On (04/03/17 15:34), Minchan Kim wrote: [..] > > c) spin_locks probably have better fairness guarantees > > In fact, it wouldn't be an imporant because zram's slot lock contention > is not heavy.
mostly agree. I think (and I may be mistaken) direct IO causes contention; but direct IO is probably not a usual zram workload. > > what do you think? can we, in this patch set, also replce bit > > spin_locks with a normal spin_lock? > > With changing only zram side from bit_spin_lock to spin_lock, > it would be crippled. I mean zsmalloc should be changed, too > and it's really hard. :( hm, good point. -ss