On 10/19/21 3:24 PM, 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. These look good to me, only the trivial comment that patch 7 has a blocK instead of block. I'll await further comments/reviews on these. -- Jens Axboe
