Patch to allow trailing dots on managed mounts

[Pierre A. Humblet]

Christopher Faylor wrote:
> 
> On Mon, Dec 20, 2004 at 11:18:05AM -0500, Pierre A. Humblet wrote:
> >
> >Christopher Faylor wrote:
> >>
> >> On Mon, Dec 20, 2004 at 10:53:34AM -0500, Pierre A. Humblet wrote:
> >> >Stripping from the Posix path can't be done during normalize_
> >> >because it would apply to all paths (not only disk).
> >>
> >> Why can't we just strip the dots from the path in
> >> path_conv::set_normalized path?
> >
> >You can, after checking the device. But why do it all
> >the time if it's only needed by chdir?
> 
> It just seems more consistent and safer to do the same thing to both the
> win32 and posix paths.  chroot probably needs it too.  After removing
> the code from normalize_posix_path, this is probably not a performance
> hit.

Good points. normalize_posix_path already finds the length, so the
performance hit will be really small.

Not sure what you mean by "removing the code from normalize_posix_path".
It's still important that normalize_{posix,win32}_path strip the final 
'.' in "xxxx/.", because :check looks for a final /
 
Pierre