Hi Paul, Fortunately this bug is bisectable and the first bad commit is: commit 9b2e4f1880b789be1f24f9684f7a54b90310b5c0 Author: Paul E. McKenney Date: Fri Sep 30 12:10:22 2011 -0700 rcu: Track idleness independent of idle tasks However, at that commit, the error messages are a bit different: [ 1.665656] WARNING: at /c/wfg/linux/kernel/rcutiny.c:67 rcu_idle_enter_common+0x103/0x1b2() [ 1.668811] WARNING: at /c/wfg/linux/kernel/rcutiny.c:112 rcu_idle_exit_common+0xfa/0x18c() [ 4.758670] WARNING: at /c/wfg/linux/kernel/trace/trace_events.c:1720 event_trace_self_tests_init+0x87/0x11b() Here is the first back trace: [ 1.660165] Testing tracer function: [ 1.664825] ------------[ cut here ]------------ [ 1.665656] WARNING: at /c/wfg/linux/kernel/rcutiny.c:67 rcu_idle_enter_common+0x103/0x1b2() [ 1.667006] Hardware name: Bochs [ 1.667583] Modules linked in: [ 1.668106] Pid: 0, comm: swapper Not tainted 3.2.0-rc5+ #41 [ 1.668811] Call Trace: [ 1.668811] [] ? printk+0x2b/0x3b [ 1.668811] [] warn_slowpath_common+0xfc/0x13b [ 1.668811] [] ? rcu_idle_enter_common+0x103/0x1b2 [ 1.668811] [] ? rcu_idle_enter_common+0x103/0x1b2 [ 1.668811] [] warn_slowpath_null+0x3e/0x4e [ 1.668811] [] rcu_idle_enter_common+0x103/0x1b2 [ 1.668811] [] rcu_irq_exit+0x1f8/0x20f [ 1.668811] [] irq_exit+0x117/0x14d [ 1.668811] [] do_IRQ+0xe5/0x106 [ 1.668811] [] ? trace_hardirqs_off_caller+0x97/0x167 [ 1.668811] [] common_interrupt+0x35/0x3c [ 1.668811] [] ? print_shortest_lock_dependencies+0xec/0x292 [ 1.668811] [] ? native_safe_halt+0x18/0x28 [ 1.668811] [] default_idle+0xd7/0x1f0 [ 1.668811] [] cpu_idle+0x1c4/0x2c8 [ 1.668811] [] rest_init+0x1db/0x1ef [ 1.668811] [] ? reciprocal_value+0x64/0x64 [ 1.668811] [] start_kernel+0x8e5/0x8f9 [ 1.668811] [] ? loglevel+0x55/0x55 [ 1.668811] [] ? reserve_ebda_region+0xb9/0xc9 [ 1.668811] [] i386_start_kernel+0x13a/0x14f Thanks, Fengguang