Xen project Mailing List

Hi guys,

After upgrading Debian Dom0 from xen-hypervisor-amd64-4.1.3~rc1+hg-20120614.a9c0a89c08f2-4 to xen-hypervisor-amd64-4.1.3-1 all PV domains with debian squeeze with stable kernel linux-image-2.6.32-45 fails to 'xm restore' with trace:

[296707.372141] INFO: task xenwatch:16 blocked for more than 120 seconds.
[296707.372869] "echo 0 > /proc/sys/kernel/hung_task_timeout_secs" disables this message.
[296707.373605] xenwatch D 0000000000000002 0 16 2 0x00000000
[296707.374353] ffffffff814891f0 0000000000000246 ffff88007ff57138 ffffffff81041238
[296707.375113] ffff88007ff57138 ffffffff81311f90 000000000000f9e0 ffff88007ff9dfd8
[296707.375884] 0000000000015780 0000000000015780 ffff88007ff57100 ffff88007ff573f8
[296707.376663] Call Trace:
[296707.377428] [<ffffffff81041238>] ? pick_next_task_fair+0xca/0xd6
[296707.378210] [<ffffffff812fba70>] ? thread_return+0x79/0xe0
[296707.378994] [<ffffffff8100e15c>] ? xen_vcpuop_set_next_event+0x0/0x60
[296707.379802] [<ffffffff812fbe2d>] ? schedule_timeout+0x2e/0xdd
[296707.380585] [<ffffffff8100e22f>] ? xen_restore_fl_direct_end+0x0/0x1
[296707.381232] [<ffffffff812fce0a>] ? _spin_unlock_irqrestore+0xd/0xe
[296707.381832] [<ffffffff810ad4e8>] ? cpupri_set+0x10c/0x135
[296707.382419] [<ffffffff812fbce4>] ? wait_for_common+0xde/0x15b
[296707.383059] [<ffffffff8104a46c>] ? default_wake_function+0x0/0x9
[296707.383671] [<ffffffff81061cb0>] ? flush_cpu_workqueue+0x5e/0x75
[296707.384287] [<ffffffff81064dda>] ? kthread_stop+0x5d/0xa2
[296707.384913] [<ffffffff81061d0b>] ? cleanup_workqueue_thread+0x44/0x51
[296707.385507] [<ffffffff81061db2>] ? destroy_workqueue+0x76/0xad
[296707.386148] [<ffffffff8108b03b>] ? stop_machine_destroy+0x2e/0x47
[296707.386791] [<ffffffff811f0969>] ? shutdown_handler+0x230/0x25c
[296707.387443] [<ffffffff812fc3a6>] ? mutex_lock+0xd/0x31
[296707.387986] [<ffffffff811f1c8f>] ? xenwatch_thread+0x117/0x14a
[296707.388564] [<ffffffff81065042>] ? autoremove_wake_function+0x0/0x2e
[296707.389097] [<ffffffff811f1b78>] ? xenwatch_thread+0x0/0x14a
[296707.389656] [<ffffffff81064d75>] ? kthread+0x79/0x81
[296707.390197] [<ffffffff81011baa>] ? child_rip+0xa/0x20
[296707.390751] [<ffffffff81010d61>] ? int_ret_from_sys_call+0x7/0x1b
[296707.391519] [<ffffffff8101151d>] ? retint_restore_args+0x5/0x6
[296707.392185] [<ffffffff8100e22f>] ? xen_restore_fl_direct_end+0x0/0x1
[296707.392757] [<ffffffff81011ba0>] ? child_rip+0x0/0x20
[296707.393303] INFO: task getty:634 blocked for more than 120 seconds.
[296707.393866] "echo 0 > /proc/sys/kernel/hung_task_timeout_secs" disables this message.
[296707.394449] getty D 0000000000000002 0 634 1 0x00000000
[296707.395084] ffffffff814891f0 0000000000000286 ffff88007ef0cde8 ffffffff81041238
[296707.395673] ffff88007ef0cde8 ffffffff81311f90 000000000000f9e0 ffff88007f239fd8
[296707.396300] 0000000000015780 0000000000015780 ffff88007ef0cdb0 ffff88007ef0d0a8
[296707.396954] Call Trace:
[296707.397518] [<ffffffff81041238>] ? pick_next_task_fair+0xca/0xd6
[296707.398103] [<ffffffff812fba70>] ? thread_return+0x79/0xe0
[296707.398675] [<ffffffff812fbe2d>] ? schedule_timeout+0x2e/0xdd
[296707.399244] [<ffffffff81044007>] ? select_task_rq_fair+0x7dc/0x843
[296707.399810] [<ffffffff812fbce4>] ? wait_for_common+0xde/0x15b
[296707.400355] [<ffffffff8104a46c>] ? default_wake_function+0x0/0x9
[296707.400920] [<ffffffff8104b09b>] ? sched_exec+0x114/0x130
[296707.401471] [<ffffffff810f4b2b>] ? do_execve+0xe8/0x2c3
[296707.402013] [<ffffffff8100f500>] ? sys_execve+0x35/0x4c
[296707.402552] [<ffffffff81010f9a>] ? stub_execve+0x6a/0xc0
...

Upgrading domU kernel to debian testing 3.2.23-1 fix this problem.
Saving under xen 4.3.1 and restoring under xen 4.1.3~rc1+hg-20120614.a9c0a89c08f2-4 also works fine.
Is there any way to save/restore debian squeeze vm's running on 4.1.3 Dom0?

[Xen-devel] Xen 4.1.3 restore problem with 2.6.32-45 domU