openbsd
7 years agocreate a gross mechanism that can handle gcc-only, clang-only,
deraadt [Mon, 17 Apr 2017 19:44:59 +0000 (19:44 +0000)]
create a gross mechanism that can handle gcc-only, clang-only,
and gcc+clang architectures
with kettenis

7 years agoUse better types for checksum() parameters and local var from
krw [Mon, 17 Apr 2017 18:31:08 +0000 (18:31 +0000)]
Use better types for checksum() parameters and local var from
dhclient. Eliminate extra prototypes for checksum() and wrapsum().

7 years agoconsictently use .Dv NULL and a few other tweaks; ok schwarze@
otto [Mon, 17 Apr 2017 18:07:15 +0000 (18:07 +0000)]
consictently use .Dv NULL and a few other tweaks; ok schwarze@

7 years agodon't manipulate hdr.len, it's used internally by libutil now; ok florian@
otto [Mon, 17 Apr 2017 16:37:39 +0000 (16:37 +0000)]
don't manipulate hdr.len, it's used internally by libutil now; ok florian@

7 years agowhitespace fixes
otto [Mon, 17 Apr 2017 16:36:35 +0000 (16:36 +0000)]
whitespace fixes

7 years agodon't manipulate hdr.len, it's used internally by libutil now; ok florian@
otto [Mon, 17 Apr 2017 16:03:15 +0000 (16:03 +0000)]
don't manipulate hdr.len, it's used internally by libutil now; ok florian@

7 years agoChange build infrastructure to allow building both gcc and clang. This
kettenis [Mon, 17 Apr 2017 15:53:21 +0000 (15:53 +0000)]
Change build infrastructure to allow building both gcc and clang.  This
doesn't actually flip the switch yet, so aarch64 continues to be the only
architecture for which we build clang.

ok jsg@, deraadt@

7 years agoRecognize nl_langinfo(CODESET) return values "646" and "" as aliases
schwarze [Mon, 17 Apr 2017 14:31:23 +0000 (14:31 +0000)]
Recognize nl_langinfo(CODESET) return values "646" and "" as aliases
for "US-ASCII", useful for different versions of NetBSD and Solaris.
Found by dtucker@ and by Tom G. Christensen <tgc at jupiterrise dot com>.
OK dtucker@ deraadt@

7 years agoTest syslogd(8) "last message repeated" feature and -r option.
bluhm [Mon, 17 Apr 2017 14:27:46 +0000 (14:27 +0000)]
Test syslogd(8) "last message repeated" feature and -r option.

7 years agoAdd syslogd(8) option -r to suppress the summary line for pipe and
bluhm [Mon, 17 Apr 2017 14:18:44 +0000 (14:18 +0000)]
Add syslogd(8) option -r to suppress the summary line for pipe and
remote loghost as they are most commonly used for automated log
processing.  With -rr the "last message repeated" feature can be
disabled completely.
OK sthen@ deraadt@ jmc@

7 years agotest display of long URIs
schwarze [Mon, 17 Apr 2017 13:29:20 +0000 (13:29 +0000)]
test display of long URIs

7 years agoimplement display of long .Lk URIs to match -Tascii
schwarze [Mon, 17 Apr 2017 13:26:30 +0000 (13:26 +0000)]
implement display of long .Lk URIs to match -Tascii

7 years agoSkip objc tests with clang.
kettenis [Mon, 17 Apr 2017 12:54:48 +0000 (12:54 +0000)]
Skip objc tests with clang.

7 years agoFix handling of trailing punctuation in .Lk.
schwarze [Mon, 17 Apr 2017 12:52:00 +0000 (12:52 +0000)]
Fix handling of trailing punctuation in .Lk.
This macro is unusual in so far as trailing punction needs to remain
inside the scope because it must be inside, not after the display
of long URIs in terminal output mode.
Improves formatting of fw_update(1), help(1), less(1), sendbug(1),
acx(4), inet6(4), ipsec(4), oce(4), isakmpd.conf(5), afterboot(8),
release(8), traceroute(8).

7 years agoChange COMPILER_VERSION tests which limited additional warnings to gcc4
jsg [Mon, 17 Apr 2017 11:02:31 +0000 (11:02 +0000)]
Change COMPILER_VERSION tests which limited additional warnings to gcc4
to instead skip them on gcc3 as clang can handle -Wpointer-sign and
-Wold-style-definition.

7 years agoWe need to update the ia6_updatetime when changing the vltime/pltime
florian [Mon, 17 Apr 2017 10:29:01 +0000 (10:29 +0000)]
We need to update the ia6_updatetime when changing the vltime/pltime
vio ioctl(2). Otherwise ifconfig shows the wrong times since
vltime/pltime are calculated from this value on export.

The handling of router advertisements does the right thing.
OK bluhm

7 years agoDon't bother moving the cursor for empty lines.
nicm [Mon, 17 Apr 2017 08:10:44 +0000 (08:10 +0000)]
Don't bother moving the cursor for empty lines.

7 years agoRemove a couple of redraw flags that no longer have any effect.
nicm [Mon, 17 Apr 2017 06:40:32 +0000 (06:40 +0000)]
Remove a couple of redraw flags that no longer have any effect.

7 years agoMemory leak, from David CARLIER.
nicm [Sun, 16 Apr 2017 20:33:46 +0000 (20:33 +0000)]
Memory leak, from David CARLIER.

7 years agoUse EL1 to clear lines when redrawing the leftmost pane, rather than
nicm [Sun, 16 Apr 2017 20:32:14 +0000 (20:32 +0000)]
Use EL1 to clear lines when redrawing the leftmost pane, rather than
spaces.

7 years agofrom brad: document rgephy
jmc [Sun, 16 Apr 2017 20:26:34 +0000 (20:26 +0000)]
from brad: document rgephy

7 years agoDisable colored diagnostics in LLVM's lld.
patrick [Sun, 16 Apr 2017 19:28:25 +0000 (19:28 +0000)]
Disable colored diagnostics in LLVM's lld.

Requested by deraadt@

7 years agoAccording to termcap(3), char PC, *BC, *UP need to be extern. Fixes an
tb [Sun, 16 Apr 2017 18:04:02 +0000 (18:04 +0000)]
According to termcap(3), char PC, *BC, *UP need to be extern. Fixes an
ld(1) error found by mestre a while ago.

fix suggested and ok naddy

7 years agoAlways evaluate expression in BUG_ON() macro to avoid unused variable
bluhm [Sun, 16 Apr 2017 17:16:21 +0000 (17:16 +0000)]
Always evaluate expression in BUG_ON() macro to avoid unused variable
warnings.  Makes non diagnostic kernel compile.
OK kettenis@

7 years agoMake sure these tests print "SKIPPED" if the necessary variables aren't set.
kettenis [Sun, 16 Apr 2017 16:59:53 +0000 (16:59 +0000)]
Make sure these tests print "SKIPPED" if the necessary variables aren't set.
The current code doesn't work since the magic .BEGIN target runs before
the regress target that prints "SKIPPED" and the .BEGIN target fails when
the variables aren't set.

ok bluhm@

7 years agosync
deraadt [Sun, 16 Apr 2017 16:17:23 +0000 (16:17 +0000)]
sync

7 years agoDefine EXT_IMPLICIT_NBIT like we do on sparc64 and mips64. Makes vaious
kettenis [Sun, 16 Apr 2017 16:14:37 +0000 (16:14 +0000)]
Define EXT_IMPLICIT_NBIT like we do on sparc64 and mips64.  Makes vaious
long double math stuff (including printf) actually work.

While there remove 'struct ieee_ldouble', which isn't defined on other
architectures.

ok deraadt@

7 years agoUse INT_MAX instead of SIZE_MAX as the maximum file size we can handle.
kettenis [Sun, 16 Apr 2017 16:04:48 +0000 (16:04 +0000)]
Use INT_MAX instead of SIZE_MAX as the maximum file size we can handle.
Fixes this test on 64-bit architectures.

ok visa@

7 years agoDefine DEF_WEAK like we do for ld.so to fix building this with clang.
kettenis [Sun, 16 Apr 2017 16:01:05 +0000 (16:01 +0000)]
Define DEF_WEAK like we do for ld.so to fix building this with clang.

ok millert@, deraadt@

7 years agoMove comments into a block and uses {} to unconfuse reading.
deraadt [Sun, 16 Apr 2017 15:11:01 +0000 (15:11 +0000)]
Move comments into a block and uses {} to unconfuse reading.

7 years agoUse %zu to print a size_t.
kettenis [Sun, 16 Apr 2017 14:40:47 +0000 (14:40 +0000)]
Use %zu to print a size_t.

7 years agoReplace fetch_and_add() with atomic_inc_int_nv() from <sys/atomic.h>
visa [Sun, 16 Apr 2017 14:28:07 +0000 (14:28 +0000)]
Replace fetch_and_add() with atomic_inc_int_nv() from <sys/atomic.h>
to make the code more similar to sparc64's.

OK mpi@, guenther@, kettenis@

7 years agoFlip previously warm pages in the buffer cache to memory above the DMA
beck [Sun, 16 Apr 2017 14:25:42 +0000 (14:25 +0000)]
Flip previously warm pages in the buffer cache to memory above the DMA
region if uvm tells us it is available.
nits from deraadt@
ok krw@ guenther@ kettenis@

7 years agoStatistics for high memory flips in the buffer cache
beck [Sun, 16 Apr 2017 14:24:03 +0000 (14:24 +0000)]
Statistics for high memory flips in the buffer cache
nits from deraadt@
ok krw@ guenther@ kettenis@

7 years agoRemove some defines marked #ifdef notyet, which really is #ifdef notever.
kettenis [Sun, 16 Apr 2017 14:20:49 +0000 (14:20 +0000)]
Remove some defines marked #ifdef notyet, which really is #ifdef notever.
Reduces the diffs with other architectures.

7 years agoAlways return nonzero from _longjmp too.
kettenis [Sun, 16 Apr 2017 13:35:14 +0000 (13:35 +0000)]
Always return nonzero from _longjmp too.

ok jsg@

7 years agoRemove unused rnd_seed variable. Upstream made the same change 4 days ago.
kettenis [Sun, 16 Apr 2017 11:07:03 +0000 (11:07 +0000)]
Remove unused rnd_seed variable.  Upstream made the same change 4 days ago.
Fixes compiling this test using clang on arm64.

7 years agoremove -g from CFLAGS at florian's request
jsg [Sun, 16 Apr 2017 10:19:54 +0000 (10:19 +0000)]
remove -g from CFLAGS at florian's request

7 years agoDifferent compilers and versions of compilers have different warnings.
jsg [Sun, 16 Apr 2017 10:16:35 +0000 (10:16 +0000)]
Different compilers and versions of compilers have different warnings.
Remove -Werror to give code a greater chance of building.

ok deraadt@ florian@

7 years agoRemove /etc/ssl/acme/. We don't need it now that we have a default acme-conf(5)
ajacoutot [Sun, 16 Apr 2017 08:50:49 +0000 (08:50 +0000)]
Remove /etc/ssl/acme/. We don't need it now that we have a default acme-conf(5)
that direclty uses /etc/ssl/{,private} by default. Adapt the httpd.conf example
accordingly.

ok florian@ benno@ millert@

7 years agobackout previous, data_len is not always initialized
otto [Sun, 16 Apr 2017 06:26:55 +0000 (06:26 +0000)]
backout previous, data_len is not always initialized

7 years agoRemove -Werror to make dhcrelay6 on gcc3 architecture successfully.
aoyama [Sun, 16 Apr 2017 00:52:54 +0000 (00:52 +0000)]
Remove -Werror to make dhcrelay6 on gcc3 architecture successfully.

ok reyk

7 years agoBuild sparc64 kernels with -ffreestanding.
kettenis [Sat, 15 Apr 2017 18:58:45 +0000 (18:58 +0000)]
Build sparc64 kernels with -ffreestanding.

7 years agoDocument RB_TIMEBAD; delete RB_* that are obsolete/unimplemented
guenther [Sat, 15 Apr 2017 18:55:27 +0000 (18:55 +0000)]
Document RB_TIMEBAD; delete RB_* that are obsolete/unimplemented

ok visa@ deraadt@

7 years agomore streamlining, with help from bluhm;
jmc [Sat, 15 Apr 2017 18:45:15 +0000 (18:45 +0000)]
more streamlining, with help from bluhm;

7 years agoAfter forced unmount of a file system that has other mount points
bluhm [Sat, 15 Apr 2017 13:56:43 +0000 (13:56 +0000)]
After forced unmount of a file system that has other mount points
in it, dangling mounts could remain.  When unmounting check the
hierarcy and unmount recursively.  Also prevent that a new mount
appears during the scan.
Joint work with natano@; testing and OK krw@

7 years agoBuild full mandoc.db(5) databases by default using makewhatis(8)
schwarze [Sat, 15 Apr 2017 13:12:08 +0000 (13:12 +0000)]
Build full mandoc.db(5) databases by default using makewhatis(8)
without -Q during the build and in weekly(8).  According to tests
by many developers, makewhatis(8) takes a few minutes at most even
on slower hardware like octeon, loongson, ALIX, RPI3, Soekris,
cubox, softiron etc., and security(8) is often worse than makewhatis(8).

In case this causes excessive weekly(8) run times on even slower
(~50 MHz-class) CPUs, consider adding "MAKEWHATISARGS=-Q" to
/etc/weekly.local on machines that feel unhappy.

OK sthen@ kettenis@ millert@ deraadt@

7 years agoTake implementation for getsocket() from arp(8). This brings
bluhm [Sat, 15 Apr 2017 11:58:51 +0000 (11:58 +0000)]
Take implementation for getsocket() from arp(8).  This brings
routing table filter and pledge(2) to ndp(8) modes -s and -d.
OK florian@

7 years agoName global variable for routing socket rtsock like in ndp(8).
bluhm [Sat, 15 Apr 2017 11:50:24 +0000 (11:50 +0000)]
Name global variable for routing socket rtsock like in ndp(8).
OK florian@

7 years agoLet ndp(8) use the process's current routing table by default.
bluhm [Sat, 15 Apr 2017 11:42:09 +0000 (11:42 +0000)]
Let ndp(8) use the process's current routing table by default.
from Pierre Emeriaud;  OK jca@ claudio@
Sync code and man page with arp(8).  OK florian@

7 years agoThe tlb flushes in pmap_set_{l1,l2,l3} use ranges that don't fully make sense.
kettenis [Sat, 15 Apr 2017 11:15:02 +0000 (11:15 +0000)]
The tlb flushes in pmap_set_{l1,l2,l3} use ranges that don't fully make sense.
But those tlb flushes shouldn't be necessary anyway, so simply remove them.
Simplify the tlb flushing code now that we no longer flush ranges, and
revive the branch that doesn't flush a specific ASID for the kernel pmap
since its mappings are global.

ok patrick@, visa@

7 years agoHandle R_SPARC_WPLT30 relocations against a local symbol. Fixes building the
kettenis [Sat, 15 Apr 2017 10:23:35 +0000 (10:23 +0000)]
Handle R_SPARC_WPLT30 relocations against a local symbol.  Fixes building the
audio/openal port with clang.

ok guenther@

7 years agoregen
florian [Sat, 15 Apr 2017 09:16:05 +0000 (09:16 +0000)]
regen

7 years agoupdate to 4.1.16rc1
florian [Sat, 15 Apr 2017 09:15:45 +0000 (09:15 +0000)]
update to 4.1.16rc1
tests & OK sthen

(if there are more changes coming for 4.1.16 release we will just
commit them on top)

7 years agoFix format string in ACPI_MEMDEBUG block
guenther [Sat, 15 Apr 2017 04:54:04 +0000 (04:54 +0000)]
Fix format string in ACPI_MEMDEBUG block

From Anton Lindqvist (anton.lindqvist(at)gmail.com)

7 years agoSync bus_dmamap_load_raw() with amd64 for better constraint checking.
visa [Sat, 15 Apr 2017 04:38:27 +0000 (04:38 +0000)]
Sync bus_dmamap_load_raw() with amd64 for better constraint checking.
Needed by xhci(4).

7 years agoNo GMX on CN73xx.
visa [Sat, 15 Apr 2017 04:18:40 +0000 (04:18 +0000)]
No GMX on CN73xx.

7 years agomark newish phony targets as phony.
espie [Fri, 14 Apr 2017 21:37:15 +0000 (21:37 +0000)]
mark newish phony targets as phony.
okay tb@

7 years agoRemove unused define satoin6().
bluhm [Fri, 14 Apr 2017 20:50:35 +0000 (20:50 +0000)]
Remove unused define satoin6().

7 years agoPass down the address family through the pr_input calls. This
bluhm [Fri, 14 Apr 2017 20:46:31 +0000 (20:46 +0000)]
Pass down the address family through the pr_input calls.  This
allows to simplify code used for both IPv4 and IPv6.
OK mikeb@ deraadt@

7 years agoadjust .Lk tests after groff and mandoc .Lk improvements
schwarze [Fri, 14 Apr 2017 19:50:54 +0000 (19:50 +0000)]
adjust .Lk tests after groff and mandoc .Lk improvements

7 years agoonce cvs's flow-control mechanism has kicked in, stop reading from
jcs [Fri, 14 Apr 2017 19:39:59 +0000 (19:39 +0000)]
once cvs's flow-control mechanism has kicked in, stop reading from
its local reader until memory usage goes back down below its
watermark.

during a checkout of a tree with big files (like www) from a fast
reader (disk) and a slow writer (net), the amount of data being
buffered can back up enough that cvs tries to allocate too many
buffers and hits its resource limit, causing death.

cvs's flow-control mechanism properly detects this early on, but the
message sent to the reader process to stop sending data takes too
long to process.

take more aggressive action and just stop reading from the reader
until the writer has ejected enough data that it can start re-using
its already-allocated buffers instead of allocating new ones.

ok deraadt

7 years agoShow long .Lk URIs in like an indented display, similar to groff.
schwarze [Fri, 14 Apr 2017 19:34:54 +0000 (19:34 +0000)]
Show long .Lk URIs in like an indented display, similar to groff.
Suggested by bentley@.

7 years agoAdd test for IP payload compression. It is sending small and big
bluhm [Fri, 14 Apr 2017 19:03:50 +0000 (19:03 +0000)]
Add test for IP payload compression.  It is sending small and big
ping packets as only the latter get actually compressed.

7 years agoDo not make the colon after the .Lk link text italic.
schwarze [Fri, 14 Apr 2017 18:24:15 +0000 (18:24 +0000)]
Do not make the colon after the .Lk link text italic.
I just pushed the same change to GNU troff.

7 years agoDo not make the colon after the .Lk link text italic.
schwarze [Fri, 14 Apr 2017 18:23:29 +0000 (18:23 +0000)]
Do not make the colon after the .Lk link text italic.
I just pushed the same change to GNU troff.

7 years agoAdd tests with the ipsec.conf SA bundle keyword.
bluhm [Fri, 14 Apr 2017 18:14:33 +0000 (18:14 +0000)]
Add tests with the ipsec.conf SA bundle keyword.

7 years agoUp to now ipsecctl(8) grouped SAs with identical src and dst to the
bluhm [Fri, 14 Apr 2017 18:06:28 +0000 (18:06 +0000)]
Up to now ipsecctl(8) grouped SAs with identical src and dst to the
flow which the first SA matched by the flow type.  This behaviour
was mostly undocumented and unexpected.  Make SA bundles explicit
in ipsec.conf(5).  Only group SAs that have the same src and dst
and also the same bundle identifier.
OK hshoexer@

7 years agoUse freezero(3) when cleaning up session tickets - not only does it require
jsing [Fri, 14 Apr 2017 15:32:41 +0000 (15:32 +0000)]
Use freezero(3) when cleaning up session tickets - not only does it require
less code, but there is also a potential performance gain since they can be
larger allocations.

7 years agoUse freezero(3) to clean up the X25519 keys - simpler, cleaner code.
jsing [Fri, 14 Apr 2017 15:26:53 +0000 (15:26 +0000)]
Use freezero(3) to clean up the X25519 keys - simpler, cleaner code.

7 years agoUse freezero(3) in the CBB clean up path, since this could hold sensitive
jsing [Fri, 14 Apr 2017 15:20:55 +0000 (15:20 +0000)]
Use freezero(3) in the CBB clean up path, since this could hold sensitive
information (such as master keys).

7 years agoSwitch i2d_SSL_SESSION() back to freezero(3) now that the size constraints
jsing [Fri, 14 Apr 2017 15:20:16 +0000 (15:20 +0000)]
Switch i2d_SSL_SESSION() back to freezero(3) now that the size constraints
have been relaxed.

7 years agoClean up server key exchange EC point handling. Encode the point directly
jsing [Fri, 14 Apr 2017 15:19:39 +0000 (15:19 +0000)]
Clean up server key exchange EC point handling. Encode the point directly
into the CBB memory, rather than mallocing and memcpying, which also makes
makes the code more consistent with the client. Add a missing check for the
first EC_POINT_point2oct() call.

ok beck@

7 years agoAvoid some false positives with cppcheck. No binary change.
bluhm [Fri, 14 Apr 2017 15:11:31 +0000 (15:11 +0000)]
Avoid some false positives with cppcheck.  No binary change.
OK jsg@

7 years agowhitespace
deraadt [Fri, 14 Apr 2017 15:02:51 +0000 (15:02 +0000)]
whitespace

7 years agoDefine __dead and __pure even if __STRICT_ANSI__ for compilers that
kettenis [Fri, 14 Apr 2017 07:22:02 +0000 (07:22 +0000)]
Define __dead and __pure even if __STRICT_ANSI__ for compilers that
support GCC's __attribute__ syntax.  Fixes several warnings when compiling
code with clang with -std=c++11 or some other strict ISO standard.

ok millert@, guenther@

7 years agocorrect path; from Klemens Nanni
deraadt [Fri, 14 Apr 2017 01:23:02 +0000 (01:23 +0000)]
correct path; from Klemens Nanni

7 years agoSVM: calculate max ASID value and save for later use. This will be used in
mlarkin [Fri, 14 Apr 2017 01:02:28 +0000 (01:02 +0000)]
SVM: calculate max ASID value and save for later use. This will be used in
an upcoming diff to handle ASID/VPID reuse/rollover.

7 years agoUpdate a man page example to reflect recent changes.
mlarkin [Fri, 14 Apr 2017 00:53:28 +0000 (00:53 +0000)]
Update a man page example to reflect recent changes.

Spotted by Ilya Kaliman, thanks.

7 years agoA little bit more trivial cleanup.
kettenis [Thu, 13 Apr 2017 23:29:02 +0000 (23:29 +0000)]
A little bit more trivial cleanup.

7 years agoXr sigprocmask(2) not the obsolete sigsetmask(3)
millert [Thu, 13 Apr 2017 21:49:11 +0000 (21:49 +0000)]
Xr sigprocmask(2) not the obsolete sigsetmask(3)

7 years agoUse the non-interrupt-safe pool allocator for the vp pool to avoid runninng
kettenis [Thu, 13 Apr 2017 20:48:29 +0000 (20:48 +0000)]
Use the non-interrupt-safe pool allocator for the vp pool to avoid runninng
out of kva in the kmem_map.  Avoids a hang when spawning a lot of processes.

7 years agoUse recallocarray in getdelim/getline to clear memory on buffer resizes,
brynet [Thu, 13 Apr 2017 18:36:51 +0000 (18:36 +0000)]
Use recallocarray in getdelim/getline to clear memory on buffer resizes,
inspired by a similar change to fgetln.

ok deraadt millert

7 years agoallow clearing less than allocated and document freezero(3) better
otto [Thu, 13 Apr 2017 18:32:55 +0000 (18:32 +0000)]
allow clearing less than allocated and document freezero(3) better

7 years agomoving some code into a switch meant that break no longer stopped the loop.
tedu [Thu, 13 Apr 2017 15:32:15 +0000 (15:32 +0000)]
moving some code into a switch meant that break no longer stopped the loop.
try harder with a goto. diagnosis and original fix by tb.

7 years agoRemove a bunch of obsolete, unused and irrelevant DHCP client types, fields,
krw [Thu, 13 Apr 2017 14:59:40 +0000 (14:59 +0000)]
Remove a bunch of obsolete, unused and irrelevant DHCP client types, fields,
and enums.

7 years agomulti-statement CHECK_RANGE() macro isn't safe for all placements, and
deraadt [Thu, 13 Apr 2017 14:48:31 +0000 (14:48 +0000)]
multi-statement CHECK_RANGE() macro isn't safe for all placements, and
needs to use "do {} while 0" idiom; all callers need repair also.
Discovered by jsg

7 years agofix wrongly indented lines
jsg [Thu, 13 Apr 2017 07:30:21 +0000 (07:30 +0000)]
fix wrongly indented lines

7 years agoAdd a NAT-T keepalive timer in case we are behind a NAT gateway.
patrick [Thu, 13 Apr 2017 07:04:09 +0000 (07:04 +0000)]
Add a NAT-T keepalive timer in case we are behind a NAT gateway.

See RFC 5996, section 2.23, NAT Traversal:
  In the case of a mismatching NAT_DETECTION_DESTINATION_IP hash, it
  means that the system receiving the NAT_DETECTION_DESTINATION_IP
  payload is behind a NAT and that system SHOULD start sending
  keepalive packets as defined in [UDPENCAPS].

With markus@, ok reyk@

7 years agoregen
guenther [Thu, 13 Apr 2017 04:09:46 +0000 (04:09 +0000)]
regen

7 years agoDelete the getlogin59 syscall, which was last used one year, two releases,
guenther [Thu, 13 Apr 2017 04:06:46 +0000 (04:06 +0000)]
Delete the getlogin59 syscall, which was last used one year, two releases,
and four libc major versions ago

ok sthen@ jsing@ deraadt@ jca@

7 years agoProvide mips64 with kernel-facing TCB_{GET,SET} macros that store it
guenther [Thu, 13 Apr 2017 03:52:25 +0000 (03:52 +0000)]
Provide mips64 with kernel-facing TCB_{GET,SET} macros that store it
in struct mdproc.  With that, all archs have those and the __HAVE_MD_TCB
macro can be unifdef'ed as always defined.

ok kettenis@ visa@ jsing@

7 years agoInspired by recent dhcrelay changes, bring in stricter CFLAGS!
krw [Wed, 12 Apr 2017 19:19:11 +0000 (19:19 +0000)]
Inspired by recent dhcrelay changes, bring in stricter CFLAGS!

7 years agoOne 'foo' per function is enough. Rename one to eliminate shadowing.
krw [Wed, 12 Apr 2017 19:17:30 +0000 (19:17 +0000)]
One 'foo' per function is enough. Rename one to eliminate shadowing.

7 years agoAdd missing function prototypes. Mark kw_cmp() as static like all
krw [Wed, 12 Apr 2017 19:12:01 +0000 (19:12 +0000)]
Add missing function prototypes. Mark kw_cmp() as static like all
its friends.

7 years agoThe kernel has to slightly different version of SipHash_Final but with
florian [Wed, 12 Apr 2017 18:35:50 +0000 (18:35 +0000)]
The kernel has to slightly different version of SipHash_Final but with
the same bug as just fixed in userland:

----------
SipHash_Final() was assuming the digest was 64-bit aligned, resulting in
misaligned memory accesses with armv7 ramdisk -Os bsd.rd ping
ok florian millert
----------

OK deraadt@

7 years agoUse calloc(3) and recallocarray(3) instead of *alloc* + memset(0).
tb [Wed, 12 Apr 2017 18:24:37 +0000 (18:24 +0000)]
Use calloc(3) and recallocarray(3) instead of *alloc* + memset(0).

ok deraadt

7 years agoSipHash_Final() was assuming the digest was 64-bit aligned, resulting in
deraadt [Wed, 12 Apr 2017 17:41:49 +0000 (17:41 +0000)]
SipHash_Final() was assuming the digest was 64-bit aligned, resulting in
misaligned memory accesses with armv7 ramdisk -Os bsd.rd ping
ok florian millert

7 years agoNew strstr() implementation from musl libc by Rich Felker. This
millert [Wed, 12 Apr 2017 16:06:12 +0000 (16:06 +0000)]
New strstr() implementation from musl libc by Rich Felker.  This
version uses the two-way string matching algorithm and is faster
than the old implementation.  With this change, ports that check
for strstr having linear complexity time strstr will no longer
replace the libc strstr with a private version.
OK deraadt@ espie@

7 years agoPrevent inosused from wrapping when soft updates is enabled while
millert [Wed, 12 Apr 2017 15:23:08 +0000 (15:23 +0000)]
Prevent inosused from wrapping when soft updates is enabled while
scanning the used inode map.  The code as written assumes inosused
is signed but this is no longer the case.  OK deraadt@