Add two new RDC PHYs from Andrius V.

(msaitoh)

Regen.

(msaitoh)

Add two new RDC PHYs from Andrius V.

(msaitoh)

use PMAP_PAGE_INIT() to initialize mutex in pmap_page.

VM_MDPAGE_INIT() in pmap_free_pdp() had initialized pp_flags,
so it unintentionally cleared PMAP_PAGE_FLAGS_PV_TRACKED.
use PMAP_PAGE_INIT to avoid using PMAP_PAGE_FLAGS_PV_TRACKED.

pointed out by tnn@, thanks

(ryo)

The specification calls this a class, not type.

(is)

lookup_fastforward(): failure to vget foundobj vnode also translates into
EOPNOTSUPP; VOP_LOOKUP() should retry it.

(ad)

For single page I/O, use direct mapping if available.

(ad)

ubc_direct_release(): remove spurious call to uvm_pagemarkdirty().

(ad)

PR kern/54759: vm.ubc_direct deadlock when read()/write() into mapping of itself

Prevent ubc_uiomove_direct() on mapped vnodes.

(ad)

fix typo in comment.

(is)

Multilink fragment protocol type.

(is)

Multilink PPP: sanity check of option values, storage of remote MRRU.

(is)

Define a few more LCP options. Recognize, sanity-check and report (but
still reject for the moment) multilink PPP configuration options received.

(is)

PR kern/55146 (100+ file system test cases failing)

- namei_oneroot(): key on negative return from lookup_fastforward()
  (EOPNOTSUPP), not positive.

- lookup_crossmount(): don't lose track of founddir.

From hannken@, with a couple of tweaks.

(ad)

Add history section

(sevan)

add wrapper to free response from xenbus_dev_request_and_reply(), so
that call in xenbus_dev_write() wouldn't need to know how it's allocated

(jdolecek)

add a small wrapper xenbus_directory_free() to free result of
xenbus_directory(), so that caller doesn't need to be aware how the memory
was allocated

(jdolecek)

mostly convert to kmem_alloc()

doing this, remove check for failed allocation with KM_SLEEP, and make
sure to not hold mutex during the call

(jdolecek)

partially convert to kmem_alloc()

plug memory leak in one xenbus_probe_device_type() error path
when read_backend_details() fails

(jdolecek)

convert from malloc() to kmem_alloc()

(jdolecek)

convert the node watch code to use kmem_alloc() instead of malloc()

(jdolecek)

make xenbus_watch_path() static, it's not used outside xenbus_client.c

(jdolecek)

revert the watch.node change, xenbus_watch_path() and xenbus_watch_path2()
need to use non-constant path

(jdolecek)

drop wd lock in wdstart1() before calling the ata_bio hook; when called
from ata thread context, that can still need to sleep for wdc attachments
in wdcwait()

(jdolecek)

Recognize more rdc devices (Andrius V.)

(christos)

regen

(christos)

Add more RDC products (Andrius V)

(christos)

change xenbus_read() interface so that caller supplies the buffer and it's
size, caller doesn't free(9) the returned value any more

(jdolecek)

Disable userconf in the RAMDISK kernel to shrink it a bit

(martin)

Ooops, editor mishap in previous

(martin)

switch from malloc() to kmem_zalloc()

(jdolecek)

no need to malloc()+copy watch.node, it's constant string - either "device"
or "backend"; just use the strings direct

(jdolecek)

convert from malloc() to kmem_alloc()

(jdolecek)

switch to kmem_zalloc() instead of malloc() for struct kernfs_mount

(jdolecek)

switch KERNFS_ALLOCENTRY() to use kmem_zalloc() instead of malloc()

(jdolecek)

Fix KASAN build on aarch64

(skrll)

remove <sys/malloc.h> include, not used here

(jdolecek)

avoid "panic: LOCKDEBUG: Mutex error: rw_vector_enter,309: spin lock held"

ok nonaka@. thanks

(ryo)

bring the section on automatic acquisition of kernel_lock up-to-date

(jdolecek)

Whitespace

(skrll)

adjust previous - set the Tx flag only when the feature supported, instead
of first setting it and then masking it when unsupported

(jdolecek)

Exit successfully after printing the search path, stop further processing.

Continuing to process had the unintended effect that `man` failed to find
a matching manual page for {EMPTY LIST OF REQUESTED MANUAL PAGES}, and
exited with 1.

Prompted by a fish shell snippet that tried and failed to distinguish between
FreeBSD man (-p takes argument) and NetBSD man (-p no argument) by comparing
`man -p` exit code.

ok riastradh, logix (which also pointed out the manual page SYNOPSIS is
saying man -p should be used this way).

(maya)

pass and use feature-ipv6-csum-offload for ipv6 csum support, matches
Linux Dom0/DomU

(jdolecek)

use conf/filesystems.config, fix missing PTYFS

(bouyer)

add known_mpsafe parameter also to pirq_establish(), and pass the parameter
to underlying event_set_handler()

(jdolecek)

convert to IFEF_MPSAFE, also enable interrupt handler without biglock

no performance difference observed compared to version before change,
for neither UP nor MP DomU

(jdolecek)

Document the phil-wifi branch

(martin)

remove restriction on interrupt level for MP-safe interrupt handlers

(jdolecek)

just pass true as known_mpsafe for interrupt establish, drop the
(level != IPL_VM); NFCI, this uses IPL_HIGH so was mpsafe already

(jdolecek)

make a pass on locking, replacing spl*() calls with mutexes:
- sc_tx_lock covers any access to tx list, tx ring, and writes to
  if_flags and if_itimer
- sc_rx_lock covers any access to rx list, tx ring, and free tx counter

fix suspend and detach to work again - recent softintr changes made
xennet_tx_complete() not actually being called, because the call
in xennet_handler() was after IFF_RUNNING check; now it's called
directly rather than triggering softint, with updates locking it's safe

enable DVF_DETACH_SHUTDOWN for xennet(4), though the call only
triggers notification to backend to close, leaves actual detach to
xenwatch_thread

(jdolecek)

#undef XENNET_DEBUG, accidentaly left enabled in previous commit

(jdolecek)

Tickets #820, #822, #823

(martin)

Pull up following revision(s) (requested by riastradh in ticket #823):

sys/dev/cgdvar.h: revision 1.19
sys/dev/cgd.c: revision 1.122
sys/dev/cgd.c: revision 1.123
sys/dev/cgd.c: revision 1.124

Defer crypto operations to a workqueue and make it utilize all CPUs.

Make device mpsafe.

Some code cleanup.

Don't wait for data buffer.

cgd: switch from malloc(9) to kmem(9)
XXX might be worthwhile to use pool_cache(9) in the write path

(martin)

Pull up following revision(s) (requested by riastradh in ticket #822):

sys/dev/dkwedge/dk.c: revision 1.99

Allow dumping to cgd(4) on a dk(4).

(Technically this also allows dumping to a dk(4) on which there
happens to be a cgd(4) configured, but I'm not sure how to
distinguish that case here.  So don't do that!)

(martin)

Pull up following revision(s) (requested by pgoyette in ticket #820):

sys/dev/wscons/wsevent.c: revision 1.43
sys/dev/wscons/wsevent.c: revision 1.44

Make default protocol version used by wscons selectable and default
to the current version.

Fixes PR 55103.

KNF (Format block comment)
NFCI

(martin)

udf_get_node(): fix typo and honor "lktype" argument.

(hannken)

Sync image size with available space reserved in the RAMDISK kernel

(martin)

postinstall: fix contents_owner to return an error on error

find returning nothing via stdout but does return an error is an error.
Fixes the case where dhcpcd chroot db directory isn't owned by _dhcpcd.

(roy)

Use full available ram disk space

(martin)

Slightly bump ramdisk space

(martin)

remove some XENNET_DEBUG code which causes kernel faults

(jdolecek)

remove some unnecessary spl*() calls, and fix the rxreq va/pa deallocation
in xennet_xenbus_detach()

(jdolecek)

save some memory - no need to embed sc inside struct xennet_rxreq

(jdolecek)

Sync "common" declarations with ifdefs in header.

(martin)

Print CPUID 0x80000007 %edx on both Intel and AMD.

(msaitoh)

Remove ci_max_ext_cpuid because it's the same as ci_cpuid_extlevel.

(msaitoh)

define M_CSUM_BLANK, eventually to be used by xennet(4)/xvif(4) to offload
checksums to NIC hw

(jdolecek)

Rename CPUID_APM_TSC to CPUID_APM_ITSC. No functional change.

(msaitoh)

Welcome to 9.99.55!

struct proc regained p_oppid.

(kamil)

Remove copy-paste garbage.

(rin)

set rxreq_gntref to GRANT_INVALID_REF immediately after revoking
access, so it's clear where it comes from

(jdolecek)

Reintroduce struct proc::p_oppid

Relying on p_opptr is not safe as there is a race between:
- spawner giving a birth to a child process and being killed
- spawnee accessng p_opptr and reporting TRAP_CHLD

PR kern/54786 by Andreas Gustafsson

(kamil)

dion't define endian as well as try to declare it as a variable.

(mrg)

CPUID Fn00000001 %edx bit 8 is printed as "TSC", so rename CPUID Fn8000_0007
%edx bit 8 from "TSC" to "ITSC" (Invariant TSC) to avoid confusion.

(msaitoh)

fix comment in typo

(snj)

Cleanup CPU attach output:
- Always print the core's vendor and product name.
- Print the CPU ID on the same line as the name. Single line of dmesg
  per core.
- Use aprint_verbose for reporting additional details.

(jmcneill)

new OpenSSL/binutils

(christos)

merge conflicts

(christos)

Prevent waiting for processes to exit forever by introducing _rc_kill_ntries,
which if set, will SIGKILL the processes that did not die yet.

(christos)

Switch USB to use non-coherent buffers for data transfers in the
same way as OpenBSD.

The use of coherent (uncacheable on ARM and other arches) mappings
for transfer buffers impacts performance, espcially where memcpys
are involved.

Audit the necessary usb_syncmem operations - a few were missing.

(skrll)

There is no "s" lock.

(christos)

- Untangle spawn_return by splitting it up to sub-functions.
- Merge the eventswitch parent notification code which was copied in two
  places (eventswitchchild)
- Fix bugs in the eventswitch parent notification code:
  1. p_slflags should be accessed holding both proc_lock and p->p_lock
  2. p->p_opptr can be NULL if the parent was PSL_CHTRACED and exited.

Fixes random crashes the posix_spawn_kill_spawner unit test which tried
to dereference a NULL pptr.

(christos)

fix the mrg arm entry text that said v7 + v7, not v7 + v8.

noticed while checking what was wrong with the prior ;)

(mrg)

remove RX flip from diagnostic, it was removed

(jdolecek)

fix date for mrg arm entry

(jdolecek)

remove xengnt_grant_transfer() and xengnt_revoke_transfer(), nothing
uses it any more

(jdolecek)

note removal of xen rx-flip

(jdolecek)

remove support for legacy rx-flip mode for xennet(4)/xvif(4), making
rx-copy (first shipped in NetBSD 6.0 in 2012) the only supported
mode

this is mostly to simplify maintenance and future development

rx-flip is not supported by Linux Dom0/DomU, and NetBSD Dom0/DomU
defaults to rx-copy for over 8 years now too, so there is little
need to keep the support for compatibility

besides compatibility there is no other reason to keep rx-flip -
page transfer is generally slower than copy due to necessary MMU/TLB
manipulation, especially on MP systems

(jdolecek)

Don't block signals when running hooks. Here ctx->sigset == defsigs ==
0x60006003, which blocks SIGTERM by default, so running something simple
from a hook, like /etc/rc.d/racoon restart does not work. The script
is then stuck waiting for the daemon to die, which it won't since it will
never receive the signal, and the hook never terminates.

(christos)

Fix depenency on common symbols in sbin.

(joerg)

Return early on 0-sized transfers (usually to/from NULL-objects)

This logic is already present in subr_copy.c:copyin_vmspace() and
rumpcopy.c:copyinstr().

This avoids memcpy() calls for NULL objects that is Undefined Behavior,
allowed in the kernel space (-fno-delete-null-pointer-checks), but not
in userland.

Reported by UBSan.

(kamil)

Conditionalize a few more functions (hopefully fixing the build)

(martin)

Support the use of DHCP in the install environment

(sevan)

no need to pollute dmesg by xennet_watchdog(), turn the message into DPRINTFN()

(jdolecek)

remove nsmb from MD list of devices to create for 'all'

(jdolecek)

We generate this.

(christos)

tools: Build tic with full fat terminfo

(roy)

terminfo: Add guards to optionally build parts of libterminfo

Reading from a database is now optional.
Compiling terminfo descriptions (including from $TERMINFO) is now optional.
Compat support is now optional.

This removes 17k on amd64 from the binary size, which allows it to be used
again on space constrained ramdisks.

(roy)

Adapt to libterminfo as part of libhack-curses

(martin)

Extend the curses hack to terminfo - build selected parts of
libterminfo as part of libhack, so we can drop unwanted features
for small install media.

(martin)

add entry for NetMos NM9900 Octa UART

notice: On the boards I've obtained, the number on the cable stamp
doesn't match the actual unit number of the attached device.
Cable's 0,1,2,..,7 corresponds to comN+0,2,4,6,1,3,5,7 :-(

(ryo)

revert to previous

(mlelstv)

PR/55137: Kouichi Hashikawa: ipfstat -f incorrect output
- make sure frag is initialized to 0
- initialize ipfr_p field

(christos)

PR/55137: Kouichi Hashikawa: ipfstat -f incorrect output
Fix incorrect byte order.

(christos)

we don't delete entries, we mark them obsolete so that postinstall can
delete the files.

(christos)

us and them, then!

(christos)

Allocate PV entries in PAGE_SIZE chunks, and cache partially allocated PV
pages with the pmap.  Worth about 2-3% sys time on build.sh for me.

(ad)

NetBSD 9.99.54 - namei changes.

(ad)

Document and set a default for KERNEL_DIR

(christos)

Switch to type generic PRIu macros

Fixes build on i386.

(kamil)

make sure _OSRELEASE is defined where it is needed.

(christos)

Fix LLVM sanitizers build after removal of netsmb

Cherry-pick upstream commit:

From 695ca346eb0c724fb79c6f166cfecd736eae6608 Mon Sep 17 00:00:00 2001
From: Kamil Rytarowski <n54@gmx.com>
Date: Sun, 5 Apr 2020 00:54:55 +0200
Subject: [PATCH] [compiler-rt] Fix build on NetBSD 9.99.52+

Add a fallback definition of the netsmb device driver that
was removed.

(kamil)

stop xfer timeouts during recovery, all xfers will be requeued anyway

this avoids race with the timeout routine when processing the xfers
for requeueing

should fix PR kern/54790 by Izumi Tsutsui

(jdolecek)

one more netsmb reference.

(mlelstv)

fix deadlock in wdcwait() when xfer timeout happens while the atabus
thread sleeps in wdcwait() - check current lwp rather than relying
on global ATACH_TH_RUN channel flag

should fix the hang part of the problem reported in
http://mail-index.netbsd.org/netbsd-users/2020/03/12/msg024249.html

thanks to Paul Ripke for providing extensive debugging info

(jdolecek)

Move the note about sys_errlist[] users to COMPATIBILITY, and strengthen it.

(dholland)

Add missing ERRORS section.

(dholland)

lib/libusbhid: Fix possible left shift changes signedness bit.

This bug was reported by UBSan runs.

lib/libusbhid/parse.c:246:20

Can result in left shift changes signedness bit as a side effect positive number
can go negative, cast it to unsigned for the operation and silence the issue.

Reviewed by: kamil@

(fox)

lib/libusbhid: Fix possible left shift changes signedness bit.

This bug was reported by UBSan runs.

lib/libusbhid/usage.c:247:27
lib/libusbhid/usage.c:244:28
lib/libusbhid/usage.c:235:13

Can result in left shift changes signedness bit as a side effect positive number
can go negative, cast it to unsigned for the operation and silence the issue.

Reviewed by: kamil@

(fox)

lib/libusbhid: Fix possible left shift changes signedness bit.

This bug was reported by UBSan runs.

lib/libusbhid/data.c:58:25
lib/libusbhid/data.c:91:7
lib/libusbhid/data.c:92:7

Can result in left shift changes signedness bit as a side effect positive number
can go negative, cast it to unsigned for the operation and silence the issue.

Reviewed by: kamil@

(fox)

Minor clarification. And a typo. Bump date.

(dholland)

Mark uvm_map_entry_cache with PR_LARGECACHE.

(ad)

Add new example kernel module: current_time

Submitted by Apurva Nandan.

(kamil)

silence lint.

(christos)

Always include the sampling member so that the struct size is consistent.

(christos)

vrelel(): clear VV_MAPPED with the vnode still locked.

(ad)

Regen.

(ad)

Merge the remaining changes from the ad-namecache branch, affecting namei()
and getcwd():

- push vnode locking back as far as possible.
- do most lookups directly in the namecache, avoiding vnode locks & refs.
- don't block new refs to vnodes across VOP_INACTIVE().
- get shared locks for VOP_LOOKUP() if the file system supports it.
- correct lock types for VOP_ACCESS() / VOP_GETATTR() in a few places.

Possible future enhancements:

- make the lookups lockless.
- support dotdot lookups by being lockless and inferring absence of chroot.
- maybe make it work for layered file systems.
- avoid vnode references at the root & cwd.

(ad)

Switch x86 to the new binutils.

(christos)

Regen everyone!

(christos)

Man page for _lwp_gettid(2).

(Not yet installed, pending adding the libc stub.)

(thorpej)

Regen for _lwp_gettid(2).

(thorpej)

preempt_needed(), preempt_point(): simplify the definition of these and
key on ci_want_resched in the interests of interactive response.

(ad)

NetBSD 9.99.53 -- struct lwp changed.

(thorpej)

Add support for lazily generating a "global thread ID" for a LWP.  This
identifier uniquely identifies an LWP across the entire system, and will
be used in future improvements in user-space synchronization primitives.

(Test disabled and libc stub not included intentionally so as to avoid
multiple libc version bumps.)

(thorpej)

- sched_idle(): if a migration took place, rather than going idle briefly
  to avoid stealing back the LWP, remember the target CPU and avoid it.

- sched_preempted(): only choose a totally idle CPU; in the interests of
  overall throughput it's better to make use of SMT / slow CPUs.

(ad)

Infrastructure for putting kernel+modules in /netbsd/kernel and
/netbsd/modules respectively instead of /netbsd and
/stand/<arch>/<version>/modules.  This is only supported for x86,
and is turned off by default. To try it, add KERNEL_DIR=yes in your
/mk.conf and install a system from that build.

(christos)

Multilink fragment protocol type.

(is)

fix netsmb, libgnuctf

(christos)

Avoid copying zero-sized objects (from the NULL address)

Detected with UBSan

(kamil)

remove also files.netsmb from here

(jdolecek)

smb is no more.

(christos)

remove smbfs

(christos)

Adjust for field rename.

(christos)

more smb removal.

(christos)

Multilink PPP: sanity check of option values, storage of remote MRRU.

(is)

We don't have smbfs anymore.

(christos)

gnuctf is only binutils=234

(christos)

Make makelists() a little bit faster.

(ad)

smbfs is no more

(jdolecek)

const for psignal

(christos)

note SMBFS and nsmb removal

(jdolecek)

mark nsmb major obsolete

(jdolecek)

fix libgnuctf

(christos)

remove nsmb(4)

(jdolecek)

fix sets for new binutils

(christos)