rcu_sched CPU stall is detected where qemu-kvm is exiting and freeing the inode

Issue

• rcu_sched CPU stall is detected where qemu-kvm is exiting and freeing the inode.

kernel: INFO: rcu_sched detected stalls on CPUs/tasks:
    ...
{
16}  (t=1320046 jiffies g=1209611683 c=1209611682 q=2898038)
Task dump for CPU 16:
qemu-kvm        R  running task        0 353541  37401 0x0008012a
Call Trace:
    ...
<EOI>  [<ffffffffb25b94fd>] ? find_get_pages+0x7d/0x1d0
[<ffffffffb25b959a>] ? find_get_pages+0x11a/0x1d0
[<ffffffffb25c468e>] pagevec_lookup+0x1e/0x30
[<ffffffffb26d4082>] remove_inode_hugepages+0x152/0x430
[<ffffffffb2461dee>] ? physflat_send_IPI_mask+0xe/0x10
[<ffffffffb2456c22>] ? native_smp_send_reschedule+0x52/0x70
[<ffffffffb26d449e>] hugetlbfs_evict_inode+0x1e/0x40
[<ffffffffb265eea4>] evict+0xb4/0x180
[<ffffffffb265f7ac>] iput+0xfc/0x190
[<ffffffffb265a020>] __dentry_kill+0x120/0x180
[<ffffffffb265a130>] dput+0xb0/0x160
[<ffffffffb2643ade>] __fput+0x17e/0x260
[<ffffffffb2643cae>] ____fput+0xe/0x10
[<ffffffffb24be7db>] task_work_run+0xbb/0xe0
[<ffffffffb249dca1>] do_exit+0x2d1/0xa40
[<ffffffffb249e48f>] do_group_exit+0x3f/0xa0
[<ffffffffb249e504>] SyS_exit_group+0x14/0x20
[<ffffffffb2b7605b>] tracesys+0xa3/0xc9