Re: New feature
Rogier Wolff (R.E.Wolff@BitWizard.nl)
Wed, 24 Sep 1997 14:40:58 +0200 (MET DST)
Ricky Beam wrote:
>
> Letting the chips fall where they may, I quote David S. Miller:
> > Date: Tue, 23 Sep 1997 21:49:12 -0400 (EDT)
> > From: "Richard B. Johnson" <root@chaos.analogic.com>
> >
> > Correct! Now, is a directory access a file access? I don't think it
> > should be.
> >
> >Why not? It has been forever at least for ext2. Have a look at the
> >end of fs/ext2/dir.c:ext2_readdir().
> >
> >I think POSIX even states that this is how things should work.
>
> This is UNIX, the filesystem consists of FILES - period. What you do with
> magic bits in the filesystem is your business.
>
> There used to be all manner of neat things one could do with setting/clearing
> the directory bit in the inode... turn a regular file into a directory and
> vice-versa. The kernel now blocks such idiot things as accessing directories
> without specifying that you accessing them as a directory: 'cat .' no longer
> works.
>
> >ATIME is the access time for an inode, when you read an inode which
> >happens to be a directory you are accessing it.
>
> A clarification... if you access an inode (any inode,) the access time is
> updated. stat()ing the inode accesses the inode, hence the atime is updated
> for every file on the drive following 'ls -laR /'
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.
Roger.
--
** R.E.Wolff@BitWizard.nl ** +31-15-2137555 ** http://www.BitWizard.nl/ **
Florida -- A 39 year old construction worker woke up this morning when a
109-car freigt train drove over him. According to the police the man was
drunk. The man himself claims he slipped while walking the dog. 080897