On Tue, Nov 22 2016, Shaohua Li wrote: > > I still don't like we have no control when to do endio, but don't have better > idea either. These patches work for raid5, but we need to find similar tricky > way to workaround raid1/raid10, which reuse bi_phys_segments too. Probably it's > time MD allocates additional data and attaches to bio. Adding a counter will > solve the issue in a consistency/clean way for raid1/10/5. For raid1/raid10 we could do a very similar thing. There is an awkwardness in raid1 w.r.t waiting for bi_phys_segments to reach 1, but that might disappear if Coly's resync changes go through. Alternately it might make sense to use bio_split so there is one r1_bio per bio. I might try the raid10 version and see what it looks like. Thanks, NeilBrown > > Thanks, > Shaohua
Attachment:
signature.asc
Description: PGP signature
