B.A.T.M.A.N January 2011

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

20 participants
35 discussions

[B.A.T.M.A.N.] B.A.T.M.A.N.-Advanced Multicast Optimizations, v3
by Linus Lüssing 30 Jan '11

30 Jan '11

Hi everyone, here's the third iteration of the multicast patches which is mainly some bugfixing, smaller but also slightly larger restructuring. Thanks for the feedback again! Cheers, Linus = Changelog v3 = * rebasing to commit [8bb3d9422707a7814ac804acbc00569dd017288c] * 4B-aligning for 'struct mcast_entry', adding __packed attribute [01/19] * converted mcast_forw_table and subsequent lists to hlists [10/19], [11/19], [12/19], [13/19] * splitted mcast_forw_table_seq_print_text into several functions [11/19] * shorter rcu-locking when printing to sysfs, only in seq_print_if_entry [11/19] * add missing spin_lock initilization [10/19] * fix logic for break_flag in tracker_packet_for_each() macro [07/19] * compat.h, should be "#define MC_LIST struct netdev_hw_addr" instead [03/19] * adding mcast duplicate check seqno lock [17/19] * fixed rcu-locking in route_mcast_tracker_packet(), in fact, using refcounting there now [07/19] * restructuring route_mcast_tracker_packet(), merging two loops * also delete dest from tracker packet if unknown originator in zero_tracker_packet() [07/19] * compactify check-code in zero_tracker_packet() [07/19] * don't use index i like nexthop_tracker_packets[i] in skb sending loop in route_mcast_tracker_packet(), it has the wrong base [07/19] * use hash_find() instead of own orig_node look-up in zero_tracker_packt() and add_router_of_dest() [07/19] * reserve dest_entries on heap instead of stack, >4k is too much for the stack [07/19] * rename dest_entries_list to dest_entries_buckets, trying to clarify its purpose [07/19] * add comment for find_mca_match() [07/19] * move comments for zero_tracker_packet() and shrink_tracker_packet() to the correct patches [10/19] * remove wrong comment for sync_orig() (duplicate of sync_table()) [10/19] * remove variable name defines for net_device multicast lists [03/19] * directly create & use SKBs for tracker_packets [07/19], [09/19] * moving stuff from mcast_proact_tracker_prepare() to build_tracker_packet_skb() [07/19] * only create methods / variables / defines / macros in patches that need them (this changed the BAT_MCAST/MCAST_TRACKER define numbers) -[01/19], -[03/19] * use rcu-locking for mcast_forw_table, +[18]

1 18

[B.A.T.M.A.N.] batman-adv 2011.0.0 released
by Marek Lindner 29 Jan '11

29 Jan '11

The B.A.T.M.A.N. team is excited to announce its latest release, 2011.0.0, mainly fulfilling popular feature requests but also delivering bug fixes and performance enhancements as well as modifications under the hood. As the kernel module always depends on the Linux kernel it was compiled against, it does not make sense to provide binaries on our website. As usual, you will find the signed tarballs in our download section: http://downloads.open-mesh.org/batman/releases/batman-adv-2011.0.0/ as well as prepackaged binaries in your distribution. Important changes ---------------- The "gateway support" coming with this release manipulates the DHCP request / reply behavior once activated. It is strongly suggested that you read our documentation before you make use of this feature to avoid surprises. Thanks ---------------- Thanks to all people sending in patches: * Andreas Langer <an.langer(a)gmx.de> * Jesper Juhl <jj(a)chaosbits.net> * Linus Lüssing <linus.luessing(a)web.de> * Marek Lindner <lindner_marek(a)yahoo.de> * Simon Wunderlich <siwu(a)hrz.tu-chemnitz.de> * Sven Eckelmann <sven(a)narfation.org> and to all those that supported us with good advice or rigorous testing: * David S. Miller <davem(a)davemloft.net> * Russell Senior <russell(a)personaltelco.net> * Tobias Klauser <tklauser(a)distanz.ch> batman-adv ---------------- This release brings the long awaited internet gateway support: Batman-adv nodes in a network can be declared 'gateways' with a certain bandwidth at their disposal. All clients learn about these gateways and can choose their preferred gateway based on link qualities / bandwidth. The decision is 'propagated' to the higher layers by manipulating the DHCP requests. Furthermore, batman-adv implements a bridge loop detection & avoidance that aim to allow multiple batman-adv nodes connected to the same LAN segment without provoking a loop. The initial layer2 fragmentation implementation that was included in the last releases was further enhanced. Batman-adv nodes forwarding other nodes' traffic inspect the to be transfered packets to automatically fragment the packet if the MTU of the outgoing interface does not allow a normal forward. Fragmented packets are reassembled whenever possible to keep the overhead as low as possible. Also new is the possibility to configure the hop penalty at runtime (via sysfs). The internally used hash implementation has been reimplemented after David Miller pointed out major deficiencies. These changes gave the batman-adv module a 10% speedup when forwarding payload traffic as these operations heavily make use of the hash code. The Linux kernel bit operation functions replaced the custom functions in batman-adv which are better debugged. The cleanup of the sysfs code greatly reduced the code size and makes this code more manageable. batctl ---------------- The support for the new gateway sysfs API was among the biggest changes for this release. It allows to easily turn the gateway mode on/off and displays all known gateways including their announced bandwidth. To monitor the bridge loop avoidance batctl offers a new parameter to display the the bridge neighbor table while highlighting the 'gateway to LAN'. The bisect component now can deal with the 32Bit sequence numbers in the OGMs and the bat-hosts parser became more robust while replacing MAC addresses with bat-host names. Happy routing, The B.A.T.M.A.N. team

1 0

[B.A.T.M.A.N.] [PATCH 1/2] batman-adv: Merge batman-adv 2011.0.0
by Sven Eckelmann 29 Jan '11

29 Jan '11

Signed-off-by: Sven Eckelmann <sven(a)narfation.org> --- batman-adv/CHANGELOG | 13 +++++++++++++ batman-adv/README | 15 ++++++++------- 2 files changed, 21 insertions(+), 7 deletions(-) diff --git a/batman-adv/CHANGELOG b/batman-adv/CHANGELOG index beff486..6265828 100644 --- a/batman-adv/CHANGELOG +++ b/batman-adv/CHANGELOG @@ -1,3 +1,16 @@ +batman-adv 2011.0.0: + +* support latest kernels (2.6.21 - 2.6.38) +* conversion of reference counting to kref +* allow merging and refragmentation of unicast packets during transfers +* add softif bridge loop detection +* make hop_penalty configurable through sysfs +* reimplement most of the batman-adv hash functionality +* support for optimized DHCP based gateway selection +* cleanup of the sysfs code + + -- Sat, 29 Jan 2011 20:08:05 +0100 + batman-adv 2010.2.0: * support latest kernels (2.6.21 - 2.6.37) diff --git a/batman-adv/README b/batman-adv/README index fb5b9ff..85e4518 100644 --- a/batman-adv/README +++ b/batman-adv/README @@ -1,4 +1,4 @@ -[state: 21-11-2010] +[state: 27-01-2011] BATMAN-ADV ---------- @@ -18,7 +18,7 @@ Batman advanced was implemented as a Linux kernel driver to re- duce the overhead to a minimum. It does not depend on any (other) network driver, and can be used on wifi as well as ethernet lan, vpn, etc ... (anything with ethernet-style layer 2). It compiles -against and should work with Linux 2.6.21 - 2.6.37. Supporting +against and should work with Linux 2.6.21 - 2.6.38. Supporting older versions is not planned, but it's probably easy to backport it. If you work on a backport, feel free to contact us. :-) @@ -83,15 +83,16 @@ All mesh wide settings can be found in batman's own interface folder: # ls /sys/class/net/bat0/mesh/ -# aggregated_ogms bonding fragmentation orig_interval -# vis_mode +# aggregated_ogms gw_bandwidth hop_penalty +# bonding gw_mode orig_interval +# fragmentation gw_sel_class vis_mode There is a special folder for debugging informations: # ls /sys/kernel/debug/batman_adv/bat0/ -# originators socket transtable_global transtable_local -# vis_data +# gateways socket transtable_global vis_data +# originators softif_neigh transtable_local Some of the files contain all sort of status information regard- @@ -250,7 +251,7 @@ CONTACT Please send us comments, experiences, questions, anything :) IRC: #batman on irc.freenode.org -Mailing-list: b.a.t.m.a.n(a)open-mesh.net (optional subscription +Mailing-list: b.a.t.m.a.n(a)open-mesh.org (optional subscription at https://lists.open-mesh.org/mm/listinfo/b.a.t.m.a.n) You can also contact the Authors: -- 1.7.2.3

2 2

[B.A.T.M.A.N.] [PATCH 1/3] batman-adv: Remove vis info on hashing errors
by Sven Eckelmann 29 Jan '11

29 Jan '11

A newly created vis info object must be removed when it couldn't be added to the hash. The old_info which has to be replaced was already removed and isn't related to the hash anymore. Signed-off-by: Sven Eckelmann <sven(a)narfation.org> --- batman-adv/vis.c | 2 +- 1 files changed, 1 insertions(+), 1 deletions(-) diff --git a/batman-adv/vis.c b/batman-adv/vis.c index 2cdf01c..9d2729c 100644 --- a/batman-adv/vis.c +++ b/batman-adv/vis.c @@ -446,7 +446,7 @@ static struct vis_info *add_packet(struct bat_priv *bat_priv, info); if (hash_added < 0) { /* did not work (for some reason) */ - kref_put(&old_info->refcount, free_info); + kref_put(&info->refcount, free_info); info = NULL; } -- 1.7.2.3

2 3

[B.A.T.M.A.N.] Dead code cleanup
by Sven Eckelmann 28 Jan '11

28 Jan '11

Hi, I started to go through the release todo and removed some dead code. Best regards, Sven

2 6

[B.A.T.M.A.N.] [PATCH] batctl: tcpdump - support LARGETAIL flag
by Marek Lindner 28 Jan '11

28 Jan '11

Signed-off-by: Marek Lindner <lindner_marek(a)yahoo.de> --- batctl/tcpdump.c | 5 +++-- 1 files changed, 3 insertions(+), 2 deletions(-) diff --git a/batctl/tcpdump.c b/batctl/tcpdump.c index bdf81bc..09a7e16 100644 --- a/batctl/tcpdump.c +++ b/batctl/tcpdump.c @@ -385,10 +385,11 @@ static void dump_batman_frag(unsigned char *packet_buff, ssize_t buff_len, int r printf("BAT %s > ", get_name_by_macaddr((struct ether_addr *)unicast_frag_packet->orig, read_opt)); - printf("%s: FRAG, seq %hu, ttl %hhu, flags [%c], ", + printf("%s: FRAG, seq %hu, ttl %hhu, flags [%c%c], ", get_name_by_macaddr((struct ether_addr *)unicast_frag_packet->dest, read_opt), ntohs(unicast_frag_packet->seqno), unicast_frag_packet->ttl, - (unicast_frag_packet->flags & UNI_FRAG_HEAD ? 'H' : '.')); + (unicast_frag_packet->flags & UNI_FRAG_HEAD ? 'H' : '.'), + (unicast_frag_packet->flags & UNI_FRAG_LARGETAIL ? 'L' : '.')); if (unicast_frag_packet->flags & UNI_FRAG_HEAD) parse_eth_hdr(packet_buff + ETH_HLEN + sizeof(struct unicast_frag_packet), -- 1.7.2.3

1 1

[B.A.T.M.A.N.] An Experimental Comparison of Routing Protocols in Multi Hop Ad Hoc Networks
by or4n9e＠gmx.de 26 Jan '11

26 Jan '11

Hi, following your call for paper submissions at http://www.open-mesh.org/wiki/Experience, I'd briefly like to make your group aware of the following 2010 scientific paper [1] about OSLR, BATMAN L2+L3 and Babel protocols (performance comparison etc, kind of a follow-up to the already listed paper at [2]). I assume it's of some interest for you guys! Thanks, orange (Piranha Project Maintainer), http://piranha.pwnz.org/ [1] http://bridgingthelayers.org/docs/murray_routing.pdf [2] http://ro.uow.edu.au/cgi/viewcontent.cgi?article=1747&context=infopapers -- GMX DSL Doppel-Flat ab 19,99 Euro/mtl.! Jetzt mit gratis Handy-Flat! http://portal.gmx.net/de/go/dsl

2 1

[B.A.T.M.A.N.] [PATCH 1/2] batman-adv: Create roughly equal sized fragments
by Sven Eckelmann 25 Jan '11

25 Jan '11

The routing algorithm must know how large two fragments are to be able to decide that it is safe to merge them or if it should resubmit without waiting for the second part. When these two fragments have a too different size, it is not possible to guess right in every situation. The user could easily configure the MTU of the attached cards so that one fragment is forwarded and the other one is added to the fragments table to wait for the missing part. For even sized packages it is possible to split it so that the resulting packages are equal sized by ignoring the old non-fragment header at the beginning of the original packet. This still creates different sized fragments for uneven sized packets. Reported-by: Russell Senior <russell(a)personaltelco.net> Reported-by: Marek Lindner <lindner_marek(a)yahoo.de> Signed-off-by: Sven Eckelmann <sven(a)narfation.org> --- This patch is completely untested and is currently only to support marec and russell-- during their debugging sessions. Maybe it doesn't even compile. batman-adv/unicast.c | 5 +++-- 1 files changed, 3 insertions(+), 2 deletions(-) diff --git a/batman-adv/unicast.c b/batman-adv/unicast.c index 677663b..84e70bc 100644 --- a/batman-adv/unicast.c +++ b/batman-adv/unicast.c @@ -230,7 +230,7 @@ int frag_send_skb(struct sk_buff *skb, struct bat_priv *bat_priv, struct unicast_frag_packet *frag1, *frag2; int uc_hdr_len = sizeof(struct unicast_packet); int ucf_hdr_len = sizeof(struct unicast_frag_packet); - int data_len = skb->len; + int data_len = skb->len - uc_hdr_len; if (!bat_priv->primary_if) goto dropped; @@ -238,10 +238,11 @@ int frag_send_skb(struct sk_buff *skb, struct bat_priv *bat_priv, frag_skb = dev_alloc_skb(data_len - (data_len / 2) + ucf_hdr_len); if (!frag_skb) goto dropped; + skb_reserve(frag_skb, ucf_hdr_len); unicast_packet = (struct unicast_packet *) skb->data; memcpy(&tmp_uc, unicast_packet, uc_hdr_len); - skb_split(skb, frag_skb, data_len / 2); + skb_split(skb, frag_skb, data_len / 2 + uc_hdr_len); if (my_skb_head_push(skb, ucf_hdr_len - uc_hdr_len) < 0 || my_skb_head_push(frag_skb, ucf_hdr_len) < 0) -- 1.7.2.3

2 7

[B.A.T.M.A.N.] [PATCH] batman-adv: Fix kernel panic when fetching vis data on a vis server
by Linus Lüssing 25 Jan '11

25 Jan '11

Commit 593180d8b00bd486187841468adb53ca1edf3950 introduced a bug leading to a kernel panic when fetching the vis data on a vis server. That commit forgot to rename one variable name, which this commit fixes now. Reported-by: Russell Senior <russell(a)personaltelco.net> Signed-off-by: Linus Lüssing <linus.luessing(a)web.de> --- vis.c | 4 ++-- 1 files changed, 2 insertions(+), 2 deletions(-) diff --git a/batman-adv/vis.c b/batman-adv/vis.c index 69178d9..2cdf01c 100644 --- a/batman-adv/vis.c +++ b/batman-adv/vis.c @@ -268,10 +268,10 @@ int vis_seq_print_text(struct seq_file *seq, void *offset) buff_pos += sprintf(buff + buff_pos, "%pM,", entry->addr); - for (i = 0; i < packet->entries; i++) + for (j = 0; j < packet->entries; j++) buff_pos += vis_data_read_entry( buff + buff_pos, - &entries[i], + &entries[j], entry->addr, entry->primary); -- 1.7.2.3

2 1

[B.A.T.M.A.N.] [PATCH] batman-adv: make broadcast seqno operations atomic
by Marek Lindner 25 Jan '11

25 Jan '11

Batman-adv could receive several payload broadcasts at the same time that would trigger access to the broadcast seqno sliding window to determine whether this is a new broadcast or not. If these incoming broadcasts are accessing the sliding window simultaneously it could be left in an inconsistent state. Therefore it is necessary to make sure this access is atomic. Reported-by: Linus Lüssing <linus.luessing(a)web.de> Signed-off-by: Marek Lindner <lindner_marek(a)yahoo.de> --- batman-adv/originator.c | 1 + batman-adv/routing.c | 15 +++++++++++---- batman-adv/types.h | 2 ++ 3 files changed, 14 insertions(+), 4 deletions(-) diff --git a/batman-adv/originator.c b/batman-adv/originator.c index cf2ec37..b1a3d92 100644 --- a/batman-adv/originator.c +++ b/batman-adv/originator.c @@ -220,6 +220,7 @@ struct orig_node *get_orig_node(struct bat_priv *bat_priv, uint8_t *addr) INIT_HLIST_HEAD(&orig_node->neigh_list); INIT_LIST_HEAD(&orig_node->bond_list); spin_lock_init(&orig_node->ogm_cnt_lock); + spin_lock_init(&orig_node->bcast_seqno_lock); spin_lock_init(&orig_node->neigh_list_lock); kref_init(&orig_node->refcount); diff --git a/batman-adv/routing.c b/batman-adv/routing.c index 06201dc..2cf595d 100644 --- a/batman-adv/routing.c +++ b/batman-adv/routing.c @@ -1427,28 +1427,32 @@ int recv_bcast_packet(struct sk_buff *skb, struct batman_if *recv_if) bcast_packet->orig)); if (!orig_node) - goto unlock; + goto rcu_unlock; kref_get(&orig_node->refcount); rcu_read_unlock(); + spin_lock_bh(&orig_node->bcast_seqno_lock); + /* check whether the packet is a duplicate */ if (get_bit_status(orig_node->bcast_bits, orig_node->last_bcast_seqno, ntohl(bcast_packet->seqno))) - goto out; + goto spin_unlock; seq_diff = ntohl(bcast_packet->seqno) - orig_node->last_bcast_seqno; /* check whether the packet is old and the host just restarted. */ if (window_protected(bat_priv, seq_diff, &orig_node->bcast_seqno_reset)) - goto out; + goto spin_unlock; /* mark broadcast in flood history, update window position * if required. */ if (bit_get_packet(bat_priv, orig_node->bcast_bits, seq_diff, 1)) orig_node->last_bcast_seqno = ntohl(bcast_packet->seqno); + spin_unlock_bh(&orig_node->bcast_seqno_lock); + /* rebroadcast packet */ add_bcast_packet_to_list(bat_priv, skb); @@ -1457,8 +1461,11 @@ int recv_bcast_packet(struct sk_buff *skb, struct batman_if *recv_if) ret = NET_RX_SUCCESS; goto out; -unlock: +rcu_unlock: rcu_read_unlock(); + goto out; +spin_unlock: + spin_unlock_bh(&orig_node->bcast_seqno_lock); out: if (orig_node) kref_put(&orig_node->refcount, orig_node_free_ref); diff --git a/batman-adv/types.h b/batman-adv/types.h index 56309bf..b7b9561 100644 --- a/batman-adv/types.h +++ b/batman-adv/types.h @@ -90,6 +90,8 @@ struct orig_node { spinlock_t ogm_cnt_lock; /* protects: bcast_own, bcast_own_sum, * neigh_node->real_bits, * neigh_node->real_packet_count */ + spinlock_t bcast_seqno_lock; /* protects bcast_bits, + * last_bcast_seqno */ atomic_t bond_candidates; struct list_head bond_list; }; -- 1.7.2.3

1 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 January 2011