AnsweredAssumed Answered

System reboot caused by RCU warning

Question asked by Richard Li on Oct 30, 2016
Latest reply on Oct 31, 2016 by igorpadykov

My custom imx6q board  always reboot when running a long time LTP test.

hardware: imx6q with 1GHz,1G ddr3

kernel: Linux 3.0.35_4.1.0

 

After running a long time LTP test, such as 4 hours, the system always reboot caused by RCU warning.

I have no idea what caused the problem, kernel bug or ddr parameter?

 

By the way, the ddr3 parameter is calibrated with ddr_stress_tester by ourselves.

 

Does any one can help me?

Thanks!

 

[189627.422520] ------------[ cut here ]------------
[189627.427279] WARNING: at kernel/rcutree_plugin.h:180 rcu_note_context_switch+0x19c/0x1f0()
[189627.435546] Modules linked in:
[189627.438733] [<c00464ac>] (unwind_backtrace+0x0/0x138) from [<c007d634>] (warn_slowpath_common+0x4c/0x64)
[189627.448311] [<c007d634>] (warn_slowpath_common+0x4c/0x64) from [<c007d668>] (warn_slowpath_null+0x1c/0x24)
[189627.458061] [<c007d668>] (warn_slowpath_null+0x1c/0x24) from [<c00c5bd8>] (rcu_note_context_switch+0x19c/0x1f0)
[189627.468261] [<c00c5bd8>] (rcu_note_context_switch+0x19c/0x1f0) from [<c04f6574>] (__schedule+0x64/0x940)
[189627.477841] [<c04f6574>] (__schedule+0x64/0x940) from [<c04f6f34>] (preempt_schedule+0x3c/0x58)
[189627.486640] [<c04f6f34>] (preempt_schedule+0x3c/0x58) from [<c0209428>] (do_msgrcv+0x1ac/0x388)
[189627.495435] [<c0209428>] (do_msgrcv+0x1ac/0x388) from [<c0209630>] (sys_msgrcv+0x2c/0x54)
[189627.503722] [<c0209630>] (sys_msgrcv+0x2c/0x54) from [<c003f880>] (ret_fast_syscall+0x0/0x30)
[189627.512339] ---[ end trace 5e09246341fe919e ]---
[189681.037049] init: untracked pid 5827 exited
[189687.896645]  P4231 P4231 P4231 P4231

Outcomes