tree 252f986cdcdce54cd1bd896ad01714d085e71068
parent 9bfa529518cb2402d8f8384b95530cb0a7d2b16e
author Jens Axboe <axboe@kernel.dk> 1590161082 -0600
committer Jens Axboe <axboe@kernel.dk> 1590521524 -0600

io_uring: support true async buffered reads, if file provides it

If the file is flagged with FMODE_BUF_RASYNC, then we don't have to punt
the buffered read to an io-wq worker. Instead we can rely on page
unlocking callbacks to support retry based async IO. This is a lot more
efficient than doing async thread offload.

The retry is done similarly to how we handle poll based retry. From
the unlock callback, we simply queue the retry to a task_work based
handler.

Signed-off-by: Jens Axboe <axboe@kernel.dk>