- remove sc_ih_clk, sc_ih_fmn, sc_ih_ipi from struct rmixl_cpu_softc;
they were unused just taking up space
- rmixl_intr_init_clk() and rmixl_intr_init_ipi() are now type void

(cliff)

Add QUOTA2 where QUOTA is enabled (and QUOTA2 commented out where QUOTA
is commented out)

(bouyer)

- rename ddb_running_on_this_cpu to ddb_running_on_this_cpu_p
according to pedicate unction naming style convention

(cliff)

Fix pasto, use GETSYSCALLS() where possible

(pooka)

- improve formatting: a space char followed by tab char follows "options"

(cliff)

Sync with HEAD

(bouyer)

- remove erroneous include of opt_multiprocessor.h
("MULTIPROCESSOR" instance was inside a comment)

(cliff)

newer pcc front end provides some definitions according to arch,
so provide the TARGMACH definition to the compiler here.

(makes no difference to the currently imported sources, but
will be easier for people wanting to try newer sources)

(plunky)

Sync with HEAD

(bouyer)

use built-in macro ___STRING() rather than rolling our own

(plunky)

Sync with HEAD

(bouyer)

Sync with HEAD

(bouyer)

Sync with HEAD

(bouyer)

Q2V -> QL

(bouyer)

Q2V_ -> QL_ rename

(bouyer)

ffs_clusteracct() moved to ffs_subr.c, so remove it from this copy of
ffs_alloc.c (otherwise we have duplicate symbols at link time).
Always use UFS_FSNEEDSWAP(fs) even if !KERNEL; so we can use for makefs too.
other !KERNEL consumers will have UFS_FSNEEDSWAP(fs) == 0 anyway.

(bouyer)

Make sure we can do host kevent since the -current (and recent nb5)
libc resolver uses it.  Error out in case of rump fd kevent (TODO).
Fixes one more problem pointed out by Alexander Nasonov.

Also, implement dup().
(TODO: implement it along the fcntl path too)

(pooka)

Add missing debug keyword for libdm.so.0.0.debug.

(njoly)

Use PRIu64 to fix build on amd64 (and presumably other 64-bit ports)

(pgoyette)

abuse config_interrupts to make sure we don't switch to 'intr' mode until
interrupts are enabled

(jmcneill)

add support for command buffering in vcons. still a WIP, same limitations
as async drawing (no ddb, needs interrupts). you can try it with options
VCONS_DRAW_INTR. as with async, there are still occasional glitches.

(jmcneill)

FLOPPYPAD + ${MACHINE} instead of hardcoding amd64.

(jym)

Added LevelOne NASCon and some Nettronix based devices.

(phx)

add std dprint to fdcall

(pooka)

Turn the "idle workers available" from a flag into a counter so
that we can finally avoid all races (*knock knock*).

(pooka)

Disable more file-system and pseudo-devices to shrink install floppies.

(tsutsui)

add a 'setmode' callback to genfb and use it to setup write-combining
MTRRs on x86 whenever switching to WSDISPLAYIO_MODE_EMUL

(jmcneill)

Build fix.

(rmind)

oh no, more tests

(pooka)

Time to start adding tests for the routing code to make that part
of the kernel more approachable.

Begin the task with an xfail test for PR kern/40455.

(pooka)

Fix typos.

(wiz)

Improve a bit.

(wiz)

Improve a bit.

(wiz)

New sentence, new line. Bump date for previous.

(wiz)

add PS/2 registers from synaptics spec:
PMS_RESET_WRAP_MODE, PMS_SET_WRAP_MODE, PMS_ACK, PMS_ERROR and PMS_RESEND

while here use #define<tab> consistently.

(cegger)

Add IPI/cpu_send_ipi to intrsw.

(matt)

Add MPC85xx variants.
Move to having a common kernel for all MPC85xx variants (CADMUS/PIXIS still
remain a problem).

(matt)

Allow code to allow inclusive/exclusive locators for subdevices.
(uses on mpc85xx to match on SVRs or not match on SVRs).

(matt)

Add P2020/P2010 truth table definition.
Remove trailing whitespace.

(matt)

Add more MPC85xx SVRs.
Fix/add debug register definitions.

(matt)

Add some P2020 / MPC856x definitions.

(matt)

Add MPC8555/41, MPC8568/67, and P2020 variations.

(matt)

Powerpc send_ipi use ci_cpuid, not cpu_index

(matt)

- add config_queues_lock to serialize access to deferred_config_queue and
interrupt_config_queue for adds, deletes and (those DIAGNOSTIC) searches
- add dc_funcs_lock to serialize dispatch of deferred config functions
many (all?) of which may not be MP safe.

(cliff)

- distinguish between cpuid and cpu index.

(cliff)

- eliminate assumptions that cpuid 0 is primary
- distinguish between cpuid and cpu index.

(cliff)

- cpucore_rmixl_attach() now examines 'userapp_cpu_map' bitmask obtained
from the firmware, and avoids attaching threads that are not enabled there.

(cliff)

- renamed cpu_rmixl_attach_once() to cpu_rmixl_attach_primary() (more clear)
- cpu_rmixl_attach_primary() obtains ci_cpuid from COP0 EBASE reg
instead of assuming always cpuid==0
- in cpu_rmixl_attach(), first call of the function determines we are attaching
primary cpu, instead of asuming core0 thread0 is always the primary cpu
- cpu_fmn_intr() uses CPU_IS_PRIMARY() instead of asuming cpu_number()==0
always indicates the primary cpu
- debug function rmixl_cpuinfo_print() argument is now a cpu index, not cpuid

(cliff)

- add cpuid column to table displayed by cpu_debug_dump()

(cliff)

- re-define cpu_number() to now mean ci_index instead of ci_cpuid
- re-define CPU_IS_PRIMARY() to use CPUF_PRIMARY instead of ci_cpuid

(cliff)

tmplen should be static

(sjg)

Switch libdevmapper library to use libdm as it;s backend lib and do not
communicate with device-mapper directly. Our lvm stack now looks like this
lvm<->libdevmapper<->libdm<->dm where only libdm knows how our dm protocol
looks like.

No objections on tech-userlevel@.

(haad)

Initial import of libdm and dmctl to tree. libdm library can be used
to access, manage and manipulate device-mapper driver. Which opens us bunch
of new possibilities like

dm-multipath device target
dm-crypt    device target
dm-ccd     compatibility layer

With this import I'm bringing in dmctl tool for working with dm driver ,too.
I plan to replace gpl2 licensed dmsetup command with our dmctl tool in near
feature. It can also by placed to /rescue where we was not able to put
dmsetup because of licensing problems.

With libdm in tree we can now write RUMP atf test suite for dm driver to
ensure LVM subsystem stability as time goes.

Reviewed by: blymn@ and rmind@
Oked: by no objections on tech-userlevel@

(haad)

Distinguish between a corrupt obj pointer and a null obj pointer.

(matt)

First pass at new crtstuff for powerpc.  Doesn't quite work yet
(obj parameter to ___start seems to get corrupted sometimes)

(matt)

BookE needs PMAP_NEEDS_PROCWR (for ptrace breakpoints)

(matt)

If we get a PGM exception for a trap, use SIGTRAP/TRAP_BRKPT instead of
SIGILL/ILL_ILLTRP so GDB will recognize it.

(matt)

- add preliminary code to support D-Link DSM-G600 rev.B NAS.
- it features 9600bps console, Moto MCU on the 2nd DUART channel,
  ACARD ATP865 PCIIDE and IP Plus IP1000A GbE.  It has MiniPCI
  slot.

(nisimura)

weprovidedarracount()sothatwedon'tneedtosqueezethespacesout.

(christos)

Add some macros to access GPMC config registers easily.

(ahoka)

add some notes on access control

(pooka)

Don't define weak aliases in the SMALL case. They conflict with the real
function name since SMALL doesn't use the normal rename logic.

(joerg)

add missing SRCDIR for synaptics.

(mrg)

Test clearing of quota entries

(bouyer)

Add -c option, which clears quota entries for the specified users/groups

(bouyer)

Implement clear command (quota2 only), which either free the specified
quota2 entry (if both disk and inode usage are 0) or revert its limits to
the default quota entry.

(bouyer)

Limit test duration based on wall time instead of loops
(took a long time on slow platforms).

(pooka)

Force gcc to generate a stack frame for the call to dlsym(RTLD_NEXT).
Without this hack at least amd64 -O2 just used jmp and The Wrong
Thing happened.

(pooka)

StorCenter models are called 250, 500 and 1T.

(phx)

Retire legacy grfinfo and hilinfo derived from HPBSD
since finally hp300 got wscons.

(tsutsui)

PR/44529: Martin Danielsson: fsck_msdos crashes when verifying corrupt file
system. Avoid using the long name index when it is 0. Refactor the code to
avoid duplication.

(christos)

Create a WAPBL transaction when setting quotas.

(bouyer)

Add an option to mount the filesystem -o log to rump quota servers.
Add some quota tests on a WAPBL-enabled filesystem.

(bouyer)

- add D-Link GSM-G600 info.  IP1000A is supposed a member of stge(4).
- move StorCenter info as it looks correct.

(nisimura)

malloc/free aren't async-signal-safe, so avoid calling them when
signals aren't blocked.

this bug made tests/rump/rumpkern/t_sp:sigsafe rarely deadlock

(pooka)

make sure we don't use "fullset" before we init it

(pooka)

types.h for time_t

(pooka)

add StorCenter heuristic by PCI vendor ID detection.

(nisimura)

Resurrect and move retired ITE files that are still used by bootloaders.

(tsutsui)

add Dallas I2C RTC driver for StorCenter, leaving it commented out for now.

(nisimura)

add IOMEGA StorCenter fixup codes.  Need more change as the satellite
microcontroller protocol is defined in 8-byte fixed length sequence.

(nisimura)

duh, _sys_read, not read.  STAY FIXED, DAMNIT!

(pooka)

Fix PR/44500 for hppa.

(skrll)

Unbreak the ssp lossage from the default -current build by removing
it.  I still don't have any idea what the ssp stuff is supposed to
do and how it's supposed to even begin to work.  If someone wants
to change this now, run tests/lib/librumphijack before commit so
that I can avoid another multihour debugging session!

(pooka)

call the non-compat pollts() from inside the library

(pooka)

This is no longer broken for macppc.

(matt)

PR port-macppc/43619
Save handler before calling sendsig_reset before it will reset to SIG_DFL
is SA_RESETHAND is set.

(matt)

not here yet either (mirror amd64 commit)

(pooka)

Use EVCNT_ATTACH_STATIC

(matt)

Fix inverted test.  cpu_setmcontext should return success. :)

(matt)

binutils is automagically doing securecrt so make sure plt calls are emitted
with the secureplt magic as well.

(matt)

options SPLASHSCREEN_PROGRESS is no more

(jmcneill)

move macros for validating fs/gs to segments.h and use them
in the linux32 code as well.

(chs)

Third argument of MODULE() is typically a string literal.
Moreover it (at least sometimes) didn't build otherwise.

(jakllsch)

Make this build.

Also, the previous commit contanined "unintended" functional changes I'm
going to ignore.

(jakllsch)

%d -> %zd for size_t printf format

(jmcneill)

KNF, no functional changes intended.

(christos)

Fix typo, from Henning Petersen in PR 44513.

(dholland)

provide  __arraycount() if not there.

(christos)

- add support for using compressed images as splash images
- retire SPLASHSCREEN_PROGRESS and SPLASHSCREEN_IMAGE options

(jmcneill)

add support for passing image files to the kernel with the 'splash' keyword:

vesa on;splash /netbsd.jpg;boot -z

(jmcneill)

add BI_MODULE_IMAGE boot module type

(jmcneill)

add a public domain image decoder

(jmcneill)

Regen for MPMan.

(njoly)

Add MPMan to the vendor list.

(njoly)

don't use __arraycount in tools

(pooka)

Fix usage().
Document new behavior and options.

(bouyer)

update macro names to match new ones
(i don't know why they changed, just fixing the build)

(pooka)

Support query of the partition size in case underlying device is
a wedge.  This still lacks the proplibistic query of the sector
size and just assumes 512.  It's good that we make asking a file's
size as simple as requiring one stat(), one open() and three (3)
different ioctls plus some proplist mumbojumbo.  I'm surprised it's
available at all by means other than #wish.

code mostly from Taylor R Campbell, rant from me.

(pooka)

Fix the 'test' target to force atf to use its own files from DESTDIR instead
of the installed ones.  This is achieved by overriding all ATF_* environment
variables and making them point inside DESTDIR.

(jmmv)

Add missing DPADD on libatf-c.a.  Spotted by Garrett Cooper.

(jmmv)

Change format to print the quota version for each filesystem in editor.
For version 1, have '-d' edit the filesystem's grace times.

(bouyer)

Delete unnecessary casts to void *. No functional change intended.  Same
assembly generated before and after this change.

(dyoung)

Call daemon() later to make sure there is a listening socket by
the time the foreground process exits.

discussed with mrg.

(pooka)

respond to L2CAP Information requests

(plunky)

librumphijack build infra

(pooka)

Add a simple test for the purpose of making sure rumphijack works
on -current in addition to 5.x.  The test serves a simple index.html
on a hijacked bozohttpd and checks the file can be retrieved.

(pooka)

Switch NetBSD/hp300 to wscons with rasops. Simply ported from OpenBSD/hp300.

- Only A1416 Kathmandu (topcat) framebuffer on 425t is tested, but
  all other variants (TigerShark, Hyperion, DaVinci, GatorBox, Renaissance)
  should also work if they are working on OpenBSD/hp300.
- sti(4) and SGC bus support are not pulled because I don't have 425e
  and I can't confirm that 362 and 382 actually have SGC bus.
  (I'll commit a DIO based dumb driver for 362 and 382 framebuffers later)
- Xorg server with wsfb driver will also be integrated soon
  once after keycode with NoSymbol problem is addressed.
  (We have to re-think what code should be used on WSDISPLAY_COMPAT_RAWKBD)
- MI HIL keyboard and mouse drivers are working fine though
  cngetc via hilkbd has some problem (still we can input commands).
- No old HP-UX like HIL ioctl compatibility (we removed COMPAT_HPUX anyway).
  grfinfo(8) and hilinfo(8) will be removed shortly.

Demonstrated on NetBSD booth at Open Source Conference 2011 Kagawa.

(tsutsui)

Note NetBSD/hp300 wscons.

(tsutsui)

Initially generate hildevs_data.h from hildevs.

(tsutsui)

Add NetBSD's RCS Id.

(tsutsui)

use unsigned char instead of uint8_t, easier for portability

(christos)

Switch NetBSD/hp300 to wscons with rasops. Simply ported from OpenBSD/hp300.

- Only A1416 Kathmandu (topcat) framebuffer on 425t is tested, but
  all other variants (TigerShark, Hyperion, DaVinci, GatorBox, Renaissance)
  should also work if they are working on OpenBSD/hp300.
- sti(4) and SGC bus support are not pulled because I don't have 425e
  and I can't confirm that 362 and 382 actually have SGC bus.
  (I'll commit a DIO based dumb driver for 362 and 382 framebuffers later)
- Xorg server with wsfb driver will also be integrated soon
  once after keycode with NoSymbol problem is addressed.
  (We have to re-think what code should be used on WSDISPLAY_COMPAT_RAWKBD)
- MI HIL keyboard and mouse drivers are working fine though
  cngetc via hilkbd has some problem (still we can input commands).
- No old HP-UX like HIL ioctl compatibility (we removed COMPAT_HPUX anyway).
  grfinfo(8) and hilinfo(8) will be removed shortly.

Demonstrated on NetBSD booth at Open Source Conference 2011 Kagawa.

(tsutsui)

Fix a race condition in the worker thread caching logic: if we got
two or more syscall requests before any worker thread ran, we might
not have enough threads to handle the requests.  In some scenarios
this could lead to a deadlock.

(pooka)

wm_gmii_reset(): restore generic reset delays to what they were before
rev 1.186. This makes the following hardware find its PHY again, and I can't
see how these larger delays could be a problem for other hardware:

wm0 at pci6 dev 7 function 0: Intel i82541GI 1000BASE-T Ethernet, rev. 5
wm0: interrupting at ioapic2 pin 0, event channel 5
wm0: 32-bit 66MHz PCI bus
wm0: 65536 word (16 address bits) SPI EEPROM
wm0: Ethernet address 00:13:72:54:ee:13
igphy0 at wm0 phy 1: Intel IGP01E1000 Gigabit PHY, rev. 0
igphy0: 10baseT, 10baseT-FDX, 100baseTX, 100baseTX-FDX, 1000baseT, 1000baseT-FDX, auto

(bouyer)

make it possible to specify client connection retry model in
RUMPHIJACK_RETRY

(pooka)

Add another connection retry model which simply does exit(1) if
the connection is severed.
(mostly for tests so that everything can be hooked to rump_server's exit)

(pooka)

Make sure to close parent's server communication socket after fork.

(pooka)

Revert my fix.

(haad)

Remove trailing TABs.

(tsutsui)

be kinder about kqueue()
(but paradoxically omit the surprise)

(pooka)

Put -l compatibility note in STANDARDS section.

(wiz)

Add -l option to copy a tree as links.
Non-standard option similar to Gnutools cp(1)
Approved by core.

(darcy)

Do Skip device checks, which ends up parsing fstab, when working on a
filesystem image with -F option set.

(njoly)

- adjust a comment for frodo (362 doesn't have frodo)
- whitespace nits

(tsutsui)

Regen from diodevs rev 1.13:
> Pull description about TurboVRX ("tigershark") display from OpenBSD.

(tsutsui)

Pull description about TurboVRX ("tigershark") display from OpenBSD.

(tsutsui)

RENASSIANCE -> RENAISSANCE

(tsutsui)

Regen from diodevs rev 1.12:
> RENASSIANCE -> RENAISSANCE

(tsutsui)

RENASSIANCE -> RENAISSANCE

(tsutsui)

add a directory which i forgot to add with sys/modules/Makefile rev.1.60

(yamt)

Add the actual function files. From FreeBSD via Stathis Kamperis

(christos)

Putter depends on puffs add missing dependience.

(haad)

Add LLVM and clang build system.  It is enabled by MKLLVM=yes.
Upstream sources can be fetched by running "make checkout" in
src/external/bsd/llvm, they will be properly imported once the
integration and missing features are sorted out.

(joerg)

Don't use -nostdinc or -nostdinc++ if this is included in a HOSTLIB.

(joerg)

Allow HOSTLIB to use C++ by not specifying -nostdinc++.

(joerg)

remquo{,f} from FreeBSD via Stathis Kamperis

(christos)

not there yet.

(christos)

fix build!

(christos)

When restoring the screen ("DPMS On") in FFBDacDPMSMode(), use the saved
state to restore the timing generator control register.  This restores the
correct hsync and vsync values, and fixes a bug that caused the screen not
to be restored for graphics modes where vsync is normally disabled.

(jdc)