PR/50701: Martijn van Buul: Adjust MS Confort Mouse 6000

(christos)

we don't have va_mask

(christos)

Don't expose kernel structs to userland.

(christos)

prefer <sys/cpu.h> instead of <machine/cpu.h>

(christos)

Include <bsd.own.mk> before setting any MK variables, so that /etc/mk.conf
is processed first, and it's settings don't override ours.

(christos)

Undo previous changes; the debug libraries were installed because of a bug
in the Makefiles... (Setting MKDEBUGLIB in /etc/mk.conf overrides some
Makefile settings)

(christos)

use __register_t

(christos)

Move the ipl_cookie_t definition earlier, since the mutex internal impl
uses it.

(christos)

Need _KERNTYPES for round_page() vaddr_t.

(christos)

This is a mess:

-> sys/mutex.h -> sys/intr.h -> machine/intr.h -> footbridge/footbridge_intr.h
-> arm/cpu.h -> sys/cpu_data.h -> sys/sched.h -> sys/time.h -> sys/timevar.h
-> sys/systm.h -> sys/param.h -> uvm_param.h -> sys/resourcevar.h

Now sys/resourcevar.h needs a concrete definition of kmutex_t, and
although we started including sys/mutex.h we never got to include
machine/mutex.h which actually defines it, since the footbridge_intr.h
took us for a long ride. Take the easy way out and include
arm/mutex.h in footbridge_intr.h so that we get the definition we
need.

(christos)

spell empty.

(christos)

Handle also the earm{ev,hf}...

(christos)

Map "earm" to armv5

(martin)

prop_dictionary_t hasn't been needed here in, um, 5 years, so remove
the type madness.

(pooka)

include proplib.h if you're going to useuseitit

(pooka)

include <stdbool.h> for bool

(pooka)

Don't include <rump/rumpvnode_if.h> from rump.h.  It's not needed
unless you're doing something special, but requires register_t.
Adjust the few places which actually need rumpvnode_if.h.

(pooka)

add -D_KERNTYPES

(pooka)

Remove unnecessary LLE_REMREF

The code around it was copied from arptimer, but LLE_REMREF
is unnecessary because it is needed only for arptimer that
is called after LLE_ADDREF.

This is a possible fix for PR#50548, PR#50702 and PR#50704.

(ozaki-r)

Note that update builds are broken if MKDTRACE got enabled for your
platform.

(martin)

add t_hypot

(christos)

Add t_hypot to fix build break.  Hi gson!

(jnemeth)

Move librumpuser compile-time options into the librumpuser source
directory.  Those options apply only for the in-tree implementation.

(pooka)

bump version for pr_output change

(pooka)

move cvslatest last

(christos)

use __register_t

(christos)

sys/conf.h is a kernel only header

(christos)

vmparam.h is a kernel header.

(christos)

Fix incorrect magic numbers in scaling.  From FreeBSD commit 23397, by
way of Rin Okuyama.  Fixes PR lib/50698.

(gson)

Add some tests of hypot() and hypotf()

(gson)

Add tool cvslatest

(christos)

Toolify

(christos)

Tool version

(christos)

use __register_t

(christos)

expose __register32_t for frame.h

(christos)

fix amiga sets

(christos)

Replace atexit() calls by destructor attributes - this shared library
is dlopen/dlclose'd !

(martin)

use 0 for udata for portability (FreeBSD has a pointer, we have uintptr_t)

(christos)

the eloop hack should not be needed anymore.

(christos)

expose __vaddr_t for segments.h since it is a userland api.

(christos)

Fix alpha sets

(christos)

sets for cvslatest.

(christos)

Add a program to find the latest timestamp in a cvs repository

(christos)

use __paddr_t since this is a userland API.

(christos)

Undo previous; the lossage is more basic.

(christos)

Remove hf and eb. Rin Okuyama

(christos)

Alpha needs _KERNTYPES

(christos)

use namespace protected types.

(christos)

expose label_t for _KMEMUSER

(christos)

mips needs _KMEMUSER for label_t in pcb.h

(christos)

cache stuff needs _KERNTYPES

(christos)

Use namespace protected types.

(christos)

expose __fpregister_t too, merge definitions.

(christos)

use namespace protected types.

(christos)

assym.h needs _KERNTYPES

(christos)

Grow!

(christos)

t_lwp_create needs kernel types.

(christos)

regen

(christos)

PR/50701: Martijn van Buul: Add MS Confort Mouse 6000

(christos)

merge GCC 5.3.  round 1.  this probably won't work  :-)

libsanitizer likely needs to be re-ported, though i've done most of
the updating we'll need i think already.

(mrg)

prepare for GCC 5:
- remove GCC 4.5 stuff, add most very basic GCC 5 stuff.
- use mknative-gcc.old for GCC 4.8
- copy mknative-gcc to mknative-gcc.old (should have done this when we
  switched to using gcc.old.)

this should not affect current builds.

(mrg)

add a comment saying this is for GCC 4.8

(mrg)

fix dreamcast sets.

(christos)

we don't need <sys/mbuf.h>

(christos)

fix shark sets

(christos)

Grr. arm is different.

(christos)

Instead of hard-coding who has cpu counters why don't you look for them?

(christos)

Fix non-cpu counter build.

(christos)

fix evbarm sets.

(christos)

expose kernel types for gdb so that we don't have to fix each arch separately
for now.

(christos)

expose the kernel types for standalone code.

(christos)

We'll define the kernel types for standalone code.

(christos)

revert previous, we'll do differently

(christos)

Add -D_KERNTYPES

(christos)

PR/5068: Ingo Schwarze: Don't use the .af (alter format) request to display
a named register.

(christos)

pmc_evid_, pmc_ctr_t etc are defined in <machine/types.h> but are not exposed
anymore by default since they violate the userland namespace. Either we
should create __pmc types, or a separate machine/pmc.h header.

(christos)

Rename the cpu_id() define to cpu_idnum() so that it does not conflict with
dtrace and friends.

(christos)

Define _KERNTYPES for things that need it.

(christos)

Hide {p,v}{addr,size}_t and register_t (and a couple more types that
are machine-specific) from userland unless _KERNEL/_KMEMUSER and a
new _KERNTYPES variables is defined. The _KERNTYPES should be fixed
for many subsystems that should not be using it (rump)...

(christos)

Rename the two register_t uses that leaked to userland __register_t so that
they can be fixed later if we want.

(christos)

XXX: use unused variables.

(christos)

no need for <sys/mbuf.h>

(christos)

move more kernel stuff under #ifdef _KERNEL

(christos)

define sooner _KMEMUSER

(christos)

don't include <sys/mbuf.h>

(christos)

bridge with the higher Makefile.inc

(christos)

add another special case so SM50x's framebuffer gets mapped with prefetching /
write combining enabled.
This needs to be generalized somewhere...

(macallan)

Convert unsigned char to wider type before left shift.

Avoids undefined behaviour if shifted quantity overflows int.

CID 980971

(riastradh)

Don't include <sys/vnode.h> to get the pager constants.

(christos)

put _KMEMUSER in the Makefile

(christos)

allow uvm_param.h since it defines types...

(christos)

protect more kernel stuff.

(christos)

Omit spurious include of <sys/isa_defs.h> in dtrace tools build.

Fixes dtrace tools cross-build from architectures with no dtrace
support:

#  compile  libctf/ctf_create.lo
[...]
/export/anoncvs/src/tools/libctf/../../external/cddl/osnet/sys/sys/isa_defs.h:50:2: error: #error "architecture not supported"
#error "architecture not supported"
  ^

(riastradh)

don't need mbuf.h!

(christos)

Fix libproc and librtld_db from Rin Okuyama

(christos)

Those were local changes not meant to be part of the revert.  SORRY!

(riastradh)

Explain what REDIRECT is for.

(christos)

expose less to userland.

(christos)

make this compile again

(christos)

don't use kernel types.

(christos)

MAXNAMLEN, not MAXNAMELEN. From Giuseppe Gatta on netbsd-docs.

(dholland)

unsigned -> unsigned int

(wiz)

cleanup pkgsrc junk

(christos)

fix compilation

(christos)

Needs sys/select.h for fd_set; fails only when _POSIX_SOURCE is defined,
because then something else doesn't include it. (I guess.)

(dholland)

unsigned -> unsigned int in previous change to accomodate C++ boneheadedness.

(dholland)

u_int -> unsigned int, in some of the output.

In the bulk of the output making this change would require reworking a
bunch of the internals to allow spaces in type names, which at the
moment doesn't seem worthwhile.

(dholland)

we don't want sys/mbuf.h

(christos)

Apparently a recent round of vandalism by the C++ standards committee
includes disallowing the standard and historic type name "unsigned". Add
a gratuitous "int" to recent changes.

C++ is really a blight on the world.

(dholland)

Use correct include guard.

(dholland)

u_int -> unsigned, u_long -> unsigned long, etc.

(dholland)

Update synopses to match u_int/u_long/etc. changes. Do not bump date as
it's not a substantive change.

(dholland)

add proftpd diffs.

(christos)

add more points.

(christos)

add more blacklist rejection points.

(christos)

Needs sys/cdefs.h for __BEGIN_DECLS. Also add time.h to get time_t,
rather than doing the machine/ansi.h #ifdef dance, as it doesn't matter
much what symbols utmp.h exposes. This could be tightened later when we
manage to do enough include cleanup to make such dances less expensive
to maintain.

(dholland)

Uses __BEGIN_DECLS so needs sys/cdefs.h; also needs stdint.h.
And, while here, add missing include guard.

(dholland)

u_int -> unsigned

(dholland)

Uses __BEGIN_DECLS and needs sys/cdefs.h.

(dholland)

Requires sys/time.h for struct timespec.

(dholland)

u_{long,short,char} -> unsigned {long,short,char}

XXX: the structure in here incurs unwanted padding on LP64 platforms.

(dholland)

Needs struct timespec.

(dholland)

u_int -> unsigned

(dholland)

u_int -> unsigned

(dholland)

Uses __GNUC_PREREQ__, needs sys/cdefs.h.

(dholland)

Uses BEGIN_DECLS, needs sys/cdefs.h.

(dholland)

Back out previous change to introduce struct encapsw.

This change was intended, but Nakahara-san had already made a better
one locally!  So I'll let him commit that one, and I'll try not to
step on anyone's toes again.

(riastradh)

Uses FILE *, needs stdio.h.

(dholland)

Uses BEGIN_DECLS and needs sys/cdefs.h.

(dholland)

Include <rpc/rpc.h> so XDR is defined.

(dholland)

Needs sys/types.h for mode_t, off_t, ssize_t, and others.

(dholland)

Needs stdint.h.

(dholland)

Erm, fix previous.
(also, as this protocol is externally defined it probably should use its
own constant and not either of these system-defined hostname lengths...)

(dholland)

These all need stdint.h.

Also, dumprestore.h needs ufs/ufs/dinode.h; in talkd.h use unsigned
char instead of u_char; and in timed.h include limits.h and use
_POSIX_HOST_NAME_MAX instead of MAXHOSTNAMELEN as the latter would
require sys/param.h.

(dholland)

u_int{16,32,64}_t -> uint{16,32,64}_t, for the benefit of userland.

(dholland)

Needs prop_array.h to compile.

(dholland)

u_int{8,16,32}_t -> uint{8,16,32}_t, also u_int -> unsigned and
u_char -> unsigned char.

(dholland)

u_int8_t -> uint8_t

(dholland)

Needs sys/time.h for struct timespec.

(dholland)

Needs stdint.h.

(dholland)

Needs sys/time.h for struct timespec.

(dholland)

turned mapped v4 addresses to real v4 addresses so that they work.

(christos)

Needs stddef.h for size_t.

(dholland)

Has __BEGIN_DECLS, needs sys/types.h.

(dholland)

Needs sys/cdefs.h for __BEGIN_DECLS and sys/types.h for mode_t.

(dholland)

u_int -> unsigned; also requires stddef.h and stdint.h to be compilable.

(dholland)

The actual header file for these functions is dm.h, not libdm.h.
Bump date.

(dholland)

- Has __BEGIN_DECLS so needs sys/cdefs.h;
- Needs stddef.h and stdint.h to be compilable on its own

(dholland)

handle v4 mapped addresses

(christos)

Enable KDTRACE_HOOKS on i386 and amd64 GENERIC.

(riz)

u_char, u_int, u_long -> unsigned char, unsigned, unsigned long

This makes these headers work without _NETBSD_SOURCE.

(dholland)

Build DTrace userland by default on i386, amd64 and earm*

(riz)

Remove unused var

(roy)

PR/50691 - Nikolai Lifanov -- add LTS

(jnemeth)

Prevent interface deletion from running into an infinite loop when altq closes.
$ cat /etc/altq.conf
interface ppp0 bandwidth 10M priq
class priq ppp0 high_class_ppp NULL priority 1
class priq ppp0 low_class_ppp NULL priority 0 default
filter ppp0 high_class_ppp 0 0 0 0 1
$ ifconfig ppp0 create
$ /etc/rc.d/altqd onestart
$ ifconfig ppp0 destroy
$ pkill altqd
XXX: pullup-7

(christos)

Size the kevent receive buffer correctly.

(roy)

Fix build for clang and kqueue

(roy)

Try to reduce singular/plural confusion. Use more markup.

(wiz)

Regen.

(msaitoh)

Add anoter new I219-LM.

(msaitoh)

Improved wording.

(pgoyette)

correct comment in literal section

(plunky)

clarify closing of descriptor returned by dirfd(3)

(plunky)

Revert previous

atm_input is not compatible with ifp->if_input and other *_input routines...

(ozaki-r)

Don't call atm_input directly

This should be the last one that was survived ifp->if_input replacements.

No functional change.

(ozaki-r)

Don't abuse struct protosw for ip_encap -- introduce struct encapsw.

Mostly mechanical change to replace it, culling some now-needless
boilerplate around all the users.

This does not substantively change the ip_encap API or eliminate
abuse of sketchy pointer casts -- that will come later, and will be
easier now that it is not tangled up with struct protosw.

(riastradh)

Fix build with RUMP_LOCKDEBUG=yes

(ozaki-r)

fix typo.

(zafer)

PR/50686: David Binderman: fix memory leak

(christos)

remove malloc casts.

(christos)

PR/50685: David Binderman: fix memory leak

(christos)

PR/50684: David Binderman: Fix memory leak

(christos)