Re: Installation problems with GRUB
> > Next, I tried obtaining a newer version of GRUB. From
> > ftp.uruk.org/public/grub, I obtained grub-0.5.tar.gz,
> > dated August 20, 1998. I created a GRUB boot floppy according to
> > the instructions included with that archive:
> > dd if=bin/stage1 of=/dev/fd0 bs=512 count=1
> > dd if=bin/stage2 of=/dev/fd0 bs=512 seek=1
> > Unfortunately, using this GRUB boot floppy gave me the same
> > result.
> That seems like a fairly old GRUB. The debian ftp archive should
> contain a more recent version. There might be some sort of issue
> if you are using new hardware. If you install the hurd debian package
> on your slink system, or even do a:
> `dpkg -x grub*.deb /foo` you can get a known working copy of the stage1,
> stage2, etc., binaries on your system (this command will extract it
> to a "foo" directory without actually installing it on your system.)
> The two-part dd commands you listed above should work fine.
> Please try again using the debian package as instructed above, then
> let us know how it goes.
Well, I tried it again last night, and I had the same problem. I used
the grub .deb file from the sid/main/binary-hurd-i386 directory, which
is version 0.5.90.
I was looking around today, and I discovered there's a newer version out,
0.5.91, which has LBA support and support for AMI/AWARD bios extensions.
I'm getting way over my head here, but those sound familiar -- I'm pretty
sure I've got an AMI bios, and my drives definitely do the LBA thang.
In case it helps, when I boot Linux, in the kern.log file (and on the
console) I get the lines:
hda: [EZD] [remap 0->1] [1584/255/63] hda1 hda2 <hda5 hda6 hda7 hda8>
hdb: [EZD] [remap 0->1] [786/64/63] hdb1 hdb2
which leads me to believe there's something odd about my disk configuration.
Also, when I boot Windows 98, it uses EZ-BIOS so that I can use more than
8 GB of my big drive. This was installed from the software I installed
when I added the drive. I am trying to use the small drive for Hurd though,
which didn't require EZ-BIOS before I added the big one.
I'll try using 0.5.91, and see if I get anywhere. Failing that, I'll
grab the latest source from the CVS repository, try compiling that, and
see what happens.
I'll let the list know how it goes...