hi all, i just finished installing a dual boot of windows 2k, and slackware 9.1 (finally got around to switchin from red hat), and at first everything was working great, but i quickly found out the it was only running off of one CPU (dual XP 2000's computer) so i re-compiled the kernal with SMP support, and i also changed from i686 to athlon based, and after i did that everything was working fine within linux, but when i tried to swith to windows all i get is loading windows and after that it goes no futher..

IDE 1 CD-Rom
IDE 2 Windows HD /dev/hdb1
IDE 3 Slackware /dev/hdc1 = /
/dev/hdc2 = /swap
/dev/hdc5 = /usr
/dev/hdc6 = /opt
/dev/hdc7 = /home

IDE 4 Shared HD between windows and Linux (vFat)

here is the last part of my lilo.conf
# Linux bootable partition config begins
image = /boot/vmlinuz
root = /dev/hda2
label = Linux
read-only
# Linux bootable partition config ends
# Windows bootable partition config begins
other = /dev/hdb1
label = Windows
tabel = /dev/hdb
# Windows bootable partition config ends

lilo is installed in the MBR of the first hd (/dev/hdb)
and i've tried re-installing lilo a couple time and no go both time.......i can mount the windows drive from within linux, but cannot boot to it for some reason....anybody have any ideas ????

What error message do you get? What happens?

i don't know whether it's a typo in your post or in your config file, but table is spelled "table," not as it's listed in your post: tabel. maybe that would mess something up. ?

oops, that was my mistake it is a typo......guess i should have used spell check before i sent that. :-)

Was this the order of the drives before? I just installed Mepis and to get Lilo past the bootmanager program embedded in my old fujitsu's firmware, I had to swap the drives. So, mt WinXP drive is /dev/hdb1 like yours is. I inserted the new line in my /etc/lilo.conf that told WIndows it was on the first drive:

boot-as=0x80

You may need to do that if it still doesn't boot after the typo is corrected. Good luck

yup, this is the same drive order from when i had red hat installed, the only diffidence is when i had red hat i used grub as my boot loader, and now i'm using lilo, and also when i first installed slackware, and windows....i could boot to each of them no problem......it just started after i re-compiled the kernel......so i'm wondering if something went wrong.....also i remember reading somewhere that lilo has to be in the MBR of a HD on the master of the primary IDE channel (could be wrong)......if i can't get it working i might just move the drive order so the CD-rom is last on the secondary channel and three drives are first, and then re-install everything....don't have anything on the system as of yet....so it's not big loss.

I think if u don't let cdrom as master n HD as slave it won't be problem, when u run lilo.
It is problem that can not be decribed easily .. bcoz once HD is installed as slave the channel is controlled by CDROM, n the speed of CD is of kbps rather than mbps, throughput is not the problem bcoz when HD got the channel reserved, it just strobes the datas, that will not be many handshaking cycles between cpu <-> cdrom mechanic, then cpu <-> HD delays ..

You can just remove the table = /dev/hdb.
That always cripple lilo from booting windoze for me.

I second the motion kiko set forth.
Windows usually wants HD to be master - and that actually makes sense.

That's why you fool it into thinking its first with the boot-as=0x80 parameter in your /etc/lilo.conf (just did it 2 days ago - it works)

This command tells Lilo to boot Windows as if its mon the first drive in the BIOS ---> "0x80". This used to be accomplished with the table=/dev/hda and the map=hda-hdb and the other way around, or something. I got it screwed up before and had to google for the answer for about 1/2 a day.

I must say that kiko does have that as a point, as do you, but there is a workaround that is useful and efficient (remember if you do it that way, you don't have to rewrite the MBR of your Windows drive. LILO won't be pointing to it anyways

The other point kiko touched on is valid as well, but for different reasons (Linux will find a way to work around OS's selfishness.LOL). The drives do make a difference as to what controllers they are on as far as I/O. If your cdrom is PIO only and you put it on the same cable as your spanky new ATA133 drive that can do UDMA5, then you're not gonna get your 50-100k transfers, you'll get 2 or 3 like te cdrom is capable of. The controller goes only as fast as the slowest device, so you'd have good luck matching the fastest drives to the same controller. You'll get better I/O that way.

vectordrake you are a GOD! i went ahead and tried you advice and added in the map-drive=0x80 to the windows part of my lilo.conf, and SUCCESS windoze is booting up normal.

Thank you all for you all for replying so quickly, this is why i love the linux community