Expose drm fb helpers for ddb enter/exit.

(riastradh)

Add enable/disable polling hooks to genfb_ops.

(riastradh)

Fix WSDISPLAYIO_GET_BUSID on new intelfb(4).

Need to pass the PCI device, not one of its children (which intelfb
is), to wsdisplayio_busid_pci.

Fixes starting X.

(riastradh)

Fix typos and improve grammar and punctuation.

(wiz)

Sort SEE ALSO. Use .An. Bump date.

(wiz)

Rework Intel framebuffer attachment.

Turns out genfb_softc must be at the start of its own device_private,
so we can't have it in the middle of a structure, and certainly can't
have potentially multiple genfbs per graphics device.  Failing to do
this is why entering ddb didn't work with an i915drmkms console.

Instead of putting a genfb_softc into each drm_fb_helper, put a
device_t into each drm_fb_helper and create a new intelfb(4) device
for each Intel framebuffer.  This will be more flexible, too, in case
we want accelerated framebuffers later on.  (XXX Need to adapt the
radeon code, which I'm disabling until that happens to avoid breaking
the build.)

While here, defer disabling VGA, vga_cndetach, and genfb_attach until
we're actually ready to do them all together along with the mode
switch.  This should reduce the amount of time during which the
screen is blanked, in case it hangs.

(riastradh)

Use .Sy for 'Just-In-Time', add a reference to bpfjit(4).

(alnsn)

man 4 bpfjit

(alnsn)

add p++, and move it and cxxcom to comp-cxx-bin

(plunky)

make p++ optional, on MKCXX

(plunky)

add PCC libexec/cxxcom

(plunky)

add p++ tool, for PCC

(plunky)

update build framework for import of pcc-20140706

(plunky)

fix merge conflicts for pcc-20140706

(plunky)

Fix last change (pa_calib_ticks should be sc_pa_calib_ticks).

(riz)

sun4v: add missing spill/fill handling for trap level 1 + cosmetic changes to comments for trap level 0 entries

(palle)

Let the assembler bail out if the immediate is too large instead of
truncating it.

(joerg)

sync with OpenBSD

(christos)

Report some additional information about the instance.

(riz)

cleanup after last commit

- add KASSERT(req != PRU_BIND) and KASSERT(req != PRU_LISTEN) inside
  tcp_usrreq() as these reqs should no longer reach here.
- remove (now unreachable) PRU_LISTEN case in switch.

(rtr)

cleanup more trash that are in the distribution.

(christos)

new tmux

(christos)

split PRU_BIND and PRU_LISTEN function out of pr_generic() usrreq
switches and put into separate functions
  xxx_bind(struct socket *, struct mbuf *)
  xxx_listen(struct socket *)

  - always KASSERT(solocked(so)) even if not implemented

  - replace calls to pr_generic() with req = PRU_BIND with call to
    pr_bind()

  - replace calls to pr_generic() with req = PRU_LISTEN with call to
    pr_listen()

  - drop struct lwp * parameter from at_pcbsetaddr(), in_pcbbind() and
    unp_bind() and always use curlwp.

rename existing functions that operate on PCB for consistency (and to
free up their names for xxx_{bind,listen}() PRUs

  - l2cap_{bind,listen}() -> l2cap_{bind,listen}_pcb()
  - sco_{bind,listen}() -> sco_{bind,listen}_pcb()
  - rfcomm_{bind,listen}() -> rfcomm_{bind,listen}_pcb()

patch reviewed by rmind

welcome to netbsd 6.99.48

(rtr)

remove ssp commented out lines, they don't see to be needed anymore.

(christos)

merge conflicts, fix build.

(christos)

cleantags to avoid conflicts.

(christos)

Add consistent prefix to a softc member name.

(tsutsui)

Add a FIRST1G page freelist to x86, for old graphics devices.

(riastradh)

Yet another change for PR#48451. Don't use prefetchable memory for BAR0, too.
Reported by taca@.

(msaitoh)

regen

(pooka)

Ensure that symbols are aliased to standard names when
using RUMP_KERNEL_IS_LIBC.

For example configure scripts do not always include headers, and if select
is visible as a __RENAME'd __select50 instead of select, configure will
not find it.

before:
=== snip ===
checking for gethostbyname... yes
checking for socket... no
checking for socket in -lsocket... no
=== snip ===

after:
=== snip ===
checking for gethostbyname... yes
checking for socket... yes
checking for main in -linet6... no
=== snip ===

Another option would be to enable the libc compat bits, and it probably
will be done at some point, but this is easier and more self-contained
for now.

(pooka)

fix a bunch of bugs & tpyos:
- keep track of registers being mapped or not to avoid double (un)mapping when
  switching between WSDISPLAYIO_MODE_DUMBFB and _MAPPED
- distinguish between _DUMBFB and _MAPPED
- actually map the framebuffer at offset 0 in _DUMBFB
- set BUS_SPACE_MAP_PREFETCHABLE when mmap()ing video memory
now Xorg + xf86-video-wsfb work on top of voodoofb

(macallan)

- Fix a bug that bge can't be initialized correctly on some sparc64 systems
  (PR#49018).
  See also: http://mail-index.netbsd.org/port-sparc64/2013/10/28/msg002036.html
- Change initializaton order as the same as the document, Linux tg3 and FreeBSD
  (part of r241436).

(msaitoh)

Fix typo. Bump date for previous.

(wiz)

Sync DIAGNOSTIC performance comments with historical reality.

Somehow, options(4) ended up with a bizarre claim that DIAGNOSTIC can
reduce performance by 15%.  While that might have been true at some
isolated point due to a bug, it's an outlier.  Since at least 2BSD,
DIAGNOSTIC has added asserts, resulting in at most
difficult-to-perceived performance degredation, and many people have
been running production systems (meaning systems they intend to use,
rather than debugging targets) with this option.    (The decision a
while ago to enable DIAGNOSTIC in -current's GENERIC reflects this; if
it really were a 15% hit such enabling by default would be unreasonable.)

(gdt)

Rename sljitarch.h to sljit_machdep.h.

(alnsn)

Omit redundant vga_is_console test.

This was a copypasta error -- before it tested iot then memt -- but I
see no evidence that passing memt to vga_is_console is ever sensible.

(riastradh)

RCS IDs do not go in header files.

(riastradh)

Tweak style, make it compile...oops.

(riastradh)

For the unbind code to work the bind code must set the offset.

(riastradh)

Also unbind the right offset.

(riastradh)

Bind i810 dcache pages at the requested offset, not at VA start.

From John D Baker in PR xsrc/48344.

XXX pullup to 6 (by patch)

(riastradh)

split PRU_SENDOOB and PRU_RCVOOB function out of pr_generic() usrreq
switches and put into separate functions
  xxx_sendoob(struct socket *, struct mbuf *, struct mbuf *)
  xxx_recvoob(struct socket *, struct mbuf *, int)

  - always KASSERT(solocked(so)) even if request is not implemented

  - replace calls to pr_generic() with req = PRU_{SEND,RCV}OOB with
    calls to pr_{send,recv}oob() respectively.

there is still some tweaking of m_freem(m) and m_freem(control) to come
for consistency.  not performed with this commit for clarity.

reviewed by rmind

(rtr)

regen

(pooka)

In the RUMP_KERNEL_IS_LIBC case, account for WEAKASM, i.e. create a
_sys_foo strong alias, and _foo and foo weak aliases.  This actually
treats all syscalls as WEAKASM, but it's probably(?) harmless, and saves
us from manual work a la libc.  Fixes at least pthread_cancelstub.

(pooka)

No functional change:
- Grouping functions.
- KNF.

(msaitoh)

No functional change:
- Add missing prototypes.
- Rename functions for consistency and clarify.
- KNF.

(msaitoh)

More general advice about "cvs [update aborted]... No such file
or directory".

Also invent a "Cause:" keyword for use in "Symptom:/Cause:/Fix" sections.

Also add a note near the top of the file saying that there are
several sections marked by "^^^".

(apb)

How to deal with src/external/mit/lua/src: No such file or directory

(apb)

Implement rumpcomp_sync_icache() hyprecall for arm and add
a barebone implementation of arm cache ops to librumpkern_sljit.

(alnsn)

Whitespace

(skrll)

Avoid calling copyout with holding mutex(IPL_NET)

Because copyout may lead a page fault that may sleep, we have to pull it
out from the critical section of mutex(IPL_NET) in bridge_ioctl_gifs.

(ozaki-r)

Build fix (use error when defined).

(htodd)

Add missing unlock

(ozaki-r)

fixed removal of Lua 5.1 entries from distrib/sets/lists

(lneto)

npf_iscached: add an assert.

(rmind)

NPF: rework of the connection saving and restoring:
- Add support for saving a snapshot of the current connections together
  with a full configuration.  Support a reverse load operation.  Eliminate
  the old 'sess-save' and 'sess-load' in favour of the new mechanism.
- Share code between load and reload operations: the latter performs
  load from npf.conf without affecting the connections.
- Simplify and fix races with connection loading.
- Bump NPF_VERSION.

(rmind)

Remove (harmless) redundant assignment of task->queue in usb_add_task.

(riastradh)

Clean up random implementation for librumpuser

Use /dev/urandom for platforms without arc4random, not srandom(),
deduplicate code, do not read excessive random bytes

Reviewed by pooka@

(justin)

rename npf man pages to libnpf.

(christos)

rename npf.3 -> libnpf.3 per rmind@'s request

(christos)

Free temporary segs on the way out of bus_dmamap_load_pglist.

(riastradh)

Cast to sparc_cache_flush() argument types. This change makes it clear
thatr sparc_cache_flush() is defined by sljit (unlike other ports).

(alnsn)

Cast to __syncicache() argument types.

(alnsn)

Implement rumpcomp_sync_icache() hyprecall for mips and add
a barebone implementation if mips cache ops to librumpkern_sljit.

(alnsn)

Add sljitarch.h to cobalt and pmax.

(alnsn)

Add parentheses around macro arguments.

(alnsn)

Define SLJIT_CACHE_FLUSH() for mips.

(alnsn)

Fix source lists for radeon dri modules.

(riastradh)

Override CPUFLAGS for compiler cpu-specific options, instead of appening
to CFLAGS. This allows the user to define CPUFLAGS without breaking the
build.

(bouyer)

Add workaround for PR#48451. Some BCM5717-5720 based systems get NMI on boot.
This problem doesn't occur when we don't use prefetchable memory in the APE
area. I don't know the reason...  The way to not to use prefetchable memory
was adviced by Riastradh. Tested with HP MicroServer Gen8.

(msaitoh)

'file system' for consistency with documentation (instead of 'filesystem').

(wiz)

Do not printf(foobar()).  Bad bad bad bad bad!

(riastradh)

Fix ttm swap storage destruction.

(riastradh)

Fix MODULE() dependencies to account for VND_COMPRESSION

(pooka)

Two tweaks: don't use a temporary register to dereference the err agrument
after xcall and don't generate ((tmp1 & 0xf) << 2) twice in emit_msh().

(alnsn)

Don't use scratch registers for X and to restore A after BPF_COPX call.

(alnsn)

1) On 64bit systems, don't add the 32bit execsw[] to the global exec array.
  exec_elf32 works on 32bit systems only, and will crash 32bit binaries on
  64bit systems.
2) Now that exec_elf32 is dormant, we can give the native ELF loaders the
  highest priority.

Binaries will load faster now (system boot, compilation, etc.).

With the help of njloy@. Discussed a bit on tech-kern@, no disagreement.

(maxv)

Enable KMEM_REDZONE on DIAGNOSTIC. It will try to catch overflows.

No comment on tech-kern@

(maxv)

restore the right colour map when X exits

(macallan)

Add SERDES support from Matthias Drochner. Not all devices were tested but all
SERDES entries are enabled.

(msaitoh)

Make Linux ww_mutex use a spin lock inside, not an adaptive lock.

(riastradh)

Make if_vioif MPSAFE

- Introduce VIOIF_MPSAFE
  - It's enabled only when NET_MPSAFE is defined in if.h or the kernel config
- Add tx and rx mutex locks
  - Locking them is performance sensitive, so it's not used when !VIOIF_MPSAFE
- Set SOFTINT_MPSAFE to vioif_rx_softint only when VIOIF_MPSAFE

(ozaki-r)

Introduce VIRTIO_F_PCI_INTR_MPSAFE for virtio

It is set by a child driver, e.g., if_vioif. If set, virtio sets
PCI_INTR_MPSAFE for pci_intr_establish.

(ozaki-r)

one more quoted comment

(christos)

fix lint build

(christos)

Conditionalize more stuff on TI_AM335X so BEAGLEBONEXM kernel
builds succeed.

(riz)

Avoid bus_space_read/write_8 on i386.

(riastradh)

Avoid bus_space_read/write_8 on i386.

Would be nice if there were a clearer way to detect it than _LP64,
and if its absence caused a compile error instead of a link error.

(riastradh)

Add kludge to use luna68k 4/8bpp framebuffers as a monochrome server.

The idea is taken from mlterm-fb --depth option implementation.

This change should not affect other framebuffer types.
Tested on 4bpp on LUNA and 1bpp/8bpp on LUNA-II.

(tsutsui)

Ignore the contents of revents when poll() returns failure.
This is one more part to the fix for PR kern/46464.  Patch
from pooka.

(gson)

Don't assume if_init is always set

if_init may be NULL, e.g., if_vlan.

PR kern/48997

(ozaki-r)

Adjustments based on suggestions from gnrp in PR 45276:
- fix argument names of VOP_REMOVE
- mention deprecated behavior of calling VOP_GETATTR unlocked
- mention rename lock

(dholland)

removed accidental importation

(lneto)

removed accidental importation

(lneto)

am335x: expect the board code to provide a set_mpu_volt() function,
to change the MPU code voltage. When changing MPU frequency also request
a voltage change. provide frequency/voltage tables for the "blank" revision
and de A revision, based on documents from www.ti.com (this means
it's not possible anymore to request a beaglebone white to run at 1Ghz,
but this was not working anyway).
beagle_machdep.c: callback to the tps65217pmic to change the MPU voltage.

Tested on beaglebone, from 275 to 720Mhz, and on beaglebone black from
300 to 1000Mhz.

(bouyer)

Add a callback to change the regulator outputs.
Tested on beaglebone.

(bouyer)

typo in comment

(dholland)

There is also no HTML version of "RELEASE_NOTES".

(tron)

Try to fix the build:
There is no "AAAREADME.html". So only attempt to install the ASCII version.

Not sure why my full build didn't catch this problem. Sorry.

(tron)

Mark #endifs that are far enough away from their corresponding #ifs to
be unclear.

(dholland)

use hdestroy1 now that hdestroy does not free the key.

(christos)

amend the new destroy function to take function pointers.

(christos)

Keep countries in sync with IATA listings.  NetBSD is not in the business
of defining in which country an airport is.

(mbalmer)

The lua 5.1 to 5.3 update added isprint to libkern.h

(prlw1)

Restrict -mno-fp-ret-in-387 to GCC.

(joerg)

Don't return ENETRESET when ioctl SIOCSIFMTU

Otherwise, just changing MTU with ifconfig shows
a confusable error message.

RP kern/48996

(ozaki-r)

Sync wizd's feature list with reality.

(tron)

Update list of Postfix README files.

(tron)

Rationalize README file handling:
1.) Install only README files that are relevant to the Postfix binaries
    distributed with NetBSD.
2.) Create a single list of the above files that is used for both the
    text versions and HTML versions.

Problem detected by wizd(8).

(tron)

KNF.

(uebayasi)

ipifunc[]: Comment IPI constant names for grep'ability.  Constify.

(uebayasi)

since we support 601 hardware now, let's build the kernel for it as well

(macallan)

Fix braino in last commit and free the node iff the insert failed.

(joerg)

Drop variable only used in return.

(joerg)

Add ASUSTeK USB-N10 NANO support.

(nonaka)

regen.

(nonaka)

Added a ASUSTeK USB-N10 NANO.

(nonaka)

Change kernfs from hashlist to vcache.

(hannken)

Added RTL8188EU firmware for urtwn(4).

(nonaka)

Remove trailing whitespace. Sort SEE ALSO. Add comma in list.

(wiz)

Grammar. Bump date.

(wiz)

add the new hdestroy functions

(christos)

Our hdestroy implementation was non-conformant because it freed the key of
each entry. Add a new function hdestroy1 that allows the user to control
what gets freed. Pointed out by Pedro Giffuni at FreeBSD.

(christos)

Note about urtwn(4)'s RTL8188EUS and RTL8188ETV support.

(nonaka)

Added support for the Realtek RTL8188EUS and RTL8188ETV chipsets from FreeBSD.

(nonaka)

Remove another KAME IPSEC residue, "struct secasvar" and "struct secpolicy".

(hannken)

Add options WSDISPLAY_COMPAT_RAWKBD for Xorg server.

(tsutsui)

Pull LUNA's keyboard driver changes from OpenBSD/luna88k.

- split keyboard mapping definitions to new files, omkbdmap.[ch]
- add WSDISPLAY_COMPAT_RAWKBD support
  (actually the name is wrong; it doesn't emit raw keycode but
  converts MD code into PS/2 one as "raw keycode on x86" for Xorg server)
- allow to enter into ddb by CTRL+ALT(zenmen)+ESC
- remove unnecessary return values

Now LUNA's keyboard works on Xorg server without tweaks by xmodmap(1) etc.

(tsutsui)

Add one more cmsize == 0 check before PUTCMAP ioctl.

(tsutsui)

Include sljit files.

(alnsn)

Add commented out bpfjit options to mips kernels.
While here, add 2 missing RCS ids.

(alnsn)

No need to handle colormap ioctls if cmsize == 0 even in WSFB_CI case.

(tsutsui)

Fix wrong VRAM size calculation on old drivers without GET_FBINFO ioctl.

The linebytes is a byte number of width so it should be
(linebytes * height), not (width * linebytes).

XXX: I'm not sure where WSDISPLAYIO_GET_FBINFO implementation was discussed
    but I wonder if we should use proper ifdef or macro to share this
    third party Xorg driver with OpenBSD...

(tsutsui)

formatting

(rmind)

Bye bye npf_session.c

(rmind)

Disable NAK timeout on bulk transfers, I couldn't find a way to clear
the NAK bit not the FIFO.
Make xfer aborts more robust to races, by setting proper status and flags
in xfer.

(bouyer)

gcc-4.8 complains about not being able to inline

(christos)

Fix gcc warnings.

(rmind)

sun4v: Introduce a NORMAL_GLOBALS macro - first used in sparc_interrupt - other relevant uses will be handled as the sun4v work is progressesing - ok martin@

(palle)

undo previous

(christos)

XXX: remove file that includes userland headers.

(christos)

lua(4): fixed ldblib.c

* for some reason it wasn't added on my last commit

(lneto)

fix usage

(dholland)

Add npf_conn.h missed in the previous commit.

(rmind)