Copy stable/9 to releng/9.3 as part of the 9.3-RELEASE cycle.

Approved by: re (implicit)
Sponsored by: The FreeBSD Foundation

Copy head to stable/9 as part of 9.0-RELEASE release cycle.

Approved by: re (implicit)

To restart, sysinstall calls execl. Since it will create a new process, we
can't check to see if sysinstall is running as init just by checking if the
PID is 0. Introduce a new option that sets the RunningAsInit flag, and update
the code to check RunningAsInit intstead of getpid().

PR: bin/38854
Submitted by: Peter Sedeffow <peter at trumanbrewery.com>
Approved by: rrs (mentor)
MFC after: 1 month

Make sysinstall WARNS=2 clean.

Approved by: cperciva (mentor)
MFC after: 1 month

Stop treating Xorg as a distribution in the mainline portion of sysinstall
and leave it to be handled in the packages section (or post-install
completely) along with all the other packages.

Don't tell the user to remove the media until its actually possible for
them to remove the media. CDs need to be unmounted before they can be
removed.

Clean out the remaining alpha-isms.

Better to just statically set the name vs. determine at run time.

Sync program name agnostic changes with SADE.

Be more consistant and use UNZIPPER everywhere.

Fix a number of documentation-lags-behind-reality bugs in sysinstall(8).
While here, fix a couple of comments too.

Submitted by: Oliver Fromme <olli -at- lurza.secnetix.de>
Approved by: re (kensmith)

Use the RB_AUTOBOOT define over 0.

Approved by: ru (mentor)

G/C some cruft.

Gratuitous "typo in comment" fix.

The /stand directory now disappears out from under sysinstall while
it is running during installations. The vsystem() function does get
used after this happens (e.g. if you manually configure the network
interface and ask it to enable the interface) so it needs to be a
little smarter about whether it uses /stand/sh or /bin/sh.

MFC after: 3 days

Make logic more readable.

Fixed assorted misuses of NULL in integer context.

Do an update mount operation to mount the mfsroot as read/write rather
than read/only when sysinstall is running as init. This fixes several
install issues.

* Negative #if's are harder to read as they don't tell exactly what arch
something applies to. So change #ifndef to an explicit list of defines.
* Treate sparc64 and ia64 as 64-bit platforms, which means larger roots.
* sparc64 should halt back to the firmware, not reset.
* sparc64 doesn't need to play MS-DOS/BIOS partition crap games.

Reviewed by: jake

Teach sysinstall that documents on the boot floppy might not be gzipped.

Reviewed by: ru

MFS: Stop reaping children. It makes evil things happen to the extraction
loop.

Move the 'type' variable inside the block where it is actually used.

DTRT in the restart case

Close all open file descriptors before restarting sysinstall.

PR: bin/30737
Submitted by: Alexey V. Neyman <alex.neyman@auriga.ru>

Silence warnings.

Add a function to free all of sysinstall's internal variables from the
environment. This fixes an annoying bug where hitting Ctrl-C and
telling sysinstall to 'restart' will do no such thing since many of
the options are still set and so you won't be prompted for them
again.

MFC after: 1 week

Whitespace change only. Use the same indention level throughout this
file and add a blank line after variable declarations.

Submitted by: jhb

If we're running as init, install a signal handler for SIGCHLD.

PR: bin/14729
Submitted by: jhb

Fix the type of the NULL arg to execl()

Idea from: Theo de Raadt <deraadt@openbsd.org>

Look for help files in /usr/src/usr.sbin/sysinstall instead of
/usr/src/release/sysinstall.

Adapt sysinstall to use the new msgNoYes() function which assumes
no as a default. Sysinstall should be both less dangerous and less
annoying as a result of this change, though that's just my opinion
(since they're the defaults which annoy ME the least :).

Changes have been made to support a concept of VAR_FIXIT_TTY being
standard or serial. This change needs to be done to the entire system that
depends on this. This way we don't have some code using OnVTY checks
and other doing
strcmp(variable_get(VAR_FIXIT_TTY), "standard") == 0
checks. Also we need to set VAR_FIXIT_TTY to "serial" if we come up on
a serial console.

Also fixed a dialog problem in that dialog was used when dialog was
disabled causing some troubles such as not letting the cursor keys
work when exiting the fixit mode on media (ie. not the fixit shell but
for example fixit on a floppy).

Submitted by: Doug Ambrisko <ambrisko@whistle.com>
PR: 22352

Teach sysinstall how to restart itself on Ctrl-C (as an addition to its old
tricks of rebooting and continuing where it was.)

Reviewed by: jkh, jhb

Initiate deorbit burn sequence for <machine/console.h>.

Replace all in-tree uses with necessary subset of <sys/{fb,kb,cons}io.h>.
This is also the appropriate fix for exo-tree sources.

Put warnings in <machine/console.h> to discourage use.
November 15th 2000 the warnings will be converted to errors.
January 15th 2001 the <machine/console.h> files will be removed.

Allow the Fix-it functionality to detect that we are on a serial console,
and DTRT rather than start the fixit shell on a non-existant vty.

PR: 19837
Submitted by: Doug Ambrisko <ambrisko@whistle.com>
Approved by: JKH

Add #includes for sysctlbyname() to quiet a warning.

Make sysinstall use the new debug.boothowto OID. It will now go
straight into debug mode if you boot -v. Also conditionalize some
annoying debugging output now that we have this ability.

Partially submitted by: msmith
Approved by: jkh [to make certain wise-acres happy ;)]

In retrospect, msgNotify() should leave its contents on the screen
longer to give the user something to look at while things are happening.
Change it to do so and insert the appropriate screen saves elsewhere.

Completely rip-out and redesign sysinstall's refresh model as well
as redoing all the menus to have proper, or at least non-hallucinogenic,
keyboard accelerators.

This requires my recent update to libdialog to work properly and will
probably also exhibit some other "interesting" behavior while the last
few missing screen clears are found (which is why I'm not going to MFC
immediately). At least now, however, sysinstall does not gratuitously
redraw random screens at the drop of a hat and drive serial console
installers out of their minds.

I shouldn't have incremented PART_OFF; it was wrong and broke label
display to boot. Also fix some various warning fluff while I'm in
here cleaning up.

o Catch up with pnp changes - kget no longer saves pnp data.
o Catch up with ncurses updates - some internal function names have changed.

Oh crud, did I ever screw the pooch! Rather than sync this with -stable,
I backed-out the changes in -current and didn't touch stable at all (I
thought I had my patch order reversed, not what actually happened).
AIEEE! I can't even blame the crack for this one since I broke my
crack pipe a few weeks ago. I think sleep deprivation gets the blame
for this one.

Medal for noticing this one goes to: Jim Bloom <bloom@acm.org>

MFC: Catch 3.2-stable sysinstall up to 4.0-current level functionality,
bringing in DHCP support. The only thing I left out were Poul-Henning's
newfs changes since I'm not sure if he's brought the rest of that support
into -stable yet. If it turns out that this is the case, I'll MFC those
changes too.

$Id$ -> $FreeBSD$

we only need to declare i if we are building for the alpha.

More Alpha ifdefage.

Another batch of fixes for dhcp support in sysinstall, now dragging
in some code from C. Stone to parse the lease information. This is still
a WIP and this commit is largely intended to allow others to sync up; the
dhclient code still only works when doing dhcp configuration post-install
and requires a bit more work on the boot floppy before it will truly
work in the minimal bootstrapping role.

Add in a hack to turn off unaligned access warnings for alpha.

Submitted by: msmith

Unbreak the fixit floppy.
Clean up some of the media handling to use common routines.

Fix incorrect initial state for two variables.

Write changes out to /etc/rc.conf again; rc.conf.site is dead!

Totally change the way variables are accounted for in sysinstall.
Now we know which variables are internal and which need to be
backed to /etc/rc.conf.site. rc.conf is not touched now.

Also kget kernel change information back properly and set up a loader.rc
file to use it.

close doc race.

Adapt sysinstall more fully to the alpha (deal with proper boot signatures,
don't present label editor, etc).

Submitted by: dfr

Deal with both foo.hlp and FOO.TXT files.

Simplify gunzip usage even more (doh!).

Call gunzip correctly.

Make serial console based installs actually work by:

1. Detecting the split /dev/ttyv0 / /dev/console case, e.g. you've
booted with the -h flag and you have a VGA card also.

2. Adding an extra "menu" for selecting terminal type and adding ANSI
to the list of compiled-in terms.

3. Opening the proper file descriptors before disowning ourselves.
Requested by: pst

Always write out rc.conf, regardless of status.
PR: 3698

Adapt to rc.conf file format.

Back out mount change here (which is the only branch it should have
occupied in the first place :)

YAMF22

Revert $FreeBSD$ to $Id$

Save /etc/sysconfig data at the very very end, so that it can be
used post-install time as a configuration tool as well.

Update comments about a kernel recompile being necessary for PS/2 mice.

OK, I've got two ideas to file in the "really seemed like a good idea
at the time, but on further reflection..." bucket with these changes.

1. Checking the media before frobbing the disks was a fine idea, and
I wish it could have worked, but that leads to a rather difficult
situation when you need to mount the media someplace and you're about
to:

a) Chroot away from your present root.
b) Newfs the root to be.

You're basically screwed since there's no place to stick the mount
point where it will be found following the newfs/chroot (and eliminating
the chroot in favor of just using the "root bias" feature would work
great for the distributions but not the pkg_add calls done by the
package installer).

2. Automatic timeout handling. I don't know why, but alarm() frequently
returns no residual even when the alarm didn't go off, which defies
the man page but hey, since when was that so unusual? Take out timeouts
but retain the code which temporarily replaces the SIGINT handler in
favor of a more media-specific handler. This way, at least, if it's hanging
you can at least whap it. I think the timeout code would have been losing
over *really slow* links anyway, so it's probably best that it go.

This should fix NFS, tape & CDROM installs again (serves me right for getting
complacent and using just the FTP installs in my testing).

Use alarm's return status to more properly denote timeout rather than keeping
an icky global around.

Add timeout support to sysinstall's general media handling,
allow SIGINT to stand for "time out now!" in certain cases.

Don't call systemShutdown() from signal handler so that the media device
is closed. The FTP library doesn't expect to be called recursively from
an interrupt. :)

Make the long-awaited change from $Id$ to $FreeBSD$

This will make a number of things easier in the future, as well as (finally!)
avoiding the Id-smashing problem which has plagued developers for so long.

Boy, I'm glad we're not using sup anymore. This update would have been
insane otherwise.

Support the use of the 2nd CDROM as a fixit aid. Also put the EHS
into a submenu in case you need to start it again (or at some other point
in the installation).

Submitted-By: joerg

Fix some bogons in my close() handling.

Cosmetic tweaks, initialize a few variables, fix a reversed conditional.

Add Joerg's user management screen for David Nugent's pw(8) program, something
which will also need to be brought in before this screen will work.

Add some commentary about how the slip startup code is bogus.

Steal Joerg's loop for more properly closing all files and graft it into
the EHS startup. My loop was functional but more bogus.

Installation cleanup, round II.

Saner screen refresh changes, part I.

If we're unable to open /dev/ttyv0, assume that we're on a single tty system
and use /dev/console.

I really think the proper test is to determine which device has been configured
to be the console (remember the RB_SERIAL flag?) and use it instead of always
trying to open /dev/ttyv0 first.

Close PR#1542. Don't just assume 24 lines, get the tty size.
Some things may still display text on the 24th line, but that's because
they've always been screens designed to fit into a minimal real-estate
and have hardwired assumptions about the dimensions. They'll be a little
harder to make dynamic.

Better error checking in helpfile expander.
Tart up some of the output a little.

Whoops! Now that I'm decompressing docs into /tmp, it's important to
make sure that /tmp is there first! :-) While I'm at it, clean up the
somewhat inexplicably bogus Mkdir() function.

Support compressed doc files again.
Add a few strategic screen clears.
Do a lot less wasted screen I/O in restoring screen contents that don't need
restoring.
Use tar instead of cp to back up /etc in installUpdate.
Don't panic when upgrade shell exits.

Fix some long-standing malloc bugs in the package handling code (freeing garbage
in one place, leaking memory in another).

Add a facility to invoke subsystems directly by naming them on
sysinstall's command-line when running post-install. A replacement
for pkg_manage might, for example, be `/stand/sysinstall configPackages'

Fix bogon where upgrade shell was entered with tty modes spammed.

Fix bug with release name checking in ftp_strat. Turned a bunch of
bogus exit()s into proper calls to systemShutdown().

Micro tweak for getting the Holographic shell message to stay on the
screen longer.

Fix more display weirdnesses that only show up during a real installation.

Add a new -fake argument to make sysinstall not actually change things while
I'm testing it.

Fix an ancient index list bug.
Make FTP re-initializion work.
Fix fix fix.

Move some of functions around in order to make this stuff easier
to work on.

Fix display refresh bug in displayHelp()
Go back to Normal menus for Media and FTP menus rather than
radio menus - the difference in behavior is confusing and sort
of detracts from the added (small) advantage of seeing what you last
picked.

Quite a few changes:

1. Fix the last display bugs (I hope) by use of dialog rebuilds at stategic
points.
2. Clean up the distributions menus so that everybody (that's reasonable)
has All and Clear options for setting/clearing things en-masse.
3. Various attempts at display optimization.
4. Change the wording of the `Don't use Write!' dialogs to make them more
explicitly define when and when not to use the option.

Bring this into sync.

I still have a _very very annoying_ display bug which occurs when a menu
item causes a submenu to be displayed - the screen repaints for the original
menu (which is restored upon return from the submenu) are off by about 4
characters. I've tried restoring the screen, the cursor position, you name
it - same deal. Grrrr! This commit is my first step in trying to get someone
else to help me look into this one since I'm just tearing my hair out at this
point!

1. Update all the copyrights to delete useless clauses 3 and 4.
2. Change more of the menu code over to new system.
3. Streamline label editor.

Major enhancements to the package adder (next step is to use the new dialog
functions to make the menus far less klunky, but one thing at a time).

Fix bug that had debug output spewing uselessly on vty2 when running multiuser.

Lots of fixes:

1. Revamp package installer to use new dependency lists and also
pkg_add's new `read from stdin' mode to prevent a copy of the package
from hitting the disk unnecessarily.

2. More fixes for running "not as init" - don't get upset if CDROM already
mounted, do the right thing instead.

3. If running as init, assume first-time install and _don't show the
(W)rite option in the fdisk screen.

4. Many other little tweaks, some of which will have to wait for fuller testing
until I can create a boot floppy (testing certain system-destroying
features of sysinstall can be a royal pain). Expect some more commits.

Miscellaneous cleanup before big round of menu restructuring (to use new
libdialog features).

Update the -current sources from the 2.1 branch.

Approved (in spirit) by: jkh

This mega-commit brings in Jordan's latest sysinstall version..
This looks like it was developed offline, and is being spammed over the
top of the existing. "That's fine by me! I dont really care how you do
it, just get it in there..." said Jordan in a conversation a short while
ago...

Merge RELENG_2_0_5 into HEAD

Remove trailing whitespace.

Sync up my work for the night. This should implement ALL possible
installation methods and provide a fairly robust set of menu options.
This should also fix a few more bugs on Poul-Henning's latest gripe
list.

Get help files form /stand/help.tgz

Shrink the boot floppy by removing all extraneous stuff. Also increase
number of inodes.

Fix various bugs reported by Poul. Implement VTY switching.

Try again to redirect the console back before rebooting.

Put the console back on ttyv0 before rebooting.

Ask name of "other" ftp site in correct place.
Make TIOCCONS redirection for console messages work properly; we no longer
open /dev/console for all I/O - we use the VTYs directly.

Clean up the configuration stuff a bit more - we should now populate
/etc with good things(tm).

Fire off ppp with all the proper initial options set.

Totally revamp the TCP/IP dialog; now network selection makes a bit more sense.

Root floppy (which actually may be able to go completely away at some point
soon!) is now loadable from ftp/nfs/dos as well as CDROM and (of course)
floppy.

Fix more problems on Poul's Gripe List.

Bring in all my fixes to Poul's gripe list as of last night.

Cleanup.

Invoke the _right_ shell. Sigh.

Whoops - that last feature I committed to put a diagnostic shell in
ttyv3 would work a lot better if I actually exec'd a shell! :-)
Reviewed by:

1. Spawn a shell on VTY4 and leave it lying around.
2. Bring in Gary's prompt stuff in extractGeneric.
3. Add another global for telling when we're running as init.

1. Bring in the first cut of Poul's ftp routines. We still don't
use them yet, but it's close (we're working on the last wrinkles
in the CD install for now).
2. Complete the CDROM installation strategy code.
3. Simplify the distribtuion loading code.
4. General error message cleanup.
5. Write the /etc/fstab file now and split those routines into config.c
6. Clean up the menus a little more.

The latest batch from me. Still some screen anomalies (One could consider
the entire libdialog one such :-( ) but functionally ok I think.

label.c: wizard mode now returns;
disks.c: clrtobot() so that deleted stuff disappears.
disks.c: offset is signed (for OnTrack diskmanager)
system.c: don't setbuf(stdout,0), it's too slow.

Make newfs options work on rootfs.
Add size argument to new_part, so it can come up with a good default for newfs.
Fix (possibly) a dialog botch after label.c's wizard mode.
Make vsystem even smarter abour crunched binaries (what a speedup!)
(You need to recompile crunchgen !)

Now cpio works for me. (it was Mostly confusion about when the chroot
happened)
Make vsystem crunch aware in an intelligent fashion.
make the boot.flp target more specific (and faster).

#if 0 out PHK's code until we can find a way of making it crunchgen
dependant so that standalone builds aren't hosed.

Change vsystem to know that the shell is crunched in.
Sanitize a couple of messages.

Clear the screen on msgNotifys.
Don't notify in vsystem() - it obscures the original message.
Put some debugging code into cpio_extract() so that I can see
why it doesn't work now. :(

Totally re-work cpio_extract.
Erase the annoying "Command output is on debugging screen" message
when it's time to switch back.

More assorted verbosity. Also don't log informational messages.

1. Incorporate Nickolay's ammended patches for Russian screenmap support
as per Andrey's letter. Make a few modifications for correctness.

2. Add Language menu back to first menu - it was too buried in the
Options menu.

3. Add size information to all distributions.

4. Add a compat20 distribution (we need to make one of these!! Any takers? Please?)

Submitted by: Nickolay N. Dudorov <nnd@gw.itfs.nsk.su> & jkh

Whoops! Put the helpful message telling you to look at the second
screen on the first screen, and not the second.. :-)

Poul says that setting fonts on monocrome displays is useless
as most mono displays aren't that intelligent anyway. Disable it in this
case.

Clear the screen after changing the font.

1. Use 8x16 fonts.
2. Don't use russian screenmap - apparently not necessary with right font.
3. Dequote bogusly quoted font name in english language setting.
4. Use setterm() and hack around an undesirable side-effect (cbreak is unset).
5. be smarter about setting OnVTY, and use it in cases where it makes sense.
Submitted by: Nickolay N. Dudorov <nnd@gw.itfs.nsk.su>

1. Primitive bad144 support (I believe an additional command is needed, but
won't know until Poul wakes up again).
2. Make vsystem() put its output on the debugging fd.
3. DTRT with root filesystem placement - now I see how this has to work
(thanks, Poul).
4. Many miscellaneous spelling errors fixed and general cleanup.

Use my own version of system() everywhere - it knows where to find the
shell!

1. Set a path to /stand by default.
2. Don't clear() when I can dialog_clear().

This will now compile and even scribble helpfully on your disks.
It remains to be seen how successfully. The distribution loading code
is still not here yet, but the partition/newfs/mount/cpio-extract cycle
is as complete as it's ever going to get, modulo possible bug fixes.
The TCP/IP setup screen is also sort of here, albeit in a highly-changing
state due to the fact that per-interface information isn't being kept
right now but is being added (thanks, Gary!).

1. Don't check CHUNK_BSD_COMPAT for now; it always returns "no" and is
getting in the way of testing right now.
2. Call system() on commands, not vsystem(). No need for vsystem() here.
3. Add a path for the shell.

Don't use curses routines until that part of the system is initialized.

Ok, we should now create all filesystems, mount them and extract the
cpio floppy at this point.

Correct a few ordering errors in how the partitions were being displayed.

Some fairly serious cleanup. The proper offset should now be used in
creating partitions. Still need to get the mount points displaying carefully,
but I need to get this into my tree on time so that I can work on that.

All the editors are finished. Now if I can just get a problem with libdisk
fixed, we should be able to fully set up the user's disk. Still to come
with next commit: filesystem setup, distribution extraction, final
configuration.

This will now write MBR entries - should be enough for testing.

Commit my latest changes before having a nap. Still not close to done,
nor is it in sync with my working sources, but it leaves me less CVS hassles
to bring in the new files at this time. Still no documentation to translate
quite yet, but soon. This stuff is actually very close now.

o Add extra menu types (radio implemented, multiple choice shortly).
o Make the framework generally more robust.
o Figured out how to nest the menu descriptions - no more grotty initialization
of menus.
o Fix bug with helpline and helpfile not being reset.
o Add stubs for the media selection code.

Coming next: Fdisk and disklabel screens using Phk's new libdisk stuff.

This commit was generated by cvs2svn to compensate for changes in r8097,
which included commits to RCS files with non-trunk default branches.

Here is my first "framework" commit of the new sysinstall. There is a LOT
more to come in the next 24 hours, this is just the first stable result of
8 hours of hacking so far. The specification format for menus is pretty
much hammered out and the beginnings (very humble) of the doc hierarchy
are present for an example. It should be quite easy to add a lot more
menus quickly to this since I did go somewhat out of my way to make the
framework easy to work with. This is NOT the glorious semi-graphical
sysinstall (or whatever its name will be) that the install-geeks are working
on, this is simply the "son of sysinstall" I've been promising to write in
the interim for 2.0.5 and 2.1R (super install doesn't come until 2.2R).