Nelson Mandela died today, aged 95. I whish farewell to a great statesman.

(mbalmer)

Add cdbr_open_mem.

(joerg)

Pull up following revision(s) (requested by pettai in ticket #1889):
distrib/sets/lists/base/mi: revision 1.1049
distrib/sets/lists/man/mi: revision 1.1449
usr.bin/gzip/Makefile: revision 1.17
usr.bin/gzip/Makefile: revision 1.18
usr.bin/gzip/zmore: revision 1.4
usr.bin/gzip/zmore.1: revision 1.4
Added zless(1) - comes in the lastest version of zmore(1) from OpenBSD
(OKed by tron@)
Add zfgrep that fell off from last update

(bouyer)

Pull up following revision(s) (requested by pettai in ticket #982):
distrib/sets/lists/base/mi: revision 1.1049
distrib/sets/lists/man/mi: revision 1.1449
usr.bin/gzip/Makefile: revision 1.17
usr.bin/gzip/Makefile: revision 1.18
usr.bin/gzip/zmore: revision 1.4
usr.bin/gzip/zmore.1: revision 1.4
Added zless(1) - comes in the lastest version of zmore(1) from OpenBSD
(OKed by tron@)
Add zfgrep that fell off from last update

(bouyer)

Use -marm for clang builds, too.

(joerg)

It is silly to kill the system when an interface failed to clear promiscuous
mode. Some return EINVAL when they are dying, but others like USB return EIO.
Downgrade to a DIAGNOSTIC printf. Same should be done for the malloc/NOWAIT,
but this is rarely hit.

(christos)

Use armv6z so {ldr,str}ex{b,h,d} won't generate warnings.

(matt)

Add a "varshm" check/fix to make sure /var/shm is mentioned in /etc/fstab
(and add a default of tmpfs with 25% of available ram limit if not).
To avoid the warning but not mount the tmpfs, just comment out the line
for /var/shm.

Also move "obsolete" to the end of the list again, as it should be run
last.

(martin)

binutils-2.24 out.

(wiz)

Use MBR_PTYPE_ defines from <sys/bootblock.h>.

(jakllsch)

Mirror my changes of src/sbin/gpt/create.c 1.4 in migrate.c;

Use less bogus CHS addresses in PMBR.
With the ending head set at 0xff one machine I have will never leave
the initial startup screen if such a disk is present. Additionally,
Wikipedia suggests without citiation that 254 is the maximium allowable
value for the head, and this seems to be the case.

(jakllsch)

Clarify comment about SER_SZERR.

(jdc)

Bump date for previous.

(wiz)

Fix err/errx confusion, pointed out by rmind.

(martin)

CID 271284: Missing error check on open

(christos)

add a note about GMP i wrote but forgot to commit a few days ago.

(mrg)

Provide variants of the -s option to allow limiting the tmpfs dynamically
at mount time to 1/Nth or to N percent of the available ram.

(martin)

add a couple of more things to worry about.

(mrg)

Fix RTC going backwards a year when saving time, because dt_mon != dt_year.

Patch by Roc Valles.

(rkujawa)

fixed lua(4) alloc (nsize can be lesser than osize)

(lneto)

- npf_do_nat: fix a race condition and simplify the logic.
- npf_session_setnat: clear the NAT association on failure.

(rmind)

repply the hacks from rev 1.2 to the gmp*.h files, and
make a note about it.

(mrg)

make trialdivtab.h depend on the Makefile.

(mrg)

consolidate the trialdivtab.h generation code.

(mrg)

Split checkout target into two parts now that llvm and clang are
imported. Requested by matt@

(joerg)

Remove the "evtchn_do_event: handler %p didn't lower ipl %d %d\n" printf.
With help from Robert Elz we've finally figured out what's going on, and
it actually isn't a bug in the handler, but related to spin mutexes.
When a spin mutex is released, the IPL isn't lowered back if the
curcpu is holding other spin mutexes. This is because mutexes may not
be released in order (and, in this case, the CPU in interrupted while
it holds a spin mutex at IPL < IPL_SCHED).
Also remove the test and resetting the IPL, it will be reset anyway
inside the loop, or at the end of the loop.

(bouyer)

Coverity fix:
979928 Use after free
debug3 line moved to before one of its arguments is freed

(spz)

Coverity complaint fixes:
bin/cat/cat.c 976654 Argument cannot be negative
                    (missing check for fileno result, stdout)
bin/cat/cat.c 976653 Improper use of negative value
                    (missing check for fileno result, stdin)

(spz)

- don't use defattr before it actually contains something useful
- clear the screen only once, when we know what colour to use
- set up the colour map only once, when we know which one to use
-> no more funky colour flashing while cgsix attaches on my U1

(macallan)

libkern build glue for compiler-rt.

(joerg)

Don't include bswap for ARM right now, the routines require ARMv6.
Separate the soft-float code and don't include it in the libgcc case.

(joerg)

+SIMPLEQ_LAST.0

(njoly)

remove accidently commited file

(martin)

Add ARM (EABI) specific sources. Split off code that requires C11,
unwind support in libc or overlaps with soft-float in preparation for
using compiler-rt in the non-clang case.

(joerg)

Don't use GOT relocations.  Use PC relative for the GOT and GOTOFF for
_DYNAMIC.  Make thumb friendly.

(matt)

fixed lua(4) build (added _KERNEL guard into lstrlib.c to avoid double usage)

(lneto)

Improve EABI handling of string functions. Most importantly, fix
__aeabi_memset, which has the arguments in the wrong order.

(joerg)

Fix pasto in the !ufs2 case (use dp->dp1 and not of dp->dp2).
This would be a problem only when allocating a new data block and the
indir block is already allocated, which explains why automated tests didn't
find it.
Problem reported on tech-kern@ and fix tested by manu@.

(bouyer)

Don't use cpsr_all/spsr_all with mrs, it doesn't take a mask.

(joerg)

Only write to the framebuffer memory 32 bits at a time.  If we try to write
larger amounts, we will generate SER_WRITE|SER_SZERR with P4 framebuffers.
(Writing 64 bits at a time is the default with memset().)
Also, set the default attributes to zero to avoid vertical stripes from
uninitialised defattr.
XXX; screen is still not cleared of old text.

(jdc)

Fix 1 bpp rasops copycols() op:

- fix inverted shift direction in MBL() and MBR() macro in BE case
  (used by GETBITS() and PUTBITS() in copycols() function in rasops_bitops.h)
- make all bitmask values unsigned and use proper uint32_t types for
  bitmap variables (to avoid arithmetic right shift)
- fix various botches in right-to-left copy op (logic is taken from hp300)

Tested on bwtwo(4) on NetBSD/sparc.

(tsutsui)

Fix off by one in copyrows() backward case.

(tsutsui)

Fix typo in secureplt support.

(skrll)

Let's use a format string, shall we?

(joerg)

Fix aliases.

(joerg)

Update TCP CB with new values on rfc1323 and mssdflt sysctl updates
From yasuoka@iij.ad.jp in kern/44254

(kefren)

merged luaconf.h of kernel and userspace Lua

(lneto)

using Lua auxlib in luacore module

(lneto)

put Lua auxlib and some stdlibs in lua(4)

(lneto)

changed lua_Number to int64_t

(lneto)

Handle more cases where a screen has not been fully initialized (inserted
into the scrq) by calling screen_fini() instead of screen_end() where
appropriate. Pointed out by chuq.

(christos)

Finish version auto-setting.

(christos)

new file(1)

(christos)

Get version from the configure script

(christos)

merge conflicts

(christos)

Put back the vnode changes I backed out yesterday; they were not the problem.
I've tested them with 2 -j 20 builds on an 8 cpu box. It crashed reliably
with the pcu changes present before.

(christos)

Enable warnings on sparc64 again

(martin)

Apply upstream fix for warnings seen on sparc64

(martin)

Add a /var/shm directory (to be used for shm_open and friends)

(martin)

In ehci_check_qh_intr don't treat a short control transfer as done if the
status phase is still inflight.  Let the hardware complete it.

PR/48358: Repeated low-speed USB control transfers returning short data
  fail on EHCI

PR/46696: uhub disables port where USB keyboard attached

(skrll)

Adjust KASSERT in device_ctrl_done to polling case.

PR/48213: ehci assertion failure during boot

(skrll)

remote_event_destroy is unused, mark it so.

(joerg)

Use format string.

(joerg)

For load/store double, name the second register explicitly.

(joerg)

Mark unused inline functions as such.

(joerg)

Use format strings.

(joerg)

Inform compiler that the volatile assignment will terminate the function.

(joerg)

Revert all previous TAILQ_NEXT() = NULL, changes, checks against
NULL vip's and removal of screens from queues. Instead introduce
a new screen_end1() function that can be used to clean screens that
are not associated with queues yet. Pointed out by chuq@

(christos)

fix indentantion

(christos)

One const ought to be enough for every variable.

(joerg)

revert fpu/pcu changes until we figure out what's wrong; they cause random
freezes

(christos)

Revert recent vnode changes per PR/48411, I still have deadlocks with
build -j 20 on an 8 cpu machine.

(christos)

add new internal functions

(christos)

- centralize opening of regular and temp files to avoid code duplication
- don't cast malloc
- use malloc sizeof(*var) instead of sizeof(type)

(christos)

These don't need count anymore.

(christos)

Disable parentheses checks since libgmp doesn't pass them.

(wiz)

fabsl alias.

(joerg)

Use explicit form of register pair operations by specifying both.

(joerg)

Like asm.h r1.25, don't use (PLT) anymore.

(joerg)

Use PLT_SYM.

(joerg)

Explicitly name the register pairs.

(joerg)

Avoid warning on sparc64

(martin)

Use PLT_SYM.

(joerg)

arm doesn't need to use (PLT) anymore.

(matt)

Add two new options:

-g which shows the GUID for a partition
-i which shows all the gory details for a particular option

XXX the output format of -i is subject to change

(jnemeth)

PR/48419 - Henning Petersen -- unbalanced parenthesis

(jnemeth)

Fix recovery mode, there were multiple issues:
1. the btree filename was not set so that we always used a transient
  in-memory db for the data
2. we did not call sync after creation with R_RECNOSYNC so that the header
  of the btree was never written
3. we did not call the right flavor of sync before copying the tree to the
  preserved files

(christos)

Don't rebuild the header files all the time if they did not change

(christos)

unwind-arm-common.h is EABI specific. Mark it as such.

(skrll)

Simplify

(skrll)

nvi has moved.

(nakayama)

pppd has moved.

(nakayama)

Wording improvements from Abhinav Upadhyay <er.abhinav.upadhyay@gmail.com>

(wiz)

Shut up a bunch of (valid) warnings from Clang.

(joerg)

Fix format string use.

(joerg)

Allow gnueabi to be matched for the relevant cases as well.

(joerg)

Mark usage as dead.

(joerg)

Include crtbegin.h first to make it possible to build with the stricter
attribute consistency checks in clang.

(joerg)

Bring back the compatibility db3->db1 code from the old vi.
This is selected with -DUSE_DB1 in the Makefile. By default
use use the db1 specific code from the current nvi which is
enabled using -DUSE_BUNDLED_DB. Unfortunately recovery seems
to be broken for both.

(christos)

Use normal style for checking for long double support. Add missing
aliases for fminl and fmaxl, fabsl is here just for symmetry.

(joerg)

Add fabsl alias if necessary.

(joerg)

make the debugging code work again.

(christos)

rcs-5.9.2 is out.

(wiz)

fix tracing.

(christos)

fix compilation of log debugging

(christos)

remove unused variable

(christos)

Change vrelel() to mark the vnode as changing after it has aquired
the vnode lock but before it calls VOP_INACTIVE().

Should fix the race between layer_node_find() trying to vget(, LK_NOWAIT)
a locked vnode when vrelel() marked it as changing and wants its lock.

PR kern/48411 (repeatable SMP crashes in amd64-current)

(hannken)

Make warnings non-fatal on sparc64.
Warings reported upstream.

(martin)

fix definition of __GMP_CC.

(mrg)

Long double related changes.

(joerg)

Add entry for LLVM, including note about CVS ACL.

(joerg)

clean up __GMP_CC, GMP_MPARAM_H_SUGGEST and CONFIG_TOP_SRCDIR.

(mrg)

Update nvi status:
. add Sven as upstream maintainer
. remove myself as responsible
. update Location

(aymeric)

GMP 5.1.3 updates for armeb and earm.

(mrg)

note GMP 5.1.3 is in tree.

(mrg)

note MPC, MPFR, and GMP updates.

(mrg)

GMP 5.1.3 files for earmeb and mips64el.

(mrg)

updates for GMP 5.1.3, and new build files for most platforms.

(mrg)

port forward our changes:
- allow working on thumb-missing platforms
- fix some minor issues

(mrg)

merge GMP 5.1.3 (part 1.)

(mrg)

add a note about MPC and MPFR potentially upsetting the update build.

(mrg)

reflect reality

(christos)

nvi has moved

(christos)

Make sure that the next pointers are NULL when we insert screens in the list.
This avoids the problem when screens are re-used that have stale pointers in
them. This was not an issue with circleq's because all the pointers used to
be updated.

(christos)

we have moved.

(christos)

Remove pppd

(christos)

These have moved to external

(christos)

pppd moved

(christos)

Unhook pppd

(christos)

hook ppp

(christos)

Merge local changes, add build glue.

(christos)

update for new binutils code our local patch to avoid warnings when macro are
used for loongson2f workarounds. Should fix mips64el kernel builds.

(bouyer)

Very slightly more consistency in ordering in SYNOPSIS.
Fix a few typos. Remove superfluous Pp.

(wiz)

Correct previous:
gmp in NetBSD is still at 5.0.2, newest version is 5.1.3.

(wiz)

Spell the null pointer as NULL, not as nul.

(joerg)