Transfer some static from my winter coat to more structures/functions
in this file.

(jakllsch)

Fix up changes commited in kttcp.c 1.29.

(jakllsch)

don't include pcap/bpf.h

(christos)

don't change the layout of the struct, it is used by fstat. (from yamt)

(christos)

PR/45730: David Holland: Avoid having 2 copies of bpf.h in /usr/include.
This adds the missing entries from libpcap to make libpcap compile with
our bpf.h.

(christos)

Don't use libpcap's bpf file, but use our own:
1. include <net/dlt.h> where needed since we have the information in a separate
  file.
2. #error in pcap/bpf.h to make sure it is not being used.
3. If we don't include <pcap/bpf.h> include <net/bpf.h> instead.
4. Install symlinks instead of copies of files.

(christos)

only set CANTRCVMORE if no error.

(christos)

simplify expression

(christos)

bootmenu doesn't parse "prompt" as part of a command list, it must
stand alone.

(mlelstv)

move the (now 1024 byte) printing buffer off the stack

(jmcneill)

Increase printing buffer of ttycons from 80 to 1024 significantly increasing
console output on large dumps.

(reinoud)

Include the fifofs Atffile

(pgoyette)

relax the test for analog devices analog pcbeep widgets

(jmcneill)

add fifofs

(christos)

add the directories

(christos)

correct install dir

(christos)

Add a fifofs test.

(christos)

Add a test for the latest fifofs fix.

(christos)

- Eliminate so_nbio and turn it into a bit SS_NBIO in so_state.
- Introduce MSG_NBIO so that we can turn non blocking i/o on a per call basis
- Use MSG_NBIO to fix the XXX: multi-threaded issues on the fifo sockets.
- Don't set SO_CANTRCVMORE, if we were interrupted (perhaps do it for all
  errors?).

(christos)

on second thought, set machine and machine_arch both from the host and
override module_machine

(jmcneill)

check return value of write, make sure we send the whole buffer to stdout

(jmcneill)

Mark all of the Rx descriptor fields 'volatile' so that the compiler
will not re-order accesses.  Some versions of GCC (such as one in NetBSD
5.x) definitely do re-order reads from these fields if they're not
marked volatile.

(dyoung)

set machine_arch to that of the host

(jmcneill)

thunk_pollchar: use read instead of getchar

(jmcneill)

get rid of urkelvisor, and use same MD majors as the host

(jmcneill)

If we need to set the PK_CHKNOSYSCALL flag in struct proc be so nice to first
take the mutex. Tnx for pointing it out to me.

(reinoud)

Put the path to the compat/common directory in a .PATH line, not in
an element of the SRCS list.  This should fix a problem in which build
products were created in the source tree.

Also add a comment about where COMPAT_50 is defined.

(apb)

Move the diagnostic check for a missing VOP_CLOSE() to the top of vrelel().
As long as we hold the vnode interlock there is no chance for this vnode
to gain new references.

Fixes false alarms observed by Thor Lancelot Simon and reported on tech-kern.

Ok: David Holland <dholland@netbsd.org>

(hannken)

New sentence, new line. Bump date for previous.

(wiz)

allow kernels w/o COMPAT_50 to build

(drochner)

Use to the MAP_NOSYSCALLS argument to mmap() to allow for NetBSD/usermode to
execute bog-standard native programs.

(reinoud)

Add int $80 and sysenter opcodes to the x86 SIGILL opcode detector

(reinoud)

Part 2 - x86 implementation of MAP_NOSYSCALLS

Currently the MAP_NOSYSCALLS is only implemented for x86 but other
architectures are easy to adapt; see the sys/arch/x86/x86/syscall.c patch.
Port maintainers are encouraged to add them for their processor ports too.
When this feature is not yet implemented for an architecture the
MAP_NOSYSCALLS is simply ignored with virtually no cpu cost..

(reinoud)

Ooops forgot the uvm_map.h

(reinoud)

Add a MAP_NOSYSCALLS flag to mmap. This flag prohibits executing of system
calls from the mapped region. This can be used for emulation perposed or for
extra security in the case of generated code.

Its implemented by adding mapping-attributes to each uvm_map_entry. These can
then be queried when needed.

Currently the MAP_NOSYSCALLS is only implemented for x86 but other
architectures are easy to adapt; see the sys/arch/x86/x86/syscall.c patch.
Port maintainers are encouraged to add them for their processor ports too.
When this feature is not yet implemented for an architecture the
MAP_NOSYSCALLS is simply ignored with virtually no cpu cost..

(reinoud)

comment and assertion

(yamt)

don't inline uvn_findpages in genfs_io.

(yamt)

Revert previous; the #include was already present, and I got confused
by a merge error.

(apb)

from http://www.freshbsd.org/commit/openbsd/ae7f934ae5bdf57dcf3431ba55fd1da93b8f1963

Initialize abridged tag word properly. x87 spec says FNINIT says tag word
contains FFFFh (all stack locations empty) which would make abridged tag
word 00h. From the Intel 64 and IA-32 Architectures Software Developer's
Manual:

"The FXSAVE instruction saves an abridged version of the x87 FPU tag word
in the FTW field (unlike the FSAVE instruction, which saves the complete
tag word). The tag information is saved in physical register order (R0
through R7), rather than in top-of- stack (TOS) order. With the FXSAVE
instruction, however, only a single bit (1 for valid or 0 for empty) is
saved for each tag."

ok rmind@

(jmcneill)

#include "opt_compat_netbsd.h"

(apb)

note new OpenPAM release

(drochner)

- make all the ifdefs match
- make it compile, and test

(christos)

Put back setitimer based code for the have-nots: (OS/X Lion)

(christos)

SRCS += ${.CURDIR}/../../../../compat/common/rndpseudo_50.c
to fix build errors like this:

DESTDIR/usr/lib/librumpdev_rnd.so: undefined reference to
`rumpns_compat_50_rnd_ioctl'

(apb)

Add COMPAT_50 and COMPAT_NETBSD32 compatibility code for rnd(4)
ioctl commands.

Tested with "rndctl -ls" using an old 32-bit version of rndctl(8)
(built for NetBSD-5.99.56/i386) and a new 64-bit kernel
(NetBSD-5.99.59/amd64).

(apb)

Return ENOTTY, not EINVAL, when the ioctl command is unrecognised.

(apb)

It's tp->t_dev, not tp->tp_dev.  Corrects commit prior to previous.

(jakllsch)

const-ify struct ucom_methods pointer within ucom(4).

(jakllsch)

Make commented-out code in ucomstop() compilable.

(jakllsch)

Add knob to turn MULTIPROCESSOR build on for XEN

(cherry)

c99: do not warn about trailing comma in enum declaration.

(njoly)

do missing ipsec->kame_ipsec renames

(drochner)

Fix error reporting in puffs_framev_enqueue_waitevent and callers.

ok christos

(riastradh)

Add eeprom.debug, and while here sort list.

(njoly)

Update LLVM/Clang snapshot to r146851. Most noticable, this fixes the
vfork issues with RUMP depending on inlining certain functions and a bad
stdio simplification.

(joerg)

Move Clang-warning flags for ah_regdomain.c into sys/conf, they are
platform independent. Use the selective -Wno-* flags instead of
-Wno-error.

(joerg)

libgcc_pic is still needed for GCC 4.1.

(joerg)

for analog devices codecs, mark the analog beeper pin as a beep generator
to help the parser. makes pcppi beep work again.

(jmcneill)

rename the IPSEC in-kernel CPP variable and config(8) option to
KAME_IPSEC, and make IPSEC define it so that existing kernel
config files work as before
Now the default can be easily be changed to FAST_IPSEC just by
setting the IPSEC alias to FAST_IPSEC.

(drochner)

as in netkey/key.c, just use cprng_fast() to get a random number
(which is used to choose an SPI), kill the dummy seeding code

(drochner)

make this build with RND_DEBUG

(drochner)

Whitespace.

(skrll)

fix killing code I just broke, and give program a chance to cleanup.

(christos)

don't produce different output if we are super user.

(christos)

Remove unused variable

(joerg)

Don't use antiquated BSD API's that require global variable, use posix timers
instead.

(christos)

Fix monotonic interval timers.

(christos)

test timer expiration.

(christos)

Add two late entries:
- x68k bootloader fix for X68060
- m68k FPE hyperbotic and trigonometric functions

(tsutsui)

Rename FRAME* macro to TRFS_FRAME* for readability and
workaround for powerpc builds. Pointed out by phx@.

XXX: <powerpc/frame.h> shouldn't pollute userland namespace anyway.

(tsutsui)

Replaced conf_args by mainbus_attach_args.

(phx)

Lowercase arch strings, and add missing ones.

(njoly)

SYMTAB_SPACE.

(kiyohara)

Restore __HAVE_MM_MD_READWRITE which was added for 5.99.53 in rev 1.14
but removed by merge botch on __HAVE_NEW_STYLE_BUS_H changes in rev 1.15.

Fixes "/dev/nvram: Device not configured" problem on installboot(8)
reported by David Ross.

(tsutsui)

WABPL is no longer considered experimental (has not been for some time)
so update its comment in config files.

(dholland)

Explicitly specify -m68020-60 to avoid possible 060SP instructions,
especially 64 bit integer mul/div ops in standalone bootloaders. From x68k.

(tsutsui)

Misc KNF and cosmetics.  No binary changes.

(tsutsui)

PR port-next68k/45719 from Jaime Fournier:
Fix RELOC() (awful VA -> PA conversion) usage introduced to appease gcc45
in rev 1.22.

(tsutsui)

adjust English usage in comment

(dholland)

SYMTAB_SPACE.

(kiyohara)

New sentence, new line. Fix some markup. Bump date for previous.

(wiz)

Use more markup. Bump date for previous.

(wiz)

New sentence, new line.
Bump date for previous.

(wiz)

Added NOR flash driver, using nor(4) and flash(4).

(phx)

Separate /dev/random pseudodevice implemenation from kernel entropy pool
implementation.  Rewrite pseudodevice code to use cprng_strong(9).

The new pseudodevice is cloning, so each caller gets bits from a stream
generated with its own key.  Users of /dev/urandom get their generators
keyed on a "best effort" basis -- the kernel will rekey generators
whenever the entropy pool hits the high water mark -- while users of
/dev/random get their generators rekeyed every time key-length bits
are output.

The underlying cprng_strong API can use AES-256 or AES-128, but we use
AES-128 because of concerns about related-key attacks on AES-256.  This
improves performance (and reduces entropy pool depletion) significantly
for users of /dev/urandom but does cause users of /dev/random to rekey
twice as often.

Also fixes various bugs (including some missing locking and a reseed-counter
overflow in the CTR_DRBG code) found while testing this.

For long reads, this generator is approximately 20 times as fast as the
old generator (dd with bs=64K yields 53MB/sec on 2Ghz Core2 instead of
2.5MB/sec) and also uses a separate mutex per instance so concurrency
is greatly improved.  For reads of typical key sizes for modern
cryptosystems (16-32 bytes) performance is about the same as the old
code: a little better for 32 bytes, a little worse for 16 bytes.

(tls)

- In cfi_probe() iterate over all chip widths for every port width to find
  x8/x16 configurations, or two x8 chips forming a x16 port width.
- The offset in cfi_cmd() is now always given for 64-bit port width.
  It will be scaled down for the actual port width to avoid problems
  with 0x2aa -> 0x554 instead of 0x555.
- Added missing cfi_reset_default() in cfi_jedec_id(). The reset is
  needed, because the chip was still in CFI-Query mode.
- Removed everything dealing with cfi_opmodes. It only complicates things
  and I found no reason for keeping them. The port width and chip width
  is known, so I just introduced a new cfi_read_qry() function which reads
  a single word using bus_space_read_N() from the QRY structure, which
  can be checked for 'Q', 'R' and 'Y'. That's all we need, before
  reading and unpacking the whole QRY and PRI structures.
- Added two new fields, cfi_unlock_addr1 and cfi_unlock_addr2 to
  the cfi struct. The unlock offsets should be kept variable, depending
  on the manufacturer or command set version.

Changes have been discussed with cliff@.

(phx)

Allow defining a bus space with pbs_limit=0, which will have the effect to
create a space which ends with 0xffffffff.
This is needed, because pbs_limit=0xffffffff ends the space with 0xfffffffe.

(phx)

Use mktime(3) instead of hand-coded equivalent.
Don't treat negative results as errors (we should be able
to handle years before 1970).

XXX: This still fails on 1969-12-31 23:59:59.

(apb)

Add some tests for dates several centures into the past and future.
These currently fail, but I'll fix that soon.

(apb)

Add some tests for timegm(3) close to the epoch.  In
particular, 1969-12-31 23:59:59 should convert to (time_t)-1
with errno = 0.

(apb)

Update ftp.NetBSD.org IPv4/IPv6 addresses.
(should be pulled up to all release branches?)

(tsutsui)

Fix panic early call rasops_init().  Move into arcvideo_attach().
Also early attach arcvideo.

(kiyohara)

Fix Bl/El unmatch.  Pointed out by kano@.

(tsutsui)

Remove trailing whitespace.

(wiz)

Add an rc.d(8) script for isibootd(8).  Taken from ndbootd(8).

(tsutsui)

Note isibootd(8) netboot server for luna68k.

(tsutsui)

Build and install isibootd(8).

(tsutsui)

Add isibootd(8) command, which is a dumb network boot server program
for the OMRON LUNA clients, like ndbootd(8) for Sun2 machines.

No particular comment on tech-userlevel@:
http://mail-index.NetBSD.org/tech-userlevel/2011/12/15/msg005872.html

(tsutsui)

Make "rndctl -L" overwrite the file with zeros, not with the
same data that was just read from the file.

(apb)

Add comments for some ioctls and structs; fix a typo in another comment.

(apb)

Commit regenerated configuration file after COPTS=-Os shuffle.

(he)

Drop makeoptions COPTS=-Os for the SMALL030 kernel now that std.atari
contains it.

(he)

Add option to include daily insecurity output in daily output.
PR bin/17029

(darcy)

The :@ example gives the impression that we encourage the ODE
convention of naming loop variables like .LINK., when in fact
single char variables are far more common in NetBSD.

(sjg)

Request always_inline for rumpclient__dofork, it won't work correctly
for vfork otherwise. Also give it the returns twice attribute to ensure
that the stack tainting is done recursively.

(joerg)

Fix a bug spotted by a user (resident flag becoming stale).
Add a few more KSEGX tests.
Only set pmap_page_colormask if virtual caches aliases are possible.

(matt)

provide a root entry if one was not found.

(christos)

support ECDSA keys used by recent ssh

(drochner)

disallow empty passphrases per default, and implement the "nullok"
option to allow it if the administator wishes, from FreeBSD

(drochner)

-remove remainders of the misguided changes in revs 1.5-1.9
-iron out more unnecessary differences to FreeBSD

(drochner)

Increase SYMTAB_SPACE.

(kiyohara)

Do not tell about specific errno, as all xattr copying function
retuenrs error from extattr_list_file and extattr_get_file.

(manu)

more clean D-cache.

(nonaka)

Whitespace.

(skrll)

Sort sections. Remove trailing whitespace.

(wiz)

Split extattr(3) into two man pages to get rid of ugly duplicate Nd.
Work done by Abhinav Upadhyay with minimal cleanup.

(wiz)

G/C pvh_aliases

(skrll)

Apparently mknative didn't work right. Fix up losses.

(mlelstv)

Document a bug when processes hang on uvn_fp2. FWIW, hung process exhibit
this stack backtrace:
sleepq_block
mtsleep
uvn_findpage
uvn_findpages
genfs_getpages
puffs_vnop_getpages
puffs_vnop_checkop
VOP_GETPAGES
uvn_get
uvm_fault_internal
trap
copyin
uiomove
ubc_uiomove
puffs_vnop_write
puffs_vnop_checkop
VOP_WRITE
vn_write
dofilewrite
sys_write
syscall

(manu)

Rework puffs_framebuf management toremove leaks and abusive reuses. On
exchange error, the puffs_framebuf is now freed immediatly, before
requeuing outstanding requests.

(manu)

make comment reflect reality

(christos)

Take softnet_lock and kernel lock in frag6_slowtimo and frag6_fasttimo,
similar to how it's done with other protocols.

If we don't do this sending ICMPv6 messages in this path can cause races
in network interface drivers.

(jakllsch)

don't leak mbufs.

(christos)

PR/45700: use dostatvfs instead of grabbing the latest cached copy of
struct statvfs from the mount point, so that chroot is handled properly.

(christos)

Remove nonexistent (probably HP-UX specific) Xref rmp(4).

(tsutsui)

increase SYMTAB_SPACE.

(oki)

SYMTAB_SPACE.

(kiyohara)

Amiga wscons/Xorg support, first phase.
Implemented wscons for CV64 and CV64/3D. Other graphics cards drivers are
prepared for it, but will not be attempted before Xorg is not running.
The wscons support is disabled by default. A GENERIC kernel should behave
like always. Use WSCONS to try out a kernel with wscons support.
Done by rkujawa@ and phx@.

(phx)

Specify COPTS="-Os" for rich gcc 4.5 as bsd.sys.mk does for userland binaries
so that gzipped atari kernels still fit into floppies.
Noticed by David Ross.

(tsutsui)

USERCONF works but it isn't very useful, so remove it

(jmcneill)

add required trace_enter/trace_exit calls in syscall() to make ktrace work

(jmcneill)

Like intrleave_to_user:, intrleave_to_kernel: should also restore PID.
Moreover, disable transration.

(kiyohara)

Store r30 and r31 beforehand to restore.

(kiyohara)

Increase MTRR_I686_NVAR_MAX from 8 to 16. Avoids
"FIXME: more than 8 MTRRs (10)" message on booting Thinkpad W520 and
similar. While here replace a magic number with MTRR_I686_NVAR_MAX * 2

(abs)

mknative-gcc for sun2 and tls / new version fix.

(mrg)

make sure m68010-netbsdelf chooses the right unwinder.

(mrg)

make sure m68010-netbsdelf chooses the right unwinder.

(mrg)

add eeprom

(macallan)

build eeprom on shark
this has been working for ages, why did I never commit it?

(macallan)

regenerate mknative-gcc files for most platforms, post skrll@ fix
and joerg@'s enable tls change.

missing: m68010 (coming shortly) and m68k (breaks the build)

XXX: didn't test i re-merged the powerpc by-hand changes properly

(mrg)

Add earlycons support.

(matt)

actually put some relevant information in the man page

(macallan)

Improve usermode timecounter. It's unreasonable to assume that we'll get
100 "SIGALRM" per second with an ITIMER_REAL at 100Hz on a HZ=100 host as
the timer may expire before a pending signal has been delivered.

Instead of setitimer, use timer_create + timer_settime and from our
intr handler use timer_getoverrun to determine how many ticks we have
missed.

(jmcneill)

allocate msgbuf with kmem_zalloc instead of thunk_malloc, now dmesg works

(jmcneill)

implement /proc/cpuinfo

(jmcneill)

add ptyfs/pty

(jmcneill)

use write instead of putchar putchar putchar putchar putchar ... for console output

(jmcneill)

implement cpu idle via sigsuspend

(jmcneill)

In lwp_create, copy l_sigstk from the template. Previously sigaltstack wasn't
inherited across fork (but SA_ONSTACK handlers was), causing the child
process to crash when the handler is invoked. ok martin@, christos@

(jmcneill)

add NoAccel option, general cleanup

(macallan)

set ld maxxfer to MAXPHYS

(jmcneill)

Cleanup i386 register printing and opcode extraction

(reinoud)

Remove the need for a siginfo structure in illegal instruction handling.

(reinoud)

OK, this is a temporary patch so usermode can run on 5.1 or old 5.99's. The
workaround mentioned here is not needed on newer 5.99 since it has just been
patched.

(reinoud)

Add missing x11 tag for xfree debug files, and add xorg debug files.

(njoly)

Fix title.

(njoly)

Indent.
Remove white-spaces and unnecessary semi-collons.

(kiyohara)

only need to setup altstack after fork, not in lwp tramp

(jmcneill)

include inttypes.h to get the fixed size integer types - from Gregor Riepl

(agc)

need to sigaltstack for each lwp

(jmcneill)