Do not fallback to .innerHTML if DOMParser() failed for whatever reason.
Signed-off-by: Jo-Philipp Wich <jo@mein.io>
(cherry picked from commit 63d9bcb6825fac92fd7dfa4ba858c8d5aafa23e8)
Instead of relying on .innerHTML which executes embedded script code to
parse a given HTML fragment, use dom.parse() which utilizies DOMParser()
internally in order to extract textContent in a safe manner.
Fixes: FS#4199
Ref: https://bugs.openwrt.org/index.php?do=details&task_id=4199
Signed-off-by: Jo-Philipp Wich <jo@mein.io>
(cherry picked from commit 993151504e8e810c083d3257555bdcdc2f00673a)
Note: This change is relevant for systems that don't
use uhttpd for LuCI.
This log can be later used for fail2ban etc.
Signed-off-by: Jan Pavlinec <jan.pavlinec@nic.cz>
(cherry picked from commit be8d749599a11229435141d85cfe5eba2dcf96d2)
Only treat the given identifier as Linux netdev name if we can find a
corresponding entry in the device info cache and do not consider strings
starting with "wlan", "ath" or "wl" to be existing devices.
This fixes incorrectly adding wireless sections as ifnames to network
interfaces when the wifi-iface section name begins with one of the
`iface_patterns_wireless` patterns.
Fixes: #5069
Signed-off-by: Jo-Philipp Wich <jo@mein.io>
(cherry picked from commit d4092b15ce92cda2d7711b57b2251ebebef32b5c)
No functional changes but required for styling rules.
Signed-off-by: Jo-Philipp Wich <jo@mein.io>
(cherry picked from commit 7fba52de91afc60a49345a75b805ff73f18f6f26)
Also convert ES6 to ES5 syntax while we're at it.
Fixes: #5476
Fixes: 5b42cd5b46 ("luci-mod-system: move default option from defaul-on trigger to none trigger")
Signed-off-by: Jo-Philipp Wich <jo@mein.io>
(cherry picked from commit fceabd74465f2aafc40edc084429f30837deff2c)
When removing a `config interface` section in `/etc/config/network`, drop
related `rule` and `rule6` sections too, as well as related `dhcp` sections
in `/etc/config/dhcp`.
Ref: https://forum.openwrt.org/t/grooming-etc-config/109764/7
Signed-off-by: Jo-Philipp Wich <jo@mein.io>
(cherry picked from commit 63034c36073bedf87db18d05d3fe8cca016c6490)
"Setup" is a noun, the thing. "Set up" is a verbal idiom, the action.
One cannot setup a DHCP server, just as one cannot fuckup a test, or
ripup some paper. One can, however, set a DHCP server up, and one can
fuck a test up, and rip some paper up.
Signed-off-by: Paul Dee <systemcrash@users.noreply.github.com>
(cherry picked from commit c0299e2c3d09c48b515447b2c6fefd00ffd2950b)
ECDH is not used for the ed25519. The scheme is called EdDSA.
Signed-off-by: Rosen Penev <rosenp@gmail.com>
(cherry picked from commit 738f36a1c31eda6d322e18c1b7fca4b6498e46a9)
.. when no dnsmasq is installed, as it
tells only that there are no settings available
Signed-off-by: Fritz D. Ansel <fdansel@yandex.ru>
(cherry picked from commit 2393289367bb3c4e366ee36cc66af9149687139b)
when buffered could not be read, there is
just a questionmark (on e8450 non-ubi)
Signed-off-by: Fritz D. Ansel <fdansel@yandex.ru>
(cherry picked from commit 61425793a78ea60b564f9767b60b001da0e4cc96)
10 lines are very few and there is much unused space
Signed-off-by: Fritz D. Ansel <fdansel@yandex.ru>
(cherry picked from commit 8fb2d82f703743d3432fc26f5d2cb14895bf57f7)
Use the system/info ubus call to obtain a TZ adjusted epoch value and
format the date time string manually instead of relying on the browsers
local representation.
Fixes: #5454
Signed-off-by: Jo-Philipp Wich <jo@mein.io>
(cherry picked from commit 07569c5af23d74bad8b78905b25d7e85bd2d3141)
The `up.gif` icon is not present by default anymore, it has been moved to
luci-compat. Probe loading.gif instead which is guaranteed to be available
by default.
Signed-off-by: Jo-Philipp Wich <jo@mein.io>
(cherry picked from commit ea9b5e87e6ec3fbab52a02ab5f18bb51292bd249)
Currently any new entry added to the black/whitelist deletes the
others. Fix this and improve the rule apply by firstly disconnect
the wifi client.
Fixes: #5343
Signed-off-by: Ansuel Smith <ansuelsmth@gmail.com>
(cherry picked from commit 085c4a08f5bf50919731304a962cfcd2b592d523)
Drop isWPSEnabled table as it's really not needed and can directly
included in the network table.
Signed-off-by: Ansuel Smith <ansuelsmth@gmail.com>
(cherry picked from commit 2a97d5533cba419f51da0b550b00a34bc9f50cce)
Fix the incorrectly displayed value for the option "ip6table".
Signed-off-by: Vladislav Grigoryev <vg.aetera@gmail.com>
(cherry picked from commit 3ebdf51f9b2e713cd655c0f97f2542c96736cae5)
Fix incorrect minimum 60GHz frequency value and ensure that a 60GHz band
choice option is available once corresponding channels are found on the
system.
Ref: http://lists.openwrt.org/pipermail/openwrt-devel/2021-July/035810.html
Suggested-by: Alex Henrie <alexhenrie24@gmail.com>
Signed-off-by: Jo-Philipp Wich <jo@mein.io>
(cherry picked from commit e5626ece12236f6be9dbb6da6eb90fcbb469a1f0)
- validate current form value IP instead of stored uci one
- ensure to not pass non-string values to the validator
- remove leftover debug code
- simplify obtaining formvalues
Signed-off-by: Jo-Philipp Wich <jo@mein.io>
(cherry picked from commit 9d15a4e70c10a43911c7f9d48240a01c23eb81b6)
- Ensure that MAC addresses are unique within the same pool
- Ensure that IP addresses are globally unique
- Ensure that IP addresses are within any DHCP pool range
Signed-off-by: Jo-Philipp Wich <jo@mein.io>
(cherry picked from commit 7e562895380d6bbe89ed88fe7576272c078f3b59)
There are many switch drivers where VLAN functionality is enabled by default.
In this situation, LuCI cannot be used to disable VLAN functionality,
because removing the line from configuration results in VLAN remaining enabled
by the driver.
When enable_vlan is set to 0 by default
by using functions in board.d,
and an unrelated change is made on the Switch webpage,
then the enable_vlan option gets deleted when changes are saved.
Therefore, the option needs to be preserved, whether set to 0 or 1.
Signed-off-by: Michael Pratt <mcpratt@pm.me>
(cherry picked from commit 51ab737a4ba2b0f2a9d479cf114837c570f8d05c)
Fix the wrong value for the "source" option of the "route6" section.
Signed-off-by: Vladislav Grigoryev <vg.aetera@gmail.com>
(cherry picked from commit 8b4015036d6f298cbfe0291ab8fdc23b4133cca7)
