commits

commits@lists.open-mesh.org

1 participants
12232 discussions

r1440 - trunk/batman-adv-kernelland
by postmaster＠open-mesh.net 09 Oct '09

09 Oct '09

Author: marek Date: 2009-10-09 17:11:51 +0000 (Fri, 09 Oct 2009) New Revision: 1440 Modified: trunk/batman-adv-kernelland/soft-interface.c Log: [batman-adv] remove no checksum flag on bat0 to fix checksumming on certain architectures It is related to the TCP checksum offload function: I am using as embedded platform PCEngines Alix 3d2, that do not support this function (and also qemu virtual machines do not support it). This patch disable the tx checksum offload and everything now works fine for me! Signed-off-by: Andrea Ghittino <a.mailevent(a)gmail.com> Modified: trunk/batman-adv-kernelland/soft-interface.c =================================================================== --- trunk/batman-adv-kernelland/soft-interface.c 2009-09-27 02:21:54 UTC (rev 1439) +++ trunk/batman-adv-kernelland/soft-interface.c 2009-10-09 17:11:51 UTC (rev 1440) @@ -114,7 +114,6 @@ #endif /* LINUX_VERSION_CODE < KERNEL_VERSION(2, 6, 29) */ dev->destructor = free_netdev; - dev->features |= NETIF_F_NO_CSUM; dev->mtu = hardif_min_mtu(); dev->hard_header_len = BAT_HEADER_LEN; /* reserve more space in the * skbuff for our header */

1 0

r1439 - trunk/batman-adv-kernelland
by postmaster＠open-mesh.net 08 Oct '09

08 Oct '09

Author: marek Date: 2009-09-27 02:21:54 +0000 (Sun, 27 Sep 2009) New Revision: 1439 Modified: trunk/batman-adv-kernelland/hard-interface.c trunk/batman-adv-kernelland/routing.c Log: [batman-adv] fixing race condition while handling lots of data The callback batman_data_ready() set the variable data_ready_cond to true before waking up the packet handling thread. When the thread has completed its work it will sleep and sets the variable back to false. When the system has to handle large quantities of traffic the callback might be called several times while the packet handling thread runs. The status variable will be overwritten once the thread is done which might leave packets in the socket buffer. If the socket buffer is full no more packets are accepted and the callback won't be triggered. Therefore the packet receive thread has to reset the variable at the very beginning. Modified: trunk/batman-adv-kernelland/hard-interface.c =================================================================== --- trunk/batman-adv-kernelland/hard-interface.c 2009-09-15 19:24:14 UTC (rev 1438) +++ trunk/batman-adv-kernelland/hard-interface.c 2009-09-27 02:21:54 UTC (rev 1439) @@ -148,7 +148,7 @@ active_ifs--; debug_log(LOG_TYPE_NOTICE, "Interface deactivated: %s\n", - batman_if->dev); + batman_if->dev); } /* (re)activate given interface. */ @@ -209,6 +209,9 @@ if (batman_if->if_num == 0) set_main_if_addr(batman_if->net_dev->dev_addr); + debug_log(LOG_TYPE_NOTICE, "Interface activated: %s\n", + batman_if->dev); + return; bind_err: Modified: trunk/batman-adv-kernelland/routing.c =================================================================== --- trunk/batman-adv-kernelland/routing.c 2009-09-15 19:24:14 UTC (rev 1438) +++ trunk/batman-adv-kernelland/routing.c 2009-09-27 02:21:54 UTC (rev 1439) @@ -625,6 +625,8 @@ wait_event_interruptible(thread_wait, (atomic_read(&data_ready_cond) || atomic_read(&exit_cond))); + atomic_set(&data_ready_cond, 0); + if (kthread_should_stop() || atomic_read(&exit_cond)) break; @@ -950,8 +952,6 @@ } rcu_read_unlock(); - atomic_set(&data_ready_cond, 0); - } kfree(packet_buff);

1 0

r1438 - in trunk/batman-experimental: . lib/bmx_http_info
by postmaster＠open-mesh.net 08 Oct '09

08 Oct '09

Author: axel Date: 2009-09-15 19:24:14 +0000 (Tue, 15 Sep 2009) New Revision: 1438 Modified: trunk/batman-experimental/batman.h trunk/batman-experimental/lib/bmx_http_info/http_info.c trunk/batman-experimental/schedule.c Log: let http_info plugin respond with simple help page to invalid http requests, some other minor changes Modified: trunk/batman-experimental/batman.h =================================================================== --- trunk/batman-experimental/batman.h 2009-09-15 19:22:26 UTC (rev 1437) +++ trunk/batman-experimental/batman.h 2009-09-15 19:24:14 UTC (rev 1438) @@ -230,8 +230,6 @@ #define ARG_NETB "netb" - - #define ARG_THROW "throw" @@ -315,8 +313,6 @@ } __attribute__((packed)); -#define OGM_OGI_MAX 255 - struct bat_packet_ogm { #if __BYTE_ORDER == __LITTLE_ENDIAN @@ -381,8 +377,8 @@ #define EXT_FIELD_TYPE ext_type #define EXT_FIELD_MSG ext_msg -#define EXT_FIELD_LEN def8 /* the size of this TLV extension header in 4 oktet blocks (including this first 4 oktets) */ -#define MIN_TLV_LEN_BYTES 4 /* the minimum size in bytes (1 octet) of a TLV extension message */ +#define EXT_FIELD_LEN_4B def8 /* the size of this TLV extension header in 4 oktet blocks (including this first 4 oktets) */ +#define MIN_TLV_LEN_1B 4 /* the minimum size in bytes (1 octet) of a TLV extension message */ // field accessor for primary interface announcement extension packets #define EXT_PIP_FIELD_RES1 def8 Modified: trunk/batman-experimental/lib/bmx_http_info/http_info.c =================================================================== --- trunk/batman-experimental/lib/bmx_http_info/http_info.c 2009-09-15 19:22:26 UTC (rev 1437) +++ trunk/batman-experimental/lib/bmx_http_info/http_info.c 2009-09-15 19:24:14 UTC (rev 1438) @@ -86,9 +86,32 @@ } else { - dbgf( DBGL_CHANGES, DBGT_INFO, "rcvd illegal %d bytes long HTTP request via fd %d", + dbg( DBGL_CHANGES, DBGT_INFO, "rcvd illegal %d bytes long HTTP request via fd %d", tcp_req_len, cn->fd); + /* + dbg_cn( cn, DBGL_ALL, DBGT_INFO, "rcvd illegal %d bytes long HTTP request via fd %d:\n%s\n", + tcp_req_len, cn->fd, tcp_req_data); + */ + check_apply_parent_option( ADD, OPT_APPLY, 0, get_option( 0,0, ARG_STATUS), 0, cn ); + dbg_printf( cn, "\nillegal HTTP request! Valid requests are:\n\n" ); + + struct list_head *list_pos; + + list_for_each( list_pos, &opt_list ) { + + struct opt_type *opt = (struct opt_type *)list_entry( list_pos, struct opt_data, list ); + + if ( opt->auth_t == A_USR && + opt->opt_t == A_PS0 && + opt->dyn_t != A_INI && + opt->cfg_t == A_ARG ) + { + dbg_printf( cn, "/%s\n\n", opt->long_name ); + + } + } + } } else { Modified: trunk/batman-experimental/schedule.c =================================================================== --- trunk/batman-experimental/schedule.c 2009-09-15 19:22:26 UTC (rev 1437) +++ trunk/batman-experimental/schedule.c 2009-09-15 19:24:14 UTC (rev 1438) @@ -845,7 +845,7 @@ if ( ((struct ext_packet*)(ext_a+ext_p))->EXT_FIELD_TYPE == ext_type ) { if ( ext_attribute[ext_type] & EXT_ATTR_TLV ) - ext_p+= 4*((struct ext_packet*)(ext_a+ext_p))->EXT_FIELD_LEN; + ext_p+= 4*((struct ext_packet*)(ext_a+ext_p))->EXT_FIELD_LEN_4B; else ext_p+= sizeof(struct ext_packet);

1 0

r1437 - trunk/batman-experimental/doc
by postmaster＠open-mesh.net 08 Oct '09

08 Oct '09

Author: axel Date: 2009-09-15 19:22:26 +0000 (Tue, 15 Sep 2009) New Revision: 1437 Added: trunk/batman-experimental/doc/Makefile Modified: trunk/batman-experimental/doc/bmx.tex Log: add completely outdated warning do doc Added: trunk/batman-experimental/doc/Makefile =================================================================== --- trunk/batman-experimental/doc/Makefile (rev 0) +++ trunk/batman-experimental/doc/Makefile 2009-09-15 19:22:26 UTC (rev 1437) @@ -0,0 +1,14 @@ +all: doc + + +doc: Makefile +# latex2html -split 0 -noinfo bmx.tex +# latex2html -split 0 -noinfo bmx.tex + pdflatex bmx.tex + pdflatex bmx.tex + rm -rv *.toc *.log *.aux + + +clean: + rm -rfv *.toc *.pdf *.log *.aux +#$$(find -type d | grep "./") \ No newline at end of file Modified: trunk/batman-experimental/doc/bmx.tex =================================================================== --- trunk/batman-experimental/doc/bmx.tex 2009-09-12 17:28:23 UTC (rev 1436) +++ trunk/batman-experimental/doc/bmx.tex 2009-09-15 19:22:26 UTC (rev 1437) @@ -24,10 +24,10 @@ \input epsf \pagestyle{fancyplain} -\setlength{\headrulewidth}{0.4pt} -\setlength{\footrulewidth}{0.4pt} -\setlength{\plainheadrulewidth}{0.4pt} -\setlength{\plainfootrulewidth}{0.4pt} +%\setlength{\headrulewidth}{0.4pt} +%\setlength{\footrulewidth}{0.4pt} +%\setlength{\plainheadrulewidth}{0.4pt} +%\setlength{\plainfootrulewidth}{0.4pt} \chead{ %{\Small @@ -84,11 +84,11 @@ \noindent -Reference -Manual -of +B.a.t.M.a.n-eXperimental Reference Manual -B.A.T.M.A.N Experimental +\large +( warning: this document is outdated! For current syntax check: \emph{bmxd ---help} ) + %(BMX) %\vspace{1.5cm} @@ -210,9 +210,9 @@ %Nodes with two interfaces propagate two different sequences of OGMs, one for each interface IP. % -Every node selects one of its neighbors as the best next hop towards a specific other node. +Every node selects one of its neighbors as the best neighbor and gateway towards a specific other node. % -This process of identifying the best next hop towards a distant node is called neighbor ranking. +This process of identifying the best neighbor towards a distant node is called neighbor ranking. % The selected neighbor is referred as the best-ranking neighbor. % @@ -1040,22 +1040,24 @@ \end{verbatim} \end{small} -\subsection{Hiding redundant topology information beyond the local neighborhood} +\subsection{Hiding local topology information beyond the neighborhood} \label{sec:hiding-topology} +Nodes may alter (i.e. reduce) the default TTL of their own OGMs to limit the number of hops that these OGMs are propagated through the mesh. +This can be done for all OGMS or just for OGMs propagating the existence of particular interfaces. +This does not affect the routing between other nodes in the mesh, but may be used to limit the range of presence (existence) of individual nodes. +For example, a node with three interfaces may be configured to send OGMs with a high TTL only for the first interface and a TTL of one for OGMs representing the second and third interface. +% +This way, the node is still reachable via the IP of it's first interface. +But it does not burden the nodes beyond its neighbor horizon with the efforts of maintaining and re-broadcasting OGMs from it's second and third interface. -Nodes may reduce the default TTL of their own OGMs to limit the number of hops that these OGMs are propagated through the mesh. This can be done for all OGMs or just for OGMs propagating the existence of particular interfaces. This does not affect the routing between other nodes in the mesh, but may be used to limit the range of presence (existence) of individual interfaces. For example, a node with three interfaces may be configured to send OGMs with a -high TTL only for the first (primary) interface and a small TTL for OGMs representing the second and third (all non-primary) interfaces. This way, the node is always reachable via the IP of it's first interface. But it does not burden the nodes beyond its neighbor horizon with the efforts of processing, maintaining, and re-broadcasting OGMs from it's second and third interface. +One side effect of the one-TTL OGMs is that packets generated on these nodes may leave the node with a source address of such a secondary interface. +%which is not known beyond the neighbor horizon of the node. +% +This has the consequence that non-neighboring nodes could not reply to this source address, simply because the OGMs for this source address have never been propagated that fare. +A Solution to counter the above problem is to automatically make an HNA announcement for all "hidden interfaces". +%Because of the policy-routing (enabled since 0.3) this approach could easily be integration in the current code. +This behavior is can be enabled using the interface-specific \emph{/a} switch. -One side effect of the one-TTL OGMs is that data-packets generated on these nodes may leave the node with a source address of such a secondary interface. This has the consequence that non-neighboring nodes could not reply to this source address, simply because the OGMs for this source address have never been propagated that fare. - -The Solution to counter the above problem is that each multi-homed node automatically makes an HNA announcement for all non-primary (hidden) interfaces and propagates this HNAs with the OGM representing its primary interface. This way the IP addresses of the non-primary interfaces are also reachable beyond the local neighbor horizon. -All HNA announcements end up in routing table 65. It tells the network stack to route packets with a destination address of a non-primary interface towards the IP address of the corresponding primary interface. - -This is now the default behavior in 0.3 - - -This behavior is can be controlled using the interface-specific \emph{/a} and \emph{/A}switch. - The following parameters can be used to reduce the TTL of OGMs representing specific interfaces. \begin{small} \begin{verbatim} @@ -1072,9 +1074,6 @@ to add the IP address of this interface to the HNA list. Also reduces the TTL for OGMs representing this interface to 1 and broadcasts the OGMs representing this interface only via this interface - - /A : attached after an interface name - to remove the IP address of this interface from the HNA list. \end{verbatim} \end{small} A small TTL for all self-initiated OGMs has some additional advantages for back-bone nodes. Nodes, which are not supposed to generate pay-load traffic itself and which were only installed to improve the connectivity and coverage of the mesh by relaying other nodes traffic. @@ -1905,117 +1904,9 @@ %Running the old and the new protocol implementations in parallel has also an advantage during the migration phase. This allows extensive testing of the new version before any functinality of the old version must be disabled. -\begin{small} \begin{verbatim} -root@ng1e:~# killall udhcpc udhcpd batmand bmxd -root@ng1e:~# udhcpc -i eth0.0 -root@ng1e:~# iptables -F -root@ng1e:~# iptables -t mangle -F -root@ng1e:~# iptables -t nat -F -root@ng1e:~# iptables -t nat -A POSTROUTING -o eth0.0 -j MASQUERADE # eth0.0 is default -root@ng1e:~# ifconfig ath0 104.130.30.201 netmask 255.0.0.0 -root@ng1e:~# ifconfig ath0:bat 105.130.30.201 netmask 255.0.0.0 -root@ng1e:~# ifconfig ath0:bmx 103.130.30.201 netmask 255.0.0.0 -root@ng1e:~# iwconfig ath0 channel 10 essid olsr.freifunk.net ap 02:CA:FF:EE:BA:BE -root@ng1e:~# batmand -g 5000 ath0:bat -root@ng2e:~# killall udhcpc udhcpd batmand bmxd -root@ng2e:~# iptables -F -root@ng2e:~# iptables -t mangle -F -root@ng2e:~# iptables -t nat -F -root@ng2e:~# ifconfig eth0.0 192.168.123.1 netmask 255.255.255.0 -root@ng2e:~# vi /etc/udhcpd.conf # control interface, router, and ip range -root@ng2e:~# udhcpd /etc/udhcpd.conf -root@ng2e:~# ifconfig ath0 104.130.30.202 netmask 255.0.0.0 -root@ng2e:~# ifconfig ath0:bat 105.130.30.202 netmask 255.0.0.0 -root@ng2e:~# ifconfig ath0:bmx 103.130.30.202 netmask 255.0.0.0 -root@ng2e:~# ifconfig eth0.1 104.130.30.212 netmask 255.0.0.0 -root@ng2e:~# ifconfig eth0.1:bat 105.130.30.212 netmask 255.0.0.0 -root@ng2e:~# ifconfig eth0.1:bmx 103.130.30.212 netmask 255.0.0.0 -root@ng2e:~# iwconfig ath0 channel 10 essid olsr.freifunk.net ap 02:CA:FF:EE:BA:BE -#make shure all interfaces are configured before batmand starts (even eth0.0) -root@ng2e:~# batmand -r 3 ath0:bat eth0.1:bat -root@ng2e:~# iptables -t nat -A POSTROUTING -o bat0 -j MASQUERADE -root@ng2e:~# iptables -t nat -A POSTROUTING -o ath0 -s 192.168.123.0/24 -d 103.0.0.0/8 -j SNAT --to 103.130.30.202 -root@ng2e:~# iptables -t nat -A POSTROUTING -o ath0 -s 192.168.123.0/24 -d 104.0.0.0/8 -j SNAT --to 104.130.30.202 -root@ng2e:~# iptables -t nat -A POSTROUTING -o ath0 -s 192.168.123.0/24 -d 105.0.0.0/8 -j SNAT --to 105.130.30.202 -root@ng2e:~# iptables -t nat -A POSTROUTING -o eth0.1 -s 192.168.123.0/24 -d 103.0.0.0/8 -j SNAT --to 103.130.30.212 -root@ng2e:~# iptables -t nat -A POSTROUTING -o eth0.1 -s 192.168.123.0/24 -d 104.0.0.0/8 -j SNAT --to 104.130.30.212 -root@ng2e:~# iptables -t nat -A POSTROUTING -o eth0.1 -s 192.168.123.0/24 -d 105.0.0.0/8 -j SNAT --to 105.130.30.212 - -# now you may connect a dhcp-client to eth0.0 of node B and surfe the web as well as access all 103/104 mesh nodes. -# do the same for node C -\end{verbatim} \end{small} - - -\begin{small} \begin{verbatim} - -root@ng1e:~# bmxd --base-port 14305 --rt-table-offset 165 --prio-rules-offset 16600 --gw-tunnel-network 10.111.0.0/16 -g 5000 --window-size 10 ath0:bmx - -root@ng2e:~# bmxd --base-port 14305 --rt-table-offset 165 --prio-rules-offset 16600 --window-size 10 ath0:bmx eth0.1:bmx - -root@ng3e:~# bmxd --base-port 14305 --rt-table-offset 165 --prio-rules-offset 16600 --window-size 10 ath0:bmx eth0.1:bmx - -root@ng3e:~# bmxd -c --base-port 14305 -d 1 -b -WARNING: You are using the experimental batman branch! ---base-port 14305 \ -BatMan-eXperimental 0.3-alpha rv751, IF: ath0:bmx 103.130.30.203, WindSize: 10, BLT: 20, OGI: 1500, UT: 0d 0h 0m -Neighbor outgoingIF bestLink (brc rcvd lseq lvld) [ viaIF RTQ RQ TQ].. -103.130.30.201 ath0:bmx 103.130.30.201 ( 10 10 15 0) [ ath0:bmx 10 10 10] -103.130.30.212 eth0.1:bmx 103.130.30.212 ( 10 10 18 1) [eth0.1:bmx 10 10 10] -103.130.30.202 ath0:bmx 103.130.30.202 ( 10 10 18 1) [ ath0:bmx 10 10 10] - -Originator outgoingIF bestNextHop (brc rcvd lseq lvld) alternative next hops... -103.130.30.201 ath0:bmx 103.130.30.201 ( 10 10 15 0) 103.130.30.202 ( 9) 103.130.30.212 ( 9) -103.130.30.212 eth0.1:bmx 103.130.30.212 ( 10 10 18 1) -103.130.30.202 ath0:bmx 103.130.30.202 ( 10 10 18 1) 103.130.30.212 ( 10) 103.130.30.201 ( 0) - -Originator HNAs... -103.130.30.202 103.130.30.212/32 -root@ng3e:~# -root@ng3e:~# bmxd -c -d 1 -b -WARNING: You are using the experimental batman branch! -BatMan-eXperimental 0.3-alpha rv751, IF: ath0:bat 105.130.30.203, WindSize: 100, BLT: 20, OGI: 1500, UT: 0d 0h14m -Neighbor outgoingIF bestLink (brc rcvd lseq lvld) [ viaIF RTQ RQ TQ].. -105.130.30.212 eth0.1:bat 105.130.30.212 (100 100 1045 1) [eth0.1:bat 100 100 100] -105.130.30.202 eth0.1:bat 105.130.30.212 (100 100 1045 0) [ ath0:bat 97 99 97] -105.130.30.201 ath0:bat 105.130.30.201 ( 99 100 2816 0) [ ath0:bat 99 98 100] - -Originator outgoingIF bestNextHop (brc rcvd lseq lvld) alternative next hops... -105.130.30.212 eth0.1:bat 105.130.30.212 (100 100 1045 1) -105.130.30.202 eth0.1:bat 105.130.30.212 (100 100 1045 0) 105.130.30.202 ( 98) 105.130.30.201 ( 99) -105.130.30.201 ath0:bat 105.130.30.201 ( 99 100 2816 0) 105.130.30.212 ( 97) 105.130.30.202 ( 95) - -Originator HNAs... -105.130.30.202 105.130.30.212/32 -root@ng3e:~# - -\end{verbatim} \end{small} - -Do some extensive testing now. All nodes can be contacted using the 103/105 netmask. - -\begin{small} \begin{verbatim} -root@ng5e:~# real-ping -f -c 100 -s 1300 105.130.30.201 -PING 105.130.30.201 (105.130.30.201) 1300(1328) bytes of data. ---- 105.130.30.201 ping statistics --- -100 packets transmitted, 100 received, 0% packet loss, time 765ms -rtt min/avg/max/mdev = 5.494/6.703/16.865/2.146 ms, pipe 2, ipg/ewma 7.729/6.875 ms - -root@ng5e:~# real-ping -f -c 100 -s 1300 103.130.30.201 -PING 103.130.30.201 (103.130.30.201) 1300(1328) bytes of data. ---- 103.130.30.201 ping statistics --- -100 packets transmitted, 100 received, 0% packet loss, time 913ms -rtt min/avg/max/mdev = 6.475/8.378/20.180/1.881 ms, pipe 2, ipg/ewma 9.228/8.751 ms -root@ng5e:~# -\end{verbatim} \end{small} - - -\begin{small} \begin{verbatim} - -\end{verbatim} \end{small} - - %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%

1 0

r1436 - trunk/batman-adv-kernelland
by postmaster＠open-mesh.net 08 Oct '09

08 Oct '09

Author: marek Date: 2009-09-12 17:28:23 +0000 (Sat, 12 Sep 2009) New Revision: 1436 Modified: trunk/batman-adv-kernelland/device.c trunk/batman-adv-kernelland/main.c Log: [batman-adv] Remove compatibility code for pre 2.6 kernel versions The earliest version of linux which is compatible to batman-adv-kernelland is 2.6.20. Their are patches available in debian for batman-adv-kernelland 0.1 to make it build against linux 2.6.15, but they cannot be ported to current trunk due to changes to the working queues. To have "support" for much older versions isn't needed and can be removed without loosing any functionality. As result the only remaining LINUX_VERSION_CODE dependend precomiler code beside compat.h is for netdevice_ops support in soft-interface.c and result code checking of unregister_chrdev in device.c. For a possible merge into linux compat.h and all appearances of LINUX_VERSION_CODE should be removed according to checkpatch.pl. Signed-off-by: Sven Eckelmann <sven.eckelmann(a)gmx.de> Modified: trunk/batman-adv-kernelland/device.c =================================================================== --- trunk/batman-adv-kernelland/device.c 2009-09-11 17:56:56 UTC (rev 1435) +++ trunk/batman-adv-kernelland/device.c 2009-09-12 17:28:23 UTC (rev 1436) @@ -26,9 +26,7 @@ #include "compat.h" -#if LINUX_VERSION_CODE >= KERNEL_VERSION(2, 6, 0) static struct class *batman_class; -#endif static int Major; /* Major number assigned to our device driver */ @@ -65,13 +63,6 @@ return 0; } -#if LINUX_VERSION_CODE < KERNEL_VERSION(2, 6, 0) - if (devfs_mk_cdev(MKDEV(tmp_major, 0), S_IFCHR | S_IRUGO | S_IWUGO, - "batman-adv", 0)) { - debug_log(LOG_TYPE_WARN, "Could not create /dev/batman-adv\n"); - return 0; - } -#else batman_class = class_create(THIS_MODULE, "batman-adv"); if (IS_ERR(batman_class)) { @@ -81,7 +72,6 @@ device_create(batman_class, NULL, MKDEV(tmp_major, 0), NULL, "batman-adv"); -#endif Major = tmp_major; return 1; @@ -94,12 +84,8 @@ if (!Major) return; -#if LINUX_VERSION_CODE < KERNEL_VERSION(2, 6, 0) - devfs_remove("batman-adv", 0); -#else device_destroy(batman_class, MKDEV(Major, 0)); class_destroy(batman_class); -#endif /* Unregister the device */ #if LINUX_VERSION_CODE < KERNEL_VERSION(2, 6, 23) Modified: trunk/batman-adv-kernelland/main.c =================================================================== --- trunk/batman-adv-kernelland/main.c 2009-09-11 17:56:56 UTC (rev 1435) +++ trunk/batman-adv-kernelland/main.c 2009-09-12 17:28:23 UTC (rev 1436) @@ -203,20 +203,12 @@ void inc_module_count(void) { -#if LINUX_VERSION_CODE < KERNEL_VERSION(2, 6, 0) - MOD_INC_USE_COUNT; -#else try_module_get(THIS_MODULE); -#endif } void dec_module_count(void) { -#if LINUX_VERSION_CODE < KERNEL_VERSION(2, 6, 0) - MOD_DEC_USE_COUNT; -#else module_put(THIS_MODULE); -#endif } int addr_to_string(char *buff, uint8_t *addr)

1 0

r1435 - trunk/batman/linux
by postmaster＠open-mesh.net 08 Oct '09

08 Oct '09

Author: marek Date: 2009-09-11 17:56:56 +0000 (Fri, 11 Sep 2009) New Revision: 1435 Modified: trunk/batman/linux/tun.c Log: [batman] add iptables MSS CLAMP rule in mangle table as suggested by Michael Rack Modified: trunk/batman/linux/tun.c =================================================================== --- trunk/batman/linux/tun.c 2009-09-11 17:56:53 UTC (rev 1434) +++ trunk/batman/linux/tun.c 2009-09-11 17:56:56 UTC (rev 1435) @@ -40,6 +40,9 @@ #define IPTABLES_ADD_MASQ "iptables -t nat -A POSTROUTING -o %s -j MASQUERADE" #define IPTABLES_DEL_MASQ "iptables -t nat -D POSTROUTING -o %s -j MASQUERADE" +#define IPTABLES_ADD_MSS "iptables -t mangle -I POSTROUTING -p tcp --tcp-flags SYN,RST SYN -o %s -j TCPMSS --clamp-mss-to-pmtu" +#define IPTABLES_DEL_MSS "iptables -t mangle -D POSTROUTING -p tcp --tcp-flags SYN,RST SYN -o %s -j TCPMSS --clamp-mss-to-pmtu" + #define IPTABLES_ADD_ACC "iptables -t nat -I POSTROUTING -s %s/%i -j ACCEPT" #define IPTABLES_DEL_ACC "iptables -t nat -D POSTROUTING -s %s/%i -j ACCEPT" @@ -101,17 +104,23 @@ } void add_nat_rule(char *dev) { - char cmd[100]; + char cmd[150]; sprintf(cmd, IPTABLES_ADD_MASQ, dev); exec_iptables_rule(cmd, ROUTE_ADD); + + sprintf(cmd, IPTABLES_ADD_MSS, dev); + exec_iptables_rule(cmd, ROUTE_ADD); } void del_nat_rule(char *dev) { - char cmd[100]; + char cmd[150]; sprintf(cmd, IPTABLES_DEL_MASQ, dev); exec_iptables_rule(cmd, ROUTE_DEL); + + sprintf(cmd, IPTABLES_ADD_MSS, dev); + exec_iptables_rule(cmd, ROUTE_DEL); } void hna_local_update_nat(uint32_t hna_ip, uint8_t netmask, int8_t route_action) {

1 0

r1434 - trunk/batman
by postmaster＠open-mesh.net 08 Oct '09

08 Oct '09

Author: marek Date: 2009-09-11 17:56:53 +0000 (Fri, 11 Sep 2009) New Revision: 1434 Modified: trunk/batman/schedule.c Log: [batman] check whether we have a valid "previous ttl" to avoid a TTL of 255 Modified: trunk/batman/schedule.c =================================================================== --- trunk/batman/schedule.c 2009-09-11 17:56:49 UTC (rev 1433) +++ trunk/batman/schedule.c 2009-09-11 17:56:53 UTC (rev 1434) @@ -230,10 +230,10 @@ /* rebroadcast ogm of best ranking neighbor as is */ if (orig_node->router->addr != neigh) { - bat_packet->tq = orig_node->router->tq_avg; - bat_packet->ttl = orig_node->router->last_ttl - 1; + if (orig_node->router->last_ttl) + bat_packet->ttl = orig_node->router->last_ttl - 1; } tq_avg = orig_node->router->tq_avg;

1 0

r1433 - trunk/batman-adv-kernelland
by postmaster＠open-mesh.net 08 Oct '09

08 Oct '09

Author: marek Date: 2009-09-11 17:56:49 +0000 (Fri, 11 Sep 2009) New Revision: 1433 Modified: trunk/batman-adv-kernelland/send.c Log: [batman-adv] check whether we have a valid "previous ttl" to avoid a TTL of 255 Modified: trunk/batman-adv-kernelland/send.c =================================================================== --- trunk/batman-adv-kernelland/send.c 2009-09-10 16:31:08 UTC (rev 1432) +++ trunk/batman-adv-kernelland/send.c 2009-09-11 17:56:49 UTC (rev 1433) @@ -301,7 +301,9 @@ /* rebroadcast ogm of best ranking neighbor as is */ if (!compare_orig(orig_node->router->addr, ethhdr->h_source)) { batman_packet->tq = orig_node->router->tq_avg; - batman_packet->ttl = orig_node->router->last_ttl - 1; + + if (orig_node->router->last_ttl) + batman_packet->ttl = orig_node->router->last_ttl - 1; } tq_avg = orig_node->router->tq_avg;

1 0

r143 - trunk/batman
by postmaster＠open-mesh.net 08 Oct '09

08 Oct '09

Author: marek Date: 2006-12-06 02:25:43 +0000 (Wed, 06 Dec 2006) New Revision: 143 Modified: trunk/batman/batman.c Log: - fix crashes in schedule_forward_packet && send_outstanding_packets Modified: trunk/batman/batman.c =================================================================== --- trunk/batman/batman.c 2006-12-06 01:49:21 UTC (rev 142) +++ trunk/batman/batman.c 2006-12-06 02:25:43 UTC (rev 143) @@ -790,42 +790,42 @@ output( "ttl exceeded \n" ); } else { - forw_node_new = alloc_memory(sizeof (struct forw_node)); - INIT_LIST_HEAD(&forw_node_new->list); + /* list_for_each(forw_pos, &forw_list) { + forw_node = list_entry(forw_pos, struct forw_node, list); + if ((int)(forw_node->when - forw_node_new->when) > 0) + break; + } */ - memcpy(&forw_node_new->pack, in, sizeof (struct packet)); + if ( unidirectional ) { - forw_node_new->pack.ttl--; - forw_node_new->when = get_time(); + forw_node_new = alloc_memory(sizeof (struct forw_node)); + INIT_LIST_HEAD(&forw_node_new->list); - if ( hna_buff_len > 0 ) { + memcpy(&forw_node_new->pack, in, sizeof (struct packet)); - forw_node_new->hna_buff = alloc_memory( hna_buff_len ); - forw_node_new->hna_buff_len = hna_buff_len; + forw_node_new->pack.ttl--; + forw_node_new->when = get_time(); - memcpy( forw_node_new->hna_buff, hna_recv_buff, hna_buff_len ); + if ( hna_buff_len > 0 ) { - } else { + forw_node_new->hna_buff = alloc_memory( hna_buff_len ); + forw_node_new->hna_buff_len = hna_buff_len; - forw_node_new->hna_buff = NULL; - forw_node_new->hna_buff_len = 0; + memcpy( forw_node_new->hna_buff, hna_recv_buff, hna_buff_len ); - } + } else { - list_for_each(forw_pos, &forw_list) { - forw_node = list_entry(forw_pos, struct forw_node, list); - if ((int)(forw_node->when - forw_node_new->when) > 0) - break; - } + forw_node_new->hna_buff = NULL; + forw_node_new->hna_buff_len = 0; - if ( unidirectional ) { + } forw_node_new->pack.flags = ( UNIDIRECTIONAL | DIRECTLINK ); if ( if_outgoing != NULL ) { forw_node_new->if_outgoing = if_outgoing; - list_add( &forw_node_new->list, (forw_node == NULL ? &forw_list : forw_pos) ); + list_add( &forw_node_new->list, /* ( forw_node == NULL ? &forw_list : forw_pos )*/ &forw_list ); } else { @@ -845,13 +845,35 @@ batman_if = list_entry(if_pos, struct batman_if, list); + forw_node_new = alloc_memory(sizeof (struct forw_node)); + INIT_LIST_HEAD(&forw_node_new->list); + + memcpy(&forw_node_new->pack, in, sizeof (struct packet)); + + forw_node_new->pack.ttl--; + forw_node_new->when = get_time(); + + if ( hna_buff_len > 0 ) { + + forw_node_new->hna_buff = alloc_memory( hna_buff_len ); + forw_node_new->hna_buff_len = hna_buff_len; + + memcpy( forw_node_new->hna_buff, hna_recv_buff, hna_buff_len ); + + } else { + + forw_node_new->hna_buff = NULL; + forw_node_new->hna_buff_len = 0; + + } + if ( ( direct_link ) && ( if_outgoing == batman_if ) ) forw_node_new->pack.flags = DIRECTLINK; else forw_node_new->pack.flags = 0x00; forw_node_new->if_outgoing = batman_if; - list_add( &forw_node_new->list, (forw_node == NULL ? &forw_list : forw_pos) ); + list_add( &forw_node_new->list, &forw_list ); } @@ -892,8 +914,8 @@ } if ( debug_level == 4 ) { - addr_to_string(forw_node->pack.orig, orig_str, ADDR_STR_LEN); - output("Forwarding packet (originator %s, seqno %d, TTL %d)\n", orig_str, forw_node->pack.seqno, forw_node->pack.ttl); + addr_to_string( forw_node->pack.orig, orig_str, ADDR_STR_LEN ); + output( "Forwarding packet (originator %s, seqno %d, TTL %d) on interface %s\n", orig_str, forw_node->pack.seqno, forw_node->pack.ttl, forw_node->if_outgoing->dev ); } if ( send_packet( send_buff, sizeof (struct packet) + forw_node->hna_buff_len, &forw_node->if_outgoing->broad, forw_node->if_outgoing->udp_send_sock ) < 0 ) {

1 0

r1432 - trunk/batman-adv-kernelland
by postmaster＠open-mesh.net 08 Oct '09

08 Oct '09

Author: marek Date: 2009-09-10 16:31:08 +0000 (Thu, 10 Sep 2009) New Revision: 1432 Modified: trunk/batman-adv-kernelland/vis.c Log: [batman-adv] can't sleep while holding a spinlock Modified: trunk/batman-adv-kernelland/vis.c =================================================================== --- trunk/batman-adv-kernelland/vis.c 2009-09-09 15:12:08 UTC (rev 1431) +++ trunk/batman-adv-kernelland/vis.c 2009-09-10 16:31:08 UTC (rev 1432) @@ -497,7 +497,7 @@ goto err; } - my_vis_info = kmalloc(1000, GFP_KERNEL); + my_vis_info = kmalloc(1000, GFP_ATOMIC); if (!my_vis_info) { debug_log(LOG_TYPE_CRIT, "Can't initialize vis packet\n"); goto err;

1 0

Jump to page:

2024

2023

2022

2021

2020

2019

2018

2017

2016

2015

2014

2013

2012

2011

2010

2009

commits