B.A.T.M.A.N

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

15 participants
4308 discussions

Re: [B.A.T.M.A.N.] Patch which fixes a uint32overflow ; fix algorithm which was unintented changed on 2014 update
by Sven Eckelmann 17 May '15

17 May '15

Please don't drop the mailing list when replying (unless it is actually private conversation). On Sunday 17 May 2015 11:21:10 Ruben Wisniewski wrote: > > should be included: > Since the overflow-bug seem to be is since epoch, the patch have to be > applied to all versions. I am talking about the part which you described yourself as regression. But I agree that the overflow bug is in there since the beginning of the gateway code. > But I think batman-adv 2013.x is that old that > nobody is using it anymore or kernels with that version. So we just > have to fix all 2014.x versions. There is a stable branch for Linux 2.6.32.x. Linux 2.6.32 came out 2009. So it is possible that David Miller decides to submit the overflow patch to stable for 2.6.32. This is not in our jurisdiction. The same may be the case for the regression - but then somebody (David Miller/stable team) must know how severe the problem is and when it was introduced. Kind regards, Sven

1 0

Re: [B.A.T.M.A.N.] Patch to add mesh_no_rebroadcast
by Sven Eckelmann 17 May '15

17 May '15

Please don't drop the mailing list when replying (unless it is actually private conversation). On Sunday 17 May 2015 11:12:08 Ruben Wisniewski wrote: > schrieb Sven Eckelmann <sven(a)narfation.org>: > > Why are you sending a patch from Linus? > > Because Simon said to me to do so yesterday. :) See my other mail with a more detailed explanation. > > And this patch doesn't apply on current master or next. > > We have this patch in our setup and it's applying to 2015.0. But v2015.0 is not master or next. And since it is a new feature, I would guess it should be applied on master. Kind regards, Sven

1 0

[B.A.T.M.A.N.] Patch which fixes a uint32overflow ; fix algorithm which was unintented changed on 2014 update
by Ruben Wisniewski 17 May '15

17 May '15

Changelog * fix overflow of uint32-value while multiplying * restore algorithm to origin calculation which expect kbit/s values Signed-off-by: Ruben Wisniewsi <ruben(a)vfn-nrw.de> --- net/batman-adv/gateway_client.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/net/batman-adv/gateway_client.c b/net/batman-adv/gateway_client.c index bb01586..6f00584 100644 --- a/net/batman-adv/gateway_client.c +++ b/net/batman-adv/gateway_client.c @@ -153,7 +153,7 @@ batadv_gw_get_best_gw_node(struct batadv_priv *bat_priv) struct batadv_neigh_node *router; struct batadv_neigh_ifinfo *router_ifinfo; struct batadv_gw_node *gw_node, *curr_gw = NULL; - uint32_t max_gw_factor = 0, tmp_gw_factor = 0; + uint64_t max_gw_factor = 0, tmp_gw_factor = 0; uint32_t gw_divisor; uint8_t max_tq = 0; uint8_t tq_avg; @@ -185,7 +185,7 @@ batadv_gw_get_best_gw_node(struct batadv_priv *bat_priv) switch (atomic_read(&bat_priv->gw_sel_class)) { case 1: /* fast connection */ tmp_gw_factor = tq_avg * tq_avg; - tmp_gw_factor *= gw_node->bandwidth_down; + tmp_gw_factor *= gw_node->bandwidth_down * 100; tmp_gw_factor *= 100 * 100; tmp_gw_factor /= gw_divisor; -- 2.4.0

2 2

[B.A.T.M.A.N.] [PATCHv5 0/3] batman-adv: Unicasting multicast reports to querier-node only
by Linus Lüssing 17 May '15

17 May '15

The last round of multicast patches send to the batman-adv mailinglist to add support for the multicast optimizations in bridged scenarios, too, unfortunately had one major conceptual flaw: It could lead to packet loss. It's not sufficient to have the unicasting of reports implemented on bridge-nodes only. Nodes without bridges need to treat reports the same way. The issue is described in detail here: https://www.open-mesh.org/projects/batman-adv/wiki/Multicast-optimizations-… Cheers, Linus ----- Changelog: v5: * Removed RFC tag: Needed exports got merged to net-next and are going to be available with Linux 4.2 * Redid compat solution - now fully backwards compatible down to 2.6.33 v4: * excluded bridge part from this patchset, they should hopefully be added to net-next soon * Added a compat solution (PATCH 3/3) * Removed Kconfig-depends as by David's suggestion the needed parsing functions for MLD are going to be forced built-ins even if IPv6 is going to be built as a module * Removed unused variable 'int ret' in batadv_mcast_is_report_ipv6() * Adjusted to new folder structure v3: * Adding Kconfig-depends and #if's (so basically adding similar dependancy constraints as the bridge code has, except that there are no depends if batman-adv gets compiled without multicast optimizations) -> the case of IPv6=M and batman-adv=y is still impossible if multicast optimizations are enabled; but I don't see the practical demand for that either - people who use IPv6 as a module will probably also want to use batman-adv as a module v2: * various bugfixes (now runtime tested, too - should(tm) work) * added netdev+bridge mailinglists

2 6

[B.A.T.M.A.N.] [PATCHv6 0/3] batman-adv: Unicasting multicast reports to querier-node only
by Linus Lüssing 17 May '15

17 May '15

The last round of multicast patches send to the batman-adv mailinglist to add support for the multicast optimizations in bridged scenarios, too, unfortunately had one major conceptual flaw: It could lead to packet loss. It's not sufficient to have the unicasting of reports implemented on bridge-nodes only. Nodes without bridges need to treat reports the same way. The issue is described in detail here: https://www.open-mesh.org/projects/batman-adv/wiki/Multicast-optimizations-… Cheers, Linus ----- Changelog: v6: * compat: copied copyright headers from original upstream c files * compat: unified ordering in compat c files: -> copyright header, then includes, then kernel specific functions v5: * Removed RFC tag: Needed exports got merged to net-next and are going to be available with Linux 4.2 * Redid compat solution - now fully backwards compatible down to 2.6.33 v4: * excluded bridge part from this patchset, they should hopefully be added to net-next soon * Added a compat solution (PATCH 3/3) * Removed Kconfig-depends as by David's suggestion the needed parsing functions for MLD are going to be forced built-ins even if IPv6 is going to be built as a module * Removed unused variable 'int ret' in batadv_mcast_is_report_ipv6() * Adjusted to new folder structure v3: * Adding Kconfig-depends and #if's (so basically adding similar dependancy constraints as the bridge code has, except that there are no depends if batman-adv gets compiled without multicast optimizations) -> the case of IPv6=M and batman-adv=y is still impossible if multicast optimizations are enabled; but I don't see the practical demand for that either - people who use IPv6 as a module will probably also want to use batman-adv as a module v2: * various bugfixes (now runtime tested, too - should(tm) work) * added netdev+bridge mailinglists

1 3

[B.A.T.M.A.N.] [PATCH] gw-bandwidth: fix potential overflow on very large input values, limit them to 100 GBit/s
by Ruben Wisniewski 16 May '15

16 May '15

Signed-off-by: Ruben Wisniewsi <ruben(a)vfn-nrw.de> --- net/batman-adv/gateway_common.c | 6 ++++++ 1 file changed, 6 insertions(+) diff --git a/net/batman-adv/gateway_common.c b/net/batman-adv/gateway_common.c index 39cf44c..6b0f4d3 100644 --- a/net/batman-adv/gateway_common.c +++ b/net/batman-adv/gateway_common.c @@ -71,10 +71,16 @@ static bool batadv_parse_gw_bandwidth(struct net_device *net_dev, char *buff, switch (bw_unit_type) { case BATADV_BW_UNIT_MBIT: + /* limit input to 100 GBit/s */ + if (ldown > 100000) + ldown = 100000 *down = ldown * 10; break; case BATADV_BW_UNIT_KBIT: default: + /* limit input to 100 GBit/s */ + if (ldown > 100000000) + ldown = 100000000 *down = ldown / 100; break; } -- 2.4.0

2 1

[B.A.T.M.A.N.] Patches for David Miller/net-next
by Sven Eckelmann 11 May '15

11 May '15

Hi, here are the patches which must be submitted (signed/polished/...) to net- next. I've also included a bundle to get the same merge in linux-merge.git as I did: $ git clone git://git.open-mesh.org/linux-merge.git $ cd linux-merge $ git remote add net-next \ git://git.kernel.org/pub/scm/linux/kernel/git/davem/net-next.git $ git remote add pkg git://git.open-mesh.org/batman-adv.git $ git remote update $ git bundle unbundle ~/merge_master.bundle refs/heads/merge/master $ git merge 685fb7ce925b138c8d53718d8622821eddc0c514 # NOW PUSH IT I've only tested it on net-next 3bb45001ac33b4f733e1e8ffb01fb07baccd528c ("Merge branch 'handle_ing_lightweight'") Kind regards, Sven

1 0

[B.A.T.M.A.N.] FW: using Batman-adv
by Joseph Zimmer 08 May '15

08 May '15

Hello, I am working with BATMAN-ADV and have successfully setup a mesh. In my implementation of the mesh, the network requires BATMAN to operate in a linear topology and ideally have node 1 hop to node 2 then through node 3 and finally node 4. Node1.Node2..Node3Node4 In this network hops are preferred rather than distant links with low RSSI values(-70). The problem that Im seeing is that the protocol chooses routes that have the least hops but have RSSI values at about -78 to -90 rather than routing through the closest node with RSSI values in the -50 range. We adjusted the hop penalty to a low value (we used 1 and 0) and we did see improvement with the hop selection but the throughput is very low due to an RSSI of -64 or less. Questions: 1) Is there a way to adjust BATMAN using batctl to select routes based on the RSSI value? a. That is, choose to hop through a node with an RSSI of -50 and never select a node that has a lower RSSI 2) Does the algorithm for route selection use RSSI in the calculation and if so can we modify the algorithm to increase the importance of RSSI in the selection? 3) How do we gain access to the code to make these modifications? a. Can you give me a little guidance on how to modify BATMAN-adv to suite my situation? Your help is greatly appreciated. Regards, Joseph Zimmer Telegrid Technology Inc. 19 Microlab Road Livingston, NJ 07039 Email: j.zimmer(a)telegrid.com Phone: 973.994.4440 X319

4 5

[B.A.T.M.A.N.] [PATCH] alfred: Fix dependency to IPv6
by Sven Eckelmann 05 May '15

05 May '15

alfred always requires IPv6 support in the kernel and not only when "Enable IPv6 support in packages" is enabled. Signed-off-by: Sven Eckelmann <sven(a)open-mesh.com> --- alfred/Makefile | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/alfred/Makefile b/alfred/Makefile index dade218..8f6855b 100644 --- a/alfred/Makefile +++ b/alfred/Makefile @@ -28,7 +28,7 @@ define Package/alfred SECTION:=net CATEGORY:=Network TITLE:=A.L.F.R.E.D. - Almighty Lightweight Fact Remote Exchange Daemon - DEPENDS:= +libc +IPV6:kmod-ipv6 +librt \ + DEPENDS:= +libc +kmod-ipv6 +librt \ +ALFRED_NEEDS_lua:lua \ +ALFRED_NEEDS_libgps:libgps endef -- 2.1.4

2 1

[B.A.T.M.A.N.] [PATCH] alfred: Make sure only link-local EUI64 is used as source
by Sven Eckelmann 05 May '15

05 May '15

The linux kernel automatically chooses the source address. This can make problems when the metric prefers an address which is not the link-local EUI64 address. For example an user can add an additional address which also can be used for link-local communication but has a higher value in the metric. This can cause alfred to send UDP link-local packets with an address which cannot be decoded by the receiver to get the MAC address of the device. The communication with this device would then fail. It is possible to define the source address when we directly bind to it. This has the problem that this address cannot receive packets with a multicast address as destination. The workaround is to create two sockets for one netsock structure. One is the socket which can send all data and receives unicast packets. The second one is receiving the multicast traffic like the announcements from master alfred server processes. Signed-off-by: Sven Eckelmann <sven(a)narfation.org> --- alfred.h | 4 ++- netsock.c | 106 ++++++++++++++++++++++++++++++++++++++++++++++++++++++-------- recv.c | 5 +-- send.c | 2 ++ server.c | 2 ++ 5 files changed, 103 insertions(+), 16 deletions(-) diff --git a/alfred.h b/alfred.h index 621805d..adf788c 100644 --- a/alfred.h +++ b/alfred.h @@ -102,6 +102,7 @@ struct interface { uint32_t scope_id; char *interface; int netsock; + int netsock_mcast; struct hashtable_t *server_hash; @@ -151,7 +152,8 @@ int alfred_client_set_data(struct globals *globals); int alfred_client_modeswitch(struct globals *globals); int alfred_client_change_interface(struct globals *globals); /* recv.c */ -int recv_alfred_packet(struct globals *globals, struct interface *interface); +int recv_alfred_packet(struct globals *globals, struct interface *interface, + int recv_sock); struct transaction_head * transaction_add(struct globals *globals, struct ether_addr mac, uint16_t id); struct transaction_head * diff --git a/netsock.c b/netsock.c index 592f3e3..6bf6016 100644 --- a/netsock.c +++ b/netsock.c @@ -84,6 +84,8 @@ void netsock_close_all(struct globals *globals) list_for_each_entry_safe(interface, is, &globals->interfaces, list) { if (interface->netsock >= 0) close(interface->netsock); + if (interface->netsock_mcast >= 0) + close(interface->netsock_mcast); list_del(&interface->list); hash_delete(interface->server_hash, free); free(interface->interface); @@ -144,6 +146,7 @@ int netsock_set_interfaces(struct globals *globals, char *interfaces) interface->scope_id = 0; interface->interface = NULL; interface->netsock = -1; + interface->netsock_mcast = -1; interface->server_hash = NULL; interface->interface = strdup(token); @@ -210,11 +213,14 @@ out: static int netsock_open(struct interface *interface) { int sock; - struct sockaddr_in6 sin6; + int sock_mc; + struct sockaddr_in6 sin6, sin6_mc; + struct ipv6_mreq mreq; struct ifreq ifr; int ret; interface->netsock = -1; + interface->netsock_mcast = -1; sock = socket(PF_INET6, SOCK_DGRAM, IPPROTO_UDP); if (sock < 0) { @@ -222,6 +228,13 @@ static int netsock_open(struct interface *interface) return -1; } + sock_mc = socket(PF_INET6, SOCK_DGRAM, IPPROTO_UDP); + if (sock_mc < 0) { + close(sock); + perror("can't open socket"); + return -1; + } + memset(&ifr, 0, sizeof(ifr)); strncpy(ifr.ifr_name, interface->interface, IFNAMSIZ); ifr.ifr_name[IFNAMSIZ - 1] = '\0'; @@ -232,12 +245,6 @@ static int netsock_open(struct interface *interface) interface->scope_id = ifr.ifr_ifindex; - memset(&sin6, 0, sizeof(sin6)); - sin6.sin6_port = htons(ALFRED_PORT); - sin6.sin6_family = AF_INET6; - sin6.sin6_addr = in6addr_any; - sin6.sin6_scope_id = ifr.ifr_ifindex; - if (ioctl(sock, SIOCGIFHWADDR, &ifr) == -1) { perror("can't get MAC address"); goto err; @@ -246,12 +253,32 @@ static int netsock_open(struct interface *interface) memcpy(&interface->hwaddr, &ifr.ifr_hwaddr.sa_data, 6); mac_to_ipv6(&interface->hwaddr, &interface->address); + memset(&sin6, 0, sizeof(sin6)); + sin6.sin6_port = htons(ALFRED_PORT); + sin6.sin6_family = AF_INET6; + memcpy(&sin6.sin6_addr, &interface->address, sizeof(sin6.sin6_addr)); + sin6.sin6_scope_id = interface->scope_id; + + memset(&sin6_mc, 0, sizeof(sin6_mc)); + sin6_mc.sin6_port = htons(ALFRED_PORT); + sin6_mc.sin6_family = AF_INET6; + memcpy(&sin6_mc.sin6_addr, &in6addr_localmcast, + sizeof(sin6_mc.sin6_addr)); + sin6_mc.sin6_scope_id = interface->scope_id; + enable_raw_bind_capability(1); if (setsockopt(sock, SOL_SOCKET, SO_BINDTODEVICE, interface->interface, strlen(interface->interface) + 1)) { perror("can't bind to device"); goto err; } + + if (setsockopt(sock_mc, SOL_SOCKET, SO_BINDTODEVICE, + interface->interface, + strlen(interface->interface) + 1)) { + perror("can't bind to device"); + goto err; + } enable_raw_bind_capability(0); if (bind(sock, (struct sockaddr *)&sin6, sizeof(sin6)) < 0) { @@ -259,6 +286,21 @@ static int netsock_open(struct interface *interface) goto err; } + if (bind(sock_mc, (struct sockaddr *)&sin6_mc, sizeof(sin6_mc)) < 0) { + perror("can't bind"); + goto err; + } + + memcpy(&mreq.ipv6mr_multiaddr, &in6addr_localmcast, + sizeof(mreq.ipv6mr_multiaddr)); + mreq.ipv6mr_interface = interface->scope_id; + + if (setsockopt(sock_mc, IPPROTO_IPV6, IPV6_ADD_MEMBERSHIP, + &mreq, sizeof(mreq))) { + perror("can't add multicast membership"); + goto err; + } + ret = fcntl(sock, F_GETFL, 0); if (ret < 0) { perror("failed to get file status flags"); @@ -271,11 +313,25 @@ static int netsock_open(struct interface *interface) goto err; } + ret = fcntl(sock_mc, F_GETFL, 0); + if (ret < 0) { + perror("failed to get file status flags"); + goto err; + } + + ret = fcntl(sock_mc, F_SETFL, ret | O_NONBLOCK); + if (ret < 0) { + perror("failed to set file status flags"); + goto err; + } + interface->netsock = sock; + interface->netsock_mcast = sock_mc; return 0; err: close(sock); + close(sock_mc); return -1; } @@ -314,6 +370,12 @@ int netsock_prepare_select(struct globals *globals, fd_set *fds, int maxsock) if (maxsock < interface->netsock) maxsock = interface->netsock; } + + if (interface->netsock_mcast >= 0) { + FD_SET(interface->netsock_mcast, fds); + if (maxsock < interface->netsock_mcast) + maxsock = interface->netsock_mcast; + } } return maxsock; @@ -324,12 +386,22 @@ void netsock_check_error(struct globals *globals, fd_set *errfds) struct interface *interface; list_for_each_entry(interface, &globals->interfaces, list) { - if (interface->netsock >= 0 && - FD_ISSET(interface->netsock, errfds)) { - fprintf(stderr, "Error on netsock detected\n"); + if ((interface->netsock < 0 || + !FD_ISSET(interface->netsock, errfds)) && + (interface->netsock_mcast < 0 || + !FD_ISSET(interface->netsock_mcast, errfds))) + continue; + + fprintf(stderr, "Error on netsock detected\n"); + + if (interface->netsock >= 0) close(interface->netsock); - interface->netsock = -1; - } + + if (interface->netsock_mcast >= 0) + close(interface->netsock_mcast); + + interface->netsock = -1; + interface->netsock_mcast = -1; } } @@ -341,7 +413,15 @@ int netsock_receive_packet(struct globals *globals, fd_set *fds) list_for_each_entry(interface, &globals->interfaces, list) { if (interface->netsock >= 0 && FD_ISSET(interface->netsock, fds)) { - recv_alfred_packet(globals, interface); + recv_alfred_packet(globals, interface, + interface->netsock); + recvs++; + } + + if (interface->netsock_mcast >= 0 && + FD_ISSET(interface->netsock_mcast, fds)) { + recv_alfred_packet(globals, interface, + interface->netsock_mcast); recvs++; } } diff --git a/recv.c b/recv.c index 59c69b4..16242bc 100644 --- a/recv.c +++ b/recv.c @@ -379,7 +379,8 @@ static int process_alfred_status_txend(struct globals *globals, return 0; } -int recv_alfred_packet(struct globals *globals, struct interface *interface) +int recv_alfred_packet(struct globals *globals, struct interface *interface, + int recv_sock) { uint8_t buf[MAX_PAYLOAD]; ssize_t length; @@ -391,7 +392,7 @@ int recv_alfred_packet(struct globals *globals, struct interface *interface) return -1; sourcelen = sizeof(source); - length = recvfrom(interface->netsock, buf, sizeof(buf), 0, + length = recvfrom(recv_sock, buf, sizeof(buf), 0, (struct sockaddr *)&source, &sourcelen); if (length <= 0) { perror("read from network socket failed"); diff --git a/send.c b/send.c index cc55489..8853970 100644 --- a/send.c +++ b/send.c @@ -182,7 +182,9 @@ ssize_t send_alfred_packet(struct interface *interface, if (ret == -EPERM) { perror("Error during sent"); close(interface->netsock); + close(interface->netsock_mcast); interface->netsock = -1; + interface->netsock_mcast = -1; } return ret; diff --git a/server.c b/server.c index c9c77b9..ac253a9 100644 --- a/server.c +++ b/server.c @@ -262,7 +262,9 @@ static void check_if_socket(struct interface *interface) close: close(interface->netsock); + close(interface->netsock_mcast); interface->netsock = -1; + interface->netsock_mcast = -1; close(sock); } -- 2.1.4

2 1

Jump to page:

2024

2023

2022

2021

2020

2019

2018

2017

2016

2015

2014

2013

2012

2011

2010

2009

2008

2007

B.A.T.M.A.N