Look at the proper offsets for MD info, depending on the size of the binary
32 or 64.

(christos)

new file(1)

(christos)

merge conflicts

(christos)

Backout previous - we do not need it anymore

(martin)

Use C++ compatible declaration for posix_spawn (instead of the C99 specific
posix one).

(martin)

from ftp.astron.com, CDF security fixes

Status:

Vendor Tag: CHRISTOS
Release Tags: FILE5_11

(christos)

fix wrong variable, import command

(christos)

Only attempt to handle as much data as we can handle.
Per my testing, allows normal MAXPHYS on domU kernels.

(jakllsch)

Define XBD_MAX_XFER instead of using (PAGE_SIZE * BLKIF_MAX_SEGMENTS_PER_REQUEST) directly.

(jakllsch)

quote SRC_DIR and DEST_DIR everywhere.  Also wrap some long lines.

(apb)

Fix a call to pwd_mkdb in the case that DEST_DIR is the empty string.

(apb)

Fix typo in usage

(martin)

fix stack backtraces through traps.

(chs)

- nuke commented-out SYMTAB_SPACE directive since this port has a
  companion bootloader which is capable of loading a plain ELF kernel
  and accompanying ELF symbol table.
- remove fxp(4) from GENERIC as no known NAS has the device.

(nisimura)

- detect SoC chipid value found at 0x5600'00b0.
- have brdsetup() to define SoC IOMUX in early startup. Device drivers
  are still responsible to arrange relevent registers which they use.
- some cleanup and clarification.

(nisimura)

- adapt DM9000 driver name change.
- move -DLIBSA_PRINTF_WIDTH_SUPPORT to CPPFLAGS variable as SACPPFLAGS
  does not work.
- add a logic to make/remove machine/ and arm/ symlinks.
- remove redundant directives in no use.

(nisimura)

Reimplement DM9000E driver.
- add a note about 16bit mode operation.
- make sure not to receive any mcast/bcast frames by clearing 64bit
  multicast filter hash value register.
- change the filename so as silimar to other implementations.

(nisimura)

Add the --trusted-keys argument to netpgpkeys(1) to print out PGP ids in a
machine-readable manner.

(agc)

re-order the fields that we print out in the pgp_sprint_pubkey() function
to be more usual.

print out the name from within pgp_sprint_pubkey() rather than tagging it
onto the end of the output from the function.

(agc)

Move entries for things that have been pulled up into the "netbsd-6"
branch to the list of changes in the previous release.

(tron)

mm_init: re-enable zero page protection.

(rmind)

pmap_kenter_pa: always print about already present mapping.
pmap_bootstrap: add comments.

(rmind)

kern/45927: fix incorrect assert.

(nonaka)

Revert pmap_pte_flush() -> xpq_flush_queue() in previous.

(bouyer)

Avoid early use of xen_kpm_sync(); locks are not available at this time.
Don't call cpu_init() twice.

Makes LOCKDEBUG kernels boot again

(bouyer)

Mention pkg_add import.

(wiz)

Update for pkg_install import.

(wiz)

melt more snowflakes until they all look like pkgsrc

(wiz)

Import pkg_install-20120221 after merging changes from here, to simplify
next import.

Status:

Vendor Tag: PKGSRC
Release Tags: pkg_install-20120221b

(wiz)

melt snowflakes

(wiz)

Import pkg_install-20120221 for joerg.
Bug fix already there, this just bumps version.

Status:

Vendor Tag: PKGSRC
Release Tags: pkg_install-20120221

(wiz)

Switch to ANSI style declaration.

(rjs)

follow the sa removal for alpha

(para)

Make time_second and time_uptime volatile, so the compiler knows they
may change during loops. Fixes the macppc build, which previously
died with:
src/sys/arch/macppc/dev/dbdma.c:62:6: error: assuming signed overflow does not occur when assuming that (X + c) < X is always false

(martin)

use glyphcache for anti-aliased fonts

(macallan)

Now that window(1) is gone, use tmux(1) for xref.

(njoly)

Fix PR 46068 using the patch by Nicolas Thauvin:
pkg_add fails to install a package when subdirectories are missing
in the path to package directory in /var/db/pkg.

(wiz)

Pass -d option to pwd_mkdb(8) in order to make databases at proper location.

(nakayama)

Avoid rebuilding services database every boot time.
The database has been renamed since the default was changed to CDB format.

(nakayama)

Use actual physmem size for allocation of bootstrap extra PT pages
to avoid wasting pages on low memory machines.
Improves atari specific part of PR/45915, tested on TT030.

(tsutsui)

Make this include file C++ save, using the new __conly_restrict.

(martin)

There are some places where C99 allows a restrict modifier (and posix
demands it), but C++ does not allow it. So add a C only version of
__restrict, which expands to nothing if compiling C++.
See http://gcc.gnu.org/bugzilla/show_bug.cgi?id=49347 for a discussion.

(martin)

Add a definition for ".St -isoC-2011".  The equivalent definition
was already in external/bsd/mdocml/dist/st.in

(apb)

Mark /usr/include/netinet/ipf_rb.h as obsolete until ipf is upgraded again.
Sort while here.

(wiz)

Numerous changes to ssss(1) and libssss(3):

+ don't use the threshold headers any more - just use our own
header

+ "ssss" is the only algorithm supported by ssss(1)

+ add back (seamless) file I/O functionality, in the case that a
memory mapping, or subsequent memory allocation, fails

+ don't store information in the ssss share header which could be used
as part of an attack (specifically the threshold and total number of
shares).

+ size the extra memory required on a much better estimate of the size
needed, rather than a constant size

+ get rid of unneeded files, now that everything has been cleaned up

(agc)

keep track of the original array length so we can pass it to kmem_free, from
enami

(christos)

fix fae free'ing, from enami.

(christos)

Prevent stack buffer overflow when copying too-large-CDB into request.

(jakllsch)

constify command data argument to scsi_command().

(jakllsch)

Restore back to double alignment.

(matt)

Change ALIGNBYTES to be AltiVec savvy

(matt)

Don't rely on being able to allocate while in a softintr.  So preallocate
all the bus_dma maps in ifinit where we can sleep.

(matt)

Fix a very old bug.  When allocating the buffer and doing just a bs= transfer,
hen we only need a single buffer equal to that blocksize in length.

(matt)

Add and use xbdminphys() to handle transfer segmentation/size limit.
Should allow us to use a normal MAXPHYS in domU kernels.

(jakllsch)

Remove another ${SIZE}

(matt)

expand a command slightly

(mrg)

Add logic to the main reconstruction loop to handle RAID5 with rotated
spares.  While here, observe that we were actually doing one more
stripe than we thought we were, and correct that too (it didn't matter
for non-RAID5_RS, but it definitely does for RAID5_RS).  Add some
bounds-checking at the beginning to handle the case where the number
of stripes in the set is smaller than the sliding reconstruction window.

XXX: this problem likely needs to be fixed for PARITY_DECLUSTERING too.

(oster)

keep the thread lock taken for the abort() callback.

(mrg)

make sure we have the thread lock when calling usb_transfer_complete()

(mrg)

comment, and effectively remove, a DIAGNOSTIC check that
is invalid for RAID5_RS.

(oster)

pull across from -current:
>add a _kernel_locked_p().

(mrg)

add a _kernel_locked_p().

(mrg)

Ticket #15

(sborrill)

Pull up the following revisions(s) (requested by joerg in ticket #15):
usr.sbin/makemandb/apropos.1: revision 1.4
usr.sbin/makemandb/makemandb.8: revision 1.2
usr.sbin/makemandb/whatis.c: revision 1.3

Add reference to whatis(1).
SQLite doesn't use the FTS index for equal ops, so force it to do a FTS
search first. Drops run time by a factor of 6 for "whatis man".

(sborrill)

Ticket #14

(sborrill)

Pull up the following revisions(s) (requested by martin in ticket #14):
include/spawn.h: revision 1.2
sys/kern/kern_exec.c: revision 1.341
sys/uvm/uvm_glue.c: revision 1.157
tests/lib/libc/gen/posix_spawn/t_fileactions.c: revision 1.3

posix_spawn: fix kernel bug when passing empty fileactions (PR kern/46038)
and add a test case for this.  Fix potential race condition, doublefreeing
of memory and memory leaks in error cases.

(sborrill)

Ticket #13

(sborrill)

Pull up the following revisions(s) (requested by tron in ticket #13):
doc/3RDPARTY: patch
doc/CHANGES: patch
external/bsd/pkg_install/dist/add/perform.c: patch
external/bsd/pkg_install/dist/delete/pkg_delete.c: patch
external/bsd/pkg_install/dist/lib/license.c: patch
external/bsd/pkg_install/dist/lib/opattern.c: patch
external/bsd/pkg_install/dist/lib/pkg_summary.5: patch
external/bsd/pkg_install/dist/lib/version.h: patch

Update "pkg_install" to version 20120128:
- pkg_install 20120128:
  - Explicitly stat(2) if mkdir failed. errno detection doesn't work e.g.
    on Solaris.
  - Provide a stable order for package names that only differe in the base
    name, not the version number.
- pkg_install 20110805:
  - Fix for pkg_delete on NFS from Anthony Mallet.

(sborrill)

Ticket #12

(sborrill)

Pull up the following revisions(s) (requested by mrg in ticket #12):
sys/dev/usb/umidi.c: revision 1.60

Fix the reversed logic in several uses of the new 'closing' member
introduced in the previous change. From PR 45909.

(sborrill)

- Make pmap_write_protect() work with pmap_kernel() too ((va & L2_FRAME)
  strips the high bits of a LP64 address)
- use pmap_protect() in pmap_pdp_ctor() to remap the PDP read-only instead
  of (ab)using pmap_kenter_pa(). No more "mapping already present" on
  console with DIAGNOSTIC kernels
- make sure to zero the whole PDP (NTOPLEVEL_PDES doens't include
  high-level entries on i386 and i386PAE, reserved by Xen). Not sure
  how it has worked before
- remove an always-true test (&& pmap != pmap_kernel(); we KASSERT that
  at the function entry).

(bouyer)

pullup from -current:
>assert kernel lock is held in a few places in inside scsipi.
>lock the kernel when calling into scsipi from umass and usscanner.
>
>with these two in place on usbmp branch, umass appears stable.

(mrg)

assert kernel lock is held in a few places in inside scsipi.
lock the kernel when calling into scsipi from umass and usscanner.

with these two in place on usbmp branch, umass appears stable.

(mrg)

When using uvm_km_pgremove_intrsafe() make sure mappings are removed
before returning the pages to the free pool. Otherwise, under Xen,
a page which still has a writable mapping could be allocated for
a PDP by another CPU and the hypervisor would refuse it (this is
PR port-xen/45975).
For this, move the pmap_kremove() calls inside uvm_km_pgremove_intrsafe(),
and do pmap_kremove()/uvm_pagefree() in batch of (at most) 16 entries
(as suggested by Chuck Silvers on tech-kern@, see also
http://mail-index.netbsd.org/tech-kern/2012/02/17/msg012727.html and
followups).

(bouyer)

SQLite doesn't use the FTS index for equal ops, so force it to do a FTS
search first. Drops run time by a factor of 6 for "whatis man".
Found by Abhinav Upadhyay.

(joerg)

Add reference to whatis(1). From Abhinav Upadhyay.

(joerg)

Remove ${SIZE}

(matt)

Posix spawn fixes:
- split the file actions allocation and freeing into separate functions
- use pnbuf
- don't play games with pointers (partially freeing stuff etc), only check
  fa and sa and free as needed using the same code.
- use copyinstr properly
- KM_SLEEP allocation can't fail
- if path allocation failed midway, we would be possibily freeing
  userland strings.
- use sizeof(*var) instead sizeof(type)

(christos)

Add a test case to call posix_spawn with empty file actions, which reproduced
the (now fixed) PR kern/46038.

(martin)

Solve previous fix (for early posix_spawn children exiting on error)
differently.

(martin)

More posix_spawn fallout:
Fix a kmem_alloc() call with zero size (PR kern/46038), allow file actions
to be passed, even if empty.
Rearange p_reflock locking for the child, avoid a double free in an
error case, avoid a memory leak in another error case - all pointed out
by yamt.
During blocking operations early in the child borrow the kernel vmspace
(as suggested by yamt).

(martin)

Bump date for previous. Use more markup.

(wiz)

Add missing __BEGIN_DECLS, from Henning Petersen in PR misc/46058.

(martin)

Check in very basic compressed file support.  httpd will now serve
a precompressed .gz file if it exists, the client claims to support
gzip and the request is not ranged.

(elric)

Use a ``certificate chain file'' rather than a ``certificate file'' so
that bozohttpd can be used with non-toplevel certs.

(elric)

minor cosmetic changes to make this more readable - no functional change

(agc)

remove the intr_lock from the mp usb api, it wasn't used.

(mrg)

don't take the kernel lock when calling into the usb code anymore

(mrg)

timeout_handle callout is MPSAFE, mark it as such.

(mrg)

fix a panic() message to refer to the correct function

(mrg)

task thread and event threads are both MPSAFE now.

(mrg)

remove a bunch of superfluous parens.

(mrg)

convert usb event code to using a mutex and condvar.

(mrg)

convert splusb() calls to using a new usb_blk_lock.

(mrg)

adjust some comments to reality.  copy some XXX comments from ehci
into uhci/ohci since it's probably right there too

(mrg)

convert to device_t.
been running on my inspiron 8500 for a couple of months.

(mrg)

several changes to the MP usb apis, and other misc changes:

- usb_transfer_complete()/usb_insert_transfer()/usb_start_next() all
  must have the thread lock held

- (*soft_intr) now is called with the thread lock held unless we are
  in polling mode.  add a usb_soft_intr() to deal with this

- XXX usbd_set_polling() api exists to increase/decrease the polling
  count, but only ukbd uses.  everyone else open codes it, but this
  should probably be changed

- (*abort) is now called with the thread lock held

- update several comments to not refer to splusb() anymore

- add many more asserts

- use more c99 struct initialisers

(mrg)

itimerfire: fix a regression, check if timer is already queued.

(rmind)

- Add NPF_DECISION_BLOCK and NPF_DECISION_PASS.  Be more defensive in the
  packet handler.  Change the default policy to block when the config is
  loaded and set it to pass when flush operation is performed.
- Use kmem_zalloc(9) instead of kmem_alloc(9) in few places.
- npf_rproc_{create,release}: use kmem_intr_{alloc,free} as the destruction
  of rule procedure might happen in the interrupt handler (under a very rare
  condition, if config reload races with the handler).
- npf_session_establish: check whether layer 3 and 4 are cached.
- npfctl_build_group: do not make groups as passing rules.
- Remove some unecessary header inclusion.

(rmind)

Add support for --combine with gcc.

(matt)

Remove extern variables defined in headers.

(matt)

Remove <sa.h> which has been obsolete since 2007.

(rmind)

remove the spl handling from usb_insert_transfer() since we're
already holding the relevant lock (which is asserted.)

(mrg)

Sort option descriptions a bit more.

(wiz)

Remove COMPAT_SA / KERN_SA.  Welcome to 6.99.3!
Approved by core@.

(rmind)

pullup umidi.c 1.60:
>fix the reversed logic in several uses of the new 'closing' member
>introduced in the previous change.
>
>from Tom Ivar Helbekkmo <tih@hamartun.priv.no> in PR 45909.  thanks!

(mrg)

fix the reversed logic in several uses of the new 'closing' member
introduced in the previous change.

from Tom Ivar Helbekkmo <tih@hamartun.priv.no> in PR 45909.  thanks!

(mrg)

PR/45735: Bug Hunting: The dump(8) manpage, as well as the program's `usage'-line,
needs improvement

(christos)

PR/46049: Tim van der Molen:
clrtobot() and clrtoeol() do not set background attributes

(christos)

Emit SA syscalls using SYS_compat_60_*

(matt)

Postfix 2.8.8 will be part of the NetBSD 6.0 release.

(tron)

Tickets 7,9,10,11

(riz)

Regen for ticket #11.

(riz)

Pull up following revision(s) (requested by mbalmer in ticket #11):
sys/dev/usb/if_udav.c: revision 1.35
sys/dev/usb/usbdevs: revision 1.608
Add new vendor QUAN and it's DM9601 USB ethernet product.
From Loganaden Velvindron <loganaden@wolfman.devio.us>.
Add support for the QUAN DM9601.
From Loganaden Velvindron <loganaden@wolfman.devio.us>.

(riz)

Pull up following revision(s) (requested by apb in ticket #10):
usr.sbin/postinstall/postinstall: revision 1.131
Use msg for indentation.

(riz)

Pull up following revision(s) (requested by apb in ticket #9):
usr.sbin/postinstall/postinstall: revision 1.130
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.

(riz)

Apply patch (requested by tron in ticket #7):

Import Postfix 2.8.8. Changes since Postfix 2.8.7:
- The Postfix sqlite client, introduced with Postfix 2.8, had an
  embarassing bug in its quoting routine. As the result of a
  last-minute code cleanup before release, this routine returned the
  unquoted text instead of the quoted text. The opportunities for
  mis-use are limited: Postfix sqlite database files are usually owned
  by root, and Postfix daemons usually run with non-root privileges so
  they can't corrupt the database. This problem was reported by Rob
  McGee (rob0).
- The Postfix 2.8.4 fix for local delivery agent database lookup
  errors was incomplete. The fix correctly added new code to detect
  database lookup errors with mailbox_transport_maps,
  mailbox_command_maps or fallback_transport_maps, but it failed to
  log the problem, and to produce a defer logfile record which is
  needed for "delayed mail" and "mail too old" delivery status
  notifications.
- The trace(8) service, used for DSN SUCCESS notifications, did not
  distinguish between notifications for a non-bounce or a bounce
  message, causing it to "reply" to mail with the null sender
  address. Problem reported by Sabahattin Gucukoglu.
- Support for Dovecot auth over TCP sockets, using code that already
  existed for testing purposes. Patrick Koetter kindly provided an
  update for the SASL_README file.
- Workaround in the LDAP client for changes in the under-documented
  OpenLDAP API, by Victor Duchovni.

I've fully tested this under NetBSD/i386 6.99.2 and build tested the
patch for NetBSD/amd64 6.0_BETA.

Kind regards

P.S. This patch does *not* include SQLite support for Postfix to
    avoid scaring Jeff Rizzo.

(riz)

"pkg_install" was updated to version 20120128.

(tron)

Regen.

(matt)

Use void * instead of sa_upcall_t for sa_register (since sa_upcall_t is
going away).

(matt)

Resolve conflicts from last import.

(tron)

Update "pkg_install" to version 20120128:
- pkg_install 20120128:
  - Explicitly stat(2) if mkdir failed. errno detection doesn't work e.g.
    on Solaris.
  - Provide a stable order for package names that only differe in the base
    name, not the version number.
- pkg_install 20110805:
  - Fix for pkg_delete on NFS from Anthony Mallet.

Status:

Vendor Tag: PKGSRC
Release Tags: pkg_install-20120128

(tron)

Add compat_60 stub for the SA syscall (all which just call sys_nosys).

(matt)

Regen.

(matt)

Make SA syscalls as COMPAT_60

(matt)

Make SA calls obsolete (use stubs in kern case, as libc needs them for now).

(rmind)

Add compat_60

(matt)

Use the release date as suggested.
XXX: I bet this is going to be broken in the next import because nobody
will remember to fix it.

(christos)

Add support for the QUAN DM9601.
From Loganaden Velvindron <loganaden@wolfman.devio.us>.

(mbalmer)

Re-gen for QUAN DM9601.

(mbalmer)

Add new vendor QUAN and it's DM9601 USB ethernet product.
From Loganaden Velvindron <loganaden@wolfman.devio.us>.

(mbalmer)

Fix boot failure on OF 2.0 machines.

Discussed with phx@ on port-macppc@:
http://mail-index.NetBSD.org/port-macppc/2012/02/18/msg001556.html

Should be pulled up to netbsd-6.

(tsutsui)

Removing remaining references to the alternate PTE space. Modify documentation appropriately

(cherry)

Adapt to constification in sys/uvm/uvm_export.h

(martin)

Expose the contents also for _KMEMUSER, fixes build of usr.sbin/crash.

(he)

Fix for MKREPRO = yes
XXX: pullup to 6

(christos)

Fix MKREPRO = yes
XXX: pullup to 6

(christos)

G/C VM_MAP_INTRSAFE flag

(rmind)

Replace simple_lock(9).

(rmind)

Remove VM_MAP_INTRSAFE and related code.  Not used since the "kmem changes".

(rmind)

Replace few simple_lock(9) cases.

(rmind)

rt_setkey: remove invalid assert, sockaddr_dup() may fail if no memory.

(rmind)

moved to external/bsd

(christos)

fix prototypes

(christos)

Tickets 4,5

(riz)

Pull up following revision(s) (requested by joerg in ticket #5):
share/mk/bsd.own.mk: revision 1.697
share/mk/bsd.README: revision 1.293
Derive MANINSTALL from MKCATPAGES and MKHTML.

(riz)

Pull up following revision(s) (requested by joerg in ticket #4):
usr.sbin/makemandb/apropos.c: revision 1.5
external/bsd/mdocml/dist/read.c: revision 1.6
external/bsd/mdocml/dist/mandoc.h: revision 1.2
usr.sbin/makemandb/makemandb.c: revision 1.3
usr.sbin/makemandb/makemandb.c: revision 1.4
usr.sbin/makemandb/makemandb.c: revision 1.5
usr.sbin/makemandb/Makefile: revision 1.2
etc/man.conf: revision 1.30
usr.sbin/makemandb/apropos.1: revision 1.3
external/bsd/mdocml/Makefile.inc: revision 1.15
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.
Also handle hyphen replacement if it was used as plain input and no
backslash sequence was used at all in the line.
Add support for apropos -s like in the old apropos. Fix capitalisation.
Add note about optional quotation. From Abhinav Upadhyay.
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.
Add support for compressed man pages in all the usual formats.
Use mmap when dealing with plain files.
Add entries for bzip2 and xz.

(riz)

m68k does not have or need db_machdep.c

(christos)

Adjust libyywrap.c location.

(njoly)

Use msg for indentation.

(njoly)

expose more for kmemuser/fix include protection

(christos)

expose more for kmemuser; fix header protection.

(christos)

expose more for kmemuser

(christos)

this has not been used for ages.

(christos)

revert sparc64 hack

(christos)

Move the ddb registers and empty command table into a separate file, like
other archs do for the benefit of crash.

(christos)

Tweak MAKEDEVSPEC target to properly report MAKEDEV failures:
- No need to ignore `rm -f' failures which should not happen.
- Do not pipe MAKEDEV output, but use a temp file.

(njoly)