Pull up following revision(s) (requested by nia in ticket #1232),
all via patch and adpated to netbsd-9:

share/man/man4/pci.4: revision 1.111
share/man/man4/pci.4: revision 1.112
share/man/man4/pci.4: revision 1.113
share/man/man4/pci.4: revision 1.114
share/man/man4/pci.4: revision 1.115
share/man/man4/pci.4: revision 1.116

add ena(4)
Put cross-references in a useful place, copying usb(4)
Caught a few drivers in the list that no longer exist this way.
Be less redundant in the DESCRIPTION
cross-reference pci(9)
add rtsx(4)
s/Disk and tape/Storage/g, more relevant cross-references

(martin)

tests/lint: explain wrong type mismatch in compound literal

When a pointer to a compound literal is used as an initializer, lint
reports a wrong type mismatch.  The details of what happens are now
documented, which allows this problem to be fixed properly.

(rillig)

Fix missing comma.

(thorpej)

Audit CFARG_IATTR in config_found() calls, and remove it in situations
where the interface attribute is not ambiguous.

(thorpej)

lint: use precise type information in message about type mismatch

While here, reword the message, avoiding operators and parentheses.

Since 2021-01-02, providing the precise type name is as easy as the
broad type classification (just replace tspec_name with type_name), and
it's definitely more useful to the human readers.

(rillig)

bind: fix workaround for bug in lint

The previous attempt resulted in a 'syntax error'.

(rillig)

lint: demonstrate bug in assignment of compound literal

Seen in external/mpl/bind/dist/lib/dns/rbtdb.c, update_rrsetstats.

(rillig)

No need to define interface attributes separately from the devices
that carry them if they are specific to that device; just declare the
interface on the device directly.

(thorpej)

sbp does not need its own (unused) interface attribute; it already
carries the "scsi" interface attribute.

(thorpej)

Remove -DHAVE_STRCASECMP from CPPFLAGS that is defined in dix-config.h.

(tsutsui)

Explicitly include dix-config.h for HAVE_foo definitions referred in os.h.

(tsutsui)

Avoid memory leak on empty lines (https://reviews.freebsd.org/D29370)

(christos)

mainbus does not need to carry the "pcibus" interface attribute; PCI
busses do not attach to mainbus, they attach to cpc@mainbus, and cpc
already carries the "pcibus" interface attribute.

(thorpej)

Mechanical conversion of config_found_sm_loc() -> config_found().
CFARG_IATTR usage needs to be audited.

(thorpej)

Brush up previous, or make things more similar to x86:

- Prevent pic_name from appearing vmstat(1) twice.
- Restore "irq" in interrupt id fields of intrctl(8).

For these purposes,

- Add is_evname member to struct intr_source.
- Bump size of is_source to INTRIDBUF, and rename it to is_intrid for clarity.

Now, outputs from vmstat(1) and intrctl(8) are like:

----
$ vmstat -ev
...
openpic irq 39                                3967  26 intr
...
$ intrctl list
interrupt id  CPU0  device name(s)
...
openpic irq 39 3967* wdc1
...
----

(rin)

New sentence, new line.

(wiz)

Fix copy-paste.

(rin)

note that ttyaction.[35] first appeared in netbsd 1.3.

(mrg)

make: fix typos in documentation of ModChain

(rillig)

Use standard section names, sort, use more macros.

(wiz)

Give config_found() the same variadic arguments treatment as
config_search().  This commit only adds the CFARG_EOL sentinel
to the existing config_found() calls.  Conversion of config_found_sm_loc()
and config_found_ia() call sites will be in subsequent commits.

(thorpej)

lint: prefix error messages with 'error:'

This makes it easier to find these errors in the build logs.

(rillig)

lint: remove redundant prototypes

No functional change.

(rillig)

lint: adjust type in documentation of add_array

No functional change.

(rillig)

tests/lint: test reachability of goto and named labels

(rillig)

tests/lint: add test for reachability of non-constant conditions

(rillig)

tests/lint: add test for reachability of nested 'if' statements

(rillig)

lint: warn about unreachable statement after joining the 'if' branches

(rillig)

lint: invert 'rchflag', call it warn_about_unreachable instead

No functional change.

(rillig)

In "rescan" routines, always pass locators and the interface attribute
straight through to config_search().  Also, for devices that carry only
one interface attribute, no need to do an ifattr_match(), because
rescan_with_cfdata() will have already validated that the parent is
eligible, which includes an interface attribute check.

(thorpej)

lint: reduce number of places where 'reached' is set

When determining the reachability of a statement, the idea was that
whenever 'reached' was set to false, 'rchflg' (the abbreviation for "do
not warn about unreachable statements") would be reset as well.

In some (trivial) cases, this was done, but many more interesting cases
simply forgot to set this second variable.  To prevent this in the
future, encapsulate this in a simple helper function.

Now even if a statement is reachable, 'rchflg' gets reset.  This does
not hurt since as long as the current statement is reachable, the value
of 'rchflg' does not matter.

No functional change.  There would be quite a big functional change
though if check_statement_reachable were to reset 'rchflg' instead of
'reached', as the comment already suggests.  In that case, with the
current code, many legitimate warnings about unreachable statements
would be skipped, especially those involving 'if' statements, since
these didn't reset 'rchflg' properly before.

(rillig)

Remove a now-unused variable.

(thorpej)

In config_search(), we already asserted that either an interface
attribute is not specified, or the specified attribute is carried
by the parent.

Add an additional assertion: That an interface attribute is specified
or that the parent has fewer than 2 interface attributes (i.e. lack of
interface attribute specification would be ambiguous).

Yes, "fewer than 2".  Zero interface attributes doesn't really make sense,
because a device cannot then be a parent of another device by definition.
But cfparent_match() would already catch this situation gracefully, and
there is obviously no ambiguity when a device has no interface attributes.

(thorpej)

CFARG_IATTR usage audit:

If a device carries only one interface attribute, there is no need
to specify it when calling config_search(); that specification is
meant only to disambiguate which interface attribute (which is a
proxy for "what kind of attach args are being used") is having
children attached.  cfparent_match() will take care of ensuring that
any potential children can attach to one of the parent's iterface
attributes, and if the parent only carries one, no disambiguation is
necessary.

(thorpej)

use a pipe instead of sched_yield()

(christos)

PR/56066: Jessica Clarke: Add tests for calling {_,}longjmp with a zero value.

(christos)

eumb does not need to carry the "mainbus" interface attribute.

(thorpej)

revert the unintentional whitespace changes

(christos)

Remove some obsolete information and add some more recent additions from
the nslookup.1 page in the distribution. Consider replacing this man page
with the one from the distribution, since this manual page describes options
available only in the original implemementation. On the other hand, this
man page is mdoc, not man... (Takahiro Kambe)

(christos)

lint: fix reachability for while (0)

(rillig)

lint: fix reachability for constant controlling expression in for loop

(rillig)

lint: fix reachability for if-then-else statements

(rillig)

lint: rename functions for handling control statements

No functional change.

(rillig)

lint: fix reachability computation in if statements

Previously, only loop statements were considered for reachability.  This
ignored the possibility of an early return in an if statement, or
unreachable branches.

(rillig)

tests/lint: sync redundant documentation in test for strict bool mode

(rillig)

tests/lint: add tests for reachability of statements

Right now, reachability is only implemented for loops, but not for
selection statements.

(rillig)

lint: add missing space when adding a new test case

(rillig)

lint: rename c_rchif to c_reached_end_of_then

No functional change.

(rillig)

tests/lint: demonstrate bug in endless loop detection

(rillig)

lint: document why an unreachable statement is set to reachable

No functional change.

(rillig)

lint: rename c_cont to c_continue

No functional change.

(rillig)

lint: document the precise meaning of control_statement.c_break

No functional change.

(rillig)

lint: rename i_infinite to i_maybe_endless

Not every loop that has 'while (1)' is an endless loop.  It may still
contain a 'return' somewhere.

(rillig)

lint: fix wrong 'falls off bottom' after return in do-while

(rillig)

lint: fix wrong 'statement not reached' in do-while loop

(rillig)

lint: remove redundant braces in doreturn

No functional change.

(rillig)

audioctl.1: describe foibles

(nia)

tests/lint: clarify that a brace-enclosed initializer needs an expr

(rillig)

lint: rename d_fargs, d_fdpos and d_fpsyms to be less abbreviated

No functional change.

(rillig)

lint: rename d_nedecl to d_nonempty_decl

No functional change.

(rillig)

lint: rename d_rdcsym to d_redeclared_symbol

No functional change.

(rillig)

lint: inline local variable in declarator_1_struct_union

No functional change.

(rillig)

lint: rename clst to case_labels

No functional change.

(rillig)

lint: rename declspecs to declaration_specifiers, just as in C99

No functional change.

(rillig)

lint: rename grammar rule 'data_def' to 'top_level_declaration'

No functional change.

(rillig)

lint: remove redundant '%prec' declarations from the grammar

No change to the generated binary.

(rillig)

Fix a comment

(skrll)

-DKERNEL_BASE_VOFFSET= has annoyed me for the last time...

Introduce KERNEL_VOFFSET_RUNTIME which prevents the addition of
-DKERNEL_BASE_VOFFSET= to the command line and use it on the
__HAVE_GENERIC_START kernels which do runtime calculation of the
offset.

(skrll)

lint: rename token T_XOR to T_BITXOR

For symmetry with the operator, which is named BITXOR.

No functional change.

(rillig)

lint: in debug output, use uniform file location references

This makes it easy to click on the location in the IDE instead of having
to manually parse the location and navigate to it.

No functional change outside debug mode.

(rillig)

lint: remove redundant operator information from the grammar

Several tokens can only ever map to a single operator and thus do not
need to encode the operator.  Indeed, they already encoded it as NOOP,
and it was not used by any grammar rule.

No functional change.

(rillig)

Adjust the kernel virtual address space so that KASAN will map the kernel
seperately from managed kernel virtual memory and not map the unused space
between the two.

(skrll)

Tweak a comment

(skrll)

Remove the unnecessary invalidation code in arm_efirt_md_map_range.

pmapboot_enter will panic if any overlapping mappings existed before and
a full TLB invalidate was done as part of turning the MMU on in locore.

(skrll)

There is no need to have the "pcibus" interface attribute on mainbus;
PCI bus instances to not attach there.

(thorpej)

seq.c: replace non-standard `\e` with standard compliant `\x1B`

(cheusov)

lint: clean up new_name_node

No functional change.

(rillig)

lint: move getopname over to tree.c

Except for the one use in print_tnode, the name of the operator is only
used in tree.c.

No functional change.

(rillig)

Don't pass the autoconfig print routine as the aux pointer to the
indirect search routine.  That argument slot is meant for "attach
args" structures, not function pointers.  Besides, no one actually
used it anyway.

(thorpej)

No need to pass the parent softc down to the indirect submatch routine;
it's available from the parent device_t there.

(thorpej)

lint: remove redundant operator properties table

It's enough to have modtab, which describes the properties of the
various operators.  There is no need to have a second table imods that
holds the same content.  Rather make modtab constant as well.

The only possible functional change is that the names of the internal
operators 'no-op', '++', '--', 'real', 'imag' and 'case' may appear in
diagnostics, where previously lint invoked undefined behavior by passing
a null pointer for a '%s' conversion specifier.

(rillig)

lint: fix argument names and table headings for operator definitions

The abbreviations in the table of operator properties had been wrong
since ops.def 1.10 from 2021-01-12, when strict bool mode was added.  In
an earlier working draft, I had named that column 'takes_others' instead
of 'requires_bool', that's where the 'o' came from.

The names of the macro arguments had been wrong since op.h 1.11 from
2021-01-09, when the order of the columns changed and the macros were
not adjusted accordingly.  Since all the properties of the operator
table are uniform, this didn't result in any bugs, it was just confusing
for human readers.

Clang-tidy suggests to enclose the macro arguments in oper.c in
parentheses but that is not possible since the arguments are either
empty or 1, and the syntactical ambiguity of the '+ 0' being either a
unary or a binary operator is needed here.

No change to the resulting binary.

(rillig)

The proliferation if config_search_*() and config_found_*() combinations
is a little absurd, so begin to tidy this up:

- Introduce a new cfarg_t enumerated type, that defines the types of
  tag-value variadic arguments that can be passed to the various
  config_*() functions (CFARG_SUBMATCH, CFARG_IATTR, and CFARG_LOCATORS,
  for now, plus a CFARG_EOL sentinel).
- Collapse config_search_*() into config_search() that takes these
  variadic arguments.
- Convert all call sites of config_search_*() to the new signature.
  Noticed several incorrect usages along the way, which will be
  audited in a future commit.

(thorpej)

lint: make lint's own code pass the strict bool mode

No functional change.

(rillig)

lint: use macro for initializing keywords table

Since today, lint's strict bool mode requires initializers to have the
correct type.  The flags in kwtab are of type bool and were initialized
with an int, for brevity.  Keep the brevity and do the conversion from
int to bool in a macro.

By defining several macros for the different kinds of keywords, reduce
the clutter of having 2 additional zeroes per line.  The macros also
remove the redundancy and thereby the possible inconsistency of filling
the wrong fields since these depend on the token type.

No functional change.  The only change to the binary is due to the
changed line numbers in the calls to lint_assert.

(rillig)

lint: use macro for encoding type information

In lint's strict bool mode, initialization must be of the correct type.
This affects the bool fields in ttab_t, which are initialized with int.
To keep the code brief, preserve these ints and let a macro do the
actual work of converting them to bool.

No change to the generated binary.

(rillig)

lint: inline access to type properties in check_bad_enum_operation

No functional change.

(rillig)

lint: in strict bool mode, check initialization as well

C99 6.7.8p11 says for initialization that "the same type constraints and
conversions as for simple assignments apply", so actually apply them.
(I had just forgotten this "operator" when I first implemented strict
bool mode.)

(rillig)

tests/lint: test initialization in strict bool mode

I somehow forgot to list INIT as one of the operators in
is_assignment_bool_or_other.

(rillig)

lint: rename ftflg to seen_fallthrough

One less abbreviation to remember when reading the code.

No functional change.

(rillig)

lint: make the calls to clear_warning_flags stand out less

Even though clear_warning_flags and its companions are implemented as
macros, they act like ordinary functions.  Do not distract the reader by
using uppercase names for them.

No functional change.

(rillig)

lint: clean up grammar for declaration after statement

No functional change.

(rillig)

lint: fix check for declaration after statement in pre-C99 mode

The new code may not be the most beautiful, but it fixes all bugs that
occurred while testing message 327.  The grammar rules are taken from
C99 6.8.2, so it's no surprise they work well.

(rillig)

Make pmapboot_enter panic if anything goes wrong and any mappings overlap
rather than only doing it in locore.S

(skrll)

cmp.c: use C99 strtoll(3) instead of legacy strtoq(3)

(cheusov)

lint: properly name C99 in message about declaration after statement

Now that C99 has been released and published, there is no reason anymore
to refer to it as C9X.

(rillig)

tests/lint: add test for declaration after statement

(rillig)

lint: use proper boolean literals instead of 0/1

The code in the C grammar is generated by yacc and is not checked by
lint's strict bool mode, therefore the replacement was done manually.

No change to the resulting change.

(rillig)

lint: update comment on setasm for C99

No functional change.

(rillig)

lint: reduce indentation and braces in merge_type_specifiers

No functional change.

(rillig)

lint: extract adjusting of the storage class into separate function

No functional change.

(rillig)

lint: document the struct for declarations more precisely

No functional change.

(rillig)

lint: remove unnecessary '%prec T_COMMA' from grammar

My previous commit message was wrong in saying that the '%prec' was
necessary.  It is not necessary.

Most probably I misspelled the name of the grammar rule as opt_comma
instead of comma_opt, which would lead to the same number of conflicts
in the grammar plus a warning, but no build failure.

(rillig)

lint: join grammar rules for initialization

The '%prec T_COMMA' is necessary to avoid lots of parse errors in the
lint1 unit tests.  Curiously, further down in the grammar, for compound
literals, the '%prec T_COMMA' is not necessary, even though the context
looks very similar.

No functional change.

(rillig)

lint: fix grammar for initialization

Previously, the grammar syntactically accepted the following code:

int var = .member = 12345;

The designation '.member =' can only be used with brace-enclosed
initializers.

(rillig)

lint: align rule names in grammar with C99

No functional change.

(rillig)

tests/lint: add more tests for initialization, based on C99 6.7.8

(rillig)

lint: fix assertion failure after error in designation

In d_c99_init.c, the initialization of array_with_designator failed.
The designator '.member' from that initialization was not cleaned up
before starting the next initialization.

(rillig)

lint: replace segmentation fault with assertion failure

(rillig)

Don't mark EFI runtime pages LX_BLKPAG_OS_READ | LX_BLKPAG_OS_WRITE as
these bits are only used by the current pmap fault code and these are
wired pages which will never fault.

(skrll)

Apply the

earlyconsgetc: return -1 instead of 0 because we will never produce input

fix to all copies.

(skrll)

Fix and improve the buffer length calculation to avoid zero length
even if blk_ms is small.
This fixes PR kern/56059.

(isaki)

lint: improve debug loggin for initialization

(rillig)

tests/lint: add comma to struct initialization

This places the error marker one line up, where it is expected.

(rillig)

lint: rename designator_pop_name to designator_shift_name

The entries are removed from the beginning, not from the end.

No functional change.

(rillig)

earlyconsgetc: return -1 instead of 0 because we will never produce input

(jmcneill)

lint: use standard form of statement macro for debug_node

No functional change.

(rillig)

lint: rename in_bit to in_bits

No functional change.

(rillig)

lint: rename tsize to type_size_in_bits

The shorter name size_in_bits was already taken by the function-like
macro with argument type tspec_t.

No functional change.

(rillig)

tests/lint: add test for incompatible pointer types in return

(rillig)

lint: add reminder that C99 does not allow '{}' in initializer

No functional change.

(rillig)

Support pmap_growkernel and KASAN shadow mapping of the new KVA.

Neither mips nor ppc booke actually use pmap_growkernel (at present).

Thanks to rin@ for testing a similar patch on ppc booke.

(skrll)

lint: replace assertion in initialization with proper error message

(rillig)

lint: rename push_member and pop_member

These two functions are supposed to model the designator that is used
for initializing structs and arrays.  The implementation is still buggy
and does not work at all for C99 designators with multiple names, see
d_init_pop_member.c.

For now, just rename the functions to head in the right direction.

No functional change.

(rillig)

Sprinkle a few more \n's and the end of some debug printfs.

(simonb)

lint: extend documentation about initialization

No functional change.

(rillig)

lint: split initstack_pop_item into separate functions

No functional change.

(rillig)

lint: improve debug logging in initstack_push

No functional change outside debug mode.

(rillig)

lint: split initstack_push into smaller functions

No functional change.

(rillig)

lint: clean up control flow in initstack_push

No functional change.

(rillig)

lint: extract extend_if_array_of_unknown_size from initstack_push

No functional change.

(rillig)

Add PTRACE_ILLEGAL_ASM using the MIPS32r6/MIPS64r6 backwards and
forwards compatible "sigrie" instruction to generate a Reserved
Instruction trap.

(simonb)

lint: replace undefined behavior during initialization with assertion

This only affects code that is already rejected by the compiler.

(rillig)

lint: improve debug logging during initialization

No functional change outside debug mode.

(rillig)

lint: make the debug log for nodes more readable

The operator NAME has the name 'name', therefore no special case is
needed.

Having the words 'with type' in the message makes the message easier to
find from the debug log.  Given that the operator name is used unquoted,
the log message 'name: int value=111' was nearly impossible to find in
the code.

Replace the '()' with an actual word, to avoid any confusion about
whether the type name might be a function type without prototype.

Reduce the amount of '=' signs, instead use commas to separate the
properties of the node.

No functional change outside debug mode.

(rillig)

lint: rename operator 'const' to 'constant'

The previous name could be too easily confused with the type qualifier
'const'.  The operator name is mainly used in the debug log, only
occasionally in the output.  Since 'constant' is not a "real" operator,
it probably doesn't occur in messages at all.

(rillig)

lint: reword message about type mismatch in initialization

Using parentheses for quotes is unusual, furthermore the previous
message didn't follow proper grammar rules, sacrificing clarity for
brevity.

(rillig)

tests/lint: add test for type mismatch in initialization

Copied and adapted from d_c99_init.c.

(rillig)

tests/lint: warning 210 has type information by now

(rillig)

lint: reduce debug logging for initialization, update documentation

No functional change outside debug mode.

(rillig)

lint: document how initialization works, improve debug logging

No functional change outside debug mode.

(rillig)

tests/lint: add more examples for initialization

(rillig)

mtree: use POSIX type uint32_t instead of u_int32_t

(cheusov)

join.c: print usage after warning "illegal option..." as it was originally intended

(cheusov)