io_uring: properly mark async work as bounded vs unbounded
authorJens Axboe <axboe@kernel.dk>
Thu, 7 Nov 2019 17:57:36 +0000 (10:57 -0700)
committerJens Axboe <axboe@kernel.dk>
Thu, 7 Nov 2019 18:57:17 +0000 (11:57 -0700)
commit5f8fd2d3e0a7aa7fc9d97226be24286edd289835
tree4f2d850b8a843da6bdc86cff648a61a0b8ad8888
parentc5def4ab849494d3c97f6c9fc84b2ddb868fe78c
io_uring: properly mark async work as bounded vs unbounded

Now that io-wq supports separating the two request lifetime types, mark
the following IO as having unbounded runtimes:

- Any read/write to a non-regular file
- Any specific networked IO
- Any poll command

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