installboot(8): fix up markup

Don't overuse .Sy - when everything is highlighted, nothing is.  Use
.Ic for options &c to get correct PostScript output (both are bold in
plain text).

Use Aq Ar inside .Pa, as both Pa and Ar are rendered as underscored
text in plain text output, and the distinction is lost.

Don't set examples in bold, but give them .Pp space around - they are
much easier to read this way.

Use consistent -width in FILES.

(uwe)

ftp 20230516 for openssl 3.0 compat SSL_OP_IGNORE_UNEXPECTED_EOF

(lukem)

installboot(8): formatting improvements

Use Ar foo instead of <foo>; this is mdoc.
Other formatting tweaks.

(lukem)

build.sh, BUILDING: consistency improvements

Consistently use "show", instead of sometimes using "display" or "print".
Fix mdoc markup (even though BUILDING is raw text).
Document help, -P, -?

(lukem)

BUILDING: update upstream, regen

doc/BUILDING.mdoc is the upstream for BUILDING, so add recent changes
in the latter to the former, formatting correctly.

Move INSTALLBOOT_UBOOT_PATHS to Environment variables.
Move INSTALLBOOT_BOARDS to "make" variables for full builds.
Add installboot(8) cross-reference.

(lukem)

try to fix powerpc compat

(christos)

include earm to <=5

(christos)

fix ia64

(christos)

use the pic version of the common library (affects hppa/alpha)

(christos)

Add the engines directory too (hannken)
Note that we don't install engines now, but we could.

(christos)

Ignore EOF from remote.

(christos)

openssl: Limit armv8 stuff to aarch64.

Strictly speaking there is armv8 aarch32, e.g. AES instructions are
exposed to 32-bit mode, but OpenSSL's `*_armv8.S' files are staunchly
aarch64 and don't work when built for aarch64 32-bit compat libraries
(which could take advantage of armv8 instructions, but only if the
code is actually aarch32, which it isn't).

The armv8 stuff is still included in aarch64 /usr/lib/libcrypto, just
not in the 32-bit compat libraries.

With any luck, this will fix the aarch64 clang build (again).

(riastradh)

Tickets #170 - #174

(martin)

Pull up following revision(s) (requested by lukem in ticket #171):

usr.bin/ftp/ssl.c: revision 1.15
usr.bin/ftp/util.c: revision 1.167
usr.bin/ftp/ftp.c: revision 1.175
usr.bin/ftp/version.h: revision 1.97

add timeout for ssl connect

Implement a timeout for SSL connection setup, using -q QUITTIME,
defaulting to 60 seconds.

SSL_connect(3) (unlike connect(2)) doesn't timeout by default.
Adapt ssl error messages destination: if unexpected error
from local API, use warn()/warnx() to stderr;
if expected error from a network operation (e.g., timeouts),
use fprintf to ttyout (which might be stdout).

Consistently use ftp_poll() instead of select();
ssl.c (using select()) was added 7 years after the
previous uses of select() were converted to poll().

Check EAGAIN as well as existing EINTR error from ftp_poll(),
for portability.

(martin)

Pull up following revision(s) (requested by lukem in ticket #174):

usr.bin/ftp/ssl.c: revision 1.14

Simplify includes

Include "ftp_var.h" instead of various system headers and "extern.h".

(martin)

Pull up following revision(s) (requested by lukem in ticket #173):

usr.bin/ftp/ftp.1: revision 1.150

ftp(1): better mark up for url vs cd example

(martin)

Pull up following revision(s) (requested by lukem in ticket #172):

usr.bin/ftp/ftp.1: revision 1.149

ftp(1): minor markup tweaks

Use .Ql instead of .Sq Li, add some missing ones.  Use .Pq instead of
explicit () for longer phrases - these are easier to read in the
postscript output b/c of extra spacing.

(martin)

Pull up following revision(s) (requested by lukem in ticket #170):

usr.bin/ftp/ssl.c: revision 1.13
usr.bin/ftp/fetch.c: revision 1.236
usr.bin/ftp/util.c: revision 1.166
usr.bin/ftp/main.c: revision 1.129
usr.bin/ftp/extern.h: revision 1.83
usr.bin/ftp/ftp.1: revision 1.148
usr.bin/ftp/cmdtab.c: revision 1.53
usr.bin/ftp/version.h: revision 1.96

Add option sslnoverify to control validation of SSL certificates.

Add netrc processing to fetch-mode (URL on command line) to enable options and autologin
via netrc.

Fix SSL cleanup in some error paths.

Certificate validation is now enabled by default. Set FTPSSLNOVERIFY=1 in environment
or configure a corresponding init macro via netrc to not validate certs (required if
you haven't installed a required CA certificate for OpenSSL).

Discussed with lukem@ on icb.

(martin)

indent: directly access the input buffer

No functional change.

(rillig)

OpenSSL entries are handled by NetBSD.dist.compat

(christos)

OpenSSL entries are handled by NetBSD.dist.compat

(christos)

indent: allow comments in column 1 to be formatted

(rillig)

indent: remove support for form feed characters inside a line

Form feeds are occasionally used to split code into pages, and this use
is still supported.  Having a form feed in the middle of a line is
exotic.

(rillig)

Simplify previous, from Riastradh

(martin)

indent: remove blank between comment and parentheses or brackets

Finally, indent formats its own source code without messing up the
layout.

(rillig)

indent: fix handling of INDENT OFF/ON comments

Previously, the 'INDENT OFF' comments were interpreted when the newline
token from the line above the comment was processed, which was earlier
than could be reasonably expected.

The 'INDENT ON' comments were interpreted equally early, which led to
the situation that the 'INDENT OFF' comments were preserved literally
but the 'INDENT ON' comments weren't.

(rillig)

doc: MAKEVERBOSE support

(lukem)

indent: move parsing of 'INDENT OFF/ON' comments to the lexer

No functional change.

(rillig)

indent: clean up detection of whether parentheses form a cast

No functional change.

(rillig)

indent: fix cast detection

In process_lparen_or_lbracket, ps.paren[...].maybe_cast was not
initialized, which may have been the cause for seemingly random spacing
around binary operators.

While here, clean up the code by reducing the number of accesses to the
parser state.

(rillig)

fix typo

(christos)

indent: fix detection of casts

A word followed by a '(' does not start a cast expression.

(rillig)

tools: if MAKEVERBOSE >=2, noisy automake builds

Explicitly build with V=1 for MAKEVERBOSE >= 2
so that Makefile.gnuhost (automake) consumers
are more consistent with native builds.

(If the upstream is using AM_SILENT_RULES(yes)
then the default is V=0).

(lukem)

add more OpenSSL-3.x

(christos)

add compat files

(christos)

add riskv64 from Robert Swindells

(christos)

+LRO Large Receive Offload, +RSC Receive Segment Coalescing (win counterpart)
PR 57408

(andvar)

indent: fix type cast in function definition

(rillig)

indent: clean up local indentation profile

The -eei option now works, the type hints are no longer necessary.

(rillig)

indent: fix duplicate space between comment and binary operator

(rillig)

indent: format its own code, extend some comments

With manual corrections, as there are still some bugs left.

No functional change.

(rillig)

indent: fix line wrapping of comments to the right of code

(rillig)

Add openssl files

(christos)

add openssl files

(christos)

indent: improve type guessing, fix formatting of declarations

(rillig)

tests/indent: test type detection

(rillig)

indent: remove accidental space before tab

(rillig)

indent: remove backslash line continuation outside preprocessing

The indenter did not handle these backslashes well, interpreting them as
unary operators, and they are an edge case anyway.  Line continuations
in string literals and character constants are kept.

(rillig)

32bit sparc needs the atomic hack too

(martin)

switch everyone to OpenSSL-3.x

(christos)

fix ppc, sh3

(christos)

tests/indent: demonstrate wrong spacing after function prototype

(rillig)

indent: fix spacing between function prototype and attributes

(rillig)

armv4 also needs the atomic hack

(martin)

indent: fix indentation of struct member names

(rillig)

Fix armv5

(martin)

fix build

(christos)

fix mips

(christos)

tests/indent: demonstrate wrong indentation of struct members

(rillig)

indent: indent multi-line conditions

No functional change.

(rillig)

indent: fix indentation of statements after controlling expression

(rillig)

tests/indent: add another test for -eei -nlp

(rillig)

indent: fix indentation of expressions in -nlp -eei mode

(rillig)

tests/indent: test extra indentation with fixed continuation indentation

(rillig)

Tickets #166 - #169

(martin)

Pull up following revision(s) (requested by he in ticket #169):

sys/arch/mips/include/vmparam.h: revision 1.67

Bump MAXTSIZ from 64MB to 128MB also for o32.

This so that the rather large cc1 from gcc12 can be run.

OK'ed by simonb@

(martin)

Pull up following revision(s) (requested by chs in ticket #168):

sys/sys/types.h: revision 1.106
sys/arch/hppa/stand/xxboot/Makefile: revision 1.4

Avoid 64bit math in the hppa first stage bootloader. Ok chs@

(martin)

Pull up following revision(s) (requested by chs in ticket #167):

sys/uvm/uvm_map.c: revision 1.406

uvm: avoid a deadlock in uvm_map_clean()

The locking order between map locks and page "busy" locks
is that the page "busy" lock comes first, but uvm_map_clean()
breaks this rule by holding a map locked (as reader) while
waiting for page "busy" locks.

If another thread is in the page-fault path holding a page
"busy" lock while waiting for the map lock (as a reader)
and at the same time a third thread is blocked waiting for
the map lock as a writer (which blocks the page-fault thread),
then these three threads will all deadlock with each other.

Fix this by marking the map "busy" (to block any modifications)
and unlocking the map lock before possibly waiting for any
page "busy" locks.

Martin Pieuchot reported that the same problem existed in OpenBSD
he applied this fix there after several people tested it.

fixes PR 56952

(martin)

Pull up following revision(s) (requested by chs in ticket #166):

sys/arch/x68k/stand/boot_ufs/Makefile: revision 1.39

Add udivdi3/umoddi3 (and ashldi3) that were required by recent UFS change.

This increases xxboot_ufs from about 4.5KB to 5.5KB (The upper limit is 8KB).

Thanks tsutsui@ for advices.

(martin)

Ammend entry for ticket #1632

(martin)

Apply patch, requested by manu in ticket #1632:

sys/arch/xen/conf/files.xen (apply patch)

Build fix to define the SELFRELOC option for XEN kernels.
This is not needed in newer branches as those include amd64/conf/files.amd64
from x86/conf/files.x86 (and get the definition via that path)

(martin)

indent: fix indentation of multi-line '?:' expressions in functions

(rillig)

indent: group parser state by topic

No functional change.

(rillig)

indent: let indent format its own code

With manual corrections, as indent does not properly indent multi-line
'?:' expressions nor multi-line controlling expressions.

(rillig)

tests/indent: -badp should not add blank lines in initializers

(rillig)

indent: document feature toggle for debugging output

(rillig)

indent: fix spacing in for loop with declaration (since 2022-02-13)

(rillig)

tests/indent: demonstrate wrong spacing in for loops

(rillig)

indent: remove redundant include lines

(rillig)

indent: clean up memory allocation

No functional change.

(rillig)

Count the number of link down events in the MAC using with LINK_DN_CNT.

- Add new event counter "link_dn_cnt" to count the number of link down
  events in the MAC.
- The LINK_DN_CNT register (at 0x0403c) is described only in the
  Denverton's datasheet, so use it only on ixgbe_mac_X550EM_a.

(msaitoh)

indent: move debugging code to separate file

No functional change.

(rillig)

indent: clean up memory and buffer management

Remove the need to explicitly initialize the buffers.  To avoid
subtracting null pointers or comparing them using '<', migrate the
buffers from the (start, end) form to the (start, len) form.  This form
also avoids inconsistencies in whether 'buf.e == buf.s' or 'buf.s ==
buf.e' is used.

Make buffer.st const, to avoid accidental modification of the buffer's
content.

Replace '*buf.e++ = ch' with buf_add_char, to avoid having to keep track
how much unwritten space is left in the buffer.  Remove all safety
margins, that is, no more unchecked access to buf.st[-1] or appending
using '*buf.e++'.

Fix line number counting in lex_word for words that contain line breaks.

No functional change.

(rillig)

uvm: avoid a deadlock in uvm_map_clean()

The locking order between map locks and page "busy" locks
is that the page "busy" lock comes first, but uvm_map_clean()
breaks this rule by holding a map locked (as reader) while
waiting for page "busy" locks.

If another thread is in the page-fault path holding a page
"busy" lock while waiting for the map lock (as a reader)
and at the same time a third thread is blocked waiting for
the map lock as a writer (which blocks the page-fault thread),
then these three threads will all deadlock with each other.

Fix this by marking the map "busy" (to block any modifications)
and unlocking the map lock before possibly waiting for any
page "busy" locks.

Martin Pieuchot reported that the same problem existed in OpenBSD
he applied this fix there after several people tested it.

fixes PR 56952

(chs)

fix typq

(pgoyette)

move hppa and powerpc to OpenSSL-3.x

(christos)

Regen

(pgoyette)

Add a Lenovo mouse so we can set the ALWAYS_ON quirk.

(pgoyette)

indent: only null-terminate the buffers if necessary

The only case where a buffer is used as a C-style string is when looking
up a keyword.

No functional change.

(rillig)

fix powerpc

(christos)

make things link

(christos)

move arm to OpenSSL-3.x

(christos)

Bump MAXTSIZ from 64MB to 128MB also for o32.

This so that the rather large cc1 from gcc12 can be run.
OK'ed by simonb@

(he)

indent: shorten variable names for indenting comments

No functional change.

(rillig)

indent: fix handling of multiple block comments in a line

(rillig)

make arm build

(christos)

undo previous

(christos)

indent: in comments, keep a leading tab

This kind of comments is used for the CVS IDs at the top of files.

(rillig)

indent: fix vertical spacing after declarations

A comment is not supposed to change the state of the 'blank line after
declaration', but it did.  The initialization of saved_just_saw_decl was
wrong though since it tried to capture the previous value after it had
already been overwritten.

(rillig)

Update loadbsd source and distribution binary to version 3.3.
- Loading the kernel to the highest priority memory segment is default now.
- New option -l to revert the to the previous behaviour of largest segment.
- New option -M to define a minimum size for the memory segment.
- Fixed some warnings and typos.
- Put assembler inline source into its own source text startit.s.
- Can be built with Bebbo's gcc6 Amiga port or with vbcc.

(phx)

indent: reduce code for scanning tokens

The input line is guaranteed to end with '\n', so there's no need to
carry another pointer around.

No functional change.

(rillig)

try to fix the build.

(christos)

indent: remove foreign RCS IDs

(rillig)

indent: miscellaneous cleanups

(rillig)

openssl compat: Try this again!

compatdir appears to be for things like /usr/lib that have
/usr/lib/eabi as a subdirectory, so for /usr/lib/openssl it would
mean we have /usr/lib/openssl/eabi as a subdirectory.  But that's not
what we have; we have /usr/lib/eabi/openssl.

Let's try using NetBSD.dist.compat.in instead, which appears to be
intended for this purpose, from what I can gather.

(riastradh)

indent: reduce binary size

No functional change.

(rillig)

kern/sys_descrip.c: Nix trailing whitespace.

(riastradh)

Check for RB_HALT in cpu_reboot.

(skrll)

x86/efiboot: Nix trailing whitespace.

(riastradh)

x86/locore.S: Nix trailing whitespace.

(riastradh)

x86/boot(8): Nix trailing whitespace.

(riastradh)

options(4): Nix trailing whitespace.

(riastradh)

mark /usr/lib/openssl and /usr/lib/openssl/modules as "compatdir".

remove them as explicit part of all arm64 builds when they're only
(currently) used and consumed by llvm builds.

this fixes gcc builds and i don't think it breaks llvm ones though
i'm still waiting for that test.

(mrg)

switch alpha to OpenSSL-3.x

(christos)

lint: add query for chained assignments

(rillig)

openssl: More set list fixes.

(riastradh)

fix alpha

(christos)

indent: prevent undefined behavior on unbalanced parentheses

(rillig)

openssl: More tests need -Wno-missing-field-initializers.

(riastradh)

indent: do not add a blank at the beginning of a line

Most calls to output_line did already reset the variable.  There may be
some untested edge cases in or after comments, but these should be fine
as well.

(rillig)

indent: do not add a space before a comment that starts a line

(rillig)

indent: replace __dead functions with return statements

No functional change.

(rillig)

switch i386 and m68k to OpenSSL-3.x

(christos)

indent: implement 'blank after declarations'

(rillig)

openssl: libcrypto bf test needs -Wno-missing-field-initializers

(riastradh)

indent: use enum instead of magic numbers for tracking declarations

No functional change.

(rillig)

tests/indent: extend test for blank line after declaration

(rillig)

Note dedication

(martin)

Add dedication to Ryo SHIMIZU.

(martin)

indent: improve names of option variables

No functional change.

(rillig)

indent: add debug logging for enum token classification

(rillig)

fix m68k/i386

(christos)

indent: fix lexing of numbers that are spread over multiple lines

(rillig)

Pull up following revision(s) (requested by manu in ticket #159):

share/man/man8/man8.x86/boot.8: revision 1.27
sys/arch/i386/stand/efiboot/version: revision 1.3
share/man/man8/man8.x86/boot.8: revision 1.28
share/man/man8/man8.x86/boot.8: revision 1.29
sys/arch/i386/stand/lib/exec.c: revision 1.79
sys/arch/i386/stand/efiboot/efiboot.c: revision 1.13
sys/arch/i386/stand/efiboot/bootx64/efibootx64.c: revision 1.6
sys/arch/i386/stand/efiboot/bootia32/efibootia32.c: revision 1.6
sys/arch/i386/stand/efiboot/boot.c: revision 1.22
sys/arch/amd64/amd64/locore.S: revision 1.219
sys/arch/i386/stand/efiboot/bootia32/startprog32.S: revision 1.3
sys/arch/i386/stand/efiboot/efiboot.h: revision 1.12
sys/arch/amd64/conf/files.amd64: revision 1.121
sys/arch/amd64/conf/std.amd64: revision 1.13
share/man/man8/man8.x86/pxeboot.8: revision 1.6
sys/arch/i386/stand/efiboot/bootx64/startprog64.S: revision 1.4
sys/arch/amd64/amd64/locore.S: revision 1.220
share/man/man8/man8.x86/dosboot.8: revision 1.4
share/man/man4/options.4: revision 1.524

Add reloc keyworkd to let EFI bootstrap load amd64 kernel at any address

EFI bootstrap assumes it can copy the amd64 kernel to its ELF load
address (that is KERNTEXTOFF - KERNBASE = 0x200000), but it can
clash with previous UEFI memory allocation, as described here:
http://mail-index.netbsd.org/tech-kern/2023/04/07/msg028833.html

This change adds a reloc keyword for controling where the EFI
boostrap will copy the kernel image. Possible values are:
default - the default and prior behavior, copy at 0x200000.
none - do not copy and use the kernel image where it was loaded.
address - specify an explicit address where to copy the kernel.

This comes with an amd64 kernel patch that makes it self-relocatable.
It first discover where it was loaded in memory, and if this is
different than the expected 0x200000, hhe the kernel relocates
itself and start over at the right address.

Merge x86 boot options in x86/boot(8) and add undocumented UEFI options

We were supposed to keep the option list in x86/boot(8), x86/dosoot(8)
and x86/pxeboot(8) in sync, but it did not happen, hence it may work
better with all the options in x86/boot(8). Also add the undocumented
UEFI boot options.

Add a SELFRELOC kernel option for the sake of documentation clarity.
Instead of telling that x86/boot(8) reloc command needs a kernel able
to self relocate, we can tell it needs a kernel built with the
SELFRELOC option. This keeps the reader from wondering what could
make a kernel able to self relocate.

Remove XXX todo marker left by mistake

Raise the version for new feature (here reloc command)
Suggested by Masanobu SAITOH

(martin)

tests/indent: add test for lexing of numbers

(rillig)

distrib/sets: format tests list

(rillig)

aarch64: New directories for compat openssl on aarch64.

Hope I got the setlist tags right, not really sure how this works.

(riastradh)

libfido2: Needs -Wno-error=pointer-sign on arm.

(riastradh)

aq(4): AQC113-116 support by ryo@ has been pull-up'ed into netbsd-10.

(rin)

indent: rename struct fields for buffers

No binary change except for assertion line numbers.

(rillig)

Note tickets #1631 - #1633

(martin)

Pull up following revision(s) (requested by chs in ticket #1633):

sys/ufs/ffs/ffs_snapshot.c: revision 1.155

ffs: apply the remaining ffs_snapshot.c part of this FreeBSD commit:
  commit 364ed814e7285c8216d8a201d3ab3674eb34ce29
  Author: Kirk McKusick <mckusick@FreeBSD.org>
  Date:  Thu Dec 9 21:24:00 2004 +0000
    Fixes a bug that caused UFS2 filesystems bigger than 2TB to
    prematurely report that they were full and/or to panic the kernel
    with the message ``ffs_clusteralloc: allocated out of group''.
    Submitted by:  Henry Whincup <henry@jot.to>
    MFC after:      1 week

all the other changes in that commit were applied previously by others:
- sborrill commmitted ffs_alloc.c rev 1.123 in 2009
- simonb committed ffs_alloc.c rev 1.110 in 2008
- the ffs_clusteralloc() part is not needed because we no longer have
  that function.

fixes PR 57307

(martin)

Note tickets #139, #141, #157 - #165

(martin)

Pull up following revision(s) (requested by chs in ticket #165):

sys/ufs/ffs/ffs_snapshot.c: revision 1.155

ffs: apply the remaining ffs_snapshot.c part of this FreeBSD commit:
  commit 364ed814e7285c8216d8a201d3ab3674eb34ce29
  Author: Kirk McKusick <mckusick@FreeBSD.org>
  Date:  Thu Dec 9 21:24:00 2004 +0000
    Fixes a bug that caused UFS2 filesystems bigger than 2TB to
    prematurely report that they were full and/or to panic the kernel
    with the message ``ffs_clusteralloc: allocated out of group''.
    Submitted by:  Henry Whincup <henry@jot.to>
    MFC after:      1 week

all the other changes in that commit were applied previously by others:
- sborrill commmitted ffs_alloc.c rev 1.123 in 2009
- simonb committed ffs_alloc.c rev 1.110 in 2008
- the ffs_clusteralloc() part is not needed because we no longer have
  that function.

fixes PR 57307

(martin)

Pull up following revision(s) (requested by chs in ticket #164):

sbin/fsck_ffs/pass2.c: revision 1.52

Apply this commit from FreeBSD:
  commit 6bae6625e0e06816c80ac4971dfccf0643abe3f0
  Author: Kirk McKusick <mckusick@FreeBSD.org>
  Date:  Wed Aug 17 14:19:59 2022 -0700
    Improve handling of missing '.' and '..' in UFS directories.
    The UFS filesystem expects to find '.' and '..' as the first two entries
    in a directory. The kernel's UFS name cache can become quite confused
    when these two entries are not present as the first two entries.
    Prior to this change, when the fsck_ffs(8) utility detected that
    '.' and/or '..' were missing, it would report them, but only offered
    to replace them if the space at the beginning of the directory was
    available. Otherwise it was left to the system administrator to
    move the offending file(s) out of the way and then rerun fsck_ffs(8)
    to create the '.' and '..' entries.
    With this change, fsck_ffs(8) will always be able to create the '.'
    and/or '..' entries. It moves any files in the way elsewhere in the
    directory block. If there is no room in the directory block to which
    to move them, they are placed in the lost+found directory.
    Reported by:  Peter Holm
    Sponsored by: The FreeBSD Foundation

(martin)

Pull up following revision(s) (requested by chs in ticket #163):

sbin/fsck_ffs/inode.c: revision 1.75

Use %zu rather than %lu to print a size_t (should fix i386 build).

(martin)

Pull up following revision(s) (requested by chs in ticket #162):

sbin/fsck_ffs/inode.c: revision 1.74
sbin/fsck_ffs/utilities.c: revision 1.68
sbin/fsck_ffs/fsck.h: revision 1.57
sbin/fsck_ffs/pass4.c: revision 1.30

catch up with sign changes in the fs.h

(martin)

Pull up following revision(s) (requested by chs in ticket #161):

sbin/fsck_ffs/setup.c: revision 1.106
sbin/fsck_ffs/pass5.c: revision 1.57

ufs: more signed/unsigned fixes

Fix the previous signed/unsigned fixes to build on 32-bit,
including applying this commit from FreeBSD:

  commit 2d34afcd04207cf3fa3d5b7f467a890eae75da41
  Author: Kirk McKusick <mckusick@FreeBSD.org>
  Date:  Sun Oct 25 21:04:07 2020 +0000
    Use proper type (ino_t) for inode numbers to avoid improper sign extention
    in the Pass 5 checks. The manifestation was fsck_ffs exiting with this error:
      ** Phase 5 - Check Cyl groups
      fsck_ffs: inoinfo: inumber 18446744071562087424 out of range
    The error only manifests itself for filesystems bigger than about 100Tb.
    Reported by:  Nikita Grechikhin <ngrechikhin at yandex.ru>
    MFC after:    2 weeks
    Sponsored by: Netflix

(martin)

Pull up following revision(s) (requested by chs in ticket #160):

usr.sbin/makefs/ffs/ffs_alloc.c: revision 1.31
sbin/tunefs/tunefs.c: revision 1.58
sbin/fsck_ffs/setup.c: revision 1.105
sbin/fsck_ffs/pass5.c: revision 1.56
usr.sbin/makefs/ffs.c: revision 1.74
usr.sbin/makefs/ffs/mkfs.c: revision 1.42
usr.sbin/makefs/Makefile: revision 1.40
sys/ufs/ffs/fs.h: revision 1.71
sbin/fsdb/fsdb.c: revision 1.54
sbin/resize_ffs/resize_ffs.c: revision 1.58
sbin/fsck_ffs/pass4.c: revision 1.29
usr.sbin/makefs/ffs/ffs_extern.h: revision 1.9
sbin/newfs/mkfs.c: revision 1.133
sys/ufs/ffs/ffs_alloc.c: revision 1.172
sbin/fsck_ffs/pass1b.c: revision 1.24
usr.sbin/dumpfs/dumpfs.c: revision 1.68
sys/ufs/ffs/ffs_extern.h: revision 1.88
usr.sbin/quotacheck/quotacheck.c: revision 1.51
sys/ufs/ffs/ffs_subr.c: revision 1.54
sbin/fsck_ffs/main.c: revision 1.91
sbin/fsck_ffs/pass1.c: revision 1.63

ufs: fixed signed/unsigned bugs affecting large file systems

Apply these commits from FreeBSD:
  commit e870d1e6f97cc73308c11c40684b775bcfa906a2
  Author: Kirk McKusick <mckusick@FreeBSD.org>
  Date:  Wed Feb 10 20:10:35 2010 +0000
    This fix corrects a problem in the file system that treats large
    inode numbers as negative rather than unsigned. For a default
    (16K block) file system, this bug began to show up at a file system
    size above about 16Tb.
    To fully handle this problem, newfs must be updated to ensure that
    it will never create a filesystem with more than 2^32 inodes. That
    patch will be forthcoming soon.
    Reported by: Scott Burns, John Kilburg, Bruce Evans
    Followup by: Jeff Roberson
    PR:          133980
    MFC after:  2 weeks

  commit 81479e688b0f643ffacd3f335b4b4bba460b769d
  Author: Kirk McKusick <mckusick@FreeBSD.org>
  Date:  Thu Feb 11 18:14:53 2010 +0000
    One last pass to get all the unsigned comparisons correct.

In additional to the changes from FreeBSD, this commit includes quite a few
related changes to appease -Wsign-compare.

(martin)