B.A.T.M.A.N March 2014

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

19 participants
50 discussions

Re: [B.A.T.M.A.N.] batman over wireless bridges
by Antonio Quartulli 11 Mar '14

11 Mar '14

(Forwarding to the ml. please keep it in the CC) On 11/03/14 22:57, Tom Sparks wrote: > > > > >> On Wednesday, 12 March 2014 7:29 AM, Antonio Quartulli <antonio(a)meshcoding.com> wrote: > >>> Hi Tom, >> >> >> On 11/03/14 13:20, Tom Sparks wrote: >>> >>> >>> I am wondering how well batman works over wireless bridges >>> like this: >>> >>> batR <--> wifiB <--> wifiB <--> batR >>> >>> >>> meanings: >>> BatR: openwrt router with batman installed >>> wifiB: wifi bridge >> >> >> What is a wifi bridge? > > Basically they are two wireless AP that communicate with each only >> How is it configured? > I don't have them yet, I am looking at the bulletM from www.ubnt.com > >> >> batman-adv uses the MAC addresses in the 802.11 header to understand >> which node is a direct neighbour, therefore the bridge in the middle may >> screw up this mechanism. It really depends on how the bridge is >> implemented (it is not a standard component). >> >> >> Cheers, >> >> -- >> Antonio Quartulli >> > -- Antonio Quartulli

1 0

[B.A.T.M.A.N.] [PATCHv2 1/2] batctl: check for batman interface
by Simon Wunderlich 11 Mar '14

11 Mar '14

From: Simon Wunderlich <simon(a)open-mesh.com> Check if the interface supplied by -m (or bat0 by default) is actually a batman interface. The check is performed by testing if the sysfs path is available. This change is required since debugfs now contains originator tables per hard interface. This might be confusing for users as batctl -m eth0 o will now (accidently) print these tables without eth0 beeing a batman interface, assuming eth0 is a hard interface used by batman-adv. Some reordering of commands were required to first allow commands which don't depend on a (correctly set up) mesh interface, like bisect or interface. Signed-off-by: Simon Wunderlich <simon(a)open-mesh.com> --- main.c | 32 +++++++++++++++++--------------- sys.c | 30 ++++++++++++++++++++++++++++++ sys.h | 1 + 3 files changed, 48 insertions(+), 15 deletions(-) diff --git a/main.c b/main.c index a37c0b7..44cf49b 100644 --- a/main.c +++ b/main.c @@ -137,7 +137,23 @@ int main(int argc, char **argv) exit(EXIT_FAILURE); } - if ((strcmp(argv[1], "ping") == 0) || (strcmp(argv[1], "p") == 0)) { + if ((strcmp(argv[1], "interface") == 0) || (strcmp(argv[1], "if") == 0)) { + + ret = interface(mesh_iface, argc - 1, argv + 1); + + } else if ((strcmp(argv[1], "tcpdump") == 0) || (strcmp(argv[1], "td") == 0)) { + + ret = tcpdump(argc - 1, argv + 1); + +#ifdef BATCTL_BISECT + } else if ((strcmp(argv[1], "bisect_iv") == 0)) { + + ret = bisect_iv(argc - 1, argv + 1); +#endif + } else if (check_mesh_iface(mesh_iface) < 0) { + fprintf(stderr, "Error - interface %s is not present or not a batman-adv interface\n", mesh_iface); + exit(EXIT_FAILURE); + } else if ((strcmp(argv[1], "ping") == 0) || (strcmp(argv[1], "p") == 0)) { ret = ping(mesh_iface, argc - 1, argv + 1); @@ -145,14 +161,6 @@ int main(int argc, char **argv) ret = traceroute(mesh_iface, argc - 1, argv + 1); - } else if ((strcmp(argv[1], "tcpdump") == 0) || (strcmp(argv[1], "td") == 0)) { - - ret = tcpdump(argc - 1, argv + 1); - - } else if ((strcmp(argv[1], "interface") == 0) || (strcmp(argv[1], "if") == 0)) { - - ret = interface(mesh_iface, argc - 1, argv + 1); - } else if ((strcmp(argv[1], "loglevel") == 0) || (strcmp(argv[1], "ll") == 0)) { ret = handle_loglevel(mesh_iface, argc - 1, argv + 1); @@ -179,12 +187,6 @@ int main(int argc, char **argv) ret = translate(mesh_iface, argc - 1, argv + 1); -#ifdef BATCTL_BISECT - } else if ((strcmp(argv[1], "bisect_iv") == 0)) { - - ret = bisect_iv(argc - 1, argv + 1); -#endif - } else { for (i = 0; i < BATCTL_SETTINGS_NUM; i++) { diff --git a/sys.c b/sys.c index fd6e107..8ba7d4f 100644 --- a/sys.c +++ b/sys.c @@ -566,3 +566,33 @@ out: free(path_buff); return res; } + +int check_mesh_iface(char *mesh_iface) +{ + char *base_dev = NULL; + char *check_iface; + char path_buff[PATH_BUFF_LEN]; + DIR *dir; + int ret = -1; + + /* use the parent interface if this is a VLAN */ + vlan_get_link(mesh_iface, &base_dev); + check_iface = mesh_iface; + if (base_dev) + check_iface = mesh_iface; + + /* try to open the mesh sys directory */ + snprintf(path_buff, PATH_BUFF_LEN, SYS_BATIF_PATH_FMT, check_iface); + dir = opendir(path_buff); + if (!dir) + goto out; + + closedir(dir); + + ret = 0; +out: + if (base_dev) + free(base_dev); + + return ret; +} diff --git a/sys.h b/sys.h index 5633822..5aec020 100644 --- a/sys.h +++ b/sys.h @@ -71,5 +71,6 @@ int interface(char *mesh_iface, int argc, char **argv); int handle_loglevel(char *mesh_iface, int argc, char **argv); int handle_sys_setting(char *mesh_iface, int setting, int argc, char **argv); int handle_gw_setting(char *mesh_iface, int argc, char **argv); +int check_mesh_iface(char *mesh_iface); #endif -- 1.7.10.4

1 1

[B.A.T.M.A.N.] [B.A.T.M.A.N.A.D.V] Deploy batman-adv using Station - Access Point mode?
by Le Tran Dat 11 Mar '14

11 Mar '14

Hi list, I want to setup a mesh network using batman-adv in which has around 4-5 Wireless Devices. In the past, I used to connect those devices via adhoc. However, due to a limitation of my Wireless Devices then I decided to give up adhoc mode. So I want to ask if anyone has experience with Batman-adv can give me a suggestion on how to deploy the mesh network without adhoc mode. For example: What I am trying to do right now is to connect those devices by using Station <-> Access Point mode. But I have no ideas how the OGM message transmit between the Mesh Node so I worry that the Station <-> Access Point can not take advance of batman-adv. Thanks, Dat

2 1

[B.A.T.M.A.N.] [PATCHv2 0/3] batctl: multicast related patches
by Linus Lüssing 11 Mar '14

11 Mar '14

The first two patches are unchanged. The third one made the -u and -m options exclusive and now uses the first bit of the first byte of the mac address to check whether it's a multicast address (instead of checing for IPv4/IPv6 multicast mac prefix), as Antonio suggested (thanks!). Cheers, Linus

2 5

[B.A.T.M.A.N.] [PATCH] batman-adv: improve DAT documentation
by Antonio Quartulli 11 Mar '14

11 Mar '14

From: Antonio Quartulli <antonio(a)open-mesh.com> Add missing documentation for BATADV_DAT_ADDR_MAX and convert an existing documentation to kerneldoc Signed-off-by: Antonio Quartulli <antonio(a)open-mesh.com> --- distributed-arp-table.h | 3 +++ types.h | 5 +++-- 2 files changed, 6 insertions(+), 2 deletions(-) diff --git a/distributed-arp-table.h b/distributed-arp-table.h index ac9be9b..d76e1d0 100644 --- a/distributed-arp-table.h +++ b/distributed-arp-table.h @@ -25,6 +25,9 @@ #include <linux/if_arp.h> +/** + * BATADV_DAT_ADDR_MAX - maximum address value in the DHT space + */ #define BATADV_DAT_ADDR_MAX ((batadv_dat_addr_t)~(batadv_dat_addr_t)0) void batadv_dat_status_update(struct net_device *net_dev); diff --git a/types.h b/types.h index d4b923c..34891a5 100644 --- a/types.h +++ b/types.h @@ -24,8 +24,9 @@ #ifdef CONFIG_BATMAN_ADV_DAT -/* batadv_dat_addr_t is the type used for all DHT addresses. If it is changed, - * BATADV_DAT_ADDR_MAX is changed as well. +/** + * batadv_dat_addr_t - it is the type used for all DHT addresses. If it is + * changed, BATADV_DAT_ADDR_MAX is changed as well. * * *Please be careful: batadv_dat_addr_t must be UNSIGNED* */ -- 1.8.4.5

2 1

[B.A.T.M.A.N.] [PATCH] batman-adv: improve the TT flags documentation
by Antonio Quartulli 11 Mar '14

11 Mar '14

From: Antonio Quartulli <antonio(a)open-mesh.com> Convert the current documentation for the TT flags in proper kerneldoc and improve it by adding an explanation for each of the flags. Signed-off-by: Antonio Quartulli <antonio(a)open-mesh.com> --- packet.h | 28 ++++++++++++++++++++++++---- 1 file changed, 24 insertions(+), 4 deletions(-) diff --git a/packet.h b/packet.h index c7f6eef..34e096d 100644 --- a/packet.h +++ b/packet.h @@ -119,10 +119,30 @@ enum batadv_tt_data_flags { BATADV_TT_FULL_TABLE = BIT(4), }; -/* BATADV_TT_CLIENT flags. - * Flags from BIT(0) to BIT(7) are sent on the wire, while flags from BIT(8) to - * BIT(15) are used for local computation only. - * Flags from BIT(4) to BIT(7) are kept in sync with the rest of the network. +/** + * enum batadv_tt_client_flags - TT client specific flags + * @BATADV_TT_CLIENT_DEL: the client has to be deleted from the table + * @BATADV_TT_CLIENT_ROAM: the client roamed to/from another node and the new + * update telling its new real location has not been received/sent yet + * @BATADV_TT_CLIENT_WIFI: this client is connected through a wifi interface. + * This information is used by the "AP Isolation" feature + * @BATADV_TT_CLIENT_ISOLA: this client is considered "isolated". This + * information is used by the Extended Isolation feature + * @BATADV_TT_CLIENT_NOPURGE: this client should never be removed from the table + * @BATADV_TT_CLIENT_NEW: this client has been added to the local table but has + * not been announced yet + * @BATADV_TT_CLIENT_PENDING: this client is marked for removal but it is kept + * in the table for one more originator interval for consistency purposes + * @BATADV_TT_CLIENT_TEMP: this global client has been detected to be part of + * the network but no nnode has already announced it + * + * Bits from 0 to 7 are called _remote flags_ because they are sent on the wire. + * Bits from 8 to 15 are called _local flags_ because they are used for local + * computations only. + * + * Bits from 4 to 7 - a subset of remote flags - are ensured to be in sync with + * the other nodes in the network. To achieve this goal these flags are included + * in the TT CRC computation. */ enum batadv_tt_client_flags { BATADV_TT_CLIENT_DEL = BIT(0), -- 1.8.4.5

2 1

[B.A.T.M.A.N.] [PATCH] batman-adv: properly align constant assignments
by Antonio Quartulli 11 Mar '14

11 Mar '14

All the assignments in packet.h are properly aligned. Also the multicast constants have to have a similar shape. Introduced by 1c090349e2f65d19a7f356cc0b61f13644c53bb6 ("batman-adv: Add IPv4 link-local/IPv6-ll-all-nodes multicast support") Signed-off-by: Antonio Quartulli <antonio(a)meshcoding.com> --- packet.h | 6 +++--- 1 file changed, 3 insertions(+), 3 deletions(-) diff --git a/packet.h b/packet.h index feaa336..c7f6eef 100644 --- a/packet.h +++ b/packet.h @@ -97,9 +97,9 @@ enum batadv_icmp_packettype { * @BATADV_MCAST_WANT_ALL_IPV6: we want all IPv6 multicast packets */ enum batadv_mcast_flags { - BATADV_MCAST_WANT_ALL_UNSNOOPABLES = BIT(0), - BATADV_MCAST_WANT_ALL_IPV4 = BIT(1), - BATADV_MCAST_WANT_ALL_IPV6 = BIT(2), + BATADV_MCAST_WANT_ALL_UNSNOOPABLES = BIT(0), + BATADV_MCAST_WANT_ALL_IPV4 = BIT(1), + BATADV_MCAST_WANT_ALL_IPV6 = BIT(2), }; /* tt data subtypes */ -- 1.8.4.5

2 1

[B.A.T.M.A.N.] [PATCHv2] batman-adv: use the name of the array as pointer
by Antonio Quartulli 11 Mar '14

11 Mar '14

Introduced by fb4756fe5f6e58ca2ab208ca634ae4c8dd6e9499 ("batman-adv: use ether_addr_copy instead of memcpy in multicast code") Signed-off-by: Antonio Quartulli <antonio(a)meshcoding.com> --- change from v1: - added Signed-off-by clause multicast.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/multicast.c b/multicast.c index e69143b..998b429 100644 --- a/multicast.c +++ b/multicast.c @@ -45,7 +45,7 @@ static int batadv_mcast_mla_softif_get(struct net_device *dev, break; } - ether_addr_copy(&new->addr, &mc_list_entry->addr); + ether_addr_copy(new->addr, mc_list_entry->addr); hlist_add_head(&new->list, mcast_list); ret++; } -- 1.8.4.5

2 1

[B.A.T.M.A.N.] [PATCHv3] batman-adv: use the proper location in ether_addr_copy()
by Antonio Quartulli 11 Mar '14

11 Mar '14

Introduced by: 513b2d107a4d88f62b9394926e377d2c6250775f ("batman-adv: prefer ether_addr_copy to memcpy") Reported-by: kbuild test robot <fengguang.wu(a)intel.com> Signed-off-by: Antonio Quartulli <antonio(a)meshcoding.com> --- Changes from v2: - added another fix in routing.c Changes from v1: - added Reported-by clause icmp_socket.c | 2 +- routing.c | 2 +- 2 files changed, 2 insertions(+), 2 deletions(-) diff --git a/icmp_socket.c b/icmp_socket.c index bf07dfd..161ef8f 100644 --- a/icmp_socket.c +++ b/icmp_socket.c @@ -230,7 +230,7 @@ static ssize_t batadv_socket_write(struct file *file, const char __user *buff, icmp_packet_rr = (struct batadv_icmp_packet_rr *)icmp_header; if (packet_len == sizeof(*icmp_packet_rr)) { addr = neigh_node->if_incoming->net_dev->dev_addr; - ether_addr_copy(icmp_packet_rr->rr, addr); + ether_addr_copy(icmp_packet_rr->rr[0], addr); } break; diff --git a/routing.c b/routing.c index 3855132..3514153 100644 --- a/routing.c +++ b/routing.c @@ -340,7 +340,7 @@ int batadv_recv_icmp_packet(struct sk_buff *skb, if (icmp_packet_rr->rr_cur >= BATADV_RR_LEN) goto out; - ether_addr_copy(&(icmp_packet_rr->rr[icmp_packet_rr->rr_cur]), + ether_addr_copy(icmp_packet_rr->rr[icmp_packet_rr->rr_cur], ethhdr->h_dest); icmp_packet_rr->rr_cur++; } -- 1.8.4.5

2 1

[B.A.T.M.A.N.] [PATCHv3] batctl: add isolation_mark support
by Antonio Quartulli 11 Mar '14

11 Mar '14

From: Antonio Quartulli <antonio(a)open-mesh.com> With this patch it is now possible to set/unset the value of the isolation_mark used by the Extended Isolation feature in batman-adv. Signed-off-by: Antonio Quartulli <antonio(a)open-mesh.com> --- Change from v2: - added section in README Change from v1: - added section in man page README | 13 +++++++++++++ man/batctl.8 | 20 ++++++++++++++++++++ sys.c | 9 +++++++-- sys.h | 1 + 4 files changed, 41 insertions(+), 2 deletions(-) diff --git a/README b/README index 5af95c8..a4507b7 100644 --- a/README +++ b/README @@ -399,6 +399,19 @@ display or modify the packet aggregation setting Usage: batctl aggregation|ag [0|1] +batctl isolation_mark +===================== + +display or modify the isolation mark. +This value is used by Extended Isolation feature. + +Usage: batctl isolation_mark|mark $value[/0x$mask] + +Example 1: batctl mark 0x00000001/0xffffffff +Example 2: batctl mark 0x00040000/0xffff0000 +Example 3: batctl mark 16 +Example 4: batctl mark 0x0f + batctl translocal ================= diff --git a/man/batctl.8 b/man/batctl.8 index 2fe5941..b808293 100644 --- a/man/batctl.8 +++ b/man/batctl.8 @@ -151,6 +151,26 @@ least XX TQ better than the currently selected gateway (XX has to be a number be .RE .RE .br +.IP "\fBisolation_mark\fP|\fBmark\fP" +If no parameter is given the current isolation mark value is displayed. +Otherwise the parameter is used to set or unset the isolation mark used by the +Extended Isolation feature. +.br +The input is supposed to be of the form $value/$mask, where $value can be any +32bit long integer (expressed in decimal or hex base) and $mask is a generic +bitmask (expressed in hex base) that selects the bits to take into consideration +from $value. It is also possible to enter the input using only $value and in +this case the full bitmask is used by default. + +.br +.br +Example 1: 0x00000001/0xffffffff +.br +Example 2: 0x00040000/0xffff0000 +.br +Example 3: 16 or 0x0F +.br +.br .PP .I \fBdebug tables: .IP diff --git a/sys.c b/sys.c index 5cebf52..fd6e107 100644 --- a/sys.c +++ b/sys.c @@ -96,8 +96,13 @@ const struct settings_data batctl_settings[BATCTL_SETTINGS_NUM] = { .opt_short = "nc", .sysfs_name = "network_coding", .params = sysfs_param_enable, - } - + }, + { + .opt_long = "isolation_mark", + .opt_short = "mark", + .sysfs_name = "isolation_mark", + .params = NULL, + }, }; static void interface_usage(void) diff --git a/sys.h b/sys.h index 2cbbcfb..5633822 100644 --- a/sys.h +++ b/sys.h @@ -46,6 +46,7 @@ enum batctl_settings_list { BATCTL_SETTINGS_BONDING, BATCTL_SETTINGS_FRAGMENTATION, BATCTL_SETTINGS_NETWORK_CODING, + BATCTL_SETTINGS_ISOLATION_MARK, BATCTL_SETTINGS_NUM, }; -- 1.8.4.5

2 1

2024

2023

2022

2021

2020

2019

2018

2017

2016

2015

2014

2013

2012

2011

2010

2009

2008

2007

B.A.T.M.A.N March 2014