Note Realtek RTS5209/RTS5229 Card Reader driver support.

(nonaka)

Fix two bugs.  (jump past entry mask, fix insv)

(matt)

Add a driver for Realtek RTS5209/RTS5229 Card Reader.
Ported from OpenBSD.

(nonaka)

Exclude mvsoctmr_* functions from FBT targets

Probing them doesn't work for DTrace/arm for some reason.

Anyway with this fix, DTrace/arm works on SHEEVAPLUG
as well as BEAGLEBONE.

(ozaki-r)

Move ipflow into ip_var.h and fix confliction

(liamjfoy)

Remove ipflow_prune and replace with ipflow_reap. ok rmind@

(liamjfoy)

Update bsd-pkg-config to version 20140318

Changes from previous version:

+ be a bit smarter in the awk command in the rmdups functionality
+ add --list-all command
+ support for PKG_CONFIG_PATH
+ optimise queries just a tad

(agc)

Once we know the bound routine, rebuilt a new callframe that can be unwound
properly.

(matt)

Add missing include files.  Don't rely on them bein pulled in as a
side effect of other includes.  Fixes build errors when getaddrinfo(1)
is added to the usr.bin tree.
- need <sys/socket.h> for address family (AF_*) and socket type (SOCK_*)
- need <stdio.h> for printf() and friends

(ginsbach)

Add a register definition for the Processor Status Word on VAX.
Change the unwind definition to specify the CFA base via %fp before
using it in the various register locations.

(joerg)

Add .cfi ops so that one can unwind through this.

(matt)

Add support for the .cfi pseudo-ops for VAX.

(matt)

Make readelf know what the vax pcrel32 reloc number is.

(matt)

Reduce unnecessary diffs to upstream.

(wiz)

Change 'i0' to be signed so that the sign bit is preserved on 'k'
  and the code at the bottom splits the multiply.
Fixes denormal results.
This might have been introduced when I deferred shifting 'k' left.

(dsl)

Fix unused variable warnings in drm2.

(riastradh)

Change previous to use uint32_t for 'interval'.
The value is a random interval in usec obtained by reducing a uint32_t
  value modulo 1000000 (multiplied by a delay in seconds of 1).
The value is then being split into secs+usec and assigned to a timeval
  (and an interval).
With -Wsign-conversion the type has to be either an unsigned 32bit
  type, or a signed 64bit one.  This is just plain stupid.
Warning about conversions between signed and unsigned types really
  only makes sense if the compiler is dynamically tracking the domain
  of the value.

(dsl)

> Can we use c99 field initializers here instead of comments?!?

Yes, yes we can.

(macallan)

use time_t for time

(christos)

Merge riastradh-drm2 to HEAD.

(riastradh)

Trim alignment out of packet length before passing it up.

(pooka)

Make all __ALIGNBYTES macros return the same type (size_t)

(christos)

Ignore LD_BIND_NOW for setuid binaries, but keep it in the environment
otherwise.

(joerg)

Remove a simplelock and replace with a kmutex

(skrll)

split 601 and generic PowerPC timecounter code a bit more

(macallan)

Avoid undefined behaviour in shifts in endian decoding routines.

If int is 32-bit and p is a uint8_t *, then p[0] is promoted to int
and p[0] << 24 can shift a one into the sign bit of an int, which is
nasal demon territory.

(riastradh)

update for agc's simultaneous changes

(dholland)

libunwind supports vax now.

(joerg)

Add basic unwind support for VAX. PSW handling and stack pointer after
resume is still incomplete.

(joerg)

Bracket timer_init() with disable_interrupts()/restore_interrupts()
properly. Pointed out by Chris Gilbert.

(martin)

Delay daemonization until RPC services have been registered.
Fixes one of the race conditions of PR misc/48282.

(gson)

Operations vmark(), vunmark() and vismarker() have been replaced by
vfs_vnode_iterator_*(), remove them.

Document vfs_vnode_iterator_*().

Make VI_MARKER private to vfs_vnode.c, vfs_mount.c and unfortunately
to ufs/lfs/lfs_segment.c.

Welcome to 6.99.37

(hannken)

Ticket #1034.

(msaitoh)

Ticket #1034.

(msaitoh)

Ticket #1034.

(msaitoh)

Pull up following revision(s) (requested by maxv in ticket #1034):
sys/kern/kern_verifiedexec.c: revision 1.132
Reorder code to avoid use-after-free on error. From Maxime Villard

(msaitoh)

Pull up following revision(s) (requested by maxv in ticket #1034):
sys/kern/kern_verifiedexec.c: revision 1.132
Reorder code to avoid use-after-free on error. From Maxime Villard

(msaitoh)

Pull up following revision(s) (requested by maxv in ticket #1034):
sys/kern/kern_verifiedexec.c: revision 1.132
Reorder code to avoid use-after-free on error. From Maxime Villard

(msaitoh)

Ticket #1027 and #1030.

(msaitoh)

Add some missing devices (crypto & gige)

(matt)

Apply patch(requested by mlelstv in ticket #1030):
      distrib/sets/lists/comp/md.amd64: patch
netbsd-6 amd64 does not build with MKDEBUG=yes because a file is
missing from the set lists. Current got fhis fixed in rev. 1.186.

(msaitoh)

Pull up following revision(s) (requested by bouyer in ticket #1027):
sys/ufs/ufs/ufs_quota2.c: revision 1.37
Patch from Edgar Fuss on tech-kern:
set grace time if lowering the limit cause the user/group to now be overquota.

(msaitoh)

In fill_fpregset correct the address to grab the register from

(skrll)

Ticket #1021, #1022, #1025 and #1026.

(msaitoh)

Comment out VERBOSE_INIT_ARM

(matt)

Pull up following revision(s) (requested by manu in ticket #1026):
share/man/man4/options.4: revision 1.433
sys/kern/kern_exec.c: revision 1.371- 1.372
Add EMUL_NATIVEROOT so that native binaries can be told to search an
"emulation" directory before the real root. This makes easier to test
an amd64 kernel on the top of an i386 root filesystem prior a full
migration.

(msaitoh)

Pull up following revision(s) (requested by manu in ticket #1025):
sys/compat/netbsd32/netbsd32_netbsd.c: revision 1.184
sys/uvm/uvm_swap.c: revision 1.166
sys/uvm/uvm_swap.h: revision 1.20
sys/compat/netbsd32/netbsd32.h: revision 1.99
Properly translate struct swapent for COMPAT_NETBSD32
Properly translate struct swapent for COMPAT_NETBSD32  (missing commit)

(msaitoh)

Fix typo

(martin)

Pull up following revision(s) (requested by manu in ticket #1021):
sys/compat/netbsd32/netbsd32_event.c: revision 1.10
Fix netbsd32 compatibility bug in kevent().

The keo_put_events() callback copies the events from kernel to userland.
It is called for sets of up to 8 events (constant chosen in kevbuf
definitition in kevent1()). The callback is called with pointer to
userland buffer, count of events to copy, and an index parameter which tracks
where we are in userland buffer when called multiple time.
COMPAT_NETBSD32's flavor of keo_put_events() is
netbsd32_kevent_put_events(). It did not honour the index parameter, which
caused invalid event data to be returned when userland requested more that 8
events. This caused many reliability problems, and the obvious startup crash
of dovecot log process when it accessed udata in the nineth event in its
buffer, which was NULL.

(msaitoh)

Add commented out MVSOC_EARLY_CONSOLE option

(matt)

Add _LOCORE around marvell_interregs_pbase

(matt)

defflag MVSOC_CONSOLE_EARLY

(matt)

Pull up following revision(s) (requested by manu in ticket #1022):
sys/compat/netbsd32/netbsd32_ioctl.h: revision 1.44
sys/compat/netbsd32/netbsd32_ioctl.c: revision 1.68
Add ATAIOCCOMMAND ioctl form COMPAT_NETBSD32
This enables SMART monitoring by a netbsd32 binary.

(msaitoh)

Base on ARMADAXP, not SHEEVAPLUG.
Simplify.
Support FPU_VFP

(matt)

Add xhci

(matt)

defflag VERBOSE_INIT_ARM

(matt)

Add MEMSIZE to opt_mvsoc.h

(matt)

Enable VFP on MV88SV58XX

(matt)

Only include "opt_mvsoc.h" if _INTR_PRIVATE is defined.

(matt)

Don't replicate most of mk.marvell in mk.armadaxp

(matt)

Default comfreq to mvTclk if CONSFREQ is not defined.

(matt)

Add a test suite.

This needs a bit of work to support running the tests from bin/, but
not much, and it runs ok (just less tidily) in place.

At the moment none of it really works because the bsd-pkg-config
script doesn't honor PKG_CONFIG_PATH.

(dholland)

Move defparam opt_mvsoc.h to more appropriate place.

(matt)

Update bsd-pkg-config to version 20140317

Changes from previous version:

+ recurse into other packages found, so that the whole dependency
  tree of cflags/libs is displayed
+ also look in system and X11R7 prefixes for pkgconfig files
+ remove duplicates on output in --cflags and --libs* output, but
  preserve the order

(agc)

KNF, bump warns, fix warnings to print the program name.

(christos)

Hack the detach path so we do it without error messages

(martin)

avoid using __force_align_arg_pointer__ on non-x86, where it does not exist.

(mrg)

Misc cleanup, suggested by Chuck Silvers.

(martin)

Fix build.

Hi dh!

(skrll)

fix memory leak
XXX: pullup 6

(christos)

rename component.c -> md_component.c

(pooka)

Handle VAX

(martin)

iTypo in comment.

(skrll)

merge xmodmap 1.0.8.

(mrg)

merge xrdb 1.1.0, xset 1.2.3, xvinfo 1.1.2, and xwd 1.0.6.

(mrg)

merge xmessage 1.0.3, xmodmap 1.0.8, xprop 1.2.2, xrandr 1.4.1,
xrdb 1.1.0, xset 1.2.3, xvinfo 1.1.2, and xwd 1.0.6.

(mrg)

update for xlsclients 1.1.3.

(mrg)

merge xfindproxy 1.0.3, xfwp 1.0.3, xhost 1.0.6, xinit 1.3.3,
xload 1.1.2, xlsclients 1.1.3, and xman 1.1.3.

(mrg)

Change zfs_sync() to use vfs_vnode_iterator.

(hannken)

Change smbfs_sync() to use vfs_vnode_iterator.

(hannken)

Change pageflush() to use vfs_vnode_iterator.

(hannken)

Change msdosfs_sync() to use vfs_vnode_iterator.

(hannken)

Change nfs_clearcommit() to use vfs_vnode_iterator.

(hannken)

Change nfs_sync() to use vfs_vnode_iterator.

(hannken)

Change lfsquota1_handle_cmd_quotaon() and lfs_q1sync()
to use vfs_vnode_iterator.

(hannken)

Change vismarker() to VI_MARKER for lfs_writevnodes().
This operation has to be changed to vfs_vnode_iterator.

(hannken)

Change quota1_handle_cmd_quotaon() and q1sync() to use vfs_vnode_iterator.

(hannken)

Change ext2fs_sync() to use vfs_vnode_iterator.

(hannken)

Change ffs_sync() to use vfs_vnode_iterator.

(hannken)

Change snapshot_expunge() to use vfs_vnode_iterator.

(hannken)

Change sysctl_kern_vnode() to use vfs_vnode_iterator.

(hannken)

Add fstrans_startnowait()/fstrans_done() to vrele_thread().

(hannken)

updates for xauth 1.0.8 and xconsole 1.0.6.

(mrg)

avoid const problems for now.

(mrg)

merge xauth 1.0.7, xclipboard 1.1.2, xclock 1.0.6, xconsole 1.0.4
and xfd 1.1.2.

(mrg)

updates for makedepend 1.0.5.

(mrg)

update for mkfontscale 1.1.1.

(mrg)

regenerate for xkeyboard-config 2.11.

(mrg)

merge mkfontscale 1.1.1 bits.

(mrg)

merge iceauth 1.0.5, imake 1.0.5, makedepend 1.0.4, mkfontscale 1.1.0,
and twm 1.0.7.

(mrg)

merge xkeyboard-config 2.11.

(mrg)

updates for libdrm 2.4.52.

(mrg)

various updates for libdrm 2.4.52.

(mrg)

merge libdrm 2.4.52.

(mrg)

merge libxcb 1.10 and libpthread-stubs build.
update the libxcb generated files.

(mrg)

add libpthread-stubs.

(mrg)

libpthread-stubs needs a config.h

(mrg)

merge libxcb 1.10.

(mrg)

turn on PTHREADS.

(mrg)

merge libpciaccess 0.13.2.  there are now no local changes! thanks wiz.

(mrg)

Add options MVSOC_INTERREGS_PBASE for *old* Armada XP board..

(kiyohara)

updates for pixman 0.32.4 and x86.

(mrg)

updates for pixman 0.32.4.

(mrg)

merge pixman 0.32.4.

(mrg)

merge xtrans 1.3.3 (deletes Xtranstli.c -- so sad, bye bye.)

(mrg)

merge xtrans 1.3.3.

(mrg)

enable FFS_EI.

(mrg)

fix the libXrender build.

(mrg)

updates for libX11 1.6.2, dri3proto and presentproto, and fix
the libXrender build.

(mrg)

add dri2proto and presentproto.

(mrg)

define HAVE__XEATDATAWORDS.  (XXX move into ${X11FLAGS.THREADLIB}).

(mrg)

Add a lot more tests for exp2() and exp2f().
exp2f(7.7) and exp2f(8.8) seem too far from their expected values
  (especially the latter).
exp2(-1023) and below are badly broken.

(dsl)

Print the result as a 'long double' - on i386 a return value that
should be infinity might just be too large for 'double' and won't
get converted until it has to be saved to memory.

(dsl)

merge libX11 1.6.2.

(mrg)

Simplify somewhat: this is C not FORTRAN-IV - we have structures!
Directly us 'ieee_double_shape_type' when ripping apart the fp number
  to avoid an extra store-load in 'STRICT_ASSIGN'.
Keep 'k' as the exponent, only do 'k << 20' when generating the fp number
  from it.
Fix infinity and underflow returns on i386 - because the value is returned
  in %st0 (x87 stack) we have to generate a long double error value.
The returned value for integers -1023 and below (which should be denormal
  fp valuesr) are 'just plain wrong' (tm).

(dsl)

merge libXrender 0.9.8.

(mrg)

merge libXrandr 1.4.2, libXrender 0.9.8 libXv 1.0.10 and libXvMC 1.0.8.

(mrg)

merge libXfont 1.4.7.

(mrg)

Fix overflow and underflow on i386.
The return value of a 'float' function is in the x87 %st(0) register.
This is an 80bit 'long double' register.
If you multiply 0x1p100f by 0x1p100f the caller sees 0x1p200 - not the
  expected infinity.
So use a 'double' value which goes through a store-load sequence to generate
  the required exception and value.

(dsl)

merge libSM 1.2.2, libXaw 1.0.12, libXfont 1.4.7, libXi 1.7.2,
libXmu 1.1.2, and libXpm 3.5.11 parts.

(mrg)

Check that the result isn't equaly to the expected value before checking
  the absolute size of the error term.
If the expected result is +/-infinity it should compare equal, but the
  result of the subtract may not be zero.
Also print the result and error values in fp hex to make it easier to
  see how may lsb bits are incorrect.

(dsl)

Disable xpci(4) PCI passthrough in XEN3_DOMU kernels by popular demand.

(jakllsch)

Introduce RUMP_CURLWP compile option to select curlwp scheme.
Default is still "hypercall".

(pooka)

remove unnecesary verbosity: s/RUMP_CURLWP_MODEL/RUMP_CURLWP/

(pooka)

note that it's a good idea to do a clean build if an option is changed

(pooka)

document compile-time options for rump kernels

(pooka)

fix tyop

(pooka)

Allow RUMP_LDSCRIPT itself to be used to specify which ldscript to use.
HAVE_SUN_LD will be deprecated in the future.

(pooka)

Bump INSTALL ramdisk size slightly to allow building with DBG=-g.

(joerg)

libunwind supports PowerPC.

(joerg)

Link with --eh-frame-hdr.

(joerg)

Remove unneeded dependencies

(reinoud)

__do_global_ctors_aux and __do_global_ctors_aux must be used locally, so
mark them as hidden.

(joerg)

Exclude sscom_init() prototype definition when its not declared later

(reinoud)

Retrict rumpfs to the provided size, like other filesystems.

(njoly)

Do not use uninitialized pointer if posix_memalign() fails.

(njoly)

When trying to extend a file, don't wait until the underlying memory
allocation succeed. Return ENOSPC upon failure.

(njoly)

Use C99 initializers

(martin)

Add roundl()

(martin)

Provide all missing variants of trunc/floor/ceil.

(martin)

Spelling, whitespace.

(wiz)

Add umcs(4)

(martin)

Add umcs(4) - a driver for moschip 78{1,2,4}0 based multiport serial
adapters. Ported from FreeBSD.

(martin)

Add umcs(4)

(martin)

Add umcs(4)

(martin)

Fix comment: ar.bsp is the RSE backing store base pointer, not ar.bspstore

(cherry)

Add stack align test stub for ia64 native csu

(cherry)

Remove the 'prot' argument from elf_load_psection(). It is not used
outside, and can be declared locally. Clearer.

ok christos@

(maxv)

Use MACHINE_CPU instead of MACHINE_ARCH with pattern matching

Include bsd.own.mk to use MACHINE_CPU.

Advised by matt@

(ozaki-r)

typo in previous

(dholland)