MFC r325834,r325997,326502: Move sys/boot to stand/

This is effectively a direct commit to stable/11, due to differences between
stable/11 and head. Changes to DTS in sys/boot/fdt/dts were often
accompanied by kernel changes. Many of these were also risc-v updates that
likely had many more dependencies to MFC.

Because of this, sys/boot/fdt/dts remains as-is while everything else in
sys/boot relocates to stand/.

r325834: Move sys/boot to stand. Fix all references to new location

r325997: Remove empty directories.

r326502: Document the sys/boot -> stand move in hier.7 and the top-level README.

MFC r324551: Move lib/libstand to sys/boot/libsa

Copy head@r302406 to stable/11 as part of the 11.0-RELEASE cycle.
Prune svn:mergeinfo from the new branch, as nothing has been merged
here.

Additional commits post-branch will follow.

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

Set the malloc alignment to 64 bytes on platforms that use the U-Boot API
device drivers. Recent versions of u-boot run with the MMU enabled, and
require DMA-based I/O to be aligned to cache line boundaries.

These changes are based on a patch originally submitted by Juergen Weiss,
but I reworked them and thus any problems are purely my fault.

Submitted by: "Juergen Weiss" <weiss@uni-mainz.de>
Reviewed by: imp, nwhitehorn, jhb

- Removing some unneeded definitions of NULL(cruft related to 1970's C).
In C90, NULL is guaranteed to be declared in <stddef.h> and also in
<string.h>. Though the correct way to define NULL in FreeBSD is to
include <sys/_null.h>, other parts of libstand still require <string.h>
to build; therefore, we keep <string.h> in stand.h and add a note about
this;
- Removing no longer used 'Prototype' definition. Quote from bde@:

'Cruft related to getting incomplete struct declarations within
prototypes forward-declared before the structs. It doesn't mean
"prototype" but only part of a prototype-related hack. No longer
used.'

- Replacing iaddr_t with uintptr_t;
- Removing use of long double to determine alignment. Use a fixed 16 byte
alignment instead;

Reviewed by: bde
Obtained from: DragonFlyBSD (partially)
MFC after: 1 month

style(9) cleanup

Fixing building bustage on 32 bits platforms when WARNS >= 2. Note that
this fix only applies to zalloc.c, the other part of libstand such like
qdivrem.c still gives compilation warnings on sparc64 tinderbox builds;
therefore, WARNS level isn't changed for now.

Submitted by: Garrett Cooper <yanegomi@gmail.com>
Reviewed by: bde

Just use uintptr_t and intptr_t rather than requiring each arch to provide
explicit int/long typedefs.

Define iaddr_t and saddr_t for arm.

No reason to disable this test on powerpc.

Trust the code more than the comment(s) and correct some false
statements about pointer data type sizes, which spread probably
by copy-and-paste.

Reimplement malloc/free debugging that includes the offending file:line
info. This turned out to be rather useful on ia64 for tracking down
malloc/free problems.
Detect duplicate free()'s - otherwise these show up as a guard1 failure
and it looks like corruption instead of something simple like a second
free() where there shouldn't be.
Deal with libz using libc headers and not seeing the malloc/free stuff that
we provide in libstand. Do similar nastiness to what is done for bzlib.

Tested on: i386, ia64 (compile, run)

Define the types iaddr_t and saddr_t for sparc64.

A quick and dirty port of libstand to ia64.

Disable the end guard for now.

The test for failing the end guard was always triggering (and was reported as
such in compiler warnings). This is a temporary band-aid until I can work
out what's really going on.

Reviewed by: obrien

Inital PowerPC loader build support.

Submitted by: Benno Rice <benno@jeamland.net>

$Id$ -> $FreeBSD$

To clarify last commit, msmith says:
The reason the Alpha has less room than the i386 is because the bootstrap
is given a 256K mapping by the firmware; to add more requires extra work.

Reduce the sbrk() increment from 64K to 4K. There's not much room on the
Alpha, and wasting potentially 64K-4 bytes of RAM just isn't an option.

Prune unused zalloc components as recommended by Matt Dillon. Extra debugging
code is still enabled (it's not very expensive).

* Enable old UFS compatibility code for booting from Digital Unix formatted
disks.
* Fix a whole raft of warnings, printf and otherwise.
* Make zalloc work for alpha (just a case of using the right typedef).
* Add some (disabled) malloc debug printing to stand.h.

Replace the old and extremely icky Mach/NetBSD allocator with a similarly
compact and much better one donated by Matt Dillon. Implement a simple
sbrk() which uses the existing setheap() api.

Remove the custom allocator from the UFS code. It wasn't working quite
right, and it shouldn't be needed with the new allocator.

Fix a serious problem with changing the value of already-existent
environment variables. Don't attempt to modify the supposedly-const
argument to putenv()

Fix an off-by-one sizing error in the zipfs code detected by the new
allocator.

Submitted by: zmalloc from Matt Dillon <dillon@backplane.com>