Try to use the existing HW address programmed into the device if possible.

(jmcneill)

Avoid "hole in RX ring" condition by ensuring a new mbuf is available
before using an existing one.

(jmcneill)

Not distrib note changes

(martin)

Tickets #1526 and #1527

(martin)

Pull up following revision(s) (requested by mlelstv in ticket #1527):

sys/dev/scsipi/cd.c: revision 1.343
sys/kern/subr_disk.c: revision 1.130

Avoid division by zero if label isn't valid.
Allow open of RAWPART even when no medium is loaded.
Keep errors silent if no medium is loaded.
Fixes PR kern/55104

(martin)

Pull up following revision(s) (requested by mrg in ticket #1526):

include/stdio.h: revision 1.98

move P_tmpdir from "/var/tmp/" to "/tmp/".

the main effect of this is to make GCC and other libiberty using
tools use /tmp instead of /var/tmp for compiler temp files,
which can be a bottleneck on larger systems.

a survey of other platforms shows only OSX also uses /var/tmp,
everyone else has switched to /tmp long ago.

cons:  some smaller systems may have a smaller /tmp than /var/tmp,
and this may cause builds to fail with out of space earlier.
point the build to /var/tmp using TMPDIR in this case.

one can argue that setting TMPDIR would work around this, but we
want to have the effect for all users without having special setup.

(martin)

Prepare for 8.2 release.

(martin)

Note Rawrite32/distrib note changes

(martin)

Pull up following revision(s):

distrib/notes/ews4800mips/xfer: revision 1.3
distrib/notes/bebox/prep: revision 1.3
distrib/notes/newsmips/xfer: revision 1.3
distrib/notes/i386/xfer: revision 1.21
distrib/notes/common/main: revision 1.547
distrib/notes/news68k/xfer: revision 1.4

Replace references to i386 install media for Rawrite32 with
links to the homepage.

(martin)

Remove outdated Rawrite32 copy

(martin)

Summary: Introduce a keyboard event ring buffer for USB console keyboards

For the console keyboard, events must be handled separately from the
interrupt handler, because of the possibility that the user chooses to
drop into DDB by use of the designated key combination.  This delayed
handling is tied to the HZ clock through the callout(9) mechanism,
meaning that another keyboard event arriving within 10ms of the
previous one might overwrite its predecessor before it's handled.
This causes dropped events for fast typists - and the situation gets
worse with modern, programmable, keyboards.

This patch introduces a ring buffer in the ukbd_softc data structure,
so that multiple pending events may be safely stored.

(tih)

KNF

(skrll)

store the request buffer in the vxfer instead of the packet, clearer

(maxv)

PTE_SYNC before TLB flush and no need to PTE_SYNC after an unmap.

(skrll)

Support PMAP_FAULTINFO on arm

(skrll)

Provide a curpcb

(skrll)

Provide a pt_entry_t and PRIxPTE

(skrll)

Fix a c&p error from 1.3 that should fixes mrg's rk3399 slow vs fast CPU
problem.

My XU4 certainly reports the right relative speeds for the 5 (yes, 5) CPUs
that are avalable.

(skrll)

Fix skipped daily in Eastern Europe

- Move /etc/daily from 03:15 -> 04:15
- Move /etc/weekly from 04:30 -> 05:30 (to keep the spacing)

This should prevent problems during daylight savings changeover (which
is usually between 01:00 -> 04:00, depending upon the region) where the
cronjobs may be executed twice, or never at all...

This expands on the previous fix in revision 1.13 from 1997.

(kim)

Allow path components that start with ~

(jmcneill)

new acpica

(christos)

merge conflicts

(christos)

Don't set DVF_DETACH_SHUTDOWN.  The MII layer wants to manage the lifecycle
of the PHY devices, and if a NIC driver chooses not to detach its PHYs
at shutdown, that's the driver's business.

PR kern/55121.

(thorpej)

autofs: Make autofs(5) timeout messages include affected process name and PID

Sync warning message with FreeBSD and DragonFlyBSD.
taken-from: FreeBSD

(tkusumi)

Deal with both .db and .cdb extensions...

(christos)

PR/55118: Andreas Gustafsson: Oops, set the right flag variable.

(christos)

Try to fix previous: include etc.evbarm/Makefile.inc to get proper sub-arch
settings and key the subdirs off KERENEL_SETS instead of ALL_KERNELS
(which is not always defined)

(martin)

tic: free the old buffer after promoting the record type

(roy)

terminfo: Add a comment for prior change.

(roy)

terminfo: use , as a record version delimiter rather than @

@ is allowed in a terminfo description, but a comma is not.

(roy)

Whitespace

(roy)

tic: use EXIT_FAILURE rather than magic numbers

(roy)

Whitespace

(roy)

tic: promote record type when merging newer into older

(roy)

- comb down rcvfilt() filter adjustment  logic.
- some mii_statchg() can be written in simpler form.

(nisimura)

Reduce default AUDIO_BLK_MS from 40msec to 10msec on all platform except m68k
(m68k uses 40msec default as before).  And remove the option from GENERIC.
- It's not good idea to set such parameter in individual GENERICs.
- 4msec is (probably no problem for most modern real hardware but)
  too aggressive to be default.
- 10msec is too severe for antique machines but it's hard to draw a line.

(isaki)

Remove trailing whitespace.

(wiz)

Add description about channel limitation introduced in audio.c 1.43.
PR kern/54973.

(isaki)

Revert 1.96, 1.97 and 1.100.  (no response from nia@)
These changes are not correct nor notable unresolvable problem.
If you find any problems, please send a report before changing manpage.

(isaki)

support for older compilers

(christos)

Comment out some of the CTASSERTS for lint until I fix lint.

(christos)

Belatedly add NanoPi R1.

(bad)

PR kern/55114: Install fails with "cpu_switchto: switching above IPL_SCHED (8)"

cpu_switch(): Remove stuff dealing with interrupt levels.  From memory it
was something to do with TLB shootdown interrupts but they have long been
outside the SPL framework.

(ad)

Add higher state of consciousness
https://twitter.com/joshwink1/status/1243360644042113024

(sevan)

use standard deferred if_start framework instead of custom variant

(jdolecek)

- change to use rcvfilt_locked() name to clarify the intent.
- stylise receive filter manipulation logic.
- use ETHER_F_ALLMULTI flag.

(nisimura)

Use the new functions and add more error checking.

(christos)

Introduce a bunch of inline functions and utilities to avoid code duplication.

(christos)

replace the conditional m_pullup() on start of ether_sw_offload_tx()
with a KASSERT(), to make it clear no mbuf manipulation is ever done here

(jdolecek)

replace the conditional m_pullup() on start of bridge_output() with
a KASSERT(), to make it clear no mbuf manipulation is ever done here

the condition should never trigger, this always runs after ether_output()
M_PREPEND()s ether_header

(jdolecek)

fix in4_cksum() panic "in4_cksum: mbuf 14 too short for IP header 20"
triggered by bridge_output() when passing packet originally for
interface supporting hw csum offload to destination interface
not supporting it

problem happens because bridge_output() is called after ether_output()
M_PREPEND() the ether_header into the mbuf chain, if there is not
enough space on the first mbuf of the chain, it ends up prepending
a new short mbuf with just ether_header

triggered by running UDP (IPv4) 'netio -u' benchmark with packet size 2 KB

XXX seems in6_undefer_cksum() should have similar fix, however I was
XXX not able to trigger the problem there

(jdolecek)

Get back to terminfo.cdb.

(rin)

As described in tech-userlevel:
- Modify the writing code to only write entries in the new
  format for the terminal descriptions that require it.
- Store new format entries as <name>@v3
- Store old format entries with clamped values as <name> for
  backwards compatibility
- Lookup first <name>@v3 and then <name> if that is not found.
- Don't create terminfo2 anymore; old programs keep working with
  clamped entries, and new programs be able to use the wide
  fields with using the original db file.

(christos)

extend mii_statchg() to use ether_mediachange(). care about hash filter selection.

(nisimura)

extend mii_statchg() to use ether_mediachange(). care about hash filter selection.

(nisimura)

Allow open of RAWPART even when no medium is loaded.
Keep errors silent if no medium is loaded.

Fixes PR kern/55104

(mlelstv)

Avoid division by zero if label isn't valid.

(mlelstv)

synaptics: Revert changes that broke gestures in cases where they worked.

It seems I have to figure out another way to make scrolling work for me.

(nia)

Add special handling for model 0x0f stepping >=2 or mode 0x0e to get Tjmax.

(msaitoh)

try to be better register descriptions

(nisimura)

Add missing line-break

(sborrill)

Fix ticket number

(sborrill)

Ticket #1524

(sborrill)

Pull up the following revisions(s) (requested by christos in ticket #1524):
etc/named.conf: revision 1.8

Stop using obsolete dnssec-lookaside (implies use of dlv.isc.org, which
no longer serves any useful purpose).

(sborrill)

Ticket #806

(sborrill)

Pull up the following revisions(s) (requested by christos in ticket #806):
etc/named.conf: revision 1.8

Stop using obsolete dnssec-lookaside (implies use of dlv.isc.org, which
no longer serves any useful purpose).

(sborrill)

more register mashup work

(nisimura)

register definition mashed. from UEFI and CycloneV PDF

(nisimura)

Let xunlong,orangepi-lite2 share bwfm config with raspberrypi,3-model-b-plus;
it seem to work fine and there's no better alternative at the moment.

(thorpej)

Add OrangePi Lite and Lite2.

(thorpej)

Add a more complete / correct device tree for the OrangePi Lite 2.

(thorpej)

Attach sunxirtc in pass4; it's required by mmcpwrseq on some boards.

(thorpej)

Add the clock providers for the sun6i_a31, sun8i_a23, sun8i_r40, sun8i_v3,
sun8i_h3, sun50i_h5 (same as H3), and sun50i_h6 RTC blocks.

(thorpej)

No need for warnx() if !LOCKDEBUG

(ad)

- Stats: add a cast to make sure 32-bit wraparound works properly.
- cache_purge_children(): just use RB_TREE_MIN().
- Change the name of the pool to "namecache".
- Update comments.

(ad)

Comments

(ad)

TAILQ -> LIST

(ad)

sleepq_t is now a LIST.  Forgot to commit earlier.

(ad)

Fix crash observed with procfs on current-users by David Hopper.  LWP refcnt
and p_zomblwp both must reach the needed state, and LSZOMB be set, under a
single hold of p_lock.

(ad)

sigpost(): check for LSZOMB, not l_refcnt == 0.

(ad)

SOBJ_SLEEPQ_FIFO is gone

(ad)

softint_overlay() (slow case) gains ~nothing but creates potential headaches.
In the interests of simplicity remove it and always use the kthreads.

(ad)

NetBSD 9.99.52 - sleepq_t & kcondvar_t shrunk

(ad)

Change sleepq_t from a TAILQ to a LIST and remove SOBJ_SLEEPQ_FIFO.  Only
select/poll used the FIFO method and that was for collisions which rarely
occur.  Shrinks sleep_t and condvar_t.

(ad)

Leave the idle LWPs in state LSIDL even when running, so they don't mess up
output from ps/top/etc.  Correctness isn't at stake, LWPs in other states
are temporarily on the CPU at times too (e.g.  LSZOMB, LSSLEEP).

(ad)

sched_curcpu_runnable_p(): cosmetic change.

(ad)

Don't overwrite ci_is_slow if it has explicitly been set.  Changed for
correctness, should have no impact.

(ad)

in xennet_tx_complete() replace the grant status condition with just KASSERT()

it's really a possible programming error rather than something actually
happening in wild - since rev. 1.57 in 2012 this condition leaked
locked mutex, and nobody noticed

(jdolecek)

Stop using IFF_OACTIVE, just keep processing until we fill the Tx ring

xennet_tx_complete() is now called only in xennet_softstart() just
before processing Tx queue; xennet_handler() just schedules the
softint for xennet_softstart() iff the Tx ring list is completely empty

no performance changes observed

XXX the detach and suspend code expects xennet_handler() to collect
XXX the Tx responses, the tsleep() loop should still work since the tsleep()
XXX would allow the softint to trigger; untested though

(jdolecek)

Add HPD

From Rocky Hotas, thanks!

(leot)

Add TLDW

From Rocky Hotas, thanks!

(leot)

Add tests for poll(POLLIN) before read().
This affects only standalone test, not atf.

(isaki)

When using SYSCTL_SETUP there is no guarantee of the order in which
the set-up functions will be called, so it is perfectly acceptable
for a compat code's routine to be called ahead of the code in other
parts of the kernel.

So make sure that the 2nd level sysctl node ``vfs.generic'' exists
before trying to add the 3rd level entries.

XXX Rather than creating the 2nd level node in two places, we could
XXX add the shared ``vfs.generic'' node to sysctl_init_base.c but
XXX this is left for another day.

(pgoyette)

Sync with sys/dev/audio/audio.c rev1.65.
> Fix to start recording immediately when open() with READ mode is called.
This affects only standalone test, not atf.

(isaki)

Fix to start recording immediately when open() with READ mode is
called (unless pause).
- Opening /dev/audio always initializes pause with false.  Therefore
  it always starts recording.
- Opening /dev/sound inherites pause from the sticky parameter.
  Therefore whether /dev/sound starts recording or not depends on the
  sticky pause parameter.
This fixes two problems:
- Opening /dev/audio didn't start recording after merging isaki-audio2.
- Opening /dev/sound didn't start recording regardless of the sticky
  pause, probably since long time ago (at least netbsd-7).

(isaki)

update register definition. from UEFI source code.

(nisimura)

correct global reset logic. still remains MII speed trouble.

(nisimura)

Only build sshramdisk when building RPI_INSTALLx

(skrll)

Bump IMAGESIZE

(skrll)

Increase MEMORY_DISK_ROOT_SIZE

(skrll)

Fix the build, use %jd and (intmax_t) cast for big numbers, off_t
and ptrdiff_t aren't always the same size.

(kre)

adjust minor comments

(nisimura)

Add "snps,dwc3" to the list of compatible strings we match against.
The code already handles "snps,dwc3" not being a subordinate of
an SoC-specific node, but the string was missing from the match
routine.

Necessitated by the sun50i-h6 device tree update on Jan 3 2020, which
elimiated the "allwinner,sun50i-h6-dwc3" node and placed all of the
clocks, resets, etc. directly under the "snps,dwc3" node.

(thorpej)

try to DRT about config messages, redo ...

(nisimura)

try to DRT about config messages

(nisimura)

DRT about config messages

(nisimura)

better boot messaging, redo

(nisimura)

better boot messaging

(nisimura)

Remove trailing whitespace.

(wiz)

new paragraph

(christos)

more info

(christos)

Revert putting a $ as the final character for blowfish. It is not required by
MCF and we want to be compatible.

(christos)

use dma32 for descriptor store. likely still missing freq designation(s)

(nisimura)

improve error messages.

(christos)

describe DeveloperBox pin assignment correctly

(nisimura)

Fix typos. Use more markup. New sentence, new line.

(wiz)

Delete the BUGS paragraph about the "missing" const qualifier for the
result type of strerror() (and strerror_l()).  While that once should
really have been present, when strerror() was invented, there was no
"const" qualifier in C to apply, and now the way the code is writtem
(really needs to be because of NLS support) the const is no longer
really appropriate.

Applications still shouldn't attempt to modify the result however.

(kre)

Document strerror_l()

While here also document (but comment it out since it isn't
available - yet) strerror_lr().  To include that, simply
uncomment the relevant lines, and (twice I think) s/returns/return/
on lines just after currently commented out lines (that is, it
currently says, "A returns" after the comments are returned, we
need it to be "A and B return" - the "and B" appears when the comment
markers are removed, removing the 's' from returns must be done manually.

In addition to adding strerror_l() some additional enhancements were
made to the general strerror() doc.

(kre)

try to decode _DSD ACPI resource

(nisimura)

- bump blowfish size, explain version
- add passwd xref

(christos)

Add missing trailing $ for blowfish

(christos)

Relax fdatasync restriction that fd be writable

The restriction that a fd passed to fdatasync(2) must be writable was
added in 2003 in order to comply with POSIX.  Since then, POSIX has
removed that requirement, and POSIX-valid programs have been therefore
encountering errors on NetBSD.

Patch by Paul Ripke after discussion on netbsd-users.  Issue
discovered with pkgsrc/databases/mongodb3 as used by pkgsrc/net/unifi.

(gdt)

add pw_gensalt

(christos)

PR/55095: David A. Holland: pw_gensalt(3) undocumented

(christos)

remove 'file-system SMBFS' and 'pseudo-device nsmb' from all kernel configs
to prepare for their eventual removal

(jdolecek)

Arrange that strerror(-1) prints "Unknown error: -1" and not the
unsigned equivalent of -1.

While here, guarantee, even when !NLS, that nothing here (not even
snprintf deciding to complain about EILSEQ or something) can ever
alter errno (ie: always save and restore it, not only in the NLS case).
The functions here must never alter errno, whatever happens.

(kre)

Protect against malloc failure corrupting errno, which is not
permitted of these functions.

(kre)

Use exact match to search testname.
This didn't affect test results.

(isaki)

Simplify #ifdefs

(skrll)

Trailing whitespace

(skrll)

Add a file type entry for CLM files.

(thorpej)

- Refactor the firmware loading code into a set of common routines in the
  core driver that understand how to find model-specific files before the
  generic files.  This greatly simplifies the firmware loading procedure
  in each of the bus front ends.  Adding additional firmware file types
  (such as CLM files) much simpler.
- Make firmware image selection table-driven rather than a set of switch()
  and if() statements.
- Add several additional firmware image entries.

(thorpej)

make sure to use mac_write() for GMACxxx registers. try to decode xMII link status report

(nisimura)

ohci@fdt, ehci@fdt are machine independent; move from sunxi to sys/dev/fdt

(tnn)

Update a comment.

(ad)

add entry for NetMos NM9900 Quad UART.

The NM9900 seems to have variations such as 4UART and 2UART+1Printer,
but I tested on 4UART board.

(ryo)

g/c UGEN_BULK_RA_WB which is not a supported option

(sevan)

Remove the argument from USB_{ATTACH,DETACH}, for consistency.

(maxv)

If we're going to loop, pausing and then retrying malloc() after it
has failed, in the hope that some other thread has free'd some memory,
but we want to bound the number of attempts, it helps if we actually
count them - otherwise we never get nearer to the limit.

In practice, malloc() for a reasonable application on a modern system
almost never fails, so the code containing this bug has probably never
been, and never will be, executed, but just in case, someday.

For this, it isn't clear if the intent was to have 10 retries (ie: 11
attempts) or 10 tries, but as the code said "retries > 10", I am
assuming the former (not that it matters, if the malloc() has failed
10 times in a row, with 10 second pauses between, the chances of an
11th succeeding aren't great).

(kre)

Revert previous

It will be addressed in a better way.

Requested by <kre>

(kamil)

add more register description

(nisimura)

reset the csum_flags in bridge_brodcast() also for bmcast path

for destination interfaces with real hardware offloading this fixes
multicast packet corruption; for xvif(4) this fix stops treating them
as having no csum

may fix PR kern/42386

(jdolecek)

add more registers supposed to exist

(nisimura)

try to show 32 GPIO line usage

(nisimura)

fix FDT 2nd addr to map EEPROM store. mention ACPI _DSD contains FDT equiv parameters.

(nisimura)

try to be consistent for the number of address match filter. tentative util the reality.

(nisimura)

small knit about multicast hash table handling. use sc_100mii as bool

(nisimura)

fix vmx(4) cannot link up at boot time.  reviewed by msaitoh@n.o, thanks.

vmx(4) could call if_link_state_change(ifp, LINK_STATE_UP) from vmxnet3_init()
before ifp->if_link_cansched was set, because dp->dom_if_up() (in6_if_up() for
INET6) could call ifp->if_init().  And then, workqueue_enqueue() was not called
at that time.
As the result, the last LQ_ITEM was stuck LINK_STATE_UP, so
if_link_state_change_work_schedule() was never called until
if_link_state_change(ifp, LINK_STATE_DOWN) was called.

To fix this issue, vmx(4) avoid calling if_link_state_change() before
ifp->if_link_cansched is set.

(knakahara)

Fix cprng_fast64 return type to match code

(skrll)

note that MKLLVMRT affects more than amdgpu.

(mrg)

Fix type confusion. Found by kASan when doing a normal attach+detach over
vHCI.

(maxv)

Use a vhci_request_t, will be required for future changes.

(maxv)

Fix unused area size found by pgoyette@.

(msaitoh)

nuke <sys/queue.h> and reorder #include

(nisimura)

Calculate DDR3's tRAS correctly.

(msaitoh)

- Define some new parameters of DDR3 SPD ROM.
- Use fine timebase parameters for time calculation on DDR3. This change
  makes PC3-XXXX vaule more correctly on newer DDR3.

(msaitoh)

KNF. No functional change.

(msaitoh)

correct inverted logic to determine RMII/MII

(nisimura)

add some more findings about SC2A11 GbE. do some code steamline practice.

(nisimura)

Avoid buffer overflow

Detected with ASan + RUMPKERNEL.

(kamil)

Fix off-by-one

Before accessing array member, check whether it is not out of valid range.

Detected with ASan + RUMPKERNEL.

(kamil)

Regen.

(ad)

Update comments

(ad)

The image size increased.

(prlw1)

Fix format string usage.

(joerg)

cache_remove(): remove from the vnode list first, so cache_revlookup()
doesn't try to re-activate an entry no longer on the LRU list.

(ad)

Oops. Fix merge error.

(ad)

Cosmetic tweak

(ad)

Correct an assertion.

(ad)

ncs_collisions is gone.

(ad)

Regen.

(ad)

- Deal with (rare) hash collisions by using memcmp() to partition further.
- Adjust some comments.

(ad)

Fix a race between cache_enter() and cache_lookup().  Make sure the cache
entry is fully constructed before it appears on a vnode list.

(ad)

Pacify a syzbot complaint about bit shifting.

Reported-by: syzbot+dd5df915b2cd2f3f3da5@syzkaller.appspotmail.com

(ad)

Reduce #ifdefs

(skrll)

Fix synopsis, use more markup.

(wiz)

Add missing man page for curses mouse functions.

(roy)

Don't attempt to detach an evcnt before attaching it.  If its not
already attached, we will panic.

It turns out that this check wasn't really needed anyway, it was
simply paranoia on my part.

Thanks to hannken@ for bringing this to my attention.

(pgoyette)

Fix the currency symbol for fi_FI.ISO8859-1

In the Finnish language, the recommended symbol for euro is the euro sign
where it is available, and the lowercase letter e otherwise.

The use of the ISO currency code EUR is not an abbreviation of the word
euro in the Finnish language, just like FIM is not an abbreviation of
the word markka.

Reference:
    https://www.kielikello.fi/-/euro-
    Euro
    Kielikello 3/1998
    Kotimaisten kielten keskus
    Institute for the Languages of Finland
    [Last retrieved 2020-03-23]

(kim)