commits June 2014

commits@lists.open-mesh.org

1 participants
39 discussions

[linux-merge]linux integration; branch, merge/master, updated. v3.14-rc1-2152-ge1b9cf4
by postmaster＠open-mesh.org 30 Jun '14

30 Jun '14

The following commit has been merged in the merge/master branch: commit e1b9cf4b021b60f664ece222c47b1c7b3c38d7d2 Merge: 150985e7766f4864b025efc3c5fac8b922d371e2 8cd753ef306c2273d0da3aff5df758880ca78e32 Author: Antonio Quartulli <antonio(a)meshcoding.com> Date: Sun Jun 29 01:25:34 2014 +0200 Merge commit '8cd753ef306c2273d0da3aff5df758880ca78e32' into merge/master diff --combined net/batman-adv/bridge_loop_avoidance.c index 6f0d9ec,a957c81..a957c81 --- a/net/batman-adv/bridge_loop_avoidance.c +++ b/net/batman-adv/bridge_loop_avoidance.c @@@ -800,11 -800,6 +800,6 @@@ static int batadv_check_claim_group(str 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 +861,13 @@@ static int batadv_bla_process_claim(str 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 +878,24 @@@ 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 +925,19 @@@ 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, -- linux integration

1 0

[batman-adv] master's head updated: Merge branch 'next' (6ae4281)
by postmaster＠open-mesh.org 30 Jun '14

30 Jun '14

Repository : ssh://git@open-mesh.org/batman-adv Branch 'master' now includes: 11be1ac batman-adv: fix a bogus warning from batadv_is_on_batman_iface() 6ae4281 Merge branch 'next'

1 0

[batman-adv] master: Merge branch 'next' (6ae4281)
by postmaster＠open-mesh.org 30 Jun '14

30 Jun '14

Repository : ssh://git@open-mesh.org/batman-adv On branch : master >--------------------------------------------------------------- commit 6ae4281474675fbca5bedcf768972a32db586eb6 Merge: 7644650 11be1ac Author: Marek Lindner <mareklindner(a)neomailbox.ch> Date: Mon Jun 30 14:12:59 2014 +0800 Merge branch 'next' >--------------------------------------------------------------- 6ae4281474675fbca5bedcf768972a32db586eb6 hard-interface.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-)

1 0

[batman-adv] next: batman-adv: fix a bogus warning from batadv_is_on_batman_iface() (11be1ac)
by postmaster＠open-mesh.org 30 Jun '14

30 Jun '14

Repository : ssh://git@open-mesh.org/batman-adv On branch : next >--------------------------------------------------------------- commit 11be1ac9cacdf843cdaaa0ae9f4dfe7206ac6d35 Author: Cong Wang <cwang(a)twopensource.com> Date: Thu May 22 11:57:17 2014 -0700 batman-adv: fix a bogus warning from batadv_is_on_batman_iface() batman tries to search dev->iflink to check if it's a batman interface, but ->iflink could be 0, which is not a valid ifindex. It should just avoid iflink == 0 case. Reported-by: Jet Chen <jet.chen(a)intel.com> Tested-by: Jet Chen <jet.chen(a)intel.com> Cc: David S. Miller <davem(a)davemloft.net> Cc: Steffen Klassert <steffen.klassert(a)secunet.com> Cc: Antonio Quartulli <antonio(a)open-mesh.com> Cc: Marek Lindner <mareklindner(a)neomailbox.ch> Signed-off-by: Cong Wang <xiyou.wangcong(a)gmail.com> Signed-off-by: Cong Wang <cwang(a)twopensource.com> Signed-off-by: David S. Miller <davem(a)davemloft.net> Signed-off-by: Marek Lindner <mareklindner(a)neomailbox.ch> >--------------------------------------------------------------- 11be1ac9cacdf843cdaaa0ae9f4dfe7206ac6d35 hard-interface.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/hard-interface.c b/hard-interface.c index b851cc5..fbda6b5 100644 --- a/hard-interface.c +++ b/hard-interface.c @@ -83,7 +83,7 @@ static bool batadv_is_on_batman_iface(const struct net_device *net_dev) return true; /* no more parents..stop recursion */ - if (net_dev->iflink == net_dev->ifindex) + if (net_dev->iflink == 0 || net_dev->iflink == net_dev->ifindex) return false; /* recurse over the parent device */

1 0

[linux-merge]linux integration; annotated tag, v3.16-rc3, created. v3.16-rc3
by postmaster＠open-mesh.org 29 Jun '14

29 Jun '14

The annotated tag, v3.16-rc3 has been created at dee9ef9e4858ca40b659cbec3204f3dd2ab4acca (tag) tagging 4c834452aad01531db949414f94f817a86348d59 (commit) replaces v3.16-rc2 tagged by Linus Torvalds on Sun Jun 29 14:11:44 2014 -0700 - Shortlog ------------------------------------------------------------ Linux 3.16-rc3 -----BEGIN PGP SIGNATURE----- Version: GnuPG v1 iQEcBAABAgAGBQJTsIEQAAoJEHm+PkMAQRiGPQgH/0Bs0Lw4MSnAMdS3B5Le8ZGp 9pFXZ3aK4jQErfA2spKtXDTRiQ9QOxzL6ZpDzrm36N98iX87402dPA+3wdea3HiK LhPggj3d9hX8/L2yaBk7eY783NcPs2dTlqtLFx6k5UwGhZc1KmCcaDmZ4poI1yca 64pYT7zCCGSlKIbuiXoO5h/tTA0BzS+dAZZIsDVBHehXR4I37KTCDKQEn9V077P8 5+MLnyjvAgiDQ1+yHrTezdQHE2iKWJ1GDK11aTN85qzGt5fCWi5JlYwdLP/rC5wi HcjPL/H+IgbZOhttnYmIE2yVJk7LPoRWwbf9LG5sKkBomqVe7Gh1GQKIYy69aOA= =cfzs -----END PGP SIGNATURE----- ----------------------------------------------------------------------- -- linux integration

1 0

[batmand] master: batmand: Don't try to schedule_own_packet with no if_incoming (3a1bed8)
by postmaster＠open-mesh.org 29 Jun '14

29 Jun '14

Repository : ssh://git@open-mesh.org/batmand On branch : master >--------------------------------------------------------------- commit 3a1bed88a500d2fe857fa889a90c52340298979b Author: Sven Eckelmann <sven(a)narfation.org> Date: Tue Jun 17 21:56:03 2014 +0200 batmand: Don't try to schedule_own_packet with no if_incoming send_outstanding_packets checks if a forw_node has a correct if_incoming. Otherwise it jumps to packet_free to deallocate the packet infrastructure. But this also schedules packets with the batman interfaces as target incoming_if. This is known to be NULL but is dereferenced in schedule_own_packet. This NULL dereference should be avoided. Signed-off-by: Sven Eckelmann <sven(a)narfation.org> Signed-off-by: Elektra Wagenrad <onelektra(a)gmx.net> >--------------------------------------------------------------- 3a1bed88a500d2fe857fa889a90c52340298979b schedule.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/schedule.c b/schedule.c index 3e65d68..03cdb78 100644 --- a/schedule.c +++ b/schedule.c @@ -359,7 +359,7 @@ void send_outstanding_packets(uint32_t curr_time) packet_free: list_del((struct list_head *)&forw_list, forw_pos, &forw_list); - if (forw_node->own) + if (forw_node->own && forw_node->if_incoming) schedule_own_packet(forw_node->if_incoming); debugFree(forw_node->pack_buff, 1501);

1 0

[batmand] master: batmand: Don't dereference orig_node before checking for NULL (036aea1)
by postmaster＠open-mesh.org 29 Jun '14

29 Jun '14

Repository : ssh://git@open-mesh.org/batmand On branch : master >--------------------------------------------------------------- commit 036aea1fe69d275aecae64cc643e8e34fac550a0 Author: Sven Eckelmann <sven(a)narfation.org> Date: Tue Jun 17 21:56:04 2014 +0200 batmand: Don't dereference orig_node before checking for NULL The update_route functions first stores the orig_node->router in an extra variable and later checks if orig_node is NULL. This is not only a potential cause of a crash but can also cause new compilers to drop the NULL check completely [1]. [1] https://gcc.gnu.org/onlinedocs/gcc-3.4.3/gcc/Optimize-Options.html#index-fd… Signed-off-by: Sven Eckelmann <sven(a)narfation.org> Signed-off-by: Elektra Wagenrad <onelektra(a)gmx.net> >--------------------------------------------------------------- 036aea1fe69d275aecae64cc643e8e34fac550a0 batman.c | 6 +++--- 1 file changed, 3 insertions(+), 3 deletions(-) diff --git a/batman.c b/batman.c index 89b3bf1..c3dfa7f 100644 --- a/batman.c +++ b/batman.c @@ -353,10 +353,9 @@ void update_routes(struct orig_node *orig_node, struct neigh_node *neigh_node, u prof_start(PROF_update_routes); debug_output(4, "update_routes() \n"); - old_router = orig_node->router; - /* also handles orig_node->router == NULL and neigh_node == NULL */ if ((orig_node != NULL) && (orig_node->router != neigh_node)) { + old_router = orig_node->router; if ( ( orig_node != NULL ) && ( neigh_node != NULL ) ) { addr_to_string( orig_node->orig, orig_str, ADDR_STR_LEN ); @@ -415,7 +414,8 @@ void update_routes(struct orig_node *orig_node, struct neigh_node *neigh_node, u orig_node->router = neigh_node; } else if (orig_node != NULL) { - hna_global_update(orig_node, hna_recv_buff, hna_buff_len, old_router); + hna_global_update(orig_node, hna_recv_buff, hna_buff_len, + orig_node->router); } prof_stop(PROF_update_routes);

1 0

[batmand] master: batmand: Fix bitarray 1 bit shift type (0c2aa5e)
by postmaster＠open-mesh.org 29 Jun '14

29 Jun '14

Repository : ssh://git@open-mesh.org/batmand On branch : master >--------------------------------------------------------------- commit 0c2aa5e93522ade7f955e071a35e0c7b015a497b Author: Sven Eckelmann <sven(a)narfation.org> Date: Tue Jun 17 21:56:00 2014 +0200 batmand: Fix bitarray 1 bit shift type The default type of an integer constant is int. This reduced the possible bits for a shift to 32. But the size of uintmax_t is most likely larger (64 bit). Thus the upper 32 bit cannot be accessed correctly with this bitarray implementation. Signed-off-by: Sven Eckelmann <sven(a)narfation.org> Signed-off-by: Elektra Wagenrad <onelektra(a)gmx.net> >--------------------------------------------------------------- 0c2aa5e93522ade7f955e071a35e0c7b015a497b bitarray.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/bitarray.c b/bitarray.c index 133f6e7..354eddd 100644 --- a/bitarray.c +++ b/bitarray.c @@ -49,7 +49,7 @@ uint8_t get_bit_status( TYPE_OF_WORD *seq_bits, uint16_t last_seqno, uint16_t cu word_offset= ( last_seqno - curr_seqno ) % WORD_BIT_SIZE; /* which position in the selected word */ word_num = ( last_seqno - curr_seqno ) / WORD_BIT_SIZE; /* which word */ - if ( seq_bits[word_num] & 1<<word_offset ) /* get position status */ + if ( seq_bits[word_num] & (1ULL << word_offset)) /* get position status */ return 1; else return 0; @@ -72,7 +72,7 @@ void bit_mark( TYPE_OF_WORD *seq_bits, int32_t n ) { word_offset= n%WORD_BIT_SIZE; /* which position in the selected word */ word_num = n/WORD_BIT_SIZE; /* which word */ - seq_bits[word_num]|= 1<<word_offset; /* turn the position on */ + seq_bits[word_num] |= 1ULL << word_offset; /* turn the position on */ } /* shift the packet array p by n places. */

1 0

[batmand] master: batmand: Free socket when setsockopt failed (4de9e4b)
by postmaster＠open-mesh.org 29 Jun '14

29 Jun '14

Repository : ssh://git@open-mesh.org/batmand On branch : master >--------------------------------------------------------------- commit 4de9e4b5b707631d065571a437391c2a0de11052 Author: Sven Eckelmann <sven(a)narfation.org> Date: Tue Jun 17 21:56:01 2014 +0200 batmand: Free socket when setsockopt failed client_to_gw_tun calls setsockopt which can fail. In this case it jumps to the error handling and cleanup code but doesn't close the udp_sock. This has to be done to avoid leaking of file descriptors. Signed-off-by: Sven Eckelmann <sven(a)narfation.org> Signed-off-by: Elektra Wagenrad <onelektra(a)gmx.net> >--------------------------------------------------------------- 4de9e4b5b707631d065571a437391c2a0de11052 posix/tunnel.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/posix/tunnel.c b/posix/tunnel.c index 08c56a2..8e7611b 100644 --- a/posix/tunnel.c +++ b/posix/tunnel.c @@ -201,7 +201,7 @@ void *client_to_gw_tun(void *arg) sock_opts = 1; if (setsockopt(udp_sock, SOL_SOCKET, SO_REUSEADDR, &sock_opts, sizeof(sock_opts)) < 0) { debug_output(0, "Error - can't set options on udp socket: %s\n", strerror(errno)); - goto out; + goto udp_out; } if (bind(udp_sock, (struct sockaddr *)&my_addr, sizeof(struct sockaddr_in)) < 0) {

1 0

[batmand] master: batmand: Don't cause NULL ptr dereference when searching through HNAs (f284128)
by postmaster＠open-mesh.org 29 Jun '14

29 Jun '14

Repository : ssh://git@open-mesh.org/batmand On branch : master >--------------------------------------------------------------- commit f2841283a3357c895b5d88c0b3be84b0dab6c665 Author: Sven Eckelmann <sven(a)narfation.org> Date: Tue Jun 17 21:56:02 2014 +0200 batmand: Don't cause NULL ptr dereference when searching through HNAs The _hna_global_add function searches through the list of HNAs to check if there is already a hna for this orig pointer. Otherwise it is automatically allocated. This was implemented by setting hna_orig_ptr to NULL after the comparison. This will most likely crash the program because the list_for_each_entry implementation uses hna_orig_ptr to find the next entry before doing the loop condition check. The next entry cannot be found by dereferencing NULL+epsilon. Signed-off-by: Sven Eckelmann <sven(a)narfation.org> Signed-off-by: Elektra Wagenrad <onelektra(a)gmx.net> >--------------------------------------------------------------- f2841283a3357c895b5d88c0b3be84b0dab6c665 hna.c | 9 +++++---- 1 file changed, 5 insertions(+), 4 deletions(-) diff --git a/hna.c b/hna.c index f0d64e1..0f28bfe 100644 --- a/hna.c +++ b/hna.c @@ -322,6 +322,7 @@ static void _hna_global_add(struct orig_node *orig_node, struct hna_element *hna struct hna_orig_ptr *hna_orig_ptr = NULL; struct orig_node *old_orig_node = NULL; struct hashtable_t *swaphash; + int found = 0; hna_global_entry = ((struct hna_global_entry *)hash_find(hna_global_hash, hna_element)); @@ -354,14 +355,14 @@ static void _hna_global_add(struct orig_node *orig_node, struct hna_element *hna return; list_for_each_entry(hna_orig_ptr, &hna_global_entry->orig_list, list) { - if (hna_orig_ptr->orig_node == orig_node) + if (hna_orig_ptr->orig_node == orig_node) { + found = 1; break; - - hna_orig_ptr = NULL; + } } /* append the given orig node to the list */ - if (!hna_orig_ptr) { + if (!found) { hna_orig_ptr = debugMalloc(sizeof(struct hna_orig_ptr), 704); if (!hna_orig_ptr)

1 0

2023

2022

2021

2020

2019

2018

2017

2016

2015

2014

2013

2012

2011

2010

2009

commits June 2014