Pull up following revision(s) (requested by tsutsui in ticket #1080):
distrib/utils/sysinst/arch/alpha/md.c: revision 1.53
Apply patch in PR port-alpha/48697 (with some tweaks).
Tru64 UNIX's disklabel is the same format as BSD disklabel,
and it seems Tru64 stores incorrect geometry values in
d_nsectors (sectors/track) and d_secpercyl (sectors/cylinder).
d_secperunit seems always reliable so use it to get
dlsec (sectors/track) and dlcylsize (sectors/cylinder) values.
Should be pulled up to netbsd-6 branches.

(riz)

Tickets 1079, 1080.

(riz)

Pull up following revision(s) (requested by tsutsui in ticket #1080):
distrib/utils/sysinst/arch/alpha/md.c: revision 1.53
Apply patch in PR port-alpha/48697 (with some tweaks).
Tru64 UNIX's disklabel is the same format as BSD disklabel,
and it seems Tru64 stores incorrect geometry values in
d_nsectors (sectors/track) and d_secpercyl (sectors/cylinder).
d_secperunit seems always reliable so use it to get
dlsec (sectors/track) and dlcylsize (sectors/cylinder) values.
Should be pulled up to netbsd-6 branches.

(riz)

Pull up following revision(s) (requested by tsutsui in ticket #1079):
distrib/utils/sysinst/arch/newsmips/md.c: revision 1.22
distrib/utils/sysinst/arch/pmax/md.c: revision 1.68
distrib/utils/sysinst/arch/vax/md.c: revision 1.40
distrib/utils/sysinst/arch/alpha/md.c: revision 1.52
distrib/utils/sysinst/arch/macppc/md.c: revision 1.46
distrib/utils/sysinst/arch/mipsco/md.c: revision 1.21
distrib/utils/sysinst/arch/landisk/md.c: revision 1.11
distrib/utils/sysinst/arch/news68k/md.c: revision 1.35
Use msg_display(3) rather than printf(3) to print dobootblk messages.
The latter won't show any characters on the sysinst screen
and other ports have used the former.

(riz)

Don't set SOFTINT_MPSAFE to vioif_rx_softint

vioif_rx_softint calls vioif_populate_rx_mbufs that is not MPSAFE.
vioif_populate_rx_mbufs is also called via vioif_ioctl and so can
be called by two LWPs simultaneously, resulting in kernel panic.

PR kern/49007

(ozaki-r)

We need just a VA, not wired physical pages too.

(riastradh)

Don't double-free obj->pages, with two different allocators to boot!

(riastradh)

Destroy the DMA subregion tag on error in drm_pci_detach.

Now that drm_pci_attach calls the driver load function, rather than
deferring it to the caller, the DMA subregion tag can be created
here.

(riastradh)

Elide up to one excess driver-named component from firmware pathnames.

(riastradh)

In radeon_device_init, use ddev->dev, not &pdev->dev, for rdev->dev.

ddev->dev is the actual device_t, and we need it for firmware load.
&pdev->dev is a dummy structure that serves no purpose except some
kludge in our Linux PCI compatibility crap.  In Linux the two should
be the same, so no need to #ifdef __NetBSD__ this.

(riastradh)

Tweak failure case for Linux firmware API.

request_firmware is guaranteed to yield null on failure;
release_firmware on null is guaranteed to be a no-op.

(riastradh)

Try to recover from vbus errors. Some devices draw too much
current at plug time, causing a trancient error.

(bouyer)

Need <sys/atomic.h> for atomic_cas_uint.

(riastradh)

Don't touch task after calling it.

(riastradh)

Fix usb task queue locking.

(riastradh)

intel_dp_aux_wait_done can run while cold.

(riastradh)

Do a post-write DMA sync pos-write, not a pre-write one.

(riastradh)

Don't forget to advance the array index as we walk the queue...

(riastradh)

Lock the right lock around the gmbus condvar.

(riastradh)

Fix wsmouse at uatp attachment for module.

(riastradh)

Put in dummy gem uvm ops for radeon (uses ttm for real mmap).

(riastradh)

Expand null macros to `do {} while (0)', not to nothing.

(riastradh)

Correct return values of Linux test-and-set operations (PR 48999).

Linux's Documentation/atomic_ops.txt says in no uncertain terms these
must return 0 or 1, not zero or nonzero.  I don't think this has
caused an issue for drm (yet), but better to have it right.

(riastradh)

Add non-atomic Linux test-and-set operations.

(riastradh)

Remove defunct declarations of alloc_page/free_page.

(riastradh)

Push -Wno-missing-prototypes into specific radeon source files.

Don't apply it willy-nilly to the whole kernel!

(riastradh)

Adjustments to the "earlyentropy" branch in response to the various
discussions beginning with my initial proposal
http://mail-index.netbsd.org/tech-kern/2014/04/08/msg016876.html and
particularly the long discussion of cprng_fast() performance (e.g.
https://mail-index.netbsd.org/tech-crypto/2014/04/21/msg000642.html).

In particular:

* Per-CPU, lockless cprng_fast replacement using Dennis Ferguson's
  "ccrand" implementation of ChaCha8.

* libkern arc4random() is gone, gone, gone.

* Entropy estimator reverted to 32-bit recordkeeping and timestamps
  per Dennis' comments and analysis.

* LZF entropy estimator removed: it required a great deal of state,
  and rejected only truly pathological input.

I have not yet reverted the changes that provide LZF in the kernel
as generic functionality; I will likely revert those changes prior
to any merge of this branch to HEAD.

(tls)

Remove needless linux_gfp.c.

Trying to make do with Linux's page allocation interface is pretty
much always wrong: either we let uvm allocate pages in uvm objects
for us, or we use bus_dma.

(riastradh)

Initialize the drm_global stuff.

(riastradh)

Add hw.ncpuonline documentation.

(njoly)

Update lists after adding support for vnd to rump (report from hannken@)

(prlw1)

Sync DEFCOPTS and CFLAGS with i386 and amd64 counterparts.
Especially -fno-omit-frame-pointer

(bouyer)

Make sure to call ifp->if_output() with KERNEL_LOCK held.
Should fix mpls-related atf tests.

(bouyer)

call this 20140717.

(mrg)

Finish KAME IPSEC removal:
- Remove field kfs_value, it is always zero. Compute the hash from kt_tag.
- Remove stray definitions kernfs_revoke_sa and kernfs_revoke_sp.

While here, remove kfs_type from allocvp(), it is always kt->kt_tag.

(hannken)

Add vnd support to rump

(prlw1)

Create the dhcpd.leases file if it doesn't exist already, don't just
warn about it not being there.
Closes PR misc/48868

(spz)

rewrite much of the SSL code:
- handle errors in many places they weren't properly
- make SSL_accept() an error the main code notices
- expand bozo_ssl_err() to include bozo_ssl_warn(), bozo_clear_ssl_queue()
- remove empty bozo_ssl_flush()

(mrg)

link against main.c and ssl-bozo.c (which provides the IO backend.)

(mrg)

use const and remove unnecessary braces.

(mrg)

Actually hook up the radeon_ttm uvm pager ops!

(riastradh)

From https://github.com/raspberrypi/linux/commit/517d5c1c9ceb7bf94c4e56e4fb97758e13f24b3b

> Fix for ALSA driver crash
> Avoids an issue when closing and opening vchiq where a message can arrive before service handle has been written

(jmcneill)

need opt_arm_debug.h for VERBOSE_INIT_ARM

(jmcneill)

Get suspend/resume sorta working for i915drm.

(riastradh)

Remove needless pci_map_rom error messages.

(riastradh)

Explain why i915_gem_release_mmap is broken.

(riastradh)

Restore error message tweak that got lost in the merge.

(riastradh)

I thought I removed this troublemaker already.

(riastradh)

First whack at radeon driver.

No hardware to test yet, but it builds.

(riastradh)

Make it build and boot on my test machines.

Screen blanks on boot on the Ivy Bridge system with

  DRM error in cpt_serr_int_handler: PCH transcoder A FIFO underrun

But after that everything is OK.  Appears to be an upstream problem.
To investigate...

I think there's a cache flushing issue somewhere -- there are little
display artefacts on my T60.

(riastradh)

Limit the minimum size of a disk sector to 512 bytes, to prevent memory
overflow on extremely low secsize. This normally conforms to the old standard
(for which there doesn't seem to be a clear spec). Since 2011, IDEMA's Advanced
Format standardizes it to 4k, so this change won't cause any trouble on
new devices.

Put the printf under DIAGNOSTIC temporarily to see if someone complains.

after a quick discussion on tech-kern

(maxv)

Delete troublesome drm_trace.h and drm_trace_points.c.

(riastradh)

fix merge conflicts

(riastradh)

sun4v: Introduce SET_MMU_CONTEXTID macro for setting the mmu context id on both sun4u and sun4v

(palle)

Avoid use of __arraycount in tools as not portable

(justin)

Hold arenas_mtx across (still unused) _malloc_pre/postfork (PR 47428).

(riastradh)

Add doc/3RDPARTY entry for MesaLib.

(riastradh)

Mention tiotg and tifb

(bouyer)

Remove an accidental C99-ism in tools which needs to be portable

(justin)

Add a commented out tifb and associated wsdisplay entry; mark the
driver as experimental and point out that video mode is hardcoded.
Add tiotg, mark as experimental but I think it's good enough to have
it in by default.

(bouyer)

Add tifb and tiotg

(bouyer)

Glue in tifb and tiotg drivers.

(bouyer)

Work in progress on a driver for the graphic controller found in TI am355x SoCs,
based on the FreeBSD driver.
Only dumb framebuffer mode supported.
At this time, the video timings are hardcoded in the driver's source.
The timings inclued here are for the 12" 1280x800 LCD+LVDS cape from
Chalk Elec.
A better way to define video modes is clearly needed.

(bouyer)

Add the bus front end for the Mentor Graphics USB OTG controllers found
in the TI am355x SoCs.

(bouyer)

Add USB control registers.

(bouyer)

Add peripheral clock defines, from FreeBSD.

(bouyer)

Add motg(4), a driver for Mentor Graphic's OTG USB controller IP,
as found in several SoCs.
Only host mode is supported, good enough for umass (USB keys), kubd and ums,
and USB2 hubs.
It fails to properly talk to USB/serial adapters at this time.

(bouyer)

- move all register definitions and hardware documenting comments to
  tcxreg.h
- expand hardware documenting comments
no functional change

(macallan)

remove debugging

(christos)

PR/48367: David A. Holland: Mark possible submake nodes so that we can
avoid closing the job pipe on exec for them in order to make recursive
makes work in parallel.

(christos)

make a minor output change match again.

(mrg)

Add new transform for gl.pc.
Now GLX_TLS is "no" since GLX_USE_TLS is not defined.

(nakayama)

The placeholder name in fontconfig.pc.in has been changed.

(nakayama)

Keep setting 'error' as appropriate (even if this place is broken enough
to crash in many other ways...)

(maxv)

PR/30756: Thomas Klausner: Deactivate threads after detaching the main thread.

(christos)

basename_r and dirname_r are non-standard and different from what we define
here on FreeBSD. Their libgen.h exposes them unconditionally, so hide ours
for now.

(christos)

make -t chroot option available always again.

(mrg)

Kill void * for bridge in struct ifnet

No functional change.

(ozaki-r)

Because of the way we do substitutions we (currently anyway) need to
set prefix=@prefix@ and includedir=@includedir@ here. Fixes PR 48991.

(dholland)

Print uint32_t field as such.

(joerg)

socklen_t is not uint8_t, so don't print it as such.

(joerg)

Darren fixed n8 and n9 recently

(martin)

NetBSD has an entry for protocol #255 in /etc/protocols labeled "reserved"
so expect that in the parsing output instead of 255.

(martin)

Document the upper limit on size of props dictionary that was recently
imposed.

(pgoyette)

Clean up ERRORS (sort, merge).

(wiz)

PR/47430: Fix debugging threads again, GETTID == ptid_get_lwp for us, not
ptid_get_tid :-)

(christos)

replace the old macros with the function names like the rest of the code did.

(christos)

Bump date for -c (exactly one year after previous bump)

(wiz)

use host rules for files.

(christos)

space after comma

(gson)

Use TOOL_VFONTEDPR

(christos)

stray slash

(christos)

fix typo

(christos)

Add vfontedpr

(christos)

descend to vgrind

(christos)

build glue

(christos)

first attempt at toolification.

(christos)

Correct the bread size of struct fsinfo from Gerald Lee at DELL dot com

(christos)

stopgap for non-netbsd build.

(christos)

Unbreak the build for i386

__debugused => XHCI_DEBUG

(ozaki-r)

Suppress another debug output

It's output when DDB.

Fix previous: DEBUG => XHCI_DEBUG

(ozaki-r)

fix type, from enh at google dot com

(christos)

Factor out the basename and dirname code into basename_r and dirname_r
which are static for now. Inspired by similar changes to android(bionic).

(christos)

Do not explicitly undef __INT8_TYPE__, clang is handled specially below.

(joerg)

Add a link to ../../net/bpfjit/t_extmem.c.

(alnsn)

remove PCC specific definitions for __packed, __aligned and __section
which result in a _Pragma() usage.  I don't think this ever worked,
as you can't concatenate string literals in this way.  A warning
would have been produced, but until recently did not result in an
error condition so this was overlooked. Use __attribute__ instead as
this is supported by PCC.

(plunky)

smbfs depends on nsmb, so add the dependency as appropriate. Fixes
# modload smbfs
on modular kernels, PR kern/40011, and probably system crashes.

(maxv)

Tell which dependency has failed

(maxv)

Add module crud to uatp(4).

Compile-tested only; hardware not available right now.

(riastradh)

Convert uatp(4) to use usb_task instead of a workqueue.

Compile-tested only; hardware not available right now.

(riastradh)

Note import of dhcpcd-6.4.2

(roy)

Remove unused file.

(roy)

fix fat fingered KASSERT(solocked(0)) -> KASSERT(solocked(so)) mistake.

spotted by Takahiro HAYASHI

(rtr)

pr_generic() for req = PRU_RCVOOB is always called with control == NULL
so don't bother with a conditional block that handles non-NULL, it
doesn't happen.

(rtr)

Make bus_space_barrier inline to avoid unnecessary argument handling.

(nakayama)

wskbd(4) has a belgian keyboard layout now.

(mbalmer)

Consistency changes only.  Start sentences with a capital letter and end
them with a full stop.

(mbalmer)

Put them one file per line.

(mbalmer)

Add Sharp CE-175TU entry, and bump date.

(ryoon)

Sharp CE-175TU USB to Zaurus Option Port 16 cable is uplcom(4) device.

(ryoon)

Regen.

(ryoon)

Add Sharp CE-175TU USB to Zaurus Option Port 16 cable.
With this cable, you can login to NetBSD/zaurus on SL-C750 via serial console.

(ryoon)

Note import of dhcpcd-6.4.1

(roy)

Sync

(roy)

Use static_cast for rb_tree_insert_node void * return value.

(joerg)

Remove unused file.

(alnsn)

Add a belgian keyboard layout, based on the french keyboard layout.
Fixes PR install/46871.

(mbalmer)

Ticket 1914.

(msaitoh)

Ticket 1914.

(msaitoh)

Ticket 1914.

(msaitoh)

Pull up following revision(s) (requested by maxv in ticket #1914):
sys/kern/sys_module.c: revision 1.15 via patch
Fix a user-controlled memory allocation. kmem_alloc(0) will panic the system.
ok christos@

(msaitoh)

Pull up following revision(s) (requested by maxv in ticket #1914):
sys/kern/sys_module.c: revision 1.15 via patch
Fix a user-controlled memory allocation. kmem_alloc(0) will panic the system.
ok christos@

(msaitoh)

Pull up following revision(s) (requested by maxv in ticket #1914):
sys/kern/sys_module.c: revision 1.15 via patch
Fix a user-controlled memory allocation. kmem_alloc(0) will panic the system.
ok christos@

(msaitoh)

Don't include atomic.c, we have aliases for it in src/common.

(joerg)

Ticket 1097 and 1098.

(msaitoh)

Ticket 1097 and 1098.

(msaitoh)

Ticket 1097 and 1098.

(msaitoh)

Pull up following revision(s) (requested by maxv in ticket #1098):
sys/kern/sys_module.c: revision 1.15
Fix a user-controlled memory allocation. kmem_alloc(0) will panic the system.
ok christos@

(msaitoh)

Pull up following revision(s) (requested by maxv in ticket #1098):
sys/kern/sys_module.c: revision 1.15
Fix a user-controlled memory allocation. kmem_alloc(0) will panic the system.
ok christos@

(msaitoh)

Pull up following revision(s) (requested by maxv in ticket #1098):
sys/kern/sys_module.c: revision 1.15
Fix a user-controlled memory allocation. kmem_alloc(0) will panic the system.
ok christos@

(msaitoh)

Pull up following revision(s) (requested by maxt in ticket #1097):
sys/kern/kern_core.c: revision 1.23
Fix a read-beyond-end string read.

(msaitoh)

Pull up following revision(s) (requested by maxt in ticket #1097):
sys/kern/kern_core.c: revision 1.23
Fix a read-beyond-end string read.

(msaitoh)

Pull up following revision(s) (requested by maxt in ticket #1097):
sys/kern/kern_core.c: revision 1.23
Fix a read-beyond-end string read.

(msaitoh)

Cleanup comments. No functional change.

(msaitoh)

Make bridge MPSAFE

- Introduce BRIDGE_MPSAFE
  - It's enabled only when NET_MPSAFE is defined
    in if.h or the kernel config
- Add iflist and rtlist mutex locks
  - Locking iflist is performance sensitive,
    so it's not used when !BRIDGE_MPSAFE
- Add bif object reference counting
  - It enables fine-grain locking for bridge member lists
    by allowing to not hold a lock during touching a bif
  - bridge_release_member is added to decrement the
    reference count
  - A condition variable is added to do bridge_delete_member
    gracefully
- Add if_bridgeif to ifnet
  - It's a shortcut to a bif object of a bridge member
  - It reduces a bif lookup cost and so lock contention on iflist
- Make bridgestp MPSAFE too

(ozaki-r)

Remove a duplicated error message

(ozaki-r)

Suppress debug output of xhci_device_intr_done

It's annoying because it is output even on keyboard inputs.

(ozaki-r)

clang doesn't like -Wno-tautological-undefined-compare, use
-Wno-tautological-compare instead.

(riz)

sun4v: add interrupt handling to sun4v trap table

(palle)

Bugfix: add missing cmp instruction to GET_MAXCWP macro

(palle)

Make named and the rest of BIND seed their RNG from /dev/urandom, not /dev/random.  Better would be to rip the enormous gonkulating RNG machinery out of libisc entirely.  Later.

(tls)

Refactor BPF_COPX code. New version doesn't load buf and buflen after copx call.

(alnsn)

Add bpfjit_cop_copx and bpfjit_copx_cop tests.

(alnsn)

make the comment at the DHCPv6 define less cryptic

(spz)

Don't use BJ_TMP2REG for 32bit packet reads. Assign this register to (buf+X)
in BPF_LD+BPF_IND and save one instruction.

(alnsn)

whitespace cleanup

(martin)

Vinnitsa isn't even in South-Western Russia. Make it Ukraine.

(asau)

Sukhum isn't in Georgia since 1993 (1990, 1992, 1993, 1994,
1999, or 2008 depending on what you count as break-up).
In any case in recent twenty years it was easier to get there
through Russia than through Georgia.

(asau)

At suggestion of christos@, use -b for b&w rather than -c

(pgoyette)

Simferopol' is in Krym rather than in Adygeya, and it is in Russia.
(Whether your government recognize it or not is another issue.
If you want to visit it these days, you have to enter Russia.
Following Ukrainians, even before the recent unrest and referendum
it was strongly advised not to discuss this topic with locals there.
At least it was advised not to call Krym a part of Ukraine.)
Fix Kerch' entry as well.

(asau)

"peripheral", not "peripherial".

(dholland)

Fix compare to actually do the right thing.

(joerg)

Disable more warnings for clang build to deal with xsrc changes.

(joerg)

Clang expects __INTx_TYPE__ to be used with signed/unsigned modifier
when defining (u)intx_t, GCC has a separate __UINTx_TYPE__. Newer Clang
will also provide the latter, but expects the behavior of the former to
remain. As ARM uses unsigned char by default, distinguish between Clang
and non-Clang here to the right type for int8_t.

(joerg)

Provide an option to restore traditional black-and-white mode.

(pgoyette)

Setup Bt458 color palette to support ANSI color text on 8bpp framebuffer.

Mostly taken from OpenBSD/luna88k.  Also sync some comments.
Tested on LUNA-II, and mlterm-fb with 8bpp wallpaper also works fine.

Thanks to Kenji Aoyama (OpenBSD/luna88k maintainer) for providing
his spare 8bpp board.

(tsutsui)

print the error from bus_space_map.

(christos)

Add the new files to unbreak the build (hi marc!)

(martin)

remove WORDS_BIGENDIAN, nothing uses it.
Pick up DHCPv6 from Makefile, so we only enable it for USE_INET6 like
before.

(christos)