In message <200111230226.fAN2Qlb08807@guild.plethora.net>, Peter Seebach writes
:
>In message <20011123051758.L17760@snark.ptc.spbu.ru>, "Valeriy E. Ushakov" wri
>t
>es:
>>On Thu, Nov 22, 2001 at 18:44:05 -0600, Peter Seebach wrote:
>>> 1.  Anyone done this?
>
>>NetBSD hybernates just fine on my Dell Latitude.
>
>Argh.
>
>>> 2.  Any obvious design flaws/gotchas?  (e.g., I can't do it naively on my
>>> system, because the swap partition may get trashed while NetBSD is asleep.)
>>> 3.  Is there some established protocol for writing code which says
>>> "then replace the entire contents of memory with this stream, and
>>> jump to a specific location"?
>
>>But isn't laptop's bios supposed to handle it itself?
>
>Not on this system, anyway.  Or, at least, it doesn't seem to; it seems to
>hibernate to part of the Windows drive when you tell Windows to hibernate.
>It might be that, if I repartitioned to have a special hibernation partition,
>it'd do this automatically - unfortunately, I already have 4 FDISK partitions.
>(Windows, weird boot diagnostics, NetBSD, and BSD/OS.)
>
>>When you turn the power on, BIOS detects a hybernated netbsd in that
>>special partition and restores it instead of booting.
>
>In Windows, anyway, my system gets as far as displaying the splash screen
>before realizing that it's got a saved state to restore.  Maybe this is just
>a chintzy bios, or maybe it's the hibernation partition...

What kind of laptop is this?  On my old IBM laptops, hibernate worked 
through the BIOS; on my current one (a T21), it's done through the OS, 
which means that it doesn't work under NetBSD,

		--Steve Bellovin, http://www.research.att.com/~smb
		Full text of "Firewalls" book now at http://www.wilyhacker.com