jwz@teak.Berkeley.EDU (Jamie Zawinski) (10/17/89)
Ok, WHY is it that warm booting boots the file system and resets the network? Usually when I warm boot, it's because I've done something stupid and crashed the machine. That does NOT mean that I want all of my open files to abort, or that I want all of my telnet connections to wedge. If I was of that mindset, I'd probably be off in the corner right now flagelating myself in search of enlightenment. It may be that warm-booting hardware-resets the ethernet board, making open connections invalid (?) but the filesystem is just this big tree-structure in memory. I can't see any reason to re-read that other than plain spitefulness. (Ok I'm done ranting now.) -- Jamie
pf@islington-terrace.csc.ti.com (Paul Fuqua) (10/17/89)
Date: Monday, October 16, 1989 9:33pm (CDT)
From: pasteur!jwz%teak.Berkeley.EDU at ucbvax.berkeley.edu (Jamie Zawinski)
Subject: warm booting
Ok, WHY is it that warm booting boots the file system and resets the network?
Usually when I warm boot, it's because I've done something stupid and crashed
the machine. That does NOT mean that I want all of my open files to abort, or
that I want all of my telnet connections to wedge. If I was of that mindset,
I'd probably be off in the corner right now flagelating myself in search of
enlightenment.
One of the guiding lights of the Explorer project described warm-booting
as throwing away an arbitrary chunk of system state, then starting over
and hoping it works. In the microcode, a warm-boot resets some
registers and does some minimal cleanup, then jumps into the boot
sequence after the disk-setup code.
Anyway, warm-booting hits all the active processes, including all those
network receivers; that's what kills the network connections. The file
system is probably rebooted to ensure consistency (I don't know nothing
about booting no file systems).
If it really bothers you, you could try moving the "Boot File System"
initialization from the warm list to the cold list and disk-save, then
see what breaks next time you boot and warm-boot (kids, don't try this
at home).
Paul Fuqua pf@csc.ti.com
{smu,texsun,cs.utexas.edu,rice}!ti-csl!pf
Texas Instruments Computer Science Center
PO Box 655474 MS 238, Dallas, Texas 75265