System hang after nvme controller reset failure

Issue

• System hang after nvme controller reset failure:

nvme nvme0: async event result 00010300
nvme nvme0: controller is down; will reset: CSTS=0x3, PCI_STATUS=0x10
nvme 0000:86:00.0: irq 56 for MSI/MSI-X
nvme nvme0: Device not ready; aborting reset
nvme nvme0: Removing after probe failure status: -19
nvme0n1: detected capacity change from 1600321314816 to 0
....
Workqueue: nvme-wq nvme_remove_dead_ctrl_work [nvme]
Call Trace:
[<ffffffffb8767bc9>] schedule+0x29/0x70
[<ffffffffb834eee5>] blk_mq_freeze_queue_wait+0x75/0xe0
[<ffffffffb80c2d00>] ? wake_up_atomic_t+0x30/0x30
[<ffffffffb834fca4>] blk_freeze_queue+0x24/0x50
[<ffffffffb834417f>] blk_cleanup_queue+0x7f/0x1b0
[<ffffffffc04e355a>] nvme_ns_remove+0x9a/0xc0 [nvme_core]
[<ffffffffc04e370b>] nvme_remove_namespaces+0x9b/0xe0 [nvme_core]
[<ffffffffc0565892>] nvme_remove+0x92/0x140 [nvme]
[<ffffffffb83c607e>] pci_device_remove+0x3e/0xc0
[<ffffffffb84a7cd2>] __device_release_driver+0x82/0xf0
[<ffffffffb84a7d63>] device_release_driver+0x23/0x30
[<ffffffffc0563831>] nvme_remove_dead_ctrl_work+0x31/0x50 [nvme]
[<ffffffffb80b9d4f>] process_one_work+0x17f/0x440
[<ffffffffb80bade6>] worker_thread+0x126/0x3c0
[<ffffffffb80bacc0>] ? manage_workers.isra.25+0x2a0/0x2a0
[<ffffffffb80c1c31>] kthread+0xd1/0xe0
[<ffffffffb80c1b60>] ? insert_kthread_work+0x40/0x40
[<ffffffffb8774c1d>] ret_from_fork_nospec_begin+0x7/0x21
[<ffffffffb80c1b60>] ? insert_kthread_work+0x40/0x40