Hi Slava, > Frankly speaking, I don't quite follow why should we update ctime > during the rename operation. Why do we need to do this? What is the > justification of this? This is not explicitly stated in the man page or anything like that, but it seems to be a rule generally followed by file systems. By the way, I did this experiment on apfs, and after the rename operation, the ctime changed. > And we still continue to operate by atime [1-4]. Should we do something > with it? I looked at other filesystems marked as NOATIME (e.g. jffs2), and similar atime operations were still preserved. Thx, Yangtao
