Re: [BUG] v3.12-rc5-139-gbdeeab6 assertion failed at drivers/net/bonding/bond_main.c (3398)

On Sat, Oct 19, 2013 at 10:27:02PM +0200, Thomas Glanzmann wrote:

Hello,
I'm referring to commit 7864a1adf7291993d74923fdd0a45459ce9da27e. I just
checked out the tip from Linus tree compiled it and run it on Debian Wheezy
with 4 1 GBIT NICs, two connected to one switch and two to another. I
configured bonding to use tlb. On the ports is one untagged and one
tagged VLAN. When I boot the kernel even if only one NIC is attached I
get multiple failed assertions:

...snip...

[   19.080783] RTNL: assertion failed at drivers/net/bonding/bond_main.c (3398)
[   19.080868] CPU: 4 PID: 0 Comm: swapper/4 Not tainted 3.12.0-rc5+ #1
[   19.080944] Hardware name: Supermicro X9SRD-F/X9SRD-F, BIOS 1.0a 10/15/2012
[   19.081022]  0000000000000000 0000000000000000 ffffffff81373aba ffff88103e1a0000
[   19.081316]  ffffffffa040aea2 ffff88103e1a0000 0000000000000000 ffff88103e1a0000
[   19.081607]  ffffffff812c54f5 00000000ffffffff ffff88107fc83c90 ffff88107f000480
[   19.081912] Call Trace:
[   19.081982]  <IRQ>  [<ffffffff81373aba>] ? dump_stack+0x41/0x51
[   19.082176]  [<ffffffffa040aea2>] ? bond_set_rx_mode+0x2d/0xa7 [bonding]
[   19.082258]  [<ffffffff812c54f5>] ? __dev_mc_add+0x48/0x59
[   19.082337]  [<ffffffff8134d4ee>] ? igmp6_group_added+0x65/0x17d
[   19.082418]  [<ffffffff810fc836>] ? kmem_cache_alloc_trace+0xbb/0xcb
[   19.082497]  [<ffffffff8134e7dd>] ? ipv6_dev_mc_inc+0x20e/0x236
[   19.082578]  [<ffffffff81339f70>] ? addrconf_join_solict+0x2e/0x33
[   19.082656]  [<ffffffff81332e22>] ? ipv6_dev_ac_inc+0x131/0x16f
[   19.082735]  [<ffffffff81336f67>] ? addrconf_join_anycast+0x38/0x3e
[   19.082814]  [<ffffffff8133a0b1>] ? __ipv6_ifa_notify+0x106/0x234
[   19.082893]  [<ffffffff8133a20d>] ? ipv6_ifa_notify+0x2e/0x37
[   19.082971]  [<ffffffff8133a660>] ? addrconf_dad_completed+0x32/0x185
[   19.083050]  [<ffffffff8133a874>] ? addrconf_dad_timer+0xc1/0x13c
[   19.083129]  [<ffffffff8133a7b3>] ? addrconf_dad_completed+0x185/0x185
[   19.083210]  [<ffffffff81040613>] ? call_timer_fn+0x4b/0xf6
[   19.083288]  [<ffffffff8133a7b3>] ? addrconf_dad_completed+0x185/0x185
[   19.083367]  [<ffffffff81040c04>] ? run_timer_softirq+0x189/0x1ce
[   19.083448]  [<ffffffff8107c7dd>] ? tick_sched_do_timer+0x25/0x25
[   19.083527]  [<ffffffff8103ad68>] ? __do_softirq+0xe8/0x201
[   19.083606]  [<ffffffff81075d98>] ? ktime_get+0x5f/0x6b
[   19.083682]  [<ffffffff8107afca>] ? clockevents_program_event+0x9a/0xb6
[   19.083761]  [<ffffffff8137df5c>] ? call_softirq+0x1c/0x30
[   19.083839]  [<ffffffff81003b7c>] ? do_softirq+0x2c/0x60
[   19.083916]  [<ffffffff8103af46>] ? irq_exit+0x3b/0x7f
[   19.083994]  [<ffffffff81023b19>] ? smp_apic_timer_interrupt+0x2c/0x37
[   19.084074]  [<ffffffff8137d48a>] ? apic_timer_interrupt+0x6a/0x70
[   19.084149]  <EOI>  [<ffffffff8129a6b1>] ? cpuidle_enter_state+0x43/0xa6
[   19.084330]  [<ffffffff8129a6aa>] ? cpuidle_enter_state+0x3c/0xa6
[   19.084408]  [<ffffffff8129a7e6>] ? cpuidle_idle_call+0xd2/0x142
[   19.084486]  [<ffffffff81008ff9>] ? arch_cpu_idle+0x6/0x17
[   19.084563]  [<ffffffff8106f7e9>] ? cpu_startup_entry+0x10d/0x180
[   19.084641]  [<ffffffff8137780c>] ? _raw_spin_unlock_irqrestore+0x6/0x7
[   19.084726]  [<ffffffff810226dd>] ? start_secondary+0x1df/0x1e5

Fixed in commit b32418705107265dfca5edfe2b547643e53a732e ("bonding: RCUify
bond_set_rx_mode()") net-next. It should get into mainline soon.

...snip...

Howto reproduce:

       - Build the kernel from Torvalds tip
       - Configure bonding with tlb with one NIC.

auto bond0
iface bond0 inet static
       address ...
       netmask ...
       network ...
       broadcast  ...
       gateway ...
       bond-mode balance-tlb
       bond-miimon 100
       slaves eth0
       # eth1 eth2 eth3

Cheers,
       Thomas