Mailing-List: contact cygwin-help AT sourceware DOT cygnus DOT com; run by ezmlm
List-Subscribe: <mailto:cygwin-subscribe AT sources DOT redhat DOT com>
List-Archive: <http://sources.redhat.com/ml/cygwin/>
List-Post: <mailto:cygwin AT sources DOT redhat DOT com>
List-Help: <mailto:cygwin-help AT sources DOT redhat DOT com>, <http://sources.redhat.com/ml/#faqs>
Sender: cygwin-owner AT sources DOT redhat DOT com
Delivered-To: mailing list cygwin AT sources DOT redhat DOT com
From: Chris Faylor <cgf AT cygnus DOT com>
Date: Wed, 18 Oct 2000 15:42:30 -0400
To: Egor Duda <cygwin AT sources DOT redhat DOT com>
Subject: Re: /bin/pwd (getcwd) and symlinks
Message-ID: <20001018154230.C5355@cygnus.com>
Reply-To: cygwin AT sources DOT redhat DOT com
Mail-Followup-To: Egor Duda <cygwin AT sources DOT redhat DOT com>
References: <20001018134646 DOT 16601 DOT qmail AT web113 DOT yahoomail DOT com> <000c01c0390d$0591e760$21c9ca95 AT mow DOT siemens DOT ru>
Mime-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Disposition: inline
User-Agent: Mutt/1.3.6i
In-Reply-To: <000c01c0390d$0591e760$21c9ca95@mow.siemens.ru>; from Andrej.Borsenkow@mow.siemens.ru on Wed, Oct 18, 2000 at 06:09:25PM +0400

I've previously posted about this.  It probably shouldn't work this way but,
as Corinna pointed out, this change was made to fix a more serious problem
with find.

It's on my list of things to fix so it will be fixed eventually.

cgf

On Wed, Oct 18, 2000 at 06:09:25PM +0400, Andrej Borsenkow wrote:
>>
>> Ok.  HP-UX /bin/pwd prints the actual directory.  But, what does /bin/pwd on
>> Linux do?  Where is any documentation for what it should do for symlink?  I
>> haven't found anything useful that describes what happens for symlink.
>>
>
>It is not "what it should do" - it is how symlinks and getcwd are (typically)
>implemented.
>
>Symlinks are implemented by storing target name as symlink file content. When
>you open symlink, kernel implicitly opens target, giving you (dev,inode) - or
>vnode - of _target_. After this step, there is no way to recover the original
>name (or inode) of symlink.
>
>getcwd is implemented by fetching inode of current directory, scanning parent
>directory (..) and comparing inodes there with current inode. When it is
>found, directory entry gives you last part of absolute name. Than getcwd goes
>up (cd ..) and repeats it. It does it until it hits root (inode(.) ==
>inode(..)).
>
>If you combine both, you see, that there is no way getcwd can actually find
>_symlink_ name.
>
>As stated - this is typical implementation. YMMV :-) But I'm really curious to
>see any Unix-like system that does it differently.

--
Want to unsubscribe from this list?
Send a message to cygwin-unsubscribe AT sourceware DOT cygnus DOT com