Fix value for RE_TDESC_CMD_LGSEND_V6 (currently unused)

(jmcneill)

Performance improvements for PCIe and 8168 based devices:
- When using the countdown timer for interrupt moderation on PCIe devices,
  use a timer rate value based on a 125MHz PCIe reference clock instead of
  33 MHz.
- For 8168 based devices, ditch the countdown timer and instead use the
  (undocumented) hardware interrupt moderation feature.
- Support TSOv4 on 8168D and later devices.

(jmcneill)

Offering for the wiz daemon.

(riastradh)

Modularize ualea(4).

Unclear why we have a separate xyz.ioconf for module and XYZ.ioconf
for rump component, but at least xyz_modcmd obviates the need for
xyz_component.c (though evidently the latter could have been replaced
anyway by RUMP_COMPONENT=ioconf in the rump component Makefile).

(riastradh)

Include options CAN and canloop pseudo-device by default.
Add awincan0 device to CUBIEBOARD

(bouyer)

First draft of a driver for the CAN controller found in the
Allwinner A20 SoC.

(bouyer)

include socketvar.h, so avoid a warning on struct sockopt *

(bouyer)

if_canloop.c is in netcan/

(bouyer)

Add can_ifinit_timings() helper function, which initialise timing parameters
to known invalid values.

(bouyer)

Remove clt_clock_freq from user-supplied timing informations, it is defined
by the brp.

(bouyer)

sun4v: Add handling of trap 0x06c @ trap level 1 - based on code from OpenBSD, but slightly adapted to NetBSD. verified using qemu

(palle)

amend 1367

(snj)

Pull up following revision(s) (requested by bouyer in ticket #1367):
sys/arch/amd64/conf/XEN3_DOM0: revision 1.134
sys/arch/i386/conf/XEN3_DOM0: revision 1.112
Be more explicit about the status of dom0 SMP support.

(snj)

Make comment match.

(riastradh)

regen

(riastradh)

gson@ informs me the Alea I and II have the same USB product id.

Thus, ualea(4) should work for both of them.  Rename USB product id
ARANEUS ALEA_II to ARANEUS ALEA to reflect this.

No functional change intended, except perhaps via usbdevs.

(riastradh)

Fix the same typo again in the same comment.

Still no functional change.

(riastradh)

Fix typo in comment.  No functional change.

(riastradh)

new OpenSSH

(christos)

bump libssh

(christos)

bump

(christos)

merge conflicts

(christos)

PR/52174: Remove root test, it is too verbose. XXX: need to come up with
something better.

(christos)

Remove redundant duplicate assignment.

from clang static analyzer

(maya)

Remove duplicate assignment.
We assign the same value unconditionally just before.

from clang static analyzer

XXX surrounding code seems fishy

(maya)

remove redundant assignments
nchan is overwritten before it is used in all cases. so is i.

from clang static analyzer
ok riastradh

(maya)

switch pmc_ncounters to unsigned int.

matches userland pmc, makes it clear to static analyzers that if the loop
in pmc_nmi (for (i = 0; i < pmc_ncounters; i++) ) is not entered, then the
condition i == pmc_ncounters (== 0) is satisfied and no null derefs occur

this change only helps analyzers read the code, null deref was not possible
before.

(maya)

First pass reducing magic numbers adding register defines.
No change to register values/writes, except VLPCI_MISC_CTL_HIADDR_DIS added.

(flxd)

always initialize iovcnt

from clang static analyzer

(maya)

KNF, whitespace.

(flxd)

change aprint_verbose() to know easily msi devices affinity to CPU#0 or not.

suggested by msaitoh@n.o.

(knakahara)

use DPRINTF instead of #define INTRDEBUG and printf().

(knakahara)

avoid calloc(1,0).
from clang static analyzer

ok christos

(maya)

Use %zu for size_t (fix build of 32bit archs)

(ozaki-r)

Add missing ':'

(msaitoh)

Swap arguments to strncmp() so that the thing we are testing, the
section name, comes first; and so that the length of the string
constant is right next to that constant.

(uwe)

Move deallocation of shstrsz inside the block where it is allocated.

(uwe)

Convert IPSEC_ASSERT to KASSERT or KASSERTMSG

IPSEC_ASSERT just discarded specified message...

(ozaki-r)

Remove __FreeBSD__ and __NetBSD__ switches

No functional changes (except for a debug printf).

Note that there remain some __FreeBSD__ for sysctl knobs which counerparts
to NetBSD don't exist. And ipsec_osdep.h isn't touched yet; tidying it up
requires actual code changes.

(ozaki-r)

Enable PCIe's interrupt as much as possilbe in ppb(4) to detect and count
status change event. HotPlug function itself have not implemented yet.

- Interrupt and each event are counted by evcnt(9). Example:

  ppb0 Interrupt                                                0    0 intr
  ppb0 Attention Button Pressed                                0    0 misc
  ppb0 Power Fault Detected                                    0    0 misc
  ppb0 MRL Sensor Changed                                      0    0 misc
  ppb0 Presence Detect Changed                                  0    0 misc
  ppb0 Command Completed                                        0    0 misc
  ppb0 Data Link Layer State Changed                            0    0 misc

- Print message if ppb_printevent is not zero. The default vaule is 0.
  The output messages:

  Attention Button Pressed
  Power Fault Detected
  MRL Sensor Changed
  Presence Detect Changed
  Command Completed
  Data Link Layer State Changed

- Remove workaround code to disable interrupt (ppb.c rev. 1.35).

Tested with Dell Latitude 2120 without if_bge.c rev. 1.304's workaround.
dmesg when bge's device timeout occured:

  ppb3: Presence Detect Changed
  ppb3: Data Link Layer State Changed
  ppb3: Presence Detect Changed

vmstat -e |grep ppb

  ppb3 Interrupt                                            2    0 intr
  ppb3 Presence Detect Changed                              2    0 misc
  ppb3 Data Link Layer State Changed                        1    0 misc

(msaitoh)

Remove dummy breaks and returns after bc_exit
bc_exit being marked nonreturn silences compiler warnings.

(maya)

libnetpgpverify major bump to 5

(agc)

Improved mixing function - more expensive in cycles though.
hw.driverN.saturate is no longer needed and has been removed.

Mixing function thanks to jmcneill@.

(nat)

Add canconfig(8), an utility to configure CAN interfaces.
Compile-tested only.

(bouyer)

Add CAN registers

(bouyer)

Make it build from build.sh (fix warnings)

(bouyer)

Add can-related directories and files; a build.sh release now completes.

(bouyer)

Add infranstructure to configure timings from userland on a can interface.
This uses the SIOCGDRVSPEC/SIOCSDRVSPEC ioctls.
Compile-tested only.

(bouyer)

hw.driverN.saturate bool->integer (default 16).

This means for a greater number than 16 (or user set value) the saturate
function is turned off.  Results in better listening if a large number of
channels are in use.

(nat)

Bumped libnetpgpverify major from 4 to 5

(agc)

Update netpgpverify sources in base from 20160617 to 20170201 (i.e. bring
over changes from master sources in pkgsrc/security/netpgpverify, version 20170201):

Changes:

Update netpgpverify (and libnetpgpverify) to 20160614
+ handle signatures created by gpg with "--no-emit-version", don't assume
there will always be a version string.
+ add a test for above
Fixes security PR  51240.
Thanks to xnox@ubuntu.com for reporting the error

Update netpgpverify and libnetpgpverify to 20160615:
Simplify the method of finding the end of the versioning information
in the signature - back up to the "\n" character at the end of the
signature start:

"-----BEGIN PGP SIGNATURE-----\n"

and then find the "\n\n" character sequence to denote the start of the
signature itself. The previous version worked, but this is more efficient.

Update netpgpverify and libnetpgpverify to 20160616
+ bring over joerg's printflike change from the netpgpverify
version in src/crypto
+ add a test for cleartext signatures with version information
to complement the one with no version information

Update netpgpverify and libnetpgpverify to 20160622 during freeze to fix PR  51262
+ take a bit of a step backwards, and don't use stdbool.h, just to appease
Solaris 10 compiler

Update netpgpverify and libnetpgpverify to 20160623
+ remove use of asprintf and vasprintf from libverify. Inspired
by work from Dimitri John Ledkov. Should allow building on Linux
without superfluous definitions.
+ also free the BIGNUM struct in PGPV_BN_clear() - from Dimitri
John Ledkov

Update netpgpverify and libnetpgpverify to 20160626
+ make the pgpv_t and pgpv_cursor_t structures opaque
+ add new accessor functions for fields in the pgpv_cursor_t struct
+ add new creation functions for the pgpv_t and pgpv_cursor_t structs

Update netpgpverify and libnetpgpverify to 20160704
+ get rid of redundant PGPV_ARRAY definition in libverify.c, brought in when
the definitions moved from verify.h
+ fix obuf_add_mem() to use a const void *, as any struct can be
dumped using it
+ remove redundant NO_SUBKEYS definition - unused
+ add an (unused as yet) ARRAY_FREE() macro

Update netpgpverify and libnetpgpverify to 20160705
External API changes
====================
+ add a pgpv_cursor_close() function to free resources associated with
a cursor
Better memory management
========================
+ restructure the way dynamic arrays are used, to avoid memory
corruption issues and memory leaks - keep all dynamic arrays in the global
data structure, and use indices in the other data structures to index them.
Means lack of data localisation, but avoids stale pointers, and leaks.
+ make signer field of signature a uint8_t array, rather than a pointer
+ use our own version of strdup(3) - don't depend on it being
available in standard library
+ keep track of whether litdata filenames and userid were allocated or not,
and free memory in pgpv_close() if it was allocated
+ free up allocated resources which were allocated in pgpv_close()

Update netpgpverify and libnetpgpverify to 20160706
+ 20160705 introduced a bug whereby a key subid would match and verify
fine, but, if formatted, would not display the correct subkey
information.  Fix to show the correct information in this case.

Update netpgpverify and libnetpgpverify to 20160707 to fix some
unusual build errors shown by old gcc versions (works fine for
gcc-5.2.1 on ubuntu and gcc-5.3.0 on NetBSD 7.99.32)
+ use ULL suffix on unsigned 64bit constants, not UL
+ don't typedef the public structs twice - second time just define it
without the typedef
Fixes PR  51327

Update netpgpverify and libnetpgpverify to 20160708
+ clear and free bignums properly - helps immensely with plugging
memory leaks

Update netpgpverify and libnetpgpverify to 20160828
+ bring over change from christos in src/crypto to check for
the end of an ASCII-armored signature
+ no need for namespace protection in array.h any more, now
that netpgp/verify.h now contains opaque structures
+ minor typo clean-up in a definition (benign, ignored by compiler)

update netpgpverify and libnetpgpverify to 20170201
+ make sure howmany() macro is defined
pointed out by cube - thanks!

(agc)

Fix month in Dd.

(wiz)

If -x (xsrc location) is passed for "check", display it also in the
"how to fix" invocation instructions.

(martin)

Section at index 0 is SHN_UNDEF.  When searching for symbol sections
skip it and start with index 1.

(uwe)

Don't hide first = 1 assignment inside for(), that just obscures the
loop.  Same object code is generated.

(uwe)

IPL_SOFTUSB suffices here to synchronize with usb xfer callback.

(riastradh)

Follow my own locking rules.

(riastradh)

Perform icache syncs for ARM_MMU_EXTENDED as well.  This helps the PT_STEP
code in pr/52119 and probably other things.

(skrll)

Convert bc(1) to mdoc.

(wiz)

Use correct header for SIZE_MAX. from a.rin

(maya)

Disable MULTIPROCESSOR for dom0. The dom0 support code is not MP-safe
(see xen_shm_machdep.c for example, but there are probably others).

(bouyer)

Use macro. No functional change.

(msaitoh)

Tweak locking rule.  Fix broken unlocked rmw.

(riastradh)

fix build failure

(knakahara)

regen

(riastradh)

New rndsource driver for Araneus Alea II TRNG USB devices.

Disabled by default in x86/GENERIC and usbdevices.config pending
review and testing without rump ugenhc in the way, but enabled in
x86/ALL for compile-testing.

(Hi gson!  Finally found a round tuit in my pocket, next to a certain
rectangular one.)

(riastradh)

Welcome to 7.99.70

(hannken)

Add vfs_trybusy() and mountlist_iterator_trynext() and use it for the syncer.

(hannken)

Update vfs_busy(), vfs_unbusy(), vfs_mountalloc() and vfs_rootmountalloc().

(hannken)

No need to keep a not yet visible mount busy.  Move vfs_busy()
from vfs_mountalloc() to vfs_rootmountalloc().

XXX: Do we really need to vfs_busy() for vfs_mountroot?

(hannken)

Remove unused argument "nextp" from vfs_busy() and vfs_unbusy().
Remove argument "keepref" from vfs_unbusy() and add vfs_ref() where needed.

(hannken)

Add vfs_ref(mp) and vfs_rele(mp) to add or remove a reference to
struct mount.  Rename vfs_destroy(mp) to vfs_rele(mp) and replace
incrementing mp->mnt_refcnt with vfs_ref(mp).

(hannken)

Cleanup after mountlist iterator:
- remove now unused field mnt_list.
- rename mount_list to mountlist and make it local to vfs_mount.c.
- make mountlist_lock local to vfs_mount.c.

Change pstat.c to retrieve vnodes by lru lists.

(hannken)

Uncomment MULTIPROCESSOR in dom0 kernels

(maya)

Fix bulk xfer buffer with usedma = false.

(riastradh)

1354, 1367

(snj)

Pull up following revision(s) (requested by khorben in ticket #1367):
sys/arch/amd64/conf/XEN3_DOM0: revision 1.126
sys/arch/i386/conf/XEN3_DOM0: revision 1.104
sys/arch/xen/x86/xen_pmap.c: revision 1.25
In the MP case,
do not attempt to pmap_tlb_shootdown() after a pmap_kenter_ma() during
boot. pmap_tlb_shootdown() assumes post boot. Instead invalidate the
entry on the local CPU only.
XXX: to DTRT, probably this assumption needs re-examination.
XXX: The tradeoff is a (predicted) single word size comparison
      penalty, so perhaps a decision needs performance stats.
xen dom0 SMP is now bootable again.
--
add the 'options MULTIPROCESSOR' in respective configs, but mark them
experimental - and thus disabled by default.

(snj)

Accept AH with NULL algorithm of zero-length key

(ozaki-r)

regen for ticket 1354

(snj)

Pull up following revision(s) (requested by manu in ticket #1354):
sys/compat/linux/arch/alpha/syscalls.master: revision 1.92 via patch
sys/compat/linux/arch/amd64/syscalls.master: revision 1.58 via patch
sys/compat/linux/arch/arm/syscalls.master: revision 1.65 via patch
sys/compat/linux/arch/i386/syscalls.master: revision 1.122 via patch
sys/compat/linux/arch/m68k/syscalls.master: revision 1.91 via patch
sys/compat/linux/arch/mips/syscalls.master: revision 1.61 via patch
sys/compat/linux/arch/powerpc/syscalls.master: revision 1.70 via patch
sys/compat/linux/common/linux_misc.c: revision 1.234
sys/compat/linux/common/linux_signal.h: revision 1.31
Add pselect6 Linux system call.
This lets Matlab R2016A run on NetBSD/amd64

(snj)

Fix order of outputs in rumpuser_open call.

Now ugenhc works again after a four-year hiatus...

(riastradh)

add opencrypto atf for CRYPTO_3DES_CBC and CRYPTO_AES_CBC(=CRYPTO_RIJNDAEL128_CBC).

(knakahara)

Set cpu_cc_freq when changing frequencies.

(jmcneill)

Omit needless XXX comment.

(riastradh)

Omit needless XXX comment.

(riastradh)

Justify the last unjustified assertion here.

Sprinkle a few more assertions to help along the way.

(Actually, it was justified; I just hadn't made explicit the relation
to the value of fdpos that all two callers specify.)

(riastradh)

Emphasize that MAX_WINDOW_SIZE is bounded by the maximum uint32_t.

Since we store window sizes in uint32_t, the maximum had better fit
in uint32_t!

(riastradh)

Fix previous; pci_intr_setattr takes a pci_intr_handle_t not the driver's
opaque handle.

(jmcneill)

Plug memory leak.

CID 978372

(riastradh)

Plug memory leak in ParseTraditionalInclude.

(riastradh)

Plug memory leak.

CID 978370

(riastradh)

Make it clearer that this is not a memory leak.

CID 978369

(riastradh)

Plug memory leak in JobPrintCommand.

CID 978366

(riastradh)

Plug memory leaks in Dir_FindFile.

CID 978364

(riastradh)

Clarify that assignment is for effect, not truth of value.

CID 977502

(riastradh)

Guarantee no arithmetic overflow.

(riastradh)

Replace dead conditional by assert.

CID 975995

(riastradh)

Check fcntl return values out of paranoia.

CID 975277
CID 975278

(riastradh)

Check return value of fseek.

CID 975275
CID 975276

(riastradh)

Assert archive member syntax.  Cite who guarantees it.

(riastradh)

nowadays function specifiers go in declaration_specifiers
(not after the declarator)

(dholland)

Factor out parts of cached_realpath.

This wacky API confused Coverity (CID 1405167) and is no good for
human readers either.

(riastradh)

Remove unmaintained link to imrryr from NetBSD specific fortune(6) entries

The original content is not accessible with the original link (missing
closing /). This page does not look to be updated since 2002 and artwork
is not freely reusable.

(kamil)

Use bmake_malloc where we don't test the return value of malloc currently.

Appeases coverity.

(maya)

Use, don't kludge, MAKE_ATTR_UNUSED.

CID 1300234
CID 1300237
CID 1300238
CID 1300245
CID 1300255
CID 1300267
CID 1300284

(riastradh)

Stress rump hyperentropy a little harder.

(riastradh)

Add support for PCI_INTR_MPSAFE

(jmcneill)

Add function ptr for pci_intr_setattr to struct arm32_pci_chipset

(jmcneill)

Can't destroy pb until we're done using it.

(riastradh)

regen to confirm no functional change

(riastradh)

Count vnode arguments correctly.

Don't count arguments that have WILLRELE/WILLPUT; count arguments
that are struct vnode *.

No functional change currently because it happens that every released
or put vnode argument comes first or after other ones.

(riastradh)

regen

(riastradh)

Back out previous.

Breaks file systems for which VOP_UNLOCK doesn't work on a reclaimed
vnode.

The only case in tree right now is sys/fs/union -- most file systems
use genfs_unlock, which does work on a reclaimed vnode.

Maybe we can work around this -- and still enable VOP_RECLAIM's
callees to assert lock ownership -- by having VOP_RECLAIM unlock the
vnode instead.

(riastradh)

Adapt for clock domain support

(jmcneill)

Init array no longer required

(jmcneill)

typo fix

(jmcneill)

Disable SDR104 until the Tegra K1 custom tuning method is implemented.
This is required to work around errata that describes periodic data CRC
errors after autotuning has completed.

(jmcneill)

ATF test attach2/t_ptrace_wait* is unstable

Mark it broken with:

PR kern/52168 attach2:t_ptrace_wait* is unsable

In order to get reliable failure, abort this test at the end (if we get
there).

(kamil)

Mark signal3/t_ptrace_wait* with timeout on sparc 32-bit

PR kern/52167 signal3/t_ptrace_wait* timeouts on sparc 32-bit

This test is also marked as failure on other platforms:

PR kern/51918 ptrace tracee can prevent tracer to get its signals by masking

Both issues are valid on sparc 32-bit, but the timeout one wins.

(kamil)

Mark syscallemu1/t_ptrace_wait* broken on 32-bit sparc

PR kern/52166

(kamil)

Add fclock(4)

(jmcneill)

Add support for simple fixed-rate clock source bindings.

(jmcneill)

Add support for multiple clock domains in clk API.

(jmcneill)

Remove list of init devices

(jmcneill)

get cpu pll clock from devicetree

(jmcneill)

Attach devices in order of "phandle" property inserted by DTC.

(jmcneill)

Get DC clock parent from devicetree

(jmcneill)

Restore atf_tc_expect_fail in signal4 of t_ptrace_wait*

Accidentally there was removed a PR, it was never fixed.

PR 51918 "ptrace tracee can prevent tracer to get its signals by masking"

(kamil)

Revert "Mark tests of tunnel/AH/IPv6 as expected failure (PR kern/52161)"

The issue was fixed by christos@

(ozaki-r)

Wording.

(wiz)

Merge "userland" to be consistent with other man pages.

(wiz)

More wording and whitespace.

(wiz)

use xfer->c_slot instead of slot = 0

fix ahci_bio_complete() calls - last param is reason, not slot

(jdolecek)

ahci_intr_port(): use same code on both the error and non-error path for
for getting the slot and skipping interrupts for non-active slots

(jdolecek)

Cross reference the manuals for listed modules.

(sevan)

regen to confirm no functional change

(riastradh)

Keep vnode locked during VOP_RECLAIM.

No bump because it wouldn't have been possible to acquire the lock in
VOP_RECLAIM anyway -- instant deadlock because vn_lock waits to
transition out of the RECLAIMING state first.  Benefit is that we can
now assert ownership of the lock in any operations called by
VOP_RECLAIM.

Discussed on tech-kern:

https://mail-index.netbsd.org/tech-kern/2017/04/01/msg021751.html

(riastradh)

cosmetic fixes:
- __func__ in printfs
- no space after sizeof
- eliminate useless casts
- u_intX_t -> uintX_t

(christos)

tweak english

(jdolecek)

Build without installation example lua kernel module: luareadhappy

(kamil)

Add authors email address.

(sevan)

probe for NCQ and WRITE DMA FUA EXT support

(jdolecek)

Re-edit
Add authors email address
Remove mb(9) as it's depricated
Bump date

TODO:
Software & Hard interrupt context description.

(sevan)

pass also ata_command via ata_xfer, callers of ata_exec_command() is now
responsible for allocation/disposal of the structure

change code to allocate ata_xfer for commands on stack same way as previously
the ata_command were, using c_slot 0; adjust asserts so that it would allow
several xfers with same c_slot, as long as only one such transfer is active
at a time

(jdolecek)

No need for membar_datadep_consumer here.

PSLIST_READER_FOREACH takes care of it already.
PSLIST_WRITER_FOREACH is exclusive so doesn't need it.

(riastradh)