Observed on both qemuarm and qemuppc (not tried other archs yet), soft lockup warnings are observed after leaving for over 1 hrs (I didn't check in the middle). root@qemuppc:~# BUG: soft lockup - CPU#0 stuck for 303s! [swapper:0] Modules linked in: x_tables ip_tables iptable_filter NIP: c0009ff8 LR: c0009ff8 CTR: c000d754 REGS: c04ddee0 TRAP: 0901 Not tainted (2.6.34.7-wr-standard) MSR: 00009032 <EE,ME,IR,DR> CR: 24000028 XER: 20000000 TASK = c04b4400[0] 'swapper' THREAD: c04dc000 GPR00: 00000000 c04ddf90 c04b4400 00800000 0090c000 00e00000 ffda8030 00049032 GPR08: 00000001 c04dc000 24000022 00000000 8e02f6f0 NIP [c0009ff8] cpu_idle+0x8c/0xf8 LR [c0009ff8] cpu_idle+0x8c/0xf8 Call Trace: [c04ddf90] [c000a05c] cpu_idle+0xf0/0xf8 (unreliable) [c04ddfb0] [c0003e94] rest_init+0xb0/0xdc [c04ddfc0] [c0485774] start_kernel+0x2e8/0x2fc [c04ddff0] [00003438] 0x3438 Instruction dump: 7c00e878 7c00f12d 40a2fff4 7c0000a6 5400045e 7c000124 801e0000 70090004 40820010 801c00a0 7c0903a6 4e800421 <7c0000a6> 60008000 7c000124 7c00f028 One thing I did earlier is to test iptables, and so far when this bug is reported NAT table is missing. Not sure whether this test action may lead to this error or it's more generic. Record here for tracking purpose first.
btw, this is the lsb image
Is this related to the iptables? I am not sure if we saw this during the demos which did not have iptables enabled.
this is not reproducible any more
verify it