Sweep unncessary radix.h inclusions

(ozaki-r)

Add serial commas. Fix minus. Sort SEE ALSO. Fix xref.

(wiz)

Bump libutil for new pidfile functions.

(roy)

Remove out-dated comments and unnecessary splsoftnet for pool_{get,put}

(ozaki-r)

more build fixes

(skrll)

Fix non UHCI_DEBUG build

(skrll)

Don't use USB_IS_SS when there's a switch (speed) to use.

(skrll)

Add psref to rump kernel

(ozaki-r)

Fix usage of pslist(9)

Pointed out by riastradh@.

(ozaki-r)

Move #include <sys/pslist.h> inside #ifdef _KERNEL for building brconfig

(ozaki-r)

Cause readers to fail noisily after PSLIST_ENTRY_DESTROY.

Using NULL only causes readers to stop iteration, because it looks
like the end of the list; using a bogus non-null pointer causes them
to crash.

(riastradh)

Remove incorrect comment about non-effect of PSLIST_ENTRY_DESTROY.

This does have an effect -- it nullifies the next pointer.  (The text
was copied from PSLIST_DESTROY, which really has no effect without
options DIAGNOSTIC.)

(riastradh)

Use pslist(9) in bridge(4)

This adds missing memory barriers to list operations for pserialize.

(ozaki-r)

Don't call pfxlist_onlink_check with holding llentry lock

From FreeBSD (as of 2016-04-11).

Should fix PR kern/51060.

(ozaki-r)

Char -> wchar_t from Ingo Schwarze.

(christos)

there is no pflags, just flags.

(christos)

more macro WIDECHAR undoing from Ingo Schwarze.

(christos)

Apply revision 274569 from FreeBSD:
Only compare visitation counters if they've both been set for the current
type graph walk.

(christos)

Simplify xhci_do_command*

(skrll)

Mark ./usr/libdata/debug/usr/bin/pcpp.debug as obsolete

pcpp.debug is already in ./usr/libdata/debug/usr/libexec/

This fixes build with MKPCC=yes and all debug flags turned on.

(kamil)

Remove duplicate entry of /usr/libdata/debug/usr/bin/p++.debug

I don't have p++.debug locally so assume that it's properly set for the
first time as obsoleted.

(kamil)

route is a uint32_t in xhci_init_slot so pass it as such

(skrll)

Note import of dhcpcd-6.10.2

(roy)

Note addition of new pidfile functions.

(roy)

dhcpcd now uses libutil.

(roy)

Sync

(roy)

Free DMA buffer after disabling slot.
Fix bug DCBA was not cleared after disabling slot.

From Takahiro HAYASHI

(skrll)

Implement pidfile_lock, pidfile_read and pidfile_clean.

Discussed on tech-net@, ok core@.

(roy)

Fix xref section.

(wiz)

Fix set_port_feature U[12]_TIMEOUT to work when speed > 4.

From Takahiro HAYASHI

(skrll)

Fix a DPRINTF where the route string and roothub port were swapped.

From Takahiro HAYASHI

(skrll)

Remove variable v0 becasue ~XHCI_PS_CLEAR does not mask speed bits.

From Takahiro HAYASHI

(skrll)

Fix assertion of upper limit of slot index.

From Takahiro HAYASHI

(skrll)

Fix lock assertion in xhci_softintr() when entering DDB on ukbd.
Move this KASSERT to xhci_softintr() from xhci_handle_event().

From Takahiro HAYASHI

(skrll)

Unwrap lines and other whitespace.

From Takahiro HAYASHI

(skrll)

Revert 1.126.2.12 partially: fix some devices ware not recognized.
uhub would give up enumeration too quickly.

From Takahiro HAYASHI

(skrll)

default uhubdebug to 0 again

(skrll)

GC PAX_ASLR_EXEC

(christos)

Fix a debug message: sync with other device name expressions.

From Takahiro HAYASHI

(skrll)

Add a comment

(skrll)

We don't need TRAP_SIGDEBUG

(christos)

Fix xpipe->xp_pipe was xpipe->pipe.

From Takahiro HAYASHI

(skrll)

Enable PIE binaries on amd64 by default.

(christos)

Fix merge botch

(skrll)

- Turn on ASLR by default
- Add PAX_MPROTECT_DEBUG flag

(christos)

Fix ASLR stack setup:
- make it work on machine where the stack grows up.
- use the same limits as mmap.

(christos)

Add a flags sysctl for ASLR to disable particular features. We seem to be
having issues with stack setup.

(christos)

Note tickets #1076

(martin)

Note tickets #1076 and #1141

(martin)

Catch up to -current (via patch), requested by mspo in #1141:

libexec/httpd/CHANGES                          up to 1.21
libexec/httpd/Makefile                        up to 1.26
libexec/httpd/auth-bozo.c                      up to 1.18
libexec/httpd/bozohttpd.8                      up to 1.58
libexec/httpd/bozohttpd.c                      up to 1.79
libexec/httpd/bozohttpd.h                      up to 1.44
libexec/httpd/cgi-bozo.c                      up to 1.32
libexec/httpd/content-bozo.c                  up to 1.13
libexec/httpd/daemon-bozo.c                    up to 1.17
libexec/httpd/dir-index-bozo.c                up to 1.25
libexec/httpd/lua-bozo.c                      up to 1.14
libexec/httpd/main.c                          up to 1.13
libexec/httpd/netbsd_queue.h                  up to 1.1
libexec/httpd/printenv.lua                    up to 1.3
libexec/httpd/ssl-bozo.c                      up to 1.22
libexec/httpd/tilde-luzah-bozo.c              up to 1.14
libexec/httpd/testsuite/Makefile              up to 1.5
libexec/httpd/testsuite/test-bigfile          up to 1.2

Import bozohttpd 20151028:
o  add CGI support for ~user translation (-E switch)
o  add redirects to ~user translation
o  fix bugs around ~user translation
o  add schema detection for absolute redirects
o  fixed few memory leaks
o  bunch of minor tweaks
o  removed -r support
o  smarter redirects
Changes in 20150320:
o  fix redirection handling
o  support transport stream (.ts) and video object (.vob) files
o  directory listings show correct file sizes for large files

(martin)

Sync with HEAD (KERNHIST_DUMP fix)

(skrll)

Sort sections. Use more markup.

(wiz)

Pull up following revision(s) (requested by christos in ticket #1076):
usr.sbin/vnconfig/vnconfig.8: revision 1.41
usr.sbin/vnconfig/vnconfig.c: revision 1.43-1.44
Since now the kernel returns ENXIO again for devices exceeding the
number of devices configured in the kernel, use that instead of
scanning /dev.
Cosmetic tweaks to vncfonfig -l output.
This should restore compatibility for old scripts
trying to find free vnodes in the new cloning world order.

(martin)

Pull up following revision(s) (requested by christos in ticket #1076):
usr.sbin/vnconfig/vnconfig.8: revision 1.41
usr.sbin/vnconfig/vnconfig.c: revision 1.43-1.44
Since now the kernel returns ENXIO again for devices exceeding the
number of devices configured in the kernel, use that instead of
scanning /dev.
Cosmetic tweaks to vncfonfig -l output.
This should restore compatibility for old scripts
trying to find free vnodes in the new cloning world order.

(martin)

Cosmetic tweaks to vncfonfig -l output, from Robert Elz (with minor tweaks
[and all bugs] by me). This should restore compatibility for old scripts
trying to find free vnodes in the new cloning world order.

(martin)

Don't call pfxlist_onlink_check with holding llentry lock

Sync nd6_free with FreeBSD (as of 2016-04-10).

Should fix PR kern/51056.

(ozaki-r)

Macro parentheses police!

One *hopes* this has no functional change.

(riastradh)

Use && to fail early if subcommands of kernel link fail.

(Remind me why we string together shell commands instead of just
using make targets?)

(riastradh)

Change some 0's to NULL's from Pedro Giffuni

(christos)

More WIDECHAR elimination (Ingo Schwarze)

(christos)

Parrot markup for PTHREAD_MUTEX_INITIALIZER synopsis.

(riastradh)

Use the proper pluralization of `frotz'.

Per the jargon file, `frotz' is a contraction of `frobnitz', which
pluralizes to `frobnitzem'; `frobbotzim' is the plural of the related
but not identical term `frobozz'.

Fixing other uses of this plural elsewhere among the man pages is
left to the reader inclined to humorous linguistic studies.

(riastradh)

fix test.

(christos)

properly check for type id overflow

(christos)

Say `loop header', not `loop head'.

`Head' means something else in this context -- a list head.

(riastradh)

Fix a vestige of an earlier word choice: s/connected/associated/1

(riastradh)

Fix wording around `list heads'.

A list head is not a list, conceptually.

(riastradh)

Tighten opening paragraph.  Split into two smaller sentences.

(riastradh)

Add missing argument to kernhist_dump.

Hi nick!

(riastradh)

Fix the ABI of idtype_t.

(riastradh)

Don't pollute <sys/condvar.h> with <sys/mutex.h>.

Instead, forward-declare struct kmutex, since we only need to
describe pointer types to it.

This breaks an include cycle on several architectures:

sys/cpu_data.h
-> sys/condvar.h
-> sys/mutex.h
-> sys/intr.h
-> machine/intr.h
-> machine/cpu.h
-> sys/cpu_data.h

(riastradh)

Add autounload logic to dtrace module prototype.

(riastradh)

Disable autoload if there are any profile probes.

(riastradh)

Use imperative mood consistently.

(riastradh)

Fix missing NAME operand to PSLIST_ENTRY_DESTROY.

(riastradh)

Prevent autounload while cyclics are still in use.

(riastradh)

Use \(em for em-dash.

(uwe)

Add couple of missing .Ns

(uwe)

Add t_pslist.debug

(martin)

Add passive references, intermediate between pserialize and refcount.

Discussed on tech-kern:

https://mail-index.netbsd.org/tech-kern/2016/01/24/msg020069.html

API is still experimental and likely to change.  (Obvious changes:
either remove extra arguments everywhere, or shrink psref_target to a
single bit, at the expense of possibly valuable diagnostic checks.)
Should do some real testing before we use this in anger!

(riastradh)

Fix name of example list head.

(riastradh)

Introduce pserialize-safe linked lists.

These are like LIST_* from queue(3), but issue the appropriate memory
barriers for pserialize readers and writers.

Discussed on tech-kern:

https://mail-index.netbsd.org/tech-kern/2016/04/03/msg020365.html

(riastradh)

Make this list diff- and sort-friendlier.

No functional change.

(riastradh)

Revert prior.

(roy)

Revert prior.

(roy)

- remove printf
- fix indent

(christos)

remove HAVE_FPARSELN

(christos)

Add regression tests for some past gdb bugs.

(gson)

Don't rely on XHCI_IMAN_INTR_PEND.  Instead do as others do and simply
clear the bit and handle any pending events.

usb 3 now works on my toshiba laptop (and probably elsewhere)

(skrll)

__diagused, not __unused -- used in an assert.

(riastradh)

Use sockaddr_dl_init

(christos)

Rework locking mechanism introduced in rev. 1.21: use __cpu_simple_lock... instead of mutex_... since this causes issues when LOCKDEBUG is defined

(palle)

include opt_compat_netbsd.h to get COMPAT_SOCKCRED70 defined.

(mrg)

Reminder to add RCSIDs in any future DRM update.

(riastradh)

remove more ifdefs

(christos)

remove more ifdefs

(christos)

_SC_PAGESIZE is not the page size; it's a symbolic code for retrieving
the page size.

(dholland)

remove useless cast.

(christos)

Add PAX_MPROTECT_DEBUG

(christos)

- tidy up error messages
- add a length argument to arpresolve()
- add KASSERT for overflow

(christos)

Don't create an RTM_MISS message for every route allocation.
GC unused code and variables.

(christos)

moar registers, less tpyos

(macallan)

Pull across ddb(4) show kernhist updates from HEAD

(skrll)

Improve ddb(4) show kernhist

1) really prints all the histories merged together (rather than just the
  "first" when no argument specified
2) dumps a single history when an argument is given, e.g.
  "show kernhist usbhist"
3) uses db_printf correctly

(skrll)

Add sc_pid to sockcred so that SOCK_DGRAM and LOCAL_CREDS socket option
can learn the process id of the AF_LOCAL sender.
Add compat glue for old structure.

(roy)

pretty-print link addresses.

(christos)

pretty-print link addresses.

(christos)

Don't interpret routing requests by interface index as arp entry additions!

(christos)

Fix rssi

(roy)

ieee80211 users in Other OS export rssi and noise as int8_t.
We should not be the odd one out for no good reason and the majority
of the ieee80211 drivers treat rssi as int8_t.

(roy)

Revert prior, no idea why it was causing me problems, but it no longer does.

(roy)

Fully avoid the nonstandard types u_char, u_int, and u_long, which
were only used inconsistently in a few places anyway.

(gson)

Fill rtm_addrs properly

This fixes that arp(8) on some archs (only 32bit?) shows "(weird)"
for every entries unexpectedly.

Confirmed on evbarm by ryo@ and i386 by me.

(ozaki-r)

Drop trailing whitespace.

(wiz)

Fill sdl with sockaddr_dl_init

And add an assertion of if_addrlen and ll_addr.

From christos@

(ozaki-r)

Remove FreeBSD RCSId tag

(skrll)

revert part of previous; not needed.

(christos)

Use WIFCONTINUED.

(christos)

Implement WIFCONTINUED using the linux value instead of the FreeBSD one...

(christos)

- clarify si_signo difference for waitid and wait6 as per POSIX.
- implement it.
- document WIFCONTINUED

(christos)

improve usage

(dholland)

don't create ktrace records if we were not asked.

(christos)

Fix ATF_REQUIRE(false)

(christos)

Add parent notification on SIGCONT as required by waitid(2)/wait6(2)

(christos)

Add wait tests.

(christos)

Add wait6(2) tests.

(christos)

Add modular dependency on zlib module.

(pgoyette)

Update dependency: zlib is only needed for the swcrypto device, not for
any other component of opencrypto.

(pgoyette)

Switch PCI xhci(4) attachment from pci_intr_map() to
pci_intr_alloc()/pci_intr_release().

This enables MSI where available.

(skrll)

Stop calling reallocarr and just do the same bounds checking but without
messing around with copying pointers which was causing ssh to crash.

(roy)

Simplify even more to make it clear how the status is set.

(christos)

Set the exit status code properly.

(christos)

more FreeBSD changes from Kurt Lidl.

(christos)

Fix pasto for GPIO_PIN_EVENTS/GPIO_PIN_LEVEL/GPIO_PIN_FALLING
(not used at this time)

(bouyer)

Fix DS1672 suppport: need to provide dm_ch_reg to avoid a KASSERT
on write. For consistency set dm_ch_value too, even if it's unused.

(bouyer)

Unbreak build of kernels without INET

(ozaki-r)

binutils 2.26 mknative for aarch64

(skrll)

Set ELFSIZE to 64

(skrll)

Skip looking for .MAKE.JOBS if either of compatMake or forceJob is true.

If -B compatMake will be true and should be honored.
If -j forceJobs will be true and the command line supplied .MAKE.JOBS
should be respected - for consistency with other usage.

(sjg)

Add test for PR/49872

(christos)

sgimip is big endian, noted by @nisimura.
minor updates.

(mrg)

PR/49872:Yorick Hardy: sed functions 'i' and 'a' discard leading white space
FIxed as suggested, behavior now matches gnu-sed.

(christos)

PR/51043: Yorick Hardy: ftp(1) should use the port number for CONNECT

(christos)

remove sys/wait.h just added, put it in the file that needed it (kvm_proc.c)

(christos)

no need to pass the coredump flag to exit1() since it is set and known
in one place.

(christos)

some ports need <sys/wait.h>

(christos)

catch up with exit1() change

(christos)

catch up with p_xstat split.

(christos)

Split p_xstat (composite wait(2) status code, or signal number depending
on context) into:
1. p_xexit: exit code
2. p_xsig: signal number
3. p_sflag & WCOREFLAG bit to indicated that the process core-dumped.

Fix the documentation of the flag bits in <sys/proc.h>

(christos)

I do like the idea of April 43, 2016, but it's just not standardized enough.

(wiz)

For the benefit of compilers not using our native C runtime environment,
install a "sysident.o" file providing the note identifying NetBSD
executables (and marking the currently installed version).

(martin)

Fix description of si_status.

(christos)

Add some diagnostic printf's when we reset permissions.

(christos)

FreeBSD patches from Kurt Lidl.

(christos)

If the per-target LDSTATIC flag is set, don't include the PIE linker
flags as we currently don't support static PIE.

(joerg)

binutils 2.26 mknative for alpha

(skrll)

Sort sections.

(wiz)

Sort sections.

(wiz)

Use more markup, fix a typo.

(wiz)

Remove some double quotes.

Parity is kept.

(wiz)

Add a test case for here documents positioned outside `` with the
redirect operator inside the ``.  Move the one similar test using
$( ) into its own test case (and expand that test as well.
The `` case was omitted before, as no shells (including NetBSD) were
parsing it correctly,  Now the NetBSD shell does, so ... (from kre@)

(christos)