From: Kristian Evensen
According to the OpenWRT Network documentation for route, the 'source' option is
"The preferred source address when sending to destinations covered by the
target". However, netifd currently stores this value in RTA_SRC on
NEWROUTE/DELROUTE.
RTA_SRC i
From: Kristian Evensen
Enable callers to pass the source IP of an IPv4 route when using
proto_add_ipv4_route(). This is useful with for example DHCP in a multihomed
scenario, as it provides an easy way to match default routes with the correct IP
address. One use case for this are applications
From: Kristian Evensen
This patch depends on "Pass source address to proto_add_ipv4_route".
I have not found a scenario that would break by setting the source address on
default, but please let me know if any special considerations should be taken.
Signed-off-by: Kristian Evensen
--
On Wed, Dec 10, 2014 at 9:10 AM, Hans Dedecker wrote:
> As a side effect of this patch I noticed multiple default routes (one
> default route with source IP set and one default route with source IP
> unset) are present in netifd if the dhcp server replies the default
> route both in the router opt
Hi Petr,
On Fri, Sep 20, 2019 at 9:56 AM Petr Štetiar wrote:
> could you please rebase to series to the current state of the tree? I would
> like to apply it, thanks!
>
> BTW don't forget to include the license change ACKs.
I will do it during or right over the weekend. Btw, can I consider
this
"flash"-name for the spi-nor node.
All changes have been tested on the WE1026-5G-16M and work fine. I.e.,
the device works as before the DTS-changes.
v1->v2:
* Added missing acked-bys.
Signed-off-by: Kristian Evensen
Acked-by: Mathias Kresin
Acked-by: Alex Maclean
Acked-by: IN
Signed-off-by: Kristian Evensen
---
.../ramips/base-files/etc/board.d/01_leds | 5 +++
.../ramips/base-files/etc/board.d/02_network | 6 ++-
.../dts/mt7620a_zbtlink_we1026-h-32m.dts | 14 +++
.../ramips/dts/mt7620a_zbtlink_we1026-h.dtsi | 42 +++
target/linux
the WE1026-H. The
second patch adds support for the WE1026-H.
The main change between v1 and v2 is the addition of acked-bys on the first
patch, triggered by the re-licensing of the DTS'.
Signed-off-by: Kristian Evensen
Kristian Evensen (2):
ramips: Update ZBT WE1026 DTS-files
ramips
Hi,
On Tue, Sep 24, 2019 at 1:28 PM Adrian Schmutzler
wrote:
>
> Hi,
>
> first of all:
> Naming scheme for ZBT devices is mixed in ramips. Some include the ZBT in
> model name, some don't. In your case, this means the following options:
> zbtlink,zbt-we1026-hcorresponding to modelZBT-WE1
Hi Adrian,
On Tue, Sep 24, 2019 at 6:22 PM wrote:
> I'm all about consistency. I just scanned the image definitions in ramips:
>
> ...
>
> The only device deviating from the pattern "zbtlink_zbt-something" is
> zbtlink_we1026-5g-16m.
>
> So, IMO the correct solution _in terms of consistency_ wou
Hi,
On Mon, Oct 21, 2019 at 6:07 PM Robert Marko wrote:
>
> Merging this today has caused a regression in ipq40xx.
> PHY reset patch was is the issue, it now forces devices to have GPIO
> for PHY reset which most devices don't have and if it's missing it
> will make the driver panic and probing w
Hi all,
On Wed, Aug 28, 2019 at 7:37 PM Kristian Evensen
wrote:
>
> firewall3 currently creates one rule for each interface that is a member of a
> zone. On for example devices with multiple interfaces, the current firewall3
> behavior quickly leads to a lot of rules. In order t
"flash"-name for the spi-nor node.
* Add label MAC.
All changes have been tested on the WE1026-5G-16M and work fine. I.e.,
the device works as before the DTS-changes.
Signed-off-by: Kristian Evensen
Acked-by: Mathias Kresin
Acked-by: Alex Maclean
Acked-by: INAGAKI Hiroshi
Acke
his
device. However, the files are structured so that adding support for the
16MB version should be easy.
* Only the LAN port is accessible from the outside of the casing and LEDs
are not visible.
Signed-off-by: Kristian Evensen
---
v2->v3:
* Rebase on top of master.
* Added label mac (thanks Adri
the WE1026-H. The
second patch adds support for the WE1026-H.
The main change between v1 and v2 is the addition of acked-bys on the first
patch, triggered by the re-licensing of the DTS'. The changes between v2 and v3
are mostly related to ensuring consistent naming.
Signed-off-by: Kristian Ev
Hi,
On Sat, Nov 2, 2019 at 3:18 PM Kristian Evensen
wrote:
> Acked-by: Mathias Kresin
> Acked-by: Alex Maclean
> Acked-by: INAGAKI Hiroshi
> Acked-by: Petr Štetiar
I was a bit too fast when sending this patch and forgot to add the
ones who have ACKed the proposed change to t
Hi Adrian,
On Sun, Nov 3, 2019 at 12:36 PM wrote:
>
> Hi Kristian,
>
> > -Original Message-
> > From: openwrt-devel [mailto:openwrt-devel-boun...@lists.openwrt.org]
> > On Behalf Of Kristian Evensen
> > Sent: Samstag, 2. November 2019 15:19
> > To:
Hi Adrian,
On Mon, Nov 4, 2019 at 11:44 AM Adrian Schmutzler
wrote:
>
> If set, label MAC address is available from one of two sources,
> device tree or board.json. So far, the function get_mac_label
> was meant for retrieving the address, while an option in uci
> system config was specified only
Hi,
On Sun, Nov 3, 2019 at 3:14 PM wrote:
> Okay, if it's not visible I do not think it's worth to deviate from normal
> procedure here.
>
> I've remove the power_led label and aliases.
>
> Feel free to test and provide an updated solution for the use as USB LED.
>
> Despite, note that the first
Hi,
On Mon, Nov 4, 2019 at 4:36 PM Adrian Schmutzler
wrote:
> However, I'm not aware of a case where board.json is used for anything else
> than setting up config files, like in a script with user-interaction etc.
While I agree that those are the most common use-cases for board.json,
board.json
Use reset-gpio instead of the custom phy-reset-gpio property to do phy
reset on the U4019. phy-reset-gpio was incorrectly introduced when we
added support for the U4019, and will be deprecated.
Signed-off-by: Kristian Evensen
---
.../arch/arm/boot/dts/qcom-ipq4019-unielec-u4019.dtsi
Hello,
I am currently performing some performance measurements, comparing the
(wired) routing throughput (WAN <-> LAN) of 19.07, 21.02 and master on
mt7621 (ZBT WG-3526). I have connected one client to my LAN and one to
the WAN, and use iperf3 to measure. I create parallel flows (in order
to take
Hello again,
On Sun, Dec 19, 2021 at 12:29 PM Kristian Evensen
wrote:
> Based on my measurements, the throughput is reduced by ~50% going from
> 19.07 and to 21.02/master (~450Mbit/s vs. ~900Mbit/s). I do not have a
> particular commit I can point to, but I believe the regressions is
&g
Hi,
On Tue, Dec 21, 2021 at 6:34 PM Kristian Evensen
wrote:
> Since the only change between my sets of tests is the software,
> something has clearly improved in either the kernel or OpenWrt (as
> would be expected :)). Are there any particular commits/patches that
> would be worth
Hi,
On Sun, Feb 6, 2022 at 9:52 AM Arınç ÜNAL wrote:
>
> On 06/02/2022 00:15, Arınç ÜNAL wrote:
> > On 05/02/2022 21:23, Hauke Mehrtens wrote:
> >> On 2/5/22 19:21, Rosen Penev wrote:
> >>> On Sat, Feb 5, 2022 at 10:12 AM Hauke Mehrtens wrote:
>
> On 2/3/22 13:06, DENG Qingfang wrote:
Hi,
On Tue, Feb 11, 2020 at 2:04 PM Petr Štetiar wrote:
> > * Cleans up the nftables-dependencies in netfilter.mk. All targets are
> > not at 4.14+, so there is no need to specify for example "ge 4.9.0" or
> > keep "lt 4.9.0" around.
> > * Fix building support for nftables sets. In 4.18 the confi
Hi everyone,
I am sorry for my late reply to this thread. My email provider flagged
it as spam, so I only saw the conversation now. It seems that you have
reached a conclusion on how to proceed, but I thought I should anyway
share my notes/observations on this issue (in case they can be
useful).
vices. Removing a device is also handled correctly.
Signed-off-by: Kristian Evensen
---
system-linux.c | 9 -
1 file changed, 8 insertions(+), 1 deletion(-)
diff --git a/system-linux.c b/system-linux.c
index d533be8..aff67d6 100644
--- a/system-linux.c
+++ b/system-linux.c
@@ -590,6 +590
Hi,
On Wed, Mar 11, 2020 at 2:13 PM Kristian Evensen
wrote:
>
> After an interface has been renamed on a "fast" device (for example
> x86_64), the interface is sometimes not handled correctly by netifd.
> Looking in the logs, I see the following messages when renaming fails
, the MTU of the
bridge device is update. The callback covers both the case when a device
is added and removed (NEWLINK events are received in both cases), so the
old work-around is removed.
Signed-off-by: Kristian Evensen
---
bridge.c | 22 +-
device.h | 2
Hello,
On Thu, Aug 6, 2020 at 1:44 PM Jaap Buurman wrote:
> However, on this mailing list a user by the name of Kristian claims
> that disabling flow control helps fix this problem, as can be read
> here:
> https://lists.openwrt.org/pipermail/openwrt-devel/2017-November/009882.html
My patch unf
From: Kristian Evensen
Multi-wan support was recently added to netifd, but limited to IPv6. This patch
enables multi-wan for IPv4 as well. In addition, the patch introduces some
changes that make the multi-wan support more robust.
1) Instead of using the interface index to decide on interface
From: Kristian Evensen
This patch implements the comments received for version 1 (see
https://lists.openwrt.org/pipermail/openwrt-devel/2013-May/020119.html).
* Routes are only added to the specified table. The only exception are the
IPv4-routes Linux generates automatically when an address
This patch implements the comments received for version 1 (see
https://lists.openwrt.org/pipermail/openwrt-devel/2013-May/020119.html), and
improves upon version 2 by respecting metrics.
* Routes are only added to the specified table. The only exception are the
IPv4-routes Linux automatically
Hi,
The current generic multi-wan implementation does not work properly when a
device is connected to multiple overlapping subnets. Reaching machines outside
the networks work, but attempts to reach machines inside the networks will
always go through the first interface that was added with the sub
Hi Rafal,
On Sat, Aug 24, 2013 at 5:19 PM, Rafał Miłecki wrote:
>
> I wonder what do you think about this solution. Is this something we
> could try to implement ourself? Is it worth it? Is there some existing
> project doing similar thing?
This is a very interesting discovery. Have you tried to
Hi Rafael,
I will just reply to both emails here. First, it seems as though my
ebtables knowledge has gotten a bit rusty and there were some details
I forgot. The tool is used for firewalling bridges, so, the LAN and
WAN interface has to make up a bridge. I succeeded with a simple
configuration on
From: Kristian Evensen
uloop_run calls uloop_setup_signals() to set up signal handling before the while
loop, but does not remove the signal handling after the loop has ended. This can
cause problems for for example applications using the ubus file descriptor in
their own event loops, and
Hi,
On Mon, Oct 21, 2013 at 5:31 PM, Felix Fietkau wrote:
> bool is a more appropriate type here than uint8_t
Thanks for letting me know.
> I think you need to use a different approach. What you're doing here
> breaks recursive uloop_run calls, which are used in a few places.
> You need to chan
From: Kristian Evensen
uloop_run calls uloop_setup_signals() to set up signal handling before the while
loop, but does not remove the signal handling after the loop has ended. This can
cause problems for for example applications using the ubus file descriptor in
their own event loops, and
Hello,
I am currently working on an embedded project based on the Atheros
AR9344 SoC. As a prototype device, we are using the TP-Link TL-WDR4300
router (http://wiki.openwrt.org/toh/tp-link/tl-wdr4300) and latest
OpenWRT trunk. The kernel is 3.10.18.
We have over the last couple of weeks experienc
Hi Michel,
Thank you very much for a quick reply and sorry for my late reply. I
was without internet and access to the device yesterday.
On Fri, Nov 29, 2013 at 6:21 PM, Michel Stempin
wrote:
> Are your devices and hubs enumerated as full or high-speed?
They are all enumerated as high-speed.
>
Hi Vittorio,
Thank you very much for your detailed replies :)
On Fri, Nov 29, 2013 at 6:56 PM, Vittorio G (VittGam)
wrote:
> Have you tried with stable AA? I have the same router but with AA, and I've
> attached quite a lot of USB devices to it (one 32GB Cruzer Fit pendrive as
> extroot, 2 power
Hi again,
On Fri, Nov 29, 2013 at 6:21 PM, Michel Stempin
wrote:
>
> What happens if you turn off the WiFi during this time?
When looking through the config, I just noticed that wifi has always
been turned off.
-Kristian
___
openwrt-devel mailing list
Hello,
I have spent some days debugging an issue I am frequently seeing with
netifd and ppp modems. When connecting the modem after having
connected and disconnected it once, i.e., the second time netifd
"sees" the modem, netifd frequently thows a sigsev.
After going through the code, adding prin
From: Kristian Evensen
This patch adds the userspace and kernelspace match for nfacct, the extended
accounting infrastructure.
Signed-off-by: Kristian Evensen
---
package/kernel/linux/modules/netfilter.mk | 21 +
package/network/utils/iptables/Makefile | 15
Hi,
On Mon, Mar 31, 2014 at 11:39 AM, banglang huang
wrote:
> And I want to learn how this file can be generated and why the
> openwrt knows the dns ip address automatically (nameserver) ??
This file is generated by the network interface daemon (netifd) in the
function interface_write_re
Hi,
On Mon, Mar 31, 2014 at 11:58 AM, Tymon wrote:
> How can it retrieve the nameserver in 'DHCP' way ?
When a nameserver is retrieved through DHCP, it is part of the reply
sent by the DHCP server. It is set (and communicated to) netifd
through the /lib/netifd/dhcp.script, which is run by udhcp
From: Kristian Evensen
Hello,
I have written a small tool which automatically adds/removes a usb LAN dongle
from uci. When the USB device is inserted, I add entries to UCI (using libuci)
and do network reload (using libubus). When the device is removed, I remove the
UCI-entries and do reload
101 - 149 of 149 matches
Mail list logo
