commit r30917 ("kernel: bypass all netfilter hooks if the sysctls for that
functionality have been disabled - eliminates the overhead of enabling
CONFIG_BRIDGE_NETFILTER in the kernel config") introduced an optimization
which should reduce/eliminate the overhead for traffic send over bridges on
kernels compiled with CONFIG_BRIDGE_NETFILTER=y. But this optimization
breaks the nf_call_iptables per bridge setting which is more fine grained
than the global sysctl net.bridge.bridge-nf-call-iptables setting.
A test reflecting a real world setup was created to identify if this really
eliminates the overhead and if per-bridge nf_call_iptables could be used in
some setups to increase the throughput. A Qualcomm Atheros QCA9558 based
system with one ethernet and an ath9k wifi 3x3 in HT40 mode was used.
Cables from the AP to the wifi station were used to reduce interference
problems during the tests.
The wlan interface was put in one bridge interface called br-wlan. This
bridge usually contains some more wlan interfaces. The eth0 was put in a
second bridge called br-lan. This usually contains some other privileged
wlan or mesh interfaces. Routing was added between br-lan and br-wlan.
Three kernels were tested:
* (default) OpenWrt kernel for this device
* (brfilter-global) OpenWrt kernel with CONFIG_BRIDGE_NETFILTER=y
* (brfilter-local) OpenWrt kernel with CONFIG_BRIDGE_NETFILTER=y and
without 644-bridge_optimize_netfilter_hooks.patch
The changes to the the netfilter settings of the bridge were done via:
* (brfilter-global) /sbin/sysctl -w net.bridge.bridge-nf-call-iptables=1
* (brfilter-lobal) echo 1 > /sys/class/net/br-lan/bridge/nf_call_iptables
and/or echo 1 > /sys/class/net/br-wan/bridge/nf_call_iptables
A station connected to the wlan0 (AP) interface was used to send traffic to
a PC connected via ethernet. iperf with 3 concurrent transmissions was used
to generate the traffic.
| kernel | br-nf-* global | nf-call* iface | download | upload |
|-----------------|----------------|----------------|----------|----------|
| default | 0 | - | 209 | 268 |
| brfilter-global | 0 | - | 185 | 243 |
| brfilter-local | 0 | - | 187 | 243 |
| brfilter-local | 0 | br-lan | 157 | 226 |
| brfilter-local | 0 | br-lan br-wlan | 139 | 161 |
| brfilter-global | 1 | - | 136 | 162 |
Download/upload results in Mibit/s
It can be seen that the patch doesn't eliminate the overhead. It can also
be seen that the throughput of brfilter-global and brfilter-local with
disabled filtering is the roughly the same. Also the throughput for
brfilter-global and brfilter-local for enabled filtering on all bridges is
roughly the same.
But also the brfilter-local throughput is higher when only br-lan requires
the filtering. This setting would not be possible with
644-bridge_optimize_netfilter_hooks.patch applied and thus can only be
compared with brfilter-global and filtering enabled for all interfaces.
Signed-off-by: Sven Eckelmann <sven@open-mesh.com>
SVN-Revision: 46835
It was corrupted in r38528. The most obvious symptom is repeated messages like this:
Tue Sep 8 08:25:18 2015 kern.warn kernel: [77141.972226] br-lan: received packet on wlan0 with own address as source address
Signed-off-by: Dmitry Ivanov <dima@ubnt.com>
SVN-Revision: 46821
Everything except for blkcipher was already built-in, so make blkcipher
built-in as well.
Signed-off-by: Felix Fietkau <nbd@openwrt.org>
SVN-Revision: 46820
Add support for the Netgear Nighthawk X4 R7500 and build
appropariate sysupgrade and factory images.
Known issues:
* 5 GHz wifi not working - there is no quantenna driver
* One of the USB ports is not working
Signed-off-by: Jonas Gorski <jogo@openwrt.org>
SVN-Revision: 46796
To use gpio leds as ide leds, we need to enable the trigger to be
included in the kernel.
Signed-off-by: Jonas Gorski <jogo@openwrt.org>
SVN-Revision: 46795
Add full ubi and sysupgrade images for AP148 and add sysupgrade support
for ipq806x to allow updating the current installation.
Signed-off-by: Jonas Gorski <jogo@openwrt.org>
SVN-Revision: 46789
Currently, multicast packets from an STA are sent to any according
multicast listener directly through the bridge multicast-to-unicast
feature. Unfortunately, so far this includes the originating STA, too,
resulting in multicast packets being echo'ed back to the originating STA
if it itself is a multicast listener for that group.
This behaviour breaks IPv6 duplicate address detection: An IPv6 Neighbor
Solicitation for IPv6 Duplicate Address Detection is being echo'ed back,
resulting in the host falsely detecting an address collision, which
makes the node unable to claim an IPv6 address and use IPv6 in general.
Mac80211 unfortunately only prevents the echoes for us for multicast
frames. For the multicast frames cast to a unicast destination we'll
need to take care of excluding the originator ourselves.
Signed-off-by: Linus Lüssing <linus.luessing@c0d3.blue>
SVN-Revision: 46765
Support for lantiq_svip_be has been removed a while ago, so EASY33016
images weren't buildable anymore. Remove the recipes as well as gzip
compressed kernel support, as EASY33016 was the last user of it.
Signed-off-by: Jonas Gorski <jogo@openwrt.org>
SVN-Revision: 46760
We register all gpio buttons and leds through DT, so no need to keep
fixes/additions for the platform data based bay.
Signed-off-by: Jonas Gorski <jogo@openwrt.org>
SVN-Revision: 46753
Add support for Comtrend VR-3026e v1.
The device is almost identical to the Comtrend VR-3025un.
Signed-off-by: Martin Tesar <tesarmar@gmail.com>
Signed-off-by: Jonas Gorski <jogo@openwrt.org>
SVN-Revision: 46752
/etc/uci-defaults/02_network had a typo, making it generate the wrong
network config.
Closes#20407.
Signed-off-by: Jonas Gorski <jogo@openwrt.org>
SVN-Revision: 46727
Add an upstream fix for /proc/net/route causing missing routes doing
several continued reads from it.
Only 4.1+ is affected.
Closes#20403.
Signed-off-by: Jonas Gorski <jogo@openwrt.org>
SVN-Revision: 46726
should improve flash access times. Should be harmless to gnerally
enable regardless if a flash supporting dual reads is attached. In
doubt, spi-nor will just fall back to serial reads.
Signed-off-by: Jonas Gorski <jogo@openwrt.org>
SVN-Revision: 46725
A call to pskb_may_pull() might reallocate skb->data. Therefore we
should only assign the src-pointer after any potential reallocations.
Signed-off-by: Linus Lüssing <linus.luessing@c0d3.blue>
Signed-off-by: Felix Fietkau <nbd@openwrt.org>
SVN-Revision: 46721
This reverts commit a080e8e1943156168913d0353a2e99d1151102aa.
It did not fix the problem but just hid some symptom. The real issue was
that IGMP/MLD report suppression was not considered for the
multicast-to-unicast feature. A recent netifd which isolates IGMP/MLD
reports between STAs by utilizing AP-isolation and bridge-hairpinning
should have fixed this.
It is perfectly fine to apply multicast-to-unicast to IPv6 Neighbor
Solicitations, too (once that feature is configured correctly).
Signed-off-by: Linus Lüssing <linus.luessing@c0d3.blue>
Signed-off-by: Felix Fietkau <nbd@openwrt.org>
SVN-Revision: 46720
Remove 131-MIPS-export-get_c0_perfcount_int.patch which was already applied
in 4.1.6. This fixes the following build error:
arch/mips/ath79/setup.c:217:77: error: redefinition of '__kstrtab_get_c0_perfcount_int'
arch/mips/ath79/setup.c:211:77: note: previous definition of '__kstrtab_get_c0_perfcount_int' was here
arch/mips/ath79/setup.c:217:350: error: redefinition of '__ksymtab_get_c0_perfcount_int'
arch/mips/ath79/setup.c:211:350: note: previous definition of '__ksymtab_get_c0_perfcount_int' was here
scripts/Makefile.build:258: recipe for target 'arch/mips/ath79/setup.o' failed
Reported-by: swalker
Signed-off-by: Jonas Gorski <jogo@openwrt.org>
SVN-Revision: 46715
Properly treat -ENOSYS as no PHY, else ehci-orion won't work without
generic phy support.
Signed-off-by: Jonas Gorski <jogo@openwrt.org>
SVN-Revision: 46711
Some CFEs seem to misconfigure the mapped memory flash access with
fast read but without a dummy byte, causing all accesses to be prefixed
with 0xff.
This of course breaks reading out the nvram, so do not just move back to
single i/o accessors, but also ensure that the dummy byte is correctly
set.
Signed-off-by: Jonas Gorski <jogo@openwrt.org>
SVN-Revision: 46707
Now that the switch driver can handle two devices with
the same MAC address in separate VLANs we can go back
to using the same address on both interfaces.
This is the Linksys firmware's default behavior.
Signed-off-by: Claudio Leite <leitec@staticky.com>
Signed-off-by: Imre Kaloz <kaloz@openwrt.org>
SVN-Revision: 46700
This also clears any bootloader-set FDB defaults. This had
caused issues creating port-based VLANs when mappings
overlapped previous VLANs. Packets destined to a port
not in the default port group flooded all ports.
Tested on a 88E6171 (Linksys EA4500) and 88E6172 ('1900AC)
Signed-off-by: Claudio Leite <leitec@staticky.com>
Signed-off-by: Imre Kaloz <kaloz@openwrt.org>
SVN-Revision: 46699
The u-boot boot counter was never reset after a successful boot,
which sometimes could make some variables become out of sync.
This patch adds support for the boot counter and enables
auto_recovery unconditionally.
Signed-off-by: Jonas Gorski <jogo@openwrt.org>
Signed-off-by: Rob Mosher <nyt-openwrt@countercultured.net>
Signed-off-by: Imre Kaloz <kaloz@openwrt.org>
SVN-Revision: 46690
The patch submitted in [46649] was mangled in the use of gmails webmail interface, tabs replaced with spaces, resulting in a patch which dit not apply.
This should fix the issue, sorry for the noise.
Signed-off-by: Ben Whitten <benwhitten@gmail.com>
SVN-Revision: 46676
no-op since 2.6.35
removed in Kernel 4.1
see https://lwn.net/Articles/380931/
Signed-off-by: Dirk Neukirchen <dirkneukirchen@web.de>
SVN-Revision: 46672
no-op since 2.6.35
removed in Kernel 4.1
see https://lwn.net/Articles/380931/
Signed-off-by: Dirk Neukirchen <dirkneukirchen@web.de>
SVN-Revision: 46671
The BT Home Hub 5A uses three PEF7071 with PHY ID 0xd565a401. Daniel's
PHY driver (for his u-boot sources) already supports that PHY because
it uses a PHY ID mask of 0xfffffff8.
Signed-off-by: Martin Blumenstingl <martin.blumenstingl@googlemail.com>
SVN-Revision: 46670
Instead of using board name provided explicitly as text in LED name, ex.:
[...]
mlwg2)
status_led="mlwg2:blue:system"
[...]
use $board variable, which allows to combine together multiple boards with same color and LED names, ex:
[...]
mlw221|\
mlwg2)
status_led="$board:blue:system"
[...]
The above approach allows to shrink size of code in base-files/etc/board.d/01_leds and base-files/etc/diag.sh scripts dramatically.
One thing to keep in mind here is that we assume to use proper and consistent LED naming scheme ("device:color:led-name").
Signed-off-by: Piotr Dymacz <pepe2k@gmail.com>
SVN-Revision: 46665
The upstream LED naming convention is "device:color:led-name", but it seems that many of supported boards in OpenWrt don't follow this approach.
The following patch fixes this inconsistency in dts{,i} files and updates base-files scripts for ramips target:
* fixes wrong indentation
* keeps case statements structure in same convention as in other scripts (no empty line after ";;", no indentation for case...esac body)
* fixes wrong LED names for some of boards (makes them the same as in dts{,i} files)
* combines boards with same configuration (ex. set_wifi_led "rt2800pci-phy0::radio" in 01_leds)
Signed-off-by: Piotr Dymacz <pepe2k@gmail.com>
SVN-Revision: 46664
Signed-off-by: Weijie Gao <hackpascal@gmail.com>
This patch adds support for TP-Link TL-WDR6500 v2.
The firmware has a U-Boot header for kernel, and a TP-LINK v2 header for
the whole firmware, so I have to create a new firmware creation method.
SVN-Revision: 46663
Signed-off-by: Weijie Gao <hackpascal@gmail.com>
This patch adds flash layout parser for TP-Link firmwares which have a 64kb
bootloader.
This is used for TP-Link TL-WDR6500 v2.
SVN-Revision: 46662
*Enable SMEM MTD parser and its dependencies (SMEM & HW spinlocks) in
the kernel config
*Replaces the MTD layout in DT by the dynamic layout provided by the
SMEM parser for AP148
Using the OF based parser is still possible on platforms which have a
fixed MTD partition layout.
Signed-off-by: Mathieu Olivari <mathieu@codeaurora.org>
SVN-Revision: 46658
This patch adds a new parser which uses the SMEM available on IPQ and
some other QCOM platforms to map the MTD partitions.
Signed-off-by: Mathieu Olivari <mathieu@codeaurora.org>
SVN-Revision: 46657
2 patches are cherry-picked from the following LKML thread:
*https://lkml.org/lkml/2015/4/11/208
The last patch (036-soc-qcom-add-smem-to-IPQ806x-platforms.patch) is
adding the corresponding DT nodes required for IPQ806x.
Signed-off-by: Mathieu Olivari <mathieu@codeaurora.org>
SVN-Revision: 46656
This change cherry-picks the following 3 changes from linux-next:
*fb7737 hwspinlock/core: add device tree support
*19a0f6 hwspinlock: qcom: Add support for Qualcomm HW Mutex block
*bd5717 hwspinlock: qcom: Correct msb in regmap_field
We're also adding a patch to add the hardware spinlock device nodes on
IPQ806x platforms (033-soc-qcom-Add-sfbp-device-to-IPQ806x-dts.patch).
Signed-off-by: Mathieu Olivari <mathieu@codeaurora.org>
SVN-Revision: 46655
The "linux,part-probe" dts parsing is a pretty neat generic feature.
It has been posted to kernel.org and could easily be reused by all
targets.
This change moves the patch to the 3.18 and 4.1 generic folders, and
makes the feature available to all platforms who may want to use it.
Signed-off-by: Mathieu Olivari <mathieu@codeaurora.org>
Acked-by: Hauke Mehrtens <hauke@hauke-m.de>
SVN-Revision: 46654
This patch is backport of a fix to a USB prototype mismatch bug found
in 3.18 kernels, originally submitted by Boris Brezillon [1].
The symptom of this bug was that devices attached to the at91 using
at91_ohci on a hub never appeared and failed to initialise.
[1] http://www.spinics.net/lists/linux-usb/msg125969.html
Signed-off-by: Ben Whitten <ben.whitten@gmail.com>
SVN-Revision: 46649
to avoid editing the dts every time the kernel size changes.
uImage is now bigger than 1MB. Pad uImage to 64k erase block size.
Signed-off-by: Günther Kelleter <guenther.kelleter@devolo.de>
SVN-Revision: 46648
Switch to generich chip irqs/irq domains.
Interrupts were broken since kernel 3.14. dLAN USB extender is now
booting again.
Signed-off-by: Günther Kelleter <guenther.kelleter@devolo.de>
SVN-Revision: 46647
Switch to new 8250 debug uart code because the old
mach-mcs814x/include/mach/debug-macro.S tries to include
asm/hardware/debug-8250.S which no longer exists since kernel 3.14
Signed-off-by: Günther Kelleter <guenther.kelleter@devolo.de>
SVN-Revision: 46646
The MT7621 uses a 2 bit wide configuration of the sdhci, spi, mdio, pcie,
wdt, uart2 and uart3 in the GPIO_MODE register. It was correctly done
for sdhci, spi, pcie and wdt, The same has to be done for uart3, uart2
and mdio.
Signed-off-by: Sven Eckelmann <sven@open-mesh.com>
SVN-Revision: 46645
The PINS conntrolled by the SPI bits in the GPIO_MODE register is always
7 and not 8 for nand mode.
Signed-off-by: Sven Eckelmann <sven@open-mesh.com>
SVN-Revision: 46644
The uart3 setting in GPIO_MODE register is before the uart2 setting. Also
don't mix uart2 and uart3 function/groups.
Signed-off-by: Sven Eckelmann <sven@open-mesh.com>
SVN-Revision: 46643
The following patch:
* fixes wrong indentation
* combines together boards with same configuration
* removes unnecessary spaces
* removes empty lines after ";;" and uses only "|\" (for consistency)
* orders boards alphabetically
in base-files/etc/board.d/02_network script.
One bug was left untouched: freestation5 board is listed twice in case statement, in ramips_setup_macs().
Signed-off-by: Piotr Dymacz <pepe2k@gmail.com>
SVN-Revision: 46634
This patch:
* fixes various mistakes (leading spaces, wrong or missing model names)
* shortens model strings (we don't need to use whole model string from dts file, make it for consistency and as optimization)
* sorts boards alphabetically
in base-files/lib/ramips.sh script.
Case statement was tested after above changes with simple script, which:
* takes model string from dts{,i} file (model = "...")
* provides this value as case input ($machine)
* returns output ($name)
Signed-off-by: Piotr Dymacz <pepe2k@gmail.com>
SVN-Revision: 46632
This patch:
* fixes wrong indentation
* fixes wrong names and descriptions (ex. Dlink -> D-Link)
* changes filenames to manufacturer names
* moves oem/unknown manufacturers boards to misc.mk profiles
* combines profiles for boards from the same manufacturers in one file (ex. Olimex)
* fix profile files structure (keep the same convention as in other targets, ex. ar71xx)
in various .mk profiles, in ramips target.
Signed-off-by: Piotr Dymacz <pepe2k@gmail.com>
SVN-Revision: 46631
It seems that support for Edimax BR-6524N was dropped long time ago (dts file for this device is missing).
This patch removes remaining code.
Signed-off-by: Piotr Dymacz <pepe2k@gmail.com>
SVN-Revision: 46630
The following patch changes 7Links PX-4885 dts{,i} filenames, board, image and profile names from "PX4885" to "PX-4885" (for consistency).
Signed-off-by: Piotr Dymacz <pepe2k@gmail.com>
SVN-Revision: 46629
The following patch changes Buffalo WMR-300 dts filename, board, image and profile names from "WMR300" to "WMR-300" (for consistency).
Signed-off-by: Piotr Dymacz <pepe2k@gmail.com>
SVN-Revision: 46628
