A= hdd: LS-120 CSMO 05 UHD Floppy, ATAPI FLOPPY drive
=0A= ide0 at 0x1f0-0x1f7,0x3f6 on irq 14
=0A= ide1 at 0x170-0x177,0x376 on irq 15
=0A= Floppy drive(s): fd0 is 1.44M
=0A= FDC 0 is a post-1991 82077
=0A= md driver 0.36.3 MAX_MD_DEV=3D4, MAX_REAL=3D8
=0A= scsi : 0 hosts.
=0A= scsi : detected total.
=0A= Partition check:
=0A= hdd: 123264kB, 963/8/32 CHS, 533 kBps, 512 sector size, 720 rpm
=0A= hdd: no media in the drive
=0A= RAMDISK: Compressed image found at block 0
=0A= VFS: Mounted root (ext2 filesystem).
=0A= (scsi0) <Adaptec AIC-7895 Ultra SCSI host adapter> found at PCI = 12/0
=0A= (scsi0) Wide Channel A, SCSI ID=3D7, 32/255 SCBs
=0A= (scsi0) Downloading sequencer code... 404 instructions downloaded
=0A= (scsi1) <Adaptec AIC-7895 Ultra SCSI host adapter> found at PCI = 12/1
=0A= (scsi1) Wide Channel B, SCSI ID=3D7, 32/255 SCBs
=0A= (scsi1) Downloading sequencer code... 404 instructions downloaded
=0A= scsi0 : Adaptec AHA274x/284x/294x (EISA/VLB/PCI-Fast SCSI) = 5.1.2/3.2.4
=0A= <Adaptec AIC-7895 Ultra SCSI host adapter>
=0A= scsi1 : Adaptec AHA274x/284x/294x (EISA/VLB/PCI-Fast SCSI) = 5.1.2/3.2.4
=0A= <Adaptec AIC-7895 Ultra SCSI host adapter>
=0A= scsi : 2 hosts.
=0A= Vendor: IBM Model: DCAS-34330W Rev: S65A
=0A= Type: Direct-Access ANSI SCSI revision: 02
=0A= Detected scsi disk sda at scsi0, channel 0, id 0, lun 0
=0A= Vendor: IBM Model: DDRS-34560 Rev: S97B
=0A= Type: Direct-Access ANSI SCSI revision: 02
=0A= Detected scsi disk sdb at scsi0, channel 0, id 5, lun 0
=0A= Vendor: IBM Model: DDRS-34560 Rev: S97B
=0A= Type: Direct-Access ANSI SCSI revision: 02
=0A= Detected scsi disk sdc at scsi0, channel 0, id 6, lun 0
=0A= (scsi0:0:0:0) Synchronous at 40.0 Mbyte/sec, offset 8.
=0A= SCSI device sda: hdwr sector=3D 512 bytes. Sectors=3D 8467200 [4134 MB] = [4.1 GB]
=0A= sda: sda1 sda2 sda3
=0A= (scsi0:0:5:0) Synchronous at 20.0 Mbyte/sec, offset 15.
=0A= SCSI device sdb: hdwr sector=3D 512 bytes. Sectors=3D 8925000 [4357 MB] = [4.4 GB]
=0A= sdb: sdb1 sdb2 sdb3 sdb4 < sdb5 sdb6 >
=0A= (scsi0:0:6:0) Synchronous at 20.0 Mbyte/sec, offset 15.
=0A= SCSI device sdc: hdwr sector=3D 512 bytes. Sectors=3D 8925000 [4357 MB] = [4.4 GB]
=0A= sdc: sdc1 sdc2 sdc3 sdc4
=0A= VFS: Mounted root (ext2 filesystem) readonly.
=0A= Trying to unmount old root ... okay
=0A= Adding Swap: 130748k swap-space (priority -1)
=0A=
=0A=
=0A= The second file is for kernel 2.2.5-15
=0A=
=0A= Linux version 2.2.5-15 (root@porky.devel.redhat.com) (gcc version = egcs-2.91.66 19990314/Linux (egcs-1.1.2 release)) #1 Mon Apr 19 23:00:46 = EDT 1999
=0A= Detected 332503754 Hz processor.
=0A= Console: colour VGA+ 80x25
=0A= Calibrating delay loop... 331.78 BogoMIPS
=0A= Memory: 127816k/131072k available (996k kernel code, 412k reserved, = 1572k data, 60k init)
=0A= VFS: Diskquotas version dquot_6.4.0 initialized
=0A= CPU: Intel Pentium II (Deschutes) stepping 01
=0A= Checking 386/387 coupling... OK, FPU using exception 16 error = reporting.
=0A= Checking 'hlt' instruction... OK.
=0A= POSIX conformance testing by UNIFIX
=0A= mtrr: v1.26 (19981001) Richard Gooch (rgooch@atnf.csiro.au)
=0A= PCI: PCI BIOS revision 2.10 entry at 0xfafd0
=0A= PCI: Using configuration type 1
=0A= PCI: Probing PCI hardware
=0A= Linux NET4.0 for Linux 2.2
=0A= Based upon Swansea University Computer Society NET3.039
=0A= NET4: Unix domain sockets 1.0 for Linux NET4.0.
=0A= NET4: Linux TCP/IP 1.0 for NET4.0
=0A= IP Protocols: ICMP, UDP, TCP, IGMP
=0A= Initializing RT netlink socket
=0A= Starting kswapd v 1.5
=0A= Detected PS/2 Mouse Port.
=0A= Serial driver version 4.27 with MANY_PORTS MULTIPORT SHARE_IRQ = enabled
=0A= ttyS00 at 0x03f8 (irq =3D 4) is a 16550A
=0A= ttyS01 at 0x02f8 (irq =3D 3) is a 16550A
=0A= pty: 256 Unix98 ptys configured
=0A= apm: BIOS version 1.2 Flags 0x07 (Driver version 1.9)
=0A= Real Time Clock Driver v1.09
=0A= RAM disk driver initialized: 16 RAM disks of 4096K size
=0A= PIIX4: IDE controller on PCI bus 00 dev 39
=0A= PIIX4: not 100% native mode: will probe irqs later
=0A= ide0: BM-DMA at 0xf000-0xf007, BIOS settings: hda:pio, hdb:pio
=0A= ide1: BM-DMA at 0xf008-0xf00f, BIOS settings: hdc:pio, hdd:pio
=0A= hdb: ATAPI CDROM, ATAPI CDROM drive
=0A= hdd: LS-120 CSMO 05 UHD Floppy, ATAPI FLOPPY drive
=0A= ide0 at 0x1f0-0x1f7,0x3f6 on irq 14
=0A= ide1 at 0x170-0x177,0x376 on irq 15
=0A= hdb: ATAPI 36X CD-ROM drive, 128kB Cache
=0A= Uniform CDROM driver Revision: 2.54
=0A= Floppy drive(s): fd0 is 1.44M
=0A= FDC 0 is a post-1991 82077
=0A= md driver 0.90.0 MAX_MD_DEVS=3D256, MAX_REAL=3D12
=0A= raid5: measuring checksumming speed
=0A= raid5: using high-speed MMX checksum routine
=0A= pII_mmx : 809.244 MB/sec
=0A= p5_mmx : 784.098 MB/sec
=0A= 8regs : 573.024 MB/sec
=0A= 32regs : 411.480 MB/sec
=0A= using fastest function: pII_mmx (809.244 MB/sec)
=0A= scsi : 0 hosts.
=0A= scsi : detected total.
=0A= md.c: sizeof(mdp_super_t) =3D 4096
=0A= Partition check:
=0A= RAMDISK: Compressed image found at block 0
=0A= autodetecting RAID arrays
=0A= autorun ...
=0A= ... autorun DONE.
=0A= VFS: Mounted root (ext2 filesystem).
=0A= (scsi0) <Adaptec AIC-7895 Ultra SCSI host adapter> found at PCI = 12/0
=0A= (scsi0) Wide Channel A, SCSI ID=3D7, 32/255 SCBs
=0A= (scsi0) Downloading sequencer code... 374 instructions downloaded
=0A= (scsi1) <Adaptec AIC-7895 Ultra SCSI host adapter> found at PCI = 12/1
=0A= (scsi1) Wide Channel B, SCSI ID=3D7, 32/255 SCBs
=0A= (scsi1) Downloading sequencer code... 374 instructions downloaded
=0A= scsi0 : Adaptec AHA274x/284x/294x (EISA/VLB/PCI-Fast SCSI) = 5.1.15/3.2.4
=0A= <Adaptec AIC-7895 Ultra SCSI host adapter>
=0A= scsi1 : Adaptec AHA274x/284x/294x (EISA/VLB/PCI-Fast SCSI) = 5.1.15/3.2.4
=0A= <Adaptec AIC-7895 Ultra SCSI host adapter>
=0A= scsi : 2 hosts.
=0A= (scsi0:0:0:0) Synchronous at 40.0 Mbyte/sec, offset 8.
=0A= Vendor: IBM Model: DCAS-34330W Rev: S65A
=0A= Type: Direct-Access ANSI SCSI revision: 02
=0A= Detected scsi disk sda at scsi0, channel 0, id 0, lun 0
=0A= (scsi0:0:5:0) Synchronous at 20.0 Mbyte/sec, offset 15.
=0A= Vendor: IBM Model: DDRS-34560 Rev: S97B
=0A= Type: Direct-Access ANSI SCSI revision: 02
=0A= Detected scsi disk sdb at scsi0, channel 0, id 5, lun 0
=0A= (scsi0:0:6:0) Synchronous at 20.0 Mbyte/sec, offset 15.
=0A= Vendor: IBM Model: DDRS-34560 Rev: S97B
=0A= Type: Direct-Access ANSI SCSI revision: 02
=0A= Detected scsi disk sdc at scsi0, channel 0, id 6, lun 0
=0A= SCSI device sda: hdwr sector=3D 512 bytes. Sectors=3D 8467200 [4134 MB] = [4.1 GB]
=0A= sda: sda1 sda2 sda3
=0A= SCSI device sdb: hdwr sector=3D 512 bytes. Sectors=3D 8925000 [4357 MB] = [4.4 GB]
=0A= sdb: sdb1 sdb2 sdb3 sdb4 < sdb5 sdb6 >
=0A= SCSI device sdc: hdwr sector=3D 512 bytes. Sectors=3D 8925000 [4357 MB] = [4.4 GB]
=0A= sdc: sdc1 sdc2 sdc3 sdc4
=0A= autodetecting RAID arrays
=0A= autorun ...
=0A= ... autorun DONE.
=0A= VFS: Mounted root (ext2 filesystem) readonly.
=0A= change_root: old root has d_count=3D1
=0A= Trying to unmount old root ... okay
=0A= Freeing unused kernel memory: 60k freed
=0A= Adding Swap: 130748k swap-space (priority -1)
=0A=
=0A=
=0A= You will see that there is a major difference in the PARTITION CHECK = SECTION
=0A=
=0A= Hope this helps
=0A=
=0A= Gerry
=0A=
=0A=
=0A=
=0A=
=0A=
=0A=
=0A=
=0A=
=0A= At 6/24/99 2:41 PM you wrote - Hello Marty ( 24/6/99 7.45 pm)
=0A= Thanks for your latest instructions to me.
=0A= I will try them out later when I have a little free time
=0A= In the meantime I will try to clear up some of the confusion
=0A=
=0A= The situation is that I only have one machine which is a triple boot = machine
=0A= At the front end I run System Commander which is a Windows 98 = program.
=0A=
=0A= System Commander allows me to boot between
=0A= a) A floppy drive
=0A= b) Windows 98
=0A= c) Red Hat Linux (Resident Kernel 2.2.5-15) v5.2 + Xfree 3.3.3.1
=0A=
=0A= >From the floppy drive I can also boot Red Hat Linux 5.2 (Resident = Kernel 2.0.36-37) + Xfree 3.3.3.1 which is the original kernel for = 5.2
=0A=
=0A= Kernels to Linux are meant to be upwards compatible so that in using = Kernel 2.2.5-15 with the v5.2 software I thought I would be OK but = apparently I am not.
=0A=
=0A= The reason for going down this path will interest you.
=0A=
=0A= I originally tried to install v6.0 from scratch. This failed because it = did not detect my LS120 drive as does v5.2. I therefore reinstalled v5.2 = and tried the alternative of using the Upgrade path to install version = 6.0. This also failed because the system came back with a list of = partitions that it had recognised and asked me to pick out my root = partition. The only problem was that the list did not contain my root = partition.
=0A=
=0A= To overcome these problems I extracted the information to build a new = kernel from the 6.0 disk 1 and did this following the instructions given = in the book by Naba Barkati "Linux Secrets" and this worked OK. I = therefore ended up with a 2.2.5-15 kernel on a v5.2 software system = which itself had already been upgraded to Xfree 3.3.3.1.
=0A=
=0A= At this point I tried once again to upgrade my v5.2 with the = 2.2.5-kernel to the software for 6.0. This time it worked and it worked = because during the upgrade process the request to select my /root = partition never appeared.
=0A=
=0A= This method of upgrading my system never really worked however and I = could still not use my LS120 drive and the system kept freezing up on = me. At this time the person at RedHat Software that I was dealing with = had a user name called 'mgalgaci' . You may know him .
=0A=
=0A= As a result of these problems I decided that I really could not use the = software provided with version 6 of RedHat Linux and returned it to my = supplier who is LinuxIT in the UK. I did however keep the boot up floppy = for the 2.2.5-15 kernel and decided to continue with it to sort out the = underlying problems with it. After all I would have liked to have been = able to install this release of the software because otherwise I am = likely to be stuck up a gum tree in the future. I therefore tried to = approach the problems involved afresh which is whwn I began dealing with = yourself.
=0A=
=0A= I note that the dmesg file for kernel 2.2.5-15 does give an email = address of
=0A= root@porky.devel.redhat.com
=0A= whereas the corresponding email address for the 2.0.36-37 kernel is
=0A= root@porky.redhat.com
=0A= This suggests to me that the 2.2.5-15 kernel is still not fully = developed and would explain why it had so many bugs and why it was not = upwardly compatible with previous software. The 2.0.36-37 kernel appears = to work differently particularly during the "partition check: ".e.g.
=0A= hdd: 123264kB, 963/8/32 CHS, 533 kBps, 512 sector size, 720 rpm
=0A=
=0A= I also could not find any part of the dmesg in either kernel which = verified the existence of the parallel port which uses IRQ 7 ? although = both serial ports were properly identified. On my machine I also have = two USB ports but again they were not picked up by either kernel? I will = check out the pointer information and let you know the result.
=0A=
=0A= Thank you for your assistance
=0A=
=0A= Gerry Hooper
=0A=
=0A= At 6/24/99 3:37 PM you wrote - Hello Again Marty
=0A= Its now 8.30 pm and I have run the tests that you requested
=0A=
=0A= Under kernel 2.2.5-15
=0A=
=0A= the #ls -l /lib/modules/preferred points to
=0A= /lib/modules2.2.5-15
=0A=
=0A= Under Kernel 2.0.36
=0A=
=0A= the #ls -l /lib/modules/preferred points to
=0A= /lib/modules/2.0.36-0.7
=0A=
=0A= If I delete the /etc/printcap file under kernel 2.2.5-15
=0A= and then start Xwindows and use the printtool I get the messages
=0A= /dev/lp0: not detected
=0A= /dev/lp1: not detected
=0A= /dev/lp2: not detected
=0A=
=0A= the system then indicates that it may be a hardware problem and that I = should get it looked at.
=0A=
=0A= Again under 2.2.5-15 I checked out the cat file > /dev/lpX
=0A= rountines but they again failed.
=0A=
=0A= My recommendation is that you should refer the whole of this matter to =
=0A= root@porky.devel.redhat.com
=0A= for his consideration. I am always happy to oblige by running any tests = that are required because it is in my interest that I should be able to = buy and install the latest release of Red Hat Linux. As matters stand I = cannot perform this very simple function
=0A=
=0A= Thanks once again for your patience ans assistance
=0A=
=0A= Gerry Hooper
=0A=
=0A=
At 6/25/99 2:41 AM you wrote - Hi Marty
=0A= Thanks for your continued support (6/24/99 5:31PM)
=0A= I have carried out your instructions and attach the results
=0A= The original faults remain unchanged
=0A= I personally believe that there are some bugs in the 2.2.5-15 complied = binary code
=0A= These bugs are related to the ability of the 2.2.5-15 kernel to detect = hardware
=0A= They do not exist in kernel 2.0.36-37
=0A=
=0A= PRINTER UPDATE
=0A=
=0A= I modified the /etc/conf.modules file to contain the added line
=0A= alias parport_lowlevel parport_pc
=0A= It had no effect. I therefore re-booted the kernel: still no effect.
=0A= The print tool still gives the message
=0A= /dev/lp0: not detected
=0A= /dev/lp1: not detected
=0A= /dev/lp2: not detected
=0A= On my machine I do not have Samba (Lan Printer) or ncpfs(Netware = Printer) installed but this should not affect the setting up of a Local = Printer
=0A=
=0A= LSMOD ETC
=0A=
=0A= 1. lsmod response is listed now
=0A=
=0A= Module Size Used by
=0A=
=0A= aic7xxx 106928 10
=0A=
=0A=
=0A=
=0A= 2. insmod /lib/modules/2.2.5-15/fs/vfat.o is listed now through 2>> = facility
=0A=
=0A=
=0A=
=0A= /lib/modules/2.2.5-15/fs/vfat.o: unresolved symbol = fat_read_super_R7101628c
=0A=
=0A= /lib/modules/2.2.5-15/fs/vfat.o: unresolved symbol = fat_date_unix2dos_R83fb36a1
=0A=
=0A= /lib/modules/2.2.5-15/fs/vfat.o: unresolved symbol = fat_read_inode_Rf23f0483
=0A=
=0A= /lib/modules/2.2.5-15/fs/vfat.o: unresolved symbol = fat_put_inode_R00b60105
=0A=
=0A= /lib/modules/2.2.5-15/fs/vfat.o: unresolved symbol = fat_dir_operations_R060b8452
=0A=
=0A= /lib/modules/2.2.5-15/fs/vfat.o: unresolved symbol = fat_put_super_R6cc696f2
=0A=
=0A= /lib/modules/2.2.5-15/fs/vfat.o: unresolved symbol = fat_fs_panic_R240a9fc6
=0A=
=0A= /lib/modules/2.2.5-15/fs/vfat.o: unresolved symbol = fat_unlock_creation_R9e32cc5b
=0A=
=0A= /lib/modules/2.2.5-15/fs/vfat.o: unresolved symbol = fat_get_entry_R57ed075c
=0A=
=0A= /lib/modules/2.2.5-15/fs/vfat.o: unresolved symbol = fat_brelse_R2190b516
=0A=
=0A= /lib/modules/2.2.5-15/fs/vfat.o: unresolved symbol = fat_add_cluster_R71aedf1a
=0A=
=0A= /lib/modules/2.2.5-15/fs/vfat.o: unresolved symbol = fat_delete_inode_R6499d474
=0A=
=0A= /lib/modules/2.2.5-15/fs/vfat.o: unresolved symbol = fat_mark_buffer_dirty_R3a53e527
=0A=
=0A= /lib/modules/2.2.5-15/fs/vfat.o: unresolved symbol = fat_notify_change_Re08c2a5f
=0A=
=0A= /lib/modules/2.2.5-15/fs/vfat.o: unresolved symbol = fat_esc2uni_Rdf8b5a1e
=0A=
=0A= /lib/modules/2.2.5-15/fs/vfat.o: unresolved symbol = fat_write_inode_Rc75806f3
=0A=
=0A= /lib/modules/2.2.5-15/fs/vfat.o: unresolved symbol = fat_lock_creation_R4d85e995
=0A=
=0A= /lib/modules/2.2.5-15/fs/vfat.o: unresolved symbol fat_scan_R84c7a6b4
=0A=
=0A= /lib/modules/2.2.5-15/fs/vfat.o: unresolved symbol = fat_cache_inval_inode_R173fe278
=0A=
=0A= /lib/modules/2.2.5-15/fs/vfat.o: unresolved symbol = fat_statfs_Rf5ee95c8
=0A=
=0A= /lib/modules/2.2.5-15/fs/vfat.o: unresolved symbol = fat_readdirx_Rd5f12b92
=0A=
=0A=
=0A=
=0A= 3. repeated lsmod is listed now: there is no change in the values
=0A=
=0A= Module Size Used by
=0A=
=0A= aic7xxx 106928 10
=0A=
=0A= 3. the results for mount -t vfat /dev/hdd1 /dev/floppy2 are listed = now
=0A=
=0A= mount: fs type vfat not supported by kernel
=0A=
=0A=
=0A=
=0A= Anything else that I can do to sort these problems out please let me = know
=0A=
=0A= I remain of the opinion that they should be thrown back to the person = responsible for the 2.2.5-15 kernel at Red Hat because there are = definite low level binary errors in this kernel
=0A=
=0A= Kind Regards
=0A=
=0A= Gerry
=0A=
=0A=
=0A=
At 6/30/99 11:57 PM you wrote - Hi Marty
=0A= Its now five days since I heard from you
=0A= Does this mean that Redhat have given up on solving the problems with = your 6.0 kernel?
=0A= I had an Email this morning from Stephen your sup-manager to tell me = that he has sent on to you the files that I sent you on the 16th June!! = Has he been away on holiday or somthing and only looks at his mail every = three weeks?
=0A= Time is running out for maintenance but if there is anything you want me = to do please let me know? How about sending the full record of this = investigation to the Redhat Development chap listed at = root@porky.devel.redhat.com on the /var/dmesg message.
=0A= Kind Regards
=0A= Gerry Hooper
=0A= ghphpr@tesco.net
=0A= Our Previous Response:
=0A= 6.0 = "forget" to write the following to your /etc/conf.modules: alias parport_lowlevel parport_pc Add the line and try printtool again. Can you send me the following information, in the order specified? # lsmod # insmod /lib/modules/2.2.5-15/fs/vfat.o # lsmod # mount -t vfat /dev/hdd1 /dev/floppy2 Thanks, Marty">=0A= At 6/20/99 9:30 PM we wrote -
=0A=
=0A= Good evening,
=0A=
=0A= Hmm, you should have the msdos module in your /lib/modules/2.2.5-15/fs = directory. Do you see it there?
=0A=
=0A= Either way, you should try the 'vfat' module instead:
=0A=
=0A= # mount -t vfat /dev/hdd /dev/floppy2
=0A=
=0A= Let me know if this works for you.
=0A=
=0A= Thanks,
=0A= Marty
=0A=
=0A=
=0A=
=0A=
=0A= At 6/24/99 1:47 AM we wrote -
=0A=
=0A= Good morning,
=0A=
=0A= I think I may see the problem. Where does the symlink = /lib/modules/preferred point?
=0A=
=0A= # ls -l /lib/modules/preferred
=0A=
=0A= I bet it points to your old module directory 2.0.36. Is this the case? = If so, first delete the link /lib/modules/preferred and try again = loading the modules you need. I'm sorry I didn't catch this the first = time.
=0A=
=0A= As for your printer, try deleting your /etc/printcap and recreating it = with printtool. Hopefully this will clear things up.
=0A=
=0A= One more thing, do you still have 5.2 installed? I'm a bit confused = about your comparisons because I get the feeling you're doing them on = the same machine. Did you do a full upgrade from 5.2 using the boot = diskette and all that?
=0A=
=0A= Thanks,
=0A= Marty
=0A=
=0A=
=0A=
=0A=
=0A= At 6/24/99 5:31 PM we wrote -
=0A=
=0A= Good afternoon,
=0A=
=0A= Your printer problem is due to the way upgrades from 5.X -> 6.0 "forget" = to write the following to your /etc/conf.modules:
=0A=
=0A= alias parport_lowlevel parport_pc
=0A=
=0A= Add the line and try printtool again.
=0A=
=0A= Can you send me the following information, in the order specified?
=0A=
=0A= # lsmod
=0A=
=0A= # insmod /lib/modules/2.2.5-15/fs/vfat.o
=0A=
=0A= # lsmod
=0A=
=0A= # mount -t vfat /dev/hdd1 /dev/floppy2
=0A=
=0A= Thanks,
=0A= Marty
=0A= Enter Your Results And Additional Comments:
=0A= =0A=
=0A= Attachments:=0A=
=0A= =0A= Add File Attachment:
=0A=
Mark My Incident As SOLVED.
=0A=
=0A= =0A=
=0A= =0A= =0A= =0A= =0A=
=0A= =0A= =0A= =0A= =0A= =0A= =0A= =0A=
=0A= =0A= =0A= =0A= =0A= =0A= =0A= =0A= =0A=

(support - 2.4)
= =0A=
=0A=
3D""3D"top"
=0A=
 
=0A=
=0A=


=0A= =0A=


=0A=
Slashdot News |=0A= Freshmeat News | Linux Week News | Red = Hat News
=0A= Applications| User = Groups | Linux Index | Developer Resources | Knowledge Base
=0A= Corporate: Products & = Services | About Us | News & Events | Partners | Customers | Legal=0A=

=0A= copyright © = 1999 Red Hat Software, Inc. All rights reserved.
=0A=

=0A= =0A= =0A= </BODY>=0A=
=0A=

------=_NextPart_000_0007_01BECE91.FC1D21E0--

-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.rutgers.edu
Please read the FAQ at http://www.tux.org/lkml/