Take vnode lock for VOP_FSYNC().

(hannken)

Flush the log to disk when ffs_sync() gets called with MNT_WAIT.

(hannken)

Take fstrans_start before syncing a file system.

(hannken)

Let syncer try fstrans_start() before running VFS_SYNC() to get rid
of the syncer lock/unlock from vfs_suspend().

(hannken)

Protect attaching and detaching lwp_info to mount with a mutex.

(hannken)

more fix about IFF_OACTIVE.

fix TODO of r1.479 and reduce IFF_OACTIVE manipulation.

(knakahara)

refactor: functionize deferred_start code holding locks.

(knakahara)

refactor: functionize txrxintr disable/enable code.

(knakahara)

Protect sysctl_net_inet_ip_pmtudto with icmp_mtx instead of softnet_lock

(ozaki-r)

Make NOMPSAFE comments informative

(ozaki-r)

Rename if_acquire_NOMPSAFE to if_acquire

It can be used in MP-safe ways. So let's remove the confusing postfix.
If it's used in a unsafe way, warn NOMPSAFE in a comment.

(ozaki-r)

Fill rmx_locks too

Otherwise userland sees garbage in it.

This should fix t_mtudisc6 failing on babylon5.

(ozaki-r)

Commit TRAP_HWWPT -> TRAP_DBREG switch in <sys/siginfo.h>

This is integral part of the previous two commits.

Sponsored by <The NetBSD Foundation>

(kamil)

Note TRAP_HWWPT -> TRAP_DBREG rename manual steps for build.sh -u

(kamil)

Add new si_code TRAP_DBREG

It replaces previous TRAP_HWWPT and is designed to be used for debug
register traps (both watchpoints and breakpoints).

TRAP_HWWPT wasn't documented in siginfo(2) neither noted in doc/CHANGES -
- document it and add new entry in CHANGES. This move is a step towards
switch the watchpoint ptrace(2) api to PT_*DBREGS.

This code was introduced recently and has no impact on stable releases.

Sponsored by <The NetBSD Foundation>

(kamil)

Make the test more stable

(ozaki-r)

Make test names self-descriptive

(ozaki-r)

allow 0 timestamp

(christos)

Allow 0 timestamp

(christos)

allow 0 timestamp

(christos)

need <sys/stat.h>

(christos)

fix msdos reproducible builds!

(christos)

start fixing msdosfs for reproducible builds; does not work yet.

(christos)

add -T for reproducible builds (used in makefs)

(christos)

PR/51973: Use proper fd for AuthorizedKeysCommand

(christos)

The default values for MKPIC are yes and no for sh3 and m68000, respectively.
Bump date.

(rin)

Switch back arch/amd64/t_ptrace_wait.c to r1.6 and restore original DBREGS

This is preparation to replace watchpoint API to dbregs.

Sponsored by <The NetBSD Foundation>

(kamil)

set pixmap pitch alignment to 8 bytes - the blitter will lock up otherwise
which isn't mentioned anywhere in the manual.
Leave in #ifdef DEBUG code to trace blitter timeouts just in case

(macallan)

On i386 (but not on amd64) we can enable SSE comparatively very late, when
probing/attaching the FPU.  This is a problem for cpu_rng with the VIA
processors because, by design, cpu_rng attaches, and the entropy subsystem
starts up, very early.

If SSE is not enabled, calls to any "PadLock" instructions (ACE, RNG)
on the VIA processors will trap, per the manual:
linux.via.com.tw/support/beginDownload.action?eleid=181&fid=261

All VIA CPUs with PadLock, or which match the model/stepping test as
possibly having PadLock, have SSE.  Just unconditionally enable it before
trying to turn the crypto block on.

Fixes crash at RNG attach time reported by Andrus V.; fix proposed by
jak@.

(tls)

New sentence, new line.
Remove superfluous macros and whitespace.
Some wording improvements.

(wiz)

New sentence, new line.
Remove superfluous Oc and merge line to make it less likely to happen again.
Fix typo.

(wiz)

Make sure to prepare /dev/xp node entry.

Missed in the initial xp(4) support import:
http://mail-index.netbsd.org/source-changes/2016/12/03/msg079493.html

(tsutsui)

update doc/CHAGES

(knakahara)

add l2tp(4) to amd64/GENERIC and amd64/ALL config.

(knakahara)

add l2tp(4) basic test.

(knakahara)

l2tp(4) support rump.

(knakahara)

add man for l2tp(4).

(knakahara)

add l2tp(4) subcomands to ifconfig(8).

originally implemented by IIJ SEIL team.

(knakahara)

add missing files.

(knakahara)

support interface name which includes digit.

(knakahara)

Use nc instead of ftp/httpd

ftp with rumphijack is unstable probably because ftp uses siglongjmp from
a signal hander. So stop using ftp and use nc instead. This fixes test
failures of t_mtudisc on some environments such as my development machine
(amd64) and anita on sparc64.

(ozaki-r)

add l2tp(4) L2TPv3 interface.

originally implemented by IIJ SEIL team.

(knakahara)

Support paccept for nc

(ozaki-r)

Do if_snd check even if IXGBE_LEGACY_TX is used. It's used by ALTQ.

(msaitoh)

txt_si softint is used only for multiqueue (ixgbe_mq_start_locked()), so don't
schedule it in ixgbe_legacy_irq().

(msaitoh)

fix about IFF_OACTIVE.

It is required manipulate IFF_OACTIVE for old ethernet controllers and ALTQ
to use if_start. To avoid race, txq_lock of 0th txq is needed to write
IFF_OACTIVE.

TODO: read side refactor

(knakahara)

Quell maybe-uninitialized false positives from gcc -Os.

reported by John D. Baker at current-users@.
http://mail-index.netbsd.org/current-users/2017/02/15/msg031132.html

(nonaka)

Add ${GPT_TIMESTAMP}

(christos)

Add -T timestamp for reproducible builds.

(christos)

there is no -f in install

(christos)

sorry, correct English grammer

(rin)

lcWrap.c is miscompiled on vax, which results in input failure via XIM.
Besides, some clients, e.g., pkgsrc/x11/kterm, receives SIGSEGV.

(rin)

add chrp

(christos)

Reintroduce ATF_TP_ADD_TC_HAVE_DBREGS for t_ptrace_wait* ATF tests

PT_GETDBREGS & PT_SETDBREGS API will replace the current one for
watchpoints.

Sponsored by <The NetBSD Foundation>

(kamil)

Fix spelling: s/alogorithms/algorithms

(abhinav)

set the default meta_directory where we are installing postfix-files.

(christos)

Fix an inverted logic.

(tsutsui)

ftp(1): split the auth processing function.

(nonaka)

- Print Data Select and Data Scale in pci_conf_print_pcipm_cap().
- The Message Data register of MSI cap is not 32bit but 16bit.
- When the PCIE_LCAP_MAX_SPEED bitfiled is 0, it means it supports 2.5GT/s only.
- Print link de-emphasis value by "-X dB".
- Print Completion Timeout Ranges Supported filed with alphabets.
- Print TPH Completer Supported fileld's meaning.
- Print PCIE_DCAP2_MAX_EETLP correctly. 0 means 4 End-End TLP Prefixes.
- If the Supported Link Speed Vector is 0, the Link Capability 2 register is not
  implemented. Don't decode LCAP2 when the vector is 0.
- The ACS's Egress Control Vector is 32bit, so print with 0x%08x.
- Print SR-IOV's device ID.
- Use __SHIFTOUT() to avoid using magic number.
- Prefix "0x" for hexadecimal value.

(msaitoh)

Remove stray closef.

(nat)

Increase buffer size reported to strlcpy() to be one larger than the
length of the string we copy in so that there is space for the '\0'.

(elric)

Avoid if_dl and if_sadl to be NULL

Calling if_deactivate_sadl and then if_sadl_setrefs exposes NULL-ed if_dl
and if_sadl to users for a moment. It's harmful because users expect that
they're always non-NULL. Fix it.

Note that a race condition still remains; if_dl and if_sald aren't updated
atomically so a user can see different data from if_dl and if_sadl.
Fortunately none uses both if_dl and if_sadl at the same time, so the race
condition doesn't hurt nobody for now. (In the first place exposing one
data with two ways is problematic?)

(ozaki-r)

Add new entry to TODO.ptrace

research ipkdb(4)

Sponsored by <The NetBSD Foundation>

(kamil)

1433

(snj)

Pull up following revision(s) (requested by maxv in ticket #1433):
sys/compat/linux/arch/amd64/linux_machdep.c: 1.50, 1.51
Don't let userland choose %rip. This is the Intel Sysret vulnerability
again.
--
Make sure %rip is in userland. This is harmless, since the return to
userland is made with iret instead of sysret in this path. While here, use
size_t.

(snj)

1433

(snj)

Pull up following revision(s) (requested by maxv in ticket #1433):
sys/compat/linux/arch/amd64/linux_machdep.c: 1.50, 1.51
Don't let userland choose %rip. This is the Intel Sysret vulnerability
again.
--
Make sure %rip is in userland. This is harmless, since the return to
userland is made with iret instead of sysret in this path. While here, use
size_t.

(snj)

1433

(snj)

Pull up following revision(s) (requested by maxv in ticket #1433):
sys/compat/linux/arch/amd64/linux_machdep.c: 1.50, 1.51
Don't let userland choose %rip. This is the Intel Sysret vulnerability
again.
--
Make sure %rip is in userland. This is harmless, since the return to
userland is made with iret instead of sysret in this path. While here, use
size_t.

(snj)

1359

(snj)

Pull up following revision(s) (requested by maxv in ticket #1359):
sys/compat/linux/arch/amd64/linux_machdep.c: 1.50, 1.51
Don't let userland choose %rip. This is the Intel Sysret vulnerability
again.
--
Make sure %rip is in userland. This is harmless, since the return to
userland is made with iret instead of sysret in this path. While here, use
size_t.

(snj)

1358, 1359

(snj)

Pull up following revision(s) (requested by maxv in ticket #1359):
sys/compat/linux/arch/amd64/linux_machdep.c: revisions 1.50, 1.51
Don't let userland choose %rip. This is the Intel Sysret vulnerability
again.
--
Make sure %rip is in userland. This is harmless, since the return to
userland is made with iret instead of sysret in this path. While here, use
size_t.

(snj)

Pull up following revision(s) (requested by chs in ticket #1358):
sys/miscfs/procfs/procfs_map.c: revision 1.45
Maps don't change that frequently between reads, so don't give up and
do what linux does (support reading from an offset).

(snj)

Handle persistent memory. Currently only debug output.

(nonaka)

x86: add e820 memory type.

(nonaka)

x86: make btinfo_memmap from btinfo_efimemmap for to reduce mem_cluster_cnt.

should fix PR/51953.

(nonaka)

also compile poly1305.c with -O0 on vax to address ssh login failure from/to
some hosts

(rin)

Don't call grow_mixer_states with sc_intr_lock held.

Addresses PR kern/51965.

(nat)

Add most of my USER_LDT code for amd64, but disable it and put a comment
about why Wine still does not work.

Nothing changes, but at least it is a step forward.

(maxv)

correct grammar

(rin)

add hack for libmandoc on vax:
mandoc(1) receives SIGILL in in_line_argn() from mdoc_macro.c

(rin)

Check %eip with USER_LDT too.

(maxv)

add hack for libssh on vax

(rin)

Add tests for ifconfig up/down

(ozaki-r)

Fix previous

(skrll)

More stuff to do...

(skrll)

Do ND in L2_output in the same manner as arpresolve

The benefits of this change are:
- The flow is consistent with IPv4 (and FreeBSD and OpenBSD)
  - old: ip6_output => nd6_output (do ND if needed) => L2_output (lookup a stored cache)
  - new: ip6_output => L2_output (lookup a cache. Do ND if cache not found)
- We can remove some workarounds in nd6_output
- We can move L2 specific operations to their own place
- The performance slightly improves because one cache lookup is reduced

(ozaki-r)

new postfix

(christos)

Resolve conflicts.

(christos)

Add postfix .so dir

(christos)

add postfix .so dir

(christos)

sorry, fix comment...

(rin)

PR port-vax/51967 GCC 5.4 aborts on vax due to SIGILL for cc1
As a workaround, compile dse.c with -O0.

(rin)

sqlite-3.17.0 out.

(wiz)

Update current versions of some 3rd party components

(maya)

Fix typo.

(abhinav)

fix number of arguments of kmem_alloc and kmem_zalloc macro. ok skrll.

(zafer)

prune duplicate xkeyboard-config and xkbcomp

(snj)

Update TODO.ptrace

Note BSD4.4-like /proc debugging removal
Explain kevent(2)-based debugging proposal

Sponsored by <The NetBSD Foundation>

(kamil)

Make sure %rip is in userland. This is harmless, since the return to
userland is made with iret instead of sysret in this path. While here, use
size_t.

(maxv)

Don't let userland choose %rip. This is the Intel Sysret vulnerability
again.

(maxv)

128 bit integers are displayed in decimal.

bn.h and bignum.c from netgpg.

(nonaka)

NVMe 1.2.1

(nonaka)

nvme(4): Limit the number of queues to the number allocated in HW.

(nonaka)

Remove more ix*_start_locked() from interrupt context like previous commit.
> Fix a bug that ix*_start_locked() is called in interrput context
> (ix*_msix_que). The function is called in softint(ix*_handle_que()). OK'd by
>k-nakahara.

(msaitoh)

Use percpuq.

(msaitoh)

G/C VM_MAX_KERNEL_BUF

(skrll)

delete the xkbcomp tool now that nothing uses it

(snj)

Protect mtudisc and redirect stuffs of icmp/icmp6 with mutex

We have to run pr_init of icmp and icmp6 prior to tcp and tcp6 ones
for mutex initialization.

(ozaki-r)

Fix a bug that ix*_start_locked() is called in interrput context
(ix*_msix_que). The function is called in softint(ix*_handle_que()). OK'd by
k-nakahara.

(msaitoh)

catch up with 2012 xkeyboard-config and stop generating *.dir

(snj)

fix typo

(knakahara)

Fix build of ALL of amd64

(ozaki-r)

remove extra crlf.

(nonaka)

prune /usr/X11R7/lib/X11/xkb/torture, a ghost from the days of xkbdata

(snj)

Use IFQ_LOCK instead of splnet for if_snd

(ozaki-r)

Replace splnet with splsoftnet

(ozaki-r)

Remove unnecessary splnet

ok @knakahara

(ozaki-r)

Add BRIDGE_IPF

(ozaki-r)

Update comments to reflect bpf MP-ification

(ozaki-r)

update for xkeyboard-config 2.20

(snj)

merge xkeyboard-config 2.20

(snj)

Use fnullop_fcntl instead of audiofcntl.

(nat)

fix off-by-one, found by asan.

(christos)

1357

(snj)

Pull up following revision(s) (requested by mrg in ticket #1357):
libexec/httpd/CHANGES: revision 1.25
libexec/httpd/bozohttpd.8: revisions 1.63-1.65
libexec/httpd/bozohttpd.c: revisions 1.85, 1.86
libexec/httpd/bozohttpd.h: revision 1.47
libexec/httpd/cgi-bozo.c: revisions 1.36, 1.37
libexec/httpd/libbozohttpd/libbozohttpd.3: revision 1.4
libexec/httpd/testsuite/Makefile: revision 1.7
libexec/httpd/testsuite/html_cmp: revision 1.5
libexec/httpd/testsuite/test-bigfile: revision 1.4
libexec/httpd/testsuite/test-simple: revisions 1.3, 1.4
libexec/httpd/testsuite/t11.in: revision 1.1
libexec/httpd/testsuite/t11.out: revision 1.1
libexec/httpd/testsuite/cgi-bin/empty: revision 1.1
Update bozohttpd to 20170201:
- fix an infinite loop in cgi processing
- fixes and clean up for the testsuite
- no longer sends encoding header for compressed formats

(snj)

bah, combine revisions for ticket 1357

(snj)

1357

(snj)

Pull up following revision(s) (requested by mrg in ticket #1357):
libexec/httpd/CHANGES: revision 1.25
libexec/httpd/bozohttpd.8: revisions 1.63-1.65
libexec/httpd/bozohttpd.c: revisions 1.85, 1.86
libexec/httpd/bozohttpd.h: revision 1.47
libexec/httpd/cgi-bozo.c: revisions 1.36, 1.37
libexec/httpd/libbozohttpd/libbozohttpd.3: revision 1.4
libexec/httpd/testsuite/Makefile: revision 1.7
libexec/httpd/testsuite/html_cmp: revision 1.5
libexec/httpd/testsuite/test-bigfile: revision 1.4
libexec/httpd/testsuite/test-simple: revisions 1.3, 1.4
libexec/httpd/testsuite/t11.in: revision 1.1
libexec/httpd/testsuite/t11.out: revision 1.1
libexec/httpd/testsuite/cgi-bin/empty: revision 1.1
Update bozohttpd to 20170201:
- fix an infinite loop in cgi processing
- fixes and clean up for the testsuite
- no longer sends encoding header for compressed formats

(snj)

netbsd_elf_signature - look at note segments (phdrs) not note
sections.  They point to the same data in the file, but sections are
for linkers and are not necessarily present in an executable.

The original switch from phdrs to shdrs seems to be just a cop-out to
avoid parsing multiple notes per segment, which doesn't really avoid
the problem b/c sections also can contain multiple notes.

(uwe)

note accept4

(maya)

es_arglen is now in units of bytes, update the emulations accordingly.

(chs)

Add new entry in doc/CHANGES

ptrace(2): Add signal mask information accessors API:
                PT_GET_SIGMASK and PT_SET_SIGMASK

Sponsored by <The NetBSD Foundation>

(kamil)

sun4v: Revert previous - paddr_t is not a pointer... noted by nakayama@

(palle)

Add a KASSERT, otherwise it looks like a NULL deref; from Mootja.

(maxv)

Memory leak, found by Mootja; not tested, but obvious enough.

(maxv)

Uninitialized var, found by Mootja; not tested, but obvious enough.

(maxv)

overlapping strcpy is UB. use memmove
from asan+ubsan

(maya)

Make gpt(8) work when compiled on a glibc-based OS.

This restores the ability to build amd64 install-image's under Linux.

(aymeric)

Add entry for prop_ingest in the NAME section.

(abhinav)

Use .Fn to refer a function name instead of Nm.
Also, use Xr to refer problib(3) in the HISTORY section instead of Nm.

While parsing the man pages, any .Nm occurrence gets replaced by the
value specified in the NAME section. Referencing individual function
names with .Nm was causing makemandb(8) to replicate the complete
NAME section for every such occurrence. This was leading to an
ugly looking snippet in apropos(1)'s output when searched for
any of these man pages.

(abhinav)

file system police.

(wiz)

Sort sections.

(wiz)

Whitespace

(skrll)

Remove redundant splnet/splx calls - ec_lock is IPL_NET.

(skrll)

Convert to kmem(9)

(skrll)

Typo in comment

(skrll)

KNF (sort #include <sys/...>) and remove a duplicate

(skrll)

RTP timestamp is now set correctly.  The streaming socket is set to non
block blocking and connection state machinre improvement.

(nat)

Update TODO.ptrace

Remove:
- add PT_SET_SIGMASK and PT_GET_SIGMASK - used by checkpointing software

This interface has been committed to HEAD.

Sponsored by <The NetBSD Foundation>

(kamil)

Introduce new interface in ptrace(2) - PT_GET_SIGMASK and PT_SET_SIGMASK

Add new interface to add ability to get/set signal mask of a tracee.
It has been inspired by Linux PTRACE_GETSIGMASK and PTRACE_SETSIGMASK, but
adapted for NetBSD API.

This interface is used for checkpointing software to set/restore context
of a process including signal mask like criu or just to track this property
in reverse-execution software like Record and Replay Framework (rr).

Add new ATF tests for this interface
====================================
getsigmask1:
    Verify that plain PT_SET_SIGMASK can be called

getsigmask2:
    Verify that PT_SET_SIGMASK reports correct mask from tracee

setsigmask1:
    Verify that plain PT_SET_SIGMASK can be called with empty mask

setsigmask2:
    Verify that sigmask is preserved between PT_GET_SIGMASK and
    PT_SET_SIGMASK

setsigmask3:
    Verify that sigmask is preserved between PT_GET_SIGMASK, process
    resumed and PT_SET_SIGMASK

setsigmask4:
    Verify that new sigmask is visible in tracee

Kernel ABI bump delayed as there are more interfaces to come in ptrace(2).

Sponsored by <The NetBSD Foundation>

(kamil)

Update TODO.ptrace with new entries

Added entries:
- add support to read debugger events via a file descriptor in procfs
  (kevent(2)), it's still useful in cases when a parent traces tracee and has
  to call waitpid(2) for its child - as this clashes with GUI toolkits
- fix more calls for netbsd32 compat

Sponsored by <The NetBSD Foundation>

(kamil)

sun4v: Fix calculation of mmu data fault address (pointer arithmetic)

(palle)

fix typo.

(christos)

make this compile again.

(christos)

Don't use slow file_exists_exact if we don't have to.

file_exists_exact function was introduced in 1.26 for the benefit of
cross-building on OS X case-insensitive file system.  It is extremely
slow on diskless machines.  That becomes especially noticeable when
you upgrade a system that has a lot of obsolete files, it can take
literally hours.

(uwe)

new byacc

(christos)