Bump kernel from 4.4.49 to 4.4.50
Compile tested: All targets
Run tested: ar71xx Archer C7 v2
Signed-off-by: Kevin Darbyshire-Bryant <kevin@darbyshire-bryant.me.uk>
This patch enables mtdsplit for the MikroTik subtarget.
It converts mach-rbspi.c to use a single "firmware" partition.
Finally, it converts the SPI NOR profile to metadata images:
the original combined-image format is gone, the images are now
"standard" sysupgrade images with metadata appended.
Note: kernel2minor apparently pads the kernel container to erase-block
boundary, but this is undocumented behaviour, so we do not rely on it
and call pad-to anyway.
Note: in platform.sh, the boards are tested last in platform_check_image()
as this should eventually become the "*)" default case when more devices
switch to metadata images.
Signed-off-by: Thibaut VARENE <hacks@slashdirt.org>
The RouterBOOT bootloader does not care where the kernel lives in the SPI
flash, all that matters is that the kernel is wrapped in the custom yaffs
container as generated by kernel2minor.
This container has a fixed signature as follows:
00000000 00 00 00 01 00 00 00 01 ff ff 6b 65 72 6e 65 6c |..........kernel|
This patch adds mtdsplit support for identifying that signature and
triggering the search for the rootfs. rootfs is expected at EB boundary since
we use wget mtd_find_rootfs_from(). We make no use of the yaffs file size
field because it contains invalid data in the image generated by kernel2minor.
Signed-off-by: Thibaut VARENE <hacks@slashdirt.org>
The fakeroot header is expected by the netgear bootloader in the last
64 bytes of the last block used by the uImage. With the current
linux-4.9 uImage being more than 128k smaller than the linux-4.4 uImage
the bootloader was unable to locate the rootfs ih_magic and was
refusing to load and start the kernel.
Fixes: FS#542
Signed-off-by: Thomas Reifferscheid <thomas@reifferscheid.org>
CONFIG_FB_EFI and CONFIG_FB_SIMPLE are needed to display console text on
EFI framebuffer.
CONFIG_FB_EFI is needed when the kernel is directly launched via EFI
shell or EFI startup.nsh script.
CONFIG_FB_SIMPLE is needed when launching the kernel via grub2 efi. In
this case, grub2 has prepared a gfxterm framebuffer and the kernel just
need to use the already prepared grub's gfxterm framebuffer to display
console text.
Signed-off-by: Alif M. Ahmad <alive4ever@live.com>
For AR71XX devices, GMAC1 always connects port 0 of the built-in switch,
as the CPU port.
This patch sets correct vlan for some devices with wrong settings:
a) mark port 0 as CPU port, tagged
b) reverse port order, marking these ports untagged
Signed-off-by: Weijie Gao <hackpascal@gmail.com>
This patch adds support for the TP-Link TL-WR802N Version 2.
Specification:
- SoC: Qualcomm Atheros QCA9533 (560 MHz, MIPS 24Kc)
- RAM: 32 MiB
- Storage: 4 MiB of FLash on board
- Wireless: Built into QCA9533 (Honey Bee), PHY modes b/g/n
- Ethernet: 1x100M (port0)
Installation through OEM Web Interface:
- Connect to TL-WR802N by Ethernet or Wi-Fi
- Go to http://192.168.0.1 (default user is "admin" & password is "admin)
- Go to "System Tools -> Firmware Upgrade"
- Browse for firmware "*.factory-us.bin" or "*.factory-eu.bin" for eu model
- Click upgrade
Installation through tftp:
Note: T_OUT, T_IN and GND on the board must be connected to USB TTL
Serial Configuration 115200 8n1
- Boot the TL-WR802N
- When "Autobooting in 1 seconds" appears type "tpl" followed by enter
- Connect to the board Ethernet port
(IPADDR: 192.168.1.1, ServerIP: 192.168.1.10)
- tftpboot 0x80000000 <Firmware Image Name>
- Record the result of "printenv bootcmd"
- Enter "erase <Result of 'printenv bootcmd'> +0x3c0000"
(e.g erase 0x9f020000 +0x3c0000)
- Enter "cp.b 0x80000000 <Result of 'printenv bootcmd'> 0x3c0000"
(e.g cp.b 0x80000000 0x9f020000 0x3c0000)
- Enter "bootm <Result of 'printenv bootcmd'>"
(e.g bootm 0x9f020000)
Signed-off-by: Thomas Roberts <tom.p.roberts@gmail.com>
Fixes the following warning on first boot if the ltq_atm modules are
not included in the image:
ls: /lib/modules/4.4.49/ltq_atm*: No such file or directory
Signed-off-by: Mathias Kresin <dev@kresin.me>
Use the nas0 interface for the netdev trigger as default. Use the ptm0
interface for xRX200 boards to match the default wan interface set in
02_network.
Signed-off-by: Mathias Kresin <dev@kresin.me>
Move the code to check if the current system is a system with vdsl
support to a dedicate function to make it reusable.
Signed-off-by: Mathias Kresin <dev@kresin.me>
QCA956x is configured like AR934x, not like the older chips.
Should fix ethernet hangs when using the WAN port without SGMII
Signed-off-by: Felix Fietkau <nbd@nbd.name>
This makes the Geode images actually useful again. The Geos profile
should include the relevant hardware for that board, and the Default
profile adds the via-rhine adapter which seems to have been present in
the net5501 and alix targets killed in commit 9e0759ea26 ("x86: merge
all geode based subtargets into one").
Signed-off-by: David Woodhouse <dwmw2@infradead.org>
Lima is a module for the (IoE) internet of everything applications,
based on Qualcomm/Atheros QCA4531.
Specification:
- 650/600/216 MHz (CPU/DDR/AHB)
- 64 MB of RAM (DDR2)
- 32 MB of FLASH
- 2T2R 2.4 GHz
- 2x 10/100 Mbps Ethernet
- USB 2.0 Host
- PCIe
- UART for serial console
- 14x GPIO
Flash instruction:
1. Download lede-ar71xx-generic-lima-squashfs-sysupgrade.bin
and rename it to 8dev_recovery.bin
2. Binary file copy to USB stick and plug into slot USB
3. Press S2 (reset) button
4. Press and hold the S1 button by 5 seconds
5. Wait ~40 seconds to complete recovery
Tested on:
- Lima development kit
Signed-off-by: Karol Dudek <karoiz@sli.pl>
commit 33b72b8e0f
"ar8216: adjust ATU flushing in case of link changes"
introduced portwise flushing on link down events. Now the ARL table could
be in a chaotic state after boot where ar8xxx_sw_get_arl_table looped
forever (depending on the entries collected while booting).
Signed-off-by: Günther Kelleter <guenther.kelleter@devolo.de>
GPIO18 and GPIO19 on OMEGA2(+) should be GPIO mode, enable PWM lead to a conflict
[ 0.290633] rt2880-pinmux pinctrl: pin io18 already requested by pinctrl; cannot claim for 10005000.pwm
[ 0.299722] rt2880-pinmux pinctrl: pin-18 (10005000.pwm) status -22
[ 0.305729] rt2880-pinmux pinctrl: could not request pin 18 (io18) from group pwm0 on device rt2880-pinmux
[ 0.315131] mtk-pwm 10005000.pwm: Error applying setting, reverse things back
Keep PWM disabled.
Signed-off-by: Furong Xu <xfr@outlook.com>
The Asus WL-500W buttons are active high,
but the software treats them as active low.
Fix the inverted logic.
Signed-off-by: Mirko Parthey <mirko.parthey@web.de>
Signed-off-by: Rafał Miłecki <rafal@milecki.pl>
BCM4704 SoC has only USB 1.1 core:
ssb: Core 3 found: USB 1.1 Hostdev (cc 0x808, rev 0x03, vendor 0x4243)
but this device actually comes with two USB 2.0 ports.
It appears embedded controller isn't used but instead there are two PCI
attached controllers (next to the BCM4321 wireless card):
1106:3038 VT82xx/62xx UHCI USB 1.1 Controller
1106:3104 USB 2.0
Choose a set of USB drivers which actually support this hardware.
Signed-off-by: Mirko Parthey <mirko.parthey@web.de>
Signed-off-by: Rafał Miłecki <rafal@milecki.pl>
When memory is tight, modules may need to be loaded into vmalloc()
space. The code then has to generate jump trampolines which enable
relocations between vmalloc space and physical address space.
The code had a bug that was freeing these trampolines even when the
module was successfully loaded.
Signed-off-by: Felix Fietkau <nbd@nbd.name>
The cAP 2nD, mAP 2nD and wAP 2nD apparently all share the same QCA953x/SPI-NOR
platform.
This patch adds preliminary kernel support based on init data used in
Mikrotik's linux-3.3.5 kernel patch. Since this is totally experimental and
untested, the necessary glue in base-files has not been added, the code is
provided as a starting point in case somebody with that hardware wants to help
getting it supported.
The cAP lite (cAPL 2nD) should also be supported once it is determined whether
or not it uses the same board identifier as the mAP lite.
Code has been successfully built and has no side effect (no impact on already
supported devices).
Signed-off-by: Thibaut VARENE <hacks@slashdirt.org>
This patch adds support for the MikroTik RouterBOARD hEX lite
https://routerboard.com/RB750r2
Specifications:
- SoC: Qualcomm QCA9531 (850MHz)
- RAM: 64MB
- Storage: 16MB NOR SPI flash
- Ethernet: 5x100M (1 PoE in)
This ethernet router is based on the same platform as the hEX PoE lite (it
shares the same board identifier), but has no USB and no PoE out.
Signed-off-by: Thibaut VARENE <hacks@slashdirt.org>
This patch adds support for the MikroTik RouterBOARD hEX PoE lite
https://routerboard.com/RB750UPr2
Specifications:
- SoC: Qualcomm QCA9531 (650MHz)
- RAM: 64MB
- Storage: 16MB NOR SPI flash
- Ethernet: 5x100M (1 PoE in, 1 PoE out)
- USB: Type A
This ethernet router is based on the same platform as the wireless router hAP.
Signed-off-by: Thibaut VARENE <hacks@slashdirt.org>
This patch implements support for the hAP lite in mach-rbspi.c
hAP lite was supported via mach-rb941.c, so this file is removed
as well as the corresponding build bits.
Signed-off-by: Thibaut VARENE <hacks@slashdirt.org>
This patch adds support for the MikroTik RouterBOARD mAP lite
https://routerboard.com/RBmAPL-2nD
Specifications:
- SoC: Qualcomm QCA9533 (650MHz)
- RAM: 64MB
- Storage: 16MB NOR SPI flash
- Wireless: builtin QCA9533, 2x2:2
- Ethernet: 1x100M
This is another 16M SPI NOR mikrotik device. The machine file is named
mach-rbspi.c because I plan to add support for several of the other spi-based
RouterBOARD devices in subsequent patches: they share most of the hardware
and thus the same codebase.
Signed-off-by: Thibaut VARENE <hacks@slashdirt.org>
Some boards were apparently forgotten when ralink,portmap was renamed
to mediatek,portmap -- probably because they used the long obsolete
ralink,port-map attribute.
If this commit breaks ethernet wan/lan assignment, this is because
the port-map attribute wasn't actually parsed, you'll have to replace
"wllll" by "llllw" in the dts file belonging to that board (and send
a patch doing that!)
Signed-off-by: Daniel Golle <daniel@makrotopia.org>
This adds the default LED and network settings for the PC Engines APU2
when running under the x86 target.
[dwmw2: Change Ethernet port setup]
Signed-off-by: Chris Blake <chrisrblake93@gmail.com>
Based on a patch from Chris Blake <chrisrblake93@gmail.com>, except let's
do it by using the LED configuration instead of hard-coding it for each
board type. And try using /bin/board_detect to do the default behaviour,
on the first boot where the config hasn't yet been generated.
Signed-off-by: David Woodhouse <dwmw2@infradead.org>
This change moves the files in 657418d to the root of the x86 target.
This is done in preperation for adding more devices under other
subtargets.
CC: David Woodhouse <dwmw2@infradead.org>
Signed-off-by: Chris Blake <chrisrblake93@gmail.com>
The DWR-512 embeds the hw slic device si3210. This device have the IRQ line
attached to the gpio1. This patch export the gpio1 with proper name and
parameters to the sysfs.
Signed-off-by: Giuseppe Lippolis <giu.lippolis@gmail.com>
this caused
'mv: can't rename '/mnt/sysupgrade.tgz': No such file or directory'
when running sysupgrade -n
Signed-off-by: John Crispin <john@phrozen.org>
Add support for the EBR-2310, which is almost identical to the DIR-615
rev E4, without the wifi.
Signed-off-by: Alexandru Gagniuc <mr.nuke.me@gmail.com>
TP-Link TL-WA850RE v2 is a wall-plug N300 Wi-Fi range extender,
based on Qualcomm/Atheros QCA9533 v2.
Short specification:
- 550/391/195 MHz (CPU/DDR/AHB)
- 1x 10/100 Mbps Ethernet
- 32 MB of RAM (DDR1)
- 4 MB of FLASH
- 2T2R 2.4 GHz
- 2x internal antennas (embedded on PCB)
- 9x LED (all can be turned off with GPIO15), 2x button
- UART (J3) header on PCB
Flash instruction: use "factory" image directly in vendor GUI.
Warning: this device does not include any kind of recovery mechanism
in the bootloader and disassembling process is not trivial.
You can access vendor firmware over serial line using:
- login: root
- password: sohoadmin
Image was tested only in US version of the device, but should work
also with the same device version sold in other countries.
Available FLASH space, with LEDE trunk, is only 240 KB.
Signed-off-by: Piotr Dymacz <pepe2k@gmail.com>
The WN3000RPv3 is a repeater with a single ethernet port. Setting up the
switch, even to disable it, is unnecessary and possibly confusing.
Configure LAN as eth0 instead.
Signed-off-by: Thibaut VARENE <hacks@slashdirt.org>
The list of v2/v3 devices is getting longer and makes reading the names of
the profiles unnecessary hard.
Signed-off-by: Sven Eckelmann <sven.eckelmann@open-mesh.com>
Fixes the following issue:
root@LEDE:/# sysupgrade /tmp/lede-mvebu-armada-388-clearfog-sdcard.img.gz
Saving metaconfig...
Image metadata not found
Use sysupgrade -F to override this check when downgrading or flashing to vendor firmware
Image check 'fwtool_check_image' failed.
root@LEDE:/#
Acked-by: Felix Fietkau <nbd@nbd.name>
Signed-off-by: Jonas Gorski <jonas.gorski@gmail.com>
The firmware hotplug script tries to read the mac address from a
partition with the name "uboot-env" which does not exist (instead it's
name is uboot_env). This broke calculation of the new checksum (after
patching the mac address) which resulted in ath9k refusing to use the
EEPROM data.
The original error reported by ath9k was:
PCI: Enabling device 0000:00:0e.0 (0000 -> 0002)
ath: phy0: Bad EEPROM checksum 0x2523
ath: phy0: Unable to initialize hardware; initialization status: -22
ath9k 0000:00:0e.0: Failed to initialize device
Fixes: a20616863d ("lantiq: use ath9k device tree bindings
binding/owl-loader")
Signed-off-by: Martin Blumenstingl <martin.blumenstingl@googlemail.com>
platform_find_rootfspart() fails if the kernel partition comes before the
rootfs partition. The proposed patch fixes this while preserving what I
understand was the original idea: stop at first match.
Signed-off-by: Thibaut VARENE <hacks@slashdirt.org>
Upstream linux commit 7523e4dc5057e "module: use a structure to
encapsulate layout." moves some of the struct module members into a
separate struct module_layout.
Signed-off-by: Martin Blumenstingl <martin.blumenstingl@googlemail.com>
Do not put the probe function into the __init section, but use the normal
section. This fixes a section mismatch warning.
Signed-off-by: Hauke Mehrtens <hauke@hauke-m.de>
The following patches were dropped because they are already applied
upstream:
0012-pinctrl-lantiq-fix-up-pinmux.patch
0013-MTD-lantiq-xway-fix-invalid-operator.patch
0014-MTD-lantiq-xway-the-latched-command-should-be-persis.patch
0015-MTD-lantiq-xway-remove-endless-loop.patch
0016-MTD-lantiq-xway-add-missing-write_buf-and-read_buf-t.patch
0017-MTD-xway-fix-nand-locking.patch
0044-pinctrl-lantiq-introduce-new-dedicated-devicetree-bi.patch
0045-pinctrl-lantiq-Fix-GPIO-Setup-of-GPIO-Port3.patch
0046-pinctrl-lantiq-2-pins-have-the-wrong-mux-list.patch
0047-irq-fixes.patch
0047-mtd-plat-nand-pass-of-node.patch
0060-usb-dwc2-Add-support-for-Lantiq-ARX-and-XRX-SoCs.patch
0120-MIPS-lantiq-add-support-for-device-tree-file-from-bo.patch
0121-MIPS-lantiq-make-it-possible-to-build-in-no-device-t.patch
122-MIPS-store-the-appended-dtb-address-in-a-variable.patch
The PHY driver was reduced to the code adding the LED configuration,
the rest is already upstream:
0023-NET-PHY-adds-driver-for-lantiq-PHY11G.patch
The SPI driver was replaced with the version pending for upstream
inclusion:
New driver:
0090-spi-add-transfer_status-callback.patch
0091-spi-lantiq-ssc-add-support-for-Lantiq-SSC-SPI-controller.patch
Old driver:
0100-spi-add-support-for-Lantiq-SPI-controller.patch
Signed-off-by: Hauke Mehrtens <hauke@hauke-m.de>
This patch add ZyXEL NSA325 2-Bay Media Server
The ZyXEL NSA325 device is a Kirkwood based NAS:
- SoC: Marvell 88F6702 1600Mhz
- SDRAM memory: 512MB DDR2 400Mhz
- Gigabit ethernet: Marvell Alaska
- Flash memory: 128MB
- 1 Power button
- 1 Power LED (blue)
- 5 Status LED (green/red)
- 1 Copy/Sync button
- 1 Reset button
- 2 SATA II ports (internal)
- 2 USB 2.0 ports (back)
- 1 USB 3.0 port (front)
- Fan (fixed speed)
- hardware watchdog in a mcu
Basically a bigger, more powerful version of NSA310,
installation is the same as they share the same flash layout.
A notable difference is that there is a hardware watchdog
in a mcu on the board, which is disabled by default in the LEDE u-boot.
The watchdog is also disabled with a GPIO activation through
raw register change when kwbooting or it would reset the board before
the new uboot was transferred.
Signed-off-by: Alberto Bursi <alberto.bursi@outlook.it>
Signed-off-by: Felix Fietkau <nbd@nbd.name> [remove dead code]
switch nsa310b mac address reading to mtd_get_mac_ascii helper as
it seems the fw_env.config file is created way later than when
network is set up, when I tested I still had that file included
in the image through /files folder.
Signed-off-by: Alberto Bursi <alberto.bursi@outlook.it>
When we merged all the Geode boards into one generic target, the default
network and LED configuration was lost. Put it back.
Fixes: 9e0759ea26 ("x86: merge all geode based subtargets into one")
Signed-off-by: David Woodhouse <dwmw2@infradead.org>
Adds preliminary kernel 4.9 support for this target.
- Refreshed/Updated all patches
Added 3 new patches:
- 093 --> Add virtual PCI MMIO mapping
- 230 --> Remove deprecated code
- 240 --> Rework AT24 eeprom code to use the new NVMEM API
Compiled & tested on cns3xxx (gw2388)
Signed-off-by: Koen Vandeputte <koen.vandeputte@ncentric.com>
Add support for the Observa Telecom VH4032N router.
This is another BCM6368 router, 128 MB RAM, 32MB flash and 3 USB
host ports.
The wifi chip is an onboard Broadcom BCM43222.
Signed-off-by: Daniel Gonzalez Cabanelas <dgcbueu@gmail.com>
[jonas.gorski: use gpio-hog instead of abusing ephy-reset]
Signed-off-by: Jonas Gorski <jonas.gorski@gmail.com>
Remove CONFIG_VLAN_8021Q overrides for two targets
These features are built into the kernel image for all targets
Signed-off-by: Felix Fietkau <nbd@nbd.name>
Probe the switch through DT instead of a platform device. This fixes
probe, as GPIO offsets are now properly accounted for.
Fixes the following issue:
[ 0.802953] 8021q: 802.1Q VLAN Support v1.8
[ 0.810518] rtl8366_smi: gpio_request failed for 18, err=-517
Signed-off-by: Jonas Gorski <jonas.gorski@gmail.com>
BCM6362 and BCM3380 seem to share the same PRID. Disable serial output
for them until we can find a way to tell them apart reliably.
Fixes: 8f3cfe4ba2 ("brcm63xx: lzma-loader: add BCM3380 support")
Signed-off-by: Jonas Gorski <jonas.gorski@gmail.com>
This patch adds support for the ADB P.DG AV4202N aka Pirelli PRG AV4202N.
Signed-off-by: Paul Wassi <p.wassi@gmx.at>
Signed-off-by: Jonas Gorski <jonas.gorski@gmail.com>
Add support for mapping some GPIO lines to IRQs. GPIO to IRQ mappings
were found out through experimentation, helped by having the GPIO as
output still toggling the IRQ input.
Based-on: http://patchwork.ozlabs.org/patch/660534/
Signed-off-by: Jonas Gorski <jonas.gorski@gmail.com>
Setting the clear bit for an interrupt seems to cause interrupts to be
deasserted again immediately. So unset the bit for BCM6318 to allow
subsequent interrupts to still work.
Signed-off-by: Jonas Gorski <jonas.gorski@gmail.com>
Default CONSOLE value for ar71xx target is "ttyS0,115200", so there
is no need to specify the same value in device profiles.
Signed-off-by: Piotr Dymacz <pepe2k@gmail.com>
This commit reverts 42f3c1f ("arc770: fix broken upstream change")
and simultaneously adds back-ported fix from Linus' tree:
a524c218bc94 ("ARC: [arcompact] brown paper bag bug in unaligned
access delay slot fixup").
Note mentioned patch will appear in stable trees soon as well so
IMHO there's not much sense in adding this separate patch in lede/master branch.
As well as we will get rid of it here in 17.01 once we bump 4.4 kernel here
later down the line.
Signed-off-by: Alexey Brodkin <abrodkin@synopsys.com>
Cc: Felix Fietkau <nbd@nbd.name>
Cc: John Crispin <john@phrozen.org>
Cc: Jo-Philipp Wich <jo@mein.io>
mtdsplit_lzma requires that the rootfs be aligned to a block boundary.
Pad the kernel partition to make this so.
Signed-off-by: Claudio Leite <leitec@gmail.com>
buildbot fails to build without regmap, which I couldn't reproduce locally.
Hopefully this fixes it.
Signed-off-by: Álvaro Fernández Rojas <noltari@gmail.com>
64 bit support for Raspberry Pi 3 is experimental. However, bcm2709 subtarget
should serve as a stable and compatible 32 bit support.
Signed-off-by: Álvaro Fernández Rojas <noltari@gmail.com>
This adds support for Raspberry Pi 3 on the bcm2709 subtarget, in order to
keep 32 bit image support when 64 bit support is added on bcm2710 subtarget.
Signed-off-by: Álvaro Fernández Rojas <noltari@gmail.com>
Initial idea was to use package with this PHY driver for devices that
need it. Unfortunately this can't work as bgmac is built-in and PHY
probing happens before loading modules - it results in PHY subsystem
picking default (generic) PHY driver.
There were two ways of solving this:
1) Making bcm53xx use bgmac as module
xor
2) Built-in Broadcom PHY driver
After some quick discussion it seems we can simply built-in the driver
as increased kenel size is relatively small (1805 B).
Signed-off-by: Rafał Miłecki <rafal@milecki.pl>
These are patches queued for 4.11. It adds support for even more hw and
removes some annoying WARN_ONCE.
Signed-off-by: Rafał Miłecki <rafal@milecki.pl>
It's needed for macaddr_add.
Fixes: 50efd403e6 ("bcm53xx: set WAN MAC address to don't share one with LAN interface")
Signed-off-by: Rafał Miłecki <rafal@milecki.pl>
After "73d923e base-files: emit tagged switch configuration by default"
some default network configurations are broken because the lan and wan
ifnames are forcibly set to untagged netdevs.
Adjust the offending set_interfaces_lan_wan() calls to use the proper
tagged device names.
Signed-off-by: Jo-Philipp Wich <jo@mein.io>
It just makes code consistent. This trivial change may be a 17.01
candidate to provide simpler backporting experience.
Signed-off-by: Rafał Miłecki <rafal@milecki.pl>
In kernel 4.9 it's already present so we don't want to overwite it (with
older & API incompatible version).
Signed-off-by: Rafał Miłecki <rafal@milecki.pl>
Refresh patches for all targets that support kernel 4.4.
Compile-tested on all targets that use kernel 4.4 and aren't marked
broken, except arc770 and arch38 due to broken toolchain.
Runtime-tested on ar71xx, octeon, ramips and x86/64.
Signed-off-by: Stijn Tintel <stijn@linux-ipv6.be>
