The debug log is only available when the debugfs is also enabled. This
must be reflected in Kconfig to avoid incompatible selection of batman-adv
options.
Signed-off-by: Sven Eckelmann <sven@narfation.org>
* luci-app-cjdns: luci admin support for cjdns-v20.2
* adds layer 2 and 3 address display column
* fixes switch label pinger
* get version from addr string
* parse addr string for peerStats publicKey
Signed-off-by: William Fleurant <william@netblazr.com>
* luci-app-cjdns: bump release, update repo-url and license year
Signed-off-by: William Fleurant <william@netblazr.com>
* luci-app-cjdns: support views for 17.01 and 18.06
- fixed user/name column
- removed latency column
- combind peerstats functions
- fix css on overview page
- table displays with `cbi_update_table` with fallback
- columns ordered similar to peerStats.js output
- normalized XHR polling to mimic wireless.htm by jow@openwrt
Signed-off-by: William Fleurant <william@netblazr.com>
Signed-off-by: William Fleurant <william@netblazr.com>
Applied Daniel Golle's patch from
https://github.com/bmx-routing/bmx7/pull/35 to "rename some list stuff
to avoid namespace collision with libubox."
Signed-off-by: Eneas U de Queiroz <cote2004-github@yahoo.com>
NodogSplash iptables entries are overwritten if the OpenWrt Firewall is restarted.
This change adds a restart hook to firewall.user to restart NoDogSplash if firewall is restarted.
Author-name: Rob White
Signed-off-by: Rob White rob@blue-wave.net
As no maintainer take the opportunity to drop from this repo and put in
main openwrt/packages repo.
Signed-off-by: Kevin Darbyshire-Bryant <ldir@darbyshire-bryant.me.uk>
I haven't used the package personally for last 3 years, so
de-maintainering myself as testing version updates is extra work I do
not need for stuff I do not use.
* Fix memory leaks 300522 and 300520 due to changed interface status and increased nlmsg size
* Fix physical channel detection: https://github.com/bmx-routing/bmx7/issues/19
* Enable DEBUG_MALLOC and CORE_LIMIT=20000 by default
* support latest kernels (3.2 - 4.17)
* coding style cleanups and refactoring
* add DAT cache and multicast flags netlink support
* avoid redundant multicast TT entries
* bugs squashed:
- update data pointers after skb_cow()
- fix header size check in batadv_dbg_arp()
- fix skbuff rcsum on packet reroute
- fix multicast-via-unicast transmission with AP isolation
- fix packet loss for broadcasted DHCP packets to a server
- fix multicast packet loss with a single WANT_ALL_IPV4/6 flag
Signed-off-by: Sven Eckelmann <sven@narfation.org>
* synchronization of batman-adv netlink and packet headers
* add DAT cache and multicast flags netlink support
* disable translation support for non-unicast mac addresses
Signed-off-by: Sven Eckelmann <sven@narfation.org>
Sysupgrade currently leads to a loss of the content of /etc/bmx7 which
contains the secret keys. After the reboot the device creates new IPs
leading to a new "identity" aka ID, shortID and IPv6 changes.
Occationally /proc/net/if_inet6 contains interface IDs with
three digits. In this case, the regex in wait_for_ll_address()
does not work anymore and alfred is not starting.
This patch changes the evaluation so that fields are used instead
of the mere position by counting characters.
Signed-off-by: Adrian Schmutzler <freifunk@adrianschmutzler.de>
Using procd allows better control and monitoring of the bmx7 process.
Keep polling for mac80211.sh process for now until we find a better
way to make sure that wireless configuration has been generated before
starting bmx7.
Setting maximum core size to 20000 looks dangerous on small devices,
but keep it as it was before, we shall discuss this separately.
Signed-off-by: Daniel Golle <daniel@makrotopia.org>
nodogsplash2: Add compatibility with mwan3 version 2
Changes in mwan3 version 2 resulted in previous recommended fix for Nodogsplash to fail.
Setting fw_mark to use bits set "higher" than used in mwan3 solves the problem. This works also for all previous versions of Nodogsplash.
Author-name: Rob White
Signed-off-by: Rob White <rob@blue-wave.net>
* Sort links by name
* Allow configuring invalid --pathLq1Threshold (due to configReload problem) and fix it in description
* Introduce --linkThroughputEfficiency=0 (70) and cleanup metric algos
* Update expectedThroughput only if estimation exists
* Calculate and use average expected throughput instead of fluctuating current value
* Rename wireless expected throughput variables
* Align ordering of link_status struct and field_format
* Fix: Interpret expected TP as Kbps
* Fix txRateExpected typos
* Add expected throughput as reported by minstrel(_ht)
