I think we'll need to queue this up for 4.21 ASAP independent of the rest, given that with separate poll queues userspace could otherwise submit I/O that will never get polled for anywhere.
- [PATCHSET v4] Support for polled aio Jens Axboe
- [PATCH 01/27] aio: fix failure to put the file poin... Jens Axboe
- Re: [PATCH 01/27] aio: fix failure to put the f... Bart Van Assche
- Re: [PATCH 01/27] aio: fix failure to put t... Jens Axboe
- Re: [PATCH 01/27] aio: fix failure to p... Bart Van Assche
- [PATCH 04/27] block: wire up block device iopoll me... Jens Axboe
- [PATCH 02/27] aio: clear IOCB_HIPRI Jens Axboe
- Re: [PATCH 02/27] aio: clear IOCB_HIPRI Christoph Hellwig
- Re: [PATCH 02/27] aio: clear IOCB_HIPRI Jens Axboe
- Re: [PATCH 02/27] aio: clear IOCB_HIPRI Christoph Hellwig
- Re: [PATCH 02/27] aio: clear IOCB_... Jens Axboe
- [PATCH 07/27] iomap: ensure that async polled IO is... Jens Axboe
- [PATCH 03/27] fs: add an iopoll method to struct fi... Jens Axboe
- [PATCH 06/27] iomap: wire up the iopoll method Jens Axboe
- [PATCH 11/27] aio: only use blk plugs for > 2 de... Jens Axboe
- Re: [PATCH 11/27] aio: only use blk plugs for &... Christoph Hellwig
- [PATCH 08/27] aio: use assigned completion handler Jens Axboe
- [PATCH 05/27] block: ensure that async polled IO is... Jens Axboe