batman-adv rev. 1352 module crashes when loading

[anonymous]

Hello, today i compiled the lastest svn version for a Openmoko Freerunner device. When i try to load the module it crashes and dumps a trace, which i attached to this ticket.

[email@…]

trace log

[marek]

Just to confirm: You don't activate the module by adding interfaces ? You just load it (modprobe batman-adv) which makes it crash ?

[email@…]

The loading with "insmod batman-adv.ko".

[marek]

The function receive_raw_packet() which you can find in the tracelog you provided can't be called without an active interface. That leads to the conclusion you configured batman-adv somehow. We need to know what you configured in order to reproduce the issue. Also, the function proc_interfaces_write() that is mentioned as well gets called when you try to set an interface. I guess something is special about the interface you are trying to activate. If you are not aware of any options may be you have a start script somewhere ?

[email@…]

Sorry for the late response. You are right the driver for that device is "non-standard". I'm working with batman-adv on Freerunner GTA02 devices. They use a ar6000 chipset which is not supported by the linux-wireless group, because it is a full-mac hardware.

I configure the batman device with a ipv4 address and a ipv6 local-link address. Nothing else, if that is what you mean.

[anonymous]

Hmm.., I have to reproduce that otherwise it gets hard. I can dig out my Freerunner but I need the image you are using. Where can I get it ? And then I need the exact steps which lead to the crash.

[email@…]

You need the kernel i use, right? The other userspace related stuff shouldn't have anything to do with it.

I'm using a recent version of the "andy-branch" from the openmoko kernel. You can build one for yourself (http://wiki.openmoko.org/wiki/Kernel#Building_the_andy-tracking_branch).

If you need more than that, please contact me.

[marek]

It is very difficult to help you if you don't let us know what _exactly_ you are doing. You might have found a bug which we need to reproduce first at our end in order to fix it. With the information you supplied that is not possible.

Please provide the kernel image & rootfs you are using plus any startup scripts / commands / setup descriptions related to batman.

[email@…]

Ok, i was a bit shy to submit this, because i use a hackable1 image, which is pretty big (a debian based distribution). I uploaded the rootfs (including the kernel) to http://www-user.tu-chemnitz.de/~kait/gta02-image.tar.gz Please leave me a message when you got it, so i can delete it. The startup script i use is /etc/init.d/batman. The bootmanager I use is Qi, so the kernel that gets booted, is the one that /boot/uImage-GTA02.bin points to.

[marek]

Ok, I got it. Thanks for your help.
Would you mind joining our IRC channel or continue this via email ? The latency might improve. ;-)

[marek]

Not sure how to reach you, so I try it here again. I have fixed various startup problems. Could you please try it again using the latest version (at least rev 1409) and let me know about the result?

[kait]

Now it is not crashing direct after loading any more. But I get the following errors when doing after activating a interface with:

$echo "eth1" > /proc/net/batman-adv/interfaces

[21474598.555000] batman-adv: B.A.T.M.A.N. advanced 0.2-beta r1410 (compatibility version 7) loaded [21474598.635000] eth1: no IPv6 routers present [21474609.580000] BUG: sleeping function called from invalid context at /home/kai/openmoko/kernel/linux-2.6/mm/slab.c:3046 [21474609.585000] in_atomic(): 1, irqs_disabled(): 0, pid: 1635, name: bat_events [21474609.590000] 3 locks held by bat_events/1635: [21474609.600000] #0: (bat_events){....}, at: [<c0063844>] run_workqueue+0xa4/0x228 [21474609.605000] #1: ((hardif_check_interfaces_wq).work){....}, at: [<c0063844>] run_workqueue+0xa4/0x228 [21474609.610000] #2: (rcu_read_lock){....}, at: [<bf077914>] hardif_check_interfaces_status+0x20/0x12c [batman_adv] [21474609.615000] [<c0034960>] (dump_stack+0x0/0x14) from [<c004b94c>] (might_sleep+0xf0/0x110) [21474609.625000] [<c004b85c>] (might_sleep+0x0/0x110) from [<c00ac890>] (kmem_cache_alloc+0x24/0xa4) [21474609.630000] r5:c6d5c000 r4:c7815000 [21474609.635000] [<c00ac86c>] (kmem_cache_alloc+0x0/0xa4) from [<c02b0e38>] (sock_alloc_inode+0x20/0x64) [21474609.640000] r7:c7a20420 r6:c7815000 r5:c6d5c000 r4:c7815000 [21474609.645000] [<c02b0e18>] (sock_alloc_inode+0x0/0x64) from [<c00c377c>] (alloc_inode+0x30/0x68) [21474609.650000] r4:c7815000 [21474609.655000] [<c00c374c>] (alloc_inode+0x0/0x68) from [<c00c3b28>] (new_inode+0x18/0xac) [21474609.660000] r4:00000011 [21474609.665000] [<c00c3b10>] (new_inode+0x0/0xac) from [<c02b1a9c>] (sock_alloc+0x24/0x98) [21474609.670000] r6:c6d5df5c r5:c6d5c000 r4:00000011 [21474609.675000] [<c02b1a78>] (sock_alloc+0x0/0x98) from [<c02b2440>] (sock_create+0x7c/0x2c8) [21474609.680000] r5:00000003 r4:00000011 [21474609.690000] [<c02b23c4>] (sock_create+0x0/0x2c8) from [<c02b26c4>] (sock_create_kern+0x38/0x44) [21474609.695000] [<c02b268c>] (sock_create_kern+0x0/0x44) from [<bf077668>] (hardif_activate_interface+0x44/0x158 [batman_adv]) [21474609.700000] [<bf077624>] (hardif_activate_interface+0x0/0x158 [batman_adv]) from [<bf077978>] (hardif_check_interfaces_status+0x84/0x12c [batman_adv]) [21474609.710000] r4:c7afb740 [21474609.715000] [<bf0778f4>] (hardif_check_interfaces_status+0x0/0x12c [batman_adv]) from [<bf077a30>] (hardif_check_interfaces_status_wq+0x10/0x18 [batman_adv]) [21474609.730000] r4:00000000 [21474609.735000] [<bf077a20>] (hardif_check_interfaces_status_wq+0x0/0x18 [batman_adv]) from [<c00638c4>] (run_workqueue+0x124/0x228) [21474609.745000] [<c00637a0>] (run_workqueue+0x0/0x228) from [<c006472c>] (worker_thread+0xe4/0xf8) [21474609.755000] [<c0064648>] (worker_thread+0x0/0xf8) from [<c0067ee4>] (kthread+0x60/0x94) [21474609.760000] r6:c0064648 r5:c7a20420 r4:c6d5c000 [21474609.765000] [<c0067e84>] (kthread+0x0/0x94) from [<c005541c>] (do_exit+0x0/0x268) [21474609.770000] r6:00000000 r5:00000000 r4:00000000 [21474609.775000] BUG: scheduling while atomic: bat_events/1635/0x00000002 [21474609.780000] 4 locks held by bat_events/1635: [21474609.785000] #0: (bat_events){....}, at: [<c0063844>] run_workqueue+0xa4/0x228 [21474609.790000] #1: ((hardif_check_interfaces_wq).work){....}, at: [<c0063844>] run_workqueue+0xa4/0x228 [21474609.795000] #2: (rcu_read_lock){....}, at: [<bf077914>] hardif_check_interfaces_status+0x20/0x12c [batman_adv] [21474609.800000] #3: (sk_lock-AF_PACKET){....}, at: [<c03362a4>] packet_do_bind+0x24/0x12c [21474609.805000] Modules linked in: batman_adv ipv6 [21474609.810000] [<c0034960>] (dump_stack+0x0/0x14) from [<c004cf00>] (schedule_bug+0x58/0x64) [21474609.820000] [<c004cea8>] (schedule_bug+0x0/0x64) from [<c036a230>] (schedule+0x68/0x354) [21474609.825000] r5:c798e320 r4:c6d5de2c [21474609.830000] [<c036a1c8>] (schedule+0x0/0x354) from [<c036a7dc>] (schedule_timeout+0x20/0xc0) [21474609.835000] [<c036a7bc>] (schedule_timeout+0x0/0xc0) from [<c036a094>] (wait_for_common+0xdc/0x17c) [21474609.845000] r7:c6d5c000 r6:7fffffff r5:c6d5de28 r4:c6d5de2c [21474609.850000] [<c0369fb8>] (wait_for_common+0x0/0x17c) from [<c036a1c4>] (wait_for_completion+0x18/0x1c) [21474609.855000] r8:c6d5c000 r7:00000543 r6:c7a36000 r5:c6c506a0 r4:c6d5de20 [21474609.860000] [<c036a1ac>] (wait_for_completion+0x0/0x1c) from [<c00653ec>] (synchronize_rcu+0x38/0x44) [21474609.870000] [<c00653b4>] (synchronize_rcu+0x0/0x44) from [<c02bf21c>] (synchronize_net+0x1c/0x28) [21474609.880000] r4:c6c50400 [21474609.885000] [<c02bf200>] (synchronize_net+0x0/0x28) from [<c02bf23c>] (dev_remove_pack+0x14/0x18) [21474609.890000] [<c02bf228>] (dev_remove_pack+0x0/0x18) from [<c03362fc>] (packet_do_bind+0x7c/0x12c) [21474609.900000] [<c0336280>] (packet_do_bind+0x0/0x12c) from [<c0337910>] (packet_bind+0x70/0xa4) [21474609.905000] r7:c7a20420 r6:c6c50400 r5:c7a36000 r4:c6d5dee4 [21474609.910000] [<c03378a0>] (packet_bind+0x0/0xa4) from [<c02b0d20>] (kernel_bind+0x18/0x1c) [21474609.915000] r6:c6d5df5c r5:00000001 r4:c7afb740 [21474609.920000] [<c02b0d08>] (kernel_bind+0x0/0x1c) from [<bf0776a8>] (hardif_activate_interface+0x84/0x158 [batman_adv]) [21474609.925000] [<bf077624>] (hardif_activate_interface+0x0/0x158 [batman_adv]) from [<bf077978>] (hardif_check_interfaces_status+0x84/0x12c [batman_adv]) [21474609.945000] r4:c7afb740 [21474609.950000] [<bf0778f4>] (hardif_check_interfaces_status+0x0/0x12c [batman_adv]) from [<bf077a30>] (hardif_check_interfaces_status_wq+0x10/0x18 [batman_adv]) [21474609.965000] r4:00000000 [21474609.970000] [<bf077a20>] (hardif_check_interfaces_status_wq+0x0/0x18 [batman_adv]) from [<c00638c4>] (run_workqueue+0x124/0x228) [21474609.980000] [<c00637a0>] (run_workqueue+0x0/0x228) from [<c006472c>] (worker_thread+0xe4/0xf8) [21474609.985000] [<c0064648>] (worker_thread+0x0/0xf8) from [<c0067ee4>] (kthread+0x60/0x94) [21474609.995000] r6:c0064648 r5:c7a20420 r4:c6d5c000 [21474610.005000] [<c0067e84>] (kthread+0x0/0x94) from [<c005541c>] (do_exit+0x0/0x268) [21474610.010000] r6:00000000 r5:00000000 r4:00000000

[kait]

Sorry i forgot the code block:

[ 2794.200000] batman-adv: B.A.T.M.A.N. advanced 0.2-beta r1410 (compatibility version 7) loaded 
[ 2819.215000] BUG: sleeping function called from invalid context at /home/kai/openmoko/kernel/linux-2.6/mm/slab.c:3046
[ 2819.220000] in_atomic(): 1, irqs_disabled(): 0, pid: 2063, name: bat_events
[ 2819.225000] 3 locks held by bat_events/2063:
[ 2819.230000]  #0:  (bat_events){....}, at: [<c0063844>] run_workqueue+0xa4/0x228
[ 2819.235000]  #1:  ((hardif_check_interfaces_wq).work){....}, at: [<c0063844>] run_workqueue+0xa4/0x228
[ 2819.245000]  #2:  (rcu_read_lock){....}, at: [<bf08a914>] hardif_check_interfaces_status+0x20/0x12c [batman_adv]
[ 2819.255000] [<c0034960>] (dump_stack+0x0/0x14) from [<c004b94c>] (__might_sleep+0xf0/0x110)
[ 2819.265000] [<c004b85c>] (__might_sleep+0x0/0x110) from [<c00ac890>] (kmem_cache_alloc+0x24/0xa4)
[ 2819.275000]  r5:c7a2c000 r4:c7815000
[ 2819.275000] [<c00ac86c>] (kmem_cache_alloc+0x0/0xa4) from [<c02b0e38>] (sock_alloc_inode+0x20/0x64)
[ 2819.285000]  r7:c7a20420 r6:c7815000 r5:c7a2c000 r4:c7815000
[ 2819.290000] [<c02b0e18>] (sock_alloc_inode+0x0/0x64) from [<c00c377c>] (alloc_inode+0x30/0x68)
[ 2819.300000]  r4:c7815000
[ 2819.300000] [<c00c374c>] (alloc_inode+0x0/0x68) from [<c00c3b28>] (new_inode+0x18/0xac)
[ 2819.310000]  r4:00000011
[ 2819.310000] [<c00c3b10>] (new_inode+0x0/0xac) from [<c02b1a9c>] (sock_alloc+0x24/0x98)
[ 2819.320000]  r6:c7a2df5c r5:c7a2c000 r4:00000011
[ 2819.325000] [<c02b1a78>] (sock_alloc+0x0/0x98) from [<c02b2440>] (__sock_create+0x7c/0x2c8)
[ 2819.335000]  r5:00000003 r4:00000011
[ 2819.335000] [<c02b23c4>] (__sock_create+0x0/0x2c8) from [<c02b26c4>] (sock_create_kern+0x38/0x44)
[ 2819.345000] [<c02b268c>] (sock_create_kern+0x0/0x44) from [<bf08a668>] (hardif_activate_interface+0x44/0x158 [batman_adv])
[ 2819.355000] [<bf08a624>] (hardif_activate_interface+0x0/0x158 [batman_adv]) from [<bf08a978>] (hardif_check_interfaces_status+0x84/0x12c [batman_adv])
[ 2819.370000]  r4:c7b94900
[ 2819.370000] [<bf08a8f4>] (hardif_check_interfaces_status+0x0/0x12c [batman_adv]) from [<bf08aa30>] (hardif_check_interfaces_status_wq+0x10/0x18 [batman_adv])
[ 2819.385000]  r4:00000000
[ 2819.390000] [<bf08aa20>] (hardif_check_interfaces_status_wq+0x0/0x18 [batman_adv]) from [<c00638c4>] (run_workqueue+0x124/0x228)
[ 2819.400000] [<c00637a0>] (run_workqueue+0x0/0x228) from [<c006472c>] (worker_thread+0xe4/0xf8)
[ 2819.410000] [<c0064648>] (worker_thread+0x0/0xf8) from [<c0067ee4>] (kthread+0x60/0x94)
[ 2819.415000]  r6:c0064648 r5:c7a20420 r4:c7a2c000
[ 2819.420000] [<c0067e84>] (kthread+0x0/0x94) from [<c005541c>] (do_exit+0x0/0x268)
[ 2819.430000]  r6:00000000 r5:00000000 r4:00000000
[ 2819.435000] BUG: scheduling while atomic: bat_events/2063/0x00000002
[ 2819.440000] 4 locks held by bat_events/2063:
[ 2819.445000]  #0:  (bat_events){....}, at: [<c0063844>] run_workqueue+0xa4/0x228
[ 2819.450000]  #1:  ((hardif_check_interfaces_wq).work){....}, at: [<c0063844>] run_workqueue+0xa4/0x228
[ 2819.460000]  #2:  (rcu_read_lock){....}, at: [<bf08a914>] hardif_check_interfaces_status+0x20/0x12c [batman_adv]
[ 2819.470000]  #3:  (sk_lock-AF_PACKET){....}, at: [<c03362a4>] packet_do_bind+0x24/0x12c
[ 2819.480000] Modules linked in: batman_adv ipv6 [last unloaded: batman_adv]
[ 2819.485000] [<c0034960>] (dump_stack+0x0/0x14) from [<c004cf00>] (__schedule_bug+0x58/0x64)
[ 2819.495000] [<c004cea8>] (__schedule_bug+0x0/0x64) from [<c036a230>] (schedule+0x68/0x354)
[ 2819.500000]  r5:c6f1c0a0 r4:c7a2de2c
[ 2819.505000] [<c036a1c8>] (schedule+0x0/0x354) from [<c036a7dc>] (schedule_timeout+0x20/0xc0)
[ 2819.515000] [<c036a7bc>] (schedule_timeout+0x0/0xc0) from [<c036a094>] (wait_for_common+0xdc/0x17c)
[ 2819.525000]  r7:c7a2c000 r6:7fffffff r5:c7a2de28 r4:c7a2de2c
[ 2819.530000] [<c0369fb8>] (wait_for_common+0x0/0x17c) from [<c036a1c4>] (wait_for_completion+0x18/0x1c)
[ 2819.540000]  r8:c7a2c000 r7:00000543 r6:c7a36000 r5:c7b3b6a0 r4:c7a2de20
[ 2819.545000] [<c036a1ac>] (wait_for_completion+0x0/0x1c) from [<c00653ec>] (synchronize_rcu+0x38/0x44)
[ 2819.555000] [<c00653b4>] (synchronize_rcu+0x0/0x44) from [<c02bf21c>] (synchronize_net+0x1c/0x28)
[ 2819.560000]  r4:c7b3b400
[ 2819.565000] [<c02bf200>] (synchronize_net+0x0/0x28) from [<c02bf23c>] (dev_remove_pack+0x14/0x18)
[ 2819.575000] [<c02bf228>] (dev_remove_pack+0x0/0x18) from [<c03362fc>] (packet_do_bind+0x7c/0x12c)
[ 2819.580000] [<c0336280>] (packet_do_bind+0x0/0x12c) from [<c0337910>] (packet_bind+0x70/0xa4)
[ 2819.590000]  r7:c7a20420 r6:c7b3b400 r5:c7a36000 r4:c7a2dee4
[ 2819.595000] [<c03378a0>] (packet_bind+0x0/0xa4) from [<c02b0d20>] (kernel_bind+0x18/0x1c)
[ 2819.605000]  r6:c7a2df5c r5:00000001 r4:c7b94900
[ 2819.610000] [<c02b0d08>] (kernel_bind+0x0/0x1c) from [<bf08a6a8>] (hardif_activate_interface+0x84/0x158 [batman_adv])
[ 2819.620000] [<bf08a624>] (hardif_activate_interface+0x0/0x158 [batman_adv]) from [<bf08a978>] (hardif_check_interfaces_status+0x84/0x12c [batman_adv])
[ 2819.635000]  r4:c7b94900
[ 2819.635000] [<bf08a8f4>] (hardif_check_interfaces_status+0x0/0x12c [batman_adv]) from [<bf08aa30>] (hardif_check_interfaces_status_wq+0x10/0x18 [batman_adv])
[ 2819.650000]  r4:00000000
[ 2819.650000] [<bf08aa20>] (hardif_check_interfaces_status_wq+0x0/0x18 [batman_adv]) from [<c00638c4>] (run_workqueue+0x124/0x228)
[ 2819.665000] [<c00637a0>] (run_workqueue+0x0/0x228) from [<c006472c>] (worker_thread+0xe4/0xf8)
[ 2819.670000] [<c0064648>] (worker_thread+0x0/0xf8) from [<c0067ee4>] (kthread+0x60/0x94)
[ 2819.680000]  r6:c0064648 r5:c7a20420 r4:c7a2c000
[ 2819.685000] [<c0067e84>] (kthread+0x0/0x94) from [<c005541c>] (do_exit+0x0/0x268)
[ 2819.690000]  r6:00000000 r5:00000000 r4:00000000

[marek]

Replying to kait:

Now it is not crashing direct after loading any more.

That is good news! :-)

But I get the following errors when doing after activating a interface with: $echo "eth1" > /proc/net/batman-adv/interfaces

Wow - the freerunner seems to trigger every (almost) impossible race condition that no-one saw before. ;-) I see the problem but I have to think about a solution ..

[marek]

Ok, r1425 should fix this issue.

Keep us posted. :-)

[kait]

Ok, looks like you fixed this one :) I tested it with r1426.

Now the nodes don't crash anymore. But (there always is a "but", isn't it?) they don't build a mesh. The nodes are sending their originator messages, but the other nodes don't receive them. At least it looks like that, from my point of view.

I have no idea what more information i could give you. The logfiles are all looking normal, no error messages at all.

[marek]

Yes, this was a regression that came with r1425. We fixed it in r1427.

[kait]

Tested with r1427 today. I get these errors when setting the ip address

[  368.980000] batman-adv: B.A.T.M.A.N. advanced 0.2-beta r1427 (compatibility version 8) loaded 
[  374.140000] BUG: sleeping function called from invalid context at /home/kai/openmoko/kernel/linux-2.6/mm/slab.c:3046
[  374.150000] in_atomic(): 1, irqs_disabled(): 0, pid: 1958, name: batman
[  374.155000] 1 lock held by batman/1958:
[  374.160000]  #0:  (vis_hash_lock){....}, at: [<bf06326c>] vis_init+0x28/0x1a0 [batman_adv]
[  374.165000] [<c0034960>] (dump_stack+0x0/0x14) from [<c004b94c>] (__might_sleep+0xf0/0x110)
[  374.175000] [<c004b85c>] (__might_sleep+0x0/0x110) from [<c00ac890>] (kmem_cache_alloc+0x24/0xa4)
[  374.185000]  r5:00000005 r4:bf06a300
[  374.185000] [<c00ac86c>] (kmem_cache_alloc+0x0/0xa4) from [<bf06329c>] (vis_init+0x58/0x1a0 [batman_adv])
[  374.195000]  r7:c6ca3b00 r6:00000000 r5:00000005 r4:bf06a300
[  374.200000] [<bf063244>] (vis_init+0x0/0x1a0 [batman_adv]) from [<bf05c56c>] (activate_module+0x54/0xec [batman_adv])
[  374.210000] [<bf05c518>] (activate_module+0x0/0xec [batman_adv]) from [<bf05db90>] (proc_interfaces_write+0x1d4/0x2c8 [batman_adv])
[  374.225000]  r6:00000000 r5:00000005 r4:bf06a300
[  374.230000] [<bf05d9bc>] (proc_interfaces_write+0x0/0x2c8 [batman_adv]) from [<c00eaa40>] (proc_reg_write+0x90/0xa4)
[  374.240000]  r8:c6c31f78 r7:c68ce7a0 r6:fffffffb r5:c7a4d240 r4:bf05d9bc
[  374.245000] [<c00ea9b0>] (proc_reg_write+0x0/0xa4) from [<c00b01dc>] (vfs_write+0xbc/0x14c)
[  374.255000] [<c00b0120>] (vfs_write+0x0/0x14c) from [<c00b07a8>] (sys_write+0x4c/0x7c)
[  374.260000]  r7:00000004 r6:00000000 r5:00000000 r4:c68ce7a0
[  374.265000] [<c00b075c>] (sys_write+0x0/0x7c) from [<c002fdc0>] (ret_fast_syscall+0x0/0x2c)
[  374.275000]  r6:00000001 r5:00000005 r4:00028258

This is the startscript I use:

#!/bin/sh
insmod /root/batman-adv.ko
sleep 5 
echo "eth1" > /proc/net/batman-adv/interfaces
sleep 10
ip -6 addr add 2001:bd8:4711::102/64 dev bat0

[marek]

It is a bit confusing because the error log you pasted seems unrelated to the IP address. In the log message you can see proc_interfaces_write() which gets called if you add an interface to /proc/.../interfaces. Are you sure it is related to the IP address ?
Using your startscript does not provoke a crash here but you might run into a race condition that I don't trigger. You are running this on top of 2.6.28 or 2.6.31 ?

[kait]

You are right. The module crashes when adding the interface. Sorry, i got that wrong.

The Kernel:

$ uname -a
$ Linux runner2 2.6.29-GTA02_andy-tracking-mokodev #5 PREEMPT Fri Aug 28 11:50:30 CEST 2009 armv4tl GNU/Linux

[marek]

I made a patch for that bug (r1432), please check it.
Are you running the kernel with debugging enabled ? If so, I'd like to know what options you have. I started the module a dozen times but I'm not able to trigger this. :-)

[kait]

Looks line r1432 is working. At least it doesn't crash anymore.

I don't know if this has anything to do with batman at all, but I have problems when trying to ping other nodes via the bat0 device.

# ping6 fe80::2ff:fff:fed2:cc99%bat0
ping: getnameinfo: Temporary failure in name resolution
64 bytes from unknown: icmp_seq=0 ttl=64 time=44.101 ms

# ping6 2001:bd8:4711::102
PING 2001:bd8:4711::102 (2001:bd8:4711::102): 56 data bytes
ping: getnameinfo: Temporary failure in name resolution
64 bytes from unknown: icmp_seq=0 ttl=64 time=1.186 ms

# ping6 fe80::212:cfff:fe8f:587%eth1
PING fe80::212:cfff:fe8f:587%eth1 (fe80::212:cfff:fe8f:587%eth1): 56 data bytes
64 bytes from runner1: icmp_seq=0 ttl=64 time=25.354 ms
64 bytes from runner1: icmp_seq=1 ttl=64 time=10.596 ms
64 bytes from runner1: icmp_seq=2 ttl=64 time=12.319 ms
[...]

I use ipv6 addresses and this problem exists over the global scope and the link local scope address, but only when using the bat0 device. When I use the wireless device directly, I have no problems. Thats the reason why I think, that maybe it is a problem with batman.

I attached the kernel config file.

[kait]

kernel config

[T_X]

Hi kait. I'm having my doubts that this might be batman-related too. Well, it should be normal that you're not able to do any recursive link-local address lookups as they are only unique for the certain link, right? But I'm also wondering, where you got this runner1 from, did you make any entries in your /etc/hosts for that? My output with batman or without looks just like this with a current batman-svn-version and the default debian 2.6.30-kernel from unstable:

ping6 fe80::2ff:74ff:fe52:fbfa%bat0
PING fe80::2ff:74ff:fe52:fbfa%bat0(fe80::2ff:74ff:fe52:fbfa) 56 data bytes
64 bytes from fe80::2ff:74ff:fe52:fbfa: icmp_seq=1 ttl=64 time=0.057 ms
64 bytes from fe80::2ff:74ff:fe52:fbfa: icmp_seq=2 ttl=64 time=0.084 ms

Hmm, maybe a 'ping6 -n' could be a workaround for the moment ;)?

[marek]

Can we close this ticket assuming there are no further problems ?