Do we really need correct st_nlink count for directories?

[Eric Blake]

According to Eric Blake on 4/24/2008 7:32 AM:
> But if it takes an entire directory read to determine a 
> correct st_nlink, in order to avoid an entire directory as an 
> optimization, then it isn't optimal. I'm all for dropping correct 
> st_nlink, and using 1 instead.

More importantly, if it will speed up the stat() of a directory in the 
case where the st_nlink optimization is not being used (which I believe is 
more often the case), then the performance improvement of not counting 
subdirs for a faster stat() definitely outweighs the loss in optimization 
on the few places where the readdir() st_nlink optimization even makes 
sense (for example, even though find is coded to use the st_nlink 
optimization, I would hazard a guess that less than 50% of find queries 
are actually able to use it, because there are only a limited subset of 
queries where you only care about identifying subdirectories, rather than 
visiting all files in the directory).

-- 
Don't work too hard, make some time for fun as well!

Eric Blake             ebb9@byu.net