Guest kernel crashes when booting a NUMA guest without explicitly specifying cpus= in -numa option

[Bharata B Rao]

Target: x86_64-softmmu

Qemu Command line: [root@hs22 qemu-1.0-rc2]# ./x86_64-softmmu/qemu-system-x86_64 -smp sockets=2,cores=4,threads=2 -numa node,nodeid=0,mem=4g -numa node,nodeid=1,mem=1g -cpu core2duo -m 5g /home/bharata/f15-lvm -nographic --enable-kvm -net nic,macaddr=54:52:00:46:26:84,model=e1000 -net tap,script=/etc/qemu-if,ifname=vnet0

Qemu version: 1.0-rc2

When guest is started with -numa option without explicitly specifying the cpus=, guest kernel crashes as below:

[ 0.252159] divide error: 0000 [#1] SMP
[ 0.252970] last sysfs file:
[ 0.252970] CPU 1
[ 0.252970] Modules linked in:
[ 0.252970]
[ 0.252970] Pid: 2, comm: kthreadd Not tainted 2.6.38.6-26.rc1.fc15.x86_64 #1 Bochs Bochs
[ 0.252970] RIP: 0010:[<ffffffff8104f4d4>] [<ffffffff8104f4d4>] select_task_rq_fair+0x44a/0x571
[ 0.252970] RSP: 0000:ffff88011767fc60 EFLAGS: 00010046
[ 0.252970] RAX: 0000000000000000 RBX: ffff88015d6ad300 RCX: 0000000000000000
[ 0.252970] RDX: 0000000000000000 RSI: 0000000000000100 RDI: 0000000000000000
[ 0.252970] RBP: ffff88011767fd10 R08: 0000000000000100 R09: ffff88015d6ad338
[ 0.252970] R10: 0000000000013840 R11: 0000000000800711 R12: 0000000000000000
[ 0.252970] R13: ffff88015fc0f810 R14: 0000000000000001 R15: 0000000000000000
[ 0.252970] FS: 0000000000000000(0000) GS:ffff88015fc00000(0000) knlGS:0000000000000000
[ 0.252970] CS: 0010 DS: 0000 ES: 0000 CR0: 000000008005003b
[ 0.252970] CR2: 00000000ffffffff CR3: 0000000001a03000 CR4: 00000000000006e0
[ 0.252970] DR0: 0000000000000000 DR1: 0000000000000000 DR2: 0000000000000000
[ 0.252970] DR3: 0000000000000000 DR6: 00000000ffff0ff0 DR7: 0000000000000400
[ 0.252970] Process kthreadd (pid: 2, threadinfo ffff88011767e000, task ffff88015d671720)
[ 0.252970] Stack:
[ 0.252970] ffffffff81475873 ffffffff81a02140 ffff88011767fce0 ffffffff8106c5a3
[ 0.252970] ffff88015d6ad318 000000010000000e 0000000000013840 0000000000013840
[ 0.252970] ffff88015d6ad318 0000007d00000001 ffff880100000000 ffff88015d6d81e8
[ 0.252970] Call Trace:
[ 0.252970] [<ffffffff81475873>] ? _raw_spin_lock_irq+0x1c/0x1e
[ 0.252970] [<ffffffff8106c5a3>] ? alloc_pid+0x2e6/0x335
[ 0.252970] [<ffffffff81048960>] select_task_rq+0x16/0x46
[ 0.252970] [<ffffffff8104e29a>] wake_up_new_task+0x3a/0xde
[ 0.252970] [<ffffffff810546ce>] do_fork+0x1f1/0x2bf
[ 0.252970] [<ffffffff8100804e>] ? load_TLS+0x10/0x14
[ 0.252970] [<ffffffff81008714>] ? __switch_to+0xc6/0x220
[ 0.252970] [<ffffffff81010c1a>] kernel_thread+0x75/0x77
[ 0.252970] [<ffffffff8106eacf>] ? kthread+0x0/0x8c
[ 0.252970] [<ffffffff8100a9e0>] ? kernel_thread_helper+0x0/0x10
[ 0.252970] [<ffffffff8106ee93>] kthreadd+0xe7/0x124
[ 0.252970] [<ffffffff8100a9e4>] kernel_thread_helper+0x4/0x10
[ 0.252970] [<ffffffff8106edac>] ? kthreadd+0x0/0x124
[ 0.252970] [<ffffffff8100a9e0>] ? kernel_thread_helper+0x0/0x10
[ 0.252970] Code: 01 45 c0 8b 8d 78 ff ff ff 48 8b 75 90 89 cf e8 4a 28 ff ff 3b 05 bd 89 ae 00 89 c1 7c c5 48 8b 45 c0 8b 4b 08 31 d2 48 c1 e0 0a
[ 0.252970] f7 f1 45 85 e4 75 08 48 3b 45 b0 72 08 eb 0d 48 89 45 b8 eb
[ 0.252970] RIP [<ffffffff8104f4d4>] select_task_rq_fair+0x44a/0x571
[ 0.252970] RSP <ffff88011767fc60>

When cpus= is specified for each node explicitly, guest boots fine.

[Thomas Huth (th-huth)]

Is this still an issue with the latest version of QEMU?

[Launchpad Janitor (janitor)]

[Expired for QEMU because there has been no activity for 60 days.]