Messages in this thread | | | Date | Fri, 24 Aug 2012 17:47:56 +0000 | From | Bart Van Assche <> | Subject | Re: Kernel hangs in truncate_inode_pages() |
| |
On 08/24/12 17:24, Mike Christie wrote: > On 08/24/2012 08:13 AM, Bart Van Assche wrote: >> If I log in with the open-iscsi initiator to an iSCSI target an power >> down the target the initiator hangs in truncate_inode_pages(). This >> happens systematically with kernel 3.6-rc3 and it's something I've >> never seen with kernel 3.4 or before. Has anyone else already run into >> this issue ? > > I have not seen this issue before and have not seen it reported. > > Are you doing IO the disk at the time of power down, or are you doing a > iscsiadm -m node ... --logout command?
There was no I/O load test running at the time of the target power down. The call stack I posted is what I see before logout. And this is what I get after iscsi logout:
BUG: soft lockup - CPU#3 stuck for 23s! [iscsid:8589] Modules linked in: ... [last unloaded: crc32c] irq event stamp: 918076162 hardirqs last enabled at (918076161): [<ffffffff8140d98f>] _raw_spin_unlock_irqrestore+0x3f/0x70 hardirqs last disabled at (918076162): [<ffffffff814167d7>] apic_timer_interrupt+0x67/0x80 softirqs last enabled at (917955754): [<ffffffff81044162>] __do_softirq+0x142/0x2a0 softirqs last disabled at (917955749): [<ffffffff81416f5c>] call_softirq+0x1c/0x30 CPU 3 Pid: 8589, comm: iscsid Not tainted 3.6.0-rc3+ #1 Gigabyte Technology Co., Ltd. Z68X-UD3H-B3/Z68X-UD3H-B3 RIP: 0010:[<ffffffff8140d991>] [<ffffffff8140d991>] _raw_spin_unlock_irqrestore+0x41/0x70 RSP: 0018:ffff8800ae7e5928 EFLAGS: 00000286 RAX: ffff880106002eb0 RBX: ffffffff810956a2 RCX: 0000000000000006 RDX: 0000000000000040 RSI: ffff880106003498 RDI: ffff880106002eb0 RBP: ffff8800ae7e5938 R08: 0000000000000000 R09: 0000000000000001 R10: 0000000000000002 R11: 0000000000000000 R12: 0000000000000006 R13: 0000000000000007 R14: ffff880106002eb0 R15: ffff880106003498 FS: 00007fa8cfc22700(0000) GS:ffff88012fac0000(0000) knlGS:0000000000000000 CS: 0010 DS: 0000 ES: 0000 CR0: 000000008005003b CR2: 00007fa8cf26617c CR3: 00000000bc87c000 CR4: 00000000000407e0 DR0: 0000000000000000 DR1: 0000000000000000 DR2: 0000000000000000 DR3: 0000000000000000 DR6: 00000000ffff0ff0 DR7: 0000000000000400 Process iscsid (pid: 8589, threadinfo ffff8800ae7e4000, task ffff880106002eb0) Stack: ffff88011af0c000 ffff88001ea0fff0 ffff8800ae7e5988 ffffffffa00aaafc 0000000000000286 ffff88001ea10000 ffff8800ae7e5968 ffff88008ae44a28 ffff88008ae44800 ffff88008ae44818 0000000000000000 0000000000000001 Call Trace: [<ffffffffa00aaafc>] scsi_remove_target+0x1bc/0x1f0 [scsi_mod] [<ffffffffa06af6c6>] __iscsi_unbind_session+0xc6/0x1a0 [scsi_transport_iscsi] [<ffffffffa06af8a5>] iscsi_remove_session+0x105/0x1e0 [scsi_transport_iscsi] [<ffffffffa06af996>] iscsi_destroy_session+0x16/0x60 [scsi_transport_iscsi] [<ffffffffa07bc549>] iscsi_session_teardown+0xa9/0xd0 [libiscsi] [<ffffffffa07c91a0>] iscsi_sw_tcp_session_destroy+0x50/0x70 [iscsi_tcp] [<ffffffffa06aeef8>] iscsi_if_rx+0xab8/0xf60 [scsi_transport_iscsi] [<ffffffff81350cad>] netlink_unicast+0x1ad/0x230 [<ffffffff8135102b>] netlink_sendmsg+0x2fb/0x350 [<ffffffff8131c458>] ? sock_update_classid+0xc8/0x140 [<ffffffff813162d2>] sock_sendmsg+0xa2/0xe0 [<ffffffff8111c9ac>] ? might_fault+0x9c/0xb0 [<ffffffff8111c963>] ? might_fault+0x53/0xb0 [<ffffffff81325356>] ? verify_iovec+0x56/0xd0 [<ffffffff81317ac2>] __sys_sendmsg+0x382/0x390 [<ffffffff8111c9ac>] ? might_fault+0x9c/0xb0 [<ffffffff8111c963>] ? might_fault+0x53/0xb0 [<ffffffff81069c63>] ? lg_local_unlock+0x23/0x50 [<ffffffff811501f6>] ? cp_new_stat+0x116/0x130 [<ffffffff813198f9>] sys_sendmsg+0x49/0x90 [<ffffffff81415d52>] system_call_fastpath+0x16/0x1b
Bart.
| |