On 28 Apr 2014, at 13:24, Ilia Zykov <net...@izyk.ru> wrote: > Sorry to disturb you. > > But I have small comment about subject. > I think it is unnecessary overload involve mount point in lookup for global > cache. > You save few bytes for hashhead for every mount point, but loose effectively > lookup. > Every lookup has one more compare for mp, and your hashtable will be more > bigger. > In any case have hashhead and hashlock for every mount point is more flexible > way, IMHO. > "struct mount" is a good place for this.
First, one of the goals here is to have one table for all vnodes. Second, regarding space this even gets smaller. A hash table allocates next_power_of_two(desiredvnodes) list heads so every hash table accounts for ~1.5 pointers per vnode. -- J. Hannken-Illjes - hann...@eis.cs.tu-bs.de - TU Braunschweig (Germany)