dragonfly.git
5 years agoMerge branch 'vendor/TNFTP'
Peter Avalos [Sat, 7 Jun 2014 16:25:39 +0000 (09:25 -0700)]
Merge branch 'vendor/TNFTP'

5 years agoSync ftp with NetBSD.
Peter Avalos [Sat, 7 Jun 2014 16:15:09 +0000 (09:15 -0700)]
Sync ftp with NetBSD.

* Fix timeout on http fetches.
* Let progressmeter deal with a timeout.

Obtained-from:   NetBSD

5 years agoudp: Dispatch UDP datagrams to the correct netisr to perform ip_output()
Sepherosa Ziehau [Sat, 7 Jun 2014 13:17:08 +0000 (21:17 +0800)]
udp: Dispatch UDP datagrams to the correct netisr to perform ip_output()

Redispatch UDP datagrams to the netisr, whose owner CPU matches the
UDP datagrams hash, to do the ip_output().  As measured bt KTR, the
udp_send() spends most of its time in ip_output().

To properly support this, following stuffs are added/changed:
- Add a network private lwkt message flag to indicate the so_port of
  the netmsg should not be matched against the current netisr's msgport.
  This flag is set when we redispatch the UDP datagrams to other netisrs.
- Don't use priority messages for netisr barrier, so at UDP socket close
  time, UDP datagrams pending on other netisr msgport could be properly
  sync'ed.

The UDP datagrams redispatch itself:
- If IP options are ever configured (supposely none), a copy of it will
  be carried along w/ the UDP datagram to the target netisr.  The copy
  is made mainly because it is not safe to access the IP options of the
  inpcb in netisr which does not own the inpcb.
  (*) On the other hand accessing inpcb's multicast options is safe since
  multicast UDP datagrams output and multicast options configuration all
  happen in netisr0.
- Add nm_priv into netmsg_pru_send, which saves flags to be passed to
  ip_output() for the UDP datagram.  This does not changes the size of
  netmsg_pru_send on x86_64 due to the implicit 4bytes padding.
- udp_addrcpu_pkt() is added to calculate the "real" CPU for the UDP
  datagrams.
- Don't use inpcb route cache for redispatched UDP datagrams.  Since:
  o  The cached route is usually not for the UDP datagrams' destination.
  o  Accessing inpcb route cache in the netisr, which is not the owner
     of the inpcb is not safe.

On i7-3770 w/ 82599ES, this increases 18bytes UDP request/response
performance by ~19% (1.12M trasactions/s  ---> 1.34M transactions/s)

This commit also makes lockless firewall state table doable, since input
and output of UDP datagrams, which have same hash, are running in the
same netisr now!

5 years agotools/udpecho: Tool to test UDP request/response rate
Sepherosa Ziehau [Sat, 7 Jun 2014 12:40:01 +0000 (20:40 +0800)]
tools/udpecho: Tool to test UDP request/response rate

5 years agolocking.9: more editing
Markus Pfeiffer [Fri, 6 Jun 2014 20:01:26 +0000 (21:01 +0100)]
locking.9: more editing

5 years agolocking.9: more updates
Markus Pfeiffer [Fri, 6 Jun 2014 19:57:46 +0000 (20:57 +0100)]
locking.9: more updates

5 years agolocking.9 updates
Markus Pfeiffer [Fri, 6 Jun 2014 18:23:52 +0000 (19:23 +0100)]
locking.9 updates

5 years agoman: Add locking(9) manpage about locking and synchronisation
Markus Pfeiffer [Thu, 5 Jun 2014 21:22:27 +0000 (22:22 +0100)]
man: Add locking(9) manpage about locking and synchronisation

5 years agoopenssl: Adjust manual pages for 1.0.1h.
Sascha Wildner [Fri, 6 Jun 2014 15:57:17 +0000 (17:57 +0200)]
openssl: Adjust manual pages for 1.0.1h.

5 years agoMerge branch 'vendor/OPENSSL'
Sascha Wildner [Fri, 6 Jun 2014 15:51:29 +0000 (17:51 +0200)]
Merge branch 'vendor/OPENSSL'

5 years agoImport OpenSSL-1.0.1h.
Sascha Wildner [Fri, 6 Jun 2014 15:48:23 +0000 (17:48 +0200)]
Import OpenSSL-1.0.1h.

* Fixes for CVE-2014-0224, CVE-2014-0221, CVE-2014-0195, CVE-2014-3470
* Harmonize version and its documentation
* Fix eckey_priv_encode so it immediately returns an error upon a
    failure in i2d_ECPrivateKey
* Fix some double frees. These are not thought to be exploitable.

See also http://www.openssl.org/news/secadv_20140605.txt

Submitted-by: Robin Hahling <robin.hahling@gw-computing.net>
5 years agopktgen: Allow switching dst and src addresses configuration
Sepherosa Ziehau [Fri, 6 Jun 2014 12:47:59 +0000 (20:47 +0800)]
pktgen: Allow switching dst and src addresses configuration

So it could simulate multiple clients and one target server use case.

5 years agobsd-family-tree: Sync with FreeBSD (for OpenBSD 5.5 and DragonFly 3.8.0).
Sascha Wildner [Fri, 6 Jun 2014 02:28:18 +0000 (04:28 +0200)]
bsd-family-tree: Sync with FreeBSD (for OpenBSD 5.5 and DragonFly 3.8.0).

5 years agolibfetch: sync with freebsd
Joris Giovannangeli [Thu, 5 Jun 2014 13:34:12 +0000 (15:34 +0200)]
libfetch: sync with freebsd

5 years agohostapd.conf.5: Add back Mt for email addresses.
Sascha Wildner [Thu, 5 Jun 2014 08:32:33 +0000 (10:32 +0200)]
hostapd.conf.5: Add back Mt for email addresses.

5 years agolibthread_xu: add support for named POSIX semaphores.
Joris Giovannangeli [Wed, 4 Jun 2014 13:48:38 +0000 (15:48 +0200)]
libthread_xu: add support for named POSIX semaphores.

 * this implementation uses files as rendezvous and simply map files
   into the process address space. It means one page per semaphore.

 * by default, semaphores are created in /tmp/sem. If the
   LIBTHREAD_SEM_PREFIX environment variable is defined, its value is
   used as prefix.

5 years agoPER FREEBSD-SA-14:13 Incorrect error handling in PAM policy parser
Robert Garrett [Wed, 4 Jun 2014 13:44:02 +0000 (15:44 +0200)]
PER FREEBSD-SA-14:13  Incorrect error handling in PAM policy parser

This is the original patch submitted to FreeBSD
by as far as I can tell Peter Wemm, and Dag-Erling Smorgrav

Testing: Robert Garrett

5 years agoudp: Wait for address memory
Sepherosa Ziehau [Wed, 4 Jun 2014 08:38:16 +0000 (16:38 +0800)]
udp: Wait for address memory

It's better than wait for the current message to be done, since there
could be piles of messages on the target msgport.  This Improves UDP
sending performance a little bit.

5 years agowpa_supplicant/hostapd: Fix a number of manual page issues.
Sascha Wildner [Tue, 3 Jun 2014 13:29:50 +0000 (15:29 +0200)]
wpa_supplicant/hostapd: Fix a number of manual page issues.

* If there were no content related change on a manual page, leave .Dd
  as it is. If there were content related change, change it to the date
  when it was changed in our tree, not when it was changed in FreeBSD
  or elsewhere.

* Comment out a number of references to manual pages which we don't
  have because the drivers have not yet been ported.

* Add .Mt back to email addresses.

* .Fx -> .Dx

* Typo fix.

5 years agousb4bsd: change if_run to use usb_pause_ls
Markus Pfeiffer [Tue, 3 Jun 2014 09:43:59 +0000 (10:43 +0100)]
usb4bsd: change if_run to use usb_pause_ls

5 years agousb4bsd: implement pause function with lock and serializer
Markus Pfeiffer [Tue, 3 Jun 2014 09:15:32 +0000 (10:15 +0100)]
usb4bsd: implement pause function with lock and serializer

5 years agomandoc: bump DragonFly version number
Franco Fichtner [Mon, 2 Jun 2014 20:03:04 +0000 (22:03 +0200)]
mandoc: bump DragonFly version number

5 years agoservice: vanilla import of /etc/rc.d/xxx wrapper
Franco Fichtner [Mon, 2 Jun 2014 19:59:29 +0000 (21:59 +0200)]
service: vanilla import of /etc/rc.d/xxx wrapper

Taken-from: FreeBSD

5 years agokernel/xdisk: Move the dereference of 'tag' below the NULL check.
Sascha Wildner [Mon, 2 Jun 2014 18:07:33 +0000 (20:07 +0200)]
kernel/xdisk: Move the dereference of 'tag' below the NULL check.

5 years agokernel/iwn: Fix panic on kldunload
Johannes Hofmann [Sun, 1 Jun 2014 18:46:39 +0000 (20:46 +0200)]
kernel/iwn: Fix panic on kldunload

5 years agokernel/acpi: Make interpreter slack mode a tunable.
Sascha Wildner [Sun, 1 Jun 2014 14:06:11 +0000 (16:06 +0200)]
kernel/acpi: Make interpreter slack mode a tunable.

Taken-from: FreeBSD

5 years agopflogd: Make pflogd recognize previous logfile
Peeter [Wed, 7 May 2014 03:17:25 +0000 (12:17 +0900)]
pflogd: Make pflogd recognize previous logfile

* scan_dump() must use pcap_sf_pkthdr instead of pcap_pkthdr

* other minor changes to make pdflogd more similar to the FreeBSD version

Changes obtained from FreeBSD

5 years agoacpica: Add a Makefile for the 'aapits' utility (part of the ACPICA tests).
Sascha Wildner [Fri, 30 May 2014 17:05:04 +0000 (19:05 +0200)]
acpica: Add a Makefile for the 'aapits' utility (part of the ACPICA tests).

It is not hooked into the build and remains unhooked and serves just to
make setting up the tests easier.

5 years agokernel/acpi: Remove some empty code.
Sascha Wildner [Fri, 30 May 2014 16:56:24 +0000 (18:56 +0200)]
kernel/acpi: Remove some empty code.

5 years agokernel/npx: add process name to error message
Markus Pfeiffer [Thu, 29 May 2014 20:51:07 +0000 (21:51 +0100)]
kernel/npx: add process name to error message

5 years agonrelease: Verify that $DPORTS_TREE indeed exists as a directory.
Sascha Wildner [Thu, 29 May 2014 18:21:03 +0000 (20:21 +0200)]
nrelease: Verify that $DPORTS_TREE indeed exists as a directory.

Also, if it exists, output its name so it can be looked up in the
logfile (in case there might have been a typo when overriding the
default).

5 years agokernel/acpi: Sync acpi_video with FreeBSD.
Sascha Wildner [Thu, 29 May 2014 16:46:50 +0000 (18:46 +0200)]
kernel/acpi: Sync acpi_video with FreeBSD.

* Add LCD brightness control notify handler.

* Add support for "cycle" and "zero" events for LCD brightness control.

* Fix recognition of video outputs that use the standardized device
  ID scheme in ACPI 3.0 and set bit 31 at the same time (such as certain
  Dell laptops).

5 years agodrm/i915: New IRQ management code from Linux
François Tigeot [Thu, 29 May 2014 10:01:31 +0000 (12:01 +0200)]
drm/i915: New IRQ management code from Linux

* Move some functions around, they were defined in different files
  than in Linux

* Add some Haswell and Valleyview support code

* Generally reduce differences with Linux 3.8.13

5 years agoinstaller: Fix disk_find() on systems that have both da1 and da10.
Sascha Wildner [Thu, 29 May 2014 06:26:47 +0000 (08:26 +0200)]
installer: Fix disk_find() on systems that have both da1 and da10.

In addition to comparing names (up to the length of the disk which
was found previously), also compare name lengths.

While here, remove the disk_find() check in disk_new(). disk_new()
didn't deal with it returning NULL. But because the list of disks
we offer as targets for installing is taken from sysctl kern.disks'
output, it will not have duplicates (else it would likely be a bug
in the kernel).

Reported-by: BSDUX on IRC
5 years agotop: Fix the interactive command 'U'.
Imre Vadasz [Wed, 28 May 2014 20:07:11 +0000 (22:07 +0200)]
top: Fix the interactive command 'U'.

* The gstate->header_text and gstate->get_userid variables need to be
  updated when toggling between displaying of UIDs and usernames.

* Removing the unnecessary HAVE_FORMAT_PROCESS_HEADER checks.

5 years agotop: Fix -Sc option combination (bug #2670)
John Marino [Wed, 28 May 2014 16:26:52 +0000 (18:26 +0200)]
top: Fix -Sc option combination (bug #2670)

It was reported that the "top -S" and "top -c" commands work as expected
but "top -Sc" showed repeated lines.  I confirmed a malfunction, but a
different one: all 20 lines showing in console were the same.

The following patch will revert to standard format if a system command is
detected, otherwise the desired "full command" option is shown.

< http://bugs.dragonflybsd.org/issues/2670 >

5 years agoWPA Supplicant version upgrade complete: 0.6.10 => 2.1
John Marino [Wed, 28 May 2014 11:31:47 +0000 (13:31 +0200)]
WPA Supplicant version upgrade complete: 0.6.10 => 2.1

This is step 4 of 4 of the process to upgrade WPA Supplicant and HostAPD
from version 0.6.10 (Dec 2010) to version 2.1 (Feb 2014).

  - Added hostapd.conf and wpa_supplicant.conf to etc
  - updated share/initrd/etc/wpa_supplicant.conf with the vendor
    version,
  - minors man fixes taken from FreeBSD head,
  - major Makefile overhaul in usr.sbin/802_11 based on FreeBSD head,
  - use the vendor's driver_wired and driver_bsd,
  - l2_packet.c is a copy of the vendor's l2_packet/l2_packet_freebsd.

All significant work contributed by: Alexandre Perrin <alex@kaworu.ch>
Thanks!

5 years agoWPA Supplicant upgrade: Add local modifications and READMEs
John Marino [Wed, 28 May 2014 11:22:00 +0000 (13:22 +0200)]
WPA Supplicant upgrade: Add local modifications and READMEs

Both vendor sources only need modification on a single file: wps_upnp.c
This is step 3 of 4 of the upgrade.

5 years agoMerge branch 'vendor/WPA_SUPPLICANT'
John Marino [Wed, 28 May 2014 15:24:05 +0000 (17:24 +0200)]
Merge branch 'vendor/WPA_SUPPLICANT'

Internally conflicted, use vendor/WPA_SUPPLICANT in all cases.

5 years agoMerge branch 'vendor/HOSTAPD'
John Marino [Wed, 28 May 2014 15:21:37 +0000 (17:21 +0200)]
Merge branch 'vendor/HOSTAPD'

Internally conflicted, use vendor/HOSTAPD in all cases

5 years agoRemoved copies of hostapd/wpa to prepare for branch merge
John Marino [Wed, 28 May 2014 10:50:04 +0000 (12:50 +0200)]
Removed copies of hostapd/wpa to prepare for branch merge

The contrib directories for hostapd and wpa_supplicant were just
copies of their respective vendor branches.  They were never merged
into the master branch.

To prepare for their first merging, let's remove the existing copies of
these two contrib subdirectories completely to avoid conflicts.

This is the first of four steps to update WPA Supplicant in base.
Step 2: Merge in hostapd and wpa_supplicant vendor branches (ver 2.1)
Step 3: Add local modifications to hostapd and wpa_supplicant
Step 4: Update makefile and conf files accordingly

5 years agowpa_supplicant vendor branch: Update version 0.6.10 => 2.1 gitea/vendor/WPA_SUPPLICANT origin/vendor/WPA_SUPPLICANT
John Marino [Tue, 27 May 2014 21:43:58 +0000 (23:43 +0200)]
wpa_supplicant vendor branch: Update version 0.6.10 => 2.1

5 years agohostapd vendor branch: Update version from 0.6.10 => 2.1 gitea/vendor/HOSTAPD origin/vendor/HOSTAPD
John Marino [Tue, 27 May 2014 19:42:16 +0000 (21:42 +0200)]
hostapd vendor branch: Update version from 0.6.10 => 2.1

5 years agoudp: Merge udp_send and udp_output
Sepherosa Ziehau [Wed, 28 May 2014 12:45:25 +0000 (20:45 +0800)]
udp: Merge udp_send and udp_output

This paves way for the udp packet redispatch to distribute ip_output
workload.

5 years agoudp: Remove no longer necessary code
Sepherosa Ziehau [Sun, 25 May 2014 12:30:15 +0000 (20:30 +0800)]
udp: Remove no longer necessary code

5 years agodrm: Change the prototype of fb_create()
François Tigeot [Tue, 27 May 2014 08:36:25 +0000 (10:36 +0200)]
drm: Change the prototype of fb_create()

Make it identical to the Linux 3.8.13 version.

5 years agoLinux headers: Improve the WARN_ON() macro
François Tigeot [Mon, 26 May 2014 19:08:39 +0000 (21:08 +0200)]
Linux headers: Improve the WARN_ON() macro

Obtained-from: OpenBSD

5 years agonewvers.sh: Adjust revision.
Sascha Wildner [Mon, 26 May 2014 17:21:18 +0000 (19:21 +0200)]
newvers.sh: Adjust revision.

Reported-by: luxh
5 years agokernel/acpi: Remove some unneeded code in acpi_hp.
Sascha Wildner [Mon, 26 May 2014 08:11:15 +0000 (10:11 +0200)]
kernel/acpi: Remove some unneeded code in acpi_hp.

5 years ago3.9-DEVELOPMENT, a.k.a. master v3.9.0
Justin C. Sherrill [Mon, 26 May 2014 04:52:17 +0000 (00:52 -0400)]
3.9-DEVELOPMENT, a.k.a. master

5 years agokernel/ig_hal: Add missing braces.
Sascha Wildner [Sun, 25 May 2014 12:02:21 +0000 (14:02 +0200)]
kernel/ig_hal: Add missing braces.

In-discussion-with: sephe

5 years agokernel: Fix various indentation issues (no functional change).
Sascha Wildner [Sun, 25 May 2014 08:38:41 +0000 (10:38 +0200)]
kernel: Fix various indentation issues (no functional change).

5 years agokernel/acpi: Use standard types.
Sascha Wildner [Sun, 25 May 2014 07:59:46 +0000 (09:59 +0200)]
kernel/acpi: Use standard types.

5 years agokernel/acpi: Fix indent, also use NULL for pointers (no functional change).
Sascha Wildner [Sun, 25 May 2014 07:17:32 +0000 (09:17 +0200)]
kernel/acpi: Fix indent, also use NULL for pointers (no functional change).

5 years agoacpi.4: Sort hpet.
Sascha Wildner [Sat, 24 May 2014 21:27:39 +0000 (23:27 +0200)]
acpi.4: Sort hpet.

5 years agoacpi_dock.4: Add Mt
Sascha Wildner [Sat, 24 May 2014 21:27:02 +0000 (23:27 +0200)]
acpi_dock.4: Add Mt

5 years agokernel/acpi: Add a docking station driver.
Sascha Wildner [Sat, 24 May 2014 21:24:38 +0000 (23:24 +0200)]
kernel/acpi: Add a docking station driver.

Tested-by: profmakx (on a ThinkPad X230)
Taken-from: FreeBSD

5 years agoacpixtract.8: Add missing Fl
Sascha Wildner [Fri, 23 May 2014 11:59:50 +0000 (13:59 +0200)]
acpixtract.8: Add missing Fl

5 years agoacpi.4: Add missing space.
Sascha Wildner [Thu, 22 May 2014 07:58:15 +0000 (09:58 +0200)]
acpi.4: Add missing space.

5 years agos/NO_OBJ/NOOBJ/ in a few Makefiles.
Sascha Wildner [Wed, 21 May 2014 19:35:57 +0000 (21:35 +0200)]
s/NO_OBJ/NOOBJ/ in a few Makefiles.

5 years agolibc: Fix strcasecmp_l() and strncasecmp_l() issues.
Sascha Wildner [Wed, 21 May 2014 19:11:54 +0000 (21:11 +0200)]
libc: Fix strcasecmp_l() and strncasecmp_l() issues.

POSIX.1-2008 specifies that those two functions should be declared by
including <strings.h>, not <string.h> (the latter only has strcoll_l()
and strxfrm_l()):

http://pubs.opengroup.org/onlinepubs/9699919799/functions/strcasecmp.html

It looks like FreeBSD got this wrong in their xlocale stuff originally.

And it looks like the only clean way to properly solve it is to split
<xlocale/_string.h> into <xlocale/_string.h> and <xlocale/_strings.h>,
which this commit does.

5 years agoudp: Add KTR to measure UDP packet creation and ip_output cost
Sepherosa Ziehau [Wed, 21 May 2014 12:44:59 +0000 (20:44 +0800)]
udp: Add KTR to measure UDP packet creation and ip_output cost

5 years agoloader: Remove some unused code.
Sascha Wildner [Wed, 21 May 2014 06:53:38 +0000 (08:53 +0200)]
loader: Remove some unused code.

5 years agokernel: Remove some unused code.
Sascha Wildner [Wed, 21 May 2014 06:53:12 +0000 (08:53 +0200)]
kernel: Remove some unused code.

5 years agogroff - horrible hack to fix .gz issues
Matthew Dillon [Wed, 21 May 2014 01:42:20 +0000 (18:42 -0700)]
groff - horrible hack to fix .gz issues

* If a gzipped manual page contains a .so forwarding entry, groff
  will attempt to open the forwarding entry directly and will not
  check to see if a file with a .gz extension is present, resulting
  in an open error.

* Hack hack hack.  Horrible hack of the input search path to check
  for a .gz extension and popen() the file with zcat.  Horrible because
  I haven't bothered to modify all the other myrid code to properly call
  pclose() instead of fclose() in that case.

5 years agoifnet: Remove marker from if_multiaddrs, once iteration is done
Sepherosa Ziehau [Wed, 21 May 2014 01:21:06 +0000 (09:21 +0800)]
ifnet: Remove marker from if_multiaddrs, once iteration is done

Reported-by: tuxillo@
5 years agokernel/acpi: Improve the disabling of Debug object output.
Sascha Wildner [Tue, 20 May 2014 19:55:52 +0000 (21:55 +0200)]
kernel/acpi: Improve the disabling of Debug object output.

It is actually a global in ACPICA (since 20100304) so add a new
sysctl debug.acpi.enable_debug_objects that can be used to toggle
it, with the default being 0.

Unlike the previous attempt (which I had to revert again), this
time also clear ACPI_LV_DEBUG_OBJECT in the AcpiDbgLevel global in
case ACPI_DEBUG is not defined. My assumption that ACPICA would
initialize AcpiDbgLevel to 0 was wrong. It really is initialized
to ACPI_NORMAL_DEFAULT. It is kind of weird that
ACPI_NORMAL_DEFAULT == ACPI_DEBUG_DEFAULT in acoutput.h, but FWIW,
leave the other bits in AcpiDbgLevel set for now.

This commit reverts (again) a previous commit by sephe which solved
it differently (in the ACPICA code).

Taken-from:    FreeBSD (with modification by me)
Dragonfly-bug: <http://bugs.dragonflybsd.org/issues/2672>

5 years agoacpica: Use only one #ifdef _KERNEL ... and use 'inline' directly.
Sascha Wildner [Tue, 20 May 2014 17:57:49 +0000 (19:57 +0200)]
acpica: Use only one #ifdef _KERNEL ... and use 'inline' directly.

5 years agokernel/acpi: Fix memory and port I/O functions in the OS services layer.
Sascha Wildner [Tue, 20 May 2014 04:17:05 +0000 (06:17 +0200)]
kernel/acpi: Fix memory and port I/O functions in the OS services layer.

This makes AcpiOsReadPort(), AcpiOsWritePort(), AcpiOsReadMemory()
and AcpiOsWriteMemory() match the behavior requested in the "ACPI
User Guide and Programmer Reference".

When the width of the location that is read from or written to is less
than the (return) value's width, the read functions must ensure the
rest of the bits are zero and the write functions must write no more
than the location's width.

This should fix at least a number of temperature reading issues we
were seeing, such as the messages we were getting lately:

acpi_tz0: _TMP value is absurd, ignored (-204.0C)
acpi_tz0: _TMP value is absurd, ignored (-204.0C)
acpi_tz0: _TMP value is absurd, ignored (-203.0C)

Found-in: FreeBSD

5 years agokernel/acpi: Remove an Itanium define.
Sascha Wildner [Tue, 20 May 2014 01:47:22 +0000 (03:47 +0200)]
kernel/acpi: Remove an Itanium define.

5 years agoin_pcbladdr: Break long line; no functional changes
Sepherosa Ziehau [Mon, 19 May 2014 12:46:01 +0000 (20:46 +0800)]
in_pcbladdr: Break long line; no functional changes

5 years agoinet: Return NULL from IN_LOOKUP_MULTI if ifnet is NULL
Sepherosa Ziehau [Mon, 19 May 2014 11:45:31 +0000 (19:45 +0800)]
inet: Return NULL from IN_LOOKUP_MULTI if ifnet is NULL

This could happen for mbufs on ip_input path, whose rcvif could be NULL.

5 years agoip_input: Only multicast forwarding may be not MPSAFE
Sepherosa Ziehau [Mon, 19 May 2014 11:37:16 +0000 (19:37 +0800)]
ip_input: Only multicast forwarding may be not MPSAFE

5 years agoip_output: if_simloop is MPSAFE
Sepherosa Ziehau [Mon, 19 May 2014 11:37:01 +0000 (19:37 +0800)]
ip_output: if_simloop is MPSAFE

5 years agokernel: Fix kernel build when _KPOSIX_PRIORITY_SCHEDULING is not set.
Sascha Wildner [Mon, 19 May 2014 00:49:25 +0000 (02:49 +0200)]
kernel: Fix kernel build when _KPOSIX_PRIORITY_SCHEDULING is not set.

Define p31b_proc() and p31b_proc_done() only when it is set.

5 years agoacpi: s/ACPI[ -]CA/ACPICA/ in a few places.
Sascha Wildner [Sun, 18 May 2014 22:24:12 +0000 (00:24 +0200)]
acpi: s/ACPI[ -]CA/ACPICA/ in a few places.

5 years agoAdjust some LDADD/DPADD.
Sascha Wildner [Sun, 18 May 2014 20:50:21 +0000 (22:50 +0200)]
Adjust some LDADD/DPADD.

5 years agokernel/acpi: Use AE_OK.
Sascha Wildner [Sun, 18 May 2014 13:31:48 +0000 (15:31 +0200)]
kernel/acpi: Use AE_OK.

5 years agoix.4: Fix description of TSO and multiple tranmission queue arbitration.
Sepherosa Ziehau [Sun, 18 May 2014 08:36:15 +0000 (16:36 +0800)]
ix.4: Fix description of TSO and multiple tranmission queue arbitration.

5 years agonetisr: Don't free mbuf, if hash function has already done it.
Sepherosa Ziehau [Sun, 18 May 2014 08:32:17 +0000 (16:32 +0800)]
netisr: Don't free mbuf, if hash function has already done it.

5 years agotest/mcast: Install to /usr/local/bin
Sepherosa Ziehau [Sun, 18 May 2014 08:30:25 +0000 (16:30 +0800)]
test/mcast: Install to /usr/local/bin

And join multicast group on the interface specified.

5 years agoifnet: Properly protect if_multiaddrs using ifnet serializers
Sepherosa Ziehau [Tue, 13 May 2014 13:59:18 +0000 (21:59 +0800)]
ifnet: Properly protect if_multiaddrs using ifnet serializers

- Protect ifnet.if_multiaddrs using ifnet serializers.  Add some
  comment in the places, where only main serailizer is necessary.
- Fix if_delallmulti().  Using TAILQ_FOREACH_MUTABLE is incorrect for
  deleting an ifmultiaddr from ifnet.if_multiaddrs.  Since deleting one
  ifmultiaddr may cause additional ifmultiaddr deletion (e.g. the AF_LINK
  ifmultiaddr for AF_INET ifmultiaddr).
- Change IN_LOOKUP_MULTI and IN6_LOOKUP_MULTI macros into inline
  functions.
- Redispatch multicast IP packets to netisr0 for further processing.
  Software based IP packet hash function is changed.  And hash value
  fixup for multicast IP packets is added to the beginning of ip_input();
  this is mainly for IP packets, whose hash is calculated by hardware.
- For wlan's multicast hardware filter updating, we no longer need to
  release wlan serializer and mess up w/ the if_ioctl setting.

In netisr0, read and test ifma_refcount for AF_INET ifmultiaddr is MPSAFE
w/o ifnet serializers, since its ifma_refcount is only altered in netisr0.

In netisr0, any operation on in_multi, which is obtained from the
corresponding ifmuliaddr's ifma_protospec, is MPSAFE w/o ifnet
serializers, since ifmultiaddr for AF_INET is only set and cleared in
netisr0.

While I'm here also redispatch IP packets w/o hash to the proper netisrs,
on ip_input() path.  And unnecessary critical sections in
in_{add,del}multi() are removed.

5 years agokernel/acpi: Include <sys/ktr.h> from acpi_ec.c instead of from acpivar.h.
Sascha Wildner [Sun, 18 May 2014 00:59:30 +0000 (02:59 +0200)]
kernel/acpi: Include <sys/ktr.h> from acpi_ec.c instead of from acpivar.h.

5 years agokernel/acpi: Use ACPI_COMPARE_NAME().
Sascha Wildner [Sun, 18 May 2014 00:01:06 +0000 (02:01 +0200)]
kernel/acpi: Use ACPI_COMPARE_NAME().

5 years agokernel/acpi: Add missing header to the Makefile.
Sascha Wildner [Sun, 18 May 2014 00:00:47 +0000 (02:00 +0200)]
kernel/acpi: Add missing header to the Makefile.

5 years agokernel/acpi: Remove unneeded kprintf when overriding DSDT.
Sascha Wildner [Sat, 17 May 2014 23:15:52 +0000 (01:15 +0200)]
kernel/acpi: Remove unneeded kprintf when overriding DSDT.

ACPI will print a notice ("Logical table override...") itself.

5 years agoacpi.4: Remove COMPATIBILITY altogether.
Sascha Wildner [Sat, 17 May 2014 23:14:02 +0000 (01:14 +0200)]
acpi.4: Remove COMPATIBILITY altogether.

It can be brought back once we support more platforms.

5 years agoacpi.4: Refresh COMPATIBILITY and update the spec reference.
Sascha Wildner [Sat, 17 May 2014 23:05:35 +0000 (01:05 +0200)]
acpi.4: Refresh COMPATIBILITY and update the spec reference.

5 years agoloader.conf: Fix ACPI DSDT overriding.
Sascha Wildner [Sat, 17 May 2014 22:58:47 +0000 (00:58 +0200)]
loader.conf: Fix ACPI DSDT overriding.

Also, fix the instructions in the acpi(4) manual page.

5 years agotest: my first commit
Imre Vadasz [Sat, 17 May 2014 17:58:44 +0000 (19:58 +0200)]
test: my first commit

5 years agokernel/netgraph7: Remove KMODDEPS.
Sascha Wildner [Sat, 17 May 2014 09:45:22 +0000 (11:45 +0200)]
kernel/netgraph7: Remove KMODDEPS.

5 years agokernel/netgraph7: Fix Makefile.
Sascha Wildner [Sat, 17 May 2014 09:42:42 +0000 (11:42 +0200)]
kernel/netgraph7: Fix Makefile.

5 years agokernel/ral: Clean up a mis-delete I did.
Sascha Wildner [Fri, 16 May 2014 09:21:54 +0000 (11:21 +0200)]
kernel/ral: Clean up a mis-delete I did.

5 years agokernel/acpi: ACPI_INTEGER is obsolete since 20100121 (replaced with UINT64).
Sascha Wildner [Fri, 16 May 2014 09:09:52 +0000 (11:09 +0200)]
kernel/acpi: ACPI_INTEGER is obsolete since 20100121 (replaced with UINT64).

5 years agokernel/ral: Bring in support for further chips (RT2700/2800/3090 series).
Sascha Wildner [Thu, 15 May 2014 09:42:35 +0000 (11:42 +0200)]
kernel/ral: Bring in support for further chips (RT2700/2800/3090 series).

Unfortunately I don't have any hardware to test and no one volunteered
when I asked on the lists. Pushing anyway because it shouldn't break
RT2500/2501/2600 support. We'll see if someone reports issues with
it.

5 years agoUpdate the pciconf(8) database.
Sascha Wildner [Thu, 15 May 2014 09:20:06 +0000 (11:20 +0200)]
Update the pciconf(8) database.

May 13, 2014 snapshot from http://pciids.sourceforge.net/

5 years agokernel/acpi: i386 build fixes.
Sascha Wildner [Thu, 15 May 2014 09:06:07 +0000 (11:06 +0200)]
kernel/acpi: i386 build fixes.

5 years agokernel/acpi: Remove a debug printf I accidentally pushed.
Sascha Wildner [Thu, 15 May 2014 06:59:13 +0000 (08:59 +0200)]
kernel/acpi: Remove a debug printf I accidentally pushed.

5 years agolibmagic(3): Some MLINKS adjustments for the most recent upgrade.
Sascha Wildner [Thu, 15 May 2014 06:55:43 +0000 (08:55 +0200)]
libmagic(3): Some MLINKS adjustments for the most recent upgrade.

5 years agounits(1): Sync with FreeBSD
Eitan Adler [Thu, 15 May 2014 05:31:38 +0000 (22:31 -0700)]
units(1): Sync with FreeBSD
- fix -f and -V option
- fix a few typos
- call el_end()
- use builtin functions where appropriate