Exclude root, since that is done in fsck_root.

1. centralize the fsck error handling
2. don't check if root is on nfs
3. reboot -n if root was modified as the manual page suggests

Print out a message explaining why the system is rebooting if the root
filesystem is modified by fsck(8).

cleanup fsck error handling:
1. explain what each error means
2. remove 130 (128 + 2) which was presumably there because fsck_msdos did
not handle SIGINT properly

branches: 1.7.4; 1.7.10; 1.7.12;
* Add a stop_boot() function in rc.subr, to terminate a multi-user boot
by killing the parent process. The parent's PID is saved in $RC_PID.
* In all rc.d/* scripts that previously tried to stop the boot,
replace in-line code with "stop_boot".
* Document this.

This should fix PR 29822.

Trap SIGINT so that if fsck is interrupted by ^C, we drop into single-user
mode. Fixes misc/29822 from der Mouse, fix from Arnaud Lacombe with
comment rework by me.

branches: 1.5.2;
Add an _rc_subr_loaded variable, set to ":" by rc.subr. Scripts can use this
for a speedup by doing:
$_rc_subr_loaded . /etc/rc.subr

branches: 1.4.4;
set fsck_flags=-p in /etc/defaults/rc.conf instead of checking if it is
empty in /etc/rc.d/fsck. From Bernd Ernesti.

Allow overrides via fsck_flags

branches: 1.2.2;
Now that "fsck -p" skips file systems that are mounted read-write,
always perform the disk check (unless /fastboot exists). Previously
this would only occur when booting directly to multi-user, so the
fsck wouldn't happen after a single user boot going into multi-user.

branches: 1.1.2;
convert fsck.sh -> fsck (using kill -TERM $$ to stop autoboot).