openwrt/staging/chunkeey.git
8 years agokirkwood: remove redundant code in etc/board.d/02_network
Paul Wassi [Mon, 24 Oct 2016 13:19:52 +0000 (15:19 +0200)]
kirkwood: remove redundant code in etc/board.d/02_network

Remove redundant code: merge boards/cases that share
the same network configuration.
Also fix the alphabetical ordering of the cases.

Signed-off-by: Paul Wassi <p.wassi@gmx.at>
8 years agoar71xx: Remove switch config for the MR12/MR16
Chris Blake [Mon, 24 Oct 2016 19:05:55 +0000 (14:05 -0500)]
ar71xx: Remove switch config for the MR12/MR16

These boards do not have a switch, so they should have never been added
to this file in the first place.

Signed-off-by: Chris Blake <chrisrblake93@gmail.com>
8 years agoar71xx: add mac partition to the MR12/MR16
Chris Blake [Mon, 24 Oct 2016 19:05:54 +0000 (14:05 -0500)]
ar71xx: add mac partition to the MR12/MR16

On the stock Meraki Firmare for the MR12/MR16, a chunk of SPI space
after u-boot-env is used to store the boards Mac address. Sadly as this
was removed on any device already on OpenWRT/LEDE, moving forward a new,
64k partition named "mac" will be used to store the mac address for the
device (which is the minimum size). This allows users to properly set
the correct MAC, without editing the ART partition (which holds the same
MAC for all devices).

The reason the space is taken from kernel instead of rootfs is currently
kernels are only 1.3MB, so that way we can leave the current rootfs
space alone for users who fully utilize the available storage space.

Once this partition is added to a device, you can set your MAC doing the
following:

mtd erase mac
echo -n -e '\x00\x18\x0a\x33\x44\x55' > /dev/mtd5
sync && reboot

Where 00:18:0a:33:44:55 is your MAC address.

This was tested, and confirmed working on both the MR12 and MR16.

Signed-off-by: Chris Blake <chrisrblake93@gmail.com>
8 years agoar71xx: Move MR12 & MR16 from legacy to generic
Chris Blake [Mon, 24 Oct 2016 19:05:53 +0000 (14:05 -0500)]
ar71xx: Move MR12 & MR16 from legacy to generic

This moves the Meraki MR12 and Meraki MR16 to the new generic target.
Tested and verified working on both devices.

Note that kernel/rootfs images are still generated. This is because they
are used for the inital flashing process due to the fun pace at which
UBoot erases/writes to SPI.

Signed-off-by: Chris Blake <chrisrblake93@gmail.com>
8 years agougps: update to latest git HEAD
John Crispin [Sat, 22 Oct 2016 20:13:06 +0000 (22:13 +0200)]
ugps: update to latest git HEAD

* fixes a few error that uClibc users were seeing
* add switch to disable time adjustment

Signed-off-by: John Crispin <john@phrozen.org>
8 years agoimx6: Add ds1307 RTC to kernel for working hctosys
Petr Štetiar [Tue, 18 Oct 2016 21:26:59 +0000 (23:26 +0200)]
imx6: Add ds1307 RTC to kernel for working hctosys

Otherwise if we use ds1307 as kernel module, hctosys fails as ds1307 is
being initialized later then hctosys:

 [    2.427349] hctosys: unable to open rtc device (rtc0)
 [    3.714263] snvs_rtc 20cc000.snvs:snvs-rtc-lp: rtc core: registered 20cc000.snvs:snvs-r as rtc1
 [    8.990061] rtc-ds1307 3-006f: rtc core: registered mcp7941x as rtc0

Signed-off-by: Petr Štetiar <ynezz@true.cz>
8 years agokirkwood: Add RTC driver to kernel for working hctosys
Paul Wassi [Sun, 23 Oct 2016 12:21:22 +0000 (14:21 +0200)]
kirkwood: Add RTC driver to kernel for working hctosys

Build the RTC driver into the kernel, (and remove the optional module), in order
to make hctosys working. (Currently the module is loaded after hctosys has failed previously)

Signed-off-by: Paul Wassi <p.wassi@gmx.at>
8 years agoar71xx: Add support to DomyWifi DW33D
Jing Lin [Tue, 18 Oct 2016 08:18:50 +0000 (16:18 +0800)]
ar71xx: Add support to DomyWifi DW33D

Signed-off-by: Jing Lin <mumuqz@163.com>
8 years agouhttpd: fix handling of special "/" prefix when matching handlers
Jo-Philipp Wich [Tue, 25 Oct 2016 14:36:47 +0000 (16:36 +0200)]
uhttpd: fix handling of special "/" prefix when matching handlers

The special prefix of "/" should match any url by definition but the final
assertion which ensures that the matched prefix ends in '\0' or '/' is causing
matches against the "/" prefix to fail.

Update to current HEAD in order to fix this particular case.

Signed-off-by: Jo-Philipp Wich <jo@mein.io>
8 years agokernel: backport MIPS's ioremap_cache from 4.5
Rafał Miłecki [Mon, 24 Oct 2016 20:52:44 +0000 (22:52 +0200)]
kernel: backport MIPS's ioremap_cache from 4.5

It's needed by (updated) bcm47xxsflash driver.

Signed-off-by: Rafał Miłecki <rafal@milecki.pl>
8 years agobrcm47xx: reorder older entries in image Makefile
Rafał Miłecki [Mon, 24 Oct 2016 18:33:07 +0000 (20:33 +0200)]
brcm47xx: reorder older entries in image Makefile

Keep all entries sorted and replace tabs with spaces while at it.

Signed-off-by: Rafał Miłecki <rafal@milecki.pl>
8 years agokernel: add fix for CVE-2016-5195
Stijn Tintel [Mon, 24 Oct 2016 02:08:07 +0000 (05:08 +0300)]
kernel: add fix for CVE-2016-5195

Add fix for CVE-2016-5195 to kernel 3.18 and 4.1.
Kernel 4.4 is already fixed since version 4.4.26 which was committed in
32c28a78f798dffb983a7f00342b471e5e4ac177.

Patches taken from Sasha Levin's linux-stable git tree:
http://git.kernel.org/cgit/linux/kernel/git/sashal/linux-stable.git/

Signed-off-by: Stijn Tintel <stijn@linux-ipv6.be>
8 years agokernel: update kernel 3.18 to version 3.18.43
Stijn Tintel [Thu, 15 Sep 2016 19:16:34 +0000 (21:16 +0200)]
kernel: update kernel 3.18 to version 3.18.43

Refresh patches for all targets supporting 3.18 and not marked broken.
Compile-tested on all targets using 3.18 and not marked broken.

Changes to generic/610-netfilter_match_bypass_default_checks.patch based
on 84d489f64f3b382d5544e342f66d8806e94e85d3.

Signed-off-by: Stijn Tintel <stijn@linux-ipv6.be>
8 years agokernel: update kernel 4.1 to version 4.1.34
Stijn Tintel [Thu, 15 Sep 2016 13:27:45 +0000 (15:27 +0200)]
kernel: update kernel 4.1 to version 4.1.34

Refresh patches for all targets supporting 4.1 and not marked broken.
Compile-tested on all targets using 4.1 and not marked broken.

Changes to generic/610-netfilter_match_bypass_default_checks.patch based
on 84d489f64f3b382d5544e342f66d8806e94e85d3.
Changes to generic/666-Add-support-for-MAP-E-FMRs-mesh-mode.patch based
on a90ee92337d60fd14b6431adcda9929b955f3408.

Signed-off-by: Stijn Tintel <stijn@linux-ipv6.be>
8 years agobrcm47xx: bump kernel to 4.4
Rafał Miłecki [Mon, 24 Oct 2016 15:03:48 +0000 (17:03 +0200)]
brcm47xx: bump kernel to 4.4

Kernel 4.4 was ready for brcm47xx for almost a year now but I kept
postponing the bump due to problems with Linksys WRT300N v1.0. OpenWrt
and LEDE with 4.4 were hanging at the booting with the:
> Starting program at 0x80001000
(the last CFE message).

This was a permanent state, "make distclean" wasn't helping, I spent
hours debugging this and I was reliably reproducing the issue every
time. I also reported it on linux-mips ML in the thread:
> BCM4704 stopped booting with 4.4 (due to vmlinux size?)

After ~month I started working on WRT300N again. I got hangs as expected
every time I switched from 4.1 to 4.4. I started experimenting with:
1) TRX content (I tried dropping rootfs partition)
2) BZ_TEXT_START of lzma-loader
3) Flashing other variants of image: lzma compressed kernel (without a
   loader), gzip compressed one, uncompressed one.

At some point I got rootfs-less image booting and after that I couldn't
reproduce problem anymore, even with a complete firmware. It seems like
hardware was in some locked/unstable state that got magically fixed.

I have LEDE working now, tested it even with "make distclean", it seems
we can bump kernel now. I'll keep testing it on WRT300N for some time.

Signed-off-by: Rafał Miłecki <rafal@milecki.pl>
8 years agobrcm47xx: build also TRX image for Linksys WRT300N V1
Rafał Miłecki [Mon, 24 Oct 2016 08:34:32 +0000 (10:34 +0200)]
brcm47xx: build also TRX image for Linksys WRT300N V1

Linksys WRT300N V1 has pretty bugged CFE bootloader (it crashes in a lot
of situations) that doesn't accept .bin image.

Signed-off-by: Rafał Miłecki <rafal@milecki.pl>
8 years agobrcm47xx: open code Makefile entries for all devices
Rafał Miłecki [Thu, 20 Oct 2016 05:59:03 +0000 (07:59 +0200)]
brcm47xx: open code Makefile entries for all devices

If we want to use some of new features like per device rootfs we will
need this to specify them there.

Signed-off-by: Rafał Miłecki <rafal@milecki.pl>
8 years agotoolchain/gdb: update to version 7.12
Felix Fietkau [Wed, 19 Oct 2016 10:30:53 +0000 (12:30 +0200)]
toolchain/gdb: update to version 7.12

Signed-off-by: Felix Fietkau <nbd@nbd.name>
8 years agobinutils: remove old unused versions
Felix Fietkau [Wed, 19 Oct 2016 10:22:45 +0000 (12:22 +0200)]
binutils: remove old unused versions

Signed-off-by: Felix Fietkau <nbd@nbd.name>
8 years agomvebu: add support for the Linksys WRT3200ACM (Rango)
Imre Kaloz [Thu, 13 Oct 2016 10:52:20 +0000 (12:52 +0200)]
mvebu: add support for the Linksys WRT3200ACM (Rango)

Signed-off-by: Imre Kaloz <kaloz@openwrt.org>
Signed-off-by: Álvaro Fernández Rojas <noltari@gmail.com>
8 years agomwlwifi: upgrade to 10.3.2.0-20161013
Imre Kaloz [Thu, 13 Oct 2016 10:45:17 +0000 (12:45 +0200)]
mwlwifi: upgrade to 10.3.2.0-20161013

Also install the 88W8964 firmware.

Signed-off-by: Imre Kaloz <kaloz@openwrt.org>
8 years agomvebu: add missing status LEDs for Linksys WRT1200AC and WRT1900ACv2
Imre Kaloz [Fri, 14 Oct 2016 11:50:41 +0000 (13:50 +0200)]
mvebu: add missing status LEDs for Linksys WRT1200AC and WRT1900ACv2

Signed-off-by: Imre Kaloz <kaloz@openwrt.org>
8 years agokernel: update kernel 4.4 to version 4.4.27
Álvaro Fernández Rojas [Sun, 23 Oct 2016 12:56:09 +0000 (14:56 +0200)]
kernel: update kernel 4.4 to version 4.4.27

Refresh patches for all targets that support kernel 4.4.
compile/run-tested on brcm2708/bcm2710 only.

Signed-off-by: Álvaro Fernández Rojas <noltari@gmail.com>
8 years agobase-files: add ucidef_set_led_usbport for full usbport support
Rafał Miłecki [Wed, 19 Oct 2016 11:05:53 +0000 (13:05 +0200)]
base-files: add ucidef_set_led_usbport for full usbport support

This helper allows using usbport trigger directly. It requires usbport
compatible syntax and supports specifying multiple USB ports, e.g.:
ucidef_set_led_usbport "usb" "USB" "devicename:colour:function" "usb1-port1" "usb2-port1"

This adds a proper object to the board.json, e.g.
"usb": {
"name": "USB",
"type": "usbport",
"sysfs": "devicename:colour:function",
"ports": [
"usb1-port1",
"usb2-port1"
]
}
and supports translating it into uci section.

Signed-off-by: Rafał Miłecki <rafal@milecki.pl>
8 years agokernel: update kernel 4.4 to version 4.4.26
Koen Vandeputte [Thu, 20 Oct 2016 08:46:11 +0000 (10:46 +0200)]
kernel: update kernel 4.4 to version 4.4.26

Refresh patches for all targets that support kernel 4.4.
compile/run-tested on cns3xxx & imx6.

Signed-off-by: Koen Vandeputte <koen.vandeputte@ncentric.com>
8 years agobusybox: adjust download mirror
Hannu Nyman [Fri, 14 Oct 2016 08:09:47 +0000 (11:09 +0300)]
busybox: adjust download mirror

* Adjust download locations:
  - use https as busybox.net permanently redirects http to https
  - gentoo mirror has neither 1.25.0 nor 1.25.1 available, so drop it
    in favor of buildroot.net that has 1.25.1

Signed-off-by: Hannu Nyman <hannu.nyman@iki.fi>
Signed-off-by: Stijn Tintel <stijn@linux-ipv6.be>
8 years agosdk: predefine SOURCE_DATE_EPOCH
Jo-Philipp Wich [Fri, 21 Oct 2016 10:21:40 +0000 (12:21 +0200)]
sdk: predefine SOURCE_DATE_EPOCH

When building packages within the SDK, there is no Git revision history
available so prepopulate SOURCE_DATE_EPOCH in version.mk, similar to
how we handle REVISION already.

Acked-by: Felix Fietkau <nbd@nbd.name>
Signed-off-by: Jo-Philipp Wich <jo@mein.io>
8 years agobcm53xx: include b43 in Tenda AC9 image
Rafał Miłecki [Thu, 20 Oct 2016 06:45:00 +0000 (08:45 +0200)]
bcm53xx: include b43 in Tenda AC9 image

There is some (poor?) support for BCM43217 in b43.

Signed-off-by: Rafał Miłecki <rafal@milecki.pl>
8 years agolibreadline: set ABI_VERSION to force rebuild of dependent packages
Felix Fietkau [Wed, 19 Oct 2016 18:26:16 +0000 (20:26 +0200)]
libreadline: set ABI_VERSION to force rebuild of dependent packages

Signed-off-by: Felix Fietkau <nbd@nbd.name>
8 years agolantiq: enable cpu temp driver for all vr9 boards
Mathias Kresin [Mon, 17 Oct 2016 20:51:31 +0000 (22:51 +0200)]
lantiq: enable cpu temp driver for all vr9 boards

Only present on v1.2 vr9 SoCs but the driver takes care to not load on
boards having a v1.1 SoC.

Signed-off-by: Mathias Kresin <dev@kresin.me>
8 years agolantiq: fix thermal sensors driver
Mathias Kresin [Mon, 17 Oct 2016 18:25:53 +0000 (20:25 +0200)]
lantiq: fix thermal sensors driver

Read the temperature including the decimale place from the CGU_GPHY1_CR
register.

Decrement the temperature read from the register by 38.0 degree celsius.
The temperature range of the sensor is -38.0 to +154 °C and the register
value 0 is equal to -38.0 °C. This fixes the report of unrealistic
temperatures as seen on all tested boards.

Give the SoC a few milliseconds to get the first temperature value. On
some rare occasions there is no temperature value in the register when
read the first time after activation. This leads to a reported
temperature of -38.0 °C on boot.

Only version 1.2 of the vr9 SoC has a temperature sensor. Add a check
to make sure the driver doesn't load on v1.1 vr9 SoCs.

Signed-off-by: Mathias Kresin <dev@kresin.me>
8 years agolantiq: Sanitize device tree files
Andreas Eberlein [Thu, 29 Sep 2016 16:22:55 +0000 (18:22 +0200)]
lantiq: Sanitize device tree files

The device tree file of ARV752DPW uses numbers/hex values for gpio states and input event codes.

This cleans it up and uses the available macros from header files. This way the functions are easier to read and comprehend.

Signed-off-by: Andreas Eberlein <foodeas@aeberlein.de>
[sanitize all device tree files]
Signed-off-by: Mathias Kresin <dev@kresin.me>
8 years agolantiq: Fix buttons for ARV752DPW
Andreas Eberlein [Thu, 29 Sep 2016 16:22:23 +0000 (18:22 +0200)]
lantiq: Fix buttons for ARV752DPW

The GPIO for reset switch is wrong in definition. Further the key codes for the two additional buttons are ineffective.

Both is fixed here.

Signed-off-by: Andreas Eberlein <foodeas@aeberlein.de>
8 years agolantiq: use new build code for DGN3500
Mathias Kresin [Sat, 1 Oct 2016 07:59:00 +0000 (09:59 +0200)]
lantiq: use new build code for DGN3500

Signed-off-by: Mathias Kresin <dev@kresin.me>
8 years agolantiq: drop lzma-loader
Mathias Kresin [Sun, 2 Oct 2016 12:27:45 +0000 (14:27 +0200)]
lantiq: drop lzma-loader

It isn't used any more.

Signed-off-by: Mathias Kresin <dev@kresin.me>
8 years agolantiq: rework VG3503J image
Mathias Kresin [Sun, 2 Oct 2016 09:10:02 +0000 (11:10 +0200)]
lantiq: rework VG3503J image

Use the new image build code and remove the lzma loader. The lzma
loader was used to cheat the signature validation of the bootloader and
I found another way to do this.

To migrate boards already using LEDE/OpenWrt to the new image the
following steps need to be done once:

  VR9 # run reset_uboot_config
  VR9 # reset
  VR9 # setenv ethaddr AA:BB:CC:DD:EE:FF
  VR9 # setenv preboot ping 1.1.1.1\;bootm 0xb001f000
  VR9 # saveenv
  VR9 # tftp 0x81000000 lede-lantiq-xrx200-VG3503J-squashfs-sysupgrade.bin
  VR9 # erase 0xb0020000 $(filesize)
  VR9 # cp.b 0x81000000 0xb0020000 $(filesize)

The mac address is printed on the label at the bottom of the case.

The following steps are need to be done during first install:

  VR9 # setenv preboot ping 1.1.1.1\;bootm 0xb001f000
  VR9 # saveenv
  VR9 # tftp 0x81000000 lede-lantiq-xrx200-VG3503J-squashfs-sysupgrade.bin
  VR9 # erase 0xb0020000 $(filesize)
  VR9 # cp.b 0x81000000 0xb0020000 $(filesize)

The image uses the uImage firmware splitter now instead of hardcoded
kernel and rootfs partitions. The firmware partition size was extended
to use flash space that was reserved for partitions required only by
the ECI firmware.

Due to the changes an upgrade to a later LEDE revision from a running
LEDE is supported now.

A default switch config was added and the device uses the same MAC
addresses as the ECI firmware now instead of the same for all VG3503J.

Signed-off-by: Mathias Kresin <dev@kresin.me>
8 years agoprocd: update to the latest version, fixes a few minor service handling issues
Felix Fietkau [Wed, 19 Oct 2016 13:33:29 +0000 (15:33 +0200)]
procd: update to the latest version, fixes a few minor service handling issues

Signed-off-by: Felix Fietkau <nbd@nbd.name>
8 years agobcm53xx: include usbport trigger for devices with USB
Rafał Miłecki [Wed, 19 Oct 2016 10:28:52 +0000 (12:28 +0200)]
bcm53xx: include usbport trigger for devices with USB

We don't have any code for enabling it automatically yet but it allows
adding entries manually to the /etc/config/system.

Signed-off-by: Rafał Miłecki <rafal@milecki.pl>
8 years agokernel: drop usbdev LED trigger
Rafał Miłecki [Wed, 19 Oct 2016 10:09:45 +0000 (12:09 +0200)]
kernel: drop usbdev LED trigger

It was LEDE's trigger that was replaced by upstream usbport one.

Signed-off-by: Rafał Miłecki <rafal@milecki.pl>
8 years agoswitch to the new usbport LED trigger
Rafał Miłecki [Wed, 19 Oct 2016 10:09:45 +0000 (12:09 +0200)]
switch to the new usbport LED trigger

This makes init.d script handle existing UCI entries using the new
trigger. It also switches all targets to use its package.

Signed-off-by: Rafał Miłecki <rafal@milecki.pl>
8 years agoLatest ath10k CT 988X firmware (beta-18).
Ben Greear [Mon, 17 Oct 2016 21:23:01 +0000 (14:23 -0700)]
Latest ath10k CT 988X firmware (beta-18).

  *  Backport much of the 10.2 firmware features from upstream QCA driver.
     This includes ANI support, adaptive CCA, tx-hang workarounds,
     and lots of other things.
     Not all of this may be enabled at this point, and more code waits to
     be backported as time and motivation allows.

  *  Fix some rate-control issues where ath10k in station mode (at least), would
     sometimes get stuck at low rates.  This appeared to be a probe related
     state machine issue in the firmware, so I added some timeout logic to kick
     the state machine if it gets stuck.  This signicantly improves throughput
     tests with many stations.

  *  Support configuring WMI WD timeout using SET_SPECIAL API.

  *  Properly configure the rx-mask on bootup to work around problem found
     by Mr. Kazior.  This should remove the need to add the driver hack he
     posted.

  *  Allow configuring pdev failed-retry threshold.  This is how many consecutive
     tx failures the firmware will allow before resetting the wifi chip (not a full
     firmware crash).

Signed-off-by: Ben Greear <greearb@candelatech.com>
8 years agoath10k-ct: Update to latest 4.7 CT ath10k driver.
Ben Greear [Mon, 17 Oct 2016 20:59:05 +0000 (13:59 -0700)]
ath10k-ct: Update to latest 4.7 CT ath10k driver.

Adds Sebastian's 160Mhz support (un-tested), remove DMA32 change that
broke some x86 systems, allow setting 10.1 CT firmware keepalive watchdog
timeout, support QCA 9887 hardware, and some other tweaks.

Signed-off-by: Ben Greear <greearb@candelatech.com>
8 years agokernel: add package for usbport LED trigger
Rafał Miłecki [Tue, 4 Oct 2016 10:32:13 +0000 (12:32 +0200)]
kernel: add package for usbport LED trigger

This is upstream alternative for LEDE's ledtrig-usbdev. It's main
advantages are:
1) Support for assigning more than 1 USB port to the LED
2) Setting proper state when activating with device already connected
3) FWIW it's an upstream driver

Signed-off-by: Rafał Miłecki <rafal@milecki.pl>
8 years agoipq806x: add back end-of-UBI marker for a few factory images (FS#228)
Felix Fietkau [Tue, 18 Oct 2016 07:29:50 +0000 (09:29 +0200)]
ipq806x: add back end-of-UBI marker for a few factory images (FS#228)

Signed-off-by: Felix Fietkau <nbd@nbd.name>
8 years agoar71xx: Add support to TP-Link EAP120
Henryk Heisig [Thu, 6 Oct 2016 21:46:38 +0000 (23:46 +0200)]
ar71xx: Add support to TP-Link EAP120

Signed-off-by: Henryk Heisig <hyniu@o2.pl>
8 years agoprocd: Allow initscripts to start one daemon instance at a time
Pierre Lebleu [Thu, 13 Oct 2016 10:52:05 +0000 (12:52 +0200)]
procd: Allow initscripts to start one daemon instance at a time

Signed-off-by: Pierre Lebleu <pierre.lebleu@technicolor.com>
8 years agoxfsprogs: install path consistent with fs tools
Alberto Bursi [Sun, 16 Oct 2016 15:21:42 +0000 (17:21 +0200)]
xfsprogs: install path consistent with fs tools

changed install path from /sbin to /usr/sbin to be consistent with other filesystem tools
ext2-3-4 and f2fs tools are in /usr/sbin, for example

Signed-off-by: Alberto Bursi <alberto.bursi@outlook.it>
8 years agokernel: update kernel 4.4 to version 4.4.25
Koen Vandeputte [Mon, 17 Oct 2016 10:58:15 +0000 (12:58 +0200)]
kernel: update kernel 4.4 to version 4.4.25

Refresh patches for all targets that support kernel 4.4.
compile/run-tested on cns3xxx & imx6.

Signed-off-by: Koen Vandeputte <koen.vandeputte@ncentric.com>
8 years agomac80211: backport some upstream a-msdu tx fixes
Felix Fietkau [Mon, 17 Oct 2016 10:18:25 +0000 (12:18 +0200)]
mac80211: backport some upstream a-msdu tx fixes

Signed-off-by: Felix Fietkau <nbd@nbd.name>
8 years agoiperf: used an updated renamed tarball instead of main upstream URL
Felix Fietkau [Mon, 17 Oct 2016 09:16:30 +0000 (11:16 +0200)]
iperf: used an updated renamed tarball instead of main upstream URL

iperf upstream added some bugfixes to the already released 2.0.9 version
without changing the filename. This conflicts with old mirrored files
and the hash that we previously used.
To avoid conflict, use a renamed tarball from mirror2.openwrt.org
containing the new upstream changes

Signed-off-by: Felix Fietkau <nbd@nbd.name>
8 years agomac80211: fix build error in ath10k with hwmon enabled
Felix Fietkau [Mon, 17 Oct 2016 08:57:43 +0000 (10:57 +0200)]
mac80211: fix build error in ath10k with hwmon enabled

Signed-off-by: Felix Fietkau <nbd@nbd.name>
8 years agobusybox: add upstream patch to fix send_to_from
Stijn Tintel [Sat, 15 Oct 2016 16:14:07 +0000 (19:14 +0300)]
busybox: add upstream patch to fix send_to_from

This fixes ntpd when queries are sent to secondary IPv4 IP addresses.
See https://bugs.busybox.net/show_bug.cgi?id=9146

Signed-off-by: Stijn Tintel <stijn@linux-ipv6.be>
8 years agobusybox: update to version 1.25.1
Stijn Tintel [Sat, 15 Oct 2016 16:06:11 +0000 (19:06 +0300)]
busybox: update to version 1.25.1

Dropped upstream patches:
000-busybox-1.25.0-gzip.patch
000-busybox-1.25.0-ip.patch
000-busybox-1.25.0-ntpd.patch

Signed-off-by: Stijn Tintel <stijn@linux-ipv6.be>
8 years agoar71xx: set EU region code for TP-Link TL-WA901ND v4
Matthias Schiffer [Sat, 15 Oct 2016 17:09:56 +0000 (19:09 +0200)]
ar71xx: set EU region code for TP-Link TL-WA901ND v4

There is no US firmware for the TL-WA901ND v4 yet, so we'll just
unconditionally set the EU region for now.

This makes LEDE flashable on these devices again. The format of the region
string is slightly different from the one used on the Archer C7 that is
generated by mktplinkfw (the second half of the region string is missing),
but it's similar enough to make it work.

Tested-by: Jannis Pinter <jannis@pinterjann.is>
Signed-off-by: Matthias Schiffer <mschiffer@universe-factory.net>
8 years agoprocd: update sha256sum
Hauke Mehrtens [Sat, 15 Oct 2016 15:56:23 +0000 (17:56 +0200)]
procd: update sha256sum

Commit f5c741b5e02 updated procd to a more recent version, but did not
change the hash of the tar. Update it to the one matching the file on
the download servers.

Signed-off-by: Hauke Mehrtens <hauke@hauke-m.de>
8 years agoltq-adsl-mei: fix build error
Mathias Kresin [Sat, 15 Oct 2016 14:29:36 +0000 (16:29 +0200)]
ltq-adsl-mei: fix build error

The file to patch doesn't exists. Due to the fact that a src directory
exists, the patch was never applied.

With 28502a9 the behaviour was changed and revealed the issue.

Signed-off-by: Mathias Kresin <dev@kresin.me>
8 years agoprocd: update to latest git HEAD revision
John Crispin [Fri, 14 Oct 2016 23:38:56 +0000 (01:38 +0200)]
procd: update to latest git HEAD revision

Signed-off-by: John Crispin <john@phrozen.org>
8 years agoscripts/freebsd.sh: Remove script
Daniel Engberg [Fri, 30 Sep 2016 23:36:11 +0000 (01:36 +0200)]
scripts/freebsd.sh: Remove script

This is outdated and bad practice,  general dependencies should be documented and leave the rest up to the user.

Signed-off-by: Daniel Engberg <daniel.engberg.lists@pyret.net>
8 years agoboot/rbcfg: drop Build/Prepare rule in favor of default one
Alexandru Ardelean [Thu, 6 Oct 2016 06:40:51 +0000 (09:40 +0300)]
boot/rbcfg: drop Build/Prepare rule in favor of default one

Signed-off-by: Alexandru Ardelean <ardeleanalex@gmail.com>
8 years agokernel/wrt55agv2-spidevs: drop Build/Prepare rule in favor of default one
Alexandru Ardelean [Thu, 6 Oct 2016 06:39:52 +0000 (09:39 +0300)]
kernel/wrt55agv2-spidevs: drop Build/Prepare rule in favor of default one

Signed-off-by: Alexandru Ardelean <ardeleanalex@gmail.com>
8 years agokernel/w1-gpio-custom: drop Build/Prepare rule in favor of default one
Alexandru Ardelean [Thu, 6 Oct 2016 06:39:20 +0000 (09:39 +0300)]
kernel/w1-gpio-custom: drop Build/Prepare rule in favor of default one

Signed-off-by: Alexandru Ardelean <ardeleanalex@gmail.com>
8 years agokernel/trelay: drop Build/Prepare rule in favor of default one
Alexandru Ardelean [Thu, 6 Oct 2016 06:39:01 +0000 (09:39 +0300)]
kernel/trelay: drop Build/Prepare rule in favor of default one

Signed-off-by: Alexandru Ardelean <ardeleanalex@gmail.com>
8 years agokernel/spi-gpio-custom: drop Build/Prepare rule in favor of default one
Alexandru Ardelean [Thu, 6 Oct 2016 06:38:42 +0000 (09:38 +0300)]
kernel/spi-gpio-custom: drop Build/Prepare rule in favor of default one

Signed-off-by: Alexandru Ardelean <ardeleanalex@gmail.com>
8 years agokernel/rtc-rv5c386a: drop Build/Prepare rule in favor of default one
Alexandru Ardelean [Thu, 6 Oct 2016 06:38:25 +0000 (09:38 +0300)]
kernel/rtc-rv5c386a: drop Build/Prepare rule in favor of default one

Signed-off-by: Alexandru Ardelean <ardeleanalex@gmail.com>
8 years agokernel/rotary-gpio-custom: drop Build/Prepare rule in favor of default one
Alexandru Ardelean [Thu, 6 Oct 2016 06:37:38 +0000 (09:37 +0300)]
kernel/rotary-gpio-custom: drop Build/Prepare rule in favor of default one

Signed-off-by: Alexandru Ardelean <ardeleanalex@gmail.com>
8 years agokernel/lantiq/ltq-*: drop Build/Prepare rule in favor of default one
Alexandru Ardelean [Thu, 6 Oct 2016 06:37:24 +0000 (09:37 +0300)]
kernel/lantiq/ltq-*: drop Build/Prepare rule in favor of default one

Signed-off-by: Alexandru Ardelean <ardeleanalex@gmail.com>
8 years agokernel/i2c-gpio-custom: drop Build/Prepare rule in favor of default one
Alexandru Ardelean [Thu, 6 Oct 2016 06:35:12 +0000 (09:35 +0300)]
kernel/i2c-gpio-custom: drop Build/Prepare rule in favor of default one

Signed-off-by: Alexandru Ardelean <ardeleanalex@gmail.com>
8 years agokernel/gpio-button-hotplug: drop Build/Prepare rule in favor of default one
Alexandru Ardelean [Thu, 6 Oct 2016 06:34:46 +0000 (09:34 +0300)]
kernel/gpio-button-hotplug: drop Build/Prepare rule in favor of default one

Signed-off-by: Alexandru Ardelean <ardeleanalex@gmail.com>
8 years agokernel/button-hotplug: drop Build/Prepare rule in favor of default one
Alexandru Ardelean [Thu, 6 Oct 2016 06:34:17 +0000 (09:34 +0300)]
kernel/button-hotplug: drop Build/Prepare rule in favor of default one

Signed-off-by: Alexandru Ardelean <ardeleanalex@gmail.com>
8 years agokernel/avila-wdt: drop Build/Prepare rule in favor of default one
Alexandru Ardelean [Thu, 6 Oct 2016 06:33:38 +0000 (09:33 +0300)]
kernel/avila-wdt: drop Build/Prepare rule in favor of default one

Signed-off-by: Alexandru Ardelean <ardeleanalex@gmail.com>
8 years agonetwork/utils/maccalc: drop Build/Prepare rule in favor of default one
Alexandru Ardelean [Thu, 6 Oct 2016 06:28:30 +0000 (09:28 +0300)]
network/utils/maccalc: drop Build/Prepare rule in favor of default one

Signed-off-by: Alexandru Ardelean <ardeleanalex@gmail.com>
8 years agonetwork/utils/rssileds: drop Build/Prepare rule in favor of default one
Alexandru Ardelean [Thu, 6 Oct 2016 06:27:59 +0000 (09:27 +0300)]
network/utils/rssileds: drop Build/Prepare rule in favor of default one

Signed-off-by: Alexandru Ardelean <ardeleanalex@gmail.com>
8 years agonetwork/utils/resolveip: drop Build/Prepare rule in favor of default one
Alexandru Ardelean [Thu, 6 Oct 2016 06:27:27 +0000 (09:27 +0300)]
network/utils/resolveip: drop Build/Prepare rule in favor of default one

Signed-off-by: Alexandru Ardelean <ardeleanalex@gmail.com>
8 years agonetwork/utils/owipcalc: drop Build/Prepare rule in favor of default one
Alexandru Ardelean [Thu, 6 Oct 2016 06:27:08 +0000 (09:27 +0300)]
network/utils/owipcalc: drop Build/Prepare rule in favor of default one

Signed-off-by: Alexandru Ardelean <ardeleanalex@gmail.com>
8 years agonetwork/ipv6/map: drop Build/Prepare rule in favor of default one
Alexandru Ardelean [Thu, 6 Oct 2016 06:26:34 +0000 (09:26 +0300)]
network/ipv6/map: drop Build/Prepare rule in favor of default one

Signed-off-by: Alexandru Ardelean <ardeleanalex@gmail.com>
8 years agonetwork/utils/iwcap: drop Build/Prepare rule in favor of default one
Alexandru Ardelean [Thu, 6 Oct 2016 06:25:56 +0000 (09:25 +0300)]
network/utils/iwcap: drop Build/Prepare rule in favor of default one

Signed-off-by: Alexandru Ardelean <ardeleanalex@gmail.com>
8 years agonetwork/services/ead: drop Build/Prepare rule in favor of default one
Alexandru Ardelean [Thu, 6 Oct 2016 06:24:49 +0000 (09:24 +0300)]
network/services/ead: drop Build/Prepare rule in favor of default one

Signed-off-by: Alexandru Ardelean <ardeleanalex@gmail.com>
8 years agonetwork/ipv6/6rd: drop Build/Prepare rule in favor of default one
Alexandru Ardelean [Thu, 6 Oct 2016 06:24:31 +0000 (09:24 +0300)]
network/ipv6/6rd: drop Build/Prepare rule in favor of default one

Signed-off-by: Alexandru Ardelean <ardeleanalex@gmail.com>
8 years agonetwork/config/swconfig: drop Build/Prepare rule in favor of default one
Alexandru Ardelean [Thu, 6 Oct 2016 06:24:10 +0000 (09:24 +0300)]
network/config/swconfig: drop Build/Prepare rule in favor of default one

Signed-off-by: Alexandru Ardelean <ardeleanalex@gmail.com>
8 years agosystem/mtd: drop Build/Prepare rule in favor of default one
Alexandru Ardelean [Thu, 6 Oct 2016 06:23:28 +0000 (09:23 +0300)]
system/mtd: drop Build/Prepare rule in favor of default one

Signed-off-by: Alexandru Ardelean <ardeleanalex@gmail.com>
8 years agolibs/gettext: drop Build/Prepare rule in favor of default one
Alexandru Ardelean [Thu, 6 Oct 2016 06:22:45 +0000 (09:22 +0300)]
libs/gettext: drop Build/Prepare rule in favor of default one

Signed-off-by: Alexandru Ardelean <ardeleanalex@gmail.com>
8 years agoutils/spidev_test: drop Build/Prepare rule in favor of default one
Alexandru Ardelean [Thu, 6 Oct 2016 06:21:58 +0000 (09:21 +0300)]
utils/spidev_test: drop Build/Prepare rule in favor of default one

Signed-off-by: Alexandru Ardelean <ardeleanalex@gmail.com>
8 years agoutils/otrx: drop Build/Prepare rule in favor of default one
Alexandru Ardelean [Thu, 6 Oct 2016 06:21:44 +0000 (09:21 +0300)]
utils/otrx: drop Build/Prepare rule in favor of default one

Signed-off-by: Alexandru Ardelean <ardeleanalex@gmail.com>
8 years agoutils/usbreset: drop Build/Prepare rule in favor of default one
Alexandru Ardelean [Thu, 6 Oct 2016 06:21:17 +0000 (09:21 +0300)]
utils/usbreset: drop Build/Prepare rule in favor of default one

Signed-off-by: Alexandru Ardelean <ardeleanalex@gmail.com>
8 years agoutils/px5g-standalone: drop Build/Prepare rule in favor of default one
Alexandru Ardelean [Thu, 6 Oct 2016 06:20:21 +0000 (09:20 +0300)]
utils/px5g-standalone: drop Build/Prepare rule in favor of default one

Signed-off-by: Alexandru Ardelean <ardeleanalex@gmail.com>
8 years agoutils/oseama: drop Build/Prepare rule in favor of default one
Alexandru Ardelean [Thu, 6 Oct 2016 06:20:01 +0000 (09:20 +0300)]
utils/oseama: drop Build/Prepare rule in favor of default one

Signed-off-by: Alexandru Ardelean <ardeleanalex@gmail.com>
8 years agoutils/nvram: drop Build/Prepare rule in favor of default one
Alexandru Ardelean [Thu, 6 Oct 2016 06:19:34 +0000 (09:19 +0300)]
utils/nvram: drop Build/Prepare rule in favor of default one

Signed-off-by: Alexandru Ardelean <ardeleanalex@gmail.com>
8 years agoutils/fbtest: drop Build/Prepare rule in favor of default one
Alexandru Ardelean [Thu, 6 Oct 2016 06:19:15 +0000 (09:19 +0300)]
utils/fbtest: drop Build/Prepare rule in favor of default one

Signed-off-by: Alexandru Ardelean <ardeleanalex@gmail.com>
8 years agolibs/libiconv: drop Build/Prepare rule in favor of default one
Alexandru Ardelean [Thu, 6 Oct 2016 06:17:49 +0000 (09:17 +0300)]
libs/libiconv: drop Build/Prepare rule in favor of default one

Signed-off-by: Alexandru Ardelean <ardeleanalex@gmail.com>
8 years agolibs/libnl-tiny: drop Build/Prepare rule in favor of default one
Alexandru Ardelean [Thu, 6 Oct 2016 06:14:34 +0000 (09:14 +0300)]
libs/libnl-tiny: drop Build/Prepare rule in favor of default one

Signed-off-by: Alexandru Ardelean <ardeleanalex@gmail.com>
8 years agonetwork/services/hostapd: move whole files outside of patches and drop Build/Prepare...
Alexandru Ardelean [Wed, 5 Oct 2016 14:42:47 +0000 (17:42 +0300)]
network/services/hostapd: move whole files outside of patches and drop Build/Prepare rule in favor of default one

This more of a demo for the previous commit that comes with
this one, where I added support for copying source from 'src' to
the build dir(s).

Signed-off-by: Alexandru Ardelean <ardeleanalex@gmail.com>
8 years agobuild: copy contents of 'src' folder to build dirs (if present)
Alexandru Ardelean [Wed, 5 Oct 2016 14:28:46 +0000 (17:28 +0300)]
build: copy contents of 'src' folder to build dirs (if present)

The normal Prepare step for a build is unpack, apply patches.
But for certain packages, patches contain whole files, which
would be nice to have separately and copied over as a last step
in the Prepare phase.

We need it for some other packages + patches, but I think
the 'hostapd' package can be used as a test for this.

As a quick note:
the reason the condition is being evaluated as
`[ ! -d ./src/ ] || $(CP) ./src/* $(HOST_BUILD_DIR)`
and not with
`[ -d ./src/ ] && $(CP) ./src/* $(HOST_BUILD_DIR)`
is that the latter would translate in a build failure if the `src`
folder is not present (the exit code would be 1).
The first one, succeeds for both cases (if `src` present or not).

Signed-off-by: Alexandru Ardelean <ardeleanalex@gmail.com>
8 years agobusybox: prevent globbing, word splitting
Jan-Tarek Butt [Fri, 14 Oct 2016 18:32:22 +0000 (20:32 +0200)]
busybox: prevent globbing, word splitting

Double quote to prevent globbing and word splitting.

Signed-off-by: Jan-Tarek Butt <tarek@ring0.de>
8 years agouboot-ar71xx: make reproducible
Dirk Neukirchen [Wed, 12 Oct 2016 12:07:44 +0000 (14:07 +0200)]
uboot-ar71xx: make reproducible

OpenWrt uses ancient u-boot thats not reproducible.

There are multiple upstream changes that introduce
reproducible builds like:
859e92b775fd8ebcfacc591eaf621b677c95b6f7
(not used here - the CMD_DATE/TIMESTAMP functionality
seems to be disabled by config)

70d39f57146a6cb94736db39c770c3d95e07bedb
f3f431a712729a1af94d01bd1bfde17a252ff02c
2d9efa1227262249d381ed5d9d341cbdba76e62d

Instead of changing the Makefile too much
this changeset just tries to use the
changes in Makefile from current upstream git f5fd45f

*Should* fix issue reported by reproducible lede page:
https://tests.reproducible-builds.org/lede/lede.html

Compile tested only
(verified w. hexdump & md5sum)

Signed-off-by: Dirk Neukirchen <dirkneukirchen@web.de>
8 years agobuild: fix cleaning configured stamp file
Yousong Zhou [Mon, 26 Sep 2016 13:14:50 +0000 (21:14 +0800)]
build: fix cleaning configured stamp file

We have packages with their own parts appended to standard STAMP_CONFIGURED
(mostly with an underscore character).  This will render the current
STAMP_CONFIGURED_WILD setting invalid and the build system may miss a rebuild
on config change

 1. Build with config A
 2. Build with config B, yet .configured_A_xx did not get cleaned
 3. Return to config A, but rebuild will not happen because stamp file
    of config A still exists

Signed-off-by: Yousong Zhou <yszhou4tech@gmail.com>
8 years agoscripts: case insensitive sort device names
Karl Palsson [Wed, 12 Oct 2016 13:51:23 +0000 (13:51 +0000)]
scripts: case insensitive sort device names

When selecting devices from the Target Devices menu, the brand choices
of naming makes it confusing to find particular devices by name, as the
sorting is case sensitve.  AirTight came after ALFA, and devolo and
jjPlus both came after Zyxel.

This does _not_ apply to the Target Profile list, as that includes
"Default - all profiles" inside the profile list.

Signed-off-by: Karl Palsson <karlp@etactica.com>
8 years agoscripts: add help text for some generated KConfigs
Karl Palsson [Wed, 12 Oct 2016 13:51:22 +0000 (13:51 +0000)]
scripts: add help text for some generated KConfigs

TARGET_MULTI_PROFILE and TARGET_PER_DEVICE_ROOTFS get some help text to
try and clarify their behaviour.

Signed-off-by: Karl Palsson <karlp@etactica.com>
8 years agoramips: add support for Planex VR500.
YuheiOKAWA [Mon, 10 Oct 2016 01:24:42 +0000 (10:24 +0900)]
ramips: add support for Planex VR500.

SOC: MT7621A
RAM: 256MiB
NOR: MX25L51245G

Non Wireless Router.

Issue: soft reboot problem. SPI Flash do not exit 4byte address mode.

Signed-off-by: Yuhei Okawa <tochiro.srchack@gmail.com>
8 years agoltq-ptm: Support 1508-byte MTU for RFC4638
David Woodhouse [Fri, 7 Oct 2016 14:02:13 +0000 (15:02 +0100)]
ltq-ptm: Support 1508-byte MTU for RFC4638

Tested with VDSL on TP-Link WD8970, I see full 1500-byte PPP data
frames, which end up being 1526 byte Ethernet frames (including
Ethernet+VLAN headers) on the wire.

Fixes: FS#210
Signed-off-by: David Woodhouse <dwmw2@infradead.org>
8 years agoscripts: fix build warning when overriding packages
Zefir Kurtisi [Wed, 12 Oct 2016 15:50:56 +0000 (17:50 +0200)]
scripts: fix build warning when overriding packages

If core packages are overridden, CONFIG_OVERRIDE_PKGS is set
based on the scan order of packages, which eventually causes
that config value to be modified on each build and with
that causes the build process to warn for configuration
being out of sync.

This commit changes the CONFIG_OVERRIDE_PKGS to be sorted
and prevents that false warning.

Signed-off-by: Zefir Kurtisi <zefir.kurtisi@neratec.com>