mandag 12. juni 2006, 14:24, skrev Falko Trojahn : > >> Then I succeeded with the following: > >> - set SATA mode in BIOS to "combined" > >> - dd if=/dev/zero of=/dev/sda count=200 > >> - reboot > >> - installation works as normal > > > > We do this on the GX620, > > but just setting SATA mode to "combined" and leave it here :-) > > Works just fine. > > Yes of course, but I wondered why this might be - something must be > left on the disk or in the bios, after success in combined mode I > can switch back to normal mode, partition the drive as a whole, and > the same thing works which didn't before. >
well, I dont know if the BIOS 'adapts' to whatever geometry is on the drive on these machines. That is: BIOS reads geometry from the drive.. I will check next time I get an GX620 for (re)install. see the what Pat writes on this: http://lists.gnu.org/archive/html/bug-parted/2004-07/msg00001.html > Well, I've read these, but I wonder where from should unattended > know which settings to set in /proc/ide/hda/settings - if. Im trying to read install.pl right now and was wondering about the same :-) ---------------- from install.pl ----------------------------------- # On Linux, we may need to correct the kernel's notion of the disk # geometry. Otherwise the disk partitioning tool will have the wrong # idea about how to create the partition, and dosemu will present the # wrong geometry to the Windows installer (resulting in a partition # which does not boot). if ($is_linux) { my $bios_head = $ENV{'LEGACY_BIOS_HEAD'}; my $bios_sect = $ENV{'LEGACY_BIOS_SECT'}; ----------------------------------------------------------------------- So, install.pl uses whats in the environment variables LEGACY_BIOS_HEAD and LEGACY_BIOS_SECT --------- master script ------------------------------- # Here we shamelessly assume that we have found the boot device (BIOS # disk 80h), and that this is the device the user wanted. Need to do # better someday. FIXME. if [ -d "$edd_dir" ] ; then declare -i LEGACY_BIOS_HEAD LEGACY_BIOS_SECT heads_file="$edd_dir/legacy_max_head"; if [ -e "$heads_file" ] ; then export LEGACY_BIOS_HEAD="$(< $heads_file)" # Heads value is useless until incremented let LEGACY_BIOS_HEAD++ fi sectors_file="$edd_dir/legacy_sectors_per_track"; [ -e "$sectors_file" ] \ && export LEGACY_BIOS_SECT="$(< $sectors_file)" ------------------------------------------------------------------ I belive this is where tings go wrong for the thinkpads.. > Thanks to Florian I finally succeeded in installing > P4P800+SATA-drive with unattended. IMHO the solution mainly came > from using parted version 1.6.22. We'll try to provide an iso and > docs asap. Cool. I will test as soon I get access to It. Let me know if I can help. BTW: Is Pat still around? _______________________________________________ unattended-devel mailing list unattended-devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/unattended-devel
