Re: New feature

Ricky Beam (root@defiant.interpath.net)
Fri, 26 Sep 1997 06:34:24 -0400 (EDT)


Letting the chips fall where they may, I quote Rogier Wolff:
>I disagree. stat-ing a file accesses the directory. As does reading
>the directory. stat-ing a file does not access that file.
>
>This is the way it is supposed to work, this is the way it works on
>2.0.30.
>
>Otherwise "ls -lrt --time=atime" would be pretty destructive: you'd
>touch all access time on the way.

...

Gee, how long are we going to argue about this before someone goes to the
kernel to see what the h*ll it's doing and then consult the POSIX docs to
see if there are any rules.

I quote the stat(2) manpage (Linux 1.175):
Not all of the Linux filesystems implement all of the time
fields. Traditionally, st_atime is changed by mknod(2),
utime(2), read(2), write(2), and truncate(2).

Traditionally, st_mtime is changed by mknod(2), utime(2),
and write(2). The st_mtime is not changed for changes in
owner, group, hard link count, or mode.

Traditionally, st_ctime is changed by writing or by set-
ting inode information (i.e., owner, group, link count,
mode, etc.).

--Ricky