Functionize the same code related to ctrl vq in vioif(4)

(yamaguchi)

Divide some elements of vioif_softc into txq, rxq, and ctrlq

(yamaguchi)

Make macros not depend on vioif_softc

(yamaguchi)

Remove recently-added entry concerning routines whose return type
had been changed from void to int.  This is now handled by the new
void-typed hooks.

(pgoyette)

Create a variant of the HOOK macros that handles hook routines of
type void, and use them where appropriate.

(pgoyette)

Fix infinite loop like in fdtbus_get_string_index.

(mlelstv)

Whitespace (indent with tabs, rather than spaces), and rearrange one
line that needed wrapping so it no longer does, and is (IMO) clearer.
WHile here, note that this was created in 2019, not 2018, for the
copyright notice...

NFCI.

(kre)

add call error checks, requested by mrg@

(christos)

Note the desire to return some callback functions to having a void
type, rather than int (changing them was necessary to use a common
return type for all callbacks).

(pgoyette)

normalise comment for fileassoc(8).  from PR#44873.

(mrg)

Note that rtsock_50 is now functional, even though the implementation
remains sub-optimal.

(pgoyette)

Add the required hooks for rtsock_50 and modify the COMPATCALL() macro
to use the hooks.  While the rtsock_50 situation is still sub-optimal
(it includes the main rtsock.c with a whole bunch of function and
variable redefinitions via macros), this at least makes it possible to
load the rtsock_50 code separately from more recent code, rather than
the previous requirement that rtsock_50 be built-in.

(pgoyette)

switch sparc pmap lock to the scheme sparc64 uses:

- local IPL_NONE mutex for general pmap locking operations, not
  kernel lock.
- for pmap_activate()/pmap_deactivate(), switch to using the
  existing ctx_lock, and push handling of it into ctx_alloc() the
  ctx_free() callers.

fixes easy to trigger deadlocks on systems with >2 cpus.  without
this patch i usually hang during boot.  with it, i was able to
push the machine hard for over 12 hours.

XXX: pullup-8, and maybe -7.

(mrg)

regen for 2.25

(snj)

Incite the "Purple Cow" wrath of Gelett Burgess upon ourselves.

(jakllsch)

.Pp before .Bd is redundant.
From Edgar Pettijohn.

(uwe)

Increase the timeout a bit, and make sure we join so that there is no
race.

(christos)

Mark the fs/vfs/t_renamerace:udf_renamerace_dirs test case as an
expected failure referencing PR kern/53865, and force failure to avoid
reports of unexpected success as it does not realiably fail under
qemu.  This makes the treatment of udf_renamerace_dirs the same as
that of udf_renamerace, only with a different PR.  Also, make
whitespace consistent between the two.

(gson)

Forgot to commit file along with identcpu.c::rev1.86.

(maxv)

On certain AMD f10h CPUs (like mine), the BIOS does not enable WC+. It
means that the guest pages that are WC+ become CD, and this degrades
performance of the guests.

Explicitly enable WC+.

While here clarify the AMD identification code.

(maxv)

Don't mark the traceme_vfork_crash_bus test case as an expected failure
when it is executing as part of the lib/libc/sys/t_ptrace_waitid test,
because that particular combination is not failing, at least not on amd64.
It is failing as part of other tests, so PR lib/53343 still applies.

(gson)

Remove the HOOK2 versions of the MODULE_HOOK macros.  There were
only a few uses, and using them led to some lack of clarity in the
code.  Instead, we now use two separate hooks, with names that
make it clear(er) what we're doing.

This also positions us to start unraveling some of the rtsock_50
mess, which will need (at least) five hooks.

(pgoyette)

Handle more corner cases, clean up a little, and add a set of instructions
in Group1.

(maxv)

Reset DR7 before loading DR0-3, to prevent a fault if the host process
has dbregs enabled.

(maxv)

Error out if the higher 32 bits of DR6 and DR7 are set. MOV DR would
fault otherwise.

(maxv)

Rearrange a bit, put all the sysctl-related stuff at the end of the
file, and enclose it in a single ``#ifdef COMPAT_RTSOCK ... #endif''
block.

XXX Arguably, this code might better belong in its own source file,
but I'll leave that for a future project.

(pgoyette)

add a note about pmap_activate() and pmap_deactivate() not being allow to block.

(mrg)

Always return from expandm() with errno unaltered, so on the
off chance it failed, there's still the possibility that whatever
processes the result will be able to deal with the %m that would
(presumably) be left in the format string.

And as a frill, don't call strerror() until we know we are
going to use its result (still call it only once, no matter
how many %m's are in the format string).

(kre)

update to reality:
- note that the base system supports 32 bit / 64 bit compat directly now
- list all the expected paths for ld.elf_so
- remove largely obsolete info
- fix list of known issues
- misc other clean up

(mrg)

check for *. integer overflow over ptrdiff. Pointed out by kre@

(christos)

clang does not like to deref a null pointer unless it is qualified volatile

(tnn)

Make expandm() return a const char * so we can do away with __UNCONST()
and more importantly, so that we don't accidentally return a value
that is a const char * in reality (pointer to read only string) as a
char *.

(kre)

Don't do %%m -> %%%m escaping, requested by gson@.

(christos)

Return the buffer that has not been freed, so it can be later,
rather than the one that already was.

(kre)

adjust to new expandm.

(christos)

Provide an extra argument to store the returned pointer so we can use
the function directly as the return format (before assigning it to a
variable) to appease clang.

(christos)

print the program name when we don't find a library.
XXX: really most _rtld_error() should do that, and most do.

(christos)

Setup default wpa_supplicant flags so that it:
  * Writes diagnostics to syslog
  * Enables interface matching to hotplug wireless interfaces
  * Uses /etc/wpa_supplicant.conf

(roy)

Mention about irisboot for sgimips pre-ARC machines.

(tsutsui)

Install /etc/wpa_supplicant.conf

(roy)

Add a default wpa_supplicant.conf which allows wpa_cli(8) to configure
wpa_supplicant(8).

(roy)

Add irisboot binaries for pre-ARC machines.  PR port-sgimips/53539

(tsutsui)

Import irisboot, yet another bootloader for pre-ARC sgimips machines.

This bootloader adds support for Personal IRIS 4D/25 (IP6),
Personal IRIS 4D/35 (IP12), and Indigo R3000 (IP12) etc. and
optionally supports R4000 Indy and Indigo R4000 (ARCS machine)
using standalone WD33C93 SCSI driver.

Submitted by Naruaki Etomi via PR port-sgimips/53539,
and reviewed by tsutsui@.  See also port-sgimips@ ML thread:
https://mail-index.netbsd.org/port-sgimips/2018/09/16/msg000778.html

(tsutsui)

Explicitly size the regmap array.

(thorpej)

partially back-off rev. 1.68 - when the drive reset for port 15 fails,
on some systems controller isn't actually able to process further commands,
it's really necessary to do explicitely reset for drive 0 too

towards resolution of PR kern/53307

(jdolecek)

Add NOLUNS quirk for more SEAGATE SCA/WIDE drives.

Tested on NetBSD/luna68k and LUNA with SCA 80pin -> NARROW 50pin and
WIDE 68pin -> NARROW 50pin connectors.

(tsutsui)

Allow zfs_vn_rdwr() with "rw == UIO_READ" too.

(hannken)

Pass unmount flags down to syscall, "zfs unmount -f" now works.

(hannken)

The ZFS onexit routines expect opening ZFS_DEV with O_EXCL to return
a cloned device with an unique minor number.

Use fd_clone() on this condition to return a cloned device descriptor.

(hannken)

Disable valid assertion "!(flags & TQ_NOQUEUE)" as "zfs send" triggers it.

(hannken)

Need FOF_UPDATE_OFFSET to update the offset on plain files.

Running "zfs send" to plain files no longer writes garbage.

(hannken)

There is no such function read_rflags(), probably meant x86_read_flags().

(hannken)

Simplify regmap initialization, and fix an regmap issue that
affected TI OMAP (LCR register would get clobbered due to
using the wrong offset for the MDR1 register) reported by Lwazi Dube
(who also found the root cause).

(thorpej)

- don't try to access a nonexistant video BIOS
- don't blindly assume VGA output
- on mobility chips assume LVDS and VGA outputs if no other info is available
Now this works again on my Pismo

(macallan)

use expandm from libwrap.

(christos)

Use a utility function to expand %m correctly...

(christos)

let's try one more time: prepare the format to pass to syslog instead
of trying to escape %m.

(christos)

Ticket #1671

(martin)

Pull up following revision(s) (requested by skrll in ticket #1671):

sys/dev/usb/usbdi.c: revision 1.181

Avoid use after free when closing interrupt pipe transfer by calling
upm_close before destroying the interupt pipe xfer.

Found by kasan on wiz's machine.

(martin)

Tickets #1158, #1161 and #1162

(martin)

Pull up following revision(s) (requested by knakahara in ticket #1162):

sys/net/if_ppp.c: revision 1.162
sys/net/if_ppp.c: revision 1.163

Fix missing mutex_exit in ppp_create().

Fix missing splx in ppp_inproc().

(martin)

Pull up following revision(s) (requested by skrll in ticket #1161):

sys/dev/usb/usbdi.c: revision 1.181

Avoid use after free when closing interrupt pipe transfer by calling
upm_close before destroying the interupt pipe xfer.

Found by kasan on wiz's machine.

(martin)

Add a bs_barrier to make ThunderX xhci work.  Not sure why this is needed,
but I'll work it out later.

(skrll)

Style

(skrll)

Pull up following revision(s) (requested by sevan in ticket #1158):
distrib/common/cgdroot.rc: revision 1.2
distrib/common/cgdroot.rc: revision 1.3
distrib/common/cgdroot.rc: revision 1.4

Don't hardcode wd0a in cgdroot.kmod, try NAME=cgd.conf and ROOT.a.

-

Break the loop after a successful mount.

-

ROOT.a doesn't resolve to a useful partition, hardcode wd0a and ld0a for
now.

(martin)

instead of handling %m inline, escape it.

(christos)

Add missing semicolon.

(msaitoh)

Some translation fixes/additions from Kamil

(martin)

workaround for PR kern/53848

(knakahara)

Don't accept OIFLIST operation unless the rtsock_70_hook is loaded,
even though the results are otherwise identical to those on current.

(pgoyette)

Remove extra PHY read in brgphy_mii_phy_auto().

(msaitoh)

Rework the various sysctl-related routines to call the correct code
for each version.  While here, extract the 5.0 specific code instead
of including in the main rtsock.c code.

Also, clean up all the sysctl-related routines to prevent building
more than one copy, no matter how many places rtsock.c gets #include'd
into!

(pgoyette)

Separate macro arguments with <comma><space> for readability.

(pgoyette)

At some point we lost the text that explains the difference between
clear() and erase(). Or possibly we never had any. Put some back.
Bump date.

(dholland)

Set I82579_LPI_CTRL register correctly in wm_set_eee_pchlan().

(msaitoh)

Fix missing splx in ppp_inproc().

(knakahara)

Fix missing mutex_exit in ppp_create().

(knakahara)

disable pause, only works on v9

(christos)

fix compilation

(christos)

Hook vboxvideo driver into the build.

(uwe)

Fix handling of RTKQ_IM_HW quirk, the check was inverted.

Found by tho@useless-ficus.net, thanks.
Fixes PR/53858.

(mlelstv)

Avoid use after free when closing interrupt pipe transfer by calling
upm_close before destroying the interupt pipe xfer.

Found by kasan on wiz's machine.

(skrll)

handle errors from loading the syscalls and bail early. establish sysctl
later.

(christos)

PR install/53857: make the pkgsrc menu wider, so the first column fits
in various translations. While there shorten some (misleading) text.

(martin)

PR/53856: F. Aragon: editline/libedit not prompting colors in readline mode

(christos)

no yield for early arm

(christos)

Gut omapip support since the socket stuff to support it is gone, and
adjust to the new isc library reality.

(christos)

consolidate hint@pause in one place. remove mf.a from memory barriers

(scole)

Remove duplicate length check (undo diff to upstream)

(maya)

Undo merge botch, duplicate length checks

(maya)

pEnt isn't initialized at this point, don't free it
Undo diff to upstream.

There's a good chance we aren't using this code at all.

(maya)

Undo unnecessary diff to vendor branch.

(maya)

delete unused variable

(christos)

PR/53851: Andreas Gustafsson: libwrap prints "m" instead of errno
Handle %m inline if needed, otherwise vasprintf strips the %m and...

(christos)

Import fix from HEAD.

Now that the test behaves normally under failure (mismatched actual vs
expected outputs), maybe I can figure out why it actually fails.

(pgoyette)

Remove extraneous && from check_output()

With this extra && the test case exits abnormally when the actual and
expected outputs don't match.

(pgoyette)

- move pmap_ctx() macro into pmap.h instead of duplicating it.
- fix description of some mach ddb commands.

(mrg)

Use the make pid when naming the temporary file to create aslcompilerparse.y,
so tow parallel make instances won't step on each others toe.
Pointed out by Christos.

(martin)

don't mark a static function with default visibility.  (avoids warnings.)

(mrg)

add new source for libdrm 2.4.96.

(mrg)

When handles returned by the device do not match our expectations,
log more details.

(martin)

build fixes for libdrm 2.4.96.

XXX: may still have build issues, and cleandir before
XXX: update builds is likely needed

(mrg)

merge libdrm 2.4.96.

(mrg)

Remove extra register read in bce_mii_write().

(msaitoh)

Optimize:

* Don't save/restore the host CR2, we don't care because we're not in a
  #PF context (and preemption switches already handle CR2 safely).

* Don't save/restore the host FS and GS, just reset them to zero after
  VMRUN. Note: DS and ES must be reset _before_ VMRUN, but that doesn't
  apply to FS and GS.

* Handle FSBASE and KGSBASE outside of the VCPU loop, to avoid the cost
  of saving/restoring them when there's no reason to leave the loop.

(maxv)

Update comment (overlooked in r1.179).
From Jos� Luis Rodr�guez Garc�a in PR kern/53849.

(martin)

KNF. No functional change.

(msaitoh)

Include newly-created files in compat_netbsd32 module build.  Missed
this when resolving conflicts in most recent sync-with-HEAD

(pgoyette)

tests/net/if_ipsec/t_ipsec disable dad. This may fix PR kern/53848

(knakahara)

config.h for vboxvideo driver.

(mrg)

build infrastructure for vboxvideo driver.

(mrg)

add vboxvideo.

(mrg)

wd(4): police my grammar police

(gutteridge)

wd(4): minor grammatical improvement

(gutteridge)

- delete trailing whitespace
- remove non-functioning urls
- update new software
- add new field to indicate last update date.

(christos)

use 32 bit counters on 32 bit platforms because we don't yet have 64 bit
atomics for them.

(christos)

Allow argument expansion in menu titles as well.

(martin)

A driver that's been in GENERIC since 2011 cannot reasonably be called
"experimental". Bump date.

(dholland)

whitespace cleanup

(scole)

Spell "occurrence" correctly.

(uwe)

new bind

(christos)

bump bind libraries.

(christos)

merge conflicts, bump major, fix build.

(christos)

Correct an (old) typo in a comment.  NFC - it is just a comment.

(kre)

Fix the code taken from FreeBSD 2 revisions back, which fixed
aliases, to actually do what it was supposed to do, and not just
come close by accident.  (How broken this was, while still seeming
to work perfectly most of the time was truly amazing!)

This corrects the behaviour of an alias defined with a blank char
as the last of its value, to correctly do an alias lookup on the
word that follows the alias.

(kre)

Update some dead (#if 0'd) code that is never called to
cope with the changes made in the previous revision, in an
attempt to avoid bit rot.

Untested (uncompiled) - though it should work.

NFC: this change doesn't get compiled, let alone used.

(kre)

A similar fix to that added in 1.169 of eval.c, but here for when
processing command substitutions.  If there is an error while processing,
the any pending queued input should be discarded.  From FreeBSD.

(kre)

When an error occurs in a builtin from which we do not exit
(a normal builtin, though those are not genrally an issue for
this problem, or a special builtin that has been prefixed by "command")
make sure that we discard any pending input that might have been
queued up, but not yet processed.

We had the mechanism to fix this from when expansion of PS1 etc
was added (which has a similar problem to deal with) - all taken
from FreeBSD - but did not bother to use it here until now...

This fixes an error detected by newly added ATF tests of the eval
builtin, where
eval 'syntax error
another command'
would go ahead and evaluate "another command" which should not
happen (note: only when there was a \n between the two).

(kre)

Add several sub-tests to the test for the "eval" builtin,
including testing correct handling of error cases.
One of these new tests located a bug which will be fixed
within minutes of this commit ...

While doing this, move the test that was used only in the
echo builtin test case (testing if the NetBSD shell was being
tested) into a utility function, and also call it in the eval
test, so NetBSD specific behaviour of what is unspecified in
the standard can be checked to ensure it is not accidentally
altered.

(kre)

avoid function param names in prototypes.

(mrg)

Regen - otherwise compat_40_mount() is wrongly included in sysent.c

Missed this during earlier conflict resolution.

(pgoyette)

More markup. Add missing El. Fix typo.

(wiz)

Make aslcompilerparse.y generation atomic to avoid parallel build races.
Solution suggested by mrg.

(martin)

- Remove an extra register read in wm_kmrn_lock_loss_workaround_ich8lan().
- Don't leave the MDICNFG register modified when the Power Management
  capability offset can't get.

(msaitoh)

No functional change.
- u_int{16,32}_t -> uint{16,32}_t
- KNF.

(msaitoh)

Fix compile error for *_pci_intr_establish().

(msaitoh)

document crashme(9) functionality.

XXX: install me

(mrg)

crashme: a framework to test kernel faults.

so far, only a basic panic() and null deref nodes are added.
with options DEBUG, one can now use:

  # sysctl -w kern.crashme_enable=1

  # sysctl -w kern.crashme.panic=1
  # sysctl -w kern.crashme.null_deref=1

to trigger a crash.  crashme_enable must be set to 1 before any
of the nodes will be writeable.

supports dynamic additional/removal of crashme nodes.

(obsoletes kern.panic_now, which will be removed later.)

(mrg)

use <sys/param.h> instead of <machine/param.h> for everything.

(christos)

PR/53845: Anthony Mallet: "bad cookie" in authoritative DNS server since bind
9.12 import. AES_CC needs to always be enabled.

(christos)

remove <machine/param.h> include, nothing from it is actually used here

(jdolecek)

remove explicit <machine/param.h> include, code including this already includes
<sys/param.h>

(jdolecek)

remove explicit <machine/param.h> include, code including this already includes
<sys/param.h>

(jdolecek)

no need to include <machine/param.h> if <sys/param.h> already included

(jdolecek)

no need for <machine/param.h>

(jdolecek)

update groff (Steffen Nurpmeso)

(christos)

no need for <machine/param.h>

(christos)

no need for <machine/param.h>

(christos)

no need for <machine/param.h>

(christos)

no need for <machine/param.h>

(christos)

no need for machine/param.h

(christos)

- put all the GPLv3 software together
- put all the Last GPLv2 version software at the end.
Perhaps it is better to use separate files (per license) at this point?

(christos)

Make TEGRA124_XUSB_BIN_STATIC and TEGRA210_XUSB_BIN_STATIC compile again.

(jakllsch)

Provide global variables with C linkage, otherwise they will create
conflicts for C++ modules use.

(joerg)

Include <sys/param.h> for DEV_BSIZE/DEV_BSHIFT

(christos)

Include <sys/param.h> for DEV_BSIZE/DEV_BSHIFT

(christos)

Include <sys/param.h> for DEV_BSIZE/DEV_BSHIFT

(christos)

Include <sys/param.h> for DEV_BSIZE/DEV_BSHIFT

(christos)

Someone should merge all the wdc copies.
Include <sys/param.h> for DEV_BSIZE/DEV_BSHIFT

(christos)

Optimize: don't keep a full copy of the guest state, rather take only what
is needed. This avoids expensive memcpy's.

Also flush the V_TPR as part of the CR-state, because there is CR8 in it.

(maxv)

Actually, according to the grammar the square brackets in the "tcp
flags" are not literal, so use .Op to show that /mask is optional.

(uwe)

Restore macro with effect.  Fix the real problem that prevented it to
have the effect.

(uwe)

New sentence, new line. Punctuation fixes. Remove macros without effect.

(wiz)

u_int{8,16,32}_t -> uint{8,16,32}_t. No functional change.

(msaitoh)

Whitespace fix.

(msaitoh)

Remove a format character from a non-format string.

(tih)