B.A.T.M.A.N

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

15 participants
4308 discussions

[B.A.T.M.A.N.] [PATCH 01/14] batmand: Remove optimisation and debugging options
by Sven Eckelmann 21 May '11

21 May '11

Distributions like Gentoo or Debian have to strip the options regarding the optimisation levels and debugging information to fulfil their policies. There is currently no valid reason why the Makefile should override it. Signed-off-by: Sven Eckelmann <sven(a)narfation.org> --- This patch depends on other patches submitted earlier: - batmand: Fix FTBFS on second build attempt - batmand: Remove obsolete creation of source packages - batmand: Remove subversion/svk revision information - batmand: Use nproc to get number of available processors Makefile | 2 +- 1 files changed, 1 insertions(+), 1 deletions(-) diff --git a/Makefile b/Makefile index 4df2872..49342ac 100644 --- a/Makefile +++ b/Makefile @@ -29,7 +29,7 @@ endif #NO_POLICY_ROUTING = -DNO_POLICY_ROUTING CC = gcc -CFLAGS += -pedantic -Wall -W -Os -g3 -std=gnu99 +CFLAGS += -pedantic -Wall -W -std=gnu99 EXTRA_CFLAGS = -DDEBUG_MALLOC -DMEMORY_USAGE -DPROFILE_DATA $(NO_POLICY_ROUTING) -DREVISION_VERSION=$(REVISION_VERSION) LDFLAGS += -lpthread -- 1.7.5.1

1 13

Re: [B.A.T.M.A.N.] batman-adv bug 149: odd broadcast behavior -- generated CRC errors (Sven Eckelmann)
by David Beberman 20 May '11

20 May '11

First, my apologies on not monitoring the irc channel. A bit consumed at work. (if anybody happens to know how to get the broadcom bcm4329 driver to change channels in ad-hoc mode, please feel free to share. iwconfig ... channel XX command fails, but driver code looks correct. Target is HTC Desire). Regarding fragmentation -- this is strictly coming from the UPD/IPv6. It it not fragmenting in the wlan driver. I have fragmentation threshold set off. I'm assuming it is responding to the configuration parameters correctly. I'm using Omnipeek on a PC to sniff the packets. The antennas are about 2 feet apart. The network traffic is relatively low, in the 1 - 10% on average, with virtually no indication of corruption from hidden nodes or other sources under regular operation. That is, I only see the CRC error flags set with the broadcast packets at 802.11g data rates. This is the 802.11 packet CRC, nothing internal to it. The Omnipeek does not know about the batman-adv protocol. Also, I have other nodes setup with batman-adv which are repeating the broadcasts. I can see that only the packets without the CRC error flags are being rebroadcast. This indicates to me that the packets flagged as CRC errors by Omnipeek are also being dropped as error'ed packets by the other nodes. I do not see this problem with packets sent directly to the bcm4329 device. However, I will continue to try to see if I can get that to happen, would then mean it is bcm4329 driver and hardware problem. At the moment, using UDP broadcast with IPv4 directly onto the wlan driver without batman-adv (bat0), I do not see the behavior. I've been using my own code for this. Noticed netcat and nc last night. Will try piping with those tools today in both modes, with and without batman-adv to reverify what I'm seeing. I do not know how to do kernel debugging with kdbg/gdb across a USB port on the target side. I do not have JTAG or serial access. Google'd for how to setup the kernel to use USB port for kdbg, but did not find anything. Just lots of info on how to debug the USB driver itself. I'm using the cyanogenmod version of Android with the HTC Desire version of the Linux kernel, and the insertion of modules like batman-adv. I did verify that the modules are being built against the correct Linux kernel sources, just in case there was something silly going on. I'm building with NETCONSOLE enabled today and hope to be able to see driver messages sent over the WLAN using netcat on the host side. That should allow me to see both batman-adv and the bcm4329 trace messages. I should also be able to send the console to ttyUSB0, assuming I can get the HTC Desire USB to behave properly. I'll change the line to skb_copy as was suggested by <T_x>? and Sven. Will post my results. Thanks for the follow up. David Beberman

1 0

[B.A.T.M.A.N.] kfree_rcu patches
by Sven Eckelmann 20 May '11

20 May '11

Hi, Paul E. McKenney <paulmck(a)linux.vnet.ibm.com> proposed some changes which will reduce the amount of "kfree only" rcu-callback functions. The batman-adv compatibility layer has to add a similar functionality to support older kernels. I wanted to propose a solution which should be discussed before the kfree_rcu functionality enters Linus' kernel and our master branch. The amount of kfree_rcu_* function was reduced to support only a maximum offset of 256 to any 'struct rcu_head' inside a struct. This should be enough for the current version and supported architectures. Offsets of 4096 are supported by Paul E. McKenney's implementation, but our workaround would increase the size of the final module too much (personal opinion). ~~~ These patches should *NOT* enter the batman-adv repo right now ~~~~~ Kind regards, Sven

2 16

[B.A.T.M.A.N.] batman-adv bug 149: odd broadcast behavior -- generated CRC errors
by Sven Eckelmann 20 May '11

20 May '11

Hi, I would assume that you don't monitor the bug as you didn't answer the questions on irc. So here the part of the IRC communication which needs comments/answers from your side. <T_X> dbeberman-away: hmm, weird. could be a batman specific bug. the first thing I'd have in mind to narrow things down would be to try changing that line: http://git.open-mesh.org/?p=batman- adv.git;a=blob;f=send.c;h=33779278f1b2bfe49155bb9dfadc421305c0c8d7;hb=refs/heads/master#l485 <T_X> from skb_clone to skb_copy <T_X> from looking at the kernel code it seems like skb_clone() is only copying the skb-header, but not the skb-headroom (which we are modifying in send_skb_packet() ) --> jn_ (~jonathan(a)dslb-084-060-236-250.pools.arcor-ip.net) has joined #batman <ecsv_> T_X: and that is a problem.... why? <ecsv_> i don't see which crc is the problem <T_X> ecsv_: you mean it should not be a problem because we for clones we are writing the exact same data in the headroom again? <ecsv_> we should... but yes, it might be a good idea to implement it differently and to test it with skb_copy <ecsv_> it didn't checked the bug report... just assumed that you guessed :) <ecsv_> but still don't see what happened to which crc <ecsv_> we only modify stuff which is related to us.... and not to the encapsulated data <ecsv_> but maybe he meant the ethernet header/wifi stuff before the batman-adv header <ecsv_> but why does he see them? <ecsv_> and why are udp broadcasts fragmented? <ecsv_> we should only have fragmentation for unicast <T_X> ecsv_: I think he's talking about the wifi fragmentation <T_X> not the fragmentation in batman <ecsv_> dbeberman-away: and the only reason why we reject mails on the list is either that you are a known spammer or that you try to send html mails <ecsv_> did you get a reply? <T_X> and I suppose he's doing the sniffing on another interface in monitor mode <T_X> ah, and missed that the "my_skb_head_push()" is in fact making the headroom of the (cloned) skb writeable Kind regards, Sven

1 1

[B.A.T.M.A.N.] port to omnet++
by alfonso ariza quintana 19 May '11

19 May '11

I have adapted the code of the version 0.3.2 to omnet. If some one is interested to check the code, it's possible to download from https://github.com/inetmanet/inetmanet/tree/master/src/networklayer/manetro…

1 0

[B.A.T.M.A.N.] [PATCH 1/9] batman-adv: Annotate functions with format strings
by Sven Eckelmann 19 May '11

19 May '11

Signed-off-by: Sven Eckelmann <sven(a)narfation.org> --- bat_debugfs.c | 1 + main.h | 3 ++- 2 files changed, 3 insertions(+), 1 deletions(-) diff --git a/bat_debugfs.c b/bat_debugfs.c index abaeec5..4577d3f 100644 --- a/bat_debugfs.c +++ b/bat_debugfs.c @@ -50,6 +50,7 @@ static void emit_log_char(struct debug_log *debug_log, char c) debug_log->log_start = debug_log->log_end - log_buff_len; } +__printf(2, 3) static int fdebug_log(struct debug_log *debug_log, char *fmt, ...) { va_list args; diff --git a/main.h b/main.h index 3ca3941..f80954b 100644 --- a/main.h +++ b/main.h @@ -138,7 +138,7 @@ void dec_module_count(void); int is_my_mac(uint8_t *addr); #ifdef CONFIG_BATMAN_ADV_DEBUG -int debug_log(struct bat_priv *bat_priv, char *fmt, ...); +int debug_log(struct bat_priv *bat_priv, char *fmt, ...) __printf(2, 3); #define bat_dbg(type, bat_priv, fmt, arg...) \ do { \ @@ -147,6 +147,7 @@ int debug_log(struct bat_priv *bat_priv, char *fmt, ...); } \ while (0) #else /* !CONFIG_BATMAN_ADV_DEBUG */ +__printf(3, 4) static inline void bat_dbg(char type __always_unused, struct bat_priv *bat_priv __always_unused, char *fmt __always_unused, ...) -- 1.7.5.1

3 22

[B.A.T.M.A.N.] Extending the Exporter
by Elmo 18 May '11

18 May '11

Hello, I'm trying to extend the Exporter of OpenMesh to put another float stored at each Vertex to an off or obj file. But the Tutorial on your website dosn't make it clear i think. Do I have to extend the the BaseExporter by a Iterator for my float like: virtual PVertexIter const_myFloat_begin() = 0; virtual PVertexIter const_myFloat_end() = 0; Where do i have to add the code to get these iterators called? Is there anything else to add? Thanks in advance Malte

2 1

[B.A.T.M.A.N.] [PATCH] batman-adv: reset broadcast flood protection on error
by Marek Lindner 15 May '11

15 May '11

The broadcast flood protection should be reset to its original value if the primary interface could not be retrieved. Signed-off-by: Marek Lindner <lindner_marek(a)yahoo.de> --- send.c | 2 +- 1 files changed, 1 insertions(+), 1 deletions(-) diff --git a/send.c b/send.c index 76daa46..3377927 100644 --- a/send.c +++ b/send.c @@ -421,7 +421,7 @@ int add_bcast_packet_to_list(struct bat_priv *bat_priv, struct sk_buff *skb) primary_if = primary_if_get_selected(bat_priv); if (!primary_if) - goto out; + goto out_and_inc; forw_packet = kmalloc(sizeof(struct forw_packet), GFP_ATOMIC); -- 1.7.2.3

2 2

[B.A.T.M.A.N.] pull request: batman-adv 2011-05-14
by Sven Eckelmann 15 May '11

15 May '11

Hi, I would like tp propose following corrections for net-next-2.6/2.6.40. Both are bug fixes. The first one fixes the regression introduced by "batman-adv: Make bat_priv->primary_if an rcu protected pointer" which prevented that the attached net_devices could be destroyed because the references to them were incorrectly counted. The second one is a bug which may "filled" the broadcast queue when at the same time primary_if got changed and a broadcast was initiated. The queue counter were never reduced because there were no actual packets attached to the the queue. thanks, Sven The following changes since commit 27aea2128ec09924dfe08e97739b2bf8b15c8619: batman-adv: remove duplicate code from function is_bidirectional_neigh() (2011-05-08 16:10:42 +0200) are available in the git repository at: git://git.open-mesh.org/ecsv/linux-merge.git batman-adv/next Marek Lindner (1): batman-adv: reset broadcast flood protection on error Sven Eckelmann (1): batman-adv: Add missing hardif_free_ref in forw_packet_free net/batman-adv/aggregation.c | 14 +++++++++++--- net/batman-adv/send.c | 19 ++++++++++++++++--- 2 files changed, 27 insertions(+), 6 deletions(-)

2 3

[B.A.T.M.A.N.] bug - unregister_netdevice: usage_count = 1
by Antonio Quartulli 14 May '11

14 May '11

Hi all, I found a bug playing with my interfaces while bat0 was active. The bug is described here: http://www.open-mesh.org/issues/148 The steps to reproduce it are reported too. In a few words, when trying to unregister a device by unloading its module, the modprobe hangs and in the dmesg I can see this message: unregister_netdevice: waiting for wlan0 to become free. Usage count = 1 I tried to do some tests and what I observed is that hard_iface->refcount doesn't reach 0. This means that hard_iface_free_rcu() is never invoked and dev_put() too. I hope the information are enough.. Regards, -- Antonio Quartulli ..each of us alone is worth nothing.. Ernesto "Che" Guevara

3 4

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