dedup.

(christos)

If either set_escdelay(3) or set_tabsize(3) are called, set _reentrant
to ensure we use the saved value for the sceen.

This effectively makes ESCDELAY and TABSIZE read-only when either
of these functions are called.

(roy)

Improve portability baseline of bta2dpd's cosdata-gen and sbc_crc-gen hostprogs.

XXX why are these being generated at build time?

(jakllsch)

broken test.

(christos)

protect variable with the macro that uses it.

(christos)

we have a cloning bpf, the config test is broken

(christos)

fix endianness of "rsshash"(currently, used for debug message only) and tabify.

(knakahara)

Fix to use softint-based if_input in ixv(4).

See:
  http://mail-index.netbsd.org/source-changes/2016/02/09/msg072521.html
  http://mail-index.netbsd.org/tech-kern/2016/01/14/msg019997.html

(msaitoh)

- Count tso_err again. It was accidentally removed in ixgbe.c rev. 1.28.
- ixgbe.h: Sort entries to reduce diff against FreeBSD.

(msaitoh)

Add :range and :_

:range replaces var value with an integer sequence one per word
in the current var value.

:_ stores the current var value in $_ so that it can be referred to
later in the modifier series.

Reviewed by: christos

(sjg)

reduce install kernel size.

(nonaka)

Fix host tools build for asn1_compile, compile_et, and slc.

(jakllsch)

fix printf args

(christos)

don't try to build links when cleaning or making obj dirs

(christos)

need libutil for pidfile

(christos)

add sqlite3 to statically linked libraries for pam.

(christos)

this is not needed anymore

(christos)

include roken-common.h directly.

(christos)

make a symlink to krb5 for the tools build.

(christos)

prefer some local include files for tools build

(christos)

missing quote

(christos)

fix printf format

(christos)

Most error paths that goto out; don't hold tun_lock.
so don't mutex_exit(tun_lock) in them, but only in
the one that needs it.

ok skrll

(maya)

Fix a typo.

(abhinav)

KNF

(skrll)

Include sqlite3 to the dependent libraries

(christos)

KNF

(skrll)

KNF

(skrll)

The scsipi layer still uses malloc(9)

(skrll)

Convert to kmem(9)

(skrll)

Fix a comment

(skrll)

Another tsleep bites the dust

(skrll)

1349, 1350

(snj)

1350

(snj)

Pull up following revision(s) (requested by maya in ticket #1350):
sys/external/bsd/drm2/dist/drm/ttm/ttm_tt.c: revision 1.8
Guarantee no zero-size uao/kmem allocations via ttm.
It may be that all callers guarantee no zero-size ttm objects, but I
can't prove that in five minutes of browsing callers.  Rather than
add a KASSERT, lacking proof, we'll add a warning message so that if
it does happen then it happens noisily, but we'll also prevent the
bad consequences of passing zero into uao_create by rounding up to a
harmless nonzero allocation.
XXX pullup-7

(snj)

Pull up following revision(s) (requested by maya in ticket #1350):
sys/external/bsd/drm2/dist/drm/ttm/ttm_tt.c: revision 1.8
Guarantee no zero-size uao/kmem allocations via ttm.
It may be that all callers guarantee no zero-size ttm objects, but I
can't prove that in five minutes of browsing callers.  Rather than
add a KASSERT, lacking proof, we'll add a warning message so that if
it does happen then it happens noisily, but we'll also prevent the
bad consequences of passing zero into uao_create by rounding up to a
harmless nonzero allocation.

(snj)

Pull up following revision(s) (requested by pgoyette in ticket #1349):
lib/libutil/opendisk.c: revision 1.14
lib/libutil/opendisk.3: revision 1.12, 1.13
PR/51216: Instead of trying to open files in the current working
directory first for paths that don't contain "/", first try the
/dev paths to avoid confusion with files in the working directory
that happen to match disk names.

(snj)

new lookup order will be in 7.1 before 8.0

(snj)

add 7.1

(snj)

fix function argument.

(christos)

call the proper size functions

(christos)

missing brace

(christos)

fix args

(christos)

handle our basename.

(christos)

override arc4* for tools build.

(christos)

Update for the new *_size() functions recently added.  Mention the
implicit size limit (128KB) for the functions which do not take an
explicit limit argument.

(pgoyette)

Always print the rule id associated with a rule.

(christos)

add sized versions of the copyin ioctls

(christos)

add sized versions of the copyin ioctls.

(christos)

- Increase copyin buffer size to 4M
- Change log output format to be like the OpenBSD's pf including in
  the header the matching rule etc, and fill in the matching info.

(christos)

add sqlite3

(christos)

add sqlite3

(christos)

Sort SYNOPSIS, descriptions, SEE ALSO.
Use more macros.
Fix sections.
New sentence, new line.
Various other improvements.

(wiz)

add sqlite3

(christos)

add sqlite3

(christos)

add sqlite3

(christos)

kerberos needs sqlite3

(christos)

krb5 needs sqlite3 now

(christos)

kerberos now has sqlite3 support.

(christos)

Fix the build (objdirs)

(christos)

Switch agr(4) to use a workqueue. This is necessary because during
a callout, it allocates memory with M_WAITOK, which triggers a
DEBUG assert.

XXX we should drain the workqueue.

ok riastradh

(maya)

Use ${TOOL_SED} and resolve paths properly.

(christos)

avoid the shadow

(christos)

new Heimdal

(christos)

bump heimdal libraries

(christos)

merge conflicts, adjust the build.

(christos)

avoid -Werror=sign-conversion

(jakllsch)

factor out common subexpressions.

(christos)

Add bta2dpd(8) - bluetooth advanced audio daemon.

(nat)

bta2dpd - Bluetooth Advand Audio Distribution (A2DP) Profile Daemon.

bta2dpd allows you to stream audio to and from bluetooth devices.
It features high quality stereo audio and is intended for use with pad(4)
device.

As discussed on current-users@ and tech-userlevel@

(nat)

We need to define COMPAT_NETBSD32 before we include other files;
otherwise things like ucontext32_t will be missing.

(christos)

Add personal TODO and research goals

Things I want to finish for 8:

- as much as possible from the LLDB, Swift, .NET and VirtualBox projects
- more c11 in libc

Things I want to research for 9:

- turn system utilities into C libraries + add bindings for Lua
- rebase Haiku stack on NetBSD + add Kit(s) accessing libsystem utilities

(kamil)

we have space for one more.

(christos)

fix size computation

(christos)

copy the terminating NUL (njoly)

(christos)

Drop inadvertent redundant CPU_CONTROL_MMU_ENABLE ((1 < 22)) for PJ4Bv7.

This was intended to be CPU_CONTROL_UNAL_ENABLE, which is already handled.

Should fix PR kern/51921.

(jakllsch)

Remove tsleep/wakeup from if_zyd

(skrll)

Remove tsleep(9) and wakeup(9) from urtwn(4)

(skrll)

Remove tsleep from uyurex(4)

(skrll)

Remove tsleep from uthum(4)

(skrll)

Remove tsleep from ulpt(4)

(skrll)

Remove tsleep from uvscom(4)

(skrll)

merge for OpenSSL 1.0.2k

(spz)

updated man pages & the x86_64 mont* asm fix for the new openssl.
Expect at least one more commit until the tree builds again.

(spz)

sun4v: implement missing handling of itsb traps 0x008 and 0x009. Based on code from OpenBSD. Tested using qemu.

(palle)

Strip first otherwise we end up with __RCSID("NetBSD")

(christos)

Don't hold softnet_lock if NET_MPSAFE.

Some functions lock softnet_lock while waiting in pserialize_perform() in pfil_add_hook().
(e.g. key_timehandler(), etc)

(ryo)

remove __HAVE_COMPAT_NETBSD32

(christos)

remove __HAVE_COMPAT_NETBSD32

(christos)

remove __HAVE_COMPAT_NETBSD32

(christos)

Remove debug printf(3)

No functional change.

Sponsored by <The NetBSD Foundation>

(kamil)

Add new tests signal{9,10} in t_ptrace_wait{,3,4,6,id,pid}

signal9:
    Verify that masking SIGTRAP in tracee does not stop tracer from
    catching PTRACE_LWP_CREATE breakpoint

signal10:
    Verify that masking SIGTRAP in tracee does not stop tracer from
    catching PTRACE_LWP_EXIT breakpoint

Both tests are broken and linked with PR kern/51918.

Sponsored by <The NetBSD Foundation>

(kamil)

Add new test signal8 in t_ptrace_wait{,3,4,6,id,pid}

signal8:
    Verify that masking SIGTRAP in tracee does not stop tracer from
    catching PTRACE_VFORK_DONE breakpoint

This test is marked as broken and linked with PR kern/51918.

Sponsored by <The NetBSD Foundation>

(kamil)

propagate deadness.

(christos)

Fix build with LLVM, which does not support -mthumb-interwork.

(rin)

Spelling fixes.

(wiz)

Vrecycle() cannot wait for the vnode lock.  On a leaf file system this lock
will always succeed as we hold the last reference and prevent further
references.  On layered file systems waiting for the lock would open a can of
deadlocks as the lower vnodes may have other active references.

(hannken)

Run vflush() when going from read/write to read only.
Set link count to zero once a node has been removed.
"Implement" MNT_GETARGS.

(hannken)

Run vflush() when going from read/write to read only.

(hannken)

Handle v_writecount from layer_open(), layer_close() and layer_revoke()
so lower file system vnodes get marked as open for writing.

(hannken)

When called with WRITECLOSE vflush() must sync the vnode and take
care of unlinked but open vnodes.

PR kern/30525 remounting ffs read-only (mount -ur) does not sync metadata.

(hannken)

Add test for read/write to readonly update mount.

(hannken)

Fix spelling of partition.

(abhinav)

Fix spelling of simultaneous.

(abhinav)

Unbreak builds of rump libraries with RUMP_LOCKDEBUG

(ozaki-r)

Revert to previous.  Audio passes the test properly again.

(nat)

Audio passes the ver 1.2 of the atf test again.  t_pad_outpuut will be
reverted in a follow up commit.

(nat)

add comment about txq->txq_lock contentions in wm_nq_transmit().

(knakahara)

rump does not have ucontext32_t

(christos)

Add new tests signal[67] in t_ptrace_wait{4,6,id,pid}

signal6:
    Verify that masking SIGTRAP in tracee does not stop tracer from
    catching PTRACE_FORK breakpoint

signal7:
    Verify that masking SIGTRAP in tracee does not stop tracer from
    catching PTRACE_VFORK breakpoint

signal6 fails due to PR kern/51918
signal6 fails due to PR kern/51918 and PR kern/51630

Sponsored by <The NetBSD Foundation>

(kamil)

Add new test signal5 in t_ptrace_wait{,3,4,6,id,pid}

signal5:
    Verify that masking SIGTRAP in tracee does not stop tracer from
    catching exec() breakpoint

signal5 is failing and linked with PR kern/51918

Sponsored by <The NetBSD Foundation>

(kamil)

Add new test signal4 in t_ptrace_wait{,3,4,6,id,pid}

signal4:
    Verify that masking SIGTRAP in tracee does not stop tracer from
    catching single step trap

signal4 is marked as failing and referenced with PR kern/51918.

Sponsored by <The NetBSD Foundation>

(kamil)

Sync with HEAD/nhusb

(skrll)

Fix logic inversion spotted by paulg

(skrll)

Make MP-safe and use kmem(9)

Mostly from rmind-smpnet

(skrll)

Add new tests signal[23] in t_ptrace_wait{,3,4,6,id,pid}

signal2:
    Verify that masking SIGTRAP in tracee stops tracer from
    catching this raised signal

signal3:
    Verify that masking SIGTRAP in tracee does not stop tracer from
    catching software breakpoints

signal3 is marked as failing and refrenced with PR kern/51918
Currently signal3 requires code for other architectures than x86_64

Sponsored by <The NetBSD Foundation>

(kamil)

Add virtual routing and forwarding

(ginsbach)

Add new test signal1 in t_ptrace_wait{,3,4,6,id,pid}

signal1:
    Verify that masking single unrelated signal does not stop tracer
    from catching other signals

Sponsored by <The NetBSD Foundation>

(kamil)

provide __HAVE_COMPAT_NETBSD32 and fix multiple include protection consistently.

(christos)

use __HAVE_COMPAT_NETBSD32

(christos)

Fix a typo, add a comma.

(wiz)

Add viornd(4) to virtio(4) man page

(ryoon)

consistent softc member name prefix

(skrll)

fix memory leak at wm_detach.

pointed out and reviewed by msaitoh@n.o.

(knakahara)

Restrict the forcing of COMPAT_NETBSD32 to _LP64 kernels - this is probably
not the right thing to do, but unbreaks the build for now.

(martin)

No COMPAT_NETBSD32 for rump

(martin)

Add comment about possibly prompting for "release" portion of module
path at boot-loader time.

(pgoyette)

Use sc_pr the mixing ring in full not just the first block.

Audio will pass the atf test again.

(nat)

For LOCKDEBUG:
Always provide the location of the caller of the lock as __func__, __LINE__.

(christos)

Don't hold the thread_lock between successive calls to sc_intr as it
breaks mixing.

This will help passing the atf test.  Changes to audio.c to ensue this
will be in a followup commit.

(nat)

always compile in the COMPAT32 code; it is tiny and if we don't it breaks
the modules.

(christos)

don't return early holding a lock!

(christos)

Fix compile failure on i386 with DEBUG_MEMLOAD.

(nonaka)

Fix compile failure on i386 with PAE.

(nonaka)

Define MODULAR so we can have syscalls autoload after the module autoloads.
XXX: This is toxic, and currently caused LOCKDEBUG crashes.

(christos)

Update TODO.ptrace

Added:
- reiterate over FreeBSD tests and add missing ones if applicable

Sponsored by <The NetBSD Foundation>

(kamil)

Update TODO.ptrace

Drop:
- add ATF tests for PIOD_READ_AUXV

Add new entry:
- research what happens when a tracee masks signals (including SIGTRAP)
  and a breakpoint is triggered

Sponsored by <The NetBSD Foundation>

(kamil)

Catch netbsd32_copyargs() up to changes in r1.441 src/sys/kern/kern_exec.c

(jakllsch)

es_arglen is already in bytes...

(christos)

The argument length is in bytes; don't use howmany()

(christos)

PR/51916: Kamil Rytarowski: Don't multiply es_arglen with ptrsz since it is
already in bytes and contains the maximum possible size:
ELF_AUX_ENTRIES * sizeof(auxv) + MAXPATHLEN + ALIGN

(christos)

Improve/fix igsfb(4) support for 15BPP.

Use howmany(depth, NBBY) instead of (depth >> 3), so that 15 bits maps
to 2 bytes instead of 1.

While here, in a nearly-identical change, don't hard-code 8BPP into the
CyberPro blitter.

(jakllsch)

Make the PIOD_READ_AUXV operation more clear

Describe how to set piod_offs.

Sponsored by <The NetBSD Foundation>

(kamil)

Add io_read_auxv1 in t_ptrace_wait{,3,4,6,id,pid}

io_read_auxv1:
    Verify PT_READ_AUXV called for tracee

Sponsored by <The NetBSD Foundation>

(kamil)

Map coprocessor registers before igsfb_hw_setup(), they are needed for
setting up the video mode.

(jakllsch)

igsfb(4): use WSDISPLAY_BORDER_COLOR for overscan color

(jakllsch)

Double size of WSDISPLAYIO_SFONT name buffer to 32 characters.

We've got a font known as "DejaVu Sans Mono", and that is 17 bytes
by the time it's NUL terminated.

(jakllsch)

Add rasops15_putchar_aa() for greyscale fonts.

Based on rasops8_putchar_aa() and rasops32_putchar_aa().

(jakllsch)

- don't use their ntoh{l,s}/hton{l,s} implementation
- fix their ntohl/htonl implementations to use unsigned int instead of
  unsigned long so they work on _LP64.

(christos)

new acpica

(christos)

implement new interfaces

(christos)

merge conflicts

(christos)

Call ixgbe_get_supported_physical_layer() before ixgbe_add_media_types() to
set new media types correctly.

(msaitoh)

Document baudrate(3).

(roy)

And more functions ....

(roy)

Note the substantial work by others on the curses library.

(roy)

Document recently added functions in curses.3

(roy)

Reference curses_slk(3).

(roy)

Add missing argument

(martin)