Add reconnect code to librumpclient.  In case the connection to
the kernel server is lost, the client will now automatically attempt
to reconnect.

Among other things, this makes it possible to "reboot" and restart
the TCP/IP stack from under firefox without any perceivable less
of service.  If pages were loading at the time the TCP/IP server
was killed, there may be some broken links, but nothing a ctrl-r
cannot fix.

(pooka)

correct history

(plunky)

Add support for newer eGalax controllers that send 5 ot 6 byte long packets.
The protocol is backwards compatible.  While here, unfold a few lines and
remove the link to the documentation, which was invalid.
This enables the touchscreen on the PROTECH PS-3100 MINI POS system.

(mbalmer)

Fix spelling: it's *Jef* Raskin, not Jeff Raskin.

(dyoung)

Fix build.

(skrll)

roundup2(x, y) does not assume that x is power of 2, but y is.

(enami)

Mark librump{kern_solaris,fs_zfs}.so.0.0.debug as zfs only and fix a typo
while I'm here.

(skrll)

Provide LWP_PC

(martin)

Provide LWP_PC

(martin)

Represent test data as bytes to make it "endian-independent".

from martin

(pooka)

Provide LWP_PC

(skrll)

Use Fx.

(wiz)

document what -a is for

(yamt)

Note pcf8563rtc(4).

(jakllsch)

Add pcf8563rtc(4) manual page. Derived from rs5c372rtc(4).

(jakllsch)

fix typo in "-traditional-cpp"

(lukem)

Ticket 1413

(bouyer)

Ticket 1413

(bouyer)

Detect and spin up non-boot CPUs.

Mostly from OpenBSD.

(skrll)

Apply patch, requested by spz in ticket 1413:
dist/bind/ patch
include/resolv.h patch
include/arpa/nameser.h: patch
lib/libc/include/isc/assertions.h: patch
lib/libc/include/isc/eventlib.h patch
lib/libc/isc/assertions.c: patch
lib/libc/nameser/ns_parse.c patch
lib/libc/resolv/herror.c: patch
lib/libc/resolv/res_init.c patch
lib/libc/resolv/res_mkquery.c: patch
lib/libc/resolv/res_query.c patch
lib/libc/resolv/res_send.c: patch
usr.sbin/bind/Makefile.inc patch
usr.sbin/bind/include/config.h: patch
usr.sbin/bind/include/dns/code.h: patch
usr.sbin/bind/include/dns/enumclass.h: patch
usr.sbin/bind/include/dns/enumtype.h: patch
usr.sbin/bind/include/dns/rdatastruct.h: patch
usr.sbin/bind/include/isc/platform.h: patch
usr.sbin/bind/include/lwres/netdb.h: patch
usr.sbin/bind/include/lwres/platform.h: patch
usr.sbin/bind/libisc/Makefile: patch
lib/libc/inet/inet_addr.c new file

Update bind to  9.4-ESV-R4, latest version of the 9.4 line.

(bouyer)

Apply patch, requested by spz in ticket 1413:
dist/bind/ patch
include/resolv.h patch
include/arpa/nameser.h: patch
lib/libc/include/isc/assertions.h: patch
lib/libc/include/isc/eventlib.h patch
lib/libc/isc/assertions.c: patch
lib/libc/nameser/ns_parse.c patch
lib/libc/resolv/herror.c: patch
lib/libc/resolv/res_init.c patch
lib/libc/resolv/res_mkquery.c: patch
lib/libc/resolv/res_query.c patch
lib/libc/resolv/res_send.c: patch
usr.sbin/bind/Makefile.inc patch
usr.sbin/bind/include/config.h: patch
usr.sbin/bind/include/dns/code.h: patch
usr.sbin/bind/include/dns/enumclass.h: patch
usr.sbin/bind/include/dns/enumtype.h: patch
usr.sbin/bind/include/dns/rdatastruct.h: patch
usr.sbin/bind/include/isc/platform.h: patch
usr.sbin/bind/include/lwres/netdb.h: patch
usr.sbin/bind/include/lwres/platform.h: patch
usr.sbin/bind/libisc/Makefile: patch
lib/libc/inet/inet_addr.c new file

Update bind to  9.4-ESV-R4, latest version of the 9.4 line.

(bouyer)

DEBUG does not imply DIAGNOSTIC; make sure we have a non-null
KASSERTMSG implementation (DIAGNOSTIC) so that the variable inside
the DEBUG section gets used.

(he)

Explicitly initialize sometimes-unused variable.

(jakllsch)

Add support for __read_mostly and __cacheline_aligned (using 32 bytes as
the cacheline as its the most prevalent sized used even though IBM970 has
a 128 byte cacheline).

(matt)

Clear ownership bit after saving contents to PCB.

(matt)

The previous commit removed a { }�block, and unintendedly introduced a
C99ism (inlined variable declaration), which hurts when pdksh is used
for bootstrapping pkgsrc. Move the two declarations to the beginning
of the block.

(hauke)

The Solaris 7 "/usr/{,xpg4/}bin/sort"s expect whitespace between the "-k"
option and its argument.

(hauke)

Some ports use %r0 as tmp2 to INIT_CPUINFO which has adverse effects so
never use tmp2 to hold an address.

(matt)

Correctly define inta using a helper macro for the #

(joerg)

Define NOMAN earlier

(joerg)

Define NOMAN earlier

(joerg)

Define NOMAN earlier.

(joerg)

Defer bsd.own.mk include after NOMAN.

(joerg)

Redo previous.
bsd.own.mk must be included after NOMAN=.

(nakayama)

Cast arguments to vaddr_t when using PRIxVADDR in the printf format string.

(mbalmer)

Make some more information (psw/fpu_state) per cpu.

(skrll)

Whitespace.

(skrll)

Correct/add some comments about eiem.

(skrll)

switch to use newvers_stand.sh as suggested by joerg@

(nisimura)

Change ^T/SIGINFO to print the PC/CPU# or PC instead of "running"/"runable"
if a port provide LWP_PC.

(matt)

Add arc64 alias

(matt)

Add a new O(log(2) implementation.  On mips32/mips64, use clz/dclz.

(matt)

mips64 gcc now defaults to soft-float

(matt)

Default mips64 to softfloat.

(matt)

Teach libkvm how kernel virtual memory is laided on 64-bit MIPS kernels.

(matt)

Instead of testing for all mips variants, just use MACHINE_CPU.

(matt)

Use uintptr_t instead of u_long.
When fetching argc use off the stack, use *(long *) instead of *(int *)
because on mips64 BE system, that int deference will only fetch 0.

(matt)

Pull the following fix from OpenBSD:
http://www.openbsd.org/cgi-bin/cvsweb/src/sys/dev/pci/if_nfe.c#rev1.97
>> Some nfe(4)/rlphy(4) combos don't work, because the PHY responds to all
>> addresses on the mii bus.  As a countereasure, only attach the first PHY we
>> encounter.  It is very unlikely we're going to ever see nfe(4) with multiple
>> PHYs.  The same is probably true for any modern NIC.

Should suppress bogus ukphy30 noted in PR kern/42950.
No bad side effect on my nfe(4) which didn't get ghosts:
---
nfe0 at pci0 dev 5 function 0: NVIDIA nForce3 Ethernet #4 (rev. 0xa2)
:
rlphy0 at nfe0 phy 1: RTL8201L 10/100 media interface, rev. 1
rlphy0: 10baseT, 10baseT-FDX, 100baseTX, 100baseTX-FDX, auto

(tsutsui)

fix a failure in the sign extension consideration.

(nisimura)

- add "ll" modifier to printf.
- be more conscious about int type propagation.
- add missing version and newvers.sh

(nisimura)

change bootloader name to "altboot".  bump version and add README.

(nisimura)

Add the ability to mount NFS filesystems in COMPAT_NETBSD32
If in the kernel and NFS_ARGS_ONLY, just export struct nfs_args and its flags.

(matt)

stop screaming: s/DEBUG/debug/g

(christos)

fix previous.

(christos)

Make printing of lwp flags similar to the process one, identifying (O)nproc
(K)ernel threads and scheduler (a)ctivations.

(christos)

Use the L_ flags instead of the P_ flags for lwps.

(christos)

Fix c&p error.

(matt)

A couple more registers and their common usage.

(skrll)

avoid advertising 0/8 transport address (conforming RFC1122
Section 3.2.1.3)

(kefren)

Save CR24 and CR27 and report in ddb.

(skrll)

use SLIST_FOREACH_SAFE when deleting a peer in loop

(kefren)

Make __SIMPLELOCK_RAW_UNLOCKED available.

(skrll)

Have the register number and usage available.

(skrll)

Whitespace.

(skrll)

Drop bootprog_maker (formerly enabled by -M) and bootprog_date (formerly
disabled by -D) from the output of newvers_stand.sh.  Change -D to the
inverted logic, so that it adds the date to bootprog_rev in ().

Change all platforms accordingly. -D is added if MKREPRO is not yes and
wasn't present before. Platforms that didn't use -D don't depend on
MKREPRO now either.

(joerg)

undo previous. Read the diff wrong.

(christos)

First attempt at MP _lock_cas.

(skrll)

add an ifndef __linux__ requested by mrg

(ahoka)

inverse inversely incorrect comment

(pooka)

When deleting a fragment header use the simple copy operation only if it fits
completely into the mbuf.

(mlelstv)

fix comment. from PR kern/44395

(cegger)

A not rewritable disc can be pseudo-overwritable. A pseudo-overwritable disc
can overwrite its logical volume integrity sequence so no need to make it
append-only for this class of discs.

(reinoud)

Fix a fatal typo that causes TLB miss panic in MachFPInterrupt().
Reported in followups of PR port-cobalt/44292.

(tsutsui)

Adjust sc_frequency in MD attachment as quirks.
It is not applicable to all 1665x variants.

XXX: It looks IOBlix board actually has 22.1184MHz OSC without clock divisor
    http://amiga.resource.cx/photos/gallery/ioblix.jpg
    so needs to investigate why we have to divide sc_frequency by four
    to get right baud rate on IOBlix com. (firmware sets MCR_PRESCALE?)

(tsutsui)

Revert part of changes in rev 1.298:
- it breaks cobalt's serial console as mentioned in PR port-cobalt/44292
- MCR_PRESCALE doesn't affect unless EFR_EFCR is set in the EFR register
- even if MCR_PRESCALE is enabled we should define appropriate sc_type
  variants and BRG values should be adjusted in comspeed() per sc_type
- sc_frequency should be adjusted in MD attachment if necessary
Tested on cobalt by several people, ok from jklos@

(tsutsui)

fix comment

(christos)

remove stray printf.

(christos)

Define HAVE_PCI_DEVICE_IS_BOOT_VGA.
Makes Xorg use of pci_device_is_boot_vga.

First discussed with macallan@ then public on tech-kern@ and tech-x11@

(cegger)

Implement pci_device_is_boot_vga() for NetBSD using new WSDISPLAYIO_GET_BUSID ioctl.
This lets Xorg start on machines with more than one PCI display device
instead of exiting with 'Primary device is not PCI'.

First discussed with macallan@ then public on tech-kern@ and tech-x11@.

(cegger)

Implement new WSDISPLAYIO_GET_BUSID ioctl.
It returns the bus id and allows userland (like Xorg) to create mapping
of ttyE? and bus id. For now only PCI is implemented.

First discussed with macallan@ then public on tech-kern@ and tech-x11@

(cegger)

Lookup the physical partition backing up the logical one when searching for the
metadata partition overlap for BD-R.

Fixes a kernel panic on mounting a BD-R formatted with UDF 2.60

(reinoud)

Whitespace nits.

(tsutsui)

happiness is a quiet lint

(pooka)

shush lint

(pooka)

Fix wrong function names in messages by using __func__.  PR kern/44431

(tsutsui)

Check sys_reboot() return value now that it's possible.

(pooka)

In case sys_reboot() was called by a remote client, put the response
in the socket before we shut down.  This way the response to the
syscall travels to the caller and they know things worked correctly
instead of having to just assume.

(pooka)

Add a comment explaning why uvm_reclaim_hook is not currently
necessary for rump.

XXX: strictly speaking, the zfs implementation does not work
correctly since it just wakes up a thread instead of performing
the scan in the context of the pagedaemon

(pooka)

convert xpmsg_lock to IPL_SCHED.  the old spl/simple_lock code ran at
splsched(), and this significantly helps with stability under load
when running with multiple active CPUs.

(mrg)

Fix build of ndiscvt(8). Not sure what broke it in the first place.

(tron)

Fix MULTIPROCESSOR curlwp/curcpu handling... missed in previous commit.

(skrll)

Fix MULTIPROCESSOR curlwp/curcpu handling.

(skrll)

XXX:

in strayintr() don't print about stray zs inters in MP case.

(mrg)

move <sys/device_if.h> outside of _KERNEL since the bulk of the previous
change depends upon "device_t" being defined.

fixes sparc/stand build.

(mrg)

Fixes from PR kern/44395 Masanori Kanaoka:
- Fix register address in ale_phy_reset(). from linux
- Fix mask value in ale_stop_mac().
- Fix multicast handling. from openbsd
- Fix phy reset handling.

(cegger)

From Roman Hoog Antink <rha@open.ch>: Fixes a null pointer dereference
that might occur after removing peers from the config and then reloading.

(tteras)

From Roman Hoog Antink <rha@open.ch>: Fixes a null pointer dereference
that might occur after removing peers from the config and then reloading.

(tteras)

Fix the corruption of ps->start_hint.

(matt)

- make sure to wait for BUSY bit goes down prior to each 512B block
read from xfer register, or corrupted datum may be retrieved.  Uncovered
by corrected PIO READ_CMD value.
- remove redundant register assignments to set transfer mode.

(nisimura)

add may alias attribute to the double union because gcc-4.5.x produces
incorrect code on ARMv7a. From nikunj badjatya

(christos)

prevent resource DoS from brace expansion (from Maksymilian Arciemowicz)

(christos)

Exclude drm sources from tags computation.

(dyoung)

Per suggestion from tsutsui, convert to todr_[gs]ettime_ymdhms flavor.

(jakllsch)

Bump SYMTAB_SPACE so that it fits again.

(he)

When requesting to disable metadata partitions, also disable low formatting.
When no meta and not low formatting we can lower the required UDF version.

(reinoud)

Fix typo in comment

(reinoud)

unifdef -U__linux__ -U__FreeBSD__.  No functional change intended.

(dyoung)

Split out the raw to indexed partitioning code protecting against roque
implementations that use `ramdom' numbers for the physical partitions breaking
lots of implementations. Known curlpit is MicroSoft Windows 7.

Not only the partition mappings need to be protected against this but also the metadata partition files.

(reinoud)

Cleanup/add some asserts.  no functional change.

(matt)

Add pcf8563rtc(4), yet another I2C real time clock.

(jakllsch)

Describe device_printf(9).

XXX Need symbolic link device_printf.9 -> kprintf.9.

(dyoung)

Move device_printf() from ath_netbsd.c to subr_prf.c for reuse in a new
driver.

(dyoung)

Add support for quotactl("getall") command, and convert repquota to new
world.

(bouyer)

buildfix: use PRIxPADDR for type paddr_t

(cegger)

Remove FILES section, it's obsolete

(bouyer)

Switch remaining platforms to modern CPP for assembler.

(joerg)

no longer used

(pooka)

Since "physmem" is largely unused except for zfs wanting to know
if it has enough memory available, just pick a number which makes
zfs happy.

We *could* use a MIN() of available host mem and rump_memlimit,
though ...

(pooka)

Don't allow module to init if root is not mounted since the modcmd
wants to traverse the file system.  This *might* fix statically
linking zfs support into the kernel (at least it fixes rump kernel
bootstrap).

(pooka)

fix kobj_open_file_vnode()

patch from haad

(pooka)

don't perform tasks already done by zfs_modcmd

(pooka)

Forgot to commit this a few days ago: need to make sure we get the
right kqueue/kevent now that rumphijack hijacks them.

(pooka)

Add comma in enumeration.

(wiz)

dont negate error value passed to strerror
the linux libpciaccess returns negated value, but no other os does...

(ahoka)

New sentence, new line.

(wiz)

Avoid a crash when pDrawable is null.

(ahoka)

Fix typo reported in PR 44429 by Ryo HAYASAKA.

(wiz)

Move the termcap man page and capabilities to the Attic as we have been
using terminfo now for quite some time.

(roy)

Describe the new terminfo environment settings.

(roy)

A place for Lua 5.1 modules.

(mbalmer)

fix bug introduced in rev. 1.3: need to set the string before
converting it into a numerical value, otherwise we might get junk or crash

(drochner)

avoid crash if certain operations are done before an input record is
read, ie in a BEGIN rule

(drochner)

Remove trailing whitespace.

(skrll)

Another comment tweak

(skrll)

Tweak comment.

(skrll)

Move the clock_intr declaration.

(skrll)

#define<tab>

(skrll)

unreserve kmixer

(jmcneill)

bump libchromeXvMC major to match the other hw libraries to make
libXvMCW happy

(jmcneill)

Add a simple dhcp client for rump kernels.  It uses code from dhcpcd
for talking to the server and for setting the interface address
and route.  However, otherwise it is quite different, since we need
to be working under the assumptions that there is no stable storage
on a rump instance, and that there are n networking stacks on a
given host.

(pooka)

Change to use MAKEOBJDIR instead of MAKEOBJDIRPREFIX

(matt)

Remove CPUFLAGS?= since it's no longer needed.

(matt)

build and install libchromeXvMCPro from xf86-video-openchrome

(jmcneill)

Use DRICreatePCIBusID if present instead of building the Bus ID string
manually.

(jmcneill)

fixed a typo, it will now compile when KMADDRESS is defined. reported by Roman Hoog Antink (rha (at) open.ch)

(vanhu)

+librumphijack.so.0.0.debug

(njoly)

- Replace fatal2() with the appropriate err(3)/errx(3) calls.

(ginsbach)

Add missing is_a600() definition, used by amiga devices.
We simply define it as 0, because an A600 will never have a PPC installed.

(phx)

- Fix an incorrect comparison (always true)
  This created an unnecessary temporary file when using the -o option
  to create a new file.

- Fix -o option bug reported by Tony Finch
  The output file was incorrectly called "[stdin]" when overwriting
  the the input file via standard in (unifdef -o file < file).

(ginsbach)

include pid of rump server in hostname

(pooka)

Add missing .debug extension.

(njoly)

Match only first instance of rlphy, following suite from Free/OpenBSD.

should fix PR kern/42950, variant of patch from "js" on irc

(pooka)

Snapshot of work in progress on a modernised disk quota system:
- new quotactl syscall (versionned for backward compat), which takes
  as parameter a path to a mount point, and a prop_dictionary
  (in plistref format) describing commands and arguments.
  For each command, status and data are returned as a prop_dictionary.
  quota commands features will be added to take advantage of this,
  exporting quota data or getting quota commands as plists.

- new on disk-format storage (all 64bit wide), integrated to metadata for
  ffs (and playing nicely with wapbl).
  Quotas are enabled on a ffs filesystem via superblock flags.
  tunefs(8) can enable or disable quotas.
  On a quota-enabled filesystem, fsck_ffs(8) will track per-uid/gid
  block and inode usages, and will check and update quotas in Pass 6.
  quota usage and limits are stored in unliked files (one for users,
  one for groups)l fsck_ffs(8) will create the files if needed, or
  free them if needed. This means that after enabling or disabling
  quotas on a filesystem; a fsck_ffs(8) run is required.
  quotacheck(8) is not needed any more, on a unclean shutdown
  fsck or journal replay will take care of fixing quotas.
  newfs(8) can create a ready-to-mount quota-enabled filesystem
  (superblock flags are set and quota inodes are created).
  Other new features or semantic changes:
  - default quota datas, applied to users or groups which don't already
    have a quota entry
  - per-user/group grace time (instead of a filesystem global one)
  - 0 really means "nothing allowed at all", not "no limit".
    If you want "no limit", set the limit to UQUAD_MAX (tools will
    understand "unlimited" and "-")

  A quota file is structured as follow:
  it starts with a header, containing a few per-filesystem values,
  and the default quota limits.
  Quota entries are linked together as a simple list, each entry has a
  pointer (as an offset withing the file) to the next.
  The header has a pointer to a list of free quota entries, and
  a hash table of in-use entries. The size of the hash table depends
  on the filesystem block size (header+hash table should fit in the
  first block). The file is not sparse and is a multiple of
  filesystem block size (when the free quota entry list is empty a new
  filesystem block is allocated). quota entries to not cross
  filesystem block boundaries.

  In memory, the kernel keeps a cache of recently used quota entries
  as a reference to the block number, and offset withing the block.
  The quota entry itself is keept in the buf cache.

fsck_ffs(8), tunefs(8) and newfs(8) supports are completed (with
related atf tests :)
The kernel can update disk usage and report it via quotactl(2).

Todo: enforce quotas limits (limits are not checked by kernel yet)
      update repquota, edquota and rpc.rquotad to the new world
      implement compat_50_quotactl ioctl.
      update quotactl(2) man page

fsck_ffs required fixes so that allocating new blocks or inodes will
properly update the superblock and cg sumaries. This was not an issue up
to now because superblock and cg sumaries check happened last, but now
allocations or frees can happen in pass 6.

(bouyer)

save_fpu_lwp() was renamed to fpu_save_lwp(), so follow suit.

(he)