Do not call zfs_init and zfs_fini directly they will be called by vfs routines.

(haad)

Do not initialize space map cv twice.

(haad)

Implemented support for boot arguments.

(phx)

Updated Synology extclk from latest Linux/Synology source.

(phx)

microtime(9) stores struct timeval.  bintime(9) stores struct bintime.

(tsutsui)

* document VOP_DELETEEXTATTR
* remove "vector element" from summary table:
  1) deleteextattr caused problems for 80col rendering
  2) it's not correct on NetBSD, we use the VOFFSET instead (!)

(pooka)

rename the allow and deny files to match the documentation.

(christos)

remove three years out-of-date comment

(pooka)

Continue the discussion w.r.t. SIGEV_THREAD by nothing that pthread_join(3)
should be out of the question and that thread stack cannot be recovered.

(jruoho)

Sort sections.

(wiz)

Refer to sigevent(3).

(jruoho)

Remove the paragraph about the used notification mechanism. Instead,
reference sigevent(3). Clarify the file offset discussion.

(jruoho)

Add sigevent(3), a manual page that describes the struct sigevent used by
the POSIX real time functions and timers.

XXX: (a) Needs to be modified once real time signals are supported.
(b) Could use an example.

(jruoho)

One more reason not to use DSA keys:

The DSA algorithm seems to require a digest value which is 20 bytes
long, which kind of implies SHA-1.

If we have a DSA signature, use SHA-1 as a hash algorithm, for backwards
compatibility. RSA signatures continue to use SHA256 by default, although
this can be given as an argument, if desired.

This fixes DSA signatures with netpgp:

% netpgp --sign --userid d4a643c5 a
pub 1024/DSA 8222c3ecd4a643c5 2010-05-19 [EXPIRES 2013-05-18]
Key fingerprint: 3e4a 5df4 033b 2333 219b 1afd 8222 c3ec d4a6 43c5
uid              Alistair Crooks (DSA TEST KEY - DO NOT USE) <agc@netbsd.org>
sub 1024/DSA 8222c3ecd4a643c5 2010-05-19 [EXPIRES 2013-05-18]
netpgp passphrase:
% netpgp --verify a.gpg
Good signature for a.gpg made Tue May 18 05:41:25 2010
using DSA key 8222c3ecd4a643c5
pub 1024/DSA 8222c3ecd4a643c5 2010-05-19 [EXPIRES 2013-05-18]
Key fingerprint: 3e4a 5df4 033b 2333 219b 1afd 8222 c3ec d4a6 43c5
uid              Alistair Crooks (DSA TEST KEY - DO NOT USE) <agc@netbsd.org>
sub 1024/DSA 8222c3ecd4a643c5 2010-05-19 [EXPIRES 2013-05-18]
%

(agc)

Add gpt(8) and sort SEE ALSO (case insensitively). Bump date.

(wiz)

Mention that dkctl(8) does not modify on-disk data, and add xrefs to
fdisk(8) and gpt(8). Bump date.

(wiz)

- The new cron keeps time internally as localtime. Revert to previous behavior
  of keeping UTC, and handle timezones the way we handled them before (CRON_TZ).
  The old behavior can be selected with defining CRON_LOCALTIME.
- Provide functions to pretty-print crontab entries.

(christos)

Forgot to commit this:
Bring back the default to allow any user to install crontabs.

(christos)

Add missing include

(martin)

configure com1 and make it attach correctly
note that config loc 'mult' is assumed to be 4, and so no need to specify

(cliff)

Split "file system" and fix a typo.

(wiz)

Slight wording improvement.

(wiz)

Whoops, default to MP locking.

(pooka)

Add uniprocessor versions of mutex/rw/cv.  They work only on virtual
unicpu configurations (i.e. RUMP_NCPU==1), but are massively faster
than the multiprocessor versions since the fast path does not have
to perform any cache coherent operations.  _Applications_ with
lock-happy kernel paths, i.e. _not_ lock microbenchmarks, measure
up to tens of percents speedup on my Core2 Duo.  Every globally
atomic state required by normal locks/atomic ops implies a hideous
speed penalty even for the fast path.

While this requires a unicpu configuration, it should be noted that
we are talking about a virtual unicpu configuration.  The host can
have as many processors as it desires, and the speed benefit of
virtual unicpu is still there.  It's pretty obvious that in terms
of scalability simple workload partitioning and replication into
multiple kernels wins hands down over complicated locking or
locklessing algorithms which depend on globally atomic state.

(pooka)

Namespace rump-only kernel biglock routines appropriately.

No functional change.

(pooka)

Move routines related to kernel locking and scheduling from
locks.c to klock.c.

No functional change.

(pooka)

Add support for DIOCCACHESYNC ioctl for dm devices. Add new sync function
pointer to dm_target_t because that is the only part of dm which know real
block device. disk_ioctl_switch parses whole device table and for every
entry it calls particular sync routine which propagates DIOCCACHESYNC
to real disk.

While I was here implement some KNF fixes and remove unneeded symbols from
dm.h.

Problem reported on port-xen@ by Hugo Silva.

(haad)

Recognize a network boot device by its PCI tag, even when it is not directly
attached to a PCI bus, but to a controller chip (e.g. sk(4) on skc(4)).
Extended BTINFO_NET to usa a PCI tag as well, now since it works.

(phx)

Fix typo in comment.

(haad)

Make it possible to use the scheduler lock as the rumpuser condvar
interlock.  This is applicable in cases where the actual interlock
is the CPU the currently running thread is scheduled on.  Borrowing
the scheduler lock as the mutex mandated by pthread_cond_wait()
does away with need to have an additional mutex.  This both optimizes
runtime execution and simplifies code, as the extra lock typically
lead to quite some trickeries to avoid the dungeon collapsing due
to zaps from the wand of deadlock.

(pooka)

Set Synology extclk from 33.333333MHz to 33.164526MHz, which reduces the
error in the mftb timer from 5% to 0.002%.
The console port for QNAP and STORCENTER is 0x4500, not 0x4600.

(phx)

Macros, not functions.

(jruoho)

Synology PCI fixup, submitted by Toru Nishimura.

(phx)

Fix .Nd.

(jruoho)

Describe also the timespec() family of macros in timeradd(3).

(jruoho)

Fix a typo.

(jruoho)

Refer to timeval(3) and only describe the legacy struct timezone here.

Improve markup.

(jruoho)

Add an introductory example.

(jruoho)

Xref timespec(3).

(jruoho)

Note timeval(9). Add %D to the .Rs/.Re.

(jruoho)

Note that struct bintime is described in timecounter(9). Improve wording.

(jruoho)

move the mutex_init() for ci_ctx_lock inside cpu_attach() so that
we don't try to mutex_init() before lockdebug is available.

fixes LOCKDEBUG crashes at startup (for USII systems.)

(mrg)

Use wput_utf8() to improve conversion of UTF-16 GPT partition names to UTF-8.
Drop static CRC32 function in favor of the one from libkern.

(jakllsch)

Excluded KUROBOX from GENERIC.NAS because console is on a different port.
Add INET6 option to GENERIC(.NAS).
Build KUROBOX kernel together with GENERIC and GENERIC.NAS.
Patch submitted by Toru Nishimura.

(phx)

Due previous, remove STANDARDS.

(jruoho)

Briefly describe and note also struct bintime.

(jruoho)

Xref timeval(3).

(jruoho)

Reference only aio(3) in SEE ALSO as it is the central placeholder. Note
timespec(3). Markup: -width Er (not -width Dv), .Dv NULL (not null), etc.

(jruoho)

Fix a typo of a typo, and use .Sq in couple of places.

(jruoho)

Writing, rereading, writing, reareading, and yet still one typo.

(jruoho)

CPPFLAGS cleanup, submited by Toru Nishimura.

(phx)

Set correct extclk for KuroBox, which is not 33MHz.
Patch submitted by Toru Nishimura.

(phx)

Add an introductory manual page for the POSIX asynchronous I/O, aio(3).

This is hopefully enough for a reader to get started with the aio.

ok rmind@

(jruoho)

put back canonicalization, but avoid doing it for files that don't contain /
From dholland

(christos)

Offset indent, .Dv.

(jruoho)

Regen for multiple inclusion protection.

(njoly)

Add multiple inclusion protection to rump_syscalls.h.

(njoly)

Remove trailing comma.

(jruoho)

Use -enum in the list for clarity.

(jruoho)

Add subtitle. Break long paragraph.

(jruoho)

Note that all stat-functions are standardized nowadays.

(jruoho)

Simplify primary CPU startup slightly by using the fact that our cpuinfo's
curpcb pointer points to the uarea/pcb of lwp0 and by remembering the pointer
to this cpuinfo, thus avoiding a few CPUINFO_VA references.

(martin)

Attach CPU only of the pretense.

(kiyohara)

Indent.
s/mainbus0/mainbus?/.

(kiyohara)

Handle the HAVE_NBTOOL_CONFIG_H case.

Fixes hp700 build with objdir.

(ahoka)

Fix .Dt.

(jruoho)

Reference timeval(3).

(jruoho)

Add a manual page that describes the basic structures in <sys/time.h>.

(jruoho)

Try to describe the struct stat instead of copy-pasting it from the header.
Use xrefs to avoid duplication of effort. Note the S_IS*() macros.

(jruoho)

Actually the previous is wrong; all of these can set the errno either via
each other or via something like malloc(3) or lseek(2).

(jruoho)

Fix the note about errno in ERRORS; telldir(3), seekdir(3), and rewinddir(3)
do not set the errno.

(jruoho)

Fix the SYNOPSIS.

(jruoho)

As was done with the man-page for <sys/types.h>, move dir(5) to dirent(3).
Add a table describing the stat <-> dirent types. Note IFTODT() and DTTOIF().

(jruoho)

.Dv.

(jruoho)

Add the Butenhof's book to SEE ALSO. (It was decent enough when I read it
years ago, but if there are better ones, please feel free to add those.)

(jruoho)

Enumerate the functions in a table instead of a big SEE ALSO.
Also some stylistic changes.

(jruoho)

Sort sections. Fix typo. Fix xref. Improve wording in one case.

(wiz)

Make netboot generate a BTINFO_NET bootinfo node for the Synology sk(4) NIC.
It will pass the MAC address, which is read from Flash ROM, into the kernel.
The kernel creates a "mac-address" device-property, which is used by sk(4),
when given, before reading the MAC from its EEPROM.

(phx)

From Matt Thomas, suggested upstream:

* atof-generic.c (atof_generic): recognize snan and qnan in
addition to nan and inf.
* atof-vax.c (flonum_gen2vax): deal properly with nan, snan, qnan,
and +Inf and -Inf codes.

This fixes PR toolchain/43303.

(martin)

Add a note that the "whiteout inode", DT_WHT, is an implementation detail
and should not be seen in normal user applications.

(jruoho)

At least one markup-typo slipped through the cracks.

(jruoho)

Rewrite much of the page.

Provide much more detailed and structured discussion. Notably, try to gather
some of the conditions around the famous undefined behavior; these functions
(or the implementations) are quite prone to such behavior.

Add compatibility notes. (The big change here is that in SUSv4 it is no
longer mandated that the directory streams are implemented by using file
descriptors.)

XXX: Please proofread this. Consider also adding some additional notes. For
    instance, I left the semantics about "." and ".." out.  The returned
    errno's should be enumerated as well.

(jruoho)

and, erm, bump date too.

(dholland)

Grammar, clarification, and note that the rusage information is total
as opposed to strict subprocess only.

(dholland)

When setting up res (results), errs (error) and outs (outputs) streams,
change the pre-defined stdio streams to be denoted by "<stdout>" and
"<stderr>", to distinguish them from file names.

In netpgpkeys(1), send the default "res" (results) stream to stdout,
rather than stderr. Requested by Anon Ymous (and makes perfect sense).

(agc)

Avoid a possible overwrite of a value in the (key, value) array, in
the event that a reference to the value is passed to the
netpgp_setvar() function as the new value.  Problem noted, cause
detected, and most of the fix contributed by, Anon Ymous.  Thanks!

(agc)

IMPLEMENTATION NOTES back to NOTES.

(jruoho)

Use .An as documented in mdoc.samples(7).

(jruoho)

Use device_xname and <sys/intr.h>

(matt)

If IPL_SCHED != IPL_HIGH we need to raise ourselves back to IPL_HIGH before
we return from a fast softint that was switchedaway from.

(matt)

fix a piece of lint

(agc)

Add IPL_DDB

(matt)

Add IPL_DDB.  This is needed for watchdog on sbmips and for IPIs used by DDB.
It's above IPL_SCHED but below IPL_HIGH.

(matt)

Make sure we have a spare cell at the sr_map to make splintr will stop.

(matt)

Follow the Fundamental Theory of Algebra. Disallow factorising of
numbers less than 2 as it is not
- naturally unique (negative numbers)
- finite (0)
- non-empty (1)

Discussed with the kristaps and wiz

(joerg)

Add register information for the EHCI 1.0 Appendix C Debug Port.

(jakllsch)

Add Debug Descriptor structures and constants,
as well as symbolic identifiers for USB packet IDs.

(jakllsch)

Set RI_NO_AUTO in ri_flg if init function is invoked from cnattach.
Untested. (3100 support seems somewhat broken, per gxemul output)

(tsutsui)

Add kernel support for MIPS TLS.  Use rdhwr rt, $29 as defined by the MIPS
TLS spec so that Linux MIPS binaries will work.  Use sysarch(MIPS_TINFOSET, v)
to set the pointer.

(matt)

nfs_statvfs should return NFS_MAXNAMLEN, not MAXNAMLEN.
(Compile-tested only, but that should be ok)

(dholland)

make it compile when DIAGNOSTIC is defined

(cliff)

Regen.

(jakllsch)

Add NetChip (now PLX) vendor ID, as well as NET228[02] device IDs.

(jakllsch)

Set RI_NO_AUTO at correct place. Confirmed on EWS4800/360AD.

(tsutsui)

Complete the information in the reference (authors, publisher, date, URL).

(jruoho)

Minor markup improvements.

(jruoho)

Move the more detailed discussion around the dirent-structure from
getdents(2) to dirent(5), and extend it with some compatibility notes.

(jruoho)

Added keyboard config files for Sharp WS003SH/WS007SH.

(nonaka)

Added keyboard config files for Sharp WS003SH/WS007SH.

(nonaka)

Remove repeated text "After filling..."

XXX Someone else can decide whether we should refer to "U.S.A." or to
XXX "the United States" in the parenthetical.

(pgoyette)

Add definitions for Power Manager Registers.
Add some fields for USBHC_UHCRHDB.

(kiyohara)

Add mutex.h.

(kiyohara)

Set RI_NO_AUTO in ri_flg if init functions are invoked from cnattach.
Tested on NWS-3470D. xafb.c for NWS-5000 is untested.

XXX: isn't it better to have failsafe default than sprinkle RI_NO_AUTO
XXX: to avoid fatal silent hangup?

(tsutsui)

Add some missing atf keywords to fix MKATF=no build.

(njoly)

Clear RI_NO_AUTO in actual attach functions during autoconf(9)
for console devices that set the flag during consinit().

(tsutsui)

Set RI_NO_AUTO in ri_flg if fb_init() is invoked from cnattach.
Tested on Explora451.

(tsutsui)

Set RI_NO_AUTO in ri_flg if *fb_common_init() is invoked from cnattach.
Untested.

(tsutsui)

merge bozohttpd 20100512

(mrg)

Set RI_NO_AUTO in ri_flg if pvr_getdevconfig() is invoked from cninit.
Tested on gxemul.

(tsutsui)

import bozohttpd 20100512.  it includes these changes:
        o  fix SSL mode.  from rtr.

Status:

Vendor Tag: bozohttpd
Release Tags: bozohttpd-20100512

(mrg)

Set RI_NO_AUTO in cnattach function. Untested.

(tsutsui)

Set RI_NO_AUTO in rip->ri_flg if tga_init() is invoked from cnattach.

(tsutsui)

TERM=x68k -> vt220.  sysinst works again.

(isaki)

Make the sibyte watchdog establish an interrupt, and if taken, drop into DDB.
Make sure these interrupt are not blocked by IPL_HIGH.

(matt)

Set RI_NO_AUTO in ri->ri_flg if hpcfb_init() is invoked from cnattach.
Fixes silent hang on hpcsh and hpcmips.

XXX: all other rasops based console devices need to set this flag??

(tsutsui)

Enable dhcpcd(8) in hpcsh sysinst.

(tsutsui)

Backout rev.1.137.  It causes troubles, see PR kern/43294.
We needs more discussion/a more general solution.

(oki)

tune ADC Averaging Control parameter.

(nonaka)

Reorder sections to canonical order.

(joerg)

- increase SYMTAB_SPACE to 380000
- configure ohci at pci, and some useful USB devices

(cliff)

Do the sync before the badaddr in pci_conf_read so make no other PCI
transactions are pending before we probe.

(matt)

Cleanup duplicated groff keyword.

(njoly)

Make badaddr64 work on O32.  Add a few syncs to force errors.

(matt)

Proper markup.

(joerg)

New sentence, new line.

(joerg)

\\ -> \e

(joerg)

Reorder sections to canonical order.

(joerg)

Resort sections to canonical order

(joerg)

Use standard section headers.

(jruoho)

Reorder sections to canonical order

(joerg)

Reorder sections to canonical order.

(joerg)

\\ -> \e

(joerg)

gcscpcib depends on functions from x86/pci/pcib.c

(phx)

Don't break lines in .It, there be dragons in groff.

(joerg)

update DPRINTF to reflect actual function name

(plunky)

Explicitly close quote.

(joerg)

Resort sections to canonical order.

(joerg)

Resort sections to canonical order.

(joerg)

RETURN VALUES -> EXIT STATUS

(joerg)

Simplify markup.

(joerg)

Escape punctuation. Use logical markup for indentation.

(joerg)

Resort sections to canonical order

(joerg)

Improve/fix/deuglify markup.

(joerg)

Reorder sections to canonical order. Use .Ex -std

(joerg)

Reorder sections to canonical order.

(joerg)

Reorder sections to canonical order.

(joerg)

Reorder sections to canonical order. Use .Ex -std.

(joerg)

Reorder sections to canonical order.

(joerg)

RETURN VALUES -> EXIT STATUS.

(jruoho)

Fix section order. Use .Ex -std

(joerg)

Reorder section. Use .Ex -std.

(joerg)

Reorder to canonical order.

(joerg)

Missing -width for .Bl -tag

(joerg)

Reorder sections to canonical order.

(joerg)

Reorder sections.

(joerg)

Programs exit, they don't return values.

(joerg)

Explicitly quote |. Don't use .%T outside the scope of .Rs/.Re.

(joerg)

Briefly note /etc/rc.local in rc(8).

(jruoho)

Use standard section headers.

(jruoho)

make things compile.

(christos)

Pick up after people who find build-testing their changes too difficult.

(pooka)

May be easy to understand 'pay_len > 0' more than 'paylaod != NULL'.

(kiyohara)

Use kmem(9) instead of malloc(9).

(kiyohara)

Add missing minus to fix build of HTML page.

(wiz)

Convert to mdoc. Reorder content to sort flags and sections.

(wiz)

Clarify the wording in the table.

(jruoho)

Fix spaco.

(wiz)

Move PMAP_KMPAGE to be used in pmap_kenter_pa flags argument.
'Looks good to me' gimpy@

Forgot to commit this piece in previous commit

(cegger)

fix previous again, hopefully ok this time

(dholland)

Use a table instead of a literal .Bd "list".

(jruoho)

Use a list for the list of bugs.

(jruoho)

Move PMAP_KMPAGE to be used in pmap_kenter_pa flags argument.
'Looks good to me' gimpy@

(cegger)

Provide a sane argument for -width.

(joerg)

Fix trailing whitespace

(joerg)