[EC2:cg1.4xlarge] CPU#0 stuck for 23s! [migration/0:6] __do_softirq+0x60/0x210
Affects | Status | Importance | Assigned to | Milestone | |
---|---|---|---|---|---|
linux (Ubuntu) |
Triaged
|
Low
|
Stefan Bader |
Bug Description
Scott noticed some warnings when starting a cg1.4xlarge (HVM) Amazon instance. Right after activating all the VCPUs we see the following error repeated a few times (3-4) then boot continues and I could not see any problems resulting from that. I repeated boot and reboot tests with spot instances and normal ones. The same delay happens on every boot.
[ 28.165770] BUG: soft lockup - CPU#0 stuck for 23s! [migration/0:6]
[ 28.169759] Modules linked in:
[ 28.169759] CPU 0
[ 28.169759] Modules linked in:
[ 28.169759]
[ 28.169759] Pid: 6, comm: migration/0 Not tainted 3.2.0-17-virtual #27-Ubuntu Xen HVM domU
[ 28.169759] RIP: 0010:[<
[ 28.169759] RSP: 0018:ffff8805dd
[ 28.169759] RAX: 0000000000000000 RBX: ffffffff8101ad19 RCX: 0000000000000001
[ 28.169759] RDX: 0000000000000002 RSI: ffffffff81c0d020 RDI: 0000000000000001
[ 28.169759] RBP: ffff8805dd403f40 R08: 0000000000000000 R09: 0000000000000020
[ 28.169759] R10: 0000000000000000 R11: 0000000000000000 R12: ffff8805dd403e58
[ 28.169759] R13: ffffffff8165ab1e R14: ffff8805dd403f40 R15: ffff8805b7933fd8
[ 28.169759] FS: 000000000000000
[ 28.169759] CS: 0010 DS: 0000 ES: 0000 CR0: 000000008005003b
[ 28.169759] CR2: 0000000000000000 CR3: 0000000001c05000 CR4: 00000000000006f0
[ 28.169759] DR0: 0000000000000000 DR1: 0000000000000000 DR2: 0000000000000000
[ 28.169759] DR3: 0000000000000000 DR6: 00000000ffff0ff0 DR7: 0000000000000400
[ 28.169759] Process migration/0 (pid: 6, threadinfo ffff8805b7932000, task ffff8805b7938000)
[ 28.169759] Stack:
[ 28.169759] ffffffff8105fc23 ffff8805b7933fd8 000000000000000a ffff8805b7933fd8
[ 28.169759] 0000000000000000 ffffffff810bed01 ffff8805b7905df8 0000000000000046
[ 28.169759] ffff8805b7933fd8 0000000000000004 ffffffff810bed01 ffff8805b7905df8
[ 28.169759] Call Trace:
[ 28.169759] <IRQ>
[ 28.169759] [<ffffffff8105f
[ 28.169759] [<ffffffff810be
[ 28.169759] [<ffffffff810be
[ 28.169759] [<ffffffff8165c
[ 28.169759] [<ffffffff81015
[ 28.169759] [<ffffffff8106d
[ 28.169759] [<ffffffff8165c
[ 28.169759] [<ffffffff8165a
[ 28.169759] <EOI>
[ 28.169759] [<ffffffff810be
[ 28.169759] [<ffffffff810be
[ 28.169759] [<ffffffff810be
[ 28.169759] [<ffffffff81054
[ 28.169759] [<ffffffff8164f
[ 28.169759] [<ffffffff8104b
[ 28.169759] [<ffffffff810be
[ 28.169759] [<ffffffff81088
[ 28.169759] [<ffffffff8165c
[ 28.169759] [<ffffffff81088
[ 28.169759] [<ffffffff8165c
tags: | added: precise |
Repeating the boot after replacing the Precise kernel by an Oneiric (3.0.0) one there actually is a longer delay about the same time in boot. The difference is that is seems seems shorter 60s vs. 112s and does not cause the softlockup warnings. But something is delaying that boot as well.