Include <linux/spinlock.h> in <drm/drm_mm.h>.

(riastradh)

Add <drm/drm_os_netbsd.h> and include it in <drm/drmP.h>.

(riastradh)

Include <linux/kernel.h> and <linux/types.h> in <drm/drm_hashtab.h>.

(riastradh)

Define __user to be empty in <linux/kernel.h>.

__user qualifies a pointer into userspace for static analyzers.

(riastradh)

Dumb implementation of a subset of Linux's idr abstraction.

An idr maintains a mapping from integers to data, like file descriptors
to struct file pointers.

(riastradh)

Avoid BITS_PER_LONG in drm_bufs.c.

(riastradh)

Add initial draft of Linux list and hlist to <linux/list.h>.

Implemented in terms of TAILQ and LIST, respectively.
Not yet tested.

(riastradh)

Add type-unsafe container_of to drm2's <linux/kernel.h>.

(riastradh)

Implement Linux spin lock in terms of kmutex_t in drm2's <linux/spinlock.h>.

Currently a spin lock will be an adaptive mutex, until I identify
cases in drm2 that actually require spinning and interrupt deferral.

(riastradh)

Define __printf in drm2's <linux/kernel.h> shim.

(riastradh)

Define dma_addr_t to be bus_addr_t in drm2's <linux/dma-mapping.h> shim.

(riastradh)

Implement atomic_t and atomic_read in drm2's <linux/atomic.h> shim.

(riastradh)

Add types u8, u16, u32 to drm2's <linux/types.h> shim.

(riastradh)

Make stub Linux header files visible to drm2.

(riastradh)

Add stubs for Linux header files for drm.

(riastradh)

Add drm2 module makefile.

(riastradh)

sync with HEAD and actually get the drm2 import

(riastradh)

sync with HEAD

(riastradh)

Another codegen bug that fixes (in this case lots of) atf tests.
rump_server doesn't die so much now.

(skrll)

Add revision to previous

(skrll)

Use . for shm bus path and rump_server url instead of /tmp
Exit after first reply in mpls_fw

(kefren)

Move all shm files from /tmp into . so ATF can automatically do cleanup
for us. Ok: kefren

(martin)

Add a test for RFC4182

(kefren)

Implement RFC4182 changes - switchable via sysctl

(kefren)

Restrict the hack for offset BAR's to x86.
A cleaner fix requires a rewrite of puc.c, but not today.

(soren)

Compile tests/lib/libc/sys/t_msgrcv with -O0 on hppa to avoid compiler
bug.

(skrll)

Fix debug code to not use uninitialised variables.

(skrll)

In recv, align data after ether_header at _ALIGN_BYTES.

(pooka)

Revert the previous commit.

(alnsn)

Move to motorola syntax
Change all branches to use j.. instead of b.. (branches now use byte offsets)
(this causes it to go from 932 to 848 bytes of text)
Add GETCURPCB macro.
(binary identical before the b -> j changes)

(matt)

Oops.

(soren)

Unbreak the build - soren, please review!

(martin)

Bump date for previous.

(wiz)

Use the standard i2c address to DIMM slot mapping to show the slot number.

(soren)

PUCCN is now implicit.

(soren)

Missing AB_QUIET newlines.

(soren)

Since puc(4) serial ports are typically not identified in the
BIOS COM[1234] table, the I/O address must be manually set using
installboot(8) in order to enable a serial console.
Print the address here so the user doesn't have to dig through
PCI configuration space to find it.

Enable suspend/resume hooks.

(soren)

Allow console on com_puc without a compile-time option so that PC servers
can become headless after the first reboot (sadly, e.g. Intel AMT presents
as a com_puc, but doesn't appear in the BIOS serial port table, so you need
a keyboard and monitor to install and set the installboot parameters first).

Fix com_puc console on devices with offset BAR's.

(soren)

The SB800 SMBus controller has four selectable SDA lines.
Expose them as four iic busses.

(soren)

Use <major>.<minor> for NetBSD version.

(njoly)

Support Linuxen where libc ioctl has cmd as int unstead of unsigned long.

(pooka)

TIMEVAL_TO_TIMESPEC (for musl)

from Justin Cormack

(pooka)

Both increment_overflow() and normalize_overflow() have two different
declarations. One must be *32(). Fixes compile error observed on sparc64.

(msaitoh)

In the non-MP case, just initialize onproc to NULL.

(matt)

If not MULTIPROCESSOR, just make cpu_tlb_info(ci) return &pmap_tlb0_info

(matt)

use mvz[wb] when possible

(matt)

Use motorola syntax
Adjust for coldfire
(no binary difference for non-coldfire

(matt)

When reading from /dev/random, block at most once in cprng_strong.

We are not obligated to return exactly as many bytes as requested,
and many applications -- notably those that use stdio or otherwise
buffered I/O to read from /dev/random -- try to read many more than
32 bytes at a time from /dev/random even if all they are about to use
is 32 bytes.

In this case, blocking until we have enough entropy to fill a large
buffer causes needless application delays, e.g. causing cgdconfig
(which reads from /dev/random with stdio) to hang at boot when trying
to configure a random-keyed device for swap.

Patch tested by Aran Clauson.  Fixes PR kern/48028.

(riastradh)

Fix build with stack-protection enabled.

(tron)

Kill a few trailing spaces.

(njoly)

Merge logic from ulfs_close(), ulfs_getattr(), and ulfs_strategy()
into the preexisting lfs_*() versions of these functions, and delete
the unused ulfs copies.

(dholland)

oops, spell b_bcount properly

(dholland)

In spec_strategy, if fscow_run() fails, set b_resid along with b_error
to avoid panic in biodone. Noticed by riastradh.

(dholland)

Remove ulfs_mknod, which is not used.

(dholland)

Merge ulfs_mknod into lfs_mknod, which was missing some bits.

(dholland)

Use Mt for email addresses.

(wiz)

Use Mt for email addresses.

(wiz)

fix another defective C&P makefile

(dholland)

why does rump have to have its own private build system?

(dholland)

G/C unused pieces.

(dholland)

Collect the pieces of lfs rename into lfs_rename.c, and sprinkle static.

(dholland)

Fixes when compiling against musl libc.

from Justin Cormack via private email

(pooka)

Remove, in do_sys_renameat(), wrong KASSERTs that check for non NULL
from/to arguments. Such values are correctly handled by later
pathbuf_maybe_copyin() calls, that will fail with EFAULT.

ok from dholland@.

(njoly)

Mention rmind-smpnet branch.

(rmind)

Initialise destination before performing stat structure conversion.
Avoid getting garbage for structure members not set explicitely.

(njoly)

revert socklen_t to int (from dh)

(christos)

don't connect on first hello, there are chances that ours is not seen yet
setproctitle with ldp id - useful for rump kernels testing
fix a memory leak in ldp_peer_new
don't holddown if already holded down
peer sockets are now non-blocking
connected routes deletes are now processed
check if peer is connected before attempting to sending label mappings

(kefren)

Note addition of non-kernel coldfire support.

(matt)

Don't use the fs* variants, just use the f* ones.

(matt)

gmp, binutils, gcc, gdb configury for coldfire

(matt)

Match m5407--netbsdelf*

(matt)

Even if we aren't making a crossgdb still make the object dir for it.

(matt)

Add a M5407_PATTERN

(matt)

Match m5407-*-*

(matt)

Add m5407--netbsdelf* support

(matt)

Use m5407--netbsdelf for coldfire (Why not just use m5407--netbsd?  Because
gas will think that's a.out since internally it expands it to m68--netbsd
and that matches a.out)

(matt)

Add an initialization function of DS1287A RTC and call it during attach.

It looks the boot firmware doesn't initialize the control registers
of DS1287A (while it resets NVRAM settings) and RTC oscillator is not
started properly after replacement.

Now my LUNA-II can keep RTC properly even after reboot.

(tsutsui)

With these changes, they now produce identical binary as before the coldfire
changes.

(matt)

Properly initialize the palette for while on black even on 1bpp framebuffer.
Preparing for demonstration in Open Source Conference 2013 Kyoto.

(tsutsui)

Configury changes to match tuple of m5407--netbsdelf as a variant of
m68k--netbself but with a coldfire cpu (m5475).  Fix m68010 to actually
properly default to m68010.

(matt)

Allow alias of m5407--netbsdelf for m68k--netbsdself

(matt)

Allow alias of m5407--netbsdelf for m68k--netbsdself

(matt)

Fix broken matching for m5200 and m5407 as alias for coldfire cpus.

(matt)

Allow stripping of the m before m5... so that it can match coldfire cpus.

(matt)

Make these consistent.  Make sure all handle earm and coldfire.

(matt)

fix error handling.

(christos)

PR/48063: Justin Cormack: incorrect definition of socklen_t

(christos)

Import dhcpcd-6.0.3

(roy)

Commit the patches in Wolfgang Stukenbrock's PR bin/41896.

(1) Do not serve UIDs below MINUID in master.passwd.* maps,
just like in passwd.* maps.

(2) Since stdethers(8) and stdhosts(8) strip comments, we do
not have to do it here, much less inconsistently.

(hauke)

Enable opt-out from shlib mapfile creation.  Not all linkers (namely
the solaris one) support such fancy features.

no default functional change

(pooka)

Sort.

(wiz)

regen

(XXX: I don't think these files were previously regen'd with the
method documented in configure.ac, and therefore the diffs are now huge)

(pooka)

From Rainer Weikusat <rweikusat@mobileactivedefense.com>: Export phase1
remote address as Radius Calling-Station-Id.

(tteras)

Fix some tests for musl libc -based systems.

from Justin Cormack and Jens Staal

(pooka)

Add the following options
-A Display the FQDN of each address on all interfaces.
-a Display alias name(s) of the host.
-d Display the DNS domain.
-f Display the FQDN for the hostname.
-I Display each IP address on all interfaces.
-i Display the IP address(es) for the hostname.

(roy)

Replace mrg's clean cpp+cpp_arg interface with xrdb version 420347005c
to minimise upstream differences.
OK wiz@

(prlw1)

Replace mrg's clean cpp+cpp_arg interface with xrdb version 420347005c
to minimise upstream differences.
OK wiz@

(prlw1)

Catch open without FREAD|FWRITE (it should not happen)

(manu)

cpuset_info is no longer volatile

(matt)

Deal with m68010 not having bsrl so usr jsr on !__pic__

(matt)

Constify some more.

(uwe)

Spell "keywords" without space.  Fix couple of typos.

(uwe)

manual_find_buildkeyword() - now that we control the format string, we
may use asterisk precision specification instead of temporary
modifying the _build string itself.

(uwe)

Fix manual_find_buildkeyword() to not use non-literal printf format.

(uwe)

Use -Wno-format-nonliteral instead of blanket -Wno-format.

(uwe)

Add a couple of basic IP/MPLS forwarding tests

(kefren)

Update ichsmb(4) manual. We support both PCH and C6xx series, too.

(msaitoh)

Print the part number for DDR3 modules.

Also closes PR kern/44665.

(soren)

sys/cdefs.h should come from rumpuser_port.h

Noticed by Justin Cormack while building against musl libc.

(pooka)

Adjust for Coldfire

(matt)

default to MACHINE_GNU_PLATFORM so that when run with nbmake-foo the right
thing happens

(matt)

Don't emit _fix _floatun for coldfire

(matt)

Cleanup old cruft that wasn't working properly
Adjust for supporting Coldfire processors (default them to softfloat).

(matt)

If __HAVE_LONG_DOUBLE is not defined,
add a strong alias for nextafterl to nextafter

(matt)

Only compile is __HAVE_LONG_DOUBLE is defined

(matt)

pte and pmap files for the soft tlb on coldfire cpus.
(uncompiled and mostly probably wrong.)

(matt)

Adjust LEA_LCL, GOT_SETUP, INTERRUPT_{SAVE,RESTORE}REG for Coldfire

(matt)

Teach to use coldfire isac byterev if available, otherwise let the compiler
figure it since there is rorw instruction to fall back on.

(matt)

Fix some definitions

(matt)

Coldfire FPU looks like the 68010 FPU (no long double)

(matt)

Coldfire uses atomic_cas_generic.c

(matt)

Clear errors before beginning a transfer.

Closes PR port-i386/46792.

(soren)

Adjust for coldfire

(matt)

Use MACHINE_CPU for m68k

(matt)

Add stub port for evbcf which has enough to do a distribution build.

(matt)

Test BSD makefile about coldfire.
XXX still hate m68k--netbsdelf-coldfire

(matt)

m68k doesn't support with-float=soft

(matt)

Use MACHINE_CPU with m68k (m68000 will not be matched due to MKPIC=no)

(matt)

Support MACHINE_ARCH of coldfire

(matt)

Use new macros from <machine/asm.h> to be more portable

(matt)

Use ${MACHINE_CPU} for m68k

(matt)

Adjust for coldfire

(matt)

Remove hardfloat now that it's someplace else.

(matt)

Fix typo.  Don't define map if it isn't used.

(matt)

Cleanup.  hardfloat stuff is now in its own directory

(matt)

Move hardfloat support to its own directory

(matt)

Coldfire doesn't do FLOATX80

(matt)

Add the Makefile

(matt)

Update to compile on coldfire.

(matt)

Add for softfloat compatibility

(matt)

${MACHINE_ARCH} != "m68000" --> ${MACHINE_ARCH} == "m68k"

(matt)

Use stack adjustment in the lnk instruction to adjust the stack.
Saves one word on the following moveml

(matt)

Use ${MACHINE_CPU} == "m68k"

(matt)

Make sure fdformat is on m68k (even sun2)

(matt)

Make the shared library related tests dependent on pic

(matt)

invert tests ${MACHINE} != "m68000" -> ${MACHINE} == "m68k"

(matt)

Always terminate qc_name with NUL because pool_init(9) uses
strcmp(3) to compare wchans.

(alnsn)

Only supplie -fPIE to crtbegin.c if MKPIC is yes.

(matt)

s/bsrl/jsr/ and let gas figure it out

(matt)

+librumpnet_netmpls.so.0.0.debug

(njoly)

Do not make FUTEX_WAIT wait indefinitely for an invalid timeout
(tv_nsec = -1).

(njoly)

add RTM_LOSING, RTM_REDIRECT

(christos)

Don't access memory outside the array if tmpdirlen == 0.

(uwe)

use -Wno-format and revert "fixstring"

(christos)

Add librumpnet_netmpls into sets' lists

(kefren)

Set the string to NUL instread of providing an new empty string (from uwe)

(christos)

Add librumpnet_netmpls that provides MPLS features into rump kernels
ok'ed pooka@

(kefren)

WARNS=6
- fix cast qual issues
- don't use snprintf on a user-provided buffer

(christos)