[RT] - netfilter warnings preceeding panic on 2.6.29-rc4-rt1 +Ingo's IRQ fixes

From: Clark Williams
Date: Thu Feb 12 2009 - 17:50:14 EST


-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

Thomas/Ingo,

After applying the IPI and HARDIRQ patches from Ingo, the kernel booted
on my T60 and I ran for some time, before encountering this warning:

- ------------[ cut here ]------------
WARNING: at include/net/netfilter/nf_conntrack.h:185 nf_ct_deliver_cached_events+0x8e/0xb0()
Hardware name:
Modules linked in: vfat fat usb_storage fuse i915 drm i2c_algo_bit cfbcopyarea cfbimgblt cfbfillrect autofs4 coretemp sunrpc nf_conntrack_netbios_ns xt_state ipt_REJECT iptable_filter ip_tables cpufreq_ondemand dm_multipath scsi_dh uinput btusb pl2303 usbserial bluetooth sg snd_hda_codec_analog snd_hda_intel e1000e snd_hda_codec iwl3945 snd_hwdep lib80211 snd_seq_dummy snd_seq_oss snd_seq_midi_event snd_seq snd_seq_device snd_pcm_oss video snd_mixer_oss snd_pcm output snd_timer iTCO_wdt i2c_i801 i2c_core iTCO_vendor_support button snd thinkpad_acpi rfkill hwmon joydev soundcore snd_page_alloc dm_snapshot dm_zero dm_mirror dm_region_hash dm_log uhci_hcd ohci_hcd ehci_hcd [last unloaded: microcode]
Pid: 29166, comm: ssh Not tainted 2.6.29-rc4-rt1-tip #51
Call Trace:
[<ffffffff8023bd1b>] warn_slowpath+0xaf/0xd6
[<ffffffffa01ea9ce>] ? ipt_do_table+0x623/0x67d [ip_tables]
[<ffffffff805449cf>] ? preempt_schedule_irq+0x4f/0x6b
[<ffffffff8020ca46>] ? retint_kernel+0x26/0x30
[<ffffffff80548852>] ? __atomic_notifier_call_chain+0x40/0x4a
[<ffffffff804bb4ae>] nf_ct_deliver_cached_events+0x8e/0xb0
[<ffffffff804f9355>] ipv4_confirm+0xe3/0xf2
[<ffffffff804b403d>] nf_iterate+0x48/0xa6
[<ffffffff804c7fbc>] ? ip_finish_output+0x0/0x31f
[<ffffffff804b410e>] nf_hook_slow+0x73/0xdb
[<ffffffff804c7fbc>] ? ip_finish_output+0x0/0x31f
[<ffffffff804b4102>] ? nf_hook_slow+0x67/0xdb
[<ffffffff804c83b1>] ip_output+0xd6/0xea
[<ffffffff804c723a>] ? __ip_local_out+0x86/0x88
[<ffffffff804c725c>] ip_local_out+0x20/0x24
[<ffffffff804c7aed>] ip_queue_xmit+0x2c1/0x367
[<ffffffff804d94db>] tcp_transmit_skb+0x676/0x6b4
[<ffffffff804dbc42>] tcp_write_xmit+0x828/0x916
[<ffffffff804dbd8b>] __tcp_push_pending_frames+0x2a/0x82
[<ffffffff804d0328>] tcp_sendmsg+0x92a/0xa4f
[<ffffffff8048c641>] ? sock_common_recvmsg+0x34/0x45
[<ffffffff8048a465>] sock_aio_write+0x134/0x14c
[<ffffffff802b8bea>] do_sync_write+0xe3/0x120
[<ffffffff8024f9ad>] ? autoremove_wake_function+0x0/0x34
[<ffffffff805447f6>] ? thread_return+0x43/0xcb
[<ffffffff80342390>] ? security_file_permission+0x11/0x13
[<ffffffff802b9475>] vfs_write+0xc0/0x139
[<ffffffff802b95a7>] sys_write+0x45/0x6c
[<ffffffff8020bf5b>] system_call_fastpath+0x16/0x1b
- ---[ end trace 8c162c27ac80924f ]---

I have no idea why nf_conntrack_netbios_ns was loaded. Looking into
that a bit.

I got three of these warnings and then the system panic'ed; hard lockup
of the laptop, so no traceback. I'm rebuilding without netfilter
conntrack support now, just to sidestep this, but I'll turn it back on
for more testing.

Attached is a dmesg from just prior to the panic. Same config as
previous.

Clark


-----BEGIN PGP SIGNATURE-----
Version: GnuPG v2.0.10 (GNU/Linux)

iEYEARECAAYFAkmUp28ACgkQHyuj/+TTEp1ubQCfbFvpwmXH6Tv/gSxrID+XuiuW
Q3YAoLMEIa08PSXxqkTcWQvcEkxzWeu/
=SMR9
-----END PGP SIGNATURE-----

Attachment: rt-2.6.29-rc4-rt1-nf_conntrack_warning.dmesg
Description: Binary data