malloc(9) -> kmem(9) (easy, straight-forward cases only, for now)

branches: 1.103.66;
#include <sys/bus.h> instead of <machine/bus.h>.

Merge the ppcoea-renovation branch to HEAD.

This branch was a major cleanup and rototill of many of the various OEA
cpu based PPC ports that focused on sharing as much code as possible
between the various ports to eliminate near-identical copies of files in
every tree. Additionally there is a new PIC system that unifies the
interface to interrupt code for all different OEA ppc arches. The work
for this branch was done by a variety of people, too long to list here.

TODO:
bebox still needs work to complete the transition to -renovation.
ofppc still needs a bunch of work, which I will be looking at.
ev64260 still needs to be renovated
amigappc was not attempted.

NOTES:
pmppc was removed as an arch, and moved to a evbppc target.

branches: 1.101.10;
merge yamt-idlelwp branch. asked by core@. some ports still needs work.

from doc/BRANCHES:

idle lwp, and some changes depending on it.

1. separate context switching and thread scheduling.
(cf. gmcgarry_ctxsw)
2. implement idle lwp.
3. clean up related MD/MI interfaces.
4. make scheduler(s) modular.

branches: 1.100.2; 1.100.4; 1.100.10;
fix fall out from caddr_t changes.

Kill caddr_t; there will be some MI fallout, but it will be fixed shortly.

branches: 1.98.8;
Convert both sparc and sparc64 to MI todr.

convert to timecounters (from branch simonb-timecounters)

branches: 1.96.4; 1.96.6; 1.96.8; 1.96.14;
ANSIify. Same binary code is produced for GENERIC.MP + KGDB + DDB.

branches: 1.95.2; 1.95.8;
Fix the various todr_gettime() and todr_settime() fallouts from
-Wcast-qual differently, by instead changing the signatore of those
"functions" to take a "volatile struct timeval*" instead of a
"struct timeval*". Many places, these functions are called with
&time, and time is declared as volatile in <sys/kernel.h>. This
way we can get rid of all the ugly casts which now also triggered
warnings, and caused more code to be added to work around the
problem.

Reviewed by thorpej.

Sprinkle some const

__KERNEL_RCSID()

branches: 1.92.2;
Move all `ID prom' stuff into the prom library.

Merge the nathanw_sa branch.

Move schedintr() to clock.c and initialise schedhz in initclocks(), so
these are available to all timer implementations.

Remove a local declaration shadowing the global "int statint", thus
preventing the latter to be initialized properly.

Patch from adam@lebsack.com in PR 16608.

Untangle clock.c:
* put various time-of-day clock, timer and eeprom drivers in separate files
* avoid unnecessary overhead in interrupt handlers
* make the MSIIEP timer attach through the regular autoconf mechanisms

* `bus_type_t' is gone.
* Use BUS_ADDR() where appropriate to encode I/O space and physical
address offset into a `bus_addr_t' value.
* Drop obio_bus_map() since it's now completely equivalent to bus_space_map()
* Use bus_space_map2() to map device space at a fixed virtual address.
* Remove the virtual address argument from sbus_sbus_addr()

Typo in comments.

Allow MD code to provide functions for reading/writing NVRAM/RTC
locations. If passed NULL, the old behaviour using bus_space_{read,write}_1()
is used. Otherwise, all access to the chip goes via the MD functions.

This is necessary for mvmeppc boards where the mk48txx NVRAM/RTC is not
directly addressable.

tabify

Add microSPARC-IIep support. Protect mostek clock code with NMK48TXX
so that this file doesn't require obio, iommu and sbus to link the kernel.
Make todr_handle and establish_hostid() non-static.

defopt "options SUN4*", put #define for each of SUN4, SUN4C, SUN4M and SUN4U
into opt_arch_sparc.h and include this wherever they are used.

branches: 1.81.4;
getprop* -> PROM_getprop*

branches: 1.80.2;
Fix non-properly ifdef'ed SUN4 only code.

branches: 1.79.2;
Account for the number of counter ticks that elapse while statintr()
processing takes place. This prevents stat clock skew by a factor
approximately proportional to stathz (not a serious problem if stathz==100;
but considerable already if stathz==1000).

branches: 1.78.2;
In cpu_reboot(), only reset the TOD clock if the time is known to be good
enough, e.g. it has been initialized already by inittodr() or set
explicitly before by resettodr(). This prevents the TOD clock from going
way backwards when typing `halt' at the `mount root filesystem' prompt.

Remove left-over turds from previous __P() cleanup.

Print host ID in hex.

De-__P() this file.

Switch to MI mk48xx and intersil7170 time-of-day clock drivers.

Pull in <dev/clock_subr.h>.

remove include of <vm/vm.h> and <machine/pmap.h>. <vm/vm.h> -> <uvm/uvm_extern.h>

Nuke dead code.

Ensure uniqueness of EEPROM device without reference to cf_unit.

branches: 1.69.12;
Turn on the counter at each CPU (we fear nothing..)

Take care of a MID not present when matching timer register sets.

Init a variable.

Adapt to changes to the cpus[] array.

Dup header.

Map the per-CPU counter registers into each cpu info area.

getpropA() => getprop()

Convert to [pv]addr_t & [pv]size_t.

Replace direct `sparc_bus_map()' calls with bus space map method.

obio_bus_probe() => bus_space_probe()

Account for changed bus attachment scheme.

Update for changes to config.

Establish the interrupt handlers on sun4 `oclocks'; pointed out
by Chuck Cranor.

handle case of no keyboard configured

Use intr_establish() to install the timer interrupt handlers.

branches: 1.54.2;
Move the per-processor counter register into cpuinfo.

Add a function to stop sun4m counters and call it early during
auto-configuration.

branches: 1.52.4;
Remove `bustype' argument from map[io]dev().

Switch to copy of file in /sys/dev/ic.

Replace `cpumod' with `cpuinfo.cpu_type' equivalent.

branches: 1.49.6;
Comply with recent autoconfiguration changes. Diffs graciously
supplied by Chris Demetriou.

Fix timer initialization so that we don't gain one extra microsecond per
second on Sun4m machines. Although this was in the noise of the unstable
Sun clock crystals before, the discrepancy amounted to about 100 ppm, and
thus made NTP perform poorly. NTP now works happily on my SS20...

backout previous kprintf change

printf -> kprintf, sprintf -> ksprintf

Remove unneeded header file.

Copyright police (s/Harvard University/Harvard College/).

Remove some comments in eeprom_uio() now that they're no longer relevant.
Add some whitespace to eeprom_take().

Various simplifications and corrections from Chuck Cranor (makes `oclock'
calibration work.. thanks!).

Fix apparent paste-o ... spurious "}"

Move delay() to locore, so we control code generation and can avoid
touching memory locations whose caching disposition can perturb timings.

Turn delay() into a calibrated loop. It's more accurate and compiles into
a `leaf' routine, avoiding unexpected large delays by overflowing the
register windows.

remove include of <sys/cpu.h>

Probe clock registers on sun4s to make sure we attach the correct device
(sun4m has clocks at obio too).

Include support for sun4m timers and counters: hardware register layout
has changed. There is a counter for each CPU (currently, just one supported).

New device attachment scheme:

- split softc size and match/attach out from cfdriver into
a new struct cfattach.

- new "attach" directive for files.*. May specify the name of
the cfattach structure, so that devices may be easily attached
to parents with different autoconfiguration semantics.

comment: LEAPYEAR is not entirely correct

Add prototypes and fix bugs:
- softclock was called with extra argument.
- missing %x formats in printf's
- kgdb_copy called with only two arguments.

Fix conf.c so that it used the _conf.h files from other places instead of
rolling its own.

Correct timetochip() prototype.
Cleanup microtime() a bit.

delay() takes an unsigned int; get the code to match.

Bring prototypes into scope.
Fix a few bogus `return's.

Account for the fact that `clockreg' might not be page-aligned.
Misc prototype goo.

Make sure we assign a value to variables we use.
Don't express size of NVRAM in terms of machine page size.

Protect clearing the Intersil chip's interrupt with splhigh(). If this
gets interrupted (by the higher priority zs hardware, for example),
the clock can come to a dead stop!

Adapt to changed mapiodev() interface.

eeprom_match: account for differing physical addresses (per Jason Thorpe).

Update from Jason: eeprom device needs 8-bit access.

Export `eeprom_va' (from Jason Thorpe).

eeprom device for Sun4 from Jason Thorpe.

kernel_pmap --> pmap_kernel()

Working "config generic" support; from Theo.

Integrate changes from Chuck Cranor for the Sun 4/100.
Support for multiple register banks on SBUS devices (based on patches
from Francis Dupont).
Highlights:
romaux defines an array of register spaces.
pay attention to `vactype'.
quirks handling 4/100 idiosyncracies.

void microtime().

microtime botch

... but we still want the sun4c/sun4m case to work

oclock support for the 4/100 and 4/200

XXmatch change (and a few __P() as well)

copyright/Id cleanup

mapiodev takes a bustype argument now

sun4/300 support works

generalize bus support for match/attach

deal with sun4 idprom

add support for the Mostek MK48T08 (found in sun4m)

revert to dual clocks

sparc runs in current today

Really suppress clock warning message, even if celebrating.

Suppress `preposterous time' message if the time base equals zero, as
is the case with NFS mounted root filesystems.

fixed: hostid was incorrectly read from register

pmap.c, machdep.c, autoconf.c, cpu.c, intr.c, : net-posted patches from Torek
autoconf.c: mountroot kludges that need fixing later
clock.c: two volatile decls
conf.c: stategy()'s return void, *tty[] not tty[], __P()'d xxdump() calls,
add mem_no, *constty[] not &cons.
genassym.c, machdep.c: sys/vmmeter.h needed
intr.c: make kernel writable temporarily in intr_fasttrap()
intr.c: do not use NETISR_ARP
locore.s: icode maps at 0 so do not relocate, don't define __main here.
machdep.c: don't have sysctl.h (yet), use vm_page_free_count for freemem,
setregs() sets pc/npc/stack, note: signal delivery code needs
fixing, add various other NetBSD function stubs.
pmap.c: do kernel_pmap/kernel_pmap_store the old way, delete
vm_page_startup_initialized weirdness.
swapgeneric.c: everything about mountroot needs fixing, help.
trap.c: sunsys -> sun_sysent, etc.
vm_machdep.c: don't use cpu_coredump() for now.

Chris Torek's sparc port. Missing lots of things.