Comment 8 for bug 1602724

------- Comment From <email address hidden> 2016-08-22 13:11 EDT-------
(In reply to comment #16)
> Test kernel at http://people.canonical.com/~rtg/eeh-lp1602724/ with upstream
> commit c21377f8366c95440d533edbe47d070f662c62ef ('nvme: Suspend all queues
> before deletion') applied.

This test kernel is not ok, it stalls the wq:

[ 540.097661] INFO: rcu_sched detected stalls on CPUs/tasks:
[ 540.103320] 1-...: (1 GPs behind) idle=d35/140000000000000/0 softirq=2385/2386 fqs=65
[ 540.103411] (detected by 11, t=5472 jiffies, g=1335, c=1334, q=793)
[ 540.103492] Task dump for CPU 1:
[ 540.103539] kworker/u32:1 D 0000000000000000 0 101 0 0x00000800
[ 540.103656] Call Trace:
[ 540.103692] [c00000017bc539c0] [c00000017bc53a00] 0xc00000017bc53a00 (unreliable)
[ 540.103805] [c00000017bc53a00] [d000000001614480] nvme_suspend_queue+0x30/0x150 [nvme]
[ 540.103914] [c00000017bc53a30] [d000000001616850] nvme_dev_disable+0x110/0x440 [nvme]
[ 540.104022] [c00000017bc53b10] [d000000001617e60] nvme_reset_work+0xe0/0x1120 [nvme]
[ 540.104132] [c00000017bc53c50] [c0000000000dd630] process_one_work+0x1e0/0x5a0
[ 540.104239] [c00000017bc53ce0] [c0000000000ddb84] worker_thread+0x194/0x680
[ 540.104331] [c00000017bc53d80] [c0000000000e6680] kthread+0x110/0x130
[ 540.104424] [c00000017bc53e30] [c000000000009538] ret_from_kernel_thread+0x5c/0xa4
[ 604.094501] INFO: rcu_sched detected stalls on CPUs/tasks:
[ 604.094699] 1-...: (1 GPs behind) idle=d35/140000000000000/0 softirq=2385/2386 fqs=82
[ 604.094700] (detected by 5, t=21472 jiffies, g=1335, c=1334, q=1283)
[ 604.094705] Task dump for CPU 1:

Can you provide the backported patch for verification?