On Tue, 19 Oct 2021 22:24:09 +0100, Pavel Begunkov wrote:
> Jens tried out a similar series with some not yet sent additions:
> 8.2-8.3 MIOPS -> ~9 MIOPS, or 8-10%.
>
> 12/16 is bulky, but it nicely drives the numbers. Moreover, with
> it we can rid of some not used anymore optimisations in
> __blkdev_direct_IO() because it awlays serve multiple bios.
> E.g. no need in conditional referencing with DIO_MULTI_BIO,
> and _probably_ can be converted to chained bio.
>
> [...]
Applied, thanks!
[01/16] block: turn macro helpers into inline functions
(no commit info)
[02/16] block: convert leftovers to bdev_get_queue
(no commit info)
[03/16] block: optimise req_bio_endio()
(no commit info)
[04/16] block: don't bloat enter_queue with percpu_ref
(no commit info)
[05/16] block: inline a part of bio_release_pages()
(no commit info)
[06/16] block: clean up blk_mq_submit_bio() merging
(no commit info)
[07/16] blocK: move plug flush functions to blk-mq.c
(no commit info)
[08/16] block: optimise blk_flush_plug_list
(no commit info)
[09/16] block: optimise boundary blkdev_read_iter's checks
(no commit info)
[10/16] block: optimise blkdev_bio_end_io()
(no commit info)
[11/16] block: add optimised version bio_set_dev()
(no commit info)
[12/16] block: add single bio async direct IO helper
(no commit info)
[13/16] block: add async version of bio_set_polled
(no commit info)
[14/16] block: skip advance when async and not needed
(no commit info)
[15/16] block: optimise blk_may_split for normal rw
(no commit info)
[16/16] block: optimise submit_bio_checks for normal rw
(no commit info)
Best regards,
--
Jens Axboe
