Copy head (r256279) to stable/10 as part of the 10.0-RELEASE cycle.

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

zfs boot: try to set vfs.root.mountfrom from currdev as a fallback

This way with the new zfsloader there is no need to explicitly set zfs
root filesystem either via vfs.root.mountfrom or fstab.
It should be automatically picked up from currdev which is by default
is set from bootfs.

Tested by: Florian Wagner <florian@wagner-flo.net> (x86)
MFC after: 1 month

Introduce a new option (BOOT_PROMPT_123) that lets enter the boot prompt
only when typing the sequence "123" (opposite to the standard 'push any
button' approach).
That results useful when using serial lines sending garbage and leading
to unwilling boot prompt appearence.

Obtained from: Sandvine Incorporated
Reviewed by: emaste, jhb
Sponsored by: Sandvine Incorporated
MFC: 1 week

sys/boot/common.c
=================
Extend the loader to parse the root file system mount options in /etc/fstab,
and set a new loader variable vfs.root.mountfrom.options with these options.
The root mount options must be a comma-delimited string, as specified in
/etc/fstab.
Only set the vfs.root.mountfrom.options variable if it has not been
set in the environment.

sys/kern/vfs_mount.c
====================
When mounting the root file system, pass the mount options
specified in vfs.root.mountfrom.options, but filter out "rw" and "noro",
since the initial mount of the root file system must be done as "ro".
While we are here, try to add a few hints to the mountroot prompt
to give users and idea what might of gone wrong during mounting
of the root file system.

Reviewed by: jhb (an earlier patch)

Extend `autoboot_delay' handling - is this variable is set to -1 don't allow
user to interrupt autoboot process at all. Currently, even when
`autoboot_delay' is set to 0, loader(8) still allows autoboot process to be
interrupted by pressing any key on the console when the loader reads kernel
and modules from the disk. In some cases (i.e. untrusted environment) such
behaviour is highly indesirable and user should not be allowed to interfere
with the autoboot process at all.

Sponsored by: PBXpress Inc.
MFC after: 3 days

Whitespace only: kill trailing spaces.

Use __FBSDID().
Also some minor style cleanups.

Implement the long-awaited module->file cache database. A userland
tool (kldxref(8)) keeps a cache of what modules and versions are inside
what .ko files. I have tested this on both Alpha and i386.

Submitted by: bp

The default kernel filename is "kernel" again, not "kernel.ko".

Submitted by: mckusick

Don't indirect through a null pointer if we run out of kernel filenames.

Start countdown only _after_ the kernel was loaded. Not very fair otherwise.

Fix autoboot. Now autoboot *always* show the correct kernel name. It
gets the name from the environment variable kernelname, which is set
when a kernel is loaded. For this reason, autoboot will _first_ try
to load a kernel, and only proceed with the wait prompt after that
succeeds. If it fails, it will abort immediately.

While I understand some may think this behavior undesirable, I think
it is, overall, the best thing to do, even if we do not consider the
aesthetic issue. Notice that anyone using the default loader.rc
already has the kernel loaded before autoboot.

On unload, unset kernelname.

Separate the code that tries to load a kernel from the list of options
to the function loadakernel(). It is used by both boot() and
autoboot().

With the committed changes to the loaders modules path code, the kernel
path does not need to be hard coded.

Fix the kernel default load from "/boot/modules" to "/boot/kernel".

Reported by: dcs

The kernel is now known as `kernel.ko' and it and its matching modules
live in ``/boot/kernel/''.

Cleanup warnings. Most of these are signed/unsigned warnings, as well as
some added const's.

Delay calling the device cleanup routines until the absolute last
moment. We were cleaning up after PXE too early and the module
dependancy code would not be able to load any files if it needed
too.

The loader was written so that if /kernel was missing /kernel.old would
be booted. Due to a bug, this wasn't happening.

There is still a lesser bug in that the loader decides which file to boot
after the 10sec count down. This means the bootfile listed in the count
down in is wrong in the case where the loader will boot /kernel.old.

Update loader logic to distinguish modules vs. files.
Add support for module metadata. The old way of dependancy
handling will be supported for a while.

Reviewed by: peter

Add a cleanup function. This is needed for PXE where you should
shutdown the UNDI and unload the stack.

$Id$ -> $FreeBSD$

Walk around the end of all the silly guessing of device types and unit
numbers that we have been doing in the past, and read /etc/fstab off the
proposed root filesystem to determine the actual device name and vfs
type for the root filesystem. These are then exported to the kernel
via the environment variable vfs.root.mountfrom.

The boot countdown timer says it will boot in 1 seconds,
which is grammatically incorrect.

PR: 12628
Submitted-By: R. Matthew Emerson <rme@nightfly.apk.net>

Typo: "can't boot 'boot'" should be "can't boot 'kernel'"

Always output a linefeed when we've decided whether to boot
or display a loader prompt; either we've said
"Booting [%s]..." or we've received a non-line-feed character
and need a '\n' anyway.

Standardise on ';' as a component separator; it seems to be a little more
common than ','.

* Extend the memory available for the heap from 256k to 512k.
* Embed the stack into the bss section for loader and netboot. This
is required for netboot since otherwise the stack would be inside our
heap.
* Install loader and netboot in /boot by default.
* Fix getbootfile so that it searches for a ',' instead of a ';'
when terminating the filename.

- Add a new command 'lsdev' to list devices which might be likely to host
filesystems.
- New 'help' command and data in the help.* files (not yet installed),
provides topic and subtopic help, indexes, etc.
- Don't crash if the user tries to set an invalid console. Be helpful
instead.
- Expand tabs (badly) on the i386 video console.
- Some minor cosmetic changes.

Tweak the output one more time again. The kernel or module pathname
is useful, and usually fits all on one line with the load sizes.

Cosmetic: After the autoboot timeout, print a \n.

- VERBOSE_LS is obsolete, as the heap is much better behaved now.
- Don't whine about nodes we can't stat(); these are usually
symlinks that lead out of the filesystem.
- Autoboot is now controlled by $autoboot_delay, which is a value
in seconds or NO to disable autoboot.
- Don't autoboot at the end of boot.conf if we have already tried.
- Add a 'read' command to complement 'echo'. Both are still hidden.
- Improve the 'source' command/function so that it is possible to
source scripts off removable media. The entire script is read and
saved before beginning execution. Script lines beginning with
'@' will not be echoed when being executed. Script execution will
normally terminate at the first error, however if the script line
begins with '-' this behaviour is overriden for that command.

boot.c
Increase the robustness of the "is it time to boot yet" test;
if the time skipped the "when" time, we would miss it.
Don't spin in an endless loop if we don't find the first possible
kernel suggested. When we run out, don't try to load an empty
kernel name.

load_aout.c
printf format warnings

Use the variable with the path in it for the error message.

Bootloader update.

- Implement a new copyin/readin interface for loading modules.
This allows the module loaders to become MI, reducing code duplication.
- Simplify the search for an image activator for the loaded kernel.
- Use the common module management code for all module metadata.
- Add an 'unload' command that throws everything away.
- Move the a.out module loader to MI code, add support for a.out
kld modules.

Submitted by: Alpha changes fixed by Doug Rabson <dfr@freebsd.org>

This commit was generated by cvs2svn to compensate for changes in r38465,
which included commits to RCS files with non-trunk default branches.

This is the new unified bootstrap, sometimes known previously as the
'three-stage' bootstrap.
There are a number of caveats with the code in its current state:
- The i386 bootstrap only supports booting from a floppy.
- The kernel and kld do not yet know how to deal with the extended
information and module summary passed in.
- PnP-based autodetection and demand loading of modules is not implemented.
- i386 ELF kernel loading is not ready yet.
- The i386 bootstrap is loaded via an ugly blockmap.

On the alpha, both net- and disk-booting (SRM console machines only) is
supported. No blockmaps are used by this code.

Obtained from: Parts from the NetBSD/i386 standalone bootstrap.