schwarze [Wed, 1 Aug 2018 13:44:15 +0000 (13:44 +0000)]
Avoid a read access one byte beyond the end of an allocated string
which occurred in situations like ".Fl a Cm --"; found by
Leah Neukirchen <leah at vuxu dot org> with valgrind on Void Linux.
tb [Wed, 1 Aug 2018 13:35:33 +0000 (13:35 +0000)]
Use a single space after colon in another errx().
Pointed out by Kris Katterjohn
cheloha [Wed, 1 Aug 2018 13:13:53 +0000 (13:13 +0000)]
Disallow negative reps, they make no sense. While here, sscanf -> strtonum.
From Kris Katterjohn on tech@.
ok tb@
patrick [Wed, 1 Aug 2018 12:34:36 +0000 (12:34 +0000)]
Implement a partial update mechanism. Since the SPI-connected display
cannot read the framebuffer memory, we have to push the framebuffer to
the display. ssdfb(4) will now be able to update only a certain region
region of the framebuffer as soon as there is infrastructure to trigger
it.
florian [Wed, 1 Aug 2018 11:47:29 +0000 (11:47 +0000)]
Clarify that we chroot to the users home directory if -u is provided.
Pointed out by Andrew Daugherity (andrew.daugherity AT gmail), thanks!
Tweaks and OK jmc
rob [Wed, 1 Aug 2018 10:42:55 +0000 (10:42 +0000)]
Remove midsentence period.
jmc [Wed, 1 Aug 2018 07:09:15 +0000 (07:09 +0000)]
document when -exec evaluates to true; from kris katterjohn
ok tb
while here, knock out a useless Tn;
tb [Wed, 1 Aug 2018 06:39:58 +0000 (06:39 +0000)]
Update a comment to include -delete, -execdir, -ls and -print0 among
the primaries that do not imply -print. From Kris Katterjohn, thanks!
ok jmc
tb [Wed, 1 Aug 2018 06:37:33 +0000 (06:37 +0000)]
Document that -delete and -execdir prevent -print from being assumed.
From Kris Katterjohn, thanks.
ok jmc
jmc [Wed, 1 Aug 2018 05:55:25 +0000 (05:55 +0000)]
update currency exchange rates;
kn [Tue, 31 Jul 2018 22:48:04 +0000 (22:48 +0000)]
Zap v4mask and v6mask in host()
Simply defer checks whether a mask has been specified to where it's set in
host_*(); this is to reduce address family specific code.
OK sashan
rob [Tue, 31 Jul 2018 19:38:09 +0000 (19:38 +0000)]
Relocate some public functions above the internal functions comment.
ok claudio@
fcambus [Tue, 31 Jul 2018 17:25:55 +0000 (17:25 +0000)]
Display color depth alongside resolution when attaching simplefb(4).
OK kettenis@, deraadt@
claudio [Tue, 31 Jul 2018 15:30:04 +0000 (15:30 +0000)]
Reshuffle order in struct rde_peer a bit. No functional change.
nicm [Tue, 31 Jul 2018 13:06:44 +0000 (13:06 +0000)]
Do not leak path or use it after free.
patrick [Tue, 31 Jul 2018 12:41:57 +0000 (12:41 +0000)]
Stop reversing bytes read from the framebuffer. That was added since
the 8x16 font showed horizontally flipped characters, but as it turns
out the issue is that with 8-bit wide fonts we use optimized rasops
code that apparently writes out the character with reversed bitorder.
nicm [Tue, 31 Jul 2018 11:49:26 +0000 (11:49 +0000)]
Move struct screen_sel into screen.c and tidy up members that are only
used by copy mode.
rob [Tue, 31 Jul 2018 11:37:18 +0000 (11:37 +0000)]
Fix some debugging output now that ber type and encoding are unsigned int.
ok claudio@
claudio [Tue, 31 Jul 2018 11:02:01 +0000 (11:02 +0000)]
Make ber type and encoding a unsigned int instead of unsigned long.
This way the size is the same on all archs and 32bit should be good enough.
OK rob@
claudio [Tue, 31 Jul 2018 11:01:29 +0000 (11:01 +0000)]
Make ber type and encoding a unsigned int instead of unsigned long.
This way the size is the same on all archs and 32bit should be good enough.
OK rob@
claudio [Tue, 31 Jul 2018 11:01:00 +0000 (11:01 +0000)]
Make ber type and encoding a unsigned int instead of unsigned long.
This way the size is the same on all archs and 32bit should be good enough.
OK rob@
claudio [Tue, 31 Jul 2018 11:00:12 +0000 (11:00 +0000)]
Make ber type and encoding a unsigned int instead of unsigned long.
This way the size is the same on all archs and 32bit should be good enough.
OK rob@
nicm [Tue, 31 Jul 2018 10:32:19 +0000 (10:32 +0000)]
Clear history on RIS like most other terminals do.
kettenis [Tue, 31 Jul 2018 10:09:25 +0000 (10:09 +0000)]
Implement setting the voltage of the regulators.
kettenis [Tue, 31 Jul 2018 10:08:24 +0000 (10:08 +0000)]
Correctly set the dividers for the clock of the "big" cores.
kettenis [Tue, 31 Jul 2018 10:07:13 +0000 (10:07 +0000)]
Fix setting the voltage; the code was using the wrong variable as a step size.
Includes some cosmetic fixes as well.
claudio [Tue, 31 Jul 2018 08:04:49 +0000 (08:04 +0000)]
There is no longer the need to be careful in rde_softreconfig_in() when
traversing the prefix list. Since a while Adj-RIB-In is fully independent
and so updating the local RIB does not modify that list.
OK benno@
djm [Tue, 31 Jul 2018 03:10:27 +0000 (03:10 +0000)]
delay bailout for invalid authenticating user until after the packet
containing the request has been fully parsed. Reported by Dariusz Tytko
and MichaĆ Sajdak; ok deraadt
djm [Tue, 31 Jul 2018 03:07:24 +0000 (03:07 +0000)]
fix some memory leaks spotted by Coverity via Jakub Jelen in bz#2366
feedback and ok dtucker@
deraadt [Tue, 31 Jul 2018 02:55:57 +0000 (02:55 +0000)]
KNF
anton [Mon, 30 Jul 2018 20:53:42 +0000 (20:53 +0000)]
Declare then assign local variable, no binary change.
kettenis [Mon, 30 Jul 2018 19:40:53 +0000 (19:40 +0000)]
If we keep the BIOS framebuffer, stick to the desired framebuffer size
instead of using the actual framebuffer size. This fixes a problem seen
on various Apple machines where the BIOS framebuffer is bigger than the
actual display.
ok jcs@, mpi@
mpi [Mon, 30 Jul 2018 19:10:50 +0000 (19:10 +0000)]
regen
mpi [Mon, 30 Jul 2018 19:09:51 +0000 (19:09 +0000)]
Unlock sendmsg(2), sendto(2), recvfrom(2) and recvmsg(2).
ok kettenis@, sthen@, deraadt@, visa@, krw@
bluhm [Mon, 30 Jul 2018 18:05:56 +0000 (18:05 +0000)]
Replace fork with sniffer thread in Python. Clear the BPF list
before sniffing, Scapy 2.4.0 remembers old packets that confuse the
test.
anton [Mon, 30 Jul 2018 17:27:37 +0000 (17:27 +0000)]
Add regress covering the recently fixed NULL pointer deref in open().
jmc [Mon, 30 Jul 2018 17:24:24 +0000 (17:24 +0000)]
catch up with the symlink removal of /dev/{audio,mixer};
ok ratchov
anton [Mon, 30 Jul 2018 17:21:37 +0000 (17:21 +0000)]
Fix a NULL-pointer dereference when calling open() on a cloned device with write
permissions and the flags include O_TRUNC|O_SHLOCK.
ok deraadt@
schwarze [Mon, 30 Jul 2018 16:02:02 +0000 (16:02 +0000)]
improve consistency of markup for some variables, still far from perfect
jcs [Mon, 30 Jul 2018 15:57:04 +0000 (15:57 +0000)]
recognize WSMOUSE_TYPE_TOUCHPAD devices
ok bru
jcs [Mon, 30 Jul 2018 15:56:30 +0000 (15:56 +0000)]
define WSMOUSE_TYPE_TOUCHPAD so non-elantech drivers can stop
claiming to be elantech devices
deraadt [Mon, 30 Jul 2018 15:21:36 +0000 (15:21 +0000)]
Some cleanups and clarification, especially regarding locking (observed by rob)
ok beck jmc
deraadt [Mon, 30 Jul 2018 15:19:12 +0000 (15:19 +0000)]
sync
deraadt [Mon, 30 Jul 2018 15:16:27 +0000 (15:16 +0000)]
rename 2nd argument of unveil from vague "flags" to "permissions";
man page change will follow
kettenis [Mon, 30 Jul 2018 14:19:12 +0000 (14:19 +0000)]
Use the MI interrupt enable/distable API instead of the MD one on i386 and
remove the MD API.
ok deraadt@
espie [Mon, 30 Jul 2018 14:13:08 +0000 (14:13 +0000)]
Tweak history
Document that LOCALBASE changes may not work.
Be less dramatic about it, as it doesn't appear to be *that* bad.
espie [Mon, 30 Jul 2018 12:47:12 +0000 (12:47 +0000)]
byebye fake.mtree
mpi [Mon, 30 Jul 2018 12:22:14 +0000 (12:22 +0000)]
Use FNONBLOCK instead of SS_NBIO to check/indicate that the I/O mode
for sockets is non-blocking.
This allows us to G/C SS_NBIO. Having to keep the two flags in sync
in a mp-safe way is complicated.
This change introduce a behavior change in sosplice(), it can now
always block. However this should not matter much due to the socket
lock being taken beforhand.
ok bluhm@, benno@, visa@
ratchov [Mon, 30 Jul 2018 11:51:42 +0000 (11:51 +0000)]
Inline trivial uaudio_id_name(). From Michael Bombardieri, thanks.
"looks good" kn@
nicm [Mon, 30 Jul 2018 11:24:55 +0000 (11:24 +0000)]
Remove a leftover unused struct.
stsp [Mon, 30 Jul 2018 11:09:17 +0000 (11:09 +0000)]
Don't ask drivers to join a wifi network before an AP has been chosen.
Should fix a panic with bwfm(4) reported by mlarkin@
ok phessler@
kettenis [Mon, 30 Jul 2018 10:56:00 +0000 (10:56 +0000)]
Add support for the GIC v3 ITS and use it to implement MSI support for
rkpcie(4).
ok patrick@
benno [Mon, 30 Jul 2018 09:59:03 +0000 (09:59 +0000)]
cleanup initialization of chngdir. inspired by Ross L Richardson.
ok tb@
benno [Mon, 30 Jul 2018 09:56:50 +0000 (09:56 +0000)]
replace exit() with return(), from Ross L Richardson.
ok tb@ (previous 3 commits to main.c as well)
benno [Mon, 30 Jul 2018 09:54:35 +0000 (09:54 +0000)]
replace warn() + exit() with err()
From Ross L Richardson.
benno [Mon, 30 Jul 2018 09:53:14 +0000 (09:53 +0000)]
line too long and whitespace. From Ross L Richardson.
benno [Mon, 30 Jul 2018 09:51:49 +0000 (09:51 +0000)]
reorder option parsing to be alphabetical/same order as usage and
manpage. From Ross L Richardson.
jmatthew [Mon, 30 Jul 2018 09:04:52 +0000 (09:04 +0000)]
When converting the bios memory map to memory clusters, clip segments at
the 512GB mark as the direct map cannot address memory past that point.
ok kettenis@ (quite a while ago)
florian [Mon, 30 Jul 2018 08:57:09 +0000 (08:57 +0000)]
update to nsd 4.1.23, from the release notes:
------------------------------------------------------------------------
NSD versions 4.1.22 and before are vulnerable in comparing TSIG
information and this can be used to discover a TSIG secret.
NSD uses TSIG to protect zone transfers. The TSIG code uses a secret
key to protect the data. The secret key is shared with both sides of
the zone transfer connection. The comparison code in NSD was not time
insensitive, causing the potential for an attacker to use timing
information to discover data about the key contents.
NSD versions from 2.2.0 to 4.1.22 are vulnerable. Upgrade to 4.1.23 or
newer to get the fix.
It was reported by Ondrej Sury (ISC).
------------------------------------------------------------------------
OK tb, sthen
kn [Mon, 30 Jul 2018 08:28:40 +0000 (08:28 +0000)]
Simplify host()
Get rid of the `cont' flag, zap obvious comments, add error label.
OK benno sashan
patrick [Mon, 30 Jul 2018 08:14:45 +0000 (08:14 +0000)]
Add ssdfb(4), a driver for the SSD1309 controller that drives an
128x64 OLED display. With the typical 8x16 font we get 4 rows with
16 characters each on it. The controller can be driven using I2C,
3-wire and 4-wire SPI. This commit includes support for the 4-wire
protocol.
ok deraadt@
ajacoutot [Mon, 30 Jul 2018 08:05:06 +0000 (08:05 +0000)]
Add colon for clarity.
ok espie@
jmatthew [Mon, 30 Jul 2018 07:34:37 +0000 (07:34 +0000)]
apply the loop settle delay to handling of loop up and loop reset events,
so hotplug can be more reliable too.
jmatthew [Mon, 30 Jul 2018 07:30:54 +0000 (07:30 +0000)]
extend the loop settle time to 200ms, and adjust the check so that we'll
actually break out once the loop has been up for that long.
jmc [Mon, 30 Jul 2018 05:23:00 +0000 (05:23 +0000)]
tweak previous;
deraadt [Mon, 30 Jul 2018 00:34:57 +0000 (00:34 +0000)]
KNF
beck [Mon, 30 Jul 2018 00:30:15 +0000 (00:30 +0000)]
document the current limitation (we don't yet find an above covering
unveil for relative operations) that I am working on in BUGS
beck [Mon, 30 Jul 2018 00:16:59 +0000 (00:16 +0000)]
Allow for us to do a CREATE style lookup on a read only filesystem
if and only if we are unveil doing it. Fixes an issue noticed
by kn@ where unveil would fail with EROFS on a read only filesystem
deraadt [Mon, 30 Jul 2018 00:11:04 +0000 (00:11 +0000)]
activate unveil for testing
ok kibbles and bits
beck [Sun, 29 Jul 2018 23:53:04 +0000 (23:53 +0000)]
revert accidentally changed prototype
beck [Sun, 29 Jul 2018 23:11:02 +0000 (23:11 +0000)]
Don't exceed UNVEIL_MAX_VNODES with a long path now that we save
the traversed vnodes - noticed and fixed by semarie@
beck [Sun, 29 Jul 2018 22:53:39 +0000 (22:53 +0000)]
Make sure we don't count looking at .. as a component
as a descending match.
Noticed by Stuart Cassoff <3d0g@bell.net>
beck [Sun, 29 Jul 2018 22:30:32 +0000 (22:30 +0000)]
Add regress for ensuring .. is appropriately *not* used as a
descending match
tb [Sun, 29 Jul 2018 20:29:32 +0000 (20:29 +0000)]
Document that X509_{NAME,REQ,REQ_INFO}_free() are all NULL safe.
From Ross L. Richardson
tb [Sun, 29 Jul 2018 20:22:02 +0000 (20:22 +0000)]
Remove NULL checks before X509_{REQ,NAME}_free() and zap an unnecessary
pair of parens.
From Ross L. Richardson
benno [Sun, 29 Jul 2018 20:15:23 +0000 (20:15 +0000)]
replace hand-rolled tmp files with mkstemp()
ok florian@ back in april, reminded by theo.
deraadt [Sun, 29 Jul 2018 19:40:41 +0000 (19:40 +0000)]
Add _PATH_AUTHPROGDIR = "/usr/libexec/auth", this path will be used
to unveil. Unfortunately the auth subsystem uses _PATH_AUTHPROG =
"/usr/libexec/auth/login_", which it auth-program is appended to -- a
rather gross idea which now shows lack of wisdom.
anton [Sun, 29 Jul 2018 14:11:05 +0000 (14:11 +0000)]
add missing markup and some minor tweaks; ok jmc@
deraadt [Sun, 29 Jul 2018 13:34:26 +0000 (13:34 +0000)]
After "termcap" initilization is finished, top appears to not open any
files ever again, so we can re-pledge tighter.
schwarze [Sun, 29 Jul 2018 13:27:44 +0000 (13:27 +0000)]
garbage collect old "manpages-check" target
which has been broken for years with nobody complaining;
OK espie@ kn@
deraadt [Sun, 29 Jul 2018 13:02:01 +0000 (13:02 +0000)]
mestre and i both concluded pledge can be tightened, "unix" handles the
coming code and "rpath wpath" isn't needed
ok claudio
deraadt [Sun, 29 Jul 2018 12:46:31 +0000 (12:46 +0000)]
remove extra space in output; from Ross L Richardson
schwarze [Sun, 29 Jul 2018 11:27:14 +0000 (11:27 +0000)]
UTF-8 support: use wcwidth(3) when calculating column widths;
written during g218; no objection when shown on tech@
bluhm [Sun, 29 Jul 2018 09:25:58 +0000 (09:25 +0000)]
New scapy 2.4.0 buffers packets at the bfp layer. Call sniff() to
clear this list when starting the sniffer thread. The TCP sequence
number check in sr1() became more strict. Use the sniffer thread
to capture retransmitted packetes which the new TCP answers() check
ignores now.
kn [Sat, 28 Jul 2018 23:36:54 +0000 (23:36 +0000)]
Use strtonum in host()
This is simpler than checking three cases for `q' and gives nicer error
messages. While here, use `v6mask' as maximum netmask instead of hardcoding
it.
OK sashan
deraadt [Sat, 28 Jul 2018 22:57:27 +0000 (22:57 +0000)]
k&r -> ansi function headers in this file, 'cause clang complained about one.
mortimer [Sat, 28 Jul 2018 21:43:21 +0000 (21:43 +0000)]
Add -fno-ret-protector for arm64.
Prompted by deraadt
schwarze [Sat, 28 Jul 2018 18:32:30 +0000 (18:32 +0000)]
Issue a STYLE message when normalizing the date format in .Dd/.TH.
Leah Neukirchen pointed out that mdoclint(1) used to warn about a
leading zero before the day number, so we know that both NetBSD and
Void Linux want the message. It does no harm on OpenBSD because
Mdocdate always does the right thing anyway.
jmc@ agrees that it makes sense in contexts not using Mdocdate.
deraadt [Sat, 28 Jul 2018 18:07:26 +0000 (18:07 +0000)]
delete unused ps_uvactive, since active checks are done against the ptr
from semarie
deraadt [Sat, 28 Jul 2018 18:06:30 +0000 (18:06 +0000)]
re-ordering for sensibility, by semarie; ok jmc
kettenis [Sat, 28 Jul 2018 15:28:51 +0000 (15:28 +0000)]
Add function to convert a PCI device "tag" into a PCIe requester ID.
ok patrick@, mlarkin@, deraadt@
tb [Sat, 28 Jul 2018 15:25:23 +0000 (15:25 +0000)]
Remove NULL checks before (most) libcrypto *_free() functions.
From Ross L. Richardson, thanks!
ok deraadt
kettenis [Sat, 28 Jul 2018 13:59:08 +0000 (13:59 +0000)]
Make use of PCI_FLAGS_MSI_ENABLED such that drivers for hardware with broken
MSI support can selectively disable the use of MSI.
ratchov [Sat, 28 Jul 2018 09:11:55 +0000 (09:11 +0000)]
Move libsndio session cookie in its own $HOME/.sndio/ directory to
make libsndio easier to use with unveil(2).
"make sense" deraadt
ratchov [Sat, 28 Jul 2018 09:07:48 +0000 (09:07 +0000)]
Rename the sndiod unix domain socket to /tmp/sndio/sockN to avoid
wondering what are these "aucat" files in /tmp.
"make sense" deraadt
ratchov [Sat, 28 Jul 2018 08:11:08 +0000 (08:11 +0000)]
sync
ratchov [Sat, 28 Jul 2018 08:09:50 +0000 (08:09 +0000)]
Remove unused /dev/audio and /dev/audioctl symlinks.
ok deraadt
kettenis [Fri, 27 Jul 2018 21:11:31 +0000 (21:11 +0000)]
Use the MI interrupt enable/distable API instead of the MD one on amd64 and
remove the MD API.
ok guenther@, deraadt@, mpi@
rob [Fri, 27 Jul 2018 19:14:45 +0000 (19:14 +0000)]
Full stop.
schwarze [Fri, 27 Jul 2018 17:47:05 +0000 (17:47 +0000)]
garbage collect the unused "#define INDENT"