Fix previous; define missing lfs_version.

(rin)

Various clean ups
- cache node where possible
- remove a compiler warning
- improve a comment (the round_page might not be needed)

(skrll)

ufs_open(): Check fs->lfs_version ifdef LIBSA_*L*FS, not LIBSA_*F*FS.

This was harmless for FFS variants, that define LIBSA_FFSv[12], not LIBSA_FFS.

(rin)

Use sel{record,remove}_knote().

(thorpej)

malloc(9) -> kmem(9)

(thorpej)

make(1): clean up ParseRawLine

(rillig)

make(1): extract ParseRawLine from ParseGetLine

(rillig)

make(1): clean up another local variable in ParseGetLine

(rillig)

make(1): clean up UnescapeBackslash

(rillig)

make(1): remove unused parameter from UnescapeBackslash

(rillig)

Merge pkg_install-20201218

(maya)

make(1): split ParseGetLine into separate functions

(rillig)

make(1): separate ParseGetLine into paragraphs

(rillig)

make(1): support using C99 bool for Boolean

(rillig)

Merge pkg-install-20201218

(maya)

make(1): spell nonexistent consistently

(rillig)

While we are inside ddb, allow all memory access.

(martin)

make(1): use symbolic time for 0 in Make_Recheck

This makes the test depsrc-optional independent from the current time
zone.

(rillig)

make(1): constify Targ_FmtTime

(rillig)

Tickets #1153 and #1154

(martin)

Apply patch, requested by roy in ticket #1154:

external/bsd/dhcpcd/dist/src/dhcp.c (apply patch)

Fix reading of UDP messages via inet socket rather than BPF.
This fixes RENEW messages not being read correctly.

(martin)

Pull up following revision(s) (requested by mrg in ticket #1153):

usr.bin/vmstat/vmstat.c: revision 1.232 (patch)

move the time nlist fetches into their own namelist and only
fetch them when necessary.  allow for fallback uses of older
time sources if others are not present.

this stops vmstat from exiting if it can't get the addresses
of these time values it often doesn't need (eg, running kernels
use the sysctl method), which has cropped up recently wit the
removal of boottime variable.

a slighly modified version of this patch (modified to handle
the old boottime variable over the new one) works against a
netbsd-9 vmstat in -current too.
XXX: pullup

(martin)

Make this compile on arm

(skrll)

Add a comment

(skrll)

Make EFI_RUNTIME compile on arm

(skrll)

KNF

(skrll)

Use sel{record,remove}_knote().

(thorpej)

Use sel{record,remove}_knote().

(thorpej)

Use sel{record,remove}_knote().

(thorpej)

Minor typo - do I really need to bump the date for this?  :)

(pgoyette)

Use sel{record,remove}_knote().

(thorpej)

Use sel{record,remove}_knote().

(thorpej)

Use sel{record,remove}_knote().

(thorpej)

Use sel{remove,record}_knote().

(thorpej)

Use sel{record,remove}_knote().

(thorpej)

Use sel{record,remove}_knote().

(thorpej)

Use sel{record,remove}_knote().

(thorpej)

Update information about thorpej-futex.

(thorpej)

remove unused and #if 0'd since introduced in 1996 struct kvm_cpustate.
sparc kvm was handled differently for the sun4m port.

(mrg)

No need for PMAP_WRITE_BACK here (also aa32 pmap doesn't know about it)

(skrll)

Change second gc.interval_min to gc.interval_max.

Obviously, second gc.interval_min should be gc.interval_max.

Should be pulled up to netbnsd-9.

(taca)

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

(thorpej)

Add suppport for Linux preadv(2) and pwritev(2).  Same as native, except
for the silly way that the offset argument is passed in.

(thorpej)

Regen for POSIX timers and timerfd.

(thorpej)

Add POSIX timer and timerfd support to COMPAT_LINUX32.

(thorpej)

Remove a redundant prototype.

(thorpej)

Add sigevent mapping for Linux-32 compat.

(thorpej)

Re-factor the code that maps the clockid_t in timer_create() and
the flags in timerfd_settime() into separate functions.

(thorpej)

interupt -> interrupt

(christos)

Regen for eventfd.

(thorpej)

Add eventfd glue to COMPAT_LINUX32.

(thorpej)

-O0 hack was reintroduced to jemalloc/rtree.c for GCC9 on alpha.

(rin)

For alpha, apply -O0 hack again to rtree.c for GCC9.
Otherwise, GDB11 crashes in strange ways.

(rin)

Correct and extend comments on VA for alpha. No binary changes.

(rin)

make(1): format function definitions consistently

(rillig)

make(1): indent nonints.h and util.c with tabs instead of spaces

(rillig)

make(1): indent make.c with tabs instead of spaces

(rillig)

make(1): extract MakeChildren from MakeStartJobs

(rillig)

make(1): indent job.h with tabs instead of spaces

(rillig)

make(1): indent for.c using tabs instead of spaces

(rillig)

make(1): clean up hash function for HashTable

Expressing a multiplication as a bit shifts and subtraction is the job
of the compiler.  For humans, a multiplication is easier to read.

(rillig)

Regen for eventfd, timerfd, POSIX timers.

(thorpej)

Add Linux eventfd, timerfd, and POSIX timer calls.

(thorpej)

it's cpu_switchto (not cpu_switch)

(skrll)

Fixup the big mutex_exit comment a little. It's actually, cpu_switchto
that posts a store fence and it's AFTER setting curlwp.

More to come.

(skrll)

Make comments are not C comments - use # instead of /* ... */

(martin)

make(1): indent hash.h and make_malloc.h with tabs instead of spaces

(rillig)

make(1): document how to detect typos in .elif directives

(rillig)

make(1): clean up ParseReadLine

In function names, the word "get" was not used consistently to look up
or compute data, in several cases "get" was a synonym for "read", just
like in the standard C library (fgetc).

The really confusing part is that there are two functions now, called
ParseGetLine and ParseReadLine, and both were underdocumented.

(rillig)

when updating the per-uid "semcnt", decrement the counter for the uid
that created the ksem, not the uid of the process freeing the ksem.
fixes PR 55509.

Reported-by: syzbot+9d04b3ef2ca180ef9b06@syzkaller.appspotmail.com

(chs)

make(1): error out on .else with extraneous text

(rillig)

make(1): error out if an '.endif' contains extraneous text

(rillig)

make(1): make structure of the code in Cond_EvalLine clearer

(rillig)

make(1): add tests for parsing .if conditions

(rillig)

make(1): add more tests for parsing .endif

(rillig)

make(1): don't pretend to expand CondParser_Eval

At that point, the variable expression has already been expanded.  To
avoid the impression that the token might be relevant, pass FALSE
instead of TRUE.  No change of behavior.

(rillig)

make(1): add test for variable expressions after a parse error

These variable expressions don't need to be expanded, and they aren't.

(rillig)

make(1): add test for parsing the end of a condition

(rillig)

Trailing whitespace

(skrll)

Add a note about completion vs ordering barrier as well.

(skrll)

Add a big comment in generic_bs_barrier about mappings and what barriers
are really required and why we cheat.  Inspired by a similar comment in
x86/bus_space.c

(skrll)

Tickets #1151 and #1152#

(martin)

Pull up following revision(s) (requested by tsutsui in ticket #1152):

sys/arch/hp300/conf/RAMDISK: revision 1.16
sys/arch/hp300/conf/INSTALL: revision 1.66

Remove options (NVNODE, NBUF, and BUFPAGES) for small RAM machines.

These options make extracting binary sets much slower even on miniroot.

Instead, use "options BUFCACHE=5" (use 5% of memory for buffercache) and
also specify "-fno-unwind-tables" for COPTS to shrink INSTALL kernel.

Worth to pullup to netbsd-9.

(martin)

Pull up following revision(s) (requested by tsutsui in ticket #1151):

distrib/miniroot/install.sub: revision 1.54
distrib/miniroot/install.sub: revision 1.55
distrib/miniroot/install.sub: revision 1.56
distrib/miniroot/install.sub: revision 1.57
distrib/miniroot/install.sub: revision 1.58
distrib/miniroot/install.sub: revision 1.59
distrib/hp300/miniroot/install.md: revision 1.20
distrib/miniroot/list: revision 1.36
distrib/mvme68k/miniroot/install.md: revision 1.9
distrib/mvme68k/miniroot/install.md: revision 1.10
distrib/sun3/miniroot/install.md: revision 1.6
distrib/sun3/miniroot/install.md: revision 1.7
distrib/sun2/miniroot/install.md: revision 1.6
distrib/sun2/miniroot/install.md: revision 1.7
distrib/amiga/miniroot/install.md: revision 1.31
distrib/mac68k/miniroot/install.md: revision 1.6
distrib/mac68k/miniroot/install.md: revision 1.7
distrib/mac68k/miniroot/install.md: revision 1.8
distrib/hp300/miniroot/install.md: revision 1.18
distrib/hp300/miniroot/install.md: revision 1.19
distrib/miniroot/install.sub: revision 1.53

test ! -n "$foo"  is just a quaint way of saying test -z "$foo"
and test ! -z "$foo" is really just test -n "$foo"  so let's just
use the simple (and more obvious) forms.

NFCI.

-

Add modules set for default sets on miniroot installation.
Should be pulled up to netbsd-9 and netbsd-8.

-

The kernel and X sets should be specified in MD miniroot scripts.

Otherwise they are not listed at least on ftp installation.

The problem was reported by Martin Trusler on port-hp300@:
https://mail-index.netbsd.org/port-hp300/2020/11/21/msg000174.html

Should be pulled up to netbsd-9 and netbsd-8.

-

Appease awk warnings on suppressing dmesg timestamps.
awk: warning: escape sequence `\[' treated as plain `['
source line number 1
context is
{ h=$0; >>>  gsub("^[ <<<
awk: warning: escape sequence `\]' treated as plain `]'
source line number 1

Should be pulled up to netbsd-9.

-

Explicitly sort set names fetched via ftp nlist.

Several binary sets are stored as symbolic links on releases and
it seems some ftpd doesn't sort nlist outputs by name in such case.

Worth to pullup to netbsd-9 and netbsd-8.

-

Use proper release version strings ("9.1" rather than "91") in banners.

Also define and use "MACHINE" variable to describe port names
(no uname(1) or sysctl(8) in miniroot binary list by default).

I guess the short format like "91" by ${DISTRIBREV} was used only
for split sets for floppies in 1990's releases.

Worth to pullup to netbsd-9.

-

Inform the default installation directory in the official ftp server.

-

Fetch files via ftp using auto-fetching with URL per each binary set.

On slower machines, it takes more than five minutes to get a large
set binary and it could cause timeout of ftp control session, so
getting multiple binary sets in a single ftp session always fails.

Briefly tested on HP 9000/425e with 9.1 tree and ftp.netbsd.org binaries.

No particular comments on tech-install@ and port-hp300@.

Maybe should be pulled up to netbsd-9.

-

Add missing md_get_partition_range() that causes installation failure.

Also reported by Martin Trusler on testing hp300 miniroot.

Should be pulled up to netbsd-9 and netbsd-8.

(martin)

Unit tests for timerfd.

(thorpej)

Regen for timerfd.

(thorpej)

Native implementation of the Linux timerfd API.

(thorpej)

Unit tests for eventfd(2).

(thorpej)

Regen for eventfd(2).

(thorpej)

Native implementation of the Linux eventfd(2) API.

(thorpej)

Sync w/ HEAD.

(thorpej)

Initialize namelen

(martin)

Use full system DSB ops for bs barrier.

(jmcneill)

The bus_space(9) man page is not clear whether barriers should enforce
ordering or completion. To be safe, use dsb here instead of dmb.

(jmcneill)

t_ossaudio: More descriptive failure cases

(nia)

MAsk -> Mask, spotted by rillig@

(jmcneill)

Add some precautions for platforms without PT_STEP support.

"nat/gdb_ptrace.h" defines PT_STEP as 9, if it is not defined. nat-ptrace.c
depends on this; inf_ptrace_target::resume() uses PT_STEP unconditionally
when its ``step'' argument is non-zero. Therefore,

- Add comment that nbsd-nat.c should include <sys/ptrace.h> directly,
  instead of "nat/gdb_ptrace.h".

- Add gdb_assert(step == 0) in nbsd_nat_target::resume() ifndef PT_STEP,
  before calling inf_ptrace_target::resume().

(rin)

Revert previous as per request from kamil; now, arm/ptrace.h does not
expose unimplemented PT_STEP.

(rin)

Obsolete PT_STEP and friends, that have never been implemented.

PT_STEP was exposed to userland for NetBSD 1.6 to 9, and
PT_SETSTEP and PT_CLEARSTEP was for 8 to 9. Therefore, they cannot be
reused for other purposes.

PT_STEP was introduced to arm/ptrace.h rev 1.2:

http://cvsweb.netbsd.org/bsdweb.cgi/src/sys/arch/arm/include/ptrace.h#rev1.2

This was for workaround against GDB, which assumes PT_STEP is 9 (even
today!!), if it is undefined. But, this should have been dealt with
differently...

(rin)

PR/55875 - Valentin -- "gpt recover -r" does nothing

Delete do nothing undocumented option, that was there from initial
import.  No idea what it was intended to do and there is no longer
an "upstream".

(jnemeth)

make(1): replace %zu with %u in printf calls

This is needed to compile bmake with GCC 2.8.1 on SunOS 5.9.

To support ancient systems like this, the whole code of usr.bin/make is
supposed to use only ISO C90 features, except for filemon, which is not
used on these systems.

(rillig)

make(1): rename Vector.priv_cap to cap

There is no use case for accessing or even modifying the capacity of a
vector, therefore there is no need to hide it using the prefix "priv_".
This way, the member names are aligned between Buffer and Vector.

(rillig)

SBBR says that the CPPC interface should be used on AArch64 instead of
P-states and T-states. Enable acpipcc and acpicppc, and disable acpicpu.

(jmcneill)

acpi: Add support for Collaborative Processor Performance Control (CPPC).

CPPC is the modern replacement for processor performance controls (P-states
and T-states). The _CPC method on a processor device contains information
about supported performance levels of the device.

This driver reports the supported performance levels and allows for
setting the desired level using the machdep.cpufreq.cpuN.* sysctls. Note
that a "performance level" value is not guaranteed to represent a CPU
frequency, so treat the numbers reported as such.

(jmcneill)

acpi: Add support for ACPI Platform Communications Channel (PCC).

(jmcneill)

Add MI support for attaching drivers to arbitrary System Description
Tables.

(jmcneill)

make(1): add str_basename to reduce duplicate code

The function basename from POSIX has a few unfortunate properties, it is
allowed to return a pointer to static memory.  This is too unreliable,
therefore this trivial own implementation.

(rillig)

make(1): constify progname

(rillig)

make(1): fix .ERROR_TARGET in compat -k mode (since 2010-04-07)

(rillig)

Disable use of UBC_FAULTBUSY in tmpfs_write() for now,
which brings back zeroing of all new tmpfs data pages.
The existing code that enables this optimization skips the zeroing
in numerous cases where it is needed, resulting in corrupted files
and data leaks from the page's previous identity.

(chs)

make(1): add tests for variable expansion in .ERROR handling

(rillig)

make(1): document variable expansion in the .END node

(rillig)

fix comment

(christos)

make(1): demonstrate wrong error handling in compat mode

(rillig)

Improve previous: generated names end with .[[:digits:]] so look for that
instead.

(christos)

make(1): add comments for strange error handling in Compat_Run

(rillig)

make(1): extract InitSignals from Compat_Run

(rillig)

make(1): extract MakeBeginNode from Compat_Run

The comment "execute the commands" had once been correct but not
anymore.  Since a few years, not only the commands of the .BEGIN and
.END nodes are executed, instead the nodes are made as usual, including
their dependencies.

(rillig)

make(1): extract UseShell from Compat_RunCommand

(rillig)

Add atppc* at acpi?

(martin)

Add missing include

(martin)

Revise comments.

(isaki)

Fix a return value of audiopoll().
fo_poll is expected to return revents rather than errno on error.

(isaki)

Avoid a dead lock in audiodetach, since rev 1.63.
audio_unlink() must be called without exlock held (and
audio_mixer_destroy() must be called with exlock held).
This makes unplugging during playing/recording work (again).
Reported by Julian Coleman on current-users:
http://mail-index.netbsd.org/current-users/2020/12/10/msg040050.html

(isaki)

Add missing newline.

(isaki)

Prune job debug output that may appear out of order.

A race between child and parent means that we cannot
guarantee whether all child output is seen before we call
JobClosePipes, thus intervening debug output can appear
before or after the last child output.

(sjg)

make(1): remove dead code from GetVarnamesToUnexport

Now that the parsing of the directives is unified and strict, there is
no need anymore for the dispatched functions to check for unknown
directives.  These functions don't even get the information to decide
that since this decision is already done.

(rillig)

Avoid noise when csh does not exist

(sjg)

Ensure we have a value for .MAKE.UID

(sjg)

make(1): replace *line with line[0]

Since a line is not an iterator and since the expression *line typically
means "the current element", not "the first character", replacing *line
with line[0] more directly expresses the idea of accessing the first
character of a string.

(rillig)

make(1): clean up comment for ParseDirective

(rillig)

make(1): clean up Var_Export

(rillig)

make(1): replace bitset VarExportFlags with enum VarExportMode

The previous flags were not combined at all.

(rillig)

make(1): adjust timestamps in unit tests

These could not be known before the previous commit.

(rillig)

make(1): error out on misspelled directives

Before, make accepted misspellings like .warnings, .export-literally and
a few others, all of which are unlikely to occur in practice.  See the
test directive-misspellings.mk for further details.

(rillig)

make(1): add test for misspelled directives

This test allows the other directive-* tests to focus on the purpose of
the individual directive, allowing these tests to continue after
parsing, without errors.

(rillig)

t_ossaudio was installed for clang, so remove gcc marking

(roy)

diffutils: execl requires a NULL sentinel

(roy)

make(1): reduce indentation of ParseDirective

(rillig)

make(1): eliminate boolean argument of Var_Export

(rillig)

Ticket #1637

(martin)

Pull up following revision(s) (requested by skrll in ticket #1637):

sys/dev/usb/ohci.c: revision 1.311
sys/dev/usb/ohci.c: revision 1.312
sys/dev/usb/ohcivar.h: revision 1.62

Restructure the abort code for TD based transfers (ctrl, bulk, intr).

In PR kern/22646 some TDs can be on the done queue when the abort start
and, if this is the case, they need to processed after the WDH interrupt.
Instead of waiting for WDH we release TDs that have been touched by the
HC and replace them with new ones.  Once WDH happens the floating TDs
will be returned to the free list.

Also addresses the issue seen in PR kern/55835

Thanks to both Andreas Gustafsson and Edgar Fu=C3=9F for testing.  Apologi=
es to
Andreas Gustafsson for not committing this to HEAD for 4y6m.w

Remove leading space that crept in in the last change

(martin)

Ticket #1150

(martin)

Pull up following revision(s) (requested by skrll in ticket #1150):

sys/dev/usb/ohci.c: revision 1.311
sys/dev/usb/ohci.c: revision 1.312
sys/dev/usb/ohcivar.h: revision 1.62

Restructure the abort code for TD based transfers (ctrl, bulk, intr).

In PR kern/22646 some TDs can be on the done queue when the abort start
and, if this is the case, they need to processed after the WDH interrupt.
Instead of waiting for WDH we release TDs that have been touched by the
HC and replace them with new ones.  Once WDH happens the floating TDs
will be returned to the free list.

Also addresses the issue seen in PR kern/55835

Thanks to both Andreas Gustafsson and Edgar Fu=C3=9F for testing.  Apologi=
es to
Andreas Gustafsson for not committing this to HEAD for 4y6m.w

Remove leading space that crept in in the last change

(martin)

make(1): extract ExportVarsExpand from Var_Export

(rillig)

make(1): fix undefined behavior when exporting ${:U }

(rillig)