io_uring: fix io_wqe->work_list corruption

commit 0020ef0 upstream.

For the first time a req punted to io-wq, we'll initialize io_wq_work's
list to be NULL, then insert req to io_wqe->work_list. If this req is not
inserted into tail of io_wqe->work_list, this req's io_wq_work list will
point to another req's io_wq_work. For splitted bio case, this req maybe
inserted to io_wqe->work_list repeatedly, once we insert it to tail of
io_wqe->work_list for the second time, now io_wq_work->list->next will be
invalid pointer, which then result in many strang error, panic, kernel
soft-lockup, rcu stall, etc.

In my vm, kernel doest not have commit cc29e1b ("block: disable
iopoll for split bio"), below fio job can reproduce this bug steadily:
[global]
name=iouring-sqpoll-iopoll-1
ioengine=io_uring
iodepth=128
numjobs=1
thread
rw=randread
direct=1
registerfiles=1
hipri=1
bs=4m
size=100M
runtime=120
time_based
group_reporting
randrepeat=0

[device]
directory=/home/feiman.wxg/mntpoint/  # an ext4 mount point

If we have commit cc29e1b ("block: disable iopoll for split bio"),
there will no splitted bio case for polled io, but I think we still to need
to fix this list corruption, it also should maybe go to stable branchs.

To fix this corruption, if a req is inserted into tail of io_wqe->work_list,
initialize req->io_wq_work->list->next to bu NULL.

Cc: stable@vger.kernel.org
Signed-off-by: Xiaoguang Wang <xiaoguang.wang@linux.alibaba.com>
Reviewed-by: Pavel Begunkov <asml.silence@gmail.com>
Signed-off-by: Jens Axboe <axboe@kernel.dk>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

fs/io-wq.h

@@ -59,6 +59,7 @@ static inline void wq_list_add_tail(struct io_wq_work_node *node,
 		list->last->next = node;
 		list->last = node;
 	}
+	node->next = NULL;
 }
 
 static inline void wq_list_cut(struct io_wq_work_list *list,