openbsd
8 years agoSpliced TCP sockets become faster when the output part is running
bluhm [Thu, 25 Aug 2016 14:13:19 +0000 (14:13 +0000)]
Spliced TCP sockets become faster when the output part is running
as its own task thread.  This is inspired by userland copy where a
process also has to go through the scheduler.  This gives the socket
buffer a chance to be filled up and tcp_output() is called less
often and with bigger chunks.
When two kernel tasks share all the workload, the current scheduler
implementation will hang userland processes on single cpu machines.
As a workaround put a yield() into the splicing thread after each
task execution.  This reduces the number of calls of tcp_output()
even more.
OK tedu@ mpi@

8 years agoCompletely revert the M_WAIT change on the cluster allocation and
bluhm [Thu, 25 Aug 2016 13:59:16 +0000 (13:59 +0000)]
Completely revert the M_WAIT change on the cluster allocation and
bring back the behaviour of rev 1.72.  Although allocating small
mbufs when allocating an mbuf cluster fails seems suboptimal, this
should not be changed as a side effect when introducing m_getuio().
OK claudio@

8 years agoSimplify ip6_getpmtu() to use a 'struct rtentry *' instead of two
mpi [Thu, 25 Aug 2016 12:30:16 +0000 (12:30 +0000)]
Simplify ip6_getpmtu() to use a 'struct rtentry *' instead of two
'struct route_in6 *'.

This is another little step towards deprecating 'struct route{,_in6}'

Inputs from and ok bluhm@

8 years agogoda@ contributed most of the OpenFlow 1.3.5 parts in ofp.h (for the
reyk [Thu, 25 Aug 2016 11:00:44 +0000 (11:00 +0000)]
goda@ contributed most of the OpenFlow 1.3.5 parts in ofp.h (for the
"still-unreleased" switch(4) driver), so it is only fair to add his
copyright.  This header file will eventually move to net/ofp.h.

8 years agoFix pfsync(4)'s carp examples: as of ip_carp.c r1.245, carpdev must be
sthen [Thu, 25 Aug 2016 10:51:37 +0000 (10:51 +0000)]
Fix pfsync(4)'s carp examples: as of ip_carp.c r1.245, carpdev must be
specified.  From Bryan Stenson.

8 years agoFix rtadvd.conf(5) example for pinfoflags, it accepts numbers not strings.
sthen [Thu, 25 Aug 2016 09:41:13 +0000 (09:41 +0000)]
Fix rtadvd.conf(5) example for pinfoflags, it accepts numbers not strings.
ok phessler

8 years agoDo not crash if display-message used without a client, issue reported by
nicm [Thu, 25 Aug 2016 09:33:14 +0000 (09:33 +0000)]
Do not crash if display-message used without a client, issue reported by
Serge Aleynikov, fix from Thomas Adam.

8 years agoEnable the UWXN bit in the SCTRL register when available. This should
kettenis [Thu, 25 Aug 2016 08:17:57 +0000 (08:17 +0000)]
Enable the UWXN bit in the SCTRL register when available.  This should
prevent the kernel from accidentally executing userland pages that are
writable.

ok jsg@, patrick@

8 years agoremove lint comments
tedu [Thu, 25 Aug 2016 05:25:21 +0000 (05:25 +0000)]
remove lint comments

8 years agoremove lint comments
tedu [Thu, 25 Aug 2016 05:23:19 +0000 (05:23 +0000)]
remove lint comments

8 years agodo not request RWX mappings, RW will do
deraadt [Thu, 25 Aug 2016 05:12:06 +0000 (05:12 +0000)]
do not request RWX mappings, RW will do

8 years agoReplace name_{uid,gid}() with the libc routines user_from_uid() and
guenther [Thu, 25 Aug 2016 01:44:55 +0000 (01:44 +0000)]
Replace name_{uid,gid}() with the libc routines user_from_uid() and
group_from_gid().  Eliminate some superfluous strncpy() calls.

ok millert@

8 years agopool_setipl for udf
dlg [Thu, 25 Aug 2016 00:06:44 +0000 (00:06 +0000)]
pool_setipl for udf

ok phessler@ krw@

8 years agopool_setipl
dlg [Thu, 25 Aug 2016 00:01:13 +0000 (00:01 +0000)]
pool_setipl

ok kettenis@

8 years agopool_setipl
dlg [Thu, 25 Aug 2016 00:00:02 +0000 (00:00 +0000)]
pool_setipl

ok kettenis@

8 years agoMove the pf address printing test into its own subdirectory. This
bluhm [Wed, 24 Aug 2016 22:31:41 +0000 (22:31 +0000)]
Move the pf address printing test into its own subdirectory.  This
makes it consistent with the regress tree structure.

8 years agoMake list-like output go to stderr when appending to an archive on stdout.
guenther [Wed, 24 Aug 2016 19:15:42 +0000 (19:15 +0000)]
Make list-like output go to stderr when appending to an archive on stdout.
Simplify the recognition of -f- and TAPE=- as meaning stdin/stdout.

ok millert@

8 years agoTreat cpio's -t option as a modifier to -i, so they're ordering independent
guenther [Wed, 24 Aug 2016 19:13:52 +0000 (19:13 +0000)]
Treat cpio's -t option as a modifier to -i, so they're ordering independent

ok millert@

8 years agoset the error indicator on malloc(3) failure;
schwarze [Wed, 24 Aug 2016 18:35:12 +0000 (18:35 +0000)]
set the error indicator on malloc(3) failure;
from Andrey Chernov <ache at freebsd dot org>;
OK millert@

8 years agoSince the removal of rwhod and friends a couple of years ago, there
renato [Wed, 24 Aug 2016 16:14:24 +0000 (16:14 +0000)]
Since the removal of rwhod and friends a couple of years ago, there
isn't any file in the base system including <protocols/rwhod.h>. So,
stop suggesting the inclusion of this header in this man page.

ok tedu@ deraadt@ mpi@

8 years agoUse writev(2) to write history records using a single syscall.
millert [Wed, 24 Aug 2016 16:09:40 +0000 (16:09 +0000)]
Use writev(2) to write history records using a single syscall.
OK deraadt@

8 years agoAvoid recursively calling c_fc(). Fixes a core dump from "r r" and
millert [Wed, 24 Aug 2016 13:32:17 +0000 (13:32 +0000)]
Avoid recursively calling c_fc().  Fixes a core dump from "r r" and
other edge cases found by gsoares@.  OK tb@ gsoares@

8 years agoReplace pmap_fault_fixup() with an access flag fault handler on armv7.
kettenis [Wed, 24 Aug 2016 13:09:52 +0000 (13:09 +0000)]
Replace pmap_fault_fixup() with an access flag fault handler on armv7.

ok tom@

8 years agopool_setipl for oce(4)
dlg [Wed, 24 Aug 2016 10:38:34 +0000 (10:38 +0000)]
pool_setipl for oce(4)

ok mikeb@

8 years agoKill ip6_forward_rt reducing differences between v4 and v6.
mpi [Wed, 24 Aug 2016 09:41:12 +0000 (09:41 +0000)]
Kill ip6_forward_rt reducing differences between v4 and v6.

A single forwarding cache is not the answer.  The answer is 42... err PF!

ok bluhm@

8 years agoUse rtalloc(9) directly instead of in6_selectsr() in icmp6_reflect().
mpi [Wed, 24 Aug 2016 09:38:29 +0000 (09:38 +0000)]
Use rtalloc(9) directly instead of in6_selectsr() in icmp6_reflect().

This is another little step towards deprecating 'struct route{,_in6}'.

ok bluhm@

8 years agopool_setipl for cardbus and drm pools.
dlg [Wed, 24 Aug 2016 09:31:56 +0000 (09:31 +0000)]
pool_setipl for cardbus and drm pools.

ok kettenis@

8 years agoshorten the pkeyparam text;
jmc [Wed, 24 Aug 2016 08:17:24 +0000 (08:17 +0000)]
shorten the pkeyparam text;

8 years agoshorten the pkey text;
jmc [Wed, 24 Aug 2016 08:07:33 +0000 (08:07 +0000)]
shorten the pkey text;

8 years agoConvert quad_t to int64_t and %q to %ll
guenther [Wed, 24 Aug 2016 03:13:45 +0000 (03:13 +0000)]
Convert quad_t to int64_t and %q to %ll
Convert bzero() to memset() and bcopy() to memcpy()

ok natano@ millert@

8 years agorename nfiles to numfiles to avoid shadowing and stretch out the name.
tedu [Tue, 23 Aug 2016 23:28:02 +0000 (23:28 +0000)]
rename nfiles to numfiles to avoid shadowing and stretch out the name.
ok deraadt

8 years agosync
deraadt [Tue, 23 Aug 2016 21:51:50 +0000 (21:51 +0000)]
sync

8 years agoDeassert all reset signals for the controller. Add support for more recent
kettenis [Tue, 23 Aug 2016 21:43:51 +0000 (21:43 +0000)]
Deassert all reset signals for the controller.  Add support for more recent
SoCs that have a separate clock for each PHY.

8 years agoAdd functions to assert/deassert all reset signals for a device.
kettenis [Tue, 23 Aug 2016 21:30:18 +0000 (21:30 +0000)]
Add functions to assert/deassert all reset signals for a device.

8 years agoSome syslogd tests failed to report errors. Make these tests more
bluhm [Tue, 23 Aug 2016 21:20:05 +0000 (21:20 +0000)]
Some syslogd tests failed to report errors.  Make these tests more
strict and adapt the check patterns.

8 years agoAdd support for the usb clock on sun5i-a13 as well.
kettenis [Tue, 23 Aug 2016 19:13:37 +0000 (19:13 +0000)]
Add support for the usb clock on sun5i-a13 as well.

8 years agoinclude acpicbkbd
jcs [Tue, 23 Aug 2016 18:58:53 +0000 (18:58 +0000)]
include acpicbkbd

8 years agoshorten pkcs12;
jmc [Tue, 23 Aug 2016 18:54:04 +0000 (18:54 +0000)]
shorten pkcs12;

8 years agoadd a man page for acpicbkbd
jcs [Tue, 23 Aug 2016 18:39:08 +0000 (18:39 +0000)]
add a man page for acpicbkbd

8 years agore-enable acpicbkbd by default now that acpiec won't try burst mode
jcs [Tue, 23 Aug 2016 18:27:08 +0000 (18:27 +0000)]
re-enable acpicbkbd by default now that acpiec won't try burst mode
on the chrome ec

8 years agodon't enter burst mode for single-byte reads and writes.
jcs [Tue, 23 Aug 2016 18:26:21 +0000 (18:26 +0000)]
don't enter burst mode for single-byte reads and writes.

avoids problems on hardware with broken or unimplemented burst mode
and isn't really necessary for such small transactions anyway.
matches what linux and freebsd have done for a long time.

tested in snaps
ok deraadt kettenis

8 years agoThe device trees for sun8i and sun9i no longer include an address in the
kettenis [Tue, 23 Aug 2016 18:16:06 +0000 (18:16 +0000)]
The device trees for sun8i and sun9i no longer include an address in the
name of the /soc node.  Leave it off, as the device path will still match
on nodes that include it.

8 years agoActually make fdt_find_node() return NULL if the node couldn't be found.
kettenis [Tue, 23 Aug 2016 18:12:09 +0000 (18:12 +0000)]
Actually make fdt_find_node() return NULL if the node couldn't be found.

ok tom@

8 years agoType. 'if' != 'lf'. Fixes error message and setting hostname from dhcp lease
krw [Tue, 23 Aug 2016 16:49:48 +0000 (16:49 +0000)]
Type. 'if' != 'lf'. Fixes error message and setting hostname from dhcp lease
during install.

Spotted & diff from Patrik Lundin. Thanks!

8 years agofix previous, a condition was modified incorrectly; ok markus@ deraadt@
otto [Tue, 23 Aug 2016 16:21:45 +0000 (16:21 +0000)]
fix previous, a condition was modified incorrectly; ok markus@ deraadt@

8 years agosync
okan [Tue, 23 Aug 2016 14:05:11 +0000 (14:05 +0000)]
sync

8 years agoUpdate the L2 content of a RTF_CACHED entry instead of going though
mpi [Tue, 23 Aug 2016 13:07:26 +0000 (13:07 +0000)]
Update the L2 content of a RTF_CACHED entry instead of going though
a create/delete/insert cycle as such entry cannot be deleted when
referenced.

Regression reported by and ok bluhm@

8 years agoThe TLB refill and XTLB refill exceptions use distinct exception vectors
visa [Tue, 23 Aug 2016 12:54:09 +0000 (12:54 +0000)]
The TLB refill and XTLB refill exceptions use distinct exception vectors
on Loongson 3A. Consequently, the kernel has to set up both vectors to
prevent panics with virtual address references.

On Loongson 2F, it is enough to set up the TLB refill vector because
the XTLB refill exception uses the same vector address.

ok miod@

8 years agopool_setipl
dlg [Tue, 23 Aug 2016 12:37:44 +0000 (12:37 +0000)]
pool_setipl

8 years agopool_setipl
dlg [Tue, 23 Aug 2016 12:37:11 +0000 (12:37 +0000)]
pool_setipl

8 years agoUse rtalloc(9) directly instead of in6_selectsr() in NS/NA output
mpi [Tue, 23 Aug 2016 11:03:10 +0000 (11:03 +0000)]
Use rtalloc(9) directly instead of in6_selectsr() in NS/NA output
routines.

This is another little step towards deprecating 'struct route{,_in6}'.

ok bluhm@

8 years agoDo not use a single global struct route_in6 to cache route lookups.
mpi [Tue, 23 Aug 2016 11:01:16 +0000 (11:01 +0000)]
Do not use a single global struct route_in6 to cache route lookups.

This is a little step towards deprecating 'struct route{,_in6}'.

ok jca@, claudio@

8 years agoMake the 'ifi' global local to dhclient.c and pass it as an argument to
mpi [Tue, 23 Aug 2016 09:26:02 +0000 (09:26 +0000)]
Make the 'ifi' global local to dhclient.c and pass it as an argument to
functions needing it.

This is the first step to support multiple interfaces in one dhclient(8)
instance.

ok krw@

8 years agodowngrade an error() to a debug2() to match similar cases
djm [Tue, 23 Aug 2016 08:17:42 +0000 (08:17 +0000)]
downgrade an error() to a debug2() to match similar cases
in addr_match_list()

8 years agoadd tests for addr_match_list()
djm [Tue, 23 Aug 2016 08:17:04 +0000 (08:17 +0000)]
add tests for addr_match_list()

8 years agoImplement support for DT_INIT_ARRAY, DT_FINI_ARRAY and DT_PREINIT_ARRAY.
kettenis [Tue, 23 Aug 2016 06:46:17 +0000 (06:46 +0000)]
Implement support for DT_INIT_ARRAY, DT_FINI_ARRAY and DT_PREINIT_ARRAY.
Don't skip DT_INIT and DT_FINI for the main executable.  This matches what
Linux and Solaris do.

ok guenther@

8 years agoremove Protocol directive from client/server configs that causes
djm [Tue, 23 Aug 2016 06:36:23 +0000 (06:36 +0000)]
remove Protocol directive from client/server configs that causes
spammy deprecation warnings

hardcode SSH_PROTOCOLS=2, since that's all we support on the server
now (the client still may support both, so it could get confused)

8 years agoInstead of doing strcmp(argv0), track the invocation mode (pax/tar/cpio)
guenther [Tue, 23 Aug 2016 06:00:28 +0000 (06:00 +0000)]
Instead of doing strcmp(argv0), track the invocation mode (pax/tar/cpio)
in a separate variable

ok deraadt@

8 years agopool_setipl for tmpfs.
dlg [Tue, 23 Aug 2016 04:28:18 +0000 (04:28 +0000)]
pool_setipl for tmpfs.

8 years agoremove duplicated test for em_82574
jsg [Tue, 23 Aug 2016 04:04:23 +0000 (04:04 +0000)]
remove duplicated test for em_82574

8 years agoOnly try to set the times on a directory once, at the end, to avoid
guenther [Tue, 23 Aug 2016 03:31:44 +0000 (03:31 +0000)]
Only try to set the times on a directory once, at the end, to avoid
duplication of warning messages

ok krw@

8 years agoConvert %q to %ll in format strings
guenther [Tue, 23 Aug 2016 03:28:01 +0000 (03:28 +0000)]
Convert %q to %ll in format strings

ok natano@ krw@

8 years agofix negated address matching where the address list consists of a
djm [Tue, 23 Aug 2016 03:24:10 +0000 (03:24 +0000)]
fix negated address matching where the address list consists of a
single negated match, e.g. "Match addr !192.20.0.1"

Report and patch from Jakub Jelen. bz#2397 ok dtucker@

8 years agofix matching for pattern lists that contain a single negated match,
djm [Tue, 23 Aug 2016 03:22:49 +0000 (03:22 +0000)]
fix matching for pattern lists that contain a single negated match,
e.g. "Host !example"

report and patch from Robin Becker. bz#1918 ok dtucker@

8 years agosync
deraadt [Mon, 22 Aug 2016 22:48:14 +0000 (22:48 +0000)]
sync

8 years agoFix two small bugs in the new reset API code.
kettenis [Mon, 22 Aug 2016 22:06:59 +0000 (22:06 +0000)]
Fix two small bugs in the new reset API code.

8 years agoTwo minor fixes from dilyan palauzov.
nicm [Mon, 22 Aug 2016 20:07:58 +0000 (20:07 +0000)]
Two minor fixes from dilyan palauzov.

8 years agoRemove unused code and definitions.
kettenis [Mon, 22 Aug 2016 19:43:49 +0000 (19:43 +0000)]
Remove unused code and definitions.

8 years agoThis code served us well for many years,
schwarze [Mon, 22 Aug 2016 19:41:18 +0000 (19:41 +0000)]
This code served us well for many years,
but we are unlikely to reactivate it; espie@ agrees.
The new code is in /usr/src/usr.bin/mandoc/mandocdb.c.

8 years agoUse new clock API.
kettenis [Mon, 22 Aug 2016 19:38:42 +0000 (19:38 +0000)]
Use new clock API.

8 years agoUse new clock API to manipulate clocks.
kettenis [Mon, 22 Aug 2016 19:31:27 +0000 (19:31 +0000)]
Use new clock API to manipulate clocks.

8 years agoAdd support for the gmac clock.
kettenis [Mon, 22 Aug 2016 19:29:32 +0000 (19:29 +0000)]
Add support for the gmac clock.

8 years agoAdd an API to set the clock frequency.
kettenis [Mon, 22 Aug 2016 19:28:27 +0000 (19:28 +0000)]
Add an API to set the clock frequency.

8 years agoMove all the platform-specific code into a new function sxiehci_attach_phy()
kettenis [Mon, 22 Aug 2016 18:31:07 +0000 (18:31 +0000)]
Move all the platform-specific code into a new function sxiehci_attach_phy()
making use of pinctrl, clock and reset APIs where appropriate.

8 years agoAdd support for the usb clock. Also implements reset signal support.
kettenis [Mon, 22 Aug 2016 18:18:35 +0000 (18:18 +0000)]
Add support for the usb clock.  Also implements reset signal support.

8 years agoAdd a reset signal API alongside the clock API.
kettenis [Mon, 22 Aug 2016 18:16:58 +0000 (18:16 +0000)]
Add a reset signal API alongside the clock API.

8 years agoUse a reachable next hop for test 13, which make it fail as it should.
mpi [Mon, 22 Aug 2016 17:43:29 +0000 (17:43 +0000)]
Use a reachable next hop for test 13, which make it fail as it should.

8 years agoVarious clean up and reorganisation of the connection info handling code.
jsing [Mon, 22 Aug 2016 17:12:35 +0000 (17:12 +0000)]
Various clean up and reorganisation of the connection info handling code.
In particular, rename tls_free_conninfo() to tls_conninfo_free() and make
it a real free function. Rename tls_get_conninfo() to
tls_conninfo_populate() and have it allocate the struct tls_conninfo (after
freeing any existing one).

ok beck@

8 years agoStick with the usual 'if NULL return NULL' idiom.
jsing [Mon, 22 Aug 2016 17:08:10 +0000 (17:08 +0000)]
Stick with the usual 'if NULL return NULL' idiom.

ok beck@

8 years agoSync counters now that ifa_ifwithroute() no longer uses ifa_ifwithnet().
mpi [Mon, 22 Aug 2016 16:55:14 +0000 (16:55 +0000)]
Sync counters now that ifa_ifwithroute() no longer uses ifa_ifwithnet().

8 years agoUse rtalloc(9) instead of ifa_ifwithnet() to find an interface
mpi [Mon, 22 Aug 2016 16:53:59 +0000 (16:53 +0000)]
Use rtalloc(9) instead of ifa_ifwithnet() to find an interface
when adding a route to gateway to ensure a most specific match.

This makes "# route add"  coherent to "# route get" even with
p2p interfaces.  Fix a problem reported by Mart Tõnso.

This also fix rttest20 after the introduction of RTF_CACHED.

ok vgross@

8 years agoReplace "union sockunion" with "union sockaddr_union" which is also
millert [Mon, 22 Aug 2016 16:27:00 +0000 (16:27 +0000)]
Replace "union sockunion" with "union sockaddr_union" which is also
used in the kernel.  This makes it possible to remove the casts to
"struct sockaddr *" when calling networking syscalls.
OK jsing@ jca@

8 years agoWhen trying to edit an existing database with makewhatis(8) -d or -u
schwarze [Mon, 22 Aug 2016 16:12:52 +0000 (16:12 +0000)]
When trying to edit an existing database with makewhatis(8) -d or -u
but reading the database fails, report the full path to the database
on standard error, and mention that the database is automatically
recreated from scratch.
Suggested by espie@.

8 years agoWhen running into a mandoc.db(5) file still using the obsolete
schwarze [Mon, 22 Aug 2016 16:05:56 +0000 (16:05 +0000)]
When running into a mandoc.db(5) file still using the obsolete
format based on SQLite 3, say so in words that mortals can
understand rather than babbling about hex magic.
Suggested by espie@.

8 years agoSync refcount with recent change.
mpi [Mon, 22 Aug 2016 16:03:03 +0000 (16:03 +0000)]
Sync refcount with recent change.

8 years agoMake the ``rt_gwroute'' pointer of RTF_GATEWAY entries immutable.
mpi [Mon, 22 Aug 2016 16:01:52 +0000 (16:01 +0000)]
Make the ``rt_gwroute'' pointer of RTF_GATEWAY entries immutable.

This means that no protection is needed to guarantee that the next hop
route wont be modified by CPU1 while CPU0 is dereferencing it in a L2
resolution functions.

While here also fix an ``ifa'' leak resulting in RTF_GATEWAY being always
invalid.

dlg@ likes it, inputs and ok bluhm@

8 years agoDo not dereference ``rt->rt_ifa'' after calling rtfree(9).
mpi [Mon, 22 Aug 2016 15:37:23 +0000 (15:37 +0000)]
Do not dereference ``rt->rt_ifa'' after calling rtfree(9).

This could result in a use after free if the route entry was holding
the last reference of the address descriptor.

ok jca@, bluhm@, claudio@

8 years agoEnable SNI support in httpd(8).
jsing [Mon, 22 Aug 2016 15:02:18 +0000 (15:02 +0000)]
Enable SNI support in httpd(8).

ok reyk@

8 years agoBump TLS_API due to the addition of server side SNI functions.
jsing [Mon, 22 Aug 2016 14:58:26 +0000 (14:58 +0000)]
Bump TLS_API due to the addition of server side SNI functions.

8 years agoBump libtls minor due to the addition of symbols.
jsing [Mon, 22 Aug 2016 14:57:49 +0000 (14:57 +0000)]
Bump libtls minor due to the addition of symbols.

8 years agoProvide an API that enables server side SNI support - add the ability to
jsing [Mon, 22 Aug 2016 14:55:59 +0000 (14:55 +0000)]
Provide an API that enables server side SNI support - add the ability to
provide additional keypairs (via tls_config_add_keypair_{file,mem}()) and
allow the server to determine what servername the client requested (via
tls_conn_servername()).

ok beck@

8 years agoCreate contexts for server side SNI - these include the additional SSL_CTX
jsing [Mon, 22 Aug 2016 14:51:37 +0000 (14:51 +0000)]
Create contexts for server side SNI - these include the additional SSL_CTX
that is required for certificate switching with libssl and the certificate
itself so that we can match against the subject and SANs. Hook up the
servername callback and switch to the appropriate SSL_CTX if we find a
matching certificate.

ok beck@

8 years agoshorten the pkcs8 text;
jmc [Mon, 22 Aug 2016 13:39:52 +0000 (13:39 +0000)]
shorten the pkcs8 text;

8 years agoUse the new clock API.
kettenis [Mon, 22 Aug 2016 11:24:45 +0000 (11:24 +0000)]
Use the new clock API.

8 years agoImplement interfaces to disable clocks and add interfaces that enable or
kettenis [Mon, 22 Aug 2016 11:23:54 +0000 (11:23 +0000)]
Implement interfaces to disable clocks and add interfaces that enable or
disable all clocks for a device.  The latter interfaces are useful for
devices that have multiple clocks that don't have specific names/purposes
such as sxiahci(4).

8 years agoCall csignal() and selwakeup() from a KERNEL_LOCK'd task.
mpi [Mon, 22 Aug 2016 10:40:36 +0000 (10:40 +0000)]
Call csignal() and selwakeup() from a KERNEL_LOCK'd task.

This will allow us make bpf_tap() KERNEL_LOCK() free.

Discussed with dlg@ and input from guenther@

8 years agoSizes for free(9) from David Hill.
mpi [Mon, 22 Aug 2016 10:33:22 +0000 (10:33 +0000)]
Sizes for free(9) from David Hill.

8 years agoRefactor the uio to mbuf code out of sosend and start to make use of
claudio [Mon, 22 Aug 2016 10:23:42 +0000 (10:23 +0000)]
Refactor the uio to mbuf code out of sosend and start to make use of
MCLGETI and large mbuf clusters. This should speed up local connections
a fair bit. OK dlg@ and bluhm@ (after reverting the M_WAIT change on the
cluster allocation)

8 years agoUse the new clock API to get the clock frequency from the device tree and the
kettenis [Mon, 22 Aug 2016 09:37:27 +0000 (09:37 +0000)]
Use the new clock API to get the clock frequency from the device tree and the
actual hardware that controls the clock.  This gets rid of all the hardcoded
clock frequencies for specific hardware.

Tested by jsg@ on the BBB, mglocker@ on the C.H.I.P. and myself on the
Banana Pi and Raspberry Pi 3.

ok jsg@

8 years agoAdd code to enable the pll6 clock.
kettenis [Mon, 22 Aug 2016 06:48:38 +0000 (06:48 +0000)]
Add code to enable the pll6 clock.