Switch i386 boot code to libsa MI softraid.

Rename softraid boot files, which are currently in an MD location. This
will allow us to bring in a MI softraid.{c,h} in lib/libsa.

Remove support for building the boot blocks with DEBUGFLAGS=-D_TEST, which is
supposed to create a userland binary in order to test non-boot related
functionality. This feature has been bitrotting in a non-compiling state
for years, and causes a too-many-ifdefs disease now that there are intrusive
EFI changes.

No functional change.

move putchar() into libsa

"sure" miod@

move getchar() into libsa where applicable

ok miod@

Fix merge errors.

Found the hard way by deraadt@

Bring softraid support to i386/stand/libsa.

Apply a bunch of style(9) and whitespace fixes to i386/amd64 libsa, making
the code actually diffable. No binary change.

kill mcd dead. ok krw matthew millert thib

remove remnants of scd device, discovered by Frederic Perrin

- properly remove NENTS now after fixing the fallout.

ok deraadt@

- partially revert previous NENTS removal for arches which got busted.

- drop NENTS(), which was yet another copy of nitems().
no binary change


ok deraadt@

wt(4) was removed a while ago, so don't support booting off it anoymore.

i386/amd64 was ok tom
zaurus ok drahn

Enter cdboot, a CD-specific second-stage bootrap.

Testing krw@ and todd@, thanks.

assistance, testing and ok weingart@

Spacing and KNF. Partly from Joris Vink <nimadeus at pandora dot be>.

ok henning@, deraadt@

ansification and knf and protos

three four kills

strcpy/strcat/sprintf removal in all bootblocks. various testing by
various people. outside of some messy things in src/gnu, only one
thing in the main tree now violates this rule: bind

remove acd references; millert@ ok.

move cnspeed func and comspeed proto in places

branches: 1.22.4;
give better support to the bios memory maps.
provide memory maps editing through the machine memory command.
rearrange probing in machdep, so it provides less output,
also giving a shot for apm to fix the memory maps.
changes to kernel are minimal, only that is required due
to the api version bits addition and such cosmetic changes.
tested on all critical kernel,boot combinations; niklas@ ok

put support for test /boot environment back; some CLEANFILES and spaces

i386 bootblocks that work for 2.3. A tale too long to tell

don't echo all the control chars, not only bs&del

\n is handled by cnputc(), don't print \r

s/kbd/pc/ for consistency with kernel. Do even better __asms for BIOS calls.
Adapt to new console probing and setting API. Do not try to probe pc0 via
the NVRAM. Fix some other buglets.

revert to prev-cons behaviour

new cons

add nbdevs for the size of table of block devises' names
remove bogus codeseg

proper includes

fix twiddle() without breaking backspace

do \b

implement 'hd' device (i guess nobody will boot from 'rd')

fix tab expansion

extern int debug;

print debugging stuff only when debug!=0

use new device names: fd(0,a) -> fd0a

Fix device name.

devboot commit

commit all my mods to the last imported libsa stuff....
including:
- disklabel support;
- better boot cmd line
- smaller size (using some compilation switches ;)
- no more relocations in /boot, it's loaded in the place;
- better disk performance (maybe were already in there)
- installboot -n does not require write perms for device
- more debugs
- missing parts in libsa (such as cd9660 and so)
- i don't like 2 files for exec_i386 (sorry, toby, let's discuss maybe?)
tricks and tails:
- joined .text and .data (saves you a page)
- prot mode switching still in biosboot (it's freezed for awhile)
- biosdisk internals changed
- biosdev is not passed propery to the kernel (i'll fix it soon)
- sure i missed smth here to note (use the source, Luke!)

Initial /boot stuff (from Mickey)

branches: 1.1.2;
file dev_i386.c was initially added on branch new.