Pull up following revision(s) (requested by spz in ticket #39):
sys/arch/ews4800mips/sbd/fb_sbdio.c: revision 1.16
sys/arch/pmax/ibus/pm.c: revision 1.13
sys/dev/hpc/bivideo.c: revision 1.34
sys/dev/ic/sti.c: revision 1.19
correct size checks so they cannot be circumvented by integer overflows
reported by CTurt, thanks for the notification

(snj)

Pull up following revision(s) (requested by nat in ticket #38):
sys/dev/audiobell.c: revision 1.21
sys/dev/audiobell.c: revision 1.22
Initialize and set all required parameters for the audiobell.
--
Revert back to a 16 bit phase counter - stops overflow when performing
calculations on pitch and phase.

(snj)

For Attention Indicator Control bit and Power Indicator Control bit, it's
allowed to be a read only value 0 if corresponding capability register bit
is 0.

(msaitoh)

Pull up following revision(s) (requested by christos in ticket #36):
external/bsd/tmux/usr.bin/tmux/utempter.c: revision 1.2
PR/52288: ben: Fix utmp cleanup did not work.

(snj)

Pull up following revision(s) (requested by roy in ticket #35):
external/bsd/dhcpcd/Makefile.inc: revision 1.2
Replace @DATADIR@ with /usr/share/examples as pointed by plunky@

(snj)

Pull up following revision(s) (requested by bouyer in ticket #34):
sys/netcan/can_pcb.c: revision 1.6
tests/net/can/t_can.c: revision 1.6
Refuse to bind to a non-CAN interface.
Also release the lock in the error branch.
--
Test bind()ing to a non-existent interface.

(snj)

Pull up following revision(s) (requested by sborrill in ticket #33):
sys/dev/pci/ehci_pci.c: revision 1.65
Deal with broken BIOSes that leave PCI interrupts disabled.
Thanks to nick@

(snj)

Simplify

We can assume that rt_ifp is always non-NULL.

(ozaki-r)

Add mink, version 20170614, to othersrc. From the README:

This is a small library and utility program which transforms input
data into output data, concealing and revealing along the way.  Rather
than just do one round of encryption, this can do multiple rounds on
the same data.  All kinds of bit and byte rotation, mirroring around a
diagonal, Xor, and AES are provided, including raw AES, and CBC, CTR,
EAX and GCM modes.

Small awk-like scripts are written to transform the data; the same
script is used in each direction.

Keys, IVs and AADs are specified in the script, or can be read from
the terminal, working from a prompt. This is true for strings used
to xor data.

Thus, to extract the original from a file that has already been
transformed, you need to know not only the keys/IVs/AADs being used,
you need to know all the arguments used for bitwise rotations,
repetitions of AES encryption, and order of transformations specified
in the script. To discourage timing attacks, random data is encrypted
a random number of times after each transformation in the script.

It may be best to describe this in a worked example:

% echo 'Input data' | mink -e 'bswap 32' > scrambled
% hexdump -C scrambled
00000000  75 70 6e 49 61 64 20 74  00 0a 61 74 00 00 00 00  |upnIad t..at....|
00000010
% mink -c 11 'bswap 32' scrambled
Input data
%

But mink is designed to work with multiple encryption transformations:

% echo 'Input data' | mink -e 'bitrotr 3; bswap 32' > encrypted
% hexdump -C encrypted
00000000  ab 83 73 4a 0b 23 01 a3  00 50 0b a3 00 00 00 00  |..sJ.#...P......|
00000010
% mink -c 11 'bitrotr 3; bswap 32' encrypted
Input data
%

And the instructions are usually entered into a file, for ease of use of
decryption:

% cat 1.mink
bitrotr 3
bswap 32
% echo 'Input data' | mink -e -f 1.mink > encrypted
% mink -f 1.mink -c 11 encrypted
Input data
%

However, the main way that mink shines is when using more advanced methods
of (symmetric) encryption. And, for these, we need keys and IVs:

% sed 15q < /usr/share/dict/words > in
% wc in
      15      15      92 in
% cat 2.mink
setkey 0 "key number one"
setkey 1 "IV?"?
bitrotr 3
aes128-ctr 0 1 2367
%

After the first 2 commands, we have a 92byte file to work with.  We
also have introduced 128 bit AES in CTR mode, using 2 keys, and
running for 2367 repetitions.  The "setkey" lines set up the strings
to be used for keys and IVs.  Hex escapes (\xAB) can be used to
express 8bit entities.  If a setkey string is followed by a question
mark, the string is taken to be a prompt, and the user is asked to
enter the string.  getpass(3) is used for this.  Strings are assigned
to numeric slots, which do not need to be contiguous or sequential,
but are easier to remember that way.  The invocation of aes128-ctr
above means "use string in slot 0 as the key", use the string input by
the user as an IV, and repeat the ctr encryption 2367 times.

% mink -e -f 2.mink -o encrypted in
IV?
% hexdump -C encrypted
00000000  c2 16 39 77 9c 98 fd 4d  52 d0 27 17 35 03 e4 f0  |..9w...MR.'.5...|
00000010  15 5c a0 78 ad 19 77 21  c4 f0 f9 e3 ee a5 b4 1c  |.\.x..w!........|
00000020  bd 37 06 ad a4 d6 53 f8  ed de 12 8c 05 95 cc 55  |.7....S........U|
00000030  c4 08 fd cb 8c 1f b5 af  7c 00 50 75 04 d1 2e 3b  |........|.Pu...;|
00000040  63 70 53 5e ba 8b 41 42  2f 3e aa 81 fe 48 5c e7  |cpS^..AB/>...H\.|
00000050  a9 7c 8a 33 13 d4 d9 af  fe 6a d4 be 3e c5 7d 9f  |.|.3.....j..>.}.|
00000060
% mink -f 2.mink -c 92 encrypted
IV?
A
a
aa
aal
aalii
aam
Aani
aardvark
aardwolf
Aaron
Aaronic
Aaronical
Aaronite
Aaronitic
Aaru
% # the user entered the wrong IV in the next invocation of mink
% mink -f 2.mink -c 92 encrypted | hexdump -C
IV?
00000000  21 82 48 b5 93 dd c9 c9  47 56 10 ee 98 43 ab 91  |!.H.....GV...C..|
00000010  53 96 b3 3d 61 9a fc 87  5a 8f 06 d9 0b 40 47 1d  |S..=a...Z....@G.|
00000020  8f 00 95 89 ef 79 7a ed  c5 fa 5b 01 0c 8f 58 8c  |.....yz...[...X.|
00000030  f8 11 52 fe d9 86 4b fa  71 a8 8b 6b 6f d6 59 31  |..R...K.q..ko.Y1|
00000040  ad 29 60 64 ba 27 e0 d2  3c 4c f2 6d 39 11 cd f4  |.)`d.'..<L.m9...|
00000050  8c b3 dc d3 b8 5e d0 48  6e ee 99 61              |.....^.Hn..a|
0000005c
%

Please also note that the number of repetitions for operations is important:

% mink -v 0=zero -v1=two -e 'aes256-ctr 0 1 1330' Makefile | hexdump -C
00000000  31 3b 82 97 22 e4 ef 26  a7 f3 61 66 02 81 94 4c  |1;.."..&..af...L|
00000010  12 9f a5 63 cf 95 68 ab  00 46 c3 cd b7 c3 35 f4  |...c..h..F....5.|
00000020  0d 99 cb 11 c8 0c df ff  88 85 54 7a e9 4c 66 7b  |..........Tz.Lf{|
00000030  cd b5 7d 27 c0 45 70 d8  cd 85 dc 2c 11 e7 df 07  |..}'.Ep....,....|
00000040  aa e2 06 d1 85 a8 96 8d  4f d6 f4 8d ab 22 4a b4  |........O...."J.|
00000050  12 60 30 31 43 63 5c 73  36 1e d6 77 21 6d ff 9e  |.`01Cc\s6..w!m..|
00000060  7c 6e 52 65 7e 4c 6b d3  4d 8c 5d 2f e3 6c 87 b9  ||nRe~Lk.M.]/.l..|
00000070  87 d6 f6 ad 84 f9 22 72  d3 46 80 7c ce fb 87 2b  |......"r.F.|...+|
00000080  b0 7f 89 13 73 5a 3e 75  b3 37 60 cf e7 a0 65 4e  |....sZ>u.7`...eN|
00000090
% mink -v 0=zero -v1=two -e 'aes256-ctr 0 1 1331' Makefile | hexdump -C
00000000  c3 0e b6 08 d1 56 0b 18  c5 2a 80 97 d2 64 85 b2  |.....V...*...d..|
00000010  d1 42 b8 92 33 00 fd c5  ae ae a8 c5 8f 50 ce 3f  |.B..3........P.?|
00000020  18 64 ea fc f8 06 17 ae  62 9e c2 eb 1d d3 4d 84  |.d......b.....M.|
00000030  15 bb b4 da 6a 0c b4 3b  fe 4e 55 41 4e 43 75 64  |....j..;.NUANCud|
00000040  62 70 ad 45 fd 06 5a 87  dc 0d 62 71 9b 0d ab d7  |bp.E..Z...bq....|
00000050  0f ba ab 64 28 10 f5 fb  fd 05 60 9e 89 38 f2 e9  |...d(.....`..8..|
00000060  e6 fd 78 76 24 f3 da 82  f3 22 94 bb 8b 30 92 d7  |..xv$...."...0..|
00000070  5f 38 94 16 f4 d9 02 2b  89 37 49 e6 f3 93 23 be  |_8.....+.7I...#.|
00000080  34 f7 43 c2 c3 69 ec 93  76 62 f0 ba c1 2c 96 00  |4.C..i..vb...,..|
00000090
%

After the data has been encrypted by mink, anyone wanting to read the
original data needs to know ALL of:

1. the exact statements of the transformation script, i.e. the bitrotr, bswap,
aes128/256 commands

2.  the order of these commands - for example, AES encryption before
using a bit rotation will produce a different result to that where the
same bitwise rotation is done before AES:

% echo 'Input data' | mink -e 'setkey 0 "key"; setkey 1 "iv" ; aes128 0 1 3; bitrotr 4' | hexdump -C
00000000  ef b3 4e 04 a1 13 47 01  3b 16 d9 66 ef 3a da 00  |..N...G.;..f.:..|
00000010
% echo 'Input data' | mink -e 'bitrotr 4; setkey 0 "key"; setkey 1 "iv" ; aes128 0 1 3' | hexdump -C
00000000  89 36 80 e8 32 4d 74 62  2c 9a 02 ea 28 a9 74 6b  |.6..2Mtb,...(.tk|
00000010

3.  The keys and IVs used for any AES operations, and any AAD used in
AEAD transformations such as EAX or GCM, or strings used in XOR
transformations

4. The number of repetitions of AES encryption to perform

Arbitrarily complex scripts can be used, which would help if proof of work
is to be a part of the data transformation.

The list of transformation methods is as follows:

aes(128|256)-cbc key iv reps
aes(128|256)-ctr key iv reps
aes(128|256)-eax key iv aad reps
aes(128|256)-gcm key iv aad reps
aes(128|256) key
bitrotl nbits
bitrotr nbits
byteswap nbits
diagonal number
reverse
rol nbytes
ror nbytes
xor string

In addition, these transformations can be further specified to apply
to a range of input data - the offset and length can be appended to
each of the instructions above, using the form off:n len:m, where n
and m are both decimal integers.

After every instruction is executed, a random number of iterations of
AES encrypting random data are performed.  This is intended to
minimise the effect of timing attacks.

Alistair Crooks
Wed Jun 14 18:08:15 PDT 2017

(agc)

constify more to avoid gcc stupidity.

(christos)

postfix-3.1.6/3.2.2 out.

(wiz)

+IME

(dholland)

put back reset of retry count to 0 when starting processing a buf

do not use NCQ for retried transfers to speed up recovery, unless FUA
is requested

(jdolecek)

Check argc, and add a message.

(maxv)

Make the PMC syscalls privileged.

(maxv)

unbreak static build

(christos)

Disable interrupts for T_NMI (inline calltrap). Note that there's still a
way to evade the NMI mode here, if a segment register faults in
INTRFASTEXIT; but we don't care. I didn't test this change, but it seems
fine enough.

(maxv)

Unbreak MKPAM=no build

Make cron(8) linkage with PAM conditional.

(kamil)

style

(maxv)

style

(maxv)

remove now unnecessary casts.

(christos)

- cast ident argument because it is usually an fd (signed) and generates
  conversion to unsigned warnings.
While here:
- name the macro parameters with meaningful names
- rename the internal inline function as _FOO instead of FOO_ because there
  is no other FOO_ name in the system headers.

(christos)

Give the direct map 32 slots (16TB of va). This matches MAXPHYSMEM, in
such a way that the direct map is no longer the limiting factor for high
memory systems.

(maxv)

Move the direct map from slot 509 to slot 460. We will increase its size
dynamically.

(maxv)

Define MAXPHYSMEM globally.

(maxv)

Stop including <sys/user.h> in userland libraries

This header in this context is freebsdism.

(kamil)

Detach <sys/user.h> from sanitizer_procmaps_netbsd.cc (GCC)

This header in this context is freebsdism.

Sponsored by <The NetBSD Foundation>

(kamil)

gcc-7.1 out.

(wiz)

Add wsbell

(pgoyette)

Add EFER_TCE. This would be an interesting feature to have, since it
reduces the indirect cost of invlpg; but I'm not convinced the way we
flush upper-levels is correct for this yet.

(maxv)

Fix a bug introduced in bus_space.c::r1.39. This check too is hard-coded.
Might have had a cumulative effect on PR/52000.

(maxv)

Fix a pretty dumb mistake I made in r1.22: the alignment needs to be in the
bss, otherwise the bootloader will use memory before __kernel_end and give
a wrong start pa to the kernel.

This issue was investigated by Anthony Mallet. Should fix PR/52000.

(maxv)

support multiple encryption drivers (port from FreeBSD).

(knakahara)

refactor crypto_newsession() like FreeBSD.

(knakahara)

must release cap->cc_lock before calling cap->cc_newsession() because of spinlock.

(knakahara)

Ensure to detach children.

(nat)

Instead of directly referencing a parent device's code (ie, pcppi_bell()
routine), let the parent device pass a pointer to the code (in the aux
config data).  This allows us to load the spkr module without requiring
the pcppi parent device to exist.  (The spkr device can also have an
audio as parent.)

(pgoyette)

tickets 28-32

(snj)

Pull up following revision(s) (requested by maya in ticket #32):
external/gpl3/gcc/usr.bin/gcc/arch/mips64el/auto-host.h: revision 1.11
external/gpl3/gcc/usr.bin/gcc/arch/mips64el/configargs.h: revision 1.21
regen after new binutils

(snj)

Pull up following revision(s) (requested by jmcneill in ticket #31):
sys/arch/evbarm/fdt/fdt_machdep.c: revision 1.7
Parse more than one entry from the /memory node's reg property.

(snj)

Pull up following revision(s) (requested by jmcneill in ticket #30):
sys/dev/fdt/fdt_intr.c: revision 1.11
Fix an issue with interrupt controller lookup wrt cascading interrupt
controllers.

(snj)

Pull up following revision(s) (requested by flxd in ticket #29):
external/gpl3/gcc/dist/gcc/config/vax/builtins.md: revision 1.6
Fix PR port-vax/51761 as suggested by Paul Koning on port-vax list.
Installation (install.ram, -Os) on my VS4000 is possible without SCSI
timeouts again.
Other variable-length bit field instructions should be checked for correct
constraints, too!

(snj)

Pull up following revision(s) (requested by pgoyette in ticket #28):
sys/arch/x86/x86/cpu.c: revision 1.131
Further reduce the loop counter so that hatching completes before the
boot processor times us out.
Add a nice big XXX comment for why the counter is so low.

(snj)

Enable DEBUG for babylon5

(ozaki-r)

Add test cases for vlan(4)

From s-yamaguchi@IIJ

(ozaki-r)

KNF

(ozaki-r)

create an nmap table for module symtabs too.
needed by dtrace.

(chs)

add an lwp_trapframe() interface to return an LWP's user trapframe.
needed by dtrace.

(chs)

use MVSATA_EDMAQ_LEN instead of magic number for number of queue slots

(jdolecek)

correct size checks so they cannot be circumvented by integer overflows

reported by CTurt, thanks for the notification

(spz)

another one

(dholland)

Now that bus_space can use devmap - use it for early console

(skrll)

Use devmap if available

(skrll)

move include of <sys/iostat.h> back to where userland will get it.
some third-party software relies on this header including that one.
pointed out by joerg.

(chs)

gdb-8.0 out.

(wiz)

PR/52295: Anthony Mallet: find -delete: "relative path not safe" with absolute
paths ... Fix from OpenBSD

(christos)

Sync with FreeBSD r316541:

> Fix a double free in ixgbe_rxeof()
>
> Submitted by: rstone
> MFC after: 1 week
> Differential Revision: https://reviews.freebsd.org/D10255

(msaitoh)

Fix the partial chain check in ixgbe_rx_discard(). This bug was addded in
rev. 1.33.

(msaitoh)

Add wsbell.

Found by wiz@.

(nat)

Avoid variable lenght buffers to help the stack protector (hopefully no
functional change)

(martin)

Fix typo in macro name.

(wiz)

Update for wsbell(4) manual pages.

(nat)

Update for inclusion of references to the new wsbell(4)

(pgoyette)

Manual pages for wsbell.

(nat)

Update the speaker man page, mentioning wsbell and new formatting for
ioctls.

Ok pgoyette@.

(nat)

Implement missing ioctls for the default bell.

(nat)

Change sc_dying from u_char to bool.

(nat)

Remove dying ftom vb_args and use sc_dying instead.

(nat)

Move duplicate definition of SETBELL macro into wsbelldata.h.

(nat)

Restore Makefile.inc so dhcpcd is installed correctly as pointed out by kre@.
This reverts my prior commit.

(roy)

Add NCQ support to siisata(4).

There are still issues that need to be address before I consider this
mergable.

(jakllsch)

Macro argument parentheses paranoia.
Annotate another register initialism.

(jakllsch)

Add a few more NCQ branch action items.

(jakllsch)

update notes

(jdolecek)

add support for NCQ on Gen IIe hardware

mvsata_quetag_get()/put() pairs adjusted to work also for non-DMA code paths,
i.e. during probe; replaced wdcintr() call with local routine which uses
currently active command tag, rather than assuming always tag 0 on that
code path

~55MB/s -> ~80MB/s sequential read using fio(1) on a test HDD, quite nice

(jdolecek)

Add heading comment for private localcount_adjust subroutine.

(riastradh)

Move forward declaration to top of file.

Keep header comment above localcount_init adjoined to it.

No functional change.

(riastradh)

Use continue to denote the no-op loop to match netbsd style
newline for extra clarity.

(maya)

Fix typo
Insert Pp after a list to add vertical space before the next paragraph
Remove Pp before Bl at two places

(abhinav)

This Makefile fragment is no longer used.

(roy)

Replace @DATADIR@ with /usr/share/examples as pointed by plunky@

(roy)

Note addition of wsbell(4)

(pgoyette)

Deal with broken BIOSes that leave PCI interrupts disabled.

Thanks to nick@

(sborrill)

Restore wscons keymaps feature on hpcarm

hpc ports need to alter keydesc data at runtime in order to load
alternate keymaps. But since keydesc is const initialized data, it
is mapped read only and the operation should fail.

It seems older compiler failed to enforce the read-only mapping and
this is why it used to work, but on recent NetBSD releases, the
feature is broken.

We fix it by duplicating the keydesc data once into a malloc'ed area
that can be modified.

(manu)

There is no longer a NWSBELL (since nothing depends on it), so there's no
need to #include wsbell.h (which is where it used to be defined, as a
result of a needs-flag in files.wscons).

Remove the #include

(pgoyette)

Revert previous - it's no longer needed since data structure sizes no
longer depend on which wscons child devices exist

(pgoyette)

Always include declarations/definitions for NWSMUX and NWSDISPLAY even
if they aren't needed.  This enables us to have structures of a fixed
size regardless of which child devices are configured, which enables
better modularization.

(pgoyette)

Nothing needs the NWSBELL flag, so don't bother creating it

(pgoyette)

For now, #include wsdisplay.h to get the definition of NWSDISPLAY

The symbol is currently used to conditionally include members of a
data structure, so without this include there could be different
structure definitions, with different sizes, and data corruption is
likely.

XXX For cleaner modularization, all of the needs-flag entries in
XXX wscons should be removed.

(pgoyette)

Use IFM_1000_KX and IFM_2500_KX.

(msaitoh)

Descend into SUBDIR to actually build the wsbell module.

Missed in previous commit - thanks kre@

(pgoyette)

Add new wsbell(4) module to the sets list.

(pgoyette)

Actually create the wsbell(4) module

(pgoyette)

Add module infrastructure for wsbell(4)

(pgoyette)

PR/52292: Shinichi Doyashiki: Fix reversed comments.

(christos)

Implement xxx_rescan() and xxx_childdet() functions; these will be
needed when child device wsbell(4) becomes a separately-loadable
module.

(pgoyette)

Remove the 'needs-flag' from spkr.c - it's not needed by anything.

(pgoyette)

Do not partially initialize a static variable.

(joerg)

Parse more than one entry from the /memory node's reg property.

(jmcneill)

Grammar and punctuation fixes in few sentences
Remove Pp before Bd at couple of places

(abhinav)

Insert missing word
Insert comma in a sentence to make the comprehension easier
Remove Pp before Bd

(abhinav)

Simplify MCT; just enable it and then attach an ARMv7 generic timer.

(jmcneill)

Use fdtbus_intr_establish to hook in block interrupts instead of
intr_establish.

(jmcneill)

PR/52288: ben: Fix utmp cleanup did not work.
XXX: pullup-8

(christos)

Revert back to a 16 bit phase counter - stops overflow when performing
calculations on pitch and phase.

(nat)

Initialize and set all required parameters for the audiobell.

(nat)

Fix an issue with interrupt controller lookup wrt cascading interrupt
controllers.

(jmcneill)

- Replace CONSADDR with SSCOM2CONSOLE in example
- Remove gtmr (Exynos5422 uses mct)
- cinclude EXYNOS.local instead of TEGRA.local

(jmcneill)

Remove ODROID-U and ODROID-XU kernel configs as they no longer work.

(jmcneill)

Remove references to the PCPPISPEAKER and VAUDIOSPEAKER options - they
don't actually exist.

XXX This man page will need a major overhaul.  But first we need to get
XXX man pages for wsbell(4) and wsbell(9) (and maybe other updates, too)

(pgoyette)

Clean up the spkr_modcmd() routine so it returns ENOTTY for all commands
other than MODULE_CMD_{INIT,FINI}

(pgoyette)

Define the NWSMUX symbol for the spkr module (for modules, we cannot
use the wsmux.h file created by the needs-flag in sys/conf/files)

(pgoyette)

The wsmux.h file is only created for kernel builds.

(pgoyette)

New device wsbell - allows for a console beep for non pckbds (usb etc).
Works for platforms without pcppi - can work with spkr at audio and spkr
at pcppi.

To use add the following to your kernel config:
wsbell* at spkr? console?

Ok pgoyette@.

(nat)

Seperate the definitions for the console beep.  This code will be shared
with a new device.

Ok pgoyette@.

(nat)

Add ioctls for setting and getting the beep volume.  Currently only
supported on spkr devices attached to audio.

Ok pgoyette@.

(nat)

Remove unused code - it will reappear in a new device.

Ok pgoyette@.

(nat)

Fix build w/o VERBOSE_INIT_ARM

(jmcneill)

Add delay and enable mct timecounter.

(jmcneill)

Correctly initialize i2cbus attach args.

(jmcneill)

Avoid divide-by-zero for unconfigured PLLs

(jmcneill)

Fix KERNEL_BASE_PHYS (how did this ever work?)

(jmcneill)

add the tests set.

(christos)

Calculate UART frequency based on bootloader config.

(jmcneill)

Fix exynos5 devmap, bootstrap, and implement early_putchar.

(jmcneill)

Use arm_fdt_cpu_hatch and add mmu entry for DTB

(jmcneill)

Get the EXYNOS kernel building again with recent FDT changes. Untested.

(jmcneill)

Do m_pullup before mtod

It may fix panicks of some tests on anita/sparc and anita/GuruPlug.

(ozaki-r)

use satafis routine to build the command block for Gen IIe to reduce code
duplication

(jdolecek)

Fix missing TC bus speed.

(flxd)

Update inode member i_flag --> i_state to keep up with kernel changes

(pgoyette)

tickets 19-27

(snj)

Pull up following revision(s) (requested by manu in ticket #27):
external/bsd/dhcp/dist/relay/dhcrelay.c: revision 1.7
Fix buggy dhcrelay(8) requirement to stay in foreground
This version of dhcrelay(8) needed to stay inforeground with -d flag in
order to service requests. Running inbackground turned it deaf to DHCP
requests.
This was caused by wrong kqueue(2) usage, where kevent(2) was used with
a file descriptor obtained by a kqueue(2) call done before fork(2).
kqueue(2) man page says "The queue is not inherited by a child created
with fork(2)". As a result, kevent(2) calls always got EBADF.
The fix is to reorder function calls in dhcrelay(8) main() function.
dhcp_context_create(), which causes kqueue(2) to be invoked, is
moved with its dependencies after fork(2). This matches the code layout
of dhclient(8) and dhcpd(8), which do not have the bug.
The fix was not submitted upstream since latest ISC DHCP code was
refactored and does not have the bug anymore.

(snj)

Pull up following revision(s) (requested by joerg in ticket #26):
libexec/ld.elf_so/rtld.c: revision 1.184
Call _rtld_debug_state before running the global initialisers, so give
gdb a chance to set break points etc.

(snj)

Pull up following revision(s) (requested by skrll in ticket #25):
sys/arch/mips/mips/locore_mips1.S: revision 1.92, 1.93
fix tlb_record_asids 2nd arg to match usage - it's a maximum asid value
and not a mask
--
Add a missing ".set at" to make previous build

(snj)

Pull up following revision(s) (requested by nat in ticket #24):
sys/dev/usb/uaudio.c: revision 1.154
Start as half as many transfers for recording.  This fixes recording and
play back when uaudio is used in full duplex mode.
The transfer scheduling probably is to blame.
Ok skrll@.

(snj)

Pull up following revision(s) (requested by skrll in ticket #23):
sys/arch/sgimips/conf/INSTALL64_IP2x: revision 1.3
Align MEMORY_ROOT_DISK_SIZE with INSTALL32_IP2x

(snj)

Pull up following revision(s) (requested by skrll in ticket #22):
sys/arch/mips/mips/mips_machdep.c: revision 1.278
sys/arch/mips/mips/pmap_machdep.c: revision 1.21
Always use XKPHYS for pool pages on _LP64; otherwise use KSEG0
--
Maintain the split of physical memory into the defined freelists, but
only force pool pages to VM_FREELIST_FIRST512M for non _LP64

(snj)

Pull up following revision(s) (requested by nonaka in ticket #21):
sys/dev/pci/if_iwm.c: revision 1.74
fix typo.
reported by ryo@.

(snj)

Pull up following revision(s) (requested by nat in ticket #20):
sys/dev/pad/pad.c: revisions 1.33-1.35
sys/dev/pad/padvar.h: revision 1.9
Express BYTESTOSLEEP as an 64 bit integer.
Use BYTESTOSLEEP in expresson of BYTES_PER_SEC.
--
sc_bytes_count needs to be set in pad_audio_open not pad_open.
--
Simplification of rate limiter.  It now works uni/multiprocessor.
--
pad blocksize 1024 -> 8192.  Helps when sleeping in rate limiter.

(snj)

Pull up following revision(s) (requested by jmcneill in ticket #19):
sys/arch/arm/fdt/files.fdt: revision 1.11
sys/arch/arm/fdt/plrtc_fdt.c: revision 1.1
sys/arch/evbarm/conf/VEXPRESS_A15: revision 1.16
sys/conf/files: revision 1.1175
sys/dev/ic/pl031.c: revision 1.1
sys/dev/ic/pl031var.h: revision 1.1
Add driver for ARM PrimeCell PL031 (RTC) Real Time Clock.
--
Add fdt glue for plrtc
--
Add plrtc driver

(snj)

Further reduce the loop counter so that hatching completes before the
boot processor times us out.

Add a nice big XXX comment for why the counter is so low.

XXX Will need to pullup to NetBSD-7 branch

(pgoyette)

Rename i_flag to i_state.

The similarity to i_flags has previously caused errors.

(maya)

Add two new sub-tests to the c_line_wrapping test case.

These should detect if the errors that caused MAKDEV to fail and
pkgsrc/pkgtools/cwrappers to fail to build (problem detected in
libnbcompat/configure) ever return.

Also fixed one of the other sub-tests so that it actually does what
it should - no idea how this one has been passing, it did not for me
when I was checking the new ones (but perhaps in the interim I have
fixed something else in sh, the problem was in the area I have been
playing, and I originally debugged the new tests using a newer version
of /bin/sh than has yet been committed.)

(kre)

fix the driver to work again - switch to using tag from xfer, adjust
to framework changes (queue dynamically allocated), and do not store
pointer to xfer inside internal structures

unhide the ATAPI code and fix to compile, keep it disabled however

no indended functional changes, logic kept as close as possible to what was
there before; tested with Adaptec 1430SA (88SX7042)

(jdolecek)

add support for i.MX6UltraLite, and
ConnectCore for iMX6UL Starter Board.

(ryo)

whitespace

(flxd)

Make VAXstation 4000 TURBOchannel adaptor work.
PIO and IRQ tested on VS4000/60 and /96 with TC-USB (slhci) and PMAD-AA (le).
DMA is untested.
vax needs space to be mapped out, so memory space is pre-mapped for TC option
drivers not bus_space'ified yet, also providing custom bus_space methods for
this kludge.

(flxd)

Apply selected patches from OpenWall:
    http://cvsweb.openwall.com/cgi/cvsweb.cgi/Owl/packages/vixie-cron/

1. Add PAM support.
2. Sanitize children process reaping
3. futimens when we have an fd
4. close_all for crontab(8)
5. use a table for spool dirs instead of duplicating code.
6. handle errors from process_exit()
7. Add ENABLE_FIX_DIRECTORIES ifdef and enable it by default for compat
8. Avoid using fd's < STDERR

Not applied:
1. no xfork (no setresuid)
2. did not do the lstat before open.
3. did not enable cron group

(christos)

add cron.

(christos)

Add cron.

(christos)

tickets 14-18

(snj)

Pull up following revision(s) (requested by nat in ticket #18):
sys/dev/audio.c: revision 1.358
Retrn error to userland from startp/startr depending on the error from
trigger/start output/input.
Found by jmcneill@.

(snj)

Pull up following revision(s) (requested by jmcneill in ticket #17):
sys/arch/arm/fdt/aaci_fdt.c: revision 1.1
sys/arch/arm/fdt/files.fdt: revision 1.10
sys/arch/evbarm/conf/VEXPRESS_A15: revisions 1.14, 1.15
sys/conf/files: revision 1.1174
sys/dev/ic/pl041.c: revisions 1.1-1.3
sys/dev/ic/pl041var.h: revision 1.1
Add driver for ARM PrimeCell Advanced Audio CODEC interface (PL041).
Don't expect this driver to work on real hardware, but QEMU emulates it.
--
Add fdt glue for ARM PrimeCell Advanced Audio CODEC interface (PL041).
--
Add aaci at fdt, commented out for now. Driver should work (tm) but QEMU
and my old Thinkpad can't seem to keep up.
--
Fix two bugs:
- Inverted test for fifo status in aaci_write_data
- Return success from trigger_output (thanks nat)
--
Enable aaci
--
bus_space_write_multi_4 takes a count, not number of bytes. With this,
audio works in qemu.

(snj)

Pull up following revision(s) (requested by he in ticket #16):
distrib/notes/macppc/prep.OPENFIRMWARE: revision 1.19
Change the deprecation notice about ofwboot.elf to say that it may
still be required in some cases, as observed by some MacMini users.

(snj)

Pull up following revision(s) (requested by kre in ticket #15):
bin/sh/sh.1: revision 1.148
Fix a typo (or rather a remnant of an earlier intent).

(snj)

Pull up following revision(s) (requested by jmcneill in ticket #14):
sys/dev/sdmmc/ld_sdmmc.c: revision 1.27
Fix a race between ld_sdmmc_start and ld_sdmmc_dobio that could result in
tasks getting lost from the task queue. The symptom of this is a NULL
deref in ld_sdmmc_start since the code assumes that a task will always be
available from the pool.
This changes the code to use pcq(9) instead of a TAILQ to manage the free
task list.

(snj)

Start as half as many transfers for recording.  This fixes recording and
play back when uaudio is used in full duplex mode.
The transfer scheduling probably is to blame.

Ok skrll@.

(nat)

Test bind()ing to a non-existent interface.

(bouyer)

Refuse to bind to a non-CAN interface.
Also release the lock in the error branch.

(bouyer)