openwrt/staging/chunkeey.git
6 years agomac80211: revert "wireless: set correct mandatory rate flags"
Matthias Schiffer [Fri, 26 Jan 2018 22:24:59 +0000 (23:24 +0100)]
mac80211: revert "wireless: set correct mandatory rate flags"

Revert upstream commit 1bd773c077de "wireless: set correct mandatory rate
flags", as it breaks 11s interoperability: nodes can only associate when
neither or both have this patch. As this is a regression from released
versions, revert to the old code for now.

Signed-off-by: Matthias Schiffer <mschiffer@universe-factory.net>
6 years agorules.mk: drop `include_mk` build rule
Alexandru Ardelean [Wed, 24 Jan 2018 10:56:39 +0000 (12:56 +0200)]
rules.mk: drop `include_mk` build rule

The only users of this were the python packages
from the `packages` feed.
The 2 python interpreters would export some mk
files (e.g. python-package.mk) and then other
python packages would include it via this rule.

But there's a few things wrong with this approach,
most of them drawing from the fact that python host
needs to be built first, to export these mk files.

By now all uses of include_mk have been corrected
in the feeds and this can be removed.

Signed-off-by: Alexandru Ardelean <ardeleanalex@gmail.com>
6 years agoopenssl: tell the build system that we are doing CROSS_COMPILE
Yousong Zhou [Fri, 26 Jan 2018 10:13:21 +0000 (18:13 +0800)]
openssl: tell the build system that we are doing CROSS_COMPILE

So that it will not try to run c_rehash with the just built binaries on
certs/demo.

Fixes openwrt/packages#5432

Reported-by: Val Kulkov <val.kulkov@gmail.com>
Signed-off-by: Yousong Zhou <yszhou4tech@gmail.com>
6 years agoiptables: make kmod-ipt-debug part of default ALL build
Yousong Zhou [Fri, 26 Jan 2018 06:40:25 +0000 (14:40 +0800)]
iptables: make kmod-ipt-debug part of default ALL build

The iptables TRACE target is only available in raw table that's why the
dependency was moved from iptables-mod-trace into kmod-ipt-debug

Fixes FS#1219

Signed-off-by: Yousong Zhou <yszhou4tech@gmail.com>
6 years agobuild: disable BUILD_PATENTED by default
Yousong Zhou [Thu, 25 Jan 2018 01:40:33 +0000 (09:40 +0800)]
build: disable BUILD_PATENTED by default

This is mainly for legal considerations and not promoting the usage of
and no redistribution of binaries of patented technologies seems to be
also the established practice in other linux distros.

Signed-off-by: Yousong Zhou <yszhou4tech@gmail.com>
6 years agowireguard: bump to 20180118
Kevin Darbyshire-Bryant [Thu, 25 Jan 2018 17:20:51 +0000 (17:20 +0000)]
wireguard: bump to 20180118

Bump to latest wireguard release snapshot:

9a93a3d version: bump snapshot
7bc0579 contrib: keygen-html: update curve25519 implementation
ffc13a3 tools: import new curve25519 implementations
0ae7356 curve25519: wire up new impls and remove donna
f90e36b curve25519: resolve symbol clash between fe types
505bc05 curve25519: import 64-bit hacl-star implementation
8c02050 curve25519: import 32-bit fiat-crypto implementation
96157fd curve25519: modularize implementation
4830fc7 poly1305: remove indirect calls
bfd1a5e tools: plug memleak in config error path
09bf49b external-tests: add python implementation
b4d5801 wg-quick: ifnames have max len of 15
6fcd86c socket: check for null socket before fishing out sport
ddb8270 global: year bump
399d766 receive: treat packet checking as irrelevant for timers

No patch refresh required.

Compile-tested-for: ar71xx
Run-tested-on: ar71xx Archer C7 v2

Signed-off-by: Kevin Darbyshire-Bryant <ldir@darbyshire-bryant.me.uk>
6 years agoRevert "mt76: update to the latest version"
Felix Fietkau [Thu, 25 Jan 2018 20:18:57 +0000 (21:18 +0100)]
Revert "mt76: update to the latest version"

This reverts commit 99eb128acaf76a69119fd2de8e194f2b2bbb0427.
Connectivity issues reported by users, needs rework

Signed-off-by: Felix Fietkau <nbd@nbd.name>
6 years agomt76: update to the latest version
Felix Fietkau [Thu, 25 Jan 2018 15:49:14 +0000 (16:49 +0100)]
mt76: update to the latest version

2b7fae4 mt76: fix returnvar.cocci warnings
939e3e0 mt76x2: dfs: avoid tasklet scheduling during mt76x2_dfs_init_params()
cf59170 mt76x2: dfs: add set_domain handler
5e4d60e mt76x2: dfs: take into account dfs region in mt76x2_dfs_init_params()
f76e25f mt76x2: fix WMM parameter configuration
34d612d mt76: retry rx polling as long as there is budget left
0f8327a mt76x2: fix TSF value in probe responses
ad3f8e9 mt76: add an intermediate struct for rx status information
58a41f1 mt76: get station pointer by wcid and pass it to mac80211
b0508d3 mt76: implement A-MPDU rx reordering in the driver code
cf3cfc4 mt76: split mt76_rx_complete
461cdf9 mt76: pass the per-vif wcid to the core for multicast rx
9b2c778 mt76: validate rx CCMP PN
302af90 mt76x2: init: disable all pending tasklets during device removal
9f685fe mt7603: init: disable tbtt tasklet during device removal
c6f8cac mt76: let mac80211 validate CCMP PN for fragmented frames
3968dae mt7603: fix 40 mhz channel bandwidth reporting
9c2e03d mt7603: fix rx LDPC reporting

Signed-off-by: Felix Fietkau <nbd@nbd.name>
6 years agoar71xx: add ew-balin platform from Embedded Wireless
Catrinel Catrinescu [Mon, 15 Jan 2018 15:45:16 +0000 (16:45 +0100)]
ar71xx: add ew-balin platform from Embedded Wireless

Add the Embedded Wireless "Balin" platform
 SoC: QCA AR9344 or AR9350
 RAM: DDR2-RAM 64MBytes
 Flash: SPI-NOR 16MBytes
 WLAN: 2 x 2 MIMO 2.4 & 5 GHz IEEE802.11 a/b/g/n
 Ethernet: 3 x 10/100 Mb/s
 USB: 1 x USB2.0 Host/Device bootstrap-pin at power-up
 PCI-Express: 1 x lane PCIe 1.2
 UART: 1 x Normal, 1 x High-Speed
 JTAG: 1 x EJTAG
 GPIO: 10 x Input/Output multiplexed

The module comes already with the current vanilla OpenWrt firmware.
To update, use "sysupgrade" image directly in vendor firmware.

Signed-off-by: Catrinel Catrinescu <cc@80211.de>
6 years agoar71xx: add unaligned access hacks for VXLAN
Matthias Schiffer [Tue, 23 Jan 2018 19:52:45 +0000 (20:52 +0100)]
ar71xx: add unaligned access hacks for VXLAN

Gives a ~5% performance gain.

Signed-off-by: Matthias Schiffer <mschiffer@universe-factory.net>
6 years agomusl: move BUILD_DIR_TOOLCHAIN/musl symlink to configure step
Felix Fietkau [Wed, 24 Jan 2018 15:46:22 +0000 (16:46 +0100)]
musl: move BUILD_DIR_TOOLCHAIN/musl symlink to configure step

Avoids Build/Prepare quilt related hacks

Signed-off-by: Felix Fietkau <nbd@nbd.name>
6 years agomusl: allow autorebuild
Felix Fietkau [Wed, 24 Jan 2018 15:43:28 +0000 (16:43 +0100)]
musl: allow autorebuild

Autorebuild is disabled for the toolchain to avoid build-order issues.
However, rebuilding musl is safe, so exclude it from that restriction.
Avoids the need for manual cleaning on kernel header <-> libc API
changes like the ones introduced recently

Signed-off-by: Felix Fietkau <nbd@nbd.name>
6 years agomac80211: mesh: drop frames appearing to be from us
Felix Fietkau [Fri, 19 Jan 2018 17:34:10 +0000 (18:34 +0100)]
mac80211: mesh: drop frames appearing to be from us

Upstream backport to fix issues arising from devices with duplicate MAC
addresses

Signed-off-by: Felix Fietkau <nbd@nbd.name>
6 years agoramips: add support for Widora Neo 32MB flash revision
Jackson Ming Hu [Mon, 8 Jan 2018 05:27:28 +0000 (13:27 +0800)]
ramips: add support for Widora Neo 32MB flash revision

Widora has updated their Widora Neo board recently.

The new model uses 32MB WSON-8 factor SPI flash
instead of the original 16MB SOP-8 factor SPI flash.

All the other hardware components are the same as
the first revision.

Detailed hardware specs listed below:

CPU: MTK MT7688AN
RAM: 128MB DDR2
ROM: 32MB WSON-8 factor SPI Flash (Winbond)
WiFi: Built-in 802.11n 150Mbps?
Ethernet: 10/100Mbps x1
Audio codec: WM8960
Other IO: USB OTG;
  USB Power+Serial (CP2104);
  3x LEDs (Power, LAN, WiFi);
  2x Keys (WPS, CPU Reset)
  1x Audio In/Out
  1x IPEX antenna port
  1x Micro SD slot

Signed-off-by: Jackson Ming Hu <huming2207@gmail.com>
Signed-off-by: Mathias Kresin <dev@kresin.me>
6 years agoramips: add flash size postfix to Widora neo
Mathias Kresin [Wed, 17 Jan 2018 07:14:41 +0000 (08:14 +0100)]
ramips: add flash size postfix to Widora neo

Rename the Widora neo by adding a flash size prefix. Move the common parts
into a dtsi to be prepare everything for upcomming support of the 32MB
version.

Migrate the Widora neo to the generic board detection as well.

Signed-off-by: Mathias Kresin <dev@kresin.me>
6 years agokernel: generic: add 4.9 config option
Hans Dedecker [Tue, 23 Jan 2018 20:58:10 +0000 (21:58 +0100)]
kernel: generic: add 4.9 config option

When CGROUPS is enabled the new option CONFIG_CGROUP_NET_CLASSID is
selectable and not handled.
Add this option to the 4.9 kernel configuration.

Signed-off-by: Hans Dedecker <dedeckeh@gmail.com>
6 years agovxlan: add options to enable and disable UDP checksums
Matthias Schiffer [Wed, 24 Jan 2018 12:51:45 +0000 (13:51 +0100)]
vxlan: add options to enable and disable UDP checksums

Signed-off-by: Matthias Schiffer <mschiffer@universe-factory.net>
6 years agonetifd: update to latest git HEAD
Matthias Schiffer [Wed, 24 Jan 2018 12:50:50 +0000 (13:50 +0100)]
netifd: update to latest git HEAD

af3cadb system-linux: VXLAN: add options to enable and disable UDP checksums

Signed-off-by: Matthias Schiffer <mschiffer@universe-factory.net>
6 years agoramips: Fix GB-PC1 cpuclock again
Rosen Penev [Sun, 21 Jan 2018 03:56:33 +0000 (19:56 -0800)]
ramips: Fix GB-PC1 cpuclock again

The intended frequency is 900 MHz, not 90.

Fixes: 7059ab48a6d5 ("ramips: fix cpuclock for the GB-PC1")
Signed-off-by: Rosen Penev <rosenp@gmail.com>
6 years agoar71xx: fix MikroTik rb-nor-flash-16M-ac image
Daniel Golle [Wed, 24 Jan 2018 00:20:41 +0000 (01:20 +0100)]
ar71xx: fix MikroTik rb-nor-flash-16M-ac image

commit e15c63a375
ar71xx: add support for MikroTik RouterBOARD wAP G-5HacT2HnD (wAP AC)

changed the existing rb-nor-flash-16M-ac image in a way that it would
now only support the rb-wapg-5hact2hnd.
The board show however rather be added to the existing boards in the
rb-nor-flash-16M image template.

Reported-by: Mathias Kresin <dev@kresin.me>
Signed-off-by: Daniel Golle <daniel@makrotopia.org>
6 years agoprocd: update to latest git HEAD
Hans Dedecker [Tue, 23 Jan 2018 10:37:37 +0000 (11:37 +0100)]
procd: update to latest git HEAD

653629f trace: check asprintf() return value
67eb7e6 trace: add missing limits.h include

Signed-off-by: Hans Dedecker <dedeckeh@gmail.com>
6 years agoramips: add support for ZBT-WE1226
Daniel Golle [Tue, 23 Jan 2018 01:46:36 +0000 (02:46 +0100)]
ramips: add support for ZBT-WE1226

MT7628NN (580MHz), 8MB SPI NOR, 64MB DDR2 RAM

Everything except for the switch LEDs works great.

Signed-off-by: Daniel Golle <daniel@makrotopia.org>
6 years agosunxi: Orange Pi R1: configure USB Ethernet controller.
Hauke Mehrtens [Thu, 28 Dec 2017 18:36:08 +0000 (19:36 +0100)]
sunxi: Orange Pi R1: configure USB Ethernet controller.

Now the USB port is the WAN port.

Signed-off-by: Hauke Mehrtens <hauke@hauke-m.de>
6 years agosunxi: Orange Pi R1: Fix USB Ethernet and activate SPI
Hauke Mehrtens [Thu, 28 Dec 2017 18:34:47 +0000 (19:34 +0100)]
sunxi: Orange Pi R1: Fix USB Ethernet and activate SPI

The USB Ethernet is not working with the patches proposed for upstream,
fix this and activate the SPI node as this board always has a SPI flash.

Both patches are also targeted for upstream kernel 4.16 and 4.17.

Signed-off-by: Hauke Mehrtens <hauke@hauke-m.de>
6 years agosunxi: use upstream patch for Orange Pi R1
Hauke Mehrtens [Sun, 10 Dec 2017 19:56:56 +0000 (20:56 +0100)]
sunxi: use upstream patch for Orange Pi R1

Instead of using our own device tree definitions use the one provided in
the upstream kernel for 4.16.

Signed-off-by: Hauke Mehrtens <hauke@hauke-m.de>
6 years agosunxi: backport stmmac network patches
Hauke Mehrtens [Sun, 10 Dec 2017 19:56:31 +0000 (20:56 +0100)]
sunxi: backport stmmac network patches

Ethernet support was initial added in kernel 4.13, but deactivated
before the final release. This is backports the changes which are
activating it again from kernel 4.15.

Signed-off-by: Hauke Mehrtens <hauke@hauke-m.de>
6 years agosunxi: Add support for kernel 4.14
Hauke Mehrtens [Sun, 10 Dec 2017 17:42:44 +0000 (18:42 +0100)]
sunxi: Add support for kernel 4.14

This is based on the code for kernel 4.9, but a lot of 4.9 patches are
backports from more recent kernel version and can be removed now.

Signed-off-by: Hauke Mehrtens <hauke@hauke-m.de>
6 years agosunxi: refresh kernel configuration
Hauke Mehrtens [Sat, 16 Dec 2017 22:47:28 +0000 (23:47 +0100)]
sunxi: refresh kernel configuration

Just refresh the sunxi kernel configuration.
This also moves the CONFIG_CRYPTO_DEV_SUN4I_SS_PRNG option to the
config-4.9 file.

Signed-off-by: Hauke Mehrtens <hauke@hauke-m.de>
6 years agoRevert "kernel: add IEEE-1284 parallel port support"
Jo-Philipp Wich [Mon, 22 Jan 2018 12:41:16 +0000 (13:41 +0100)]
Revert "kernel: add IEEE-1284 parallel port support"

This reverts commit 666e9cf2220b11ccd024cad13ad54ca71d40c5b3.

The change has not been build-tested on non-x86 targets and leads to
stalled kernel builds due to unset configuration symbols there.

Signed-off-by: Jo-Philipp Wich <jo@mein.io>
6 years agoprocd: update to latest git HEAD
John Crispin [Mon, 22 Jan 2018 11:53:34 +0000 (12:53 +0100)]
procd: update to latest git HEAD

846e20c procd: add timing to start/stop logging

Signed-off-by: John Crispin <john@phrozen.org>
6 years agoar71xx: add support for the MikroTik RB911-2Hn/5Hn boards
Gabor Juhos [Thu, 18 Jan 2018 12:50:32 +0000 (13:50 +0100)]
ar71xx: add support for the MikroTik RB911-2Hn/5Hn boards

The patch adds support for the MikroTik RB911-2Hn (911 Lite2)
and the RB911-5Hn (911 Lite5) boards:

  https://mikrotik.com/product/RB911-2Hn
  https://mikrotik.com/product/RB911-5Hn

The two boards are using the same hardware design, the only difference
between the two is the supported wireless band.

Specifications:
  * SoC: Atheros AR9344 (600MHz)
  * RAM: 64MiB
  * Storage: 16 MiB SPI NOR flash
  * Ethernet: 1x100M (Passive PoE in)
  * Wireless: AR9344 built-in wireless MAC, single chain
              802.11b/g/n (911-2Hn) or 802.11a/g/n (911-5Hn)

Notes:
  * Older versions of these boards might be equipped with a NAND
    flash chip instead of the SPI NOR device. Those boards are not
    supported (yet).
  * The MikroTik RB911-5HnD (911 Lite5 Dual) board also uses the
    same hardware. Support for that can be added later with little
    effort probably.

Installation:

1. Setup a DHCP/BOOTP Server with the following parameters:
   * DHCP-Option 66 (TFTP server name): pointing to a local TFTP
     server within the same subnet of the DHCP range
   * DHCP-Option 67 (Bootfile-Name): matching the initramfs filename
     of the to be booted image. The usable intramfs files are:
       - openwrt-ar71xx-mikrotik-vmlinux-initramfs.elf
       - openwrt-ar71xx-mikrotik-vmlinux-initramfs-lzma.elf
       - openwrt-ar71xx-mikrotik-rb-nor-flash-16M-initramfs-kernel.bin

2. Press the reset button on the board and keep that pressed.

3. Connect the board to your local network via its ethernet port.

4. Release the button after the LEDs on the board are turned off.
   Now the board should load and start the initramfs image from
   the TFTP server.

5. Upload the sysupgrade image to the board with scp:
     $ scp openwrt-ar71xx-mikrotik-rb-nor-flash-16M-squashfs-sysupgrade.bin root@192.168.1.1:/tmp/fw.bin

5. Log in to the running system listening on 192.168.1.1 via ssh
   as root (without password):
     $ ssh root@192.168.1.1

7. Flash the uploaded firmware file from the ssh session via the
   sysupgrade command:
     root@OpenWrt:~# sysupgrade /tmp/fw.bin

Signed-off-by: Gabor Juhos <juhosg@freemail.hu>
6 years agoar71xx: make leds-gpio usable with single-ended GPIOs
Gabor Juhos [Thu, 18 Jan 2018 12:50:31 +0000 (13:50 +0100)]
ar71xx: make leds-gpio usable with single-ended GPIOs

Add patches for the leds-gpio driver to make it usable with
open-drain and open-source kind of GPIO lines.

This type of functionality is required by various MikroTik boards.

Signed-off-by: Gabor Juhos <juhosg@freemail.hu>
6 years agoar71xx: mach-rbspi: return rb_info from rbspi_platform_setup
Gabor Juhos [Thu, 18 Jan 2018 12:50:30 +0000 (13:50 +0100)]
ar71xx: mach-rbspi: return rb_info from rbspi_platform_setup

Modify the rbspi_platform_setup() function to return the pointer of the
rb_info structure. This allows board specific setup routines to access
the various fields of the information. It is useful for investigating
the hardware option bits for example.

Also update the board setup codes, to ensure that those handle the new
return value correctly.

Signed-off-by: Gabor Juhos <juhosg@freemail.hu>
6 years agoar71xx: add definitions for RouterBOARD hardware option bits
Gabor Juhos [Thu, 18 Jan 2018 12:50:29 +0000 (13:50 +0100)]
ar71xx: add definitions for RouterBOARD hardware option bits

Add bit definitions for the 'hardware options' tag which is used in
the MikroTik devices' hardware configurations. These values can be
used in board setup codes, to do different initialization sequences.
The values were obtained from the RouterOS 6.41-rc38 patches.

Additionally, introduce two helper functions what make the processing
of the hardware options easy.

Signed-off-by: Gabor Juhos <juhosg@freemail.hu>
6 years agouqmi: silence error on pin verification
Koen Vandeputte [Fri, 19 Jan 2018 10:31:52 +0000 (11:31 +0100)]
uqmi: silence error on pin verification

If a device only supports the 2nd verification method (uim),
the first method will fail as expected reporting an error:

"Command not supported"

Silence both separate methods and only report an error regarding
pin verification if both fail.

Signed-off-by: Koen Vandeputte <koen.vandeputte@ncentric.com>
6 years agokernel: add IEEE-1284 parallel port support
Daniel Gimpelevich [Thu, 18 Jan 2018 11:52:12 +0000 (03:52 -0800)]
kernel: add IEEE-1284 parallel port support

The kmod-lp package included both lp.ko and ppdev.ko, but ECP device
drivers may or may not require lp NOT to be loaded, needing only ppdev.
Additionally, There were no packages for any parport interface modules,
such as uss720 or parport_pc, provided here. It has not been otherwise
possible to use PC-style parport hardware for kmod-lp.

Signed-off-by: Daniel Gimpelevich <daniel@gimpelevich.san-francisco.ca.us>
6 years agopistachio: make patches apply again
Hauke Mehrtens [Sat, 20 Jan 2018 22:36:05 +0000 (23:36 +0100)]
pistachio: make patches apply again

Support for Winbond NAND flash detection was added into the generic
patches and this conflicted with this patch adding Gigadevice support.

Fixes: 02050f7e7d5b ("kernel/4.{4, 9}: add manufacturer ID for Winbond NANDs")
Signed-off-by: Hauke Mehrtens <hauke@hauke-m.de>
6 years agoipq806x: make patches apply again
Hauke Mehrtens [Sat, 20 Jan 2018 22:21:52 +0000 (23:21 +0100)]
ipq806x: make patches apply again

Some part of this patch was added to the generic patches as it was
needed also for some other target. Do not add it here any more.

Fixes: 02050f7e7d5b ("kernel/4.{4, 9}: add manufacturer ID for Winbond NANDs")
Signed-off-by: Hauke Mehrtens <hauke@hauke-m.de>
6 years agox86: Add CONFIG_RETPOLINE to fix build
Hauke Mehrtens [Sat, 20 Jan 2018 21:44:40 +0000 (22:44 +0100)]
x86: Add CONFIG_RETPOLINE to fix build

This adds support for the new configuration option CONFIG_RETPOLINE and
refreshes the configuration.

Fixes: d8565a06dc ("kernel: bump 4.9 to 4.9.77")
Fixes: 9ddfac8015 ("kernel: bump 4.14 to 4.14.14")
Signed-off-by: Hauke Mehrtens <hauke@hauke-m.de>
6 years agokernel: use upstream patches for musl
Hauke Mehrtens [Sat, 3 Jun 2017 10:59:55 +0000 (12:59 +0200)]
kernel: use upstream patches for musl

This replaces the current patches used to make the kernel headers
compatible with musl with the version which was accepted upstream. This
is included in upstream kernel 4.15.
This was compile tested with iproute2 build on all supported kernel
versions with musl and one one with glibc.

Signed-off-by: Hauke Mehrtens <hauke@hauke-m.de>
6 years agoubus: fix PKG_MIRROR_HASH
Hauke Mehrtens [Fri, 19 Jan 2018 22:59:30 +0000 (23:59 +0100)]
ubus: fix PKG_MIRROR_HASH

Fixes: dd975d15a71f ("ubus: fix wrong PKG_SOURCE_DATE")
Signed-off-by: Hauke Mehrtens <hauke@hauke-m.de>
6 years agonftables: update to 0.8.1
Philip Prindeville [Wed, 20 Dec 2017 17:22:52 +0000 (10:22 -0700)]
nftables: update to 0.8.1

Note this requires libnftnl-1.0.8 or higher, so that update needs
to be merged first.

Signed-off-by: Philip Prindeville <philipp@redfish-solutions.com>
6 years agolibnftnl: update to 1.0.9
Philip Prindeville [Wed, 20 Dec 2017 22:06:02 +0000 (15:06 -0700)]
libnftnl: update to 1.0.9

Also, drop unsupported configure options.

Don't use git retrieve but released tarball instead.

Signed-off-by: Philip Prindeville <philipp@redfish-solutions.com>
6 years agokernel: backport fix for nftables on big Endian
Hauke Mehrtens [Sat, 20 Jan 2018 19:14:47 +0000 (20:14 +0100)]
kernel: backport fix for nftables on big Endian

nftables 0.8.1 generates some new commands which will not work without
this on big endian systems. This patch is included in Linux 4.11 and
later.

My rule matching a TCP port was not working:
nft add rule ip foo bar ct state new tcp dport 22 accept

Signed-off-by: Hauke Mehrtens <hauke@hauke-m.de>
6 years agotools/tar: update to 1.30
Hannu Nyman [Fri, 19 Jan 2018 16:55:28 +0000 (18:55 +0200)]
tools/tar: update to 1.30

update GNU tar to 1.30

Signed-off-by: Hannu Nyman <hannu.nyman@iki.fi>
6 years agokernel: bump 4.4 to 4.4.112
Kevin Darbyshire-Bryant [Thu, 18 Jan 2018 15:01:19 +0000 (15:01 +0000)]
kernel: bump 4.4 to 4.4.112

Refresh patches.
Remove upstreamed patches:

target/linux/generic/patches-4.4/030-2-smsc75xx-use-skb_cow_head-to-deal-with-cloned-skbs.patch
target/linux/generic/patches-4.4/030-3-cx82310_eth-use-skb_cow_head-to-deal-with-cloned-skb.patch
target/linux/generic/patches-4.4/030-4-sr9700-use-skb_cow_head-to-deal-with-cloned-skbs.patch
target/linux/generic/patches-4.4/030-5-lan78xx-use-skb_cow_head-to-deal-with-cloned-skbs.patch

CVEs completely or partially addressed:

CVE-2017-5715
CVE-2017-5753
CVE-2017-17741
CVE-2017-1000410

Compile-tested: ar71xx Archer C7 v2
Run-tested: ar71xx Archer C7 v2

Signed-off-by: Kevin Darbyshire-Bryant <ldir@darbyshire-bryant.me.uk>
6 years agokernel: bump 4.9 to 4.9.77
Kevin Darbyshire-Bryant [Thu, 18 Jan 2018 13:51:13 +0000 (13:51 +0000)]
kernel: bump 4.9 to 4.9.77

Refresh patches.
Remove upstreamed patches:

target/linux/generic/backport-4.9/023-2-smsc75xx-use-skb_cow_head-to-deal-with-cloned-skbs.patch
target/linux/generic/backport-4.9/023-3-cx82310_eth-use-skb_cow_head-to-deal-with-cloned-skb.patch
target/linux/generic/backport-4.9/023-4-sr9700-use-skb_cow_head-to-deal-with-cloned-skbs.patch
target/linux/generic/backport-4.9/023-5-lan78xx-use-skb_cow_head-to-deal-with-cloned-skbs.patch

CVEs completely or partially addressed:

CVE-2017-5715
CVE-2017-5753
CVE-2017-17741
CVE-2017-1000410

Compile-tested: ar71xx Archer C7 v2
Run-tested: ar71xx Archer C7 v2

Signed-off-by: Kevin Darbyshire-Bryant <ldir@darbyshire-bryant.me.uk>
Tested-by: Koen Vandeputte <koen.vandeputte@ncentric.com>
6 years agokernel: bump 4.14 to 4.14.14
Kevin Darbyshire-Bryant [Thu, 18 Jan 2018 14:13:53 +0000 (14:13 +0000)]
kernel: bump 4.14 to 4.14.14

Refresh patches.

CVEs completely or partially addressed:

CVE-2017-5715
CVE-2017-5753
CVE-2017-17741
CVE-2017-1000410

Signed-off-by: Kevin Darbyshire-Bryant <ldir@darbyshire-bryant.me.uk>
Tested-by: Koen Vandeputte <koen.vandeputte@ncentric.com>
6 years agokernel/4.{4, 9}: add manufacturer ID for Winbond NANDs
Gabor Juhos [Thu, 18 Jan 2018 21:13:09 +0000 (22:13 +0100)]
kernel/4.{4, 9}: add manufacturer ID for Winbond NANDs

Some MikroTik devices are using a Winbond NAND flash. Linux treats
it as an unknown NAND before version 4.11:

  nand: device found, Manufacturer ID: 0xef, Chip ID: 0xf1
  nand: Unknown NAND 128MiB 3,3V 8-bit
  nand: 128 MiB, SLC, erase size: 128 KiB, page size: 2048, OOB size: 64

Backport a patch from 4.11 to show the manufacturer correctly:

  nand: device found, Manufacturer ID: 0xef, Chip ID: 0xf1
  nand: Winbond NAND 128MiB 3,3V 8-bit
  nand: 128 MiB, SLC, erase size: 128 KiB, page size: 2048, OOB size: 64

Tested on a MikroTik R951Ui-2HnD board.

Signed-off-by: Gabor Juhos <juhosg@freemail.hu>
6 years agodnsmasq: backport validation fix in dnssec security fix
Kevin Darbyshire-Bryant [Sat, 20 Jan 2018 08:46:28 +0000 (08:46 +0000)]
dnsmasq: backport validation fix in dnssec security fix

A DNSSEC validation error was introduced in the fix for CVE-2017-15107

Backport the upstream fix to the fix (a simple typo)

Signed-off-by: Kevin Darbyshire-Bryant <ldir@darbyshire-bryant.me.uk>
6 years agodnsmasq: backport dnssec security fix
Kevin Darbyshire-Bryant [Fri, 19 Jan 2018 17:16:08 +0000 (17:16 +0000)]
dnsmasq: backport dnssec security fix

CVE-2017-15107

An interesting problem has turned up in DNSSEC validation. It turns out
that NSEC records expanded from wildcards are allowed, so a domain can
include an NSEC record for *.example.org and an actual query reply could
expand that to anything in example.org  and still have it signed by the
signature for the wildcard. So, for example

!.example.org NSEC zz.example.org

is fine.

The problem is that most implementers (your author included, but also
the Google public DNS people, powerdns and Unbound) then took that
record to prove the nothing exists between !.example.org and
zz.example.org, whereas in fact it only provides that proof between
*.example.org and zz.example.org.

This gives an attacker a way to prove that anything between
!.example.org and *.example.org doesn't exists, when it may well do so.

Signed-off-by: Kevin Darbyshire-Bryant <ldir@darbyshire-bryant.me.uk>
6 years agoipq806x: remove merged ipq4019 patch
Christian Lamparter [Thu, 18 Jan 2018 12:26:08 +0000 (13:26 +0100)]
ipq806x: remove merged ipq4019 patch

The patch 0022-dts-ipq4019-support-ARMv7-PMU.patch
was merged into 4.8-rc1.

Signed-off-by: Christian Lamparter <chunkeey@gmail.com>
[refresh patches]
Signed-off-by: Mathias Kresin <dev@kresin.me>
6 years agoipq806x: remove dependency on non-existent kmod
Christian Lamparter [Wed, 17 Jan 2018 22:38:50 +0000 (23:38 +0100)]
ipq806x: remove dependency on non-existent kmod

During the integration of the ipq40xx target,
the phy drivers were included into the ipq806x's
target kernel config.

Signed-off-by: Christian Lamparter <chunkeey@gmail.com>
6 years agoipq-wifi: align AVM FRITZ!Box 4040's board-2.bin package
Christian Lamparter [Wed, 17 Jan 2018 22:38:49 +0000 (23:38 +0100)]
ipq-wifi: align AVM FRITZ!Box 4040's board-2.bin package

This patch renames the AVM FRITZ!Box 4040's board-2.bin
file and package to match the 'vendor_product' format.

Signed-off-by: Christian Lamparter <chunkeey@gmail.com>
6 years agoipq806x: overhaul AVM FRITZ!box 4040 device-tree file
Christian Lamparter [Wed, 17 Jan 2018 22:38:48 +0000 (23:38 +0100)]
ipq806x: overhaul AVM FRITZ!box 4040 device-tree file

This patch aligns the device-tree file with the latest
guidelines.
 - No longer include qcom-ipq4019-ap.dk01.1.dtsi. This
   file is only partially upstream and therefore subjected
   to changes that might not be compatible with the board.

   As a result, the definitions from the file have been
   copied into this dts.

 - exclusively use decimal GPIO addresses.

 - reorganize the reserved-memory layout to waste less
   memory. There's no point in keeping the u-boot loader
   around. This should also make it possible    to create
   an image that will boot with the original EVA/ADAM2 loader
   without needing to install the modified u-boot loader.
   And finally mark the "tz-apps" as reusable.
   There isn't a way to  upload apps to the trust-zone in OpenWrt
   yet. But it might see some use in the future as a "secure"
   key-store/TPM.

 - sort the first-level nodes alphabetically.
 - sort nodes with an address by the address.

Signed-off-by: Christian Lamparter <chunkeey@gmail.com>
6 years agofirmware: ath10k-firmware: update QCA988x firmware to 10.2.4-1.0-00033
Christian Lamparter [Thu, 21 Dec 2017 18:13:04 +0000 (19:13 +0100)]
firmware: ath10k-firmware: update QCA988x firmware to 10.2.4-1.0-00033

This patch updates ath10k-firmware to use the
firmware-5.bin_10.2.4-1.0-00033 firmware for the QCA988x.

Signed-off-by: Christian Lamparter <chunkeey@gmail.com>
6 years agofirmware: ath10k-firmware: update QCA9887 firmware to 10.2.4-1.0-00033
Christian Lamparter [Thu, 21 Dec 2017 18:13:03 +0000 (19:13 +0100)]
firmware: ath10k-firmware: update QCA9887 firmware to 10.2.4-1.0-00033

This patch updates ath10k-firmware to use the
firmware-5.bin_10.2.4-1.0-00033 firmware for the QCA9887.

Signed-off-by: Christian Lamparter <chunkeey@gmail.com>
6 years agofirmware: ath10k-firmware: update QCA9888 firmware to 10.4-3.4-00104
Christian Lamparter [Thu, 21 Dec 2017 18:13:02 +0000 (19:13 +0100)]
firmware: ath10k-firmware: update QCA9888 firmware to 10.4-3.4-00104

This patch updates ath10k-firmware to use the
firmware-5.bin_10.4-3.4-00104 firmware for the QCA9888.

Signed-off-by: Christian Lamparter <chunkeey@gmail.com>
6 years agofirmware: ath10k-firmware: update QCA9984 firmware to 10.4-3.4-00104
Christian Lamparter [Thu, 21 Dec 2017 18:13:01 +0000 (19:13 +0100)]
firmware: ath10k-firmware: update QCA9984 firmware to 10.4-3.4-00104

This patch updates ath10k-firmware to use the
firmware-5.bin_10.4-3.4-00104 firmware for the QCA9984.

Signed-off-by: Christian Lamparter <chunkeey@gmail.com>
6 years agofirmware: ath10k-firmware: update QCA4019 firmware to 10.4-3.4-00104
Christian Lamparter [Thu, 21 Dec 2017 18:13:00 +0000 (19:13 +0100)]
firmware: ath10k-firmware: update QCA4019 firmware to 10.4-3.4-00104

This patch updates ath10k-firmware to use the
firmware-5.bin_10.4-3.4-00104 firmware for the QCA4019.

Signed-off-by: Christian Lamparter <chunkeey@gmail.com>
6 years agofirmware: ath10k-firmware: update to 2017-12-20
Christian Lamparter [Thu, 21 Dec 2017 18:12:59 +0000 (19:12 +0100)]
firmware: ath10k-firmware: update to 2017-12-20

This update automatically includes a new firmware for the QCA6174:
firmware-6.bin_WLAN.RM.4.4.1-00079-QCARMSWPZ-1

Signed-off-by: Christian Lamparter <chunkeey@gmail.com>
6 years agoodhcp6c: add sendopts config support and update to latest git HEAD
Hans Dedecker [Tue, 16 Jan 2018 14:59:03 +0000 (15:59 +0100)]
odhcp6c: add sendopts config support and update to latest git HEAD

Add sendopts config support allowing to add options in sent DHCPv6 packets.

Options can be configured as follows :
uci set network.wan6.sendopts="sntpservers:3001:3001::1,3001:3001::2 11:00000000000000000000006674692F 0x3e8:ABCDEF"

Based on a patch by Frank Andrieu <fandrieu@gmail.com>

See https://git.openwrt.org/?p=project/odhcp6c.git;a=commit;h=510aaf6d528210c5e8a6159f9b80b32615e88c5f
for a more detailed description.

Latest git changes :
1f93bd4 dhcpv6: rework option passthrough logic
a477e95 odhcp6c: rework userclass and vendorclass command handling
510aaf6 odhcp6c: add -x opt:val support
ab75be1 treewide: update copyrights to 2018
f3a4609 odhcp6c: let odhcp6c_add_state return a success/failure indication

Signed-off-by: Hans Dedecker <dedeckeh@gmail.com>
6 years agoramips: tl-wr840n-v5: increase firmware partition for 4Mmtk layot
Serg Studzinskii [Sun, 24 Dec 2017 16:00:13 +0000 (18:00 +0200)]
ramips: tl-wr840n-v5: increase firmware partition for 4Mmtk layot

According to console log during TP-Link TL-WR840N v5 OEM firmware update
procedure 0x3e0000-0x3f0000 64kB "config" partition, which is used to store
router's configuration settings, is erased and recreated again during every
OEM firmware update procedure, thus does not contain any valuable factory data.
So it is conviniant to use this extra 64kB erase block for jffs overlay due
limited flash size on this device like it used on TP-Link's ar71xx boards.

Signed-off-by: Serg Studzinskii <serguzhg@gmail.com>
6 years agotools/mpfr: update to 4.0.0
Hannu Nyman [Wed, 17 Jan 2018 19:06:19 +0000 (21:06 +0200)]
tools/mpfr: update to 4.0.0

Update mpfr to 4.0.0

Signed-off-by: Hannu Nyman <hannu.nyman@iki.fi>
6 years agotools/mpc: Update to 1.1.0
Hannu Nyman [Wed, 17 Jan 2018 19:01:35 +0000 (21:01 +0200)]
tools/mpc: Update to 1.1.0

Update mpc to 1.1.0

(note: adds compatibility with mpfr 4.0.0)

Signed-off-by: Hannu Nyman <hannu.nyman@iki.fi>
6 years agoutil-linux: add fstrim support
Dirk Brenken [Fri, 12 Jan 2018 11:57:39 +0000 (12:57 +0100)]
util-linux: add fstrim support

This PR adds optional fstrim support

Signed-off-by: Dirk Brenken <dev@brenken.org>
6 years agokirkwood: fix sysupgrade
尤晓杰 [Wed, 17 Jan 2018 13:03:50 +0000 (21:03 +0800)]
kirkwood: fix sysupgrade

The platform_check_image() stub need to return 0 for success, otherwise
the sysupgrade will fail with:

  Image check 'platform_check_image' failed.

Fixes: aa6f5f1787a6 ("kirkwood: use image metadata")
Signed-off-by: 尤晓杰 <yxj790222@163.com>
[reworded commit message}
Signed-off-by: Mathias Kresin <dev@kresin.me>
6 years agoipq806x: fix ZyXEL NBG6817 sysupgrade
Mathias Kresin [Wed, 17 Jan 2018 19:40:55 +0000 (20:40 +0100)]
ipq806x: fix ZyXEL NBG6817 sysupgrade

Use the first found directory in the tar archive instead of relying on
a directory named according to the userspace boardname.

It allows to change the boardname without adding another compatibility
layer to the zyxel sysupgrade.

Fixes: 33f09cf15178 ("ipq806x: convert to dt-based board-detection")
Signed-off-by: Mathias Kresin <dev@kresin.me>
6 years agobase-files: gpio switch: check if direction can be set
Mathias Kresin [Mon, 15 Jan 2018 00:23:36 +0000 (01:23 +0100)]
base-files: gpio switch: check if direction can be set

Obviously not all GPIO controller allow to change the direction. The issue
is around since the beginning of the script but only due to the recent
changes error messages are more visible.

Add a check if a change of the direction is supported by the GPIO
controller and fallback to setting only the value if not.

Fixes: FS#1271
Signed-off-by: Mathias Kresin <dev@kresin.me>
6 years agobase-files: gpio switch: fix inverted logic
Mathias Kresin [Sun, 14 Jan 2018 23:28:03 +0000 (00:28 +0100)]
base-files: gpio switch: fix inverted logic

GPIOs are exported as active high to the sysfs, hence the logic need to be
inverted.

Fixes: e66c47fb14f5 ("base-files: gpio switch: set output value with
       direction")
Signed-off-by: Mathias Kresin <dev@kresin.me>
6 years agoubox: update to latest git HEAD
Hans Dedecker [Wed, 17 Jan 2018 13:23:13 +0000 (14:23 +0100)]
ubox: update to latest git HEAD

e7a63fb ubox: Remove unnecessary memset calls

Signed-off-by: Hans Dedecker <dedeckeh@gmail.com>
6 years agoubus: fix wrong PKG_SOURCE_DATE
Hans Dedecker [Wed, 17 Jan 2018 13:08:53 +0000 (14:08 +0100)]
ubus: fix wrong PKG_SOURCE_DATE

Fix wrong PKG_SOURCE_DATE introduced in e14cac0

Signed-off-by: Hans Dedecker <dedeckeh@gmail.com>
6 years agoubus: update to latest git HEAD
Hans Dedecker [Wed, 17 Jan 2018 12:45:02 +0000 (13:45 +0100)]
ubus: update to latest git HEAD

5bae22e ubus/lua: pass notification name to callback
212ceb1 valgrind complained about these
d57907c fix invalid close() call

Signed-off-by: Hans Dedecker <dedeckeh@gmail.com>
6 years agoath9k: discard undersized packets
Felix Fietkau [Wed, 17 Jan 2018 11:32:22 +0000 (12:32 +0100)]
ath9k: discard undersized packets

Sometimes the hardware will push small packets that trigger a WARN_ON
in mac80211. Discard them early to avoid this issue.

Reported-by: Stijn Tintel <stijn@linux-ipv6.be>
Signed-off-by: Felix Fietkau <nbd@nbd.name>
6 years agobuild: remove use of STAGING_DIR_HOST/usr (fixes cmake build error on macOS)
Felix Fietkau [Wed, 17 Jan 2018 10:53:10 +0000 (11:53 +0100)]
build: remove use of STAGING_DIR_HOST/usr (fixes cmake build error on macOS)

This directory is never created, nor is it used anywhere. Using it in
HOST_LDFLAGS causes a linker warning to be emitted on macOS, which
messes with cmake configure tests.

Signed-off-by: Felix Fietkau <nbd@nbd.name>
6 years ago.gitignore: add .project & .cproject for eclipse users
Catrinel Catrinescu [Thu, 11 Jan 2018 09:35:46 +0000 (10:35 +0100)]
.gitignore: add .project & .cproject for eclipse users

For eclipse users, .project is always created inside the project.
For users of CDT within Eclipse, add also .cproject, thanks Dirk
Feytons.

Signed-off-by: Catrinel Catrinescu <cc@80211.de>
6 years agoar71xx: drop mdio-gpio.h include from mach-om5pacv2.c
Gabor Juhos [Tue, 9 Jan 2018 13:59:14 +0000 (14:59 +0100)]
ar71xx: drop mdio-gpio.h include from mach-om5pacv2.c

The setup code of the OpenMesh OM5P ACv2 device uses nothing
from the mdio-gpio.h header, so remove the inclusion of that.

Also remove the kernel version check which is needed only
because the mdio-gpio.h header exists in a different path
in older kernels.

Compile tested only.

Cc: Sven Eckelmann <sven.eckelmann@openmesh.com>
Signed-off-by: Gabor Juhos <juhosg@freemail.hu>
6 years agocns3xxx: switch to kernel 4.14
Koen Vandeputte [Thu, 11 Jan 2018 15:04:40 +0000 (16:04 +0100)]
cns3xxx: switch to kernel 4.14

Signed-off-by: Koen Vandeputte <koen.vandeputte@ncentric.com>
6 years agocns3xxx: refresh patches
Koen Vandeputte [Thu, 11 Jan 2018 15:04:39 +0000 (16:04 +0100)]
cns3xxx: refresh patches

Signed-off-by: Koen Vandeputte <koen.vandeputte@ncentric.com>
6 years agocns3xxx: remove deprecated uart flags
Koen Vandeputte [Thu, 11 Jan 2018 15:04:38 +0000 (16:04 +0100)]
cns3xxx: remove deprecated uart flags

Flag UPF_NO_TXEN_TEST is deprecated.

Remove it.

Signed-off-by: Koen Vandeputte <koen.vandeputte@ncentric.com>
6 years agocns3xxx: fix ethernet IRQ imbalance in 4.14
Koen Vandeputte [Thu, 11 Jan 2018 15:04:37 +0000 (16:04 +0100)]
cns3xxx: fix ethernet IRQ imbalance in 4.14

Kernel 4.14 contains more warnings regarding NAPI usage, which showed
following warning:

[ 4016.420000] ------------[ cut here ]------------
[ 4016.430000] WARNING: CPU: 1 PID: 14 at kernel/irq/manage.c:525
__enable_irq+0x3c/0x78
[ 4016.440000] Unbalanced enable for IRQ 51
[ 4016.440000] Modules linked in: ath9k ath9k_common pppoe ppp_async
ath9k_hw ath pppox ppp_generic nf_conntrack_ipv6 mac80211 iptable_nat
ipt_REJECT ipt_MASQUERADE cfg80211 xt_time xt_tcpudp xt_state xt_nat
xt_multiport xt_mark xt_mac xt_limit xt_conntrack xt_comment xt_TCPMSS
xt_REDIRECT xt_LOG slhc nf_reject_ipv4 nf_nat_redirs
[ 4016.500000] CPU: 1 PID: 14 Comm: ksoftirqd/1 Tainted: G        W
       4.14.11 #0
[ 4016.500000] Hardware name: Gateworks Corporation Laguna Platform
[ 4016.510000] Backtrace:
[ 4016.510000] [<c010a28c>] (dump_backtrace) from [<c010a56c>]
(show_stack+0x18/0x1c)
[ 4016.520000]  r7:c059df4b r6:00000000 r5:60000093 r4:00000000
[ 4016.530000] [<c010a554>] (show_stack) from [<c048b04c>]
(dump_stack+0x94/0xb4)
[ 4016.530000] [<c048afb8>] (dump_stack) from [<c011760c>]
(__warn+0xc8/0xf8)
[ 4016.540000]  r7:c059df4b r6:c015144c r5:00000000 r4:cf863e08
[ 4016.550000] [<c0117544>] (__warn) from [<c011767c>]
(warn_slowpath_fmt+0x40/0x48)
[ 4016.550000]  r9:cf863eb0 r8:00000036 r7:00000000 r6:c0636360
r5:00000033 r4:cf80a500
[ 4016.560000] [<c0117640>] (warn_slowpath_fmt) from [<c015144c>]
(__enable_irq+0x3c/0x78)
[ 4016.570000]  r3:00000033 r2:c059e0c5
[ 4016.570000] [<c0151410>] (__enable_irq) from [<c01514ec>]
(enable_irq+0x64/0x7c)
[ 4016.580000] [<c0151488>] (enable_irq) from [<c0335c14>]
(eth_poll+0x28c/0x558)
[ 4016.590000]  r5:d08926c0 r4:cf894488
[ 4016.590000] [<c0335988>] (eth_poll) from [<c03b1718>]
(net_rx_action+0xfc/0x2e8)
[ 4016.600000]  r10:cf863eb8 r9:cf863eb0 r8:00000001 r7:0f7b5000
r6:c0636360 r5:cfdeb360
[ 4016.610000]  r4:cf894488
[ 4016.610000] [<c03b161c>] (net_rx_action) from [<c0101510>]
(__do_softirq+0xe0/0x228)
[ 4016.620000]  r10:00000100 r9:c070204c r8:c0702040 r7:00000003
r6:00000008 r5:40000003
[ 4016.620000]  r4:cf862000
[ 4016.630000] [<c0101430>] (__do_softirq) from [<c011ae28>]
(run_ksoftirqd+0x34/0x58)
[ 4016.630000]  r10:c0133964 r9:00000000 r8:00000001 r7:00000000
r6:c070b44c r5:cf862000
[ 4016.640000]  r4:cf8037e0
[ 4016.640000] [<c011adf4>] (run_ksoftirqd) from [<c0133adc>]
(smpboot_thread_fn+0x178/0x190)
[ 4016.650000] [<c0133964>] (smpboot_thread_fn) from [<c013099c>]
(kthread+0x11c/0x138)
[ 4016.660000]  r9:cf8037e0 r8:cf802cdc r7:cf839df8 r6:cf803800
r5:00000000 r4:cf802cc0
[ 4016.670000] [<c0130880>] (kthread) from [<c0107218>]
(ret_from_fork+0x14/0x3c)
[ 4016.680000]  r10:00000000 r9:00000000 r8:00000000 r7:00000000
r6:00000000 r5:c0130880
[ 4016.680000]  r4:cf803800 r3:cf862000
[ 4016.690000] ---[ end trace 51114df08f429115 ]---

This is triggered by calling eth_schedule_poll(sw) after the IRQ has
been re-enabled.

Rework the network code to only enable IRQ's again if NAPI agrees it's
safe to do so.

Also only re-enable IRQ *after* cleaning up the RX ring and re-enabling
DMA, which otherwise resulted in ugly warnings regarding dirty page
fragments.
These popped up nearly immediately when building the kernel with O2 iso
Os.

--> Note that this change fixes stability issues, at the cost of ~8%
throughput performance.

While at it, also change the iface name being used in warning prints,
making it more obvious: "switch%d" --> "cns3xxx_eth"

The changes have been tested on 4 boards, each moving ~30TB of data

Signed-off-by: Koen Vandeputte <koen.vandeputte@ncentric.com>
6 years agocns3xxx: update patches for 4.14
Koen Vandeputte [Thu, 11 Jan 2018 15:04:36 +0000 (16:04 +0100)]
cns3xxx: update patches for 4.14

Signed-off-by: Koen Vandeputte <koen.vandeputte@ncentric.com>
6 years agocns3xxx: refresh kernel 4.14 config
Koen Vandeputte [Thu, 11 Jan 2018 15:04:35 +0000 (16:04 +0100)]
cns3xxx: refresh kernel 4.14 config

Signed-off-by: Koen Vandeputte <koen.vandeputte@ncentric.com>
6 years agocns3xxx: copy patches for kernel 4.14
Koen Vandeputte [Thu, 11 Jan 2018 15:04:34 +0000 (16:04 +0100)]
cns3xxx: copy patches for kernel 4.14

Signed-off-by: Koen Vandeputte <koen.vandeputte@ncentric.com>
6 years agocns3xxx: copy config for kernel 4.14
Koen Vandeputte [Thu, 11 Jan 2018 15:04:33 +0000 (16:04 +0100)]
cns3xxx: copy config for kernel 4.14

Signed-off-by: Koen Vandeputte <koen.vandeputte@ncentric.com>
6 years agoconfig: don't define the same symbol twice
Koen Vandeputte [Sat, 13 Jan 2018 20:35:48 +0000 (21:35 +0100)]
config: don't define the same symbol twice

In commit fce35bce0f6e ("config: support new symbol intro'd in kernel
4.12")
I forgot to remove the initial debug test line.

This clearly is wrong as the same symbol is defined conditionally in the
line below as it should be.

I looked over it as I just checked if the symbol was present now upon
testing it.

Fixes: fce35bce0f6e ("config: support new symbol intro'd in kernel
4.12")

Signed-off-by: Koen Vandeputte <koen.vandeputte@ncentric.com>
6 years agoar71xx: fix format of the 701-MIPS-ath79-add-routerboard-detection patches
Gabor Juhos [Mon, 15 Jan 2018 22:04:53 +0000 (23:04 +0100)]
ar71xx: fix format of the 701-MIPS-ath79-add-routerboard-detection patches

Commit e15c63a37574 ("ar71xx: add support for MikroTik RouterBOARD wAP
G-5HacT2HnD (wAP AC") broke the format of the patches. In unified diff
format, the unchanged, contextual lines must be preceded by a space
character.

Refresh the patches with quilt to fix them.

Fixes: e15c63a37574 ("ar71xx: add support for MikroTik RouterBOARD wAPG-5HacT2HnD (wAP AC")
Signed-off-by: Gabor Juhos <juhosg@freemail.hu>
6 years agoar71xx: fix format of the 741-MIPS-ath79-add-PCI-for-QCA9556-SoC patches
Gabor Juhos [Mon, 15 Jan 2018 22:04:37 +0000 (23:04 +0100)]
ar71xx: fix format of the 741-MIPS-ath79-add-PCI-for-QCA9556-SoC patches

The patches introduced in commit 20e68f6d395d ("ar71xx: kernel: enable
PCI on QCA9556 SoC") have non standard format. In unified diff format,
the unchanged, contextual lines must be preceded by a space character.

Refresh the patches with quilt to fix them.

Fixes: 20e68f6d395d ("ar71xx: kernel: enable PCI on QCA9556 SoC")
Signed-off-by: Gabor Juhos <juhosg@freemail.hu>
6 years agoar71xx: mach-rbspi: simplify reset button registration
Gabor Juhos [Mon, 15 Jan 2018 19:50:43 +0000 (20:50 +0100)]
ar71xx: mach-rbspi: simplify reset button registration

The setup code defines four individual structures for the
Reset buttons of the supported boards. The only difference
between the definitions is the GPIO number used for the
button.

Replace the different structures with one generic variant,
and add a helper function to simplify the button registration.

Signed-off-by: Gabor Juhos <juhosg@freemail.hu>
6 years agoimx6: add support for the intel gbit driver on GW5520
Koen Vandeputte [Sat, 13 Jan 2018 20:41:59 +0000 (21:41 +0100)]
imx6: add support for the intel gbit driver on GW5520

The Gateworks GW5520 board uses 2x intel gigabit mac's, instead of the
internal i.MX6 FEC.

Add support for these.

Signed-off-by: Scott V. Kamp <outbackdingo@gmail.com>
Signed-off-by: Koen Vandeputte <koen.vandeputte@ncentric.com>
6 years agomac80211: remove support code for authsae
Felix Fietkau [Fri, 15 Dec 2017 13:46:42 +0000 (14:46 +0100)]
mac80211: remove support code for authsae

Signed-off-by: Felix Fietkau <nbd@nbd.name>
6 years agoauthsae: remove package
Felix Fietkau [Fri, 15 Dec 2017 13:45:44 +0000 (14:45 +0100)]
authsae: remove package

It is no longer actively maintained and does not work well in many
configurations. Fully replaced by wpad-mesh

Signed-off-by: Felix Fietkau <nbd@nbd.name>
6 years agoipq8065: adjust SS USB PHY power settings
Pavel Kubelun [Wed, 5 Jul 2017 16:23:23 +0000 (19:23 +0300)]
ipq8065: adjust SS USB PHY power settings

According to QSDK ipq8065 has different SS USB PHY power settings.

Adjust it according to QSDK settings.

It also corresponds to oem Netgear R7800 tarball for soc ver >= 2.
https://github.com/paul-chambers/netgear-r7800/blob/eeac2e10190f6f45e32e4c7012c4babc351898d8/git_home/linux.git/sourcecode/arch/arm/mach-msm/board-ipq806x.c#L2494

Signed-off-by: Pavel Kubelun <be.dissent@gmail.com>
6 years agoipq806x: usb: add phy 1 master reset control
Pavel Kubelun [Thu, 15 Jun 2017 10:58:34 +0000 (13:58 +0300)]
ipq806x: usb: add phy 1 master reset control

In current state only phy 0 master reset is controlled.
Add the phy 1 reset.

Signed-off-by: Pavel Kubelun <be.dissent@gmail.com>
6 years agoipq806x: usb: Control USB master reset
Pavel Kubelun [Thu, 15 Jun 2017 10:00:18 +0000 (13:00 +0300)]
ipq806x: usb: Control USB master reset

Picking commit from QSDK
https://source.codeaurora.org/quic/qsdk/oss/kernel/linux-msm/commit/?h=eggplant&id=a86bda9f8a7965f0cedd347a9c04800eb9f41ea3

Commit message:
"During removal of the glue layer(dwc3-of-simple), USB master reset is set to active and during insertion it is de-activated."

Change-Id: I537dc810f6cb2a46664ee674840145066432b957
Signed-off-by: Vasudevan Murugesan <vmuruges@codeaurora.org>
(cherry picked from commit 4611e13580a216812f85f0801b95442d02eeb836)"

Signed-off-by: Pavel Kubelun <be.dissent@gmail.com>
6 years agoipq806x: dwc3: Fix power_on and power_off sequence
Pavel Kubelun [Thu, 15 Jun 2017 08:48:17 +0000 (11:48 +0300)]
ipq806x: dwc3: Fix power_on and power_off sequence

Picking commit from QSDK
https://source.codeaurora.org/quic/qsdk/oss/kernel/linux-msm/commit/drivers/phy/phy-qcom-dwc3.c?h=eggplant&id=d316437c9cdb70023a760342678f32e27241725a

The commit fixes:
- dwc3 phy module unloading
- possibly fixes FS#177 when some devices were improperly enumerated as HS while being SS thus stucking the driver during boot.

Signed-off-by: Pavel Kubelun <be.dissent@gmail.com>
6 years agoipq806x: Update HSIO recommended usb phy settings
Pavel Kubelun [Thu, 15 Jun 2017 08:35:26 +0000 (11:35 +0300)]
ipq806x: Update HSIO recommended usb phy settings

Picking commit from QSDK
https://source.codeaurora.org/quic/qsdk/oss/kernel/linux-msm/commit/drivers/phy/phy-qcom-dwc3.c?h=eggplant&id=cf82fdf4bdd081cd81bb081f7815b915bc8bb851

The comit adjusts USB dwc3 phy default values as per QSDK recomendation and allows to set it through DT.

Commit message:
"SoC version based values will be recovered from the device node.
If device node does not have such values, defaults are applied.

Change-Id: Ia77b5b7fe95ce1a433885df1761091bced98d989
Signed-off-by: Gokul Sriram Palanisamy <gpalan@codeaurora.org>"
Signed-off-by: Pavel Kubelun <be.dissent@gmail.com>
6 years agoRevert "ipq8064: Fix dwc3 module unloading"
Pavel Kubelun [Thu, 15 Jun 2017 07:45:17 +0000 (03:45 -0400)]
Revert "ipq8064: Fix dwc3 module unloading"

Preparing for proper fixes thus reverting commits:
8db079a9ff1756059250b801617a20baba214684 "ipq8064: Fix dwc3 module unloading"
c75f059b0c4d09dd0da60e14c4933a9f645266d1 "ipq8064: Fix dwc3 module unloading"

Signed-off-by: Pavel Kubelun <be.dissent@gmail.com>
6 years agoipq806x: update clk and cpufreq drivers
Pavel Kubelun [Sat, 9 Dec 2017 17:15:05 +0000 (20:15 +0300)]
ipq806x: update clk and cpufreq drivers

A newer clk and cpufreq drivers for ipq806x platform had been sent upstream.

A change that i have noticed is that now it's possible to set min, cur and max frequencies from sysfs (previously it was bugged and caused nothing).

Following patches are removed:
 - 0036-clk-Avoid-sending-high-rates-to-downstream-clocks-du.patch - seems it was dropped from the patchset by current committer.
 - 0044-clk-qcom-krait-Remove-CLK_IS_ROOT.patch - already applied to the driver itself in the corresponding patch.
 - 0057-clk-qcom-Add-regmap-mux-div-clocks-support.patch - seem to be irrelevant to ipq806x.

Signed-off-by: Pavel Kubelun <be.dissent@gmail.com>
6 years agoipq806x: force 2nd pci slot into gen1 mode
Pavel Kubelun [Thu, 6 Jul 2017 14:20:18 +0000 (17:20 +0300)]
ipq806x: force 2nd pci slot into gen1 mode

According to QSDK and OEM tarballs (checked c2600, r7500v2, r7800) 2nd pci slot (pci1, 2,4 GHz card)) on ap148 based boards should operate in gen1 mode.
EA8500 is an exception and according to GPL pcie0 should operate in gen1 mode.

In previous commit we've added the support for this option, so enable it in DT for affected devices.

QSDK ref:
https://source.codeaurora.org/quic/qsdk/oss/kernel/linux-msm/commit/?h=release/endive_preview_cc&id=f3b07fe309027c52fc163149500cedddd707c506

While at it move the phy transmit termination offset value into dtsi file as it's platform specific.

Signed-off-by: Pavel Kubelun <be.dissent@gmail.com>