Gives the user the control to select the correct WAN IPv4 address to be used by the 6rd tunnel when mutiple WAN interfaces are configured
Signed-off-by: Hans Dedecker <dedeckeh@gmail.com>
SVN-Revision: 40566
Depend on lantiq board.
ARV7510 depends on ARV4510 because (according to OpenWrt wiki) both boards are identical except for the unsupported WiFi. Moreover, ARV7510 isn't especifically supported and ARV4510 images valid.
Signed-off-by: Álvaro Fernández Rojas <noltari@gmail.com>
SVN-Revision: 40544
By default kobs-ng will use the size of /dev/mtd0 for the IMX bootstreams.
This patch allows overriding that size in order to reserve space at the end
of the mtd partition for u-boot.img in the case of using an SPL bootloader.
For example, to flash an SPL bootloader that loads u-boot.img from 0xe0000
(14MB) offset in a 16MB /dev/mtd0:
kobs-ng init -v -x --search_exponent=1 --chip_0_size=0xe00000 SPL
nandwrite --start=0xe00000 --pad /dev/mtd0 u-boot.img
Signed-off-by: Tim Harvey <tharvey@gateworks.com>
SVN-Revision: 40532
The /sbin/block binary now requires and additional shared library, therfore set
LD_LIBRARY_PATH accordingly when invoking it from the temporary overlay.
Signed-off-by: Jo-Philipp Wich <jow@openwrt.org>
SVN-Revision: 40526
This patch implements support for 802.11s protected mesh wireless networks (using authsae) in the netifd framework.
Until meshd-nl80211 implements a proper -P option for the PID file, this uses shell backgrounding in order to be able to get the PID for the process.
Signed-off-by: Vittorio Gambaletta <openwrt@vittgam.net>
SVN-Revision: 40497
This is required to make object "session" procedure "access" dump ACLs
when no "object" or no "function" parameters are passed.
Signed-off-by: Rafał Miłecki <zajec5@gmail.com>
SVN-Revision: 40496
Patches created from tree:
git@github.com:danielschwierzeck/u-boot-lantiq.git
v2013.10..u-boot-lantiq-v2013.10-openwrt4
Signed-off-by: Daniel Schwierzeck <daniel.schwierzeck@gmail.com>
SVN-Revision: 40482
r39995 introduced a new parameter wps_pbc_in_m1 to wifi wps config, but
apparently did not provide a default value 0.
When that option's non-existing value is later evaluated in
/lib/netifd/hostapd.sh, it causes the "bad number" error to be logged in
syslog if user has not set the wps_pbc_in_m1 option. The error materialises
only if user has enabled wps.
Sat Apr 12 13:25:01 2014 daemon.notice netifd: radio1 (1254): sh: bad number
Sat Apr 12 13:25:01 2014 daemon.notice netifd: radio0 (1253): sh: bad number
Discussion in bug 15508: https://dev.openwrt.org/ticket/15508#comment:3
Error is caused by line 282:
https://dev.openwrt.org/browser/trunk/package/network/services/hostapd/files/netifd.sh#L282
My patch sets the parameter's default value to 0, which does nothing. The
default might also be set a bit later in the function, but this felt like the
most clear place to do that.
Signed-off-by hnyman <hannu.nyman@iki.fi>
SVN-Revision: 40469
* atm module needs to be loaded before linux-atm
* use absolute firmware paths
* extended validation
* add a script for mounting an optional firmware partition
Signed-off-by: John Crispin <blogic@openwrt.org>
SVN-Revision: 40460
zsmalloc was promoted from staging, while bluetooth gained a dependency
to a library shared with 802.15.4 for 6lowpan support.
Signed-off-by: Jonas Gorski <jogo@openwrt.org>
SVN-Revision: 40379
Some boards have the WLAN EEPROM stored in flash in big-endian format,
whereas the driver requires the EEPROM in little-endian format. The
conv=swab option in dd is particularly useful in this case. This patch
backports this feature from busybox-1.22 based on the following commits:
commit b941316ae5313be523b64f0a9151ee4decb2b35b dd: support conv=swab
commit 8395bd3f52f8ed46fa3ffc316b2d113afa748bae dd: fix example in a
comment. No code changes.
commit 5b9910f0a4a1b7976c46e6f849aaa263180e5521 dd: fail if swab is
attempted on odd-sized block
commit 0ff0b320a2dbb5e0b5fa245ffd9b2648d7026843 dd: code shrink
v2: include the commit messages from upstream
Signed-off-by: Matti Laakso <malaakso@elisanet.fi>
SVN-Revision: 40327
The ltq-adsl-mei driver allocates memory for the ADSL firmware in a
rather stupid way, leading to several 128k allocations, which fail when
many services are enabled. This patch tries to allocate 64 kiB chunks,
and only falls back to larger allocations if the returned pointers are
not correctly aligned. Fixes out-of-memory errors on Danube boards with
32 MiB RAM.
Signed-off-by: Matti Laakso <malaakso@elisanet.fi>
SVN-Revision: 40325
fixes incremental build with change to CONFIG_DROPBEAR_ECC
drop --with-shared which is unknown to configure
Patch by Catalin Patulea <cat@vv.carleton.ca>
SVN-Revision: 40300
Without timeout mechanism, if ssh client disconnected without sending
FIN or RST, forked dropbear servers would hang there for
KEX_RETRY_TIMEOUT seconds (8 hours).
TCP keepalive is not implemented in dropbear yet, thus the name
SSHKeepAlive.
300 seconds in this patch is selected from the default value of
ServerAliveInterval for Debian ssh client (See man ssh_config).
Signed-off-by: Yousong Zhou <yszhou4tech@gmail.com>
SVN-Revision: 40299
The antenna gain attribute has been removed from the upstream regulatory database.
Incorperate patch from Luis R. Rodriguez to skip the parsing of the gain attribute
in genregdb.awk in order to emit proper rules again.
Also refresh fuzzy patches while we're at it.
Signed-off-by: Jo-Philipp Wich <jow@openwrt.org>
SVN-Revision: 40296
NAND flash is very likely to contain bad blocks.
Currently, mtd and therefore sysupgrade fails when it encounters a single bad block, potentially leaving an unbootable system.
This patch allows the mtd utility to skip bad blocks in NAND flash and complete sysupgrade successfully.
Patch by: Matthew Redfearn <matt.redfearn@nxp.com>
Signed-off-by: Felix Fietkau <nbd@openwrt.org>
SVN-Revision: 40021
Option pbc_in_m1 is being used as a WPS capability discovery
workaround for PBC with Windows 7.
Add possibility to enable this workaround from UCI.
To enable it, turn on wps and set wps_pbc_in_m1 parameter to 1.
Signed-off-by: Pawel Kulakowski <pawel.kulakowski@tieto.com>
SVN-Revision: 39995
This patch removes dependancy of PPP from chat application as chat application can be used for other serial communication as well that is not dependant on PPP and therefore one should be able to install chat without PPP. There also are no dependencies within chat application for PPP.
Signed-off-by: Oskari Rauta <oskari.rauta at gmail.com>
SVN-Revision: 39992
To be prepared to support keepalived 1.2.10 and higher we need libnl3 or
higher. The attached patch
add some defines so that it can be build be libnl-tiny.
Patch by Thomas Heil
Signed-off-by: Felix Fietkau <nbd@openwrt.org>
SVN-Revision: 39991
Add Bosch C_CAN/D_CAN driver as a kernel package. The D_CAN controller is used
on BeagleBoneBlack.
Signed-off-by: Jan Kardell <jan.kardell@telliq.com>
SVN-Revision: 39990
Build all of the can support as modules, otherwise the kernel image is
changed when a can kernel package is selected, because the can protocol
is then built in into the kernel image.
Signed-off-by: Jan Kardell <jan.kardell@telliq.com>
SVN-Revision: 39989
This patch enables including the firmware for Intel Dual Band Wireless-N 7260 and Intel Dual Band Wireless-AC 7260 cards in iwlagn.
Signed-off-by: Matthew Reeve <mreeve@tenxnetworks.com>
SVN-Revision: 39988
Trunk and the latest backport to Attitude Adjustment have the
kmod-rt2800-soc package hidden for the ramips_rt288x target.
This makes the dependency match the dependency given in the
TITLE field.
Signed-off-by: Claudio Leite <leitec@staticky.com>
SVN-Revision: 39976
A bunch of packages such as sispctl look for libusb-config to find it, install
libusb-config in staging_dir/*/host/bin which is in the TARGET_PATH.
Signed-off-by: Florian Fainelli <florian@openwrt.org>
SVN-Revision: 39975
The kernel Kconfig dependency lists X86, and we will fail building this module
on other platforms such as Malta.
Signed-off-by: Florian Fainelli <florian@openwrt.org>
SVN-Revision: 39974
Minor fix to existing patch which did not correctly implement -n support for the export command in ash. The VEXPORT flag of a variable was set in the first call to setvareq() and then remained set in the subsequent call, because the variable already existed in the environment.
Although /lib/functions.sh by default does not export variables (NO_EXPORT=1), this issue was leading to dirty environments.
Without this fix:
(ash) export A=1; export -n B=2; /bin/ash -c "echo \$A, \$B"
1, 2
With this fix:
(ash) export A=1; export -n B=2; /bin/ash -c "echo \$A, \$B"
1,
Signed-off-by: Christophe Lermytte <openwrt@lermytte.be>
SVN-Revision: 39968
