Comment 14 for bug 791364

Situation is still the same with:
Dom0 as of today with xen-hypervisor-4.1-amd64 4.1.1-2ubuntu1 and linux-headers-3.0.0-9-server 3.0.0-9.15

oneiric hvm domU with 1 CPU and 1GB RAM, linux-image-server 3.0.0-9.15 or linux-image-generic 3.1.0-rc4 don't boot.
I've added kernel parameters debug loglevel=8 initcall_debug and these are the last messages after which only rcu_sched_state detected stall on CPU0 messages are printed:

[ 10.048487] Freeing unused kernel memory: 1376k freed
Loading, please wait...
[ 10.110763] udevd[81]: starting version 173
Begin: Loading essential drivers ... [ 10.145811] calling platform_pci_module_init+0x0/0x1000 [xen_platform_pci] @ 118
[ 10.154464] xen: --> irq=28, pirq=22
[ 10.158913] xen-platform-pci 0000:00:03.0: PCI INT A -> GSI 28 (level, low) -> IRQ 28
[ 10.170597] Grant table initialized
[ 10.204926] initcall platform_pci_module_init+0x0/0x1000 [xen_platform_pci] returned 0 after 49400 usecs
[ 10.226935] calling xenbus_probe_frontend_init+0x0/0x1000 [xenbus_probe_frontend] @ 132
[ 10.248089] initcall xenbus_probe_frontend_init+0x0/0x1000 [xenbus_probe_frontend] returned 0 after 11533 usecs
[ 10.269725] calling netif_init+0x0/0x1000 [xen_netfront] @ 132
[ 10.276463] Initialising Xen virtual ethernet driver.
[ 10.304325] initcall netif_init+0x0/0x1000 [xen_netfront] returned 0 after 27187 usecs
[ 10.331757] calling xlblk_init+0x0/0x1000 [xen_blkfront] @ 147
[ 10.349331] blkfront device/vbd/51712 num-ring-pages 1 nr_ents 32.
[ 10.364719] initcall xlblk_init+0x0/0x1000 [xen_blkfront] returned 0 after 26133 usecs
done.
Begin: Running /scripts/init-premount ... [ 70.390060] INFO: rcu_sched_state detected stall on CPU 0 (t=6000 jiffies)
[ 250.690061] INFO: rcu_sched_state detected stall on CPU 0 (t=24030 jiffies)

BTW: A debian domU with linux-image 3.0.0-1 boots fine.