eric [Tue, 14 May 2019 12:08:54 +0000 (12:08 +0000)]
fix typos in log messages
nicm [Tue, 14 May 2019 07:37:50 +0000 (07:37 +0000)]
Fix sizing of main-vertical and main-horizontal layouts, GitHub issue 1736.
anton [Tue, 14 May 2019 06:05:45 +0000 (06:05 +0000)]
Add support for `boot device' to vm.conf grammar which is the `-B device'
counterpart from vmctl.
ok mlarkin@
jca [Tue, 14 May 2019 05:58:36 +0000 (05:58 +0000)]
Enable TLSv1.0 and TLSv1.1
Like the old ftp. Discussed with at least with tedu@ "We should match
existing behavior at least for now"
bluhm [Tue, 14 May 2019 05:48:58 +0000 (05:48 +0000)]
Use propper regress target to integrate better into test framework.
visa [Tue, 14 May 2019 03:27:43 +0000 (03:27 +0000)]
Add lock order checking for smr_barrier(9). This is similar to the
checking done in taskq_barrier(9) and timeout_barrier(9).
OK mpi@
sunil [Tue, 14 May 2019 03:20:11 +0000 (03:20 +0000)]
Remove leftover elf.c
sunil [Tue, 14 May 2019 03:16:55 +0000 (03:16 +0000)]
Use elf(3) api instead of an ad-hoc elf parser.
Ok mpi@
sunil [Tue, 14 May 2019 02:32:08 +0000 (02:32 +0000)]
Remove an unused and leftover label.
sunil [Tue, 14 May 2019 02:30:00 +0000 (02:30 +0000)]
tls_write(3): Handle short writes.
Ok jca@
naddy [Mon, 13 May 2019 23:18:21 +0000 (23:18 +0000)]
Add tee(1) to the ramdisk filesystem and use it to provide both a
moving progress bar during auto upgrade/install and a clean log
afterwards. ok deraadt@
florian [Mon, 13 May 2019 23:13:24 +0000 (23:13 +0000)]
Remove strict mode for now, it is in the way.
tedu [Mon, 13 May 2019 22:58:52 +0000 (22:58 +0000)]
print a newline for attach. from f holop
beck [Mon, 13 May 2019 22:55:27 +0000 (22:55 +0000)]
Add a kernel implementation of realpath() as __realpath().
We want this so that we can stop allowing readlink() on traversed
vnodes in unveil().
This includes all the kernel side and the system call.
This is not yet used in libc for realpath, so nothing calls this yet.
The libc wrapper will be committed later.
Testing by many, and ports build by naddy@
ok deraadt@
bcook [Mon, 13 May 2019 22:48:30 +0000 (22:48 +0000)]
Remove unused pad check, which is handled by tls1_cbc_remove_padding() now.
Fixes COV-174858
ok tb@
bcook [Mon, 13 May 2019 22:36:01 +0000 (22:36 +0000)]
Acquire mutex before incrementing the refcount. Fixes COV-186144
ok tb@
tb [Mon, 13 May 2019 22:25:27 +0000 (22:25 +0000)]
Add missing word in comment.
jsg [Mon, 13 May 2019 21:50:50 +0000 (21:50 +0000)]
enable in_irq() based check
beck [Mon, 13 May 2019 21:47:09 +0000 (21:47 +0000)]
Document our currently incosnsistent behaviour and our current
sins which have been judged by Dr Posix as sinful...
jsg [Mon, 13 May 2019 21:44:13 +0000 (21:44 +0000)]
add in_irq() in_interrupt() in_task()
in_irq() uses ci_idepth on archs that have it
in_interrupt() is the same as in_irq() for now, linux has in_irq() for
hard interrupt context and in_interrupt() is also for soft interrupt
and nmi context.
mpi [Mon, 13 May 2019 21:29:28 +0000 (21:29 +0000)]
free(9) sizes.
From Jan Klemkow, ok tedu@
mpi [Mon, 13 May 2019 21:27:59 +0000 (21:27 +0000)]
free(9) sizes.
From miod@
claudio [Mon, 13 May 2019 21:13:04 +0000 (21:13 +0000)]
Don't use a hardcoded list of how attribute are dumped and instead move
to a loop-switch construct in up_generate_attr(). This way attributes
are always dumped in ascending order as suggested by the RFC and adding
special attributes is simpler than in the current way.
The MP attributes are a special case since those are handled at a later
stage of creating UPDATE messages.
OK benno@
drahn [Mon, 13 May 2019 20:55:22 +0000 (20:55 +0000)]
Move idepth tracking out of interrupt controller into common layer.
'looks good' kettenis@
claudio [Mon, 13 May 2019 20:47:19 +0000 (20:47 +0000)]
KERN_PFSTATUS is not handled by sysctl so warn about it if requested but
ignore if sysctl -a is used.
OK bluhm@
nicm [Mon, 13 May 2019 20:10:23 +0000 (20:10 +0000)]
Add support for overline (SGR 53), from Ricardo Banffy.
nicm [Mon, 13 May 2019 20:07:02 +0000 (20:07 +0000)]
Always include Lock in the menu.
guenther [Mon, 13 May 2019 20:00:32 +0000 (20:00 +0000)]
Move 'how this works' details from namespace.h to DETAILS
bluhm [Mon, 13 May 2019 19:42:28 +0000 (19:42 +0000)]
Link regress sigpthread to build.
bluhm [Mon, 13 May 2019 19:40:22 +0000 (19:40 +0000)]
Test the interaction of signals with multiple posix threads. It
covers blocking with signal mask, killing process or thread, invoking
handler or waiting for signal.
bluhm [Mon, 13 May 2019 19:21:31 +0000 (19:21 +0000)]
When killing a process, the signal is handled by any thread that
does not block the signal. If all threads block the signal, we
delivered it to the main thread. This does not conform to POSIX.
If any thread unblocks the signal, it should be delivered immediately
to this thread.
Mark such signals pending at the process instead of a single thread.
Then any thread can handle it later.
OK kettenis@ guenther@
mpi [Mon, 13 May 2019 18:20:13 +0000 (18:20 +0000)]
Do not check for IFF_RUNNING inside bstp_initialization().
This allows to set such flag after completing the initialization of a
bridge and still have bstp_tick() be scheduled from the begining.
Fix a regression reported by and ok markus@
mpi [Mon, 13 May 2019 18:14:05 +0000 (18:14 +0000)]
Deal with the case where bridge_getbif() can return NULL.
Since `bif' are removed from the interface list before calling smr_barrier()
and the hash queue is cleaned up afterward, it is possible to find an ifidx
with bridge_rtlookup() that won't match to any `bif'.
Fix a panic reported by Hrvoje Popovski, ok visa@
deraadt [Mon, 13 May 2019 17:31:51 +0000 (17:31 +0000)]
dup2(n,n) would rlimit check before handling the n==n shortcut,
and incorrectly return EBADF when n>curlim.
ok millert guenther tedu
guenther [Mon, 13 May 2019 17:18:10 +0000 (17:18 +0000)]
The fd used by nlist() isn't application visible, so mark it close-on-exec
to avoid leaking it
ok deraadt@
schwarze [Mon, 13 May 2019 17:00:12 +0000 (17:00 +0000)]
Delete tentacles of LC_NUMERIC support.
Our libc doesn't support that now and won't in the future.
Minus ~90 lines of code, no functional change.
OK tb@
jsg [Mon, 13 May 2019 16:23:15 +0000 (16:23 +0000)]
add idr_is_empty()
tb [Mon, 13 May 2019 16:04:49 +0000 (16:04 +0000)]
ephermal -> ephemeral
From Hiltjo Posthuma
pd [Mon, 13 May 2019 15:40:34 +0000 (15:40 +0000)]
vmm: add host side pvclock
Emulate kvm pvclock in vmm(4). Compatible with pvclock(4) in OpenBSD. Linux
does not attach to this (yet).
Fixes by reyk@ and tested extensively by reyk@, tb@ and phessler@
ok mlarkin@ phessler@ reyk@
reyk [Mon, 13 May 2019 15:19:16 +0000 (15:19 +0000)]
Add Connection: close when switching to "unlimited" reading mode.
Ask the server to close the connection after the request since we
don't read any further request headers. This fixes an issue with
OPTIONS and optional body, as well as similar cases.
Reported and tested by Rivo Nurges
OK benno@
schwarze [Mon, 13 May 2019 14:11:20 +0000 (14:11 +0000)]
explicitly mention that RES_NOALIASES has no effect;
jmc@ noticed that the text wasn't completely clear;
OK jmc@
denis [Mon, 13 May 2019 13:47:36 +0000 (13:47 +0000)]
fix export default-route.
OK claudio@
abieber [Mon, 13 May 2019 12:31:56 +0000 (12:31 +0000)]
regen (missed this when I did the colemak update)
rob [Mon, 13 May 2019 10:00:29 +0000 (10:00 +0000)]
Hookup ber regress.
ok bluhm@
reyk [Mon, 13 May 2019 09:54:07 +0000 (09:54 +0000)]
Fix filter rules with "forward to" statement in persistent connections.
OK bentley@ mikeb@
nicm [Mon, 13 May 2019 08:56:07 +0000 (08:56 +0000)]
Fix column width for copy mode commands.
martijn [Mon, 13 May 2019 07:33:23 +0000 (07:33 +0000)]
According to RFC3414 section 4 applications should be able to discover the
snmpEngineBoots and snmpEngineTime by sending an AuthPriv request with the
requested values set to zero and with a valid user.
Move the engine_boots and engine_time down after the user check and remove
the 0-check, so we can reply with the appropriate usmStatsNotInTimeWindows.
This allows us to use p5-Net-SNMP against snmpd with seclevel enc.
OK rob@
martijn [Mon, 13 May 2019 07:24:50 +0000 (07:24 +0000)]
According to RFC3414 section 4 applications should be able to discover the
snmpEngineID by sending a noAuthNoPriv request. Move the seclevel check to
after the usm_decode phase, so we can reply with the mandatory
usmStatsUnknownEngineIDs instead of usmStatsUnsupportedSecLevels.
This brings us one step closer to using p5-Net-SNMP with seclevel enc.
OK tb@, rob@
otto [Mon, 13 May 2019 06:04:55 +0000 (06:04 +0000)]
Mention introduction of *_conceal.
patrick [Mon, 13 May 2019 03:13:29 +0000 (03:13 +0000)]
Add the needed ICC_PMR_EL1 register bit defines for the previous
commit to unbreak the build.
from kettenis@
ok drahn@
bluhm [Mon, 13 May 2019 02:54:54 +0000 (02:54 +0000)]
The call to fseek(fp, -1, SEEK_END) also sets the reported size to
this value. To match the expectation of the test again, move this
line before the the code that sets the final position.
OK yasuoka@
ajacoutot [Mon, 13 May 2019 02:26:42 +0000 (02:26 +0000)]
Fix usage() (missing '-x').
Adapt to recent sysupgrade(8) change
ok florian@ kmos@ kn@
ajacoutot [Mon, 13 May 2019 02:25:16 +0000 (02:25 +0000)]
Use the same heuristic as the installer to find a proper prefetch area for the
sets instead of hardcoding /home. This leads the way to a knob for manually
choosing a sets directory if we want that.
Create /auto_upgrade.conf that will get picked up by the installer for the
unattended upgrade mode.
Similar inputs from naddy@ and kn@
ok florian@ kmos@ kn@
guenther [Mon, 13 May 2019 00:07:05 +0000 (00:07 +0000)]
Add support for -znoretpolineplt and make -zretpolineplt the default.
Affects i386 and amd64 only.
ok deraadt@ kettenis@
guenther [Sun, 12 May 2019 23:32:42 +0000 (23:32 +0000)]
Indentation fix
tb [Sun, 12 May 2019 22:48:03 +0000 (22:48 +0000)]
Fix double free by nulling out pointers after free.
from florian, ok jca
guenther [Sun, 12 May 2019 22:23:38 +0000 (22:23 +0000)]
Delete cpu_idle_{enter,leave}_fcn() as unused. Add RETGUARD checks to
cpu_idle_cycle()
ok mpi@ kettenis@
rob [Sun, 12 May 2019 22:17:25 +0000 (22:17 +0000)]
Add an enum test. Modify some output.
rob [Sun, 12 May 2019 22:12:16 +0000 (22:12 +0000)]
Use .OBJDIR.
kmos [Sun, 12 May 2019 21:56:27 +0000 (21:56 +0000)]
Make the "xx bytes received in y time" message go to the right place
by using the login_info() function the other messages use.
OK florian@
espie [Sun, 12 May 2019 21:49:52 +0000 (21:49 +0000)]
zap confusing unneeded parameters
okay jca@, jasper@
jca [Sun, 12 May 2019 21:30:48 +0000 (21:30 +0000)]
Repair ftp -o - and thus pkg_add: print informational messages on stderr
ok florian@ espie@
guenther [Sun, 12 May 2019 21:27:47 +0000 (21:27 +0000)]
s/availible/available/
florian [Sun, 12 May 2019 21:03:17 +0000 (21:03 +0000)]
It will show up in 6.6.
tedu [Sun, 12 May 2019 21:01:46 +0000 (21:01 +0000)]
remove .PATH. this was used in previous eras to build bpf compiler stuff,
that's all used via libpcap now.
found by deraadt ok tb
jasper [Sun, 12 May 2019 20:58:19 +0000 (20:58 +0000)]
add rcs ids
pd [Sun, 12 May 2019 20:56:34 +0000 (20:56 +0000)]
vmm: add a x86 page table walker
Add a first cut of x86 page table walker to vmd(8) and vmm(4). This function is
not used right now but is a building block for future features like HPET, OUTSB
and INSB emulation, nested virtualisation support, etc.
With help from Mike Larkin
ok mlarkin@
rob [Sun, 12 May 2019 20:56:11 +0000 (20:56 +0000)]
Remove a now obsolete comment about BITSTRING.
rob [Sun, 12 May 2019 20:55:09 +0000 (20:55 +0000)]
Add test cases for enum, and an additional test for null.
kmos [Sun, 12 May 2019 20:44:39 +0000 (20:44 +0000)]
Move us from old ftp(1) to Sunil's new ftp(1). The necessary modifications
have been made to make it behave. Any new misbehaviors can be fixed in tree.
OK florian@ deraadt@ "Have you committed ftp yet?"
rob [Sun, 12 May 2019 20:25:10 +0000 (20:25 +0000)]
Enable support for the writting of BITSTRING by treating it as an OCTETSTRING.
ok claudio@
rob [Sun, 12 May 2019 20:19:37 +0000 (20:19 +0000)]
Test mandatory use of minimal contents octets for int and enum.
rob [Sun, 12 May 2019 20:13:08 +0000 (20:13 +0000)]
Enforce smallest number of contents octets for int (and enum).
ok claudio@
jsg [Sun, 12 May 2019 20:12:25 +0000 (20:12 +0000)]
remove unused STUB definition
mpi [Sun, 12 May 2019 19:53:22 +0000 (19:53 +0000)]
Switch the list of span interfaces and interfaces to SMR.
This removes the KERNEL_LOCK() around the list iteration in bridge_enqueue().
Since the NET_LOCK() isn't protecting any data structure, release it early
in all the code paths coming from the Network Stack to prevent possible
deadlock situations with smr_barrier().
bridge_input() is still KERNEL_LOCK()ed as well as bridge_filterrule().
ok visa@
anton [Sun, 12 May 2019 19:43:34 +0000 (19:43 +0000)]
no need to store the wmesg passed to rwsleep() as a static variable anymore
rob [Sun, 12 May 2019 19:29:41 +0000 (19:29 +0000)]
Mention #include of <sys/types.h> in synopsis.
Discussed with claudio@ and tedu@.
tim [Sun, 12 May 2019 18:52:43 +0000 (18:52 +0000)]
Stop generating keys for nsd-control(8). These are unused since nsd was
switched to using a local control socket by default.
OK florian@ sthen@
nicm [Sun, 12 May 2019 18:18:30 +0000 (18:18 +0000)]
Oops, removed too much in last change.
nicm [Sun, 12 May 2019 18:16:33 +0000 (18:16 +0000)]
Fix some indentation and dead assignments.
stsp [Sun, 12 May 2019 18:12:37 +0000 (18:12 +0000)]
Fix 'ifconfig nwflags; These flags ended up overlapping with other flags
in ieee80211com's ic_flags because we haven't been paying attention to
them (they're not in the same place in the code and hence easy to miss).
Move them to a dedicated variable to avoid this problem in the future.
Add a new 'stayauth' nwflag which can be set to let net80211 ignore
deauth frames. This can be useful when deauth frames are being
persistently spoofed by an attacker. Idea from beck@
ok beck@ phessler@
rob [Sun, 12 May 2019 18:11:51 +0000 (18:11 +0000)]
Fail early if a (universal) primitive type identifies as constructed, or if a
boolean has a contents length other than 1.
ok claudio@
rob [Sun, 12 May 2019 17:50:32 +0000 (17:50 +0000)]
Enforce minimal number of octets for tag > 30.
"sure" claudio@
bluhm [Sun, 12 May 2019 17:44:25 +0000 (17:44 +0000)]
Remove redundant settings.
rob [Sun, 12 May 2019 17:42:14 +0000 (17:42 +0000)]
In long form encoding, explicitly prohibit an initial length octet of 0xff
which is reserved for future use.
ok claudio@
jsg [Sun, 12 May 2019 17:08:14 +0000 (17:08 +0000)]
mention render node device files
schwarze [Sun, 12 May 2019 16:55:11 +0000 (16:55 +0000)]
Delete documentation of the former HOSTALIASES features
which was removed in libc/asr/asr.c rev. 1.50 in 2015.
Smaller version of a patch
posted by Hiltjo Posthuma <hiltjo at codemadness dot org>.
No objection was raised when shown on tech@.
espie [Sun, 12 May 2019 16:38:14 +0000 (16:38 +0000)]
document log reading errors
sashan [Sun, 12 May 2019 16:38:02 +0000 (16:38 +0000)]
pushing NET_LOCK() further down from if_clone_{create,destroy}()
OK mpi@
kettenis [Sun, 12 May 2019 16:36:30 +0000 (16:36 +0000)]
Between the broken GIC-500 on the RK3399 and differences in the secure mode
configuration done by TF-A between boards we have to shift the priorities
that we use for both the architected ICC_PMR_EL1 register and the
memory mapped priority registers on the GIC in different ways. Make this
explicit in the code and try to handle all the cases we care about.
This includes QEMU and RK3399 boards that still use the TF-A version
provided by Rockchip.
Seems to make the rockpro64 run stable with a GENERIC kernel.
ok drahn@, patrick@
akoshibe [Sun, 12 May 2019 16:24:44 +0000 (16:24 +0000)]
Unused switch_port variable.
deraadt [Sun, 12 May 2019 16:10:08 +0000 (16:10 +0000)]
The list + list.local approach is only needed if an architecture builds
multiple bsd.rd. hppa builds one. Fold into one file.
deraadt [Sun, 12 May 2019 16:05:44 +0000 (16:05 +0000)]
The list + list.local approach is only needed if an architecture builds
multiple bsd.rd. hppa builds one. Fold into one file.
jca [Sun, 12 May 2019 15:56:56 +0000 (15:56 +0000)]
Add ZLIB_CONST
Introduced in zlib-1.2.5.2, used by a few ports. ok deraadt@
tb [Sun, 12 May 2019 15:56:31 +0000 (15:56 +0000)]
Fix signed overflow in X509_CRL_print().
fixes oss-fuzz #14558
ok beck jsing
kettenis [Sun, 12 May 2019 15:52:52 +0000 (15:52 +0000)]
Make sure the extended PM register descriptions actually exist before
using them. Fixes machines with ancient ACPI 1.0 (such as recent
hypervisors).
ok deraadt@
tb [Sun, 12 May 2019 15:52:46 +0000 (15:52 +0000)]
Revert the other hunk of r1.36 as well: in the case of CCM, ccm.key is
assigned from aesni_ccm_init_key() via CRYPTO_ccm128_init(), so it needs
to be copied over...
Pointed out by Guido Vranken.
ok jsing
robert [Sun, 12 May 2019 15:30:18 +0000 (15:30 +0000)]
exit early if the errata name is invalid to avoid building a syspatch
that cannot be installed
jsg [Sun, 12 May 2019 15:06:58 +0000 (15:06 +0000)]
Add first drm render node to list of devices to change when logging in
on ttyC0. While here add drm0 to loongson and add the complete set of
wscons and drm devices to arm64.
ok kettenis@
robert [Sun, 12 May 2019 14:57:30 +0000 (14:57 +0000)]
use readlink(1) on each file that we think is different to ensure that
only the resolved path is included
ok ajacoutot@, tb@