openbsd
8 years agoOnly probe the UART type if sc_uarttype is set to COM_UART_UNKNOWN.
kettenis [Tue, 16 Aug 2016 08:15:04 +0000 (08:15 +0000)]
Only probe the UART type if sc_uarttype is set to COM_UART_UNKNOWN.
This allows glue drivers to set the type based on other information (such as
Oopen Firmware or device tree properties) without going through to risky
code paths.

ok deraadt@, visa@

8 years agoshorten genrsa; ok jsing
jmc [Tue, 16 Aug 2016 06:40:24 +0000 (06:40 +0000)]
shorten genrsa; ok jsing

8 years agotweak previous;
jmc [Tue, 16 Aug 2016 06:39:46 +0000 (06:39 +0000)]
tweak previous;

8 years agoadd a special edition doas that only lets root drop privs.
tedu [Tue, 16 Aug 2016 04:55:33 +0000 (04:55 +0000)]
add a special edition doas that only lets root drop privs.
maybe the installer can use something like this, wink wink.
ok deraadt

8 years agofread hasn't taken a char * in about 900 years
tedu [Tue, 16 Aug 2016 04:48:04 +0000 (04:48 +0000)]
fread hasn't taken a char * in about 900 years

8 years agoremove unneeded casts
tedu [Tue, 16 Aug 2016 04:44:38 +0000 (04:44 +0000)]
remove unneeded casts

8 years agoRewrite: ALL HAIL ELF
guenther [Tue, 16 Aug 2016 04:14:34 +0000 (04:14 +0000)]
Rewrite: ALL HAIL ELF
Mention the KERN_NOSUIDCOREDUMP sysctl(3)

ok deraadt@

8 years agoWhen running 'ifconfig scan' in hostap mode display the current Tx rate our
stsp [Mon, 15 Aug 2016 22:16:46 +0000 (22:16 +0000)]
When running 'ifconfig scan' in hostap mode display the current Tx rate our
AP is using to send frames to an associated node. This used to always display
the node's highest supported Rx rate, which isn't all that interesting.
ok mpi@

8 years agoExpose more 802.11n information to userspace:
stsp [Mon, 15 Aug 2016 22:14:19 +0000 (22:14 +0000)]
Expose more 802.11n information to userspace:
A flag which indicates whether HT has been negotiated with a node,
and the current Tx MCS value we use for a node.

This grows struct ieee80211_nodereq. Applications using it must be recompiled.

ok mpi@

8 years agonormalize some comments
tedu [Mon, 15 Aug 2016 22:01:59 +0000 (22:01 +0000)]
normalize some comments

8 years agoDon't take shortcuts cleaning/invalidating the caches. The Cortex-A7 and most
kettenis [Mon, 15 Aug 2016 21:08:56 +0000 (21:08 +0000)]
Don't take shortcuts cleaning/invalidating the caches.  The Cortex-A7 and most
other ARMv7 CPUs have a L2 cache that is larger than 32KB.  And some of those
even have L1 caches larger than that.  So bailing out after 32KB will leave
the cache partially dirty.

Fixes the SATA problems on Allwinner A20 based boards.

ok patrick@

8 years agoNow that com(4) uses a different iot for the initial console and the
patrick [Mon, 15 Aug 2016 21:04:32 +0000 (21:04 +0000)]
Now that com(4) uses a different iot for the initial console and the
main attachment, the serial port is now longer recognized as console.
To fix this, store the OFW node of the initial console and check it
in the attachment driver.

ok kettenis@

8 years agoFirst stab at porting the awinmmc(4) driver from NetBSD. The driver will
kettenis [Mon, 15 Aug 2016 21:03:27 +0000 (21:03 +0000)]
First stab at porting the awinmmc(4) driver from NetBSD.  The driver will
be known as sximmc(4) and is somewhat functional.  Still needs proper clock
support and some further cleanup.

8 years agosync
sthen [Mon, 15 Aug 2016 19:31:57 +0000 (19:31 +0000)]
sync

8 years agoFor some reason the RTL8211E PHY on the Banana Pi responds to both address 0
kettenis [Mon, 15 Aug 2016 18:31:28 +0000 (18:31 +0000)]
For some reason the RTL8211E PHY on the Banana Pi responds to both address 0
and 1.  As a result rgephy(4) attaches twice.  Prevent this from happening
by passing the PHY address, which is part of the device tree, down from the
bus-specific glue to the generic driver code.

8 years agoFix typo -> s/drirect/direct/.
mglocker [Mon, 15 Aug 2016 17:49:43 +0000 (17:49 +0000)]
Fix typo -> s/drirect/direct/.

From Daniel Bolgheroni

8 years ago%*qd -> %*lld + (long long) for off_t.
krw [Mon, 15 Aug 2016 16:57:53 +0000 (16:57 +0000)]
%*qd -> %*lld + (long long) for off_t.

Avoid some casts by changing printsize() to take an int (the only
type actually passed via that parameter) instead of size_t.

Tweaks & ok guenther@

8 years agoMove server_match() from parse.y to server.c; use env instead of conf,
jsing [Mon, 15 Aug 2016 16:12:34 +0000 (16:12 +0000)]
Move server_match() from parse.y to server.c; use env instead of conf,
which is actually the same thing (cluebat from reyk@).

8 years agoThe tls_conninfo serial is also unused.
jsing [Mon, 15 Aug 2016 15:44:58 +0000 (15:44 +0000)]
The tls_conninfo serial is also unused.

8 years agoGroup conninfo fields by connection and peer cert based information,
jsing [Mon, 15 Aug 2016 15:41:50 +0000 (15:41 +0000)]
Group conninfo fields by connection and peer cert based information,
sort and remove unused fingerprint.

8 years agoFix some style(9) issues.
jsing [Mon, 15 Aug 2016 14:47:41 +0000 (14:47 +0000)]
Fix some style(9) issues.

8 years agoAllow com(4) to make use of the simplebus address translation.
patrick [Mon, 15 Aug 2016 14:17:34 +0000 (14:17 +0000)]
Allow com(4) to make use of the simplebus address translation.
This typically works for every simplebus client, but com(4) did not
make use of the supplied bus tag.  Instead it references the global
a4x bus tag.  This is needed as the ARM com(4) controllers use 4-byte
spaced registers, while com(4) operates on 1-byte.  To be able to make
use of the address translation, copy the a4x bus tag and replace the
cookie and map function with the one supplied by simplebus.

ok kettenis@

8 years agoUse lowercase 'tls' in debug and log messages for consistency.
jsing [Mon, 15 Aug 2016 14:14:55 +0000 (14:14 +0000)]
Use lowercase 'tls' in debug and log messages for consistency.

Requested by reyk@

8 years agoExplicitly pass in an SSL_CTX * to the functions that operate on one,
jsing [Mon, 15 Aug 2016 14:04:23 +0000 (14:04 +0000)]
Explicitly pass in an SSL_CTX * to the functions that operate on one,
instead of assuming that they should use the one associated with the TLS
context. This allows these functions to be used with the additional
SSL contexts that are needed to support server-side SNI.

Also rename tls_configure_keypair() to tls_configure_ssl_keypair(), so that
these functions have a common prefix.

ok reyk@

8 years agoMake httpd stricter with respect to TLS configuration - in particular, do
jsing [Mon, 15 Aug 2016 13:48:24 +0000 (13:48 +0000)]
Make httpd stricter with respect to TLS configuration - in particular, do
not allow TLS and non-TLS to be configured on the same port, do not allow
TLS options to be specified without a TLS listener and ensure that the TLS
options are the same when a server is specified on the same address/port.
Currently, these configurations are permitted but do not work as intended.

Also factor out and reuse the server matching code, which was previously
duplicated.

ok reyk@

8 years agoMove the com(4) fdt attachment driver to a more generic location and
patrick [Mon, 15 Aug 2016 13:42:49 +0000 (13:42 +0000)]
Move the com(4) fdt attachment driver to a more generic location and
rename it to fit the typical scheme of such an attachment driver.

ok kettenis@ jsg@

8 years agoRemove IWM_CMD_SYNC, which expands to zero while all the other
stsp [Mon, 15 Aug 2016 12:59:53 +0000 (12:59 +0000)]
Remove IWM_CMD_SYNC, which expands to zero while all the other
IWM_CMD_* values are bit flags. The same was done 2 years ago
in Linux git commit a10229271946731959b2269370d0492d88cfab23.

8 years agoCatch up with the SSH1 code removal and delete all mention of
naddy [Mon, 15 Aug 2016 12:32:04 +0000 (12:32 +0000)]
Catch up with the SSH1 code removal and delete all mention of
protocol 1 particularities, key files and formats, command line
options, and configuration keywords from the server documentation
and examples.  ok jmc@

8 years agoRemove more SSH1 server code:
naddy [Mon, 15 Aug 2016 12:27:56 +0000 (12:27 +0000)]
Remove more SSH1 server code:
* Drop sshd's -k option.
* Retire configuration keywords that only apply to protocol 1, as well as
  the "protocol" keyword.
* Remove some related vestiges of protocol 1 support.

ok markus@

8 years agoreplace the last uses of m_copym2 with m_dup_pkt.
dlg [Mon, 15 Aug 2016 11:35:25 +0000 (11:35 +0000)]
replace the last uses of m_copym2 with m_dup_pkt.

ok mpi@ visa@

8 years agoUse pinctrl to configure the pins used by the EMAC.
kettenis [Mon, 15 Aug 2016 09:20:47 +0000 (09:20 +0000)]
Use pinctrl to configure the pins used by the EMAC.

ok jsg@

8 years agoUse the regulator API to supply power to the target.
kettenis [Mon, 15 Aug 2016 09:16:26 +0000 (09:16 +0000)]
Use the regulator API to supply power to the target.

Tested by Daniel Bolgheroni on a Cubietech Cubieboard2.

ok jsg@

8 years agoChecking for RTF_BROADCAST here makes no sense.
mpi [Mon, 15 Aug 2016 08:52:03 +0000 (08:52 +0000)]
Checking for RTF_BROADCAST here makes no sense.

8 years agoNeuter fuse_vptofh() and fuse_fhtovp(). I implemented those functions
natano [Mon, 15 Aug 2016 07:39:46 +0000 (07:39 +0000)]
Neuter fuse_vptofh() and fuse_fhtovp(). I implemented those functions
under the assumption, that fuse_vget() has reasonable semantics, while
this is not the case. fusefs_vget() only functions correctly, if the
file in questions has recently been accessed and is still in the vnode
cache of the userspace daemon associated with the mount point.

As a matter of fact the fuse api doesn't feature a reasonable way to map
inode numbers to a handle at all (see struct fuse_operations).

ok guenther

8 years agoNo need to reset si_selpid after calling selwakeup() the function
mpi [Mon, 15 Aug 2016 07:20:14 +0000 (07:20 +0000)]
No need to reset si_selpid after calling selwakeup() the function
already does it.

8 years agoadd a bit of spacing to previous, to keep the notes about deprecated
jmc [Mon, 15 Aug 2016 07:18:44 +0000 (07:18 +0000)]
add a bit of spacing to previous, to keep the notes about deprecated
functions out the way of the main body;

ok guenther

8 years agoIntroduce bpf_put() and bpf_get() instead of mixing macro and functions
mpi [Mon, 15 Aug 2016 07:17:10 +0000 (07:17 +0000)]
Introduce bpf_put() and bpf_get() instead of mixing macro and functions
for the reference counting.

ok dlg@

8 years agoCheck if ``bd_bif'' is NULL inside bpf_catchpacket() to match bpfread()
mpi [Mon, 15 Aug 2016 07:12:11 +0000 (07:12 +0000)]
Check if ``bd_bif'' is NULL inside bpf_catchpacket() to match bpfread()
and bpfwrite(), all of which will need to grabe a lock to protect the
buffers.

ok dlg@

8 years agoMerge bpfilter_create() into bpfopen() and make it such that the
mpi [Mon, 15 Aug 2016 07:03:47 +0000 (07:03 +0000)]
Merge bpfilter_create() into bpfopen() and make it such that the
descriptor is referenced before it is inserted in the global list.

ok dlg@

8 years agoDrain FIFOs on init same as com(4) does it and reduce the RX FIFO interrupt
mglocker [Mon, 15 Aug 2016 04:32:57 +0000 (04:32 +0000)]
Drain FIFOs on init same as com(4) does it and reduce the RX FIFO interrupt
trigger to 1 character.  This fixes a hang seen on the allwinner,sun5i-r8
board during boot when the driver switches to interrupts and no terminal
is attached.

ok kettenis

8 years agoTrust the login_getcaptime() declaration and don't cast the arguments to
guenther [Mon, 15 Aug 2016 04:28:31 +0000 (04:28 +0000)]
Trust the login_getcaptime() declaration and don't cast the arguments to
their own expected type

8 years agoDrop -z nocombreloc. It was a rev 1.1 commit like on arm, so not clear
guenther [Mon, 15 Aug 2016 03:09:02 +0000 (03:09 +0000)]
Drop -z nocombreloc.  It was a rev 1.1 commit like on arm, so not clear
why it was needed then; perhaps fixed by the bump in binutils?
Removing it enables the full relro layout for ld.so.

8 years agosync
deraadt [Mon, 15 Aug 2016 02:03:56 +0000 (02:03 +0000)]
sync

8 years agoReduce qabs() and qdiv() to aliases of llabs() and lldiv().
guenther [Sun, 14 Aug 2016 23:18:03 +0000 (23:18 +0000)]
Reduce qabs() and qdiv() to aliases of llabs() and lldiv().
Merge the manual pages and call them deprecated there.

ok and manpage tweak jmc@, ok natano@

8 years agoConvert quad_t to long long
guenther [Sun, 14 Aug 2016 23:08:09 +0000 (23:08 +0000)]
Convert quad_t to long long

8 years agoConvert %q to %ll
guenther [Sun, 14 Aug 2016 23:01:13 +0000 (23:01 +0000)]
Convert %q to %ll

8 years agoWhen upcasting to off_t for multiplication, the cast needs to be *before*
guenther [Sun, 14 Aug 2016 22:57:31 +0000 (22:57 +0000)]
When upcasting to off_t for multiplication, the cast needs to be *before*
the multiplication to avoid truncation/overflow

ok krw@

8 years agoConvert %q to %ll with long long casts for printf()
guenther [Sun, 14 Aug 2016 22:56:29 +0000 (22:56 +0000)]
Convert %q to %ll with long long casts for printf()
Delete pointless casts to off_t, void*, and uid_t

ok krw@

8 years agoUse int64_t and %ll for CTLTYPE_QUAD nodes.
guenther [Sun, 14 Aug 2016 22:54:56 +0000 (22:54 +0000)]
Use int64_t and %ll for CTLTYPE_QUAD nodes.
Use memcpy() instead of up-casting from char* to long long*

ok krw@

8 years agoConvert %q to %ll and cast opaque system values to long long.
guenther [Sun, 14 Aug 2016 22:47:26 +0000 (22:47 +0000)]
Convert %q to %ll and cast opaque system values to long long.
Use uint64_t for the KERN_INTRCNT_CNT sysctl()

ok krw@

8 years agoConvert %q to %ll. Cast off_t to long long for printf-style args.
guenther [Sun, 14 Aug 2016 22:35:54 +0000 (22:35 +0000)]
Convert %q to %ll.  Cast off_t to long long for printf-style args.
Eliminate pointless casts to void* or off_t

ok krw@

8 years agoFlip 'u_quad_t' fields & variables to uint64_t, and %qu to %llu.
krw [Sun, 14 Aug 2016 22:29:01 +0000 (22:29 +0000)]
Flip 'u_quad_t' fields & variables to uint64_t, and %qu to %llu.

ok guenther@

8 years agoKill '#if defined(DEBUG) && 0' blocks that used %q
guenther [Sun, 14 Aug 2016 21:47:16 +0000 (21:47 +0000)]
Kill '#if defined(DEBUG) && 0' blocks that used %q

ok bentley@ on principle

8 years ago%qd -> %lld for an off_t already being cast to (long long)!
krw [Sun, 14 Aug 2016 21:13:08 +0000 (21:13 +0000)]
%qd -> %lld for an off_t already being cast to (long long)!

8 years agoAnother %qd -> %lld + (long long) for off_t's.
krw [Sun, 14 Aug 2016 21:07:40 +0000 (21:07 +0000)]
Another %qd -> %lld + (long long) for off_t's.

8 years agosync
deraadt [Sun, 14 Aug 2016 20:59:52 +0000 (20:59 +0000)]
sync

8 years ago%qd -> %lld + (long long) for off_t's.
krw [Sun, 14 Aug 2016 20:53:19 +0000 (20:53 +0000)]
%qd -> %lld + (long long) for off_t's.

ok guenther@ natano@

8 years agoDon't reinvent rlim_t; when printing it use %llu and cast to
guenther [Sun, 14 Aug 2016 19:46:31 +0000 (19:46 +0000)]
Don't reinvent rlim_t; when printing it use %llu and cast to
unsigned long long

ok natano@ tedu@

8 years agoDon't call sysconf() in a loop conditional when the loop won't affect it
guenther [Sun, 14 Aug 2016 19:45:24 +0000 (19:45 +0000)]
Don't call sysconf() in a loop conditional when the loop won't affect it

ok natano@ tedu@

8 years agotweak previous;
jmc [Sun, 14 Aug 2016 19:32:43 +0000 (19:32 +0000)]
tweak previous;

8 years agoimxtemp(4)
kettenis [Sun, 14 Aug 2016 19:15:22 +0000 (19:15 +0000)]
imxtemp(4)

8 years agoAdd imxtemp(4), a temperature sensor for the i.MX6 SoC. Based on code written
kettenis [Sun, 14 Aug 2016 19:08:44 +0000 (19:08 +0000)]
Add imxtemp(4), a temperature sensor for the i.MX6 SoC.  Based on code written
by patrick@.

ok deraadt@

8 years agoDrop -z nocombreloc. That's a rev 1.1 commit so not clear why it was needed
guenther [Sun, 14 Aug 2016 18:46:03 +0000 (18:46 +0000)]
Drop -z nocombreloc.  That's a rev 1.1 commit so not clear why it was needed
then; perhaps fixed by the bump in binutils?  Anyway, builds seem good without
it and that enables the full relro layout for ld.so.

ok kettenis@

8 years agoConvert remaining calls to strtoq/strtouq in base with strtoll/strtoull.
guenther [Sun, 14 Aug 2016 18:34:48 +0000 (18:34 +0000)]
Convert remaining calls to strtoq/strtouq in base with strtoll/strtoull.
Fix a type mismatch in ftp's "page" command and could make transfers restart
at the wrong position.

ok and a ull->ll tweak from natano@, ok tedu@

8 years agoRemove many unnecessary casts. Verified by comparing generated code on
guenther [Sun, 14 Aug 2016 18:30:33 +0000 (18:30 +0000)]
Remove many unnecessary casts.  Verified by comparing generated code on
both ILP32 and LP64.

ok millert@

8 years agoRefer to /etc/passwd consistently as the "legacy password file" and
tb [Sun, 14 Aug 2016 14:57:15 +0000 (14:57 +0000)]
Refer to /etc/passwd consistently as the "legacy password file" and
remove some references to differences between versions 6 and 7.

ok jmc, millert, tedu

8 years agoMerge identical amd64 and i386 stanzas.
krw [Sun, 14 Aug 2016 13:47:52 +0000 (13:47 +0000)]
Merge identical amd64 and i386 stanzas.

Requested & ok jsing@

8 years agoreplace strtoq with strtoll. ok openbsd
tedu [Sun, 14 Aug 2016 13:40:55 +0000 (13:40 +0000)]
replace strtoq with strtoll. ok openbsd

8 years agoremove uneeded sa1111_reg.h include
jsg [Sun, 14 Aug 2016 13:29:22 +0000 (13:29 +0000)]
remove uneeded sa1111_reg.h include
ok kettenis@

8 years agoRemove code for Intel 80219/80321 xscale processors used by armish.
jsg [Sun, 14 Aug 2016 11:30:54 +0000 (11:30 +0000)]
Remove code for Intel 80219/80321 xscale processors used by armish.
Generic xscale support and support for pxa2x0 used by zaurus remains.

8 years agoFix setting the SMP bit in the Auxiliary Control Register. The old code was
kettenis [Sun, 14 Aug 2016 10:36:47 +0000 (10:36 +0000)]
Fix setting the SMP bit in the Auxiliary Control Register.  The old code was
toggling the bit, clearing it when already set.  On Cortex-A7 setting the SMP
bit is essential since without it the CPU doesn't actually use its caches.

The SMP bit supposed to be set before turning on the caches and the MMU, so
move the setting of the Auxiliary Control Register before setting the
System Control Register.

ok jsg@

8 years agoAllow a bus_dmamap_sync() of length zero.
kettenis [Sun, 14 Aug 2016 10:32:17 +0000 (10:32 +0000)]
Allow a bus_dmamap_sync() of length zero.

ok tedu@

8 years agoXr encrypt(1) in the explanation of the -p option.
tb [Sun, 14 Aug 2016 09:52:08 +0000 (09:52 +0000)]
Xr encrypt(1) in the explanation of the -p option.

ok jmc

8 years agoshorten genpkey; ok guenther
jmc [Sun, 14 Aug 2016 09:02:52 +0000 (09:02 +0000)]
shorten genpkey; ok guenther

8 years agoRemove some annoying debug printfs to get rid of u_quad_t in cnmac.
visa [Sun, 14 Aug 2016 08:49:37 +0000 (08:49 +0000)]
Remove some annoying debug printfs to get rid of u_quad_t in cnmac.

8 years agoUtilize the TLB Execute-Inhibit bit with non-executable mappings on CPUs
visa [Sun, 14 Aug 2016 08:23:52 +0000 (08:23 +0000)]
Utilize the TLB Execute-Inhibit bit with non-executable mappings on CPUs
that support the Execute-Inhibit exception. This makes user space W^X
effective on Octeon Plus and later Octeon versions.

Feedback from miod@, thanks!
No objection from deraadt@

8 years agou_quad_t -> uint64_t
dlg [Sun, 14 Aug 2016 04:51:29 +0000 (04:51 +0000)]
u_quad_t -> uint64_t

de is amazing(tm). it really should be cut back and simplified a bit.

8 years agoReplace u_quad_t with unsigned long long and replace "uqd" with "ull" in
guenther [Sun, 14 Aug 2016 04:47:52 +0000 (04:47 +0000)]
Replace u_quad_t with unsigned long long and replace "uqd" with "ull" in
function names to match.  Pull some tangled assignments out of conditions
and use >>= where possible.

ok millert@

8 years agoMark the relro/got section as read-only before setting up the environment
guenther [Sun, 14 Aug 2016 04:30:39 +0000 (04:30 +0000)]
Mark the relro/got section as read-only before setting up the environment
to minimize the amount of code run without it RO.

8 years agoTurning off relro shouldn't insert padding into the middle of the relro
guenther [Sun, 14 Aug 2016 04:25:27 +0000 (04:25 +0000)]
Turning off relro shouldn't insert padding into the middle of the relro
section; remove the PAD_GOT0 use that I should have deleted before.  This
caused ld.so crashes on arm when building binutils.

critical debugging observations by kettenis@

8 years agochange some types in bio from u_quad_t to uint64_t, and fix casts in
dlg [Sun, 14 Aug 2016 04:08:03 +0000 (04:08 +0000)]
change some types in bio from u_quad_t to uint64_t, and fix casts in
drivers that fill that field in too.

quad types are going away.

8 years agoreplace u_quad_t with uint64_t.
dlg [Sun, 14 Aug 2016 01:52:47 +0000 (01:52 +0000)]
replace u_quad_t with uint64_t.

guenther is proposing the removal of u_quad_t. uint64_t is more portable.

tested on a perc5 and perc6, no functional change.

8 years agouse -nosplit for AUTHORS: looks much better, to the point where i wonder why we don...
jmc [Sat, 13 Aug 2016 23:35:00 +0000 (23:35 +0000)]
use -nosplit for AUTHORS: looks much better, to the point where i wonder why we don;t just enforce this...

8 years agoAdd the shim to make dwge(4) attach to simplebus(4). This brings us
kettenis [Sat, 13 Aug 2016 22:07:01 +0000 (22:07 +0000)]
Add the shim to make dwge(4) attach to simplebus(4).  This brings us
working gigabit on the Allwinner A20.  Probably won't work yet on other
Allwinner SoCs due to differences in how the clocks get set up.

Based on an earlier diff from patrick@.

8 years agoAdd GMAC clock support. Based on an earlier diff from patrick@.
kettenis [Sat, 13 Aug 2016 21:48:44 +0000 (21:48 +0000)]
Add GMAC clock support.  Based on an earlier diff from patrick@.

8 years agoMissed a couple qaddr_t casts
guenther [Sat, 13 Aug 2016 21:28:09 +0000 (21:28 +0000)]
Missed a couple qaddr_t casts

8 years agoPassing in &_DYNAMIC appears to have stuck, so eliminate the extra
guenther [Sat, 13 Aug 2016 20:57:04 +0000 (20:57 +0000)]
Passing in &_DYNAMIC appears to have stuck, so eliminate the extra
variable

8 years agoEliminate pointless casts to qaddr_t of a value being assigned to a void*
guenther [Sat, 13 Aug 2016 20:53:17 +0000 (20:53 +0000)]
Eliminate pointless casts to qaddr_t of a value being assigned to a void*

ok kettenis@ krw@ natano@ dlg@ espie@

8 years agodwge(4)
kettenis [Sat, 13 Aug 2016 20:51:48 +0000 (20:51 +0000)]
dwge(4)

8 years agoAdd the core of the code for dwge(4), a driver for the Synopsis Designware
kettenis [Sat, 13 Aug 2016 20:35:57 +0000 (20:35 +0000)]
Add the core of the code for dwge(4), a driver for the Synopsis Designware
GMAC core which is used on the Allwinner A20 SoCs and later SoCs in the
Allwinner family.  Since this core is very likely to show up the base of
other devices, it gets to live here in dev/ic.

Ported from NetBSD (where it is named awge(4)) by patrick@.  Some further
fixes by me.

ok deraadt@

8 years agoshorten gendsa;
jmc [Sat, 13 Aug 2016 18:23:39 +0000 (18:23 +0000)]
shorten gendsa;

8 years ago%lld -> %llu for u_int64_t struct fields.
krw [Sat, 13 Aug 2016 17:51:25 +0000 (17:51 +0000)]
%lld -> %llu for u_int64_t struct fields.

8 years agoremove ssh1 server code; ok djm@
markus [Sat, 13 Aug 2016 17:47:40 +0000 (17:47 +0000)]
remove ssh1 server code; ok djm@

8 years agoAs the relevant struct's (e.g. tcpstat) use [u_]int64_t types, the
krw [Sat, 13 Aug 2016 17:36:04 +0000 (17:36 +0000)]
As the relevant struct's (e.g. tcpstat) use [u_]int64_t types, the
proper printf() format is %ll[ud], not the antediluvian %q[ud].
Reminded by guenther@'s proposed retirement of quad types.

ok tedu@

8 years agomodern interfaces should use modern speelings, so spell quad_t as int64_t.
tedu [Sat, 13 Aug 2016 17:05:02 +0000 (17:05 +0000)]
modern interfaces should use modern speelings, so spell quad_t as int64_t.

8 years agosync
deraadt [Sat, 13 Aug 2016 15:37:48 +0000 (15:37 +0000)]
sync

8 years agoRemove entries that are no longer needed.
kettenis [Sat, 13 Aug 2016 13:55:25 +0000 (13:55 +0000)]
Remove entries that are no longer needed.

8 years agoAvoid leaking memory if tls_config_set_alpn() is called multiple times
jsing [Sat, 13 Aug 2016 13:15:53 +0000 (13:15 +0000)]
Avoid leaking memory if tls_config_set_alpn() is called multiple times
(this was in the original commit, but got reverted in the recommit).

8 years agoLet libtls load the CA, certificate and key files for nc(1), now that it
jsing [Sat, 13 Aug 2016 13:09:10 +0000 (13:09 +0000)]
Let libtls load the CA, certificate and key files for nc(1), now that it
does this at the time the tls_config_set_*_file() function is called.

ok bluhm@

8 years agoLoad CA, certificate and key files into memory when the appropriate
jsing [Sat, 13 Aug 2016 13:05:51 +0000 (13:05 +0000)]
Load CA, certificate and key files into memory when the appropriate
tls_config_set_*_file() function is called. This allows us to immediately
propagate useful error messages, play more nicely with privsep/pledge and
have a single code path. Instead of always loading the default CA when
tls_config_new() is called, defer and only load the default CA when
tls_configure() is invoked, if a CA has not already been specified.

ok beck@ bluhm@