Reiner Herrmann [Sun, 15 Oct 2017 10:56:56 +0000 (12:56 +0200)]
all: remove build_host and build_date to enable reproducible building
The hostname and date/time varies from system to system, so it is
not possible to reproduce the binary when this data is embedded.
Signed-off-by: Reiner Herrmann <reiner@reiner-h.de>
Signed-off-by: Ferry Huberts <ferry.huberts@pelagic.nl>
Reiner Herrmann [Sun, 15 Oct 2017 10:43:14 +0000 (12:43 +0200)]
all: make sure all make filelists are sorted
Originally by Reiner, adjusted by Ferry to sort all filelists.
Signed-off-by: Reiner Herrmann <reiner@reiner-h.de>
Signed-off-by: Ferry Huberts <ferry.huberts@pelagic.nl>
Ferry Huberts [Tue, 30 Jan 2018 20:03:31 +0000 (21:03 +0100)]
Merge pull request #49 from wiedi/osx-clock-10.12
mach: clock_gettime and friends appeared in Mac OSX 10.12
Sebastian Wiedenroth [Tue, 30 Jan 2018 19:25:14 +0000 (20:25 +0100)]
mach: clock_gettime and friends appeared in Mac OSX 10.12
Signed-off-by: Sebastian Wiedenroth <wiedi@frubar.net>
Ferry Huberts [Wed, 6 Dec 2017 17:03:40 +0000 (18:03 +0100)]
info: fix the previous commit: windows doesn't support MSG_DONTWAIT
Signed-off-by: Ferry Huberts <ferry.huberts@pelagic.nl>
Henning Rogge [Wed, 6 Dec 2017 16:44:17 +0000 (17:44 +0100)]
info: do not block when sending data
Even though checks are done that we can actually send, the send
can still block.
Reported-By: Joe Ayers <joe@ayerscasa.com>
Signed-off-by: Henning Rogge <hrogge@gmail.com>
Signed-off-by: Ferry Huberts <ferry.huberts@pelagic.nl>
Ferry Huberts [Wed, 6 Dec 2017 16:35:33 +0000 (17:35 +0100)]
info: java: upgrade gradle wrapper
The magical incantation is:
gradle wrapper --gradle-version 4.4
Signed-off-by: Ferry Huberts <ferry.huberts@pelagic.nl>
Ferry Huberts [Tue, 21 Nov 2017 09:07:08 +0000 (10:07 +0100)]
process_routes: fix potential null pointer dereference warnings on gcc 7
Signed-off-by: Ferry Huberts <ferry.huberts@pelagic.nl>
Ferry Huberts [Tue, 21 Nov 2017 08:43:09 +0000 (09:43 +0100)]
src: fix format truncation errors on gcc 7
Signed-off-by: Ferry Huberts <ferry.huberts@pelagic.nl>
Ferry Huberts [Tue, 21 Nov 2017 08:37:11 +0000 (09:37 +0100)]
olsr_spf: fix potential null pointer dereference warnings on gcc 7
Signed-off-by: Ferry Huberts <ferry.huberts@pelagic.nl>
Ferry Huberts [Tue, 21 Nov 2017 08:36:03 +0000 (09:36 +0100)]
nameservice: fix potential null pointer dereference warnings on gcc 7
Signed-off-by: Ferry Huberts <ferry.huberts@pelagic.nl>
Ferry Huberts [Tue, 21 Nov 2017 07:05:59 +0000 (08:05 +0100)]
all: link with -lm where <math.h> is included.
Otherwise we'll get crashes on OpenWRT
Signed-off-by: Ferry Huberts <ferry.huberts@pelagic.nl>
Ferry Huberts [Tue, 21 Nov 2017 07:01:53 +0000 (08:01 +0100)]
hashing: fix implicit fall-through warning on gcc 7
Signed-off-by: Ferry Huberts <ferry.huberts@pelagic.nl>
Ferry Huberts [Sun, 5 Nov 2017 14:38:21 +0000 (15:38 +0100)]
info: java: use Java 1.8
Signed-off-by: Ferry Huberts <ferry.huberts@pelagic.nl>
Ferry Huberts [Sun, 5 Nov 2017 14:36:01 +0000 (15:36 +0100)]
info: java: upgrade gradle wrapper
The magical incantation is:
gradle wrapper --gradle-version 4.3
Signed-off-by: Ferry Huberts <ferry.huberts@pelagic.nl>
Ferry Huberts [Wed, 9 Aug 2017 07:16:24 +0000 (09:16 +0200)]
info: java: upgrade gradle wrapper
The magical incantation is:
gradle wrapper --gradle-version 4.1
Signed-off-by: Ferry Huberts <ferry.huberts@pelagic.nl>
Ferry Huberts [Thu, 27 Jul 2017 12:32:44 +0000 (14:32 +0200)]
info: java: upgrade workspace
Signed-off-by: Ferry Huberts <ferry.huberts@pelagic.nl>
Ferry Huberts [Mon, 3 Jul 2017 21:04:59 +0000 (23:04 +0200)]
Merge pull request #32 from stspdotname/fix-sendto-ipv4
Fix olsr_sendto() for IPv4 on BSD.
Stefan Sperling [Thu, 22 Jun 2017 16:28:55 +0000 (18:28 +0200)]
Fix olsr_sendto() for IPv4 on BSD.
Sending multicast packets requires the IP_MULTICAST_IF socket option.
Without this socket option, packets sent to 255.255.255.255 are dropped.
With this change, olsr_sendto() can become a direct call to sendto() and
the -DSPOOF workaround which relies on libnet to send packets is no
longer required.
Tested on OpenBSD.
Signed-off-by: Stefan Sperling <stsp@stsp.name>
Ferry Huberts [Mon, 3 Jul 2017 20:19:44 +0000 (22:19 +0200)]
Merge pull request #37 from stspdotname/bcast-addr-len-and-type
Fix initialization of a struct sockaddr_in.
Stefan Sperling [Mon, 3 Jul 2017 20:08:41 +0000 (22:08 +0200)]
Fix initialization of a struct sockaddr_in.
While reading an IPv4 broadcast address from the configuration file,
do not forget about initializing the family sockaddr_in struct field.
Otherwise the sockaddr_in struct is invalid, and a BSD kernel will reject it.
Signed-off-by: Stefan Sperling <stsp@stsp.name>
Ferry Huberts [Mon, 3 Jul 2017 13:01:24 +0000 (15:01 +0200)]
info: java: fix tests by letting mock bind to localhost
Signed-off-by: Ferry Huberts <ferry.huberts@pelagic.nl>
Ferry Huberts [Fri, 30 Jun 2017 06:58:47 +0000 (08:58 +0200)]
Merge pull request #31 from stspdotname/add-ipv6-routes-correctly
Fix adding IPv6 routes on BSD.
Ferry Huberts [Fri, 30 Jun 2017 06:33:30 +0000 (08:33 +0200)]
Merge pull request #35 from stspdotname/icmp6-rediraccept
Do not toggle net.inet6.icmp6.rediraccept on OpenBSD.
Stefan Sperling [Thu, 22 Jun 2017 16:18:26 +0000 (18:18 +0200)]
Fix adding IPv6 routes on BSD.
The KAME hack (storing the interface index inside the prefix) is only used
for link-local addresses. So only apply it in case the gateway is link-local.
For gateway routes (reachable on-link), install a cloning route which tells
the kernel to perform NDP address resolution to find the gateway.
The previous code was installing a bogus route to fe:80::1 for any destination.
I don't see how this could ever have worked.
Tested in Berlin's Freifunk mesh.
Signed-off-by: Stefan Sperling <stsp@stsp.name>
Ferry Huberts [Fri, 23 Jun 2017 08:16:22 +0000 (10:16 +0200)]
Merge pull request #36 from stspdotname/rtm-delete-flags-v6
Use correct RTM flags while deleting an IPv6 route on BSD.
Stefan Sperling [Fri, 23 Jun 2017 08:02:55 +0000 (10:02 +0200)]
Use correct RTM flags while deleting an IPv6 route on BSD.
Same problem as fixed for IPv4 in
65978bd6996c8f62ac3f0adf971692fc8f2dc9cb.
I forgot to include the IPv6 code path in that commit.
Signed-off-by: Stefan Sperling <stsp@stsp.name>
Ferry Huberts [Thu, 22 Jun 2017 20:29:57 +0000 (22:29 +0200)]
Merge pull request #34 from stspdotname/rtm-delete-flags
Use correct RTM flags while deleting a route on BSD.
Ferry Huberts [Thu, 22 Jun 2017 20:18:27 +0000 (22:18 +0200)]
Merge pull request #25 from neeels/master
fix segfault when printing "You must be root"
Ferry Huberts [Thu, 22 Jun 2017 20:15:24 +0000 (22:15 +0200)]
Merge pull request #26 from neeels/segfault_cleanup
fix: don't attempt to cleanup uninitialized multiGwTunnels
Neels Hofmeyr [Fri, 16 Jun 2017 01:37:25 +0000 (03:37 +0200)]
fix: don't attempt to cleanup uninitialized multiGwTunnels
If olsr_cnf->smart_gw_use_count is 1, no multi-gw tunnels are initialized, and
the sgwTunnel4InterfaceNames and sgwTunnel6InterfaceNames are actually NULL. Do
not attempt to cleanup those if NULL.
Symptom: stopping olsrd e.g. via SIGINT leads to a segfault during cleanup.
The reason is that smart_gw_use_count is 1, which attempts to then dereference
one of sgwTunnel{4,6}InterfaceNames[0], amounting to <NULL>[0].
Signed-off-by: Neels Hofmeyr <neels@hofmeyr.de>
Neels Hofmeyr [Thu, 15 Jun 2017 21:36:37 +0000 (23:36 +0200)]
fix segfault when printing "You must be root"
When olsr_cnf is NULL, always print OLSR_PRINTF messages to avoid segfault.
I compiled olsrd and ran ./olsrd --help, resulting in a segfault.
That's because main.c prints an error message even before the config is loaded,
so that OLSR_PRINTF tries to lookup the logging level in a NULL olsr_cnf.
Signed-off-by: Neels Hofmeyr <neels@hofmeyr.de>
Stefan Sperling [Thu, 22 Jun 2017 17:09:19 +0000 (19:09 +0200)]
Do not toggle net.inet6.icmp6.rediraccept on OpenBSD.
OpenBSD no longer supports the net.inet6.icmp6.rediraccept sysctl.
It was removed in 2014: http://marc.info/?l=openbsd-cvs&m=
140914827732101&w=2
Signed-off-by: Stefan Sperling <stsp@stsp.name>
Stefan Sperling [Thu, 22 Jun 2017 16:57:08 +0000 (18:57 +0200)]
Use correct RTM flags while deleting a route on BSD.
If a route to a given destination already exists, the add_del_route()
function attempts to delete this route before inserting a new one.
While doing so, it asks olsr_rt_flags() for route message flags to use.
It should ask for route message flags that apply to a deletion. However,
it only did so if the 'add' parameter is false. This looks like a case of
badly copy-pasted code, since this section of code is obviously trying
to delete a route regardless of the value of 'add'.
Signed-off-by: Stefan Sperling <stsp@stsp.name>
Ferry Huberts [Thu, 22 Jun 2017 16:18:00 +0000 (18:18 +0200)]
Merge pull request #30 from stspdotname/route-deletion-needs-netmask
Fix route deletion on BSD.
Stefan Sperling [Thu, 22 Jun 2017 16:11:17 +0000 (18:11 +0200)]
Fix route deletion on BSD.
Specify a netmask when deleting routes. If no netmask is passed
in the routing message then deletion of the route can fail.
This fix covers both IPv4 and IPv6 routes.
Tested on OpenBSD only, but this change should not hurt other BSDs.
Signed-off-by: Stefan Sperling <stsp@stsp.name>
Ferry Huberts [Thu, 22 Jun 2017 16:07:29 +0000 (18:07 +0200)]
Merge pull request #29 from stspdotname/detect-wifi-on-bsd
Implement a simpler way of detecting wifi interfaces on BSD.
Ferry Huberts [Thu, 22 Jun 2017 16:06:59 +0000 (18:06 +0200)]
Merge pull request #28 from stspdotname/missing-olsr-random-include
Add a missing include for olsr_random() on BSD.
Stefan Sperling [Mon, 12 Jun 2017 18:58:56 +0000 (20:58 +0200)]
Implement a simpler way of detecting wifi interfaces on BSD.
The media type of an interface can be used to tell wireless interfaces
apart. The SIOCGIFMEDIA ioctl works the same way across all BSDs,
whereas the net80211 ioctls used previously differ.
Signed-off-by: Stefan Sperling <stsp@stsp.name>
Stefan Sperling [Mon, 12 Jun 2017 18:58:22 +0000 (20:58 +0200)]
Add a missing include for olsr_random() on BSD.
Signed-off-by: Stefan Sperling <stsp@stsp.name>
Ferry Huberts [Fri, 16 Jun 2017 18:31:56 +0000 (20:31 +0200)]
info: java: upgrade gradle wrapper
The magical incantation is:
gradle wrapper --gradle-version 4.0
Signed-off-by: Ferry Huberts <ferry.huberts@pelagic.nl>
Ferry Huberts [Tue, 6 Jun 2017 10:20:22 +0000 (11:20 +0100)]
Merge branch 'release-0.9.6'
* release-0.9.6:
Update version after release of v0.9.6.2
Release v0.9.6.2
Signed-off-by: Ferry Huberts <ferry.huberts@pelagic.nl>
Ferry Huberts [Tue, 6 Jun 2017 10:17:32 +0000 (11:17 +0100)]
Update version after release of v0.9.6.2
Signed-off-by: Ferry Huberts <ferry.huberts@pelagic.nl>
Ferry Huberts [Tue, 6 Jun 2017 10:17:26 +0000 (11:17 +0100)]
Release v0.9.6.2
Signed-off-by: Ferry Huberts <ferry.huberts@pelagic.nl>
Ferry Huberts [Mon, 5 Jun 2017 22:31:15 +0000 (00:31 +0200)]
Merge branch 'release-0.9.6'
* release-0.9.6:
link_set: fix typo in
f8d4e24
Ferry Huberts [Mon, 5 Jun 2017 22:21:28 +0000 (00:21 +0200)]
link_set: fix typo in
f8d4e24
This broke LinkQualityMult
Signed-off-by: Ferry Huberts <ferry.huberts@pelagic.nl>
Ferry Huberts [Mon, 5 Jun 2017 22:21:28 +0000 (00:21 +0200)]
link_set: fix typo in
f8d4e24
This broke LinkQualityMult
Signed-off-by: Ferry Huberts <ferry.huberts@pelagic.nl>
Ferry Huberts [Mon, 29 May 2017 09:41:26 +0000 (11:41 +0200)]
pud: only check gpsd status; no clear of info on bad fix or invalid signal
Signed-off-by: Ferry Huberts <ferry.huberts@pelagic.nl>
Ferry Huberts [Mon, 29 May 2017 09:42:57 +0000 (11:42 +0200)]
info: java: gradle: upgrade wrapper
The magical incantation is:
gradle wrapper --gradle-version 3.5
Signed-off-by: Ferry Huberts <ferry.huberts@pelagic.nl>
Ferry Huberts [Mon, 6 Mar 2017 21:59:30 +0000 (22:59 +0100)]
info: java: gradle: upgrade wrapper
The magical incantation is:
gradle wrapper --gradle-version 3.4.1
Signed-off-by: Ferry Huberts <ferry.huberts@pelagic.nl>
Ferry Huberts [Thu, 23 Feb 2017 17:58:43 +0000 (18:58 +0100)]
info: java: update jackson bundles
Signed-off-by: Ferry Huberts <ferry.huberts@pelagic.nl>
Ferry Huberts [Thu, 23 Feb 2017 17:58:15 +0000 (18:58 +0100)]
info: java: update workspace
Signed-off-by: Ferry Huberts <ferry.huberts@pelagic.nl>
Ferry Huberts [Mon, 20 Feb 2017 21:53:21 +0000 (22:53 +0100)]
info: java: gradle: upgrade wrapper
The magical incantation is:
gradle wrapper --gradle-version 3.4
Signed-off-by: Ferry Huberts <ferry.huberts@pelagic.nl>
Ferry Huberts [Fri, 10 Feb 2017 10:35:24 +0000 (11:35 +0100)]
Merge branch 'release-0.9.6'
* release-0.9.6:
Update version after release of v0.9.6.1
Release v0.9.6.1
Signed-off-by: Ferry Huberts <ferry.huberts@pelagic.nl>
Henning Rogge [Fri, 10 Feb 2017 10:20:16 +0000 (11:20 +0100)]
Update version after release of v0.9.6.1
Signed-off-by: Henning Rogge <hrogge@gmail.com>
Henning Rogge [Fri, 10 Feb 2017 10:20:11 +0000 (11:20 +0100)]
Release v0.9.6.1
Signed-off-by: Henning Rogge <hrogge@gmail.com>
Ferry Huberts [Wed, 8 Feb 2017 11:40:44 +0000 (12:40 +0100)]
Merge branch 'release-0.9.6'
* release-0.9.6:
hello reception: only remove UNSPEC duplicates
hello reception: remove duplicate neighbors
lq_packet: do not report duplicate neighbours in HELLO messages
hello reception: use link status ordering
lq_packet: make the hello link order a define
info: automatically detect whether the reply should have HTTP headers
info: java: upgrade gradle wrapper
Signed-off-by: Ferry Huberts <ferry.huberts@pelagic.nl>
Ferry Huberts [Fri, 3 Feb 2017 13:45:09 +0000 (14:45 +0100)]
hello reception: only remove UNSPEC duplicates
Only UNSPEC links can be duplicates, so only try to remove those.
This is quite a bit faster at the cost of some complexity.
Signed-off-by: Ferry Huberts <ferry.huberts@pelagic.nl>
Iwan G. Flameling [Fri, 3 Feb 2017 13:33:55 +0000 (14:33 +0100)]
hello reception: remove duplicate neighbors
See the explanations in the descriptions of commits
2f38613 and
5089a79.
To fix the issue more thoroughly we need to ensure that no duplicate
neighbours are sent or received because that confuses the receiver.
This is especially true when the HELLO messages are fragmented, and
even more so when the fragments have some delay between them since
that causes link instability.
This commit fixes that for the receiving side.
Signed-off-by: Iwan G. Flameling <iwanovich@gmail.com>
Signed-off-by: Ferry Huberts <ferry.huberts@pelagic.nl>
Iwan G. Flameling [Fri, 3 Feb 2017 13:17:51 +0000 (14:17 +0100)]
lq_packet: do not report duplicate neighbours in HELLO messages
See the explanations in the descriptions of commits
2f38613 and
5089a79.
To fix the issue more thoroughly we need to ensure that no duplicate
neighbours are sent or received because that confuses the receiver.
This is especially true when the HELLO messages are fragmented, and
even more so when the fragments have some delay between them since
that causes link instability.
This commit fixes that for the sending side.
Signed-off-by: Iwan G. Flameling <iwanovich@gmail.com>
Signed-off-by: Ferry Huberts <ferry.huberts@pelagic.nl>
Ferry Huberts [Mon, 30 Jan 2017 16:53:19 +0000 (17:53 +0100)]
hello reception: use link status ordering
See the explanation in the description of commit
2f38613.
This commit fixes the receive side of the described problem.
From this commit on 'broken' senders and 'fixed' receivers can
interoperate correctly.
The only case left unfixed is 'broken' senders and 'broken' receivers
(which is obviously unfixable), all other combinations are fixed.
'broken' =
v0.6.6.1-136-g3d2fd73 <= olsrd <
v0.9.5-82-g2f38613
'fixed' = olsrd <
v0.6.6.1-136-g3d2fd73
OR
olsrd >=
v0.9.5-82-g2f38613 (senders )
olsrd >= this commit (receivers)
Signed-off-by: Ferry Huberts <ferry.huberts@pelagic.nl>
Ferry Huberts [Mon, 30 Jan 2017 16:24:07 +0000 (17:24 +0100)]
lq_packet: make the hello link order a define
This will be re-used later
Signed-off-by: Ferry Huberts <ferry.huberts@pelagic.nl>
Ferry Huberts [Tue, 7 Feb 2017 12:28:06 +0000 (13:28 +0100)]
info: automatically detect whether the reply should have HTTP headers
This is the case when a HTTP request is done.
The request can still override whether or not HTTP headers are sent
by employing the 'http' and 'plain' request prefixes.
Signed-off-by: Ferry Huberts <ferry.huberts@pelagic.nl>
Ferry Huberts [Tue, 7 Feb 2017 12:28:06 +0000 (13:28 +0100)]
info: automatically detect whether the reply should have HTTP headers
This is the case when a HTTP request is done.
The request can still override whether or not HTTP headers are sent
by employing the 'http' and 'plain' request prefixes.
Signed-off-by: Ferry Huberts <ferry.huberts@pelagic.nl>
Ferry Huberts [Mon, 6 Feb 2017 15:55:17 +0000 (16:55 +0100)]
info: java: upgrade gradle wrapper
The magical incantation is:
gradle wrapper --gradle-version 3.3
Signed-off-by: Ferry Huberts <ferry.huberts@pelagic.nl>
Ferry Huberts [Mon, 6 Feb 2017 15:55:17 +0000 (16:55 +0100)]
info: java: upgrade gradle wrapper
The magical incantation is:
gradle wrapper --gradle-version 3.3
Signed-off-by: Ferry Huberts <ferry.huberts@pelagic.nl>
Ferry Huberts [Wed, 25 Jan 2017 06:45:02 +0000 (07:45 +0100)]
Merge branch 'release-0.9.6'
* release-0.9.6:
changelog: fixup
Update version after release of v0.9.6
Release v0.9.6
release: insert changelog just before the marker of the previous release
release: minor improvements to the release script
changelog: minor layout fixes
Signed-off-by: Ferry Huberts <ferry.huberts@pelagic.nl>
Ferry Huberts [Wed, 25 Jan 2017 06:41:08 +0000 (07:41 +0100)]
changelog: fixup
Signed-off-by: Ferry Huberts <ferry.huberts@pelagic.nl>
Henning Rogge [Tue, 24 Jan 2017 19:50:48 +0000 (20:50 +0100)]
Update version after release of v0.9.6
Signed-off-by: Henning Rogge <hrogge@gmail.com>
Henning Rogge [Tue, 24 Jan 2017 19:50:32 +0000 (20:50 +0100)]
Release v0.9.6
Signed-off-by: Henning Rogge <hrogge@gmail.com>
Ferry Huberts [Fri, 20 Jan 2017 16:20:23 +0000 (17:20 +0100)]
release: insert changelog just before the marker of the previous release
So that the notes that are added manually stay on top.
Signed-off-by: Ferry Huberts <ferry.huberts@pelagic.nl>
Ferry Huberts [Fri, 20 Jan 2017 15:28:28 +0000 (16:28 +0100)]
release: minor improvements to the release script
Signed-off-by: Ferry Huberts <ferry.huberts@pelagic.nl>
Ferry Huberts [Fri, 20 Jan 2017 15:30:57 +0000 (16:30 +0100)]
changelog: minor layout fixes
Signed-off-by: Ferry Huberts <ferry.huberts@pelagic.nl>
Ferry Huberts [Fri, 20 Jan 2017 12:34:45 +0000 (13:34 +0100)]
Update version after branch of v0.9.6
Signed-off-by: Ferry Huberts <ferry.huberts@pelagic.nl>
Ferry Huberts [Fri, 20 Jan 2017 12:34:45 +0000 (13:34 +0100)]
Branch v0.9.6
Signed-off-by: Ferry Huberts <ferry.huberts@pelagic.nl>
Ferry Huberts [Fri, 20 Jan 2017 12:15:56 +0000 (13:15 +0100)]
lq_packet: change link status ordering in hello messages
A while ago we noticed that neighbours of nodes with multiple interfaces
on the same medium report infinite costs on their links to those nodes.
This problem is 100% reproducible, and was traced down to
commit
3d2fd73a5528a9c7cdccd088f2dcca80b37e66b9
Author: Henning Rogge <henning.rogge@fkie.fraunhofer.de>
Date: Mon Mar 31 11:18:30 2014 +0200
Fix processing error for fragmented hellos
Below is the setup in which we noticed the problem.
wlan0 wlan0
172.31.175.97/16 172.31.175.61/16
(((*))) ------------------------------------------------- (((*)))
| |
| |
| |
____|___ 172.29.175.97/15 ________ 172.29.175.61/15 ____|____
| |-eth1.2580---------| |--------eth1.2580-| |
| Node 97 | | Switch | | Node 61 |
|_________|-eth2.2580---------|________| |_________|
172.28.175.97/15
In this setup node 97 will report normal link costs for its (wired) links
to node 61 (see the first table below), while node 61 will report infinite
link costs for both its (wired) links to node 97 (see the second table
below).
Table: Links (node 97)
Local IP Remote IP Hyst. LQ NLQ Cost
172.29.175.97 172.29.175.61 0.000 1.000 1.000 0.100
172.28.175.97 172.29.175.61 0.000 1.000 1.000 0.100
172.31.175.97 172.31.175.61 0.000 1.000 1.000 1.000
Table: Links (node 61)
Local IP Remote IP Hyst. LQ NLQ Cost
172.29.175.61 172.29.175.97 0.000 1.000 0.000 INFINITE
172.29.175.61 172.28.175.97 0.000 1.000 0.000 INFINITE
172.31.175.61 172.31.175.97 0.000 1.000 1.000 1.000
Checking the HELLO messages that are received on node 61 from node 97,
we see the following:
[node 61] # tcpdump -vni eth1.2580 udp port 698
tcpdump: listening on eth1.2580, link-type EN10MB (Ethernet), capture size 262144 bytes
06:21:23.528204 IP (tos 0xc0, ttl 1, id 42455, offset 0, flags [DF], proto UDP (17), length 80)
172.28.175.97.698 > 255.255.255.255.698: OLSRv4, seq 0xf7c0, length 52
Hello-LQ Message (0xc9), originator 172.31.175.97, ttl 1, hop 0
vtime 3.000s, msg-seq 0x533d, length 48
hello-time 1.000s, MPR willingness 3
link-type Symmetric, neighbor-type Symmetric, len 12
neighbor 172.29.175.61, link-quality 0.00%, neighbor-link-quality 0.00%
link-type Unspecified, neighbor-type Symmetric, len 20
neighbor 172.31.175.61, link-quality 0.00%, neighbor-link-quality 0.00%
neighbor 172.29.175.61, link-quality 0.00%, neighbor-link-quality 0.00%
Node 61 receives HELLO messages from node 97 that report (amongst others):
1- a SYMMETRIC link-type to node 61 (172.29.175.61)
2- an UNSPECIFIED link-type to node 61 (172.29.175.61)
Clearly, this is 'confusing' and the root cause of why node 61 reports
infinite costs for the links, as shown above.
On node 61 - the receiving side - an entry with link type SYMMETRIC is
processed, but then later overwritten with UNSPECIFIED from the duplicate
entry.
For now we fix this problem by letting the sending side order the link
type statuses differently. In a future follow-up patch we'll fix both
the sending and receiving sides more thoroughly.
Signed-off-by: Ferry Huberts <ferry.huberts@pelagic.nl>
Signed-off-by: Henning Rogge <henning.rogge@fkie.fraunhofer.de>
Ferry Huberts [Fri, 20 Jan 2017 12:33:31 +0000 (13:33 +0100)]
changelog: add a remark about updated plugin versions
Signed-off-by: Ferry Huberts <ferry.huberts@pelagic.nl>
Ferry Huberts [Mon, 16 Jan 2017 20:44:15 +0000 (21:44 +0100)]
txtinfo: bump version
Needed because of extensive rework, using the 'info' base code, which
added one or more parameters.
Signed-off-by: Ferry Huberts <ferry.huberts@pelagic.nl>
Ferry Huberts [Mon, 16 Jan 2017 20:47:29 +0000 (21:47 +0100)]
netjson: bump version
Needed because of extensive rework, using the 'info' base code, which
added one or more parameters.
Signed-off-by: Ferry Huberts <ferry.huberts@pelagic.nl>
Ferry Huberts [Mon, 16 Jan 2017 20:43:28 +0000 (21:43 +0100)]
jsoninfo: bump version
Needed because of extensive rework, using the 'info' base code, which
added one or more parameters.
Signed-off-by: Ferry Huberts <ferry.huberts@pelagic.nl>
Ferry Huberts [Mon, 16 Jan 2017 20:44:39 +0000 (21:44 +0100)]
pud: bump version
Needed because of extensive rework, using gpsd
Signed-off-by: Ferry Huberts <ferry.huberts@pelagic.nl>
Ferry Huberts [Mon, 16 Jan 2017 21:57:50 +0000 (22:57 +0100)]
nameservice: fix typos of filewrite-interval plugin parameter
Signed-off-by: Ferry Huberts <ferry.huberts@pelagic.nl>
Ferry Huberts [Mon, 16 Jan 2017 21:55:01 +0000 (22:55 +0100)]
nameservice: fix plugin versions
Signed-off-by: Ferry Huberts <ferry.huberts@pelagic.nl>
Ferry Huberts [Sat, 14 Jan 2017 13:23:15 +0000 (14:23 +0100)]
Merge pull request #18 from SvenRoederer/olsr-nameservice-version
nameservice: bump version to 0.4
Sven Roederer [Sat, 14 Jan 2017 03:46:52 +0000 (04:46 +0100)]
nameservice: bump version to 0.4
for the new parameter "filewrite-interval" we should bump the version, so init-scripts can give correct parameters.
Automatic generate config-files (like in OpenWrt from UCI-settings) have no knowledge if this parameter can be added to configfile when there is all the same version.
Signed-off-by: Sven Roederer <freifunk@it-solutions.geroedel.de>
Ferry Huberts [Fri, 13 Jan 2017 13:33:50 +0000 (14:33 +0100)]
Merge pull request #16 from SvenRoederer/rename_sighup-interval
nameservice: rename parameter "sighup_interval" to "filewrite_interval"
Sven Roederer [Thu, 12 Jan 2017 20:11:05 +0000 (21:11 +0100)]
nameservice: rename parameter "sighup_interval" to "filewrite_interval"
Playing around with the new option "sighup_interval" I found, that this
also controls the interval of calling the "name-change-script".
Checking the code points out that this is the maintimer for writing
any status-file to disk. So it's more consistent to change the parameter-
name to what it's really doing.
Also update the README to reference this parameter in the others
affected by it.
Signed-off-by: Sven Roederer <freifunk@it-solutions.geroedel.de>
Ferry Huberts [Tue, 10 Jan 2017 19:32:12 +0000 (20:32 +0100)]
logging: work around bogus Coverity warning
Signed-off-by: Ferry Huberts <ferry.huberts@pelagic.nl>
Ferry Huberts [Tue, 10 Jan 2017 14:36:45 +0000 (15:36 +0100)]
cfgparser: use superfasthash for the configuration checksum
Saves space, by request of LEDE/OpenWRT
Signed-off-by: Ferry Huberts <ferry.huberts@pelagic.nl>
Ferry Huberts [Tue, 10 Jan 2017 11:10:19 +0000 (12:10 +0100)]
cfgparser: pull in openssl sha256 code to avoid linking to it
Obsoletes/removes the NOCONFIGCHECKSUM build option
Signed-off-by: Ferry Huberts <ferry.huberts@pelagic.nl>
Ferry Huberts [Tue, 10 Jan 2017 09:48:43 +0000 (10:48 +0100)]
cfgparser: use correct SHA256_DIGEST_LENGTH
Not SHA_DIGEST_LENGTH, which is for SHA1 checksums
Signed-off-by: Ferry Huberts <ferry.huberts@pelagic.nl>
Ferry Huberts [Tue, 10 Jan 2017 11:01:24 +0000 (12:01 +0100)]
Merge pull request #14 from bittorf/master
logging: show interface name during blocked socket error
Bastian Bittorf [Tue, 10 Jan 2017 08:46:59 +0000 (09:46 +0100)]
logging: show interface name during blocked socket error
message was:
OLSR: sendto IPv4 Operation not permitted
message now:
OLSR: sendto IPv4 'Operation not permitted' on interface wlan1
(added also for the IPv6 case)
Signed-off-by: Bastian Bittorf <bittorf@bluebottle.com>
Ferry Huberts [Mon, 9 Jan 2017 19:45:34 +0000 (20:45 +0100)]
Merge pull request #12 from booo/sighup-interval
nameservice: add new sighup-interval parameter
Philipp Borgers [Mon, 9 Jan 2017 17:05:56 +0000 (18:05 +0100)]
nameservice: add new sighup-interval parameter
We add the 'sighup-interval' parameter to control the time interval
between sighups to the file specified by 'sighup-pid-file'.
Signed-off-by: Philipp Borgers <borgers@mi.fu-berlin.de>
Ferry Huberts [Fri, 6 Jan 2017 11:59:04 +0000 (12:59 +0100)]
files: update configuration generation script
Signed-off-by: Ferry Huberts <ferry.huberts@pelagic.nl>
Ferry Huberts [Fri, 6 Jan 2017 11:22:32 +0000 (12:22 +0100)]
files: get rid of net-tools usage
net-tools has been deprecated well over a decade now.
Signed-off-by: Ferry Huberts <ferry.huberts@pelagic.nl>
Ferry Huberts [Wed, 4 Jan 2017 17:12:35 +0000 (18:12 +0100)]
make: remove -Wmissing-include-dirs
It's causing grief for LEDE/OpenWRT and doesn't seem that useful in the
first place.
Signed-off-by: Ferry Huberts <ferry.huberts@pelagic.nl>
Ferry Huberts [Wed, 4 Jan 2017 14:12:50 +0000 (15:12 +0100)]
gui: clean-ups
Especially the gui olsr_protocol.h file was REALLY out of sync.
Signed-off-by: Ferry Huberts <ferry.huberts@pelagic.nl>
Thorsten Maerz [Sun, 18 Dec 2016 14:25:57 +0000 (15:25 +0100)]
make: strip CC
Windows can't handle CC with leading/trailing whitespace.
Fixes msys/mingw build on win32.
Signed-off-by: Thorsten Maerz <info@netztorte.de>
Signed-off-by: Ferry Huberts <ferry.huberts@pelagic.nl>