bug#15835: [PATCH] chmod: change in the inode's st_ctime

[Pádraig Brady]

tag 15835 wontfix
close 15835
stop

On 11/08/2013 01:07 PM, Guilherme de Almeida Suckevicz wrote:
> Hello,
> 
> I saw this issue on TODO file:
> 
> "Modify chmod so that it does not change an inode's st_ctime
>   when the selected operation would have no other effect.
>   First suggested by Hans Ecke <http://hans.ecke.ws> in
>     http://thread.gmane.org/gmane.comp.gnu.coreutils.bugs/2920
>   Discussed more recently on <http://bugs.debian.org/497514>."
> 
> I made a patch just to chmod, for now, to this issue and I would like
> to ask if it need something else or if you can make some comments about
> this patch, I really appreciate.
> 
> 
> --- src/chmod.c    2013-11-08 10:19:48.608214043 -0200
> +++ src/chmod-orig.c    2013-11-08 10:16:30.378505318 -0200
> @@ -266,7 +266,7 @@
>        new_mode = mode_adjust (old_mode, S_ISDIR (old_mode) != 0,
> umask_value,
>                                change, NULL);
> 
> -      if (mode_changed (file, old_mode, new_mode) && ! S_ISLNK (old_mode))
> +      if (! S_ISLNK (old_mode))
>          {
>            if (chmodat (fts->fts_cwd_fd, file, new_mode) == 0)
>              chmod_succeeded = true;

This has been discussed on and off, most recently at:
http://lists.gnu.org/archive/html/coreutils/2013-10/msg00028.html

The current thought is that it would be best done in the kernel,
as in userspace we haven't visibility of side effects of the call.

thanks,
Pádraig.