byebye PG_HOLE as it turned out to be unnecessary.

(yamt)

comments

(yamt)

MIPS uses a variable page size now.
(pulled up from matt-nb5-mips64)

(matt)

Let address/entry be 64bit if they are just sign-extended 32-bit values.
(change pulled up from matt-nb5-mips64).

(matt)

document postinstall pwd_mkdb item

(apb)

Add "pwd_mkdb" item, which checks whether /etc/pwd.db is in the
new format, and runs "pwd_mkdb -V 1 /etc/master.passwd" to fix it.

(apb)

s/tmux/tmux(1)/ in the entry about removal of window(1)

(apb)

Move fmtprint and print_needed since they are dependent on ELFSIZE.

(matt)

MIPS uses a variable page size now

(matt)

ACPICA 20120215 is out, including support for ACPI 5.0.

(jruoho)

don't rely on an application (such as ssss(1)) being able to mmap(2) the
shares it presents via libssss - allocate space using calloc(3) if the mmap
fails.

this change allows us to use ssss(1) to combine shares from different
websites, mounted using httpdev(8).

(agc)

return the number of characters read/written in the small target operation,
not the size that was requested.

(agc)

Ticket #6.

(riz)

Pull up following revision(s) (requested by martin in ticket #6):
distrib/common/Makefile.distrib: revision 1.9
distrib/i386/kmod/Makefile: revision 1.11
distrib/amd64/kmod/Makefile: revision 1.2
Use a new variable, MODULEVER, for kernel modules, tracking what's been
done elsewhere.

(riz)

Use a new variable, MODULEVER, for kernel modules, tracking what's been
done elsewhere.

(riz)

Ticket #6.

(riz)

Pull up following revision(s) (requested by christos in ticket #6):
distrib/sets/sets.subr: revision 1.136
share/mk/bsd.kmodule.mk: revision 1.34
sys/conf/osrelease.sh: revision 1.122
Add new -k (for kernel modules) flag to return MM.99.pp for current and MM.mm
for the release branch.
use osrelease -k to avoid that the path is 6.0_BETA
use osrelease.sh -k instead of hand-rolled logic.

(riz)

use osrelease.sh -k instead of hand-rolled logic.

(christos)

Derive MANINSTALL from MKCATPAGES and MKHTML.

(joerg)

use osrelease -k to avoid that the path is 6.0_BETA

(christos)

Add new -k (for kernel modules) flag to return MM.99.pp for current and MM.mm
for the release branch.

(christos)

Move the ksegx tlb init code into its own function.
Fix a problem with concurrent shootdowns by tracking what cpus want a
shootdown for a pmap, and if anoter cpu wants a shootdown, perform the
shootdown on ourselves.

(matt)

Add _dlauxinfo helper function to provide access to the ELF auxilary
vector. This can normally be found via __ps_strings, but libc is
initialised too early when linked dynamically and doesn't have access to
it yet, so provide an alternative mechanism via ld.elf_so.
Bump libc minor.

(joerg)

Add entries for bzip2 and xz.

(joerg)

Use mmap when dealing with plain files.

(joerg)

Do not create obsolete ./usr/include/quota directory anymore.

(njoly)

add cpuconf.h to the sets

(christos)

Added cpuconf.h needed by libkvm build.

(christos)

Add support for compressed man pages in all the usual formats.

(joerg)

Merge mandoc.h r1.99 and read.c 1.28 from upstream:
Add mparse_readmem, which allows application to run the parser directly
from memory, e.g. after de-compressing a document.

(joerg)

remove debug spam

(macallan)

use -n too, so that we don't get a dup note.

(christos)

Tickets 2,3

(riz)

Pull up following revision(s) (requested by joerg in ticket #2):
etc/weekly: revision 1.26
etc/daily: revision 1.81
UPDATING: revision 1.235
Run makemandb with the normal 022 umask, not the restrictive 077.
Extend entry to makemandb permission issue.

(riz)

Pull up following revision(s) (requested by christos in ticket #3):
sys/conf/osrelease.sh: revision 1.121
Don't eat _BETA. I am sure dsl will come up with something more elegant.

(riz)

mips_nfreelist is the one past the highest freelist used.

(matt)

Allow KSEG1 even on _LP64.

(matt)

enable_cp2 needs a early clobber

(matt)

window.debug is now obsolete.

(njoly)

cache glyphs in video memory when using anti-aliased fonts on r128fb and
radeonfb

(macallan)

generic support for caching glyphs in video memory
for speeding up anti-aliased fonts on slow CPUs

(macallan)

Don't eat _BETA. I am sure dsl will come up with something more elegant.

(christos)

Correct LDOUBLE_SIZE

(skrll)

Document -n.

(wiz)

The default blocksize is 4096 bytes.

(wiz)

Fix tags in set lists for window removal.

(wiz)

fix a description of PMAP_PREFER argument.

(yamt)

Add KASSERTs to uvm_pagealloc_pgfl to verify the page is actually free and has
the contents that it should.
Redo the KASSERTs for the pageq in uvm_pagefree.

(matt)

Add extern int mips_ksegx_tlb_slot;

(matt)

Use (intptr_t) when casting pointers to get proper sign extension.

(matt)

Add mips_ksegx_tlb_slot.
When casting pointers, use intptr_t so we get proper sign extension.

(matt)

When booting a N32 MP kernel, make sure to setup the ksegx mapping.

(matt)

Change db_expr_t to an register_t so we can see the full register contents
on N32 kernels.

(matt)

PTR_DSRL -> PTR_DSLL typo

(matt)

Make ps /l also printout the uarea.

(matt)

If we are building an INSTALL kernel and the ramdisk.fs exists in the object
tree, insert it into the kernel.  This needs to be done before the board
frag so it written into the kernel before the uboot images are created.

(matt)

Allow use of large (>256MB) bats for iobats.  If XBSEN isn't present, they
will be "downsized" into multiple 256MB bats.  Tested by riz.

(matt)

make sure we initialise the circa structures when decoding a file, and no
header is present.

(agc)

Add the ability to autoconfigure raid components on raw disks.
This change causes components on raw disks, as opposed to components inside
partitions or wedges, to be autoconfigured if the raid set is configured
for autoconfiguration.

Approved by oster@ and mrg@ for submission after the NetBSD-6 tag.  I've
been running these changes in production at my day job for over a year
without a problem.

See http://mail-index.NetBSD.org/tech-kern/2010/11/09/msg009167.html
for the original discussion of this patch and for a version of this patch
that works with NetBSD-5.x systems.

(buhrow)

Remove window(1).

window(1) was replaced by tmux(1) on March 2011.  The source code has
been reimported into othersrc/usr.bin/window and an up-to-date package
is available in pkgsrc/misc/window.

This removal was approved by core@, with the condition that it be done
once netbsd-6 had been brached.  And the branching has just happened!

(jmmv)

Track the victims selected by the pagedaemon and what happens to then.
Keep a hint for what page group has the most free pages for a given color.

(matt)

Recurse into 'window'.

(jmmv)

Initial import of window-20120215.

This is a snapshot of the window(1) source code taken from src on 20120215.
window(1) will be dropped from src, so this will remain here for posterity's
sake.  This copy of the sources will also be the canonical source to generate
the distfiles used by pkgsrc/misc/window, if any new ones have to be created.

Status:

Vendor Tag: TNF
Release Tags: window-20120215

(jmmv)

Don't hardcode a blocksize of 512 bytes.

Add a blocksize argument to httpdev (-b blocksize).

Bump default blocksize to 4KB.

(agc)

Add an argument to circa_file(3), and a command line flag (-n) to
circa(1), to denote that we do/do not want the circa header placed in the
output.  This is useful in cases where the sectorsize is pre-defined,
and so we do not need to place the value in a header in the output.

Bump shlib major version for the change to the API for libcirca(3).

(agc)

Pass t_renamerace and t_rmdirrace tests.

Adapt dholland@'s fix to ufs_rename to fix PR kern/43582.  Address several
other MP locking issues discovered during the course of investigating the
same problem.

Removed extraneous vn_lock() calls on the Ifile, since the Ifile writes
are controlled by the segment lock.

Fix PR kern/45982 by deemphasizing the estimate of how much metadata
will fill the empty space on disk when the disk is nearly empty
(t_renamerace crates a lot of inode blocks on a tiny empty disk).

(perseant)

- make this work for arm32
- factor out duplicated ifdefs

(christos)

don't need arm/cpu.h, we just need the cpureg.h...

(christos)

cast printf so it can be used from userland, and include the proper include
files for userland if not kernel.

(christos)

move stuff between the two files, so db_interface contains the functions it
needs for userland programs to link.

(christos)

Arrange so that we can be used from userland, by not accessing pointers
directly and using db_read_bytes().
XXX: arm26 not handled.

(christos)

expose cpu structures to kmemuser and prevent opt_*.h from being included
if kmemuser

(christos)

expose functions needed to compute stack traces to kmemuser

(christos)

expose stuff that crash needs to kmemuser

(christos)

include <sys/types.h>; things included from here need it.

(christos)

Add support for apropos -s like in the old apropos. Fix capitalisation.
Add note about optional quotation. From Abhinav Upadhyay.

(joerg)

Create CHANGES-6.0 and clear out LAST_MINUTE.

(riz)

Welcome to 6.0_BETA!

(riz)

Also handle hyphen replacement if it was used as plain input and no
backslash sequence was used at all in the line.

(joerg)

Be a bit more friendly to man pages using the roff .so command by
changing the current directory to the parent of the man -p entry, e.g.
/usr/share/man for /usr/share/man1.

(joerg)

Extend entry to makemandb permission issue.

(joerg)

Run makemandb with the normal 022 umask, not the restrictive 077.

(joerg)

Add NetBSD 7.0, and make the default 6.0.

(riz)

Note the new netbsd-6 branch, and move CHANGES to CHANGES.prev.

(riz)

Welcome to 6.99.1.

(riz)

Let address/entry be 64bit if they are just sign-extended 32-bit values.

(matt)

Back out the recent import of IPFilter 5.1.1 for the upcoming branch,
which will now have IPFilter 4.1.34.  IPFilter 5.1.1 will be restored
post-branch.

ok: core, releng.

(riz)

Add a N32 MP config.

(matt)

Add VIA VX900 host bridge to a buggy PCI mode 1 quirk table.  PR/46018
Ok releng@

(tsutsui)

Regen from pcidevs rev 1.1102:
> Add VIA VX900 Host Bridge, to handle its quirk. PR/46018
> Ok releng@

(tsutsui)

Add VIA VX900 Host Bridge, to handle its quirk. PR/46018
Ok releng@

(tsutsui)

fix for IPSEC tunnel + NAT-T + esp_frag:
Output packets larger than "esp_frag" are fragmented first
and then reinjected into ip_output for encapsulation
and transfer. The problem was that each packet got a new
ip_id value assigned, so that fragments couldn't be matched
by the receiver. Offset information was overwritten too.
approved by releng

(drochner)

callers of targetop() check for the number of characters returned, so return
the number of characters returned from targetop().

(agc)

Make sure that membar_producer gets the proper namespace mangling for libc.

(joerg)

Prevent recursive IO signals. This prevents crashes when pieces of text are
pasted in say an Xterm. Each char might create a SIGIO running out of
stackspace very quickly!

(reinoud)

Clean up cat pages that are older than the corresponding man pages.
Remove cat page directories that are empty.

(joerg)

Port amiga calls config_console() from machdep.c now, so we have to do
the same. Otherwise no console is configured anymore.
While being at it adapt files.amigappc to the latest changes in files.amiga.
Ok by releng@.

(phx)

Fix fallout from the new tests exercising all error paths: do not deactivate
the pmap of a vmspace-less child of a posix spawn operation that never
made it to userland.

(martin)

Remove trailing whitespace.

(wiz)

httpdev is a (very) small utility which allows a remote file to appear
as a local one, just like iscsi-initiator(8).  However, unlike iSCSI,
httpdev uses HTTP as a transport, so the remote end only requires a
web server.  The local end just requires the ability to issue
partial-range HTTP commands and read responses.  To do that, I wrote a
minimalist HTTP client library, although any such library could be
used.  Again, in contrast to iSCSI, no special protocol is used, there
is no encapsulation of SCSI commands, and no framing of blocks, and no
needless synthesis of SCSI commands in the initiator, or
interpretation of SCSI commands in the target.  Discovery requires no
login - either one can access a file on a webserver, or one can't.

To illustrate its use, the following commands should show how to use
the daily ISO images on nyftp to be read as an ISO image on the local
machine.

# httpdev http://nyftp.netbsd.org/pub/NetBSD-daily/HEAD/201202061530Z/iso/NetBSD-5.99.64-amd64.iso /mnt &
# ls -al /mnt/nyftp.netbsd.org/
total 18
drwxr-xr-x  2 agc  agc        1024 Feb 14 05:51 .
drwxr-xr-x  2 root  wheel          0 Dec 31  1969 ..
lrw-r--r--  1 agc  agc          12 Feb 14 05:51 address -> 128.59.23.63
lrw-r--r--  1 agc  agc          16 Feb 14 05:51 hostname -> nyftp.netbsd.org
lrw-r--r--  1 agc  agc            4 Feb 14 05:51 port -> http
lrw-r--r--  1 agc  agc          14 Feb 14 05:51 product -> NetBSD httpdev
-rw-r--r--  1 agc  agc    356603904 Feb 14 05:51 storage
lrw-r--r--  1 agc  agc          16 Feb 14 05:51 targetname -> nyftp.netbsd.org
lrw-r--r--  1 agc  agc            6 Feb 14 05:51 vendor -> NetBSD
lrw-r--r--  1 agc  agc          18 Feb 14 05:51 version -> bozohttpd-20100920
# file /mnt/nyftp.netbsd.org/storage
/mnt/nyftp.netbsd.org/storage: # ISO 9660 CD-ROM filesystem data 'NETBSD_59964                    ' (bootable)
# vnconfig vnd0 /mnt/nyftp.netbsd.org/storage
# mount -rt cd9660 /dev/vnd0a /cdrom
# ls -al /cdrom
total 6467
drwxr-xr-x  17 611  wheel    4096 Dec 31  1969 .
drwxr-xr-x  23 root  wheel      512 Jan  2 19:53 ..
drwxr-xr-x  4 611  wheel    2048 Feb  6 09:21 amd64
drwxr-xr-x  2 611  wheel    6144 Feb  6 09:04 bin
-r--r--r--  1 611  wheel    75120 Feb  6 09:25 boot
-rw-r--r--  1 611  wheel      555 Feb  6 09:25 boot.cfg
drwxr-xr-x  2 611  wheel    2048 Feb  6 09:25 dev
drwxr-xr-x  26 611  wheel    12288 Feb  6 09:25 etc
-r-xr-xr-x  1 611  wheel    3082 Feb  6 09:25 install.sh
drwxr-xr-x  2 611  wheel    10240 Feb  6 09:05 lib
drwxr-xr-x  3 611  wheel    2048 Feb  6 08:41 libdata
drwxr-xr-x  4 611  wheel    2048 Feb  6 09:25 libexec
drwxr-xr-x  2 611  wheel    2048 Feb  6 08:41 mnt
drwxr-xr-x  2 611  wheel    2048 Feb  6 09:25 mnt2
-rw-r--r--  1 611  wheel  6175009 Feb  6 09:25 netbsd
drwxr-xr-x  2 611  wheel    18432 Feb  6 09:05 sbin
drwxr-xr-x  3 611  wheel    2048 Feb  6 08:49 stand
-rwxr-xr-x  1 611  wheel  180679 Feb  6 09:25 sysinst
-rwxr-xr-x  1 611  wheel    29983 Feb  6 09:25 sysinstmsgs.de
-rwxr-xr-x  1 611  wheel    28954 Feb  6 09:25 sysinstmsgs.es
-rwxr-xr-x  1 611  wheel    29458 Feb  6 09:25 sysinstmsgs.fr
-rwxr-xr-x  1 611  wheel    25761 Feb  6 09:25 sysinstmsgs.pl
drwxr-xr-x  2 611  wheel    2048 Feb  6 09:25 targetroot
drwxr-xr-x  2 611  wheel    2048 Feb  6 08:41 tmp
drwxr-xr-x  8 611  wheel    2048 Feb  6 09:25 usr
drwxr-xr-x  2 611  wheel    2048 Feb  6 09:25 var
#

httpdev uses refuse(3) and the virtdir(3) routines, and a small http
client library I wrote.  If that was replaced with libcurl, then
httpdev could also take advantage of https (assuming we trust the
CAs...)

As for size:

% size ./httpdev
  text  data    bss    dec      hex  filename
  15065  1328  1128  17521  0x4471  ./httpdev
%

If someone were wanting to distribute a new release of an operating
system, httpdev could be very useful for accessing ISO images on a
central webserver, for example. Just saying...

Status:

Vendor Tag: CROOKS
Release Tags: httpdev-base

(agc)

make BATs >256MB work, now macppc works again on 745x CPUs as well
ok riz

(macallan)

fix crash during early startup - we can't call OpenFirmware between zapping
the BATs and complete setting them up again
ok releng

(macallan)

When making BATU, use (BAT_XBL|BAT_BL) for the extended bat lengths.

(matt)

Add 4KiB more of SYMTAB_SPACE to fix MV2120_INSTALL build.

(jakllsch)

another blah() -> blah(void)
ok riz

(macallan)

fix 'old style function definition' warning:
blah() -> blah(void)
ok riz

(macallan)

prototypes

(christos)

fix proto

(christos)

add prototypes.

(christos)

Regen from revision 1.87 of doc/BUILDING.mdoc.

(apb)

Minor changes to wording in the documentation for
build.sh install-image and live-image.

(apb)

Document build.sh "live-image" and "install-image" targets.

These were documented in revision 1.97 of src/BUILDING, but were
not documented in src/doc/BUILDING.sh.  The original log message
was:

revision 1.97
date: 2012/01/22 03:53:32;  author: tsutsui;  state: Exp;  lines: +39 -0
Add "live-image" and "install-image" target support to build.sh.

"live-image" target builds pre-installed disk images that can be used on
emulators or boot from USB memory sticks to try NetBSD without installation.
Currently amd64, i386, pmax, sparc, sparc64, sun2, sun3, and vax
(which have working emulators and don't require extra tools like preparing
msdosfs or partitioning MD label structures) support this target.

"install-image" target builds an bootable installation disk image that can
be used as an install CD but burned into USB memory sticks etc.
Current only amd64 and i386 (which would support USB boot) have this target.

For more details (and known issue) see following posts on netbsd-bugs@:
http://mail-index.NetBSD.org/netbsd-bugs/2011/08/06/msg023639.html
http://mail-index.NetBSD.org/netbsd-bugs/2011/09/23/msg024207.html
http://mail-index.NetBSD.org/netbsd-bugs/2011/12/07/msg025166.html
http://mail-index.NetBSD.org/netbsd-bugs/2011/12/08/msg025178.html

No particular comments about implementation, and
"go ahead" comments from mrg@, riz@ and christos@.

Closes PR toolchain/45153 and PR misc/45155.

(apb)

Build release notes if groff is present (MKGROFF!=no), not if man pages
are requested (MKMAN!=no).

(joerg)

since we have to drop locks to call into the usb code, we need to make
sure that only one caller ends up doing something in close_out_jack().
add a "closing" member that is set when closing and error out in various
places.

with this in place i can read and write from umidi without triggering
any locking or other obvious issues, though the writing is currently
broken (it worked in 5.99.60-era.)  it runs the correct time, but no
output occurs no the synth itself.  more work needed here.

(mrg)

void function _arc4randbytes cannot return a value.

(njoly)

add aubtfwl.4 to the build

(plunky)

a manual page for aubtfwl(4), detailing where to find
a firmware file and where to put it

(plunky)

Mark usage as __dead.

(joerg)

Add workaround for ARM cores that don't properly implement the MRC instruction
used for reading the thread pointer.  This is the corresponding change which
was made to <arm/mcontext.h>

(matt)

Kill dup lines.

(njoly)

Bump MV2120 SYMTAB_SPACE.
Add kernels variants for root on the primary partitions of "Drive 1".
Add MV2120 to evbarm release.

(jakllsch)

/home/paul/COMMIT

(pgoyette)

Whitespace in previous.

Ride previous releng OK.

(skrll)

Obsolete /usr/share/man/cat3f that never contained any files, nor will it.

(wiz)

fix incomplete device_t/softc split which led to crash on attachment,
closes PR kern/45874 by Hauke Fath
approved by releng

(drochner)

Pass RI_NO_AUTO to rasops_init so that rasops doesn't attempt to allocate
memory as we're too early in kernel startup for this.

My A7000 boots now.

OK releng. ack 3 nak 0.

(skrll)

Remove unused a valiable 'siopdbg'.

(kiyohara)

The default configuration for TLS support is supposed to use Sun
as, but we use GNU as.  So in this case it generates wrong
relocations in non-PIC mode.

This change fixes lib/libc/tls/t_tls_dynamic test failure on 32-bit
sparc as described in PR toolchain/45985.

Approved by releng.

(nakayama)

Fix typo.

(wiz)

Add missing word in comment.

(wiz)

Comment out test n3_6 - the data files are missing

(martin)

Fix various LP64 thinkos.

(matt)

mem_cluster_cnt on mips is now a u_int.

(matt)

Add more KASSERTs (more! more! more!).
When returning page to the free pool, make sure to dequeue the pages before
hand or free page queue corruption will happen.

(matt)

Fix the build after ipfilter tests were added.

(riz)

Add a few more posix_spawn testcases, which should detect everything that
was wrong in the initial version, causing the failure reported in
PR kern/45991.

(martin)

PAT flags are not under control of Xen domains currently, so there is no
point in enabling them.

Avoids:
- a warning logged by hypervisor when a domain attempts to modify the PAT
MSR.
- an error during domain resuming, where a PAT flag has been set on a page
while the hypervisor does not allow it.

ok releng@.

(jym)

Use separate pending and paging tailq entries.
Add a queue check routine to validate the queues aren't corrupt.

(matt)

When I split non-quota-specific functions out from ../ffs/quotas_common.sh
I forgot to ajust this test.
Fix PR bin/46015

(bouyer)

Move posix_spawn tests to lib/libc/gen/posix_spawn - they test both libc
and kernel, but that is an implementation detail unrelated to the tests
themselfs.
Ok: releng

(martin)

Build fix - convert to new-style definitions.
OK by releng.

(jdc)

Update dangling references to quotactl(2), mostly now libquota(3), and to
quotactl(8), mostly now quotarestore(8). ok riz@

(dholland)

Fix wrong xref; wrote quotactl(8), meant quotacheck(8). ok riz@

(dholland)

Add cpu_spawn_return() as per other archs to make it compile. No tests done
yet.

(reinoud)

Note removal of quotactl(8) and addition of quotadump(8)/quotarestore(8)
for backup purposes. While here, also improve the earlier quota-related
entries.

At quota2 merge time [20110306]:
  - main entry is for "ffs", not "kernel"
  - add entry to note addition of repquota -x
  - add entry to note addition of quotactl(8)
  - add entry to note changes to quotactl(2)
  - avoid explicitly saying to use repquota -x and quotactl(8) to migrate
    as it's no longer quite accurate

At libquota addition time [20120201]:
  - main entry is for "libquota", not "kernel"
  - add entry to note that quotactl(2) is gone and __quotactl(2) is now
    a private interface
  - add entry to note that the repquota -x dump format changed (from xml
    to tabular)
  - continue to make no reference to the vfs-level changes as they aren't
    user-facing

Prompted/suggested by apb@; ok by martin for releng.

(dholland)

Update ipf test suite with test cases from IPFilter 5.1.1

(darrenr)

Small typo in macro (.xr -> .Xr).

(njoly)

Add mm_md_direct_mapped_virt (inverse of mm_md_direct_mapped_phys).  Add a
third argument, vsize_t *, which, if not NULL, returns the amount of virtual
space left in that direct mapped segment.
Get rid most of the individual direct_mapped assert and use the above
routines instead.
Improve kernel core dump code.

(matt)

Update reference output to match yesterday's fix. ok martin@ for releng

(dholland)

revert prior change to unbind_jacks_from_mididev(), it's wrong.
slightly clean up free_all_jacks().

(mrg)

Make this compilable again.  Use the new defines and switcht to
CFATTACH_DECL_NEW.

(matt)

Modify __lwp_getprivate_fast to call _lwp_getprivate if the mrc instruction
returns 0 (some ARM cores are broken and don't raise an undefined exception
when encountering an unknown or privileged mrc instruction in user mode).

(matt)

Deal with the limited range of VAX floating point numbers when compiling
for VAX.

(matt)