commits June 2014

commits@lists.open-mesh.org

1 participants
39 discussions

[batmand] master: batmand: Use memleak/error path free implementation of hash_resize (c8f485b)
by postmaster＠open-mesh.org 29 Jun '14

29 Jun '14

Repository : ssh://git@open-mesh.org/batmand On branch : master >--------------------------------------------------------------- commit c8f485bd88c4e2113e897fda11abbbc25f232915 Author: Sven Eckelmann <sven(a)narfation.org> Date: Tue Jun 17 21:55:59 2014 +0200 batmand: Use memleak/error path free implementation of hash_resize The current implementation of hash_resize uses hash_add directly to initialize a new hash table. But hash_add has two possible situations when it returns an error and hash_resize would leak the data: * data already exists * malloc fails The check for the duplicated data is not really harmful (beside increasing the time to re-add elements) but the malloc can potentially return an error. This malloc is unnecessary and just takes extra time. Instead the bucket from the old hash table can be re-used. Signed-off-by: Sven Eckelmann <sven(a)narfation.org> Signed-off-by: Elektra Wagenrad <onelektra(a)gmx.net> >--------------------------------------------------------------- c8f485bd88c4e2113e897fda11abbbc25f232915 hash.c | 72 +++++++++++++++++++++++++++++++++++++++------------------------- 1 file changed, 44 insertions(+), 28 deletions(-) diff --git a/hash.c b/hash.c index 7271b89..1c551dc 100644 --- a/hash.c +++ b/hash.c @@ -63,6 +63,40 @@ void hash_delete(struct hashtable_t *hash, hashdata_free_cb free_cb) hash_destroy(hash); } +/* adds data to the hashtable and reuse bucket. + * returns 0 on success, -1 on error */ +static int hash_add_bucket(struct hashtable_t *hash, void *data, + struct element_t *bucket, int check_duplicate) +{ + int index; + struct element_t *bucket_it, *prev_bucket = NULL; + + index = hash->choose(data, hash->size); + bucket_it = hash->table[index]; + + while (bucket_it != NULL) { + if (check_duplicate && + hash->compare(bucket_it->data, data)) + return -1; + + prev_bucket = bucket_it; + bucket_it = bucket_it->next; + } + + /* init the new bucket */ + bucket->data = data; + bucket->next = NULL; + + /* and link it */ + if (prev_bucket == NULL) + hash->table[index] = bucket; + else + prev_bucket->next = bucket; + + hash->elements++; + return 0; +} + /* free only the hashtable and the hash itself. */ void hash_destroy(struct hashtable_t *hash) { @@ -186,19 +220,8 @@ struct hashtable_t *hash_new(int size, hashdata_compare_cb compare, /* adds data to the hashtable. returns 0 on success, -1 on error */ int hash_add(struct hashtable_t *hash, void *data) { - int index; - struct element_t *bucket, *prev_bucket = NULL; - - index = hash->choose(data, hash->size); - bucket = hash->table[index]; - - while (bucket != NULL) { - if (hash->compare(bucket->data, data)) - return -1; - - prev_bucket = bucket; - bucket = bucket->next; - } + int ret; + struct element_t *bucket; /* found the tail of the list, add new element */ bucket = debugMalloc(sizeof(struct element_t), 304); @@ -206,18 +229,11 @@ int hash_add(struct hashtable_t *hash, void *data) if (!bucket) return -1; - /* init the new bucket */ - bucket->data = data; - bucket->next = NULL; - - /* and link it */ - if (prev_bucket == NULL) - hash->table[index] = bucket; - else - prev_bucket->next = bucket; + ret = hash_add_bucket(hash, data, bucket, 1); + if (ret < 0) + debugFree(bucket, 1307); - hash->elements++; - return 0; + return ret; } /* finds data, based on the key in keydata. returns the found data on success, @@ -307,10 +323,10 @@ struct hashtable_t *hash_resize(struct hashtable_t *hash, int size) /* copy the elements */ for (i = 0; i < hash->size; i++) { - bucket = hash->table[i]; - while (bucket != NULL) { - hash_add(new_hash, bucket->data); - bucket = bucket->next; + while (hash->table[i]) { + bucket = hash->table[i]; + hash->table[i] = bucket->next; + hash_add_bucket(new_hash, bucket->data, bucket, 0); } } /* remove hash and eventual overflow buckets but not the

1 0

[batmand] master: batmand: Import hash table version from alfred (9ff4b0e)
by postmaster＠open-mesh.org 29 Jun '14

29 Jun '14

Repository : ssh://git@open-mesh.org/batmand On branch : master >--------------------------------------------------------------- commit 9ff4b0e570390bb254069a922ce12b505bfefb7b Author: Sven Eckelmann <sven(a)narfation.org> Date: Tue Jun 17 21:55:58 2014 +0200 batmand: Import hash table version from alfred All new functionality, cleanup and similar things to the hash implementation were done in alfred. So import the newest version to keep it in sync. Signed-off-by: Sven Eckelmann <sven(a)narfation.org> Signed-off-by: Elektra Wagenrad <onelektra(a)gmx.net> >--------------------------------------------------------------- 9ff4b0e570390bb254069a922ce12b505bfefb7b hash.c | 88 ++++++++++++++++++++++++++++++++++++++++------------------------ hash.h | 69 +++++++++++++++++++++++++++++--------------------- 2 files changed, 96 insertions(+), 61 deletions(-) diff --git a/hash.c b/hash.c index 2a9f557..7271b89 100644 --- a/hash.c +++ b/hash.c @@ -20,11 +20,11 @@ */ -#include <stdio.h> /* NULL */ #include "hash.h" +#include <stdlib.h> +#include <stdio.h> #include "allocate.h" - /* clears the hash */ void hash_init(struct hashtable_t *hash) { @@ -70,9 +70,17 @@ void hash_destroy(struct hashtable_t *hash) debugFree(hash, 1303); } +/* free hash_it_t pointer when stopping hash_iterate early */ +void hash_iterate_free(struct hash_it_t *iter_in) +{ + debugFree(iter_in, 1304); +} + /* iterate though the hash. first element is selected with iter_in NULL. - * use the returned iterator to access the elements until hash_it_t returns NULL. */ -struct hash_it_t *hash_iterate(struct hashtable_t *hash, struct hash_it_t *iter_in) + * use the returned iterator to access the elements until hash_it_t returns + * NULL. */ +struct hash_it_t *hash_iterate(struct hashtable_t *hash, + struct hash_it_t *iter_in) { struct hash_it_t *iter; @@ -81,14 +89,17 @@ struct hash_it_t *hash_iterate(struct hashtable_t *hash, struct hash_it_t *iter_ iter->index = -1; iter->bucket = NULL; iter->prev_bucket = NULL; - } else - iter= iter_in; + } else { + iter = iter_in; + } - /* sanity checks first (if our bucket got deleted in the last iteration): */ + /* sanity checks first (if our bucket got deleted in the last + * iteration): */ if (iter->bucket != NULL) { if (iter->first_bucket != NULL) { - /* we're on the first element and it got removed after the last iteration. */ + /* we're on the first element and it got removed after + * the last iteration. */ if ((*iter->first_bucket) != iter->bucket) { /* there are still other elements in the list */ @@ -105,9 +116,10 @@ struct hash_it_t *hash_iterate(struct hashtable_t *hash, struct hash_it_t *iter_ } else if (iter->prev_bucket != NULL) { - /* we're not on the first element, and the bucket got removed after the last iteration. - * the last bucket's next pointer is not pointing to our actual bucket anymore. - * select the next. */ + /* we're not on the first element, and the bucket got + * removed after the last iteration. The last bucket's + * next pointer is not pointing to our actual bucket + * anymore. Select the next. */ if (iter->prev_bucket->next != iter->bucket) iter->bucket = iter->prev_bucket; @@ -124,7 +136,8 @@ struct hash_it_t *hash_iterate(struct hashtable_t *hash, struct hash_it_t *iter_ return iter; } } - /* if not returned yet, we've reached the last one on the index and have to search forward */ + /* if not returned yet, we've reached the last one on the index and + * have to search forward */ iter->index++; /* go through the entries of the hash table */ @@ -142,12 +155,13 @@ struct hash_it_t *hash_iterate(struct hashtable_t *hash, struct hash_it_t *iter_ } /* nothing to iterate over anymore */ - debugFree(iter, 1304); + hash_iterate_free(iter); return NULL; } /* allocates and clears the hash */ -struct hashtable_t *hash_new(int size, hashdata_compare_cb compare, hashdata_choose_cb choose) +struct hashtable_t *hash_new(int size, hashdata_compare_cb compare, + hashdata_choose_cb choose) { struct hashtable_t *hash; @@ -155,8 +169,8 @@ struct hashtable_t *hash_new(int size, hashdata_compare_cb compare, hashdata_cho if (!hash) return NULL; - hash->size= size; - hash->table= debugMalloc(sizeof(struct element_t *) * size, 303); + hash->size = size; + hash->table = debugMalloc(sizeof(struct element_t *)*size, 303); if (!hash->table) { debugFree(hash, 1305); @@ -183,11 +197,11 @@ int hash_add(struct hashtable_t *hash, void *data) return -1; prev_bucket = bucket; - bucket= bucket->next; + bucket = bucket->next; } /* found the tail of the list, add new element */ - bucket = debugMalloc(sizeof(struct element_t),304); + bucket = debugMalloc(sizeof(struct element_t), 304); if (!bucket) return -1; @@ -197,17 +211,17 @@ int hash_add(struct hashtable_t *hash, void *data) bucket->next = NULL; /* and link it */ - if (prev_bucket == NULL) { + if (prev_bucket == NULL) hash->table[index] = bucket; - } else { + else prev_bucket->next = bucket; - } hash->elements++; return 0; } -/* finds data, based on the key in keydata. returns the found data on success, or NULL on error */ +/* finds data, based on the key in keydata. returns the found data on success, + * or NULL on error */ void *hash_find(struct hashtable_t *hash, void *keydata) { int index; @@ -226,20 +240,21 @@ void *hash_find(struct hashtable_t *hash, void *keydata) return NULL; } -/* remove bucket (this might be used in hash_iterate() if you already found the bucket - * you want to delete and don't need the overhead to find it again with hash_remove(). - * But usually, you don't want to use this function, as it fiddles with hash-internals. */ +/* remove bucket (this might be used in hash_iterate() if you already found + * the bucket you want to delete and don't need the overhead to find it again + * with hash_remove(). But usually, you don't want to use this function, as it + * fiddles with hash-internals. */ void *hash_remove_bucket(struct hashtable_t *hash, struct hash_it_t *hash_it_t) { void *data_save; - data_save = hash_it_t->bucket->data; /* save the pointer to the data */ + /* save the pointer to the data */ + data_save = hash_it_t->bucket->data; - if (hash_it_t->prev_bucket != NULL) { + if (hash_it_t->prev_bucket != NULL) hash_it_t->prev_bucket->next = hash_it_t->bucket->next; - } else if (hash_it_t->first_bucket != NULL) { + else if (hash_it_t->first_bucket != NULL) (*hash_it_t->first_bucket) = hash_it_t->bucket->next; - } debugFree(hash_it_t->bucket, 1306); @@ -261,7 +276,12 @@ void *hash_remove(struct hashtable_t *hash, void *data) while (hash_it_t.bucket != NULL) { if (hash->compare(hash_it_t.bucket->data, data)) { - hash_it_t.first_bucket = (hash_it_t.bucket == hash->table[hash_it_t.index] ? &hash->table[ hash_it_t.index ] : NULL); + int bucket_same; + bucket_same = (hash_it_t.bucket == + hash->table[hash_it_t.index]); + hash_it_t.first_bucket = (bucket_same ? + &hash->table[hash_it_t.index] : + NULL); return hash_remove_bucket(hash, &hash_it_t); } @@ -272,7 +292,8 @@ void *hash_remove(struct hashtable_t *hash, void *data) return NULL; } -/* resize the hash, returns the pointer to the new hash or NULL on error. removes the old hash on success. */ +/* resize the hash, returns the pointer to the new hash or NULL on error. + * removes the old hash on success. */ struct hashtable_t *hash_resize(struct hashtable_t *hash, int size) { struct hashtable_t *new_hash; @@ -292,8 +313,9 @@ struct hashtable_t *hash_resize(struct hashtable_t *hash, int size) bucket = bucket->next; } } - - hash_delete(hash, NULL); /* remove hash and eventual overflow buckets but not the content itself. */ + /* remove hash and eventual overflow buckets but not the + * content itself. */ + hash_delete(hash, NULL); return new_hash; } diff --git a/hash.h b/hash.h index 66d7162..1682eb8 100644 --- a/hash.h +++ b/hash.h @@ -28,8 +28,8 @@ typedef int (*hashdata_choose_cb)(void *, int); typedef void (*hashdata_free_cb)(void *); struct element_t { - void *data; /* pointer to the data */ - struct element_t *next; /* overflow bucket pointer */ + void *data; /* pointer to the data */ + struct element_t *next; /* overflow bucket pointer */ }; struct hash_it_t { @@ -40,56 +40,69 @@ struct hash_it_t { }; struct hashtable_t { - struct element_t **table; /* the hashtable itself, with the buckets */ - int elements; /* number of elements registered */ - int size; /* size of hashtable */ - hashdata_compare_cb compare; /* callback to a compare function. - * should compare 2 element datas for their keys, - * return 0 if same and not 0 if not same */ - hashdata_choose_cb choose; /* the hashfunction, should return an index based - * on the key in the data of the first argument - * and the size the second */ + struct element_t **table; /* the hashtable itself, with the + * buckets */ + int elements; /* number of elements registered */ + int size; /* size of hashtable */ + hashdata_compare_cb compare; /* callback to a compare function. + * should compare 2 element datas for + * their keys, return 0 if same and not + * 0 if not same */ + hashdata_choose_cb choose; /* the hashfunction, should return an + * index based on the key in the data + * of the first argument and the size + * the second */ }; /* clears the hash */ -void hash_init(struct hashtable_t *hash); +void hash_init(struct hashtable_t *hash); /* allocates and clears the hash */ -struct hashtable_t *hash_new(int size, hashdata_compare_cb compare, hashdata_choose_cb choose); +struct hashtable_t *hash_new(int size, hashdata_compare_cb compare, + hashdata_choose_cb choose); -/* remove bucket (this might be used in hash_iterate() if you already found the bucket - * you want to delete and don't need the overhead to find it again with hash_remove(). - * But usually, you don't want to use this function, as it fiddles with hash-internals. */ -void *hash_remove_bucket(struct hashtable_t *hash, struct hash_it_t *hash_it_t); +/* remove bucket (this might be used in hash_iterate() if you already found + * the bucket you want to delete and don't need the overhead to find it again + * with hash_remove(). But usually, you don't want to use this function, as it + * fiddles with hash-internals. */ +void *hash_remove_bucket(struct hashtable_t *hash, + struct hash_it_t *hash_it_t); /* remove the hash structure. if hashdata_free_cb != NULL, * this function will be called to remove the elements inside of the hash. * if you don't remove the elements, memory might be leaked. */ -void hash_delete(struct hashtable_t *hash, hashdata_free_cb free_cb); +void hash_delete(struct hashtable_t *hash, hashdata_free_cb free_cb); /* free only the hashtable and the hash itself. */ -void hash_destroy(struct hashtable_t *hash); +void hash_destroy(struct hashtable_t *hash); /* adds data to the hashtable. returns 0 on success, -1 on error */ -int hash_add(struct hashtable_t *hash, void *data); +int hash_add(struct hashtable_t *hash, void *data); /* removes data from hash, if found. returns pointer do data on success, * so you can remove the used structure yourself, or NULL on error . * data could be the structure you use with just the key filled, * we just need the key for comparing. */ -void *hash_remove(struct hashtable_t *hash, void *data); +void *hash_remove(struct hashtable_t *hash, void *data); -/* finds data, based on the key in keydata. returns the found data on success, or NULL on error */ -void *hash_find(struct hashtable_t *hash, void *keydata); +/* finds data, based on the key in keydata. returns the found data on success, + * or NULL on error */ +void *hash_find(struct hashtable_t *hash, void *keydata); -/* resize the hash, returns the pointer to the new hash or NULL on error. removes the old hash on success */ -struct hashtable_t *hash_resize(struct hashtable_t *hash, int size); +/* resize the hash, returns the pointer to the new hash or NULL on error. + * removes the old hash on success */ +struct hashtable_t *hash_resize(struct hashtable_t *hash, int size); /* print the hash table for debugging */ -void hash_debug( struct hashtable_t *hash); +void hash_debug(struct hashtable_t *hash); /* iterate though the hash. first element is selected with iter_in NULL. - * use the returned iterator to access the elements until hash_it_t returns NULL. */ -struct hash_it_t *hash_iterate(struct hashtable_t *hash, struct hash_it_t *iter_in); + * use the returned iterator to access the elements until hash_it_t + * returns NULL. */ +struct hash_it_t *hash_iterate(struct hashtable_t *hash, + struct hash_it_t *iter_in); + +/* free hash_it_t pointer when stopping hash_iterate early */ +void hash_iterate_free(struct hash_it_t *iter_in); #endif

1 0

[alfred] master: alfred-gpsd: Avoid underrun when reading from gpsd (71b36fa)
by postmaster＠open-mesh.org 26 Jun '14

26 Jun '14

Repository : ssh://git@open-mesh.org/alfred On branch : master >--------------------------------------------------------------- commit 71b36fa34beba4beecd2ce397e52b4a8e190a1dd Author: Sven Eckelmann <sven(a)narfation.org> Date: Wed Jun 25 13:35:58 2014 +0200 alfred-gpsd: Avoid underrun when reading from gpsd The gpsd output reading function is ignoring \r characters. This is done by moving the current position (cnt) one position back in the character buffer. It is jumping to the -1 character (max number for size_t) when it was reading the first character at position 0. This is not problematic when the cnt is increased directly after it by 1. Overflows/underflows are defined for *unsigned* types and thus it just jumps back to 0. Unfortunatelly, it is trying to access the memory for another check before increasing the position again. This check is done on memory outside of the buffer and therefore invalid. Instead doing two check after each other, it is in this situation better to do both at once and just handle the current character. Signed-off-by: Sven Eckelmann <sven(a)narfation.org> Signed-off-by: Simon Wunderlich <sw(a)simonwunderlich.de> >--------------------------------------------------------------- 71b36fa34beba4beecd2ce397e52b4a8e190a1dd gpsd/alfred-gpsd.c | 9 +++++---- 1 file changed, 5 insertions(+), 4 deletions(-) diff --git a/gpsd/alfred-gpsd.c b/gpsd/alfred-gpsd.c index d6cdfd6..87943bd 100644 --- a/gpsd/alfred-gpsd.c +++ b/gpsd/alfred-gpsd.c @@ -315,15 +315,16 @@ static void gpsd_read_gpsd(struct globals *globals) return; } - if (buf[cnt] == '\r') + switch (buf[cnt]) { + case '\r': cnt--; - - if (buf[cnt] == '\n') { + break; + case '\n': eol = true; buf[cnt] = '\0'; break; } - } while (cnt++ < sizeof(buf) - 1); + } while (cnt++ < sizeof(buf) - 1 && !eol); if (!eol) { gps_close(&globals->gpsdata);

1 0

[alfred] master: alfred-gpsd: Calculate size of global buffer and not of a single char (c67658a)
by postmaster＠open-mesh.org 26 Jun '14

26 Jun '14

Repository : ssh://git@open-mesh.org/alfred On branch : master >--------------------------------------------------------------- commit c67658a715ad71394b3d4f3a82023b11a085a1fc Author: Sven Eckelmann <sven(a)narfation.org> Date: Wed Jun 25 13:35:57 2014 +0200 alfred-gpsd: Calculate size of global buffer and not of a single char The global buffer is 65536 bytes in size and not only 1 byte. This statement evaluated to a negative value which was casted to an unsigned value (aka... very large value). This is a regression was introduced in 71dbc00fd879f1e592b07d8397f724fb3f69ac64 ("alfred: Use strncpy instead of strcpy for string"). Signed-off-by: Sven Eckelmann <sven(a)narfation.org> Signed-off-by: Simon Wunderlich <sw(a)simonwunderlich.de> >--------------------------------------------------------------- c67658a715ad71394b3d4f3a82023b11a085a1fc gpsd/alfred-gpsd.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/gpsd/alfred-gpsd.c b/gpsd/alfred-gpsd.c index 84a0ded..d6cdfd6 100644 --- a/gpsd/alfred-gpsd.c +++ b/gpsd/alfred-gpsd.c @@ -301,7 +301,7 @@ static void gpsd_read_gpsd(struct globals *globals) size_t cnt; bool eol = false; char buf[4096]; - const size_t tpv_size = sizeof(*globals->buf) - + const size_t tpv_size = sizeof(globals->buf) - sizeof(*globals->push) - sizeof(struct alfred_data) - sizeof(*globals->gpsd_data); @@ -450,7 +450,7 @@ static int gpsd_server(struct globals *globals) int max_fd, ret; const size_t overhead = sizeof(*globals->push) + sizeof(struct alfred_data); - const size_t tpv_size = sizeof(*globals->buf) - + const size_t tpv_size = sizeof(globals->buf) - sizeof(*globals->push) - sizeof(struct alfred_data) - sizeof(*globals->gpsd_data);

1 0

[batman-adv] master: batman-adv: increase default hop penalty (7644650)
by postmaster＠open-mesh.org 24 Jun '14

24 Jun '14

Repository : ssh://git@open-mesh.org/batman-adv On branch : master >--------------------------------------------------------------- commit 7644650bb766bd4c7b6be5d97e6b1c3ed93a38d7 Author: Simon Wunderlich <simon(a)open-mesh.com> Date: Tue Jun 17 12:16:03 2014 +0200 batman-adv: increase default hop penalty The default hop penalty is currently set to 15, which is applied like that for multi interface devices (e.g. dual band APs). Single band devices will still use an effective penalty of 30 (hop penalty + wifi penalty). After receiving reports of too long paths in mesh networks with dual band APs which were fixed by increasing the hop penalty, we'd like to suggest to increase that default value in the default setting as well. We've evaluated that increase in a handful of medium sized mesh networks (5-20 nodes) with single and dual band devices, with changes for the better (shorter routes, higher throughput) or no change at all. This patch changes the hop penalty to 30, which will give an effective penalty of 60 on single band devices (hop penalty + wifi penalty). Signed-off-by: Simon Wunderlich <simon(a)open-mesh.com> Signed-off-by: Marek Lindner <mareklindner(a)neomailbox.ch> >--------------------------------------------------------------- 7644650bb766bd4c7b6be5d97e6b1c3ed93a38d7 soft-interface.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/soft-interface.c b/soft-interface.c index e783afb..9bf382d 100644 --- a/soft-interface.c +++ b/soft-interface.c @@ -757,7 +757,7 @@ static int batadv_softif_init_late(struct net_device *dev) atomic_set(&bat_priv->gw.bandwidth_down, 100); atomic_set(&bat_priv->gw.bandwidth_up, 20); atomic_set(&bat_priv->orig_interval, 1000); - atomic_set(&bat_priv->hop_penalty, 15); + atomic_set(&bat_priv->hop_penalty, 30); #ifdef CONFIG_BATMAN_ADV_DEBUG atomic_set(&bat_priv->log_level, 0); #endif

1 0

[batman-adv] master: Merge branch 'next' (7de3b43)
by postmaster＠open-mesh.org 24 Jun '14

24 Jun '14

Repository : ssh://git@open-mesh.org/batman-adv On branch : master >--------------------------------------------------------------- commit 7de3b43302cc971666fb6526e328659ab04f633e Merge: 8c12d63 8cd753e Author: Marek Lindner <mareklindner(a)neomailbox.ch> Date: Tue Jun 24 23:28:45 2014 +0800 Merge branch 'next' >--------------------------------------------------------------- 7de3b43302cc971666fb6526e328659ab04f633e bridge_loop_avoidance.c | 44 ++++++++++++++++++++++++++++++++++---------- 1 file changed, 34 insertions(+), 10 deletions(-)

1 0

[batman-adv] master's head updated: Merge branch 'next' (7de3b43)
by postmaster＠open-mesh.org 24 Jun '14

24 Jun '14

Repository : ssh://git@open-mesh.org/batman-adv Branch 'master' now includes: 8cd753e batman-adv: drop QinQ claim frames in bridge loop avoidance 7de3b43 Merge branch 'next'

1 0

[batman-adv] next: batman-adv: drop QinQ claim frames in bridge loop avoidance (8cd753e)
by postmaster＠open-mesh.org 24 Jun '14

24 Jun '14

Repository : ssh://git@open-mesh.org/batman-adv On branch : next >--------------------------------------------------------------- commit 8cd753ef306c2273d0da3aff5df758880ca78e32 Author: Simon Wunderlich <simon(a)open-mesh.com> Date: Mon Jun 23 15:55:36 2014 +0200 batman-adv: drop QinQ claim frames in bridge loop avoidance Since bridge loop avoidance only supports untagged or simple 802.1q tagged VLAN claim frames, claim frames with stacked VLAN headers (QinQ) should be detected and dropped. Transporting the over the mesh may cause problems on the receivers, or create bogus entries in the local tt tables. Reported-by: Antonio Quartulli <antonio(a)open-mesh.com> Signed-off-by: Simon Wunderlich <simon(a)open-mesh.com> Signed-off-by: Marek Lindner <mareklindner(a)neomailbox.ch> >--------------------------------------------------------------- 8cd753ef306c2273d0da3aff5df758880ca78e32 bridge_loop_avoidance.c | 44 ++++++++++++++++++++++++++++++++++---------- 1 file changed, 34 insertions(+), 10 deletions(-) diff --git a/bridge_loop_avoidance.c b/bridge_loop_avoidance.c index 6f0d9ec..a957c81 100644 --- a/bridge_loop_avoidance.c +++ b/bridge_loop_avoidance.c @@ -800,11 +800,6 @@ static int batadv_check_claim_group(struct batadv_priv *bat_priv, bla_dst = (struct batadv_bla_claim_dst *)hw_dst; bla_dst_own = &bat_priv->bla.claim_dest; - /* check if it is a claim packet in general */ - if (memcmp(bla_dst->magic, bla_dst_own->magic, - sizeof(bla_dst->magic)) != 0) - return 0; - /* if announcement packet, use the source, * otherwise assume it is in the hw_src */ @@ -866,12 +861,13 @@ static int batadv_bla_process_claim(struct batadv_priv *bat_priv, struct batadv_hard_iface *primary_if, struct sk_buff *skb) { - struct batadv_bla_claim_dst *bla_dst; + struct batadv_bla_claim_dst *bla_dst, *bla_dst_own; uint8_t *hw_src, *hw_dst; - struct vlan_ethhdr *vhdr; + struct vlan_hdr *vhdr, vhdr_buf; struct ethhdr *ethhdr; struct arphdr *arphdr; unsigned short vid; + int vlan_depth = 0; __be16 proto; int headlen; int ret; @@ -882,9 +878,24 @@ static int batadv_bla_process_claim(struct batadv_priv *bat_priv, proto = ethhdr->h_proto; headlen = ETH_HLEN; if (vid & BATADV_VLAN_HAS_TAG) { - vhdr = vlan_eth_hdr(skb); - proto = vhdr->h_vlan_encapsulated_proto; - headlen += VLAN_HLEN; + /* Traverse the VLAN/Ethertypes. + * + * At this point it is known that the first protocol is a VLAN + * header, so start checking at the encapsulated protocol. + * + * The depth of the VLAN headers is recorded to drop BLA claim + * frames encapsulated into multiple VLAN headers (QinQ). + */ + do { + vhdr = skb_header_pointer(skb, headlen, VLAN_HLEN, + &vhdr_buf); + if (!vhdr) + return 0; + + proto = vhdr->h_vlan_encapsulated_proto; + headlen += VLAN_HLEN; + vlan_depth++; + } while (proto == htons(ETH_P_8021Q)); } if (proto != htons(ETH_P_ARP)) @@ -914,6 +925,19 @@ static int batadv_bla_process_claim(struct batadv_priv *bat_priv, hw_src = (uint8_t *)arphdr + sizeof(struct arphdr); hw_dst = hw_src + ETH_ALEN + 4; bla_dst = (struct batadv_bla_claim_dst *)hw_dst; + bla_dst_own = &bat_priv->bla.claim_dest; + + /* check if it is a claim frame in general */ + if (memcmp(bla_dst->magic, bla_dst_own->magic, + sizeof(bla_dst->magic)) != 0) + return 0; + + /* check if there is a claim frame encapsulated deeper in (QinQ) and + * drop that, as this is not supported by BLA but should also not be + * sent via the mesh. + */ + if (vlan_depth > 1) + return 1; /* check if it is a claim frame. */ ret = batadv_check_claim_group(bat_priv, primary_if, hw_src, hw_dst,

1 0

[linux-merge]linux integration; annotated tag, v3.16-rc2, created. v3.16-rc2
by postmaster＠open-mesh.org 22 Jun '14

22 Jun '14

The annotated tag, v3.16-rc2 has been created at d2d9c4d67d98c18cbe96b84ba8cb3edc2a1eb59e (tag) tagging a497c3ba1d97fc69c1e78e7b96435ba8c2cb42ee (commit) replaces v3.16-rc1 tagged by Linus Torvalds on Sat Jun 21 19:03:04 2014 -1000 - Shortlog ------------------------------------------------------------ Linux 3.16-rc2 -----BEGIN PGP SIGNATURE----- Version: GnuPG v1 iQEcBAABAgAGBQJTpmOIAAoJEHm+PkMAQRiGAagH+wQEtx2mX1vh4zVRgtN2ciOd HbRbR8BUcZ9xIfnxudaoxiN1mS/4yW+Ju09CiP1w0u504AGajKOfJMSfAbusClwJ 4PX4wdqiLgcnDlWEtoVMwcVS2O0IXidQrs1xVudUW6EGB4WohLEV2biTuNwavBoK d/QckhRV4QEUD91hVhnylXAWerAVp2E5oNHvke96x3vjOZyh3v0C+mva5GI9/wS/ R4tnXxNpkUwEPecAGdiIbwlZfk/uTyesxBzlRPPlOOeKRkSeoP1kFPv6PJKdoGyc 3oet5hqzSgnzZqiVLP5tQluE7f3SlFHo8Q2R4KXDFYVHqXICvrVll3XOum8sGZ0= =Z1wt -----END PGP SIGNATURE----- ----------------------------------------------------------------------- -- linux integration

1 0

[linux-merge]linux integration; annotated tag, v3.16-rc1, created. v3.16-rc1
by postmaster＠open-mesh.org 16 Jun '14

16 Jun '14

The annotated tag, v3.16-rc1 has been created at b79831e695f2e100b056794ed586eca25ee13ddb (tag) tagging 7171511eaec5bf23fb06078f59784a3a0626b38f (commit) replaces v3.15 tagged by Linus Torvalds on Sun Jun 15 17:45:40 2014 -1000 - Shortlog ------------------------------------------------------------ Linux 3.16-rc1 -----BEGIN PGP SIGNATURE----- Version: GnuPG v1 iQEcBAABAgAGBQJTnmhkAAoJEHm+PkMAQRiGYyoIAJ+dCxQjx0Jmu5VLs48yvUkx AVbnJeEq0ScgFPBAlfrGnnXczOyZGD+wveNRwb3bN4f6hGkWHPncfExAeTU35QQ+ 92kCLScPU6kn4tnTqjac4ZrUhBCfgg5hFFamXOPidVDG4f5wYwql3IvP4O3eMOcZ IOx7kgweqoPm4A/LjXQ3L21kghs88NXySWMwwfWFdXaV++ey07slCWLzGF5rMDh/ xfBDfgTS4gdrbGeE+1z/qkoWyHwKnCad8Uh2Fu5CcprElwCjLLhrLPccYSRKO2IR 2ZSj/mcMb1FhH7AOyXBYMVbjhOH5MCIlHvJYYp7kwHfs66UTmnkczOJxq1ynKP0= =Whty -----END PGP SIGNATURE----- Antonio Quartulli (2): batman-adv: add blank line between declarations and the rest of the code batman-adv: remove semi-colon after macro definition David S. Miller (2): Merge git://git.kernel.org/pub/scm/linux/kernel/git/davem/net Merge git://git.kernel.org/pub/scm/linux/kernel/git/davem/net Octavian Purdila (1): net: add __pskb_copy_fclone and pskb_copy_for_clone Simon Wunderlich (1): batman-adv: Start new development cycle Wilfried Klaebe (1): net: get rid of SET_ETHTOOL_OPS ----------------------------------------------------------------------- -- linux integration

1 0

2024

2023

2022

2021

2020

2019

2018

2017

2016

2015

2014

2013

2012

2011

2010

2009

commits June 2014