B.A.T.M.A.N December 2011

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

22 participants
40 discussions

[B.A.T.M.A.N.] [PATCH 1/2] batman-adv: Remove date from README
by Sven Eckelmann 24 Dec '11

24 Dec '11

The date has to be updated when a patch touches the README. Therefore, nearly every feature will modify this date. It can happens quite often that not only one feature is currently in development or waiting on the mailinglist. This creates merge conflicts when applying a patchset. The date itself doesn't provide any additional information when this file is only available in a release tarball or as part of a SCM repository. Signed-off-by: Sven Eckelmann <sven(a)narfation.org> --- README | 2 -- 1 files changed, 0 insertions(+), 2 deletions(-) diff --git a/README b/README index ddc9569..58d3473 100644 --- a/README +++ b/README @@ -1,5 +1,3 @@ -[state: 13-11-2011] - BATMAN-ADV ---------- -- 1.7.7.3

2 3

[B.A.T.M.A.N.] [PATCH] batman-adv: refactor tt_global_del() to avoid misalignment
by Marek Lindner 24 Dec '11

24 Dec '11

From: Sven Eckelmann <sven(a)narfation.org> Signed-off-by: Sven Eckelmann <sven(a)narfation.org> --- translation-table.c | 42 ++++++++++++++++++++++-------------------- 1 files changed, 22 insertions(+), 20 deletions(-) diff --git a/translation-table.c b/translation-table.c index d542b9d..a1be7ef 100644 --- a/translation-table.c +++ b/translation-table.c @@ -657,29 +657,31 @@ void tt_global_del(struct bat_priv *bat_priv, struct tt_local_entry *tt_local_entry = NULL; tt_global_entry = tt_global_hash_find(bat_priv, addr); - if (!tt_global_entry) + if (!tt_global_entry || tt_global_entry->orig_node != orig_node) goto out; - if (tt_global_entry->orig_node == orig_node) { - if (roaming) { - /* if we are deleting a global entry due to a roam - * event, there are two possibilities: - * 1) the client roamed from node A to node B => we mark - * it with TT_CLIENT_ROAM, we start a timer and we - * wait for node B to claim it. In case of timeout - * the entry is purged. - * 2) the client roamed to us => we can directly delete - * the global entry, since it is useless now. */ - tt_local_entry = tt_local_hash_find(bat_priv, - tt_global_entry->common.addr); - if (!tt_local_entry) { - tt_global_entry->common.flags |= TT_CLIENT_ROAM; - tt_global_entry->roam_at = jiffies; - goto out; - } - } - _tt_global_del(bat_priv, tt_global_entry, message); + if (!roaming) + goto out_del; + + /* if we are deleting a global entry due to a roam + * event, there are two possibilities: + * 1) the client roamed from node A to node B => we mark + * it with TT_CLIENT_ROAM, we start a timer and we + * wait for node B to claim it. In case of timeout + * the entry is purged. + * 2) the client roamed to us => we can directly delete + * the global entry, since it is useless now. */ + tt_local_entry = tt_local_hash_find(bat_priv, + tt_global_entry->common.addr); + if (!tt_local_entry) { + tt_global_entry->common.flags |= TT_CLIENT_ROAM; + tt_global_entry->roam_at = jiffies; + goto out; } + +out_del: + _tt_global_del(bat_priv, tt_global_entry, message); + out: if (tt_global_entry) tt_global_entry_free_ref(tt_global_entry); -- 1.7.5.4

2 2

[B.A.T.M.A.N.] [PATCH] batman-adv: convert time_after instances to has_timed_out
by Marek Lindner 22 Dec '11

22 Dec '11

To increase readability the has_timed_out() functions has been introduced. This patch converts existing time_after() calls to use this wrapper function (if applicable). Signed-off-by: Marek Lindner <lindner_marek(a)yahoo.de> --- originator.c | 11 ++++------- routing.c | 3 +-- soft-interface.c | 4 ++-- vis.c | 3 +-- 4 files changed, 8 insertions(+), 13 deletions(-) diff --git a/originator.c b/originator.c index 847ff7e..1161f27 100644 --- a/originator.c +++ b/originator.c @@ -282,8 +282,7 @@ static bool purge_orig_neighbors(struct bat_priv *bat_priv, hlist_for_each_entry_safe(neigh_node, node, node_tmp, &orig_node->neigh_list, list) { - if ((time_after(jiffies, - neigh_node->last_valid + PURGE_TIMEOUT * HZ)) || + if ((has_timed_out(neigh_node->last_valid, PURGE_TIMEOUT)) || (neigh_node->if_incoming->if_status == IF_INACTIVE) || (neigh_node->if_incoming->if_status == IF_NOT_IN_USE) || (neigh_node->if_incoming->if_status == IF_TO_BE_REMOVED)) { @@ -327,9 +326,7 @@ static bool purge_orig_node(struct bat_priv *bat_priv, { struct neigh_node *best_neigh_node; - if (time_after(jiffies, - orig_node->last_valid + 2 * PURGE_TIMEOUT * HZ)) { - + if (has_timed_out(orig_node->last_valid, 2 * PURGE_TIMEOUT)) { bat_dbg(DBG_BATMAN, bat_priv, "Originator timeout: originator %pM, last_valid %lu\n", orig_node->orig, (orig_node->last_valid / HZ)); @@ -372,8 +369,8 @@ static void _purge_orig(struct bat_priv *bat_priv) continue; } - if (time_after(jiffies, orig_node->last_frag_packet + - msecs_to_jiffies(FRAG_TIMEOUT))) + if (has_timed_out(orig_node->last_frag_packet, + FRAG_TIMEOUT)) frag_list_free(&orig_node->frag_list); } spin_unlock_bh(list_lock); diff --git a/routing.c b/routing.c index 5bc41c8..87b95a2 100644 --- a/routing.c +++ b/routing.c @@ -232,8 +232,7 @@ int window_protected(struct bat_priv *bat_priv, int32_t seq_num_diff, { if ((seq_num_diff <= -TQ_LOCAL_WINDOW_SIZE) || (seq_num_diff >= EXPECTED_SEQNO_RANGE)) { - if (time_after(jiffies, *last_reset + - msecs_to_jiffies(RESET_PROTECTION_MS))) { + if (has_timed_out(*last_reset, RESET_PROTECTION_MS)) { *last_reset = jiffies; bat_dbg(DBG_BATMAN, bat_priv, diff --git a/soft-interface.c b/soft-interface.c index bd8c7cf..51788db 100644 --- a/soft-interface.c +++ b/soft-interface.c @@ -396,8 +396,8 @@ void softif_neigh_purge(struct bat_priv *bat_priv) hlist_for_each_entry_safe(softif_neigh, node_tmp, node_tmp2, &softif_neigh_vid->softif_neigh_list, list) { - if ((!time_after(jiffies, softif_neigh->last_seen + - msecs_to_jiffies(SOFTIF_NEIGH_TIMEOUT))) && + if ((!has_timed_out(softif_neigh->last_seen, + SOFTIF_NEIGH_TIMEOUT)) && (atomic_read(&bat_priv->mesh_state) == MESH_ACTIVE)) continue; diff --git a/vis.c b/vis.c index ac7e661..4f4b2a0 100644 --- a/vis.c +++ b/vis.c @@ -714,8 +714,7 @@ static void purge_vis_packets(struct bat_priv *bat_priv) if (info == bat_priv->my_vis_info) continue; - if (time_after(jiffies, - info->first_seen + VIS_TIMEOUT * HZ)) { + if (has_timed_out(info->first_seen, VIS_TIMEOUT)) { hlist_del(node); send_list_del(info); kref_put(&info->refcount, free_info); -- 1.7.5.4

1 1

[B.A.T.M.A.N.] switch routing algorithm at runtime (v2)
by Marek Lindner 22 Dec '11

22 Dec '11

Hi, next round for the selectable routing algorithm patches. The major change which comes with this patchset is the module parameter that allows to set the routing algorithm. This setting only affects batX interfaces created after the change has been made. The list of available routing algorithms moved to the batman-adv debugfs root directory, so that it is visible even before the first batX interface has been created. The sysfs entry still shows the configured routing algorithm but does not allow to change it. Big kudos to Sven for helping me out with the compat macros. Cheers, Marek

1 8

[B.A.T.M.A.N.] [PATCH] batman-adv: checkpatch cleanup - remove lines longer than 80 chars
by Marek Lindner 20 Dec '11

20 Dec '11

The long line was introduced with b26e478f8f. Signed-off-by: Marek Lindner <lindner_marek(a)yahoo.de> --- net/batman-adv/translation-table.c | 2 +- 1 files changed, 1 insertions(+), 1 deletions(-) diff --git a/net/batman-adv/translation-table.c b/net/batman-adv/translation-table.c index ab8dea8..58361ed 100644 --- a/net/batman-adv/translation-table.c +++ b/net/batman-adv/translation-table.c @@ -679,7 +679,7 @@ void tt_global_del(struct bat_priv *bat_priv, * 2) the client roamed to us => we can directly delete * the global entry, since it is useless now. */ tt_local_entry = tt_local_hash_find(bat_priv, - tt_global_entry->common.addr); + tt_global_entry->common.addr); if (!tt_local_entry) { tt_global_entry->common.flags |= TT_CLIENT_ROAM; tt_global_entry->roam_at = jiffies; -- 1.7.5.4

3 5

[B.A.T.M.A.N.] pull request: batman-adv 2011-12-12
by Marek Lindner 13 Dec '11

13 Dec '11

Hi David, here comes the next batch I'd like to get the pulled into net-next-2.6/3.3. We have a couple of minor fixes for unregister_netdevice(), compat flags, return value checks and a readme update. Plus the regression fix from Dan that addresses a bug introduced with the previous batch. Security researcher Paul Kot checked our kernelland / user space communication and found a subtle bug which would make batman-adv write to memory locations it should not write to. However, he did not find a way to exploit this bug, therefore we classified it as "non-critical". Please let us know if that is not ok with you. Thanks, Marek The following changes since commit 76e8d7b0d1b0091929e8cd18551a87b6bb6ef2c1: batman-adv: Fix range check for expected packets (2011-11-20 13:08:35 +0100) are available in the git repository at: git://git.open-mesh.org/linux-merge.git batman-adv/next Antonio Quartulli (3): batman-adv: create a common substructure for tt_global/local_entry batman-adv: generalise tt_local_reset_flags() batman-adv: format multi-line if in the correct way Dan Carpenter (1): batman-adv: remove extra negation in gw_out_of_range() Marek Lindner (2): batman-adv: report compat_version in version field in case of version mismatch batman-adv: readme update (mention ap isolation and new log level) Paul Kot (1): batman-adv: bat_socket_read missing checks Simon Wunderlich (2): batman-adv: check return value for hash_add() batman-adv: use unregister_netdevice() when softif_create fails Sven Eckelmann (2): batman-adv: Directly check read of icmp packet in copy_from_user batman-adv: Only write requested number of byte to user buffer Documentation/networking/batman-adv.txt | 7 +- net/batman-adv/gateway_client.c | 2 +- net/batman-adv/icmp_socket.c | 14 +- net/batman-adv/routing.c | 3 +- net/batman-adv/soft-interface.c | 2 +- net/batman-adv/translation-table.c | 336 ++++++++++++++++++------------- net/batman-adv/types.h | 14 +- net/batman-adv/vis.c | 6 +- 8 files changed, 216 insertions(+), 168 deletions(-)

2 12

[B.A.T.M.A.N.] bat_socket_read missing checks
by Paul 12 Dec '11

12 Dec '11

Writing a icmp_packet_rr and then reading icmp_packet can lead to kernel memory corruption, if __user *buf is just below TASK_SIZE. -- Regards, Paul

3 7

[B.A.T.M.A.N.] [PATCH] batctl: fix log level 'all'
by Marek Lindner 10 Dec '11

10 Dec '11

Signed-off-by: Marek Lindner <lindner_marek(a)yahoo.de> --- sys.c | 2 +- 1 files changed, 1 insertions(+), 1 deletions(-) diff --git a/sys.c b/sys.c index e510bee..105bb51 100644 --- a/sys.c +++ b/sys.c @@ -213,7 +213,7 @@ int handle_loglevel(char *mesh_iface, int argc, char **argv) log_level = 0; break; } else if (strcmp(argv[i], "all") == 0) { - log_level = 15; + log_level = 7; break; } else if (strcmp(argv[i], "batman") == 0) log_level |= (1 << 0); -- 1.7.5.4

1 1

[B.A.T.M.A.N.] [PATCH] batman-adv: Generate config compatible with kbuild macros
by Sven Eckelmann 10 Dec '11

10 Dec '11

Linux 3.0-rc1-30-g2a11c8e introduced the macros IS_ENABLED, IS_BUILTIN and IS_MODULE that can be used to identify which state a tristate option had. The compat-autoconf.h needs to provide additional precompiler variables that these marcros can use. Signed-off-by: Sven Eckelmann <sven(a)narfation.org> --- gen-compat-autoconf.sh | 9 ++++++++- 1 files changed, 8 insertions(+), 1 deletions(-) diff --git a/gen-compat-autoconf.sh b/gen-compat-autoconf.sh index 440accc..1a020c1 100755 --- a/gen-compat-autoconf.sh +++ b/gen-compat-autoconf.sh @@ -12,15 +12,22 @@ gen_config() { VALUE="${2}" echo "#undef ${KEY}" + echo "#undef __enabled_${KEY}" + echo "#undef __enabled_${KEY}_MODULE" case "${VALUE}" in y) echo "#define ${KEY} 1" + echo "#define __enabled_${KEY} 1" + echo "#define __enabled_${KEY}_MODULE 0" ;; m) echo "#define ${KEY} 1" + echo "#define __enabled_${KEY} 0" + echo "#define __enabled_${KEY}_MODULE 1" ;; n) - # leave it undefined + echo "#define __enabled_${KEY} 0" + echo "#define __enabled_${KEY}_MODULE 0" ;; *) echo "#define ${KEY} \"${VALUE}\"" -- 1.7.7.3

2 1

[B.A.T.M.A.N.] [PATCH] batman-adv: warn if added interface is part of a bridge
by Marek Lindner 10 Dec '11

10 Dec '11

Signed-off-by: Marek Lindner <lindner_marek(a)yahoo.de> --- Thanks to Sven we even have a 'nice' compat.h workaround. :) compat.h | 1 + hard-interface.c | 8 ++++++++ 2 files changed, 9 insertions(+), 0 deletions(-) diff --git a/compat.h b/compat.h index 964c066..dcd92d1 100644 --- a/compat.h +++ b/compat.h @@ -52,6 +52,7 @@ #if LINUX_VERSION_CODE < KERNEL_VERSION(2, 6, 36) #define __rcu +#define IFF_BRIDGE_PORT 0 || (hard_iface->net_dev->br_port ? 1 : 0) #endif /* < KERNEL_VERSION(2, 6, 36) */ diff --git a/hard-interface.c b/hard-interface.c index d3e0e32..68b667c 100644 --- a/hard-interface.c +++ b/hard-interface.c @@ -281,6 +281,14 @@ int hardif_enable_interface(struct hard_iface *hard_iface, if (!atomic_inc_not_zero(&hard_iface->refcount)) goto out; + /* hard-interface is part of a bridge */ + if (hard_iface->net_dev->priv_flags & IFF_BRIDGE_PORT) + pr_err("You are about to enable batman-adv on '%s' which " + "already is part of a bridge. Unless you know exactly " + "what you are doing this is probably wrong and won't " + "work the way you think it would.\n", + hard_iface->net_dev->name); + soft_iface = dev_get_by_name(&init_net, iface_name); if (!soft_iface) { -- 1.7.5.4

1 1

2023

2022

2021

2020

2019

2018

2017

2016

2015

2014

2013

2012

2011

2010

2009

2008

2007

B.A.T.M.A.N December 2011