B.A.T.M.A.N May 2016

b.a.t.m.a.n@lists.open-mesh.org

29 participants
110 discussions

How to read batadv-vis jsondoc correctly
by Patrick Bosch 17 May '16

17 May '16

Hi all As the title suggest, I'm wondering how to read the jsondoc output correctly. It seems that every possible link is included in that output, but which one is the one that is used? If there is one link, it should be the one with the lowest metric, but what if a node has more than one link? For example the following setup: Node 1 | | Node 2 ------- Node 3 Here, node two would have two links. Now, if there would be some more nodes that are in the jsondoc output, how can I find out which are the links that are in use? Thanks in advance for the help! Kind regards Patrick

1 0

Re: [B.A.T.M.A.N.] Node movement
by Sven Eckelmann 16 May '16

16 May '16

On Monday 16 May 2016 19:05:26 Andrei Palade wrote: > I installed batman-adv on 4-Raspberry Pi Model 3, and I now want to > emulate an out of reach situation. Instead of physically moving the > PIs into another room, I want to use some mechanism that will do this > for me. I ran the following command on node D: > > ip6tables -A INPUT -m mac --mac-source XX:XX:XX:XX:XX:XX -j DROP This is the wrong layer. What you are now doing is to add a layer 3 firewall based on layer 2 addresses. The batman-adv stuff happens in completely different layers/headers because batman-adv encapsulates the traffic in its own headers. So batman-adv is below this IPv6 firewall and thus the firewall never sees the packet on the intermediate node. The mac address you used here is the one directly below the the IPv6 header and has nothing to do with batman-adv. And this address will not change when batman-adv forwards packets. So your rule will drop any IPv6 packet from XX:XX:XX:XX:XX:XX independent of the path it took to reach this node. But your rule doesn't block any batman-adv traffic (OGM, ELP, ...) because these are all send as ethernet frames (layer 2). But what you want in your test setup to filter these messages > Did I approached this correctly? Unfortunately, this is not correct. You have to modify the layer below batman-adv. So for example add an ebtables rule on a "wrapper" bridge only containing your original ethernet/wifi interface which is then added to batman-adv as interface (this may require some tinkering with the wifi driver to allow adding of a wifi interface to a bridge - search for IFF_DONT_BRIDGE). Or you should look if your ethernet/wifi driver allows mac filters. This is not the first time this was asked. Here some example search terms: * http://search.gmane.org/?query=ebtables&author=&group=gmane.org.freifunk.ba… * http://search.gmane.org/?query=force+topology&author=&group=gmane.org.freif… * http://search.gmane.org/?query=blocking+ogm&author=&group=gmane.org.freifun… You could also try to use netem + ingress TC filters. Ingress shaping is a little bit tricky but there are examples available which use ifb to map ingress to the ifb interface's egress (which then can be used like normally). * https://serverfault.com/questions/350023/tc-ingress-policing-and-ifb-mirror… * http://tdistler.com/2011/06/10/netem-wan-emulation-how-to-setup-a-netem-box * http://andthatsjazz.org/lartc/mac.html > Basically what I want to do is from > A, B, C, D, to be able to route packages from A to D through C, > instead of physically moving D out of reach of A. How do I do this, > and how can I check if it works? You have to check via `batctl o` or `batctl ping -R ...` to find out over which neighbor the target is reachable. Kind regards, Sven

1 0

Node movement
by Andrei Palade 16 May '16

16 May '16

I installed batman-adv on 4-Raspberry Pi Model 3, and I now want to emulate an out of reach situation. Instead of physically moving the PIs into another room, I want to use some mechanism that will do this for me. I ran the following command on node D: ip6tables -A INPUT -m mac --mac-source XX:XX:XX:XX:XX:XX -j DROP where the MAC source is the MAC address of wlan0 of node A. For instance, I have A, B, C, D each running batman-adv, I want to route all the traffic from node A to node D through node C. Now I can't login into D, and the route tables do not show any change (i.e., batctl o). Did I approached this correctly? Basically what I want to do is from A, B, C, D, to be able to route packages from A to D through C, instead of physically moving D out of reach of A. How do I do this, and how can I check if it works? -- Andrei

1 0

[B.A.T.M.A.N.] pull request [net]: batman-adv 20160515
by Antonio Quartulli 16 May '16

16 May '16

Hello David, although we are extremely late in the release cycle we have 4 fixes which would really be worth merging before releasing linux-4.6. As you can read in the git tag below, each of them can lead to a kernel crash or to an unstable system. We came up with several fixes after having tested our new B.A.T.M.A.N. V code at the Wireless Battle Mesh in Porto (PT) at the beginning of the month, however, what I am sending here is the minimum subset that we though being extremely important to avoid easy kernel crashes. The change footprint is also rather small. Please pull or let me know if you rather prefer to get this through net-next. If you decide to pull, you will hit some conflicts when merging net into net-next, but I can send you some instructions to ease the process. Thanks a lot! Antonio The following changes since commit b91506586206140154b0b44cccf88c8cc0a4dca5: Merge branch 'xgene-fixes' (2016-05-13 21:12:07 -0400) are available in the git repository at: git://git.open-mesh.org/linux-merge.git tags/batman-adv-fix-for-davem for you to fetch changes up to 6b892c1cb0805acee5d4ddd9e7878ed076c1b7c7: batman-adv: Fix refcnt leak in batadv_v_neigh_* (2016-05-14 15:51:39 +0800) ---------------------------------------------------------------- During the Wireless Battle Mesh v9 in Porto (PT) at the beginning of May, we managed to uncover and fix some important bugs in our new B.A.T.M.A.N. V algorithm. These are the most critical fixes we came up with aimed to avoid easy kernel crashes: - avoid potential crash due to NULL pointer dereference in B.A.T.M.A.N. V routine when a neigh_ifinfo object is not found, by Sven Eckelmann - avoid crash due to double kref_put on neigh_node object in B.A.T.M.A.N. V routine leading to use-after-free, by Sven Eckelmann (this crash can be always replicated) - avoid use-after-free of skb when counting outgoing bytes, by Florian Westphal - fix neigh_ifinfo object reference counting imbalance when using B.A.T.M.A.N. V, by Sven Eckelmann. Such imbalance may lead to the impossibility of releasing the related netdev object on shutdown. ---------------------------------------------------------------- Florian Westphal (1): batman-adv: fix skb deref after free Sven Eckelmann (3): batman-adv: Avoid nullptr derefence in batadv_v_neigh_is_sob batman-adv: Fix double neigh_node_put in batadv_v_ogm_route_update batman-adv: Fix refcnt leak in batadv_v_neigh_* net/batman-adv/bat_v.c | 30 ++++++++++++++++++++++++++---- net/batman-adv/bat_v_ogm.c | 4 +++- net/batman-adv/routing.c | 4 +++- 3 files changed, 32 insertions(+), 6 deletions(-)

2 5

[B.A.T.M.A.N.] [PATCH next] batman-adv: Fix build against recent Debian Stretch kernels
by Sven Eckelmann 15 May '16

15 May '16

The kernels for Debian stretch require some special CFLAGS settings which are only correctly defined when NOSTDINC_FLAGS is defined inside the execution of the Makefile via kbuild. But batman-adv sets it currently outside to insert compatibility include headers and compat-sources. This can be avoided by making the top Makefile kbuild compatible and redefining the NOSTDINC_FLAGS when kbuild include this Makefile. The actual build of the batman-adv module is then done by adding the subdirectory to obj-y. Signed-off-by: Sven Eckelmann <sven(a)narfation.org> --- Makefile | 10 ++++++---- 1 file changed, 6 insertions(+), 4 deletions(-) diff --git a/Makefile b/Makefile index 365a7cc..d662573 100644 --- a/Makefile +++ b/Makefile @@ -43,7 +43,7 @@ RM ?= rm -f REVISION= $(shell if [ -d "$(PWD)/.git" ]; then \ echo $$(git --git-dir="$(PWD)/.git" describe --always --dirty --match "v*" |sed 's/^v//' 2> /dev/null || echo "[unknown]"); \ fi) -export NOSTDINC_FLAGS := \ +NOSTDINC_FLAGS := \ -I$(PWD)/compat-include/ \ -I$(PWD)/include/ \ -include $(PWD)/compat.h \ @@ -54,11 +54,13 @@ NOSTDINC_FLAGS += -DBATADV_SOURCE_VERSION=\"$(REVISION)\" endif include $(PWD)/compat-sources/Makefile - +obj-y += net/batman-adv/ export batman-adv-y BUILD_FLAGS := \ - M=$(PWD)/net/batman-adv \ + M=$(PWD) \ + PWD=$(PWD) \ + REVISION=$(REVISION) \ CONFIG_BATMAN_ADV=m \ CONFIG_BATMAN_ADV_DEBUG=$(CONFIG_BATMAN_ADV_DEBUG) \ CONFIG_BATMAN_ADV_BLA=$(CONFIG_BATMAN_ADV_BLA) \ @@ -66,7 +68,7 @@ BUILD_FLAGS := \ CONFIG_BATMAN_ADV_NC=$(CONFIG_BATMAN_ADV_NC) \ CONFIG_BATMAN_ADV_MCAST=$(CONFIG_BATMAN_ADV_MCAST) \ CONFIG_BATMAN_ADV_BATMAN_V=$(CONFIG_BATMAN_ADV_BATMAN_V) \ - INSTALL_MOD_DIR=updates/net/batman-adv/ + INSTALL_MOD_DIR=updates/ all: config $(MAKE) -C $(KERNELPATH) $(BUILD_FLAGS) modules -- 2.8.1

1 0

[B.A.T.M.A.N.] [PATCH 1/7] batman-adv: remove useless inline attribute for sysfs helper function
by Antonio Quartulli 15 May '16

15 May '16

the compiler can optimize functions within the same C file and therefore there is no need to make it explicit. Remove the useless inline attribute for __batadv_store_uint_attr() Signed-off-by: Antonio Quartulli <a(a)unstable.cc> --- net/batman-adv/sysfs.c | 12 ++++++------ 1 file changed, 6 insertions(+), 6 deletions(-) diff --git a/net/batman-adv/sysfs.c b/net/batman-adv/sysfs.c index 414b207..b99ea80 100644 --- a/net/batman-adv/sysfs.c +++ b/net/batman-adv/sysfs.c @@ -389,12 +389,12 @@ static int batadv_store_uint_attr(const char *buff, size_t count, return count; } -static inline ssize_t -__batadv_store_uint_attr(const char *buff, size_t count, - int min, int max, - void (*post_func)(struct net_device *), - const struct attribute *attr, - atomic_t *attr_store, struct net_device *net_dev) +static ssize_t __batadv_store_uint_attr(const char *buff, size_t count, int min, + int max, + void (*post_func)(struct net_device *), + const struct attribute *attr, + atomic_t *attr_store, + struct net_device *net_dev) { int ret; -- 2.8.2

3 17

[B.A.T.M.A.N.] [PATCH 1/2] batman-adv: Include main.h in all files
by Sven Eckelmann 13 May '16

13 May '16

main.h includes statements which (re)define preprocessor variables which influence the compiled code. This makes it necessary to include it in all files. For example, it redefines pr_fmt used to the module as prefix for each pr_* message. Reported-by: Antonio Quartulli <a(a)unstable.cc> Signed-off-by: Sven Eckelmann <sven(a)narfation.org> --- net/batman-adv/bat_algo.h | 2 +- net/batman-adv/bat_v_ogm.h | 4 ++-- net/batman-adv/netlink.h | 2 ++ 3 files changed, 5 insertions(+), 3 deletions(-) diff --git a/net/batman-adv/bat_algo.h b/net/batman-adv/bat_algo.h index 03dafd3..b727762 100644 --- a/net/batman-adv/bat_algo.h +++ b/net/batman-adv/bat_algo.h @@ -18,7 +18,7 @@ #ifndef _NET_BATMAN_ADV_BAT_ALGO_H_ #define _NET_BATMAN_ADV_BAT_ALGO_H_ -struct batadv_priv; +#include "main.h" int batadv_iv_init(void); diff --git a/net/batman-adv/bat_v_ogm.h b/net/batman-adv/bat_v_ogm.h index d849c75..4c4d45c 100644 --- a/net/batman-adv/bat_v_ogm.h +++ b/net/batman-adv/bat_v_ogm.h @@ -18,10 +18,10 @@ #ifndef _BATMAN_ADV_BATADV_V_OGM_H_ #define _BATMAN_ADV_BATADV_V_OGM_H_ +#include "main.h" + #include <linux/types.h> -struct batadv_hard_iface; -struct batadv_priv; struct sk_buff; int batadv_v_ogm_init(struct batadv_priv *bat_priv); diff --git a/net/batman-adv/netlink.h b/net/batman-adv/netlink.h index fa152a8..39044cc 100644 --- a/net/batman-adv/netlink.h +++ b/net/batman-adv/netlink.h @@ -18,6 +18,8 @@ #ifndef _NET_BATMAN_ADV_NETLINK_H_ #define _NET_BATMAN_ADV_NETLINK_H_ +#include "main.h" + void batadv_netlink_register(void); void batadv_netlink_unregister(void); -- 2.8.1

1 1

[B.A.T.M.A.N.] [PATCH] batman-adv: Fix mcast_flags debugfs output
by Linus Lüssing 13 May '16

13 May '16

We recently switched to the atomic variants set_bit() / clear_bit() / test_bit() for originator capabilities. Unfortunately, this was missed for the new multicast debugfs entry, resulting in an empty multicast flags table. Fixing this by using test_bit(). Fixes: f1f0fbb2c956 ("batman-adv: Add debugfs table for mcast flags") Signed-off-by: Linus Lüssing <linus.luessing(a)c0d3.blue> --- net/batman-adv/multicast.c | 8 ++++---- 1 file changed, 4 insertions(+), 4 deletions(-) diff --git a/net/batman-adv/multicast.c b/net/batman-adv/multicast.c index abb12f1..2dd5aec 100644 --- a/net/batman-adv/multicast.c +++ b/net/batman-adv/multicast.c @@ -1205,12 +1205,12 @@ int batadv_mcast_flags_seq_print_text(struct seq_file *seq, void *offset) rcu_read_lock(); hlist_for_each_entry_rcu(orig_node, head, hash_entry) { - if (!(orig_node->capa_initialized & - BATADV_ORIG_CAPA_HAS_MCAST)) + if (!test_bit(BATADV_ORIG_CAPA_HAS_MCAST, + &orig_node->capa_initialized)) continue; - if (!(orig_node->capabilities & - BATADV_ORIG_CAPA_HAS_MCAST)) { + if (!test_bit(BATADV_ORIG_CAPA_HAS_MCAST, + &orig_node->capabilities)) { seq_printf(seq, "%pM -\n", orig_node->orig); continue; } -- 1.7.10.4

1 0

[B.A.T.M.A.N.] [PATCH v2 1/5] batman-adv: make the GW selection class algorithm specific
by Antonio Quartulli 13 May '16

13 May '16

The B.A.T.M.A.N. V algorithm uses a different metric compared to its predecessor and for this reason the logic used to compute the best Gateway is also changed. This means that the GW selection class fed to this logic has a semantics that depends on the algorithm being used. Make the parsing and printing routine of the GW selection class routing algorithm specific. Each algorithm can now parse (and print) this value independently. If no API is provided by any algorithm, the default is to use the current mechanism of considering such value like an integer between 1 and 255. Signed-off-by: Antonio Quartulli <a(a)unstable.cc> --- Changes from v1: - add missing includes - create non-anonymous struct to store GW specific calls (makes it easier to provide working kernel-doc) net/batman-adv/bat_v.c | 34 ++++++++++++++++++++++++++++++++++ net/batman-adv/sysfs.c | 36 ++++++++++++++++++++++++++++++++++-- net/batman-adv/types.h | 13 +++++++++++++ 3 files changed, 81 insertions(+), 2 deletions(-) diff --git a/net/batman-adv/bat_v.c b/net/batman-adv/bat_v.c index c16cd44..e4eb221 100644 --- a/net/batman-adv/bat_v.c +++ b/net/batman-adv/bat_v.c @@ -21,8 +21,10 @@ #include <linux/atomic.h> #include <linux/bug.h> #include <linux/cache.h> +#include <linux/errno.h> #include <linux/init.h> #include <linux/jiffies.h> +#include <linux/kernel.h> #include <linux/netdevice.h> #include <linux/rculist.h> #include <linux/rcupdate.h> @@ -33,6 +35,8 @@ #include "bat_v_elp.h" #include "bat_v_ogm.h" +#include "gateway_client.h" +#include "gateway_common.h" #include "hard-interface.h" #include "hash.h" #include "originator.h" @@ -324,6 +328,32 @@ err_ifinfo1: return ret; } +static ssize_t batadv_v_store_sel_class(struct batadv_priv *bat_priv, + char *buff, size_t count) +{ + u32 old_class, class; + + if (!batadv_parse_throughput(bat_priv->soft_iface, buff, + "B.A.T.M.A.N. V GW selection class", + &class)) + return -EINVAL; + + old_class = atomic_read(&bat_priv->gw.sel_class); + atomic_set(&bat_priv->gw.sel_class, class); + + if (old_class != class) + batadv_gw_reselect(bat_priv); + + return count; +} + +static ssize_t batadv_v_show_sel_class(struct batadv_priv *bat_priv, char *buff) +{ + u32 class = atomic_read(&bat_priv->gw.sel_class); + + return sprintf(buff, "%u.%u MBit\n", class / 10, class % 10); +} + static struct batadv_algo_ops batadv_batman_v __read_mostly = { .name = "BATMAN_V", .bat_iface_activate = batadv_v_iface_activate, @@ -336,6 +366,10 @@ static struct batadv_algo_ops batadv_batman_v __read_mostly = { .bat_neigh_cmp = batadv_v_neigh_cmp, .bat_neigh_is_similar_or_better = batadv_v_neigh_is_sob, .bat_neigh_print = batadv_v_neigh_print, + .gw = { + .bat_store_sel_class = batadv_v_store_sel_class, + .bat_show_sel_class = batadv_v_show_sel_class, + }, }; /** diff --git a/net/batman-adv/sysfs.c b/net/batman-adv/sysfs.c index 233abcf..ad37cb0 100644 --- a/net/batman-adv/sysfs.c +++ b/net/batman-adv/sysfs.c @@ -513,6 +513,38 @@ static ssize_t batadv_store_gw_mode(struct kobject *kobj, return count; } +static ssize_t batadv_show_gw_sel_class(struct kobject *kobj, + struct attribute *attr, char *buff) +{ + struct batadv_priv *bat_priv = batadv_kobj_to_batpriv(kobj); + + if (bat_priv->bat_algo_ops->gw.bat_show_sel_class) + return bat_priv->bat_algo_ops->gw.bat_show_sel_class(bat_priv, + buff); + + return sprintf(buff, "%i\n", atomic_read(&bat_priv->gw.sel_class)); +} + +static ssize_t batadv_store_gw_sel_class(struct kobject *kobj, + struct attribute *attr, char *buff, + size_t count) +{ + struct batadv_priv *bat_priv = batadv_kobj_to_batpriv(kobj); + + if (buff[count - 1] == '\n') + buff[count - 1] = '\0'; + + if (bat_priv->bat_algo_ops->gw.bat_store_sel_class) + return bat_priv->bat_algo_ops->gw.bat_store_sel_class(bat_priv, + buff, + count); + + return __batadv_store_uint_attr(buff, count, 0, BATADV_TQ_MAX_VALUE, + batadv_post_gw_reselect, attr, + &bat_priv->gw.sel_class, + bat_priv->soft_iface); +} + static ssize_t batadv_show_gw_bwidth(struct kobject *kobj, struct attribute *attr, char *buff) { @@ -624,8 +656,8 @@ BATADV_ATTR_SIF_UINT(orig_interval, orig_interval, S_IRUGO | S_IWUSR, 2 * BATADV_JITTER, INT_MAX, NULL); BATADV_ATTR_SIF_UINT(hop_penalty, hop_penalty, S_IRUGO | S_IWUSR, 0, BATADV_TQ_MAX_VALUE, NULL); -BATADV_ATTR_SIF_UINT(gw_sel_class, gw.sel_class, S_IRUGO | S_IWUSR, 1, - BATADV_TQ_MAX_VALUE, batadv_post_gw_reselect); +static BATADV_ATTR(gw_sel_class, S_IRUGO | S_IWUSR, batadv_show_gw_sel_class, + batadv_store_gw_sel_class); static BATADV_ATTR(gw_bandwidth, S_IRUGO | S_IWUSR, batadv_show_gw_bwidth, batadv_store_gw_bwidth); #ifdef CONFIG_BATMAN_ADV_MCAST diff --git a/net/batman-adv/types.h b/net/batman-adv/types.h index 32c6d0e..f4fdf08 100644 --- a/net/batman-adv/types.h +++ b/net/batman-adv/types.h @@ -1259,6 +1259,17 @@ struct batadv_forw_packet { }; /** + * struct batadv_algo_gw_ops - mesh algorithm callbacks (GW specific) + * @bat_store_sel_class: parse and stores a new GW selection class + * @bat_show_sel_class: prints the current GW selection class + */ +struct batadv_algo_gw_ops { + ssize_t (*bat_store_sel_class)(struct batadv_priv *bat_priv, char *buff, + size_t count); + ssize_t (*bat_show_sel_class)(struct batadv_priv *bat_priv, char *buff); +}; + +/** * struct batadv_algo_ops - mesh algorithm callbacks * @list: list node for the batadv_algo_list * @name: name of the algorithm @@ -1283,6 +1294,7 @@ struct batadv_forw_packet { * the orig_node due to a new hard-interface being added into the mesh * @bat_orig_del_if: ask the routing algorithm to apply the needed changes to * the orig_node due to an hard-interface being removed from the mesh + * @gw: callbacks related to GW mode */ struct batadv_algo_ops { struct hlist_node list; @@ -1312,6 +1324,7 @@ struct batadv_algo_ops { int max_if_num); int (*bat_orig_del_if)(struct batadv_orig_node *orig_node, int max_if_num, int del_if_num); + struct batadv_algo_gw_ops gw; }; /** -- 2.8.2

2 10

[B.A.T.M.A.N.] [PATCH 1/2] batman-adv: remove forward declaration by including proper header
by Antonio Quartulli 12 May '16

12 May '16

If main.h is included, the forward declaration for struct batadv_priv is not required. Cc: Sven Eckelmann <sven(a)narfation.org> Signed-off-by: Antonio Quartulli <a(a)unstable.cc> --- Sven, is there any special region for not having the include in this file ? It seems to be compiling just fine. Cheers, net/batman-adv/bat_algo.h | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/net/batman-adv/bat_algo.h b/net/batman-adv/bat_algo.h index 03dafd3..b727762 100644 --- a/net/batman-adv/bat_algo.h +++ b/net/batman-adv/bat_algo.h @@ -18,7 +18,7 @@ #ifndef _NET_BATMAN_ADV_BAT_ALGO_H_ #define _NET_BATMAN_ADV_BAT_ALGO_H_ -struct batadv_priv; +#include "main.h" int batadv_iv_init(void); -- 2.8.2

2 8

← Newer
1
2
3
4
5
6
7
8
9
10
11
Older →

2024

2023

2022

2021

2020

2019

2018

2017

2016

2015

2014

2013

2012

2011

2010

2009

2008

2007

B.A.T.M.A.N May 2016