Badness in local_bh_enable at kernel/softirq.c:109
Call Trace:
[<c0127ac0>] local_bh_enable+0x90/0xa0
[<d0897ee5>] ppp_async_push+0xe5/0x240 [ppp_async]
[<d089772e>] ppp_asynctty_wakeup+0x2e/0x60 [ppp_async]
[<c01f0aa5>] uart_flush_buffer+0xa5/0xf0
[<c01d2b9e>] do_tty_hangup+0x56e/0x640
[<c0134d66>] worker_thread+0x226/0x3d0
[<c01d2630>] do_tty_hangup+0x0/0x640
[<c011d5e0>] default_wake_function+0x0/0x30
[<c0109916>] ret_from_fork+0x6/0x14
[<c011d5e0>] default_wake_function+0x0/0x30
[<c0134b40>] worker_thread+0x0/0x3d0
[<c01074e5>] kernel_thread_helper+0x5/0x10
This happens in a SMP p3 2x800 system, config.gz attached.
It was triggered (never happened before) using the "wondersharper" from
lartc.org (which uses tc)