This commit replace rproc_{shutdown,boot}() with rproc_{stop,start}(), which
skip destroy the virtio device at stop but reinitialize it again at start:
[ 1502.733760] remoteproc remoteproc0: crash detected in
f9210000.toppwr:tl421-rproc: type watchdog
[ 1502.744019] remoteproc remoteproc0: handling crash #1 in
f9210000.toppwr:tl421-rproc
[ 1502.820741] remoteproc remoteproc0: recovering
f9210000.toppwr:tl421-rproc
[ 1502.831588] remoteproc remoteproc0: stopped remote processor
f9210000.toppwr:tl421-rproc
[ 1502.844458] kobject (ffffffc0b8d1b898): tried to init an initialized
object, something is seriously wrong.
[ 1502.855415] CPU: 6 PID: 169 Comm: kworker/6:1 Tainted: G W
4.9.27-04453-g1a03c96 #253
[ 1502.865442] Hardware name: Banks (DT)
[ 1502.869557] Workqueue: events rproc_crash_handler_work
[ 1502.875319] Call trace:
[ 1502.878066] [<ffffff800808bd9c>] dump_backtrace+0x0/0x1cc
[ 1502.884106] [<ffffff800808bf7c>] show_stack+0x14/0x1c
[ 1502.889761] [<ffffff80083fef08>] dump_stack+0xa8/0xe0
[ 1502.895415] [<ffffff8008402b24>] kobject_init+0x8c/0x9c
[ 1502.901262] [<ffffff800853758c>] device_initialize+0x3c/0xe8
[ 1502.907590] [<ffffff80085397d4>] device_register+0x14/0x28
[ 1502.913727] [<ffffff80084b777c>] register_virtio_device+0xc4/0x114
[ 1502.920641] [<ffffff8008878b20>] rproc_add_virtio_dev+0x7c/0x108
[ 1502.927360] [<ffffff8008875cbc>] rproc_vdev_do_probe+0x14/0x1c
[ 1502.933883] [<ffffff8008875a60>] rproc_start+0xac/0x1ac
[ 1502.939728] [<ffffff8008877a68>] rproc_trigger_recovery+0x2f8/0x324
[ 1502.946740] [<ffffff8008877b24>] rproc_crash_handler_work+0x90/0xb0
[ 1502.953752] [<ffffff80080cd570>] process_one_work+0x204/0x704
[ 1502.960179] [<ffffff80080cdac4>] worker_thread+0x54/0x4a8
[ 1502.966225] [<ffffff80080d4aec>] kthread+0xec/0x100
[ 1502.971683] [<ffffff8008083890>] ret_from_fork+0x10/0x40
When the crash happen, It?s better to destroy and recreate all virtio device
and it?s children(rpmsg device) again to match the remote side state like
the original behavior.
Thanks
Xiang
