Add a southeastern china airport.

(mbalmer)

explicitly include sys/resource.h

(para)

Rearange VA layout for 64 bit kernels, as suggested by Lars Heidieker:
move the dynamic kernel memory past 4GB (basically unlimiting it) and
use the now freed address range past kernel data upto 2GB for modules.

(martin)

This is a working version of memcpy implemented using NEON instructions.
Still needs tuning as it is still about 15% than the non-NEON version.

(matt)

There is no point in trying to tell userland access from kernel access by
looking at the address on sparc64 - so, let us believe the comment and always
deal with kernel access here.
While there, fix write access to kernel text for bigger kernels.

(martin)

Tickets 758, 759

(riz)

sys/dev/sdmmc/sdhc.c patch
sys/dev/sdmmc/sdhcreg.h patch

Support SDHC version 3 clocks.
[skrll, ticket #759]

(riz)

Pull up following revision(s) (requested by hannken in ticket #758):
sys/kern/vfs_wapbl.c: revision 1.53
sys/kern/vfs_wapbl.c: revision 1.54
wapbl_biodone: Release the buffer before reclaiming the log.
  wapbl_flush() may wait for the log to become empty and
  all buffers should be unbusy before it returns.
Try to coalesce writes to the journal in MAXPHYS sized and aligned blocks.
Speeds up wapbl_flush() on raid5 by a factor of 3-4.
Discussed on tech-kern.
Needs pullup to NetBSD-6.

(riz)

Note that Postfix 2.9.5 was imported.

(tron)

Install documentation and manual pages for Postfix's memcache client support.

(tron)

include limits.h for CHAR_MIN

(dsl)

Include "sys/ioctl.h" for the prototype of ioctl(2) which fixes the build.

(tron)

Fix build of Postfix 2.9.5.

(tron)

This is included into user-programs by signal.h, it shouldn't be
pulling in machine/fpu.h - which doesn't describe anything userspace
(directly) needs.

(dsl)

Only expose the bulk of sys/proc.h and sys/lwp.h if _KERNEL or _KMEMUSER
is defined.
i386 and amd64 build ok.

(dsl)

Expose the DTYPE_xxx (names) defines even if _KERNEL isn't defined.
There is no other excuse for including this file in userspace.

(dsl)

Only define 'struct kinfo_proc' if _KERNEL or _KMEMUSER is defined.
All 'normal' programs should be using kinfo_proc2 instead.

(dsl)

#include sys/ioctl.h

(dsl)

Resolve conflicts from last import.

(tron)

Import Postfix 2.9.5. Major changes since version 2.8.x:
- Support for long, non-repeating, queue IDs (queue file names). The
  main benefit of non-repeating names is simpler logfile analysis. See
  the description of "enable_long_queue_ids" in postconf(5) for
  details.
- Memcache client support, and support to share postscreen(8) and
  verify(8) caches via the proxymap server. Details about memcache
  support are in memcache_table(5) and MEMCACHE_README.
- Gradual degradation: if a database is unavailable (can't open, most
  read or write errors) a Postfix daemon will log a warning and
  continue providing the services that don't depend on that table,
  instead of immediately terminating with a fatal error. To terminate
  immediately when a database file can't be opened, specify
  "daemon_table_open_error_is_fatal = yes".
- Revised postconf(1) command. It warns about unused parameter
  name=value settings in main.cf or master.cf (likely mistakes),
  understands "dynamic" parameter names such as names that depend on
  the name of a master.cf entry (finally, "postconf -n" shows all
  parameter settings), and it can display main.cf and master.cf in a
  more user-friendly format (postconf -nf, postconf -Mf).
- Read/write deadline support in the SMTP client and server to defend
  against application-level DOS attacks that very slowly write or read
  data one byte at a time.

Status:

Vendor Tag: VENEMA
Release Tags: PFIX-2-9-5

(tron)

Expose the pseudo-errors if _KMEMUSER is defined (as well as _KERNEL).
CVS :----------------------------------------------------------------------

(dsl)

Put the kernel function prototypes inside #ifdef _KERNEL

(dsl)

These files don't need compat/osf1/osf1_cvt.h

(dsl)

amd64's machine/reg now includes fpu.h (to define 'struct fxsave64)
so no need to include signal.h here - which shouldn't result in
the definitions in fpu.h being included anyway.

(dsl)

There is some very dubious code to locate the values of
tcb.rcv_adv, tcb.snd_nxt and tcb.rcv_nxt for a socket.
so set CPPFLAGS+=      -D_KMEMUSER

(dsl)

fix mtree sets

(christos)

Test the netbsd and mtree flavors.

(christos)

Don't define __BIT/__BITS if __ASSEMBLER__ is defined.
(since they use C syntax they won't work anyways).

(matt)

I've NFI why this code has userspace stuff playing with 'struct proc *'
  nor why it uses &p0 as a magic constant (rather than NULL).
Re-instate the definition of p0, but enable the 'fake' definition
  of 'struct proc' if the relevant part of sys/proc.h seems to have
  been #if'ed away.
Should fix the build.

(dsl)

# There is some very dubious code to locate the values of
# tcb.rcv_adv, tcb.snd_nxt and tcb.rcv_nxt for a socket.
so add: CPPFLAGS+=    -D_KMEMUSER

(dsl)

Deal with _LIBC (aliases, etc).
Add missing #endif.

(matt)

Rename strlen_armv6.S to strlen.S since this is no longer armv6 dependent.

(matt)

Really expand the name, not the definition, as the comment says.

(apb)

we define HAS_ISBLANK in the Makefile

(christos)

Remove check for GLOB_NOCHECK - the behaviour has been changed again, so it
will return a modified pattern.

(martin)

pstat is a kmem groveller, set -D_KMEMUSER in CPPFLAGS

(dsl)

Explicitly include sys/resource.h

(dsl)

Explicitly include sys/ioctl.h

(dsl)

Replace alloca(256) with an array.
The code ended up using alloca() from libc - which 'just can't work' (tm).
It is also pointless!

(dsl)

Use NULL not 0 for pointer assignment.

(skrll)

Correct a comment

(skrll)

KNF a comment.

(skrll)

Add an assembly version of strcpy/strncpy/strlcpy.
(they all use a common source with defines to determine which to build).

(matt)

If EDID is supplied, parse and print it. While here, try not to crash
if is_console is false.

(jmcneill)

Read EDID data on Beagleboard and pass it to omapfb driver if found.

(jmcneill)

add a function to lookup the i2c controller tag for the specified device_t

(jmcneill)

Comment out the unused 'struct proc p0;'
This is userspace, it isn't supposed to know what a 'struct proc' is.

(dsl)

This needs an explicit include of sys/ioctl.h.
XXX: There is some crappy code in here :-)

(dsl)

Explicitly include sys/ioctl.h

(dsl)

Explicitly include sys/ioctl.h

(dsl)

Don't rely on sys/lwp.h includeing sys/resource.h for us.

(dsl)

Don't rely on something including sys/ioctl.h for us.

(dsl)

Hook up generic_bs_sr_4 for bus_space_set_region_4 in the OMAP bus space.

(jakllsch)

remove dead assignment (Christoph Mallon)

(christos)

Fix pasto that affected bind -k (Christoph Mallon)

(christos)

add OMAP3 SCM (temp sensor)

(jmcneill)

Add OMAP3530 temperature sensor support

(jmcneill)

fix incomplete comment

(apb)

add a compile option to override the fallback to software drawing of bitmap
fonts on R3xx ( options RADEONFB_ALWAYS_ACCEL_PUTCHAR ) since this works just
fine on my RV350.
Next step: don't map VRAM if we don't need it.

(macallan)

units(1): add -l and -L options.  [apb 20130101]

(apb)

Add -l and -L options to units(1).  "-l" simply lists all unit
definitions, while "-L" alsoreduces them to depend only on a few
primitive units (such as m, kg, sec).

(apb)

Correctly ignore duplicate definitions.  "continue" applies only to the
innermost loop.

(apb)

Happy 2013 everybody

(spz)

add a sync to make it less likely to boot with /etc pointing to nothing after
a crash.

(christos)

add TPS65950 RTC support

(jmcneill)

reset todr clock on shutdown

(jmcneill)

add support for TPS65950 watchdog timer

(jmcneill)

avoid strict aliasing problems

(christos)

Move the two fields used to save some i387 state on the last fpu trap
  into their own sub-structure of the pcb (from 'struct savefpu').
They only (seem) to be used in some code that generates core dumps
  for 32bit processes (code that might be broken as well!).
'struct safefpu' is now identical to 'struct fxsave64'. One (or both)
  needs extending to support AVX - might need to be dynamically sized.
Removed all the __aligned(16) except for the one in struct pcb itself.
  Only the copy used for the fsave instruction need be aligned.

(dsl)

new awk, ping update

(christos)

Add support for '%n' being a shorthand for 'fg %n'.

(dsl)

add tps65950 driver

(jmcneill)

TI TPS65950 driver, right now about all it can do is control LEDs

(jmcneill)

beagleboard supports 8-bit SD mode

(jmcneill)

add I2C support

(jmcneill)

Add OMAP3530 I2C support

(jmcneill)

fix some typos & comments

(macallan)

R4xx support is untested at best, so don't pretend to support them.

(macallan)

clean up some more confusion between (ws)displays and CRTCs.
Now palette initialization no longer stomps over the port table, which gives
us a fighting chance to intentionally enable the right outputs.
How on earth did this ever work?

(macallan)

At each function level print a useful offset (as also shown in rlv)

OK from gimpy

(skrll)

Make this work on all ARMs but keep the armv6 optimizations.  It as fast as
the existing strlen for small string and once strings are 8 bytes or more in
length it starts getting significantly faster.  For really long strings,
compared to the existing strlen, this uses about 1/2 of the cycles for the
non-armv6 version and about 1/3 of the cycles for the armv6 version.

(matt)

Always re-enable the VFP when loading for a kernel LWP.

(matt)

Correct one of the kernel_size calculations.

(matt)

Remove a debugging printf

(matt)

Switch to using vfp_kernel_{acquire,release} so that softints don't cause
the VFP to become disabled.

(matt)

print the PC of the VFP kernel fault in the panic message.

(matt)

bump for arch linux/rpi support.

(pooka)

Take into account armv6 hacks from common/lib/libc/arch/arm/atomic to
allow this to build with -march=armv6k

(pooka)

fix function name in error messages.

(christos)

PR/46579: Takahiro HAYASHI: ifconfig clone destroy should not print useless
diagnostic and exit with an error code. Get the interface name and flags
opportunistically to allow the code to return normally if it does not need
to do anything.

(christos)

typo in comment, from Abhinav Upadhyay in PR 47385

(dholland)

size_t needs to printed with %zu

(pooka)

No need to check both TARGET_64BIT and ix86_preferred_stack_boundary >= 64,
if the former is true the latter is also true.

(dsl)

Fix failure message for ecalloc.

(dholland)

If malloc, calloc, or realloc returns NULL when a size of 0 was
requested, which is allowed by pertinent standards, honor it instead
of bombing.

Do not do this for calloc(x, y) where x != 0 && y != 0 but x*y == 0;
in that case bomb.

(dholland)

If -march=native is specified, only enable AVX if the processor supports
  it and the OS has enabled XGETBV for application use.
It might need to also check XCR0[2] (having executed XGETBV) to check that
  the kernel actually supports saving the YMM registers, but I suspect the
  kernel might defer setting that until the first fault.
See vol 1 section 13.5 of the Intel SDM (intel_x86_325462.pdf).
Fixes toolchain/45673

(dsl)

Remove a couple of items.

(skrll)

Fix bug where menu mark was not displayed on menus with O_ONEVALUE set
Set a default menu mark

(blymn)

FUSE seems to allow short writes without errors but PUFFS doesn't. Work
around this by returning ENOSPC in case of a short write to avoid protocol
errors. This change is based on problem analysis provided by Antti Kantee.

This fixes PR lib/45129 by myself.

(tron)

Add support for R3xx.
Whi

(macallan)

Always call brelse() on error for breadn() too.

(hannken)

deduct phdrlen after computing max size.

(christos)

1. Allocate the max packet size before accounting for phdrlen, harmless.
2. In the clear-route-cache sendto, don't send 0 bytes (if -s was specified
  with < 8, phdrlen would be 0).
3. Always send ICMP_MINLEN packets; this is what everyone else does. Makes
  ping -s n where n < 8 work.
4. The condition for checking the data bytes was completely wrong. only check
  the data bytes if we got all of them.
5. The condition for printing a newline was wrong; before it would not print
  a newline before printing the data bytes, and it would append to the previous
  error message.

(christos)

- use warn/err
- if both users are on the same machine, use the loopback address. This
  allows us to use talkd from inetd listening only to the loopback.

(christos)

better messages, more quoting.

(christos)

don't leak a vnode on error

(christos)

Always call brelse() on error. Otherwise a possible error from bread() will
cause the buffer to stay lock and we end up blocking forever in
VOP_CLOSE->spec_close->vinvalbuf->bbysy since the buffer is marked busy
but there is no I/O pending.
This caused my laptop to hang on boot_findwedge because:
    findroot: unable to read block 358331527 of dev dk0 (22)

(christos)

kill Id RCS keyword.

(christos)

The sanity check prevented messages that carry only ancillary data.

(mlelstv)

add missing prototype for crc32v

(jmcneill)

add support for generating boot.scr scripts with -T script

(jmcneill)

remove useless cast.

(christos)

Update for comment of OVERO.  It support Overo COMS.

(kiyohara)

Change size of some devices.  from DM37xx and OMAP 35xx datasheet.
Support Overo *STORM.  Tested only Overo EarthSTORM(Silverload).

(kiyohara)

Change order for TLB operation.
We can boot Overo EarthSTORM.

(kiyohara)

- defer session cleanup to not force detachments
- use more and explicit locking
- improve connection recovery
- use larger timeouts
- handle ccb buffer underflow correctly
- simplify throttling code

Sessions can now temporarily exist without a valid
connection, you also need to update iscsid(8).

(mlelstv)

If an untraced process sleeps in recvmsg/sendmsg, the syscall does not
allocate an iov structure for ktrace. When tracing is then enabled
and the process wakes up, it crashes the kernel.

Undo the last commit which introduced this error path.

Avoid the mentioned kmem_alloc assertion by adding a sanity check analog
to similar code in sys_generic.c for I/O on file handles instead of
sockets.

This also causes the syscall to return EMSGSIZE if the msg_iovlen member
of the msg structure is less than or equal to 0, as defined in
recvmsg(2)/sendmsg(2).

(mlelstv)

Handle invalid messages more gracefully, handle sessions with no valid
connections, and be more verbose about errors.

(mlelstv)

A few slight speedups (remove one instruction from the main loop).

(matt)

Enable the watchdog driver on BeagleBone.

(riz)

In kernel tickle mode, tickle once immediately upon setting the mode.
Works around a peculiarity where the first watchdog period (only) on the
BeagleBone is only half as long as expected.

(riz)

merge 2012-12-20

(christos)

from http://www.cs.princeton.edu/~bwk/btl.mirror/

Status:

Vendor Tag: NAWK
Release Tags: NAWK20121220

(christos)

no need to print irq twice

(jmcneill)

dont set SDHC_FLAG_SINGLE_ONLY for OMAP3530 either

(jmcneill)

enable WDT2 (MPU) for watchdog instead of WDT3 (IVA2). remove WDT1, it doesnt exist on OMAP3530

(jmcneill)

Add acpi0 at mainbus - it is not realy optional.

(martin)

Rename "tappit-hen" to "tappithen" because unit names cannot contain
hyphens.  Also change the definition from "3 imperial" (a dimensionless
quantity) to "3 bottle" (2.25 litres).  Wikipedia and several other web
sites say 3 bottles or 2.25 litres, but
<http://www.christies.com/lotfinder/wine/chateau-lafite-rothschild-vintage-1811-1-tappit-hen-per-4992576-details.aspx>
says:

    There seems to be no firm definition of the bottle size referred to
    as "tappit-hen". Believed to be of Scottish origin, a tappit-hen
    can vary, according to which definition one follows, from 1 1/2
    bottles to a tregnum or 3-bottle bottle. It is safest to say that it
    approximates to a magnum.

and ends with:

    1 Tappit-Hen (2.25 liters) per lot

(apb)

Allow a number and a unit to be juxtaposed without an intervening space.
Now "litres/100km" works as desired, instead of silently being
treated as "litres/100".

(apb)

Add "wide" to description. From OpenBSD via Ingo Schwarze.

(wiz)

Fix references to undefined units:
* "southamericarand" should be "southafricarand";
* "sidereal day" should be "siderealday";
* "perusol" should be "perunewsol".

(apb)

Improve the description of the [count] from-unit to-unit arguments.

(apb)

Fix a bug that a multicast filter entry is incorrectly overrode
on some cases.

(msaitoh)

Move the initialization of n to after the error branch.

From Taylor R Campbell <riastradh@netbsd.org>

(hannken)

strlen implementation for armv6 and later.  Uses clz and uqadd8 to really
speed the search for NUL.  as fast as normal strlen at about a length of
6 or 7 and 2-3 times faster starting around 10.

(matt)

Simplify the xen event handler callback by:
- moving the interrupt handler callback traversal into a separate
  function.
- using evt_iterate_bits() to scan through the pending bitfield
- removing cross-cpu pending actions - events recieved on the wrong
  vcpu are re-routed via hypervisor_send_event().
- simplifying nested while() loops by encapsulating them in
  equivalent functions.

Many thanks for multiple reviews by bouyer@ and jym@

(cherry)

Debug.  This becomes faster than the normal strlen at about 80 characters.

(matt)

Enable DCache Streaming Switch and Write Allocate.
This change improve system performance significantly.

(msaitoh)

Convert to mdoc and fix stuff while here.

(wiz)

sprinkle const, no functional change intended.

(christos)

Instead of changing the code to match the documentation, change the documentation
to match the code. NOCHECK is used only by csh(1) and csh(1) is too broken for
words (histchars quoting is "special").

(christos)

ia64 mknative results for gcc and binutils

(martin)

Make it all compile

(martin)

Consistent/Correct error message from failing usbd_set_config.

Use aprint_error_dev.

(skrll)

- reset the chip if the tx engine gets stuck after a link state change,
  from OpenBSD
- no need to do a full reset of the chip when enabling or disabling
  promiscuous mode

(jmcneill)

Hide inline functions from userland

(martin)

Since cvs does not provide a version number for the entire tree, provide
identifiers which can be used by external parties to check if the NetBSD
source tree is of required vintage.

inspired by a buildrump.sh bug report due to an out-of-date src

(pooka)

better patch (no functional change); centralize freeing in one place.

(christos)

- hit me with your parenthesis stick
- remove extra void *

(christos)

Add some usb devices.

(skrll)

Note another ia64 compiler workaround

(martin)

Compile bignum.c with -O1 only on ia64 to avoid a gcc bug

(martin)

Add fpclassify{,d}

(martin)

Provide sig_atomic_t.

(martin)

Note libelf ia64 hack

(martin)

ia64 compiler bug workaround: reduce -O to 1 for one file

(martin)

Add WCHAR_MIN and WCHAR_MAX.
Why don't we use the compiler supplied __WCHAR_MIN__ and __WCHAR_MAX__ for
this?

(martin)

Document libc hesiod.c compiler workaround on ia64

(martin)

Avoid gcc 4.5.3 compiler bug on ia64 by compiling hesiod.c with -O1 only.

(martin)

Add a MID for IA64 (also it never used a.out, we need it for kernel crash
dumps)

(martin)

Remove unused header evtchn.h from intr.h

(cherry)

Add support for WARN_REFERENCESE and __KERNEL_RCSID for assembler files.

(martin)

prevent double free on error.

(christos)

Install headers needed for userland build.
Add missing accessors for pthread and TLS support - just dummies for now.

(martin)

Build glue for ia64

(martin)

Note NEON works on arm now.
Note new PCU changes.

(matt)