Remove the code related to the deprecated madwifi driver.

Signed-off-by: Hannu Nyman <hannu.nyman@iki.fi>

Use the new hostapd functionality (in LEDE) to detect 802.11r and 802.11w
more properly. Leave the old logic in place for Openwrt.

Adjust 802.11w default value from "0" to "" in order to avoid unnecessary
"option ieee80211w 0" lines in /etc/config/wireless

Signed-off-by: Hannu Nyman <hannu.nyman@iki.fi>

Adds support for the fwmark option.

FwMark is a 32-bit fwmark for outgoing packets.
If set to 0 or "off", this option is disabled.

Signed-off-by: Dan Luedtke <mail@danrl.com>

Signed-off-by: Luiz Angelo Daros de Luca <luizluca@gmail.com>

Signed-off-by: Florian Eckert <Eckert.Florian@googlemail.com>

Signed-off-by: Florian Eckert <Eckert.Florian@googlemail.com>

Signed-off-by: Dan Luedtke <mail@danrl.com>

Signed-off-by: Hannu Nyman <hannu.nyman@iki.fi>

Openwrt theme apparently accepsts slightly different html than
Bootstrap or Material. Cahnge code to <br /> to make it compatible.

This fixes #1011

Signed-off-by: Hannu Nyman <hannu.nyman@iki.fi>

The default hostapd/wpad mini version does not support ieee80211w
and the option's visibility in LuCI may lead it to be placed in user's
/etc/config/wireless preventing wifi from starting.

Check for hostapd/wpad full before exposing the option.

Signed-off-by: Hannu Nyman <hannu.nyman@iki.fi>

luci ignores only 127.0.0.1 in real time connections.
The dnsmasq.init script sets up resolv.conf(s) as such.
With alternate DNS servers configured 'localhost' will
resolve to '127.0.0.1' or '::1'. The connections graph
will spam itself. openwrt/luci#996

Signed-off-by: Eric Luehrsen <ericluehrsen@hotmail.com>

Added some Japanese translations of recently added strings.

Signed-off-by: INAGAKI Hiroshi <musashino.open@gmail.com>

Synchronized translations with Lua sources.

Signed-off-by: INAGAKI Hiroshi <musashino.open@gmail.com>

Use ListValue type and enable translation of the option values.

Signed-off-by: Hannu Nyman <hannu.nyman@iki.fi>

Add support for IEEE802.11w Management Frame Protection options.

Mention in the help text that the full version of wpad/hostapd is
needed.

Currently supported by ath9k and ath10k, and in LEDE also by
mwlwifi and mt76.

Signed-off-by: Hannu Nyman <hannu.nyman@iki.fi>

Apply hexstring datatype to mobility domain and r1_key_holder.

Signed-off-by: Hannu Nyman <hannu.nyman@iki.fi>

Add datatype 'hexstring' for input validaiton datatypes.

It will accept any hexadecimal string.

(no length validation, as rangelength can be used for that.)

Signed-off-by: Hannu Nyman <hannu.nyman@iki.fi>

* Add WPA/WPA2-PSK mixed mode to the list of allowed modes
* Add datatype validation to some fields
* Add help texts
* Reformat longest lines

Signed-off-by: Hannu Nyman <hannu.nyman@iki.fi>

Adds support for 802.11r fast transition to luci.
Corrects the deletion on nasid in issue #844

The R0KH & R1KH options are not perfect as it would be better
to separate out each option and do some sanity checking on
the values, but that is beyond my capabilities.

Signed-off-by: Warren Linton <warren@linton.id.au>

(squashed from three original commits in #994. commit message reworded)

The same wifi option 'nasid' is used for two purposes,
RADIUS NAS ID and 802.11r NAS ID / R0KH-ID.

In LuCI the 'nasid' option has depended on WPA-EAP/WPA2-EAP, which
can lead to accidental deletion of the option, if the user modifies
any wifi settings while he has WPA(2)-PSK & 802.11r mode.

802.11r can also be with WPA(2)-PSK, so add them to the dependency
list of allowed encryptions.

(Note: after this fix the 'nasid' field is shown unnecessarily
to "normal" WPA(2)-PSK users. It might later be set to depend on
802.11r being used, after 802.11r has been added to LuCI.)

Reference to bugs:
https://github.com/openwrt/luci/issues/844
https://dev.openwrt.org/ticket/20764
https://bugs.lede-project.org/index.php?do=details&task_id=274

Signed-off-by: Hannu Nyman <hannu.nyman@iki.fi>

Signed-off-by: Jo-Philipp Wich <jo@mein.io>

The previous commit erroneously used "ports" instead of "port" as name for the
option widget, causing wrong uci values to be written.

Also work around some cbi idiosyncrasies regarding MultiValue widgets which
prevented rendering the correct initial selection state.

Signed-off-by: Jo-Philipp Wich <jo@mein.io>

Added and modified some Japanese translations.

Signed-off-by: INAGAKI Hiroshi <musashino.open@gmail.com>

Synchronized translations with Lua sources.

Signed-off-by: INAGAKI Hiroshi <musashino.open@gmail.com>

The Overview page and Network>Interfaces page currently do not give much information about IPv6, particularly with Prefix Delegated setups. In these setups, ISP will delegate a prefix to the router. Currently LuCI doesn't display this Prefix Delegation from the ISP anywhere. A number of changes was added to this commit:

1) self:_ubus("ipv6-prefix") was extracted and put into protocol.ip6prefix.
2) Network>Interfaces page, if a .ip6prefix is present, show it under Status. (IPv6-PD).
3) On the Overview page, "Type" and "Prefix Delegated" has been added to the IPv6 Network Overview Status:
  - Type will display the .proto, similar to the IPv4 case. If a .ip6prefix is present, it'll display a "-pd" at the end of the Type: i.e. dhcpv6-pd vs. dhcpv6.
  - If no .ip6prefix is present, it'll do what it does currently, and just show Address, or :: if no address is present.
  - If .ip6prefix is present, it'll show the "Prefix Delegated", it'll also hide "Address" if no address is present, else it'll show ifc6.ip6addr as well.

Signed-off-by: Cody R. Brown <dev@codybrown.ca>

Added and modified some translations for Japanese.

Signed-off-by: INAGAKI Hiroshi <musashino.open@gmail.com>

Synchronized translations with Lua sources.

Signed-off-by: INAGAKI Hiroshi <musashino.open@gmail.com>

Move the password setting into the .parse() callback to ensure that it is
updated regardless of whether "Save" or "Save & Apply" has been pressed.

Signed-off-by: Jo-Philipp Wich <jo@mein.io>

Signed-off-by: Jo-Philipp Wich <jo@mein.io>

Signed-off-by: Jo-Philipp Wich <jo@mein.io>

Reloading the page ensures that the allowed tx power values and channel select
lists are properly populated with the new effective values.

Signed-off-by: Jo-Philipp Wich <jo@mein.io>

Sometimes txpower is not set so we should have an option with empty
value to let it stay unset.

Signed-off-by: Rafał Miłecki <rafal@milecki.pl>

The missing parens lead to a wrong expression precedence, causing a runtime
error when attempting to compare nil with a number.

Signed-off-by: Jo-Philipp Wich <jo@mein.io>

These validators accept IP address specifications in the form

 - address
 - address/bits
 - address/netmask

Signed-off-by: Jo-Philipp Wich <jo@mein.io>

Signed-off-by: Jo-Philipp Wich <jo@mein.io>

The expiry time in a dnsmasq lease file line may be 0 (i.e.
expiry date = 01/01/1970 00:00:00 GMT) to denote an infinite
lease time, so adjust the code to properly support that.

The expiry attribute of the lease object will be set to "false"
in case of an infinite lease. This is to mimic the odhcp code below.
If the expiry date is not equal to 0, then just do exactly what was
done before (return the os.diff of current time and ts).

Signed-off-by: Cody R. Brown <dev@codybrown.ca>

This should result in the MAC address display being the same
using odhcpd for v4 DHCP as when using DNSMasq for v4 DHCP.

Signed-off-by: Daniel Dickinson <lede@cshore.thecshore.com>

Recently a string to be translated were added to Lua source, so
I synchronized po templates and po file of each language with Lua
source.

Signed-off-by: INAGAKI Hiroshi <musashino.open@gmail.com>

Fixed some display problems where the UUID label of the drop down
list does not exist and the positions of other labels are misplaced
in mount settings page.

Signed-off-by: INAGAKI Hiroshi <musashino.open@gmail.com>

Fixed an issue where the network name is displayed as "%q" on wlan
connection destination network settings page.

Signed-off-by: INAGAKI Hiroshi <musashino.open@gmail.com>