Age | Commit message (Collapse) | Author |
|
Some channels may restrict its use to indoor only. Add support for
this and alert the user if a channel with this limitation is used.
Closes #5695
Signed-off-by: Christian Marangi <ansuelsmth@gmail.com>
Signed-off-by: Paul Donald <newtwen+github@gmail.com>
|
|
Signed-off-by: Paul Donald <newtwen+github@gmail.com>
|
|
Include support in luci to enable the Private PSK.
When Private PSK is enabled, clients can use a default password (common),
or have their own private password that is associate with the client MAC.
The password is retrieved from RADIUS server, asking for the client MAC,
and when such client MAC is not defined, RADIUS should return a default
password.
RADIUS can also return other parameters like VLANs, which can be used to
put clients dinamically in different vlans based on their MAC, or the
default configuration.
Private PSK is not compatible yet with SAE encryption, therefore cannot
be used yet with WPA3 or WPA3-mixed.
It implements the UI for the PPSK already in master: https://github.com/openwrt/openwrt/pull/3509
Closes #4513
Signed-off-by: Manuel Giganto <mgigantoregistros@gmail.com>
Signed-off-by: Paul Donald <newtwen+github@gmail.com>
|
|
This follows https://github.com/openwrt/openwrt/pull/14975
Signed-off-by: Paul Donald <newtwen+github@gmail.com>
|
|
This follows https://github.com/openwrt/openwrt/pull/14975
Signed-off-by: Paul Donald <newtwen+github@gmail.com>
|
|
As a convenience at load/save, the hex field de/encodes hex respectively
This follows https://github.com/openwrt/openwrt/pull/14975
Signed-off-by: Paul Donald <newtwen+github@gmail.com>
|
|
Tested with filogic banana rpi4 / QNAP qualcommax ipq8074
Signed-off-by: Rudy Andram <rmandrad@gmail.com>
Tested-By: Daniel Pawlik <pawlik.dan@gmail.com>
|
|
The previous implementation of binding the protocol validation to the
name validation caused problems when creating a new interface where an
already existing interface had the new one's protocol as its name, as the
protocol would be used when validating if an interface of that name already
existed.
Consider the following case:
1. Interface 'gre' with protocol PPPoE is created
2. Interface 'foo' with protocol GRE tunnel over IPv4 (which resolves to
'gre') tries to be created
Creating interface 'foo' would error out during protocol validation as
'gre' would be passed to the name validation, which in turn would check if
an interface with name 'gre' already exists, which it does.
Instead of reusing the validation logic, simply trigger the name
validation manually as the protocol changes to properly pass the name of
the interface instead of the protocol. This also gives the benefit of
keeping all error states related to the name contained in the appropriate
UI element.
Fixes #7146
Tested-by: Tiago Gaspar <tiagogaspar8@gmail.com>
Signed-off-by: Daniel Nilsson <dannil+github@protonmail.com>
|
|
Closes #7252
Thanks to @efahl
Signed-off-by: Paul Donald <newtwen+github@gmail.com>
|
|
Signed-off-by: Chen Minqiang <ptpt52@gmail.com>
|
|
Signed-off-by: Paul Donald <newtwen+github@gmail.com>
|
|
|
|
dnsmasq: Add options for new EDNS0 options
|
|
Fixes: #7260
Ref: https://git.openwrt.org/5c2e5d5ea4e0dd9a0ed66154d818b96bf81664f4
Signed-off-by: Jo-Philipp Wich <jo@mein.io>
|
|
Signed-off-by: Carsten Schuette <schuettecarsten@googlemail.com>
|
|
Make it possible to set Operating Channel Validation through luci.
Signed-off-by: Michael Yartys <mail@yartys.no>
|
|
instead of returning a space separated string
follow-up to 5ab0cb1f15bcb31015ae8345d302c6d023ba51e2
Closes #7209
Signed-off-by: Paul Donald <newtwen+github@gmail.com>
|
|
Closes #7199
Signed-off-by: Paul Donald <newtwen+github@gmail.com>
|
|
Signed-off-by: Florian Eckert <fe@dev.tdt.de>
|
|
And include ccmp-256 bit
Signed-off-by: Paul Donald <newtwen+github@gmail.com>
|
|
Closes #7193
See https://github.com/openwrt/netifd/commit/db3934d2f740bdfe8537933741f71b439a109422
Signed-off-by: Paul Donald <newtwen+github@gmail.com>
|
|
treewide: add new css class button-row
|
|
Signed-off-by: Florian Eckert <fe@dev.tdt.de>
|
|
Signed-off-by: Sergio Nemirowski <sergio@outerface.net>
|
|
Signed-off-by: Florian Eckert <fe@dev.tdt.de>
|
|
Signed-off-by: Florian Eckert <fe@dev.tdt.de>
|
|
Signed-off-by: Paul Donald <newtwen+github@gmail.com>
|
|
Signed-off-by: Paul Donald <newtwen+github@gmail.com>
|
|
An user can insert a strange value right now, with this commit, only integer value can be insert.
Signed-off-by: Andrea Pesaresi <andreapesaresi82@gmail.com>
|
|
Signed-off-by: Paul Donald <newtwen+github@gmail.com>
|
|
Due to the changing on script from this commit https://github.com/openwrt/openwrt/commit/7ebcf2fb9c5db8b75666761f5e767c91abb1f437 I suggest
these changing to select "old style" or "new style" with steering_flow
Signed-off-by: Andrea Pesaresi <andreapesaresi82@gmail.com>
Signed-off-by: Paul Donald <newtwen+github@gmail.com>
|
|
Set a matching HT mode and channel number on joining a wireless network.
Fixes: #7066
Signed-off-by: Jo-Philipp Wich <jo@mein.io>
|
|
Fixes: #7074
Fixes: 61cef9baad ("luci-mod-network: don't trigger uci save on removing bridge vlans")
Signed-off-by: Jo-Philipp Wich <jo@mein.io>
|
|
luci-base: conversion of ucitrack handling from uci to json
|
|
On the configuration page of '/admin/network/dhcp' on the IP sets tab,
it is possible to make changes directly in the table. However, this is
not displayed correctly when using the theme 'luci-theme-material' if the
options can be changed directly in this view.
The problem is solved by not editing the configuration in the overview page
but in the overlay page.
Signed-off-by: Florian Eckert <fe@dev.tdt.de>
|
|
Move the json file to where it belongs.
Signed-off-by: Florian Eckert <fe@dev.tdt.de>
|
|
Move the json file to where it belongs.
Signed-off-by: Florian Eckert <fe@dev.tdt.de>
|
|
Move the json file to where it belongs.
Signed-off-by: Florian Eckert <fe@dev.tdt.de>
|
|
Move the json file to where it belongs.
Signed-off-by: Florian Eckert <fe@dev.tdt.de>
|
|
In order to unclutter the default dnmasq configuration view, move the
Hostnames (A+PTR), SRV, MX and CNAME sections into a new dedicated
"DNS Records" tab.
Signed-off-by: Jo-Philipp Wich <jo@mein.io>
|
|
- Hide anonymous section title if only one instance is present
- Hide section remove button if only one instance is present
- In case multiple sections are present, label the first and
subsequent anonymous sections as `Default instance` and
`Unnamed instance #N` respectively
- Label named sections as `Instance "XXX"`
- Label delete buttons as `Remove instance "XXX"`,
`Remove default instance` or `Remove instance #N` depending on whether
the associated instance is the first anonymous one, a subsequent
anonymous one or a named one
- Label add button as `Add server instance`
- Add placeholder text `New instance nameā¦` to the section name input
Ref: https://forum.openwrt.org/t/x/194048
Signed-off-by: Jo-Philipp Wich <jo@mein.io>
|
|
add a null-check before parsing networks
loop through available networks on all interfaces except loopback
Closes #7047
Signed-off-by: Paul Donald <newtwen+github@gmail.com>
|
|
Appears under DHCPv6-service when set to 'server' or 'hybrid'.
Option 'ntp' is serviced by odhcpd. It sends a DHCPv6 NTP Server option
56 for each 'ntp' list entry.
Once uci loads the system section, it provides suggestion presets of
system configured upstream NTP servers, and the LAN IP(v6) of the router
which are then statically configured by the user.
It would make more sense if odhcpd fetches them directly.
Alas, it does not (yet).
Signed-off-by: Paul Donald <newtwen+github@gmail.com>
|
|
ULA as defined in RFC4193 mentioned a possible future where the L bit
is 0 (implying some global coordinator be used). Locally generated ULA
always have the L bit set to 1, so the prefix is always fd00::/8 (though
ULA with the L bit set 0 or 1 are treated identically).
Technically, fc00::/7 isn't a range, but a prefix which implies a range,
conveniently partitioned by and at the L bit. A range would be correctly
denoted as fc00::-fdff:ffff:ffff:ffff:ffff:ffff:ffff:ffff.
The rewording avoids mention of fc00::/7. The randomly generated prefix
in all cases starts fd.
The sentences were split to ease i18n.
Signed-off-by: Paul Donald <newtwen+github@gmail.com>
|
|
Do not trigger a uci save operation on removing bridge VLANs as that might
invalidate the ephemeral section ID of a just added network device.
Fixes: #6990
Signed-off-by: Jo-Philipp Wich <jo@mein.io>
|
|
This resulted in the sysfs parameter to be invalid as it contained a space.
Signed-off-by: Daniel Nilsson <daniel.nilsson94@outlook.com>
|
|
Add a clarifying note about two different meanings of "network"on
the wireless join page. In the dialog title the word "network" refers
to the wireless SSID, while in the config item the same "network"
refers to the name to be used in OpenWrt network configuration.
Reference to forum discussion at
https://forum.openwrt.org/t/cant-join-a-scanned-network/189859
Signed-off-by: Hannu Nyman <hannu.nyman@iki.fi>
|
|
Relax the CNAME domain validation to allow wildcard names.
Fixes: #6801
Signed-off-by: Jo-Philipp Wich <jo@mein.io>
|
|
See following for more detail:
https://w1.fi/cgit/hostap/commit/?id=af35e7af7f8bb1ca9f0905b4074fb56a264aa12b
Signed-off-by: Paul Donald <newtwen@gmail.com>
|
|
saved value will now be only an integer.
Signed-off-by: Paul Donald <newtwen@gmail.com>
|