Update Spleen kernel fonts to version 1.8.0, bringing the following
improvements:

- Improve ampersand character, making it more consistent with other
  sizes (5x8 version)

(fcambus)

Check the return value of iic_acquire_bus(). This function may fail.

One of the case is driver's detaching phase on shutdown. mutex_tryenter()
might fail and return with EBUSY. To avoid calling iic_release_bus() without
taking lock, check the return value of iic_acquire_bus().

(msaitoh)

kyua-cli: fix build of .cpp files

Fix assignment of SRCS for C++ sources that use .cpp instead of .cc.

(lukem)

Fix addition of C++ .cc source to SRCS

Fix assumption that all of the source files are .c, as some are .cc.

(lukem)

use ggc-none.c not ggc-none.o in SRCS

(lukem)

Remove evcnt(9) counter for tlbhit; there's no way to observe TLB hits
from kernel, and it just counted intersection b/w tlbmiss and tlbenter.

(rin)

No longer depend on libl.

(rin)

KNF. No functional change.

(msaitoh)

Allow kernel to compile by preventing "error: stack usage is ..."
too-large errors.  What was there previously was not correct and this
will need to be revisited for debugger to ever become functional
anyway.

(scole)

Remove Xr to extattr(2) which we don't have.

(wiz)

reality check

(christos)

Fix struct entry size (thanks kre@)

(christos)

Add font files for Spleen, which can be loaded into the wsfont pool
or a wscons display device using wsfontload(8).

For example, Spleen 16x32 can be loaded and enabled as follow:

wsfontload -N spleen-16x32 -w 16 -h 32 spleen-16x32.fnt
wsconsctl -dw font=spleen-16x32

(fcambus)

Fix mm_md_kernacc() for 64 bit kernels (including n32):
- FAULT for any physical address less than start of cached XKPHY address.
- Pass any remaining physical address less then end of RAM.
- Pass any remaining physical address within the KEGS0 kernel address range.
Ignore all remaining addresses and fall back to uvm_kernacc() for checking
virtual address ranges.

Fixes pmap(1) (and probably other kmem grovellers).

(simonb)

Use prop_data_value (not prop_data_data_nocopy)

(skrll)

Remove support for storing settings in EFI environment variables now that
we have boot.cfg support.

(jmcneill)

make(1): fix evaluation of unreachable conditions

Since 2015-10-11, make had evaluated unreachable conditions even though
the manual page said it didn't.

(rillig)

make(1): demonstrate bug when evaluating conditions

(rillig)

Protect against multiple inclusion.

(simonb)

Fix /private/tmp mode to match etc/rc.d/perusertmp

(kim)

nameser_compat.h: correct preprocessor error syntax

(gutteridge)

Add missing debug entry for new t_mbtowc

XXX Please note that there are now TWO tests named t_mbtowc - this one in
XXX tests/lib/libc/stdlib/ and an older one in tests/lib/libc/locale
XXX I don't know if the functionality is duplicated or not.

(pgoyette)

fix bug introduced in conversion to kmem(9), passing address of the local
variable instead of pointer to allocated memory

should fix PR kern/55426 by Andreas Gustafsson

(jdolecek)

eval: Make radix optional even if minimum result width is specified.
This is what GNU m4 does too.

(uwe)

Improve markup for better PostScript output.
Make sure quotes are right and copy-pastable even in UTF-8 output.

(uwe)

command_version: print config file path

(jmcneill)

Make sure we pass full pages to fdt_memory_remove_range

(jmcneill)

fix the build

(christos)

Introduce genfs_pathconf() and use it for the default case in all filesystems.

(christos)

print the index of the problematic symbol

(christos)

Load boot.cfg from the EFI system partition instead of the first FFS
partition on that device.

(jmcneill)

Support loading boot.cfg from non file-system based devices.

(jmcneill)

Use UXA (not SNA) by default on Intel Gen3/4/5 chipsets (except 915).

This works around PR/54995 (screen corruption on N455)
and PR/55198 (X freezes/hangs on 965GM).  Ok'ed by mrg@.

Should be pulled up to netbsd-9.

(tsutsui)

constify, NFCI

(jdolecek)

make sure to strip leading '$' from NetBSD rcs id so that it won't be
replaces by cvs, and regen the headers again

(jdolecek)

regen again

(jdolecek)

regen - NFCI, confirmed by comparing compiled binary object before and
after

(jdolecek)

make this compile with normal WARNS

(jdolecek)

make this work with objdir, drop unused -ll

(jdolecek)

Ignore the supplied size, and always use the argument size that we know.
Found by maxv@

(christos)

mark several init-only functions as noinline to avoid bumping into the
stack limit

(jdolecek)

Ups. Fix shell parameter.

(jruoho)

belately note the stackgap removals in 2007

(jdolecek)

Add also a test that writes random garbage to every ifconfig(8) option that
takes parameters. Based on quick testing, iwn(4) and wm(4) pass, but
urtwn(4) panics. Use at your own risk; in some cases, it may be possible
that horrors are written directly to the hardware.

(jruoho)

prop_dictionary_[sg]et_cstring{,_nocopy} -> prop_dictionary_[sg]et_string{,_nocopy}

(jmcneill)

prop_data_create_data+prop_dictionary_set+prop_object_release -> prop_dictionary_set_data

(jmcneill)

prop_dictionary_get_cstring -> prop_dictionary_get_string

(jmcneill)

prop_data_data_nocopy -> prop_data_value

(jmcneill)

prop_data_data_nocopy -> prop_data_value

(jmcneill)

prop_data_data_nocopy -> prop_data_value

(jmcneill)

citrus SUBDIR is gone

(mlelstv)

Add the default TNF copyright (2005), cf. PR misc/55419.

(jruoho)

Add t_t_mbtowc.

(jruoho)

Start moving the remaining tests from src/regress to src/tests.

(jruoho)

Enable audio codec

(jmcneill)

avoid excessive stack usage in mp_cpu_start(), this is called after VM
init so kmem(9) can be used

(jdolecek)

Move the test for mktemp(3) to the right place.

(jruoho)

small constify

(jdolecek)

mark ahc_loadseq()/ahd_loadseq() as __noinline to avoid 'stack usage
might get unbounded' - seems the variable-length array using const variable
is misinterpreted as unbounded when inlined

(jdolecek)

Add a simple test case that writes random garbage to (almost) every sysctl node.
This test reproduced already at least five unique panics in a few quick runs.
The test is skipped by default as it is not likely safe even without the panics.

(jruoho)

apply __HIDE_DELAY so zfs and friends at least compile

(macallan)

Fix NULL deref on attach failure. Found via vHCI fuzzing.

Reported-by: syzbot+9fdcdc21799e5d6d75ee@syzkaller.appspotmail.com

(maxv)

Yet another idiotic compat syscall that was developed with literally zero
test made. Simply invoking this syscall with _valid parameters_ triggers a
fatal fault, because the kernel tries to write to userland addresses.

With specially-crafted parameters it is easy to completely escalate
privileges into the kernel.

Also the size of the allocation is just obviously wrong, but it looks like
the callers are even more wrong, so not gonna fix it for now.

Reported-by: syzbot+b05096f3114b2820d81c@syzkaller.appspotmail.com

(maxv)

Add a basic test for enabling/disabling network interface capabilities.

(jruoho)

Rename module directory for ibm4xx from /stand/powerpc-4xx to
/stand/powerpc-ibm4xx to match with what we set in evbppc_machdep.c:

https://nxr.netbsd.org/xref/src/sys/arch/evbppc/evbppc/evbppc_machdep.c#86

(And we use ibm4xx not 4xx for directory in sys/arch/powerpc.)

Note that we had never enabled MODULAR for ibm4xx yet; kernel with
MODULAR option was unable to build. This should be why this discrepancy
had never been found until today...

(rin)

Remove stale comments.

(rin)

Add a test case for PR kern/53767.

(jruoho)

Add test cases for different 802.11 options. These include cases for
PR kern/35045, PR kern/45745, and PR kern/55424.

(jruoho)

fpu_dumpfpn(): Add KASSERT to check argument != NULL.

(rin)

Return +INF without exception for +INF, as required by the architecture.
Also, avoid NULL dereference for fpu_dumpfpn() aka DUMPFPN().

Taken from (remaining part of) FreeBSD Revision 343078:
https://svnweb.freebsd.org/base?view=revision&revision=343078

(rin)

Remove extra newline in DPRINTF().

(rin)

fpu_dumpfpn(): Add missing \n in printf() format.

(rin)

Fix typo in DPRINTF.

(rin)

Use atf_pass to avoid test failure in case there are no suitable interfaces.

(jruoho)

Set FPSCR[VXSQRT] for non-zero negative operand.

Taken from (a part of) FreeBSD Revision 343078:
https://svnweb.freebsd.org/base?view=revision&revision=343078

(rin)

Sort headers. No functional changes intended.

(rin)

Restrict {MIN,MAX}_PAGE_SIZE for MODULAR || _MODULE, which makes
non-MODULAR kernel a little bit efficient.

They are also exposed to userland for jemalloc.

(rin)

Add new file containing a couple of UART prototypes.

Missing file pointed out by rin@.  Thanks!

(simonb)

Fix MODULAR support for ibm4xx by providing __nothing as VM_MDPAGE_INIT().

(rin)

Stop allocating struct cpu_info in BSS; No need to db_read_bytes()
against cpu_info, just ci_data.cpu_callout is enough.

Save 1408 bytes of BSS for, e.g., aarch64.

(rin)

Fix typo in name of evcnt(4) counter.

(rin)

Fix build failure due to -Werror=stack-usage.

Use db_read_bytes() against particular member of structure in use,
by which we can avoid to have whole structure in the stack.

Now, stack usage shrinks: 4240 --> 1296

(rin)

Drop redundant quoting in the nested printf example.

(uwe)

Use ASCII single quote.

(uwe)

Try to improve markup to get a nicer PostScript output.

While here edit slightly for consistency: get the octal character
ranges in order, use \num to refer to octal everywhere, etc.

(uwe)

Adjust prior to enforce minimum socket length includes sa_family

Not that the code strictly needs it, but if the macro is ever used
elsewhere then it makes sense as every sockaddr must have it.
The rest of the structure is dictated by the family and in some cases,
truncated on purpose so this is fine.

(roy)

Remove superfluous Pp.

(wiz)

Sort options and their descriptions.
Replace tab with space in table.
Make non-standard section subsections.
Comment out Xr to procsystime(1) which does not exist on NetBSD.

(wiz)

Remove !cold KASSERT - it does not compile on all kernels, and it is not the
right thing to test for anyway. XXX should we panic instead? Are "compatible"
strings this long happening in real devices?

(martin)

Avoid creating a full temporary softc struct on the stack.

(martin)

Avoid unbounded stack usage warning

(skrll)

Avoid large stack usage with kmem_alloc - we're in device_register so
there is no problem doing this.

(skrll)

Reference PRs consistently.

(jruoho)

Hack to get the "has ANSI C headers" check to pass on newer versions
of clang as bundled with Xcode.  If someone really cares about ensuring
NetBSD's host tools build on a system without <stdlib.h>, then I'm sure
that enterprising individual will also find a super #correct fix for this
specific problem.

(thorpej)

Add dtoverlay command to specify device tree overlays from the boot
loader command line.  Add support for specifying device tree overlays
in boot.cfg, with the syntax:

dtoverlay=/path/to/overlay.dtbo
dtoverlay=hd0e:/overlays/example.dtbo

Multiple overlays can be specified, and they are loaded in the order
they appear in boot.cfg.

Remove support for efiboot.plist.

(thorpej)

Fix fallout from migration from alloca() to malloc() in the previous.

of_compatible() is used at least for macppc in very early boot stage
where malloc() is still not available.

Therefore, for small (<= OFW_MAX_STACK_BUF_SIZE = 256) buffer, use
statically allocated one in the stack. For larger one, we continue to
use malloc() but KASSERT(!cold) is added for sure.

Fix boot failure for macppc reported by martin:
http://mail-index.netbsd.org/port-macppc/2020/06/25/msg002756.html

(rin)

Test creating thousands of bridge(4)'s.  Unlike with tap(4) (PR kern/55417),
this test succeeeds. It is even possible to have the ultimate ifconfig(8)
output with more than 65,000 devices.

(jruoho)

uvm_emap_size was removed a while ago

(jdolecek)

rearrange code to remove need for the scratch space variable, simply put
the data to destination bootinfo buffer directly

XXX compile tested only, needs confirmation that it still works

(jdolecek)

Verify that PR kern/52150 is no longer present.

(jruoho)

Register NVMM as an actual pseudo-device. Without PMF handler, to
explicitly disallow ACPI suspend if NVMM is running.

Should fix PR/55406.

(maxv)

Instead of UINT_MAX + 2, test what is claimed.

(jruoho)

Actually create the tap(4) with a negative device number. Doh.

(jruoho)

Include all necessary letters in the word "Cryptographic".

(thorpej)

If ubc_winshift gets constified, the extern declaration must be too.

(mlelstv)

Reference also PR kern/52526.

(jruoho)

Fix NULL deref. The original code before Jaromir's cleanup had an #ifndef
block that wrongly contained the 'else' statement, causing the NULL check
to have no effect.

Reported-by: syzbot+c41bbfe5a7ff07bf0f99@syzkaller.appspotmail.com

(maxv)

Add a test case for PR kern/52744, which no longer appears to be present.

(jruoho)

filemon_read: report error if lseek fails

Reviewed by: riastradh

(sjg)

Fix pastos.

(jruoho)

Add a test case for kern/52771.

(jruoho)

Fix PR kern/53333.

(jruoho)

Add a test case for PR kern/53410.

(jruoho)

Try to improve markup.

(uwe)

enable gcc stack usage limit for kernel functions, set to 3.5 KiB for now
as that seems to be enough to accomodate the current biggest stack usages

there are about six functions which use over 3KiB local stack, and
about a dozen between 2-3 KiB, so pushing this further needs more work
if desired

compile tested on amd64, i386, sparc64, sparc, powerpc (evbppc - BookE),
m68k (mac68k)

(jdolecek)

avoid variable-length array

(jdolecek)

Add test cases for PR kern/53546 and PR kern/55417. Both are skipped as
both reproduce panics.

(jruoho)

remove experimental direct pipe code (using uvm_loan()) I added in 2001 - it's
slower than the non-direct variant on MP systems, if anybody wants
to hack on this further it's available in Attic

(jdolecek)

Try to improve markup.

(uwe)

make ubc_winshift / ubc_winsize constant, and based on whatever is bigger
of (1 << UBC_WINSHIFT, MAX_PAGE_SIZE)

given that default UBC_WINSHIFT is 13, this changes behaviour only
for mips and powerpc (BookE/OEA), which will now have twice as much
memory used for UBC windows; if this ever becomes a problem, it's
possible to reduce ubc_nwins in MD code similar to what is done on sparc

this eliminates variable-length arrays in ubc_fault(),
ubc_uiomove(), and ubc_zerorange() so that the stack usage can be
determined and checked in compile time

(jdolecek)

Fix a tyop in a comment.

(simonb)

no need for alloca() in of_compatible(), malloc() is fine

(jdolecek)

Reference also PR kern/53860.

(jruoho)

Reference PRs consistently.

(jruoho)

Require root privileges.

(jruoho)

use maximum-size fixed size array instead of variable-length array
in uvm_aio_aiodone() so that the stack usage can be determined and
checked in compile time; this is not called recursively not
particularly deep in call stack, so there is no need to save every
last drop of stack space here

(jdolecek)

space, tab

(sevan)

Fix column layout

(sevan)

Don't assume everything is a wedge and has a parent device. On non-wedges
this crashes.

Instead, just talk to the referenced device and rely on the dk driver
to pass requests correctly.

(mlelstv)

Fix tyop in an acient comment.

(simonb)

Reduce ixgbe's busy loop using with workqueue and kpause.

- Use workqueue instead of softint to make some functions sleepable.
- Use new workqueue and enqueue it in ixgbe_local_timer() and
  ixgbe_recovery_mode_timer() to make them sleepable.
- Make new ixgbe_delay() and use it. This functions sleeps if the time is
  more than equals 1 tick. If it's not, do delay().

(msaitoh)

Use unsigned to avoid undefined behavior in ixgbe_fc_enable_generic().
Same as ixgbe_common.c rev. 1.24. Found by KUBSan.

(msaitoh)

Use a date that is actually a real date.

(uwe)

Bump date for previous.

(uwe)

Fix --error-output to be more like GNU m4.  It's a long version of -o.

(uwe)

Fix --error-output to be more like GNU m4.

GNU m4 --error-output is the same as -o despite the name.  It does NOT
affect warnings, error messages, and 'errprint' output so drop the
misguided bit of code that tried to freopen stderr without closing it
on failure.  Drop -e (which was our local invention) and make merge
--error-output with -o so that both set traceout.  Make trace_file()
preserve the old traceout on error and return error status so that the
caller can emit appropriate warning.

Do not yet support disabling tracing with an empty name, the rest of
the code is not ready, we don't do -o positionally and we don't have
`debugfile'.

(uwe)

don't try allocating 16KB of scratch space on stack

it's too early for kmem_alloc(), so use static variable in BSS; it's used
post reloc, so don't need to use the RELOC() macros

XXX compile-tested only on i386

(jdolecek)

reduce stack usage in ipf_nat_ioctl()

also, in SIOCADNAT, make sure to not leak kernel data

(jdolecek)

disable inlining for couple functions which allocate usb_device_request_t
on stack to reduce total run_init() stack usage from 7KB to 2KB

(jdolecek)

actually stop using the stack variable

(jdolecek)

avoid allocating almost 5k struct ibm561data on stack in ibm561_cninit();
it's too early for kmem_alloc(), so use static variable in BSS

(jdolecek)

rearrange to avoid allocating softc on stack in ug_isa_match()

(jdolecek)

avoid stack usage in match routine

(jdolecek)

No truss here

(sevan)

mdocify

(sevan)

remove unused x86_stos

(maxv)

mdocify

(sevan)

reduce stack usage in compat_60_ptmget_ioctl() - allocate struct ptmget
via kmem_alloc()

(jdolecek)

Note that -h is an extension to POSIX.  Bump date.

(riastradh)

New mv -h option.

`mv -h source target' just issues rename(source, target) without
discriminating on whether target resolves to a directory; this way
you can atomically replace a symlink to a directory.

(riastradh)

Try not to lose error output with --error-output.

Try to avoid the trap we set up ourselves while avoiding freopen(3).
When exit flushes and closes open streams it may close sfp first and
when it comes about to flush and close stderr, the descriptor is
already gone and we lose any buffered error output.  This actually
happens on some hosts, breaking --trace output used by autoconf.

(uwe)

reduce stack usage in dsl_scan_recurse() - allocate memory for
temporary zbookmark_phys_t using kmem_alloc() rather than stack;
this recuses several times usually, and this saves 2x
sizeof(zbookmark_phys_t) == 64 bytes per recursion

part of fix for PR kern/55402 by Frank Kardel

(jdolecek)

change dsl_scan_visitbp() to allocate blkptr_t dynamically rather than
on-stack - this function is called recursively, and the 120 bytes per call
add up; also remove unused variable

part of fix for PR kern/55402 by Frank Kardel

(jdolecek)

reduce stack usage in vdev_queue_io_to_issue() - zio_t is about 1KB, and
the function potentially recurses into itself

part of fix for PR kern/55402 by Frank Kardel

(jdolecek)

Add input files

(martin)

Fix directories

(martin)

Regen: Add AC_MSG_RESULT yes/no to the uio checks.

(uwe)

Add AC_MSG_RESULT yes/no to the uio checks.

(uwe)

fix tyop in KASSERT() condition which made it an assignment rather than a check

the field indeed is supposed to be set to GRANT_INVALID_REF at this moment,
the grant must be already revoked by this time

pointed out by Taylor R Campbell

(jdolecek)

Fix pasto; use {,U}LLONG{MIN,MAX} correctly in the {u,}llong{min,max}
range checks.

PR lib/55414

(thorpej)

Redo cnmac attachments - cnmacM @ gmxN @ pip0 @ iobus

Missed one file - thanks martin@.

(simonb)

Also install new tests.

(jruoho)

Ensure sockaddrs have valid lengths for RO_MISSFILTER.

Thanks to maxv@ for spotting this.

(roy)

As bin/55344 was fixed, note the flags also in rc.conf(5).

(jruoho)

remove special handling for symbolic links for COMPAT_43 lstat, it's
not necessary; this removes the only places in kernel which did namei
LOOKUP with LOCKPARENT

fixes diagnostic KASSERT() in namei() code

Reported-by: syzbot+628382ecf1438e53d08d@syzkaller.appspotmail.com

(jdolecek)

Fix references in comments.

(jruoho)

Check that fstat(1) works (cf. PR kern/55407).

(jruoho)