otto [Wed, 27 Dec 2017 10:05:23 +0000 (10:05 +0000)]
step one in avoiding unneccesary init of chunk_info;
some cleanup; tested by sthen@ on a ports build
sthen [Wed, 27 Dec 2017 00:00:49 +0000 (00:00 +0000)]
sync (if_bwfm_pci.h)
kettenis [Tue, 26 Dec 2017 23:47:41 +0000 (23:47 +0000)]
regen
kettenis [Tue, 26 Dec 2017 23:47:17 +0000 (23:47 +0000)]
Add two IDT PCIe switches found in a SPARC T4-1 and tweak another one to
not include the revision of the chip.
patrick [Tue, 26 Dec 2017 21:01:02 +0000 (21:01 +0000)]
Cherry-pick a change from LLD to make the behavior of the -v option
more closer to GNU linkers. This should help with autoconf/libtool
compatibility in ports.
Requested by Brad
"no objection" kettenis@
patrick [Tue, 26 Dec 2017 20:59:44 +0000 (20:59 +0000)]
Update to compiler-rt 5.0.1.
ok kettenis@
patrick [Tue, 26 Dec 2017 19:34:27 +0000 (19:34 +0000)]
Cherry-pick a change from LLVM that marks specific pseudo memory
instructions to have side effects so the optimizer does not reorder
them across fnstcw/fldcw sequences. Fixes a bug seen in sqlite3 on
i386.
ok kettenis@
jca [Tue, 26 Dec 2017 19:12:22 +0000 (19:12 +0000)]
Missed in previous
Spotted by patrick@, thanks
millert [Tue, 26 Dec 2017 19:10:31 +0000 (19:10 +0000)]
The hack for /dev/fd support on systems without it was removed a
long time ago. Update the comments and replace test_stat() calls
with stat() since test_stat() now just calls stat() directly. Also
rename the "mode" parameter to "amode" in test_eaccess() to match
access(2) and make it clear that this is the access mode and not
the file mode. OK jca@
kettenis [Tue, 26 Dec 2017 17:16:33 +0000 (17:16 +0000)]
Correct register numbers for AXP809.
bluhm [Tue, 26 Dec 2017 17:03:46 +0000 (17:03 +0000)]
As the divert-to pf state is not connected, it is not linked to the
incpb and so not removed when the socket is closed. Packets matching
reply and reply-to states are linked to the sockets immediately.
So these states must go away when the socket is closed. Check this
with the output of "pfctl -s states" after running the Perl scripts.
jca [Tue, 26 Dec 2017 15:24:38 +0000 (15:24 +0000)]
sync
jca [Tue, 26 Dec 2017 15:23:33 +0000 (15:23 +0000)]
Use a version script to limit exported symbols
ok ratchov@ kettenis@
kettenis [Tue, 26 Dec 2017 15:14:22 +0000 (15:14 +0000)]
sync
kettenis [Tue, 26 Dec 2017 15:11:17 +0000 (15:11 +0000)]
Add ARM EABI runtime aliases to the GCC runtime functions that we include
in libc.
ok patrick@, jsg@, guenther@
kettenis [Tue, 26 Dec 2017 13:22:36 +0000 (13:22 +0000)]
drm/i915: call kunmap_px on pt_vaddr
From Matthew Auld
44a7102484db0ddfa6f855b57ffe0566f739b55a in mainline linux
Pointed out by Martin Ziemer
kettenis [Tue, 26 Dec 2017 12:58:59 +0000 (12:58 +0000)]
Sync Allwinner A64 pin defenitions with Linux.
Add support for the 2nd A64 controller.
Add Allwinner A33 support.
kevlo [Tue, 26 Dec 2017 09:31:51 +0000 (09:31 +0000)]
Implement support for the "next-generation" clock bindings for the
Allwinner A23/A33.
ok kettenis@
patrick [Tue, 26 Dec 2017 00:51:29 +0000 (00:51 +0000)]
Fix channel initialization by not using the same loop variable in nested
for-loops. Oops.
anton [Mon, 25 Dec 2017 16:01:35 +0000 (16:01 +0000)]
Markup SIGHUP and rework sentence; ok jmc@
patrick [Mon, 25 Dec 2017 12:09:20 +0000 (12:09 +0000)]
sync
tedu [Mon, 25 Dec 2017 02:11:19 +0000 (02:11 +0000)]
memcpy from the right place. at this point, the used variable is not
relevant. from Mark Karpilovskij.
ok millert
patrick [Mon, 25 Dec 2017 00:41:59 +0000 (00:41 +0000)]
Tedu files that got removed in LLVM 5.0.1.
patrick [Sun, 24 Dec 2017 23:29:52 +0000 (23:29 +0000)]
Update to LLVM 5.0.1 which apparently includes plenty of bugfixes.
Requested by Brad
Tested by me on amd64
Tested by bluhm on i386
Tested by kettenis on arm64, armv7 and sparc64
kettenis [Sun, 24 Dec 2017 23:24:45 +0000 (23:24 +0000)]
Re-enable the code that unmaps the startup code. This was temporarily
disabled while lld was growing up, but forgotten about when gap.o was
added back.
ok deraadt@
patrick [Sun, 24 Dec 2017 23:19:13 +0000 (23:19 +0000)]
Merge LLVM 5.0.1 release.
patrick [Sun, 24 Dec 2017 23:15:17 +0000 (23:15 +0000)]
Import LLVM 5.0.1 release including clang, lld and lldb.
cheloha [Sun, 24 Dec 2017 22:12:49 +0000 (22:12 +0000)]
Consolidate printf(3) calls at the end of main().
Makes reading the classification algorithm easier.
ok tb@
patrick [Sun, 24 Dec 2017 19:50:56 +0000 (19:50 +0000)]
Add a PCI attachment driver for bwfm(4). It's not finished, but it's
already past the point where development can occur out of the tree.
With this I can successfully scan for access points and tell the chip
to attach to an SSID. RX path should work as well, but since I forgot
to bring the antenna with me to my parents, the reception is a bit
horrible in the metal enclosure.
There are a few reasons this driver is rather big. First we set up the
ARM Cores, uploading the firmware and kicking it off. Then we need to
read all needed information from the registers. Once that is done we
have to set up countless buffers. There are 2 TX rings and 3 RX rings,
plus N TX rings for the actual data that is yet to be implemented.
Merry Christmas!
ok kettenis@
kettenis [Sun, 24 Dec 2017 19:43:51 +0000 (19:43 +0000)]
For systems where the cpu node in the device tree has a "clocks" property,
implement hw.cpuspeed using the clock framework.
ok patrick@
kettenis [Sun, 24 Dec 2017 19:42:51 +0000 (19:42 +0000)]
For systems where the cpu node in the device tree has a "clocks" property,
implement hw.cpuspeed using the clock framework.
ok patrick@
kettenis [Sun, 24 Dec 2017 18:24:06 +0000 (18:24 +0000)]
Implement A10/A20 CPU clock.
cheloha [Sun, 24 Dec 2017 16:59:50 +0000 (16:59 +0000)]
Use more libm.
adj360() is a modulo operation: use fmod(3) to simplify things.
Adding 0.5 to the phase to steer implicit truncation is a hack:
use lround(3) to simplify things.
Product of a discussion with pjanzen@/jca@/tb@ regarding floating-point
rounding and the importance of correctly classifying a gibbous moon.
ok pjanzen@ jca@ tb@
fcambus [Sun, 24 Dec 2017 16:19:27 +0000 (16:19 +0000)]
Somes fixes for the VIA PadLock drivers.
- Do not ignore the return value of m_copyback() on i386
- Always free sc->op_buf before returning on both amd64 and i386,
pointed out by mikeb@, thanks!
OK mikeb@
bluhm [Sun, 24 Dec 2017 14:18:19 +0000 (14:18 +0000)]
There was a corner case where linking the inp to the state key did
work in pf. The function pf_inp_link() takes the state key from
the mbuf and not the one pf_find_state() has just found. Introduce
a new function pf_state_key_link_inpcb() that links the given state
key and inpcb together with some sanity checks.
OK sashan@
kettenis [Sun, 24 Dec 2017 12:55:52 +0000 (12:55 +0000)]
Allow SD/MMC controller drivers to allocate their own DMA mapping structure
to account for any relevant hardware constraints.
ok stsp@
kettenis [Sun, 24 Dec 2017 10:32:25 +0000 (10:32 +0000)]
Generate SIGBUS/BUS_ADRALN when we see a SP Alignment Exception from userland.
ok jsg@
millert [Sun, 24 Dec 2017 01:50:50 +0000 (01:50 +0000)]
Fix one possible buffer overflow and one underflow. Also some minor
cleanups. From Jan Kokemueller. OK deraadt@
tb [Sun, 24 Dec 2017 00:11:43 +0000 (00:11 +0000)]
Avoid comparing equal lines twice when running without -i.
From kshe, ok jca
cheloha [Sat, 23 Dec 2017 20:58:14 +0000 (20:58 +0000)]
As we only use the .tv_sec field, simplify gettimeofday(2) -> time(3).
While here, use err(3) instead of errx(3) if adjtime(2) fails.
Discussed/tweaked with/by tb@ and jca@.
ok tb@ jca@
cheloha [Sat, 23 Dec 2017 20:53:07 +0000 (20:53 +0000)]
As we only use the .tv_sec field, simplify gettimeofday(2) -> time(3).
ok tb@ jca@
cheloha [Sat, 23 Dec 2017 20:04:23 +0000 (20:04 +0000)]
Use the monotonic clock for logging progress in cdio(1) and ftp(1).
Keeps the progress log from blipping or stalling if, e.g., the
system time is changed in the midst of a rip or a transfer.
ok tb@ jca@
mlarkin [Sat, 23 Dec 2017 18:35:41 +0000 (18:35 +0000)]
regen
mlarkin [Sat, 23 Dec 2017 18:34:54 +0000 (18:34 +0000)]
Add missing Sunrise Point-H I2C controller PCI IDs
sthen [Sat, 23 Dec 2017 16:20:21 +0000 (16:20 +0000)]
Make it clear that MFS can be used for the noperm permission holding DESTDIR.
(If anyone needs a bit more handholding than is in release(8), the faq page
https://www.openbsd.org/faq/faq5.html#Release has more information about
setting this up).
Much simplified from, but inspired by, various diffs from ians@. ok tb@ jmc@
kettenis [Sat, 23 Dec 2017 15:04:46 +0000 (15:04 +0000)]
Remove stray space.
kettenis [Sat, 23 Dec 2017 12:55:26 +0000 (12:55 +0000)]
Sync Allwinner H3/H5 pin definitions with Linux.
espie [Sat, 23 Dec 2017 12:35:47 +0000 (12:35 +0000)]
the trick to get signal names requires skipping over POSIX stuff we don't
have now.
as seen by landry@
kettenis [Sat, 23 Dec 2017 12:28:45 +0000 (12:28 +0000)]
Avoid null-pointer dereference wen a device tree refers to a pin-function
that doesn't exist.
bru [Sat, 23 Dec 2017 10:50:15 +0000 (10:50 +0000)]
scrolling: take care of noise.
patrick [Sat, 23 Dec 2017 10:30:54 +0000 (10:30 +0000)]
Since ikev2_init_recv() is supposed to only handle responses to an
exchange that we initiatiated, we are not allowed to respond to such
a msg. Also we don't need the DH check in ikev2_sa_initiator_dh() as
it's only called when we initiate, so the check would not run, or when
we get a Create Child SA response, where an error should only lead to
us having another attempt at an exchange.
Found by and ok markus@
fcambus [Sat, 23 Dec 2017 10:30:25 +0000 (10:30 +0000)]
Remove dead assignment.
Now that we call rasops_putchar_rotated(), we don't need ri anymore.
OK kettenis@
kettenis [Sat, 23 Dec 2017 10:23:34 +0000 (10:23 +0000)]
Use regmap interface to tweak "SYSREG" PHY configuration registers and
remove exsysreg(4). The exsysreg(4) driver would no longer attach
as syscon(4) would match its node in the device tree. As a result the
driver corss-call would panic.
Makes te Odroid-XU4 work again.
jsg [Sat, 23 Dec 2017 01:44:24 +0000 (01:44 +0000)]
enable snooping on Intel 200 Series HDA
jsg [Sat, 23 Dec 2017 01:42:31 +0000 (01:42 +0000)]
match on intel 200 series
jsg [Sat, 23 Dec 2017 01:34:35 +0000 (01:34 +0000)]
regen
jsg [Sat, 23 Dec 2017 01:33:51 +0000 (01:33 +0000)]
add ids for intel coffee lake-s and series 200/300 pch
schwarze [Sat, 23 Dec 2017 00:52:33 +0000 (00:52 +0000)]
clarify that -i only affects the comparison of ASCII characters;
OK jmc@ tb@
nicm [Fri, 22 Dec 2017 23:16:41 +0000 (23:16 +0000)]
Fix memory leak in screen_redraw_make_pane_status.
anton [Fri, 22 Dec 2017 21:19:10 +0000 (21:19 +0000)]
Since export accepts several variables, put them on a single line; ok tb@
rpe [Fri, 22 Dec 2017 19:41:49 +0000 (19:41 +0000)]
Cope with the fact, that the DHCLIENT route label is gone which
helped to find out if the default route was configured by dhclient.
Check if the interface connected to the default route is in the
dhcp interface group which accomplishes the same functionality.
With feedback from and OK krw@
kettenis [Fri, 22 Dec 2017 15:52:36 +0000 (15:52 +0000)]
If we use the cookie-based interrupt API on sun4v, whe shouldn't put the
interrupts in the lookup table. On machines like the Oracle T7-2 this
would make the code believe that there is interrupt sharing, but since
we bypass intr_list_handler() when using the cookie-based interrupt API,
we wouldn't properly acknowledge the interrupt.
This fixes the last remaining problem with interrupts on the Oracle T7-2.
bru [Fri, 22 Dec 2017 15:21:04 +0000 (15:21 +0000)]
1. Use unfiltered deltas for scrolling. 2. (Re-)Simplify the edge area
setup.
nicm [Fri, 22 Dec 2017 10:18:51 +0000 (10:18 +0000)]
Do not try to set default value on user options (they don't have one),
from Charles Howard in GitHub issue 1161.
nicm [Fri, 22 Dec 2017 10:16:36 +0000 (10:16 +0000)]
Remove duplicate WheelUp/WheelDown entries in list, GitHub issue 1184.
bluhm [Thu, 21 Dec 2017 23:27:17 +0000 (23:27 +0000)]
Add some comments what the tests are doing.
millert [Thu, 21 Dec 2017 21:04:38 +0000 (21:04 +0000)]
Expand u_short and u_int to unsigned short and unsigned int
respectively to avoid compilation errors when one of the POSIX or
X/OPEN version macros is defined. Also sync the field descriptions
with kqueue.2. OK deraadt@
jmc [Thu, 21 Dec 2017 21:03:49 +0000 (21:03 +0000)]
typo;
rpe [Thu, 21 Dec 2017 19:22:14 +0000 (19:22 +0000)]
Adopt comments to previous change
krw [Thu, 21 Dec 2017 15:14:45 +0000 (15:14 +0000)]
Stop creating a dhclient.conf in the install script, now
that 'send host-name' is the default behaviour for
dhclient.
ok rpe@
mpi [Thu, 21 Dec 2017 12:09:38 +0000 (12:09 +0000)]
Call if_start() directly to send the EAPOL key.
Using if_enqueue() here, from interrupt context, might result in
the packet beeing enqueued, incorrectly encrypted, on the TX ring.
This race has been recently exposed by the re-introduction of the
TX mitigation. It exists because the net80211 stack sets
IEEE80211_NODE_TXPROT on the node while processing the 3rd message,
assuming the answer has already been transmitted. However a CPU
returns from if_enqueue() it cannot assume that the send queue is
empty. So call if_start() to flush this queue.
Encrypting the 4th message of the 4way handshake with the new key
breaks WPA handshake as found the hardway by anton@.
Race analysed by dlg@, a lot of net80211 inputs and suggetions from
stsp@.
ok stsp@, dlg@
tb [Thu, 21 Dec 2017 10:05:59 +0000 (10:05 +0000)]
Add an -i option for case insensitive comparison of lines.
From Claus Assmann with minor tweaks by me.
ok millert
tb [Thu, 21 Dec 2017 07:29:15 +0000 (07:29 +0000)]
Two width specifiers changed from %02 to %0 in previous. Change them back.
ok deraadt
deraadt [Thu, 21 Dec 2017 06:55:22 +0000 (06:55 +0000)]
Do not use %hx for chars, ok kettenis
djm [Thu, 21 Dec 2017 05:46:35 +0000 (05:46 +0000)]
another libssh casualty
jmatthew [Thu, 21 Dec 2017 05:09:56 +0000 (05:09 +0000)]
Close the right file descriptor and clean up the tls context in aldap_close().
ok zhuk@ deraadt@
dlg [Thu, 21 Dec 2017 03:59:32 +0000 (03:59 +0000)]
drivers might want to call ifiq_barrier.
dlg [Thu, 21 Dec 2017 03:58:27 +0000 (03:58 +0000)]
add some ifq and ifiq barriers in ixl_down.
move rxfill in ixl_up so the qtail is only written once.
djm [Thu, 21 Dec 2017 03:01:49 +0000 (03:01 +0000)]
missed one (unbreak after ssh/lib removal)
dlg [Thu, 21 Dec 2017 01:55:44 +0000 (01:55 +0000)]
now that we have multiple input queues in ifnet structs, use them.
for now we still only have one set of tx and rx rings, but sets us up to
bump that number.
dlg [Thu, 21 Dec 2017 01:11:47 +0000 (01:11 +0000)]
prototype if_attach_iqueues so drivers can configure multiple iqs.
djm [Thu, 21 Dec 2017 00:41:22 +0000 (00:41 +0000)]
unbreak unit tests after removal of src/usr.bin/ssh/lib
kettenis [Thu, 21 Dec 2017 00:32:55 +0000 (00:32 +0000)]
Implement missing bits to support lazy binding. Note that the code
deliberately does not save the floating-point argument registers
before calling _dl_bind(). Doing so would force an FPU context switch
upon every function call through the PLT. But since we compile ld.so
with -march=armv8-a+nofp+nosimd this is safe since nothing in the _dl_bind()
codepath uses he FPU registers.
ok guenther@, drahn@
djm [Thu, 21 Dec 2017 00:00:28 +0000 (00:00 +0000)]
revert stricter key type / signature type checking in userauth path;
too much software generates inconsistent messages, so we need a
better plan.
bluhm [Wed, 20 Dec 2017 23:37:56 +0000 (23:37 +0000)]
Skip this regress test if no remote test machine is specified as
environment variable.
tb [Wed, 20 Dec 2017 21:08:06 +0000 (21:08 +0000)]
Fix adding and removing files with git-style a/ b/ diffs: only skip
six letters if they actually match "--- a/" and "+++ b/" instead of
laxer checks.
ok tom, tedu, millert
krw [Wed, 20 Dec 2017 18:51:14 +0000 (18:51 +0000)]
Always 'send' host-name, rather than relying on dhclient.conf to ask
for it. Removes the need to install a dhclient.conf for a default
configuration.
Install script simplification to follow.
General enthusiasm. ok millert@ benno@ tom@ ian@
stsp [Wed, 20 Dec 2017 18:20:59 +0000 (18:20 +0000)]
Two background scan fixes for iwn(4) (4965 devices only):
1) Use only one antenna while scanning on 5GHz. Fixes very low RSSI values.
2) During a background scan while associated to a 5GHz AP, send probe
requests on 2GHz channels with an OFDM rate (6Mbps) because the buggy
firmware does not like sending with a CCK rate (1Mbps) in this state.
CCK rates are not valid for 5GHz, which could explain this firmware bug.
Taken from FreeBSD r222679.
jca@ tested and confirmed the first problem but lacks a 5GHz access point to
associate to and test the second one with.
I am going to *blindly trust* FreeBSD on the second one! It is not a big deal
if it doesn't actually apply to our firmware version as the change itself is
rather inconsequential in practice. And 4965 hardware is quite rare nowadays.
ok jca@
mpi [Wed, 20 Dec 2017 11:22:29 +0000 (11:22 +0000)]
Sync C mutex implementations.
ok kettenis@
mpi [Wed, 20 Dec 2017 11:08:44 +0000 (11:08 +0000)]
Switch x86 architectures to the common C mutex implementation.
This is a step towards MI mutexes.
ok kettenis@
krw [Wed, 20 Dec 2017 02:02:31 +0000 (02:02 +0000)]
Add a sanity check on the number of chunks being
assembled into a volume, to enure the on-disk
metadata and the in-memory metadata agree.
zhuk [Tue, 19 Dec 2017 20:44:53 +0000 (20:44 +0000)]
More NUL (not NULL) checks.
With support from guenther@ and millert@; final okay espie@
patrick [Tue, 19 Dec 2017 18:06:41 +0000 (18:06 +0000)]
Pass calibrated TSC frequency to vmm(4) guests. On machines where TSC
cannot be calculated from the CPUID register, because the CPU does not
support it, but can be calibrated from another timer, the vmm(4) guest
doesn't have a chance to read or calibrate its own TSC frequency since
it has no access to a precise enough time counter. In that case, fake
the existence of the register and supply our calibrated TSC frequency.
mlarkin@ tells me this also greatly helps Linux guests.
ok mlarkin@
martijn [Tue, 19 Dec 2017 17:46:28 +0000 (17:46 +0000)]
Remove EXAMPLES section.
The example itself was taken directly from the POSIX specification by jmc@,
who doesn't remember adding it. This example isn't very clear about what
it actually does, so just remove it.
If someone can come up with one or more examples that give a more clear
introduction to the workings of sed feel free to send them in.
Prompted by a diff by kshe59 <at> zoho <dot> eu
OK jmc@
deraadt [Tue, 19 Dec 2017 16:41:43 +0000 (16:41 +0000)]
curproc access isn't necessarily as cheap as a local variable access,
so only get it once
ok guenther
nicm [Tue, 19 Dec 2017 15:00:39 +0000 (15:00 +0000)]
Report better error from server when socket create fails, GitHub issue
1201.
espie [Tue, 19 Dec 2017 13:51:13 +0000 (13:51 +0000)]
somehow, I lost that static.
From Michal W. Bombardieri, thx
stefan [Tue, 19 Dec 2017 10:04:59 +0000 (10:04 +0000)]
Remove unused ps_stackgap from process struct
Nothing uses this field since Linux compat was removed.
ok mpi@ deraadt@ guenther@
mpi [Tue, 19 Dec 2017 09:35:56 +0000 (09:35 +0000)]
Remove a 27 year old #ifdef notdef chunk involving SB_LOCK.
ok bluhm@
mpi [Tue, 19 Dec 2017 09:32:15 +0000 (09:32 +0000)]
Inline socket buffer related defines, no functional change.
ok bluhm@