Maciej Żenczykowski wrote:
> I believe '/usr/bin/find' does use the link count for some
> optimization (it assumes a directory has nlink-2 subdirectories, -2
> since 1 for the link from parent to itself, and 1 for the '.'), but
> should fall back to a less optimized mode if it sees a mismatch...
It can speed up some find operations quite a lot, when find doesn't
have to stat(2) all the entries in the leaf directories to know if it
needs to recurse deeper or not.
And it is sometimes used by users of find(1) too, to find leaf directories.
'find . -links 2' is the easiest way I know of for that.
/Bellman
_______________________________________________
Tux3 mailing list
[email protected]
http://mailman.tux3.org/cgi-bin/mailman/listinfo/tux3
