G/C GOLDENGATE

(skrll)

Don't leak kernel pointers to userland in kern.file2, same as kern.proc2.

(maxv)

Wrap arm_fdt_cpu_okay with #ifdef MULTIPROCESSOR

(jmcneill)

Include rtsock_50.c in the build for compat_50 module.  It might be
reachable via the rtsock code, but there's stuff in the compat_09
module that refers to the _50 stuff.  Including this allows us to
modload the compat_09 module.

(pgoyette)

Fix editing mistake, remove ) from func call that is now gone.

(kre)

add support for kabylake and skylake GT4 (untested) GPUs.
largely taken from openbsd and linux 4.13 trees (which
have this code identical), with mimimal porting to netbsd.

i have not installed (and thus tested) the newer referenced
firmware files.

only real local change is to fix IS_BROXTON() macro to check
the things valid in this era of drm.  previous match would
attach on KBL, and then a loop would never exit.

tested on kabylake P630.  needs mesa 11.x or newer for GL
to work.

ok @riastradh.

(mrg)

Add Makefiles for version-specific compat_netbsd32_sysvipc* modules

(pgoyette)

More clean-up of module dependencies

(pgoyette)

Regen

(pgoyette)

More MODULAR tags

(pgoyette)

Now that compat_netbsd32 is no longer monolithic, it no longer
requires compat_09 (and all higher) compat code.  Adjust the
requirements list accordingly.

(pgoyette)

always call lockdebug_dismiss() from DDB -- there are always some
minimal lockdebug checks in place, even without LOCKDEBUG.

adjust lockdebug_abort() to ignore problems after ld_panic is set
so that there's a chance of this working.

this fixes ddb 'reboot' on softiron od1000.

(mrg)

more efficient to use decode_result :-) Pointed out by kre@

(christos)

Another dependency typo

(pgoyette)

Fix some module requirements

(pgoyette)

PR standards/34646

Make test(1) always use the POSIX "number of args" evaluation rules
when they apply.

Only fall back to the old expression evaluation when there are more
than 4 args, or when the args given cannot work as a test expression
using the POSIX rules.  That is when the result is unspecified.

Also fix old bug where a string of whitespace is considered to be a
valid number (at least one digit is needed amongst it somewhere...)

XXX pullup -8

(kre)

Use correct module name for dependency

(pgoyette)

Regen

(pgoyette)

Missed a couple of MODULAR tags

(pgoyette)

PR/53597: Yasuhiro Horimoto: Avoid segmentation fault in bad history file.

(christos)

I've been on an Ed Wood binge.

(sevan)

Fix lockdebug diagnostic error of trying to acquire an rw_lock from a
pserialized active context. From riastradh@

(christos)

Don't have modules depend on bpf; they will find the dependency dynamically
at runtime. Otherwise builtin modules will complain about non-builtin bpf.

(christos)

Add some more module glue

(pgoyette)

Complete removal of the "monolithic" all-version compat_sysvipc module

(pgoyette)

add new dir to lists

(para)

Remove this check, it has never protected against mmap on page zero, and
has since been replaced by the code in exec_vm_minaddr.

(maxv)

add nta named directory.

(christos)

Put the nta files in a subdirectory instead of requiring the namedb root
directory to be writable by named... Others have expressed the same concern,
but upstream refused: https://bugzilla.redhat.com/show_bug.cgi?id=1487823
https://bugs.isc.org/Public/Bug/Display.html?id=46242

(christos)

if_msk: detach/reattach fixes

- Fill in sc->sk_bsize upon pci_mapreg_map() so that we'll correctly
  bus_space_unmap() on detach.
- Clear sc->sk_intrhand after interrupt disestablish, to prevent
  disestablishing unestablished interrupt.

(jakllsch)

Try to properly guess a path on the ftp server for pkg binaries
and use that as default.

Assume for $N.99 (current) we can use $N.0. Otherwise strip all
RC*/STABLE* annotations and just use the plain numbers.

(martin)

use pa_dmat64 if pci_dma64_available().

(mrg)

- link to pci_mapreg_type() from pci_mapreg_map() for the type argument
- note that 64 bit memory mappings are different to others
- bump date

(mrg)

Typo - missing trailing quote

(pgoyette)

Adjust module requirements to reflect the new more granular modules

(pgoyette)

Add ATS Relaxed Ordering supported bit described in Address Translation
Relaxed Ordering ECN.

(msaitoh)

Fix comment.

(msaitoh)

- Try m_defrag() to reduce the number of DMA segment if bus_dmamap_load_mbuf()
  returned EFBIG. When m_defrag() is called, txqNNdefrag event counter is
  incremented. If the 2nd try of bus_dmamap_load_mbuf() failed, txqNNtoomanyseg
  event counter is incremented.

- Reduce the max number of DMA segemts from 256 to 64 (it's the same valuse
  as other BSD's (EM_MAX_SCATTER) and more than before if_wm.c rev. 1.75's
  value (40)) because we do m_defrag() now.

(msaitoh)

- Split txdrop evcnt into pcqdrop and descdrop.
- Simplify evcnt name by removing duplicated "tx" or "rx".
- s/tu/underrun/

(msaitoh)

Split compat_netbsd32_sysvipc into version-specific modules

(pgoyette)

use pci_mapreg_type() instead of assuming PCI_MAPREG_TYPE_MEM.
suggested by riastradh@.

fixes hdaudio on thinkpad P51 where the registers are mapped
over 4GB range.  probably fixes intel X10-based systems that
map hdaudio high as well, but i was able to work around that
in the BIOS.  not so for the P51.

XXX: pullup-7, pullup-8.

(mrg)

Regen

(pgoyette)

Prepare for splitting compat_netbsd32_sysvipc into version-specific modules

(pgoyette)

Fix a bug that SIOC[GZ]IFDATA returned old ifreq structure.

(msaitoh)

Add new compat_netbsd32_16 module to sets

(pgoyette)

Add a compat_netbsd32_16 to handle the sigcode stuff.

(pgoyette)

Add a netbsd32_compat40 module, move the 4.0 mount stuff there.

(pgoyette)

regen

(pgoyette)

The netbsd32 40 mount code belongs in a netbsd32 40 module

(pgoyette)

Regen with updated module information for ntp 30 and 50 compat calls

(pgoyette)

Move the NTP compat_30 and compat_50 routines into the version-specific
modules.

(pgoyette)

Move netbsd32_rndpseudo_50.c out of the main netbsd32 module and into
the module that handles netbsd32_50 compatability.

(pgoyette)

Put the ``extern struct emul'' into the netbsd32_netbsd.h header
so it doesn't have to be separately declared in each source file.

Thanks mrg@

(pgoyette)

Add Exynos5 SMP support.

(jmcneill)

Make this driver MPSAFE

(jmcneill)

Remove unused variable

(pgoyette)

More fixing of placement of version number in module name

(pgoyette)

Typo

(pgoyette)

Update

(pgoyette)

Make compat_netbsd32_43 into a stand-alone module.

(pgoyette)

Use the real version numbers, not just the ones from the template.

(pgoyette)

Regen

(pgoyette)

Insert "MODULAR compat_netbsd32_43" as neeed

(pgoyette)

Define vector for compat_netbsd32 version of sys_modctl()

(pgoyette)

Split the host name used for ftp transfers from the one used for http
transfers. This is slightly inconsistent, as directories are still
shared - but this allows us to default to cdn/nycdn for http (which don't
support ftp).

(martin)

Get ready for compat_netbsd32_43 module

(pgoyette)

Fix module names - the version info comes prior to the .kmod suffix!

(pgoyette)

Fix some typos

(pgoyette)

And regen again

(pgoyette)

More over-zealousness - what was I thinking?

(pgoyette)

Revert previous - it was wrong.

(pgoyette)

Regen

(pgoyette)

Don't be overzealous with changes!  Return a few syscalls to their
original COMPAT_xx status (rather than STD)

(pgoyette)

Don't include the 60-specific code in the monolithic compat_netbsd32 module

(pgoyette)

Also need to declare the emul_netbsd32 !

(pgoyette)

Install the syscalls into the netbsd32 table, not the default/native table.

(pgoyette)

Add -DCOMPAT_NETBSD32 for all the version-specific modules

(pgoyette)

Whitespace cleanup from last update.  NFC.

(kre)

Use netbsd32 syscall names in the establish/disestablish tables

(pgoyette)

add dead...

(christos)

Regen syscall files and fix includes for syscallvar.h

(pgoyette)

Update for module-provided compat syscalls

(pgoyette)

More work on splitting the compat_netbsd32 module

(pgoyette)

armgic_ipi_send: use GIC interface number, not CPU index, when setting
the target(s) for an IPI.

(jmcneill)

Re-add support for cpu nodes with status = "disabled", lost in previous commit

(jmcneill)

psci_fdt_bootstrap -> arm_fdt_cpu_bootstrap

(jmcneill)

Ticket #1019

(martin)

Pull up following revision(s) (requested by nakayama in ticket #1019):

sbin/raidctl/rf_configure.h: revision 1.2
sbin/raidctl/rf_configure.c: revision 1.27
sbin/raidctl/rf_configure.c: revision 1.28
sbin/raidctl/rf_configure.c: revision 1.29
sbin/raidctl/raidctl.8: revision 1.73
sbin/raidctl/rf_configure.c: revision 1.30
sbin/raidctl/rf_configure.c: revision 1.31
sbin/raidctl/rf_configure.c: revision 1.32

support NAME=<wedge name> syntax for disks and spares

-

stop using magic constants
wrap long lines
use warn{,x}
make static
knf

-

White space and comment formatting.  NFC.

-

With char bug[SIZE]  using sizeof(bug[0]) is kind of boring, use
sizeof(bug) instead...

-

Avoid needless pointer calisthenics: &foo[0] -> foo

-

Several more cleanups:
1. Don't force use of "for" when "while" works better.
2. No need to check c != '\0' when we also check (c == ' ' || c == '\t')
3. Use the size of the buffer we're using, rather than a different one
  (not really a concern, they're the same size)
4. Don't use fscanf() to read file data, use fgets() & sscanf().
5. After using a pointer as a char *, validate alignment before switching
  to int * (can only fail if kernel #define gets set stupidly)  Or #6...
6. Validate sparemap file name isn't too long for assigned space.
7. recognise that strlen() returns size_t - don't shove it into an int.
8. On out of mem, be more clear which allocation failed in warning msg.

ATF tests all pass.  But I don't think they use sparemap files.

(martin)

reduce the battlefield

(maxv)

changed kcopy() to asm to avoid replacement memcpy() to kasan_memcpy() when defined KASAN.

(ryo)

Replace KDASSERT by panic.

(maxv)

Tickets #1015 - #1018

(martin)

Pull up following revision(s) (requested by knakahara in ticket #1018):

sys/netinet6/in6_l2tp.c: revision 1.17
sys/netinet/in_l2tp.c: revision 1.16

fix: l2tp(4) cannot receive packets after reset session without reset tunnel. Pointed out by k-goda@IIJ

When the following operations are done after established session, the l2tp0
cannot receive packets until done deletetunnel && tunnel "src" "dst".

====================
ifconfig l2tp0 deletesession
ifconfig l2tp0 deletecookie
ifconfig l2tp0 session 200 100
====================

XXX pullup-8

(martin)

Pull up following revision(s) (requested by tsutsui in ticket #1017):

sys/arch/sgimips/conf/INSTALL32_IP2x: revision 1.10
sys/arch/sgimips/stand/boot/Makefile: revision 1.20
sys/arch/sgimips/sgimips/machdep.c: revision 1.146
sys/arch/sgimips/hpc/wdsc.c: revision 1.35
distrib/sgimips/instkernel/Makefile: revision 1.13

Disable misc options to shrink an INSTALL kernel for IP2x.

The ARC BIOS on Indy seems to have ~8MB limit.

Fixes PR port-sgimips/53378 from Naruaki Etomi.
Should be pulled up to netbsd-7 and netbsd-8.

-

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.

-

Use elf2ecoff(1) rather than objcopy(1) to generat ecoff for old machines.
objcopy(1) was used instead of elf2ecoff(1) since Makefile rev 1.9:
http://cvsweb.netbsd.org/bsdweb.cgi/src/sys/arch/sgimips/stand/boot/Makefile#rev1.9
but it looks ELF binaries generated by recent binutils are too complicated
to convert them to ecoff for old machines by objcopy(1).

Reported by Naruaki Etomi in PR port-sgimips/53519.
Should be pulled up to at least netbsd-8.

-

Fix wdsc(4) probe failure on HPC1.5 machines (Indigo R3k/R4k, IP6/IP10/IP12).

The alignment adjustment code was removed in rev 1.36 for common
mips bus_space changes, but the code was not bus_space related
but used for uint32_t register acccess during probe.

Reported and tested by Naruaki Etomi in PR port-sgimips/53522.
Should be pulled up to netbsd-8.

-

Use elf2ecoff(1) rather than objcopy(1) to generate ecoff kernel.

Reported by Naruaki Etomi in PR port-sgimips/53518.

Should be pulled up to netbsd-7 and netbsd-8.

(martin)

Pull up following revision(s) (requested by kre in ticket #1016):

usr.bin/find/function.c: revision 1.77

Inspired by PR pkg/53543

When calculating the length of the args that can be
appended in a "find .... -exec something {} +"
usage, remember to allow for the arg pointers, which
form part of what is allowed in ARG_MAX.

From a fairly empty installation of HEAD on amd64
and with a "/tmp/args" command that simply prints
its arg count, and the length of the arg strings,
with this mod I see ..

netbsd# find / -exec /tmp/args {} +
Argc 5000 Arglen 107645
Argc 5000 Arglen 151324
Argc 5000 Arglen 187725
Argc 5000 Arglen 206591
Argc 5000 Arglen 172909
Argc 5000 Arglen 186264
Argc 5000 Arglen 167906
Argc 2881 Arglen 98260

The upper limit of 5000 args is in the code.

Using the biggest of those, 5000
args, plus 206591 bytes of strings
uses 246591 bytes total (this excludes
the command name, so add a few more).

That's fairly close to the ARG_MAX
of 262144.

On another system (with longer paths) I see:
(this is just a small part of the output, using a
different version of the dummy command, and a
slightly different invocation)

Args: 4546 Len 218030
Args: 4878 Len 217991
Args: 4813 Len 218028
Args: 4803 Len 218029

There, 4878*8 + 217991 == 257015 which is about
as close as we'd want to come to the arg limit.

XXX pullup -8

(martin)

Pull up following revision(s) via patch (requested by kre in ticket #1015):

bin/sh/expand.c: revision 1.124
bin/sh/expand.c: revision 1.127
bin/sh/parser.c: revision 1.148
bin/sh/parser.c: revision 1.149
bin/sh/syntax.c: revision 1.6
bin/sh/syntax.h: revision 1.9 (partial)

First pass at fixing some of the more arcane pattern matching
possibilities that we do not currently handle all that well.

This mostly means (for now) making sure that quoted pattern
magic characters (as well as quoted sh syntax magic chars)
are properly marked, so they remain known as being quoted,
and do not turn into pattern magic.  Also, make sure that an
unquoted \ in a pattern always quotes whatever comes next
(which, unlike in regular expressions, includes inside []
matches),

-

Part 2 of pattern matching (glob etc) fixes.
Attempt to correctly deal with \ (both when it is a literal,
in appropriate cases, and when it appears as CTLESC when it was
detected as a quoting character during parsing).

In a pattern, in sh, no quoted character can ever be anything other
than a literal character.  This is quite different than regular
expressions, and even different than other uses of glob matching,
where shell quoting is not an issue.

In something like
ls ?\*.c
the ? is a meta-character, the * is a literal (it was quoted).  This
is nothing new, sh has handled that properly for ever.

But the same happens with
VAR='?\*.c'
and
ls $VAR
which has not always been handled correctly.  Of course, in
ls "$VAR"
nothing in VAR is a meta-character (the entire expansion is quoted)
so even the '\' must match literally (or more accurately, no matching
happens - VAR simply contains an "unusual" filename).  But if it had
been
ls *"$VAR"
then we would be looking for filenames that end with the literal 5
characters that make up $VAR.

The same kinds of things are requires of matching patterns in case
statements, and sub-strings with the % and # operators in variable
expansions.

While here, the final remnant of the ancient !! pattern matching
hack has been removed (the code that actually implemented it was
long gone, but one small piece remained, not doing any real harm,
but potentially wasting time - if someone gave a pattern which would
once have invoked that hack.)

(martin)

Rename _pmap_alloc_pdp -> pmap_alloc_pdp, and make it public.

(maxv)

Add a test where printf is run twice in the same shell, and the
first invocation uses \c in a %b arg - make sure that 2nd invocation
is not affected by that  (it was until recently, for a very long time).

(kre)

A truly ancient bug found by Edgar Fuss

When printf is running builtin in a sh, global vars aren't reset to
0 between invocations.  This affects "rval" which remembers state
from a previous %b \c and thereafter always exits after the first
format conversion, until we get a conversion that generates an
error (which resets the flag almost by accident)

printf %b abc\\c
abc (no \n)
printf %s%s hello world
hello (no \n, of course, no world ...)
printf %s%s hello world
hello
printf %s%s hello world
hello
printf %d hello
printf: hello: expected numeric value
0 (no \n)
printf %s%s hello world
helloworld (no \n, and we are back!)

This affects both /bin/sh and /bin/csh (and has for a very long time).

XXX pullup -8

(kre)

Correctly align the size+redzone for KASAN, on amd64 it happens to be
always 8byte-aligned but on other architectures it may not be.

(maxv)

cleanup aarch64 mpstart and fdt bootstrap
* arm_cpu_hatch_arg is a bad idea. avoid serializing CPU startup, and eliminate arm_cpu_hatch_arg.
  in mpstart, resolve own cpu index using array of cpu_mpidr[] (aarch64)
* add support fdt enable-method "spin-table"
* add support fdt enable-method "brcm,bcm2836-smp" (for 32bit RaspberryPi)
* use arm_fdt_cpu_bootstrap() instead of psci_fdt_bootstrap()
* rename "arm/fdt/psci_fdt.h" to "arm/fdt/psci_fdtvar.h" because of conflict of include file for needs-flag
* add devmap for cpu spin-table of raspberrypi3/aarch64
* no need to force hatch APs for raspberrypi3/arm32 ifndef MULTIPROCESSOR.
* fix to work pmap_extract(kerneltext/data/bss) even if before calling pmap_bootstrap

idea to use cpu_mpidr[] by jmcneill@. reviewd by skrll@. thanks.

(ryo)

Remove from the generic module those files which are being included
in their own version-specific modules.

(pgoyette)

Typo in comment from Joerg's gtmr diff

(skrll)

More compat_netbsd32_xx modules

(pgoyette)

Initial cut at a compat_netbsd32_09 and compat_netbsd32_12 modules.

(pgoyette)

Update sc_mptargets atomically, as PEs will be started up in parallel w/o locking here

(jmcneill)

Another compat_80

(pgoyette)

Add compat_80 to the list of compats

(pgoyette)

Note that the compat_netbsd32 module will be split into version-specific
components.

(pgoyette)

Update for removal of monolithic compat modules, and for introduction
of version-specific compat_netbsd32_xx modules

(pgoyette)

Don't descend into the build directories for monolithic compat and
compat_sysv modules.

Add sub-dirs for the soon-to-arrive version-specific compat_netbsd32_xx
modules.

(pgoyette)

Remove the monolithic compat and compat_sysv modules.  Now that we
don't have aliases, we need to use only the version-specific module
names when resolving dependencies.

(pgoyette)

Don't use printx before setting up stack.

(skrll)

Remove the last usage of xen_pirq_alloc() and pirq_establish()
outside of the x86 interrupt and xen events framework.

This allows us to finally unify the interrupt path for both Xen
and x86 as changes 'internal' to the subsystem.

This change has been kindly tested on real hardware by gson@

The change is not cosmetic and may thus affect users on various
hardware configurations - especially involving legacy hardware.

I look forward to bug reports.

(cherry)

Fix typos and DEBUG_MMU output.  From Rin Okuyama.

(skrll)

Make the use of 'irqs' in the range 0 < irq < 255 by xen
as a handle for internal use explicit.

This allows us to pass up the handle as "legacy" irq while
establishing interrupt handlers for xen.

No functional change.

(cherry)

In the NIOAPIC case, we do not need to support "legacy" irqs,
ie; We don't need to simultaneously pass back the irq in the
range 0 < irq < 16 (which are sometimes described as "legacy"
in src

This was non-obvious, until the semantics of "legacy" were
used in inconsistent ways in Xen (to also mean interrupts in
the 0 < irq < 256 range) which causes problems with attempting
to unify the sys/arch/x86/isa/isa_machdep.c:isa_intr_establish_xname()
function between XEN and !XEN

This commit should not affect current functionality on any
either native or Xen. It is needed for future code reorg, and
published now as a preview.

(cherry)

Add compat_netbsd80.config

(pgoyette)

In the case of no compat modules (ENOSYS returned from the stubs), reset
the retcode to zero so we don't return the wrong value when there is
nothing to do for the particular command.  This was causing spurious
failures for RAIDFRAME_SET_COMPONENT_LABEL which is wrapped in

#if 0
...
#endif

(pgoyette)

Boot APs with status "disabled" if they have an enable-method property

(jmcneill)

Attach cpu nodes with status "disabled" if they have an enable-method
property. This is a valid configuration according to the devicetree
specification.

(jmcneill)

Add fdt_add_bus_match, which is like fdt_add_bus but allows for the caller to filter child nodes on their own

(jmcneill)

Add instructions for building NetBSD on modern Solaris 11 hosts

(palle)

Add a reboot command to the bootloader

(jmcneill)

Add support for setting environment variables. Currently the following env
vars are supported: "fdtfile", "initrd", and "rootdev".

(jmcneill)

Take interrupt-map-mask into consideration when mapping PCI interrupts.

(jmcneill)

Add "dtb" command for loading a custom .dtb file.

(jmcneill)

Fix previous; PSCI_CPU_ON requires an MPIDR!

(jmcneill)

Ditch arm,arm-v8 compatible string (everything uses arm,armv8)

(jmcneill)

When freeing the required list, make sure we calculate its size based
on the size of the contained elements, not on the size of a pointer to
those elements.

(pgoyette)

The compat_netbsd32 module requires compat_09 and above.  Since we
removed the alias names, we need to explicitly refer to the oldest
non-32 required compat module.

(pgoyette)

Add a todo item for a future major bump (rename many symbols)

(maya)

Pass clock provider's phandle to fdtbus_clock_controller_func.decode()
and update callers.

This allows to accomodate clock managers whose clocks are identified
directly by a clock instead of a pair (clock provider, index).

ok jmcneill@ on port-arm

(aymeric)

include sys/param.h for MIN/MAX, needed when building in pkgsrc
tolerate set version failing in another location too.

this now works well enough for mesa 18.2.0 to display glxgears followed
soon by X crashing and refusing to start again until the machine is
rebooted, at least in my case.

PR pkg/51795, pkg/51786

(maya)

Let's get the sense of strncmp correct, shall we?

PR pkg/51795
PR kern/51786

XXX pullup-7
XXX pullup-8

(riastradh)