oweals/openwrt.git
11 years agokernel: add kirkwood thermal module
Luka Perkov [Sat, 12 Oct 2013 16:26:55 +0000 (16:26 +0000)]
kernel: add kirkwood thermal module

Signed-off-by: Luka Perkov <luka@openwrt.org>
SVN-Revision: 38371

11 years agoar7: drop linux 3.9 support
Jonas Gorski [Sat, 12 Oct 2013 10:50:47 +0000 (10:50 +0000)]
ar7: drop linux 3.9 support

Linux 3.9 is EOL since quite a while, no need to keep support for it.

Signed-off-by: Jonas Gorski <jogo@openwrt.org>
SVN-Revision: 38370

11 years agoar7: add support for linux 3.10 and switch to it
Jonas Gorski [Sat, 12 Oct 2013 10:50:44 +0000 (10:50 +0000)]
ar7: add support for linux 3.10 and switch to it

Run tested on WAG354Gv2.

Signed-off-by: Jonas Gorski <jogo@openwrt.org>
SVN-Revision: 38369

11 years agomvswitch: fix compilation for linux 3.10
Jonas Gorski [Sat, 12 Oct 2013 10:50:39 +0000 (10:50 +0000)]
mvswitch: fix compilation for linux 3.10

Signed-off-by: Jonas Gorski <jogo@openwrt.org>
SVN-Revision: 38368

11 years agoacx-mac80211: update to latest version
Jonas Gorski [Sat, 12 Oct 2013 10:50:35 +0000 (10:50 +0000)]
acx-mac80211: update to latest version

Fixes compilation for linux 3.10.

Signed-off-by: Jonas Gorski <jogo@openwrt.org>
SVN-Revision: 38367

11 years agoar7-atm: fix compilation for linux 3.10
Jonas Gorski [Sat, 12 Oct 2013 10:50:19 +0000 (10:50 +0000)]
ar7-atm: fix compilation for linux 3.10

Update proc code to new proc_create to allow compilation with
linux 3.10+.

Signed-off-by: Jonas Gorski <jogo@openwrt.org>
SVN-Revision: 38366

11 years agodnsmasq: add option broadcast to host sections
Jo-Philipp Wich [Fri, 11 Oct 2013 18:33:15 +0000 (18:33 +0000)]
dnsmasq: add option broadcast to host sections

There are certain consumer devices which are outliers in protocol conformance.
An example is Samsung bluray players, which require broadcast DHCP responses
(on Ethernet only, strangely not on Wifi).

By specifying:

config host
...
option broadcast 1

this will enable the response to be sent as an Ethernet broadcast and not as
a unicast.

Signed-off-by: Philip Prindeville <philipp@redfish-solutions.com>
SVN-Revision: 38365

11 years agomake the generated images actually flashable on the Unifi devices.
Imre Kaloz [Fri, 11 Oct 2013 15:41:42 +0000 (15:41 +0000)]
make the generated images actually flashable on the Unifi devices.

Signed-off-by: Imre Kaloz <kaloz@openwrt.org>
SVN-Revision: 38364

11 years agoath9k: fix rx issues introduced in my last commit (fixes #14307)
Felix Fietkau [Thu, 10 Oct 2013 21:10:45 +0000 (21:10 +0000)]
ath9k: fix rx issues introduced in my last commit (fixes #14307)

Signed-off-by: Felix Fietkau <nbd@openwrt.org>
SVN-Revision: 38362

11 years agofirewall: small improvements in nat reflection
Jo-Philipp Wich [Thu, 10 Oct 2013 18:15:10 +0000 (18:15 +0000)]
firewall: small improvements in nat reflection

- do not insert duplicate rules when setting up reflection to a zone containing multiple interfaces
- set up reflection for any protocol, not just TCP and UDP

SVN-Revision: 38361

11 years agoIncorrect checkout of git submodules
Jo-Philipp Wich [Thu, 10 Oct 2013 15:01:50 +0000 (15:01 +0000)]
Incorrect checkout of git submodules

When checking out git packages, buildroot doesn't seem to track the revisions
correctly of any submodules referenced by that project. As a result, the
submodule stays at whatever revision was referenced by the head of the master
branch. Running a 'git submodule update' after the checkout fixes this problem.

Signed-off-by: Owen Kirby <osk@exegin.com>
SVN-Revision: 38359

11 years agoppp: correct module loaded check in proto_pptp_setup
Jo-Philipp Wich [Thu, 10 Oct 2013 14:58:12 +0000 (14:58 +0000)]
ppp: correct module loaded check in proto_pptp_setup

proto_pptp_setup is responsible for loading the required modules to establish
a pptp connection to a foreign peer. The function checks whether all required
modules are already loaded, before actually loading them.

It seems that the filter being used to accomplish this, is not restrictive
enough in some cases. For instance when pptp nat helper modules are present on
a system, and already loaded before a pptp connection is enabled. Then the
search filter (possibly) returns the following for module=pptp, where actually
no matches are expected, resulting in the pptp.ko module not being loaded,
thereby failing to establish the pptp connection.

# module="pptp" ; grep "$module" /proc/modules
nf_nat_pptp 1312 0 - Live 0x86ce7000
nf_conntrack_pptp 3072 1 nf_nat_pptp, Live 0x86cb9000
nf_nat_proto_gre 784 1 nf_nat_pptp, Live 0x86cba000
nf_conntrack_proto_gre 2368 1 nf_conntrack_pptp, Live 0x86cbf000
nf_nat 9792 13 nf_nat_rtsp,nf_nat_tftp,nf_nat_sip,nf_nat_pptp,nf_nat_h323,nf_nat_proto_gre,nf_nat_amanda,nf_nat_irc,nf_nat_ftp,ipt_REDIRECT,ipt_NETMAP,ipt_MASQUERADE,iptable_nat, Live 0x86ca8000
nf_conntrack 37264 31 nf_nat_rtsp,nf_conntrack_rtsp,nf_nat_tftp,nf_conntrack_tftp,nf_nat_snmp_basic,nf_conntrack_snmp,nf_nat_sip,nf_conntrack_sip,nf_nat_pptp,nf_conntrack_pptp,nf_nat_h323,nf_conntrack_h323,nf_conntrack_proto_gre,nf_nat_amanda,nf_conntrack_amanda,nf_conntrack_broadcast,nf_nat_irc,nf_conntrack_irc,nf_nat_ftp,nf_conntrack_ftp,ipt_MASQUERADE,iptable_nat,nf_nat,xt_helper,xt_connmark,xt_connbytes,xt_conntrack,xt_CT,xt_NOTRACK,xt_state,nf_conntrack_ipv4, Live 0x86c90000

The search filter can be made more accurate/restrictive, by requiring the
occurance of the exact name of the module at the beginning of a line in
/proc/modules.

# module="pptp" ; grep "^$module " /proc/modules
pptp 13296 2 - Live 0x86e80000

Signed-off-by: Tijs Van Buggenhout <tvb@able.be>
SVN-Revision: 38358

11 years agodropbear: update to 2013.59 (released 4 october 2013)
Jo-Philipp Wich [Thu, 10 Oct 2013 14:42:05 +0000 (14:42 +0000)]
dropbear: update to 2013.59 (released 4 october 2013)

- drop mirror www.mirrors.wiretapped.net (not working anymore)
- drop patch 300-ipv6_addr_port_split.patch, included upstream
- refresh patches
- various upstream changes: http://matt.ucc.asn.au/dropbear/CHANGES

Signed-off-by: Catalin Patulea <cat@vv.carleton.ca>
SVN-Revision: 38356

11 years agox86: kvm: disable squashfs and jffs2 images by default
Jo-Philipp Wich [Thu, 10 Oct 2013 14:40:52 +0000 (14:40 +0000)]
x86: kvm: disable squashfs and jffs2 images by default

SVN-Revision: 38355

11 years agoath9k: add some cleanups and minor fixes
Felix Fietkau [Thu, 10 Oct 2013 14:31:22 +0000 (14:31 +0000)]
ath9k: add some cleanups and minor fixes

Signed-off-by: Felix Fietkau <nbd@openwrt.org>
SVN-Revision: 38354

11 years agokernel: add default config for ARM_ERRATA symbols for kernel 3.3
Hauke Mehrtens [Thu, 10 Oct 2013 11:13:54 +0000 (11:13 +0000)]
kernel: add default config for ARM_ERRATA symbols for kernel 3.3

Commit r37470 did not add a default conig for the ARM_ERRATA symbols for kernel 3.3.

Signed-off-by: Hauke Mehrtens <hauke@hauke-m.de>
SVN-Revision: 38352

11 years agoopenvpn: rework initscript (fixes #14299)
Luka Perkov [Wed, 9 Oct 2013 11:25:09 +0000 (11:25 +0000)]
openvpn: rework initscript (fixes #14299)

Signed-off-by: Luka Perkov <luka@openwrt.org>
SVN-Revision: 38351

11 years agoprocd: fix command handling in procd_add_instance()
Felix Fietkau [Wed, 9 Oct 2013 11:07:23 +0000 (11:07 +0000)]
procd: fix command handling in procd_add_instance()

Signed-off-by: Felix Fietkau <nbd@openwrt.org>
SVN-Revision: 38350

11 years agoIPv6 support should depend on if we've enabled it
Imre Kaloz [Wed, 9 Oct 2013 11:03:00 +0000 (11:03 +0000)]
IPv6 support should depend on if we've enabled it

Signed-off-by: Imre Kaloz <kaloz@openwrt.org>
SVN-Revision: 38349

11 years agoralink: fixes mpr a1/a2 network setup
John Crispin [Wed, 9 Oct 2013 10:08:52 +0000 (10:08 +0000)]
ralink: fixes mpr a1/a2 network setup

Signed-off-by: Suschman <suschman@gmx.net>
Signed-off-by: John Crispin <blogic@openwrt.org>
SVN-Revision: 38348

11 years agomac80211: rtlwifi: Align private space in rtl_priv struct
Hauke Mehrtens [Wed, 9 Oct 2013 09:13:41 +0000 (09:13 +0000)]
mac80211: rtlwifi: Align private space in rtl_priv struct

Signed-off-by: Hauke Mehrtens <hauke@hauke-m.de>
SVN-Revision: 38346

11 years agoralink: add support for the mt7530 eval board
John Crispin [Tue, 8 Oct 2013 21:10:15 +0000 (21:10 +0000)]
ralink: add support for the mt7530 eval board

Signed-off-by: John Crispin <blogic@openwrt.org>
SVN-Revision: 38345

11 years agoralink: add missing header file to mt7530 driver
John Crispin [Tue, 8 Oct 2013 21:10:00 +0000 (21:10 +0000)]
ralink: add missing header file to mt7530 driver

Signed-off-by: John Crispin <blogic@openwrt.org>
SVN-Revision: 38344

11 years agolantiq: fix up diag.sh
John Crispin [Tue, 8 Oct 2013 15:28:33 +0000 (15:28 +0000)]
lantiq: fix up diag.sh

init no longer exists as a parameter

Signed-off-by: John Crispin <blogic@openwrt.org>
SVN-Revision: 38342

11 years agokernel: dont rely on library modules to be depended on
John Crispin [Tue, 8 Oct 2013 15:08:33 +0000 (15:08 +0000)]
kernel: dont rely on library modules to be depended on

fixes https://dev.openwrt.org/ticket/14257

Signed-off-by: John Crispin <blogic@openwrt.org>
SVN-Revision: 38341

11 years agokernel: remove debug code from module insert script
John Crispin [Tue, 8 Oct 2013 15:08:23 +0000 (15:08 +0000)]
kernel: remove debug code from module insert script

Signed-off-by: John Crispin <blogic@openwrt.org>
SVN-Revision: 38340

11 years agohostapd: recognize 8021x as an authentication mode
Felix Fietkau [Tue, 8 Oct 2013 11:09:52 +0000 (11:09 +0000)]
hostapd: recognize 8021x as an authentication mode

Currently, in order to configure the authentication daemon in
8021x mode, we need to set wireless.@wifi-iface[0].encryption="wpa"
Though it works it confuses folks as 8021x is using WEP
encryption and not WPA. Therefore the terminology itself is
confusing. This change adds 8021x as a recognized string for 8021x
authentication.

Signed-off-by: Mathieu Olivari <mathieu@qca.qualcomm.com>
Signed-off-by: Luis R. Rodriguez <mcgrof@qca.qualcomm.com>
SVN-Revision: 38339

11 years agohostapd: add external registrar support
Felix Fietkau [Tue, 8 Oct 2013 11:09:48 +0000 (11:09 +0000)]
hostapd: add external registrar support

Setting wireless.@wifi-iface[N].ext_registrar=1 will enable UPNP
advertising and add an external registrar to the interface this vif
belongs to (br-lan if the vif is included in the LAN bridge). By
enabling this we append upnp_iface=xxx to the hostapd config file.

Signed-off-by: Sujith Manoharan <c_manoha@qca.qualcomm.com>
Signed-off-by: Mathieu Olivari <mathieu@qca.qualcomm.com>
Signed-off-by: Luis R. Rodriguez <mcgrof@qca.qualcomm.com>
SVN-Revision: 38338

11 years agohostapd: enable WPS2 support on hostapd-full.config
Felix Fietkau [Tue, 8 Oct 2013 11:09:44 +0000 (11:09 +0000)]
hostapd: enable WPS2 support on hostapd-full.config

Enable CONFIG_WPS2 for hostapd. This is required to support
options like Virtual Push Button in WPS.

Signed-off-by: Sujith Manoharan <c_manoha@qca.qualcomm.com>
Signed-off-by: Luis R. Rodriguez <mcgrof@qca.qualcomm.com>
SVN-Revision: 38337

11 years agohostapd: fix hostapd RSN preauthentication PMKSA caching
Felix Fietkau [Tue, 8 Oct 2013 11:09:40 +0000 (11:09 +0000)]
hostapd: fix hostapd RSN preauthentication PMKSA caching

In 2009 OpenWrt's hostapd config added an "auth_cache" boolean
to be used to address a reported issue #12129 [0] on a forum [1].
The reported issue on the ticket is different that the one
described on the forum. The commit was r33359. This change broke
proper RSN preauthentication [2] [3] [4] expectations on hostapd's
configuration for WPA2 and this in turn disabled PMKSA caching and
Opportunistic Key Caching. This change:

  * Leaves the "auth_cache" to be used only for WPA networks for those
    looking to use this as a workaround to a reported issue but annotates
    a warning over its usage.

  * Separate "auth_cache" from WPA2 RSN preauthentication, leaving
    WPA2 RSN preauthentication to enabled only with "rsn_preauth" with
    the expected and recommended settings.

  * Adds a new WPA2 RSN preauthentication "rsn_preauth_testing" to
    be used when evaluating funcionality for WPA2 RSN preauthentication
    with the expected and recommended settings with the only difference
    so far with what should be enabled by default to disable Opportunistic
    Key Caching.

Disabling the PMKSA cache should mean the STA could not roam off and back
onto the AP that had PMKSA caching disabled and would require a full
authentication cycle. This fixes this for WPA2 networks with
RSN preauthentication enabled.

This change should be applied to AA as well as trunk.

  TL DR;

The issue described on the forum has to do with failure of a STA
being able to try to authenticate again with the AP if it failed
its first try. This may have been an issue with hostapd in 2009
but as per some tests I cannot reproduce this today on a WPA2
network.

The issue described on the ticket alludes to a security issue with the
design of using a Radius server to authenticate to an AP. The issue
vaguely alludes to the circumstances of zapping a user, deleting their
authentication credentials to log in to the network, and that if
RSN preauthentication is enabled with PMKSA caching that the user
that was zapped would still be able to authenticate.

Lets treat these as separate issues.

I cannot reproduce the first issue reported on the forums of not
being able to authenticate anymore on a WPA2 network.

The issue reported on the ticket modified WPA2 RSN preauthentication
by adding two fields to the hostapd configuration if auth_cache
was enabled:

  * disable_pmksa_caching=1
  * okc=0

The first one disables PMKSA authentication cache.
The second one disables Opportunistic Key Caching.

The issue reported on the ticket was fixed by implementing a workaround
in hostapd's configuration. Disabling PMKSA caching breaks proper use
of WPA2 RSN pre authentication. The usage of disable_pmksa_caching=1
prevents hostapd from adding PMKSA entries into its cache when a successful
802.1x authentication occurs. In practice RSN preauthentication would
trigger a STA to perform authentication with other APs on the same SSID,
it would then have its own supplicant PMKSA cache held. If a STA roams
between one AP to another no new authenitcation would need to be performed
as the new AP would already have authenticated the STA. The purpose of the
PMKSA cache on the AP side would be for the AP to use the same PMKID for
a STA when the STA roams off onto another BSSID and later comes back to it.

Disabling Opportunistic Key Caching could help the reported issue
as well but its not the correct place to address this. Opportunistic
Key Caching enables an AP with different interfaces to share the
PMKSA cache. Its a technical enhancement and disabling it would
be useful to let a testing suite properly test for RSN preauthentication
given that otherwise Opportunistic Key Caching would enable an
interface being tested to derive its own derive the PMKSA entry.
In production though okc=1 should be enabled to help with RSN
preauthentication.

The real fix for this particular issue outside of the scope of hostapd's
configuration and it should not be dealt with as a workaround to
its configuration and breaking expected RSN preauthentication and
technical optimizations. Revert this change and enable users to pick
and choose to enable or disable disable_pmksa_caching and okc expecting them
to instead have read clearly more what these do.

As for the core issure ported, the correct place to fix this is to
enable a sort of messaging between the RADIUS server and its peers
so that if caching for authentication is enabled that cache can be
cleared upon user credential updates. Updating a user password
(not just zapping a user) is another possible issue that would need
to be resolved here. Another part of the solution might be to reduce
the cache timing to account for any systematic limitations (RADIUS
server not able to ask peers to clear cache might be
one).

[0] https://dev.openwrt.org/changeset/33359
[1] https://forum.openwrt.org/viewtopic.php?id=19596
[2] http://wireless.kernel.org/en/users/Documentation/hostapd#IEEE_802.11i.2FRSN.2FWPA2_pre-authentication
[3] http://wireless.kernel.org/en/users/Documentation/wpa_supplicant#RSN_preauthentication
[4] http://wiki.openwrt.org/doc/recipes/rsn_preauthentication

Signed-off-by: Luis R. Rodriguez <mcgrof@do-not-panic.com>
SVN-Revision: 38336

11 years agohostapd: Add WPS unconfigured & WPS pin method support
Felix Fietkau [Tue, 8 Oct 2013 11:09:36 +0000 (11:09 +0000)]
hostapd: Add WPS unconfigured & WPS pin method support

Signed-off-by: Mathieu Olivari <mathieu@qca.qualcomm.com>
Signed-off-by: Luis R. Rodriguez <mcgrof@qca.qualcomm.com>
SVN-Revision: 38335

11 years agohostapd: Add eap_reauth_period config option
Felix Fietkau [Tue, 8 Oct 2013 11:09:27 +0000 (11:09 +0000)]
hostapd: Add eap_reauth_period config option

This adds the eap_reauth_period to be used for modifying
the RADIUS server reauthentication authentication period,
a parameter that gets passed directly to the hostapd
configuration file.

Signed-off-by: Mathieu Olivari <mathieu@qca.qualcomm.com>
Signed-off-by: Luis R. Rodriguez <mcgrof@qca.qualcomm.com>
SVN-Revision: 38334

11 years agomac80211: wl12xx: build them again
Hauke Mehrtens [Mon, 7 Oct 2013 22:16:17 +0000 (22:16 +0000)]
mac80211: wl12xx: build them again

WL_TI is bool but was set to =m, which deactivated it. Now it is set to
=y and should be activated in the config so the wl12xx driver gets
build.

This closes #14212.

Signed-off-by: Hauke Mehrtens <hauke@hauke-m.de>
SVN-Revision: 38333

11 years agomac80211: brcmfmac: BRCMFMAC_USB was not set in config
Hauke Mehrtens [Mon, 7 Oct 2013 21:49:28 +0000 (21:49 +0000)]
mac80211: brcmfmac: BRCMFMAC_USB was not set in config

Somehow BRCMFMAC_USB was not set in the build and brcmfmac.ko did not had usb support.

Signed-off-by: Hauke Mehrtens <hauke@hauke-m.de>
SVN-Revision: 38332

11 years agoopenvpn: switch to new procd init script style
Luka Perkov [Mon, 7 Oct 2013 21:28:14 +0000 (21:28 +0000)]
openvpn: switch to new procd init script style

Signed-off-by: Luka Perkov <luka@openwrt.org>
SVN-Revision: 38331

11 years agopolarssl: update to 1.2.9
Luka Perkov [Mon, 7 Oct 2013 21:28:12 +0000 (21:28 +0000)]
polarssl: update to 1.2.9

Signed-off-by: Luka Perkov <luka@openwrt.org>
SVN-Revision: 38330

11 years agolantiq: fix ltq_ptm compile
John Crispin [Mon, 7 Oct 2013 20:13:32 +0000 (20:13 +0000)]
lantiq: fix ltq_ptm compile

Signed-off-by: John Crispin <blogic@openwrt.org>
SVN-Revision: 38329

11 years agolantiq: experia v8 fixes
John Crispin [Mon, 7 Oct 2013 15:03:00 +0000 (15:03 +0000)]
lantiq: experia v8 fixes

Signed-off-by: John Crispin <blogic@openwrt.org>
SVN-Revision: 38328

11 years agolantiq: fix fritz3370 dts file
John Crispin [Mon, 7 Oct 2013 15:02:56 +0000 (15:02 +0000)]
lantiq: fix fritz3370 dts file

Signed-off-by: John Crispin <blogic@openwrt.org>
SVN-Revision: 38327

11 years agolantiq: fix fritz image generation
John Crispin [Mon, 7 Oct 2013 15:02:52 +0000 (15:02 +0000)]
lantiq: fix fritz image generation

Signed-off-by: John Crispin <blogic@openwrt.org>
SVN-Revision: 38326

11 years agolantiq: fix eva partition split for squashfs
John Crispin [Mon, 7 Oct 2013 15:02:48 +0000 (15:02 +0000)]
lantiq: fix eva partition split for squashfs

Signed-off-by: John Crispin <blogic@openwrt.org>
SVN-Revision: 38325

11 years agoralink: fix asiarf image generation
John Crispin [Mon, 7 Oct 2013 15:02:44 +0000 (15:02 +0000)]
ralink: fix asiarf image generation

Signed-off-by: John Crispin <blogic@openwrt.org>
SVN-Revision: 38324

11 years agoralink: split mt7530 driver out of the gsw driver
John Crispin [Mon, 7 Oct 2013 15:02:39 +0000 (15:02 +0000)]
ralink: split mt7530 driver out of the gsw driver

Signed-off-by: John Crispin <blogic@openwrt.org>
SVN-Revision: 38323

11 years agoralink: build recovery images for dcs930 ip cam
John Crispin [Mon, 7 Oct 2013 15:02:35 +0000 (15:02 +0000)]
ralink: build recovery images for dcs930 ip cam

Signed-off-by: John Crispin <blogic@openwrt.org>
SVN-Revision: 38322

11 years agoralink: DIR610A1 Fix remaining issues
John Crispin [Mon, 7 Oct 2013 15:02:31 +0000 (15:02 +0000)]
ralink: DIR610A1 Fix remaining issues

This patch fixes:

 - Ethernet and Wireless LEDs
 - Wireless MAC Address
 - Remove whitespaces from dts file

Signed-off-by: Flavio Silveira <fggs@terra.com.br>
SVN-Revision: 38321

11 years agoralink: MT7620 eval board pinmux fixes
John Crispin [Mon, 7 Oct 2013 15:02:27 +0000 (15:02 +0000)]
ralink: MT7620 eval board pinmux fixes

Signed-off-by: John Crispin <blogic@openwrt.org>
SVN-Revision: 38320

11 years agoralink: make cevt systick selectable via kconfig
John Crispin [Mon, 7 Oct 2013 15:02:23 +0000 (15:02 +0000)]
ralink: make cevt systick selectable via kconfig

Signed-off-by: John Crispin <blogic@openwrt.org>
SVN-Revision: 38319

11 years agoralink: fix mt7620a pinmux
John Crispin [Mon, 7 Oct 2013 15:02:18 +0000 (15:02 +0000)]
ralink: fix mt7620a pinmux

Signed-off-by: John Crispin <blogic@openwrt.org>
SVN-Revision: 38318

11 years agotools: add tool to generate dcs93xl images
John Crispin [Mon, 7 Oct 2013 15:01:05 +0000 (15:01 +0000)]
tools: add tool to generate dcs93xl images

Signed-off-by: John Crispin <blogic@openwrt.org>
SVN-Revision: 38317

11 years agoodhcp6c: improve server selection, introduce "forceprefix" option
Steven Barth [Mon, 7 Oct 2013 08:43:33 +0000 (08:43 +0000)]
odhcp6c: improve server selection, introduce "forceprefix" option

SVN-Revision: 38314

11 years agoiproute2: update to 3.11
Luka Perkov [Mon, 7 Oct 2013 08:26:41 +0000 (08:26 +0000)]
iproute2: update to 3.11

Signed-off-by: Luka Perkov <luka@openwrt.org>
SVN-Revision: 38313

11 years agobrcm47xx: Cocci spatch "noderef"
Hauke Mehrtens [Mon, 7 Oct 2013 00:01:01 +0000 (00:01 +0000)]
brcm47xx: Cocci spatch "noderef"

sizeof when applied to a pointer typed expression gives the size of the
pointer.
Found by coccinelle spatch "misc/noderef.cocci"

Signed-off-by: Thomas Meyer <thomas@m3y3r.de>
SVN-Revision: 38312

11 years agobrcm47xx: move clock detection patches
Hauke Mehrtens [Mon, 7 Oct 2013 00:00:10 +0000 (00:00 +0000)]
brcm47xx: move clock detection patches

Signed-off-by: Hauke Mehrtens <hauke@hauke-m.de>
SVN-Revision: 38311

11 years agobrcm47xx: move and rename nvram gpio patch
Hauke Mehrtens [Sun, 6 Oct 2013 23:57:11 +0000 (23:57 +0000)]
brcm47xx: move and rename nvram gpio patch

Signed-off-by: Hauke Mehrtens <hauke@hauke-m.de>
SVN-Revision: 38310

11 years agobrcm47xx: update early printk patches
Hauke Mehrtens [Sun, 6 Oct 2013 23:55:35 +0000 (23:55 +0000)]
brcm47xx: update early printk patches

Early printk was disabled because it caused hangs on some devices, the
old patches were using the CFE console now we use the normal serial
console, it is at a constant address.

Signed-off-by: Hauke Mehrtens <hauke@hauke-m.de>
SVN-Revision: 38309

11 years agobrcm47xx: update board detection patches
Hauke Mehrtens [Sun, 6 Oct 2013 23:51:24 +0000 (23:51 +0000)]
brcm47xx: update board detection patches

Add patches like they are currently in the mainline mips tree and add
some more pending patches for board detection.

* Now the board name is shown under machine in /proc/cpuinfo.

Signed-off-by: Hauke Mehrtens <hauke@hauke-m.de>
SVN-Revision: 38308

11 years agokernel: b53: add switch name into printk when it fails.
Hauke Mehrtens [Sun, 6 Oct 2013 22:35:28 +0000 (22:35 +0000)]
kernel: b53: add switch name into printk when it fails.

Signed-off-by: Hauke Mehrtens <hauke@hauke-m.de>
SVN-Revision: 38307

11 years agobrcm47xx: b44: fix some problems with the phy
Hauke Mehrtens [Sun, 6 Oct 2013 18:31:32 +0000 (18:31 +0000)]
brcm47xx: b44: fix some problems with the phy

* do not try initialize a unused phy
* some improvements to the phylib patch
* do not turn the phy off when mac is off

Signed-off-by: Hauke Mehrtens <hauke@hauke-m.de>
SVN-Revision: 38306

11 years agoath9k: fix tx queueing issues after background scans
Felix Fietkau [Sat, 5 Oct 2013 12:12:21 +0000 (12:12 +0000)]
ath9k: fix tx queueing issues after background scans

Signed-off-by: Felix Fietkau <nbd@openwrt.org>
SVN-Revision: 38304

11 years agobuild: extend menu with few ubifs options
Luka Perkov [Fri, 4 Oct 2013 10:49:47 +0000 (10:49 +0000)]
build: extend menu with few ubifs options

Signed-off-by: Luka Perkov <luka@openwrt.org>
SVN-Revision: 38303

11 years agoopkg: add support for SHA256 verification
Steven Barth [Fri, 4 Oct 2013 05:56:20 +0000 (05:56 +0000)]
opkg: add support for SHA256 verification

based on a patch by Evan Hunt <each@isc.org>
Signed-off-by: Steven Barth <steven@midlink.org>
SVN-Revision: 38302

11 years agokirkwood: add support for UBI (3.10.x)
Luka Perkov [Fri, 4 Oct 2013 03:45:52 +0000 (03:45 +0000)]
kirkwood: add support for UBI (3.10.x)

Signed-off-by: Luka Perkov <luka@openwrt.org>
SVN-Revision: 38301

11 years agonetifd: update to latest version, adds macvlan support
Felix Fietkau [Thu, 3 Oct 2013 14:51:34 +0000 (14:51 +0000)]
netifd: update to latest version, adds macvlan support

Signed-off-by: Felix Fietkau <nbd@openwrt.org>
SVN-Revision: 38299

11 years agoimx6: update upstream pcie patches
Luka Perkov [Thu, 3 Oct 2013 13:06:42 +0000 (13:06 +0000)]
imx6: update upstream pcie patches

Signed-off-by: Luka Perkov <luka@openwrt.org>
SVN-Revision: 38298

11 years agokirkwood: enable dts compilation
Luka Perkov [Thu, 3 Oct 2013 02:50:03 +0000 (02:50 +0000)]
kirkwood: enable dts compilation

In 3.10 we might build those, so for now just add this functionality.

Signed-off-by: Luka Perkov <luka@openwrt.org>
SVN-Revision: 38297

11 years agokirkwood: fix initramfs images
Luka Perkov [Thu, 3 Oct 2013 02:50:02 +0000 (02:50 +0000)]
kirkwood: fix initramfs images

Signed-off-by: Luka Perkov <luka@openwrt.org>
SVN-Revision: 38296

11 years agokirkwood: update patches and configuration (3.10)
Luka Perkov [Thu, 3 Oct 2013 02:50:00 +0000 (02:50 +0000)]
kirkwood: update patches and configuration (3.10)

Signed-off-by: Luka Perkov <luka@openwrt.org>
SVN-Revision: 38295

11 years agobrcm47xx: add support for ipv6 in default network config
Hauke Mehrtens [Wed, 2 Oct 2013 23:22:18 +0000 (23:22 +0000)]
brcm47xx: add support for ipv6 in default network config

Modify /etc/init.d/netconfig to use UCI defaults for building the
default network config, which includes appropriate defaults for IPV6.

Signed-off-by: Nathan Hintz <nlhintz@hotmail.com>
SVN-Revision: 38294

11 years agobroadcom-diag: fix 'autoload' number of diag.ko
Hauke Mehrtens [Wed, 2 Oct 2013 23:14:08 +0000 (23:14 +0000)]
broadcom-diag: fix 'autoload' number of diag.ko

Change the autoload number from '5' to '05' so the module loads earlier (as intended).

Signed-off-by: Nathan Hintz <nlhintz@hotmail.com>
SVN-Revision: 38293

11 years agokernel: bgmac: fix BCM4707 patch
Hauke Mehrtens [Wed, 2 Oct 2013 23:12:46 +0000 (23:12 +0000)]
kernel: bgmac: fix BCM4707 patch

The bool logic was the wrong way around.

Signed-off-by: Hauke Mehrtens <hauke@hauke-m.de>
SVN-Revision: 38292

11 years agokernel: refresh patches
Hauke Mehrtens [Wed, 2 Oct 2013 22:26:43 +0000 (22:26 +0000)]
kernel: refresh patches

Signed-off-by: Hauke Mehrtens <hauke@hauke-m.de>
SVN-Revision: 38291

11 years agokernel: bcma: update to wireless-testing master-2013-10-01
Hauke Mehrtens [Wed, 2 Oct 2013 22:26:15 +0000 (22:26 +0000)]
kernel: bcma: update to wireless-testing master-2013-10-01

Signed-off-by: Hauke Mehrtens <hauke@hauke-m.de>
SVN-Revision: 38290

11 years agokernel: bgmac: move other bgmac patches to generic
Hauke Mehrtens [Wed, 2 Oct 2013 22:25:26 +0000 (22:25 +0000)]
kernel: bgmac: move other bgmac patches to generic

The bgmac driver will be used on the brcm47xx and the bcm53xx target.

Signed-off-by: Hauke Mehrtens <hauke@hauke-m.de>
SVN-Revision: 38289

11 years agokernel: bgmac: move bgmac patches already applied in mainline kernel to generic
Hauke Mehrtens [Wed, 2 Oct 2013 22:24:01 +0000 (22:24 +0000)]
kernel: bgmac: move bgmac patches already applied in mainline kernel to generic

The bgmac driver will be used on the brcm47xx and the bcm53xx target.
These are only the patches already applied in current net-next/master
branch.

Signed-off-by: Hauke Mehrtens <hauke@hauke-m.de>
SVN-Revision: 38288

11 years agopackage/index: fix index creating when building without signing
Steven Barth [Wed, 2 Oct 2013 16:38:28 +0000 (16:38 +0000)]
package/index: fix index creating when building without signing

Signed-off-by: Alexander Couzens <lynxis@fe80.eu>
SVN-Revision: 38287

11 years agoodhcp6c: validate hop-limit on RAs
Steven Barth [Wed, 2 Oct 2013 14:16:05 +0000 (14:16 +0000)]
odhcp6c: validate hop-limit on RAs

SVN-Revision: 38285

11 years agoAdd package signing infrastructure
Steven Barth [Wed, 2 Oct 2013 12:12:10 +0000 (12:12 +0000)]
Add package signing infrastructure

Add package signing key and certificate configuration options to the
"Image configuration" submenu. If enabled, the Packages.gz list will
be signed as file Packages.sig. The passphrase for the signing key can
be sourced from a file or entered by the user. The signing certificate
is automatically added to the firmware image if opkg-smime is selected.

Signed-off-by: Evan Hunt <each@isc.org>
Signed-off-by: Steven Barth <steven@midlink.org>
SVN-Revision: 38284

11 years ago6relayd: remove compat ULA setting by default (too many bad sideeffects)
Steven Barth [Wed, 2 Oct 2013 06:14:39 +0000 (06:14 +0000)]
6relayd: remove compat ULA setting by default (too many bad sideeffects)

SVN-Revision: 38283

11 years agoimx6: add default profile
Luka Perkov [Wed, 2 Oct 2013 00:33:49 +0000 (00:33 +0000)]
imx6: add default profile

Signed-off-by: Luka Perkov <luka@openwrt.org>
SVN-Revision: 38282

11 years agokirkwood: drop PROFILE from image names
Luka Perkov [Wed, 2 Oct 2013 00:33:48 +0000 (00:33 +0000)]
kirkwood: drop PROFILE from image names

Signed-off-by: Jonas Gorski <jogo@openwrt.org>
SVN-Revision: 38281

11 years agokirkwood: add linux 3.10 support
Luka Perkov [Wed, 2 Oct 2013 00:33:47 +0000 (00:33 +0000)]
kirkwood: add linux 3.10 support

Backport appropriate patches to allow using device tree only board
defintions.

Signed-off-by: Jonas Gorski <jogo@openwrt.org>
SVN-Revision: 38280

11 years agokirkwood: prepare for 3.10
Luka Perkov [Wed, 2 Oct 2013 00:33:45 +0000 (00:33 +0000)]
kirkwood: prepare for 3.10

Signed-off-by: Luka Perkov <luka@openwrt.org>
SVN-Revision: 38279

11 years agokirkwood: explicitly set CPU_TYPE to xscale
Luka Perkov [Wed, 2 Oct 2013 00:33:43 +0000 (00:33 +0000)]
kirkwood: explicitly set CPU_TYPE to xscale

Signed-off-by: Luka Perkov <luka@openwrt.org>
SVN-Revision: 38278

11 years agoc80211: add iwldvm to iwlwifi, this is needed to support the old wifi card.
Hauke Mehrtens [Tue, 1 Oct 2013 22:41:46 +0000 (22:41 +0000)]
c80211: add iwldvm to iwlwifi, this is needed to support the old wifi card.

This closes #14258

Signed-off-by: Hauke Mehrtens <hauke@hauke-m.de>
SVN-Revision: 38277

11 years agokernel: add industrialio-triggered-buffer.ko only when the config option is selected.
Hauke Mehrtens [Tue, 1 Oct 2013 22:40:32 +0000 (22:40 +0000)]
kernel: add industrialio-triggered-buffer.ko only when the config option is selected.

Signed-off-by: Hauke Mehrtens <hauke@hauke-m.de>
SVN-Revision: 38276

11 years agoep93xx: fix kmod-input-keyboard-ep93xx dependencies
Hauke Mehrtens [Tue, 1 Oct 2013 22:39:28 +0000 (22:39 +0000)]
ep93xx: fix kmod-input-keyboard-ep93xx dependencies

Signed-off-by: Hauke Mehrtens <hauke@hauke-m.de>
SVN-Revision: 38275

11 years agoep93xx: fix sound dependencies for ep93xx
Hauke Mehrtens [Tue, 1 Oct 2013 22:38:40 +0000 (22:38 +0000)]
ep93xx: fix sound dependencies for ep93xx

Signed-off-by: Hauke Mehrtens <hauke@hauke-m.de>
SVN-Revision: 38274

11 years agoat91: fix missing drivers/iio/adc/at91-adc.ko in kmod-at91-adc
Hauke Mehrtens [Tue, 1 Oct 2013 22:37:17 +0000 (22:37 +0000)]
at91: fix missing drivers/iio/adc/at91-adc.ko in kmod-at91-adc

Signed-off-by: Hauke Mehrtens <hauke@hauke-m.de>
SVN-Revision: 38273

11 years agonetifd: don't remove & readd addresses that only have a changed lifetime
Steven Barth [Tue, 1 Oct 2013 17:30:05 +0000 (17:30 +0000)]
netifd: don't remove & readd addresses that only have a changed lifetime

SVN-Revision: 38269

11 years agoprocd: fix incorrect use of sizeof() in vsnprintf()
Jo-Philipp Wich [Tue, 1 Oct 2013 14:41:07 +0000 (14:41 +0000)]
procd: fix incorrect use of sizeof() in vsnprintf()

SVN-Revision: 38268

11 years agoprocd: update to latest git head
John Crispin [Mon, 30 Sep 2013 20:48:51 +0000 (20:48 +0000)]
procd: update to latest git head

this fixes the syslog problem when using eglibc on x86.

Signed-off-by: John Crispin <blogic@openwrt.org>
SVN-Revision: 38267

11 years agolantiq: add support for VGV7519 KPN Experia box v8
John Crispin [Mon, 30 Sep 2013 20:48:46 +0000 (20:48 +0000)]
lantiq: add support for VGV7519 KPN Experia box v8

Signed-off-by: Maikel Bloemendal <mbloemendal@gmail.com>
Signed-off-by: John Crispin <blogic@openwrt.org>
SVN-Revision: 38266

11 years agoralink: dts fixes
John Crispin [Mon, 30 Sep 2013 20:48:39 +0000 (20:48 +0000)]
ralink: dts fixes

Signed-off-by: John Crispin <blogic@openwrt.org>
SVN-Revision: 38265

11 years agoralink: fix DIR-320-B1 board detection
John Crispin [Mon, 30 Sep 2013 20:48:34 +0000 (20:48 +0000)]
ralink: fix DIR-320-B1 board detection

https://dev.openwrt.org/ticket/14226

Signed-off-by: John Crispin <blogic@openwrt.org>
SVN-Revision: 38264

11 years agoralink: Change size of image for DIR-320 B1 to 8MB
John Crispin [Mon, 30 Sep 2013 20:48:27 +0000 (20:48 +0000)]
ralink: Change size of image for DIR-320 B1 to 8MB

http://patchwork.openwrt.org/patch/4102/

Signed-off-by: Vladimir Stolyarov <xakep6666@yandex.ru>
SVN-Revision: 38263

11 years agokernel: lib-textsearch needs to be auto probed for amanda nat to work
John Crispin [Mon, 30 Sep 2013 20:48:12 +0000 (20:48 +0000)]
kernel: lib-textsearch needs to be auto probed for amanda nat to work

https://dev.openwrt.org/ticket/14232

Signed-off-by: John Crispin <blogic@openwrt.org>
SVN-Revision: 38262

11 years agouClibc: fix scoped-IPv6-addresses in getnameinfo
Steven Barth [Mon, 30 Sep 2013 08:41:00 +0000 (08:41 +0000)]
uClibc: fix scoped-IPv6-addresses in getnameinfo

SVN-Revision: 38261

11 years agoscripts/gen-dependencies.sh: use the cross readelf (#12940)
Felix Fietkau [Sun, 29 Sep 2013 20:10:42 +0000 (20:10 +0000)]
scripts/gen-dependencies.sh: use the cross readelf (#12940)

Signed-off-by: Felix Fietkau <nbd@openwrt.org>
SVN-Revision: 38259

11 years agomac80211: add AP inactivity handling / probing fixes to improve stability
Felix Fietkau [Sun, 29 Sep 2013 19:42:00 +0000 (19:42 +0000)]
mac80211: add AP inactivity handling / probing fixes to improve stability

Signed-off-by: Felix Fietkau <nbd@openwrt.org>
SVN-Revision: 38257

11 years agomac80211: really fix the long standing wds ap tx power regression (#14153)
Felix Fietkau [Sun, 29 Sep 2013 12:53:21 +0000 (12:53 +0000)]
mac80211: really fix the long standing wds ap tx power regression (#14153)

Signed-off-by: Felix Fietkau <nbd@openwrt.org>
SVN-Revision: 38255