From: I Lee Hetherington (ilh_at_sls.lcs.mit.edu)
Date: Mon Jul 06 1998 - 13:32:43 EDT
Alexander Krumeich wrote:
> IBM has provided several fixes for OS/2 and NT users so that they can use
> the suspend/resume features or their TP600s when they use those OSes'
> native file systems -- HPFS and NTFS. The trouble seems to be that the
> BIOS (or whoever is responsible for switching to hibernation mode) can't
> read the hibernation file. Either the file is not written properly in
> the first place or it can't be read. My guess is that the BIOS needs a
> FAT formatted partition to write the file because it's written to an
> area on the hard disk that is hard coded into the BIOS, probably
> something like the first n sectors of the first primary partition. I
> guess further that the write/read operation fails when the BIOS finds a
> partition that is not FAT formatted.
I believe this is wishful thinking. For one, my problems with suspend/resume to
not involve hibernation in any way. Suspend is distinct from hibernate, and
suspend does not use the disk.
> All these people who have experienced the problem: What is your hard drive
> configuration? Is Linux the only OS on your hard disk? Do you boot Linux from
> a _primary_ partition? Is a fat partition visible when you suspend the
> machine?
I have a FAT-16 Win95 partition with a hibernation file. Linux boots from a
primary partition. Basically, this is exactly the same setup that worked on a
760C, a 760E, and a Pentium 770.
> I wonder what would happen if Linux is booted from an extended partition with
> the Win95 partition mounted, so that the BIOS can see the FAT partition and
> write the hibernation file to it. Anybody willing to try this out? Any
> comments?
I believe that the hibernation file is not at all an issue.
This problem also occurs on Pentium II TP770s.
--Lee Hetherington
This archive was generated by hypermail 2.1.3 : Thu Jan 23 2003 - 09:54:12 EST