you do not need suse cds......just found the ftp fix here in wiki

http://portal.suse.com/sdb/en/2004/0...booting91.html

enjoy

[COLOR=royalblue]Will do, u can follow the post at :http://forums.suselinuxsupport.de/in...showtopic=6744

Sylv1

Unfortunately yes, I did this several times but :

Floppy Disk
To install the driver update from floppy disk, proceed as follows:
Download the file parted.img.gz and save it to a directory of your choice, for example, in /tmp.
Insert a floppy disk and use the following command to write the image to the medium:
gunzip -c parted.img.gz >/dev/fd0
Caution: The previous content of the floppy disk will be lost.

ok

Insert the installation CD/DVD (important: if you use an AMD-64 system, insert the 32-bit side of the DVD). Boot the first installation CD or DVD up to the point where you can choose one of the different installation variants.

using the boot.iso CD image

Press F6. A message asking you to keep the driver update ready will be displayed.

ok

Use the arrow key to select the menu item "Installation".

ok

Enter the boot parameter "fixpart=1" and press ENTER.

ok

When the message "Please choose the Driver Update medium" is displayed, insert the driver update disk you created and press "OK".

ok

In the following menu, select "floppy" and confirm with "OK".

ok, but no new driver is found because previous step already loaded it...

After completing the driver update, press "Back".

ok, then I get a prompt : Please insert CD1
As I dont have CD1, I get an error : CD1 not found continuing with manual installation

The following dialog for repairing the partition table displays your hard disk (usually /dev/hda) and the status of the partition table (broken). Select the hard disk and click "OK" to repair the partition table.

not ok, I never get this dialog

After repairing the partition table, exit the menu with "Back". Press the key combination CTRL-ALT-DEL to reboot the computer and remove all media from the drives.

Now you should be able to boot both Windows and Linux.

:-(((

my eyes can't handle your bb code for yellow, sorry to ask but can you click on edit and amend to red or something else?

Plse amend the yellow.

I am thinking out loud you may want to see the Fedora partiion issues which appear very similar
http://www.linuxquestions.org/questi...7&highlight=FC

Sorry bad colors, here is a new version

you are going to hate me, heh heh

there may be another way by using blocklist commands. you may need to do some reading but let me share the little I know.

1) after checking your bios order get your grub device map to match....a gimme
2) mount your XP manually and if successful put it in your /etc/fstab if you like
open terminal su to root password
mkdir /mnt/xp
mount -t ntfs /dev/ sda1 /mnt/xp
3) I assume XP is formatted in 512 blocks thats 512 byte sizes.
one file of exactly 512 bytes fills one block and no residual so its 1 block
one file of exactly 513 bytes fills one full block and residual into another block so its 1 + 1
then find each file on your xp that is SUPPOSED to boot your system.....more reading for you heh heh
then convert them to blocks.

4) now the theory is, if XP jumps from MBR to file one then to file two hopefully the blocklists will be
0 is start and 0 +1, means start at zero and go one block = all of mbr
1 + 20 means start at beginning of 1 and read the next 20 blocks.

I hope I have it anyhow.
5) then instead of your grub line for XP being chainloader +1 it becomes
chainloader 0 +1, 1 + 20, 21 + 200

well you get my drift.....we are jumping to mbr for sda, reading all of it then jumping to the next block after mbr, (in windows speak called the hidden sector) then reading all of that and then jumping to another area and reading all of that.

6) this is really messy. How do we know we have sequential blocks? You may have to use dd command to read the output

mbr is easy.....dd if=/dev/sda bs=512 count=1
the next block if it is truly to be used
dd if=/dev/sda bs=512 seek=1 count=X means we are counting in blocks of 512 bytes jumping (seek) 1 block then outputting X blocks.

When you read man dd it mentions the of= option but we don't need it as we print to screen.

How will you know that the output is what you need I have NO idea. heh heh.

from the manual

How to specify block lists

A block list is used for specifying a file that doesn't appear in the filesystem, like a chainloader. The syntax is [offset]+length[,[offset]+length].... Here is an example:

0+100,200+1,300+300


This represents that GRUB should read blocks 0 through 99, block 200, and blocks 300 through 599. If you omit an offset, then GRUB assumes the offset is zero.

########end of quote
comment if you have a sequential list to read
0 + 100 is better that 0 + 1, 1 +99 for mbr and hidden sector plus

Good news finally managed to get the parted fix dialog :

The following dialog for repairing the partition table displays your hard disk (usually /dev/hda) and the status of the partition table (broken). Select the hard disk and click "OK" to repair the partition table.

Bad news, the status of the windows disk partition table (sda & sdb) are both OK....

Good News
I have tested the below solution on my linux and it now rests upon your broad shoulders to prove it works for you, heh heh

linux slackware 10 amended /boot/grub/menu.lst

title mbr
rootnoverify (hd0)
chainloader 0+1

Upon reboot it took me back to my splash yippee. Now the explains.

tried rootnoverify (hd0,0) and it gave me error 13...not happy
the syntax of the chainloader is chainloader SPACE zero plus one WITHOUT GAPS so its not 0 + 1 ok?

upon reboot I also tried some stuff in a terminal. When I got the commands correct the grub prompt appeared without error messages. Thats how I found out about the no spaces. However my correct version did not work in a booted terminal which I assume is because I am booted.

2) That simplifies all that horrible blocklist stuff to this

title xp
hide (hd0,0) # your /boot partition for linux
rootnoverify (hd1) # the whole drive to get mbr
makeactive
chainloader 0+1

3) post results please. I am still concerned about your fourth drive being a linux drive so if it fails first time use the E for edit functions at the menu line to amend hd1 to hd2 or hd3 please

Good luck

hide (hd0,0)
rootnoverify (hd1)
makeactive
chainloader 0+1

Result : Error 12 : Invalid device requested

hide (hd0,0)
rootnoverify (hd1,0)
makeactive
chainloader 0+1

Result : Hangs

any chance you have tested rootnoverify (hd2) then (hd3) ?

getting also invalid device for hd2 and hd3, not sure if the whole drive syntax is ok.

darn it. The mbr works for me and I don't think I have special hardware.....actually 5 years old but not raid or sata.

there are some FC2 and suse fixes in my error section here
http://www.linuxquestions.org/questi...hreadid=237511
see if the links offer any comfort.

can you re-install using the correct geometry and then hopefully your ntldr should work?

Now if you were happy to re-install would you consider a Mandrake distro instead of suse 9.1?

As you can see I am running oout of options.

Actually I just had another thought about your partition table error for suse 9.1

when I open a terminal and type su and root passwd then grub the command
geometry (hd0)
gives output C/H/S of 1027/255/63

Then as I already knew that was correct for me I did this

geometry (hd0) 1027 255 63
there is one space gaps between each string. And the output was the same as you would expect. ie no error messages

Translating to your situation you will need a rescue cd that can accurately report your chs if you don't already know it.
eg fdisk /dev/hda assuming your suse is first drive IDE change to /dev/sda for scsi

then you do the grub geometry command with the correct values and only then do you do
root (hd0,0) # your suse drive /boot parttion
setup (hd0) to embed the grub with a possible changed blocklist or address of where the stage2 and stage1.5 files are

And then you redo the ntloader boot.ini stuff

of course with all the other stuff I have forgotten your bios order, and types so change the guilty if not right.

good luck