module unloading hangs after using ftrace with 3.11.0-rc1

From: Gleb Natapov
Date: Sun Jul 21 2013 - 07:52:44 EST


Hi Steven,

After tracing KVM using trace-cmd with kvm trace points enables I tried
to unload kvm modules and rmmod hanged in tracing_reset_all_online_cpus:

[ 1234.665038] rmmod D 00000000ffffffff 0 7819 7664 0x00000000
[ 1234.665038] ffff880129661d38 0000000000000086 ffff880374c7b4d0 ffff880129661fd8
[ 1234.665038] 0000000000004000 0000000000012a40 ffff88017af63180 ffff880179fe3180
[ 1234.665038] ffff880129661c78 ffffffff810ad4f9 0000000000000008 0000000000000002
[ 1234.665038] Call Trace:
[ 1234.665038] [<ffffffff810ad4f9>] ? rcu_is_cpu_idle+0x38/0x57
[ 1234.665038] [<ffffffff81139e51>] ? d_set_d_op+0xb5/0xb5
[ 1234.665038] [<ffffffff813afbcc>] schedule+0x64/0x66
[ 1234.665038] [<ffffffff813afe9a>] schedule_preempt_disabled+0x18/0x24
[ 1234.665038] [<ffffffff813ae900>] __mutex_lock_slowpath+0x156/0x1c2
[ 1234.665038] [<ffffffff813adbad>] mutex_lock+0x18/0x29
[ 1234.665038] [<ffffffff810bfaad>] tracing_reset_all_online_cpus+0x12/0x49
[ 1234.665038] [<ffffffff810cd105>] trace_module_notify+0x27b/0x2a2
[ 1234.665038] [<ffffffff81060aab>] notifier_call_chain+0x37/0x63
[ 1234.665038] [<ffffffff81060be9>] __blocking_notifier_call_chain+0x4b/0x60
[ 1234.665038] [<ffffffff81060c12>] blocking_notifier_call_chain+0x14/0x16
[ 1234.665038] [<ffffffff8108cf62>] SyS_delete_module+0x228/0x274
[ 1234.665038] [<ffffffff813b1d82>] system_call_fastpath+0x16/0x1b

I haven't verified how reproducible this is yet.

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