audio.4: Mention the blk_ms sysctl

(nia)

Lower blk_ms on more ports that are even slightly likely to play video

(nia)

Sort sections.

(wiz)

Ticket #1521

(sborrill)

Pull up the following revisions(s) (requested by jdolecek in ticket #1521):
sys/arch/xen/xen/xennet_checksum.c: revision 1.4

Make the packet length check less strict, allow the physical packet longer
than IP payload. This fixes problem where checksum was not recomputed for
short packets coming from Windows domU on same physical host as it seems
Windows does some padding.

(sborrill)

Ticket #1726

(sborrill)

Pull up the following revisions(s) (requested by jdolecek in ticket #1726):
sys/arch/xen/xen/xennet_checksum.c: revision 1.4

Make the packet length check less strict, allow the physical packet longer
than IP payload. This fixes problem where checksum was not recomputed for
short packets coming from Windows domU on same physical host as it seems
Windows does some padding.

(sborrill)

synaptics: Increase default scale_z for more precise scrolling

(nia)

Ticket #781

(sborrill)

Pull up the following revisions(s) (requested by jdolecek in ticket #781):
sys/arch/xen/xen/xennet_checksum.c: revision 1.4

Make the packet length check less strict, allow the physical packet longer
than IP payload. This fixes problem where checksum was not recomputed for
short packets coming from Windows domU on same physical host as it seems
Windows does some padding.

(sborrill)

Catch up with major bump in terminfo. Unbreak build.

(rin)

Fix build; add -f flag to "rm miniroot.fs" in case of its absence.

(rin)

pms.4: Typo.

(nia)

audio.4: O_NONBLOCK isn't the actual problem

(nia)

audio.4: Remove reference to using O_NONBLOCK

It doesn't work properly, but this turns out to not be a problem in most
code (code where it is generally uses threads)...

Don't provide misleading information about using it, or programmers
might start wondering why their code doesn't work.

Noted by Yorick Hardy on current-users

(nia)

redo the commit correctly

(nisimura)

include files.sociox for generic64

(nisimura)

add a missing line

(nisimura)

Include <sys/rwlock.h> for krwlock_t required by uvm_pagewait().

(rin)

PR misc/54886: bump threshold for automatic/default creation of a tmpfs /tmp
up slightly (to 384 MB ram). This will make sure the default install has
a > 64 MB /tmp available (number pulled out of thin air, 64 MB is the minimum
required by the ZFS tests).

(martin)

Do the "have transmit resources available" check before checking that
there is a packet in the send queue.

(thorpej)

Unfortunately all the clockctl ioctls contain pointers to structs instead
of the structs themselves, so they need special handling. Undo previous
and do the permissions checks explicitly. It would be better to fix the
clockctl ioctls to contain the structs themselves...

(christos)

Don't open-code ioctls, pass back to the 64 bit ones (found by maxv@),

(christos)

Fix compiler-rt build with MKIPFILTER=no

Cherry-pick upstream patch:

From 7f6b25ad1bb3f8057655a9bad2a3b69621f4a543 Mon Sep 17 00:00:00 2001
From: Kamil Rytarowski <n54@gmx.com>
Date: Mon, 16 Mar 2020 00:39:02 +0100
Subject: [PATCH] [compiler-rt] [netbsd] Add support for the MKIPFILTER=no
distribution

Add fallback definition for the IPFilter ioctl commands.

https://github.com/llvm/llvm-project/commit/7f6b25ad1bb3f8057655a9bad2a3b69621f4a543.patch

PR bin/55070 by Roy Marples

(kamil)

Add and use a new function, mowner_init_owner(), that initializes an
MBUFTRACE mowner structure (so that providers of it don't have to
grovel the internals).

(thorpej)

NetBSD 9.99.50 -- ifmedia / mii locking changes.

(thorpej)

Define and implement a locking protocol for the ifmedia / mii layers:
- MP-safe drivers provide a mutex to ifmedia that is used to serialize
  access to media-related structures / hardware regsiters.  Converted
  drivers use the new ifmedia_init_with_lock() function for this.  The
  new name is provided to ease the transition.
- Un-converted drivers continue to call ifmedia_init(), which will supply
  a compatibility lock to be used instead.  Several media-related entry
  points must be aware of this compatibility lock, and are able to acquire
  it recursively a limited number of times, if needed.  This is a SPIN
  mutex with priority IPL_NET.
- This same lock is used to serialize access to PHY registers and other
  MII-related data structures.

The PHY drivers are modified to acquire and release the lock, as needed,
and assert the lock is held as a diagnostic aid.

The "usbnet" framework has had an overhaul of its internal locking
protocols to fit in with the media / mii changes, and the drivers adapted.

USB wifi drivers have been changed to provide their own adaptive mutex
to the ifmedia later via a new ieee80211_media_init_with_lock() function.
This is required because the USB drivers need an adaptive mutex.

Besised "usbnet", a few other drivers are converted: vmx, wm, ixgbe / ixv.

mcx also now calls ifmedia_init_with_lock() because it needs to also use
an adaptive mutex.  The mcx driver still needs to be fully converted to
NET_MPSAFE.

(thorpej)

synaptics: Check for single-finger event jitter in all cases.

I now have smooth two-finger scrolling.

(nia)

Don't bother with IFF_OACTIVE.

(thorpej)

Don't bother with IFF_OACTIVE.

(thorpej)

tests: ndp_rtm: Check correct source address in RTA_AUTHOR

Fixes PR kern/55074.

(roy)

uvm_physseg: cluster fields used during RB tree lookup for PHYS_TO_VM_PAGE().

(ad)

Skip tests when we know there is not enough space available

(martin)

Fix a comment.

(ad)

- pmap_enter(): Remove cosmetic differences between the EPT & native cases.
  Remove old code to free PVEs that should not be there that caused panics
  (merge error moving between source trees on my part).

- pmap_destroy(): pmap_remove_all() doesn't work for EPT yet, so need to catch
  up on deferred PTP frees manually in the EPT case.

- pp_embedded: Remove it.  It's one more variable to go wrong and another
  store to be made.  Just check for non-zero PTP pointer & non-zero VA
  instead.

(ad)

pmap_enter(): look directly in the tree for old PVE when installing an
unmanaged mapping, because there is no existing pmap_page to check in
the shortcut path (it traps).

pv_pte_next(): don't assert pp_embedded because it could have been removed
(during pmap_pp_remove()).

(ad)

Fix a comment

(skrll)

Move some comments

(skrll)

KNF the KNF

(skrll)

KNF

(skrll)

Use SYSCTL_SETUP mechanism to establish our sysctl tree, rather than
using explicit calls to netbsd32_sysctl_{init,fini}()

(pgoyette)

Remove useless #include

(skrll)

Use howmany.  NFCI.

(skrll)

Remove duplicate ehci @ fdt

(skrll)

Now that we have split the quota code out of the compat_netbsd32_50
module, it no longer depends on compat_50_quota.

Should address PR kern/55073

(pgoyette)

Hardcode a check for ZFS and too little free space in workdir to skip
ZFS tests if the atf working directory is size restricted (ZFS has a
hardcoded minimal size of 64MB for a storage pool and 128 MB disk size).

(martin)

driver description file for Socionext products

(nisimura)

Fix build for ports using uvm/pmap; pmap_remove_all() returns a boolean.

(rin)

Skip tests where the old or the new image size seem to not fit into
the working directory.

(martin)

Fix build with UVMHIST.

(rin)

Skip this test if the atf working directory has not enough space
for the created image.

(martin)

Use roundup.  NFCI.

(skrll)

werase, wclrtobot, wclrtoeol - make code even more similar.
Rename attr to battr to match bch and WINDOW::battr.
No functional change intended.

(uwe)

__NEED_ERASE - don't shadow __LDATA::attr with macro parameter name.

(uwe)

route: Enable receive error reporting on the routing socket

(roy)

synaptics: Detect multiple fingers outside the gesture period

I suspect this code doesn't make any sense if we want two-finger scrolling

(nia)

synaptics: Automatically scroll when multiple fingers are detected

This will mean we automatically get two-finger scrolling on multitouch
pads.

This works, but the scrolling is janky. Why does it eventually move the
mouse cursor (only one finger detected, but two are still there), and why
does it jump up slightly?

(nia)

Update a comment.

(ad)

synaptics: fix syntax

(nia)

synaptics: Allow the max_speed sysctls to be set

(nia)

Don't require a write lock for page enqueue/activate/deactivate.

(ad)

Re: kern/55071 (Panic shortly after running X11 due to kernel diagnostic assertion "mutex_owned(&pp->pp_lock)")

pmap_pp_remove(): get rid of a "goto" to make it clearer what's going on.

(ad)

Make uvm_pagemarkdirty() responsible for putting vnodes onto the syncer
work list.  Proposed on tech-kern@.

(ad)

Catch up with reality.

(ad)

Make page waits (WANTED vs BUSY) interlocked by pg->interlock.  Gets RW
locks out of the equation for sleep/wakeup, and allows observing+waiting
for busy pages when holding only a read lock.  Proposed on tech-kern.

(ad)

rump - page/object dirtyness tracking corrections.

(ad)

synaptics: Set up_down_emulation=3 by default.

It's a less surprising default for modern hardware.

(nia)

Unused variable.

(ad)

PR kern/55071 (Panic shortly after running X11 due to kernel diagnostic assertion "mutex_owned(&pp->pp_lock)")

- Fix a locking bug in pmap_pp_clear_attrs() and in pmap_pp_remove() do the
  TLB shootdown while still holding the target pmap's lock.

Also:

- Finish PV list locking for x86 & update comments around same.

- Keep track of the min/max index of PTEs inserted into each PTP, and use
  that to clip ranges of VAs passed to pmap_remove_ptes().

- Based on the above, implement a pmap_remove_all() for x86 that clears out
  the pmap in a single pass.  Makes exit() / fork() much cheaper.

(ad)

- Hide the details of SPCF_SHOULDYIELD and related behind a couple of small
  functions: preempt_point() and preempt_needed().

- preempt(): if the LWP has exceeded its timeslice in kernel, strip it of
  any priority boost gained earlier from blocking.

(ad)

uvm_map_lookup_entry(): save the hint even on failure, since code elsewhere
relies on it pointing to the previous entry.

(ad)

Remove unused RW lock defs.

(ad)

provide both miniroot.fs (for people wanting to install NetBSD without
more 3rdparty software than what's provided on our ISO) and .fs.tz (for
people downloading over slow links).

(is)

Trailing whitespace

(skrll)

OR into bp->b_cflags; don't overwrite.

(ad)

OR into bp->b_cflags; don't overwrite.

(ad)

OR into bp->b_cflags; don't overwrite.

(ad)

OR into bp->b_cflags; don't overwrite.

(ad)

- uvmspace_exec(), uvmspace_free(): if pmap_remove_all() returns true the
  pmap is emptied.  Pass UVM_FLAG_VAONLY when clearing out the map and avoid
  needless extra work to tear down each mapping individually.

- uvm_map_lookup_entry(): remove the code to do a linear scan of map entries
  for small maps, in preference to using the RB tree.  It's questionable,
  and I think the code is almost never triggered because the average number
  of map entries has probably exceeded the hard-coded threshold for quite
  some time.

- vm_map_entry: get it aligned on a cacheline boundary, and cluster fields
  used during rbtree lookup at the beginning.

(ad)

pmap_remove_all(): Return a boolean value to indicate the behaviour.  If
true, all mappings have been removed, the pmap is totally cleared out, and
UVM can then avoid doing the work to call pmap_remove() for each map entry.
If false, either nothing has been done, or some helpful arch-specific voodoo
has taken place.

(ad)

uvm_pdpolicy: Require a write lock on the object only for dequeue.
No sense in requiring that for enqueue/activate/deactivate.

(ad)

Put ACPI idle under ACPICPU_ENABLE_C3 until the wrinkles are ironed out.
This seems well written and basically all good, but currently doesn't enter
a low power state, and imposes a big performance penalty.  Proposed on
port-i386 & port-amd64.

(ad)

tmpfs_inactive(): do like other file systems and truncate the file if it
has been deleted.  Otherwise VFS will try to write cached data "back to
disc", which in the case of a UAO means needless page deactivations and
the resulting TLB shootdowns.

(ad)

tmpfs_reg_resize(): do nothing if newsize == oldsize.

(ad)

sparc cpu_intr_p(): try to work around l_cpu not being set early on by
using curcpu().

(ad)

synaptics: Make up_down_emulation useful for single-button clickpads

On devices such as the Thinkpad X250, the clickpad can be pressed
to generate mouse button events 1 and 2. There are also additional
physical buttons which the pms(4) driver recognizes as "up/down" buttons
(mouse buttons 3 and 4). Allow these to be remapped to buttons 1 and 2
and used like normal touchpad buttons with the following sysctl:

# sysctl -w hw.synaptics.up_down_emulation=3

While here, adjust the existing "middle button emulation"
(hw.synaptics.up_down_emulation=1) so it works with single-button
clickpads.

XXX: 3 may be a more useful default than the current default,
depending on hardware availability of touchpads with "up/down buttons".

Update the documentation accordingly.

(nia)

make the packet length check less strict, allow the physical packet longer
than IP payload

this fixes problem where checksum was not recomputed for short
packets coming from Windows DOMU on same physical host, seems Windows
does some padding; reported by sborrill@

XXX pullup-9

(jdolecek)

Sort sections. Fix date.

(wiz)

add an example of using midirecord(1) to learn about how device
specific MIDI messages are formed, allowing their use from MIDI
input files, not just the device controls.

using this method, i can now control the voice mode type on my
new synth.

(mrg)

hopefully fix builds

(skrll)

On amd64, mark the whole tree as NX. No real functional change, just to
prevent possible future surprises, and to make it a little harder to map
executable pages in ROP chains.

(maxv)

style

(maxv)

fix memory leaks

(maxv)

wrong size passed to copyout

(maxv)

fix more broken kernhist formats (now I got them all).

(christos)

revert the 0x% -> %# change for fixed width formats pointed out by uwe.

(christos)

more changes from sc.dying in PR/55068

(christos)

t_glob.c: clean up test code

In struct vfs_file, using an int as a boolean is an anachronism and has
been replaced with a single-character file type, like in ls(1).

Some other redundant test code has been removed as well since it was
either unreachable or existed only for performance reasons.

(rillig)

t_glob.c: add test cases for hidden directory and file

The existing test code was geared towards every little bit of
performance. It even duplicated the file definitions in vfs_stat in order
to avoid a few strcmp calls. This made the test code fragile. Therefore,
vfs_stat has been rewritten completely to not duplicate any information
from the vfs.

In vfs_stat, the returned st_mode is now more realistic. It had been 0
before. The file mode is only logged when it makes sense. In the ENOENT
case it is not logged anymore.

The debug logging for opendir/closedir now logs the same pointer, so that
the corresponding calls can be matched easily. Failed vfs_opendir calls
are logged as well, to get a more complete picture of which callbacks are
called.

(rillig)

t_glob.c: use distinct names for test structures

Before, the structures and functions defined by the test used the same
prefix as the code to be tested. This made it difficult to draw a line
between these parts.

(rillig)

t_glob.c: move expected globbing result directly into the test cases

This makes the tests more self-contained. The example directory tree that
is common to all the tests is still defined elsewhere, but in the same
file. Setting up the example directory structure in each test would make
the tests even more independent and read.

(rillig)

Fix printf format

(martin)

Fix a bug that atphy(4) doesn't work with Attansic L2 rev. 1.
Reported by Rocky Hotas.

- On ASUS M2N-MX SE Plus (NVIDIA MCP61 with Attansic L2 rev. 1), changing
  debug port 0x29's value makes the next PHY read fail with error.  Read any
  register to ignore this problem if the PHY is Attansic L2 revision 1.
  I don't know if this problem is from L2 rev. 1 itself or from the
  combination because I have only one machine which has L2 rev. "1".
  At least, ASUS eee pc 900 (Attansic L2 rev. "2") has no this problem.
- Add comment. AR8021 document has no description about the power saving
  control register(debug port 0x29).
- Add comment. AR8031 document says the lower 14 bits are reserved and the
  default value is 0x36d0. Shouldn't we clear those bits?
- I have no document neither L1(F1) nor L2(F2), so I don't know whether the
  debug port access is correct or not.

Tested with the following machines:
- ASUS P5B SE,        L1 rev. 5,            age(4)
- ASUS K50IJ,          L1 rev. 9,            ale(4)
- ASUS eee pc 900,    L2 rev. 2,            lii(4)
- ASUS M2N-MX SE Plus, L2 rev. 1,            nfe(4)
- Intel DP55WB,        82578(AR8021 rev. 2), wm(4)
- Dell inspiron 14z,  AR0835 rev. 9,        alc(4)

(msaitoh)

PR/55068: sc.dying: Fix printf formats:
- no %s/%p for kernel log
- 0x% -> %#
- always %j for kernel log

(christos)

Re-enable the routing perm check now that the kernel has been fixed to return
valid information.

(christos)

Use the socket credentials that are established during the socket creation
instead of the current process credentials (which can change via
set{e,}{u,g}id(2)) and by passing the fd to a different process. This makes
the routing socket behave like other file descriptors. Proposed in tech-kern.

(christos)

Enhance the DIAGNOSTICs around pmap_grow_map

(skrll)

Re-indent a function call.  NFCI.

(skrll)

Oops... remove a stray <space>

(skrll)

Fixup some comments

(skrll)

Note terminfo change

(roy)

terminfo: promote numeric parameters from short to int

POSIX mandates implementations must support upto a short but may exceed it.
When NetBSD terminfo was implemented, no terminfo description used over
a short, but because ncurses has supported ints for some time, some now do.

Infact, such a terminfo description was imported where colour pairs for
screen-256color went up to 65536 which exposed a bug in the existing
implementation where it set to zero. Because the number might mean
something more than a range, we need to be able to store it accurately.

This requires a version bump because whilst the API hasn't changed thanks
to C int promotion, the ABI has. Also the underlying database structure
has changed as well - we now store the numeric paramter inside a uint32_t
field rather than a uint16_t one.
Whilst this change can still read the old style database, the old one
cannot read the new one and thus we now maintain the database as
terminfo2.cdb, leaving the old library and database alone so old programs
still work fine.

libcurses, libfrom, libmenu and libpanel have also been bumped to
accomoate this change.

(roy)

Provide the (small) miniroot uncompressed, so that it can be used from
an AmigaOS without additional software to install NetBSD.
Closes PR port-amiga/45443

(is)

Ticket #1520

(martin)

Pull up following revision(s) (requested by knakahara in ticket #1520):

sys/netipsec/key.c: revision 1.271
sys/net/if_ipsec.c: revision 1.28
sys/net/if_ipsec.c: revision 1.29

Fix ipsecif(4) SPDADD pfkey message has garbage.  Pointed out by ohishi@IIJ.

"setkey -x" output is the following.
========== before ==========
sadb_msg{ version=2 type=14 errno=0 satype=0
  len=15 reserved=0 seq=0 pid=0
sadb_ext{ len=56 type=18 }
sadb_x_policy{ type=2 dir=1 id=9 }
{ len=40 proto=50 mode=1 level=3 reqid=16393
sockaddr{ len=0 family=0  }
sockaddr{ len=0 family=0  }
}
========== before ==========

========== after ==========
sadb_msg{ version=2 type=14 errno=0 satype=0
  len=11 reserved=0 seq=0 pid=0
sadb_ext{ len=24 type=18 }
sadb_x_policy{ type=2 dir=1 id=9 }
{ len=8 proto=50 mode=1 level=3 reqid=16393
}
========== after ==========

reduce unnecessary reqid of NAT-T ipsecif(4), suggested by ohishi@IIJ.

Fix kern/55066.  Pointed out and fixed by Chuck Zmudzinski, thanks.
ok'ed by ozaki-r@n.o

(martin)

Ticket #780

(martin)

Pull up following revision(s) (requested by knakahara in ticket #780):

sys/netipsec/key.c: revision 1.271
sys/net/if_ipsec.c: revision 1.28
sys/net/if_ipsec.c: revision 1.29

Fix ipsecif(4) SPDADD pfkey message has garbage.  Pointed out by ohishi@IIJ.

"setkey -x" output is the following.
========== before ==========
sadb_msg{ version=2 type=14 errno=0 satype=0
  len=15 reserved=0 seq=0 pid=0
sadb_ext{ len=56 type=18 }
sadb_x_policy{ type=2 dir=1 id=9 }
{ len=40 proto=50 mode=1 level=3 reqid=16393
sockaddr{ len=0 family=0  }
sockaddr{ len=0 family=0  }
}
========== before ==========

========== after ==========
sadb_msg{ version=2 type=14 errno=0 satype=0
  len=11 reserved=0 seq=0 pid=0
sadb_ext{ len=24 type=18 }
sadb_x_policy{ type=2 dir=1 id=9 }
{ len=8 proto=50 mode=1 level=3 reqid=16393
}
========== after ==========

reduce unnecessary reqid of NAT-T ipsecif(4), suggested by ohishi@IIJ.

Fix kern/55066.  Pointed out and fixed by Chuck Zmudzinski, thanks.
ok'ed by ozaki-r@n.o

(martin)

regen.

(nisimura)

add ASMedia Technology

(nisimura)

Fix kern/55066.  Pointed out and fixed by Chuck Zmudzinski, thanks.

ok'ed by ozaki-r@n.o

(knakahara)

Use I40E_PFINT_ITRN registers to enable ITR for MSI-X interrupts

(yamaguchi)

Fix locking against myself at ixl_link_state_update()
called by ixl_get_link_status()

Though the timing to call the function is change,
there is no order to get link status at initialization.

(yamaguchi)

Ticket #779

(martin)

Pull up following revision(s) (requested by thorpej in ticket #779):

sys/arch/mips/mips/bus_dma.c: revision 1.39

Allow len == 0 in bus_dmamap_sync().

XXX pullup-9

(martin)

Improve error check:

- We check PHY register read error correctly (timeout and NFE_PHY_ERROR), so
  don't check NFE_PHY_DATA register's value with 0xffffffff or 0. At least,
  some registers may have 0.
- Check NFE_PHY_ERROR bit in nfe_miibus_writereg().
- Improve debug printf

(msaitoh)

Regen.

(msaitoh)

0x001374 is non-bitreversed value of Attansic OUI(0x00c82e).

Attansic/Atheros correctly uses ID1 and ID2 register, so delete all 0x001374
related entries.

(msaitoh)

- Add Quake Technologies and Aeluros' OUI
- Add Teranetics TN1010 10GBase-T PHY

(msaitoh)

Use BUS_ADDR_{LO,HI}32().

(thorpej)

Allow len == 0 in bus_dmamap_sync().

XXX pullup-9

(thorpej)

Support MBUFTRACE.

(thorpej)

curses: wrap the erase logic in a macro

Easier to use, it's in one place and now hopefully everyone is happy.

X

(roy)

reduce unnecessary reqid of NAT-T ipsecif(4), suggested by ohishi@IIJ.

(knakahara)

Fix build without kbd at zstty, simplify.

(rin)

Adjust the logic for enabling the 64-bit data path when a 64-bit slot
is detected:
- If DATA64_EN isn't set in CFG after a reset, don't use 64-bit data path
  at all (it's been disabled by an EEPROM setting).
- Provide a hook for force-disabling the 64-bit data path.
- Otherwise, perform the "known 64-bit cards" check as done previously
  (because dodgy-vendor-EEPROM-settings still applies).

(thorpej)

- Add the D-Link products, identified by subsystem ID.
- On some variations, the internal PHY is ghosted at #0 and #1.  Work
  around this by ignoring PHY #0 accesses unless we don't find one, and
  then look for one there as a fall-back if we don't detect anything else.
- Fix access width when setting the TxDMAUrgentThresh register.
- Support MBUFTRACE.

(thorpej)

xbd backend not supporting cache flush is not an autoconfiguration error

(jdolecek)

pmap_tlb_miss_lock needs to be globally visible.

(thorpej)

If we're going to consult the MULTIPROCESSOR option, we should pull in
"opt_multiprocessor.h".

(thorpej)

regen for xkeyboard-config 2.29

(snj)

Disable routing perms check for now.

(christos)

move debugging code after the NULL check.

(christos)

Handle fds that are pointing to routing sockets. If the fd has access to
make changes via the routing socket, grant full permission to make filter
changes.

(christos)

Fix typo

(martin)

Reduce diff with upstream

Remove local no longer needed modifications.

(kamil)

curses: use perror rather than err in initscr

libhack lacks err and perror is more portable.

(roy)

Add new netbsd32 quota modules to sets lists

(pgoyette)

regen

(pgoyette)

Split out the quota code from the rest of compat_netbsd32 module.  This
allows loading of compat_netbsd32 on kernels that don't have ``options
QUOTA'' enabled.

(pgoyette)

terminfo: truncate numeric parameters to a short

Because that is what our API demands.
We should probably change to int when we next bump the API.

(roy)