io_uring: consider the overflow of sequence for timeout req
authoryangerkun <yangerkun@huawei.com>
Tue, 15 Oct 2019 13:59:29 +0000 (21:59 +0800)
committerJens Axboe <axboe@kernel.dk>
Tue, 15 Oct 2019 14:55:50 +0000 (08:55 -0600)
commit5da0fb1ab34ccfe6d49210b4f5a739c59fcbf25e
treec9eaf517afbc8c7dc8e2f9633ec17d6f0416e647
parent7a7c5e715e722c86d602c56a09e77f000364e263
io_uring: consider the overflow of sequence for timeout req

Now we recalculate the sequence of timeout with 'req->sequence =
ctx->cached_sq_head + count - 1', judge the right place to insert
for timeout_list by compare the number of request we still expected for
completion. But we have not consider about the situation of overflow:

1. ctx->cached_sq_head + count - 1 may overflow. And a bigger count for
the new timeout req can have a small req->sequence.

2. cached_sq_head of now may overflow compare with before req. And it
will lead the timeout req with small req->sequence.

This overflow will lead to the misorder of timeout_list, which can lead
to the wrong order of the completion of timeout_list. Fix it by reuse
req->submit.sequence to store the count, and change the logic of
inserting sort in io_timeout.

Signed-off-by: yangerkun <yangerkun@huawei.com>
Signed-off-by: Jens Axboe <axboe@kernel.dk>
fs/io_uring.c