XXX: don't KASSERT() if sc->lock is NULL, this is going to happen
XXX: during autoconfig match() where we've faked up a sc anyway.

(mrg)

add comment about r1.6 -> r1.7 change.

(tnozaki)

make this compile again.  i think it is even right.

(mrg)

Recent changes to the at91 code enable this kernel config to build after
the hypothetical ds1672rtc(4) is commented out and symtab space is bumped.
Fixes PR#43834.

(jakllsch)

compile again.

(nonaka)

enough with the compile time warnings.

(christos)

simplify code by removing a useless "#define STATIC static" in both debug
and normal usage, and all of its uses.

(agc)

give the correct count of arguments to the sub-functions

(agc)

An undocumented behavior of the sysctl kern.arandom node used to allow
sucking up to 8192 bytes out of the kernel arc4random() generator at a
time.  Supposedly some very old application code uses this to rekey
other instances of RC4 in userspace (a truly great idea).  Reduce the
limit to 256 bytes -- and note that it will probably be reduced to
sizeof(int) in the future, since this node is so documented.

(tls)

Add very, very temporary and ugly work around for my braino about detaching bad devices.  Will fix as soon as I have more time with the hardware where I can reproduce this.

(tls)

Fix a backwards memcmp() that was causing all hardware sources to detach.

(tls)

adapt to audiomp api changes

(jmcneill)

add gcscaudio_get_locks()

(jmcneill)

adapt to audiomp api changes

(jmcneill)

adapt to audiomp api changes

(jmcneill)

First step of random number subsystem rework described in
<20111022023242.BA26F14A158@mail.netbsd.org>.  This change includes
the following:

An initial cleanup and minor reorganization of the entropy pool
code in sys/dev/rnd.c and sys/dev/rndpool.c.  Several bugs are
fixed.  Some effort is made to accumulate entropy more quickly at
boot time.

A generic interface, "rndsink", is added, for stream generators to
request that they be re-keyed with good quality entropy from the pool
as soon as it is available.

The arc4random()/arc4randbytes() implementation in libkern is
adjusted to use the rndsink interface for rekeying, which helps
address the problem of low-quality keys at boot time.

An implementation of the FIPS 140-2 statistical tests for random
number generator quality is provided (libkern/rngtest.c).  This
is based on Greg Rose's implementation from Qualcomm.

A new random stream generator, nist_ctr_drbg, is provided.  It is
based on an implementation of the NIST SP800-90 CTR_DRBG by
Henric Jungheim.  This generator users AES in a modified counter
mode to generate a backtracking-resistant random stream.

An abstraction layer, "cprng", is provided for in-kernel consumers
of randomness.  The arc4random/arc4randbytes API is deprecated for
in-kernel use.  It is replaced by "cprng_strong".  The current
cprng_fast implementation wraps the existing arc4random
implementation.  The current cprng_strong implementation wraps the
new CTR_DRBG implementation.  Both interfaces are rekeyed from
the entropy pool automatically at intervals justifiable from best
current cryptographic practice.

In some quick tests, cprng_fast() is about the same speed as
the old arc4randbytes(), and cprng_strong() is about 20% faster
than rnd_extract_data().  Performance is expected to improve.

The AES code in src/crypto/rijndael is no longer an optional
kernel component, as it is required by cprng_strong, which is
not an optional kernel component.

The entropy pool output is subjected to the rngtest tests at
startup time; if it fails, the system will reboot.  There is
approximately a 3/10000 chance of a false positive from these
tests.  Entropy pool _input_ from hardware random numbers is
subjected to the rngtest tests at attach time, as well as the
FIPS continuous-output test, to detect bad or stuck hardware
RNGs; if any are detected, they are detached, but the system
continues to run.

A problem with rndctl(8) is fixed -- datastructures with
pointers in arrays are no longer passed to userspace (this
was not a security problem, but rather a major issue for
compat32).  A new kernel will require a new rndctl.

The sysctl kern.arandom() and kern.urandom() nodes are hooked
up to the new generators, but the /dev/*random pseudodevices
are not, yet.

Manual pages for the new kernel interfaces are forthcoming.

(tls)

Ticket #1695

(sborrill)

Pull up the following revisions(s) (requested by rmind in ticket #1695):
sys/kern/kern_event.c: revision 1.74

kqueue_register: avoid calling fd_getfile() with filedesc_t::fd_lock held.
Fixes PR/45479 by KOGULE Ryo.

(sborrill)

Ticket #1695

(sborrill)

Pull up the following revisions(s) (requested by rmind in ticket #1695):
sys/kern/kern_event.c: revision 1.74

kqueue_register: avoid calling fd_getfile() with filedesc_t::fd_lock held.
Fixes PR/45479 by KOGULE Ryo.

(sborrill)

Ticket #1695

(sborrill)

Pull up the following revisions(s) (requested by rmind in ticket #1695):
sys/kern/kern_event.c: revision 1.74

kqueue_register: avoid calling fd_getfile() with filedesc_t::fd_lock held.
Fixes PR/45479 by KOGULE Ryo.

(sborrill)

port ad-audiomp2 branch changes to 5.99.56
todo:
- port ad-audiomp2 pcppi changes
- these drivers need to be fixed:
  - emuxki
  - gcscaudio
  - hdafg
  - uaudio
  - umidi
  - MD audio drivers

(jmcneill)

grammar fix

(jnemeth)

fix memory leak, pointed by nonaka-san(again^3).

(tnozaki)

remove useless free(), pointed by nonaka-san(again^2).

(tnozaki)

return EINVAL when module validation failed, pointed by nonaka-san(again).

(tnozaki)

return EINVAL when module validation failed, pointed by nonaka-san(again).

(tnozaki)

remove unused variable, pointed by nonaka-san, thanks.

(tnozaki)

reliability fix, merge libc/regex fix to avoid memory exhaust problem.

(tnozaki)

One more.

(mbalmer)

Add a few.

(mbalmer)

Apply the better patch in the PR.

(christos)

Better wording.

(mbalmer)

disable kloader(4).

(nonaka)

enable kloader(4).

(nonaka)

[merging from cherry-xenmp] bring in bouyer@'s changes via:
http://mail-index.netbsd.org/source-changes/2011/10/22/msg028271.html
From the Log:
Log Message:
Various interrupt fixes, mainly:
keep a per-cpu mask of enabled events, and use it to get pending events.
A cpu-specific event (all of them at this time) should not be ever masked
by another CPU, because it may prevent the target CPU from seeing it
(the clock events all fires at once for example).

(cherry)

Add workaround for infrequently encountered DMA engine limitation.

(jakllsch)

grab major 203 for the iSCSI communications device (between kernel
driver and /sbin/iscsid)

(agc)

PR/45633: Christian Biere: Don't access byte after NUL when setting magic.

(christos)

Tickets 1436, 1438, 1439

(bouyer)

Tickets 1432, 1436, 1438, 1439

(bouyer)

Pull up following revision(s) (requested by drochner in ticket #1439):
dist/openpam/lib/openpam_configure.c: revision 1.6
Don't allow '/' characters in the "service" argument to pam_start()
The "service" is blindly appended to config directories ("/etc/pam.d/"),
and if a user can control the "service" it can get PAM to read config
files from any location.
This is not a problem with most software because the "service" is
usually a constant string. The check protects 3rd party software
from being abused.
(CVE-2011-4122)

(bouyer)

Pull up following revision(s) (requested by drochner in ticket #1439):
dist/openpam/lib/openpam_configure.c: revision 1.6
Don't allow '/' characters in the "service" argument to pam_start()
The "service" is blindly appended to config directories ("/etc/pam.d/"),
and if a user can control the "service" it can get PAM to read config
files from any location.
This is not a problem with most software because the "service" is
usually a constant string. The check protects 3rd party software
from being abused.
(CVE-2011-4122)

(bouyer)

Pull up following revision(s) (requested by christos in ticket #1438):
sys/kern/kern_event.c: revision 1.73
PR/45618: Motoyuki OHMORI: kqueue EVFILT_TIMER with smaller timeout value
makes DIAGNOSTIC kernel panic:
        KASSERT((c->c_flags & CALLOUT_PENDING) !=3D 0);
If the computed ticks are <= 0 set it to 1

(bouyer)

Pull up following revision(s) (requested by christos in ticket #1438):
sys/kern/kern_event.c: revision 1.73
PR/45618: Motoyuki OHMORI: kqueue EVFILT_TIMER with smaller timeout value
makes DIAGNOSTIC kernel panic:
        KASSERT((c->c_flags & CALLOUT_PENDING) !=3D 0);
If the computed ticks are <= 0 set it to 1

(bouyer)

Pull up following revision(s) (requested by christos in ticket #1436):
sys/uvm/uvm_meter.c: revision 1.57 via patch
if you are going to dereference a variable, check the variable itself, not
it cousin.

(bouyer)

Pull up following revision(s) (requested by christos in ticket #1436):
sys/uvm/uvm_meter.c: revision 1.57 via patch
if you are going to dereference a variable, check the variable itself, not
it cousin.

(bouyer)

Pull up following revision(s) (requested by dholland in ticket #1432):
lib/librmt/rmtlib.c: revision 1.23
lib/librmt/rmtlib.c: revision 1.24
Improve isrmt() check: it cannot be a rmt fd if there are no pipes
open for the fd.  Prevents collision with rumphijack.
Also, prevent potential hyperspace memory access.
Does someone want to write tests for this facility?
PR/38413: Takahiro Kambe: mt(1) print some junk output when using remote tape
Not all fields are valid in the ioctl to get tape info in the rmt protocol.
Zero out the struct so that we don't print junk.
While here, KNF.

(bouyer)

ticket 1696

(bouyer)

ticket 1696

(bouyer)

ticket 1696

(bouyer)

Pull up following revision(s) (requested by drochner in ticket #1696):
dist/openpam/lib/openpam_configure.c: revision 1.6
Don't allow '/' characters in the "service" argument to pam_start()
The "service" is blindly appended to config directories ("/etc/pam.d/"),
and if a user can control the "service" it can get PAM to read config
files from any location.
This is not a problem with most software because the "service" is
usually a constant string. The check protects 3rd party software
from being abused.
(CVE-2011-4122)

(bouyer)

Pull up following revision(s) (requested by drochner in ticket #1696):
dist/openpam/lib/openpam_configure.c: revision 1.6
Don't allow '/' characters in the "service" argument to pam_start()
The "service" is blindly appended to config directories ("/etc/pam.d/"),
and if a user can control the "service" it can get PAM to read config
files from any location.
This is not a problem with most software because the "service" is
usually a constant string. The check protects 3rd party software
from being abused.
(CVE-2011-4122)

(bouyer)

Pull up following revision(s) (requested by drochner in ticket #1696):
dist/openpam/lib/openpam_configure.c: revision 1.6
Don't allow '/' characters in the "service" argument to pam_start()
The "service" is blindly appended to config directories ("/etc/pam.d/"),
and if a user can control the "service" it can get PAM to read config
files from any location.
This is not a problem with most software because the "service" is
usually a constant string. The check protects 3rd party software
from being abused.
(CVE-2011-4122)

(bouyer)

Fix passing of floating point registers.

(mlelstv)

The compiler is allowed to use intermediate higher precision for float
arithmetic, which may cause differences smaller than float precision
but still much larger than eps = 1e-30.
Forcing intermediate results to volatile variables removes the excess
precision.

(mlelstv)

fix build when ALTQ is defined

(jmcneill)

Note pow(4)/rtcalarm(8).

(isaki)

Obsolete pow(4) and rtcalarm(8).  Discussed in port-x68k@.
Probably the combination of pow(4) ioctl and rtcalarm(8) does
not work for a long time, and nobody uses them.
I'll rewrite a part about power switch handler as a new device.

(isaki)

bye xbox

(christos)

Ticket #1694

(sborrill)

Pull up the following revisions(s) (requested by bouyer in ticket #1694):
sys/dev/pci/if_age.c: revision 1.40 via patch

- age_init() is called from age_watchdog() which is in interrupt context; we
can't sleep here or we get a DIAGNOSTIC panic when age_watchdog() fires.
- More correct bus_dma(9) usage in age_encap()
- Introduce a age_shutdown() to be called by pmf(9) at shutdown time,
to stop the DMA engine.
- Be consistent in WAIT/NOWAIT use in init routines
- Use BUS_DMA_COHERENT where appropriate
- Rework the interrupt routine a bit, and ACK but do not disable interrupts
here. There seems to be a race where interrupts would not be properly
reenabled after this, leading do watchdog timeouts.

(sborrill)

Ticket #1693

(sborrill)

Pull up the following revisions(s) (requested by christos in ticket #1693):
sys/kern/kern_event.c: revision 1.73

PR/45618: Motoyuki OHMORI: kqueue EVFILT_TIMER with smaller timeout value
makes DIAGNOSTIC kernel panic. If the computed ticks are <= 0 set it to 1.

(sborrill)

Ticket #1692

(sborrill)

Ticket #1692

(sborrill)

Ticket #1692

(sborrill)

Pull up the following revisions(s) (requested by christos in ticket #1692):
dist/bind/bin/named/query.c            patch
dist/bind/lib/dns/rbtdb.c              patch

Cache lookup could return RRSIG data associated with nonexistent records,
leading to an assertion failure.  Fixes CVE-2011-4313.

(sborrill)

Pull up the following revisions(s) (requested by christos in ticket #1692):
dist/bind/bin/named/query.c            patch
dist/bind/lib/dns/rbtdb.c              patch

Cache lookup could return RRSIG data associated with nonexistent records,
leading to an assertion failure.  Fixes CVE-2011-4313.

(sborrill)

Pull up the following revisions(s) (requested by christos in ticket #1692):
dist/bind/bin/named/query.c patch
dist/bind/lib/dns/rbtdb.c patch

Cache lookup could return RRSIG data associated with nonexistent records,
leading to an assertion failure.  Fixes CVE-2011-4313.

(sborrill)

Ticket #1691

(sborrill)

Ticlet #1691

(sborrill)

Pull up the following revisions(s) (requested by christos in ticket #1691):
sys/uvm/uvm_meter.c: revision 1.57

Dereference correct variable and thus stop a sysctl crash.

(sborrill)

Pull up the following revisions(s) (requested by christos in ticket #1691):
sys/uvm/uvm_meter.c: revision 1.57

Dereference correct variable and thus stop a sysctl crash.

(sborrill)

Ticket #1691

(sborrill)

Pull up the following revisions(s) (requested by christos in ticket #1691):
sys/uvm/uvm_meter.c: revision 1.57

Dereference correct variable and thus stop a sysctl crash.

(sborrill)

Ticket #1690

(sborrill)

Pull up the following revisions(s) (requested by riz in ticket #1690):
distrib/sets/lists/base/mi patch
distrib/sets/lists/man/mi patch
sbin/Makefile patch
sbin/resize_ffs/Makefile patch
sbin/resize_ffs/TODO patch
sbin/resize_ffs/resize_ffs.8 patch
sbin/resize_ffs/resize_ffs.c patch

Fix resize_ffs(8) so that growing and shrinking all FFSv1
(including byteswapped) and growing FFSv2 file systems is supported.

(sborrill)

remove Xbox support

(jmcneill)

Ticket #1689

(sborrill)

Pull up the following revisions(s) (requested by is in ticket #1689):
gnu/dist/gcc4/gcc/version.c: revision 1.19
gnu/dist/gcc4/gcc/config/arm/arm.md: revision 1.1.1.2-1.2

Remove pattern for arith_adjacentmem in arm.md, thus fixing
PR toolchain/45576 at a very small run-time cost.

(sborrill)

Switch from COPYDIR to COPY/LINK when populating usr/mdec directory, to
avoid file duplication.

(njoly)

Ticket #1688

(sborrill)

Pull up the following revisions(s) (requested by tron in ticket #1688):
sys/kern/kern_module.c: revision 1.84

Return ENOEXEC instead of 0 if the info of a kernel module doesn't have the
expected size. This prevents kernel panics when loading broken modules.

(sborrill)

Obey MNT_RELATIME, the only addition is that mkdir in ufs sets IN_ACCESS too.

(christos)

- collect the long (and sometimes incomplete) lists of basic flags into
  the header file and use that.
- sort the list of basic flags
- add MNT_RELATIME, ST_RELATIME
- mask all the op flags, for symmetry.

The real bit difference is (which is harmless):
    - mount was missing MNT_EXTATTR
    - update sets MNT_RDONLY twice
    - ops also could or in MNT_GETARGS, but this is impossible because the
      code would have chosen to do getargs then.

(christos)

Ticket #1687

(sborrill)

Pull up the following revisions(s) (requested by dholland in ticket #1687):
sys/arch/i386/i386/vector.S: revision 1.54

Keep interrupts disabled in NMI handler. Addresses PR/43007.

(sborrill)

document MNT_RELATIME

(christos)

Exploit hidden __cerror

(joerg)

Add coverity annotations about unreachable code (Kamil Dudka)

(christos)

Add missing *

(christos)

Initialize termbuf (Kamil Dudka)

(christos)

Initialize res (Kamil Dudka)

(christos)

initialize t to protect against empty string (Kamil Dudka)

(christos)

check for negative return of ct_visual_char (Kamil Dudka)

(christos)

Off by one in allocation could cause buffer overflow (Kamil Dudka)

(christos)

remove unrecheable code (Kamil Dudka)

(christos)

Mark __cerror as hidden and exploit that to simplify assembler.

(joerg)

Explicitly mark __cerror as hidden

(joerg)

remove one more kassert

(christos)

Hook up TLS support on ARM. Tested by martin@, kernel support from matt@

(joerg)

Update LLVM/Clang snapshot to r144933. Besides the usual bugfixes and
improvements, this brings the first of the address sanity checker and
the new tainted variable logic for the analyzer.

(joerg)

remove incorrect assertions (the len passed is the size of the bsd structure
not the linux one)

(christos)

remove incorrect assertions (the len passed is the size of the bsd structure
not the linux one)

(christos)

note bozohttpd 20111118

(mrg)

merge bozohttpd 20111118

(mrg)

update to bozohttpd 20111118.  nothing major is missing here but
the changes since the prior import were:

        o  add -P <pidfile> option, from jmmv@netbsd.org
        o  avoid crashes with http basic auth, from pooka@netbsd.org
        o  add support for REDIRECT_STATUS variable, from tls@netbsd.org
        o  support .mp4 files in the default map
        o  directory indexes with files with : are now displayed properly, from
          reed@netbsd.org
        o  allow -I option to be useful in non-inetd mode as well

Status:

Vendor Tag: bozohttpd
Release Tags: bozohttpd-20111118

(mrg)

Provide linux_siginfo_t for all archs, to unbreak kdump build.

(martin)

Release dtoa lock before returning, pointed out by enami.

(martin)

[merging from cherry-xenmp]
- Make clock MP aware.
- Bring in fixes that bouyer@ brought in via:
  cvs rdiff -u -r1.54.6.4 -r1.54.6.5 src/sys/arch/xen/xen/clock.c

Thanks to riz@ for testing on dom0

(cherry)

regen

(christos)

unlock before returning on error, thanks enami.

(christos)

remove clauses 3 and 4

(christos)

include the new siginfo.h file

(christos)

add sigtimedwait support

(christos)

- add sigtimedwait support.
- merge the siginfo population code.

(christos)

add sigtimedwait

(christos)

add a copyin function for sigset so that we can use sigtimedwait1 from
emulations.

(christos)

PR/45627: Martin Husemann: Plug memory leak

(christos)

Add some dependencies

(joerg)

- use mutex obj for pageable object
- add a function to wait for a mutex obj being available
- replace some "livelock" kpauses with it

(yamt)

share a lock among pmap uobjs

(yamt)

Fix a lint warning noted by njoly.

(wiz)

Bump date for previous.

(wiz)

kqueue_register: avoid calling fd_getfile() with filedesc_t::fd_lock held.

Fixes PR/45479 by KOGULE Ryo.

(rmind)

allow the -I option to be useful in non-daemon mode, by letting it force
the returned port number

(mrg)

Also match ICH8, ICH9 and ICH10 devices.  Tested on ICH10.

(riz)

Reorder to make GCC happy.

(joerg)

Use __dead

(joerg)

Use __dead

(joerg)

FreeBSD bug report 161344: TLS area for the main thread is set up to
early, if e.g. pointers to functions are used as initializers.

(joerg)

tidy up.

(christos)

Don't return ENOSYS if nothing needs to be done for the CPU.

XXX Shouldn't this be a runtime test on the CPU?

(joerg)

fixed some crashes in LIST_FOREACH where current element could be removed during the loop

(vanhu)

fixed some crashes in LIST_FOREACH where current element could be removed during the loop

(vanhu)

Fix last envsys2 conversion (1.10).  From Tetsuya Isaki in PR#45622.

(jakllsch)

Make the test for MKISCSI actually work by including bsd.own.mk and
testing the expanded variable, not just the variable name(!)  Fixes the
build for sun2.

(he)

identifycpu() is not just cosmetic for the banner but initializes how FPU
contexts are saved.
Also drop code that checks for fputype before it is determined.

(mlelstv)

Replace section number with actual one as we did in see also list.

(enami)

Copy node expiration date before comparing it, otherwise the comparison
does not work (no idea why) and cached node is never used.

(manu)

add kevent test

(christos)

PR/45618: Motoyuki OHMORI: kqueue EVFILT_TIMER with smaller timeout value
makes DIAGNOSTIC kernel panic:
KASSERT((c->c_flags & CALLOUT_PENDING) != 0);
If the computed ticks are <= 0 set it to 1

(christos)

Add a test for PR/45618: Motoyuki OHMORI: kqueue EVFILT_TIMER with smaller
timeout value makes kernel busy or panic

(christos)

Remove trailing whitespace. Merge two EINVAL entries.

(wiz)

+ 3218. [security]      Cache lookup could return RRSIG data associated with
+                      nonexistent records, leading to an assertion
+                      failure. [RT #26590]

so that someone looking can tell more easily we've applied it.

(mrg)

this works with much less memory too.

(christos)

add require.memory=500M

(christos)

PR/45619: jmmv: Allow atf tests to request a minimum amount of memory

(christos)

patch for: https://www.isc.org/software/bind/advisories/cve-2011-tbd

(christos)

Add the iscsi kernel module and directories for ibm4xx and booke variants.
Re-sort.

(he)

Semicolons are not optional :)

(he)

use ARG_CHAR_T instead of CHAR_T for integer promotion.

(tnozaki)