remove unused "_DIAGNOSTIC" option and opt_diagnostic.h.
note that this is unrelated to the widely used "DIAGNOSTIC" option.

(chs)

remove a const to allow building with QUEUEDEBUG.

(chs)

Instead of defining VCONS_DONT_READ as the raw value of 0x30, instead
define it as (VCONS_NO_COPYCOLS|VCONS_NO_COPYROWS) for clarify. NFC.

(jmcneill)

lint: remove custom stdbool.h for strict bool mode again

Adding it broke the official builds.

(rillig)

Remove MAXCPUS dependency.

(jmcneill)

make(1): fix a few inconsistencies for lint's strict bool mode

(rillig)

Add VND_COMPRESSION option

(jmcneill)

Add PPP_BSDCOMP, PPP_DEFLATE, PPP_FILTER options

(jmcneill)

add directory /usr/libdata/lint/strict-bool

This is needed for usr.bin/xlint/xlint.

(rillig)

Add UFS_ACL and UFS_EXTATTR options

(jmcneill)

Add QUOTA and QUOTA2 options

(jmcneill)

Add COMPAT_OSSAUDIO

(jmcneill)

lint: refactor strict bool mode and improve comments

The previous comment above typeok_strict_bool_compatible was too hard to
understand.

(rillig)

lint: in strict bool mode, perform the normal checks as well

(rillig)

lint: simplify typeok_strict_bool_compatible

(rillig)

lint: add missing warning for bool() == int

(rillig)

lint: simplify typeok_strict_bool_assign

When that function is called, the nodes are already before_conversion.

(rillig)

lint: remove redundant calls to before_conversion

(rillig)

lint: add test for missing check in strict bool mode

(rillig)

lint: replace integer constant expressions with true and false

LINTFLAGS=-gST make lint, with manual review.

The error messages from lint are all correct, they are not complete
though.  The return value of a function returning bool may still be
compared to the integer 0.

(rillig)

lint: in strict bool mode, integer constants do not have type bool

Previously, lint1 allowed integer constants such as 0 and 1 to be used
as bool constants.  This was only half-baked since after fixing all
error messages from that strict mode, there may still be integer
literals in the code that should be replaced with true or false.  This
would stop a migration from int to bool in the middle, leaving
inconsistent code around.

To find the remaining type inconsistencies, treat integers and bool as
completely incompatible, even for compile time constants.

(rillig)

trailing whitespace

(jmcneill)

ID_AA64PFR0_EL1.GIC=0 means that the CPU interface system registers are
not implemented. This does not necessarily mean that there is no GIC in the
system, as GICv2 uses MMIO instead of system registers for the CPU
interface.

While here, add description for GIC=3, which means that the v4.1 system
register interface is supported.

(jmcneill)

trailing whitespace

(jmcneill)

lint: prepare to make strict bool mode even stricter

Currently, strict bool mode still allows integer constant expressions to
be converted implicitly to bool.  This is something that other languages
such as Go, Java, C#, Pascal don't allow.

By providing a custom implementation of <stdbool.h> that defines false
and true to custom bool constant identifiers, lint will cover these
cases as well.

To prepare for this, reword the rules and restructure the tests in
d_c99_bool_strict.c.

(rillig)

lint: normalize whitespace in tests

(rillig)

oops, fix mistake in previous commit. removed unwanted member that had been added.

(ryo)

lint: replace 0 and 1 with false and true, where appropriate

Change in behavior: Passing the option -h exactly 4294967296 times or
any multiple thereof is no longer equivalent to passing it never at all,
it is now equivalent to passing it once.  See main2.c, hflag++ for the
actual change.

Other than that, no functional change intended.

A very large portion of the code already conformed to the requirements
of the strict bool mode.  The only missing thing was using the constant
literals false and true instead of 0 and 1.  For sure there are some
integer literals left that can be converted.  For now, all literals that
appeared in the form " = 0" or " = 1" have been replaced.

(rillig)

entropy: Record number of time and data samples for userland.

This more or less follows the semantics of the RNDGETESTNUM and
RNDGETESTNAME ioctls to restore useful `rndctl -lv' output.

Specifically: We count the number of time or data samples entered
with rnd_add_*.  Previously it would count the total number of 32-bit
words in the data, rather than the number of rnd_add_* calls that
enter data, but I think the number of calls makes more sense here.

(riastradh)

entropy: Sample cycle counter or timecounter in hardclock.

Only do so when we're short on entropy, in order to minimize
performance impact.

The sampling should stay close to the time of the actual hardclock
timer interrupt, so that the oscillator driving it determines when we
sample the cycle counter or timecounter, which we hope is driven by
an independent oscillator.

If we used a callout, there might be many other influences -- such as
spin lock delays possibly synchronized with this core's cycle counter
-- that could get between the timer interrupt and the sample.

In the glorious tickless future, this should instead be wired up to
the timer interrupt handler, however that manifests in the future
tickless API.

(riastradh)

Note that TPM 1.2 devices can attach to ACPI, as well.

(thorpej)

Enable tpm @ acpi (now that it can match TPM 1.2 devices, which are not,
as the comment implies, experimental).

(thorpej)

Match PNP0C31 as a TPM 1.2 device.  Works on my ThinkPad X260, and
eliminates the last of the devices that attach to "isa".

(thorpej)

Add missing printing-of-newlines in the attach routine.

(thorpej)

lint: fix LINTFLAGS for lint2

These have never worked since they report a syntax error in
<sys/cdefs.h> just because there is an __inline over there.

I wonder why it has been necessary at all to have custom LINTFLAGS for
lint itself.  It's just an ordinary program.

(rillig)

use fdtbus_intr_establish_xname

(jmcneill)

lint: merge duplicate code for non-zero detection

(rillig)

use fdtbus_intr_establish_xname

(jmcneill)

lint: add test for bool bit fields as operands of '|'

(rillig)

use fdtbus_intr_establish_xname

(jmcneill)

trailing whitespace

(jmcneill)

use fdtbus_intr_establish_xname

(jmcneill)

Add 'const char *xname' param to fdtbus_intr_establish_byname

(jmcneill)

use fdtbus_intr_establish_xname

(jmcneill)

use fdtbus_intr_establish_xname

(jmcneill)

use fdtbus_intr_establish_xname

(jmcneill)

lint: in strict bool mode, allow bool as operands of the comma operator

(rillig)

lint: demonstrate wrong message for comma operator

(rillig)

Install image built with all of MKDEBUG, MKKDEBUG, and MKDEBUGLIB all
set to "yes" is a bit larger, so bump the image size accordingly.

(pgoyette)

use fdtbus_intr_establish_xname

(ryo)

use fdtbus_intr_establish_xname

(ryo)

Ticket #1185

(martin)

Pull up following revision(s) (requested by knakahara in ticket #1185):

share/man/man9/pci_msi.9: revision 1.18

Fix PR kern/55920 .  Add pci_msi_count() and pci_msix_count() description.

(martin)

add fdtbus_intr_establish_xname() function

(ryo)

rnd(4): Fix formatting of authors paragraph with `.An -nosplit'.

(riastradh)

Tweak wording for consistency: `if empty or not set', not `if unset'.

(riastradh)

Clarify that `entropy' may be left unset in rc.conf.

(riastradh)

- Match 400-, 495-, and 500-series controllers.
- Apply OpenBSD rev 1.38, which ignores the SMBALERT# interrupt that
  fires spuriously on some systems (ignored by Linux and FreeBSD, as
  well).

(thorpej)

Improve error messages.
- prefix MD device name if it's considered to be related to the MD driver.
- revise some messages.

(isaki)

Change (harmless) zero-length debug messages.
sys/modules is compiled with -Wzero-length-format and this
makes sys/modules compilable even if AUDIO_DEBUG is defined.

(isaki)

Remove unused LIMITCHECK() macro (now contained in netbsd32_rlimit.c).

(simonb)

Add 'const char *xname' arg to fdtbus_interrupt_controller_func .establish

(jmcneill)

Initial attempt at documenting the KERNEL_DIR option.

XXX EXPERIMENTAL, Work-In-Progress ! XXX

(pgoyette)

Remove some blank lines to group the VND and FSS ioctls together.

(simonb)

lint: add tests for newly added messages for strict bool mode

(rillig)

Update sets lists to accomodate KERNEL_DIR.  When both KERNEL_DIR and
MKDEBUG are defined, we create an empty $DESTDIR/usr/libdata/debug/netbsd/
directory.

Should fix ``build.sh release'' issue reported in kern/55923

XXX Since nothing ever seems to populate this directory, perhaps we should
XXX simply prevent its creation?  If we do create it, should its creation
XXX perhaps be conditioned on MKKDEBUG rather than MKDEBUG?

XXX There is still another problem with ``build.sh install-image'' but I
XXX open a new PR for that.

(pgoyette)

/home/paul/XXX.txt

(pgoyette)

Use acpi_dsm() in acpi_usb_post_reset().

(thorpej)

In acpi_pci_ignore_boot_config(), use acpi_dsm() and add some comments
describing the (somewhat confusing) semantics of the _DSM method used
here.

(thorpej)

- Add a new routine, acpi_pack_compat_list(), that packs the _HID
  and _CID device ID strings into an OpenFirmware-like string list,
  analogous to how the "compatible" property is represented in
  OpenFirmware and the Device Tree specifications.
- Add generic routines for evaluating Device Specific Methods (_DSM).

(thorpej)

Handle the KFILTER_BYFILTER and KFILTER_BYNAME ioctls for compat32.
Passes "atf kernel/kqueue/t_ioctl".

(simonb)

Handle FSSIOCSET and FSSIOCGET; vndconfig(8) works with compat32 now.
XXX: FSSIOCSET50 and FSSIOCGET50 are not (yet) handled.

(simonb)

lint: add type details to message for enum type mismatch

Before December 2020, it was cumbersome to add type information to a
message since the caller had to explicitly allocate buffers for the type
names.  That's probably the reason why this crucial detail had been left
out of the warning.

(rillig)

lint: fix enum type in the lexical analysis part

Detected by Clang, reported by riastradh:

> scan.l:144:29: error: implicit conversion from enumeration type
>  'tspec_t' to different enumeration type 'op_t'
> return operator(T_ASTERISK, NOTSPEC);
>        ~~~~~~~~            ^~~~~~~

and by lint as well, with a less detailed and less helpful message:

> scan.l(144): warning: enum type mismatch, arg #2 [156]

Since scan.c is generated from scan.l, it is not included in the default
"make lint" though.

The value of these two constants is the same, furthermore that value is
never actually used in the code, therefore no functional change.

(rillig)

Add "TX " to "Queue No Descriptor Available" to make it more understandable.

(msaitoh)

Unmount and unconfigure in the reverse order that they were done.  Fixes
problem with leaving a dangling snapshot configured.

(simonb)

Handle rev 1.62 of kern/sys_lwp.c
  make _lwp_park return the remaining time to sleep in the "ts" argument
  if it is a relative timestamp, as discussed in tech-kern.
for compat32.  Thanks skrll@ for some cleanup tips.

(simonb)

Use netbsd32_* types instead of relying on __packed in the vnd ioctl
structures.  __packed didn't work on archs where 64-bit types were
64-bit aligned with a 32-bit ABI (eg MIPS n32).

OK mrg@

(simonb)

/dev/random: Fix nonblocking read.

The flag passed to cdev_read is IO_NDELAY, not FNONBLOCK/O_NONBLOCK,
and although the latter two coincide, IO_NDELAY has a different
value.

(riastradh)

entropy: Use a separate condvar for rndsource list lock.

Otherwise, two processes both waiting for entropy will dance around
waking each other up (by releasing the rndsource list lock) and going
back to sleep (waiting for entropy).

Witnessed on the armv7 testbed when /etc/security presumably ran
twice over a >day-long test, until the metaphorical plug got pulled:

net/if_ipsec/t_ipsec_natt (509/888): 2 test cases
    ipsecif_natt_transport_null: [ 37123.2631856] entropy: pid 1005 (dd) blocking due to lack of entropy
[256.523317s] Failed: atf-check failed; see the output of the test for details
    ipsecif_natt_transport_rijndaelcbc: [274.370791s] Failed: atf-check failed; see the output of the test for details
[532.486697s]
...
    puffs_lstat_symlink: [ 123442.1606517] entropy: pid 9499 (dd) blocking due to lack of entropy
[ 123442.1835067] entropy: pid 1005 (dd) blocking due to lack of entropy
[ 123442.1944600] entropy: pid 9499 (dd) blocking due to lack of entropy
[ 123442.1944600] entropy: pid 1005 (dd) blocking due to lack of entropy
...

(riastradh)

Fix for virtios viomb memory balloon driver. The inflate_done() and
deflate_done() issued a wrong extent to bus_dmamap_sync() giving rise to
panics on aarch64s mmio backend.

(reinoud)

Reset options on every loop.

(christos)

in pmap_writetext(), restore the context also when we return early.

(chs)

If an option is not found in the standard table, log it and don't try to
dereference it. This prevents crashes from XenServer VM that PXE boots and
includes option 175 in the DHCP request. Reported by Stephen Borrill.

(christos)

Improve English in comments

(skrll)

Improve english in a comment

(skrll)

skip sigbus_adraln on MIPS

(skrll)

Don't advertise AUDIO_PROP_CAPTURE - no capture method is provided

[  1.0000000] audio0 at vraiu0: missing capture method

(skrll)

entropy(7): Add missing URL to reference.

(riastradh)

threadpool(9): Tidy up thread naming.

- `dispatcher', not `overseer' -- much more appropriate metaphor.
- Just omit `/-1' from unbound thread names.
- Just omit `@-1' from dynamic-priority (PRI_NONE) thread names.

(riastradh)

threadpool(9): Make threadpool_percpu_ref_remote non-sleepable.

Needed for threadpool-based workqueue_enqueue to run in interrupt
context.

(riastradh)

lint: update usage for lint1 and lint2

(Forgotten in the previous commit.)

(rillig)

lint: add new check for strict bool mode

In strict bool mode, bool is considered incompatible with all other
scalar types, just as in Java, C#, Pascal.

The controlling expressions in if statements, while loops, for loops and
the '?:' operator must be of type bool.  The logical operators work on
bool instead of int, the bitwise operators accept both integer and bool.
The arithmetic operators don't accept bool.

Since <stdbool.h> implements bool using C preprocessor macros instead of
predefining the identifiers "true" and "false", the integer constants 0
and 1 may be used in all contexts that require a bool expression.
Except from these, no implicit conversion between bool and scalar types
is allowed.

See usr.bin/tests/xlint/lint1/d_c99_bool_strict.c for more details.

The command line option -T has been chosen because all obvious choices
(-b or -B for bool, -s or -S for strict) are already in use.  The -T may
stand for "types are checked strictly".

The default behavior of lint doesn't change.  The strict bool check is
purely optional.

An example program for strict bool mode is usr.bin/make, which has been
using explicit comparisons such as p != NULL, ch != '\0' or n > 0 in
most places for a long time now, even before the refactoring in 2020.

(rillig)

Typo in comment

(skrll)

Don't use arc4random for the tools build on NetBSD either.
compat_defs.h explicitly forces _POSIX_SOURCE and this hides the
prototype in stdlib.h.

(joerg)

Turn on DIAGNOSTIC to help track down problem in hpcmips test runs

(skrll)

Fix PR kern/55920 .  Add pci_msi_count() and pci_msix_count() description.

(knakahara)

Trailing whitespace

(skrll)

Fix spelling of preferred in comments.  While here improve grammar in
one of the comments.

(skrll)

Improve a comment

(skrll)

Correct the spelling of preferred.

(skrll)

Make this compile without -DINET6 (Kurt Schreiner)

(christos)

Only try to create up-to NAME_MAX filenames.
XXX: this should be moved to stdio/t_mktemp.c where the rest of the tests
are and the code lives.

(christos)

lint: split typeok into several functions

(rillig)

lint: sprinkle a few const modifiers throughout the code

(rillig)

Trailing whitespace

(skrll)

Small simplification

(skrll)

Tidy up more error handling confusion

(skrll)

KNF

(skrll)

Trailing whitespace

(skrll)

For "ps/a" print all of p_comm instead of just the first 10 chars, add a
space after PID, left justify the command name.

(simonb)

PR/55715: pmax testbed panics with "assertion "asid == curcpu()->ci_pmap_asid_cur" failed"

It's GXemul that has the bug! Unfortunately, there's no way (currently) to
detect if we're running under GXemul emulation, so disable for all mips
for now.  Hopefully, GXemul will get fixed soon.

(skrll)

in pgt_page_alloc(), wait and retry the page allocation if PR_WAITOK.
fixes PR 55895.

(chs)

entropy: Downgrade consolidation warning from printf to LOG_DEBUG.

Candidate fix for PR kern/55458.  This message is rather technical,
and so is unlikely to be acted on by anyone not debugging the kernel
anyway.  Most likely, on any system where it is a real problem, there
will be another (less technical) entropy warning anyway.

(riastradh)

lint: extend test for strict _Bool handling

There is no danger in allowing (flags & FLAG) as a controlling
expression, provided that it is immediately compared to zero, such as in
an if statement or as the operand of a logical operator.

(rillig)

make(1): make a few more bool expressions more precise

The previous version of lint(1) from a few hours ago didn't catch all
occurrences.  And even the current one doesn't catch everything.
Function arguments and return types still need some work.  The "return
quietly" from shouldDieQuietly still implicitly converts from int to
_Bool.

No functional change.

(rillig)

Various entropy integration improvements.

- New /etc/security check for entropy in daily security report.

- New /etc/rc.d/entropy script runs (after random_seed and rndctl) to
  check for entropy at boot -- in rc.conf, you can:

  . set `entropy=check' to halt multiuser boot and enter single-user
    mode if not enough entropy

  . set `entropy=wait' to make multiuser boot wait until enough entropy

  Default is to always boot without waiting -- and rely on other
  channels like security report to alert the operator if there's a
  problem.

- New man page entropy(7) discussing the higher-level concepts and
  system integration with cross-references.

- New paragraph in afterboot(8) about entropy citing entropy(7) for
  more details.

This change addresses many of the issues discussed in security/55659.
This is a first draft; happy to take improvements to the man pages and
scripted messages to improve clarity.

I considered changing motd to include an entropy warning with a
reference to the entropy(7) man page, but it's a little trickier:
- Not sure it's appropriate for all users to see at login rather than
  users who have power to affect the entropy estimate (maybe it is,
  just haven't decided).
- We only have a mechanism for changing once at boot; the message would
  remain until next boot even if an operator adds enough entropy.
- The mechanism isn't really conducive to making a message appear
  conditionally from boot to boot.

(riastradh)

lint: add enum flags to the test about strict boolean mode

(rillig)

lint: allow custom flags to be specified in all unit tests

Previously, only the tests called msg_*.c could use this feature.  Since
it is useful for other tests as well, enable it everywhere.

(rillig)

make(1): consistently use boolean expressions in conditions

Most of the make code already followed the style of explicitly writing
(ptr != NULL) instead of the shorter (ptr) in conditions.

The remaining 50 instances have been found by an experimental,
unpublished check in lint(1) that treats bool expressions as
incompatible to any other scalar type, just as in Java, C#, Pascal and
several other languages.

The only unsafe operation on Boolean that is left over is (flags &
FLAG), for an enum implementing a bit set.  If Boolean is an ordinary
integer type (the default), some high bits may get lost.  But if Boolean
is the same as _Bool (by compiling with -DUSE_C99_BOOLEAN), C99 6.3.1.2
defines that a conversion from any scalar to the type _Bool acts as a
comparison to 0, which cannot lose any bits.

(rillig)

PR/55715: pmax testbed panics with "assertion "asid == curcpu()->ci_pmap_asid_cur" failed"

disable the sigbus_adraln test on qemu for now

(skrll)

rc.subr(8): Fix markup -- use .It to set off a new item.

(riastradh)

Disable nouveau for now. It can cause long pauses at boot and sometimes
fails to initialize the display at all.

(jmcneill)

lint: split check_global_variable into several functions

This reduces the indentation and allows the messages in the comments to
contain the complete text instead of only an abbreviated prefix.

No functional change.

(rillig)

lint: fix assertion failure for syntax error in enum declaration

(rillig)

lint: add test for triggering assertion failures in lint1

(rillig)

lint: add test for treating _Bool as non-scalar type

This strict mode is not yet implemented.  The plan is to use it for
usr.bin/make, to get rid of the many possible variants of defining the
Boolean type in make.h.  These variants did find some bugs, but not
reliably so.  Using static analysis seems more promising for this.

In an early stage of developing this test, lint1 crashed in the enum
definition in line 213, where the node for the '?:' had been NULL.  This
can happen in other situations as well, such as with syntax errors, but
these should be rare, as lint is usually only run if the compiler has
accepted the source code.  Still, there should not be any assertion
failures while running lint1.

(rillig)

Add whitespace after comma

(ryoon)

lint: remove redundant parentheses around return value

(rillig)

lint: change return type of eqptrtype to bool

(rillig)

lint: rename functions with short names

(rillig)

lint: use message 232 for unused label in function

This message had been defined but not used before.

(rillig)

Explicitly cast offset calculation, size_t may be smaller than off_t.

(martin)

Fix a typo in debug message

(ryoon)

lint: explain missing error for _Bool constant expressions

(rillig)

lint: fix conversion of non-constant scalar to _Bool

(rillig)

fstyp: Sync HAMMER1/2 with DragonFly BSD

taken-from: DragonFly BSD

(tkusumi)