Copy stable/10@r272459 to releng/10.1 as part of
the 10.1-RELEASE process.

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

MFC: Merge in the changes in -current:

Support ! operator in "files" files.
Improve error detection and reporting
Cleanup code to make it easier to maintain.
Remove mandatory keyword: it has been used for 17 years.
Bump version number (we should have bumped for -I too, but didn't)

r261501 | imp | 2014-02-04 17:26:11 -0700 (Tue, 04 Feb 2014) | 5 lines
Fix ! by not clearing not at the bottom of the loop.
Add a blank line
Submitted by: bde (blank line)

r261493 | imp | 2014-02-04 11:28:58 -0700 (Tue, 04 Feb 2014) | 5 lines
Implement the '!' operator for files* files. It means 'include this
only if the specified option is NOT specified.' Bump version because
old config won't be able to cope with files* files that have this
construct in them.

r261446 | imp | 2014-02-03 12:14:36 -0700 (Mon, 03 Feb 2014) | 5 lines
Convert the loop by gotos into a for loop to improve readability. I
did this only with the inner loop for the token parsing, and not the
outer loop which was understandable enough when the extra layers of
looping went away...

r261445 | imp | 2014-02-03 12:10:33 -0700 (Mon, 03 Feb 2014) | 4 lines
Fix a bug introduced in r261437 that failed to honor "optional
profiling-routine" to work, since profiling-routine is not really an
option or a device, but a special case elsewhere in the code.

r261444 | imp | 2014-02-03 11:56:41 -0700 (Mon, 03 Feb 2014) | 2 lines
Slight cleanup to the error messaging to compress code vertically...

r261442 | imp | 2014-02-03 11:31:51 -0700 (Mon, 03 Feb 2014) | 2 lines
Better error messages when EOF is hit in the middle of a phrase.

r261438 | imp | 2014-02-03 09:54:53 -0700 (Mon, 03 Feb 2014) | 5 lines
Move the check for standard keyword + optional inclusion specifier to
its proper location. Otherwise you could have 'file.c standard pci'
without an error. This construct isn't in our tree, and has no well
defined meaning.

r261437 | imp | 2014-02-03 09:47:10 -0700 (Mon, 03 Feb 2014) | 4 lines
Don't believe we have a requirement until after we've checked all the
known key words. This will make error messages slightly better in
weird corner cases, but should otherwise be a nop.

r261436 | imp | 2014-02-03 09:46:01 -0700 (Mon, 03 Feb 2014) | 3 lines
In the 17 years since r30796, the mandatory keyword has never been used
in any files as far as I can tell, and is currently unused. Retire it.

r261435 | imp | 2014-02-03 08:10:44 -0700 (Mon, 03 Feb 2014) | 6 lines
Slightly deobfuscate read_file() and likely pessimize the runtime
performance by epsilon.
(Translation: elminate bogus macros that hid 'returns' making it hard
to read and moved a block of code inline rather than at the end of the
fuction where it was effectively a 'gosub' kind of goto).

MFC r263429

Fix kern/187712: config(8) does not respect KERNCONFDIR.
The impact of this bug is that you cannot build a kernel if both of the
following are true:
1) The kernel config file is in a non-default location
2) The kernel config file uses the "include" statement from config(5).

usr.sbin/config/main.c
usr.sbin/config/config.8
usr.sbin/config/config.h
usr.sbin/config/lang.l
Added a "-I path" option to config(8). By analogy to cc(1), it adds
an extra path in which the "include" statement will search for
files.

Makefile.inc1
Pass "-I ${KERNCONFDIR}" to config(8).

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

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

Work around build breakages with GCC 4.2.

Reported by: tinderbox

Improve compatibility with recent flex from flex.sourceforge.net.

Make the internal assertion correct--only fail when '\0' is found in
places other than the end of the test section. Otherwise, with kernel
compiled with Clang which happens to be setting ELF section alignment
differently config(8) was throwing assert() failure unnecessarily

Reported by: Kimmo Paasiala <kpaasial (at) gmail.com>
Tested by: Kimmo Paasiala <kpaasial (at) gmail.com>
MFC after: 10 days

Add the Clang specific -Wmissing-variable-declarations to WARNS=6.

This compiler flag enforces that that people either mark variables
static or use an external declarations for the variable, similar to how
-Wmissing-prototypes works for functions.

Due to the fact that Yacc/Lex generate code that cannot trivially be
changed to not warn because of this (lots of yy* variables), add a
NO_WMISSING_VARIABLE_DECLARATIONS that can be used to turn off this
specific compiler warning.

Announced on: toolchain@

Loosen restrictions for quoted strings. Now we can use more complex strings
and "escaped" quote characters.

MFC after: 1 month

- Fix the error message when a dependency string is not provided to
reference a missing dependency rather than a missing compile command.
- Don't append a newline to the auto-generated compile command. The
compile command has a newline appended when it is later output to the
Makefile.

MFC after: 2 weeks

Fix world after byacc import:
- old yacc(1) use to magicially append stdlib.h, while new one don't
- new yacc(1) do declare yyparse by itself, fix redundant declaration of
'yyparse'

Approved by: des (mentor)

Remove trailing whitespace per mdoc lint warning

Disussed with: gavin
No objection from: doc
Approved by: joel
MFC after: 3 days

Fix copyright year

Spotted by: pluknet

fgets(3) returns a pointer, so compare against NULL, not integer 0.

Replace index() and rindex() calls with strchr() and strrchr().

The index() and rindex() functions were marked LEGACY in the 2001
revision of POSIX and were subsequently removed from the 2008 revision.
The strchr() and strrchr() functions are part of the C standard.

This makes the source code a lot more consistent, as most of these C
files also call into other str*() routines. In fact, about a dozen
already perform strchr() calls.

Generate ${NORMAL_CTFCONVERT} invocation without '@' modifier:
- ${NORMAL_CC} is also invoked without '@'
- Userland CTF support was changed previously to echo ctfconvert invocations too

The generated Makefile for the kernel was not running ctfconvert on
object files corresponding to source files that had the compile-with
option set in conf/files. This means that any fbt probes for functions
in that object file would not have correct argument types.

The fix is to run ctfconvert on any target file that does not have the
no-obj option set in files.

PR: bin/160275
Reported by: Paul Ambrose (ambrosehua AT gmail DOT com)
MFC after: 1 week

Improve portability of config(8).

- Use strlen(dp->d_name) instead of the unportable dp->d_namlen. Rename
i to len to make it slightly more descriptive and prevent negative
indexing of the array.
- Replace index() by strchr().

This supposedly fixes compilation on GNU systems.

Submitted by: Robert Millan <rmh debian org> (original patch)
MFC after: 3 weeks

- Merge changes to the base system to support OFED. These include
a wider arg2 for sysctl, updates to vlan code, IFT_INFINIBAND,
and other miscellaneous small features.

Generate MACHINE= and MACHINE_ARCH= lines based on the machine
directive. Once this is MFC'd, we can move these out of the template
files where they are (incosnsitently) defined.

MFC after: 1 week

Remove the advertising clause from UCB copyrighted files in usr.sbin. This
is in accordance with the information provided at
ftp://ftp.cs.berkeley.edu/pub/4bsd/README.Impt.License.Change

Move most of the remaining USD/PSD/SMM papers into share/doc

Don't generate input() since it's not used.

Remove references to MAKEDEV(8)

Fix typos.

PR: bin/148894
Submitted by: olgeni

Give the user a hint as to what config(8) is unhappy with.

Remove support for creating the 'machine' symlink as well as creating the
extra MACHINE_ARCH symlink for certain platforms (such as pc98). The
support for creating these symlinks was added to sys/conf/kern.post.mk in
changeset 152964. The intention of that commit was to remove this code
from config(8), but config(8) was never updated.

Approved by: imp

mdoc: drop redundant .Pp calls, kill EOL whitespace

Allow a kernel config to specify a set but empty value via
'makeoptions OPTION=' for consistency with the make commandline.

Previously 'makeoptions WERROR=' would result in a syntax error; now
it produces the same effect as 'makeoptions WERROR'. Both forms now
result in 'WERROR=' in the generated Makefile.

Given the lag between introducing the options-compat and fixing the
bug in the config file was long, and merged to stable, we have to bump
the version here so that stable/8 kernels can have their required
version number bumped to make sure the right config is used.

MFC after: 3 days

Turns out that it is a bad idea to have a missing compat option be a
fatal condition. While it works out really well for diagnosing the
case where you want it, but don't have it, it works really badly for
the case where you don't have it and don't want it. Remove the printf
and exit pair. Replate it with simple return to silently ignore this
condition.

This is needed to fix the COMPAT_IA32 being required in options.* in
stable, but we need to run this change through -current first...

MFC after: 3 days

Put warnings out to stderr rather than stdout.

MFC after: 3 days

Enhance config to handle MACHINEs with multiple architectures:

- Passing -m to config will now print the MACHINE and MACHINE_ARCH
given in the passed kernel configuration file and then exit.
- If an option is defined in options.MACHINE with the same name as the
architecture of the kernel being configured, that option will be
considered set. This allows conditional compilation based on CPU
architecture.

Config version is now 600010.

Reviewed by: imp

style(9) fixes:

o make cmd scoped to the whole do_rules function, since it really is
scoped to the whole fucnion. Making it static was the wrong way to
fix referencing it outside of the block in which it was declared
(and conforms to the style of the rest of the file).
o remove a couple of meaningless blank lines
o properly wrap one line.

sparc64, and possibly other architectures, pads the length of the
section holding the config file to sh_addralign bytes using NULs.
This bogusly triggers an assert. Break out of the loop when we hit an
NUL within that many bytes of the end.

MFC after: 3 days

Require the option that's mapped be listed in the options file. This
will allow people with old config options to either have it just work
(if config is new enough), or get a version error (if their config is
about 7.0 or newer) rather than getting a cryptic error about
duplicated options in the options file, or getting an error about an
unknown option, at which point they'd update their config file only to
learn they need a new config, only to learn they didn't really need to
update their config file... All this because our version checking was
in the wrong place for the past decade...

# hopefully this is the last change, and we'll be able to config with an
# 8.0 GENERIC file on stable/8 after I merge this change and add the
# compat options.

MFC after: 3 days

Redo how we add compat options so as to be compatible with old
versions of config. Remove support for the syntax OLD = NEW form the
options file, and instead have a new file $S/conf/options-compat.
This file will be parsed as OLD NEW on each line. Bump version of
config. Since nothing in -current ever used this, there's no hazards
for current users, so I'm not bumping the version in the
Makefiles.$MACHINE. No need, really, for this version bump in
-current, but this was introduced into -stable before I realized the
version check was ineffective there, so the verison bump doesn't hurt
here and keeps the two branches in sync, versionwise, after the MFC.

MFC after: 3 days

Move checking the version up from Makefile generation to just after
we've parsed the config file. Makefile generation is too late if
we've introduce changes to the syntax of the metafiles to warn about
version skew, since we have to try to parse them and we get an parse
error that's rather baffling to the user rather than a 'your config is
too old, upgrade' which we should get.

We have to defer doing it until after we've read the user's config
file because we define machinename there. The version required to
compile the kernel is encoded in Makefile.machinename. There's no
real reason for this to be the case, but changing it now would
introduce some logistical issues that I'd rather avoid for the moment.
I intend to revisit this if we're still using config in FreeBSD 10.

This also means that we cannot introduce any config metafile changes
that result in a syntax error or other error for the user until 9.0 is
released. Otherwise, we break the upgrade path, or at least reduce
the usefulness of the error messages we generate.

# This implies that the config file option mapping will need to be redone.

MFC after: 3 days

Revert r206179 (by imp) and do something similar which is more consistent
with all other corresponding CTF places by changing the corresponding
code which is generated by config(8). Or in short, move the '@' from
the variable definition to the use of the variable. [1]

While I'm here break up a long line. [2]

Discussed with: imp [1,2], bde [2]

Bump minor version of config to reflect the new option remapping
feature. The kernel makefiles have specifically not been bumped
because nothing uses this new feature and doing so forces everybody to
recompile for no good reason. This chnage will be MFC'd where the
kernel version numbers for amd64 and ia64 will be bumped, since those
are the only two that have use the option remapping feature. Once
merged, this will give a better error message to folks that are using
buildkernel without buildworld or kernel-toolchain to update their
kernels.

MFC after: 3 days

Allow option aliasing. Lines of the form:

OLD_OPT = NEW_OPT

in options* files will now map OLD_OPT to NEW_OPT with a friendly
message. This is indented for situations where we need to preserve an
interface in the config file in an upwards compatible fashion on a
stable branch.

Reviewed by: nwhitehorn@
MFC after: 3 days

- Handle calloc() allocation failures.
- Fixed a comment.
- 2 -> EXIT_FAILURE in some places.
- errx() -> err() where appropriate.

PR: 144644
Submitted by: Garrett Cooper

The last big commit: let usr.sbin/ use WARNS=6 by default.

ANSIfy almost all applications that use WARNS=6.

I was considering committing all these patches one by one, but as
discussed with brooks@, there is no need to do this. If we ever
need/want to merge these changes back, it is still possible to do this
per application.

Move the comment to it's correct place.

Remove leftover of alpha support for config(8)--we have MAP_FAILED
globally defined.

Make config -x <kernel> only return non-zero characters,
so that:

config -x <kernel> | grep <something>

just works.

Reported by: Danny Braniss <danny@cs.huji.ac.il>

Delete commented out ancient history.

Allow multiple makeoption lines to be used with the += operator, this permits
the following syntax in the kernel config.

makeoptions MODULES_OVERRIDE=foo
makeoptions MODULES_OVERRIDE+=bar
makeoptions MODULES_OVERRIDE+=baz

Bump config minor version to 600007.

Allow kernel config files to include files other than those in the CWD,
using unquoted paths.

Revice the way the CTF conversion is done per object. Avoid creating a second
shell (which was the problem with the original implementation) and avoid
letting make see an empty definition (which was the problem with the current
implementation).

Change the CTF conversion makefile code to use a new line to avoid
spawning another shell.

Requested by: Ed Schouten

M config/mkmakefile.c

Add the CTF conversion to the generated makefile. In the case where
NO_CTF or !WITH_CTF, the macro is empty.

Fix a bug introduced by DEFAULTS feature. When the config file
doesn't exist, we make a directory and then say "oops, that file isn't
there" leaving the directory behind. Add a stat for the config file
so that we detect this before making the directory. This is
semi-lame, but less lame than having this bug.

Honour the logical current working directory ($PWD) when using config's
-d destdir option. For an automounted src tree using the logical cwd
in the Makefile keeps amd(8)'s mount timeout refreshed. Code to check
$PWD's validity cribbed from pwd(1).

Discussed on hackers@.

Spring forward into the late 1970's by following item 5 in the yacc
manual's Appendix D ("Old Features Supported but not Encouraged") in
the Seventh Edition Unix Programmer's Manual (January, 1979) by
retiring the " = {" method of of action specification in favor of a
plain "{". It is no longer necessary for this bootstrap program to
be compatible with 6th Edition systems. Some yaccs in the wild do not
support this old syntax any more, and compatibility with those systems
is more important these days (as there are easily 7 orders of magnitude
more of them than real v6 systems today).

Reviewed by: jhb@ and dds@ (the latter gave the reference).

Fixed "make checkdpadd" (missing library dependencies).

Approved by: re (kensmith)

Fix some problems that affect multiple file inclusion. Bruce found
this bug and submitted these patches to dunstan@. He sent them to me
to test, and I discovered they were needed for the atmel kernel config
files. Since we were playing with them in the terminal room after the
developer's summit today, I thought I'd go ahead and commit them to
allow those folks that now have atmel hardware (thanks Andre) a chance
to try it out w/o my help. Since dunstan@ is asleep right now, risk
stepping on his toes a little by going ahead and committing this
change.

Submitted by: dunstan@, bde@
Tested by: bde@

Handle !INCLUDE_CONFIG_FILE entirely in the kernel. This should make some
developers happy, since it will let them to use old config(8) with newer
kernels.

Reviewed by: imp
Approved by: imp

Bump config(8) version and build requirement for config(8) to 600006. This
is caused by my latest changes to config(8). You're supposed to install new
config(8) in order to prevent yourself from seeing a warning about old
version of that tool.

You should configure the kernel with a new config(8) then.

Oked by: rwatson, cognet (mentor)

This is loser's commit message:

Remember about tricky cases, where options contain unfriendly characters,
from the ANSI-C string point of view ('"' in this case). The x09 build
breakage was caused by SC_CUT_SEPCHARS options.

I did test this patch number of times; each time unprofessionally and
inappropriately.

OKed by: cognet (mentor)

Add ${.CURDIR} before kernconf.tmpl in order to let the build see newly
added kernconft.tmpl file.

Approved by: cognet (mentor)

Improve INCLUDE_CONFIG_FILE support.

This change will let us to have full configuration of a running kernel
available in sysctl:

sysctl -b kern.conftxt

The same configuration is also contained within the kernel image. It can be
obtained with:

config -x <kernelfile>

Current functionality lets you to quickly recover kernel configuration, by
simply redirecting output from commands presented above and starting kernel
build procedure. "include" statements are also honored, which means options
and devices from included files are also included.

Please note that comments from configuration files are not preserved by
default. In order to preserve them, you can use -C flag for config(8). This
will bring configuration file and included files literally; however,
redirection to a file no longer works directly.

This commit was followed by discussion, that took place on freebsd-current@.
For more details, look here:

http://lists.freebsd.org/pipermail/freebsd-current/2007-March/069994.html
http://lists.freebsd.org/pipermail/freebsd-current/2007-May/071844.html

Development of this patch took place in Perforce, hierarchy:

//depot/user/wkoszek/wkoszek_kconftxt/

Support from: freebsd-current@ (links above)
Reviewed by: imp@
Approved by: imp@

Don't leak files.
This also eliminates the need for ifp init.

Submitted by: ru@

Document multiple hints lines.
Tweak machine description to more accurately reflect second argument being
the cpu architecture. Note correspondence with MACHINE and MACHINE_ARCH.

End my resistance to jmg's multiple hints files and bring in support
for having multiple hints files generate a correct hints.c (eg, with
all the specified ones catenated together).

style(9) pedantry

Since multiple hints lines do not work, make doing that a fatal error.

Comment out the first paragraph; while it tells the true (we're using
the old BSD config(8) utility in FreeBSD), it does not look good.

PR: docs/100328

Added rudimentary support for the "include" directive (inside "files").
This will be used to split sys/conf/files into multiple files similar
to how this is done in NetBSD.

Remove unused variables.

Remove an unused variable.

Use calloc() instead of zeroing the memory our own.

Document the two argument form of the "machine" directive.

Reviewed by: imp
MFC After: 3 days

Allow newer config versions to config older versions with the same
major number.

Reviewed by: ru@, jhb@, arch@ (a few months ago)

# this is subject to refinement based on experience.

- Change the exit message from "make cleandepend; make depend" to
"make cleandepend && make depend".

PR: bin/91765
Submitted by: Soeren Straarup <xride@x12.dk>
Approved by: cpercival
MFC after: 3 days

- Avoid adding devices multiple times to the device list.
- Avoid adding options multiple times to the option list.

Based on a patch by: Matt Emmerton <matt@gsicomp.on.ca>

Clean up most of the "XXX"-tagged items:

- The code that creates hints.c and env.c from the skeleton files
moved into separate functions.

- Sanity checks for missing "ident" and "cputype" directives moved
into main(), alongside the existing check for "machine".

PR: bin/90310
Submitted by: Matt Emmerton <matt@gsicomp.on.ca>

config.5 was repocopied from share/man/man5/ to usr.sbin/config/.

Repocopied by: peter

Document "makeoption", "nocpu", and "nomakeoptions" directives.
(Indeed this should be repo-copied to src/usr.sbin/config/.)

Add "makeoption" as an alias to "makeoptions", for symmetry.

Style: use S_ISDIR() (submitted by bde@) and eq() where appropriate.

The DEFAULTS changes caused the user specified config file to be opened
much later than before, and it is now after we do a mkdir ../compile/FILE.
As a result, if you do 'config DOESNOTEXIST', it now creates the directory
../config/DOESNOTEXIST. It did not do that before. If DEFAULTS does not
exist, it still fails early before any permanent changes.

This shameless hack restores the old behavior of ensuring the config file
actually exists before mkdiring its counterpart directory.

Now I can rmdir ../compile/D and it will stay dead, after my fingers keep
sabotaging me with 'config D<tab><enter>'. (Some of my kernel names
started with D, which used to be 1-character unique and my fingers knew
this very well...)

600004 is a better new version than 700000 based on some future commits to
this file. With ru@'s approval, change it to this version. In this case we
had to bump the version because the old parser would choke on | in the new
'or' syntax and consider that a device.

Approved by: ru@

- Allow duplicate "machine" directives with the same arguments.
- Move existing "machine" directives to DEFAULTS.

Make config(8) understand ORed dependecies in "files*" and
improve tracking of known devices. Bump config(8) version.

Clean some code that became obfuscated over the years:
Don't keep duplicate files in the files list just to
mark the device as "known" later. XXX: Since the
device list isn't unique (there can be two "device foo"
directives, as this the case with LINT+DEFAULTS), we
have to traverse it all to mark all copies of the same
device as "used", but this is not worse than it was.

Minor comment tweak to prevent gcc from being upset about the substring
/* appearing in a comment.

Add some rationale about when to bump and not bump the config version.
Clarify that it is not like the shlib versions, and not like param.h's
__FreeBSD_version/osreldate either.

When config(8) was actively changing a while back, the interface between
config and the build system (eg: /sys/conf/files.* and Makefile.*) was
changing rapidly. configvers is a version number of that interface.

User specified config files do not have a version number. The decision
about whether a user supplied config file is syntactically valid or not
belongs to the parser and sanity checks, not an arbitary number.

since nocpu isn't used in the kernel config base, we don't need to
bump the version. Peter Wemm, John Baldwin and I hammered this out
after the last time I needlessly incremented the version.

Use the eq() macro for comparing strings (style), and remove a "break"
statement to null the effect of several identical "cpu" directives.

By a popular demand, add "nomakeoptions" as an alias to "nomakeoption".

Implement the "nocpu" directive.

Requested by: rwatson

Bump config(8) version for the DEFAULTS change.

Optionally include a DEFAULTS config file if it is present in the current
directory before the specified config file. This is implemented by
opening DEFAULTS as stdin if it exists, and if so resetting stdin to the
actual config file when DEFAULTS is fully parsed via yywrap(). In short,
this lets us create DEFAULTS kernel configs in /sys/<arch>/conf that can
enable certain options or devices by default and allow users to disable
them via 'nooptions' or 'nodevice' rather than having to create kludge
NO_FOO options.

Requested by: scottl
Reviewed by: scottl

Put rev. 1.65 into usage().

Add an option to tell what version of config(8) this is.

Ask that "make cleandepend" be run before "make depend", now that
we don't do this automatically.

Suggested by: bde

Allow one to specify a second parameter to the machine line. This
allows us to specify the machine_arch as well as machine. If
specified then a second link will be made, similar to machine, from
$MACHINE_ARCH to $S/$MACHINE_ARCH/include.

This is for ports where MACHINE != MACHINE_ARCH (pc98 today, others in
the future?).

Reviewed by: arch@, nyan@

style(9) nits

Unstaticize config. It was being optimized away.

MFC after: 3 days

Bump the version number for the addition of devices / nodevices.

Fully document (no)?(device|option)s?.

This page should probably be repocopied to src/usr.sbin/config/.

"device" and "nodevice" lines can actually specify more than one device
(separated by commas), so add "devices" and "nodevices" as aliases.

MFC after: 2 weeks

Sort sections.

The ioconf.c is no longer generated.

Submitted by: Craig Rodrigues

Allow multiple devices to be specified on one device / nodevice line.
Also allow "device" / "nodevice" to be spelled "devices" / "nodevices".

MFC after: 2 weeks

If the file specified in an "include" line does not exist in the current
directory, and its name does not begin with a period or a forward slash,
go look for it in ../../conf.

Wished for by: scottl
MFC after: 2 weeks

We accept both "option" and "options" so also accept both "nooption"
and "nooptions".

Kill count device support from config. I've changed the last few
remaining consumers to have the count passed as an option. This is
i4b, pc98/wdc, and coda.

Bump configvers.h from 500013 to 600000.

Remove heuristics that tried to parse "device ed5" as 5 units of the ed
device. This broke things like the snd_emu10k1 device, which required
quotes to make it parse right. The no-longer-needed quotes have been
removed from NOTES, GENERIC etc. eg, I've removed the quotes from:
device snd_maestro
device "snd_maestro3"
device snd_mss

I believe everything will still compile and work after this.

Per letter dated July 22, 1999, delete clause 3 from code directly
from Berkeley.

Static device counts will not be supported in 6.x so place the generated
macros (N<driver>) under BURN_BRIDGES.

Discussed with: peter, imp, scottl, ...

Assorted markup, spelling, and grammar fixes.

Document the "files" directive.

Reminded by: jmg

Fix a few glitches in my previous commit.
This makes config(8) WARNS?=6 compliant.

Add a new "files" directive, which allows to include a files.foo file directly
from a kernel config file.
Bump config version to reflect this change.

Whitespace nit.

Fixed spurious syntax errors for including files that don't begin with
a SEMICOLON token (a newline or semicolon, or one of these preceded
by a comment and/or whitespace). The input stream was switched too
early and the parser was expecting a SEMICOLON in the included file
instead of after the filename in the include directive.

Submitted by: Stefan Farfeleder <stefan@fafoe.narf.at>
Kept alive by: Adam C. Migus <adam@migus.org>

Add a reference to config(5) in the SEE ALSO section.

Suggested by: dcs
Approved by: ceri (mentor)

Eliminate non-existent word.

Submitted by: jwd

"towards" -> "toward". According to dictionary.com, the use of "towards"
is common in British English, while "toward" is the preferred form in
American English. Use the American form for consistency.

Correct the date on the manual page.

Submitted by: Tom Rhodes <trhodes@freebsd.org>,
underway@comcast.net (Gary W. Swearingen)

Disallow multiple 'machine' directives in a kernel configuration
file.

Reviewed by: ru, bde

New section 5 manual page detailing our kernel configuration file
format.

Reviewed by: Ruslan Ermilov <ru@freebsd.org>, Jens
Schweikhardt <schweikh@schweikhardt.net>

Not particularly pretty hack to generate rules to make .ln files
from .c files. Actually, this is overkill, as the .ln file targets
are assumed from .? (any) files. This is not a problem in practice,
merely a bit untidy, as the linting rules DTRT. See the sys/conf/*
and sys/mk/* files for usage.

Check the return values of opendir() and unlink() in cleanheaders().
If unlink() fails, just print a warning for now.

Best we can do is WARNS=1 due to lex.

Do not refer to the non-existant BDECFLAGS.

Don't convert the kernel ident to uppercase when writing the Makefile.

Grammar fix

PR: 51164
Submitted by: Jeff Ito <jeffi@rcn.com>
MFC after: 3 days

Bump the config version to force people to upgrade their config(8)
so the fix for emitting multiple instances of .o files will prevent
link errors on LINT.

Avoid emitting duplicate makefile entries.

Print FYI messages on stderr. Previously, they were printed on stdout,
and due to buffering they would sometimes come out after the actual
error message when mkheaders() failed due to an unknown device, so you'd
get an error messages followed by 20 or 30 lines of harmless warnings.

There are lots of other warning messages in config(8) that are printed
on stdout, but these were the most egregious (at least with LINT).

Implemented "nooption" and "nomakeoption" config(8) tokens.
Fixed memory leak in the "nodevice" option implementation.

Use these instead of sed(1) in MD NOTES.

Use a single makefile (sys/conf/makeLINT.mk) to generate
LINT for all architectures. (Previous versions missed
the LINT dependency on Makefile, and i386 version also
missed the dependency on ${NOTES}.)

Fixed bugs in the previous NOTES conversion using the
"nodevice" token and sed(1):

- i386 LINT lost "device pst".

- pc98 LINT lost SC_*, MAXCONS and KBD_DISABLE_KEYMAP_LOAD
options, and got needless DPT_* options.

- Added nooptions PPC_DEBUG, PPC_PROBE_CHIPSET, KBD_INSTALL_CDEV
to sparc64 LINT so that it has a chance to config(8).

This basically returns us to where we were before.

Allow for boolean make options (``makeoptions NO_MODULES'').
Fixed the potential bug in rmopt().

Bring the suggested command for extracting a configuration file from the
kernel into line with that suggested in LINT.

PR: bin/48157
Submitted by: naddy
MFC After: 4.8

Implemented a simple "nodevice" config(8) command that cancels
the effect of the "device" command, and use it to generate the
OLDCARD from GENERIC.

Suggested by: bde

Convert to using <sys/queue.h> macros.

Remove this Makefile, these docs are built by the Makefiles in src/share/doc/smm.

Uniformly refer to a file system as "file system".

Approved by: re

mdoc(7) police: Removed redundant .Ns calls.

Remove stray function prototype. I wrote and removed this function but
forgot the prototype.

Submitted by: dd

Check for missing static unit 'count' declarations as well.

Check that we are not supplying 'device foo N' to devices that do not take
a statuc unit count.

The .Nm utility

Moved the setting of all profiling-related variables except the key one
(PROFLEVEL) to kern.pre.mk so that it is easier to manage. Bumped config
version to match.

Moved the check for cputype being configured to a less bogus place in
mkmakefile.c.

We don't need to hit the user over the head so strong now that we've
fixed all the known warnings suffiently to not require NO_WERROR.

Ok'ed by: peter

Clean up hex() and octal() to return and work with unsigned integers since
they scan values of unsigned types, and since they do not need otherwise,
have them take const char * arguments.

Consistancy check: s/file system/filesystem/g

Add a hint about -DNO_WERROR

Commit some infrastructure for turning on -Werror for kernel compiles.
It doesn't actually do it yet though. This adds a flag to config so
that we can exclude certain vendor files from this even when the rest
of the kernel has it on. make -DNO_WERROR would also bypass all of it.

Bump configvers for different handling of maxusers.

Allow maxusers to be specified as 0 in the kernel config, which will
cause the system to auto-size to between 32 and 512 depending on the
amount of memory.

MFC after: 1 week

mdoc(7) police:

- Bump document date.
- Remove hard sentence breaks.
- Fix markup.

Use includes to get prototypes for hints and env arrays.

Submitted by: bde

Enable hardwiring of things like tunables from embedded enironments
that do not start from loader(8).

mdoc(7) police: s/BSD/.Bx/ where appropriate.

Perform a major cleanup of the usr.sbin Makefiles.
These are not perfectly in agreement with each other style-wise, but they
are orders of orders of magnitude more consistent style-wise than before.

Remove whitespace at EOL.

Put on my peril-sensitive sunglasses and remove the POLA-violating
stealth hints loading. 'make release' has been fixed to not need this
now anyway. If you want static hints, specify it explicitly.

Hey! Why did it suddenly get so dark??

Remove the old machine symlink first.

Introduce an "include" directive. It takes one argument, a filename
to be included into this one. This works the same way as #include
does in C; as far as the user is concerned, the included file is
inlined into the current one.

Since config(8) is no longer limited to working on one user-supplied
file, printing just a line number in an error message is not
sufficient. The new global variable yyfile represents the file
currently being parsed, and must be printed as well.

Reviewed by: imp
Obtained from: OpenBSD

mdoc(7) police: removed HISTORY info from the .Os call.

Bring FILES section up to date.

submitted by: Mark Peek <mark-ml@whistle.com>

Bump config version in config.

Move kernel compile directory from sys/compile/FOO to
sys/compile/${MACHINE}/FOO.

Reviewed by: obrien, peter and the USENIX terminal room secret kernel cabal

- Backout botched attempt to introduce MANSECT feature.
- MAN[1-9] -> MAN.

Set the default manual section for usr.sbin/ to 8.

newbus_ioconf() is long gone

Some more tidying up. we dont use config-dependent anyware. Eliminate
some duplicate code (cut/paste bug?). tidy up some other minor stuff.

s/special/compilewith/ (so I stop confusing myself) and GC an unused
function I missed before.

Untangle some special magic that happened for conflicting defintions for
local files.* and options.* files on the third pass.

Remove some more dead code. :-/

GC some leftover stuff (device-driver suffix)

Move the 'dont forget "make depend"' to be the last thing that people see,
after the warnings.

Futher cleanups. Since we have two options lists, one for proper options
and one for Makefile options, pass in the list head and use a common
newopt() routine.
Fix the 'config vmunix' support glue which was broken for a few minutes.

Collect together a handful of copies of the option generator code into a
single newopt(char *name, char *value) function. Change newdev() to
do the same thing rather than depending on the evil 'cur' device hack.

${BDECFLAGS} work. And fix a real error in the process. A "MAXUSERS"
string could have been passed to free(); There are some warnings here
I am not sure how to fix as they are in the lex scanner code, etc.

Devices are NOT compiled with the "special c2 option -i"..

Remove the need to list each and every cpu platform. Config will now
take your word for the 'machine' switch.

Try and make it clearer that the static units thing is a FYI only.

Unwind a bit more cruft - we only have one type of device now.

Oops. An old version of a local change leaked in with the last commit.
Since it is here, clean it up a bit.

Redo the stray header file cleanup code to not depend on timestamps
or access times or anything. Just bite the bullet and keep a list of
header files that we know about.

AARGH! This does not mix too well with the -o noatime mount option.
:-(

I confess. This is a truely nasty hack. I am so ashamed!
(I think config(8) source does bad things to your brain :-)

Clean up likely stray *.h files in the build directory.
Eg: if isa.h ceases being generated, zap it.
The heuristics to figure out a 'likely' file are pretty revolting.

Add support for configuring PowerPC kernels.

Reviewed by: peter

Clean up some obsolete stuff. config -r has not been needed since around
FreeBSD 3.x or so when the 'make depend' picked up the opt_foo.h files.
Convert warnings into actual errors in the hope that buildkernel users
will pay more attention. :-(

Stop turning 'mandatory' into an implicit 'count' line. i386/npx was the
only consumer of this and it is no longer needed.

Implement option strings that we can use in #ifdefs (where unavoidable)
as a replacement for the evil #define NFOO. If 'device npx' is in the
static kernel, a synthetic option '#define DEV_NPX 1' will be available
to stick in an opt_xxx.h file. "#if NNPX > 0" can be replaced with
"#ifdef DEV_NPX" and we can get rid of the overloaded meaning of the
device count mechanism.

Try and stop config(8) from freaking out due to unnecessary paranoia
when using -d. Use realpath(3) to locate the top of the tree rather than
trying to manually trim back the results of a getcwd().

Requested by: alfred

.Nm Config -> .Nm, there is no such program "Config" in FreeBSD.

PR: 23404
Submitted by: Gerald Pfeifer <pfeifer@dbai.tuwien.ac.at>

Make the xxxFILES= list generation generic. This makes it easier to add
things like MFILES= or CONFFILES= without having to modify config code.

Fix buffer overflows in filenames. If you had a path > 80 characters
for your /usr/obj/path/to/my/files path to the kernel, then weird
things happened. make buildkernel would fail because config was
dumping core or generating bad file names (depending on the lenght of
the path).

While I was here, also use strlcpy, strlcat and snprintf (or asprintf)
as necessary. Minor format policing for the snprintf calls as well.

mdoc(7) police: use the new features of the Nm macro.

Use Fx macro wherever possible.

Avoid use of direct troff requests in mdoc(7) manual pages.

Make it possible to specify profiling in the kernel config file.

Do so for LINT.

Add ia64 support.

If a ${KERNEL}.hints file exists, and no hints are specified explicitly,
then include the hints with a marker indicating that it is a fallback.
The kernel side of this is to come shortly.

Argh! I broke the static hints parser at the last minute on freefall when
I added the $FreeBSD$ (commented) line.
Fix:
1: s/break/continue/
2: will somebody please shoot me! :-]

Borrow phk's axe and apply the next stage of config(8)'s evolution.

Use Warner Losh's "hint" driver to decode ascii strings to fill the
resource table at boot time.

config(8) no longer generates an ioconf.c table - ie: the configuration
no longer has to be compiled into the kernel. You can reconfigure your
isa devices with the likes of this at loader(8) time:
set hint.ed.0.port=0x320

userconfig will be rewritten to use this style interface one day and will
move to /boot/userconfig.4th or something like that.

It is still possible to statically compile in a set of hints into a kernel
if you do not wish to use loader(8). See the "hints" directive in GENERIC
as an example.

All device wiring has been moved out of config(8). There is a set of
helper scripts (see i386/conf/gethints.pl, and the same for alpha and pc98)
that extract the 'at isa? port foo irq bar' from the old files and produces
a hints file. If you install this file as /boot/device.hints (and update
/boot/defaults/loader.conf - You can do a build/install in sys/boot) then
loader will load it automatically for you. You can also compile in the
hints directly with: hints "device.hints" as well.

There are a few things that I'm not too happy with yet. Under this scheme,
things like LINT would no longer be useful as "documentation" of settings.
I have renamed this file to 'NOTES' and stored the example hints strings
in it. However... this is not something that config(8) understands, so
there is a script that extracts the build-specific data from the
documentation file (NOTES) to produce a LINT that can be config'ed and
built. A stack of man4 pages will need updating. :-/

Also, since there is no longer a difference between 'device' and
'pseudo-device' I collapsed the two together, and the resulting 'device'
takes a 'number of units' for devices that still have it statically
allocated. eg: 'device fe 4' will compile the fe driver with NFE set
to 4. You can then set hints for 4 units (0 - 3). Also note that
'device fe0' will be interpreted as "zero units of 'fe'" which would be
bad, so there is a config warning for this. This is only needed for
old drivers that still have static limits on numbers of units.
All the statically limited drivers that I could find were marked.

Please exercise EXTREME CAUTION when transitioning!

Moral support by: phk, msmith, dfr, asmodai, imp, and others

A checkpoint of a part of a work-in-progress. Some more cleanups for
config(8). This commit allows control of the creation of the
#include "foo.h" files. We now only create them explicitly when needed.
BTW; these are mostly bad because they usually imply static limits on
numbers of units for devices. eg: struct mysoftc sc[NFOO];
These static limits have Got To Go.

Remove single-space hard sentence breaks. These degrade the quality
of the typeset output, tend to make diffs harder to read and provide
bad examples for new-comers to mdoc.

Mark the 'conflicts' keyword as obsolete, and don't generate (unused)
resource table entries for it.

Use Ns and Sx. -current is 4 not 3.

Remove the cam-specific device wiring code. This was a duplicate of
the data in the resource tables, and cam is getting it directly now.

Bump configvers.h as a precaution. Although existing config files will
work unmodified still, new config files won't work on the old ones.

Clean up something in config(8) that has annoyed me for ages. Remove
the need to specify the unit number of unwired devices. ie: instead
of saying "device fxp0" we can say "device fxp" which is much closer
to what it actually means. The former (fxp0) implied something about
reserving the 0th unit, but it does not and never did - it was a
figment of config(8)'s imagination that we had to work around..
"device fxp0" simply means "compile in the fxp device driver", so we
may as well just write it as "device fxp" which is closer to what it
really means.

Doing this also saves us from filling up the ioconf.c tables with
meaningless entries.

Fix a bungle with the CAM static wiring tables. Write CAMCONF_UNSPEC
instead of -2. This (I believe) caused static wirings to not match.
This should fix Bill Pechter's problem but we'll see.

Problem discovered by: Bill Pechter <pechter@shell.monmouth.com>

Remove even known options if they are in the wrong options header. This
fixes movement of options. Stale copies were left behind.

Bump configversion. The controller/device changes are upwards but not
downwards compatable. If you try and config a s/controller/device/ kernel
with an old config(8), the results will be less than satisfactory.

Support getting *.$MACHINE from sys/conf as well as sys/$MACHINE/conf.
This would mean that we could move files.alpha, files.i386, files.pc98
etc all next to conf/files, and the various Makefiles next to each
other. This should go a long way towards committers "seeing" the
Alpha etc stuff and remembering to update that too as it would be
right next to the i386 config files. Note this does not include
the GENERIC etc files as they can't be shared. I haven't actually
moved the files, but the support is here for it. It still supports
the per-machine conf directories so that folks working on a new arch
can just distribute a subdir of files.

Apply the axe to some more cruft in config(8). In particular:
- redo the "at" configuration system so that it just syntax checks
to make sure the device you're configuring something "at" appears to
exist. Nuke a bunch of complexity that was responsible for creating
"clones" of wildcard devices and some wierd stuff in a few places
including the scbus config tables etc.
- merge "controller" and "device" - there is no difference as far as
the kernel is concernend, it's just something there to make life
difficult for config file writers. "controller" is now an alias for
"device".
- emit full scsi config into the resource tables. We could trivially
change cam to use that rather than it's own "special" table for wiring
and static configuration. ATA could use this too for static wiring.
- try and emulate some of the quirks of the old system where it made
sense. Some were too strange though and I'd be very suprised if they
were features and not outright bugs. nexus handling is still strange.
One thing in particular is that some of the wierd entries in the
newbus devtables is now gone as it was a quirk side effect of the
wildcard/question-mark cloning above.

GENERIC and LINT still build etc.

Fixed removal of unknown options. For options files with only a single
known option, unknown options following the known option were not
removed. Now I think only unknown options in unknown options files
are not removed. This is harmless because unknown options files should
not be used, but removing the files would be cleaner.

Add a 'warning' option for the files* files. This is intended to enable
giving a dire warning about certain drivers going away in the future.

Fix another quirk in the unknown device detection, and also deal with
unknown 'controller' lines.

A better version of the previous checkin. If the user specifies
a custom file that could override a FreeBSD file under a different
configuration, but doesn't under this one, give a different warning.

When specifying additiona user-specified kernel compilation source
files in a 'files.XXX' file, config allows non-FreeBSD source files
with the same name as a FreeBSD source file to override the latter,
and in this situation it issues a warning.

However, if one of the user-specified files is actually a FreeBSD
source file (perhaps your kernel has some custom option that requires
that file), config mistakenly thinks it's a completely new file
and goes ahead and overrides all previous information for that file
(and issues the warning).

Fix this.

With help from: julian

Allow config to generate proper ioconf.c files when devices request DMA
channel 0.

Submitted by: dfr

Re-support "tape" as an alias for device and clean up a bit more cruft
left over from the original system (d_dk was for preassigned iostat unit
numbers for disks)

Add option "-d destdir" which instructs config to use another output
directory than the default one. If the option is not given, then the
output of config is exactly as before. Only when an alternate output
directory has been specified will config modify its behavior.

Additional changed:
o Remove the now conflicting and unused NODEV define. It
conflicts with NODEV in sys/param.h.
o Rename the now conflicting MACHINE token to ARCH. It
conflicts with MACHINE in sys/param.h.
o Fix some easy style bugs.
o Fix some easy grammar bugs in the manpage.

Approved by: peter, archie

Further cleanup. Also remove the following unused or defunct tokens:
and, bio, cam, master, minor, net, priority, sequential, size, slave, trace

Zap references to 'major' token, it's not used.
Zap references to devices.i386 - it's not used. (neither is devices.pc98
or devices.alpha)

$Id$ -> $FreeBSD$

Fix a bunch of broken cross-references

Add $Id$, to make it simpler for members of the translation teams to
track.

The Id line is normally at the bottom of the main comment block in the
man page, separated from the rest of the manpage by an empty comment,
like so;

.\" $Id$
.\"

If the immediately preceding comment is a @(#) format ID marker than the
the $Id$ will line up underneath it with no intervening blank lines.
Otherwise, an additional blank line is inserted.

Approved by: bde

Delete special handling for 'device-driver' suffix, it's not used in
the kernel source now.

Don't include trailing whitespace in ID tokens before comments. ie:
options FOO=10 # comment
would give FOO the value of "10 " and that caused unwanted
touches on the opt_*.h files.

I hope I've got this right..

Detect and remove defunct or unknown options from opt_*.h files. This
can happen when options are removed from the options files.

Check and warn about unknown devices in the config file along the same
lines as the pseudo-device checking. Previously 'device xyz0' would have
been silently accepted without comment.

Understand 'config kernelname' and pretend it was a
'makeoptions KERNEL=kernelname'. Warn about any trailing stuff as it's
not handled here. This is a simple bandaid, hopefully to head off some
complaints from certain people.

Put on my viking helmet from the closet, and get out the war axe and
"retire" some more stuff.

We don't have composite devices here either.

config(8) lobotomy, please see commit msg in sys.

(I have no idea why cvs didn't take these changes before.)

Make options like NO_F00F_HACK work (with context sensitive lexical rules).

Bump configvers; when the updates to generic/lint get committed, the old
config has severe indigestion.

More cleanups, tweaks and features.
- make this work: options FOO123=456 *without quotes*
- grumble (but accept) vector xxxintr, and tty/net/bio/cam flags.
- complain if a device is specified twice (eg: 2 x psm0)
- don't require quotes around: port IO_COM2
- recognize negative numbers. (ie: options CAM_DEBUG_UNIT=-1)
- GC some more unused stuff (we don't have composite disks from config(8)).
- various other nits (snprintf paranoia etc)

The Alpha probably wouldn't appreciate getting the pc98 isa port
definitions. Change it from machine != I386 to machine == PC98.

Slightly reorder the all: to make sure it's before any alternate kernel
names for debugging etc. all: should now always be the first target.

Use pc98/pc98/pc98.h instead of isa/isareg.h in PC98 kernel.

Submitted by: Takahashi Yoshihiro <nyan@wyvern.cc.kogakuin.ac.jp>

Further cleanups. i386_ioconf.c and alpha_ioconf.c were essentially the
same and were merged into a single newbus_ioconf.c. CG'd some more unused
code.

Get out the blow torch and hack away all the unused stuff. Note that
I zapped the MACHINE_MIPS stuff, it isn't likely to be useful apart from
recognition of the machine name. It would be reasonable to expect new
ports would look something like the alpha/i386 from a config perspective.

Corresponding minimal changes for kernel configuration after new-bus
commit.

Made booting with -a work for all configurations. Previously it
only worked for configurations with "swap on generic".

usr.sbin/config/config.y:
- ignore all "swap [on] device ...' specifications except for
warning about them. They haven't done anything related to swap
for almost 4 years, and were previously silently ignored,
except for "swap on generic" which stopped swap${KERNEL}.c
from being generated. Code to support swapping is now deader
than before.

usr.sbin/config/mkswapconf.c:
- don't generate a dummy setconf() function in swap${KERNEL}.c.

sys/i386/conf/files.i386:
- swapgeneric.c is now standard. It should be merged into autoconf.c
so that it doesn't conflict with swap${KERNEL}.c for kernels named
"generic".

sys/i386/i386/autoconf.c:
- don't call setroot() for mfs roots. Since setroot() doesn't do anything
harmful, this was just a waste of time, except possibly for booting with
-a it may have helped prevent an undesireable call to setconf() by
finding a bogus rootdev.
- honor -a for ffs roots. -a now overrides all other ways of specifying
the root device. Previously, -r had precedence over -a, and the -a
handling was usually a no-op.
- don't honor -a for non-ffs roots, since it would currently just get in
the way of a clean panic.

sys/i386/i386/swapgeneric.c:
- don't declare things that are now always declared in swap${KERNEL}.c.
Don't decide things that are now decided in autoconf.c. Code to
support the "generic" case is now dead instead of useless.

Clean up the -g/DEBUG handling. This logic can go in the Makefile
so that config -g can work the same as: makeoptions DEBUG="-g"

Back out default debug kernel. The flags revert to historical behaviour.

Requested-by: ache
bde
dg

Modify targets for debug kernels: when -g was specified, make will
now build a debug kernel called kernel.debug, and create a stripped
version called kernel at the same time. The two targets install and
install.debug are otherwise unchanged.

Requested-by: dillon

Update man page accordingly.

add -s to usage

PR: 11056
Submitted by: Nickolay N. Dudorov <nnd@mail.nsk.ru>

add -s to synopsis

1. Modify config to issue different code for debugging.
2. Config complains if you use -g:

Debugging is enabled by default, there is no ned to specify the -g option

3. Config warns you if you don't use -s:

Building kernel with full debugging symbols. Do
"config -s BSD" for historic partial symbolic support.
To install the debugging kernel, do make install.debug

(BSD was the name of the config file I used; I print out the same
name).

4. Modify Makefile.i386, Makefile.alpha, Makefile.pc98 and config to
work if a kernel name other than 'kernel' is specified. This is
not absolutely necessary, but useful, and it was relatively easy.
I now have a kernel called /crapshit :-)

5. Modify Makefile.i386, Makefile.alpha, Makefile.pc98 "clean" target
to remove both the debug and normal kernel.

6. Modify all to install the stripped kernel by default and the debug
kernel if you enter "make install.debug".

7. Update version number of Makefiles and config.

1. Modify config to issue different code for debugging.
2. Config complains if you use -g:

Debugging is enabled by default, there is no ned to specify the -g option

3. Config warns you if you don't use -s:

Building kernel with full debugging symbols. Do
"config -s BSD" for historic partial symbolic support.
To install the debugging kernel, do make install.debug

(BSD was the name of the config file I used; I print out the same
name).

4. Modify Makefile.i386, Makefile.alpha, Makefile.pc98 and config to
work if a kernel name other than 'kernel' is specified. This is
not absolutely necessary, but useful, and it was relatively easy.
I now have a kernel called /crapshit :-)

5. Modify Makefile.i386, Makefile.alpha, Makefile.pc98 "clean" target
to remove both the debug and normal kernel.

6. Modify all to install the stripped kernel by default and the debug
kernel if you enter "make install.debug".

7. Update version number of Makefiles and config.

Oops, really don't generate compiler warnings about missing braces.

Don't generate compiler warnings about missing braces.

Fix the handling of certain devices.

Previously the foolowing lines would have broken:
controller fdc0 at isa? disable port ? bio
controller fdc0 at isa? disable port 0x100 bio

While this would work:
controller fdc0 at isa? disable port "IO_FD1" bio

The first of the three lines is useful for making placeholder devices
for PCMCIA-floppies, and the second is useful for non-standard hardware.
The failure is a "(null)" string in ioconf.c that the compiler pukes on.

Thanks to: Bruce Evans (bde@freebsd.org)

Eliminate compiler warning when compiling "config.c" in a kernel.

* Change 'struct resource' to 'struct config_resource'.
* Bump config version.

Ignore `vector xxxintr' specifications except for checking their syntax.
Interrupt handlers are now configured in drivers.

Didn't update config/SMM.doc. It doesn't have any i386 examples (not
even `isa').

Bumped CONFIGVERS. This is not necessary for -current yet, but using
the new config with old system sources gives null pointers for all
vectors.

Removed unused/unusable "ha" keyword. Device classes should never have
been visible here.

Fixed disordering of keyword table by "cam".

Don't generate unused file ioconf.h.

Didn't bump CONFIGVERS, since ioconf.h was already unused when
CONFIGVERS was last bumped (although essentially the same (CAM)
commit batch that bumped CONFIGVERS also added bogus includes of
ioconf.h).

Snip out some dead code from the alpha part.

Bump config version. Also mention that even after rebuilding a new
config, you may need to resynchronize your config file with syntax
or name changes documented in GENERIC or LINT.

Build CAM hardwiring info.

add missing sys/queue.h

Reviewed by: Doug Rabson
Submitted by: nsouch
'local' token added to support new bus architecture .c files generated
by .m files.

Dump out ISA device resources for alpha.

Removed all traces of PARAM in Makefile.i386. Incremented CONFIGVERS
to reflect the dependency of Makefile.i386 on nothing being put in
PARAM.

Config versioning is too closely coupled with the Makefile.i386.

Convert the maxusers directive to a normal MAXUSERS option (normally
define MAXUSERS in opt_param.h as directed in /sys/conf/options;
if it's not mentioned there, then define it in IDENT; never define
it in PARAM). MAXUSERS probably should be a completely normal option.

Don't define PARAM now that it is empty.

Cleaned up similar conversion of cpu directives to XXX_CPU options.

Warn about redefined options. Keep using the last value processed.
Options are processed reverse file order, so the first definition in
the config file wins (except for directives that are converted to
option).

Fixed off-by-1 errors in option line numbers. yyline is 0-based, but
was used as if it is 1-based. This happened to give the correct result
for options without values because of a compensating error in newline
lexing. Didn't fix the latter, so line numbers in yyerror() may still
be 1 too high in some cases.

MF22: more explanatory message from config on version mismatch detection.

Don't generate declarations for isa device structs in "ioconf.h".

Don't generate declarations for isa interrupt handlers at all.
Isa interrupt handlers are now declared in <i386/isa/isa_device.h>
but should be converted take a `void *' arg and staticized as
soon as possible.

Updated CONFIGVERS. New configs are very incompatible with
previous versions.

Add (mostly stub) alpha support. Incidentally, it doesn't build on stable
unless I manually construct y.tab.h. Is this normal?

Backed out previous backout. Put y.tab.h back in SRCS.

Backed out previous commit. It was tested, but not for the case where
a separate object tree doesn't exist. Crufty makefiles will have to
put y.tab.h in SRCS so that we know not to create foo.h from foo.y.

Fixed races in `make -jN' using new yacc rules.

Support PC-98 machine.

Don't generate code with syntax errors (hard newlines in string constants
in config.c).

Scaffolding for High Availability code. The actual code will probably
come tomorrow.

Submitted by: Simon Shapiro <shimon@simon-shapiro.org>

Add message when using obsolete command line option.

Suggested by: joerg

Style & message change.

Submitted by: bde

Fixed `make -jN' for large N, as usual.

Style police service brought to you by: bde

Re-introduce '-n' - now a no-op for backwards compatibility only.

Requested by: David Quattlebaum <sasdrq@unx.sas.com>

Make '-n' the default, and introduce a new flag '-r' to get old
behaviour. Also indicate which option(s) are unknown if there are any
old-style options.

All our options are new-style now - enable the warning if unrecognized
(that is, old-style) options are found.

Fix incorrect format string detected by "-Wformat".

Fix some breakage from my last set of changes.

PR: bin/4892
Submitted by: Vasim Valejev <vasim@uddias.diaspro.com>

Bump configvers to 300003 to account for the crd->card (and related)
namespace changes.

Allow for a keyword in the "files" file named "mandatory". The first
candidate for this is "npx0", more are likely to follow.

Check for pseudo-devices that are being configured, but don't appear
in any "files" file. The ``pseudo-device bpf 2'' already hit me too
often.

For safety's sake, explicitly depend all objects on the configvers.h header
to make sure that it's all recompiled even if there is no 'make depend'.
This is overkill, but should be one less thing that someone can do wrong.

Recognize a %VERSREQ=nnnnn string in the system Makefile. Both config(8)
and the kernel will have a 'config interface version number'. If an
incompatable change is made to the kernel that requires a rebuild of
config(8) (such as the cam devtab stuff), then the version number would be
bumped in both places. If a user neglects to rebuild config, then they
will get a nagging (but non-fatal) warning that they need to rebuild
config.

Add isa_devtab_cam.

Wrap too long lines.
Requested by: Bruce.

Police from style.9 and Bruce.

Use err(3), add usage(). -Wall clean.

Oops, the previous commit shouldn't have touch the Makefile.

Don't say that swapping is configured by config.

compare return value from getopt against -1 rather than EOF, per the final
posix standard on the topic.

Reserve a placeholder for 4 SMP ipi interrupts in the same way that
it's done for pci. This is so that systat and vmstat can get at the
interrupt counts for the Inter-Processor Interrupts when running a smp
kernel. This doesn't affect the normal kernel, but makes life easier for
the smp people who don't have to track two versions of config.

Remove support for "port none" and "port auto", it gains nothing,
non-standard and not used. "port auto" is equal to "port?" or missing "port"
keyword now. "port none" is really probe routine task (return -1 for
no ports).

Add #define's for port "none" and "auto"
Now port ommiting is equal to port "none" not to port 0

Print negative values only for two cases used
in config for iobase: -1 (auto) and -2 (none)
Other values are printed as big port numbers

Fix iobase printing for autodetect and none cases
(negative numbers was printed as very big ports)

Sweep through the tree fixing mmap() usage:

- Use MAP_FAILED instead of the constant -1 to indicate
failure (required by POSIX).
- Removed flag arguments of '0' (required by POSIX).
- Fixed code which expected an error return of 0.
- Fixed code which thought any address with the high bit set
was an error.
- Check for failure where no checks were present.

Discussed with: bde

Tidy up the generated config.c file. Use #include "opt_config.h", comment
out text after #endif line, add missing \n at end of file, only install new
config.c if it's different to the last one which preserves the timestamp.

Part #2 of the config cleanup. More aggressive, replaced an NIH
version of strdup() by a macro, killed many calls to strdup(), thus
potentially wasting less malloc'ed space (their args were never be
free()ed desptie despite of being malloc'ed). Probably still a huge
memory leak at all... Also killed two totally useless variables.

I've tested it as i could, but wouldn't be surprised if unexpected
problems showed up. So watch out this space!

Round #1 of cleaning up the config(8) mess. This is only the more
conservative part of the tidyup, like fixing potential buffer overflow
conditions. It is believed to be safe to go into 2.2.

Pointed out by: lozenko@cc.acnit.ac.ru (Evgeny A. Lozenko)

Moved nonstandard compiler profiling options out of config. Just print
the profiling level in config and decide what to do in makefiles.

Makefile.i386:
Align functions to 16-byte boundaries if profiling is enabled. This
will allow a fourfold reduction in the size of the profiling buffers.

Bugfix: all device counts >= 256 was broken, they truncated by % 255
because of u_char count field size. It hits when device
header file already present.

Bring back the `config file in the kernel' feature from the 1.x days. This
is conditionalized by the INCLUDE_CONFIG_FILE option in your kernel config
file and is not turned on by default.

Submitted-By: Bill Pechter <pechter@shell.monmouth.com>

Backout yacc changes.

yacc rule changes.

Implemented a new keyword `disable'. This should be useful for controlling
dangerous drivers in GENERIC.

Removed non-comments on #endifs in config.y.

Improved output formatting in mkioconf.c.

Count PCI irqs in up to 4 ISAish counters named `pci irqnn' instead of
in the clk0 counter.

Reviewed by: se

Make a little more effort to avoid touching certain generated files if
they were not changed. This makes 'make depend' more useful.

Changed the default/min/max number of users to 8/2/512 for all machine
types. This is closer to the reality of reasonable values.

Implemented non-statistical kernel profiling. This is based on
looking at a high resolution clock for each of the following events:
function call, function return, interrupt entry, interrupt exit,
and interesting branches. The differences between the times of
these events are added at appropriate places in a ordinary histogram
(as if very fast statistical profiling sampled the pc at those
places) so that ordinary gprof can be used to analyze the times.

gmon.h:
Histogram counters need to be 4 bytes for microsecond resolutions.
They will need to be larger for the 586 clock.
The comments were vax-centric and wrong even on vaxes. Does anyone
disagree?

gprof4.c:
The standard gprof should support counters of all integral sizes
and the size of the counter should be in the gmon header. This
hack will do until then. (Use gprof4 -u to examine the results
of non-statistical profiling.)

config/*:
Non-statistical profiling is configured with `config -pp'.
`config -p' still gives ordinary profiling.

kgmon/*:
Non-statistical profiling is enabled with `kgmon -B'. `kgmon -b'
still enables ordinary profiling (and distables non-statistical
profiling) if non-statistical profiling is configured.

Implement support for conf/options and i386/conf/options.i386
Note that this code is dormant unless the options files exist.
Also, parsing of quoted options in the config files is improved.

What this allows, is all the options in LINT to be specified to be
configured as #defines in a file rather than on the CC command line at
kernel build time. This means that 'make depend' will catch dependencies
on actual *options*, meaning that you can run 'config' and 'make depend'
in complete safety WITHOUT removing the compile directory each time.

Unfortunately, this requires a pass over the source to get the individual
files to #include the new .h files that would be generated by config.
This has a small compile time penalty (appears up to about 2% slower)
from a "fresh" build. Of course, you should not be needing to do complete
rebuilds very often once this was completed, so it would be an overall
win for most people.

Since this code is dormant and we've got a lot of other things happening
on the kernel tree at the moment (prototypes, devfs, static declarations
etc) I am not planning on doing any changes to activate this feature just
yet.

If CONFIG_NO_CLOBBER_EVER is defined (e.g., in /etc/make.conf), don't make
it possible for config to ever blow away a work directory. Default behavior
remains broken.

Support the configuration of "od" devices.

Submitted by: akiyama@kme.mei.co.jp (Shunsuke Akiyama)

Added support for a %SFILES token to auto-generate a SFILES= file list in
the same way that is done for CFILES. Files ending in .s or .S that match
the option criteria will be included in this list.

Don't pre-processor define 'ident'. This has subtle consequences for people
who don't carefully consider their choice for the machine name. The same
functionality can still be had with an "option", so nothing is lost.

Fix CLEANFILES. Some temporary files were missing.

Put declarations in a header file ("ioconf.h")

Generate prototypes for SCSI functions and function pointers.

Fix redundant declarations of interrupt handlers.

Generate 4.4-style includes (<> instead of "").

Clean up formatting of both the source and the output a bit.

Add missing quote to yyerror call.

Allow the specification of the controller bus when wiring down scsi buses.
This is performed by using a line similar to:

controller scbus0 at ahc0 bus 1

to wire scbus0 to the second bus on an adaptec 2742T controller.

Reviewed by: Peter Dufault(dufault@hda.com), Rod Grimes(rgrimes@FreeBSD.org)

Killed TIMEZONE, DST, and HZ keywords. They have generated a config error
for more than a year now. They've been replaced with userland methods for
changing (see adjkerntz).

Remove trailing whitespace.

Fix 3 printf's that had the wrong number of arguments.
Submitted by: gibbs

Updated to work with Poul-Henning's recent kernel changes in the swap
device table layout...basically, don't output the cruft anymore - it
is now dynamic.

Reviewed by: John Dyson and David Greenman
Submitted by: Poul-Henning Kamp

Don't automatically default dumps to be on a swap device; if the user
wants dumps, he can either configure it explicitly (`dumps on' whatever) or
use the dumpon(8) utility.

Revise this to actually print all the various isa_device field values
it really should have been printing all this time. Also fix my rather
bogus handling of the id_conflicts value by moving it to the end of
isa_device and dealing with that correctly now.

Add a new `conflicts' flag for telling when a device is in conflict with
others. The flag can be put in descriptive locations, e.g.:

device sb0 at isa? port 0x220 irq 7 conflicts drq 1 vector sbintr
or
device psm0 at isa? port "IO_KBD" conflicts tty irq 12 vector psmintr

But is nonetheless boolean only. You can't turn conflict checking off for
only a given type of conflict. I didn't deem it worth the trouble at this
stage, and it's far better than the ALLOW_CONFLICT_* that preceeded it.

Change the perennially annoying reminder to "make depend" (which may or
may not be desired if you're just going to blow the kernel away again later)
and substitute one that tells the user where the new kernel build
directory actually IS, which can at least be argued to be useful information
in all cases.
Reviewed by: davidg

Change warning message for when a device is wired to a floating
host adapter to something hopefully clearer. Take into account
that "wnum()" writes into a static buffer in the warning.

Fixed bug with wiring down given SCSI bus.

Added configuration for SCSI devices wired in place. The documentation
is in "man 4 scsi".

Implement "clean" entries for device config entries.

Submitted by: Pointed out by Bruce Evans <bde@zeta.org.au>

Make good on my promise to finally clean up the config clobbering.
If you invoke config with the `-n' flag or have NO_CONFIG_CLOBBER in
your environment, config will behave the same way it used to. This is
now _documented_ as well. Rip out all the CONFIG_DONT_CLOBBER cruft;
some of it wasn't even correct anyway.

Write setconf() in KNF, in K&R C, and in Standard C.Standard C.

Do away with 'options SWAP_GENERIC' once and for all: I get ill
just thinking about it.

Two changes need to be made to allow 'config kernel swap generic' to
work properly without requiring any compile-time flags:

/usr/src/usr.sbin/config/mkswapconf.c: we need to define a dummy stub
for the setconf() function to replace the one in swapgeneric.c that
isn't available in non-generic configurations.

/usr/src/sys/i386/i386/autoconf.c: the -a boot flag causes setroot()
to be skipped and lets setconf() prompt the user for a root device.
If you skip setroot() in a non-generic kernel, you could get severely
hosed. To avoid this, we silently ignore the -a flag if rootdev != NODEV.
(rootdev is always initialized to NODEV in swapgeneric.c, so if
we find that rootdev is something other than NODEV, we know we're
not using a generic configuration.)

config.y:
Support slice numbers in device names. The syntax is `<driver name>
[<unit number>] ['s' <slice number>] [<partition letter>]'. Only
`['s' <slice number>]' is new here. The slice number defaults to 0
so that there is no change in the output from config if this new
feature is not used.

Replace some magic disk numbers by `dk' slice and label macros.

mkswapconf.c:
Improve the output formatting:
Generate <> style includes.
Print minor numbers in hex so that slice numbers are easy to see and edit.
Print the rootdev and dumpdev names in comments like the swapdev names.

Always return null-terminated identifiers from yylex().

Restore the field widths that were clobbered by the previous commit so that
the columns in ioconf.c line up if there are no long names.

Removed the name length limitations on the device-driver names entirely.
Now device-names can be more than 3 chars.

Fix Sharnoff complaint #121 (cross-reference to config.new which does not
exist on the i386 platform).

Allow config to fully handle the aic7770 driver dependancies.

Finish removing `#ifdef STATCLOCK'. Not removing it properly here
probably broke (only) the association of interrupts with strings
in vmstat output.

Remove ifdef for STATCLOCK. STATCLOCK is standard.

Don't core on ``root on qq0 and fx0 and yy0'' specifications; just
take the first one and ignore all the rest, giving a warning message.
Fixes PR #3.

Put a simple algorithmus in there that checks the last generated prototype
for ioconf.c with the current and suppress the generation if they are
equal. This now prevents all the warnings from the c-compiler about
sio or snd or two adaptecs. This works only if the sio lines are grouped
together in the config files, but it is better than nothing.

Add new keywords to config. The options availible in file.i386 are now:

/*
* filename [ standard | optional ] [ config-dependent ]
* [ dev* | profiling-routine ] [ device-driver] [ no-obj ]
* [ compile-with "compile rule" [no-implicit-rule] ]
* [ dependancy "dependancy-list"]
*/

I added

no-obj - This entry does not create anything linkable to the kernel.
dependancy - Add additional dependancy rules to a target.
no-implicit-rule - Don't assume .c -> .o type rules. Config is really
dumb in this area and assumes that everything is a .c file
irregarless of extention. This was the best choice really
since there may even be .c file that you don't want to follow
the standard rules.

This was all done so that the building to the aic7770 assembler and using
the aic7770 assembler in the building of the aic7770 driver could be config
dependant. I can now have an entry like this for the driver:

aic7770 optional ahc device-driver \
compile-with "${CC} $> -o $@" \
dependancy "$S/gnu/misc/aic7770/aic7770.c" \
no-obj no-implicit-rule
aic7770_seq.h optional ahc device-driver \
compile-with "${.CURDIR}/aic7770 -o $@ $S/gnu/misc/aic7770/aic7770.seq"\
dependancy "$S/gnu/misc/aic7770/aic7770.seq aic7770" \
no-obj no-implicit-rule
i386/isa/aic7770.c optional ahc device-driver \
dependancy "aic7770_seq.h"

I also added '\' escaping to newlines so that this doesn't look as gross as
it could have.

Reviewed by: jkh

Treat formfeeds like any other whitespace.

Don't include map.h - it's been obsoleted.

Whoops - one two many zeros in the initializer. This should have everybody
enabled by default now.

Changes to initialize the id_enabled flag to 1.
Also initialize some fields that were never initialized before, and
simply defaulted to 0. I've never looked at this code before, now
I know why. Config needs to die. Horribly.

Added a slot in the swap-dev-table for NFS-diskless to abuse.

Undo the damage done to my previous changes. Whoever added the
conditional did it backwards, thus flipping the behavior back off
again by default (and only re-enableable through a very counter-intuitive
option setting!). I'm glad I caught this and would merely like to state
again for the record that if you're going to go and modify my changes then
you should at least:

1. Do it correctly, since to do otherwise is kind of a slap in the
face.

2. TELL me.

This is not me just being compulsive, this is simple courtesy.
I'm speaking just of my own preferences here, not necessarily trying
to impose my standards on the group at large (e.g. some other folks might
not even care).
Submitted by: jkh

Revert back to old config behavior if compiled with -DCONFIG_DONT_CLOBBER.

Eradicate my #1 (ok, maybe #2) peeve by making config now blow away
and recreate any previous ../../compile/<blah> directory before laying
down new files. The depends just aren't smart enough to save
us from the grief that config's old behavior has always caused.
Submitted by: jkh

Oops...forgot to list the changes....

/usr/src/usr.sbin.config:
o -DSTATCLOCK gives kludges to support the rtc non-device as well as
old kludges to support the clk non-device.

/usr/src/usr.sbin/config.8:
o Document the trivialness of the new vector.h.

/usr/src/usr.sbin/mkglue.c:
o Only print DEVICE_NAMES and NR_DEVICES in vector.h. These are
only required to support vmstat. The vmstat interface will need
to be improved for dynamic loading.

/usr/src/usr.sbin/mkioconf.c:
o Print device ids to be used as indexes into DEVICE_NAMES.
o Print secondary interrupt handler entry points (xxxintr()) instead
of primary ones (VdevU()). Primary ones are now XintrI() and
XfastintrI() and are independent of the config so they are not
handled here.
o Minor cleanups.

Submitted by: Bruce Evans

Support for Bruce Evans' new dynamic interrupt support.

Submitted by: Bruce Evans

Generate vector.h line for statclock. (See next commit for details.)

Now that the timezone specification is no longer allowed, don't issue a
warning that it is missing.

Disallow timezone specification, but still allow it to be present
if it is zero (so that we can parse old config files).

Converted 'vmunix' to 'kernel'.

Upgrade config to be compatible with our i386 port, pull in 95% of the
changes that have been made in FreeBSD 1.x, except for possibly the nfs
diskless support this is a completed config.

This commit was generated by cvs2svn to compensate for changes in r1553,
which included commits to RCS files with non-trunk default branches.