openwrt/staging/blocktrron.git
7 years agoramips: add support for Asus RT-AC51U
Ørjan Malde [Fri, 5 May 2017 15:54:26 +0000 (17:54 +0200)]
ramips: add support for Asus RT-AC51U

Specification:
 - SoC: MediaTek MT7620A (580 MHz)
 - RAM: 64 MiB (Winbond W9751G6JB-25)
 - Flash: 16 MiB (Spansion S25FL128SAIF00)
 - LAN: x4 100M
 - WAN: x1 100M
 - Others: USB 2.0, reset button, wps button and 9 LEDs

Issues:
 - 5 GHz band is not functional (missing driver support)

Installation:

Asus windows recovery tool:
 - install the Asus firmware restoration utility
 - unplug the router, hold the reset button while powering it on
 - release when the power LED flashes slowly
 - specify a static IP on your computer:
     IP address: 192.168.1.75;
     Subnet mask 255.255.255.0
 - Start the Asus firmware restoration utility, specify the sysupgrade
   image, and press upload

TFTP Recovery method:
 - set computer to a static ip, 192.168.1.75
 - connect computer to the LAN 1 port of the router
 - hold the reset button while powering on the router for a few seconds
 - send firmware image using a tftp client; i.e from linux:
 $ tftp
 tftp> binary
 tftp> connect 192.168.1.1
 tftp> put lede-ramips-mt7620-rt-ac51u-squashfs-sysupgrade.bin
 tftp> quit

Signed-off-by: Ørjan Malde <foxyred333@gmail.com>
7 years agoramips: add new device ZyXEL Keenetic as kn
Alexey Belyaev [Thu, 27 Apr 2017 18:02:05 +0000 (00:02 +0600)]
ramips: add new device ZyXEL Keenetic as kn

This device exactly same as NBG-419N but with USB port and USB Led.

Specification:

- SoC: Ralink RT3052 (MIPS24Kc) @384MHz
- RAM: 32 MiB
- Flash: 8 MiB
- WLAN: WiSoC 2T2R/300Mbps (2.4GHz)
- LAN: 4x100M
- WAN: 1x100M
- USB: 1x2.0

Installation via serial console (57600 8N1) from TFTP server
 - rename the firmware to something shorter, for example
   "sysupgrade.bin" (max. 32 chars)
 - copy firmware TFTP server's directory
 - when you power on device, and see U-Boot log, immediatly push "2"
   once.
 - You will see this message:
     2: System Load Linux Kernel then write to Flash via TFTP.
        Warning!! Erase Linux in Flash then burn new one. Are you sure?
 - Push "y", and enter: device IP, then TFTP server's IP, and then
   image firmware file name.

The firmware will be downloaded within ~30 seconds and flashed to the
device (It will take about 2 minutes).

Signed-off-by: Alexey Belyaev <spider@spider.vc>
[squash commits, compact commit message, fix compatible string, remove
superfluous pinmuxes]
Signed-off-by: Mathias Kresin <dev@kresin.me>
7 years agodnsmasq: use append_interface_name when using option --interface-name
Hans Dedecker [Mon, 8 May 2017 20:38:58 +0000 (22:38 +0200)]
dnsmasq: use append_interface_name when using option --interface-name

Signed-off-by: Hans Dedecker <dedeckeh@gmail.com>
7 years agodnsmasq: add interface-name uci list.
Daniel Danzberger [Tue, 2 May 2017 15:05:07 +0000 (17:05 +0200)]
dnsmasq: add interface-name uci list.

This patch adds the interface-name option for each dhcp config
in /etc/config/dhcp.

With the interface_name option users can define a DNS name for each dhcp section
that will be resolved by dnsmasq with the underlaying interface address.

For example:
config dhcp 'lan'
option interface 'lan'
...
list interface_name 'home.lan'
...

Signed-off-by: Daniel Danzberger <daniel@dd-wrt.com>
Signed-off-by: Hans Dedecker <dedeckeh@gmail.com> [PKG_RELEASE increase]
7 years agoramips: use GPIO_ACTIVE macros in UBNT-ERX.dts
Sven Roederer [Fri, 5 May 2017 20:18:44 +0000 (22:18 +0200)]
ramips: use GPIO_ACTIVE macros in UBNT-ERX.dts

- reorder includes
- use GPIO_ACTIVE_LOW macro instead of hardcoded "1"

Signed-off-by: Sven Roederer <devel-sven@geroedel.de>
7 years agoramips: strip padding from TP-Link sysupgrade images
Henryk Heisig [Fri, 5 May 2017 16:07:04 +0000 (18:07 +0200)]
ramips: strip padding from TP-Link sysupgrade images

This reduce size of sysupgrade firmware.

Signed-off-by: Henryk Heisig <hyniu@o2.pl>
7 years agoramips: add factory firmware for Tp-Link C20i/C50
Henryk Heisig [Fri, 5 May 2017 15:07:28 +0000 (17:07 +0200)]
ramips: add factory firmware for Tp-Link C20i/C50

TP-Link firmware doesn't accept sysupgrade.bin with metadata.

Signed-off-by: Henryk Heisig <hyniu@o2.pl>
7 years agobrcm63xx: fix invalid Asmax AR 1004g DTS reference
Marcin Jurkowski [Fri, 21 Apr 2017 00:36:18 +0000 (02:36 +0200)]
brcm63xx: fix invalid Asmax AR 1004g DTS reference

Build profile for Asmax AR 1004g refers to an invalid DTS "rg100a". The
correct DTS for this device is "ar1004g".

Signed-off-by: Marcin Jurkowski <marcin1j@gmail.com>
7 years agolantiq: fix avm fritz box mac addresses
Mathias Kresin [Sun, 26 Mar 2017 08:53:35 +0000 (10:53 +0200)]
lantiq: fix avm fritz box mac addresses

It has been shown that the Fritz boxes have the correct mac address set
in the wireless calibration data/eeeprom. Use this mac address as base
for the ethernet and xdsl interface increment/decrement the address to
match the values stored in the tffs.

Signed-off-by: Mathias Kresin <dev@kresin.me>
7 years agox86: fix lifting kernel CPU requirements and always enable PAE
Daniel Golle [Fri, 5 May 2017 12:16:44 +0000 (14:16 +0200)]
x86: fix lifting kernel CPU requirements and always enable PAE

commit 89878f60f4 x86: lift kernel minimum CPU requirement to Pentium MMX
caused kconfig havoc. Fix this and make sure PAE is enabled even on legacy
CPUs as the minimum required CPU has been Pentium MMX for a while now and
hence PAE is supported even on the x86_legacy target.

Signed-off-by: Daniel Golle <daniel@makrotopia.org>
7 years agobuild: allow val.% targets to bypass the prepare steps
Felix Fietkau [Fri, 5 May 2017 11:12:23 +0000 (13:12 +0200)]
build: allow val.% targets to bypass the prepare steps

Significantly reduces time spent processing those targets and should
also silence some log clutter which could confuse buildbot

Signed-off-by: Felix Fietkau <nbd@nbd.name>
7 years agox86/legacy: add missing config symbol
Jo-Philipp Wich [Fri, 5 May 2017 08:19:31 +0000 (10:19 +0200)]
x86/legacy: add missing config symbol

Signed-off-by: Jo-Philipp Wich <jo@mein.io>
7 years agox86: lift kernel minimum CPU requirement to Pentium MMX
Daniel Golle [Fri, 5 May 2017 07:53:44 +0000 (09:53 +0200)]
x86: lift kernel minimum CPU requirement to Pentium MMX

Anything older than that isn't supported since commit f4f8f4a180366,
hence also switch to Pentium MMX when building the kernel.

Signed-off-by: Daniel Golle <daniel@makrotopia.org>
7 years agox86: generic: use HIGHMEM64G instead of HIGHMEM4G
Daniel Golle [Fri, 5 May 2017 07:51:17 +0000 (09:51 +0200)]
x86: generic: use HIGHMEM64G instead of HIGHMEM4G

commit 4b4f73937371 switched on HIGHMEM4G which implicitely disabled
PAE and hence also NX and other useful and security-relevant features.
Re-enable PAE by switching to HIGHMEM64G.

Signed-off-by: Daniel Golle <daniel@makrotopia.org>
7 years agox86: keep /boot mounted for kexec
Daniel Golle [Fri, 5 May 2017 03:56:10 +0000 (05:56 +0200)]
x86: keep /boot mounted for kexec

Signed-off-by: Daniel Golle <daniel@makrotopia.org>
7 years agodnsmasq: make tftp root if not existing
Alberto Bursi [Tue, 2 May 2017 17:31:17 +0000 (19:31 +0200)]
dnsmasq: make tftp root if not existing

If there's a TFTP root directory configured, create it with mkdir -p
(which does not throw an error if the folder exists already)
before starting dnsmasq. This is useful for TFTP roots in /tmp, for example.

Originally submitted by nfw user aka Nathaniel Wesley Filardo

Signed-off-by: Alberto Bursi <alberto.bursi@outlook.it>
7 years agodnsmasq: fix dhcp_option usage warning
Hans Dedecker [Thu, 4 May 2017 20:33:37 +0000 (22:33 +0200)]
dnsmasq: fix dhcp_option usage warning

Don't display unnecessary dhcp_option usage warning in case
dhcp_option is empty

Signed-off-by: Hans Dedecker <dedeckeh@gmail.com>
7 years agoramips: purge default packages on MT7621
Daniel Golle [Thu, 4 May 2017 12:07:47 +0000 (14:07 +0200)]
ramips: purge default packages on MT7621

In order to have a smaller initramfs image remove all packages not
needed on all devices and add them explicitely for those actually
needing them. Also remove wpad-mini from ramips default package set
and add it to all sub-targets except for MT7621.
While at it reorder packages alphabetically and replace kmod-mt76 with
kmod-mt7603 and/or kmod-mt76x2 depending on the chip actually used on
a specific board.

Hopefully fixes FS#758

Signed-off-by: Daniel Golle <daniel@makrotopia.org>
7 years agosunxi: clean up modules definitions
Matthias Schiffer [Thu, 4 May 2017 05:00:06 +0000 (07:00 +0200)]
sunxi: clean up modules definitions

Module definitions for kmod-wdt-sunxi and kmod-eeprom-sunxi are removed
(wdt-sunxi was builtin anyways; nvmem-sunxi, which is the new name of
eeprom-sunxi is changed to builtin). As kmod-eeprom-sunxi was specified
in DEFAULT_PACKAGES, but not available on kernel 4.4, it was breaking the
image builder.

Support for kmod-sunxi-ir is added for kernel 4.4 (it is unclear why it
was disable before, it builds fine with with kernel 4.4).

Condtionals only relevant for pre-4.4 kernels are removed from modules.mk,
as sunxi does't support older kernels anymore.

Fixes FS#755.

Signed-off-by: Matthias Schiffer <mschiffer@universe-factory.net>
7 years agoramips: enable ramdisk for mt7621
Paul Spooren [Wed, 3 May 2017 22:47:34 +0000 (00:47 +0200)]
ramips: enable ramdisk for mt7621

Fixes #758

Signed-off-by: Paul Spooren <paul@spooren.de>
7 years agoocteon: port missing 4.9 patches
Felix Fietkau [Wed, 3 May 2017 21:43:01 +0000 (23:43 +0200)]
octeon: port missing 4.9 patches

Signed-off-by: Felix Fietkau <nbd@nbd.name>
7 years agoarmvirt: 64: enable usb support
Yousong Zhou [Wed, 3 May 2017 15:17:25 +0000 (23:17 +0800)]
armvirt: 64: enable usb support

It was originally only enabled for the "32" subtarget along with the
"usb" feature which is now also shared by the "64" target

This should fix the phase1 build of armvirt/64

Signed-off-by: Yousong Zhou <yszhou4tech@gmail.com>
7 years agoipq806x: fix EA8500 switch configuration
Jo-Philipp Wich [Wed, 3 May 2017 11:59:55 +0000 (13:59 +0200)]
ipq806x: fix EA8500 switch configuration

Do not assign the CPU port twice, this confuses LuCI and possible other
programs relying on topology information in board.json.

Ref: https://github.com/openwrt/luci/issues/1086

Signed-off-by: Jo-Philipp Wich <jo@mein.io>
7 years agohostapd: add legacy_rates option to disable 802.11b data rates.
Nick Lowe [Mon, 27 Mar 2017 09:50:23 +0000 (10:50 +0100)]
hostapd: add legacy_rates option to disable 802.11b data rates.

Setting legacy_rates to 0 disables 802.11b data rates.
Setting legacy_rates to 1 enables 802.11b data rates. (Default)

The basic_rate option and supported_rates option are filtered based on this.

The rationale for the change, stronger now than in 2014, can be found in:

https://mentor.ieee.org/802.11/dcn/14/11-14-0099-00-000m-renewing-2-4ghz-band.pptx

The balance of equities between compatibility with b clients and the
detriment to the 2.4 GHz ecosystem as a whole strongly favors disabling b
rates by default.

Signed-off-by: Nick Lowe <nick.lowe@gmail.com>
Signed-off-by: Felix Fietkau <nbd@nbd.name> [cleanup, defaults change]
7 years agohostapd: fix reload frequency change patch
Abhilash Tuse [Thu, 23 Mar 2017 06:04:32 +0000 (11:34 +0530)]
hostapd: fix reload frequency change patch

When sta is configured, hostapd receives 'stop' and 'update' command from
wpa_supplicant. In the update command, hostapd gets sta parameters with
which it configures ap.

Problem is, with the default wireless configuration:
mode:11g freq:2.4GHz channel:1
If sta is connected to 5GHz network, then ap does not work. Ideally with
340-reload_freq_change.patch hostapd should reload the frequency changes
and start ap in 5GHz, but ap becomes invisible in the network.

This issue can be reproduced with following /etc/config/wireless:
config wifi-device  radio0
        option type     mac80211
        option channel  1
        option hwmode   11g
        option path     'virtual/uccp420/uccwlan'
        option htmode   'none'

config wifi-iface 'ap'
        option device 'radio0'
        option encryption 'none'
        option mode 'ap'
        option network 'ap'
        option ssid 'MyTestNet'
        option encryption none

config wifi-iface 'sta'
       option device radio0
       option network sta
       option mode sta
       option ssid TestNet-5G
       option encryption psk2
       option key 12345

This change updates current_mode structure based on configured hw_mode
received from wpa_supplicant. Also prepare rates table after frequency
selection.

Signed-off-by: Abhilash Tuse <Abhilash.Tuse@imgtec.com>
Signed-off-by: Felix Fietkau <nbd@nbd.name> [cleanup, patch refresh]
7 years agobase-files: implement ucidef_set_hostname(), ucidef_set_ntpserver()
Jo-Philipp Wich [Wed, 3 May 2017 11:32:33 +0000 (13:32 +0200)]
base-files: implement ucidef_set_hostname(), ucidef_set_ntpserver()

Commit 2036ae4 (base-files: support hostname and ntp servers through board.d)
was supposed to implement these procedures but lacked the required changes
to uci-defaults.sh.

Add the missing procedures now to fix config generation on targets relying
on hostname or NTP server presetting.

Fixes FS#754.

Reported-by: Cristian Morales Vega <cristian@samknows.com>
Signed-off-by: Jo-Philipp Wich <jo@mein.io>
7 years agoopkg: alternatives: use ERROR level for symlink failure
Yousong Zhou [Wed, 3 May 2017 09:39:28 +0000 (17:39 +0800)]
opkg: alternatives: use ERROR level for symlink failure

Changes since last version

    04e279e pkg_alternatives: use ERROR level for symlink failure

Signed-off-by: Yousong Zhou <yszhou4tech@gmail.com>
7 years agobusybox: alternatives: accomodate CONFIG_BUSYBOX_CUSTOM=y
Yousong Zhou [Wed, 3 May 2017 09:32:40 +0000 (17:32 +0800)]
busybox: alternatives: accomodate CONFIG_BUSYBOX_CUSTOM=y

When busybox customisation is enabled, we should depend on config
symbols CONFIG_BUSYBOX_CONFIG_xxx to form alternatives specs

Reported-by: Hannu Nyman <hannu.nyman@iki.fi>
Signed-off-by: Yousong Zhou <yszhou4tech@gmail.com>
7 years agobuild: reset ALTERNATIVES field in Package/Default
Yousong Zhou [Wed, 3 May 2017 09:30:08 +0000 (17:30 +0800)]
build: reset ALTERNATIVES field in Package/Default

Otherwise ipkg packages may wrongly take on alternatives specs of
another package sharing the same package Makefile

Fixes FS#753

Signed-off-by: Yousong Zhou <yszhou4tech@gmail.com>
7 years agomediatek: fix sysupgrade on eMMC boards
John Crispin [Wed, 3 May 2017 08:31:55 +0000 (10:31 +0200)]
mediatek: fix sysupgrade on eMMC boards

Signed-off-by: John Crispin <john@phrozen.org>
7 years agodnsmasq: bump to 2.77test5
Kevin Darbyshire-Bryant [Sun, 30 Apr 2017 11:42:30 +0000 (12:42 +0100)]
dnsmasq: bump to 2.77test5

A number of small tweaks & improvements on the way to a final release.
Most notable:

Improve DHCPv4 address-in-use check.
Remove the recently introduced RFC-6842 (Client-ids in DHCP replies)
support as it turns out some clients are getting upset.

Signed-off-by: Kevin Darbyshire-Bryant <kevin@darbyshire-bryant.me.uk>
7 years agobusybox: add alternatives specs for those also provided by procps-ng
Yousong Zhou [Tue, 2 May 2017 16:52:53 +0000 (00:52 +0800)]
busybox: add alternatives specs for those also provided by procps-ng

Signed-off-by: Yousong Zhou <yszhou4tech@gmail.com>
7 years agoiproute2: bump PKG_RELEASE
Yousong Zhou [Tue, 2 May 2017 14:45:18 +0000 (22:45 +0800)]
iproute2: bump PKG_RELEASE

Signed-off-by: Yousong Zhou <yszhou4tech@gmail.com>
7 years agoiproute2: add ip-tiny, ip-full as alternatives of /sbin/ip
Yousong Zhou [Sat, 4 Mar 2017 11:37:43 +0000 (19:37 +0800)]
iproute2: add ip-tiny, ip-full as alternatives of /sbin/ip

They will not be in conflict anymore ;)

Signed-off-by: Yousong Zhou <yszhou4tech@gmail.com>
7 years agobusybox: add as an alternative of /sbin/ip
Yousong Zhou [Tue, 2 May 2017 14:11:17 +0000 (22:11 +0800)]
busybox: add as an alternative of /sbin/ip

Signed-off-by: Yousong Zhou <yszhou4tech@gmail.com>
7 years agoopkg: alternatives support
Yousong Zhou [Mon, 20 Mar 2017 11:52:40 +0000 (19:52 +0800)]
opkg: alternatives support

Changes since last version

    546bc72 pkg: alternatives support
    7a96972 libbb: xreadlink: fix memory leak on failure case
    3f13edd pkg_run_script: use pkg->dest in half installed case

Signed-off-by: Yousong Zhou <yszhou4tech@gmail.com>
7 years agobuild: ipkg: new field Alternatives
Yousong Zhou [Sun, 5 Mar 2017 08:19:46 +0000 (16:19 +0800)]
build: ipkg: new field Alternatives

It's a list of specs of the following form seprated by commas to
describe alternatives provided by this package

    <prio>:<path>:<altpath>

<path> will be a symbolic link to <altpath> of the highest <prio>

Signed-off-by: Yousong Zhou <yszhou4tech@gmail.com>
7 years agobuild: cleanup tmp/ dir of target rootfs
Yousong Zhou [Sun, 5 Mar 2017 08:17:17 +0000 (16:17 +0800)]
build: cleanup tmp/ dir of target rootfs

Signed-off-by: Yousong Zhou <yszhou4tech@gmail.com>
7 years agoimx6: remove linux 4.4 support
Felix Fietkau [Wed, 1 Feb 2017 09:18:11 +0000 (10:18 +0100)]
imx6: remove linux 4.4 support

Signed-off-by: Felix Fietkau <nbd@nbd.name>
7 years agotarget/arc770: switch to 4.9 kernel
Alexey Brodkin [Mon, 6 Mar 2017 18:30:08 +0000 (21:30 +0300)]
target/arc770: switch to 4.9 kernel

With update of ARC tools to arc-2016.09 based on GCC v6.x
we have to bump Linux kernel version so both toolchain and
the kernel use the same ARC ABIv4.

Signed-off-by: Alexey Brodkin <Alexey.Brodkin@synopsys.com>
7 years agotarget/archs38: switch to 4.9 kernel
Alexey Brodkin [Mon, 6 Mar 2017 18:30:07 +0000 (21:30 +0300)]
target/archs38: switch to 4.9 kernel

With update of ARC tools to arc-2016.09 based on GCC v6.x
we have to bump Linux kernel version so both toolchain and
the kernel use the same ARC ABIv4.

Signed-off-by: Alexey Brodkin <Alexey.Brodkin@synopsys.com>
7 years agotoolchain/arc: update to the most recent release arc-2016.09
Alexey Brodkin [Mon, 27 Feb 2017 18:27:50 +0000 (21:27 +0300)]
toolchain/arc: update to the most recent release arc-2016.09

arc-2016.09 is the most recent toolchain for ARC cores and
it is based on top of upstream Binutils 2.27 and GCC 6.2.1.

With updated major version of GCC we copied all GCC 6.x patches
for ARC as well as Bintils 2.27 patches.

Note that toochain sports ARCv4 ABI and so must be used
with 4.8+ Linux kernels. Even though it will build v4.4 kernel
perfectly fine on attempt to run user-space apps they won't
work with older kernel. That said previuosly sent RFC patches with
Linux kernel update are required:
 [1] https://patchwork.ozlabs.org/patch/726686/
 [2] https://patchwork.ozlabs.org/patch/726687/

Signed-off-by: Alexey Brodkin <Alexey.Brodkin@synopsys.com>
7 years agolinux-4.9: Fix building of DWC3 module
Alexey Brodkin [Tue, 2 May 2017 10:36:19 +0000 (13:36 +0300)]
linux-4.9: Fix building of DWC3 module

DWC3 driver uses of_usb_get_phy_mode() which is
implemented in drivers/usb/phy/of.c and in bare minimal
configuration it might not be pulled in kernel binary.

In case of ARC or ARM this could be easily reproduced with
"allnodefconfig" +CONFIG_USB=m +CONFIG_USB_DWC3=m.

On building all ends-up with:
---------------------->8------------------
  Kernel: arch/arm/boot/Image is ready
  Kernel: arch/arm/boot/zImage is ready
  Building modules, stage 2.
  MODPOST 5 modules
ERROR: "of_usb_get_phy_mode" [drivers/usb/dwc3/dwc3.ko] undefined!
make[1]: *** [__modpost] Error 1
make: *** [modules] Error 2
---------------------->8------------------

This was affecting Lede in case of CONFIG_ALL_KMODS=y.

The patch is already scheduled for v4.12 and once it is backported to
4.9.y it must be removed from here.

Note this patch makes sense not only for ARC instead it fixes
a generic issue which may affect more arches and platforms thus
putting it in "generic/" folder.

Signed-off-by: Alexey Brodkin <Alexey.Brodkin@synopsys.com>
Signed-off-by: Felix Fietkau <nbd@nbd.name> [patch rename]
7 years agoperf: disable build for ARC, it is currently broken
Felix Fietkau [Tue, 2 May 2017 13:03:59 +0000 (15:03 +0200)]
perf: disable build for ARC, it is currently broken

Signed-off-by: Felix Fietkau <nbd@nbd.name>
7 years agotarget: mpc85xx: make kernel_menuconfig
Alexandru Ardelean [Mon, 10 Apr 2017 14:48:24 +0000 (17:48 +0300)]
target: mpc85xx: make kernel_menuconfig

It's the best method I know (so far) to refresh
the kernel config to a default state.

Signed-off-by: Alexandru Ardelean <ardeleanalex@gmail.com>
7 years agotarget: mpc85xx: tl_wdr4900_v1: drop 'fsl_rstcr_restart' hook
Alexandru Ardelean [Mon, 10 Apr 2017 15:29:59 +0000 (18:29 +0300)]
target: mpc85xx: tl_wdr4900_v1: drop 'fsl_rstcr_restart' hook

Since commit:
http://github.com/torvalds/linux/commit/7120438e5d82f445acbfe131a1b58eab7e83fa33

Seems that fsl_rstcr_restart() has been converted
to a reset handler and dropped as hook/callback.

Apply the same to the `tl_wdr4900_v1` target.

Signed-off-by: Alexandru Ardelean <ardeleanalex@gmail.com>
7 years agotarget: mpc85xx: refresh kernel patches
Alexandru Ardelean [Mon, 10 Apr 2017 14:44:10 +0000 (17:44 +0300)]
target: mpc85xx: refresh kernel patches

Re-applied patches:
001-powerpc-85xx-add-gpio-keys-to-of-match-table.patch
100-powerpc-85xx-tl-wdr4900-v1-support.patch

Dropped patch:
200-spi-fsl-espi-preallocate-local-buffer.patch
now part of kernel upstream hash
https://github.com/torvalds/linux/commit/1423877b73ed5f4982eaba8bed359605b9918a2b

Signed-off-by: Alexandru Ardelean <ardeleanalex@gmail.com>
7 years agotarget: mpc85xx: switch to kernel 4.9
Alexandru Ardelean [Mon, 10 Apr 2017 13:28:39 +0000 (16:28 +0300)]
target: mpc85xx: switch to kernel 4.9

Signed-off-by: Alexandru Ardelean <ardeleanalex@gmail.com>
7 years agobuild: remove absolute path to perl and replace with /usr/bin/env perl
Bastian Köcher [Mon, 27 Mar 2017 15:21:12 +0000 (17:21 +0200)]
build: remove absolute path to perl and replace with /usr/bin/env perl

Signed-off-by: Bastian Köcher <git@kchr.de>
7 years agokernel: allow selecting RTC drivers on targets without explicit RTC support
Felix Fietkau [Wed, 12 Apr 2017 07:15:38 +0000 (09:15 +0200)]
kernel: allow selecting RTC drivers on targets without explicit RTC support

Keep them disabled by default to avoid pulling in extra kernel bloat

Signed-off-by: Felix Fietkau <nbd@nbd.name>
7 years agouboot-omap: rename patch to 101-disable-thumb-omap3.patch
Alexander Couzens [Fri, 28 Apr 2017 12:34:46 +0000 (14:34 +0200)]
uboot-omap: rename patch to 101-disable-thumb-omap3.patch

It's disabling thumb for the whole omap3 family.

Signed-off-by: Alexander Couzens <lynxis@fe80.eu>
7 years agouboot-omap: enable thumb for overo
Alexander Couzens [Fri, 28 Apr 2017 12:27:45 +0000 (14:27 +0200)]
uboot-omap: enable thumb for overo

Even it's unknown if it's booting with thumb, because the lack of hardware.
Enable it for now, so the build succeed, because without thumb it can not fit
into spl.

Signed-off-by: Alexander Couzens <lynxis@fe80.eu>
7 years agomediatek: fix mt7530 mcm reset
John Crispin [Sat, 29 Apr 2017 07:48:35 +0000 (09:48 +0200)]
mediatek: fix mt7530 mcm reset

Early SoC versions with an ECO of 1 required the gpio reset.
Mass production SoCs no longer need this work around.

Signed-off-by: John Crispin <john@phrozen.org>
7 years agoodhcpd: update to git HEAD version (FS#656,FS#595)
Hans Dedecker [Fri, 28 Apr 2017 19:45:10 +0000 (21:45 +0200)]
odhcpd: update to git HEAD version (FS#656,FS#595)

9268ca6 ndp: don't trigger IPv6 ping when neighbor entry is invalid
2b3355f ndp: fix adding proxy neighbor entries
7dff5b4 ndp: fix wrong interface name in syslog message
a54afb5 dhcpv6-ia: Fix segfault when writing DHCPv4 leases in state file
c0e9dbf ubus: don't segfault when there're no leases

Signed-off-by: Hans Dedecker <dedeckeh@gmail.com>
7 years agodnsmasq: support dhcp_option config as a list
Hans Dedecker [Wed, 26 Apr 2017 20:37:27 +0000 (22:37 +0200)]
dnsmasq: support dhcp_option config as a list

Configuring dhcp_option as an option does not allow the usage of white
spaces in the option value; fix this by supporting dhcp_option as a list
config while still supporting the option config to maintain backwards
compatibility

Signed-off-by: Jo-Philipp Wich <jo@mein.io>
Signed-off-by: Hans Dedecker <dedeckeh@gmail.com>
7 years agox86: enable 4G high memory support for generic (32bit) subtarget
Jo-Philipp Wich [Thu, 27 Apr 2017 13:15:29 +0000 (15:15 +0200)]
x86: enable 4G high memory support for generic (32bit) subtarget

This is required to fully support a number of 32bit x86 systems equipped
with more than one GB ram, e.g. certain Atom CPU machines.

Ref: https://forum.lede-project.org/t/17-01-1-x86-doesnt-detect-all-of-the-ram/3295

Signed-off-by: Jo-Philipp Wich <jo@mein.io>
7 years agoramips: ZTE-Q7: fix switch driver initialization in DTS (FS#716)
Jo-Philipp Wich [Tue, 25 Apr 2017 12:56:57 +0000 (14:56 +0200)]
ramips: ZTE-Q7: fix switch driver initialization in DTS (FS#716)

Add the changes suggested by FS#716 to fix the switch driver initialization
on the ZTE Q7.

Also remove the `pinctrl-names` field obsoleted by the changes.

Reported-by: Harry Lau <harrylwc@gmail.com>
Signed-off-by: Jo-Philipp Wich <jo@mein.io>
7 years agobusybox: nslookup_lede: mimic output format of old Busybox applet
Jo-Philipp Wich [Tue, 25 Apr 2017 12:21:08 +0000 (14:21 +0200)]
busybox: nslookup_lede: mimic output format of old Busybox applet

When invoking "nslookup_lede" with a domain argument and without explicit
query type, issue both A and AAAA queries and display the resulting IP
addresses in a numbered list style, similar to how the old BusyBox nslookup
used to output the records.

This is required for compatibility with certain scripts.

Ref: https://forum.lede-project.org/t/nslookup-ipv6-in-lede-17-01-1

Signed-off-by: Jo-Philipp Wich <jo@mein.io>
7 years agobusybox: nslookup_lede: fix compatibility with v1.25
Jo-Philipp Wich [Tue, 25 Apr 2017 12:03:11 +0000 (14:03 +0200)]
busybox: nslookup_lede: fix compatibility with v1.25

The ":*" optstring syntax was only recently introduced with BusyBox v1.26,
older versions need a corresponding hint in the "opt_complementary" variable
to denote flag values that should be stored as llist entries.

Add the required opt_complementary entry to fix random SIGBUS, SIGILL or
SIGSEGV related crashes on BusyBox 1.25.x when attempting to use the "-q"
flag of the "nslookup_lede" applet.

Ref: https://forum.lede-project.org/t/nslookup-ipv6-in-lede-17-01-1

Signed-off-by: Jo-Philipp Wich <jo@mein.io>
7 years agoramips: fix syntax error in board.d/01_leds
Mathias Kresin [Thu, 27 Apr 2017 07:38:25 +0000 (09:38 +0200)]
ramips: fix syntax error in board.d/01_leds

Fixes the following error:

  syntax error: unexpected newline (expecting ")")

Fixes: FS#739
Signed-off-by: Mathias Kresin <dev@kresin.me>
7 years agoramips: ZBT-WE826: include the correct dtsi
Mathias Kresin [Wed, 26 Apr 2017 17:45:38 +0000 (19:45 +0200)]
ramips: ZBT-WE826: include the correct dtsi

Fix a copy/paste error and include the ZBT-WE826 dtsi instead of the
ZBT-WG3526 one.

Fix the syntax error in the ZBT-WE826 dtsi to prevent an compile error.

Signed-off-by: Mathias Kresin <dev@kresin.me>
7 years agodropbear: fix procd interface trigger install
Hans Dedecker [Wed, 26 Apr 2017 14:44:02 +0000 (16:44 +0200)]
dropbear: fix procd interface trigger install

Install procd interface triggers only for interfaces which are enabled
so dropbear instances running on (an) enabled interface(s) are not
restarted due to an interface trigger of an interface which is disabled.

Signed-off-by: Hans Dedecker <dedeckeh@gmail.com>
7 years agobusybox: Enable sendfile by default
Daniel Engberg [Fri, 21 Apr 2017 08:36:39 +0000 (10:36 +0200)]
busybox: Enable sendfile by default

Enable sendfile system call by default.
http://lists.busybox.net/pipermail/busybox-cvs/2014-November/034831.html

Signed-off-by: Daniel Engberg <daniel.engberg.lists@pyret.net>
7 years agoinclude/packages-defaults.mk: Remove LARGEFILE option
Daniel Engberg [Fri, 21 Apr 2017 08:23:20 +0000 (10:23 +0200)]
include/packages-defaults.mk: Remove LARGEFILE option

Remove LARGEFILE option, support was removed back in 2011 (OpenWrt rev 25208).

Signed-off-by: Daniel Engberg <daniel.engberg.lists@pyret.net>
7 years agoimx6: ventana: add GW553x board identification
Tim Harvey [Thu, 20 Apr 2017 21:51:37 +0000 (14:51 -0700)]
imx6: ventana: add GW553x board identification

Signed-off-by: Tim Harvey <tharvey@gateworks.com>
7 years agoimx6: disable MSI interrupts
Tim Harvey [Thu, 20 Apr 2017 20:28:20 +0000 (13:28 -0700)]
imx6: disable MSI interrupts

The IMX6 PCIe host controller does not fire legacy interrupts if MSI is
enabled. A patch is being worked on to enable MSI at runtime only when
it is needed but meanwhile this patch will disable MSI for the imx6 kernel.

This fixes the no interrupt issue on cards/drivers that use legacy interrupts
such as ath9k.

Signed-off-by: Tim Harvey <tharvey@gateworks.com>
Tested-by: Koen Vandeputte <koen.vandeputte@ncentric.com>
7 years agoimx6: apply upstream mmc fix for DDR50 cards on UHS-I capable boards
Tim Harvey [Thu, 20 Apr 2017 18:46:01 +0000 (11:46 -0700)]
imx6: apply upstream mmc fix for DDR50 cards on UHS-I capable boards

https://git.kernel.org/pub/scm/linux/kernel/git/ulfh/mmc.git/commit/?h=fixes&id=9f327845358d3dd0d8a5a7a5436b0aa5c432e757

Signed-off-by: Tim Harvey <tharvey@gateworks.com>
7 years agobuild: fix symlinked .config handling
Sergey Ryazanov [Thu, 20 Apr 2017 02:10:45 +0000 (05:10 +0300)]
build: fix symlinked .config handling

When running "make menuconfig" with symlinked .config (e.g. to
env/.config) it renames symlink to .config.old, creates new .config file
and writes updated configuration here.

This breaks the desired workflow when changes in the configuration could
be checked using "scripts/env diff" and commited with
"scripts/env save". Since the env/.config file is not updated.

Fix this issue by exporting KCONFIG_OVERWRITECONFIG=1, which forces
mconf to overwrite the .config content, instead of renaming it and
creating a new file. This variable is set only if .config is a symlink,
otherwise the variable is not exported and the old behaviour is
preserved.

Signed-off-by: Sergey Ryazanov <ryazanov.s.a@gmail.com>
7 years agoapm821xx: Merge the Cisco Meraki MX60 & MX60W
Chris Blake [Fri, 31 Mar 2017 09:32:55 +0000 (04:32 -0500)]
apm821xx: Merge the Cisco Meraki MX60 & MX60W

Both of these boards share the same DTS, and hardware. The only
difference would be the PCI-E slot, and ath9k card found on the MX60W.
Due to the similarities, it would be more efficient to merge these
profiles.

Signed-off-by: Chris Blake <chrisrblake93@gmail.com>
7 years agoapm821xx: Add default packages to NAND target
Chris Blake [Fri, 31 Mar 2017 09:31:21 +0000 (04:31 -0500)]
apm821xx: Add default packages to NAND target

This moves core router packages to the NAND target, to ensure they are
applied to all images. This change is being done due to an issue found
when flashing the MX60W image, which came without these when built as a
multi image.

Signed-off-by: Chris Blake <chrisrblake93@gmail.com>
7 years agoar71xx: fold 641-MIPS-ath79-fix-AR934x-OTP-offsets.patch into the patch that it fixes
Felix Fietkau [Tue, 18 Apr 2017 07:59:54 +0000 (09:59 +0200)]
ar71xx: fold 641-MIPS-ath79-fix-AR934x-OTP-offsets.patch into the patch that it fixes

Signed-off-by: Felix Fietkau <nbd@nbd.name>
7 years agomac80211: Backport AR934x OTP Patch
Chris Blake [Mon, 27 Mar 2017 16:39:07 +0000 (11:39 -0500)]
mac80211: Backport AR934x OTP Patch

Merged upstream in https://patchwork.kernel.org/patch/9572541/, this
patch fixes the OTP offset used by the AR934x and AR955X to properly
enable reading from the OTP.

Signed-off-by: Christian Lamparter <chunkeey@googlemail.com>
Signed-off-by: Chris Blake <chrisrblake93@gmail.com>
7 years agoar71xx: Add OTP support for AR934x
Chris Blake [Mon, 27 Mar 2017 16:38:49 +0000 (11:38 -0500)]
ar71xx: Add OTP support for AR934x

The patch 523-MIPS-ath79-OTP-support only supports the OTP offsets for
AR933x chips, which has changed on newer platforms such as the AR934x.
The follwoing change is to add support for reading the OTP on the
AR934x. Tested on an Aerohive AP-121.

Signed-off-by: Christian Lamparter <chunkeey@googlemail.com>
Signed-off-by: Chris Blake <chrisrblake93@gmail.com>
7 years agotools: host/include/getline.h: Remove
Rosen Penev [Wed, 12 Apr 2017 20:18:49 +0000 (13:18 -0700)]
tools: host/include/getline.h: Remove

Seems to be useful only for old and unsupported operating systems. Removed and adjusted the mtd-utils Makefile.

Signed-off-by: Rosen Penev <rosenp@gmail.com>
7 years agokernel: remove out of tree direct-io disable hack
Felix Fietkau [Wed, 12 Apr 2017 07:35:14 +0000 (09:35 +0200)]
kernel: remove out of tree direct-io disable hack

Direct-IO support has to be enabled for the release build anyway, so
this hack is not worth keeping

Signed-off-by: Felix Fietkau <nbd@nbd.name>
7 years agoutils/mdadm: Update to 4.0
Daniel Engberg [Fri, 24 Mar 2017 11:17:24 +0000 (12:17 +0100)]
utils/mdadm: Update to 4.0

Update mdadm to 4.0
Remove 000-compile.patch as it's fixed upstream
Refresh patches
Add mdadm.h-Undefine-dprintf-before-redefining.patch
Source: http://git.openembedded.org/openembedded-core/tree/meta/recipes-extended/mdadm/files
Add RAID 0,1 and 10 as depends to make mdadm usable.

Signed-off-by: Daniel Engberg <daniel.engberg.lists@pyret.net>
7 years agoramips: use the ZBT-WE826 power LED for status indication
Mathias Kresin [Wed, 12 Apr 2017 05:56:40 +0000 (07:56 +0200)]
ramips: use the ZBT-WE826 power LED for status indication

Signed-off-by: Mathias Kresin <dev@kresin.me>
7 years agoramips: add ZBT-WE826 32MB variant
Mathias Kresin [Wed, 12 Apr 2017 05:53:56 +0000 (07:53 +0200)]
ramips: add ZBT-WE826 32MB variant

The ZBT-WG826 is available with 16 or 32 MByte of flash. Split the
device tree source file, rename the currently supported 16 MByte
version and add the 32 MByte variant.

Signed-off-by: Mathias Kresin <dev@kresin.me>
7 years agoramips: rename Digineo AC1200 Pro to ZBT-WG3526 32MB
Mathias Kresin [Sat, 1 Apr 2017 08:46:44 +0000 (10:46 +0200)]
ramips: rename Digineo AC1200 Pro to ZBT-WG3526 32MB

The Digineo AC1200 Pro is the 32MB flash variant of the ZBT-WG3526 with
unpopulated/exposed sdhci slot. Rename to board to the OEM/ODM name and
add the sdhci kernel module to use it for multiple clones.

Signed-off-by: Mathias Kresin <dev@kresin.me>
7 years agoramips: prepare ZBT-WG3526 for hardware variants
Mathias Kresin [Fri, 7 Apr 2017 21:52:27 +0000 (23:52 +0200)]
ramips: prepare ZBT-WG3526 for hardware variants

The ZBT-WG3526 is available with 16 or 32 MByte of flash. Rename the
current supported 16MByte version to indicate which flash size variant
is supported.

Signed-off-by: Mathias Kresin <dev@kresin.me>
7 years agoramips: add support for ZBT-WE1326
Jiawei Wang [Thu, 20 Apr 2017 03:07:26 +0000 (11:07 +0800)]
ramips: add support for ZBT-WE1326

Specification:
- SoC: MT7621AT, MT7603EN and MT7612EN
- Flash: 16 MiB (W25Q128FVSG)
- RAM: 512 MiB (EM6GE16EWXD-12H)
- Ethernet: 1 x WAN (10/100/1000Mbps) and 4 x LAN (10/100/1000 Mbps)
- Others: USB 2.0, micro SD slot, reset button and 8 x LEDs

Issues:
- Two LEDs for 2.4 GHz and 5 GHz Wi-Fi do not work, can't find GPIOs.
- The pwr LED is not GPIO controllable

How to install:
- The original firmware is OpenWrt, so both LuCI or sysupgrade can be used.
- Do not keep settings, for sysupgrade please use the -n option.

Signed-off-by: Jiawei Wang <buaawjw@gmail.com>
7 years agoramips: fix ZyXEL Keenetic Omni/Omni II GPIOs
Alexey Belyaev [Sat, 22 Apr 2017 20:46:23 +0000 (02:46 +0600)]
ramips: fix ZyXEL Keenetic Omni/Omni II GPIOs

Each GPIO bank has only 24 gpios. Use and enable the correct banks for
GPIOs > 24.

Signed-off-by: Alexey Belyaev <spider@spider.vc>
7 years agoramips: fix ZyXEL Keenetic Omni switch config
Alexey Belyaev [Sat, 22 Apr 2017 20:46:23 +0000 (02:46 +0600)]
ramips: fix ZyXEL Keenetic Omni switch config

The wan port is connected to switch port 0. Fix the mediatek,portmap as
well as the default switch config.

Signed-off-by: Alexey Belyaev <spider@spider.vc>
7 years agolantiq: DGN3500: make switch LEDs blink on activity
Kevin Darbyshire-Bryant [Thu, 20 Apr 2017 18:14:57 +0000 (19:14 +0100)]
lantiq: DGN3500: make switch LEDs blink on activity

Make switch port LEDs blink on activity as they do with stock firmware.

The u-boot loader sets suitable values in the LED control register that
unfortunately get blown away by the linux switch driver init.  Restore
that value by using driver 'initvals'.  0x0431 is the led control
register, the value 0x0960. Each nybble represents each LED's mode
(LED3,2,1,0) in this case OFF, 1000LINK+Activity, 100/10LINK+Activity,
OFF.  Only LED1&2 are physically present.

Enabling the blinkenlights exposed the fact that an RRCP packet was
being emitted by the switch on all ports every second.  Further initvals
copied from stock firmware's u-boot driver disabled this 'feature'.

Signed-off-by: Kevin Darbyshire-Bryant <kevin@darbyshire-bryant.me.uk>
7 years agogeneric: rtl8366rb: add support for initvals from DTS
Kevin Darbyshire-Bryant [Thu, 20 Apr 2017 18:14:35 +0000 (19:14 +0100)]
generic: rtl8366rb: add support for initvals from DTS

Allow per-device initvals in the DTS file for rlt8366rb
switches.

Shamelessly copies 30494598f891fc4436a49eca55fc31d03c087841

Signed-off-by: Kevin Darbyshire-Bryant <kevin@darbyshire-bryant.me.uk>
7 years agonetifd: return error status in reload_service
Hans Dedecker [Fri, 21 Apr 2017 09:44:57 +0000 (11:44 +0200)]
netifd: return error status in reload_service

Based on a patch by Alexandru Ardelean.
netifd ubus reload call returns the actual reload error status;
return error status as well in reload_service

Signed-off-by: Hans Dedecker <dedeckeh@gmail.com>
7 years agonetifd: update to git HEAD version
Hans Dedecker [Mon, 24 Apr 2017 11:28:02 +0000 (13:28 +0200)]
netifd: update to git HEAD version

11cb9cf ubus: add interface method to trigger renew event
4375d1b system-linux: allow "throw" route type
5fbd904 netifd: propagate error code on netifd_reload()
6e0acec interface-ip: fix device name for IPv6 link-local DNS server

Signed-off-by: Hans Dedecker <dedeckeh@gmail.com>
7 years agoodhcpd: update to git HEAD version
Hans Dedecker [Mon, 24 Apr 2017 16:39:59 +0000 (18:39 +0200)]
odhcpd: update to git HEAD version

570069d ubus: rework dumping IPv6 and IPv4 leases
4e579c4 dhcpv6-ia: simplify logic to write statefile and dhcpv6 logging

Signed-off-by: Hans Dedecker <dedeckeh@gmail.com>
7 years agoomap: rework image generation and profiles
Alexander Couzens [Sat, 15 Apr 2017 17:00:10 +0000 (19:00 +0200)]
omap: rework image generation and profiles

Migrate to new image generation and
introduce sdcard generation (based on sunxi).
Profile now generated via image/Makefile instead
of profiles/ directory.

Signed-off-by: Alexander Couzens <lynxis@fe80.eu>
7 years agopackage/uboot-omap: add am335x_boneblack
Alexander Couzens [Sun, 16 Apr 2017 17:26:45 +0000 (19:26 +0200)]
package/uboot-omap: add am335x_boneblack

Signed-off-by: Alexander Couzens <lynxis@fe80.eu>
7 years agopackage/uboot-omap: add default uEnv to boot via mmc card
Alexander Couzens [Wed, 19 Apr 2017 02:48:03 +0000 (04:48 +0200)]
package/uboot-omap: add default uEnv to boot via mmc card

Signed-off-by: Alexander Couzens <lynxis@fe80.eu>
7 years agopackage/uboot-omap: re-introduce patches dropped by update to v2017.01
Alexander Couzens [Tue, 18 Apr 2017 18:15:59 +0000 (20:15 +0200)]
package/uboot-omap: re-introduce patches dropped by update to v2017.01

The beagleboard doesn't boot with Thumb enabled, but without Thumb the
SPL (first stage bootloader) is too big to fit into SRAM.

Signed-off-by: Alexander Couzens <lynxis@fe80.eu>
7 years agopackage/uboot-omap: install image to STAGING_DIR_IMAGE
Alexander Couzens [Sat, 15 Apr 2017 16:49:57 +0000 (18:49 +0200)]
package/uboot-omap: install image to STAGING_DIR_IMAGE

to use them later by image/Makefile to create full working sdcard images

Signed-off-by: Alexander Couzens <lynxis@fe80.eu>
7 years agopackage/uboot-omap: assign BUILD_DEVICES most boards
Alexander Couzens [Sat, 15 Apr 2017 16:52:34 +0000 (18:52 +0200)]
package/uboot-omap: assign BUILD_DEVICES most boards

Signed-off-by: Alexander Couzens <lynxis@fe80.eu>
7 years agoomap: correct CPU_TYPE: omaps are cortext-a8
Alexander Couzens [Tue, 18 Apr 2017 17:58:20 +0000 (19:58 +0200)]
omap: correct CPU_TYPE: omaps are cortext-a8

Signed-off-by: Alexander Couzens <lynxis@fe80.eu>
7 years agoomap/Kconfig: refresh by oldconfig
Alexander Couzens [Sat, 22 Apr 2017 13:14:00 +0000 (15:14 +0200)]
omap/Kconfig: refresh by oldconfig

Signed-off-by: Alexander Couzens <lynxis@fe80.eu>
7 years agoomap/Kconfig: add usb options
Alexander Couzens [Sat, 22 Apr 2017 13:13:35 +0000 (15:13 +0200)]
omap/Kconfig: add usb options

including phy and required dma options

Signed-off-by: Alexander Couzens <lynxis@fe80.eu>
7 years agoomap/Kconfig: enable SMP
Alexander Couzens [Sat, 22 Apr 2017 13:12:13 +0000 (15:12 +0200)]
omap/Kconfig: enable SMP

the pandaboard is a multicore omap

Signed-off-by: Alexander Couzens <lynxis@fe80.eu>
7 years agoomap/Kconfig: add omap2plus_cpufreq
Alexander Couzens [Sat, 22 Apr 2017 12:58:36 +0000 (14:58 +0200)]
omap/Kconfig: add omap2plus_cpufreq

Signed-off-by: Alexander Couzens <lynxis@fe80.eu>
7 years agomac80211: rt2x00: add incomplete support for external PA on MT7620
Daniel Golle [Sun, 23 Apr 2017 23:44:29 +0000 (01:44 +0200)]
mac80211: rt2x00: add incomplete support for external PA on MT7620

Do as the vendor driver does -- however, some devices apparently
patched support for external PA into the driver similar to how it was
done on Rt3352, using EEPROM_NIC_CONF1 rather than EEPROM_NIC_CONF2,
hence we check for both fields. Somehow the vendor driver also no
longer offers the option of only one of the TX paths having an external
PA (which was probably to weird to ever be implemented in practise,
though it doesn't seem like a particularly bad idea to me). Do the same
in rt2x00 and enable support for external PA on both TX paths whenever
it is set for TX0 in EEPROM.

Signed-off-by: Daniel Golle <daniel@makrotopia.org>