Remove arch check now that rump smp is available on all host archs.

(pooka)

Add SMP support for all architectures.

tested on sparc64 by martin

(pooka)

The atexit test has been atf-ified.

(pgoyette)

Now that PR/44311 is fixed we can remove the expected-failure from this
test.

(pgoyette)

Fix to build failure.  Include uvm/uvm_extern.h for pmap_extract().

(kiyohara)

Turn off VERBOSE_ARM_INIT

(matt)

Check that file attributes can be changed (where applicable).

(pooka)

Need to deal with msg_iov in netbsd32_to_msghdr.

(matt)

Make the SA support as optional as is possible.

(matt)

check from_cidr_to_union() result

(kefren)

Centralise the PDC procedure calls into one file and use a minimal set of
kernel data for PDC results.

Remove the filler components from the PDC structures as a result.

Other tidyups while here.

(skrll)

Stop using aprint_debug(9) in an interrupt handler.

(jruoho)

Fix fd leak in error case. Found by cppcheck.

(wiz)

Fix fd leaks. Found by cppcheck.

(wiz)

Close masterfd after reading from it. Found by cppcheck.

(wiz)

Fix possible NULL pointer dereference in #ifdef AS_DEBUG_FILE.
Found by cppcheck.

(wiz)

unifdef -U RCSSTAT
This code was added in 1.2 by brezak with changes for the commit message:
"Change <sys/dir.h> to <dirent.h>"
and looks like it never worked.

(wiz)

Fix more memory leaks in case malloc fails

(kefren)

Fix fd leaks in error cases. Found by cppcheck.

(wiz)

Fix fd leak in error case. Found by cppcheck.

(wiz)

Use pclose() after popen(). Found by cppcheck.

(wiz)

Fix fd leaks in error cases. Found by cppcheck.

(wiz)

Fix memory leak in error case. Found by cppcheck.

(wiz)

Need to check for VNOVAL when setting attrs.

(pooka)

Fix fd leaks in error cases. Found by cppcheck.

(wiz)

Fix fd leak in error cases. Found by cppcheck.

(wiz)

Fix memory leak in error case. Found by cppcheck.

XXX: so_tag has same problem?

(wiz)

Fix fd leaks in error cases. Found by cppcheck.

(wiz)

Fix fd leaks in error cases. Found by cppcheck.

(wiz)

Fix fd leaks. Found by cppcheck.
While here, fix typo in warning message.

(wiz)

Fix fd leak in error case. Found by cppcheck.

(wiz)

Add missing breaks in switch statement. Found by cppcheck.

(wiz)

Fix fd leak in error case. Found by cppcheck.

(wiz)

Add audio for uaudio

(matt)

Close fd in error case. Found by cppcheck.

(wiz)

Add ukphy (makes MIIVERBOSE happier)

(matt)

Use format string for err(). From Henning Petersen in PR 44326.

(wiz)

Supply a hook at allow an RAMDISK image to automagically inserted into a
newly built kernel.

(matt)

Add better color matching selecting free pages.  KM pages will now allocated
so that VA and PA have the same color.  On a page fault, choose a physical
page that has the same color as the virtual address.

When allocating kernel memory pages, allow the MD to specify a preferred
VM_FREELIST from which to choose pages.  For machines with large amounts
of memory (> 4GB), all kernel memory to come from <4GB to reduce the amount
of bounce buffering needed with 32bit DMA devices.

(matt)

Add a MD hook to indicate a change of vmspace due to exec.  (This is useful
to update any cpu flag due to a change to/from a 64bit and a 32bit address
space).  This can set the state needed for copyout/copyin before setregs
is invoked.

(matt)

Print the number of page colors in use with db> show uvm

(matt)

Tsort functions and remove a small #if 0 block leftover from earlier cleanup.
No functional change.

(dholland)

Make the alternate name for "manet" caps so it's not a duplicate.
Gives the impression of closing PR 44311.

(dholland)

fsck should be in acronyms.comp, not the regular acronyms file.

(dholland)

If we're going to have non-acronyms like "1337" and "noob" in here,
at least define them properly. In particular, don't define "noob" as "noob".

(dholland)

Revert the changes made in the commit:

  "Do not queue functions via sysmon_taskq(9) in the pmf(9) resume hooks.
  There is a small and unlikely race when the drivers are loaded as modules;
  suspend, resume, queue a function, and immediately unload the module."

Queueing the checks in the resume hooks was observed to improve the resume
time by several seconds (!) on a system with poorly implemented AML code.
The observed race condition is a wider issue related to modules; things like
sysmon_taskq(9) should be modified to allow cancelling of queued functions.

(jruoho)

Fix previous.

(jruoho)

withdraw my name from the copyright notice as there is little code
substance made by me.

(nisimura)

Add options P1003_1B_SEMAPHORE.

(tsutsui)

If using a kernel ldscript, make sure it's added to the list of kernel
dependencies.

(matt)

add support for autostarting watchdogs (wdog was started by firmware
and can't be disabled).  Add critical pool hooks for kernel tickled
watchdogs.

(matt)

When determining max/min sym value, ignore ABS symbols which don't
have a type.  This stops ksym thinking that the _KERNEL_OPT symbols
could actually be useful.

(matt)

Switch back to malloc since kmem isn't softintr safe.

(matt)

Use LIST_FOREACH when possible.

(matt)

Make this support the DS1553 as well.  Add support for the centry reg which
newer RTCs seem to have.

(matt)

Add the new mpc85xx/booke device names.

(matt)

Add a generic Motorola/Freescale i2c driver.  This was taken from sandpoint
and heavily modified.  It works on most Freescale PowerPC and ARM SoCs.

(matt)

Add a generic Motorola/Freescale i2c driver.  This was taken from sandpoint
and heavily modified.  It works on most Freescale PowerPC and ARM SoCs.

(matt)

Don't try to change the size of non-reg nodes or when vattr doesn't
contain a new size.

from njoly

(pooka)

Don't run FP Exception tests under qemu - they don't work.

While here, properly skip certain tests on vax architecture.  (It is
not legal to have a test-program with zero test cases, so each test
case needs to check-and-skip.)

(pgoyette)

Ooopppsss - locale/ tests are not yet ready - don't SUBIDR yet

(pgoyette)

Reformat comment.

(skrll)

Atf-ify the atexit test

(pgoyette)

Fixed strvisx usage

(stacktic)

kqueue signal filters are not supported, so be explicit about it.

(pooka)

Apparently, the ad.{MACHINE_ARCH} files are not processed when

${MACHINE} == ${MACHINE_ARCH}

So rename the files for affected ports accordingly.

This is confirmed to fix build breakage on port-i386 reported by
gson@ gson.org in private Email thread.

(pgoyette)

update comment

(pooka)

destroy mutex/cv before freeing the struct holding them, fixes
LOCKDEBUG panic

(drochner)

Fix usage (8th bit begone!)

(wiz)

Document import of dhcpcd-5.2.10

(roy)

Don't build the getline shim anymore as the NetBSD libc getline(3) has
proven itself now.

(roy)

Import dhcpcd-5.2.10 with the following change:
* -r, --request <address> now works again

Status:

Vendor Tag: roy
Release Tags: dhcpcd-5-2-10

(roy)

Add -c to control the number of CPUs configured in the kernel.

(pooka)

use rawpart

(pooka)

make this work when rawpart != d

(pooka)

Give this a chance of working on non-4kB pagesize.
from martin's sparc64 test run

(pooka)

remove another extraneous static

(dholland)

fix this properly.

(dholland)

Scan for devices that are "hot-pluggable".

(jruoho)

Add (commented out) options MODULAR.  Confirmed working on XM6i.

(tsutsui)

Xref module(7) in comments.

(tsutsui)

avoid a double free - from Anthony Bentley.

(agc)

Enable options DDB.  Ok from isaki@

(tsutsui)

PR/44310: Alexander Nasonov: write to /dev/bpf truncates size_t to int

(christos)

The servent test has been atf-ified

(pgoyette)

Remove atf-ified tests: servent protoent md5sha

(pgoyette)

Sets list magic for the servent/protoent/hash tests

(pgoyette)

Add the new directory for hash test data

(pgoyette)

Atf-ify servent and protoent tests, and the remaining hash test.

(pgoyette)

Prefix section headers for 1 to 9 with "NetBSD " to match output
for non-HTML pages (i.e. groff ones). Noted by njoly.

(wiz)

If we survive the check for wrong signal code, reset our expectation to
"pass" before checking the signal errno.

(pgoyette)

Zero the in-memory super block after allocation.  It contains a pointer
that, if not properly initialized, could be improperly deallocated.

(jakllsch)

Fix $NetBSD$ keyword

(pgoyette)

avoid cloberring warning.

(christos)

kern/44310: Alexander Nasonov: write to /dev/bpf truncates size_t to int

(christos)

Simplify and avoid kernel segv when the list is NULL.

(christos)

Kernels modules are c99 (gnu99) just like the kernel.
Add(move) -std=gnu99 to CPPFLAGS (from COPTS).

(matt)

Tidy up common comments.  (Yes, too many dup code...)

(tsutsui)

Pull code that maps the kernel segment table cache invalidated for 040/060,
to reduce diffs from other m68k ports.

Compile test only. (any working machine for these port?)

(tsutsui)

One more reference to HAVE_SIGINFO bites the dust

(pgoyette)

Pull code that maps the kernel segment table cache invalidated for 040/060,
to reduce diffs from other m68k ports.

Tested on LC630 (with 68LC040) upto single user.

(tsutsui)

Fix a double free[*], pointed out by Anthony Bentley.

[*] This was actually a triple free. We go all the way to 11.

(agc)

Remove last references to __HAVE_SIGINFO

(skrll)

__HAVE_SIGINFO has not been required/used for a long time now.

Prompted by uwe@

(skrll)

Stopgap buildfix: gcc complains about test_ops possibly being clobbered
by sigjmp and I can't find a proper way to add -Wno-uninitialized in the
proper place in CFLAGS.

(martin)

do the alphabet

(pooka)

document rump_pub_lwproc_rfork()

(pooka)

rump_lwproc_newproc -> rump_lwproc_rfork

(pooka)

+ rump_lwproc_newproc -> rump_lwproc_rfork()
+ add a tess for rump_lwproc_rfork()

(pooka)

regen: rump_lwproc_rfork

(pooka)

There is a use case where preserving the parent's fd table is
relevant, so to accommodate that change rump_lwproc_newproc() to
rump_lwproc_rfork().  The new interface has the rfork() fd table
semantics.  The equivalent of rump_lwproc_newproc() is
rump_lwproc_rfork(RUMP_RFCFDG).

(pooka)

New sentence, new line. Bump date for previous.

(wiz)

Cap h.loc.count at MAXKEYCODE instead of bailing out.  This lets some obscure/
roque keyboards attach.

From loganaden@devio.us (Loganaden Velvindron), thanks!

(mbalmer)

Add references RFC3068 and RFC3964; mark draft-itojun-... as expired.

(is)

Make this build again by removing the const qualifier on the
local pcb0 variable in mips_init_lwp0_uarea().

(he)

Four and a half year after the end of 6bone, it should not surprise that the
web page dedicated to it is gone. Fortunately, 6bone itself is not referenced
in this manual page. Thanks to aniou@ircnet for the heads-up.

(is)

Adjust error messages for the previous commit.

(jruoho)

add usr/tests/lib/libc/ieeefp and ./usr/libdata/debug/usr/tests/lib/libc/ieeefp

(dbj)

add the normal DKWEDGE* options, and enable drvctl.

(mrg)

fix a comment

(mrg)

now that it's optional, kill schizo_pci_find_ino().
prepare to enable the tomatillo iocache.

(mrg)

layer_inactive: With specnodes introduced during vmlocking2
                it is safe to cache device nodes.

Tested with nullfs only as unionfs with device nodes panics.

(hannken)

Switch mvmeppc bootloader to using MI libsa and libkern for maintainability.
Mostly taken from mvme68k.

Compile test only, but no functional changes.

(tsutsui)

Whitespace.

(skrll)

add missing ./usr/tests/lib/libc/ieeefp

(dbj)

- build a GENERIC kernel on release for validation on releng daily build
- comment out targets which should be put once after src/distrib/amigappc
  is prepared

(tsutsui)

Misc cosmetics to reduce diff among hp300 derived m68k ports.
XXX: too many quirks to merge...

(tsutsui)

Add a dummy <machine/bus.h> which nows seems to be mandatory in MI code
including <sys/bus.h> even if a port doesn't use it. Taken from sbmips.

(tsutsui)

Explicitly include <machine/bus.h> here for bus_space_tag_t and bus_dma_tag_t.

(tsutsui)

Make sure to build GENERIC kernel on release build.

This means there is no kernel binary in past releases. Oh well...

(tsutsui)

Fix one more fatal typo. Umm.

(tsutsui)

Xref __insn_barrier(3) and mb(9) with each other.

(jruoho)

Misc cosmetics to reduce diffs among these ports.

(tsutsui)

Remove references to SAVESTART and SAVENAME in comments.

(dholland)

Some cosmetics to reduce diffs from other m68k ports.

(tsutsui)

Note hw.acpi.power and hw.acpi.wake.

(jruoho)

Remove "last kernel PT page" stuff derived from hp300 where PA != VA.

Compile test only, but same changes as other PA == VA m68k ports like
luna68k, news68k, and x68k used for a decade.

(tsutsui)

Use "real" device names for the hw.acpi.wake and hw.acpi.power sysctl(8)
trees. For instance, instead of hw.acpi.wake.DURT, we have hw.acpi.wake.wm0
for wm(4). This is a temporary solution, but without solving the big
abstraction questions, this is the best we can do.

(jruoho)

Print a warning if the BIOS thinks that the timer is unreliable. However, do
not touch the quality of the timecounter(9) based on this information.

(jruoho)

Switch mvme68k standalone problams to using MI libsa and libkern
for maintainablility. Also bump version for note.

Compile test only, but no functional changes.

(tsutsui)

Make pmap_aliasmask unsigned and declare it in <m68k/pmap_motorola.h>.

XXX: does anyone still have working HP320 or HP350 with 020+HP_MMU+VAC?

(tsutsui)

Pull code that maps the kernel segment table cache invalidated for 040/060,
to reduce diffs from other m68k ports.

Tested on HP362 (030) and HP382 (040), and no visible performance difference.
(what should we do per "for the 68040 not strictly necessary but
recommended by Motorola" comment?)

(tsutsui)

5.99.43: R.I.P. SAVESTART.

(dholland)

Final regen for tonight.

(dholland)

Remove remaining references to SAVESTART.

(dholland)

Remove references to SAVESTART in zfs.

(dholland)

Regen yet again.

(dholland)

Remove the special refcount behavior (adding an extra reference to the
parent dir) associated with SAVESTART in relookup().

Check all call sites to make sure that SAVESTART wasn't set while
calling relookup(); if it was, adjust the refcount behavior. Remove
related references to SAVESTART.

The only code that was reaching the extra ref was msdosfs_rename,
where the refcount behavior was already fairly broken and/or gross;
repair it.

Add a dummy 4th argument to relookup to make sure code that hasn't
been inspected won't compile. (This will go away next time the
relookup semantics change, which they will.)

(dholland)

Remove disabled code for VAC.  No such news68k models.

(tsutsui)

Regen again.

(dholland)

Add an INRELOOKUP namei flag. Sigh. (We don't need more namei flags.)

However, because of a protocol deficiency puffs relies on being able
to keep track of VOP_LOOKUP calls by inspecting their contents, and
this at least allows it to use something vaguely principled instead of
making wild guesses based on whether SAVESTART is set.

Update libp2k to use INRELOOKUP instead of SAVESTART.

(dholland)

Regen.

(dholland)

Pull code that maps the kernel segment table cache invalidated for 040/060,
to reduce diffs from other m68k ports.
(though news68k doesn't have 040/060 models)

(tsutsui)

Remove unused nameidata field ni_startdir.

(dholland)

The unmatchedvarparens test actually generates a blank link as output,
so make sure we have a matching blank line in the "golden" output match
file.

(pgoyette)

The ieeefp stuff has been atf-ified, with the exception of ieeefp/testfloat
(which was never hooked into the build anyway).

(pgoyette)

Add the ieeefp tests to everything EXCEPT for vax and m68000 architectures.

(pgoyette)

Atf-ify several of the ieeefp tests.

Note that t_round is currently mostly disabled, due to PR/44293.  Also,
testfloat has not been converted.

(pgoyette)

Make these 64-bit clean.

(matt)

Relocate cdboot (aka bootxx_cd9660) to 0x1000 instead of 0x600.
A similar change was made in v1.38 of Makefile.bootxx about 13 months ago,
no reason these shouldn't be consistent with each other, like they were before.

(jakllsch)

Correct comments referencing PRIMARY_LOAD_ADDRESS to match current reality.

(jakllsch)

Typo - it is h_getopt_long and not h_getoptlong

(pgoyette)

The last of the libc/stdlib tests have been atf-ified

(pgoyette)

Convert a few more tests from regress to atf

(pgoyette)

Build t_ldexp test always.  The test itself determines if it is running
on an unsupported platform (vax) and exits.

(pgoyette)

clean up lint (on amd64)

(agc)