tcp_tw_bucket is broken at "tcp_timewait_kill"

Hua Qin (qinhua@networks.ecse.rpi.edu)
Tue, 21 Jan 2003 17:27:38 -0500 (EST)


Hi,

I was running 2.4.18 kernel, after very heavy networking traffic, my
machine was panic.
at
START: tcp_timewait_kill at c021e08c
[f0adfe94] tcp_twkill__thr at c021e87e
[f0adfeac] tasklet_action at c0124644
[f0adfec8] do_softirq at c0124441
[f0adfee8] .text.lock.tcp at c020f692
[f0adff0c] do_generic_file_read at c01345ea
[f0adff38] inet_setsockopt at c0228063
[f0adff54] sys_setsockopt at c01e7c44
[f0adff80] sys_socketcall at c01e8303
[f0adffc0] system_call at c0108f53

I looked at tcp_tw_death_row:
tcp_tw_death_row = $13 =
{0xd4b4d380, 0x2, 0xf3089100, 0xd4728a80, 0xe0a51300, 0xd99c2580,
0xdab52a00, 0xeb568280}

You can see the second one "0x2" is not a structure of tcp_tw_bucket.

So my question is how this wrong info could possibly being added into
tcp_tw_death_row. or is this because this memory was not clean somewhere?

Thanks!
Hua

-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at http://vger.kernel.org/majordomo-info.html
Please read the FAQ at http://www.tux.org/lkml/