Eric Blake wrote: > According to Jim Meyering on 9/7/2009 3:09 AM: >> ls -i now prints consistent inode numbers also for mount points. >> This makes ls -i DIR less efficient on systems with dysfunctional readdir, >> because ls must stat every file in order to obtain a guaranteed-valid >> inode number. [bug introduced in coreutils-6.0] > > On systems with working d_type, is it sufficient to limit the stat() to > situations where we know that the entry might be a directory (DT_DIR or > DT_UNKNOWN)? In other words, are we guaranteed that mount points can only > occur atop directories, and that we can avoid the stat() for regular files?
Recently I saw a note in which someone mentioned bind mounts on non-directories, so at least in general, that may not be sufficient. However, I don't have the reference handy and can't dig right now.