openbsd
6 years agoWhen 'log updates' is enabled make sure that the case of an invalid path
claudio [Wed, 29 Aug 2018 11:46:28 +0000 (11:46 +0000)]
When 'log updates' is enabled make sure that the case of an invalid path
can be distinguished from a withdraw that was caused by a deny filter.
OK sthen@

6 years agoWhen checking for AS 0 make sure the full ASPATH is validated and don't
claudio [Wed, 29 Aug 2018 11:43:15 +0000 (11:43 +0000)]
When checking for AS 0 make sure the full ASPATH is validated and don't
return on the first 0. The other AS_ERR_SOFT cases do the same and ensure
that an ASPATH with an AS_ERR_SOFT is still valid enough for subsequent
manipulation or inspection.
OK benno@

6 years agoHistorically /etc/netstart (and the equivalent code in the install
krw [Wed, 29 Aug 2018 11:30:48 +0000 (11:30 +0000)]
Historically /etc/netstart (and the equivalent code in the install
script) did 'ifconfig <if> down' before starting dhclient(8). This was
a way of ensuring old running copies of dhclient were killed before a
new one started. Current dhclient does not need this assist, so change
"ifconfig <if> down" to "ifconfig <if> up" pending further script
optimizations.

Similar to a 2014 attempt by halex@. Prompted by a misc@ report
from Kristjan Komlosi reporting hanging diskless setups.

ok halex@ kn@

6 years agoKeep any text killed in the command prompt with C-w and yank it with
nicm [Wed, 29 Aug 2018 09:50:32 +0000 (09:50 +0000)]
Keep any text killed in the command prompt with C-w and yank it with
C-y, only use the top buffer if no text has previously been killed. This
and previous change promped by discussion with kn@.

6 years agoAdd C-Left and C-Right as aliases for M-b and M-f.
nicm [Wed, 29 Aug 2018 08:56:51 +0000 (08:56 +0000)]
Add C-Left and C-Right as aliases for M-b and M-f.

6 years agoAdjust comment to make it clear that a NULL nexthop is actually eligable
claudio [Wed, 29 Aug 2018 08:51:49 +0000 (08:51 +0000)]
Adjust comment to make it clear that a NULL nexthop is actually eligable

6 years agoPrevent ospfd from starting if another ospfd is already using the same
remi [Wed, 29 Aug 2018 08:43:16 +0000 (08:43 +0000)]
Prevent ospfd from starting if another ospfd is already using the same
socket. The check on the control socket happens early so that the rc
scripts detect a failure.

OK claudio@ florian@

6 years agoAdd set-case-replaced to toggle case-preserving replace on or off.
reyk [Wed, 29 Aug 2018 07:50:16 +0000 (07:50 +0000)]
Add set-case-replaced to toggle case-preserving replace on or off.

By default, replacing "foo" with "bar" turns "FOO" into "BAR".
With case-replace turned off, "FOO" will turn into "bar".

OK florian@ tb@

6 years agoInitialize device numbers for newly allocated ptys. Prevents a panic caused by
anton [Wed, 29 Aug 2018 06:04:46 +0000 (06:04 +0000)]
Initialize device numbers for newly allocated ptys. Prevents a panic caused by
the following: a new pty is allocated in which the kernel console output is
redirected to, poll(2):ing from /dev/console at this point would be delegated to
the device with the major number taken from the pty due to the earlier
redirection. Since the pty does not have its correct device major assigned, the
wrong device ends up being used.

ok deraadt@ millert@

6 years agoFirst pass in bringing i386 in sync with amd64. This does not yet work, but is
pd [Wed, 29 Aug 2018 04:51:12 +0000 (04:51 +0000)]
First pass in bringing i386 in sync with amd64. This does not yet work, but is
being committed now so we can work on the rest in-tree.

ok mlarkin@

6 years agomove the todr chip handle into a prtc_softc structure
dlg [Wed, 29 Aug 2018 02:03:12 +0000 (02:03 +0000)]
move the todr chip handle into a prtc_softc structure

no realy change, this avoid having to allocate the todr struct in attach

ok deraadt@

6 years agoRun Wycheproof AES-CMAC testvectors against libcrypto.
tb [Tue, 28 Aug 2018 18:28:30 +0000 (18:28 +0000)]
Run Wycheproof AES-CMAC testvectors against libcrypto.

6 years agoremove some extra parens and fix some other formatting issues
tb [Tue, 28 Aug 2018 18:25:33 +0000 (18:25 +0000)]
remove some extra parens and fix some other formatting issues
pointed out by gofmt (thanks anton)

6 years agoUnclutter bgpd.conf example config file
job [Tue, 28 Aug 2018 17:52:16 +0000 (17:52 +0000)]
Unclutter bgpd.conf example config file

We're aiming to work towards an examples/bgpd.conf file that is a good starting
point for real world deployments rather than a showcase of all capabilities
bgpd(8) has.

OK deraadt@ claudio@

6 years agoRemove extra "and" in "These functions and have been available"
tb [Tue, 28 Aug 2018 17:47:29 +0000 (17:47 +0000)]
Remove extra "and" in "These functions and have been available"

6 years agozap trailing whitespace
tb [Tue, 28 Aug 2018 17:45:50 +0000 (17:45 +0000)]
zap trailing whitespace

6 years agoDisplay per-TDB counters in verbose mode.
mpi [Tue, 28 Aug 2018 15:17:56 +0000 (15:17 +0000)]
Display per-TDB counters in verbose mode.

ok sthen@

6 years agoAdd per-TDB counters and a new SADB extension to export them to
mpi [Tue, 28 Aug 2018 15:15:02 +0000 (15:15 +0000)]
Add per-TDB counters and a new SADB extension to export them to
userland.

Inputs from markus@, ok sthen@

6 years agoDrop SSLv2, SSLv3 support.
cheloha [Tue, 28 Aug 2018 14:30:48 +0000 (14:30 +0000)]
Drop SSLv2, SSLv3 support.

No need to check for SSLv2/3 sessions when printing the tally mark.
Also do SSLv23_client_method -> TLS_client_method.

ok jsing@

6 years agoTweak messages emitted by editor_resize() so "Partition x shrunk ..."
krw [Tue, 28 Aug 2018 12:40:54 +0000 (12:40 +0000)]
Tweak messages emitted by editor_resize() so "Partition x shrunk ..."
is only printed if the disklabel gets modified. Also, include the new
size.

On failure just report "Amount too big.".

6 years agofix misplaced parenthesis inside if-clause. it's harmless and the only issue is
mestre [Tue, 28 Aug 2018 12:25:53 +0000 (12:25 +0000)]
fix misplaced parenthesis inside if-clause. it's harmless and the only issue is
showing an unknown error (since it's not defined) during fatal(), if it ever
an error occurs inside that condition.

OK deraadt@ markus@ djm@

6 years agofix build with DEBUG_PK enabled
mestre [Tue, 28 Aug 2018 12:17:45 +0000 (12:17 +0000)]
fix build with DEBUG_PK enabled

OK dtucker@

6 years agoavoid uninitialised variable use
jsg [Tue, 28 Aug 2018 09:35:35 +0000 (09:35 +0000)]
avoid uninitialised variable use
ok patrick@ kettenis@

6 years agoavoid uninitialised variable use in an error path
jsg [Tue, 28 Aug 2018 09:33:18 +0000 (09:33 +0000)]
avoid uninitialised variable use in an error path
ok kettenis@

6 years agoFix stat of path components - this was breaking chrome - because the
beck [Tue, 28 Aug 2018 02:51:55 +0000 (02:51 +0000)]
Fix stat of path components - this was breaking chrome - because the
final directory in a name unveil was not being added with UNVEIL_INSPECT
if it was not already unveiled. Now passes just added regress test for
this case

6 years agoAdd regress test of stat all the way to a nonexistant terminal
beck [Tue, 28 Aug 2018 02:49:47 +0000 (02:49 +0000)]
Add regress test of stat all the way to a nonexistant terminal

6 years agoCheck for SSL_write(3) error.
cheloha [Tue, 28 Aug 2018 02:14:22 +0000 (02:14 +0000)]
Check for SSL_write(3) error.

jsing@ notes that this is not a complete solution, as we don't
account for retries or partial writes, but that this is a step
in a right direction.

May want to revisit this later to provide a complete solution.

ok jsing@

6 years agosync
deraadt [Tue, 28 Aug 2018 01:23:34 +0000 (01:23 +0000)]
sync

6 years agoadd support for storing the time of day on OPL based machines.
dlg [Tue, 28 Aug 2018 00:00:42 +0000 (00:00 +0000)]
add support for storing the time of day on OPL based machines.

this let's me pull the date back from 2023 to 2018, and have it
stay like that after a reboot.

ok deraadt@

6 years agoReduce excessive right padding in tagged list heads.
schwarze [Mon, 27 Aug 2018 23:12:57 +0000 (23:12 +0000)]
Reduce excessive right padding in tagged list heads.
The 1.2em was an approximate equivalent of the 2n traditionally used
for terminal display, but it is much too wide for HTML rendering.
Issue reported by mikeb@.

6 years agohitemp(4)
kettenis [Mon, 27 Aug 2018 21:56:31 +0000 (21:56 +0000)]
hitemp(4)

6 years agotweak failure messages
tb [Mon, 27 Aug 2018 21:27:39 +0000 (21:27 +0000)]
tweak failure messages

6 years agodedup AES-CBC-PKCS5 encryption and decryption checks
tb [Mon, 27 Aug 2018 21:24:13 +0000 (21:24 +0000)]
dedup AES-CBC-PKCS5 encryption and decryption checks

6 years agohitemp(4)
kettenis [Mon, 27 Aug 2018 21:12:28 +0000 (21:12 +0000)]
hitemp(4)

6 years agoAdd hitemp(4), a driver for the temperature sensors on the HiSilicon Hi3660
kettenis [Mon, 27 Aug 2018 21:09:47 +0000 (21:09 +0000)]
Add hitemp(4), a driver for the temperature sensors on the HiSilicon Hi3660
and Hi3670 SoCs.

6 years ago2x missing "..."
tb [Mon, 27 Aug 2018 21:02:25 +0000 (21:02 +0000)]
2x missing "..."

6 years agoRun Wycheproof AES-CCM testvectors against libcrypto.
tb [Mon, 27 Aug 2018 20:58:02 +0000 (20:58 +0000)]
Run Wycheproof AES-CCM testvectors against libcrypto.

6 years agosort SEE ALSO;
jmc [Mon, 27 Aug 2018 20:36:49 +0000 (20:36 +0000)]
sort SEE ALSO;

6 years agoAdd 64-bit DMA support. Enable DMA on HiSilicon SoCs.
kettenis [Mon, 27 Aug 2018 20:13:16 +0000 (20:13 +0000)]
Add 64-bit DMA support.  Enable DMA on HiSilicon SoCs.

ok patrick@

6 years agorkdwusb(4)
kettenis [Mon, 27 Aug 2018 20:09:29 +0000 (20:09 +0000)]
rkdwusb(4)

6 years agohiclock(4), hidwusb(4) and hireset(4)
kettenis [Mon, 27 Aug 2018 20:05:56 +0000 (20:05 +0000)]
hiclock(4), hidwusb(4) and hireset(4)

6 years agoAdd hiclock(4). Make sure hireset(4) attaches early.
kettenis [Mon, 27 Aug 2018 20:05:06 +0000 (20:05 +0000)]
Add hiclock(4).  Make sure hireset(4) attaches early.

6 years agomove bwfm* at pci? to the wireless section
jcs [Mon, 27 Aug 2018 19:32:48 +0000 (19:32 +0000)]
move bwfm* at pci? to the wireless section

6 years agoIf the maximum for a community is defined via the large flag then
claudio [Mon, 27 Aug 2018 19:32:37 +0000 (19:32 +0000)]
If the maximum for a community is defined via the large flag then
this maximum should also be passed to strtonum() instead of UINT_MAX
or the error handling does not work.

6 years agon2s and l2n3 finally bite the dust!
jsing [Mon, 27 Aug 2018 17:11:32 +0000 (17:11 +0000)]
n2s and l2n3 finally bite the dust!

6 years agoConvert ssl3_get_cert_verify() to CBS and clean up somewhat.
jsing [Mon, 27 Aug 2018 17:04:34 +0000 (17:04 +0000)]
Convert ssl3_get_cert_verify() to CBS and clean up somewhat.

ok inoguchi@

6 years agoDedup DTLS header writing code and convert to CBB.
jsing [Mon, 27 Aug 2018 16:56:46 +0000 (16:56 +0000)]
Dedup DTLS header writing code and convert to CBB.

There are three versions of the DTLS header writing code, which primarily
differ by the fragment offset and fragment length values that differ.
Rework dtls1_write_message_header() such that it can be used in all three
cases and convert it to CBB in the process.

ok inoguchi@ tb@

6 years agoAdd some missing statics.
jsing [Mon, 27 Aug 2018 16:48:12 +0000 (16:48 +0000)]
Add some missing statics.

6 years agoSimplify new session ticket encoding/generation.
jsing [Mon, 27 Aug 2018 16:42:48 +0000 (16:42 +0000)]
Simplify new session ticket encoding/generation.

The original code did a crazy encode/malloc/encode/decode/modify/encode
dance, in order to encode a session in the form needed to encrypt then add
to a session ticket. By modifying the encoding functions slightly, we can
do this entire dance as a single encode.

Inspired by similar changes in BoringSSL.

ok inoguchi@ tb@

6 years agoRename struct kd -> kcov_dev. The terse name was initially fine since it wasn't
anton [Mon, 27 Aug 2018 15:57:39 +0000 (15:57 +0000)]
Rename struct kd -> kcov_dev. The terse name was initially fine since it wasn't
used outside of dev/kcov.c. Nowadays, struct proc includes a kcov pointer and it
therefore deserves a more descriptive name.

Prodded by visa@; ok deraadt@ visa@

6 years agoEnable hireset(4).
kettenis [Mon, 27 Aug 2018 15:56:22 +0000 (15:56 +0000)]
Enable hireset(4).

6 years agoFix formatting and grammatical issues with the description of how to use
jsing [Mon, 27 Aug 2018 15:42:39 +0000 (15:42 +0000)]
Fix formatting and grammatical issues with the description of how to use
i2d_SSL_SESSION. Also rework the example code so that it is clearer and
uses more appropriate names.

Input from and ok schwarze@, tb@

6 years agoAdd hirest(4), a driver to support reset signal controller blocks on
kettenis [Mon, 27 Aug 2018 14:12:59 +0000 (14:12 +0000)]
Add hirest(4), a driver to support reset signal controller blocks on
HiSilicon SoCs.

6 years agoAdd glue for the USB3 controller on the HiKey 970.
kettenis [Mon, 27 Aug 2018 13:56:11 +0000 (13:56 +0000)]
Add glue for the USB3 controller on the HiKey 970.

6 years agodisklabel(8) no longer touches bootstrap code so stop mentioning it.
krw [Mon, 27 Aug 2018 13:09:51 +0000 (13:09 +0000)]
disklabel(8) no longer touches bootstrap code so stop mentioning it.
The MD writedisklabel() function ensures the disklabel does not
intrude on bootstraps.

Add a Xr to installboot(8), which does bootstrappy stuff these days.

ok jmc@

6 years agoUndefined plans can never match. Check for that before trying to use the plan.
claudio [Mon, 27 Aug 2018 12:15:20 +0000 (12:15 +0000)]
Undefined plans can never match. Check for that before trying to use the plan.
OK millert@ gsoares@

6 years agoWrap some overly long lines. No functional change
claudio [Mon, 27 Aug 2018 11:43:36 +0000 (11:43 +0000)]
Wrap some overly long lines. No functional change

6 years agoMemory leaks, from Gang Fan in GitHub issue 1453.
nicm [Mon, 27 Aug 2018 11:03:34 +0000 (11:03 +0000)]
Memory leaks, from Gang Fan in GitHub issue 1453.

6 years agoompinmux has been replaced by pinctrl
jsg [Mon, 27 Aug 2018 10:17:35 +0000 (10:17 +0000)]
ompinmux has been replaced by pinctrl

6 years agoswitch from ompinmux to pinctrl
jsg [Mon, 27 Aug 2018 10:05:18 +0000 (10:05 +0000)]
switch from ompinmux to pinctrl
ok kettenis@

6 years agohandle 16 bit register width required for omap3/omap4
jsg [Mon, 27 Aug 2018 10:03:35 +0000 (10:03 +0000)]
handle 16 bit register width required for omap3/omap4
ok kettenis@

6 years agoEnable radeondrm(4). Note that on arm64 we do not yet have a way to
kettenis [Mon, 27 Aug 2018 10:03:26 +0000 (10:03 +0000)]
Enable radeondrm(4).  Note that on arm64 we do not yet have a way to
determine whether radeondrm(4) is the console or not.  So leave out the
bits that try to make sure the console ends up with drm0/wsdisplay0 that
we have on other platforms.

6 years agomerge_filterset() needs to produce a stable sorted filterset to make sure
claudio [Mon, 27 Aug 2018 09:49:00 +0000 (09:49 +0000)]
merge_filterset() needs to produce a stable sorted filterset to make sure
the RDE can compare the sets on reload and skip those that did not change.
For large communities the check is wrong and incomplete, replace it with
a simple memcmp() of the structs which will result in a stable order.
OK phessler@

6 years agoAdd an interface that allows drivers to claim a framebuffer and check
kettenis [Mon, 27 Aug 2018 09:30:07 +0000 (09:30 +0000)]
Add an interface that allows drivers to claim a framebuffer and check
whether another driver has already claimed a framebuffer.  Use this in
radeondrm(4) and simplefb(4) to prevent the latter from attaching if
radeondrm(4) is attached to the hardware that provides the framebuffer
set up by the firmware.

ok mlarkin@

6 years agoAdd protocol and cipher patterns in regress appstest.sh
inoguchi [Mon, 27 Aug 2018 06:50:13 +0000 (06:50 +0000)]
Add protocol and cipher patterns in regress appstest.sh

6 years agoHiSilicon SoCs pre-divide the clock by a factor 8.
kettenis [Mon, 27 Aug 2018 06:43:28 +0000 (06:43 +0000)]
HiSilicon SoCs pre-divide the clock by a factor 8.

6 years agoDo not add ::1 on non-default lo(4) as intended.
mpi [Sun, 26 Aug 2018 22:30:00 +0000 (22:30 +0000)]
Do not add ::1 on non-default lo(4) as intended.

Reported by and ok sthen@

6 years agosync
deraadt [Sun, 26 Aug 2018 22:23:22 +0000 (22:23 +0000)]
sync

6 years agodisklabel(8) has not used the primary/secondary boot files
krw [Sun, 26 Aug 2018 22:06:50 +0000 (22:06 +0000)]
disklabel(8) has not used the primary/secondary boot files
for years. Remove mention of same from 'FILES' section.

6 years agosyncd
deraadt [Sun, 26 Aug 2018 20:44:03 +0000 (20:44 +0000)]
syncd

6 years agoless uppercase in Nd;
jmc [Sun, 26 Aug 2018 20:36:57 +0000 (20:36 +0000)]
less uppercase in Nd;

6 years agopinctrl(4)
kettenis [Sun, 26 Aug 2018 19:54:39 +0000 (19:54 +0000)]
pinctrl(4)

6 years agoAdd pinctrl(4), a generic pin mux driver.
kettenis [Sun, 26 Aug 2018 19:50:08 +0000 (19:50 +0000)]
Add pinctrl(4), a generic pin mux driver.

ok patrick@

6 years agoHandle fork() errors. The Perl fork() wrapper returns undef rather
naddy [Sun, 26 Aug 2018 19:09:55 +0000 (19:09 +0000)]
Handle fork() errors.  The Perl fork() wrapper returns undef rather
than -1 on error.  ok jca@

6 years agoUnveil pflogd(8). Similiar to florian@'s recent changes to ifconfig(8),
brynet [Sun, 26 Aug 2018 18:26:51 +0000 (18:26 +0000)]
Unveil pflogd(8). Similiar to florian@'s recent changes to ifconfig(8),
the priviledged parent cannot be pledged due to certain ioctls, but we
can use unveil(2) to lock down its access to the filesystem.

To be able to use hostnames/dns in tcpdump-like filter expressions,
we unveil /etc/{resolv.conf,hosts,services} "r", kept in sync with the
kernel bypass for pledge("dns")

Additionally, we need to unveil /dev/bpf "r" and the output log file
"rwc".

The unpriviledged child is pledged "stdio recvfd" and thus does not need
any unveils.

With feedback/testing from florian@, deraadt@

ok florian@ deraadt@

6 years agopflogd(8): don't try to rename(2) broken/invalid pflog files, instead,
brynet [Sun, 26 Aug 2018 18:24:46 +0000 (18:24 +0000)]
pflogd(8): don't try to rename(2) broken/invalid pflog files, instead,
suspend logging until the log file has been moved out of the way, and
we have received either SIGHUP or SIGALRM.

ok florian@ deraadt@

6 years agoplgpio(4)
kettenis [Sun, 26 Aug 2018 18:10:16 +0000 (18:10 +0000)]
plgpio(4)

6 years agofix the same "an non" issue found by tb in EVP_EncryptInit.3;
jmc [Sun, 26 Aug 2018 18:04:54 +0000 (18:04 +0000)]
fix the same "an non" issue found by tb in EVP_EncryptInit.3;

6 years agoThe three -B forms of disklabel(8) were nuked years ago. So remove
krw [Sun, 26 Aug 2018 17:47:54 +0000 (17:47 +0000)]
The three -B forms of disklabel(8) were nuked years ago. So remove
stray reference to them.

6 years agoCheck return value of EVP_CipherInit_ex()
tb [Sun, 26 Aug 2018 17:43:39 +0000 (17:43 +0000)]
Check return value of EVP_CipherInit_ex()

6 years agoRun Wycheproof AES-CBC-PKCS5 testvectors against libcrypto.
tb [Sun, 26 Aug 2018 17:38:16 +0000 (17:38 +0000)]
Run Wycheproof AES-CBC-PKCS5 testvectors against libcrypto.

6 years agoRemove some redundant info from log.Fatalf
tb [Sun, 26 Aug 2018 17:35:40 +0000 (17:35 +0000)]
Remove some redundant info from log.Fatalf

6 years agosimplify returned value
tb [Sun, 26 Aug 2018 17:34:40 +0000 (17:34 +0000)]
simplify returned value

6 years agoSome of the functions in this manual need <openssl/dsa.h>, others
tb [Sun, 26 Aug 2018 17:03:32 +0000 (17:03 +0000)]
Some of the functions in this manual need <openssl/dsa.h>, others
need <openssl/x509.h>. The functions {d2i,i2d}_DSA_params_{bio,fp}(3)
were missing from the manual, so document them. The return values
of the i2d_* functions are left undocumented, as these still need to
be audited.

ok schwarze (lots of input and help as usual)

6 years agotypo: an nonce -> a nonce
tb [Sun, 26 Aug 2018 16:55:30 +0000 (16:55 +0000)]
typo: an nonce -> a nonce

6 years agoAdd plgpio(4), a driver for the ARM PrimeCell GPIO (PL061) peripheral.
kettenis [Sun, 26 Aug 2018 16:52:16 +0000 (16:52 +0000)]
Add plgpio(4), a driver for the ARM PrimeCell GPIO (PL061) peripheral.

ok jsg@, patrick@

6 years agoSupport nesting of elements with next-line scope.
schwarze [Sun, 26 Aug 2018 16:18:38 +0000 (16:18 +0000)]
Support nesting of elements with next-line scope.
For example, ksh93(1) needs this for .B\n.SM.

6 years agoAdd support for HiSilicon Hi3660 (Kirin960) and Hi3670 (Kirin970).
kettenis [Sun, 26 Aug 2018 14:12:47 +0000 (14:12 +0000)]
Add support for HiSilicon Hi3660 (Kirin960) and Hi3670 (Kirin970).

6 years agoStop s_server if error occurs in regress appstest.sh
inoguchi [Sun, 26 Aug 2018 13:28:13 +0000 (13:28 +0000)]
Stop s_server if error occurs in regress appstest.sh

6 years agoAdd q: format prefix to escape sh(1) special characters. Suggested by
nicm [Sun, 26 Aug 2018 09:28:42 +0000 (09:28 +0000)]
Add q: format prefix to escape sh(1) special characters. Suggested by
someone ages ago and then more recently in GitHub issue 1449.

6 years agoAdd regress for kcov. It will only run if /dev/kcov can be opened successfully.
anton [Sun, 26 Aug 2018 08:12:09 +0000 (08:12 +0000)]
Add regress for kcov. It will only run if /dev/kcov can be opened successfully.

6 years agoAdd test cases for private versus non-private futexes.
visa [Sun, 26 Aug 2018 06:50:30 +0000 (06:50 +0000)]
Add test cases for private versus non-private futexes.

While here, fix timeout handling in futex_twait(), and fix the return
value of functions awakener() and signaled().

OK mpi@

6 years agoreturn -1 from the interrupt handler if we didn't find any work to do
jmatthew [Sun, 26 Aug 2018 06:40:03 +0000 (06:40 +0000)]
return -1 from the interrupt handler if we didn't find any work to do

6 years agosync
deraadt [Sat, 25 Aug 2018 22:54:09 +0000 (22:54 +0000)]
sync

6 years agoAdd code to print the characteristics of the caches that can be discovered
kettenis [Sat, 25 Aug 2018 20:45:28 +0000 (20:45 +0000)]
Add code to print the characteristics of the caches that can be discovered
through the CLIDR_EL1 register.

ok patrick@

6 years agoAdd GPT support. Mostly copied from amd64.
kettenis [Sat, 25 Aug 2018 20:43:39 +0000 (20:43 +0000)]
Add GPT support.  Mostly copied from amd64.

ok krw@

6 years agoAdd umt(4) for USB Windows Precision Touchpad devices
jcs [Sat, 25 Aug 2018 20:31:31 +0000 (20:31 +0000)]
Add umt(4) for USB Windows Precision Touchpad devices

Based on imt(4)

Rename HIDMT_INPUT_MODE_MT to HIDMT_INPUT_MODE_MT_TOUCHPAD

ok deraadt

6 years agoremove unused header
gilles [Sat, 25 Aug 2018 19:05:23 +0000 (19:05 +0000)]
remove unused header

from Freddy Dissaux

6 years agoUse __HAVE_ACPI to decide whether ACPI support should be considered.
kettenis [Sat, 25 Aug 2018 18:42:43 +0000 (18:42 +0000)]
Use __HAVE_ACPI to decide whether ACPI support should be considered.

ok deraadt@, krw@, jca@

6 years agoMove HID->bus constant conversion for HID report types out of ihidev
jcs [Sat, 25 Aug 2018 18:32:05 +0000 (18:32 +0000)]
Move HID->bus constant conversion for HID report types out of ihidev
into hidmt.

The HID code uses hid_feature, hid_input, and hid_output constants
to refer to report types internally that then need to be converted
to their bus-level counterparts before actually getting sent out (so
hid_feature becomes UHID_FEATURE_REPORT for USB,
I2C_HID_REPORT_TYPE_FEATURE for i2c).

This conversion was hard-coded in ihidev but ihidev_[gs]et_report
should assume the type passed is already an i2c-level define, not a
hid one.  This is how uhidev does it.

Add a conversion routine callback that any hidmt callers need to set
so that hidmt can convert hid constants to the bus-level versions.

Also add a similar conversion function to uhidev.

ok deraadt