ratchov [Thu, 30 Jun 2016 21:37:29 +0000 (21:37 +0000)]
Use CLOCK_UPTIME instead of CLOCK_MONOTONIC, as the later makes jumps
during suspend/resume cycles which triggers watchdog time-outs and
in turn prevents sndiod from resuming.
afresh1 [Thu, 30 Jun 2016 21:16:13 +0000 (21:16 +0000)]
Update perl Time::HiRes to 1.9739
Which provides hires `utime`
requested by espie@ OK millert@
mglocker [Thu, 30 Jun 2016 19:54:13 +0000 (19:54 +0000)]
Also mention the term 'USB device' in the title to make clear that the
reference count is on a device level.
ok jmc
gilles [Thu, 30 Jun 2016 18:41:39 +0000 (18:41 +0000)]
update aliases documentation to reflect reality
tedu [Thu, 30 Jun 2016 15:59:30 +0000 (15:59 +0000)]
don't need the .xr to sysctl now either
tedu [Thu, 30 Jun 2016 15:58:06 +0000 (15:58 +0000)]
nptys sysctls were removed
tedu [Thu, 30 Jun 2016 15:56:59 +0000 (15:56 +0000)]
ntpys sysctl was removed
tedu [Thu, 30 Jun 2016 15:54:53 +0000 (15:54 +0000)]
992 ptys is the hard max.
mpi [Thu, 30 Jun 2016 12:36:27 +0000 (12:36 +0000)]
Use ``rt_addr'' rather than ``rt_ifa'' to get the source address
corresponding to a route.
ok florian@ on a previous version, input and ok bluhm@
bcook [Thu, 30 Jun 2016 12:19:51 +0000 (12:19 +0000)]
Tighten behavior of _rs_allocate failure for portable arc4random implementations.
In the event of a failure in _rs_allocate for rsx, we still have a reference to
freed memory for rs on return. Not a huge deal since we subsequently abort in
_rs_init, but it looks strange on its own.
ok deraadt@
bcook [Thu, 30 Jun 2016 12:17:29 +0000 (12:17 +0000)]
Tighten behavior of _rs_allocate on Windows.
For Windows, we are simply using calloc, which has two annoyances:
the memory has more permissions than needed by default, and it comes
from the process heap, which looks like a memory leak since this memory
is rightfully never freed.
This switches _rs_alloc on Windows to use VirtualAlloc, which restricts the
memory to READ|WRITE and keeps the memory out of the process heap.
ok deraadt@
bcook [Thu, 30 Jun 2016 11:10:29 +0000 (11:10 +0000)]
bump to 2.4.2
florian [Thu, 30 Jun 2016 10:17:18 +0000 (10:17 +0000)]
Prevent NULL deref, pointed out by llvm's scan-build.
OK jca, millert
otto [Thu, 30 Jun 2016 09:00:48 +0000 (09:00 +0000)]
adapt S option: add C, rm F (not relevant with 0 cache and disables
chunk rnd), rm P: is default
mpi [Thu, 30 Jun 2016 08:19:03 +0000 (08:19 +0000)]
Restore the automagically added /64 route on p2p interfaces in order to
send traffic to link-local addresses without default route.
Fix a regression reported by Michael Lechtermann, ok stsp@, sthen@
dtucker [Thu, 30 Jun 2016 05:17:05 +0000 (05:17 +0000)]
Explicitly check for 100% completion to avoid potential floating point
rounding error, which could cause progressmeter to report 99% on completion.
While there invert the test so the 100% case is clearer. with & ok djm@
mlarkin [Thu, 30 Jun 2016 02:29:22 +0000 (02:29 +0000)]
remove some unused variables (that were commented out anyway)
bcook [Thu, 30 Jun 2016 02:02:06 +0000 (02:02 +0000)]
Remove flags for disabling constant-time operations.
This removes support for DSA_FLAG_NO_EXP_CONSTTIME, DH_FLAG_NO_EXP_CONSTTIME,
and RSA_FLAG_NO_CONSTTIME flags, making all of these operations unconditionally
constant-time.
Based on the original patch by César Pereid. ok beck@
tb [Wed, 29 Jun 2016 20:38:39 +0000 (20:38 +0000)]
If /tmp/vi.recover doesn't exist, don't create it. Warn once
that it doesn't exist, afterwards fail silently.
ok millert
chris [Wed, 29 Jun 2016 17:54:08 +0000 (17:54 +0000)]
Document arptimeout, arpdown. Prodded by and ok jmc@
jmc [Wed, 29 Jun 2016 17:14:28 +0000 (17:14 +0000)]
sort the -o list;
jca [Wed, 29 Jun 2016 14:19:38 +0000 (14:19 +0000)]
Spring cleanup
- pointless casts, kill caddr_t or replace it with char *
- signed counters
- simplify if_getmtu, only one method is needed and SIOCGIFMTU is the
cheapest
- we no longer have drivers for IFT_FDDI
- hide details of iflist management
- if (dflag) log_debug -> log_debug
- dead code and comments
- etc etc
Input from and ok florian@
visa [Wed, 29 Jun 2016 13:39:03 +0000 (13:39 +0000)]
Make the fdt parser skip nop tokens. This is needed on some
octeon boards.
Feedback and ok kettenis@
eric [Wed, 29 Jun 2016 06:46:06 +0000 (06:46 +0000)]
Explicitely enclose SMTP transactions between BEGIN and COMMIT/ROLLBACK
filter events. Bump filter API version.
ok gilles@ jung@
mlarkin [Wed, 29 Jun 2016 06:05:15 +0000 (06:05 +0000)]
Don't write a 1 to the RIRR bit in the IOAPIC redirection register. This bit
is R/O, and although it should not matter what value is written there,
Hyper-V's emulated IOAPIC interprets a write of 1 in some unexpected way and
subsequently blocks interrupt delivery. This primarily manifests itself as
de(4) timeouts when using Hyper-V VMs with the "Legacy Network Adapter"
interface.
This diff has been in snaps for almost a month with no reported fallout.
Based on an idea originally from mikeb with further input from kettenis and
deraadt.
jca [Tue, 28 Jun 2016 19:45:53 +0000 (19:45 +0000)]
Add missing words to help the reader.
ok mikeb@
jca [Tue, 28 Jun 2016 18:22:50 +0000 (18:22 +0000)]
whitespace nit
tedu [Tue, 28 Jun 2016 18:00:59 +0000 (18:00 +0000)]
revert previous. better fix applied to fts_open.
espie [Tue, 28 Jun 2016 17:35:34 +0000 (17:35 +0000)]
oops, $pkg can be more than a pkgname. We have the right information
in the plist, so use it instead.
jca [Tue, 28 Jun 2016 17:35:14 +0000 (17:35 +0000)]
Add -M and -m options to specify the outgoing and incoming minimum TTL
Req by and ok blumh@
jca [Tue, 28 Jun 2016 17:32:58 +0000 (17:32 +0000)]
IP_MINTTL and IPV6_MINHOPCOUNT work on TCP and UDP sockets
Input from and ok bluhm@
millert [Tue, 28 Jun 2016 17:25:08 +0000 (17:25 +0000)]
For the EINVAL case there can be more than a single option.
millert [Tue, 28 Jun 2016 17:21:48 +0000 (17:21 +0000)]
fts_open() requires that the list passed as argument to contain at least
one path. When the list is empty (contain only a NULL pointer), return
EINVAL instead of pretending to succeed, which will cause a NULL pointer
deference in a later fts_read() call. From FreeBSD.
chris [Tue, 28 Jun 2016 17:18:24 +0000 (17:18 +0000)]
Add sysctl for arp timers: net.inet.ip.arptimeout (expire timer for resolved
entries) and net.inet.ip.arpdown (expire timer for unresolved entries)
ok mpi@
millert [Tue, 28 Jun 2016 17:12:29 +0000 (17:12 +0000)]
Do not return an error in fts_open(3) if one of the paths in argv
is empty. Otherwise, programs using fts(3) will report an error
if one of the paths is empty instead of just treating it as a
non-existent file. OK guenther@
jca [Tue, 28 Jun 2016 16:59:14 +0000 (16:59 +0000)]
Use IPV6_MINHOPCOUNT to finish implementing ttl-security for IPv6.
ok renato@ deraadt@
espie [Tue, 28 Jun 2016 15:38:36 +0000 (15:38 +0000)]
option -z that uses is-branch info to produce "complete"
stem--[flavor][%branch] listing.
feedback and suggestions semarie@
espie [Tue, 28 Jun 2016 15:30:29 +0000 (15:30 +0000)]
document is-branch
tedu [Tue, 28 Jun 2016 15:20:13 +0000 (15:20 +0000)]
don't report errors for 'rm -rf ""'. report by rkitover.
ok benno deraadt
tedu [Tue, 28 Jun 2016 14:47:00 +0000 (14:47 +0000)]
introduce rwlock for socketbuf instead of the old flag and tsleep dance.
ok mikeb bluhm
jca [Tue, 28 Jun 2016 11:23:57 +0000 (11:23 +0000)]
IPV6_MINHOPCOUNT now supports UDP (unicast and multicast)
jca [Tue, 28 Jun 2016 11:22:53 +0000 (11:22 +0000)]
Add UDP unicast and multicast support for IP_MINTTL/IPV6_MINHOPCOUNT
Requested by renato@, ok blumh@
jmc [Tue, 28 Jun 2016 07:17:59 +0000 (07:17 +0000)]
fix a couple of errors in the page;
from henning petersen, netbsd pr lib/51284
ok semarie
tb [Tue, 28 Jun 2016 06:40:11 +0000 (06:40 +0000)]
Back out previous; otto saw a potential race that could lead to a
double unmap and I experienced a much more unstable firefox.
discussed with otto on icb
jmc [Tue, 28 Jun 2016 06:31:30 +0000 (06:31 +0000)]
more standard Nd, and zap an unbalanced ";
ok jmatthew
mlarkin [Tue, 28 Jun 2016 05:50:55 +0000 (05:50 +0000)]
Don't panic on null vmxon region or vmxon failure - this will be handled
in the calling function which will return EIO back up the stack to vmd(8).
Fixes a panic when CPUs fail to spin up for other reasons during boot,
noticed by reyk.
mlarkin [Tue, 28 Jun 2016 05:37:50 +0000 (05:37 +0000)]
Enable UMIP feature if present. miod originally pointed this out to me.
ok deraadt@
jmatthew [Tue, 28 Jun 2016 04:41:37 +0000 (04:41 +0000)]
enable ure(4) where we already have url(4)
afresh1 [Tue, 28 Jun 2016 04:30:02 +0000 (04:30 +0000)]
Stop linking perl to libpthread
No longer necessary with TIB work
from guenther@ ok sthen@
afresh1 [Tue, 28 Jun 2016 04:28:18 +0000 (04:28 +0000)]
Stop linking perl to libutil, never necessary
from guenther@ ok sthen@
semarie [Tue, 28 Jun 2016 04:27:58 +0000 (04:27 +0000)]
in pledgereq (the array for doing correspondance between textual promise as in pledge(2) and internal PLEDGE_* flag), the flags member should be a uint64_t and not a int.
ok deraadt@
deraadt [Tue, 28 Jun 2016 01:45:28 +0000 (01:45 +0000)]
sync
deraadt [Tue, 28 Jun 2016 00:01:10 +0000 (00:01 +0000)]
If an error path if close() is called, save errno so that original error
is shown by errx
ok millert krw
deraadt [Mon, 27 Jun 2016 23:58:08 +0000 (23:58 +0000)]
Be more careful initializing and tracking socket s through main, this is
so complicated that a future refactoring could easily in introduce a bug.
ok millert krw
deraadt [Mon, 27 Jun 2016 23:54:25 +0000 (23:54 +0000)]
Improve .Nd, and Xr from usb(4)
jmatthew [Mon, 27 Jun 2016 23:50:31 +0000 (23:50 +0000)]
add ure(4) man page
jmatthew [Mon, 27 Jun 2016 23:38:01 +0000 (23:38 +0000)]
the flag is USBD_EXCLUSIVE_USE, not USBD_EXCLUSIVE_ACCESS
jca [Mon, 27 Jun 2016 20:57:41 +0000 (20:57 +0000)]
Missing "break;" in switch statement; repairs IP_MINTTL.
jca [Mon, 27 Jun 2016 19:55:02 +0000 (19:55 +0000)]
Repair kill(2) on zombie processes.
kill(2) is supposed to find zombie processes, this probably got broken
when the process reaper was introduced. As a temporary(tm) workaround,
walk the list of zombie processes if we can't find the target pid in the
main process list.
Problem with zsh initially reported by Geoff Wozniak on misc@, analysis
done by naddy@. ok kettenis@ tedu@
renato [Mon, 27 Jun 2016 19:18:54 +0000 (19:18 +0000)]
Fix inverted logic in recv_hello().
Bug introduced by rev1.48 two weeks ago. We were not respecting the
advertised transport connection preference (LDPoIPv4 or LDPoIPv6),
the fix is pretty obvious.
renato [Mon, 27 Jun 2016 19:08:39 +0000 (19:08 +0000)]
Remove superfluous call to l2vpn_pw_exit().
renato [Mon, 27 Jun 2016 19:06:33 +0000 (19:06 +0000)]
Handle ibuf_add() errors.
tweaks from claudio@
stsp [Mon, 27 Jun 2016 19:01:02 +0000 (19:01 +0000)]
Revert if_iwm.c r1.166. We are still seeing links dropping upon HT protection
updates with some iwn chips, such as the 2200. Reported by mlarkin and krw.
jmc [Mon, 27 Jun 2016 17:36:33 +0000 (17:36 +0000)]
minor tweaks; ok tedu
espie [Mon, 27 Jun 2016 16:52:30 +0000 (16:52 +0000)]
make proper use of fstatat, as suggested by guenther@
okay millert@
jsing [Mon, 27 Jun 2016 16:52:01 +0000 (16:52 +0000)]
Regen.
jsing [Mon, 27 Jun 2016 16:50:07 +0000 (16:50 +0000)]
Restore the sys_o58_kill system call.
jsing [Mon, 27 Jun 2016 16:49:45 +0000 (16:49 +0000)]
Restore the sys_o58_kill system call.
By keeping both the new (sys_kill/sys_thrkill) and old (sys_o58_kill)
system calls for the OpenBSD 6.0 release, code that uses either of these
mechanisms will work on both of the supported OpenBSD releases. This
provides a clean transition for runtimes that make direct system calls
(namely the Go programming language).
This requires a minimal amount of non-intrusive code and does not block
development progress within OpenBSD.
ok deraadt@ guenther@
jca [Mon, 27 Jun 2016 16:36:08 +0000 (16:36 +0000)]
Kill outdated comment. Reading the code is enough.
ok sthen@ bluhm@
jca [Mon, 27 Jun 2016 16:33:48 +0000 (16:33 +0000)]
Implement IPV6_MINHOPCOUNT support.
Useful to implement GTSM support in daemons such as bgpd(8). Diff from
2013 revived by renato@. Input from bluhm@, ok bluhm@ deraadt@
bluhm [Mon, 27 Jun 2016 15:59:51 +0000 (15:59 +0000)]
Copy inp_hops from the listening socket to the accepted one and use
its value for the SYN+ACK packet. This makes the IPV6_UNICAST_HOPS
socket option usable for incoming TCP connections.
tested by renato@; OK jca@
tedu [Mon, 27 Jun 2016 15:47:38 +0000 (15:47 +0000)]
somehow nopass snuck onto the :wheel example. i think it's better without.
tedu [Mon, 27 Jun 2016 15:41:17 +0000 (15:41 +0000)]
revise environment handling.
Add a setenv keyword for manipulating the environment. keepenv now means
only retain everything. (for one release, the old use of keepenv will still
work.)
Allow setting variables to new or existing values, and also removing vars
when keepenv is used.
ok djm martijn tb
tedu [Mon, 27 Jun 2016 15:33:40 +0000 (15:33 +0000)]
defer munmap to after unlocking malloc. this can (unfortunately) be an
expensive syscall, and we don't want to tie up other threads. there's no
need to hold the lock, so defer it to afterwards.
from Michael McConville
ok deraadt
deraadt [Mon, 27 Jun 2016 14:43:04 +0000 (14:43 +0000)]
whitespace
deraadt [Mon, 27 Jun 2016 14:26:26 +0000 (14:26 +0000)]
sync
jsg [Mon, 27 Jun 2016 13:06:41 +0000 (13:06 +0000)]
Fix a memory leak in an error path.
ok mikeb@
jsg [Mon, 27 Jun 2016 12:32:39 +0000 (12:32 +0000)]
Correct the order of memset arguments.
ok mikeb@
bluhm [Mon, 27 Jun 2016 12:25:27 +0000 (12:25 +0000)]
The variable swapping between inp, newinp and oldinpcb in syn_cache_get()
was overly complicated. Simplify the code without functional change.
OK jca@
jmatthew [Mon, 27 Jun 2016 11:42:47 +0000 (11:42 +0000)]
add ure(4), a driver for Realtek RTL8152 10/100 USB Ethernet adapters,
ported from FreeBSD.
ok dereaadt@
espie [Mon, 27 Jun 2016 08:38:15 +0000 (08:38 +0000)]
surprised nobody noticed before me
let scp:// work with PKG_CACHE
basically, we do an extra fork, so we end up cleaning up the connection twice
that's okay for the handles, but not for the commands, since we do a spurious
ABORT BYE BYE on cleanup.
espie [Mon, 27 Jun 2016 06:10:04 +0000 (06:10 +0000)]
make the fallback code more accurate (in particular, it should return names
based on the actual device, not any kind of inode equality which won't hold
for duplicates of the dev tree in a chroot)
no bump as it doesn't change any API.
okay and improvements guenther@
semarie [Mon, 27 Jun 2016 04:26:41 +0000 (04:26 +0000)]
dovutimens: call vrele(9) before returning EINVAL
ok guenther@
semarie [Mon, 27 Jun 2016 04:14:38 +0000 (04:14 +0000)]
sys_revoke: call vrele() before returning ENOTTY
ok guenther@
tedu [Sun, 26 Jun 2016 21:04:08 +0000 (21:04 +0000)]
increase the minimum for auto rounds to 6. that was the previous low bound
for login.conf, and we don't want to go lower.
tedu [Sun, 26 Jun 2016 20:32:31 +0000 (20:32 +0000)]
upgrade selected login.conf to use auto rounds for bcrypt. the installer
already does this, so we don't want to go backwards on password changes.
ok krw
tedu [Sun, 26 Jun 2016 19:53:40 +0000 (19:53 +0000)]
somebody forgot to look at format string warnings
tb [Sun, 26 Jun 2016 15:17:43 +0000 (15:17 +0000)]
Document how to create full boot and installation images containing
release tarballs and ports, suitable for installs without network.
Idea and patch from Bryan Everly, thanks!
With input from deraadt, jmc and myself.
ok jmc
semarie [Sun, 26 Jun 2016 14:27:14 +0000 (14:27 +0000)]
use error code path instead of return early without calling VOP_ABORTOP() and
vrele()/vput().
ok deraadt@
jsg [Sun, 26 Jun 2016 10:12:29 +0000 (10:12 +0000)]
sxie(4) is only for integrated 'EMAC' 10/100 Ethernet. Make this a bit
more obvious. The Gigabit 'GMAC' controller used by some A20 systems
is not supported.
jsg [Sun, 26 Jun 2016 10:04:18 +0000 (10:04 +0000)]
remove devices from soc abstraction device lists that now attach to fdt
jsg [Sun, 26 Jun 2016 09:06:35 +0000 (09:06 +0000)]
Dynamically attach cpsw(4) with the FDT.
jsg [Sun, 26 Jun 2016 07:25:05 +0000 (07:25 +0000)]
Dynamically attach tiiic using the FDT.
Only match on omap4 compatible controllers such as the one in the
am335x on bbb. omap3 has different register offsets and a different
compatible string.
This reuses kettenis' fdt i2c scan function from imxiic(4).
mglocker [Sun, 26 Jun 2016 06:50:26 +0000 (06:50 +0000)]
Oups, remove '#define I2S_DEBUG' again.
mglocker [Sun, 26 Jun 2016 06:48:39 +0000 (06:48 +0000)]
Make i2s compile with I2S_DEBUG; %x -> %p.
jsg [Sun, 26 Jun 2016 05:30:21 +0000 (05:30 +0000)]
add a missing argument specified in format string so this builds again
jsg [Sun, 26 Jun 2016 05:27:44 +0000 (05:27 +0000)]
destatic
jsg [Sun, 26 Jun 2016 05:16:33 +0000 (05:16 +0000)]
Dynamically attach omdog(4) using the FDT.
jsg [Sun, 26 Jun 2016 01:56:07 +0000 (01:56 +0000)]
Don't list cubieboard2, pmap can't handle allwinner a20.
espie [Sat, 25 Jun 2016 18:02:59 +0000 (18:02 +0000)]
recognize @option is-branch
so that it trickles to snapshots and then can be used in ports