mention panic uses.

(christos)

fix dev_t debugging printf lossage.

(christos)

We need to install libamu.

(christos)

regen for christos_time_t merge

(rtr)

Replace time_t values in libsa sources with the following two types
to avoid unnecessary 64 bit ops which would make binaries larger:

satime_t (currently unsigned int):
numbers in seconds returned by the machine dependent getsecs() function
which are used to measure relative time

saseconds_t (currently int):
numbers in seconds used to specify timeout to network drivers

Per discussion on current-users.

(tsutsui)

Uset uint64_t rather than (signed) long to read the PCC register.

(tsutsui)

Fix getsecs() to return more proper values:
- use not only second but also minute and hour registers
- use bcdtobin() to read mkclock registers

(tsutsui)

more shared library major number bumps.

(dogcow)

Move declarations into a common header.

(tsutsui)

Fix blatantly wrong exposition of .WAIT example.
PR bin/40372 from Gao Ya'nan.

(dholland)

use PRI macro formats to printf type dev_t

(cegger)

Return ENOENT instead of panicking when irq doesn't equal line
(mpacpi_findintr_linkdev: irq mismatch). This doesn't fix the cause of
kern/38540, but stops the bogus panic. It's pretty definite that the device
with the mismatched irq will not function.
----------------------------------------------------------------------

(sborrill)

use PRIu64 format to printf DISKPART

(cegger)

use PRIu64 format to printf major() value

(cegger)

cast to char to printf DISKPART
XXX Shouldn't sun3 use the MI drivers instead?

(cegger)

use PRIu64 format to printf major() and minor() values

(cegger)

use PRIu64 format to printf major() and minor() values

(cegger)

cast to char to printf DISKPART

(cegger)

use PRIu64 format to printf major() and minor() values

(cegger)

Enable and fix prototype warnings.

(tsutsui)

use PRIx64 format to printf type dev_t

(cegger)

- WARNSfy
- make sure machine and ${MACHIN_ARCH} symlinks are always created
- several tweaks to allow build without installed DESTDIR

(tsutsui)

Warning options should be in CFLAGS, not CPPFLAGS.

(tsutsui)

WARNSfy

(tsutsui)

- WARNSfy
- make sure machine and ${MACHINE_ARCH} symlinks are always created
- add incpath to AFLAGS too for build without DESTDIR

(tsutsui)

- WARNSfy
- add src/common/dist/zlib to incpath for build without DESTDIR

(tsutsui)

- WARNSfy
- make sure machine and ${MACHIN_ARCH} symlinks are always created
- add src/common/dist/zlib to INCPATH and remove unnecessary options
- remove -I${DESTDIR}/usr/include from INCPATH

(tsutsui)

Use WARNS=2 rather than WARNS=4 since it's too strict for MI libsa yet.

(tsutsui)

WARNSfy

(tsutsui)

WARNSfy

(tsutsui)

- add some make variables to suppress errors on build without DESTDIR
- make sure machine and ${MACHINE_ARCH} symlinks are always created

(tsutsui)

Make sure machine and ${MACHINE_ARCH} symlinks are always created.

(tsutsui)

Add some make variables to suppress errors on build without DESTDIR.

(tsutsui)

Create machine and ${MAHCINE_ARCH} symlinks in ${.OBJDIR}
to allow build without DESTDIR.

(tsutsui)

- add prototype warning options
- catch up changes for common powerpc/stand/mkbootimage

XXX: daily build for rs6000 should be enabled on releng server?

(tsutsui)

- add prototype warning options
- add some make variables to suppress errors on build without DESTDIR

(tsutsui)

WARNSfy

(tsutsui)

Use <machine/stdarg.h> rather than <stdarg.h> to allow build without DESTDIR.

(tsutsui)

WARNSfy

(tsutsui)

WARNSfy.

(tsutsui)

Add parantheses to macro to suppress possible compiler warnings.

(tsutsui)

Enable prototype warning options.

(tsutsui)

Enable prototype warning options.

(tsutsui)

Retain compatibility with old firm_event.  Old Xamiga binary now works with
-current kernel.

(mhitch)

Use userspace version unless _HARDKERNEL.  Otherwise we use atomic
cas to implement spinlocks and spinlocks to implement atomic cas....
which might suck.

Since the userspace version uses ll/sc, which doesn't exist on
R2000/R3000, rump will not work on those platforms.  *snif* (well,
pthread in general AFAICT).

(pooka)

include sys/param.h for _HARDKERNEL instead of homegrown def.

(pooka)

.. and finally, we need __ASSEMBLER__ instead of _LOCORE, since we
might want this header from userspace assembly files as well.

(pooka)

protect kernel inline with _KERNEL

(christos)

* Add adapted version of the userspace atomic_cas ops for platforms
  lacking special instructions.  We always use the spinlocked version
  (could use RAS on UP NetBSD host, but it's not portable).
* Add platform-based symbol quirk tables for selectively not renaming
  toolchain symbols for some platforms.  Although, this should really
  depend on the (toolchain,platform)-tuple and not just the platform.

This allows arm, hppa, mips and sh3 to build succesfully with an
isolated rump kernel namespace.  ... now, one arch remains: vax.
you must compile vax.  then, only then will you MI be.

(pooka)

Shove in some hppa love to help with spcopy.S (why does that need
to be in libkern?).

(pooka)

s/set/get/

(christos)

retain compatibility with old firm_event. Perhaps we should bump the protocol
version with an ioctl?

(christos)

use PRI format macros to printf major and minor values

(cegger)

adaption for making kgdb_dev a dev_t

(cegger)

got a couple wrong

(jnemeth)

Compat a.out stat structures also need to use u_int32_t dev_t.

(mhitch)

fix dev_t printf formats.

(mhitch)

bump versions for time_t

(christos)

fix dev_t printf formats.

(christos)

... except we mostly need _HARD/SOFTKERNEL in assembly and we can't
include param.h from asm as such.  So wrap in #ifndef logore.

(pooka)

New sentence, new line.

(wiz)

regen.

(wiz)

Make ready for HTML output, add comma.

(wiz)

Mark up arguments with Ar and drop trailing whitespace.

(wiz)

Implement DIOCCACHESYNC for RAIDframe too.

(oster)

Add a space.

(wiz)

New sentence, new line.

(wiz)

Deal with new dev_t build lossage.  Fix from lukem.

(oster)

Currently _KERNEL can mean either a) use kernel C namespace or b)
code runs in priviledged mode.  Introduce _HARDKERNEL and _SOFTKERNEL
to be able to say if the code runs in priviledged mode.

Currently used by rump, but Xen should probably set them too.

(pooka)

fix lutimes.

(christos)

regen: mount -o rump

(pooka)

Note mount -o rump support.

(pooka)

Compat a.out stat structures should use compat struct timespec50.
Noticed by mhitch@.

(tsutsui)

Decode the fpu status bits into siginfo code values when signaling fpu
traps. This makes us pass all of our ieee regression tests.

(martin)

Use PRIu64 format to printf major() and minor() values.

(martin)

typo

(sketch)

remove !NTP stub.

(christos)

document -o rump

(pooka)

*utimes should allow NULL tvp.

(christos)

Protect against nfssilly in RUMP_THREADS=0

(pooka)

Now that COMPAT_50 is needed inside the kernel for the time_t changes,
piggyback it and guard the copying of syscall return values on it. Our
libc has been fixed quite some time ago, time to get rid of the bogus copy.

(martin)

Support mount -o rump, which uses the rump server instead of using
the kernel service.  E.g. "mount -t efs -o rump /dev/sgidev /mnt"
runs effectively "rump_efs /dev/sgidev /mnt".

(pooka)

Default to threading, non-threaded debug mode can be compiled separately.
Useful if running this from fstab.

(pooka)

-o resport, not -o resvport

(pooka)

Fix noresport mount option parsing.

(pooka)

regen.

(rumble)

Fix time_t change fallout.

(rumble)

Make this compile.

(rumble)

regen.

(christos)

fix lutimes.

(christos)

don't do size computations because g++ does not like it.

(christos)

fix format instead of casting wildly

(macallan)

mention 32 bit time_t remnants

(christos)

Add casts to int to rdunit() and rdpart() macro to suppress a warning
since they are unlikely 64 bits.

XXX: actually these should be DISKUNIT() and DISKPART() in <sys/disklabel.h>.

(tsutsui)

Add a note about make(1) .for loop processing.

(dsl)

Oops, allow 2008 in dates as well.

(wiz)

don't truncate the print.

(christos)

provide more details.

(christos)

deal with 64bit time_t
Noticed by Wojciech Galazka

(macallan)

__rump_unavailable -> rump__unavailable

(pooka)

It's "show uvmhist" not "uvmhist".

(skrll)

- fix leaked lock, thanks ad@ for noticing.
- remove unneeded cast.

(christos)

fix COMPAT dependency.

(christos)

Instead of stashing the body of every iteration of a .for loop, pass
Parse_SetInput() the name of a function to get the loop data blocks.
This should only be a change to the program logic.

(dsl)

- insert a space after the recognized string if it was an exact match
- initialize properly the string used for completion.
From Alex Bligh alex at alex dot org dot uk
- Make char constants consistent

(christos)

fix christos-time_t info

(christos)

Avoid crash under LOCKDEBUG by destroying a mutex correctly.
Also: move mutex init/destroy into attach/detach functions and
handle case where line discipline is closed before the irframe
device. Fixes PR kern/40271.

(mlelstv)

Register with pmf.

(mlelstv)

shared library bumps; undoubtably there will be many more

(jnemeth)

Regen for following christos-time_t merge.

(nakayama)

Follow christos-time_t merge.

(nakayama)

Make it compile.

(dogcow)

Use PRIu64 to print 64 bit dev_t.

(tsutsui)

Regen from syscalls.master rev 1.64:
> Update for compat_50 stuff. (compile test only)

(tsutsui)

Update for compat_50 stuff. (compile test only)

(tsutsui)

Treat all of ^__ as compiler namespace and do not rename those
symbols.  This fixes e.g. fortify and sh3 millicode issues.  This
is still not perfect, however, since e.g. the MIPS _gp_disp is not
in ^__.

XXX: we provide a few ^__ symbols in the kernel, but those can be
easily renamed.  The harder problem is how to keep the kernel clean
of those afterwards.

(pooka)

Use PRIu64 to print 64 bit dev_t.

(tsutsui)

Regen from syscalls.master rev 1.72:
> Update for compat_50 stuff. (compile test only)

(tsutsui)

Update for compat_50 stuff. (compile test only)

(tsutsui)

Warn that autoselecting nfs based on : or @ in the device path will
be removed in a future release.

(pooka)

Regen from syscalls.master rev 1.31:
> Update for compat_50 stuff. (compile test only)

(tsutsui)

Update for compat_50 stuff. (compile test only)

(tsutsui)

Regen to pull proper "created from" revision strings.

(tsutsui)

<sys/syscallargs.h> requires <sys/sched.h> for cpuset_t.

(tsutsui)

Use 64bit long minor nuber in dm_dev structure.

(haad)

Fix aprint_debug after christos-time_t merge.

(haad)

make this compile

(cegger)

make this compile

(cegger)

make this compile

(cegger)

make this compile

(cegger)

make this compile

(cegger)

Make this compile.

(nakayama)

make this compile

(cegger)

make this compile

(cegger)

make this compile

(cegger)

make this compile

(cegger)

make this compile

(cegger)

make kgdb_dev a dev_t

(cegger)

make this compile

(cegger)

make this compile

(cegger)

make this compile

(cegger)

make this compile

(cegger)

make this compile

(cegger)

make this compile

(cegger)

make this compile

(cegger)

Redo previous for 32-bit kernel.

(nakayama)

p_timer has been changed to struct itimerspec.

(nakayama)

Initialize sc->sc_dev before we use it.

(matt)

Fix inverted logic on FPU type check in fpfault().  From OpenBSD.

(tsutsui)

another bsd.own.mk missing

(christos)

we need route_enqueue not to be static

(christos)

add another directory the cvs did not merge.

(christos)

need bsd.own.mk for $NETBSDSRCDIR

(christos)

merge another directory that cvs missed.

(christos)

cvs does not -j new directories, do it manually.

(christos)

fix typo

(dholland)

get the correct date for today.

(christos)

mentioned files that need changing.

(christos)

add whitespace

(jmcneill)

mention christos-time_t

(christos)

mention christos-time_t

(christos)

mention how to deal with christos-time_t

(christos)

this change was somehow missed.

(christos)

christos-time_t: welcome to 5.99.7

(christos)