Trailing whitespace.

sgimips: call mips_vector_init earlier to fix boot.

PR port-sgimips/58269

Move the at-shutdown call to resettodr() from cpu_reboot() to kern_reboot().

It's a small step, but it's a step.

Remove unnecessary <sys/malloc.h>.

fix typos in "environment" word.

Fix all remaining typos, mainly in comments but also in few definitions and log messages, reported by me in PR kern/54889.
Also fixed some additional typos in comments, found on review of same files or typos.

branches: 1.150.6;
uvm_availmem(): give it a boolean argument to specify whether a recent
cached value will do, or if the very latest total must be fetched. It can
be called thousands of times a second and fetching the totals impacts not
only the calling LWP but other CPUs doing unrelated activity in the VM
system.

Rename uvm_free() -> uvm_availmem().

uvmexp.free -> uvm_free()

Initialize ci_divisor_delay by temporary cpufreq for early delay(9) calls.

This is actually required for Indigo R3k.
Reported by Naruaki Etomi in PR port-sgimips/53520.

Fix build failure without options DDB and with pseudo-device ksyms.

There is a typo for ksyms (NKSYMS vs NKSYM).
Also use consistent "#if NKSYMS" rather than "#if NKSYMS > 0" etc.
and add comments after corresponding #endif for readability.

Fixes PR port-sgimips/53521 from Naruaki Etomi.
Should be pulled up to (at least) netbsd-8.

branches: 1.145.2; 1.145.4;
Cleanup and clarify the ELFSIZE mess:

We now have 2 variables automatically set in elf_machdep.h:

ARCH_ELFSIZE: the size for userland binaries
KERN_ELFSIZE: the size for the kernel binaries

DB_ELFSIZE has been deleted and KERN_ELFSIZE should have always the
same values DB_ELFSIZE used to have.

In sys/exec_elf.h, if ELFSIZE is not set, it is set to KERN_ELFSIZE
for the kernel and ARCH_ELFSIZE for userland. These defaults should
eliminate the need for most manual ELFSIZE setting.

branches: 1.144.6;
kill variable stack allocation

branches: 1.143.2;
switch all ports to use uvm_init.c:uvm_md_init()

uvm_setpagesize() is now subsumed within this funciton.

Prefer modern simple past tense and past participle of catch

The "catched" form is obsolete and nonstandard, prefer "caught".

branches: 1.141.2;
Cleanup includes

branches: 1.140.6;
- use cpu_{g,s}etmodel
- fix unused

- remove unused variables
- use (void) when we don't care about the return value

branches: 1.138.2;
split device_t/softc for all remaining drivers.
replace "struct device *" with "device_t".
use device_xname(), device_unit(), etc.

branches: 1.137.2;
Fix -fno-common fallout.

branches: 1.136.2;
Fix DEBUG build.

#include <sys/bus.h> instead of <machine/bus.h>.

Merge forward matt-nb5-mips64
Adapt to new interrupt/spl framework
Deal with arcbios calls being O32 and kernels being N32/N64.
Need to save/restore T8 across arcbios calls.

branches: 1.133.2; 1.133.4; 1.133.6;
Remove separate mb_map. The nmbclusters is computed at boot time based
on the amount of physical memory and limited by NMBCLUSTERS if present.
Architectures without direct mapping also limit it based on the kmem_map
size, which is used as backing store. On i386 and ARM, the maximum KVA
used for mbuf clusters is limited to 64MB by default.

The old default limits and limits based on GATEWAY have been removed.
key_registered_sb_max is hard-wired to a value derived from 2048
clusters.

branches: 1.132.2;
Fix various build with the 32bit mips ports after the mips64 merge.

Merge from matt-nb5-mips64
Merge mips-specific arch files.

- Use uvm_lwp_setuarea() instead of directly setting address to lwp_t::l_addr.
- Replace most remaining uses of l_addr with uvm_lwp_getuarea() or lwp_getpcb().
- Amend assembly in ports where it accesses PCB via struct user.
- Rename L_ADDR to L_PCB in few places. Reduce sys/user.h inclusions.

Kill proc0paddr. Use lwp0.l_addr instead.

Remove all declarations of physmem from sys/arch. Add an include of
<sys/systm.h> to the one file that did not already contain it.
This now means that physmem can be changed by updating systm.h and uvm_page.c
(excluding fixing printfs)

Ansify function definitions w/o arguments. Generated with sed.

Use "defopt MODULAR" in sys/conf/files, and #include "opt_modular.h"
in all kernel sources that use the MODULAR option.
Proposed in tech-kern on 18 Jan 2009.

Enable Personal IRIS 4D/20 and 4D/25 support:
- Adapt int(4) to handle the INT1 chip
- Move generic rtc clocks out of hpc/ and into dev/
- Handle the very strangely wired eeprom and other bits in arcemu
- Sprinkle MACH_SGI_IP6 as necessary
- Enable IP6/IP10 devices in GENERIC32_IP12. Yes, the naming is poor but
there's no winning with kernel/hw compatibility on sgimips...

Tested on my 4D/25. Doesn't (appear to) break macallan@'s IP22.

branches: 1.124.4;
As discussed on tech-kern: mutex_init is too heavyweight for early bootstrap
phases, so move the initialization of the ksyms mutex back into main via
a function called ksyms_init. Rename the existing (but quite different)
ksyms_init* variations into ksyms_addsyms_elf() and ksyms_addsyms_explicit()
and adapt machdep code accordingly.

Remove LKMs and switch to the module framework, pass 1.

Proposed on tech-kern@.

It is not appropriate to call pmf_system_shutdown(9) from
doshutdownhooks(9): shutdown hooks registered by shutdownhook_establish(9)
expect to be called with interrupts disabled, but shutdown hooks
registered with pmf_device_register1(9) expect to be called with
interrupts enabled. So I have made two changes:

1 Do not call pmf_system_shutdown() from doshutdownhooks(). Instead,
change every call to doshutdownhooks() to a call to doshutdownhooks()
followed by a call to pmf_system_shutdown(). No functional change
is intended by this change.

2 Make i386 re-enable interrupts briefly while it calls
pmf_system_shutdown(). I leave it to others either to fix the
other ports, or to factor out some MI shutdown code, as joerg@
suggests, and fix that. Note that a functional change *is* intended
by this change.

I hope that this patch will stop us from flip-flopping between
calling doshutdownhooks() and pmf_system_shutdown() sometimes with
and sometimes without interrupts enabled.

branches: 1.121.2; 1.121.8;
Two fixes to make -current kernel be loadable by 4.0 bootloader:
- if bootinfo is not passed from bootloader, check argv[1] first
which might be the loaded kernel path specified for bootloader
- when using argv[0] as the kernel path (loaded directly by the PROM)
assume boot partition is 0 (partition a) since the kernel might be
in an SGI volume header partition and it can't be a root partition

Ok'ed by rumble@, and tested on IP22 and IP32.

branches: 1.120.2;
Replce exec_map with a pool. Proposed on tech-kern@, reviewed by chs@.

branches: 1.119.4; 1.119.6; 1.119.8;
Adjust boot device detection code to make netboot work properly:
- make a bootloader pass bootpath which includes device path on ARC BIOS
via BTINFO_BOOTPATH
- make a kernel use BTINFO_BOOTPATH in bootinfo to see boot device first,
then check argv[0] passed from ARC BIOS if there is no valid bootinfo
- check OSLoadPartition variables in the ARCBIOS environment and whole
argv[] arguments for IP12
- initialize mach_type before it's referred
- bump boot version

Tested on IP32 with various kernel/bootloader combinations.

branches: 1.118.6;
Replace bootinfo stuff on sgimips (which has been disabled in bootloader)
with ones used on arc (which has been working for a year) and
enable it in bootloader. Tested on IP32.

Interrupt handling changes, in discussion since February:

- Reduce available SPL levels for hardware devices to none, vm, sched, high.
- Acquire kernel_lock only for interrupts at IPL_VM.
- Implement threaded soft interrupts.

branches: 1.116.2;
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.

use crime_reboot() and mcclock_poweroff() when appropriate.
With this my O2 resets and powers down again even when using crmfb,
apparently it freaks out the ARCS firmware - with crmfb the machine would
just sit there blinking instead of rebooting.

branches: 1.114.8; 1.114.10; 1.114.12;
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.

IP12's PROM passes us a '-a' flag in argv when booting directly from the
disk volume header. On NetBSD this means RB_ASKNAME, however in SGI terms
it appears to mean 'autoboot'.

Add an appropriate kludge to keep us from doing the wrong thing.

XXX - This whole argv/env/prom parameter mess needs to be cleaned up.

A few changes to allow booting to multi-user with IP12 kernels squished
into oversized volume headers:

Query ARCBIOS for OSLoadFilename and OSLoadOptions first, rather than only
using the environment variables passed since we want to go through arcemu.

Add an ugly hack to arcemu that munges the nvram 'bootfile' variable. We
need this because bootfile will point to the volume header, which then
translates into a bogus root device of 'sd0i', or similar.

Clean up extracting environment variables passed by the prom. No functional
change.

branches: 1.110.2; 1.110.4; 1.110.10;
Use (char *) to appease -Wpointer-arith.

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

TRUE -> true, FALSE -> false

Since IP12 doesn't use ARCS, we could read the console setting from the
nvram, but it may not be valid. For example, if the console is set to
graphics, but no keyboard is attached, we should not attempt to use a
graphical console.

To fix this, we extend arcemu to parse the prom's environment array, which
is the third argument to the kernel. If the console mode is graphics, but
no keyboard is found, the 'gfx' environment variable will contain 'dead'.

Note that since IP12 has no bootloader, we assume that the third parameter
is the environment table, not bootinfo.

branches: 1.106.2;
Merge newlock2 to head.

Fix typo in comment.

Move the mips1_clock_intr handling into int.c where it belongs.

IP12 (and perhaps other systems) appear to be buggy and incapable of
reliably supporting badaddr(). False negatives appear to occur
approximately 1.8 percent of the time, although neither false positives
nor consecutive false negatives occur. We take advantage of the latter
property and always use a wrapper that makes multiple checks.

My IP12 no longer sees occasional ghost devices and related panics during
boot.

Handle fpu interrupts (MIPS interrupt 0) on IP12.

Specify the member names of struct platform for clarity.

Guiness -> Guinness.

merge yamt-splraiseipl branch.

- finish implementing splraiseipl (and makeiplcookie).
http://mail-index.NetBSD.org/tech-kern/2006/07/01/0000.html
- complete workqueue(9) and fix its ipl problem, which is reported
to cause audio skipping.
- fix netbt (at least compilation problems) for some ports.
- fix PR/33218.

branches: 1.98.2;
sgimips converted to timecounters. Systems that can do them get mips_cp0
based logic. We also use the common mips3_clock functions for those systems,
removing some port-specific code.

Tested by martin@.

branches: 1.97.2;
Fix delay() so that gcc4 does not optimize it into a noop. Lifted from the
newsmips implementation by tsutsui@

Remove declarations for cpu_model[] since it's in <sys/systm.h>.

branches: 1.95.4; 1.95.6; 1.95.8; 1.95.10; 1.95.12;
merge ktrace-lwp.

Restore missing call to r5k_enable_sdcache(). Its' removal was a thinko
on my part -- the COP0 SC presence test fails on r5k IP22, but
r5k_enable_sdcache() will DTRT regardless.

Pointed out by tsutsui@

branches: 1.93.2;
Avoid a global name "netmask" - it causes lots of shadow warnings all over
the place. Collect all interrupt masks into a single array instead.
Constify a bit and fix a bogus argument to config_rootfound().

Move the MI printing of `copyright' to the MD cpu_startup() code
where the printing of `version' is already performed.
This has the benefit of allowing the copyright to be available
via dmesg(8) on platforms which need the `msgbuf' to be setup
in cpu_startup() before printed output is remembered.

mips_sdcache_forceinv has been removed.

XXX Should we check MIPS3_CONFIG_SE in mips3_cp0_config_read() and
XXX call r5k_enable_sdcache() accordingly here?

branches: 1.90.2;
Set mips_sdcache_forceinv to 1 for r5k processors.

We now support secondary cache operation on r5ksc out-of-the-box.

branches: 1.89.2; 1.89.4;
Nuke MIPS64; we'll use MIPS3 to indicate the presence of MIPS3 instructions
and _LP64 to indicate that 64-bit is desired.

Pointed out by simonb@

MIPS3_CONFIG_SC is an unreliable indicator for SC presence; r5k_enable_sdcache()
will gracefully exit if there is no secondary cache. Rework #ifdef'd out
r5k cache code to take this into account, and remove pointless r10k case.

Use the mips3 bus functions on IP30. Recognize the IP30 as a supported* SGI
platform.

* For extremely small values of "supported". arch/mips/* needs a lot of work
yet for MIPS64.

s,#ifdef MIPS3,#if defined(MIPS3) || defined(MIPS64),g

KNF some and make printf output a bit nicer.

Add IP12 stuff. Most bits included are from Steve Rumble.

After this change it is possible to run IP12.

If argv[1] doesn't exist, don't try to use it.

Tidy up total/avail memory message as well as other ports.
As per discussion on port-sgimips.

branches: 1.81.2;
Sprinkle some #ifdef MIPS3 to be able to compile the port without
including MIPS3 code.

for upcoming IP12

Print an informative message when unimpl_intr() is invoked, rather than panic().

Completely remove IP2X and IP32, which renders opt_machtype.h obsolete.

g/c obsolete function.

Attach int0 at mainbus (precedent being the ip32 CRIME device, which handled
interrupts and timer calibration yet is also attached at mainbus).

Introduce the INDY_R4600_CACHE config option, which more accurately describes
the code enabled by this option.

This renders #ifdef IPxx completely obsolete, thus we theoretically can unify
the GENERIC files if a workable load address relocation scheme can be found.

Remove another #ifdef IP32

Large rototill of the lowlevel code:

* move all device-specific bitbanging to the device drivers in question --
i.e., the ip32 init code doesn't access the CRIME registers directly but
instead trusts the CRIME driver to DTRT. machdep.c is cleaner thereby.
* introduce int.c, which handles the interrupt-controller goo formerly
in ip2x.c.
* unify the CPU interrupt handler, with CRIME, INT2, and the various timer
handlers hooked in via pointers in the platform struct.
* eliminate ip2x.c and ip32.c. After the interrupt handler cleanup, the
actual platform-dependent code was so minimal (rev identification, mask level
initialization) that it made sense to move it into machdep.c
* "#ifdef DEBUG printf("foo"); #endif" becomes aprint_debug.

Tested on my ip22 and ip32. Discussed with pooka@, rafal@.

Call sgimips_bus_dma_init() in mach_init to set up the proper MIPS1/MIPS3
bus_dma_sync() routine.

Add a call to ip1x_init().

Wrap more mips3_ operations in MIPS3 defines.

Streamline bootinfo logic.

Rework the memory detection routine to handle cases where kernel doesn't
fit inside one memory chunk.
Leave 1 page before kernel untouched as that's where our initial kernel
stack before we switch to proc0 stack is (fixes boot problems on my Indy
with small kernels).

Add trailing newline. Noticed by Steve Rumble.

Replace the traditional buffer memory management -- based on fixed per buffer
virtual memory reservation and a private pool of memory pages -- by a scheme
based on memory pools.

This allows better utilization of memory because buffers can now be allocated
with a granularity finer than the system's native page size (useful for
filesystems with e.g. 1k or 2k fragment sizes). It also avoids fragmentation
of virtual to physical memory mappings (due to the former fixed virtual
address reservation) resulting in better utilization of MMU resources on some
platforms. Finally, the scheme is more flexible by allowing run-time decisions
on the amount of memory to be used for buffers.

On the other hand, the effectiveness of the LRU queue for buffer recycling
may be somewhat reduced compared to the traditional method since, due to the
nature of the pool based memory allocation, the actual least recently used
buffer may release its memory to a pool different from the one needed by a
newly allocated buffer. However, this effect will kick in only if the
system is under memory pressure.

Fix typo ("ARCBIOS_MEM_ExecptionBlock"). From Steve Rumble.

Part two in the Great Renaming. ip22.c becomes ip2x.c, and all generic ip2x
routines become ip2x_*.

Change all IP22 preprocessor conditionals to IP2X preprocessor conditionals.

Part one of two -- part two will be renaming ip22_*() to ip2x_*().

IP20 should handle the watchdog timer in the same manner as for IP22.

For the purposes of machdep.c, consider IP20 to be a variant of IP22. ip20.c
and #ifdef IP20 will go away (short-term), ip22_* will be renamed to ip2x_*
(slighly longer-term).

The ARCBIOS->FlushAllCaches() vector is NULL on at least one SGI platform
(ip20) and has no actual operational effect, thus it should not be invoked here.

Remove the extra call to consinit().

www.netbsd.org -> www.NetBSD.org

Apply a bunch of patches written by Christopher SEKIYA
(with several cosmetic changes by me) which fixes O2 (IP32) support.

Now my R5000 O2 works fine in multiuser with on-board AIC7880 SCSIs
and several PCI network cards (but only on serial console yet).
L2 cache on R5000/Rm5200 is still disabled for now, but it will be
fixed later, hopefully.

See recent discussion on port-sgimips for details.

TAB/space cleanup.

Cast through (void *) to appease gcc3.

Use common mips generic software interrupt routines.

Changes from Izumi Tsutsui, tested by Havard Eidnes.

__KERNEL_RCSID()

branches: 1.54.2;
Call ksyms_init() instead of ddb_init() in case of
NKSYMS || defined(DDB) || defined(LKM)

Use PAGE_SIZE rather than NBPG.

Oops, make sure to re-enable the IP32 watchdog correctly, else we end up
getting random reboots after returning from DDB.

Parse out flags passed to the kernel on boot; from Scott Taylor, tweaked to
use the MI boot_flags.h interface by me.

Merge the nathanw_sa branch.

Get rid of some extraneous declarations/comments.

Add the MIPS3_CONFIG_SE (name taken from Rm52xx manual) bit, which is the
external cache enable bit -- this allows software to enable or disable the
(external) L2 cache on the R5k and Rm527x and the (external) L3 cache on
the Rm7k. If the (external) cache is disabled, treat it as if there were
no cache for the purposes of the cache setup code.

Also, update sgimips code to use the new name.

There's no cache_r5k.h yet, so don't try to include it.

Postpone DDB/KGDB init until we know the machine type, so the debugger can
be attached to the right place, twiddle the right watchdog registers, etc.

Checkpoint of O2 work by Chris Sekiya and myself. This is the sgimips bit;
still needs some arch/mips support code before it will fully work.

Move console setup after we've determined machine type so the console init
code can guess where to find a console depending on machine type.

Use crimereg.h definitions for the crime dog. Also, move watchdog
disable down the very last possible place in cpu_reboot.

from Chris Sekiya

Fix signed/unsigned comparison warnings.

remove trailing \n in panic(). approved perry.

Don't include <sys/map.h>.

Make nbuf, nswbuf, and bufpages unsigned. Make all operations on these
variables unsigned, and update places where their values are printed.

Merge cpu_sysctl() for all mips ports, based on powerpc and m68k precedent.

For now, only pmax implements CPU_BOOTED_KERNEL. Need to revisit.

branches: 1.37.2;
Update copyright dates.

Fix up clock interrupt accounting for the sgimips port -- make sure
to schedule clock interrupts at a fixed interval, rather scheduling
the next one based on the time of the arrival/servicing of the previous
clock interrupt. Also, pick up a trick from the sbmips port to convert
a division in ip22_clkread to a multiplication, since those are much
cheaper -- the details of that are described in Simon's commit (see
Message-Id: <20020306073437.1D2A8B004@cvs.netbsd.org>). Thanks to
Jason Thorpe and Dominic Sweetman's "See MIPS Run" (where I found
mention of this very subject while looking for something totally un-
related! 8-) for the clue about the source of the timekeeping problems.

For the IP32, where we have no clock-calibration code yet, use the CPU
frequency provided by ARCS instead; it beats a hard-coded value!

As an added bonus, most of the CPU-clock related stuff is now collected
together in cpu_info_store, rather than as a collection of unorganized
global variables.

Fix nasty crashes on IP22 machines with console running at 38.4kbps (which
generally translates to "high interrupt load") -- the old code re-enabled
interrupts in the machine-specific hardware interrupt handler causing the
handler to be re-entered, possible multiple times. Could lead to kernel
stack overflows, and all sorts of mysterious crashes/hangs as a result.

While here, fix up the IP32 interrupt handler code to also not re-enable
interrupts.

Thanks for ideas/comments go to Chuq and Stephen Ma.

Replace lots of 8x<space> with <tabs> and other miscellaneous indentation
fixes.
Wrap a couple of long lines.
Use <return-type>\n<function name> as per KNF in a few places.

Respect OSLoadPartition even if it doesn't come before SystemPartition in argv.
Add a few comments about the above.

- Check for provided bootinfo from the bootloader. If we find it,
save off DDB symbol table information.
- Make loading of memory work reliably with the bootloader; for each
candidate memory type from ARCS (which now includes LoadedProgram),
check to see if the kernel is within that chunk, and load the pages
around it if it is.

Also check SystemPartition when finding boot device. Especially
when netbooting, it is practical to load the kernel directly.

branches: 1.30.2;
When loadfile() leaves the ELF header at the end of the kernel, it enables
erroneous code that tries to compute the end of the symbol table. Disable
that for now so that kernels don't hang when booted from the disk.

Make microtime(9) do interpolation to get better-than-hz resolution, if
the platform supplies a clkread function that does something other than
return 0 (which is the default unless overridden by the platorm code).
Supply such a function for the IP22; even if it isn't perfect, it goes
a long way to making ntp usable.

While I'm at it, move the ticks-per-hz variable out of the struct platform
since it's really private to the per-platform interrupt/clock code.

XXX: No clkread function supplied for IP32, since it has other problems --
like a hardcoded ticks-per-hz, but the same code as on the IP22 could be
used.

branches: 1.28.2;
Use boot arguments to try to determine the boot device. The OSLoadPartition
string, if present, will override the second argument (which may be the
path/kernel being loaded). This will provide a way to netboot the kernel
and allow the root device be set to a disk partition.

Run vfs_shutdown() if NOSYNC is *not* set, rather than set.

If OSLoadOptions is set to auto, boot straight to multi-user. Requested
one too many times and it seems to do no harm, so enable it.

Obey the RB_POWERDOWN flag passed to cpu_reboot; make it work on the Indy
by enabling the FPU before jumping back to ARCS (gross, but necessary!).

Update pmap_update to now take the updated pmap as an argument.
This will allow improvements to the pmaps so that they can more easily defer expensive operations, eg tlb/cache flush, til the last possible moment.

Currently this is a no-op on most platforms, so they should see no difference.

Reviewed by Jason.

branches: 1.23.2; 1.23.4;
Count and display the number of CPUs present on the system.

Fill in cpu_model[]

- Use the identifier strings that are fetched by the MI ARC BIOS
code.
- Garbage-collect some non-useful stuff from the mainbus_attach_args.

bcopy -> memcpy, bzero -> memset

Use sys/dev/arcbios.

- Finish off KGDB kernel debugger support for sgimips

Implement a db_trap_callback which turns off the watchdog on entry to DDB
and turns it back on on exit. Hook the callback in early in the boot, so
it's there as soon as DDB is available.

replace vm_map{,_entry}_t with struct vm_map{,_entry} *.

Adjustments for dynamic system configuration. From Rafal K. Boni.

Do the first allocsys() pass (the pass that computes sizes) before
calling pmap_bootstrap(); pmap_bootstrap() needs some of the size
information computed by allocsys().

Sprinkle pmap_update() calls after calls to:
- pmap_enter()
- pmap_remove()
- pmap_protect()
- pmap_kenter_pa()
- pmap_kremove()
as described in pmap(9).

These calls are relatively conservative. It may be possible to
optimize these a little more.

Call pmap_bootstrap() before calling uvm_pageboot_alloc(), so that
the kernel virtual address space bounds will be set up.

Use uvm_pageboot_alloc() for early memory allocation, rather than
calling pmap_steal_memory() directly. On these platforms, since
uvm_pageboot_alloc() is a wrapper around pmap_steal_memory(), there
is no functional change. This is merely for API consistency.

eliminate the KERN_* error codes in favor of the traditional E* codes.
the mapping is:

KERN_SUCCESS 0
KERN_INVALID_ADDRESS EFAULT
KERN_PROTECTION_FAILURE EACCES
KERN_NO_SPACE ENOMEM
KERN_INVALID_ARGUMENT EINVAL
KERN_FAILURE various, mostly turn into KASSERTs
KERN_RESOURCE_SHORTAGE ENOMEM
KERN_NOT_RECEIVER <unused>
KERN_NO_ACCESS <unused>
KERN_PAGES_LOCKED <unused>

branches: 1.9.2;
Require the machine-dependant DDB commands to be in db_machine_command_table[]
and link it directly to db_command_table[] so that it's not necessary
to do this at runtime. Make db_machine_command_table[] const on all ports.
g/c now unneded stuff, like db_machine_commands_install(), db_machine_init()

Patch written by enami.

Make softclock a generic soft interrupt of the API is available,
adding the requisite void * argument to softclock().

Consistently use pmap_steal_memory() to allocate USPACE.

branches: 1.6.2;
clean up and consistency for CP0 Count, Compare, Wired, and Config
access function names and prototypes.

Add an align argument to uvm_map() and some callers of that
routine. Works similarly fto pmap_prefer(), but allows callers
to specify a minimum power-of-two alignment of the region.
How we ever got along without this for so long is beyond me.

Use cpu_intr() definition from mips/cpu.h instead of local one.

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

remove/move more mach vm header files:

<vm/pglist.h> -> <uvm/uvm_pglist.h>
<vm/vm_inherit.h> -> <uvm/uvm_inherit.h>
<vm/vm_kern.h> -> into <uvm/uvm_extern.h>
<vm/vm_object.h> -> nothing
<vm/vm_pager.h> -> into <uvm/uvm_pager.h>

also includes a bunch of <vm/vm_page.h> include removals (due to redudancy
with <vm/vm.h>), and a scattering of other similar headers.

branches: 1.1.2; 1.1.4;
Initial commit. Needs work.

Move the at-shutdown call to resettodr() from cpu_reboot() to kern_reboot().

It's a small step, but it's a step.

Remove unnecessary <sys/malloc.h>.

fix typos in "environment" word.

Fix all remaining typos, mainly in comments but also in few definitions and log messages, reported by me in PR kern/54889.
Also fixed some additional typos in comments, found on review of same files or typos.

branches: 1.150.6;
uvm_availmem(): give it a boolean argument to specify whether a recent
cached value will do, or if the very latest total must be fetched. It can
be called thousands of times a second and fetching the totals impacts not
only the calling LWP but other CPUs doing unrelated activity in the VM
system.

Rename uvm_free() -> uvm_availmem().

uvmexp.free -> uvm_free()

Initialize ci_divisor_delay by temporary cpufreq for early delay(9) calls.

This is actually required for Indigo R3k.
Reported by Naruaki Etomi in PR port-sgimips/53520.

Fix build failure without options DDB and with pseudo-device ksyms.

There is a typo for ksyms (NKSYMS vs NKSYM).
Also use consistent "#if NKSYMS" rather than "#if NKSYMS > 0" etc.
and add comments after corresponding #endif for readability.

Fixes PR port-sgimips/53521 from Naruaki Etomi.
Should be pulled up to (at least) netbsd-8.

branches: 1.145.2; 1.145.4;
Cleanup and clarify the ELFSIZE mess:

We now have 2 variables automatically set in elf_machdep.h:

ARCH_ELFSIZE: the size for userland binaries
KERN_ELFSIZE: the size for the kernel binaries

DB_ELFSIZE has been deleted and KERN_ELFSIZE should have always the
same values DB_ELFSIZE used to have.

In sys/exec_elf.h, if ELFSIZE is not set, it is set to KERN_ELFSIZE
for the kernel and ARCH_ELFSIZE for userland. These defaults should
eliminate the need for most manual ELFSIZE setting.

branches: 1.144.6;
kill variable stack allocation

branches: 1.143.2;
switch all ports to use uvm_init.c:uvm_md_init()

uvm_setpagesize() is now subsumed within this funciton.

Prefer modern simple past tense and past participle of catch

The "catched" form is obsolete and nonstandard, prefer "caught".

branches: 1.141.2;
Cleanup includes

branches: 1.140.6;
- use cpu_{g,s}etmodel
- fix unused

- remove unused variables
- use (void) when we don't care about the return value

branches: 1.138.2;
split device_t/softc for all remaining drivers.
replace "struct device *" with "device_t".
use device_xname(), device_unit(), etc.

branches: 1.137.2;
Fix -fno-common fallout.

branches: 1.136.2;
Fix DEBUG build.

#include <sys/bus.h> instead of <machine/bus.h>.

Merge forward matt-nb5-mips64
Adapt to new interrupt/spl framework
Deal with arcbios calls being O32 and kernels being N32/N64.
Need to save/restore T8 across arcbios calls.

branches: 1.133.2; 1.133.4; 1.133.6;
Remove separate mb_map. The nmbclusters is computed at boot time based
on the amount of physical memory and limited by NMBCLUSTERS if present.
Architectures without direct mapping also limit it based on the kmem_map
size, which is used as backing store. On i386 and ARM, the maximum KVA
used for mbuf clusters is limited to 64MB by default.

The old default limits and limits based on GATEWAY have been removed.
key_registered_sb_max is hard-wired to a value derived from 2048
clusters.

branches: 1.132.2;
Fix various build with the 32bit mips ports after the mips64 merge.

Merge from matt-nb5-mips64
Merge mips-specific arch files.

- Use uvm_lwp_setuarea() instead of directly setting address to lwp_t::l_addr.
- Replace most remaining uses of l_addr with uvm_lwp_getuarea() or lwp_getpcb().
- Amend assembly in ports where it accesses PCB via struct user.
- Rename L_ADDR to L_PCB in few places. Reduce sys/user.h inclusions.

Kill proc0paddr. Use lwp0.l_addr instead.

Remove all declarations of physmem from sys/arch. Add an include of
<sys/systm.h> to the one file that did not already contain it.
This now means that physmem can be changed by updating systm.h and uvm_page.c
(excluding fixing printfs)

Ansify function definitions w/o arguments. Generated with sed.

Use "defopt MODULAR" in sys/conf/files, and #include "opt_modular.h"
in all kernel sources that use the MODULAR option.
Proposed in tech-kern on 18 Jan 2009.

Enable Personal IRIS 4D/20 and 4D/25 support:
- Adapt int(4) to handle the INT1 chip
- Move generic rtc clocks out of hpc/ and into dev/
- Handle the very strangely wired eeprom and other bits in arcemu
- Sprinkle MACH_SGI_IP6 as necessary
- Enable IP6/IP10 devices in GENERIC32_IP12. Yes, the naming is poor but
there's no winning with kernel/hw compatibility on sgimips...

Tested on my 4D/25. Doesn't (appear to) break macallan@'s IP22.

branches: 1.124.4;
As discussed on tech-kern: mutex_init is too heavyweight for early bootstrap
phases, so move the initialization of the ksyms mutex back into main via
a function called ksyms_init. Rename the existing (but quite different)
ksyms_init* variations into ksyms_addsyms_elf() and ksyms_addsyms_explicit()
and adapt machdep code accordingly.

Remove LKMs and switch to the module framework, pass 1.

Proposed on tech-kern@.

It is not appropriate to call pmf_system_shutdown(9) from
doshutdownhooks(9): shutdown hooks registered by shutdownhook_establish(9)
expect to be called with interrupts disabled, but shutdown hooks
registered with pmf_device_register1(9) expect to be called with
interrupts enabled. So I have made two changes:

1 Do not call pmf_system_shutdown() from doshutdownhooks(). Instead,
change every call to doshutdownhooks() to a call to doshutdownhooks()
followed by a call to pmf_system_shutdown(). No functional change
is intended by this change.

2 Make i386 re-enable interrupts briefly while it calls
pmf_system_shutdown(). I leave it to others either to fix the
other ports, or to factor out some MI shutdown code, as joerg@
suggests, and fix that. Note that a functional change *is* intended
by this change.

I hope that this patch will stop us from flip-flopping between
calling doshutdownhooks() and pmf_system_shutdown() sometimes with
and sometimes without interrupts enabled.

branches: 1.121.2; 1.121.8;
Two fixes to make -current kernel be loadable by 4.0 bootloader:
- if bootinfo is not passed from bootloader, check argv[1] first
which might be the loaded kernel path specified for bootloader
- when using argv[0] as the kernel path (loaded directly by the PROM)
assume boot partition is 0 (partition a) since the kernel might be
in an SGI volume header partition and it can't be a root partition

Ok'ed by rumble@, and tested on IP22 and IP32.

branches: 1.120.2;
Replce exec_map with a pool. Proposed on tech-kern@, reviewed by chs@.

branches: 1.119.4; 1.119.6; 1.119.8;
Adjust boot device detection code to make netboot work properly:
- make a bootloader pass bootpath which includes device path on ARC BIOS
via BTINFO_BOOTPATH
- make a kernel use BTINFO_BOOTPATH in bootinfo to see boot device first,
then check argv[0] passed from ARC BIOS if there is no valid bootinfo
- check OSLoadPartition variables in the ARCBIOS environment and whole
argv[] arguments for IP12
- initialize mach_type before it's referred
- bump boot version

Tested on IP32 with various kernel/bootloader combinations.

branches: 1.118.6;
Replace bootinfo stuff on sgimips (which has been disabled in bootloader)
with ones used on arc (which has been working for a year) and
enable it in bootloader. Tested on IP32.

Interrupt handling changes, in discussion since February:

- Reduce available SPL levels for hardware devices to none, vm, sched, high.
- Acquire kernel_lock only for interrupts at IPL_VM.
- Implement threaded soft interrupts.

branches: 1.116.2;
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.

use crime_reboot() and mcclock_poweroff() when appropriate.
With this my O2 resets and powers down again even when using crmfb,
apparently it freaks out the ARCS firmware - with crmfb the machine would
just sit there blinking instead of rebooting.

branches: 1.114.8; 1.114.10; 1.114.12;
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.

IP12's PROM passes us a '-a' flag in argv when booting directly from the
disk volume header. On NetBSD this means RB_ASKNAME, however in SGI terms
it appears to mean 'autoboot'.

Add an appropriate kludge to keep us from doing the wrong thing.

XXX - This whole argv/env/prom parameter mess needs to be cleaned up.

A few changes to allow booting to multi-user with IP12 kernels squished
into oversized volume headers:

Query ARCBIOS for OSLoadFilename and OSLoadOptions first, rather than only
using the environment variables passed since we want to go through arcemu.

Add an ugly hack to arcemu that munges the nvram 'bootfile' variable. We
need this because bootfile will point to the volume header, which then
translates into a bogus root device of 'sd0i', or similar.

Clean up extracting environment variables passed by the prom. No functional
change.

branches: 1.110.2; 1.110.4; 1.110.10;
Use (char *) to appease -Wpointer-arith.

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

TRUE -> true, FALSE -> false

Since IP12 doesn't use ARCS, we could read the console setting from the
nvram, but it may not be valid. For example, if the console is set to
graphics, but no keyboard is attached, we should not attempt to use a
graphical console.

To fix this, we extend arcemu to parse the prom's environment array, which
is the third argument to the kernel. If the console mode is graphics, but
no keyboard is found, the 'gfx' environment variable will contain 'dead'.

Note that since IP12 has no bootloader, we assume that the third parameter
is the environment table, not bootinfo.

branches: 1.106.2;
Merge newlock2 to head.

Fix typo in comment.

Move the mips1_clock_intr handling into int.c where it belongs.

IP12 (and perhaps other systems) appear to be buggy and incapable of
reliably supporting badaddr(). False negatives appear to occur
approximately 1.8 percent of the time, although neither false positives
nor consecutive false negatives occur. We take advantage of the latter
property and always use a wrapper that makes multiple checks.

My IP12 no longer sees occasional ghost devices and related panics during
boot.

Handle fpu interrupts (MIPS interrupt 0) on IP12.

Specify the member names of struct platform for clarity.

Guiness -> Guinness.

merge yamt-splraiseipl branch.

- finish implementing splraiseipl (and makeiplcookie).
http://mail-index.NetBSD.org/tech-kern/2006/07/01/0000.html
- complete workqueue(9) and fix its ipl problem, which is reported
to cause audio skipping.
- fix netbt (at least compilation problems) for some ports.
- fix PR/33218.

branches: 1.98.2;
sgimips converted to timecounters. Systems that can do them get mips_cp0
based logic. We also use the common mips3_clock functions for those systems,
removing some port-specific code.

Tested by martin@.

branches: 1.97.2;
Fix delay() so that gcc4 does not optimize it into a noop. Lifted from the
newsmips implementation by tsutsui@

Remove declarations for cpu_model[] since it's in <sys/systm.h>.

branches: 1.95.4; 1.95.6; 1.95.8; 1.95.10; 1.95.12;
merge ktrace-lwp.

Restore missing call to r5k_enable_sdcache(). Its' removal was a thinko
on my part -- the COP0 SC presence test fails on r5k IP22, but
r5k_enable_sdcache() will DTRT regardless.

Pointed out by tsutsui@

branches: 1.93.2;
Avoid a global name "netmask" - it causes lots of shadow warnings all over
the place. Collect all interrupt masks into a single array instead.
Constify a bit and fix a bogus argument to config_rootfound().

Move the MI printing of `copyright' to the MD cpu_startup() code
where the printing of `version' is already performed.
This has the benefit of allowing the copyright to be available
via dmesg(8) on platforms which need the `msgbuf' to be setup
in cpu_startup() before printed output is remembered.

mips_sdcache_forceinv has been removed.

XXX Should we check MIPS3_CONFIG_SE in mips3_cp0_config_read() and
XXX call r5k_enable_sdcache() accordingly here?

branches: 1.90.2;
Set mips_sdcache_forceinv to 1 for r5k processors.

We now support secondary cache operation on r5ksc out-of-the-box.

branches: 1.89.2; 1.89.4;
Nuke MIPS64; we'll use MIPS3 to indicate the presence of MIPS3 instructions
and _LP64 to indicate that 64-bit is desired.

Pointed out by simonb@

MIPS3_CONFIG_SC is an unreliable indicator for SC presence; r5k_enable_sdcache()
will gracefully exit if there is no secondary cache. Rework #ifdef'd out
r5k cache code to take this into account, and remove pointless r10k case.

Use the mips3 bus functions on IP30. Recognize the IP30 as a supported* SGI
platform.

* For extremely small values of "supported". arch/mips/* needs a lot of work
yet for MIPS64.

s,#ifdef MIPS3,#if defined(MIPS3) || defined(MIPS64),g

KNF some and make printf output a bit nicer.

Add IP12 stuff. Most bits included are from Steve Rumble.

After this change it is possible to run IP12.

If argv[1] doesn't exist, don't try to use it.

Tidy up total/avail memory message as well as other ports.
As per discussion on port-sgimips.

branches: 1.81.2;
Sprinkle some #ifdef MIPS3 to be able to compile the port without
including MIPS3 code.

for upcoming IP12

Print an informative message when unimpl_intr() is invoked, rather than panic().

Completely remove IP2X and IP32, which renders opt_machtype.h obsolete.

g/c obsolete function.

Attach int0 at mainbus (precedent being the ip32 CRIME device, which handled
interrupts and timer calibration yet is also attached at mainbus).

Introduce the INDY_R4600_CACHE config option, which more accurately describes
the code enabled by this option.

This renders #ifdef IPxx completely obsolete, thus we theoretically can unify
the GENERIC files if a workable load address relocation scheme can be found.

Remove another #ifdef IP32

Large rototill of the lowlevel code:

* move all device-specific bitbanging to the device drivers in question --
i.e., the ip32 init code doesn't access the CRIME registers directly but
instead trusts the CRIME driver to DTRT. machdep.c is cleaner thereby.
* introduce int.c, which handles the interrupt-controller goo formerly
in ip2x.c.
* unify the CPU interrupt handler, with CRIME, INT2, and the various timer
handlers hooked in via pointers in the platform struct.
* eliminate ip2x.c and ip32.c. After the interrupt handler cleanup, the
actual platform-dependent code was so minimal (rev identification, mask level
initialization) that it made sense to move it into machdep.c
* "#ifdef DEBUG printf("foo"); #endif" becomes aprint_debug.

Tested on my ip22 and ip32. Discussed with pooka@, rafal@.

Call sgimips_bus_dma_init() in mach_init to set up the proper MIPS1/MIPS3
bus_dma_sync() routine.

Add a call to ip1x_init().

Wrap more mips3_ operations in MIPS3 defines.

Streamline bootinfo logic.

Rework the memory detection routine to handle cases where kernel doesn't
fit inside one memory chunk.
Leave 1 page before kernel untouched as that's where our initial kernel
stack before we switch to proc0 stack is (fixes boot problems on my Indy
with small kernels).

Add trailing newline. Noticed by Steve Rumble.

Replace the traditional buffer memory management -- based on fixed per buffer
virtual memory reservation and a private pool of memory pages -- by a scheme
based on memory pools.

This allows better utilization of memory because buffers can now be allocated
with a granularity finer than the system's native page size (useful for
filesystems with e.g. 1k or 2k fragment sizes). It also avoids fragmentation
of virtual to physical memory mappings (due to the former fixed virtual
address reservation) resulting in better utilization of MMU resources on some
platforms. Finally, the scheme is more flexible by allowing run-time decisions
on the amount of memory to be used for buffers.

On the other hand, the effectiveness of the LRU queue for buffer recycling
may be somewhat reduced compared to the traditional method since, due to the
nature of the pool based memory allocation, the actual least recently used
buffer may release its memory to a pool different from the one needed by a
newly allocated buffer. However, this effect will kick in only if the
system is under memory pressure.

Fix typo ("ARCBIOS_MEM_ExecptionBlock"). From Steve Rumble.

Part two in the Great Renaming. ip22.c becomes ip2x.c, and all generic ip2x
routines become ip2x_*.

Change all IP22 preprocessor conditionals to IP2X preprocessor conditionals.

Part one of two -- part two will be renaming ip22_*() to ip2x_*().

IP20 should handle the watchdog timer in the same manner as for IP22.

For the purposes of machdep.c, consider IP20 to be a variant of IP22. ip20.c
and #ifdef IP20 will go away (short-term), ip22_* will be renamed to ip2x_*
(slighly longer-term).

The ARCBIOS->FlushAllCaches() vector is NULL on at least one SGI platform
(ip20) and has no actual operational effect, thus it should not be invoked here.

Remove the extra call to consinit().

www.netbsd.org -> www.NetBSD.org

Apply a bunch of patches written by Christopher SEKIYA
(with several cosmetic changes by me) which fixes O2 (IP32) support.

Now my R5000 O2 works fine in multiuser with on-board AIC7880 SCSIs
and several PCI network cards (but only on serial console yet).
L2 cache on R5000/Rm5200 is still disabled for now, but it will be
fixed later, hopefully.

See recent discussion on port-sgimips for details.

TAB/space cleanup.

Cast through (void *) to appease gcc3.

Use common mips generic software interrupt routines.

Changes from Izumi Tsutsui, tested by Havard Eidnes.

__KERNEL_RCSID()

branches: 1.54.2;
Call ksyms_init() instead of ddb_init() in case of
NKSYMS || defined(DDB) || defined(LKM)

Use PAGE_SIZE rather than NBPG.

Oops, make sure to re-enable the IP32 watchdog correctly, else we end up
getting random reboots after returning from DDB.

Parse out flags passed to the kernel on boot; from Scott Taylor, tweaked to
use the MI boot_flags.h interface by me.

Merge the nathanw_sa branch.

Get rid of some extraneous declarations/comments.

Add the MIPS3_CONFIG_SE (name taken from Rm52xx manual) bit, which is the
external cache enable bit -- this allows software to enable or disable the
(external) L2 cache on the R5k and Rm527x and the (external) L3 cache on
the Rm7k. If the (external) cache is disabled, treat it as if there were
no cache for the purposes of the cache setup code.

Also, update sgimips code to use the new name.

There's no cache_r5k.h yet, so don't try to include it.

Postpone DDB/KGDB init until we know the machine type, so the debugger can
be attached to the right place, twiddle the right watchdog registers, etc.

Checkpoint of O2 work by Chris Sekiya and myself. This is the sgimips bit;
still needs some arch/mips support code before it will fully work.

Move console setup after we've determined machine type so the console init
code can guess where to find a console depending on machine type.

Use crimereg.h definitions for the crime dog. Also, move watchdog
disable down the very last possible place in cpu_reboot.

from Chris Sekiya

Fix signed/unsigned comparison warnings.

remove trailing \n in panic(). approved perry.

Don't include <sys/map.h>.

Make nbuf, nswbuf, and bufpages unsigned. Make all operations on these
variables unsigned, and update places where their values are printed.

Merge cpu_sysctl() for all mips ports, based on powerpc and m68k precedent.

For now, only pmax implements CPU_BOOTED_KERNEL. Need to revisit.

branches: 1.37.2;
Update copyright dates.

Fix up clock interrupt accounting for the sgimips port -- make sure
to schedule clock interrupts at a fixed interval, rather scheduling
the next one based on the time of the arrival/servicing of the previous
clock interrupt. Also, pick up a trick from the sbmips port to convert
a division in ip22_clkread to a multiplication, since those are much
cheaper -- the details of that are described in Simon's commit (see
Message-Id: <20020306073437.1D2A8B004@cvs.netbsd.org>). Thanks to
Jason Thorpe and Dominic Sweetman's "See MIPS Run" (where I found
mention of this very subject while looking for something totally un-
related! 8-) for the clue about the source of the timekeeping problems.

For the IP32, where we have no clock-calibration code yet, use the CPU
frequency provided by ARCS instead; it beats a hard-coded value!

As an added bonus, most of the CPU-clock related stuff is now collected
together in cpu_info_store, rather than as a collection of unorganized
global variables.

Fix nasty crashes on IP22 machines with console running at 38.4kbps (which
generally translates to "high interrupt load") -- the old code re-enabled
interrupts in the machine-specific hardware interrupt handler causing the
handler to be re-entered, possible multiple times. Could lead to kernel
stack overflows, and all sorts of mysterious crashes/hangs as a result.

While here, fix up the IP32 interrupt handler code to also not re-enable
interrupts.

Thanks for ideas/comments go to Chuq and Stephen Ma.

Replace lots of 8x<space> with <tabs> and other miscellaneous indentation
fixes.
Wrap a couple of long lines.
Use <return-type>\n<function name> as per KNF in a few places.

Respect OSLoadPartition even if it doesn't come before SystemPartition in argv.
Add a few comments about the above.

- Check for provided bootinfo from the bootloader. If we find it,
save off DDB symbol table information.
- Make loading of memory work reliably with the bootloader; for each
candidate memory type from ARCS (which now includes LoadedProgram),
check to see if the kernel is within that chunk, and load the pages
around it if it is.

Also check SystemPartition when finding boot device. Especially
when netbooting, it is practical to load the kernel directly.

branches: 1.30.2;
When loadfile() leaves the ELF header at the end of the kernel, it enables
erroneous code that tries to compute the end of the symbol table. Disable
that for now so that kernels don't hang when booted from the disk.

Make microtime(9) do interpolation to get better-than-hz resolution, if
the platform supplies a clkread function that does something other than
return 0 (which is the default unless overridden by the platorm code).
Supply such a function for the IP22; even if it isn't perfect, it goes
a long way to making ntp usable.

While I'm at it, move the ticks-per-hz variable out of the struct platform
since it's really private to the per-platform interrupt/clock code.

XXX: No clkread function supplied for IP32, since it has other problems --
like a hardcoded ticks-per-hz, but the same code as on the IP22 could be
used.

branches: 1.28.2;
Use boot arguments to try to determine the boot device. The OSLoadPartition
string, if present, will override the second argument (which may be the
path/kernel being loaded). This will provide a way to netboot the kernel
and allow the root device be set to a disk partition.

Run vfs_shutdown() if NOSYNC is *not* set, rather than set.

If OSLoadOptions is set to auto, boot straight to multi-user. Requested
one too many times and it seems to do no harm, so enable it.

Obey the RB_POWERDOWN flag passed to cpu_reboot; make it work on the Indy
by enabling the FPU before jumping back to ARCS (gross, but necessary!).

Update pmap_update to now take the updated pmap as an argument.
This will allow improvements to the pmaps so that they can more easily defer expensive operations, eg tlb/cache flush, til the last possible moment.

Currently this is a no-op on most platforms, so they should see no difference.

Reviewed by Jason.

branches: 1.23.2; 1.23.4;
Count and display the number of CPUs present on the system.

Fill in cpu_model[]

- Use the identifier strings that are fetched by the MI ARC BIOS
code.
- Garbage-collect some non-useful stuff from the mainbus_attach_args.

bcopy -> memcpy, bzero -> memset

Use sys/dev/arcbios.

- Finish off KGDB kernel debugger support for sgimips

Implement a db_trap_callback which turns off the watchdog on entry to DDB
and turns it back on on exit. Hook the callback in early in the boot, so
it's there as soon as DDB is available.

replace vm_map{,_entry}_t with struct vm_map{,_entry} *.

Adjustments for dynamic system configuration. From Rafal K. Boni.

Do the first allocsys() pass (the pass that computes sizes) before
calling pmap_bootstrap(); pmap_bootstrap() needs some of the size
information computed by allocsys().

Sprinkle pmap_update() calls after calls to:
- pmap_enter()
- pmap_remove()
- pmap_protect()
- pmap_kenter_pa()
- pmap_kremove()
as described in pmap(9).

These calls are relatively conservative. It may be possible to
optimize these a little more.

Call pmap_bootstrap() before calling uvm_pageboot_alloc(), so that
the kernel virtual address space bounds will be set up.

Use uvm_pageboot_alloc() for early memory allocation, rather than
calling pmap_steal_memory() directly. On these platforms, since
uvm_pageboot_alloc() is a wrapper around pmap_steal_memory(), there
is no functional change. This is merely for API consistency.

eliminate the KERN_* error codes in favor of the traditional E* codes.
the mapping is:

KERN_SUCCESS 0
KERN_INVALID_ADDRESS EFAULT
KERN_PROTECTION_FAILURE EACCES
KERN_NO_SPACE ENOMEM
KERN_INVALID_ARGUMENT EINVAL
KERN_FAILURE various, mostly turn into KASSERTs
KERN_RESOURCE_SHORTAGE ENOMEM
KERN_NOT_RECEIVER <unused>
KERN_NO_ACCESS <unused>
KERN_PAGES_LOCKED <unused>

branches: 1.9.2;
Require the machine-dependant DDB commands to be in db_machine_command_table[]
and link it directly to db_command_table[] so that it's not necessary
to do this at runtime. Make db_machine_command_table[] const on all ports.
g/c now unneded stuff, like db_machine_commands_install(), db_machine_init()

Patch written by enami.

Make softclock a generic soft interrupt of the API is available,
adding the requisite void * argument to softclock().

Consistently use pmap_steal_memory() to allocate USPACE.

branches: 1.6.2;
clean up and consistency for CP0 Count, Compare, Wired, and Config
access function names and prototypes.

Add an align argument to uvm_map() and some callers of that
routine. Works similarly fto pmap_prefer(), but allows callers
to specify a minimum power-of-two alignment of the region.
How we ever got along without this for so long is beyond me.

Use cpu_intr() definition from mips/cpu.h instead of local one.

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

remove/move more mach vm header files:

<vm/pglist.h> -> <uvm/uvm_pglist.h>
<vm/vm_inherit.h> -> <uvm/uvm_inherit.h>
<vm/vm_kern.h> -> into <uvm/uvm_extern.h>
<vm/vm_object.h> -> nothing
<vm/vm_pager.h> -> into <uvm/uvm_pager.h>

also includes a bunch of <vm/vm_page.h> include removals (due to redudancy
with <vm/vm.h>), and a scattering of other similar headers.

branches: 1.1.2; 1.1.4;
Initial commit. Needs work.

Remove unnecessary <sys/malloc.h>.

fix typos in "environment" word.

Fix all remaining typos, mainly in comments but also in few definitions and log messages, reported by me in PR kern/54889.
Also fixed some additional typos in comments, found on review of same files or typos.

branches: 1.150.6;
uvm_availmem(): give it a boolean argument to specify whether a recent
cached value will do, or if the very latest total must be fetched. It can
be called thousands of times a second and fetching the totals impacts not
only the calling LWP but other CPUs doing unrelated activity in the VM
system.

Rename uvm_free() -> uvm_availmem().

uvmexp.free -> uvm_free()

Initialize ci_divisor_delay by temporary cpufreq for early delay(9) calls.

This is actually required for Indigo R3k.
Reported by Naruaki Etomi in PR port-sgimips/53520.

Fix build failure without options DDB and with pseudo-device ksyms.

There is a typo for ksyms (NKSYMS vs NKSYM).
Also use consistent "#if NKSYMS" rather than "#if NKSYMS > 0" etc.
and add comments after corresponding #endif for readability.

Fixes PR port-sgimips/53521 from Naruaki Etomi.
Should be pulled up to (at least) netbsd-8.

branches: 1.145.2; 1.145.4;
Cleanup and clarify the ELFSIZE mess:

We now have 2 variables automatically set in elf_machdep.h:

ARCH_ELFSIZE: the size for userland binaries
KERN_ELFSIZE: the size for the kernel binaries

DB_ELFSIZE has been deleted and KERN_ELFSIZE should have always the
same values DB_ELFSIZE used to have.

In sys/exec_elf.h, if ELFSIZE is not set, it is set to KERN_ELFSIZE
for the kernel and ARCH_ELFSIZE for userland. These defaults should
eliminate the need for most manual ELFSIZE setting.

branches: 1.144.6;
kill variable stack allocation

branches: 1.143.2;
switch all ports to use uvm_init.c:uvm_md_init()

uvm_setpagesize() is now subsumed within this funciton.

Prefer modern simple past tense and past participle of catch

The "catched" form is obsolete and nonstandard, prefer "caught".

branches: 1.141.2;
Cleanup includes

branches: 1.140.6;
- use cpu_{g,s}etmodel
- fix unused

- remove unused variables
- use (void) when we don't care about the return value

branches: 1.138.2;
split device_t/softc for all remaining drivers.
replace "struct device *" with "device_t".
use device_xname(), device_unit(), etc.

branches: 1.137.2;
Fix -fno-common fallout.

branches: 1.136.2;
Fix DEBUG build.

#include <sys/bus.h> instead of <machine/bus.h>.

Merge forward matt-nb5-mips64
Adapt to new interrupt/spl framework
Deal with arcbios calls being O32 and kernels being N32/N64.
Need to save/restore T8 across arcbios calls.

branches: 1.133.2; 1.133.4; 1.133.6;
Remove separate mb_map. The nmbclusters is computed at boot time based
on the amount of physical memory and limited by NMBCLUSTERS if present.
Architectures without direct mapping also limit it based on the kmem_map
size, which is used as backing store. On i386 and ARM, the maximum KVA
used for mbuf clusters is limited to 64MB by default.

The old default limits and limits based on GATEWAY have been removed.
key_registered_sb_max is hard-wired to a value derived from 2048
clusters.

branches: 1.132.2;
Fix various build with the 32bit mips ports after the mips64 merge.

Merge from matt-nb5-mips64
Merge mips-specific arch files.

- Use uvm_lwp_setuarea() instead of directly setting address to lwp_t::l_addr.
- Replace most remaining uses of l_addr with uvm_lwp_getuarea() or lwp_getpcb().
- Amend assembly in ports where it accesses PCB via struct user.
- Rename L_ADDR to L_PCB in few places. Reduce sys/user.h inclusions.

Kill proc0paddr. Use lwp0.l_addr instead.

Remove all declarations of physmem from sys/arch. Add an include of
<sys/systm.h> to the one file that did not already contain it.
This now means that physmem can be changed by updating systm.h and uvm_page.c
(excluding fixing printfs)

Ansify function definitions w/o arguments. Generated with sed.

Use "defopt MODULAR" in sys/conf/files, and #include "opt_modular.h"
in all kernel sources that use the MODULAR option.
Proposed in tech-kern on 18 Jan 2009.

Enable Personal IRIS 4D/20 and 4D/25 support:
- Adapt int(4) to handle the INT1 chip
- Move generic rtc clocks out of hpc/ and into dev/
- Handle the very strangely wired eeprom and other bits in arcemu
- Sprinkle MACH_SGI_IP6 as necessary
- Enable IP6/IP10 devices in GENERIC32_IP12. Yes, the naming is poor but
there's no winning with kernel/hw compatibility on sgimips...

Tested on my 4D/25. Doesn't (appear to) break macallan@'s IP22.

branches: 1.124.4;
As discussed on tech-kern: mutex_init is too heavyweight for early bootstrap
phases, so move the initialization of the ksyms mutex back into main via
a function called ksyms_init. Rename the existing (but quite different)
ksyms_init* variations into ksyms_addsyms_elf() and ksyms_addsyms_explicit()
and adapt machdep code accordingly.

Remove LKMs and switch to the module framework, pass 1.

Proposed on tech-kern@.

It is not appropriate to call pmf_system_shutdown(9) from
doshutdownhooks(9): shutdown hooks registered by shutdownhook_establish(9)
expect to be called with interrupts disabled, but shutdown hooks
registered with pmf_device_register1(9) expect to be called with
interrupts enabled. So I have made two changes:

1 Do not call pmf_system_shutdown() from doshutdownhooks(). Instead,
change every call to doshutdownhooks() to a call to doshutdownhooks()
followed by a call to pmf_system_shutdown(). No functional change
is intended by this change.

2 Make i386 re-enable interrupts briefly while it calls
pmf_system_shutdown(). I leave it to others either to fix the
other ports, or to factor out some MI shutdown code, as joerg@
suggests, and fix that. Note that a functional change *is* intended
by this change.

I hope that this patch will stop us from flip-flopping between
calling doshutdownhooks() and pmf_system_shutdown() sometimes with
and sometimes without interrupts enabled.

branches: 1.121.2; 1.121.8;
Two fixes to make -current kernel be loadable by 4.0 bootloader:
- if bootinfo is not passed from bootloader, check argv[1] first
which might be the loaded kernel path specified for bootloader
- when using argv[0] as the kernel path (loaded directly by the PROM)
assume boot partition is 0 (partition a) since the kernel might be
in an SGI volume header partition and it can't be a root partition

Ok'ed by rumble@, and tested on IP22 and IP32.

branches: 1.120.2;
Replce exec_map with a pool. Proposed on tech-kern@, reviewed by chs@.

branches: 1.119.4; 1.119.6; 1.119.8;
Adjust boot device detection code to make netboot work properly:
- make a bootloader pass bootpath which includes device path on ARC BIOS
via BTINFO_BOOTPATH
- make a kernel use BTINFO_BOOTPATH in bootinfo to see boot device first,
then check argv[0] passed from ARC BIOS if there is no valid bootinfo
- check OSLoadPartition variables in the ARCBIOS environment and whole
argv[] arguments for IP12
- initialize mach_type before it's referred
- bump boot version

Tested on IP32 with various kernel/bootloader combinations.

branches: 1.118.6;
Replace bootinfo stuff on sgimips (which has been disabled in bootloader)
with ones used on arc (which has been working for a year) and
enable it in bootloader. Tested on IP32.

Interrupt handling changes, in discussion since February:

- Reduce available SPL levels for hardware devices to none, vm, sched, high.
- Acquire kernel_lock only for interrupts at IPL_VM.
- Implement threaded soft interrupts.

branches: 1.116.2;
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.

use crime_reboot() and mcclock_poweroff() when appropriate.
With this my O2 resets and powers down again even when using crmfb,
apparently it freaks out the ARCS firmware - with crmfb the machine would
just sit there blinking instead of rebooting.

branches: 1.114.8; 1.114.10; 1.114.12;
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.

IP12's PROM passes us a '-a' flag in argv when booting directly from the
disk volume header. On NetBSD this means RB_ASKNAME, however in SGI terms
it appears to mean 'autoboot'.

Add an appropriate kludge to keep us from doing the wrong thing.

XXX - This whole argv/env/prom parameter mess needs to be cleaned up.

A few changes to allow booting to multi-user with IP12 kernels squished
into oversized volume headers:

Query ARCBIOS for OSLoadFilename and OSLoadOptions first, rather than only
using the environment variables passed since we want to go through arcemu.

Add an ugly hack to arcemu that munges the nvram 'bootfile' variable. We
need this because bootfile will point to the volume header, which then
translates into a bogus root device of 'sd0i', or similar.

Clean up extracting environment variables passed by the prom. No functional
change.

branches: 1.110.2; 1.110.4; 1.110.10;
Use (char *) to appease -Wpointer-arith.

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

TRUE -> true, FALSE -> false

Since IP12 doesn't use ARCS, we could read the console setting from the
nvram, but it may not be valid. For example, if the console is set to
graphics, but no keyboard is attached, we should not attempt to use a
graphical console.

To fix this, we extend arcemu to parse the prom's environment array, which
is the third argument to the kernel. If the console mode is graphics, but
no keyboard is found, the 'gfx' environment variable will contain 'dead'.

Note that since IP12 has no bootloader, we assume that the third parameter
is the environment table, not bootinfo.

branches: 1.106.2;
Merge newlock2 to head.

Fix typo in comment.

Move the mips1_clock_intr handling into int.c where it belongs.

IP12 (and perhaps other systems) appear to be buggy and incapable of
reliably supporting badaddr(). False negatives appear to occur
approximately 1.8 percent of the time, although neither false positives
nor consecutive false negatives occur. We take advantage of the latter
property and always use a wrapper that makes multiple checks.

My IP12 no longer sees occasional ghost devices and related panics during
boot.

Handle fpu interrupts (MIPS interrupt 0) on IP12.

Specify the member names of struct platform for clarity.

Guiness -> Guinness.

merge yamt-splraiseipl branch.

- finish implementing splraiseipl (and makeiplcookie).
http://mail-index.NetBSD.org/tech-kern/2006/07/01/0000.html
- complete workqueue(9) and fix its ipl problem, which is reported
to cause audio skipping.
- fix netbt (at least compilation problems) for some ports.
- fix PR/33218.

branches: 1.98.2;
sgimips converted to timecounters. Systems that can do them get mips_cp0
based logic. We also use the common mips3_clock functions for those systems,
removing some port-specific code.

Tested by martin@.

branches: 1.97.2;
Fix delay() so that gcc4 does not optimize it into a noop. Lifted from the
newsmips implementation by tsutsui@

Remove declarations for cpu_model[] since it's in <sys/systm.h>.

branches: 1.95.4; 1.95.6; 1.95.8; 1.95.10; 1.95.12;
merge ktrace-lwp.

Restore missing call to r5k_enable_sdcache(). Its' removal was a thinko
on my part -- the COP0 SC presence test fails on r5k IP22, but
r5k_enable_sdcache() will DTRT regardless.

Pointed out by tsutsui@

branches: 1.93.2;
Avoid a global name "netmask" - it causes lots of shadow warnings all over
the place. Collect all interrupt masks into a single array instead.
Constify a bit and fix a bogus argument to config_rootfound().

Move the MI printing of `copyright' to the MD cpu_startup() code
where the printing of `version' is already performed.
This has the benefit of allowing the copyright to be available
via dmesg(8) on platforms which need the `msgbuf' to be setup
in cpu_startup() before printed output is remembered.

mips_sdcache_forceinv has been removed.

XXX Should we check MIPS3_CONFIG_SE in mips3_cp0_config_read() and
XXX call r5k_enable_sdcache() accordingly here?

branches: 1.90.2;
Set mips_sdcache_forceinv to 1 for r5k processors.

We now support secondary cache operation on r5ksc out-of-the-box.

branches: 1.89.2; 1.89.4;
Nuke MIPS64; we'll use MIPS3 to indicate the presence of MIPS3 instructions
and _LP64 to indicate that 64-bit is desired.

Pointed out by simonb@

MIPS3_CONFIG_SC is an unreliable indicator for SC presence; r5k_enable_sdcache()
will gracefully exit if there is no secondary cache. Rework #ifdef'd out
r5k cache code to take this into account, and remove pointless r10k case.

Use the mips3 bus functions on IP30. Recognize the IP30 as a supported* SGI
platform.

* For extremely small values of "supported". arch/mips/* needs a lot of work
yet for MIPS64.

s,#ifdef MIPS3,#if defined(MIPS3) || defined(MIPS64),g

KNF some and make printf output a bit nicer.

Add IP12 stuff. Most bits included are from Steve Rumble.

After this change it is possible to run IP12.

If argv[1] doesn't exist, don't try to use it.

Tidy up total/avail memory message as well as other ports.
As per discussion on port-sgimips.

branches: 1.81.2;
Sprinkle some #ifdef MIPS3 to be able to compile the port without
including MIPS3 code.

for upcoming IP12

Print an informative message when unimpl_intr() is invoked, rather than panic().

Completely remove IP2X and IP32, which renders opt_machtype.h obsolete.

g/c obsolete function.

Attach int0 at mainbus (precedent being the ip32 CRIME device, which handled
interrupts and timer calibration yet is also attached at mainbus).

Introduce the INDY_R4600_CACHE config option, which more accurately describes
the code enabled by this option.

This renders #ifdef IPxx completely obsolete, thus we theoretically can unify
the GENERIC files if a workable load address relocation scheme can be found.

Remove another #ifdef IP32

Large rototill of the lowlevel code:

* move all device-specific bitbanging to the device drivers in question --
i.e., the ip32 init code doesn't access the CRIME registers directly but
instead trusts the CRIME driver to DTRT. machdep.c is cleaner thereby.
* introduce int.c, which handles the interrupt-controller goo formerly
in ip2x.c.
* unify the CPU interrupt handler, with CRIME, INT2, and the various timer
handlers hooked in via pointers in the platform struct.
* eliminate ip2x.c and ip32.c. After the interrupt handler cleanup, the
actual platform-dependent code was so minimal (rev identification, mask level
initialization) that it made sense to move it into machdep.c
* "#ifdef DEBUG printf("foo"); #endif" becomes aprint_debug.

Tested on my ip22 and ip32. Discussed with pooka@, rafal@.

Call sgimips_bus_dma_init() in mach_init to set up the proper MIPS1/MIPS3
bus_dma_sync() routine.

Add a call to ip1x_init().

Wrap more mips3_ operations in MIPS3 defines.

Streamline bootinfo logic.

Rework the memory detection routine to handle cases where kernel doesn't
fit inside one memory chunk.
Leave 1 page before kernel untouched as that's where our initial kernel
stack before we switch to proc0 stack is (fixes boot problems on my Indy
with small kernels).

Add trailing newline. Noticed by Steve Rumble.

Replace the traditional buffer memory management -- based on fixed per buffer
virtual memory reservation and a private pool of memory pages -- by a scheme
based on memory pools.

This allows better utilization of memory because buffers can now be allocated
with a granularity finer than the system's native page size (useful for
filesystems with e.g. 1k or 2k fragment sizes). It also avoids fragmentation
of virtual to physical memory mappings (due to the former fixed virtual
address reservation) resulting in better utilization of MMU resources on some
platforms. Finally, the scheme is more flexible by allowing run-time decisions
on the amount of memory to be used for buffers.

On the other hand, the effectiveness of the LRU queue for buffer recycling
may be somewhat reduced compared to the traditional method since, due to the
nature of the pool based memory allocation, the actual least recently used
buffer may release its memory to a pool different from the one needed by a
newly allocated buffer. However, this effect will kick in only if the
system is under memory pressure.

Fix typo ("ARCBIOS_MEM_ExecptionBlock"). From Steve Rumble.

Part two in the Great Renaming. ip22.c becomes ip2x.c, and all generic ip2x
routines become ip2x_*.

Change all IP22 preprocessor conditionals to IP2X preprocessor conditionals.

Part one of two -- part two will be renaming ip22_*() to ip2x_*().

IP20 should handle the watchdog timer in the same manner as for IP22.

For the purposes of machdep.c, consider IP20 to be a variant of IP22. ip20.c
and #ifdef IP20 will go away (short-term), ip22_* will be renamed to ip2x_*
(slighly longer-term).

The ARCBIOS->FlushAllCaches() vector is NULL on at least one SGI platform
(ip20) and has no actual operational effect, thus it should not be invoked here.

Remove the extra call to consinit().

www.netbsd.org -> www.NetBSD.org

Apply a bunch of patches written by Christopher SEKIYA
(with several cosmetic changes by me) which fixes O2 (IP32) support.

Now my R5000 O2 works fine in multiuser with on-board AIC7880 SCSIs
and several PCI network cards (but only on serial console yet).
L2 cache on R5000/Rm5200 is still disabled for now, but it will be
fixed later, hopefully.

See recent discussion on port-sgimips for details.