Relay-Version: version B 2.10 5/3/83; site utzoo.UUCP
Path: utzoo!utgpu!water!watmath!clyde!cbosgd!ihnp4!alberta!sask!reid
From: reid@sask.UUCP
Newsgroups: comp.os.minix
Subject: Re: Request for help with rmdir bug
Message-ID: <763@sask.UUCP>
Date: Mon, 8-Jun-87 02:29:15 EDT
Article-I.D.: sask.763
Posted: Mon Jun  8 02:29:15 1987
Date-Received: Wed, 10-Jun-87 07:16:27 EDT
References: <1195@botter.cs.vu.nl>
Reply-To: reid@sask.UUCP (Irving Reid)
Distribution: world
Organization: The Church of the Least Fixed Point
Lines: 38

In article <1195@botter.cs.vu.nl> ast@cs.vu.nl (Andy Tanenbaum) writes:
> ...
>rmdir fails.  I checked the code of rmdir.  What happens is that rmdir
>first unlinks . and .. from foo, then it tries to access foo/../foo to
>unlink the directory, but that path is no longer valid because .. is now
>gone.  I tried the same test on our VAX with 4.1 BSD and it failed
>there too.  
>
>The question is: does anyone have an idea how to make rmdir work without
>modifying the kernel.  Unlinking the directory before removing . and ..
>does not sound like a good idea.
>
>Andy Tanenbaum (ast@cs.vu.nl)


I have no idea how any Unices do it, but how about:

/* rmdir.pseudo-c: */

rmdir(directory)
{
	chdir("directory/..");
	dname <- name of directory with any path stripped off;
	unlink(dname/..);
	unlink(dname/.);
	unlink(dname);
}

This would fail at the root directory, because /. and /.. are both links
to / (at least on the Unices I've used), but should work everywhere else.

(by the way, Ultrix 1.2 (almost 4.2BSD) handles the bad rmdir correctly)

-- 
 - !Baboo -   (reid@sask.uucp or {alberta, ihnp4, utcsri}!sask!reid)

Rabbi, is there a proper blessing for the Tsar?
Certainly, my son.  May God bless and keep the Tsar...  Far away from us.