Make all of the EISA chipset functions call through real functions,
rather the macros.

(thorpej)

indent: extract probably_typedef into separate function

This condition is complicated enough that it warrants being split into
several clauses, maybe even with an explanation.

No functional change.

(rillig)

indent: reduce code and data size for lexing of numbers

Instead of having a table of strings (121 pointers + 121 data
relocations), reduce that table to the actual character data and use a
secondary table for looking up the correct row in the main table.

No functional change.

(rillig)

Call cnpollc(9) before cngetc(9) as the cons(9) man page says.

Affects only inside #ifdef DEBUG part on "trap during panic" in trap.c
derived from hp300.

(tsutsui)

indent: rename option variable to be more expressive

No functional change.

(rillig)

tests/indent: demonstrate mergin of comments

This may or may not have been intended. Especially with the option
'-nfc1' the result looks confusing. It's not a case that occurs every
day, though.

(rillig)

Disable drmkms module builds.

They haven't been functional (and won't be before fixing genfb abi).
In a new drm update, we haven't prioritized fixing the build either.

(maya)

indent: convert remaining ibool to bool

No functional change intended.

(rillig)

indent: convert parser_state from ibool to bool

indent.c:400:5: error: suggest parentheses around assignment used as
    truth value
io.c:271:32: error: ���~��� on a boolean expression

No functional change intended.

(rillig)

indent: convert options from ibool to bool

No functional change intended.

(rillig)

indent: prepare for lint's strict bool mode

Before C99, C had no boolean type. Instead, indent used int for that,
just like many other programs. Even with C99, bool and int can be used
interchangeably in many situations, such as querying '!i' or '!ptr' or
'cond == 0'.

Since January 2021, lint provides the strict bool mode, which makes bool
a non-arithmetic type that is incompatible with any other type. Having
clearly separate types helps in understanding the code.

To migrate indent to strict bool mode, the first step is to apply all
changes that keep the resulting binary the same. Since sizeof(bool) is
1 and sizeof(int) is 4, the type ibool serves as an intermediate type.
For now it is defined to int, later it will become bool.

The current code compiles cleanly in C99 and C11 mode, as well as in
lint's strict bool mode. There are a few tricky places:

In args.c in 'struct pro', there are two types of options: boolean and
integer. Boolean options point to a bool variable, integer options
point to an int variable. To keep the current structure of the code,
the pointer has been changed to 'void *'. To ensure type safety, the
definition of the options is done via preprocessor magic, which in C11
mode ensures the correct pointer types. (Add CFLAGS+=-std=gnu11 at the
very bottom of the Makefile.)

In indent.c in process_preprocessing, a boolean variable is
post-incremented. That variable is only assigned to another variable,
and that variable is only used in a boolean context. To provoke a
different behavior between the '++' and the '= true', the source code
to be indented would need 1 << 32 preprocessing directives, which is
unlikely to happen in practice.

In io.c in dump_line, the variables ps.in_stmt and ps.in_decl only ever
get the values 0 and 1. For these values, the expressions 'a & ~b' and
'a && !b' are equivalent, in all versions of C. The compiler may
generate different code for them, though.

In io.c in parse_indent_comment, the assignment to inhibit_formatting
takes place in integer context. If the compiler is smart enough to
detect the possible values of on_off, it may generate the same code
before and after the change, but that is rather unlikely.

The second step of the migration will be to replace ibool with bool,
step by step, just in case there are any hidden gotchas in the code,
such as sizeof or pointer casts.

No change to the resulting binary.

(rillig)

Refactor and cleanup sio (uPD7201) drivers.

- remove confusing syscn*() functions (and its header) and prepare
  explicit siottycninit(), siottycnget() and siottycnput() functions
- use exported struct consdev and cn_tab to initialize siotty console
- enable and handle E/S interrupts to make BREAK signal detected properly
- make CSR (status regsiters) access functions inline static
- make single byte read/write (i.e. cnputc() and cngetc()) functions
  static inline and take struct sio_register rather than sio channel
- use proper integer type (uint16_t) for getsiocsr() as siotty.c
- handle channel dependent CR2A and CR2B registers properly
- use more explicit definitions for RR_* macro used by getsiocsr()
- define and use proper RR0 (read register) values
  (there is no isStatusReg(r) macro used on 4.4BSD/luna68k)

Tested on LUNA with both wscons console and serial console.

(tsutsui)

Sync with OpenBSD r1.101:

When processing a received packet, only sync the amount of bytes
mcx(4) told us has arrived.  The DMA map's mapsize on RX packets
is the length of the allocated buffer.  For mcx(4), this can be
more than around 9000 bytes, as each buffer will be at least as
big as the maximum supported MTU.  There's no need to sync the
whole buffer, if it's only a small packet.

(jmcneill)

indent: use standard definition for bool, true, false

(rillig)

indent: merge duplicate code for initializing buffers

No functional change.

(rillig)

indent: clean up initialization of options

The default values in 'struct pro' were redundant but all consistent,
even with the commented defaults in main_parse_command_line.

No functional change.

(rillig)

indent: remove ifdef for lint

NetBSD lint does not need them anymore, FreeBSD does not have lint.

(rillig)

tests/indent: test formatting of while and do-while loops

(rillig)

tests/indent: test deeply nested struct declarations

(rillig)

tests/indent: test deeply nested struct declarations

(rillig)

tests/indent: clean up test driver

For most of the test data, it was not necessary to copy the files to the
working directory, even though the comment said so.  The test data had
been taken from the cleaned files anyway.

Use sed instead of cp to copy the files, to allow for comments in all
kind of files, even the type lists for the option '-U'.

(rillig)

tests/indent: test edge cases for the tokenizer

(rillig)

indent: move statistical values into a separate struct

No functional change.

(rillig)

indent: make lex_char_or_string simpler

The previous code was so tricky that every second line needed a comment
that explains what's going on.  Replace the complicated code with the
usual straight-forward string-copying patterns.

No functional change.

(rillig)

tests/indent: extend tests for strings and characters

(rillig)

tests/indent: test lexing of character constants and string literals

(rillig)

tests/indent: fix ShellCheck warnings, use simple variables

(rillig)

Move DRM-driver firmware from base to its own set, gpufw

This set is only installed on amd64,i386,evbarm.
This set is installed on minimal installs and on install media, in
case someone needs it for basic driver functionality.

Comments:
Switched to a single MK tunable for it - that is probably unneeded.

An upcoming DRM update will include even fatter firmware, and we'd
like to minimize the impact of it.

(maya)

indent: add nonnull memory allocation functions

The only functional change is a single error message.

(rillig)

indent: group global variables for token buffer

No functional change.

(rillig)

indent: inline macro 'token'

No functional change.

(rillig)

indent: group global variables for code buffer

No functional change.

(rillig)

indent: rename variables of type token_type

The previous variable name 'code' conflicts with the buffer of the same
name.

No functional change.

(rillig)

indent: group global variables for label buffer into struct

No functional change.

(rillig)

indent: extract parse_indent_comment from fill_buffer

No functional change.

(rillig)

indent: group global variables for the comment buffer

No functional change.

(rillig)

indent: rename local variable in fill_buffer

The local variable name 'com' prevented grouping the global variables
combuf, s_com, e_com and l_com into a struct named 'com'.

No functional change.

(rillig)

tests/indent: demonstrate inconsistent formatting of comments

(rillig)

indent: fix token duplication after C99 comment

The code that keeps blank lines after C99 comments still looks wrong,
but at least it's better than before.

(rillig)

mandoc 1.14.6 out.

(wiz)

new libfido2, pam-u2f

(christos)

bump libfido2

(christos)

merge conflicts between 1.0.8 and 1.2.0

(christos)

Merge conflicts, avoid ssp issues

(christos)

merge conflicts

(christos)

Add copyright for no-memcpy toeplitz hash, pointed out by wiz@n.o, thanks.

(knakahara)

Catch up with spl changes.

It's 2021 and ugly is in (with apologies to PWEI)

(skrll)

Make the KOBO and NETWALKER kernels compile again.

(skrll)

tests/indent: demonstrate token repetition after line-end comment

(rillig)

indent: fix space-tab in indentation

(rillig)

Fix build failure for i386 INSTALL_XEN3PAE_DOMU, sorry.

(knakahara)

Add RSS toeplitz hash functions which calculate from mbuf.

(knakahara)

Import asymmetric toeplitz hash without memcpy implemented by ryo@n.o.

This implementation has better performance than memcpy'ed one.
(30%-60% improvement in micro benchmark)

import from
    https://github.com/ryo/l2pkt/blob/master/l2pkt/toeplitz_hash.c

(knakahara)

Move XEN3_DOM0 as close as possible to GENERIC.
Document why some options are disabled
Set NO_PCI_MSI_MSIX to work around crashes reported in multiple PR

(manu)

make: fix memory leak in error case of the ':?' modifier

(rillig)

sun4v: update TODO with current status when running on a T5-based system

(palle)

Regen.

(msaitoh)

Add some Intel devices:

- Tiger Lake
- I225V and I225LM
- WiFi 6 AX201

(msaitoh)

use lwp_trapframe() macro. NFC.

(ryo)

Spell the number of nanoseconds as 10^9.  Forgotten in previous.

(uwe)

Minor markup tweaks.

(uwe)

Spell the number of nanoseconds as 10^9.

10^9 is 1e9 (it's "e" that spells 10 here), not 10e9.  The target
audience of this man page is not likely to be very fluent in floating
point, so avoid significand/exponent spelling and use the spelling
that it is familiar with.

(uwe)

Markup fixes.

(uwe)

printf(3): mention snprintb(3)

(wiz)

remove stray char.

(christos)

ena: fix packet reordering issue

A reorder occures when
- memory allocation fails in bus_dmamap_load_mbuf() or
- submission queue is full

This patch makes ena(4) to
- allocate memory in advance (BUS_DMA_ALLOCNOW flag in bus_dmamap_create())
- check if the queue is vacant before pcq_get()

Patch from KUSABA Takeshi <t-kusaba@iij.ad.jp>

(jmcneill)

Tickets #1347 and #1348

(martin)

Pull up following revision(s) (requested by skrll in ticket #1348):

bin/cp/utils.c: revision 1.47

PR/54564: Jan Schaumann: cp of a fifo yields an empty file

Don't short-circuit 0 sized stat entries if they don't belong to regular
files.

Also don't try to mmap non-regular files.

(martin)

Pull up following revision(s) (requested by kre in ticket #1347):

bin/mkdir/mkdir.c: revision 1.39

PR bin/56398

The final component of both a/b/c and a/b/c/ is "c", so that's the one
to which the -m arg applies.

(martin)

Ticket #1697

(martin)

Pull up following revision(s) (requested by kre in ticket #1697):

bin/mkdir/mkdir.c: revision 1.39

PR bin/56398

The final component of both a/b/c and a/b/c/ is "c", so that's the one
to which the -m arg applies.

(martin)

Add support COMPAT_LINUX for aarch64

(ryo)

add aarch64, and sort architecture names by alphabetical order.

(ryo)

Since trampoline ABI ver0 is also used in other emulation environments (e.g. linux emulation),
checking (emul->e_sigobject != NULL) to determine if it is allowed or not.

(ryo)

add support COMPAT_LINUX for aarch64

(ryo)

Print the cache information in similar formats and arm and aarch64, e.g.

arm before
[  1.0000000] cpu0: 32KB/64B 2-way L1 PIPT Instruction cache
[  1.0000000] cpu0: 32KB/64B 2-way write-back-locking-C L1 PIPT Data cache
[  1.0000000] cpu0: 2304KB/64B 16-way write-through L2 PIPT Unified cache

arm after
[  1.0000000] cpu0: L1 32KB/64B 2-way (256 set) PIPT Instruction cache
[  1.0000000] cpu0: L1 32KB/64B 2-way (256 set) write-back-locking-C PIPT Data cache
[  1.0000000] cpu0: L2 2304KB/64B 16-way (2304 set) write-through PIPT Unified cache

aarch64 before
[  1.0000030] cpu1: L1 48KB/64B*256L*3W PIPT Instruction cache
[  1.0000030] cpu1: L1 32KB/64B*256L*2W PIPT Data cache
[  1.0000030] cpu1: L2 2048KB/64B*2048L*16W PIPT Unified cache

aarch64 after
[  1.0000030] cpu1: L1 48KB/64B 3-way (256 set) PIPT Instruction cache
[  1.0000030] cpu1: L1 32KB/64B 2-way (256 set) PIPT Data cache
[  1.0000030] cpu1: L2 2048KB/64B 16-way (2048 set) PIPT Unified cache

(skrll)

wscons: add definitions for B��PO, a standardized, Dvorak-style optimized
keyboard layout for French

(nia)

fail on error, not on success.

(mlelstv)

fix typo in comment

(ryo)

ksymsmmap: Add missing uao_reference(9) call for ks->ks_uobj.

Fix failure for savecore(8) and subsequent kernel panic, introduced to
kern_ksyms.c rev 1.03, at least for sh3 and alpha.

For sh3 and alpha, savecore(8) supports coff and ecoff, respectively, via
libkvm via nlist(3). nlist(3) routines for coff and ecoff use mmap(2) and
munmap(2) for /dev/ksyms.

This munmap(2) decrements reference count for ks->ks_uobj. Unless it is
incremented in ksymsmmap(), ks->ks_uobj will be freed unexpectedly.

(rin)

make: reduce relocations, thereby reducing .text size

No functional change.

(rillig)

make: reduce relocations and thereby .text size, make opname const

No functional change.

(rillig)

make: rename and invert CondParser.lhsStrict to leftUnquotedOK

The new name is easier to grasp than the rather abstract word 'strict'.

No functional change.

(rillig)

make: do not allow unquoted 'left == right' after modifier ':?'

Having a static variable for state that clearly belongs in the parser
looked suspicious, and indeed it was wrong.

When the distinction between .if conditions and expressions of the form
${condition:?:} was added in cond.c 1.68 from 2015-05-05, a new unit
test was added, but it didn't cover this edge case.  At that time, the
state of the condition parser consisted of a few global variables
instead of a separate data type, as would have been appropriate for
parsing nested conditions.

(rillig)

tests/make: demonstrate unintended unquoted string in conditions

(rillig)

make: remove unnecessary negation from CondParser_StringExpr

No functional change.

(rillig)

make: remove unnecessary const from parameters

These were leftovers from earlier refactorings, when extracting code to
separate functions.

No functional change.

(rillig)

make: fix out-of-bounds memory read (since previous commit)

(rillig)

make: reduce indentation in CondParser_FuncCall

No functional change.

(rillig)

make: fix documentation about operator associativity in conditions

(rillig)

centralize version handling

(christos)

UsePrivilegeSeparation is no more

(christos)

don't opencode kauth_cred_get()

(christos)

don't opencode kauth_cred_get()

(christos)

don't opencode kauth_cred_get()

(christos)

don't opencode kauth_cred_get()

(christos)

don't opencode kauth_cred_get()

(christos)

don't opencode kauth_cred_get()

(christos)

don't opencode kauth_cred_get()

(christos)

don't opencode kauth_cred_get()

(christos)

don't opencode kauth_cred_get()

(christos)

remove extra changes

(christos)

don't opencode kauth_cred_get()

(christos)

don't opencode kauth_cred_get()

(christos)

don't opencode kauth_cred_get()

(christos)

don't opencode kauth_cred_get()

(christos)

undo previous, wrong file.

(christos)

don't opencode kauth_cred_get()

(christos)

don't opencode kauth_cred_get()

(christos)

don't opencode kauth_cred_get()

(christos)

don't opencode kauth_cred_get()

(christos)

don't opencode kauth_cred_get()

(christos)

use KAUTH_MACHDEP_UNMANAGEDMEM instead of KAUTH_GENERIC_ISSUSER like all the
other framebuffers.

(christos)

don't opencode kauth_cred_get()

(christos)

don't opencode kauth_cred_get()

(christos)

don't opencode kauth_cred_get()

(christos)

use l->l_cred since it is available.

(christos)

use l->l_cred since it is available.

(christos)

Remove FreeBSD blocks that did not compile anyway, and use l->l_cred
consistently.

(christos)

Fix C++ case for eventfd.h and timerd.h

(ryoon)

linux_sys_fstatat64: Add support for AT_EMPTY_PATH flag.

This is required for rtld shipped with glibc 2.33.

Now, userland of Fedora 34 is working fine on NetBSD/amd64.

(rin)

Fix the message that appears if you patch -R an unapplied patch.

(dholland)

Make _splraise/_spllower/splx functions available to modules again.

(jmcneill)

eventfd_destroy(): Add missing kmem_free().

(thorpej)

Use coherent DMA mappings when possible.

(jmcneill)

Use coherent DMA mappings when possible.

(jmcneill)

KNF, i.e. #include <sys/param.h> then #include <sys/types.h>

Fixes builds

(skrll)

Add a comment that .set_leds function won't work correctly on cngetc(9).

I'm afraid all other keyboard drivers that have .set_leds function
(like sgimips/dev/zs_kbd.c) have the same issue.
Maybe we needs 'polled' flag in .set_leds function, or an independent
.set_leds accessop in struct wskbd_consops for cons(9).

(tsutsui)

Update the status of the thorpej-futex branch.

(thorpej)

Note eventfd, timerfd, POSIX timers, and preadv()/pwritev() additions
to COMPAT_LINUX / COMPAT_LINUX32.

(thorpej)

Regen for preadv(2) / pwritev(2).

(thorpej)

Add preadv(2) and pwritev(2) system calls to COMPAT_LINUX and COMPAT_LINUX32.

(thorpej)

Regen for eventfd(2).

(thorpej)

Add the eventfd(2) system call to COMPAT_NETBSD32.

(thorpej)

Regen for timerfd.

(thorpej)

Add timerfd system calls to COMPAT_NETBSD32.

(thorpej)

Regen for eventfd.

(thorpej)

Add the eventfd system calls to COMPAT_LINUX and COMPAT_LINUX32.

(thorpej)